{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 29704,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00033665499596014,
      "grad_norm": 1.6060822637333099,
      "learning_rate": 9.999999552565099e-06,
      "loss": 0.8102286338806153,
      "num_input_tokens_seen": 246864,
      "step": 5,
      "train_runtime": 31.1268,
      "train_tokens_per_second": 7930.927
    },
    {
      "epoch": 0.00067330999192028,
      "grad_norm": 1.5464818271407035,
      "learning_rate": 9.999997734860948e-06,
      "loss": 0.6559724807739258,
      "num_input_tokens_seen": 484672,
      "step": 10,
      "train_runtime": 58.5215,
      "train_tokens_per_second": 8281.953
    },
    {
      "epoch": 0.0010099649878804202,
      "grad_norm": 1.255466245784032,
      "learning_rate": 9.999994518923374e-06,
      "loss": 0.6377837657928467,
      "num_input_tokens_seen": 734864,
      "step": 15,
      "train_runtime": 93.1261,
      "train_tokens_per_second": 7891.061
    },
    {
      "epoch": 0.00134661998384056,
      "grad_norm": 1.1184952383437727,
      "learning_rate": 9.999989904753278e-06,
      "loss": 0.6039176940917969,
      "num_input_tokens_seen": 991408,
      "step": 20,
      "train_runtime": 119.2296,
      "train_tokens_per_second": 8315.117
    },
    {
      "epoch": 0.0016832749798007002,
      "grad_norm": 1.0011051994904234,
      "learning_rate": 9.999983892351947e-06,
      "loss": 0.5771594524383545,
      "num_input_tokens_seen": 1251968,
      "step": 25,
      "train_runtime": 154.1985,
      "train_tokens_per_second": 8119.197
    },
    {
      "epoch": 0.0020199299757608404,
      "grad_norm": 1.2246072868296953,
      "learning_rate": 9.999976481721066e-06,
      "loss": 0.49715890884399416,
      "num_input_tokens_seen": 1521544,
      "step": 30,
      "train_runtime": 185.9221,
      "train_tokens_per_second": 8183.77
    },
    {
      "epoch": 0.0023565849717209803,
      "grad_norm": 0.967162497026964,
      "learning_rate": 9.999967672862704e-06,
      "loss": 0.5369577884674073,
      "num_input_tokens_seen": 1766464,
      "step": 35,
      "train_runtime": 217.6999,
      "train_tokens_per_second": 8114.215
    },
    {
      "epoch": 0.00269323996768112,
      "grad_norm": 1.007217041879572,
      "learning_rate": 9.999957465779326e-06,
      "loss": 0.5171497344970704,
      "num_input_tokens_seen": 2003464,
      "step": 40,
      "train_runtime": 248.8593,
      "train_tokens_per_second": 8050.588
    },
    {
      "epoch": 0.0030298949636412605,
      "grad_norm": 1.0693422532991088,
      "learning_rate": 9.999945860473788e-06,
      "loss": 0.5321939945220947,
      "num_input_tokens_seen": 2227840,
      "step": 45,
      "train_runtime": 275.2799,
      "train_tokens_per_second": 8093.0
    },
    {
      "epoch": 0.0033665499596014004,
      "grad_norm": 0.8716906595548243,
      "learning_rate": 9.999932856949333e-06,
      "loss": 0.5009552478790283,
      "num_input_tokens_seen": 2492176,
      "step": 50,
      "train_runtime": 303.1086,
      "train_tokens_per_second": 8222.057
    },
    {
      "epoch": 0.0037032049555615404,
      "grad_norm": 0.8688986319111557,
      "learning_rate": 9.999918455209599e-06,
      "loss": 0.48927459716796873,
      "num_input_tokens_seen": 2769840,
      "step": 55,
      "train_runtime": 334.7211,
      "train_tokens_per_second": 8275.069
    },
    {
      "epoch": 0.004039859951521681,
      "grad_norm": 1.0357301270366783,
      "learning_rate": 9.99990265525861e-06,
      "loss": 0.5103227615356445,
      "num_input_tokens_seen": 3026456,
      "step": 60,
      "train_runtime": 365.9944,
      "train_tokens_per_second": 8269.133
    },
    {
      "epoch": 0.004376514947481821,
      "grad_norm": 1.094362374024335,
      "learning_rate": 9.999885457100789e-06,
      "loss": 0.5361778259277343,
      "num_input_tokens_seen": 3298480,
      "step": 65,
      "train_runtime": 400.0672,
      "train_tokens_per_second": 8244.814
    },
    {
      "epoch": 0.0047131699434419605,
      "grad_norm": 0.9628072964488236,
      "learning_rate": 9.999866860740943e-06,
      "loss": 0.4942628383636475,
      "num_input_tokens_seen": 3543504,
      "step": 70,
      "train_runtime": 432.5636,
      "train_tokens_per_second": 8191.868
    },
    {
      "epoch": 0.005049824939402101,
      "grad_norm": 1.107329098345885,
      "learning_rate": 9.999846866184274e-06,
      "loss": 0.517137098312378,
      "num_input_tokens_seen": 3820720,
      "step": 75,
      "train_runtime": 467.3875,
      "train_tokens_per_second": 8174.631
    },
    {
      "epoch": 0.00538647993536224,
      "grad_norm": 0.9176599751094567,
      "learning_rate": 9.99982547343637e-06,
      "loss": 0.488954496383667,
      "num_input_tokens_seen": 4044016,
      "step": 80,
      "train_runtime": 497.2019,
      "train_tokens_per_second": 8133.548
    },
    {
      "epoch": 0.005723134931322381,
      "grad_norm": 1.0170831790027959,
      "learning_rate": 9.999802682503217e-06,
      "loss": 0.46966943740844724,
      "num_input_tokens_seen": 4302400,
      "step": 85,
      "train_runtime": 528.2697,
      "train_tokens_per_second": 8144.325
    },
    {
      "epoch": 0.006059789927282521,
      "grad_norm": 0.8743435022610639,
      "learning_rate": 9.999778493391186e-06,
      "loss": 0.486570930480957,
      "num_input_tokens_seen": 4569792,
      "step": 90,
      "train_runtime": 561.3368,
      "train_tokens_per_second": 8140.91
    },
    {
      "epoch": 0.0063964449232426605,
      "grad_norm": 0.9441790271707035,
      "learning_rate": 9.999752906107043e-06,
      "loss": 0.5141637802124024,
      "num_input_tokens_seen": 4836152,
      "step": 95,
      "train_runtime": 591.4444,
      "train_tokens_per_second": 8176.85
    },
    {
      "epoch": 0.006733099919202801,
      "grad_norm": 1.0376672831473004,
      "learning_rate": 9.999725920657944e-06,
      "loss": 0.5164967536926269,
      "num_input_tokens_seen": 5086648,
      "step": 100,
      "train_runtime": 619.9272,
      "train_tokens_per_second": 8205.234
    },
    {
      "epoch": 0.007069754915162941,
      "grad_norm": 1.004153819574143,
      "learning_rate": 9.999697537051432e-06,
      "loss": 0.495100212097168,
      "num_input_tokens_seen": 5325248,
      "step": 105,
      "train_runtime": 649.8136,
      "train_tokens_per_second": 8195.04
    },
    {
      "epoch": 0.007406409911123081,
      "grad_norm": 1.0909152530318467,
      "learning_rate": 9.999667755295449e-06,
      "loss": 0.47860193252563477,
      "num_input_tokens_seen": 5574808,
      "step": 110,
      "train_runtime": 679.7499,
      "train_tokens_per_second": 8201.263
    },
    {
      "epoch": 0.007743064907083221,
      "grad_norm": 1.206437715541023,
      "learning_rate": 9.999636575398319e-06,
      "loss": 0.4889515399932861,
      "num_input_tokens_seen": 5822872,
      "step": 115,
      "train_runtime": 708.7104,
      "train_tokens_per_second": 8216.151
    },
    {
      "epoch": 0.008079719903043361,
      "grad_norm": 0.9595483061848518,
      "learning_rate": 9.999603997368766e-06,
      "loss": 0.42153682708740237,
      "num_input_tokens_seen": 6064504,
      "step": 120,
      "train_runtime": 740.173,
      "train_tokens_per_second": 8193.36
    },
    {
      "epoch": 0.008416374899003501,
      "grad_norm": 0.9867096961228317,
      "learning_rate": 9.999570021215895e-06,
      "loss": 0.46492815017700195,
      "num_input_tokens_seen": 6309816,
      "step": 125,
      "train_runtime": 771.9206,
      "train_tokens_per_second": 8174.178
    },
    {
      "epoch": 0.008753029894963642,
      "grad_norm": 1.057224526614425,
      "learning_rate": 9.99953464694921e-06,
      "loss": 0.480121374130249,
      "num_input_tokens_seen": 6544376,
      "step": 130,
      "train_runtime": 806.8311,
      "train_tokens_per_second": 8111.209
    },
    {
      "epoch": 0.009089684890923782,
      "grad_norm": 0.8374817320593494,
      "learning_rate": 9.999497874578606e-06,
      "loss": 0.479904842376709,
      "num_input_tokens_seen": 6815760,
      "step": 135,
      "train_runtime": 839.4275,
      "train_tokens_per_second": 8119.534
    },
    {
      "epoch": 0.009426339886883921,
      "grad_norm": 1.0674262110001975,
      "learning_rate": 9.999459704114361e-06,
      "loss": 0.49007296562194824,
      "num_input_tokens_seen": 7076656,
      "step": 140,
      "train_runtime": 866.5036,
      "train_tokens_per_second": 8166.909
    },
    {
      "epoch": 0.00976299488284406,
      "grad_norm": 2.3567685182114624,
      "learning_rate": 9.999420135567153e-06,
      "loss": 0.5142139434814453,
      "num_input_tokens_seen": 7324512,
      "step": 145,
      "train_runtime": 899.0251,
      "train_tokens_per_second": 8147.172
    },
    {
      "epoch": 0.010099649878804202,
      "grad_norm": 1.3503281569321224,
      "learning_rate": 9.999379168948047e-06,
      "loss": 0.50706787109375,
      "num_input_tokens_seen": 7588840,
      "step": 150,
      "train_runtime": 931.7772,
      "train_tokens_per_second": 8144.479
    },
    {
      "epoch": 0.010436304874764341,
      "grad_norm": 0.996622653073187,
      "learning_rate": 9.999336804268496e-06,
      "loss": 0.46264991760253904,
      "num_input_tokens_seen": 7840912,
      "step": 155,
      "train_runtime": 963.0988,
      "train_tokens_per_second": 8141.338
    },
    {
      "epoch": 0.01077295987072448,
      "grad_norm": 0.9327702685395796,
      "learning_rate": 9.999293041540351e-06,
      "loss": 0.487264347076416,
      "num_input_tokens_seen": 8078056,
      "step": 160,
      "train_runtime": 991.9895,
      "train_tokens_per_second": 8143.287
    },
    {
      "epoch": 0.011109614866684622,
      "grad_norm": 0.9476850423562907,
      "learning_rate": 9.999247880775849e-06,
      "loss": 0.48233590126037595,
      "num_input_tokens_seen": 8311224,
      "step": 165,
      "train_runtime": 1021.4487,
      "train_tokens_per_second": 8136.702
    },
    {
      "epoch": 0.011446269862644761,
      "grad_norm": 1.0006053765809395,
      "learning_rate": 9.999201321987617e-06,
      "loss": 0.4486554145812988,
      "num_input_tokens_seen": 8553712,
      "step": 170,
      "train_runtime": 1055.3552,
      "train_tokens_per_second": 8105.055
    },
    {
      "epoch": 0.011782924858604901,
      "grad_norm": 0.9154403948691177,
      "learning_rate": 9.999153365188676e-06,
      "loss": 0.5011663913726807,
      "num_input_tokens_seen": 8802216,
      "step": 175,
      "train_runtime": 1083.3108,
      "train_tokens_per_second": 8125.291
    },
    {
      "epoch": 0.012119579854565042,
      "grad_norm": 0.9149206133553959,
      "learning_rate": 9.99910401039244e-06,
      "loss": 0.44477195739746095,
      "num_input_tokens_seen": 9045200,
      "step": 180,
      "train_runtime": 1114.6348,
      "train_tokens_per_second": 8114.945
    },
    {
      "epoch": 0.012456234850525182,
      "grad_norm": 0.9634778866634535,
      "learning_rate": 9.999053257612708e-06,
      "loss": 0.455213737487793,
      "num_input_tokens_seen": 9322312,
      "step": 185,
      "train_runtime": 1141.629,
      "train_tokens_per_second": 8165.798
    },
    {
      "epoch": 0.012792889846485321,
      "grad_norm": 0.9499587950981273,
      "learning_rate": 9.999001106863672e-06,
      "loss": 0.5154810428619385,
      "num_input_tokens_seen": 9576240,
      "step": 190,
      "train_runtime": 1172.2045,
      "train_tokens_per_second": 8169.428
    },
    {
      "epoch": 0.013129544842445462,
      "grad_norm": 0.9102269577531555,
      "learning_rate": 9.998947558159916e-06,
      "loss": 0.4880210876464844,
      "num_input_tokens_seen": 9839848,
      "step": 195,
      "train_runtime": 1205.4183,
      "train_tokens_per_second": 8163.015
    },
    {
      "epoch": 0.013466199838405602,
      "grad_norm": 0.9531067086919549,
      "learning_rate": 9.99889261151642e-06,
      "loss": 0.4294342041015625,
      "num_input_tokens_seen": 10095912,
      "step": 200,
      "train_runtime": 1232.4753,
      "train_tokens_per_second": 8191.573
    },
    {
      "epoch": 0.013802854834365741,
      "grad_norm": 0.9552326060314473,
      "learning_rate": 9.998836266948542e-06,
      "loss": 0.4705368995666504,
      "num_input_tokens_seen": 10343200,
      "step": 205,
      "train_runtime": 1268.2009,
      "train_tokens_per_second": 8155.806
    },
    {
      "epoch": 0.014139509830325882,
      "grad_norm": 1.1419096227710246,
      "learning_rate": 9.998778524472043e-06,
      "loss": 0.43369522094726565,
      "num_input_tokens_seen": 10591232,
      "step": 210,
      "train_runtime": 1300.125,
      "train_tokens_per_second": 8146.319
    },
    {
      "epoch": 0.014476164826286022,
      "grad_norm": 1.0114914251642533,
      "learning_rate": 9.99871938410307e-06,
      "loss": 0.4530453681945801,
      "num_input_tokens_seen": 10850216,
      "step": 215,
      "train_runtime": 1333.4434,
      "train_tokens_per_second": 8136.99
    },
    {
      "epoch": 0.014812819822246161,
      "grad_norm": 0.880181413021885,
      "learning_rate": 9.99865884585816e-06,
      "loss": 0.4969773292541504,
      "num_input_tokens_seen": 11110232,
      "step": 220,
      "train_runtime": 1368.411,
      "train_tokens_per_second": 8119.076
    },
    {
      "epoch": 0.015149474818206303,
      "grad_norm": 0.9151536806190882,
      "learning_rate": 9.998596909754245e-06,
      "loss": 0.4789984703063965,
      "num_input_tokens_seen": 11367264,
      "step": 225,
      "train_runtime": 1399.9372,
      "train_tokens_per_second": 8119.839
    },
    {
      "epoch": 0.015486129814166442,
      "grad_norm": 1.005314625830885,
      "learning_rate": 9.998533575808642e-06,
      "loss": 0.4484864234924316,
      "num_input_tokens_seen": 11612704,
      "step": 230,
      "train_runtime": 1427.5765,
      "train_tokens_per_second": 8134.558
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 0.9272254915755589,
      "learning_rate": 9.998468844039064e-06,
      "loss": 0.4425339221954346,
      "num_input_tokens_seen": 11853528,
      "step": 235,
      "train_runtime": 1456.999,
      "train_tokens_per_second": 8135.578
    },
    {
      "epoch": 0.016159439806086723,
      "grad_norm": 0.8486553854858784,
      "learning_rate": 9.998402714463613e-06,
      "loss": 0.49678611755371094,
      "num_input_tokens_seen": 12123000,
      "step": 240,
      "train_runtime": 1486.0852,
      "train_tokens_per_second": 8157.675
    },
    {
      "epoch": 0.016496094802046862,
      "grad_norm": 0.8102165805370695,
      "learning_rate": 9.998335187100782e-06,
      "loss": 0.4884615421295166,
      "num_input_tokens_seen": 12370424,
      "step": 245,
      "train_runtime": 1513.1187,
      "train_tokens_per_second": 8175.448
    },
    {
      "epoch": 0.016832749798007002,
      "grad_norm": 0.9689436630796923,
      "learning_rate": 9.998266261969453e-06,
      "loss": 0.4717600345611572,
      "num_input_tokens_seen": 12618704,
      "step": 250,
      "train_runtime": 1547.6581,
      "train_tokens_per_second": 8153.418
    },
    {
      "epoch": 0.01716940479396714,
      "grad_norm": 0.8325781185766161,
      "learning_rate": 9.998195939088905e-06,
      "loss": 0.44782218933105467,
      "num_input_tokens_seen": 12902624,
      "step": 255,
      "train_runtime": 1582.5394,
      "train_tokens_per_second": 8153.114
    },
    {
      "epoch": 0.017506059789927284,
      "grad_norm": 0.8815418134728363,
      "learning_rate": 9.998124218478797e-06,
      "loss": 0.4844538688659668,
      "num_input_tokens_seen": 13150192,
      "step": 260,
      "train_runtime": 1613.5911,
      "train_tokens_per_second": 8149.643
    },
    {
      "epoch": 0.017842714785887424,
      "grad_norm": 0.818050745682155,
      "learning_rate": 9.998051100159192e-06,
      "loss": 0.461271858215332,
      "num_input_tokens_seen": 13426104,
      "step": 265,
      "train_runtime": 1643.3825,
      "train_tokens_per_second": 8169.799
    },
    {
      "epoch": 0.018179369781847563,
      "grad_norm": 0.9154862827643608,
      "learning_rate": 9.997976584150533e-06,
      "loss": 0.4711128234863281,
      "num_input_tokens_seen": 13683944,
      "step": 270,
      "train_runtime": 1670.3247,
      "train_tokens_per_second": 8192.386
    },
    {
      "epoch": 0.018516024777807703,
      "grad_norm": 0.7659672009557318,
      "learning_rate": 9.997900670473662e-06,
      "loss": 0.45070972442626955,
      "num_input_tokens_seen": 13960472,
      "step": 275,
      "train_runtime": 1699.5474,
      "train_tokens_per_second": 8214.229
    },
    {
      "epoch": 0.018852679773767842,
      "grad_norm": 0.8593763147468456,
      "learning_rate": 9.997823359149804e-06,
      "loss": 0.4660331726074219,
      "num_input_tokens_seen": 14221416,
      "step": 280,
      "train_runtime": 1731.7244,
      "train_tokens_per_second": 8212.286
    },
    {
      "epoch": 0.01918933476972798,
      "grad_norm": 0.8822300220908027,
      "learning_rate": 9.997744650200579e-06,
      "loss": 0.48664083480834963,
      "num_input_tokens_seen": 14482512,
      "step": 285,
      "train_runtime": 1759.0961,
      "train_tokens_per_second": 8232.928
    },
    {
      "epoch": 0.01952598976568812,
      "grad_norm": 0.9068046592298699,
      "learning_rate": 9.997664543648e-06,
      "loss": 0.46320815086364747,
      "num_input_tokens_seen": 14730096,
      "step": 290,
      "train_runtime": 1786.5889,
      "train_tokens_per_second": 8244.816
    },
    {
      "epoch": 0.019862644761648264,
      "grad_norm": 0.7986389157148576,
      "learning_rate": 9.997583039514469e-06,
      "loss": 0.4519622802734375,
      "num_input_tokens_seen": 14984736,
      "step": 295,
      "train_runtime": 1816.5718,
      "train_tokens_per_second": 8248.909
    },
    {
      "epoch": 0.020199299757608404,
      "grad_norm": 1.5807835204219811,
      "learning_rate": 9.997500137822776e-06,
      "loss": 0.4740860939025879,
      "num_input_tokens_seen": 15262304,
      "step": 300,
      "train_runtime": 1844.8323,
      "train_tokens_per_second": 8273.004
    },
    {
      "epoch": 0.020535954753568543,
      "grad_norm": 0.9457290976568937,
      "learning_rate": 9.997415838596106e-06,
      "loss": 0.4551679611206055,
      "num_input_tokens_seen": 15498720,
      "step": 305,
      "train_runtime": 1874.405,
      "train_tokens_per_second": 8268.608
    },
    {
      "epoch": 0.020872609749528682,
      "grad_norm": 0.8914041218990765,
      "learning_rate": 9.997330141858032e-06,
      "loss": 0.4780369758605957,
      "num_input_tokens_seen": 15760752,
      "step": 310,
      "train_runtime": 1902.5488,
      "train_tokens_per_second": 8284.02
    },
    {
      "epoch": 0.021209264745488822,
      "grad_norm": 1.2877965535552354,
      "learning_rate": 9.997243047632519e-06,
      "loss": 0.46323351860046386,
      "num_input_tokens_seen": 16013504,
      "step": 315,
      "train_runtime": 1934.3076,
      "train_tokens_per_second": 8278.675
    },
    {
      "epoch": 0.02154591974144896,
      "grad_norm": 0.8333018673513566,
      "learning_rate": 9.997154555943923e-06,
      "loss": 0.4365668296813965,
      "num_input_tokens_seen": 16266536,
      "step": 320,
      "train_runtime": 1962.8071,
      "train_tokens_per_second": 8287.384
    },
    {
      "epoch": 0.021882574737409104,
      "grad_norm": 1.5177972290977464,
      "learning_rate": 9.997064666816992e-06,
      "loss": 0.4556727409362793,
      "num_input_tokens_seen": 16520136,
      "step": 325,
      "train_runtime": 1995.2454,
      "train_tokens_per_second": 8279.751
    },
    {
      "epoch": 0.022219229733369244,
      "grad_norm": 1.031918659395949,
      "learning_rate": 9.996973380276858e-06,
      "loss": 0.5267070770263672,
      "num_input_tokens_seen": 16798064,
      "step": 330,
      "train_runtime": 2028.3129,
      "train_tokens_per_second": 8281.791
    },
    {
      "epoch": 0.022555884729329383,
      "grad_norm": 1.1130862528330914,
      "learning_rate": 9.996880696349053e-06,
      "loss": 0.43999366760253905,
      "num_input_tokens_seen": 17056688,
      "step": 335,
      "train_runtime": 2058.4759,
      "train_tokens_per_second": 8286.076
    },
    {
      "epoch": 0.022892539725289523,
      "grad_norm": 0.9665000438493415,
      "learning_rate": 9.996786615059497e-06,
      "loss": 0.5216714382171631,
      "num_input_tokens_seen": 17310792,
      "step": 340,
      "train_runtime": 2091.9533,
      "train_tokens_per_second": 8274.942
    },
    {
      "epoch": 0.023229194721249662,
      "grad_norm": 0.8668593444026271,
      "learning_rate": 9.996691136434498e-06,
      "loss": 0.4200946807861328,
      "num_input_tokens_seen": 17579456,
      "step": 345,
      "train_runtime": 2120.6579,
      "train_tokens_per_second": 8289.624
    },
    {
      "epoch": 0.023565849717209802,
      "grad_norm": 0.7961527825821757,
      "learning_rate": 9.996594260500753e-06,
      "loss": 0.49302024841308595,
      "num_input_tokens_seen": 17847440,
      "step": 350,
      "train_runtime": 2154.931,
      "train_tokens_per_second": 8282.14
    },
    {
      "epoch": 0.023902504713169945,
      "grad_norm": 0.9896352122271752,
      "learning_rate": 9.996495987285358e-06,
      "loss": 0.447528076171875,
      "num_input_tokens_seen": 18096840,
      "step": 355,
      "train_runtime": 2184.3527,
      "train_tokens_per_second": 8284.761
    },
    {
      "epoch": 0.024239159709130084,
      "grad_norm": 1.0911199704261594,
      "learning_rate": 9.996396316815793e-06,
      "loss": 0.46645488739013674,
      "num_input_tokens_seen": 18349440,
      "step": 360,
      "train_runtime": 2213.8156,
      "train_tokens_per_second": 8288.604
    },
    {
      "epoch": 0.024575814705090224,
      "grad_norm": 1.002126016267506,
      "learning_rate": 9.99629524911993e-06,
      "loss": 0.4479473114013672,
      "num_input_tokens_seen": 18609784,
      "step": 365,
      "train_runtime": 2241.9974,
      "train_tokens_per_second": 8300.538
    },
    {
      "epoch": 0.024912469701050363,
      "grad_norm": 0.9237818044489297,
      "learning_rate": 9.996192784226033e-06,
      "loss": 0.4305563926696777,
      "num_input_tokens_seen": 18842344,
      "step": 370,
      "train_runtime": 2274.4711,
      "train_tokens_per_second": 8284.275
    },
    {
      "epoch": 0.025249124697010503,
      "grad_norm": 0.9125126321156503,
      "learning_rate": 9.996088922162755e-06,
      "loss": 0.44871225357055666,
      "num_input_tokens_seen": 19130424,
      "step": 375,
      "train_runtime": 2305.9475,
      "train_tokens_per_second": 8296.123
    },
    {
      "epoch": 0.025585779692970642,
      "grad_norm": 1.0779630864313732,
      "learning_rate": 9.995983662959141e-06,
      "loss": 0.457213020324707,
      "num_input_tokens_seen": 19380112,
      "step": 380,
      "train_runtime": 2339.7827,
      "train_tokens_per_second": 8282.868
    },
    {
      "epoch": 0.025922434688930785,
      "grad_norm": 0.8795324831088377,
      "learning_rate": 9.995877006644627e-06,
      "loss": 0.4747770309448242,
      "num_input_tokens_seen": 19659832,
      "step": 385,
      "train_runtime": 2371.1182,
      "train_tokens_per_second": 8291.376
    },
    {
      "epoch": 0.026259089684890925,
      "grad_norm": 1.2733371419860156,
      "learning_rate": 9.995768953249038e-06,
      "loss": 0.4678105354309082,
      "num_input_tokens_seen": 19927752,
      "step": 390,
      "train_runtime": 2401.4599,
      "train_tokens_per_second": 8298.182
    },
    {
      "epoch": 0.026595744680851064,
      "grad_norm": 1.0771192490792671,
      "learning_rate": 9.995659502802594e-06,
      "loss": 0.43525123596191406,
      "num_input_tokens_seen": 20185416,
      "step": 395,
      "train_runtime": 2427.7823,
      "train_tokens_per_second": 8314.343
    },
    {
      "epoch": 0.026932399676811204,
      "grad_norm": 0.9456577836805365,
      "learning_rate": 9.995548655335899e-06,
      "loss": 0.4762571334838867,
      "num_input_tokens_seen": 20420368,
      "step": 400,
      "train_runtime": 2457.5364,
      "train_tokens_per_second": 8309.284
    },
    {
      "epoch": 0.027269054672771343,
      "grad_norm": 0.9613263954000164,
      "learning_rate": 9.995436410879951e-06,
      "loss": 0.4969775199890137,
      "num_input_tokens_seen": 20695408,
      "step": 405,
      "train_runtime": 2491.0116,
      "train_tokens_per_second": 8308.034
    },
    {
      "epoch": 0.027605709668731482,
      "grad_norm": 0.8721036755115033,
      "learning_rate": 9.99532276946614e-06,
      "loss": 0.4239232540130615,
      "num_input_tokens_seen": 20943248,
      "step": 410,
      "train_runtime": 2521.0302,
      "train_tokens_per_second": 8307.417
    },
    {
      "epoch": 0.027942364664691625,
      "grad_norm": 0.8535437979381594,
      "learning_rate": 9.995207731126246e-06,
      "loss": 0.4835047721862793,
      "num_input_tokens_seen": 21214136,
      "step": 415,
      "train_runtime": 2554.1967,
      "train_tokens_per_second": 8305.6
    },
    {
      "epoch": 0.028279019660651765,
      "grad_norm": 0.8093584091690936,
      "learning_rate": 9.99509129589244e-06,
      "loss": 0.44570207595825195,
      "num_input_tokens_seen": 21460264,
      "step": 420,
      "train_runtime": 2585.7759,
      "train_tokens_per_second": 8299.352
    },
    {
      "epoch": 0.028615674656611904,
      "grad_norm": 0.8584569526937949,
      "learning_rate": 9.994973463797279e-06,
      "loss": 0.4463783264160156,
      "num_input_tokens_seen": 21728504,
      "step": 425,
      "train_runtime": 2620.1859,
      "train_tokens_per_second": 8292.734
    },
    {
      "epoch": 0.028952329652572044,
      "grad_norm": 0.9111710595238781,
      "learning_rate": 9.99485423487372e-06,
      "loss": 0.42282819747924805,
      "num_input_tokens_seen": 21953136,
      "step": 430,
      "train_runtime": 2647.289,
      "train_tokens_per_second": 8292.686
    },
    {
      "epoch": 0.029288984648532183,
      "grad_norm": 0.9256101679008836,
      "learning_rate": 9.994733609155097e-06,
      "loss": 0.46303110122680663,
      "num_input_tokens_seen": 22198064,
      "step": 435,
      "train_runtime": 2682.6492,
      "train_tokens_per_second": 8274.68
    },
    {
      "epoch": 0.029625639644492323,
      "grad_norm": 0.9311365802167839,
      "learning_rate": 9.994611586675149e-06,
      "loss": 0.4240147590637207,
      "num_input_tokens_seen": 22473912,
      "step": 440,
      "train_runtime": 2714.8989,
      "train_tokens_per_second": 8277.992
    },
    {
      "epoch": 0.029962294640452466,
      "grad_norm": 1.0309604350680803,
      "learning_rate": 9.994488167468e-06,
      "loss": 0.4294229030609131,
      "num_input_tokens_seen": 22731320,
      "step": 445,
      "train_runtime": 2742.1003,
      "train_tokens_per_second": 8289.748
    },
    {
      "epoch": 0.030298949636412605,
      "grad_norm": 0.9146379679611907,
      "learning_rate": 9.994363351568158e-06,
      "loss": 0.467713737487793,
      "num_input_tokens_seen": 22983176,
      "step": 450,
      "train_runtime": 2771.929,
      "train_tokens_per_second": 8291.401
    },
    {
      "epoch": 0.030635604632372745,
      "grad_norm": 0.7536837429642986,
      "learning_rate": 9.994237139010532e-06,
      "loss": 0.42033610343933103,
      "num_input_tokens_seen": 23218120,
      "step": 455,
      "train_runtime": 2804.0237,
      "train_tokens_per_second": 8280.287
    },
    {
      "epoch": 0.030972259628332884,
      "grad_norm": 0.9184771084419948,
      "learning_rate": 9.994109529830414e-06,
      "loss": 0.46703596115112306,
      "num_input_tokens_seen": 23497752,
      "step": 460,
      "train_runtime": 2834.2872,
      "train_tokens_per_second": 8290.533
    },
    {
      "epoch": 0.03130891462429303,
      "grad_norm": 0.6890643044891042,
      "learning_rate": 9.993980524063494e-06,
      "loss": 0.42495107650756836,
      "num_input_tokens_seen": 23767984,
      "step": 465,
      "train_runtime": 2861.5776,
      "train_tokens_per_second": 8305.902
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 0.832343252188738,
      "learning_rate": 9.993850121745842e-06,
      "loss": 0.44729032516479494,
      "num_input_tokens_seen": 24026728,
      "step": 470,
      "train_runtime": 2891.9437,
      "train_tokens_per_second": 8308.159
    },
    {
      "epoch": 0.031982224616213306,
      "grad_norm": 0.8578622699663953,
      "learning_rate": 9.99371832291393e-06,
      "loss": 0.4422321796417236,
      "num_input_tokens_seen": 24272592,
      "step": 475,
      "train_runtime": 2924.4544,
      "train_tokens_per_second": 8299.87
    },
    {
      "epoch": 0.032318879612173446,
      "grad_norm": 1.0041325148031799,
      "learning_rate": 9.993585127604612e-06,
      "loss": 0.4317974090576172,
      "num_input_tokens_seen": 24513872,
      "step": 480,
      "train_runtime": 2960.0943,
      "train_tokens_per_second": 8281.45
    },
    {
      "epoch": 0.032655534608133585,
      "grad_norm": 0.8098548208751043,
      "learning_rate": 9.993450535855136e-06,
      "loss": 0.44673452377319334,
      "num_input_tokens_seen": 24756024,
      "step": 485,
      "train_runtime": 2993.3653,
      "train_tokens_per_second": 8270.298
    },
    {
      "epoch": 0.032992189604093725,
      "grad_norm": 0.8421248331986749,
      "learning_rate": 9.993314547703142e-06,
      "loss": 0.42482285499572753,
      "num_input_tokens_seen": 25009952,
      "step": 490,
      "train_runtime": 3023.6724,
      "train_tokens_per_second": 8271.383
    },
    {
      "epoch": 0.033328844600053864,
      "grad_norm": 0.8536026422226947,
      "learning_rate": 9.993177163186655e-06,
      "loss": 0.44283294677734375,
      "num_input_tokens_seen": 25259040,
      "step": 495,
      "train_runtime": 3057.8953,
      "train_tokens_per_second": 8260.27
    },
    {
      "epoch": 0.033665499596014004,
      "grad_norm": 0.940216615669213,
      "learning_rate": 9.993038382344097e-06,
      "loss": 0.42020812034606936,
      "num_input_tokens_seen": 25501808,
      "step": 500,
      "train_runtime": 3088.7586,
      "train_tokens_per_second": 8256.329
    },
    {
      "epoch": 0.03400215459197414,
      "grad_norm": 0.8554522448912235,
      "learning_rate": 9.992898205214279e-06,
      "loss": 0.4436030387878418,
      "num_input_tokens_seen": 25762616,
      "step": 505,
      "train_runtime": 3119.0839,
      "train_tokens_per_second": 8259.674
    },
    {
      "epoch": 0.03433880958793428,
      "grad_norm": 0.9340489729372833,
      "learning_rate": 9.992756631836397e-06,
      "loss": 0.4472458839416504,
      "num_input_tokens_seen": 26015680,
      "step": 510,
      "train_runtime": 3149.5513,
      "train_tokens_per_second": 8260.123
    },
    {
      "epoch": 0.03467546458389442,
      "grad_norm": 0.7819033786030177,
      "learning_rate": 9.992613662250044e-06,
      "loss": 0.43979463577270506,
      "num_input_tokens_seen": 26277416,
      "step": 515,
      "train_runtime": 3182.5309,
      "train_tokens_per_second": 8256.767
    },
    {
      "epoch": 0.03501211957985457,
      "grad_norm": 1.6863482900130655,
      "learning_rate": 9.992469296495201e-06,
      "loss": 0.4432481288909912,
      "num_input_tokens_seen": 26545824,
      "step": 520,
      "train_runtime": 3213.3783,
      "train_tokens_per_second": 8261.033
    },
    {
      "epoch": 0.03534877457581471,
      "grad_norm": 0.7820476704960201,
      "learning_rate": 9.992323534612238e-06,
      "loss": 0.41016135215759275,
      "num_input_tokens_seen": 26795936,
      "step": 525,
      "train_runtime": 3244.0026,
      "train_tokens_per_second": 8260.146
    },
    {
      "epoch": 0.03568542957177485,
      "grad_norm": 0.8042886832278499,
      "learning_rate": 9.99217637664192e-06,
      "loss": 0.4433448791503906,
      "num_input_tokens_seen": 27055000,
      "step": 530,
      "train_runtime": 3276.1751,
      "train_tokens_per_second": 8258.106
    },
    {
      "epoch": 0.03602208456773499,
      "grad_norm": 0.8272913904487212,
      "learning_rate": 9.992027822625397e-06,
      "loss": 0.4453550338745117,
      "num_input_tokens_seen": 27297128,
      "step": 535,
      "train_runtime": 3303.5962,
      "train_tokens_per_second": 8262.852
    },
    {
      "epoch": 0.036358739563695126,
      "grad_norm": 0.9043955725253953,
      "learning_rate": 9.99187787260421e-06,
      "loss": 0.4626297950744629,
      "num_input_tokens_seen": 27546760,
      "step": 540,
      "train_runtime": 3330.945,
      "train_tokens_per_second": 8269.953
    },
    {
      "epoch": 0.036695394559655266,
      "grad_norm": 0.7770682506621583,
      "learning_rate": 9.991726526620294e-06,
      "loss": 0.4282529354095459,
      "num_input_tokens_seen": 27838392,
      "step": 545,
      "train_runtime": 3364.3478,
      "train_tokens_per_second": 8274.528
    },
    {
      "epoch": 0.037032049555615405,
      "grad_norm": 0.8135023901307253,
      "learning_rate": 9.991573784715973e-06,
      "loss": 0.45714287757873534,
      "num_input_tokens_seen": 28081600,
      "step": 550,
      "train_runtime": 3395.389,
      "train_tokens_per_second": 8270.51
    },
    {
      "epoch": 0.037368704551575545,
      "grad_norm": 0.9718506266664045,
      "learning_rate": 9.99141964693396e-06,
      "loss": 0.47224225997924807,
      "num_input_tokens_seen": 28339944,
      "step": 555,
      "train_runtime": 3429.1842,
      "train_tokens_per_second": 8264.34
    },
    {
      "epoch": 0.037705359547535684,
      "grad_norm": 0.9375000762154142,
      "learning_rate": 9.99126411331736e-06,
      "loss": 0.43854055404663084,
      "num_input_tokens_seen": 28597872,
      "step": 560,
      "train_runtime": 3461.8335,
      "train_tokens_per_second": 8260.903
    },
    {
      "epoch": 0.038042014543495824,
      "grad_norm": 0.8337877797826708,
      "learning_rate": 9.991107183909665e-06,
      "loss": 0.4600650787353516,
      "num_input_tokens_seen": 28866216,
      "step": 565,
      "train_runtime": 3498.1393,
      "train_tokens_per_second": 8251.877
    },
    {
      "epoch": 0.03837866953945596,
      "grad_norm": 0.8800628690261549,
      "learning_rate": 9.990948858754763e-06,
      "loss": 0.4492210388183594,
      "num_input_tokens_seen": 29098584,
      "step": 570,
      "train_runtime": 3532.3257,
      "train_tokens_per_second": 8237.798
    },
    {
      "epoch": 0.0387153245354161,
      "grad_norm": 0.7397073663108815,
      "learning_rate": 9.990789137896927e-06,
      "loss": 0.4599879741668701,
      "num_input_tokens_seen": 29361616,
      "step": 575,
      "train_runtime": 3558.5992,
      "train_tokens_per_second": 8250.891
    },
    {
      "epoch": 0.03905197953137624,
      "grad_norm": 0.889333022759583,
      "learning_rate": 9.990628021380823e-06,
      "loss": 0.45560779571533205,
      "num_input_tokens_seen": 29617696,
      "step": 580,
      "train_runtime": 3593.8841,
      "train_tokens_per_second": 8241.138
    },
    {
      "epoch": 0.03938863452733639,
      "grad_norm": 1.232828132033305,
      "learning_rate": 9.990465509251508e-06,
      "loss": 0.4456137180328369,
      "num_input_tokens_seen": 29881880,
      "step": 585,
      "train_runtime": 3617.697,
      "train_tokens_per_second": 8259.918
    },
    {
      "epoch": 0.03972528952329653,
      "grad_norm": 0.9120493131640759,
      "learning_rate": 9.990301601554425e-06,
      "loss": 0.4629720687866211,
      "num_input_tokens_seen": 30106600,
      "step": 590,
      "train_runtime": 3647.7351,
      "train_tokens_per_second": 8253.505
    },
    {
      "epoch": 0.04006194451925667,
      "grad_norm": 0.834101012569425,
      "learning_rate": 9.990136298335414e-06,
      "loss": 0.39983794689178465,
      "num_input_tokens_seen": 30365064,
      "step": 595,
      "train_runtime": 3677.8686,
      "train_tokens_per_second": 8256.158
    },
    {
      "epoch": 0.04039859951521681,
      "grad_norm": 0.8424491008969366,
      "learning_rate": 9.989969599640698e-06,
      "loss": 0.41225323677062986,
      "num_input_tokens_seen": 30644872,
      "step": 600,
      "train_runtime": 3710.6899,
      "train_tokens_per_second": 8258.538
    },
    {
      "epoch": 0.040735254511176947,
      "grad_norm": 0.8850617310396283,
      "learning_rate": 9.989801505516895e-06,
      "loss": 0.4890312194824219,
      "num_input_tokens_seen": 30922752,
      "step": 605,
      "train_runtime": 3745.5512,
      "train_tokens_per_second": 8255.861
    },
    {
      "epoch": 0.041071909507137086,
      "grad_norm": 0.85203946043574,
      "learning_rate": 9.989632016011012e-06,
      "loss": 0.4639169692993164,
      "num_input_tokens_seen": 31187696,
      "step": 610,
      "train_runtime": 3777.3003,
      "train_tokens_per_second": 8256.61
    },
    {
      "epoch": 0.041408564503097225,
      "grad_norm": 1.3512099680322465,
      "learning_rate": 9.989461131170448e-06,
      "loss": 0.4730778694152832,
      "num_input_tokens_seen": 31410488,
      "step": 615,
      "train_runtime": 3803.6829,
      "train_tokens_per_second": 8257.914
    },
    {
      "epoch": 0.041745219499057365,
      "grad_norm": 0.8031180473102617,
      "learning_rate": 9.98928885104299e-06,
      "loss": 0.44115333557128905,
      "num_input_tokens_seen": 31662440,
      "step": 620,
      "train_runtime": 3840.0905,
      "train_tokens_per_second": 8245.233
    },
    {
      "epoch": 0.042081874495017504,
      "grad_norm": 0.7680598395727002,
      "learning_rate": 9.989115175676812e-06,
      "loss": 0.4178950309753418,
      "num_input_tokens_seen": 31933096,
      "step": 625,
      "train_runtime": 3869.9727,
      "train_tokens_per_second": 8251.504
    },
    {
      "epoch": 0.042418529490977644,
      "grad_norm": 0.8784087026320627,
      "learning_rate": 9.988940105120483e-06,
      "loss": 0.46561431884765625,
      "num_input_tokens_seen": 32168584,
      "step": 630,
      "train_runtime": 3899.8768,
      "train_tokens_per_second": 8248.615
    },
    {
      "epoch": 0.04275518448693778,
      "grad_norm": 0.8366731811125396,
      "learning_rate": 9.988763639422965e-06,
      "loss": 0.46678800582885743,
      "num_input_tokens_seen": 32442616,
      "step": 635,
      "train_runtime": 3929.4963,
      "train_tokens_per_second": 8256.177
    },
    {
      "epoch": 0.04309183948289792,
      "grad_norm": 0.9476931134629513,
      "learning_rate": 9.988585778633604e-06,
      "loss": 0.4279820919036865,
      "num_input_tokens_seen": 32675440,
      "step": 640,
      "train_runtime": 3963.2792,
      "train_tokens_per_second": 8244.547
    },
    {
      "epoch": 0.04342849447885807,
      "grad_norm": 0.8798377345919048,
      "learning_rate": 9.988406522802134e-06,
      "loss": 0.44605340957641604,
      "num_input_tokens_seen": 32919464,
      "step": 645,
      "train_runtime": 3996.8474,
      "train_tokens_per_second": 8236.357
    },
    {
      "epoch": 0.04376514947481821,
      "grad_norm": 0.8648076853686417,
      "learning_rate": 9.98822587197869e-06,
      "loss": 0.41315131187438964,
      "num_input_tokens_seen": 33177400,
      "step": 650,
      "train_runtime": 4023.6794,
      "train_tokens_per_second": 8245.538
    },
    {
      "epoch": 0.04410180447077835,
      "grad_norm": 0.8594019998599847,
      "learning_rate": 9.988043826213785e-06,
      "loss": 0.425563907623291,
      "num_input_tokens_seen": 33423360,
      "step": 655,
      "train_runtime": 4052.1788,
      "train_tokens_per_second": 8248.244
    },
    {
      "epoch": 0.04443845946673849,
      "grad_norm": 0.880253134696758,
      "learning_rate": 9.98786038555833e-06,
      "loss": 0.41025876998901367,
      "num_input_tokens_seen": 33692664,
      "step": 660,
      "train_runtime": 4083.4335,
      "train_tokens_per_second": 8251.062
    },
    {
      "epoch": 0.04477511446269863,
      "grad_norm": 0.7514462728553865,
      "learning_rate": 9.987675550063625e-06,
      "loss": 0.44844589233398435,
      "num_input_tokens_seen": 33964368,
      "step": 665,
      "train_runtime": 4117.4951,
      "train_tokens_per_second": 8248.794
    },
    {
      "epoch": 0.04511176945865877,
      "grad_norm": 0.9453705013206156,
      "learning_rate": 9.987489319781355e-06,
      "loss": 0.4434016704559326,
      "num_input_tokens_seen": 34230992,
      "step": 670,
      "train_runtime": 4148.0647,
      "train_tokens_per_second": 8252.28
    },
    {
      "epoch": 0.045448424454618906,
      "grad_norm": 0.7595208498480539,
      "learning_rate": 9.987301694763602e-06,
      "loss": 0.4736198902130127,
      "num_input_tokens_seen": 34486256,
      "step": 675,
      "train_runtime": 4180.6038,
      "train_tokens_per_second": 8249.109
    },
    {
      "epoch": 0.045785079450579046,
      "grad_norm": 0.7703016939116394,
      "learning_rate": 9.987112675062834e-06,
      "loss": 0.39811923503875735,
      "num_input_tokens_seen": 34734840,
      "step": 680,
      "train_runtime": 4214.532,
      "train_tokens_per_second": 8241.684
    },
    {
      "epoch": 0.046121734446539185,
      "grad_norm": 0.975366424682886,
      "learning_rate": 9.986922260731908e-06,
      "loss": 0.42193756103515623,
      "num_input_tokens_seen": 34978448,
      "step": 685,
      "train_runtime": 4244.296,
      "train_tokens_per_second": 8241.284
    },
    {
      "epoch": 0.046458389442499325,
      "grad_norm": 0.7837136895418859,
      "learning_rate": 9.986730451824073e-06,
      "loss": 0.4591407299041748,
      "num_input_tokens_seen": 35213016,
      "step": 690,
      "train_runtime": 4274.4771,
      "train_tokens_per_second": 8237.97
    },
    {
      "epoch": 0.046795044438459464,
      "grad_norm": 0.8253315647260433,
      "learning_rate": 9.986537248392969e-06,
      "loss": 0.45913124084472656,
      "num_input_tokens_seen": 35468448,
      "step": 695,
      "train_runtime": 4306.8114,
      "train_tokens_per_second": 8235.431
    },
    {
      "epoch": 0.047131699434419604,
      "grad_norm": 0.8548173630714615,
      "learning_rate": 9.986342650492626e-06,
      "loss": 0.433709716796875,
      "num_input_tokens_seen": 35729496,
      "step": 700,
      "train_runtime": 4339.1721,
      "train_tokens_per_second": 8234.173
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.7697252758565,
      "learning_rate": 9.98614665817746e-06,
      "loss": 0.41391630172729493,
      "num_input_tokens_seen": 35954464,
      "step": 705,
      "train_runtime": 4371.6269,
      "train_tokens_per_second": 8224.504
    },
    {
      "epoch": 0.04780500942633989,
      "grad_norm": 0.8038626787215232,
      "learning_rate": 9.98594927150228e-06,
      "loss": 0.4218019485473633,
      "num_input_tokens_seen": 36234152,
      "step": 710,
      "train_runtime": 4405.7911,
      "train_tokens_per_second": 8224.21
    },
    {
      "epoch": 0.04814166442230003,
      "grad_norm": 0.8273105170702114,
      "learning_rate": 9.985750490522286e-06,
      "loss": 0.3976966142654419,
      "num_input_tokens_seen": 36485376,
      "step": 715,
      "train_runtime": 4432.0047,
      "train_tokens_per_second": 8232.251
    },
    {
      "epoch": 0.04847831941826017,
      "grad_norm": 0.9272904611553914,
      "learning_rate": 9.985550315293066e-06,
      "loss": 0.46237959861755373,
      "num_input_tokens_seen": 36750808,
      "step": 720,
      "train_runtime": 4463.5911,
      "train_tokens_per_second": 8233.462
    },
    {
      "epoch": 0.04881497441422031,
      "grad_norm": 0.785763359771224,
      "learning_rate": 9.985348745870598e-06,
      "loss": 0.44240965843200686,
      "num_input_tokens_seen": 37031488,
      "step": 725,
      "train_runtime": 4495.5139,
      "train_tokens_per_second": 8237.431
    },
    {
      "epoch": 0.04915162941018045,
      "grad_norm": 0.6773603779900301,
      "learning_rate": 9.98514578231125e-06,
      "loss": 0.4050912857055664,
      "num_input_tokens_seen": 37307288,
      "step": 730,
      "train_runtime": 4527.9327,
      "train_tokens_per_second": 8239.364
    },
    {
      "epoch": 0.04948828440614059,
      "grad_norm": 0.8808712034560708,
      "learning_rate": 9.984941424671781e-06,
      "loss": 0.45537652969360354,
      "num_input_tokens_seen": 37557960,
      "step": 735,
      "train_runtime": 4555.1404,
      "train_tokens_per_second": 8245.182
    },
    {
      "epoch": 0.049824939402100726,
      "grad_norm": 0.7938259406398527,
      "learning_rate": 9.984735673009337e-06,
      "loss": 0.4165522575378418,
      "num_input_tokens_seen": 37808184,
      "step": 740,
      "train_runtime": 4585.2253,
      "train_tokens_per_second": 8245.655
    },
    {
      "epoch": 0.050161594398060866,
      "grad_norm": 0.8090886000091149,
      "learning_rate": 9.984528527381458e-06,
      "loss": 0.4253381729125977,
      "num_input_tokens_seen": 38063400,
      "step": 745,
      "train_runtime": 4618.0598,
      "train_tokens_per_second": 8242.292
    },
    {
      "epoch": 0.050498249394021005,
      "grad_norm": 0.7990883532468535,
      "learning_rate": 9.984319987846072e-06,
      "loss": 0.44441680908203124,
      "num_input_tokens_seen": 38303272,
      "step": 750,
      "train_runtime": 4644.9976,
      "train_tokens_per_second": 8246.134
    },
    {
      "epoch": 0.050834904389981145,
      "grad_norm": 0.9604121980001694,
      "learning_rate": 9.984110054461495e-06,
      "loss": 0.42880849838256835,
      "num_input_tokens_seen": 38592816,
      "step": 755,
      "train_runtime": 4678.5541,
      "train_tokens_per_second": 8248.877
    },
    {
      "epoch": 0.051171559385941284,
      "grad_norm": 0.8495083389051551,
      "learning_rate": 9.983898727286435e-06,
      "loss": 0.43865156173706055,
      "num_input_tokens_seen": 38850160,
      "step": 760,
      "train_runtime": 4706.1039,
      "train_tokens_per_second": 8255.27
    },
    {
      "epoch": 0.05150821438190143,
      "grad_norm": 0.8589616724081981,
      "learning_rate": 9.983686006379985e-06,
      "loss": 0.43404226303100585,
      "num_input_tokens_seen": 39104512,
      "step": 765,
      "train_runtime": 4735.9031,
      "train_tokens_per_second": 8257.034
    },
    {
      "epoch": 0.05184486937786157,
      "grad_norm": 0.9140959685348827,
      "learning_rate": 9.983471891801638e-06,
      "loss": 0.410258150100708,
      "num_input_tokens_seen": 39370704,
      "step": 770,
      "train_runtime": 4767.7006,
      "train_tokens_per_second": 8257.797
    },
    {
      "epoch": 0.05218152437382171,
      "grad_norm": 0.7679760920838427,
      "learning_rate": 9.983256383611267e-06,
      "loss": 0.43230419158935546,
      "num_input_tokens_seen": 39633960,
      "step": 775,
      "train_runtime": 4800.7333,
      "train_tokens_per_second": 8255.814
    },
    {
      "epoch": 0.05251817936978185,
      "grad_norm": 0.8149363339130661,
      "learning_rate": 9.983039481869138e-06,
      "loss": 0.4209223747253418,
      "num_input_tokens_seen": 39891616,
      "step": 780,
      "train_runtime": 4834.4102,
      "train_tokens_per_second": 8251.599
    },
    {
      "epoch": 0.05285483436574199,
      "grad_norm": 0.9056767091906156,
      "learning_rate": 9.98282118663591e-06,
      "loss": 0.4073034286499023,
      "num_input_tokens_seen": 40152016,
      "step": 785,
      "train_runtime": 4861.8603,
      "train_tokens_per_second": 8258.571
    },
    {
      "epoch": 0.05319148936170213,
      "grad_norm": 0.8557950379552083,
      "learning_rate": 9.982601497972624e-06,
      "loss": 0.45415463447570803,
      "num_input_tokens_seen": 40413744,
      "step": 790,
      "train_runtime": 4893.1312,
      "train_tokens_per_second": 8259.281
    },
    {
      "epoch": 0.05352814435766227,
      "grad_norm": 0.7961146428374651,
      "learning_rate": 9.982380415940717e-06,
      "loss": 0.42005047798156736,
      "num_input_tokens_seen": 40659480,
      "step": 795,
      "train_runtime": 4923.0308,
      "train_tokens_per_second": 8259.034
    },
    {
      "epoch": 0.05386479935362241,
      "grad_norm": 0.8134347239682204,
      "learning_rate": 9.982157940602015e-06,
      "loss": 0.4501035690307617,
      "num_input_tokens_seen": 40936760,
      "step": 800,
      "train_runtime": 4953.8143,
      "train_tokens_per_second": 8263.685
    },
    {
      "epoch": 0.05420145434958255,
      "grad_norm": 0.8702599280837396,
      "learning_rate": 9.981934072018732e-06,
      "loss": 0.4244490146636963,
      "num_input_tokens_seen": 41228728,
      "step": 805,
      "train_runtime": 4984.2814,
      "train_tokens_per_second": 8271.75
    },
    {
      "epoch": 0.054538109345542686,
      "grad_norm": 0.8733965867797476,
      "learning_rate": 9.981708810253472e-06,
      "loss": 0.4621896266937256,
      "num_input_tokens_seen": 41484072,
      "step": 810,
      "train_runtime": 5014.5205,
      "train_tokens_per_second": 8272.789
    },
    {
      "epoch": 0.054874764341502826,
      "grad_norm": 0.9425435622417606,
      "learning_rate": 9.981482155369229e-06,
      "loss": 0.4334461212158203,
      "num_input_tokens_seen": 41754768,
      "step": 815,
      "train_runtime": 5049.1957,
      "train_tokens_per_second": 8269.588
    },
    {
      "epoch": 0.055211419337462965,
      "grad_norm": 0.8095686998774289,
      "learning_rate": 9.981254107429385e-06,
      "loss": 0.40474438667297363,
      "num_input_tokens_seen": 41998664,
      "step": 820,
      "train_runtime": 5081.7858,
      "train_tokens_per_second": 8264.548
    },
    {
      "epoch": 0.05554807433342311,
      "grad_norm": 0.7865838509987403,
      "learning_rate": 9.981024666497716e-06,
      "loss": 0.40341939926147463,
      "num_input_tokens_seen": 42262672,
      "step": 825,
      "train_runtime": 5111.8206,
      "train_tokens_per_second": 8267.636
    },
    {
      "epoch": 0.05588472932938325,
      "grad_norm": 0.899005421287083,
      "learning_rate": 9.98079383263838e-06,
      "loss": 0.44587202072143556,
      "num_input_tokens_seen": 42501696,
      "step": 830,
      "train_runtime": 5142.7222,
      "train_tokens_per_second": 8264.436
    },
    {
      "epoch": 0.05622138432534339,
      "grad_norm": 0.8599365102515701,
      "learning_rate": 9.980561605915932e-06,
      "loss": 0.42188568115234376,
      "num_input_tokens_seen": 42774576,
      "step": 835,
      "train_runtime": 5172.3887,
      "train_tokens_per_second": 8269.791
    },
    {
      "epoch": 0.05655803932130353,
      "grad_norm": 0.7320432465055605,
      "learning_rate": 9.980327986395316e-06,
      "loss": 0.4249128341674805,
      "num_input_tokens_seen": 43028736,
      "step": 840,
      "train_runtime": 5200.2323,
      "train_tokens_per_second": 8274.387
    },
    {
      "epoch": 0.05689469431726367,
      "grad_norm": 1.0018610311035414,
      "learning_rate": 9.980092974141855e-06,
      "loss": 0.4386225700378418,
      "num_input_tokens_seen": 43297048,
      "step": 845,
      "train_runtime": 5232.0468,
      "train_tokens_per_second": 8275.356
    },
    {
      "epoch": 0.05723134931322381,
      "grad_norm": 0.8534352581111604,
      "learning_rate": 9.979856569221276e-06,
      "loss": 0.424209451675415,
      "num_input_tokens_seen": 43560840,
      "step": 850,
      "train_runtime": 5263.5893,
      "train_tokens_per_second": 8275.881
    },
    {
      "epoch": 0.05756800430918395,
      "grad_norm": 0.8384006697407111,
      "learning_rate": 9.979618771699688e-06,
      "loss": 0.45295119285583496,
      "num_input_tokens_seen": 43821848,
      "step": 855,
      "train_runtime": 5296.7731,
      "train_tokens_per_second": 8273.31
    },
    {
      "epoch": 0.05790465930514409,
      "grad_norm": 0.9302554581430714,
      "learning_rate": 9.979379581643588e-06,
      "loss": 0.42333412170410156,
      "num_input_tokens_seen": 44069416,
      "step": 860,
      "train_runtime": 5328.4686,
      "train_tokens_per_second": 8270.559
    },
    {
      "epoch": 0.05824131430110423,
      "grad_norm": 0.9521318781144877,
      "learning_rate": 9.979138999119866e-06,
      "loss": 0.4512813568115234,
      "num_input_tokens_seen": 44311120,
      "step": 865,
      "train_runtime": 5357.4848,
      "train_tokens_per_second": 8270.881
    },
    {
      "epoch": 0.05857796929706437,
      "grad_norm": 0.7319514869310989,
      "learning_rate": 9.978897024195801e-06,
      "loss": 0.44945154190063474,
      "num_input_tokens_seen": 44601856,
      "step": 870,
      "train_runtime": 5391.2903,
      "train_tokens_per_second": 8272.947
    },
    {
      "epoch": 0.058914624293024506,
      "grad_norm": 0.8474284842933076,
      "learning_rate": 9.978653656939058e-06,
      "loss": 0.4129044532775879,
      "num_input_tokens_seen": 44858720,
      "step": 875,
      "train_runtime": 5424.345,
      "train_tokens_per_second": 8269.887
    },
    {
      "epoch": 0.059251279288984646,
      "grad_norm": 0.863562174696954,
      "learning_rate": 9.978408897417699e-06,
      "loss": 0.44593143463134766,
      "num_input_tokens_seen": 45122128,
      "step": 880,
      "train_runtime": 5457.9302,
      "train_tokens_per_second": 8267.26
    },
    {
      "epoch": 0.059587934284944785,
      "grad_norm": 0.8933217483904758,
      "learning_rate": 9.978162745700165e-06,
      "loss": 0.445267391204834,
      "num_input_tokens_seen": 45391032,
      "step": 885,
      "train_runtime": 5489.1001,
      "train_tokens_per_second": 8269.303
    },
    {
      "epoch": 0.05992458928090493,
      "grad_norm": 0.7707420130617046,
      "learning_rate": 9.977915201855293e-06,
      "loss": 0.4131511688232422,
      "num_input_tokens_seen": 45646528,
      "step": 890,
      "train_runtime": 5519.0976,
      "train_tokens_per_second": 8270.651
    },
    {
      "epoch": 0.06026124427686507,
      "grad_norm": 0.8234697171759858,
      "learning_rate": 9.977666265952309e-06,
      "loss": 0.4340487480163574,
      "num_input_tokens_seen": 45902328,
      "step": 895,
      "train_runtime": 5552.5941,
      "train_tokens_per_second": 8266.826
    },
    {
      "epoch": 0.06059789927282521,
      "grad_norm": 0.8035456737784816,
      "learning_rate": 9.977415938060823e-06,
      "loss": 0.4263769149780273,
      "num_input_tokens_seen": 46171792,
      "step": 900,
      "train_runtime": 5585.6012,
      "train_tokens_per_second": 8266.217
    },
    {
      "epoch": 0.06093455426878535,
      "grad_norm": 0.8459156866115244,
      "learning_rate": 9.977164218250846e-06,
      "loss": 0.43036413192749023,
      "num_input_tokens_seen": 46419648,
      "step": 905,
      "train_runtime": 5617.6152,
      "train_tokens_per_second": 8263.23
    },
    {
      "epoch": 0.06127120926474549,
      "grad_norm": 0.7571335691900759,
      "learning_rate": 9.976911106592765e-06,
      "loss": 0.40351362228393556,
      "num_input_tokens_seen": 46669960,
      "step": 910,
      "train_runtime": 5646.0184,
      "train_tokens_per_second": 8265.995
    },
    {
      "epoch": 0.06160786426070563,
      "grad_norm": 0.7484863710588767,
      "learning_rate": 9.976656603157363e-06,
      "loss": 0.43393654823303224,
      "num_input_tokens_seen": 46922968,
      "step": 915,
      "train_runtime": 5678.5539,
      "train_tokens_per_second": 8263.19
    },
    {
      "epoch": 0.06194451925666577,
      "grad_norm": 0.9882923844815226,
      "learning_rate": 9.976400708015811e-06,
      "loss": 0.41126203536987305,
      "num_input_tokens_seen": 47179104,
      "step": 920,
      "train_runtime": 5709.7918,
      "train_tokens_per_second": 8262.841
    },
    {
      "epoch": 0.06228117425262591,
      "grad_norm": 0.7672124762587683,
      "learning_rate": 9.97614342123967e-06,
      "loss": 0.420442008972168,
      "num_input_tokens_seen": 47462584,
      "step": 925,
      "train_runtime": 5738.0966,
      "train_tokens_per_second": 8271.486
    },
    {
      "epoch": 0.06261782924858605,
      "grad_norm": 0.7122057608131882,
      "learning_rate": 9.97588474290089e-06,
      "loss": 0.405885124206543,
      "num_input_tokens_seen": 47713744,
      "step": 930,
      "train_runtime": 5769.7962,
      "train_tokens_per_second": 8269.572
    },
    {
      "epoch": 0.0629544842445462,
      "grad_norm": 0.9754758453509172,
      "learning_rate": 9.975624673071806e-06,
      "loss": 0.41826162338256834,
      "num_input_tokens_seen": 47956672,
      "step": 935,
      "train_runtime": 5796.4673,
      "train_tokens_per_second": 8273.431
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 0.8422777261715483,
      "learning_rate": 9.97536321182515e-06,
      "loss": 0.3948413848876953,
      "num_input_tokens_seen": 48210088,
      "step": 940,
      "train_runtime": 5827.3192,
      "train_tokens_per_second": 8273.116
    },
    {
      "epoch": 0.06362779423646647,
      "grad_norm": 0.9886489802404038,
      "learning_rate": 9.975100359234037e-06,
      "loss": 0.4421844482421875,
      "num_input_tokens_seen": 48467352,
      "step": 945,
      "train_runtime": 5855.8755,
      "train_tokens_per_second": 8276.705
    },
    {
      "epoch": 0.06396444923242661,
      "grad_norm": 0.8829170203081893,
      "learning_rate": 9.974836115371972e-06,
      "loss": 0.4475844383239746,
      "num_input_tokens_seen": 48718952,
      "step": 950,
      "train_runtime": 5884.5988,
      "train_tokens_per_second": 8279.061
    },
    {
      "epoch": 0.06430110422838675,
      "grad_norm": 0.8807520014170532,
      "learning_rate": 9.974570480312852e-06,
      "loss": 0.40182924270629883,
      "num_input_tokens_seen": 48977944,
      "step": 955,
      "train_runtime": 5917.6971,
      "train_tokens_per_second": 8276.521
    },
    {
      "epoch": 0.06463775922434689,
      "grad_norm": 1.0042397252828101,
      "learning_rate": 9.974303454130962e-06,
      "loss": 0.43555402755737305,
      "num_input_tokens_seen": 49211752,
      "step": 960,
      "train_runtime": 5950.3965,
      "train_tokens_per_second": 8270.332
    },
    {
      "epoch": 0.06497441422030703,
      "grad_norm": 0.8211497926021195,
      "learning_rate": 9.97403503690097e-06,
      "loss": 0.40522375106811526,
      "num_input_tokens_seen": 49468072,
      "step": 965,
      "train_runtime": 5984.3709,
      "train_tokens_per_second": 8266.211
    },
    {
      "epoch": 0.06531106921626717,
      "grad_norm": 0.7447840056602173,
      "learning_rate": 9.973765228697942e-06,
      "loss": 0.3863669872283936,
      "num_input_tokens_seen": 49735352,
      "step": 970,
      "train_runtime": 6016.964,
      "train_tokens_per_second": 8265.855
    },
    {
      "epoch": 0.06564772421222731,
      "grad_norm": 0.734382027474158,
      "learning_rate": 9.973494029597327e-06,
      "loss": 0.41649589538574217,
      "num_input_tokens_seen": 49971552,
      "step": 975,
      "train_runtime": 6047.5554,
      "train_tokens_per_second": 8263.099
    },
    {
      "epoch": 0.06598437920818745,
      "grad_norm": 0.8705339042164193,
      "learning_rate": 9.973221439674968e-06,
      "loss": 0.42822914123535155,
      "num_input_tokens_seen": 50213864,
      "step": 980,
      "train_runtime": 6078.4382,
      "train_tokens_per_second": 8260.981
    },
    {
      "epoch": 0.06632103420414759,
      "grad_norm": 0.7688450534962282,
      "learning_rate": 9.972947459007089e-06,
      "loss": 0.42665982246398926,
      "num_input_tokens_seen": 50476392,
      "step": 985,
      "train_runtime": 6104.5741,
      "train_tokens_per_second": 8268.618
    },
    {
      "epoch": 0.06665768920010773,
      "grad_norm": 0.7734197536139262,
      "learning_rate": 9.972672087670313e-06,
      "loss": 0.4246163845062256,
      "num_input_tokens_seen": 50739672,
      "step": 990,
      "train_runtime": 6136.5421,
      "train_tokens_per_second": 8268.447
    },
    {
      "epoch": 0.06699434419606787,
      "grad_norm": 0.7231187438444321,
      "learning_rate": 9.972395325741644e-06,
      "loss": 0.40192508697509766,
      "num_input_tokens_seen": 51009216,
      "step": 995,
      "train_runtime": 6167.6988,
      "train_tokens_per_second": 8270.381
    },
    {
      "epoch": 0.06733099919202801,
      "grad_norm": 1.2158892060384614,
      "learning_rate": 9.972117173298476e-06,
      "loss": 0.35872478485107423,
      "num_input_tokens_seen": 51269272,
      "step": 1000,
      "train_runtime": 6196.4777,
      "train_tokens_per_second": 8273.938
    },
    {
      "epoch": 0.06766765418798815,
      "grad_norm": 0.8722865344607229,
      "learning_rate": 9.971837630418597e-06,
      "loss": 0.42832045555114745,
      "num_input_tokens_seen": 51533616,
      "step": 1005,
      "train_runtime": 6228.7261,
      "train_tokens_per_second": 8273.54
    },
    {
      "epoch": 0.06800430918394829,
      "grad_norm": 0.8224314177862213,
      "learning_rate": 9.971556697180179e-06,
      "loss": 0.4298830986022949,
      "num_input_tokens_seen": 51783064,
      "step": 1010,
      "train_runtime": 6259.2042,
      "train_tokens_per_second": 8273.107
    },
    {
      "epoch": 0.06834096417990843,
      "grad_norm": 0.7864028382509498,
      "learning_rate": 9.971274373661784e-06,
      "loss": 0.41619267463684084,
      "num_input_tokens_seen": 52045168,
      "step": 1015,
      "train_runtime": 6290.0161,
      "train_tokens_per_second": 8274.25
    },
    {
      "epoch": 0.06867761917586857,
      "grad_norm": 0.7344914716133121,
      "learning_rate": 9.97099065994236e-06,
      "loss": 0.3913536071777344,
      "num_input_tokens_seen": 52328936,
      "step": 1020,
      "train_runtime": 6321.1066,
      "train_tokens_per_second": 8278.445
    },
    {
      "epoch": 0.0690142741718287,
      "grad_norm": 0.9558784127424379,
      "learning_rate": 9.970705556101253e-06,
      "loss": 0.42629356384277345,
      "num_input_tokens_seen": 52569584,
      "step": 1025,
      "train_runtime": 6353.6965,
      "train_tokens_per_second": 8273.858
    },
    {
      "epoch": 0.06935092916778884,
      "grad_norm": 0.8283750500505537,
      "learning_rate": 9.970419062218186e-06,
      "loss": 0.44330644607543945,
      "num_input_tokens_seen": 52848376,
      "step": 1030,
      "train_runtime": 6388.9263,
      "train_tokens_per_second": 8271.871
    },
    {
      "epoch": 0.06968758416374898,
      "grad_norm": 0.83674403298764,
      "learning_rate": 9.970131178373276e-06,
      "loss": 0.45426087379455565,
      "num_input_tokens_seen": 53114080,
      "step": 1035,
      "train_runtime": 6420.9877,
      "train_tokens_per_second": 8271.949
    },
    {
      "epoch": 0.07002423915970914,
      "grad_norm": 0.8449709466379397,
      "learning_rate": 9.969841904647033e-06,
      "loss": 0.47233004570007325,
      "num_input_tokens_seen": 53372888,
      "step": 1040,
      "train_runtime": 6451.5595,
      "train_tokens_per_second": 8272.866
    },
    {
      "epoch": 0.07036089415566928,
      "grad_norm": 0.8123544492021688,
      "learning_rate": 9.969551241120349e-06,
      "loss": 0.4291078567504883,
      "num_input_tokens_seen": 53661448,
      "step": 1045,
      "train_runtime": 6483.9681,
      "train_tokens_per_second": 8276.02
    },
    {
      "epoch": 0.07069754915162942,
      "grad_norm": 0.8030332992232665,
      "learning_rate": 9.969259187874507e-06,
      "loss": 0.39517912864685056,
      "num_input_tokens_seen": 53943464,
      "step": 1050,
      "train_runtime": 6515.0763,
      "train_tokens_per_second": 8279.79
    },
    {
      "epoch": 0.07103420414758956,
      "grad_norm": 0.8566048740464368,
      "learning_rate": 9.968965744991178e-06,
      "loss": 0.4297319412231445,
      "num_input_tokens_seen": 54215216,
      "step": 1055,
      "train_runtime": 6546.0126,
      "train_tokens_per_second": 8282.174
    },
    {
      "epoch": 0.0713708591435497,
      "grad_norm": 0.8034664520535688,
      "learning_rate": 9.968670912552422e-06,
      "loss": 0.4204543113708496,
      "num_input_tokens_seen": 54467560,
      "step": 1060,
      "train_runtime": 6577.9957,
      "train_tokens_per_second": 8280.267
    },
    {
      "epoch": 0.07170751413950983,
      "grad_norm": 0.740319827669326,
      "learning_rate": 9.968374690640692e-06,
      "loss": 0.41419224739074706,
      "num_input_tokens_seen": 54727720,
      "step": 1065,
      "train_runtime": 6610.6136,
      "train_tokens_per_second": 8278.765
    },
    {
      "epoch": 0.07204416913546997,
      "grad_norm": 0.8243733407326276,
      "learning_rate": 9.968077079338822e-06,
      "loss": 0.4449822425842285,
      "num_input_tokens_seen": 54991984,
      "step": 1070,
      "train_runtime": 6637.5661,
      "train_tokens_per_second": 8284.962
    },
    {
      "epoch": 0.07238082413143011,
      "grad_norm": 0.9436554111472695,
      "learning_rate": 9.967778078730038e-06,
      "loss": 0.43380327224731446,
      "num_input_tokens_seen": 55249600,
      "step": 1075,
      "train_runtime": 6666.8123,
      "train_tokens_per_second": 8287.259
    },
    {
      "epoch": 0.07271747912739025,
      "grad_norm": 0.8017952060500069,
      "learning_rate": 9.967477688897957e-06,
      "loss": 0.43855953216552734,
      "num_input_tokens_seen": 55515632,
      "step": 1080,
      "train_runtime": 6697.839,
      "train_tokens_per_second": 8288.589
    },
    {
      "epoch": 0.07305413412335039,
      "grad_norm": 0.7009019669020908,
      "learning_rate": 9.967175909926578e-06,
      "loss": 0.4424342155456543,
      "num_input_tokens_seen": 55795320,
      "step": 1085,
      "train_runtime": 6729.6403,
      "train_tokens_per_second": 8290.981
    },
    {
      "epoch": 0.07339078911931053,
      "grad_norm": 0.850660173514265,
      "learning_rate": 9.966872741900296e-06,
      "loss": 0.46137075424194335,
      "num_input_tokens_seen": 56062184,
      "step": 1090,
      "train_runtime": 6758.3808,
      "train_tokens_per_second": 8295.209
    },
    {
      "epoch": 0.07372744411527067,
      "grad_norm": 0.8901116780839742,
      "learning_rate": 9.96656818490389e-06,
      "loss": 0.4384910106658936,
      "num_input_tokens_seen": 56318576,
      "step": 1095,
      "train_runtime": 6790.0804,
      "train_tokens_per_second": 8294.243
    },
    {
      "epoch": 0.07406409911123081,
      "grad_norm": 0.8531343198557408,
      "learning_rate": 9.966262239022529e-06,
      "loss": 0.4633244514465332,
      "num_input_tokens_seen": 56589296,
      "step": 1100,
      "train_runtime": 6823.5366,
      "train_tokens_per_second": 8293.25
    },
    {
      "epoch": 0.07440075410719095,
      "grad_norm": 2.3665939638265696,
      "learning_rate": 9.965954904341768e-06,
      "loss": 0.4465372085571289,
      "num_input_tokens_seen": 56850608,
      "step": 1105,
      "train_runtime": 6853.954,
      "train_tokens_per_second": 8294.571
    },
    {
      "epoch": 0.07473740910315109,
      "grad_norm": 0.9442004499597122,
      "learning_rate": 9.965646180947555e-06,
      "loss": 0.4268515110015869,
      "num_input_tokens_seen": 57094880,
      "step": 1110,
      "train_runtime": 6886.9964,
      "train_tokens_per_second": 8290.244
    },
    {
      "epoch": 0.07507406409911123,
      "grad_norm": 0.8490889544030135,
      "learning_rate": 9.96533606892622e-06,
      "loss": 0.41021275520324707,
      "num_input_tokens_seen": 57336704,
      "step": 1115,
      "train_runtime": 6920.6663,
      "train_tokens_per_second": 8284.853
    },
    {
      "epoch": 0.07541071909507137,
      "grad_norm": 0.7859708375749179,
      "learning_rate": 9.965024568364487e-06,
      "loss": 0.4229478359222412,
      "num_input_tokens_seen": 57610312,
      "step": 1120,
      "train_runtime": 6947.4606,
      "train_tokens_per_second": 8292.283
    },
    {
      "epoch": 0.07574737409103151,
      "grad_norm": 0.7673112194155692,
      "learning_rate": 9.964711679349467e-06,
      "loss": 0.40393891334533694,
      "num_input_tokens_seen": 57857104,
      "step": 1125,
      "train_runtime": 6980.6491,
      "train_tokens_per_second": 8288.213
    },
    {
      "epoch": 0.07608402908699165,
      "grad_norm": 0.6820384071098183,
      "learning_rate": 9.964397401968656e-06,
      "loss": 0.4428199291229248,
      "num_input_tokens_seen": 58120832,
      "step": 1130,
      "train_runtime": 7010.4099,
      "train_tokens_per_second": 8290.647
    },
    {
      "epoch": 0.07642068408295179,
      "grad_norm": 0.7910366471962886,
      "learning_rate": 9.964081736309942e-06,
      "loss": 0.4140345573425293,
      "num_input_tokens_seen": 58404504,
      "step": 1135,
      "train_runtime": 7043.8435,
      "train_tokens_per_second": 8291.567
    },
    {
      "epoch": 0.07675733907891193,
      "grad_norm": 0.8086003408263853,
      "learning_rate": 9.963764682461599e-06,
      "loss": 0.4157541751861572,
      "num_input_tokens_seen": 58685680,
      "step": 1140,
      "train_runtime": 7077.3984,
      "train_tokens_per_second": 8291.985
    },
    {
      "epoch": 0.07709399407487207,
      "grad_norm": 0.8205385131599382,
      "learning_rate": 9.96344624051229e-06,
      "loss": 0.42887463569641116,
      "num_input_tokens_seen": 58941568,
      "step": 1145,
      "train_runtime": 7111.0513,
      "train_tokens_per_second": 8288.728
    },
    {
      "epoch": 0.0774306490708322,
      "grad_norm": 0.8153423714476794,
      "learning_rate": 9.96312641055107e-06,
      "loss": 0.40596814155578614,
      "num_input_tokens_seen": 59197784,
      "step": 1150,
      "train_runtime": 7137.8413,
      "train_tokens_per_second": 8293.514
    },
    {
      "epoch": 0.07776730406679234,
      "grad_norm": 0.6785446829067378,
      "learning_rate": 9.962805192667375e-06,
      "loss": 0.4016563892364502,
      "num_input_tokens_seen": 59447880,
      "step": 1155,
      "train_runtime": 7169.1454,
      "train_tokens_per_second": 8292.185
    },
    {
      "epoch": 0.07810395906275248,
      "grad_norm": 0.8607697765701847,
      "learning_rate": 9.962482586951033e-06,
      "loss": 0.41080322265625,
      "num_input_tokens_seen": 59732424,
      "step": 1160,
      "train_runtime": 7199.3053,
      "train_tokens_per_second": 8296.971
    },
    {
      "epoch": 0.07844061405871264,
      "grad_norm": 0.821523573743212,
      "learning_rate": 9.962158593492258e-06,
      "loss": 0.4302816867828369,
      "num_input_tokens_seen": 59967456,
      "step": 1165,
      "train_runtime": 7225.6423,
      "train_tokens_per_second": 8299.256
    },
    {
      "epoch": 0.07877726905467278,
      "grad_norm": 0.7968632978996432,
      "learning_rate": 9.961833212381657e-06,
      "loss": 0.43314199447631835,
      "num_input_tokens_seen": 60231992,
      "step": 1170,
      "train_runtime": 7254.8226,
      "train_tokens_per_second": 8302.338
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 0.8890470976952313,
      "learning_rate": 9.961506443710221e-06,
      "loss": 0.41593079566955565,
      "num_input_tokens_seen": 60490408,
      "step": 1175,
      "train_runtime": 7279.9714,
      "train_tokens_per_second": 8309.155
    },
    {
      "epoch": 0.07945057904659306,
      "grad_norm": 0.8154048336068187,
      "learning_rate": 9.961178287569326e-06,
      "loss": 0.4058260440826416,
      "num_input_tokens_seen": 60768184,
      "step": 1180,
      "train_runtime": 7314.7548,
      "train_tokens_per_second": 8307.617
    },
    {
      "epoch": 0.0797872340425532,
      "grad_norm": 0.7807839566377628,
      "learning_rate": 9.960848744050747e-06,
      "loss": 0.43466963768005373,
      "num_input_tokens_seen": 61032072,
      "step": 1185,
      "train_runtime": 7343.4789,
      "train_tokens_per_second": 8311.057
    },
    {
      "epoch": 0.08012388903851334,
      "grad_norm": 0.7190531262642533,
      "learning_rate": 9.960517813246633e-06,
      "loss": 0.4408872604370117,
      "num_input_tokens_seen": 61276384,
      "step": 1190,
      "train_runtime": 7375.74,
      "train_tokens_per_second": 8307.829
    },
    {
      "epoch": 0.08046054403447347,
      "grad_norm": 0.9597284200602787,
      "learning_rate": 9.96018549524953e-06,
      "loss": 0.41578149795532227,
      "num_input_tokens_seen": 61532632,
      "step": 1195,
      "train_runtime": 7405.6058,
      "train_tokens_per_second": 8308.926
    },
    {
      "epoch": 0.08079719903043361,
      "grad_norm": 0.8100494624546771,
      "learning_rate": 9.959851790152371e-06,
      "loss": 0.42075152397155763,
      "num_input_tokens_seen": 61792016,
      "step": 1200,
      "train_runtime": 7437.1454,
      "train_tokens_per_second": 8308.566
    },
    {
      "epoch": 0.08113385402639375,
      "grad_norm": 0.7887078935410129,
      "learning_rate": 9.959516698048475e-06,
      "loss": 0.37525961399078367,
      "num_input_tokens_seen": 62027184,
      "step": 1205,
      "train_runtime": 7466.1753,
      "train_tokens_per_second": 8307.759
    },
    {
      "epoch": 0.08147050902235389,
      "grad_norm": 0.776793707129068,
      "learning_rate": 9.959180219031551e-06,
      "loss": 0.4041851043701172,
      "num_input_tokens_seen": 62271992,
      "step": 1210,
      "train_runtime": 7496.5422,
      "train_tokens_per_second": 8306.762
    },
    {
      "epoch": 0.08180716401831403,
      "grad_norm": 0.7732908080951378,
      "learning_rate": 9.958842353195688e-06,
      "loss": 0.42111873626708984,
      "num_input_tokens_seen": 62536976,
      "step": 1215,
      "train_runtime": 7531.6971,
      "train_tokens_per_second": 8303.172
    },
    {
      "epoch": 0.08214381901427417,
      "grad_norm": 0.8645493771530781,
      "learning_rate": 9.958503100635377e-06,
      "loss": 0.4235030174255371,
      "num_input_tokens_seen": 62779400,
      "step": 1220,
      "train_runtime": 7558.3198,
      "train_tokens_per_second": 8305.999
    },
    {
      "epoch": 0.08248047401023431,
      "grad_norm": 0.7427416527961508,
      "learning_rate": 9.958162461445484e-06,
      "loss": 0.45887441635131837,
      "num_input_tokens_seen": 63051376,
      "step": 1225,
      "train_runtime": 7593.2492,
      "train_tokens_per_second": 8303.61
    },
    {
      "epoch": 0.08281712900619445,
      "grad_norm": 0.8479958649620551,
      "learning_rate": 9.957820435721271e-06,
      "loss": 0.43243608474731443,
      "num_input_tokens_seen": 63307328,
      "step": 1230,
      "train_runtime": 7623.4308,
      "train_tokens_per_second": 8304.309
    },
    {
      "epoch": 0.08315378400215459,
      "grad_norm": 0.7736440662634388,
      "learning_rate": 9.95747702355838e-06,
      "loss": 0.45577640533447267,
      "num_input_tokens_seen": 63565912,
      "step": 1235,
      "train_runtime": 7655.9302,
      "train_tokens_per_second": 8302.833
    },
    {
      "epoch": 0.08349043899811473,
      "grad_norm": 0.7352531094411469,
      "learning_rate": 9.957132225052848e-06,
      "loss": 0.44762301445007324,
      "num_input_tokens_seen": 63817216,
      "step": 1240,
      "train_runtime": 7687.7817,
      "train_tokens_per_second": 8301.122
    },
    {
      "epoch": 0.08382709399407487,
      "grad_norm": 0.7503810818888368,
      "learning_rate": 9.956786040301098e-06,
      "loss": 0.40407328605651854,
      "num_input_tokens_seen": 64070160,
      "step": 1245,
      "train_runtime": 7721.3291,
      "train_tokens_per_second": 8297.815
    },
    {
      "epoch": 0.08416374899003501,
      "grad_norm": 0.8722115446679295,
      "learning_rate": 9.956438469399936e-06,
      "loss": 0.41785545349121095,
      "num_input_tokens_seen": 64309904,
      "step": 1250,
      "train_runtime": 7750.8377,
      "train_tokens_per_second": 8297.155
    },
    {
      "epoch": 0.08450040398599515,
      "grad_norm": 0.8311703565686764,
      "learning_rate": 9.956089512446562e-06,
      "loss": 0.4245553970336914,
      "num_input_tokens_seen": 64534584,
      "step": 1255,
      "train_runtime": 7777.2403,
      "train_tokens_per_second": 8297.877
    },
    {
      "epoch": 0.08483705898195529,
      "grad_norm": 0.8499974838401965,
      "learning_rate": 9.955739169538559e-06,
      "loss": 0.42313251495361326,
      "num_input_tokens_seen": 64778024,
      "step": 1260,
      "train_runtime": 7808.0295,
      "train_tokens_per_second": 8296.334
    },
    {
      "epoch": 0.08517371397791543,
      "grad_norm": 0.8034907224463095,
      "learning_rate": 9.955387440773902e-06,
      "loss": 0.4202866077423096,
      "num_input_tokens_seen": 65010152,
      "step": 1265,
      "train_runtime": 7835.147,
      "train_tokens_per_second": 8297.247
    },
    {
      "epoch": 0.08551036897387557,
      "grad_norm": 1.0387802525454377,
      "learning_rate": 9.955034326250947e-06,
      "loss": 0.4267317295074463,
      "num_input_tokens_seen": 65264024,
      "step": 1270,
      "train_runtime": 7866.4372,
      "train_tokens_per_second": 8296.516
    },
    {
      "epoch": 0.0858470239698357,
      "grad_norm": 0.8361164545920644,
      "learning_rate": 9.954679826068444e-06,
      "loss": 0.4206550598144531,
      "num_input_tokens_seen": 65516832,
      "step": 1275,
      "train_runtime": 7898.4028,
      "train_tokens_per_second": 8294.947
    },
    {
      "epoch": 0.08618367896579585,
      "grad_norm": 0.9099309337742033,
      "learning_rate": 9.954323940325526e-06,
      "loss": 0.42358860969543455,
      "num_input_tokens_seen": 65758080,
      "step": 1280,
      "train_runtime": 7930.0792,
      "train_tokens_per_second": 8292.235
    },
    {
      "epoch": 0.086520333961756,
      "grad_norm": 0.7311777286002299,
      "learning_rate": 9.953966669121715e-06,
      "loss": 0.4463528633117676,
      "num_input_tokens_seen": 66042328,
      "step": 1285,
      "train_runtime": 7964.3129,
      "train_tokens_per_second": 8292.282
    },
    {
      "epoch": 0.08685698895771614,
      "grad_norm": 0.8457784271660618,
      "learning_rate": 9.953608012556924e-06,
      "loss": 0.42635326385498046,
      "num_input_tokens_seen": 66304176,
      "step": 1290,
      "train_runtime": 7995.2293,
      "train_tokens_per_second": 8292.967
    },
    {
      "epoch": 0.08719364395367628,
      "grad_norm": 0.8047060565859071,
      "learning_rate": 9.953247970731449e-06,
      "loss": 0.4460731506347656,
      "num_input_tokens_seen": 66543808,
      "step": 1295,
      "train_runtime": 8025.1251,
      "train_tokens_per_second": 8291.934
    },
    {
      "epoch": 0.08753029894963642,
      "grad_norm": 0.9031276607213266,
      "learning_rate": 9.952886543745973e-06,
      "loss": 0.44219532012939455,
      "num_input_tokens_seen": 66798984,
      "step": 1300,
      "train_runtime": 8058.1742,
      "train_tokens_per_second": 8289.593
    },
    {
      "epoch": 0.08786695394559656,
      "grad_norm": 1.0032118219135107,
      "learning_rate": 9.952523731701566e-06,
      "loss": 0.4293045997619629,
      "num_input_tokens_seen": 67069656,
      "step": 1305,
      "train_runtime": 8089.657,
      "train_tokens_per_second": 8290.791
    },
    {
      "epoch": 0.0882036089415567,
      "grad_norm": 0.7102813399759771,
      "learning_rate": 9.952159534699693e-06,
      "loss": 0.38901257514953613,
      "num_input_tokens_seen": 67327272,
      "step": 1310,
      "train_runtime": 8118.2854,
      "train_tokens_per_second": 8293.287
    },
    {
      "epoch": 0.08854026393751684,
      "grad_norm": 0.8239833872894313,
      "learning_rate": 9.951793952842197e-06,
      "loss": 0.44980688095092775,
      "num_input_tokens_seen": 67578992,
      "step": 1315,
      "train_runtime": 8149.0654,
      "train_tokens_per_second": 8292.852
    },
    {
      "epoch": 0.08887691893347698,
      "grad_norm": 0.984230793661546,
      "learning_rate": 9.95142698623131e-06,
      "loss": 0.40392866134643557,
      "num_input_tokens_seen": 67794920,
      "step": 1320,
      "train_runtime": 8179.1549,
      "train_tokens_per_second": 8288.744
    },
    {
      "epoch": 0.08921357392943711,
      "grad_norm": 0.7631557024672336,
      "learning_rate": 9.951058634969657e-06,
      "loss": 0.44867525100708006,
      "num_input_tokens_seen": 68035712,
      "step": 1325,
      "train_runtime": 8209.1367,
      "train_tokens_per_second": 8287.803
    },
    {
      "epoch": 0.08955022892539725,
      "grad_norm": 0.7964537095267661,
      "learning_rate": 9.950688899160244e-06,
      "loss": 0.4250572681427002,
      "num_input_tokens_seen": 68308224,
      "step": 1330,
      "train_runtime": 8241.646,
      "train_tokens_per_second": 8288.177
    },
    {
      "epoch": 0.0898868839213574,
      "grad_norm": 0.7367894844018001,
      "learning_rate": 9.950317778906469e-06,
      "loss": 0.4231928825378418,
      "num_input_tokens_seen": 68573840,
      "step": 1335,
      "train_runtime": 8272.3753,
      "train_tokens_per_second": 8289.498
    },
    {
      "epoch": 0.09022353891731753,
      "grad_norm": 0.786484288201692,
      "learning_rate": 9.949945274312109e-06,
      "loss": 0.4117992401123047,
      "num_input_tokens_seen": 68833648,
      "step": 1340,
      "train_runtime": 8299.9363,
      "train_tokens_per_second": 8293.274
    },
    {
      "epoch": 0.09056019391327767,
      "grad_norm": 0.9558085214484208,
      "learning_rate": 9.94957138548134e-06,
      "loss": 0.43130979537963865,
      "num_input_tokens_seen": 69056960,
      "step": 1345,
      "train_runtime": 8328.944,
      "train_tokens_per_second": 8291.202
    },
    {
      "epoch": 0.09089684890923781,
      "grad_norm": 0.9055333248086785,
      "learning_rate": 9.949196112518715e-06,
      "loss": 0.3931777238845825,
      "num_input_tokens_seen": 69327136,
      "step": 1350,
      "train_runtime": 8359.5255,
      "train_tokens_per_second": 8293.19
    },
    {
      "epoch": 0.09123350390519795,
      "grad_norm": 0.935808927233124,
      "learning_rate": 9.94881945552918e-06,
      "loss": 0.43684825897216795,
      "num_input_tokens_seen": 69580480,
      "step": 1355,
      "train_runtime": 8389.6357,
      "train_tokens_per_second": 8293.624
    },
    {
      "epoch": 0.09157015890115809,
      "grad_norm": 0.9069993718596747,
      "learning_rate": 9.948441414618064e-06,
      "loss": 0.4528181076049805,
      "num_input_tokens_seen": 69821056,
      "step": 1360,
      "train_runtime": 8418.3987,
      "train_tokens_per_second": 8293.864
    },
    {
      "epoch": 0.09190681389711823,
      "grad_norm": 0.7635046462137408,
      "learning_rate": 9.948061989891085e-06,
      "loss": 0.4195862770080566,
      "num_input_tokens_seen": 70087912,
      "step": 1365,
      "train_runtime": 8450.6176,
      "train_tokens_per_second": 8293.821
    },
    {
      "epoch": 0.09224346889307837,
      "grad_norm": 0.8738324969387051,
      "learning_rate": 9.94768118145435e-06,
      "loss": 0.41327805519104005,
      "num_input_tokens_seen": 70338744,
      "step": 1370,
      "train_runtime": 8479.7425,
      "train_tokens_per_second": 8294.915
    },
    {
      "epoch": 0.09258012388903851,
      "grad_norm": 0.7067302427081529,
      "learning_rate": 9.947298989414349e-06,
      "loss": 0.4209153652191162,
      "num_input_tokens_seen": 70604928,
      "step": 1375,
      "train_runtime": 8510.9138,
      "train_tokens_per_second": 8295.81
    },
    {
      "epoch": 0.09291677888499865,
      "grad_norm": 1.072585964181785,
      "learning_rate": 9.946915413877963e-06,
      "loss": 0.424015474319458,
      "num_input_tokens_seen": 70875408,
      "step": 1380,
      "train_runtime": 8539.3585,
      "train_tokens_per_second": 8299.852
    },
    {
      "epoch": 0.09325343388095879,
      "grad_norm": 0.7602102033883376,
      "learning_rate": 9.946530454952454e-06,
      "loss": 0.4193732738494873,
      "num_input_tokens_seen": 71143312,
      "step": 1385,
      "train_runtime": 8573.1526,
      "train_tokens_per_second": 8298.384
    },
    {
      "epoch": 0.09359008887691893,
      "grad_norm": 0.9025628144035989,
      "learning_rate": 9.946144112745477e-06,
      "loss": 0.4529858589172363,
      "num_input_tokens_seen": 71402080,
      "step": 1390,
      "train_runtime": 8603.1996,
      "train_tokens_per_second": 8299.48
    },
    {
      "epoch": 0.09392674387287907,
      "grad_norm": 0.8336274756374049,
      "learning_rate": 9.94575638736507e-06,
      "loss": 0.42154393196105955,
      "num_input_tokens_seen": 71678344,
      "step": 1395,
      "train_runtime": 8635.3196,
      "train_tokens_per_second": 8300.601
    },
    {
      "epoch": 0.09426339886883921,
      "grad_norm": 0.8124017188355276,
      "learning_rate": 9.945367278919662e-06,
      "loss": 0.45946245193481444,
      "num_input_tokens_seen": 71950360,
      "step": 1400,
      "train_runtime": 8667.1322,
      "train_tokens_per_second": 8301.519
    },
    {
      "epoch": 0.09460005386479936,
      "grad_norm": 0.717915625375132,
      "learning_rate": 9.944976787518064e-06,
      "loss": 0.41889386177062987,
      "num_input_tokens_seen": 72217200,
      "step": 1405,
      "train_runtime": 8700.3656,
      "train_tokens_per_second": 8300.479
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 0.7550323089547183,
      "learning_rate": 9.944584913269474e-06,
      "loss": 0.41649599075317384,
      "num_input_tokens_seen": 72494504,
      "step": 1410,
      "train_runtime": 8729.6166,
      "train_tokens_per_second": 8304.432
    },
    {
      "epoch": 0.09527336385671964,
      "grad_norm": 0.8361285833266869,
      "learning_rate": 9.944191656283481e-06,
      "loss": 0.4168684959411621,
      "num_input_tokens_seen": 72751208,
      "step": 1415,
      "train_runtime": 8761.827,
      "train_tokens_per_second": 8303.201
    },
    {
      "epoch": 0.09561001885267978,
      "grad_norm": 0.7927875491711355,
      "learning_rate": 9.943797016670059e-06,
      "loss": 0.4052313804626465,
      "num_input_tokens_seen": 73001944,
      "step": 1420,
      "train_runtime": 8789.3137,
      "train_tokens_per_second": 8305.762
    },
    {
      "epoch": 0.09594667384863992,
      "grad_norm": 0.7444195485562319,
      "learning_rate": 9.943400994539565e-06,
      "loss": 0.40229191780090334,
      "num_input_tokens_seen": 73254720,
      "step": 1425,
      "train_runtime": 8816.6973,
      "train_tokens_per_second": 8308.635
    },
    {
      "epoch": 0.09628332884460006,
      "grad_norm": 0.8490823812968631,
      "learning_rate": 9.943003590002744e-06,
      "loss": 0.4262238502502441,
      "num_input_tokens_seen": 73508688,
      "step": 1430,
      "train_runtime": 8849.8071,
      "train_tokens_per_second": 8306.248
    },
    {
      "epoch": 0.0966199838405602,
      "grad_norm": 0.8303900113551835,
      "learning_rate": 9.942604803170733e-06,
      "loss": 0.41085233688354494,
      "num_input_tokens_seen": 73761544,
      "step": 1435,
      "train_runtime": 8876.4041,
      "train_tokens_per_second": 8309.845
    },
    {
      "epoch": 0.09695663883652034,
      "grad_norm": 0.7613372475671172,
      "learning_rate": 9.94220463415505e-06,
      "loss": 0.44302005767822267,
      "num_input_tokens_seen": 74011864,
      "step": 1440,
      "train_runtime": 8903.0349,
      "train_tokens_per_second": 8313.105
    },
    {
      "epoch": 0.09729329383248048,
      "grad_norm": 0.7603375216766292,
      "learning_rate": 9.9418030830676e-06,
      "loss": 0.41818509101867674,
      "num_input_tokens_seen": 74273080,
      "step": 1445,
      "train_runtime": 8934.2328,
      "train_tokens_per_second": 8313.314
    },
    {
      "epoch": 0.09762994882844062,
      "grad_norm": 1.0007932990859163,
      "learning_rate": 9.941400150020676e-06,
      "loss": 0.4348613739013672,
      "num_input_tokens_seen": 74508688,
      "step": 1450,
      "train_runtime": 8960.733,
      "train_tokens_per_second": 8315.022
    },
    {
      "epoch": 0.09796660382440076,
      "grad_norm": 0.7721254517157227,
      "learning_rate": 9.940995835126957e-06,
      "loss": 0.4412411689758301,
      "num_input_tokens_seen": 74779136,
      "step": 1455,
      "train_runtime": 8989.2199,
      "train_tokens_per_second": 8318.757
    },
    {
      "epoch": 0.0983032588203609,
      "grad_norm": 0.746315457013498,
      "learning_rate": 9.940590138499508e-06,
      "loss": 0.3903627395629883,
      "num_input_tokens_seen": 75054192,
      "step": 1460,
      "train_runtime": 9020.0724,
      "train_tokens_per_second": 8320.797
    },
    {
      "epoch": 0.09863991381632103,
      "grad_norm": 0.9368904246400032,
      "learning_rate": 9.940183060251783e-06,
      "loss": 0.42001729011535643,
      "num_input_tokens_seen": 75275896,
      "step": 1465,
      "train_runtime": 9048.0617,
      "train_tokens_per_second": 8319.56
    },
    {
      "epoch": 0.09897656881228117,
      "grad_norm": 0.8254781079086709,
      "learning_rate": 9.939774600497616e-06,
      "loss": 0.3700115203857422,
      "num_input_tokens_seen": 75535280,
      "step": 1470,
      "train_runtime": 9079.9734,
      "train_tokens_per_second": 8318.888
    },
    {
      "epoch": 0.09931322380824131,
      "grad_norm": 0.826590823806469,
      "learning_rate": 9.939364759351236e-06,
      "loss": 0.4386242389678955,
      "num_input_tokens_seen": 75784288,
      "step": 1475,
      "train_runtime": 9114.0391,
      "train_tokens_per_second": 8315.115
    },
    {
      "epoch": 0.09964987880420145,
      "grad_norm": 0.7485282266596478,
      "learning_rate": 9.938953536927249e-06,
      "loss": 0.38956236839294434,
      "num_input_tokens_seen": 76044200,
      "step": 1480,
      "train_runtime": 9143.953,
      "train_tokens_per_second": 8316.338
    },
    {
      "epoch": 0.09998653380016159,
      "grad_norm": 0.6981588701530992,
      "learning_rate": 9.938540933340657e-06,
      "loss": 0.4284168243408203,
      "num_input_tokens_seen": 76322464,
      "step": 1485,
      "train_runtime": 9179.8697,
      "train_tokens_per_second": 8314.112
    },
    {
      "epoch": 0.10032318879612173,
      "grad_norm": 0.725560179105872,
      "learning_rate": 9.938126948706839e-06,
      "loss": 0.4093619346618652,
      "num_input_tokens_seen": 76595160,
      "step": 1490,
      "train_runtime": 9215.8989,
      "train_tokens_per_second": 8311.198
    },
    {
      "epoch": 0.10065984379208187,
      "grad_norm": 0.6700768009230947,
      "learning_rate": 9.937711583141567e-06,
      "loss": 0.40921902656555176,
      "num_input_tokens_seen": 76857416,
      "step": 1495,
      "train_runtime": 9243.635,
      "train_tokens_per_second": 8314.631
    },
    {
      "epoch": 0.10099649878804201,
      "grad_norm": 0.7740048246367501,
      "learning_rate": 9.937294836760995e-06,
      "loss": 0.41411256790161133,
      "num_input_tokens_seen": 77123240,
      "step": 1500,
      "train_runtime": 9274.3147,
      "train_tokens_per_second": 8315.789
    },
    {
      "epoch": 0.10133315378400215,
      "grad_norm": 0.8100365480230818,
      "learning_rate": 9.936876709681668e-06,
      "loss": 0.4696906089782715,
      "num_input_tokens_seen": 77373056,
      "step": 1505,
      "train_runtime": 9305.4174,
      "train_tokens_per_second": 8314.84
    },
    {
      "epoch": 0.10166980877996229,
      "grad_norm": 0.6718705433659631,
      "learning_rate": 9.93645720202051e-06,
      "loss": 0.3592568635940552,
      "num_input_tokens_seen": 77636776,
      "step": 1510,
      "train_runtime": 9339.512,
      "train_tokens_per_second": 8312.723
    },
    {
      "epoch": 0.10200646377592243,
      "grad_norm": 0.7744863237769691,
      "learning_rate": 9.936036313894836e-06,
      "loss": 0.4570298194885254,
      "num_input_tokens_seen": 77879128,
      "step": 1515,
      "train_runtime": 9370.4992,
      "train_tokens_per_second": 8311.097
    },
    {
      "epoch": 0.10234311877188257,
      "grad_norm": 0.8383706763585702,
      "learning_rate": 9.935614045422349e-06,
      "loss": 0.43991889953613283,
      "num_input_tokens_seen": 78137712,
      "step": 1520,
      "train_runtime": 9400.726,
      "train_tokens_per_second": 8311.881
    },
    {
      "epoch": 0.10267977376784271,
      "grad_norm": 0.7527909928089213,
      "learning_rate": 9.935190396721132e-06,
      "loss": 0.3979470729827881,
      "num_input_tokens_seen": 78397912,
      "step": 1525,
      "train_runtime": 9431.0892,
      "train_tokens_per_second": 8312.71
    },
    {
      "epoch": 0.10301642876380286,
      "grad_norm": 0.7393366732562359,
      "learning_rate": 9.934765367909658e-06,
      "loss": 0.42573184967041017,
      "num_input_tokens_seen": 78653048,
      "step": 1530,
      "train_runtime": 9460.228,
      "train_tokens_per_second": 8314.075
    },
    {
      "epoch": 0.103353083759763,
      "grad_norm": 0.8360807315869992,
      "learning_rate": 9.934338959106783e-06,
      "loss": 0.4019515037536621,
      "num_input_tokens_seen": 78926304,
      "step": 1535,
      "train_runtime": 9491.5087,
      "train_tokens_per_second": 8315.465
    },
    {
      "epoch": 0.10368973875572314,
      "grad_norm": 0.9635867308816715,
      "learning_rate": 9.933911170431753e-06,
      "loss": 0.4399405002593994,
      "num_input_tokens_seen": 79153144,
      "step": 1540,
      "train_runtime": 9522.2337,
      "train_tokens_per_second": 8312.456
    },
    {
      "epoch": 0.10402639375168328,
      "grad_norm": 0.8306904841969804,
      "learning_rate": 9.933482002004198e-06,
      "loss": 0.42841367721557616,
      "num_input_tokens_seen": 79418304,
      "step": 1545,
      "train_runtime": 9553.3998,
      "train_tokens_per_second": 8313.093
    },
    {
      "epoch": 0.10436304874764342,
      "grad_norm": 0.9439777319651309,
      "learning_rate": 9.933051453944134e-06,
      "loss": 0.4567451477050781,
      "num_input_tokens_seen": 79696368,
      "step": 1550,
      "train_runtime": 9586.4418,
      "train_tokens_per_second": 8313.446
    },
    {
      "epoch": 0.10469970374360356,
      "grad_norm": 0.6947196726636788,
      "learning_rate": 9.932619526371958e-06,
      "loss": 0.3873995304107666,
      "num_input_tokens_seen": 79955480,
      "step": 1555,
      "train_runtime": 9616.489,
      "train_tokens_per_second": 8314.415
    },
    {
      "epoch": 0.1050363587395637,
      "grad_norm": 0.8149676060322595,
      "learning_rate": 9.932186219408463e-06,
      "loss": 0.4045438766479492,
      "num_input_tokens_seen": 80202696,
      "step": 1560,
      "train_runtime": 9646.0416,
      "train_tokens_per_second": 8314.571
    },
    {
      "epoch": 0.10537301373552384,
      "grad_norm": 0.8492545239695563,
      "learning_rate": 9.931751533174819e-06,
      "loss": 0.4321916580200195,
      "num_input_tokens_seen": 80447264,
      "step": 1565,
      "train_runtime": 9672.2487,
      "train_tokens_per_second": 8317.328
    },
    {
      "epoch": 0.10570966873148398,
      "grad_norm": 0.8386933019602745,
      "learning_rate": 9.931315467792584e-06,
      "loss": 0.44005279541015624,
      "num_input_tokens_seen": 80728200,
      "step": 1570,
      "train_runtime": 9704.9856,
      "train_tokens_per_second": 8318.219
    },
    {
      "epoch": 0.10604632372744412,
      "grad_norm": 0.7500574771479777,
      "learning_rate": 9.930878023383705e-06,
      "loss": 0.4337003707885742,
      "num_input_tokens_seen": 80984752,
      "step": 1575,
      "train_runtime": 9735.2236,
      "train_tokens_per_second": 8318.736
    },
    {
      "epoch": 0.10638297872340426,
      "grad_norm": 0.7687784778977624,
      "learning_rate": 9.93043920007051e-06,
      "loss": 0.3846965551376343,
      "num_input_tokens_seen": 81247328,
      "step": 1580,
      "train_runtime": 9764.0962,
      "train_tokens_per_second": 8321.029
    },
    {
      "epoch": 0.1067196337193644,
      "grad_norm": 0.9244691006568833,
      "learning_rate": 9.929998997975712e-06,
      "loss": 0.4034592151641846,
      "num_input_tokens_seen": 81534608,
      "step": 1585,
      "train_runtime": 9794.4619,
      "train_tokens_per_second": 8324.562
    },
    {
      "epoch": 0.10705628871532454,
      "grad_norm": 0.7935947029412047,
      "learning_rate": 9.929557417222417e-06,
      "loss": 0.39935166835784913,
      "num_input_tokens_seen": 81773232,
      "step": 1590,
      "train_runtime": 9823.7062,
      "train_tokens_per_second": 8324.071
    },
    {
      "epoch": 0.10739294371128467,
      "grad_norm": 0.7673354544883247,
      "learning_rate": 9.929114457934109e-06,
      "loss": 0.4187006950378418,
      "num_input_tokens_seen": 82028336,
      "step": 1595,
      "train_runtime": 9851.3225,
      "train_tokens_per_second": 8326.632
    },
    {
      "epoch": 0.10772959870724481,
      "grad_norm": 0.809107568349099,
      "learning_rate": 9.92867012023466e-06,
      "loss": 0.43503632545471194,
      "num_input_tokens_seen": 82290216,
      "step": 1600,
      "train_runtime": 9881.009,
      "train_tokens_per_second": 8328.119
    },
    {
      "epoch": 0.10806625370320495,
      "grad_norm": 0.7260523224664193,
      "learning_rate": 9.92822440424833e-06,
      "loss": 0.44855165481567383,
      "num_input_tokens_seen": 82556384,
      "step": 1605,
      "train_runtime": 9913.9412,
      "train_tokens_per_second": 8327.302
    },
    {
      "epoch": 0.1084029086991651,
      "grad_norm": 0.7456140715511358,
      "learning_rate": 9.927777310099759e-06,
      "loss": 0.47580275535583494,
      "num_input_tokens_seen": 82831032,
      "step": 1610,
      "train_runtime": 9945.6731,
      "train_tokens_per_second": 8328.349
    },
    {
      "epoch": 0.10873956369512523,
      "grad_norm": 0.6966235182528311,
      "learning_rate": 9.927328837913976e-06,
      "loss": 0.4053511619567871,
      "num_input_tokens_seen": 83092240,
      "step": 1615,
      "train_runtime": 9978.2964,
      "train_tokens_per_second": 8327.297
    },
    {
      "epoch": 0.10907621869108537,
      "grad_norm": 0.8217861548198141,
      "learning_rate": 9.926878987816397e-06,
      "loss": 0.4055288791656494,
      "num_input_tokens_seen": 83362264,
      "step": 1620,
      "train_runtime": 10008.7882,
      "train_tokens_per_second": 8328.907
    },
    {
      "epoch": 0.10941287368704551,
      "grad_norm": 0.7653981052710573,
      "learning_rate": 9.92642775993282e-06,
      "loss": 0.397510290145874,
      "num_input_tokens_seen": 83644792,
      "step": 1625,
      "train_runtime": 10035.4134,
      "train_tokens_per_second": 8334.962
    },
    {
      "epoch": 0.10974952868300565,
      "grad_norm": 0.7578341106908358,
      "learning_rate": 9.925975154389428e-06,
      "loss": 0.3877716541290283,
      "num_input_tokens_seen": 83890080,
      "step": 1630,
      "train_runtime": 10066.363,
      "train_tokens_per_second": 8333.703
    },
    {
      "epoch": 0.11008618367896579,
      "grad_norm": 0.7707683942279778,
      "learning_rate": 9.925521171312792e-06,
      "loss": 0.40496349334716797,
      "num_input_tokens_seen": 84153976,
      "step": 1635,
      "train_runtime": 10099.0899,
      "train_tokens_per_second": 8332.828
    },
    {
      "epoch": 0.11042283867492593,
      "grad_norm": 0.7292562521470469,
      "learning_rate": 9.925065810829868e-06,
      "loss": 0.4357038974761963,
      "num_input_tokens_seen": 84381880,
      "step": 1640,
      "train_runtime": 10124.5187,
      "train_tokens_per_second": 8334.409
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 0.8204033131822,
      "learning_rate": 9.924609073067994e-06,
      "loss": 0.43710765838623045,
      "num_input_tokens_seen": 84613328,
      "step": 1645,
      "train_runtime": 10156.722,
      "train_tokens_per_second": 8330.771
    },
    {
      "epoch": 0.11109614866684622,
      "grad_norm": 0.7663727702387767,
      "learning_rate": 9.924150958154897e-06,
      "loss": 0.41872057914733884,
      "num_input_tokens_seen": 84877888,
      "step": 1650,
      "train_runtime": 10186.5771,
      "train_tokens_per_second": 8332.327
    },
    {
      "epoch": 0.11143280366280636,
      "grad_norm": 0.7942018933426384,
      "learning_rate": 9.923691466218686e-06,
      "loss": 0.4052696228027344,
      "num_input_tokens_seen": 85131784,
      "step": 1655,
      "train_runtime": 10216.5617,
      "train_tokens_per_second": 8332.723
    },
    {
      "epoch": 0.1117694586587665,
      "grad_norm": 0.7985290300811083,
      "learning_rate": 9.923230597387856e-06,
      "loss": 0.4305591583251953,
      "num_input_tokens_seen": 85419384,
      "step": 1660,
      "train_runtime": 10250.7288,
      "train_tokens_per_second": 8333.006
    },
    {
      "epoch": 0.11210611365472664,
      "grad_norm": 1.4217572618185863,
      "learning_rate": 9.922768351791289e-06,
      "loss": 0.4476034641265869,
      "num_input_tokens_seen": 85674472,
      "step": 1665,
      "train_runtime": 10281.3775,
      "train_tokens_per_second": 8332.976
    },
    {
      "epoch": 0.11244276865068678,
      "grad_norm": 0.7646869825581092,
      "learning_rate": 9.92230472955825e-06,
      "loss": 0.39969208240509035,
      "num_input_tokens_seen": 85930496,
      "step": 1670,
      "train_runtime": 10311.4557,
      "train_tokens_per_second": 8333.498
    },
    {
      "epoch": 0.11277942364664692,
      "grad_norm": 0.7898609240284816,
      "learning_rate": 9.92183973081839e-06,
      "loss": 0.39197335243225095,
      "num_input_tokens_seen": 86184832,
      "step": 1675,
      "train_runtime": 10340.8965,
      "train_tokens_per_second": 8334.368
    },
    {
      "epoch": 0.11311607864260706,
      "grad_norm": 0.7598032395642297,
      "learning_rate": 9.921373355701743e-06,
      "loss": 0.4598580837249756,
      "num_input_tokens_seen": 86459784,
      "step": 1680,
      "train_runtime": 10375.9593,
      "train_tokens_per_second": 8332.703
    },
    {
      "epoch": 0.1134527336385672,
      "grad_norm": 0.7491091093526783,
      "learning_rate": 9.920905604338732e-06,
      "loss": 0.4579578399658203,
      "num_input_tokens_seen": 86723520,
      "step": 1685,
      "train_runtime": 10403.3507,
      "train_tokens_per_second": 8336.114
    },
    {
      "epoch": 0.11378938863452734,
      "grad_norm": 0.7761767193711012,
      "learning_rate": 9.920436476860158e-06,
      "loss": 0.3870235443115234,
      "num_input_tokens_seen": 86960256,
      "step": 1690,
      "train_runtime": 10434.3407,
      "train_tokens_per_second": 8334.044
    },
    {
      "epoch": 0.11412604363048748,
      "grad_norm": 0.760078494191473,
      "learning_rate": 9.919965973397214e-06,
      "loss": 0.41385498046875,
      "num_input_tokens_seen": 87228424,
      "step": 1695,
      "train_runtime": 10461.1175,
      "train_tokens_per_second": 8338.347
    },
    {
      "epoch": 0.11446269862644762,
      "grad_norm": 0.8399991434344031,
      "learning_rate": 9.919494094081475e-06,
      "loss": 0.45699753761291506,
      "num_input_tokens_seen": 87459200,
      "step": 1700,
      "train_runtime": 10490.7039,
      "train_tokens_per_second": 8336.829
    },
    {
      "epoch": 0.11479935362240776,
      "grad_norm": 0.8539877349848104,
      "learning_rate": 9.919020839044899e-06,
      "loss": 0.4074555397033691,
      "num_input_tokens_seen": 87748232,
      "step": 1705,
      "train_runtime": 10522.8336,
      "train_tokens_per_second": 8338.841
    },
    {
      "epoch": 0.1151360086183679,
      "grad_norm": 0.7251661451050594,
      "learning_rate": 9.918546208419832e-06,
      "loss": 0.39636778831481934,
      "num_input_tokens_seen": 88001936,
      "step": 1710,
      "train_runtime": 10554.3031,
      "train_tokens_per_second": 8338.015
    },
    {
      "epoch": 0.11547266361432804,
      "grad_norm": 0.8225783440826857,
      "learning_rate": 9.918070202339e-06,
      "loss": 0.40996241569519043,
      "num_input_tokens_seen": 88278936,
      "step": 1715,
      "train_runtime": 10584.6594,
      "train_tokens_per_second": 8340.272
    },
    {
      "epoch": 0.11580931861028818,
      "grad_norm": 0.7554495580709869,
      "learning_rate": 9.91759282093552e-06,
      "loss": 0.40645623207092285,
      "num_input_tokens_seen": 88547896,
      "step": 1720,
      "train_runtime": 10614.7629,
      "train_tokens_per_second": 8341.957
    },
    {
      "epoch": 0.11614597360624832,
      "grad_norm": 0.8766967218955909,
      "learning_rate": 9.917114064342888e-06,
      "loss": 0.3911132335662842,
      "num_input_tokens_seen": 88782328,
      "step": 1725,
      "train_runtime": 10644.478,
      "train_tokens_per_second": 8340.693
    },
    {
      "epoch": 0.11648262860220845,
      "grad_norm": 0.7084523348690432,
      "learning_rate": 9.916633932694988e-06,
      "loss": 0.3835343599319458,
      "num_input_tokens_seen": 89039816,
      "step": 1730,
      "train_runtime": 10677.2698,
      "train_tokens_per_second": 8339.193
    },
    {
      "epoch": 0.1168192835981686,
      "grad_norm": 0.8523521456079981,
      "learning_rate": 9.916152426126086e-06,
      "loss": 0.38129143714904784,
      "num_input_tokens_seen": 89303968,
      "step": 1735,
      "train_runtime": 10706.699,
      "train_tokens_per_second": 8340.943
    },
    {
      "epoch": 0.11715593859412873,
      "grad_norm": 0.7576776984545847,
      "learning_rate": 9.915669544770837e-06,
      "loss": 0.41629552841186523,
      "num_input_tokens_seen": 89593040,
      "step": 1740,
      "train_runtime": 10741.1891,
      "train_tokens_per_second": 8341.073
    },
    {
      "epoch": 0.11749259359008887,
      "grad_norm": 0.728187231019083,
      "learning_rate": 9.915185288764272e-06,
      "loss": 0.39937624931335447,
      "num_input_tokens_seen": 89877672,
      "step": 1745,
      "train_runtime": 10774.7128,
      "train_tokens_per_second": 8341.538
    },
    {
      "epoch": 0.11782924858604901,
      "grad_norm": 0.8426151020843394,
      "learning_rate": 9.914699658241815e-06,
      "loss": 0.38359196186065675,
      "num_input_tokens_seen": 90140712,
      "step": 1750,
      "train_runtime": 10806.6463,
      "train_tokens_per_second": 8341.229
    },
    {
      "epoch": 0.11816590358200915,
      "grad_norm": 0.8429872913645912,
      "learning_rate": 9.91421265333927e-06,
      "loss": 0.4293849945068359,
      "num_input_tokens_seen": 90414104,
      "step": 1755,
      "train_runtime": 10841.655,
      "train_tokens_per_second": 8339.511
    },
    {
      "epoch": 0.11850255857796929,
      "grad_norm": 0.7657350475402445,
      "learning_rate": 9.913724274192829e-06,
      "loss": 0.40404348373413085,
      "num_input_tokens_seen": 90681136,
      "step": 1760,
      "train_runtime": 10874.6853,
      "train_tokens_per_second": 8338.737
    },
    {
      "epoch": 0.11883921357392943,
      "grad_norm": 0.8015552044857566,
      "learning_rate": 9.91323452093906e-06,
      "loss": 0.3948627471923828,
      "num_input_tokens_seen": 90940000,
      "step": 1765,
      "train_runtime": 10908.3237,
      "train_tokens_per_second": 8336.753
    },
    {
      "epoch": 0.11917586856988957,
      "grad_norm": 0.7117034825735634,
      "learning_rate": 9.912743393714927e-06,
      "loss": 0.3683266878128052,
      "num_input_tokens_seen": 91231056,
      "step": 1770,
      "train_runtime": 10938.095,
      "train_tokens_per_second": 8340.671
    },
    {
      "epoch": 0.11951252356584972,
      "grad_norm": 0.75155829599063,
      "learning_rate": 9.912250892657769e-06,
      "loss": 0.4033808708190918,
      "num_input_tokens_seen": 91477848,
      "step": 1775,
      "train_runtime": 10965.8711,
      "train_tokens_per_second": 8342.05
    },
    {
      "epoch": 0.11984917856180986,
      "grad_norm": 0.81832175455793,
      "learning_rate": 9.911757017905312e-06,
      "loss": 0.41222515106201174,
      "num_input_tokens_seen": 91731256,
      "step": 1780,
      "train_runtime": 10995.6212,
      "train_tokens_per_second": 8342.526
    },
    {
      "epoch": 0.12018583355777,
      "grad_norm": 0.7006697172731424,
      "learning_rate": 9.911261769595668e-06,
      "loss": 0.40465412139892576,
      "num_input_tokens_seen": 92004328,
      "step": 1785,
      "train_runtime": 11028.7345,
      "train_tokens_per_second": 8342.238
    },
    {
      "epoch": 0.12052248855373014,
      "grad_norm": 0.8864252294479316,
      "learning_rate": 9.910765147867332e-06,
      "loss": 0.40481910705566404,
      "num_input_tokens_seen": 92276744,
      "step": 1790,
      "train_runtime": 11059.6487,
      "train_tokens_per_second": 8343.551
    },
    {
      "epoch": 0.12085914354969028,
      "grad_norm": 0.7876251405437599,
      "learning_rate": 9.91026715285918e-06,
      "loss": 0.41905860900878905,
      "num_input_tokens_seen": 92533312,
      "step": 1795,
      "train_runtime": 11089.8395,
      "train_tokens_per_second": 8343.972
    },
    {
      "epoch": 0.12119579854565042,
      "grad_norm": 0.8494454851529082,
      "learning_rate": 9.909767784710476e-06,
      "loss": 0.4465792179107666,
      "num_input_tokens_seen": 92833048,
      "step": 1800,
      "train_runtime": 11121.4387,
      "train_tokens_per_second": 8347.216
    },
    {
      "epoch": 0.12153245354161056,
      "grad_norm": 0.7463694322262185,
      "learning_rate": 9.909267043560868e-06,
      "loss": 0.4266531944274902,
      "num_input_tokens_seen": 93102624,
      "step": 1805,
      "train_runtime": 11151.5667,
      "train_tokens_per_second": 8348.838
    },
    {
      "epoch": 0.1218691085375707,
      "grad_norm": 0.7752946788315056,
      "learning_rate": 9.908764929550388e-06,
      "loss": 0.4193249702453613,
      "num_input_tokens_seen": 93364528,
      "step": 1810,
      "train_runtime": 11180.6304,
      "train_tokens_per_second": 8350.56
    },
    {
      "epoch": 0.12220576353353084,
      "grad_norm": 0.8634677242215414,
      "learning_rate": 9.908261442819444e-06,
      "loss": 0.41878414154052734,
      "num_input_tokens_seen": 93589632,
      "step": 1815,
      "train_runtime": 11209.1351,
      "train_tokens_per_second": 8349.407
    },
    {
      "epoch": 0.12254241852949098,
      "grad_norm": 0.74830543384422,
      "learning_rate": 9.907756583508842e-06,
      "loss": 0.40120534896850585,
      "num_input_tokens_seen": 93874864,
      "step": 1820,
      "train_runtime": 11237.7498,
      "train_tokens_per_second": 8353.529
    },
    {
      "epoch": 0.12287907352545112,
      "grad_norm": 0.7091337282069106,
      "learning_rate": 9.907250351759761e-06,
      "loss": 0.41159734725952146,
      "num_input_tokens_seen": 94132424,
      "step": 1825,
      "train_runtime": 11269.953,
      "train_tokens_per_second": 8352.513
    },
    {
      "epoch": 0.12321572852141126,
      "grad_norm": 1.1559965925468667,
      "learning_rate": 9.906742747713766e-06,
      "loss": 0.4303712844848633,
      "num_input_tokens_seen": 94399632,
      "step": 1830,
      "train_runtime": 11302.5586,
      "train_tokens_per_second": 8352.059
    },
    {
      "epoch": 0.1235523835173714,
      "grad_norm": 0.9199249798214647,
      "learning_rate": 9.906233771512808e-06,
      "loss": 0.423277473449707,
      "num_input_tokens_seen": 94646592,
      "step": 1835,
      "train_runtime": 11332.2929,
      "train_tokens_per_second": 8351.937
    },
    {
      "epoch": 0.12388903851333154,
      "grad_norm": 0.8136871896083027,
      "learning_rate": 9.905723423299222e-06,
      "loss": 0.44223723411560056,
      "num_input_tokens_seen": 94921232,
      "step": 1840,
      "train_runtime": 11367.3332,
      "train_tokens_per_second": 8350.352
    },
    {
      "epoch": 0.12422569350929168,
      "grad_norm": 0.8008387867374361,
      "learning_rate": 9.905211703215723e-06,
      "loss": 0.3883598566055298,
      "num_input_tokens_seen": 95177072,
      "step": 1845,
      "train_runtime": 11393.584,
      "train_tokens_per_second": 8353.567
    },
    {
      "epoch": 0.12456234850525182,
      "grad_norm": 0.7678175917226583,
      "learning_rate": 9.904698611405412e-06,
      "loss": 0.4303601264953613,
      "num_input_tokens_seen": 95449896,
      "step": 1850,
      "train_runtime": 11427.601,
      "train_tokens_per_second": 8352.575
    },
    {
      "epoch": 0.12489900350121196,
      "grad_norm": 0.6604047419132149,
      "learning_rate": 9.904184148011774e-06,
      "loss": 0.41721553802490235,
      "num_input_tokens_seen": 95707104,
      "step": 1855,
      "train_runtime": 11456.5196,
      "train_tokens_per_second": 8353.942
    },
    {
      "epoch": 0.1252356584971721,
      "grad_norm": 0.7459303155902302,
      "learning_rate": 9.903668313178681e-06,
      "loss": 0.4423821926116943,
      "num_input_tokens_seen": 95970792,
      "step": 1860,
      "train_runtime": 11488.2448,
      "train_tokens_per_second": 8353.826
    },
    {
      "epoch": 0.12557231349313225,
      "grad_norm": 0.8294489681483559,
      "learning_rate": 9.903151107050377e-06,
      "loss": 0.393432092666626,
      "num_input_tokens_seen": 96203680,
      "step": 1865,
      "train_runtime": 11521.9544,
      "train_tokens_per_second": 8349.597
    },
    {
      "epoch": 0.1259089684890924,
      "grad_norm": 0.6997536196207081,
      "learning_rate": 9.902632529771502e-06,
      "loss": 0.4230193614959717,
      "num_input_tokens_seen": 96464984,
      "step": 1870,
      "train_runtime": 11553.9874,
      "train_tokens_per_second": 8349.064
    },
    {
      "epoch": 0.12624562348505253,
      "grad_norm": 0.957462890298792,
      "learning_rate": 9.902112581487074e-06,
      "loss": 0.4157410621643066,
      "num_input_tokens_seen": 96729912,
      "step": 1875,
      "train_runtime": 11583.8098,
      "train_tokens_per_second": 8350.44
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.8919983908613689,
      "learning_rate": 9.901591262342494e-06,
      "loss": 0.4416505813598633,
      "num_input_tokens_seen": 96968048,
      "step": 1880,
      "train_runtime": 11619.6646,
      "train_tokens_per_second": 8345.168
    },
    {
      "epoch": 0.1269189334769728,
      "grad_norm": 0.7825201502609969,
      "learning_rate": 9.901068572483548e-06,
      "loss": 0.4087967872619629,
      "num_input_tokens_seen": 97225200,
      "step": 1885,
      "train_runtime": 11651.5344,
      "train_tokens_per_second": 8344.412
    },
    {
      "epoch": 0.12725558847293295,
      "grad_norm": 0.7089079848229705,
      "learning_rate": 9.900544512056402e-06,
      "loss": 0.4591719627380371,
      "num_input_tokens_seen": 97502744,
      "step": 1890,
      "train_runtime": 11685.331,
      "train_tokens_per_second": 8344.029
    },
    {
      "epoch": 0.12759224346889309,
      "grad_norm": 0.691286620265617,
      "learning_rate": 9.900019081207612e-06,
      "loss": 0.4261483192443848,
      "num_input_tokens_seen": 97772680,
      "step": 1895,
      "train_runtime": 11719.5967,
      "train_tokens_per_second": 8342.666
    },
    {
      "epoch": 0.12792889846485322,
      "grad_norm": 0.8009377332816625,
      "learning_rate": 9.89949228008411e-06,
      "loss": 0.40768961906433104,
      "num_input_tokens_seen": 98047792,
      "step": 1900,
      "train_runtime": 11747.6827,
      "train_tokens_per_second": 8346.139
    },
    {
      "epoch": 0.12826555346081336,
      "grad_norm": 0.6967772267246226,
      "learning_rate": 9.898964108833216e-06,
      "loss": 0.41776623725891116,
      "num_input_tokens_seen": 98314624,
      "step": 1905,
      "train_runtime": 11780.9013,
      "train_tokens_per_second": 8345.255
    },
    {
      "epoch": 0.1286022084567735,
      "grad_norm": 0.7316298633356212,
      "learning_rate": 9.898434567602629e-06,
      "loss": 0.4049860954284668,
      "num_input_tokens_seen": 98576384,
      "step": 1910,
      "train_runtime": 11813.5093,
      "train_tokens_per_second": 8344.378
    },
    {
      "epoch": 0.12893886345273364,
      "grad_norm": 0.8356839601083614,
      "learning_rate": 9.897903656540437e-06,
      "loss": 0.4167312145233154,
      "num_input_tokens_seen": 98833312,
      "step": 1915,
      "train_runtime": 11841.7046,
      "train_tokens_per_second": 8346.207
    },
    {
      "epoch": 0.12927551844869378,
      "grad_norm": 0.8775231657795305,
      "learning_rate": 9.897371375795103e-06,
      "loss": 0.42856712341308595,
      "num_input_tokens_seen": 99093752,
      "step": 1920,
      "train_runtime": 11871.3614,
      "train_tokens_per_second": 8347.295
    },
    {
      "epoch": 0.12961217344465392,
      "grad_norm": 0.7120448976075279,
      "learning_rate": 9.896837725515484e-06,
      "loss": 0.4356838226318359,
      "num_input_tokens_seen": 99336888,
      "step": 1925,
      "train_runtime": 11903.1899,
      "train_tokens_per_second": 8345.401
    },
    {
      "epoch": 0.12994882844061406,
      "grad_norm": 0.660400937964174,
      "learning_rate": 9.89630270585081e-06,
      "loss": 0.38520548343658445,
      "num_input_tokens_seen": 99586968,
      "step": 1930,
      "train_runtime": 11935.3123,
      "train_tokens_per_second": 8343.893
    },
    {
      "epoch": 0.1302854834365742,
      "grad_norm": 0.9771836514282706,
      "learning_rate": 9.895766316950694e-06,
      "loss": 0.4136802673339844,
      "num_input_tokens_seen": 99843960,
      "step": 1935,
      "train_runtime": 11965.7896,
      "train_tokens_per_second": 8344.118
    },
    {
      "epoch": 0.13062213843253434,
      "grad_norm": 0.7539967395793588,
      "learning_rate": 9.89522855896514e-06,
      "loss": 0.4093136787414551,
      "num_input_tokens_seen": 100108688,
      "step": 1940,
      "train_runtime": 11998.5911,
      "train_tokens_per_second": 8343.37
    },
    {
      "epoch": 0.13095879342849448,
      "grad_norm": 0.7598656245274686,
      "learning_rate": 9.89468943204453e-06,
      "loss": 0.4118832588195801,
      "num_input_tokens_seen": 100351920,
      "step": 1945,
      "train_runtime": 12024.9012,
      "train_tokens_per_second": 8345.343
    },
    {
      "epoch": 0.13129544842445462,
      "grad_norm": 0.8802035558512582,
      "learning_rate": 9.89414893633963e-06,
      "loss": 0.38923780918121337,
      "num_input_tokens_seen": 100608152,
      "step": 1950,
      "train_runtime": 12056.119,
      "train_tokens_per_second": 8344.987
    },
    {
      "epoch": 0.13163210342041476,
      "grad_norm": 0.6993588667272593,
      "learning_rate": 9.893607072001584e-06,
      "loss": 0.40406494140625,
      "num_input_tokens_seen": 100876392,
      "step": 1955,
      "train_runtime": 12089.5357,
      "train_tokens_per_second": 8344.108
    },
    {
      "epoch": 0.1319687584163749,
      "grad_norm": 0.7162615293211249,
      "learning_rate": 9.893063839181925e-06,
      "loss": 0.41468162536621095,
      "num_input_tokens_seen": 101126448,
      "step": 1960,
      "train_runtime": 12121.6208,
      "train_tokens_per_second": 8342.651
    },
    {
      "epoch": 0.13230541341233504,
      "grad_norm": 0.7148397414305018,
      "learning_rate": 9.892519238032566e-06,
      "loss": 0.42147326469421387,
      "num_input_tokens_seen": 101396560,
      "step": 1965,
      "train_runtime": 12154.3278,
      "train_tokens_per_second": 8342.424
    },
    {
      "epoch": 0.13264206840829518,
      "grad_norm": 0.7672706965482106,
      "learning_rate": 9.891973268705804e-06,
      "loss": 0.4453685760498047,
      "num_input_tokens_seen": 101648168,
      "step": 1970,
      "train_runtime": 12185.6548,
      "train_tokens_per_second": 8341.625
    },
    {
      "epoch": 0.13297872340425532,
      "grad_norm": 0.7337870041127343,
      "learning_rate": 9.891425931354316e-06,
      "loss": 0.4200016975402832,
      "num_input_tokens_seen": 101919192,
      "step": 1975,
      "train_runtime": 12215.0817,
      "train_tokens_per_second": 8343.718
    },
    {
      "epoch": 0.13331537840021546,
      "grad_norm": 0.7772656204454288,
      "learning_rate": 9.890877226131165e-06,
      "loss": 0.40060834884643554,
      "num_input_tokens_seen": 102190112,
      "step": 1980,
      "train_runtime": 12245.3234,
      "train_tokens_per_second": 8345.236
    },
    {
      "epoch": 0.1336520333961756,
      "grad_norm": 0.8591686612574334,
      "learning_rate": 9.890327153189793e-06,
      "loss": 0.42541799545288084,
      "num_input_tokens_seen": 102434528,
      "step": 1985,
      "train_runtime": 12278.5421,
      "train_tokens_per_second": 8342.564
    },
    {
      "epoch": 0.13398868839213574,
      "grad_norm": 0.7597017042322285,
      "learning_rate": 9.889775712684026e-06,
      "loss": 0.41676950454711914,
      "num_input_tokens_seen": 102692776,
      "step": 1990,
      "train_runtime": 12306.5786,
      "train_tokens_per_second": 8344.543
    },
    {
      "epoch": 0.13432534338809587,
      "grad_norm": 0.7431688392390488,
      "learning_rate": 9.889222904768074e-06,
      "loss": 0.42887296676635744,
      "num_input_tokens_seen": 102949176,
      "step": 1995,
      "train_runtime": 12336.7322,
      "train_tokens_per_second": 8344.931
    },
    {
      "epoch": 0.13466199838405601,
      "grad_norm": 0.7961038504309405,
      "learning_rate": 9.888668729596528e-06,
      "loss": 0.4121106624603271,
      "num_input_tokens_seen": 103177504,
      "step": 2000,
      "train_runtime": 12366.0366,
      "train_tokens_per_second": 8343.62
    },
    {
      "epoch": 0.13499865338001615,
      "grad_norm": 1.0046132576358522,
      "learning_rate": 9.88811318732436e-06,
      "loss": 0.4177651882171631,
      "num_input_tokens_seen": 103421200,
      "step": 2005,
      "train_runtime": 12394.5481,
      "train_tokens_per_second": 8344.088
    },
    {
      "epoch": 0.1353353083759763,
      "grad_norm": 0.8118066742923531,
      "learning_rate": 9.887556278106927e-06,
      "loss": 0.42879652976989746,
      "num_input_tokens_seen": 103701232,
      "step": 2010,
      "train_runtime": 12426.3426,
      "train_tokens_per_second": 8345.274
    },
    {
      "epoch": 0.13567196337193643,
      "grad_norm": 0.7870621017032995,
      "learning_rate": 9.886998002099964e-06,
      "loss": 0.4209719181060791,
      "num_input_tokens_seen": 103961272,
      "step": 2015,
      "train_runtime": 12458.3865,
      "train_tokens_per_second": 8344.682
    },
    {
      "epoch": 0.13600861836789657,
      "grad_norm": 0.848913021773478,
      "learning_rate": 9.886438359459595e-06,
      "loss": 0.44187397956848146,
      "num_input_tokens_seen": 104208248,
      "step": 2020,
      "train_runtime": 12489.7062,
      "train_tokens_per_second": 8343.531
    },
    {
      "epoch": 0.1363452733638567,
      "grad_norm": 1.0447069405048255,
      "learning_rate": 9.885877350342319e-06,
      "loss": 0.38938353061676023,
      "num_input_tokens_seen": 104477024,
      "step": 2025,
      "train_runtime": 12525.2742,
      "train_tokens_per_second": 8341.296
    },
    {
      "epoch": 0.13668192835981685,
      "grad_norm": 0.9285864493238545,
      "learning_rate": 9.885314974905024e-06,
      "loss": 0.3803610324859619,
      "num_input_tokens_seen": 104759448,
      "step": 2030,
      "train_runtime": 12554.883,
      "train_tokens_per_second": 8344.12
    },
    {
      "epoch": 0.137018583355777,
      "grad_norm": 0.9166936556761263,
      "learning_rate": 9.884751233304973e-06,
      "loss": 0.4000194549560547,
      "num_input_tokens_seen": 105003776,
      "step": 2035,
      "train_runtime": 12581.0869,
      "train_tokens_per_second": 8346.161
    },
    {
      "epoch": 0.13735523835173713,
      "grad_norm": 0.7822665039624449,
      "learning_rate": 9.884186125699816e-06,
      "loss": 0.4202897071838379,
      "num_input_tokens_seen": 105254512,
      "step": 2040,
      "train_runtime": 12610.4257,
      "train_tokens_per_second": 8346.626
    },
    {
      "epoch": 0.13769189334769727,
      "grad_norm": 0.7995486759092666,
      "learning_rate": 9.883619652247584e-06,
      "loss": 0.4045741081237793,
      "num_input_tokens_seen": 105509560,
      "step": 2045,
      "train_runtime": 12640.6911,
      "train_tokens_per_second": 8346.819
    },
    {
      "epoch": 0.1380285483436574,
      "grad_norm": 0.6055698250670518,
      "learning_rate": 9.883051813106687e-06,
      "loss": 0.37367587089538573,
      "num_input_tokens_seen": 105767840,
      "step": 2050,
      "train_runtime": 12670.2025,
      "train_tokens_per_second": 8347.762
    },
    {
      "epoch": 0.13836520333961755,
      "grad_norm": 0.8331923001022876,
      "learning_rate": 9.882482608435924e-06,
      "loss": 0.43723330497741697,
      "num_input_tokens_seen": 106024680,
      "step": 2055,
      "train_runtime": 12701.7428,
      "train_tokens_per_second": 8347.255
    },
    {
      "epoch": 0.1387018583355777,
      "grad_norm": 0.7801185191201977,
      "learning_rate": 9.881912038394468e-06,
      "loss": 0.41613245010375977,
      "num_input_tokens_seen": 106278568,
      "step": 2060,
      "train_runtime": 12734.9345,
      "train_tokens_per_second": 8345.435
    },
    {
      "epoch": 0.13903851333153783,
      "grad_norm": 0.7322534065345657,
      "learning_rate": 9.881340103141878e-06,
      "loss": 0.4059797763824463,
      "num_input_tokens_seen": 106556464,
      "step": 2065,
      "train_runtime": 12766.3018,
      "train_tokens_per_second": 8346.698
    },
    {
      "epoch": 0.13937516832749797,
      "grad_norm": 0.7530947493580775,
      "learning_rate": 9.880766802838092e-06,
      "loss": 0.395172119140625,
      "num_input_tokens_seen": 106801288,
      "step": 2070,
      "train_runtime": 12797.8825,
      "train_tokens_per_second": 8345.231
    },
    {
      "epoch": 0.1397118233234581,
      "grad_norm": 0.6973899544113225,
      "learning_rate": 9.880192137643434e-06,
      "loss": 0.40581817626953126,
      "num_input_tokens_seen": 107054904,
      "step": 2075,
      "train_runtime": 12825.0165,
      "train_tokens_per_second": 8347.35
    },
    {
      "epoch": 0.14004847831941827,
      "grad_norm": 0.8389898119163851,
      "learning_rate": 9.879616107718608e-06,
      "loss": 0.4061131000518799,
      "num_input_tokens_seen": 107317064,
      "step": 2080,
      "train_runtime": 12854.5372,
      "train_tokens_per_second": 8348.575
    },
    {
      "epoch": 0.1403851333153784,
      "grad_norm": 0.6609868932065065,
      "learning_rate": 9.879038713224695e-06,
      "loss": 0.41812806129455565,
      "num_input_tokens_seen": 107575976,
      "step": 2085,
      "train_runtime": 12883.7093,
      "train_tokens_per_second": 8349.767
    },
    {
      "epoch": 0.14072178831133855,
      "grad_norm": 0.7141145567939953,
      "learning_rate": 9.878459954323165e-06,
      "loss": 0.39441738128662107,
      "num_input_tokens_seen": 107848176,
      "step": 2090,
      "train_runtime": 12914.8677,
      "train_tokens_per_second": 8350.699
    },
    {
      "epoch": 0.1410584433072987,
      "grad_norm": 0.7743407529518335,
      "learning_rate": 9.877879831175865e-06,
      "loss": 0.3957905054092407,
      "num_input_tokens_seen": 108112480,
      "step": 2095,
      "train_runtime": 12951.4206,
      "train_tokens_per_second": 8347.538
    },
    {
      "epoch": 0.14139509830325883,
      "grad_norm": 0.7481819508931179,
      "learning_rate": 9.877298343945025e-06,
      "loss": 0.41532192230224607,
      "num_input_tokens_seen": 108369592,
      "step": 2100,
      "train_runtime": 12982.1178,
      "train_tokens_per_second": 8347.605
    },
    {
      "epoch": 0.14173175329921897,
      "grad_norm": 0.8370657935258616,
      "learning_rate": 9.876715492793256e-06,
      "loss": 0.40746293067932127,
      "num_input_tokens_seen": 108626672,
      "step": 2105,
      "train_runtime": 13015.1455,
      "train_tokens_per_second": 8346.174
    },
    {
      "epoch": 0.1420684082951791,
      "grad_norm": 0.7349321589707577,
      "learning_rate": 9.87613127788355e-06,
      "loss": 0.3871361494064331,
      "num_input_tokens_seen": 108870536,
      "step": 2110,
      "train_runtime": 13044.5559,
      "train_tokens_per_second": 8346.052
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.7818667140651504,
      "learning_rate": 9.87554569937928e-06,
      "loss": 0.4447975158691406,
      "num_input_tokens_seen": 109143896,
      "step": 2115,
      "train_runtime": 13076.6482,
      "train_tokens_per_second": 8346.473
    },
    {
      "epoch": 0.1427417182870994,
      "grad_norm": 0.7630325652992723,
      "learning_rate": 9.874958757444203e-06,
      "loss": 0.4441415786743164,
      "num_input_tokens_seen": 109427040,
      "step": 2120,
      "train_runtime": 13107.3242,
      "train_tokens_per_second": 8348.541
    },
    {
      "epoch": 0.14307837328305953,
      "grad_norm": 0.765183022751257,
      "learning_rate": 9.874370452242454e-06,
      "loss": 0.42551164627075194,
      "num_input_tokens_seen": 109712224,
      "step": 2125,
      "train_runtime": 13140.6664,
      "train_tokens_per_second": 8349.061
    },
    {
      "epoch": 0.14341502827901967,
      "grad_norm": 0.7710166844972318,
      "learning_rate": 9.873780783938553e-06,
      "loss": 0.4402188301086426,
      "num_input_tokens_seen": 109988912,
      "step": 2130,
      "train_runtime": 13168.1017,
      "train_tokens_per_second": 8352.678
    },
    {
      "epoch": 0.1437516832749798,
      "grad_norm": 1.0241167585813946,
      "learning_rate": 9.873189752697396e-06,
      "loss": 0.3823207378387451,
      "num_input_tokens_seen": 110257848,
      "step": 2135,
      "train_runtime": 13196.8275,
      "train_tokens_per_second": 8354.875
    },
    {
      "epoch": 0.14408833827093995,
      "grad_norm": 0.8254779597912437,
      "learning_rate": 9.872597358684265e-06,
      "loss": 0.4399838447570801,
      "num_input_tokens_seen": 110520392,
      "step": 2140,
      "train_runtime": 13227.823,
      "train_tokens_per_second": 8355.146
    },
    {
      "epoch": 0.1444249932669001,
      "grad_norm": 0.7406798870502527,
      "learning_rate": 9.872003602064816e-06,
      "loss": 0.40682315826416016,
      "num_input_tokens_seen": 110805768,
      "step": 2145,
      "train_runtime": 13261.7334,
      "train_tokens_per_second": 8355.301
    },
    {
      "epoch": 0.14476164826286023,
      "grad_norm": 0.6213969075233692,
      "learning_rate": 9.8714084830051e-06,
      "loss": 0.38729090690612794,
      "num_input_tokens_seen": 111079544,
      "step": 2150,
      "train_runtime": 13293.992,
      "train_tokens_per_second": 8355.62
    },
    {
      "epoch": 0.14509830325882037,
      "grad_norm": 0.7212577451053205,
      "learning_rate": 9.870812001671533e-06,
      "loss": 0.3914778709411621,
      "num_input_tokens_seen": 111356264,
      "step": 2155,
      "train_runtime": 13327.8237,
      "train_tokens_per_second": 8355.172
    },
    {
      "epoch": 0.1454349582547805,
      "grad_norm": 0.8443456119280635,
      "learning_rate": 9.870214158230922e-06,
      "loss": 0.41277437210083007,
      "num_input_tokens_seen": 111636960,
      "step": 2160,
      "train_runtime": 13359.9158,
      "train_tokens_per_second": 8356.113
    },
    {
      "epoch": 0.14577161325074064,
      "grad_norm": 0.7043485937302846,
      "learning_rate": 9.869614952850454e-06,
      "loss": 0.4010777473449707,
      "num_input_tokens_seen": 111906848,
      "step": 2165,
      "train_runtime": 13391.4025,
      "train_tokens_per_second": 8356.619
    },
    {
      "epoch": 0.14610826824670078,
      "grad_norm": 0.6524957596530463,
      "learning_rate": 9.86901438569769e-06,
      "loss": 0.41638774871826173,
      "num_input_tokens_seen": 112186552,
      "step": 2170,
      "train_runtime": 13424.1312,
      "train_tokens_per_second": 8357.081
    },
    {
      "epoch": 0.14644492324266092,
      "grad_norm": 0.8112845976373292,
      "learning_rate": 9.86841245694058e-06,
      "loss": 0.41776609420776367,
      "num_input_tokens_seen": 112408752,
      "step": 2175,
      "train_runtime": 13450.1995,
      "train_tokens_per_second": 8357.404
    },
    {
      "epoch": 0.14678157823862106,
      "grad_norm": 0.8606617904943867,
      "learning_rate": 9.86780916674745e-06,
      "loss": 0.4430191993713379,
      "num_input_tokens_seen": 112645952,
      "step": 2180,
      "train_runtime": 13478.7895,
      "train_tokens_per_second": 8357.275
    },
    {
      "epoch": 0.1471182332345812,
      "grad_norm": 0.8842396794972677,
      "learning_rate": 9.86720451528701e-06,
      "loss": 0.40068764686584474,
      "num_input_tokens_seen": 112882360,
      "step": 2185,
      "train_runtime": 13514.5322,
      "train_tokens_per_second": 8352.665
    },
    {
      "epoch": 0.14745488823054134,
      "grad_norm": 0.746978013520797,
      "learning_rate": 9.866598502728348e-06,
      "loss": 0.41660342216491697,
      "num_input_tokens_seen": 113134120,
      "step": 2190,
      "train_runtime": 13541.2838,
      "train_tokens_per_second": 8354.756
    },
    {
      "epoch": 0.14779154322650148,
      "grad_norm": 0.8031996733913954,
      "learning_rate": 9.865991129240931e-06,
      "loss": 0.4073335647583008,
      "num_input_tokens_seen": 113379888,
      "step": 2195,
      "train_runtime": 13571.2963,
      "train_tokens_per_second": 8354.389
    },
    {
      "epoch": 0.14812819822246162,
      "grad_norm": 0.8036035707290867,
      "learning_rate": 9.865382394994614e-06,
      "loss": 0.40634775161743164,
      "num_input_tokens_seen": 113636928,
      "step": 2200,
      "train_runtime": 13604.3182,
      "train_tokens_per_second": 8353.004
    },
    {
      "epoch": 0.14846485321842176,
      "grad_norm": 0.6611978015323592,
      "learning_rate": 9.864772300159622e-06,
      "loss": 0.3928510665893555,
      "num_input_tokens_seen": 113898728,
      "step": 2205,
      "train_runtime": 13632.3298,
      "train_tokens_per_second": 8355.045
    },
    {
      "epoch": 0.1488015082143819,
      "grad_norm": 0.7883125282683513,
      "learning_rate": 9.864160844906571e-06,
      "loss": 0.4255541801452637,
      "num_input_tokens_seen": 114163760,
      "step": 2210,
      "train_runtime": 13663.888,
      "train_tokens_per_second": 8355.145
    },
    {
      "epoch": 0.14913816321034204,
      "grad_norm": 0.7504352713128878,
      "learning_rate": 9.863548029406449e-06,
      "loss": 0.41379313468933104,
      "num_input_tokens_seen": 114422912,
      "step": 2215,
      "train_runtime": 13694.7569,
      "train_tokens_per_second": 8355.235
    },
    {
      "epoch": 0.14947481820630218,
      "grad_norm": 0.7073722468012431,
      "learning_rate": 9.86293385383063e-06,
      "loss": 0.3818849563598633,
      "num_input_tokens_seen": 114656216,
      "step": 2220,
      "train_runtime": 13725.4312,
      "train_tokens_per_second": 8353.56
    },
    {
      "epoch": 0.14981147320226232,
      "grad_norm": 0.7893042155355129,
      "learning_rate": 9.862318318350865e-06,
      "loss": 0.4220993995666504,
      "num_input_tokens_seen": 114923104,
      "step": 2225,
      "train_runtime": 13754.1623,
      "train_tokens_per_second": 8355.515
    },
    {
      "epoch": 0.15014812819822246,
      "grad_norm": 0.7021058292909881,
      "learning_rate": 9.861701423139287e-06,
      "loss": 0.37604031562805174,
      "num_input_tokens_seen": 115197616,
      "step": 2230,
      "train_runtime": 13786.1926,
      "train_tokens_per_second": 8356.014
    },
    {
      "epoch": 0.1504847831941826,
      "grad_norm": 0.702955568548746,
      "learning_rate": 9.861083168368411e-06,
      "loss": 0.41863412857055665,
      "num_input_tokens_seen": 115457504,
      "step": 2235,
      "train_runtime": 13816.1114,
      "train_tokens_per_second": 8356.729
    },
    {
      "epoch": 0.15082143819014274,
      "grad_norm": 0.8438110088949446,
      "learning_rate": 9.860463554211125e-06,
      "loss": 0.3947760581970215,
      "num_input_tokens_seen": 115716824,
      "step": 2240,
      "train_runtime": 13850.1178,
      "train_tokens_per_second": 8354.934
    },
    {
      "epoch": 0.15115809318610288,
      "grad_norm": 0.9309158170507835,
      "learning_rate": 9.859842580840706e-06,
      "loss": 0.4434503078460693,
      "num_input_tokens_seen": 115982712,
      "step": 2245,
      "train_runtime": 13878.5918,
      "train_tokens_per_second": 8356.951
    },
    {
      "epoch": 0.15149474818206302,
      "grad_norm": 0.648377479053166,
      "learning_rate": 9.859220248430806e-06,
      "loss": 0.41293888092041015,
      "num_input_tokens_seen": 116276400,
      "step": 2250,
      "train_runtime": 13912.8462,
      "train_tokens_per_second": 8357.485
    },
    {
      "epoch": 0.15183140317802316,
      "grad_norm": 0.7545057347295436,
      "learning_rate": 9.85859655715546e-06,
      "loss": 0.37118537425994874,
      "num_input_tokens_seen": 116526992,
      "step": 2255,
      "train_runtime": 13941.0615,
      "train_tokens_per_second": 8358.545
    },
    {
      "epoch": 0.1521680581739833,
      "grad_norm": 0.7729212085450236,
      "learning_rate": 9.857971507189077e-06,
      "loss": 0.39316248893737793,
      "num_input_tokens_seen": 116776808,
      "step": 2260,
      "train_runtime": 13969.8482,
      "train_tokens_per_second": 8359.204
    },
    {
      "epoch": 0.15250471316994343,
      "grad_norm": 0.8826568709372734,
      "learning_rate": 9.857345098706455e-06,
      "loss": 0.4055880069732666,
      "num_input_tokens_seen": 117060496,
      "step": 2265,
      "train_runtime": 14004.1843,
      "train_tokens_per_second": 8358.966
    },
    {
      "epoch": 0.15284136816590357,
      "grad_norm": 0.8013723715222983,
      "learning_rate": 9.856717331882765e-06,
      "loss": 0.4058811187744141,
      "num_input_tokens_seen": 117309896,
      "step": 2270,
      "train_runtime": 14029.3044,
      "train_tokens_per_second": 8361.776
    },
    {
      "epoch": 0.1531780231618637,
      "grad_norm": 0.6868269410063399,
      "learning_rate": 9.85608820689356e-06,
      "loss": 0.38027167320251465,
      "num_input_tokens_seen": 117569008,
      "step": 2275,
      "train_runtime": 14062.3332,
      "train_tokens_per_second": 8360.562
    },
    {
      "epoch": 0.15351467815782385,
      "grad_norm": 0.7213676412656342,
      "learning_rate": 9.855457723914772e-06,
      "loss": 0.40387115478515623,
      "num_input_tokens_seen": 117839528,
      "step": 2280,
      "train_runtime": 14088.6895,
      "train_tokens_per_second": 8364.123
    },
    {
      "epoch": 0.153851333153784,
      "grad_norm": 0.8104279540970285,
      "learning_rate": 9.854825883122713e-06,
      "loss": 0.39426159858703613,
      "num_input_tokens_seen": 118083448,
      "step": 2285,
      "train_runtime": 14118.3056,
      "train_tokens_per_second": 8363.854
    },
    {
      "epoch": 0.15418798814974413,
      "grad_norm": 0.6458074533172264,
      "learning_rate": 9.85419268469408e-06,
      "loss": 0.390856409072876,
      "num_input_tokens_seen": 118324016,
      "step": 2290,
      "train_runtime": 14150.3507,
      "train_tokens_per_second": 8361.914
    },
    {
      "epoch": 0.15452464314570427,
      "grad_norm": 0.7105372666117817,
      "learning_rate": 9.85355812880594e-06,
      "loss": 0.3970534324645996,
      "num_input_tokens_seen": 118585120,
      "step": 2295,
      "train_runtime": 14176.234,
      "train_tokens_per_second": 8365.065
    },
    {
      "epoch": 0.1548612981416644,
      "grad_norm": 0.6495283791072095,
      "learning_rate": 9.852922215635747e-06,
      "loss": 0.37650208473205565,
      "num_input_tokens_seen": 118854080,
      "step": 2300,
      "train_runtime": 14206.2183,
      "train_tokens_per_second": 8366.342
    },
    {
      "epoch": 0.15519795313762455,
      "grad_norm": 0.7158718038291049,
      "learning_rate": 9.85228494536133e-06,
      "loss": 0.41134071350097656,
      "num_input_tokens_seen": 119127784,
      "step": 2305,
      "train_runtime": 14235.6231,
      "train_tokens_per_second": 8368.287
    },
    {
      "epoch": 0.1555346081335847,
      "grad_norm": 0.6718223727373428,
      "learning_rate": 9.851646318160901e-06,
      "loss": 0.4059895992279053,
      "num_input_tokens_seen": 119377120,
      "step": 2310,
      "train_runtime": 14270.8045,
      "train_tokens_per_second": 8365.129
    },
    {
      "epoch": 0.15587126312954483,
      "grad_norm": 0.6902729718726833,
      "learning_rate": 9.85100633421305e-06,
      "loss": 0.3960001468658447,
      "num_input_tokens_seen": 119612160,
      "step": 2315,
      "train_runtime": 14300.2105,
      "train_tokens_per_second": 8364.364
    },
    {
      "epoch": 0.15620791812550497,
      "grad_norm": 0.8505781560330091,
      "learning_rate": 9.850364993696749e-06,
      "loss": 0.42215538024902344,
      "num_input_tokens_seen": 119857400,
      "step": 2320,
      "train_runtime": 14328.793,
      "train_tokens_per_second": 8364.794
    },
    {
      "epoch": 0.15654457312146514,
      "grad_norm": 1.0148694707978583,
      "learning_rate": 9.849722296791339e-06,
      "loss": 0.4030285835266113,
      "num_input_tokens_seen": 120109528,
      "step": 2325,
      "train_runtime": 14361.9128,
      "train_tokens_per_second": 8363.059
    },
    {
      "epoch": 0.15688122811742528,
      "grad_norm": 0.7981048449843531,
      "learning_rate": 9.849078243676558e-06,
      "loss": 0.4219195365905762,
      "num_input_tokens_seen": 120353008,
      "step": 2330,
      "train_runtime": 14393.9855,
      "train_tokens_per_second": 8361.34
    },
    {
      "epoch": 0.15721788311338541,
      "grad_norm": 0.6425016509838443,
      "learning_rate": 9.848432834532506e-06,
      "loss": 0.4011368751525879,
      "num_input_tokens_seen": 120601872,
      "step": 2335,
      "train_runtime": 14428.1451,
      "train_tokens_per_second": 8358.793
    },
    {
      "epoch": 0.15755453810934555,
      "grad_norm": 0.6659597600221548,
      "learning_rate": 9.847786069539673e-06,
      "loss": 0.41481552124023435,
      "num_input_tokens_seen": 120857936,
      "step": 2340,
      "train_runtime": 14457.2145,
      "train_tokens_per_second": 8359.697
    },
    {
      "epoch": 0.1578911931053057,
      "grad_norm": 0.7145682563488723,
      "learning_rate": 9.847137948878926e-06,
      "loss": 0.41647539138793943,
      "num_input_tokens_seen": 121127688,
      "step": 2345,
      "train_runtime": 14486.1607,
      "train_tokens_per_second": 8361.614
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.7882319191306095,
      "learning_rate": 9.846488472731505e-06,
      "loss": 0.45836877822875977,
      "num_input_tokens_seen": 121419136,
      "step": 2350,
      "train_runtime": 14513.7961,
      "train_tokens_per_second": 8365.774
    },
    {
      "epoch": 0.15856450309722597,
      "grad_norm": 0.6739881193150407,
      "learning_rate": 9.845837641279038e-06,
      "loss": 0.4054107666015625,
      "num_input_tokens_seen": 121703856,
      "step": 2355,
      "train_runtime": 14547.4183,
      "train_tokens_per_second": 8366.011
    },
    {
      "epoch": 0.1589011580931861,
      "grad_norm": 0.7738343199961882,
      "learning_rate": 9.845185454703528e-06,
      "loss": 0.4128101825714111,
      "num_input_tokens_seen": 121954040,
      "step": 2360,
      "train_runtime": 14578.3944,
      "train_tokens_per_second": 8365.396
    },
    {
      "epoch": 0.15923781308914625,
      "grad_norm": 0.784723540487718,
      "learning_rate": 9.844531913187356e-06,
      "loss": 0.3837453842163086,
      "num_input_tokens_seen": 122206376,
      "step": 2365,
      "train_runtime": 14608.6241,
      "train_tokens_per_second": 8365.358
    },
    {
      "epoch": 0.1595744680851064,
      "grad_norm": 0.8179769514831658,
      "learning_rate": 9.84387701691328e-06,
      "loss": 0.4048604965209961,
      "num_input_tokens_seen": 122438744,
      "step": 2370,
      "train_runtime": 14634.4272,
      "train_tokens_per_second": 8366.487
    },
    {
      "epoch": 0.15991112308106653,
      "grad_norm": 0.9076743131807125,
      "learning_rate": 9.843220766064443e-06,
      "loss": 0.4239856243133545,
      "num_input_tokens_seen": 122684984,
      "step": 2375,
      "train_runtime": 14661.0495,
      "train_tokens_per_second": 8368.09
    },
    {
      "epoch": 0.16024777807702667,
      "grad_norm": 0.9159090534129751,
      "learning_rate": 9.842563160824364e-06,
      "loss": 0.4201195240020752,
      "num_input_tokens_seen": 122942904,
      "step": 2380,
      "train_runtime": 14695.7455,
      "train_tokens_per_second": 8365.884
    },
    {
      "epoch": 0.1605844330729868,
      "grad_norm": 0.760292004678385,
      "learning_rate": 9.841904201376939e-06,
      "loss": 0.4280568599700928,
      "num_input_tokens_seen": 123217032,
      "step": 2385,
      "train_runtime": 14728.957,
      "train_tokens_per_second": 8365.632
    },
    {
      "epoch": 0.16092108806894695,
      "grad_norm": 0.7051110448790836,
      "learning_rate": 9.841243887906443e-06,
      "loss": 0.39680407047271726,
      "num_input_tokens_seen": 123480128,
      "step": 2390,
      "train_runtime": 14764.7214,
      "train_tokens_per_second": 8363.187
    },
    {
      "epoch": 0.1612577430649071,
      "grad_norm": 0.7824900240036943,
      "learning_rate": 9.840582220597532e-06,
      "loss": 0.3842333793640137,
      "num_input_tokens_seen": 123735288,
      "step": 2395,
      "train_runtime": 14796.5395,
      "train_tokens_per_second": 8362.448
    },
    {
      "epoch": 0.16159439806086723,
      "grad_norm": 1.0609385653526975,
      "learning_rate": 9.839919199635238e-06,
      "loss": 0.4375351905822754,
      "num_input_tokens_seen": 123999616,
      "step": 2400,
      "train_runtime": 14826.8586,
      "train_tokens_per_second": 8363.175
    },
    {
      "epoch": 0.16193105305682737,
      "grad_norm": 0.721511205340686,
      "learning_rate": 9.839254825204973e-06,
      "loss": 0.41885924339294434,
      "num_input_tokens_seen": 124266200,
      "step": 2405,
      "train_runtime": 14859.2285,
      "train_tokens_per_second": 8362.897
    },
    {
      "epoch": 0.1622677080527875,
      "grad_norm": 1.126728462487787,
      "learning_rate": 9.838589097492527e-06,
      "loss": 0.4225168704986572,
      "num_input_tokens_seen": 124519656,
      "step": 2410,
      "train_runtime": 14889.9634,
      "train_tokens_per_second": 8362.657
    },
    {
      "epoch": 0.16260436304874765,
      "grad_norm": 0.8029190080555629,
      "learning_rate": 9.83792201668407e-06,
      "loss": 0.41396346092224123,
      "num_input_tokens_seen": 124786528,
      "step": 2415,
      "train_runtime": 14919.9354,
      "train_tokens_per_second": 8363.745
    },
    {
      "epoch": 0.16294101804470779,
      "grad_norm": 0.9805722145998425,
      "learning_rate": 9.837253582966146e-06,
      "loss": 0.42626590728759767,
      "num_input_tokens_seen": 125053880,
      "step": 2420,
      "train_runtime": 14947.6551,
      "train_tokens_per_second": 8366.12
    },
    {
      "epoch": 0.16327767304066793,
      "grad_norm": 0.8095745048805549,
      "learning_rate": 9.836583796525686e-06,
      "loss": 0.39048171043395996,
      "num_input_tokens_seen": 125271816,
      "step": 2425,
      "train_runtime": 14977.9408,
      "train_tokens_per_second": 8363.754
    },
    {
      "epoch": 0.16361432803662807,
      "grad_norm": 0.7288285128540476,
      "learning_rate": 9.835912657549986e-06,
      "loss": 0.44376583099365235,
      "num_input_tokens_seen": 125533632,
      "step": 2430,
      "train_runtime": 15006.3253,
      "train_tokens_per_second": 8365.381
    },
    {
      "epoch": 0.1639509830325882,
      "grad_norm": 1.3995790457343225,
      "learning_rate": 9.835240166226734e-06,
      "loss": 0.39373595714569093,
      "num_input_tokens_seen": 125784640,
      "step": 2435,
      "train_runtime": 15036.2518,
      "train_tokens_per_second": 8365.425
    },
    {
      "epoch": 0.16428763802854834,
      "grad_norm": 0.7229300838745704,
      "learning_rate": 9.834566322743987e-06,
      "loss": 0.4231441497802734,
      "num_input_tokens_seen": 126035560,
      "step": 2440,
      "train_runtime": 15066.6918,
      "train_tokens_per_second": 8365.178
    },
    {
      "epoch": 0.16462429302450848,
      "grad_norm": 0.6689751907400037,
      "learning_rate": 9.833891127290186e-06,
      "loss": 0.41379289627075194,
      "num_input_tokens_seen": 126319976,
      "step": 2445,
      "train_runtime": 15099.6551,
      "train_tokens_per_second": 8365.752
    },
    {
      "epoch": 0.16496094802046862,
      "grad_norm": 0.6349070556732227,
      "learning_rate": 9.833214580054145e-06,
      "loss": 0.3910360813140869,
      "num_input_tokens_seen": 126576440,
      "step": 2450,
      "train_runtime": 15129.8179,
      "train_tokens_per_second": 8366.025
    },
    {
      "epoch": 0.16529760301642876,
      "grad_norm": 0.6999597480811056,
      "learning_rate": 9.832536681225058e-06,
      "loss": 0.4366586685180664,
      "num_input_tokens_seen": 126841720,
      "step": 2455,
      "train_runtime": 15160.1904,
      "train_tokens_per_second": 8366.763
    },
    {
      "epoch": 0.1656342580123889,
      "grad_norm": 0.780197883502905,
      "learning_rate": 9.831857430992497e-06,
      "loss": 0.3875967741012573,
      "num_input_tokens_seen": 127087496,
      "step": 2460,
      "train_runtime": 15195.0424,
      "train_tokens_per_second": 8363.747
    },
    {
      "epoch": 0.16597091300834904,
      "grad_norm": 0.6353037517620753,
      "learning_rate": 9.831176829546416e-06,
      "loss": 0.39896020889282224,
      "num_input_tokens_seen": 127352280,
      "step": 2465,
      "train_runtime": 15221.4427,
      "train_tokens_per_second": 8366.637
    },
    {
      "epoch": 0.16630756800430918,
      "grad_norm": 0.8767774509889421,
      "learning_rate": 9.830494877077137e-06,
      "loss": 0.4484898090362549,
      "num_input_tokens_seen": 127640616,
      "step": 2470,
      "train_runtime": 15251.4896,
      "train_tokens_per_second": 8369.059
    },
    {
      "epoch": 0.16664422300026932,
      "grad_norm": 0.7985713845806428,
      "learning_rate": 9.82981157377537e-06,
      "loss": 0.41009368896484377,
      "num_input_tokens_seen": 127889752,
      "step": 2475,
      "train_runtime": 15279.6432,
      "train_tokens_per_second": 8369.944
    },
    {
      "epoch": 0.16698087799622946,
      "grad_norm": 0.7753361167046846,
      "learning_rate": 9.829126919832198e-06,
      "loss": 0.38720943927764895,
      "num_input_tokens_seen": 128144384,
      "step": 2480,
      "train_runtime": 15312.7899,
      "train_tokens_per_second": 8368.454
    },
    {
      "epoch": 0.1673175329921896,
      "grad_norm": 0.6507117723730391,
      "learning_rate": 9.82844091543908e-06,
      "loss": 0.4242893695831299,
      "num_input_tokens_seen": 128430480,
      "step": 2485,
      "train_runtime": 15341.0921,
      "train_tokens_per_second": 8371.665
    },
    {
      "epoch": 0.16765418798814974,
      "grad_norm": 0.7208515671621564,
      "learning_rate": 9.82775356078786e-06,
      "loss": 0.4002225399017334,
      "num_input_tokens_seen": 128682592,
      "step": 2490,
      "train_runtime": 15368.9218,
      "train_tokens_per_second": 8372.91
    },
    {
      "epoch": 0.16799084298410988,
      "grad_norm": 0.6531794052808017,
      "learning_rate": 9.82706485607075e-06,
      "loss": 0.34480421543121337,
      "num_input_tokens_seen": 128926400,
      "step": 2495,
      "train_runtime": 15402.2847,
      "train_tokens_per_second": 8370.602
    },
    {
      "epoch": 0.16832749798007002,
      "grad_norm": 0.7399507314850664,
      "learning_rate": 9.826374801480346e-06,
      "loss": 0.39345555305480956,
      "num_input_tokens_seen": 129162680,
      "step": 2500,
      "train_runtime": 15429.3373,
      "train_tokens_per_second": 8371.24
    },
    {
      "epoch": 0.16866415297603016,
      "grad_norm": 0.6069415456506547,
      "learning_rate": 9.825683397209617e-06,
      "loss": 0.3916600704193115,
      "num_input_tokens_seen": 129378064,
      "step": 2505,
      "train_runtime": 15455.3407,
      "train_tokens_per_second": 8371.091
    },
    {
      "epoch": 0.1690008079719903,
      "grad_norm": 0.7198769215417986,
      "learning_rate": 9.824990643451915e-06,
      "loss": 0.3956932067871094,
      "num_input_tokens_seen": 129643992,
      "step": 2510,
      "train_runtime": 15483.7167,
      "train_tokens_per_second": 8372.925
    },
    {
      "epoch": 0.16933746296795044,
      "grad_norm": 0.6540614990917913,
      "learning_rate": 9.824296540400965e-06,
      "loss": 0.3926855087280273,
      "num_input_tokens_seen": 129924984,
      "step": 2515,
      "train_runtime": 15518.5643,
      "train_tokens_per_second": 8372.23
    },
    {
      "epoch": 0.16967411796391058,
      "grad_norm": 1.0709443932016105,
      "learning_rate": 9.823601088250872e-06,
      "loss": 0.4282047271728516,
      "num_input_tokens_seen": 130167840,
      "step": 2520,
      "train_runtime": 15549.0295,
      "train_tokens_per_second": 8371.445
    },
    {
      "epoch": 0.17001077295987072,
      "grad_norm": 0.737278525653745,
      "learning_rate": 9.822904287196114e-06,
      "loss": 0.4239660263061523,
      "num_input_tokens_seen": 130415352,
      "step": 2525,
      "train_runtime": 15574.1391,
      "train_tokens_per_second": 8373.84
    },
    {
      "epoch": 0.17034742795583085,
      "grad_norm": 0.7528106144605464,
      "learning_rate": 9.822206137431553e-06,
      "loss": 0.38682529926300047,
      "num_input_tokens_seen": 130670824,
      "step": 2530,
      "train_runtime": 15606.3373,
      "train_tokens_per_second": 8372.933
    },
    {
      "epoch": 0.170684082951791,
      "grad_norm": 0.7062825540218458,
      "learning_rate": 9.821506639152421e-06,
      "loss": 0.4119154930114746,
      "num_input_tokens_seen": 130914104,
      "step": 2535,
      "train_runtime": 15635.0416,
      "train_tokens_per_second": 8373.122
    },
    {
      "epoch": 0.17102073794775113,
      "grad_norm": 0.7461395535091879,
      "learning_rate": 9.820805792554334e-06,
      "loss": 0.38445315361022947,
      "num_input_tokens_seen": 131182624,
      "step": 2540,
      "train_runtime": 15662.6314,
      "train_tokens_per_second": 8375.516
    },
    {
      "epoch": 0.17135739294371127,
      "grad_norm": 0.8173784417747686,
      "learning_rate": 9.820103597833278e-06,
      "loss": 0.3918740272521973,
      "num_input_tokens_seen": 131446384,
      "step": 2545,
      "train_runtime": 15691.5671,
      "train_tokens_per_second": 8376.881
    },
    {
      "epoch": 0.1716940479396714,
      "grad_norm": 0.7731489787109445,
      "learning_rate": 9.819400055185622e-06,
      "loss": 0.39285638332366946,
      "num_input_tokens_seen": 131695536,
      "step": 2550,
      "train_runtime": 15725.8291,
      "train_tokens_per_second": 8374.473
    },
    {
      "epoch": 0.17203070293563155,
      "grad_norm": 0.8041337980451387,
      "learning_rate": 9.818695164808108e-06,
      "loss": 0.41050448417663576,
      "num_input_tokens_seen": 131931592,
      "step": 2555,
      "train_runtime": 15754.0928,
      "train_tokens_per_second": 8374.433
    },
    {
      "epoch": 0.1723673579315917,
      "grad_norm": 0.7358691200630271,
      "learning_rate": 9.817988926897856e-06,
      "loss": 0.38992316722869874,
      "num_input_tokens_seen": 132182696,
      "step": 2560,
      "train_runtime": 15787.7022,
      "train_tokens_per_second": 8372.51
    },
    {
      "epoch": 0.17270401292755186,
      "grad_norm": 0.7279643489800007,
      "learning_rate": 9.817281341652367e-06,
      "loss": 0.3985586166381836,
      "num_input_tokens_seen": 132445536,
      "step": 2565,
      "train_runtime": 15818.4562,
      "train_tokens_per_second": 8372.848
    },
    {
      "epoch": 0.173040667923512,
      "grad_norm": 1.37083685676661,
      "learning_rate": 9.816572409269508e-06,
      "loss": 0.37847585678100587,
      "num_input_tokens_seen": 132714608,
      "step": 2570,
      "train_runtime": 15849.365,
      "train_tokens_per_second": 8373.497
    },
    {
      "epoch": 0.17337732291947214,
      "grad_norm": 0.761489635169747,
      "learning_rate": 9.815862129947537e-06,
      "loss": 0.3996464252471924,
      "num_input_tokens_seen": 132979088,
      "step": 2575,
      "train_runtime": 15876.535,
      "train_tokens_per_second": 8375.826
    },
    {
      "epoch": 0.17371397791543228,
      "grad_norm": 0.7895936640662752,
      "learning_rate": 9.815150503885078e-06,
      "loss": 0.3617348909378052,
      "num_input_tokens_seen": 133249320,
      "step": 2580,
      "train_runtime": 15908.9755,
      "train_tokens_per_second": 8375.732
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 0.7876020358954116,
      "learning_rate": 9.814437531281133e-06,
      "loss": 0.39170303344726565,
      "num_input_tokens_seen": 133508480,
      "step": 2585,
      "train_runtime": 15939.1911,
      "train_tokens_per_second": 8376.114
    },
    {
      "epoch": 0.17438728790735256,
      "grad_norm": 0.7939756360500844,
      "learning_rate": 9.813723212335085e-06,
      "loss": 0.4480318546295166,
      "num_input_tokens_seen": 133750808,
      "step": 2590,
      "train_runtime": 15969.318,
      "train_tokens_per_second": 8375.487
    },
    {
      "epoch": 0.1747239429033127,
      "grad_norm": 0.7617261738142489,
      "learning_rate": 9.813007547246691e-06,
      "loss": 0.3778937816619873,
      "num_input_tokens_seen": 133999448,
      "step": 2595,
      "train_runtime": 15996.7806,
      "train_tokens_per_second": 8376.651
    },
    {
      "epoch": 0.17506059789927284,
      "grad_norm": 0.8140838247834798,
      "learning_rate": 9.812290536216083e-06,
      "loss": 0.44868316650390627,
      "num_input_tokens_seen": 134238904,
      "step": 2600,
      "train_runtime": 16030.103,
      "train_tokens_per_second": 8374.176
    },
    {
      "epoch": 0.17539725289523297,
      "grad_norm": 0.5998128774810307,
      "learning_rate": 9.811572179443773e-06,
      "loss": 0.37663042545318604,
      "num_input_tokens_seen": 134512952,
      "step": 2605,
      "train_runtime": 16061.5084,
      "train_tokens_per_second": 8374.864
    },
    {
      "epoch": 0.17573390789119311,
      "grad_norm": 0.772272803181737,
      "learning_rate": 9.810852477130645e-06,
      "loss": 0.41533937454223635,
      "num_input_tokens_seen": 134750840,
      "step": 2610,
      "train_runtime": 16093.2792,
      "train_tokens_per_second": 8373.113
    },
    {
      "epoch": 0.17607056288715325,
      "grad_norm": 0.67748331442532,
      "learning_rate": 9.810131429477964e-06,
      "loss": 0.37744667530059817,
      "num_input_tokens_seen": 135032152,
      "step": 2615,
      "train_runtime": 16122.4433,
      "train_tokens_per_second": 8375.415
    },
    {
      "epoch": 0.1764072178831134,
      "grad_norm": 0.7040449505529823,
      "learning_rate": 9.809409036687364e-06,
      "loss": 0.3992603302001953,
      "num_input_tokens_seen": 135288336,
      "step": 2620,
      "train_runtime": 16152.8487,
      "train_tokens_per_second": 8375.509
    },
    {
      "epoch": 0.17674387287907353,
      "grad_norm": 0.9052426347376175,
      "learning_rate": 9.808685298960867e-06,
      "loss": 0.40716190338134767,
      "num_input_tokens_seen": 135558168,
      "step": 2625,
      "train_runtime": 16183.181,
      "train_tokens_per_second": 8376.485
    },
    {
      "epoch": 0.17708052787503367,
      "grad_norm": 0.6833808404010542,
      "learning_rate": 9.807960216500858e-06,
      "loss": 0.4122157096862793,
      "num_input_tokens_seen": 135815520,
      "step": 2630,
      "train_runtime": 16216.3452,
      "train_tokens_per_second": 8375.224
    },
    {
      "epoch": 0.1774171828709938,
      "grad_norm": 0.6996741899455167,
      "learning_rate": 9.807233789510106e-06,
      "loss": 0.38341798782348635,
      "num_input_tokens_seen": 136054488,
      "step": 2635,
      "train_runtime": 16246.1155,
      "train_tokens_per_second": 8374.586
    },
    {
      "epoch": 0.17775383786695395,
      "grad_norm": 0.61209458765418,
      "learning_rate": 9.806506018191754e-06,
      "loss": 0.37890195846557617,
      "num_input_tokens_seen": 136320752,
      "step": 2640,
      "train_runtime": 16282.8464,
      "train_tokens_per_second": 8372.047
    },
    {
      "epoch": 0.1780904928629141,
      "grad_norm": 1.1970055253124194,
      "learning_rate": 9.80577690274932e-06,
      "loss": 0.3950566291809082,
      "num_input_tokens_seen": 136581760,
      "step": 2645,
      "train_runtime": 16312.2023,
      "train_tokens_per_second": 8372.981
    },
    {
      "epoch": 0.17842714785887423,
      "grad_norm": 0.6467291391371658,
      "learning_rate": 9.8050464433867e-06,
      "loss": 0.3785853385925293,
      "num_input_tokens_seen": 136840544,
      "step": 2650,
      "train_runtime": 16343.4724,
      "train_tokens_per_second": 8372.795
    },
    {
      "epoch": 0.17876380285483437,
      "grad_norm": 0.7798261076705045,
      "learning_rate": 9.804314640308165e-06,
      "loss": 0.4197660446166992,
      "num_input_tokens_seen": 137101880,
      "step": 2655,
      "train_runtime": 16376.7208,
      "train_tokens_per_second": 8371.754
    },
    {
      "epoch": 0.1791004578507945,
      "grad_norm": 0.8477138317574437,
      "learning_rate": 9.803581493718361e-06,
      "loss": 0.40551438331604006,
      "num_input_tokens_seen": 137358272,
      "step": 2660,
      "train_runtime": 16409.6233,
      "train_tokens_per_second": 8370.593
    },
    {
      "epoch": 0.17943711284675465,
      "grad_norm": 0.6964825532386814,
      "learning_rate": 9.80284700382231e-06,
      "loss": 0.4023623466491699,
      "num_input_tokens_seen": 137632176,
      "step": 2665,
      "train_runtime": 16443.0448,
      "train_tokens_per_second": 8370.237
    },
    {
      "epoch": 0.1797737678427148,
      "grad_norm": 0.6596248686024557,
      "learning_rate": 9.802111170825408e-06,
      "loss": 0.3968611717224121,
      "num_input_tokens_seen": 137900680,
      "step": 2670,
      "train_runtime": 16474.153,
      "train_tokens_per_second": 8370.73
    },
    {
      "epoch": 0.18011042283867493,
      "grad_norm": 0.6534217207809051,
      "learning_rate": 9.80137399493343e-06,
      "loss": 0.39821763038635255,
      "num_input_tokens_seen": 138154128,
      "step": 2675,
      "train_runtime": 16502.2913,
      "train_tokens_per_second": 8371.815
    },
    {
      "epoch": 0.18044707783463507,
      "grad_norm": 0.7286816113039558,
      "learning_rate": 9.800635476352525e-06,
      "loss": 0.3950523853302002,
      "num_input_tokens_seen": 138375248,
      "step": 2680,
      "train_runtime": 16530.0566,
      "train_tokens_per_second": 8371.13
    },
    {
      "epoch": 0.1807837328305952,
      "grad_norm": 0.767522434284386,
      "learning_rate": 9.799895615289218e-06,
      "loss": 0.3725940704345703,
      "num_input_tokens_seen": 138620352,
      "step": 2685,
      "train_runtime": 16563.8593,
      "train_tokens_per_second": 8368.844
    },
    {
      "epoch": 0.18112038782655535,
      "grad_norm": 0.7664084034889604,
      "learning_rate": 9.799154411950407e-06,
      "loss": 0.4332607269287109,
      "num_input_tokens_seen": 138908592,
      "step": 2690,
      "train_runtime": 16594.546,
      "train_tokens_per_second": 8370.738
    },
    {
      "epoch": 0.18145704282251549,
      "grad_norm": 0.7277320875671444,
      "learning_rate": 9.79841186654337e-06,
      "loss": 0.3874448537826538,
      "num_input_tokens_seen": 139161920,
      "step": 2695,
      "train_runtime": 16625.2741,
      "train_tokens_per_second": 8370.504
    },
    {
      "epoch": 0.18179369781847562,
      "grad_norm": 0.7476758499465007,
      "learning_rate": 9.797667979275752e-06,
      "loss": 0.42276611328125,
      "num_input_tokens_seen": 139410760,
      "step": 2700,
      "train_runtime": 16652.1386,
      "train_tokens_per_second": 8371.943
    },
    {
      "epoch": 0.18213035281443576,
      "grad_norm": 0.7252005045308942,
      "learning_rate": 9.796922750355584e-06,
      "loss": 0.37654871940612794,
      "num_input_tokens_seen": 139659912,
      "step": 2705,
      "train_runtime": 16686.2972,
      "train_tokens_per_second": 8369.737
    },
    {
      "epoch": 0.1824670078103959,
      "grad_norm": 0.8422270935228626,
      "learning_rate": 9.796176179991266e-06,
      "loss": 0.41723108291625977,
      "num_input_tokens_seen": 139915320,
      "step": 2710,
      "train_runtime": 16716.7058,
      "train_tokens_per_second": 8369.79
    },
    {
      "epoch": 0.18280366280635604,
      "grad_norm": 0.7233910411902493,
      "learning_rate": 9.795428268391572e-06,
      "loss": 0.39775815010070803,
      "num_input_tokens_seen": 140151824,
      "step": 2715,
      "train_runtime": 16748.1669,
      "train_tokens_per_second": 8368.189
    },
    {
      "epoch": 0.18314031780231618,
      "grad_norm": 0.8373792898776288,
      "learning_rate": 9.794679015765655e-06,
      "loss": 0.397812032699585,
      "num_input_tokens_seen": 140396792,
      "step": 2720,
      "train_runtime": 16779.4299,
      "train_tokens_per_second": 8367.197
    },
    {
      "epoch": 0.18347697279827632,
      "grad_norm": 0.745707884947975,
      "learning_rate": 9.793928422323041e-06,
      "loss": 0.4034909248352051,
      "num_input_tokens_seen": 140653016,
      "step": 2725,
      "train_runtime": 16811.505,
      "train_tokens_per_second": 8366.474
    },
    {
      "epoch": 0.18381362779423646,
      "grad_norm": 0.9589105962042331,
      "learning_rate": 9.793176488273629e-06,
      "loss": 0.38966946601867675,
      "num_input_tokens_seen": 140898760,
      "step": 2730,
      "train_runtime": 16841.1358,
      "train_tokens_per_second": 8366.345
    },
    {
      "epoch": 0.1841502827901966,
      "grad_norm": 0.7198763464328362,
      "learning_rate": 9.792423213827697e-06,
      "loss": 0.37418951988220217,
      "num_input_tokens_seen": 141150768,
      "step": 2735,
      "train_runtime": 16874.5044,
      "train_tokens_per_second": 8364.736
    },
    {
      "epoch": 0.18448693778615674,
      "grad_norm": 0.7212796282182203,
      "learning_rate": 9.791668599195897e-06,
      "loss": 0.37967514991760254,
      "num_input_tokens_seen": 141418432,
      "step": 2740,
      "train_runtime": 16910.2336,
      "train_tokens_per_second": 8362.891
    },
    {
      "epoch": 0.18482359278211688,
      "grad_norm": 0.7589171540893651,
      "learning_rate": 9.790912644589248e-06,
      "loss": 0.42047696113586425,
      "num_input_tokens_seen": 141656000,
      "step": 2745,
      "train_runtime": 16941.1427,
      "train_tokens_per_second": 8361.656
    },
    {
      "epoch": 0.18516024777807702,
      "grad_norm": 0.7164920908348585,
      "learning_rate": 9.790155350219159e-06,
      "loss": 0.3918336868286133,
      "num_input_tokens_seen": 141924296,
      "step": 2750,
      "train_runtime": 16972.6254,
      "train_tokens_per_second": 8361.953
    },
    {
      "epoch": 0.18549690277403716,
      "grad_norm": 0.6890387800904519,
      "learning_rate": 9.789396716297399e-06,
      "loss": 0.40384387969970703,
      "num_input_tokens_seen": 142207408,
      "step": 2755,
      "train_runtime": 17007.457,
      "train_tokens_per_second": 8361.474
    },
    {
      "epoch": 0.1858335577699973,
      "grad_norm": 0.7534096614963695,
      "learning_rate": 9.788636743036122e-06,
      "loss": 0.3845983028411865,
      "num_input_tokens_seen": 142450176,
      "step": 2760,
      "train_runtime": 17039.8887,
      "train_tokens_per_second": 8359.807
    },
    {
      "epoch": 0.18617021276595744,
      "grad_norm": 0.7365680898144167,
      "learning_rate": 9.787875430647847e-06,
      "loss": 0.40271692276000975,
      "num_input_tokens_seen": 142711336,
      "step": 2765,
      "train_runtime": 17069.3043,
      "train_tokens_per_second": 8360.7
    },
    {
      "epoch": 0.18650686776191758,
      "grad_norm": 0.8003469712256231,
      "learning_rate": 9.787112779345477e-06,
      "loss": 0.38366668224334716,
      "num_input_tokens_seen": 142941288,
      "step": 2770,
      "train_runtime": 17099.3736,
      "train_tokens_per_second": 8359.446
    },
    {
      "epoch": 0.18684352275787772,
      "grad_norm": 0.753059701451915,
      "learning_rate": 9.786348789342283e-06,
      "loss": 0.42626008987426756,
      "num_input_tokens_seen": 143206512,
      "step": 2775,
      "train_runtime": 17131.519,
      "train_tokens_per_second": 8359.242
    },
    {
      "epoch": 0.18718017775383786,
      "grad_norm": 0.7346356178563641,
      "learning_rate": 9.785583460851912e-06,
      "loss": 0.392515230178833,
      "num_input_tokens_seen": 143467552,
      "step": 2780,
      "train_runtime": 17163.9922,
      "train_tokens_per_second": 8358.635
    },
    {
      "epoch": 0.187516832749798,
      "grad_norm": 0.7209082496519486,
      "learning_rate": 9.784816794088386e-06,
      "loss": 0.4191441535949707,
      "num_input_tokens_seen": 143748088,
      "step": 2785,
      "train_runtime": 17197.5016,
      "train_tokens_per_second": 8358.661
    },
    {
      "epoch": 0.18785348774575814,
      "grad_norm": 0.7240885351183632,
      "learning_rate": 9.784048789266101e-06,
      "loss": 0.39263737201690674,
      "num_input_tokens_seen": 143985168,
      "step": 2790,
      "train_runtime": 17232.099,
      "train_tokens_per_second": 8355.637
    },
    {
      "epoch": 0.18819014274171827,
      "grad_norm": 0.6764920910567673,
      "learning_rate": 9.78327944659983e-06,
      "loss": 0.40295066833496096,
      "num_input_tokens_seen": 144228888,
      "step": 2795,
      "train_runtime": 17263.3254,
      "train_tokens_per_second": 8354.641
    },
    {
      "epoch": 0.18852679773767841,
      "grad_norm": 0.6776495818432625,
      "learning_rate": 9.78250876630471e-06,
      "loss": 0.39644339084625246,
      "num_input_tokens_seen": 144463472,
      "step": 2800,
      "train_runtime": 17294.6827,
      "train_tokens_per_second": 8353.057
    },
    {
      "epoch": 0.18886345273363855,
      "grad_norm": 0.8911259002803558,
      "learning_rate": 9.781736748596267e-06,
      "loss": 0.3934288740158081,
      "num_input_tokens_seen": 144713968,
      "step": 2805,
      "train_runtime": 17324.801,
      "train_tokens_per_second": 8352.995
    },
    {
      "epoch": 0.18920010772959872,
      "grad_norm": 0.6122027336733192,
      "learning_rate": 9.78096339369039e-06,
      "loss": 0.39001753330230715,
      "num_input_tokens_seen": 144962448,
      "step": 2810,
      "train_runtime": 17352.3861,
      "train_tokens_per_second": 8354.035
    },
    {
      "epoch": 0.18953676272555886,
      "grad_norm": 0.7146382331727181,
      "learning_rate": 9.780188701803345e-06,
      "loss": 0.42749762535095215,
      "num_input_tokens_seen": 145209952,
      "step": 2815,
      "train_runtime": 17382.8351,
      "train_tokens_per_second": 8353.64
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 0.6911594446794961,
      "learning_rate": 9.779412673151772e-06,
      "loss": 0.4079616069793701,
      "num_input_tokens_seen": 145487768,
      "step": 2820,
      "train_runtime": 17414.7767,
      "train_tokens_per_second": 8354.271
    },
    {
      "epoch": 0.19021007271747914,
      "grad_norm": 0.6741984501536816,
      "learning_rate": 9.778635307952686e-06,
      "loss": 0.40035390853881836,
      "num_input_tokens_seen": 145745272,
      "step": 2825,
      "train_runtime": 17443.4494,
      "train_tokens_per_second": 8355.301
    },
    {
      "epoch": 0.19054672771343928,
      "grad_norm": 0.7681893708781832,
      "learning_rate": 9.777856606423472e-06,
      "loss": 0.38430256843566896,
      "num_input_tokens_seen": 146018784,
      "step": 2830,
      "train_runtime": 17474.002,
      "train_tokens_per_second": 8356.345
    },
    {
      "epoch": 0.19088338270939942,
      "grad_norm": 0.7285452444976679,
      "learning_rate": 9.777076568781897e-06,
      "loss": 0.4286024570465088,
      "num_input_tokens_seen": 146244704,
      "step": 2835,
      "train_runtime": 17502.6792,
      "train_tokens_per_second": 8355.561
    },
    {
      "epoch": 0.19122003770535956,
      "grad_norm": 0.6430038367443645,
      "learning_rate": 9.77629519524609e-06,
      "loss": 0.3614053726196289,
      "num_input_tokens_seen": 146494032,
      "step": 2840,
      "train_runtime": 17534.4476,
      "train_tokens_per_second": 8354.642
    },
    {
      "epoch": 0.1915566927013197,
      "grad_norm": 0.6646763428030137,
      "learning_rate": 9.775512486034564e-06,
      "loss": 0.3895431995391846,
      "num_input_tokens_seen": 146738056,
      "step": 2845,
      "train_runtime": 17565.1402,
      "train_tokens_per_second": 8353.936
    },
    {
      "epoch": 0.19189334769727984,
      "grad_norm": 0.7183092801717325,
      "learning_rate": 9.774728441366197e-06,
      "loss": 0.4221662521362305,
      "num_input_tokens_seen": 146983760,
      "step": 2850,
      "train_runtime": 17595.4235,
      "train_tokens_per_second": 8353.522
    },
    {
      "epoch": 0.19223000269323998,
      "grad_norm": 0.7217521089749688,
      "learning_rate": 9.773943061460249e-06,
      "loss": 0.39041340351104736,
      "num_input_tokens_seen": 147228680,
      "step": 2855,
      "train_runtime": 17625.6723,
      "train_tokens_per_second": 8353.082
    },
    {
      "epoch": 0.19256665768920012,
      "grad_norm": 0.8973697757259017,
      "learning_rate": 9.773156346536345e-06,
      "loss": 0.4044965744018555,
      "num_input_tokens_seen": 147482624,
      "step": 2860,
      "train_runtime": 17659.4803,
      "train_tokens_per_second": 8351.47
    },
    {
      "epoch": 0.19290331268516026,
      "grad_norm": 0.6696468006850818,
      "learning_rate": 9.772368296814489e-06,
      "loss": 0.3839643239974976,
      "num_input_tokens_seen": 147743808,
      "step": 2865,
      "train_runtime": 17689.2816,
      "train_tokens_per_second": 8352.166
    },
    {
      "epoch": 0.1932399676811204,
      "grad_norm": 0.7087996337058212,
      "learning_rate": 9.771578912515058e-06,
      "loss": 0.4366742134094238,
      "num_input_tokens_seen": 148001784,
      "step": 2870,
      "train_runtime": 17721.2591,
      "train_tokens_per_second": 8351.652
    },
    {
      "epoch": 0.19357662267708053,
      "grad_norm": 0.6388324137611209,
      "learning_rate": 9.7707881938588e-06,
      "loss": 0.38060293197631834,
      "num_input_tokens_seen": 148262200,
      "step": 2875,
      "train_runtime": 17756.4763,
      "train_tokens_per_second": 8349.753
    },
    {
      "epoch": 0.19391327767304067,
      "grad_norm": 0.6949693408716128,
      "learning_rate": 9.769996141066836e-06,
      "loss": 0.42219114303588867,
      "num_input_tokens_seen": 148550328,
      "step": 2880,
      "train_runtime": 17790.1003,
      "train_tokens_per_second": 8350.168
    },
    {
      "epoch": 0.1942499326690008,
      "grad_norm": 0.6487053766708384,
      "learning_rate": 9.76920275436066e-06,
      "loss": 0.4113940715789795,
      "num_input_tokens_seen": 148854032,
      "step": 2885,
      "train_runtime": 17822.0289,
      "train_tokens_per_second": 8352.249
    },
    {
      "epoch": 0.19458658766496095,
      "grad_norm": 0.720715568997178,
      "learning_rate": 9.768408033962145e-06,
      "loss": 0.41673879623413085,
      "num_input_tokens_seen": 149111488,
      "step": 2890,
      "train_runtime": 17853.0501,
      "train_tokens_per_second": 8352.158
    },
    {
      "epoch": 0.1949232426609211,
      "grad_norm": 0.7028627332528502,
      "learning_rate": 9.767611980093525e-06,
      "loss": 0.4128150939941406,
      "num_input_tokens_seen": 149365088,
      "step": 2895,
      "train_runtime": 17882.5028,
      "train_tokens_per_second": 8352.583
    },
    {
      "epoch": 0.19525989765688123,
      "grad_norm": 0.6718839679674657,
      "learning_rate": 9.76681459297742e-06,
      "loss": 0.4071784973144531,
      "num_input_tokens_seen": 149617136,
      "step": 2900,
      "train_runtime": 17913.1869,
      "train_tokens_per_second": 8352.346
    },
    {
      "epoch": 0.19559655265284137,
      "grad_norm": 0.7592372601052363,
      "learning_rate": 9.766015872836814e-06,
      "loss": 0.4535517692565918,
      "num_input_tokens_seen": 149879720,
      "step": 2905,
      "train_runtime": 17946.0524,
      "train_tokens_per_second": 8351.682
    },
    {
      "epoch": 0.1959332076488015,
      "grad_norm": 0.8336950907673071,
      "learning_rate": 9.765215819895066e-06,
      "loss": 0.4052119255065918,
      "num_input_tokens_seen": 150175768,
      "step": 2910,
      "train_runtime": 17978.1408,
      "train_tokens_per_second": 8353.242
    },
    {
      "epoch": 0.19626986264476165,
      "grad_norm": 0.6534029579844971,
      "learning_rate": 9.76441443437591e-06,
      "loss": 0.36791534423828126,
      "num_input_tokens_seen": 150464000,
      "step": 2915,
      "train_runtime": 18012.1189,
      "train_tokens_per_second": 8353.487
    },
    {
      "epoch": 0.1966065176407218,
      "grad_norm": 0.6818508723524315,
      "learning_rate": 9.76361171650345e-06,
      "loss": 0.4099129676818848,
      "num_input_tokens_seen": 150709512,
      "step": 2920,
      "train_runtime": 18042.9779,
      "train_tokens_per_second": 8352.807
    },
    {
      "epoch": 0.19694317263668193,
      "grad_norm": 0.76055128931276,
      "learning_rate": 9.762807666502164e-06,
      "loss": 0.4047811985015869,
      "num_input_tokens_seen": 150954016,
      "step": 2925,
      "train_runtime": 18072.5653,
      "train_tokens_per_second": 8352.661
    },
    {
      "epoch": 0.19727982763264207,
      "grad_norm": 0.9379282410974104,
      "learning_rate": 9.762002284596898e-06,
      "loss": 0.4093360900878906,
      "num_input_tokens_seen": 151195528,
      "step": 2930,
      "train_runtime": 18102.257,
      "train_tokens_per_second": 8352.303
    },
    {
      "epoch": 0.1976164826286022,
      "grad_norm": 0.6849638137532126,
      "learning_rate": 9.761195571012882e-06,
      "loss": 0.41064019203186036,
      "num_input_tokens_seen": 151444384,
      "step": 2935,
      "train_runtime": 18131.4846,
      "train_tokens_per_second": 8352.564
    },
    {
      "epoch": 0.19795313762456235,
      "grad_norm": 0.6864859316191297,
      "learning_rate": 9.760387525975705e-06,
      "loss": 0.3879849433898926,
      "num_input_tokens_seen": 151720560,
      "step": 2940,
      "train_runtime": 18164.6143,
      "train_tokens_per_second": 8352.534
    },
    {
      "epoch": 0.1982897926205225,
      "grad_norm": 0.7088448483175217,
      "learning_rate": 9.759578149711337e-06,
      "loss": 0.3525996208190918,
      "num_input_tokens_seen": 151982504,
      "step": 2945,
      "train_runtime": 18198.7501,
      "train_tokens_per_second": 8351.261
    },
    {
      "epoch": 0.19862644761648263,
      "grad_norm": 0.6824353182589592,
      "learning_rate": 9.758767442446117e-06,
      "loss": 0.3809727907180786,
      "num_input_tokens_seen": 152239816,
      "step": 2950,
      "train_runtime": 18228.8438,
      "train_tokens_per_second": 8351.589
    },
    {
      "epoch": 0.19896310261244277,
      "grad_norm": 0.6268685981115638,
      "learning_rate": 9.757955404406755e-06,
      "loss": 0.36884849071502684,
      "num_input_tokens_seen": 152505712,
      "step": 2955,
      "train_runtime": 18259.5165,
      "train_tokens_per_second": 8352.122
    },
    {
      "epoch": 0.1992997576084029,
      "grad_norm": 0.779179692105006,
      "learning_rate": 9.757142035820337e-06,
      "loss": 0.38723068237304686,
      "num_input_tokens_seen": 152736368,
      "step": 2960,
      "train_runtime": 18289.9391,
      "train_tokens_per_second": 8350.841
    },
    {
      "epoch": 0.19963641260436304,
      "grad_norm": 0.8641768010544513,
      "learning_rate": 9.756327336914318e-06,
      "loss": 0.40354108810424805,
      "num_input_tokens_seen": 152972728,
      "step": 2965,
      "train_runtime": 18327.0446,
      "train_tokens_per_second": 8346.83
    },
    {
      "epoch": 0.19997306760032318,
      "grad_norm": 1.0976059772596602,
      "learning_rate": 9.755511307916526e-06,
      "loss": 0.39396238327026367,
      "num_input_tokens_seen": 153238760,
      "step": 2970,
      "train_runtime": 18358.5832,
      "train_tokens_per_second": 8346.982
    },
    {
      "epoch": 0.20030972259628332,
      "grad_norm": 0.7096576023494939,
      "learning_rate": 9.754693949055161e-06,
      "loss": 0.40532970428466797,
      "num_input_tokens_seen": 153506592,
      "step": 2975,
      "train_runtime": 18389.3995,
      "train_tokens_per_second": 8347.559
    },
    {
      "epoch": 0.20064637759224346,
      "grad_norm": 0.7031859231919584,
      "learning_rate": 9.753875260558795e-06,
      "loss": 0.4254766464233398,
      "num_input_tokens_seen": 153735192,
      "step": 2980,
      "train_runtime": 18416.7169,
      "train_tokens_per_second": 8347.589
    },
    {
      "epoch": 0.2009830325882036,
      "grad_norm": 0.7534410085378326,
      "learning_rate": 9.75305524265637e-06,
      "loss": 0.3901668548583984,
      "num_input_tokens_seen": 154010392,
      "step": 2985,
      "train_runtime": 18449.5292,
      "train_tokens_per_second": 8347.66
    },
    {
      "epoch": 0.20131968758416374,
      "grad_norm": 0.7763918443757308,
      "learning_rate": 9.752233895577205e-06,
      "loss": 0.40813560485839845,
      "num_input_tokens_seen": 154248848,
      "step": 2990,
      "train_runtime": 18481.2788,
      "train_tokens_per_second": 8346.222
    },
    {
      "epoch": 0.20165634258012388,
      "grad_norm": 0.6459529738818899,
      "learning_rate": 9.751411219550984e-06,
      "loss": 0.40120444297790525,
      "num_input_tokens_seen": 154501280,
      "step": 2995,
      "train_runtime": 18510.5371,
      "train_tokens_per_second": 8346.667
    },
    {
      "epoch": 0.20199299757608402,
      "grad_norm": 0.8576644101372248,
      "learning_rate": 9.750587214807766e-06,
      "loss": 0.4019810199737549,
      "num_input_tokens_seen": 154745296,
      "step": 3000,
      "train_runtime": 18540.765,
      "train_tokens_per_second": 8346.22
    },
    {
      "epoch": 0.20232965257204416,
      "grad_norm": 0.6723058149224194,
      "learning_rate": 9.749761881577981e-06,
      "loss": 0.4057606220245361,
      "num_input_tokens_seen": 155018688,
      "step": 3005,
      "train_runtime": 18570.594,
      "train_tokens_per_second": 8347.535
    },
    {
      "epoch": 0.2026663075680043,
      "grad_norm": 0.6530123193051957,
      "learning_rate": 9.748935220092434e-06,
      "loss": 0.38790552616119384,
      "num_input_tokens_seen": 155279200,
      "step": 3010,
      "train_runtime": 18595.1252,
      "train_tokens_per_second": 8350.533
    },
    {
      "epoch": 0.20300296256396444,
      "grad_norm": 0.7809463118682924,
      "learning_rate": 9.748107230582295e-06,
      "loss": 0.37615840435028075,
      "num_input_tokens_seen": 155517784,
      "step": 3015,
      "train_runtime": 18626.8811,
      "train_tokens_per_second": 8349.105
    },
    {
      "epoch": 0.20333961755992458,
      "grad_norm": 0.7460635478863334,
      "learning_rate": 9.747277913279109e-06,
      "loss": 0.45763769149780276,
      "num_input_tokens_seen": 155773600,
      "step": 3020,
      "train_runtime": 18654.3814,
      "train_tokens_per_second": 8350.51
    },
    {
      "epoch": 0.20367627255588472,
      "grad_norm": 0.5917670120339513,
      "learning_rate": 9.746447268414794e-06,
      "loss": 0.4240309715270996,
      "num_input_tokens_seen": 156046696,
      "step": 3025,
      "train_runtime": 18681.7838,
      "train_tokens_per_second": 8352.88
    },
    {
      "epoch": 0.20401292755184486,
      "grad_norm": 1.0604433745209954,
      "learning_rate": 9.745615296221635e-06,
      "loss": 0.40243353843688967,
      "num_input_tokens_seen": 156329272,
      "step": 3030,
      "train_runtime": 18712.0614,
      "train_tokens_per_second": 8354.466
    },
    {
      "epoch": 0.204349582547805,
      "grad_norm": 0.6290186577565335,
      "learning_rate": 9.74478199693229e-06,
      "loss": 0.4073160171508789,
      "num_input_tokens_seen": 156594280,
      "step": 3035,
      "train_runtime": 18739.9147,
      "train_tokens_per_second": 8356.19
    },
    {
      "epoch": 0.20468623754376514,
      "grad_norm": 0.7465556476274179,
      "learning_rate": 9.743947370779791e-06,
      "loss": 0.4155287742614746,
      "num_input_tokens_seen": 156864800,
      "step": 3040,
      "train_runtime": 18769.9587,
      "train_tokens_per_second": 8357.227
    },
    {
      "epoch": 0.20502289253972528,
      "grad_norm": 0.6628061655192444,
      "learning_rate": 9.743111417997536e-06,
      "loss": 0.4174698829650879,
      "num_input_tokens_seen": 157139592,
      "step": 3045,
      "train_runtime": 18800.3101,
      "train_tokens_per_second": 8358.351
    },
    {
      "epoch": 0.20535954753568542,
      "grad_norm": 0.7383125969323396,
      "learning_rate": 9.742274138819298e-06,
      "loss": 0.4106617450714111,
      "num_input_tokens_seen": 157404208,
      "step": 3050,
      "train_runtime": 18831.5839,
      "train_tokens_per_second": 8358.522
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.7348265803194098,
      "learning_rate": 9.74143553347922e-06,
      "loss": 0.45098257064819336,
      "num_input_tokens_seen": 157675584,
      "step": 3055,
      "train_runtime": 18863.6941,
      "train_tokens_per_second": 8358.68
    },
    {
      "epoch": 0.20603285752760572,
      "grad_norm": 0.5965433477514039,
      "learning_rate": 9.740595602211814e-06,
      "loss": 0.38972673416137693,
      "num_input_tokens_seen": 157939032,
      "step": 3060,
      "train_runtime": 18893.2431,
      "train_tokens_per_second": 8359.551
    },
    {
      "epoch": 0.20636951252356586,
      "grad_norm": 0.7196036989181128,
      "learning_rate": 9.739754345251963e-06,
      "loss": 0.3838178634643555,
      "num_input_tokens_seen": 158178432,
      "step": 3065,
      "train_runtime": 18920.1821,
      "train_tokens_per_second": 8360.302
    },
    {
      "epoch": 0.206706167519526,
      "grad_norm": 0.6568440203086982,
      "learning_rate": 9.738911762834924e-06,
      "loss": 0.34606289863586426,
      "num_input_tokens_seen": 158441136,
      "step": 3070,
      "train_runtime": 18948.06,
      "train_tokens_per_second": 8361.866
    },
    {
      "epoch": 0.20704282251548614,
      "grad_norm": 0.8082725502723218,
      "learning_rate": 9.738067855196323e-06,
      "loss": 0.3873929977416992,
      "num_input_tokens_seen": 158698352,
      "step": 3075,
      "train_runtime": 18980.4485,
      "train_tokens_per_second": 8361.149
    },
    {
      "epoch": 0.20737947751144628,
      "grad_norm": 0.8091161835830041,
      "learning_rate": 9.737222622572153e-06,
      "loss": 0.4343234062194824,
      "num_input_tokens_seen": 158938704,
      "step": 3080,
      "train_runtime": 19010.3007,
      "train_tokens_per_second": 8360.662
    },
    {
      "epoch": 0.20771613250740642,
      "grad_norm": 0.7308224447915826,
      "learning_rate": 9.736376065198784e-06,
      "loss": 0.3971827507019043,
      "num_input_tokens_seen": 159210472,
      "step": 3085,
      "train_runtime": 19042.191,
      "train_tokens_per_second": 8360.932
    },
    {
      "epoch": 0.20805278750336656,
      "grad_norm": 0.6143033842402703,
      "learning_rate": 9.735528183312951e-06,
      "loss": 0.3893140316009521,
      "num_input_tokens_seen": 159482856,
      "step": 3090,
      "train_runtime": 19075.2082,
      "train_tokens_per_second": 8360.74
    },
    {
      "epoch": 0.2083894424993267,
      "grad_norm": 0.7048445707572839,
      "learning_rate": 9.734678977151762e-06,
      "loss": 0.38056011199951173,
      "num_input_tokens_seen": 159728080,
      "step": 3095,
      "train_runtime": 19104.5513,
      "train_tokens_per_second": 8360.734
    },
    {
      "epoch": 0.20872609749528684,
      "grad_norm": 0.7323233064640992,
      "learning_rate": 9.733828446952696e-06,
      "loss": 0.39778549671173097,
      "num_input_tokens_seen": 159989800,
      "step": 3100,
      "train_runtime": 19135.8939,
      "train_tokens_per_second": 8360.717
    },
    {
      "epoch": 0.20906275249124698,
      "grad_norm": 0.6640344030850117,
      "learning_rate": 9.732976592953598e-06,
      "loss": 0.37990655899047854,
      "num_input_tokens_seen": 160259776,
      "step": 3105,
      "train_runtime": 19167.6811,
      "train_tokens_per_second": 8360.937
    },
    {
      "epoch": 0.20939940748720712,
      "grad_norm": 0.7076904263114212,
      "learning_rate": 9.73212341539269e-06,
      "loss": 0.3949297904968262,
      "num_input_tokens_seen": 160505080,
      "step": 3110,
      "train_runtime": 19198.2349,
      "train_tokens_per_second": 8360.408
    },
    {
      "epoch": 0.20973606248316726,
      "grad_norm": 0.6912488838043216,
      "learning_rate": 9.731268914508556e-06,
      "loss": 0.3976278781890869,
      "num_input_tokens_seen": 160754216,
      "step": 3115,
      "train_runtime": 19228.7695,
      "train_tokens_per_second": 8360.089
    },
    {
      "epoch": 0.2100727174791274,
      "grad_norm": 0.6938652016683334,
      "learning_rate": 9.730413090540158e-06,
      "loss": 0.3723747730255127,
      "num_input_tokens_seen": 161018032,
      "step": 3120,
      "train_runtime": 19262.268,
      "train_tokens_per_second": 8359.246
    },
    {
      "epoch": 0.21040937247508754,
      "grad_norm": 0.6326147960309044,
      "learning_rate": 9.729555943726825e-06,
      "loss": 0.40009498596191406,
      "num_input_tokens_seen": 161258512,
      "step": 3125,
      "train_runtime": 19293.3169,
      "train_tokens_per_second": 8358.258
    },
    {
      "epoch": 0.21074602747104768,
      "grad_norm": 0.858184329669963,
      "learning_rate": 9.728697474308253e-06,
      "loss": 0.41200008392333987,
      "num_input_tokens_seen": 161507776,
      "step": 3130,
      "train_runtime": 19326.2799,
      "train_tokens_per_second": 8356.899
    },
    {
      "epoch": 0.21108268246700782,
      "grad_norm": 0.7151848640513205,
      "learning_rate": 9.72783768252451e-06,
      "loss": 0.385913610458374,
      "num_input_tokens_seen": 161744088,
      "step": 3135,
      "train_runtime": 19357.2052,
      "train_tokens_per_second": 8355.756
    },
    {
      "epoch": 0.21141933746296795,
      "grad_norm": 0.730670110817753,
      "learning_rate": 9.726976568616036e-06,
      "loss": 0.3971771478652954,
      "num_input_tokens_seen": 161985352,
      "step": 3140,
      "train_runtime": 19385.6475,
      "train_tokens_per_second": 8355.942
    },
    {
      "epoch": 0.2117559924589281,
      "grad_norm": 0.8607714590227785,
      "learning_rate": 9.726114132823638e-06,
      "loss": 0.3927351474761963,
      "num_input_tokens_seen": 162253288,
      "step": 3145,
      "train_runtime": 19411.3201,
      "train_tokens_per_second": 8358.694
    },
    {
      "epoch": 0.21209264745488823,
      "grad_norm": 0.7025055694038168,
      "learning_rate": 9.725250375388493e-06,
      "loss": 0.42121095657348634,
      "num_input_tokens_seen": 162501144,
      "step": 3150,
      "train_runtime": 19441.448,
      "train_tokens_per_second": 8358.49
    },
    {
      "epoch": 0.21242930245084837,
      "grad_norm": 0.839312433290983,
      "learning_rate": 9.724385296552147e-06,
      "loss": 0.431109619140625,
      "num_input_tokens_seen": 162763504,
      "step": 3155,
      "train_runtime": 19471.59,
      "train_tokens_per_second": 8359.025
    },
    {
      "epoch": 0.2127659574468085,
      "grad_norm": 0.7759391881198237,
      "learning_rate": 9.72351889655652e-06,
      "loss": 0.39732041358947756,
      "num_input_tokens_seen": 163034760,
      "step": 3160,
      "train_runtime": 19505.2786,
      "train_tokens_per_second": 8358.494
    },
    {
      "epoch": 0.21310261244276865,
      "grad_norm": 0.7309310657162112,
      "learning_rate": 9.722651175643895e-06,
      "loss": 0.3964901685714722,
      "num_input_tokens_seen": 163292368,
      "step": 3165,
      "train_runtime": 19534.7076,
      "train_tokens_per_second": 8359.089
    },
    {
      "epoch": 0.2134392674387288,
      "grad_norm": 0.7843657052755575,
      "learning_rate": 9.721782134056928e-06,
      "loss": 0.3539254665374756,
      "num_input_tokens_seen": 163541248,
      "step": 3170,
      "train_runtime": 19564.2252,
      "train_tokens_per_second": 8359.199
    },
    {
      "epoch": 0.21377592243468893,
      "grad_norm": 0.6567322718508465,
      "learning_rate": 9.720911772038645e-06,
      "loss": 0.40668630599975586,
      "num_input_tokens_seen": 163812464,
      "step": 3175,
      "train_runtime": 19598.8746,
      "train_tokens_per_second": 8358.259
    },
    {
      "epoch": 0.21411257743064907,
      "grad_norm": 0.7956717083996423,
      "learning_rate": 9.720040089832436e-06,
      "loss": 0.405721378326416,
      "num_input_tokens_seen": 164031696,
      "step": 3180,
      "train_runtime": 19630.1381,
      "train_tokens_per_second": 8356.115
    },
    {
      "epoch": 0.2144492324266092,
      "grad_norm": 0.7591217391088979,
      "learning_rate": 9.719167087682069e-06,
      "loss": 0.3783890247344971,
      "num_input_tokens_seen": 164290672,
      "step": 3185,
      "train_runtime": 19660.9481,
      "train_tokens_per_second": 8356.193
    },
    {
      "epoch": 0.21478588742256935,
      "grad_norm": 0.7748232038188828,
      "learning_rate": 9.718292765831673e-06,
      "loss": 0.38986365795135497,
      "num_input_tokens_seen": 164556344,
      "step": 3190,
      "train_runtime": 19694.7464,
      "train_tokens_per_second": 8355.342
    },
    {
      "epoch": 0.2151225424185295,
      "grad_norm": 0.6982767857393934,
      "learning_rate": 9.71741712452575e-06,
      "loss": 0.40111527442932127,
      "num_input_tokens_seen": 164825224,
      "step": 3195,
      "train_runtime": 19726.0332,
      "train_tokens_per_second": 8355.721
    },
    {
      "epoch": 0.21545919741448963,
      "grad_norm": 0.7384045182939197,
      "learning_rate": 9.716540164009174e-06,
      "loss": 0.3712003707885742,
      "num_input_tokens_seen": 165064048,
      "step": 3200,
      "train_runtime": 19753.4137,
      "train_tokens_per_second": 8356.229
    },
    {
      "epoch": 0.21579585241044977,
      "grad_norm": 0.7364554589645362,
      "learning_rate": 9.715661884527178e-06,
      "loss": 0.3975597143173218,
      "num_input_tokens_seen": 165342536,
      "step": 3205,
      "train_runtime": 19781.5812,
      "train_tokens_per_second": 8358.408
    },
    {
      "epoch": 0.2161325074064099,
      "grad_norm": 1.0976607962192748,
      "learning_rate": 9.714782286325374e-06,
      "loss": 0.3973705768585205,
      "num_input_tokens_seen": 165585960,
      "step": 3210,
      "train_runtime": 19812.1177,
      "train_tokens_per_second": 8357.812
    },
    {
      "epoch": 0.21646916240237005,
      "grad_norm": 0.7268089150591103,
      "learning_rate": 9.713901369649738e-06,
      "loss": 0.3864430904388428,
      "num_input_tokens_seen": 165839016,
      "step": 3215,
      "train_runtime": 19841.6505,
      "train_tokens_per_second": 8358.126
    },
    {
      "epoch": 0.2168058173983302,
      "grad_norm": 0.6987649116063281,
      "learning_rate": 9.713019134746618e-06,
      "loss": 0.37000174522399903,
      "num_input_tokens_seen": 166126616,
      "step": 3220,
      "train_runtime": 19878.625,
      "train_tokens_per_second": 8357.048
    },
    {
      "epoch": 0.21714247239429033,
      "grad_norm": 0.6854652087560159,
      "learning_rate": 9.712135581862722e-06,
      "loss": 0.3915052652359009,
      "num_input_tokens_seen": 166400616,
      "step": 3225,
      "train_runtime": 19910.7771,
      "train_tokens_per_second": 8357.314
    },
    {
      "epoch": 0.21747912739025047,
      "grad_norm": 0.741579298088583,
      "learning_rate": 9.711250711245139e-06,
      "loss": 0.40279717445373536,
      "num_input_tokens_seen": 166643544,
      "step": 3230,
      "train_runtime": 19942.7044,
      "train_tokens_per_second": 8356.116
    },
    {
      "epoch": 0.2178157823862106,
      "grad_norm": 0.7140918544939816,
      "learning_rate": 9.710364523141317e-06,
      "loss": 0.4020575523376465,
      "num_input_tokens_seen": 166903080,
      "step": 3235,
      "train_runtime": 19974.0793,
      "train_tokens_per_second": 8355.984
    },
    {
      "epoch": 0.21815243738217074,
      "grad_norm": 0.6723979607912556,
      "learning_rate": 9.709477017799076e-06,
      "loss": 0.411285924911499,
      "num_input_tokens_seen": 167178344,
      "step": 3240,
      "train_runtime": 20003.124,
      "train_tokens_per_second": 8357.612
    },
    {
      "epoch": 0.21848909237813088,
      "grad_norm": 0.6878840440307943,
      "learning_rate": 9.708588195466604e-06,
      "loss": 0.39572815895080565,
      "num_input_tokens_seen": 167450336,
      "step": 3245,
      "train_runtime": 20034.3122,
      "train_tokens_per_second": 8358.177
    },
    {
      "epoch": 0.21882574737409102,
      "grad_norm": 0.6715193922938915,
      "learning_rate": 9.707698056392458e-06,
      "loss": 0.396125340461731,
      "num_input_tokens_seen": 167715352,
      "step": 3250,
      "train_runtime": 20064.8517,
      "train_tokens_per_second": 8358.664
    },
    {
      "epoch": 0.21916240237005116,
      "grad_norm": 0.8683477873194472,
      "learning_rate": 9.706806600825562e-06,
      "loss": 0.4095271110534668,
      "num_input_tokens_seen": 167965288,
      "step": 3255,
      "train_runtime": 20090.4766,
      "train_tokens_per_second": 8360.443
    },
    {
      "epoch": 0.2194990573660113,
      "grad_norm": 0.8975521803908426,
      "learning_rate": 9.70591382901521e-06,
      "loss": 0.4214353561401367,
      "num_input_tokens_seen": 168221104,
      "step": 3260,
      "train_runtime": 20126.8649,
      "train_tokens_per_second": 8358.038
    },
    {
      "epoch": 0.21983571236197144,
      "grad_norm": 0.6170411450103044,
      "learning_rate": 9.705019741211061e-06,
      "loss": 0.3842911243438721,
      "num_input_tokens_seen": 168478896,
      "step": 3265,
      "train_runtime": 20154.417,
      "train_tokens_per_second": 8359.403
    },
    {
      "epoch": 0.22017236735793158,
      "grad_norm": 0.8022358317310007,
      "learning_rate": 9.704124337663145e-06,
      "loss": 0.3800793170928955,
      "num_input_tokens_seen": 168760648,
      "step": 3270,
      "train_runtime": 20181.5651,
      "train_tokens_per_second": 8362.119
    },
    {
      "epoch": 0.22050902235389172,
      "grad_norm": 0.8790433282685974,
      "learning_rate": 9.703227618621859e-06,
      "loss": 0.3934182643890381,
      "num_input_tokens_seen": 169032128,
      "step": 3275,
      "train_runtime": 20211.6462,
      "train_tokens_per_second": 8363.105
    },
    {
      "epoch": 0.22084567734985186,
      "grad_norm": 0.8388987289984772,
      "learning_rate": 9.702329584337965e-06,
      "loss": 0.4118658065795898,
      "num_input_tokens_seen": 169251176,
      "step": 3280,
      "train_runtime": 20239.7614,
      "train_tokens_per_second": 8362.311
    },
    {
      "epoch": 0.221182332345812,
      "grad_norm": 1.0179688119425085,
      "learning_rate": 9.701430235062599e-06,
      "loss": 0.432068920135498,
      "num_input_tokens_seen": 169502296,
      "step": 3285,
      "train_runtime": 20270.7947,
      "train_tokens_per_second": 8361.897
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 0.7505087661484053,
      "learning_rate": 9.700529571047259e-06,
      "loss": 0.41720094680786135,
      "num_input_tokens_seen": 169753272,
      "step": 3290,
      "train_runtime": 20302.3031,
      "train_tokens_per_second": 8361.282
    },
    {
      "epoch": 0.22185564233773228,
      "grad_norm": 0.8702295214492828,
      "learning_rate": 9.699627592543814e-06,
      "loss": 0.3856488227844238,
      "num_input_tokens_seen": 169995344,
      "step": 3295,
      "train_runtime": 20331.0939,
      "train_tokens_per_second": 8361.348
    },
    {
      "epoch": 0.22219229733369245,
      "grad_norm": 0.6338554347495682,
      "learning_rate": 9.698724299804498e-06,
      "loss": 0.39733426570892333,
      "num_input_tokens_seen": 170259632,
      "step": 3300,
      "train_runtime": 20365.4126,
      "train_tokens_per_second": 8360.235
    },
    {
      "epoch": 0.22252895232965259,
      "grad_norm": 0.702810928760001,
      "learning_rate": 9.697819693081915e-06,
      "loss": 0.3875314235687256,
      "num_input_tokens_seen": 170515824,
      "step": 3305,
      "train_runtime": 20394.3176,
      "train_tokens_per_second": 8360.948
    },
    {
      "epoch": 0.22286560732561272,
      "grad_norm": 0.658991217240677,
      "learning_rate": 9.696913772629034e-06,
      "loss": 0.38004002571105955,
      "num_input_tokens_seen": 170780736,
      "step": 3310,
      "train_runtime": 20427.8105,
      "train_tokens_per_second": 8360.208
    },
    {
      "epoch": 0.22320226232157286,
      "grad_norm": 0.6995958948914607,
      "learning_rate": 9.696006538699195e-06,
      "loss": 0.41832480430603025,
      "num_input_tokens_seen": 171052384,
      "step": 3315,
      "train_runtime": 20458.3924,
      "train_tokens_per_second": 8360.989
    },
    {
      "epoch": 0.223538917317533,
      "grad_norm": 0.7924819211741762,
      "learning_rate": 9.695097991546102e-06,
      "loss": 0.403350830078125,
      "num_input_tokens_seen": 171274112,
      "step": 3320,
      "train_runtime": 20487.5443,
      "train_tokens_per_second": 8359.914
    },
    {
      "epoch": 0.22387557231349314,
      "grad_norm": 0.7977808302300994,
      "learning_rate": 9.694188131423828e-06,
      "loss": 0.37870349884033205,
      "num_input_tokens_seen": 171525104,
      "step": 3325,
      "train_runtime": 20516.4921,
      "train_tokens_per_second": 8360.352
    },
    {
      "epoch": 0.22421222730945328,
      "grad_norm": 0.7561619376182757,
      "learning_rate": 9.693276958586809e-06,
      "loss": 0.43142070770263674,
      "num_input_tokens_seen": 171783672,
      "step": 3330,
      "train_runtime": 20550.1904,
      "train_tokens_per_second": 8359.225
    },
    {
      "epoch": 0.22454888230541342,
      "grad_norm": 0.7458082782581279,
      "learning_rate": 9.692364473289856e-06,
      "loss": 0.37647936344146726,
      "num_input_tokens_seen": 172035480,
      "step": 3335,
      "train_runtime": 20578.9793,
      "train_tokens_per_second": 8359.767
    },
    {
      "epoch": 0.22488553730137356,
      "grad_norm": 0.6951240708875135,
      "learning_rate": 9.69145067578814e-06,
      "loss": 0.3792394161224365,
      "num_input_tokens_seen": 172304080,
      "step": 3340,
      "train_runtime": 20609.042,
      "train_tokens_per_second": 8360.606
    },
    {
      "epoch": 0.2252221922973337,
      "grad_norm": 0.7209613223828127,
      "learning_rate": 9.690535566337205e-06,
      "loss": 0.4340526580810547,
      "num_input_tokens_seen": 172569512,
      "step": 3345,
      "train_runtime": 20638.0228,
      "train_tokens_per_second": 8361.727
    },
    {
      "epoch": 0.22555884729329384,
      "grad_norm": 0.7529264808294087,
      "learning_rate": 9.689619145192955e-06,
      "loss": 0.4183077812194824,
      "num_input_tokens_seen": 172818736,
      "step": 3350,
      "train_runtime": 20669.4466,
      "train_tokens_per_second": 8361.072
    },
    {
      "epoch": 0.22589550228925398,
      "grad_norm": 0.6718755517720847,
      "learning_rate": 9.688701412611663e-06,
      "loss": 0.38783390522003175,
      "num_input_tokens_seen": 173086040,
      "step": 3355,
      "train_runtime": 20705.6375,
      "train_tokens_per_second": 8359.368
    },
    {
      "epoch": 0.22623215728521412,
      "grad_norm": 0.7822324024479517,
      "learning_rate": 9.687782368849973e-06,
      "loss": 0.4020689010620117,
      "num_input_tokens_seen": 173340536,
      "step": 3360,
      "train_runtime": 20736.6614,
      "train_tokens_per_second": 8359.134
    },
    {
      "epoch": 0.22656881228117426,
      "grad_norm": 0.7108713869941112,
      "learning_rate": 9.686862014164893e-06,
      "loss": 0.37965824604034426,
      "num_input_tokens_seen": 173616464,
      "step": 3365,
      "train_runtime": 20764.3957,
      "train_tokens_per_second": 8361.258
    },
    {
      "epoch": 0.2269054672771344,
      "grad_norm": 0.7795029860075315,
      "learning_rate": 9.685940348813795e-06,
      "loss": 0.3761666536331177,
      "num_input_tokens_seen": 173870048,
      "step": 3370,
      "train_runtime": 20792.5327,
      "train_tokens_per_second": 8362.139
    },
    {
      "epoch": 0.22724212227309454,
      "grad_norm": 0.8041771071219079,
      "learning_rate": 9.685017373054421e-06,
      "loss": 0.39457848072052004,
      "num_input_tokens_seen": 174128368,
      "step": 3375,
      "train_runtime": 20821.6173,
      "train_tokens_per_second": 8362.865
    },
    {
      "epoch": 0.22757877726905468,
      "grad_norm": 0.7339178969479689,
      "learning_rate": 9.68409308714488e-06,
      "loss": 0.38925650119781496,
      "num_input_tokens_seen": 174370712,
      "step": 3380,
      "train_runtime": 20854.3634,
      "train_tokens_per_second": 8361.354
    },
    {
      "epoch": 0.22791543226501482,
      "grad_norm": 0.7179239908780196,
      "learning_rate": 9.68316749134364e-06,
      "loss": 0.39716358184814454,
      "num_input_tokens_seen": 174645672,
      "step": 3385,
      "train_runtime": 20886.3115,
      "train_tokens_per_second": 8361.729
    },
    {
      "epoch": 0.22825208726097496,
      "grad_norm": 0.7551933851542518,
      "learning_rate": 9.682240585909545e-06,
      "loss": 0.37615461349487306,
      "num_input_tokens_seen": 174890664,
      "step": 3390,
      "train_runtime": 20913.7442,
      "train_tokens_per_second": 8362.475
    },
    {
      "epoch": 0.2285887422569351,
      "grad_norm": 0.6346424089480469,
      "learning_rate": 9.681312371101803e-06,
      "loss": 0.36069483757019044,
      "num_input_tokens_seen": 175182392,
      "step": 3395,
      "train_runtime": 20946.238,
      "train_tokens_per_second": 8363.43
    },
    {
      "epoch": 0.22892539725289524,
      "grad_norm": 0.6846074648871665,
      "learning_rate": 9.680382847179984e-06,
      "loss": 0.41106176376342773,
      "num_input_tokens_seen": 175420832,
      "step": 3400,
      "train_runtime": 20979.1628,
      "train_tokens_per_second": 8361.67
    },
    {
      "epoch": 0.22926205224885537,
      "grad_norm": 0.8059482707453711,
      "learning_rate": 9.679452014404025e-06,
      "loss": 0.3877140522003174,
      "num_input_tokens_seen": 175660472,
      "step": 3405,
      "train_runtime": 21005.9061,
      "train_tokens_per_second": 8362.433
    },
    {
      "epoch": 0.22959870724481551,
      "grad_norm": 0.7671740601460406,
      "learning_rate": 9.678519873034234e-06,
      "loss": 0.42437567710876467,
      "num_input_tokens_seen": 175929608,
      "step": 3410,
      "train_runtime": 21040.9076,
      "train_tokens_per_second": 8361.313
    },
    {
      "epoch": 0.22993536224077565,
      "grad_norm": 0.788248490096606,
      "learning_rate": 9.677586423331277e-06,
      "loss": 0.41995997428894044,
      "num_input_tokens_seen": 176193152,
      "step": 3415,
      "train_runtime": 21073.9061,
      "train_tokens_per_second": 8360.726
    },
    {
      "epoch": 0.2302720172367358,
      "grad_norm": 0.5988390542351302,
      "learning_rate": 9.676651665556194e-06,
      "loss": 0.3823801040649414,
      "num_input_tokens_seen": 176465856,
      "step": 3420,
      "train_runtime": 21106.1392,
      "train_tokens_per_second": 8360.878
    },
    {
      "epoch": 0.23060867223269593,
      "grad_norm": 0.779009460670599,
      "learning_rate": 9.675715599970383e-06,
      "loss": 0.3943739652633667,
      "num_input_tokens_seen": 176700544,
      "step": 3425,
      "train_runtime": 21134.8333,
      "train_tokens_per_second": 8360.631
    },
    {
      "epoch": 0.23094532722865607,
      "grad_norm": 0.7328961051108558,
      "learning_rate": 9.674778226835617e-06,
      "loss": 0.4165501117706299,
      "num_input_tokens_seen": 176941760,
      "step": 3430,
      "train_runtime": 21166.4357,
      "train_tokens_per_second": 8359.544
    },
    {
      "epoch": 0.2312819822246162,
      "grad_norm": 0.6820136831646443,
      "learning_rate": 9.673839546414026e-06,
      "loss": 0.4000999450683594,
      "num_input_tokens_seen": 177193984,
      "step": 3435,
      "train_runtime": 21194.974,
      "train_tokens_per_second": 8360.189
    },
    {
      "epoch": 0.23161863722057635,
      "grad_norm": 0.6552676167939503,
      "learning_rate": 9.672899558968107e-06,
      "loss": 0.37542541027069093,
      "num_input_tokens_seen": 177429920,
      "step": 3440,
      "train_runtime": 21229.1617,
      "train_tokens_per_second": 8357.839
    },
    {
      "epoch": 0.2319552922165365,
      "grad_norm": 0.7096995962517556,
      "learning_rate": 9.67195826476073e-06,
      "loss": 0.4421196937561035,
      "num_input_tokens_seen": 177701568,
      "step": 3445,
      "train_runtime": 21265.2333,
      "train_tokens_per_second": 8356.436
    },
    {
      "epoch": 0.23229194721249663,
      "grad_norm": 0.8014577677798737,
      "learning_rate": 9.671015664055122e-06,
      "loss": 0.38599681854248047,
      "num_input_tokens_seen": 177939392,
      "step": 3450,
      "train_runtime": 21296.2252,
      "train_tokens_per_second": 8355.443
    },
    {
      "epoch": 0.23262860220845677,
      "grad_norm": 0.7369785386939176,
      "learning_rate": 9.670071757114877e-06,
      "loss": 0.4032459259033203,
      "num_input_tokens_seen": 178195712,
      "step": 3455,
      "train_runtime": 21331.601,
      "train_tokens_per_second": 8353.602
    },
    {
      "epoch": 0.2329652572044169,
      "grad_norm": 0.6461331375144627,
      "learning_rate": 9.669126544203957e-06,
      "loss": 0.3743563652038574,
      "num_input_tokens_seen": 178442896,
      "step": 3460,
      "train_runtime": 21363.6561,
      "train_tokens_per_second": 8352.638
    },
    {
      "epoch": 0.23330191220037705,
      "grad_norm": 0.7187172607678786,
      "learning_rate": 9.668180025586689e-06,
      "loss": 0.4013843536376953,
      "num_input_tokens_seen": 178726512,
      "step": 3465,
      "train_runtime": 21397.719,
      "train_tokens_per_second": 8352.596
    },
    {
      "epoch": 0.2336385671963372,
      "grad_norm": 0.7792730054358077,
      "learning_rate": 9.667232201527761e-06,
      "loss": 0.38550705909729005,
      "num_input_tokens_seen": 178957040,
      "step": 3470,
      "train_runtime": 21426.5725,
      "train_tokens_per_second": 8352.108
    },
    {
      "epoch": 0.23397522219229733,
      "grad_norm": 0.7932358011474754,
      "learning_rate": 9.666283072292234e-06,
      "loss": 0.393849515914917,
      "num_input_tokens_seen": 179187296,
      "step": 3475,
      "train_runtime": 21456.6858,
      "train_tokens_per_second": 8351.117
    },
    {
      "epoch": 0.23431187718825747,
      "grad_norm": 0.6543177876206827,
      "learning_rate": 9.665332638145522e-06,
      "loss": 0.39416956901550293,
      "num_input_tokens_seen": 179448264,
      "step": 3480,
      "train_runtime": 21488.1326,
      "train_tokens_per_second": 8351.04
    },
    {
      "epoch": 0.2346485321842176,
      "grad_norm": 0.737561389857538,
      "learning_rate": 9.664380899353414e-06,
      "loss": 0.4324801445007324,
      "num_input_tokens_seen": 179712536,
      "step": 3485,
      "train_runtime": 21519.5829,
      "train_tokens_per_second": 8351.116
    },
    {
      "epoch": 0.23498518718017775,
      "grad_norm": 0.8100522932660853,
      "learning_rate": 9.663427856182062e-06,
      "loss": 0.4058741569519043,
      "num_input_tokens_seen": 179991584,
      "step": 3490,
      "train_runtime": 21550.26,
      "train_tokens_per_second": 8352.177
    },
    {
      "epoch": 0.23532184217613789,
      "grad_norm": 0.6873800579452303,
      "learning_rate": 9.662473508897981e-06,
      "loss": 0.4004487991333008,
      "num_input_tokens_seen": 180248152,
      "step": 3495,
      "train_runtime": 21578.8686,
      "train_tokens_per_second": 8352.994
    },
    {
      "epoch": 0.23565849717209802,
      "grad_norm": 0.8291718170269404,
      "learning_rate": 9.661517857768052e-06,
      "loss": 0.43254861831665037,
      "num_input_tokens_seen": 180494520,
      "step": 3500,
      "train_runtime": 21610.3821,
      "train_tokens_per_second": 8352.213
    },
    {
      "epoch": 0.23599515216805816,
      "grad_norm": 0.7905535525695502,
      "learning_rate": 9.660560903059517e-06,
      "loss": 0.39968295097351075,
      "num_input_tokens_seen": 180729616,
      "step": 3505,
      "train_runtime": 21639.6933,
      "train_tokens_per_second": 8351.764
    },
    {
      "epoch": 0.2363318071640183,
      "grad_norm": 0.6278379944695095,
      "learning_rate": 9.659602645039988e-06,
      "loss": 0.3991802215576172,
      "num_input_tokens_seen": 180996584,
      "step": 3510,
      "train_runtime": 21669.4792,
      "train_tokens_per_second": 8352.604
    },
    {
      "epoch": 0.23666846215997844,
      "grad_norm": 0.7382809251900815,
      "learning_rate": 9.658643083977435e-06,
      "loss": 0.36096744537353515,
      "num_input_tokens_seen": 181252640,
      "step": 3515,
      "train_runtime": 21697.3658,
      "train_tokens_per_second": 8353.67
    },
    {
      "epoch": 0.23700511715593858,
      "grad_norm": 0.6891594265848534,
      "learning_rate": 9.657682220140202e-06,
      "loss": 0.3820670127868652,
      "num_input_tokens_seen": 181478048,
      "step": 3520,
      "train_runtime": 21726.3497,
      "train_tokens_per_second": 8352.901
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 0.7607126654599659,
      "learning_rate": 9.656720053796987e-06,
      "loss": 0.4131017684936523,
      "num_input_tokens_seen": 181727232,
      "step": 3525,
      "train_runtime": 21754.9657,
      "train_tokens_per_second": 8353.368
    },
    {
      "epoch": 0.23767842714785886,
      "grad_norm": 0.9296121081573177,
      "learning_rate": 9.655756585216859e-06,
      "loss": 0.39019145965576174,
      "num_input_tokens_seen": 181967888,
      "step": 3530,
      "train_runtime": 21783.954,
      "train_tokens_per_second": 8353.299
    },
    {
      "epoch": 0.238015082143819,
      "grad_norm": 0.8666470261941652,
      "learning_rate": 9.654791814669246e-06,
      "loss": 0.40968170166015627,
      "num_input_tokens_seen": 182206464,
      "step": 3535,
      "train_runtime": 21814.6685,
      "train_tokens_per_second": 8352.475
    },
    {
      "epoch": 0.23835173713977914,
      "grad_norm": 0.697704851869172,
      "learning_rate": 9.653825742423949e-06,
      "loss": 0.3615856647491455,
      "num_input_tokens_seen": 182475816,
      "step": 3540,
      "train_runtime": 21841.9449,
      "train_tokens_per_second": 8354.376
    },
    {
      "epoch": 0.2386883921357393,
      "grad_norm": 0.9785079969259252,
      "learning_rate": 9.652858368751118e-06,
      "loss": 0.3855022430419922,
      "num_input_tokens_seen": 182719408,
      "step": 3545,
      "train_runtime": 21870.2311,
      "train_tokens_per_second": 8354.709
    },
    {
      "epoch": 0.23902504713169945,
      "grad_norm": 0.7334973940362367,
      "learning_rate": 9.651889693921284e-06,
      "loss": 0.39143996238708495,
      "num_input_tokens_seen": 182998112,
      "step": 3550,
      "train_runtime": 21902.58,
      "train_tokens_per_second": 8355.094
    },
    {
      "epoch": 0.2393617021276596,
      "grad_norm": 0.714407413005922,
      "learning_rate": 9.650919718205331e-06,
      "loss": 0.4570009231567383,
      "num_input_tokens_seen": 183246736,
      "step": 3555,
      "train_runtime": 21931.5415,
      "train_tokens_per_second": 8355.397
    },
    {
      "epoch": 0.23969835712361973,
      "grad_norm": 0.6954132353896942,
      "learning_rate": 9.649948441874509e-06,
      "loss": 0.4050957679748535,
      "num_input_tokens_seen": 183507952,
      "step": 3560,
      "train_runtime": 21958.5527,
      "train_tokens_per_second": 8357.015
    },
    {
      "epoch": 0.24003501211957987,
      "grad_norm": 0.7487573906272144,
      "learning_rate": 9.648975865200434e-06,
      "loss": 0.3559511661529541,
      "num_input_tokens_seen": 183741504,
      "step": 3565,
      "train_runtime": 21986.9527,
      "train_tokens_per_second": 8356.843
    },
    {
      "epoch": 0.24037166711554,
      "grad_norm": 0.6895788109757781,
      "learning_rate": 9.648001988455082e-06,
      "loss": 0.41594576835632324,
      "num_input_tokens_seen": 184005232,
      "step": 3570,
      "train_runtime": 22016.02,
      "train_tokens_per_second": 8357.788
    },
    {
      "epoch": 0.24070832211150014,
      "grad_norm": 0.7672309505986075,
      "learning_rate": 9.647026811910794e-06,
      "loss": 0.4151009559631348,
      "num_input_tokens_seen": 184253816,
      "step": 3575,
      "train_runtime": 22050.8617,
      "train_tokens_per_second": 8355.856
    },
    {
      "epoch": 0.24104497710746028,
      "grad_norm": 2.4724102977226567,
      "learning_rate": 9.646050335840279e-06,
      "loss": 0.43484859466552733,
      "num_input_tokens_seen": 184517728,
      "step": 3580,
      "train_runtime": 22079.0059,
      "train_tokens_per_second": 8357.157
    },
    {
      "epoch": 0.24138163210342042,
      "grad_norm": 0.6779145990618378,
      "learning_rate": 9.6450725605166e-06,
      "loss": 0.3880629062652588,
      "num_input_tokens_seen": 184768368,
      "step": 3585,
      "train_runtime": 22109.4154,
      "train_tokens_per_second": 8356.999
    },
    {
      "epoch": 0.24171828709938056,
      "grad_norm": 0.7366745602185495,
      "learning_rate": 9.644093486213192e-06,
      "loss": 0.4208224296569824,
      "num_input_tokens_seen": 185046608,
      "step": 3590,
      "train_runtime": 22141.6167,
      "train_tokens_per_second": 8357.412
    },
    {
      "epoch": 0.2420549420953407,
      "grad_norm": 0.7618292324598931,
      "learning_rate": 9.64311311320385e-06,
      "loss": 0.409834098815918,
      "num_input_tokens_seen": 185321248,
      "step": 3595,
      "train_runtime": 22177.5888,
      "train_tokens_per_second": 8356.24
    },
    {
      "epoch": 0.24239159709130084,
      "grad_norm": 0.7051981971524487,
      "learning_rate": 9.642131441762732e-06,
      "loss": 0.3856398582458496,
      "num_input_tokens_seen": 185572792,
      "step": 3600,
      "train_runtime": 22205.4398,
      "train_tokens_per_second": 8357.087
    },
    {
      "epoch": 0.24272825208726098,
      "grad_norm": 0.710814621820656,
      "learning_rate": 9.641148472164359e-06,
      "loss": 0.36703431606292725,
      "num_input_tokens_seen": 185826336,
      "step": 3605,
      "train_runtime": 22238.1321,
      "train_tokens_per_second": 8356.203
    },
    {
      "epoch": 0.24306490708322112,
      "grad_norm": 0.7330648276364242,
      "learning_rate": 9.640164204683616e-06,
      "loss": 0.3862058162689209,
      "num_input_tokens_seen": 186095056,
      "step": 3610,
      "train_runtime": 22271.9968,
      "train_tokens_per_second": 8355.562
    },
    {
      "epoch": 0.24340156207918126,
      "grad_norm": 0.7576168227538247,
      "learning_rate": 9.639178639595748e-06,
      "loss": 0.4220901966094971,
      "num_input_tokens_seen": 186364320,
      "step": 3615,
      "train_runtime": 22302.8031,
      "train_tokens_per_second": 8356.094
    },
    {
      "epoch": 0.2437382170751414,
      "grad_norm": 0.7839310869472589,
      "learning_rate": 9.638191777176367e-06,
      "loss": 0.39240612983703616,
      "num_input_tokens_seen": 186636944,
      "step": 3620,
      "train_runtime": 22335.8015,
      "train_tokens_per_second": 8355.955
    },
    {
      "epoch": 0.24407487207110154,
      "grad_norm": 0.6637731499806915,
      "learning_rate": 9.637203617701446e-06,
      "loss": 0.42187299728393557,
      "num_input_tokens_seen": 186921680,
      "step": 3625,
      "train_runtime": 22366.4232,
      "train_tokens_per_second": 8357.245
    },
    {
      "epoch": 0.24441152706706168,
      "grad_norm": 0.7247803176102977,
      "learning_rate": 9.63621416144732e-06,
      "loss": 0.37567548751831054,
      "num_input_tokens_seen": 187174720,
      "step": 3630,
      "train_runtime": 22399.4702,
      "train_tokens_per_second": 8356.212
    },
    {
      "epoch": 0.24474818206302182,
      "grad_norm": 0.9086280533938993,
      "learning_rate": 9.635223408690688e-06,
      "loss": 0.38011651039123534,
      "num_input_tokens_seen": 187400144,
      "step": 3635,
      "train_runtime": 22424.2551,
      "train_tokens_per_second": 8357.029
    },
    {
      "epoch": 0.24508483705898196,
      "grad_norm": 0.6451506249329655,
      "learning_rate": 9.63423135970861e-06,
      "loss": 0.3992666006088257,
      "num_input_tokens_seen": 187669376,
      "step": 3640,
      "train_runtime": 22456.5917,
      "train_tokens_per_second": 8356.984
    },
    {
      "epoch": 0.2454214920549421,
      "grad_norm": 0.7773364053620899,
      "learning_rate": 9.633238014778511e-06,
      "loss": 0.40976648330688475,
      "num_input_tokens_seen": 187950704,
      "step": 3645,
      "train_runtime": 22489.16,
      "train_tokens_per_second": 8357.391
    },
    {
      "epoch": 0.24575814705090224,
      "grad_norm": 0.7392791281613345,
      "learning_rate": 9.632243374178173e-06,
      "loss": 0.38730487823486326,
      "num_input_tokens_seen": 188214176,
      "step": 3650,
      "train_runtime": 22520.5306,
      "train_tokens_per_second": 8357.449
    },
    {
      "epoch": 0.24609480204686238,
      "grad_norm": 0.7587894172533268,
      "learning_rate": 9.63124743818575e-06,
      "loss": 0.4167830467224121,
      "num_input_tokens_seen": 188480576,
      "step": 3655,
      "train_runtime": 22552.6271,
      "train_tokens_per_second": 8357.367
    },
    {
      "epoch": 0.24643145704282252,
      "grad_norm": 0.6230013062491372,
      "learning_rate": 9.630250207079745e-06,
      "loss": 0.3775656223297119,
      "num_input_tokens_seen": 188736920,
      "step": 3660,
      "train_runtime": 22582.6513,
      "train_tokens_per_second": 8357.607
    },
    {
      "epoch": 0.24676811203878266,
      "grad_norm": 0.6890951281385129,
      "learning_rate": 9.629251681139038e-06,
      "loss": 0.40346660614013674,
      "num_input_tokens_seen": 189001208,
      "step": 3665,
      "train_runtime": 22616.8085,
      "train_tokens_per_second": 8356.67
    },
    {
      "epoch": 0.2471047670347428,
      "grad_norm": 0.6426713410781273,
      "learning_rate": 9.628251860642857e-06,
      "loss": 0.3999300956726074,
      "num_input_tokens_seen": 189237816,
      "step": 3670,
      "train_runtime": 22646.9416,
      "train_tokens_per_second": 8355.999
    },
    {
      "epoch": 0.24744142203070293,
      "grad_norm": 0.6733221765228146,
      "learning_rate": 9.627250745870805e-06,
      "loss": 0.3979135751724243,
      "num_input_tokens_seen": 189485496,
      "step": 3675,
      "train_runtime": 22678.6921,
      "train_tokens_per_second": 8355.222
    },
    {
      "epoch": 0.24777807702666307,
      "grad_norm": 0.7382700112656383,
      "learning_rate": 9.626248337102834e-06,
      "loss": 0.44342265129089353,
      "num_input_tokens_seen": 189752832,
      "step": 3680,
      "train_runtime": 22709.6284,
      "train_tokens_per_second": 8355.611
    },
    {
      "epoch": 0.2481147320226232,
      "grad_norm": 0.7920343273441371,
      "learning_rate": 9.625244634619269e-06,
      "loss": 0.40932269096374513,
      "num_input_tokens_seen": 190010984,
      "step": 3685,
      "train_runtime": 22740.7514,
      "train_tokens_per_second": 8355.528
    },
    {
      "epoch": 0.24845138701858335,
      "grad_norm": 0.791556176746072,
      "learning_rate": 9.62423963870079e-06,
      "loss": 0.40236454010009765,
      "num_input_tokens_seen": 190249656,
      "step": 3690,
      "train_runtime": 22771.5181,
      "train_tokens_per_second": 8354.72
    },
    {
      "epoch": 0.2487880420145435,
      "grad_norm": 0.6313102978045809,
      "learning_rate": 9.623233349628441e-06,
      "loss": 0.3652036190032959,
      "num_input_tokens_seen": 190496768,
      "step": 3695,
      "train_runtime": 22800.3563,
      "train_tokens_per_second": 8354.991
    },
    {
      "epoch": 0.24912469701050363,
      "grad_norm": 0.6428512962641932,
      "learning_rate": 9.62222576768363e-06,
      "loss": 0.3714164972305298,
      "num_input_tokens_seen": 190753136,
      "step": 3700,
      "train_runtime": 22833.5877,
      "train_tokens_per_second": 8354.059
    },
    {
      "epoch": 0.24946135200646377,
      "grad_norm": 0.7383497914088234,
      "learning_rate": 9.62121689314812e-06,
      "loss": 0.4416037082672119,
      "num_input_tokens_seen": 191001376,
      "step": 3705,
      "train_runtime": 22862.1385,
      "train_tokens_per_second": 8354.484
    },
    {
      "epoch": 0.2497980070024239,
      "grad_norm": 0.744466826188234,
      "learning_rate": 9.620206726304045e-06,
      "loss": 0.41219215393066405,
      "num_input_tokens_seen": 191239608,
      "step": 3710,
      "train_runtime": 22892.5173,
      "train_tokens_per_second": 8353.804
    },
    {
      "epoch": 0.2501346619983841,
      "grad_norm": 0.7133326946859817,
      "learning_rate": 9.61919526743389e-06,
      "loss": 0.39624106884002686,
      "num_input_tokens_seen": 191508912,
      "step": 3715,
      "train_runtime": 22927.624,
      "train_tokens_per_second": 8352.759
    },
    {
      "epoch": 0.2504713169943442,
      "grad_norm": 0.6653116700581178,
      "learning_rate": 9.618182516820507e-06,
      "loss": 0.4067848205566406,
      "num_input_tokens_seen": 191787896,
      "step": 3720,
      "train_runtime": 22961.0905,
      "train_tokens_per_second": 8352.735
    },
    {
      "epoch": 0.25080797199030436,
      "grad_norm": 0.6988046082924974,
      "learning_rate": 9.61716847474711e-06,
      "loss": 0.37706267833709717,
      "num_input_tokens_seen": 192035360,
      "step": 3725,
      "train_runtime": 22994.9072,
      "train_tokens_per_second": 8351.213
    },
    {
      "epoch": 0.2511446269862645,
      "grad_norm": 0.748869878337807,
      "learning_rate": 9.616153141497273e-06,
      "loss": 0.4019660472869873,
      "num_input_tokens_seen": 192281376,
      "step": 3730,
      "train_runtime": 23025.3561,
      "train_tokens_per_second": 8350.854
    },
    {
      "epoch": 0.25148128198222464,
      "grad_norm": 0.7704696482574708,
      "learning_rate": 9.61513651735493e-06,
      "loss": 0.3697951793670654,
      "num_input_tokens_seen": 192527256,
      "step": 3735,
      "train_runtime": 23056.0858,
      "train_tokens_per_second": 8350.388
    },
    {
      "epoch": 0.2518179369781848,
      "grad_norm": 0.9214425701041231,
      "learning_rate": 9.614118602604376e-06,
      "loss": 0.39662995338439944,
      "num_input_tokens_seen": 192744336,
      "step": 3740,
      "train_runtime": 23084.5366,
      "train_tokens_per_second": 8349.5
    },
    {
      "epoch": 0.2521545919741449,
      "grad_norm": 0.76214898650489,
      "learning_rate": 9.613099397530267e-06,
      "loss": 0.362180233001709,
      "num_input_tokens_seen": 193012368,
      "step": 3745,
      "train_runtime": 23113.9888,
      "train_tokens_per_second": 8350.457
    },
    {
      "epoch": 0.25249124697010505,
      "grad_norm": 0.7505224401346884,
      "learning_rate": 9.612078902417623e-06,
      "loss": 0.39076662063598633,
      "num_input_tokens_seen": 193296648,
      "step": 3750,
      "train_runtime": 23141.5898,
      "train_tokens_per_second": 8352.782
    },
    {
      "epoch": 0.2528279019660652,
      "grad_norm": 0.834265648755087,
      "learning_rate": 9.611057117551821e-06,
      "loss": 0.40118160247802737,
      "num_input_tokens_seen": 193516936,
      "step": 3755,
      "train_runtime": 23172.2545,
      "train_tokens_per_second": 8351.235
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.7031514286853218,
      "learning_rate": 9.610034043218598e-06,
      "loss": 0.40969038009643555,
      "num_input_tokens_seen": 193770240,
      "step": 3760,
      "train_runtime": 23200.9624,
      "train_tokens_per_second": 8351.819
    },
    {
      "epoch": 0.2535012119579855,
      "grad_norm": 0.8453507793082906,
      "learning_rate": 9.609009679704057e-06,
      "loss": 0.39991116523742676,
      "num_input_tokens_seen": 194030280,
      "step": 3765,
      "train_runtime": 23236.3985,
      "train_tokens_per_second": 8350.273
    },
    {
      "epoch": 0.2538378669539456,
      "grad_norm": 0.7770682456316003,
      "learning_rate": 9.607984027294654e-06,
      "loss": 0.38907608985900877,
      "num_input_tokens_seen": 194266744,
      "step": 3770,
      "train_runtime": 23263.6629,
      "train_tokens_per_second": 8350.652
    },
    {
      "epoch": 0.25417452194990575,
      "grad_norm": 0.6890091112172937,
      "learning_rate": 9.606957086277213e-06,
      "loss": 0.39150118827819824,
      "num_input_tokens_seen": 194528184,
      "step": 3775,
      "train_runtime": 23292.1928,
      "train_tokens_per_second": 8351.647
    },
    {
      "epoch": 0.2545111769458659,
      "grad_norm": 0.7933239572854167,
      "learning_rate": 9.605928856938913e-06,
      "loss": 0.40138912200927734,
      "num_input_tokens_seen": 194777904,
      "step": 3780,
      "train_runtime": 23324.5208,
      "train_tokens_per_second": 8350.778
    },
    {
      "epoch": 0.25484783194182603,
      "grad_norm": 0.7240878715457888,
      "learning_rate": 9.604899339567298e-06,
      "loss": 0.3626518249511719,
      "num_input_tokens_seen": 195024592,
      "step": 3785,
      "train_runtime": 23358.6093,
      "train_tokens_per_second": 8349.153
    },
    {
      "epoch": 0.25518448693778617,
      "grad_norm": 0.7133451819608898,
      "learning_rate": 9.603868534450263e-06,
      "loss": 0.3432164192199707,
      "num_input_tokens_seen": 195310504,
      "step": 3790,
      "train_runtime": 23390.2726,
      "train_tokens_per_second": 8350.074
    },
    {
      "epoch": 0.2555211419337463,
      "grad_norm": 0.7589543128250481,
      "learning_rate": 9.602836441876074e-06,
      "loss": 0.34273595809936525,
      "num_input_tokens_seen": 195576472,
      "step": 3795,
      "train_runtime": 23420.4829,
      "train_tokens_per_second": 8350.659
    },
    {
      "epoch": 0.25585779692970645,
      "grad_norm": 0.9602442493532105,
      "learning_rate": 9.601803062133353e-06,
      "loss": 0.38918561935424806,
      "num_input_tokens_seen": 195829032,
      "step": 3800,
      "train_runtime": 23453.2385,
      "train_tokens_per_second": 8349.765
    },
    {
      "epoch": 0.2561944519256666,
      "grad_norm": 0.6503294145591316,
      "learning_rate": 9.600768395511078e-06,
      "loss": 0.35957984924316405,
      "num_input_tokens_seen": 196080976,
      "step": 3805,
      "train_runtime": 23486.6584,
      "train_tokens_per_second": 8348.611
    },
    {
      "epoch": 0.25653110692162673,
      "grad_norm": 0.6617385302485213,
      "learning_rate": 9.599732442298592e-06,
      "loss": 0.3880226850509644,
      "num_input_tokens_seen": 196352920,
      "step": 3810,
      "train_runtime": 23520.1078,
      "train_tokens_per_second": 8348.3
    },
    {
      "epoch": 0.25686776191758687,
      "grad_norm": 0.7006354190463834,
      "learning_rate": 9.598695202785596e-06,
      "loss": 0.39406642913818357,
      "num_input_tokens_seen": 196588664,
      "step": 3815,
      "train_runtime": 23547.5987,
      "train_tokens_per_second": 8348.565
    },
    {
      "epoch": 0.257204416913547,
      "grad_norm": 0.6542664868657896,
      "learning_rate": 9.59765667726215e-06,
      "loss": 0.38561620712280276,
      "num_input_tokens_seen": 196854816,
      "step": 3820,
      "train_runtime": 23580.0879,
      "train_tokens_per_second": 8348.35
    },
    {
      "epoch": 0.25754107190950715,
      "grad_norm": 0.7497040005033747,
      "learning_rate": 9.596616866018677e-06,
      "loss": 0.4339291572570801,
      "num_input_tokens_seen": 197105288,
      "step": 3825,
      "train_runtime": 23610.6197,
      "train_tokens_per_second": 8348.162
    },
    {
      "epoch": 0.2578777269054673,
      "grad_norm": 0.9335001659661932,
      "learning_rate": 9.595575769345955e-06,
      "loss": 0.38537116050720216,
      "num_input_tokens_seen": 197326752,
      "step": 3830,
      "train_runtime": 23638.9769,
      "train_tokens_per_second": 8347.517
    },
    {
      "epoch": 0.2582143819014274,
      "grad_norm": 0.7842937980302592,
      "learning_rate": 9.594533387535122e-06,
      "loss": 0.3952141046524048,
      "num_input_tokens_seen": 197595464,
      "step": 3835,
      "train_runtime": 23670.8956,
      "train_tokens_per_second": 8347.612
    },
    {
      "epoch": 0.25855103689738757,
      "grad_norm": 0.6586447367288829,
      "learning_rate": 9.593489720877678e-06,
      "loss": 0.3993344783782959,
      "num_input_tokens_seen": 197848848,
      "step": 3840,
      "train_runtime": 23700.5873,
      "train_tokens_per_second": 8347.846
    },
    {
      "epoch": 0.2588876918933477,
      "grad_norm": 0.6890006785712562,
      "learning_rate": 9.592444769665482e-06,
      "loss": 0.37561450004577634,
      "num_input_tokens_seen": 198119720,
      "step": 3845,
      "train_runtime": 23732.4189,
      "train_tokens_per_second": 8348.063
    },
    {
      "epoch": 0.25922434688930784,
      "grad_norm": 0.8229789485173136,
      "learning_rate": 9.59139853419075e-06,
      "loss": 0.36376736164093015,
      "num_input_tokens_seen": 198358864,
      "step": 3850,
      "train_runtime": 23763.8714,
      "train_tokens_per_second": 8347.077
    },
    {
      "epoch": 0.259561001885268,
      "grad_norm": 0.6908227639681479,
      "learning_rate": 9.590351014746059e-06,
      "loss": 0.3969999313354492,
      "num_input_tokens_seen": 198625496,
      "step": 3855,
      "train_runtime": 23796.0933,
      "train_tokens_per_second": 8346.979
    },
    {
      "epoch": 0.2598976568812281,
      "grad_norm": 0.6898945220475121,
      "learning_rate": 9.589302211624343e-06,
      "loss": 0.38819007873535155,
      "num_input_tokens_seen": 198874528,
      "step": 3860,
      "train_runtime": 23826.942,
      "train_tokens_per_second": 8346.624
    },
    {
      "epoch": 0.26023431187718826,
      "grad_norm": 0.7036598775140925,
      "learning_rate": 9.5882521251189e-06,
      "loss": 0.386980938911438,
      "num_input_tokens_seen": 199133768,
      "step": 3865,
      "train_runtime": 23855.9108,
      "train_tokens_per_second": 8347.355
    },
    {
      "epoch": 0.2605709668731484,
      "grad_norm": 0.7263057955453165,
      "learning_rate": 9.58720075552338e-06,
      "loss": 0.4062182903289795,
      "num_input_tokens_seen": 199382200,
      "step": 3870,
      "train_runtime": 23885.3588,
      "train_tokens_per_second": 8347.465
    },
    {
      "epoch": 0.26090762186910854,
      "grad_norm": 0.769345165560629,
      "learning_rate": 9.586148103131797e-06,
      "loss": 0.3964177846908569,
      "num_input_tokens_seen": 199628104,
      "step": 3875,
      "train_runtime": 23916.024,
      "train_tokens_per_second": 8347.044
    },
    {
      "epoch": 0.2612442768650687,
      "grad_norm": 0.8132489393307266,
      "learning_rate": 9.585094168238518e-06,
      "loss": 0.4205606460571289,
      "num_input_tokens_seen": 199869456,
      "step": 3880,
      "train_runtime": 23949.2709,
      "train_tokens_per_second": 8345.534
    },
    {
      "epoch": 0.2615809318610288,
      "grad_norm": 0.6795135733424765,
      "learning_rate": 9.58403895113828e-06,
      "loss": 0.3953943967819214,
      "num_input_tokens_seen": 200133840,
      "step": 3885,
      "train_runtime": 23981.6508,
      "train_tokens_per_second": 8345.29
    },
    {
      "epoch": 0.26191758685698896,
      "grad_norm": 0.7302022189134585,
      "learning_rate": 9.582982452126164e-06,
      "loss": 0.378597354888916,
      "num_input_tokens_seen": 200381384,
      "step": 3890,
      "train_runtime": 24014.7438,
      "train_tokens_per_second": 8344.098
    },
    {
      "epoch": 0.2622542418529491,
      "grad_norm": 0.649944822648016,
      "learning_rate": 9.58192467149762e-06,
      "loss": 0.36894831657409666,
      "num_input_tokens_seen": 200628040,
      "step": 3895,
      "train_runtime": 24043.5411,
      "train_tokens_per_second": 8344.363
    },
    {
      "epoch": 0.26259089684890924,
      "grad_norm": 0.8021694603047451,
      "learning_rate": 9.580865609548452e-06,
      "loss": 0.4031081199645996,
      "num_input_tokens_seen": 200863640,
      "step": 3900,
      "train_runtime": 24075.8561,
      "train_tokens_per_second": 8342.949
    },
    {
      "epoch": 0.2629275518448694,
      "grad_norm": 0.6532547539772996,
      "learning_rate": 9.579805266574824e-06,
      "loss": 0.37460665702819823,
      "num_input_tokens_seen": 201107616,
      "step": 3905,
      "train_runtime": 24107.0866,
      "train_tokens_per_second": 8342.261
    },
    {
      "epoch": 0.2632642068408295,
      "grad_norm": 0.5471900755243656,
      "learning_rate": 9.578743642873258e-06,
      "loss": 0.3846429824829102,
      "num_input_tokens_seen": 201342400,
      "step": 3910,
      "train_runtime": 24134.786,
      "train_tokens_per_second": 8342.415
    },
    {
      "epoch": 0.26360086183678966,
      "grad_norm": 0.6596835329184816,
      "learning_rate": 9.57768073874063e-06,
      "loss": 0.39458982944488524,
      "num_input_tokens_seen": 201607960,
      "step": 3915,
      "train_runtime": 24167.838,
      "train_tokens_per_second": 8341.994
    },
    {
      "epoch": 0.2639375168327498,
      "grad_norm": 0.7306338089576386,
      "learning_rate": 9.576616554474182e-06,
      "loss": 0.40792217254638674,
      "num_input_tokens_seen": 201845064,
      "step": 3920,
      "train_runtime": 24194.2161,
      "train_tokens_per_second": 8342.699
    },
    {
      "epoch": 0.26427417182870994,
      "grad_norm": 0.6759350162928663,
      "learning_rate": 9.575551090371512e-06,
      "loss": 0.39172210693359377,
      "num_input_tokens_seen": 202078760,
      "step": 3925,
      "train_runtime": 24224.9508,
      "train_tokens_per_second": 8341.761
    },
    {
      "epoch": 0.2646108268246701,
      "grad_norm": 0.7459149429897974,
      "learning_rate": 9.574484346730564e-06,
      "loss": 0.40038251876831055,
      "num_input_tokens_seen": 202336240,
      "step": 3930,
      "train_runtime": 24257.2794,
      "train_tokens_per_second": 8341.259
    },
    {
      "epoch": 0.2649474818206302,
      "grad_norm": 0.7499421445828943,
      "learning_rate": 9.573416323849658e-06,
      "loss": 0.40518741607666015,
      "num_input_tokens_seen": 202578456,
      "step": 3935,
      "train_runtime": 24285.4283,
      "train_tokens_per_second": 8341.564
    },
    {
      "epoch": 0.26528413681659035,
      "grad_norm": 0.7999660927513103,
      "learning_rate": 9.572347022027461e-06,
      "loss": 0.3859898567199707,
      "num_input_tokens_seen": 202842416,
      "step": 3940,
      "train_runtime": 24314.627,
      "train_tokens_per_second": 8342.403
    },
    {
      "epoch": 0.2656207918125505,
      "grad_norm": 0.6340554547352982,
      "learning_rate": 9.571276441563e-06,
      "loss": 0.3864926338195801,
      "num_input_tokens_seen": 203080968,
      "step": 3945,
      "train_runtime": 24341.4113,
      "train_tokens_per_second": 8343.024
    },
    {
      "epoch": 0.26595744680851063,
      "grad_norm": 0.7224328236351402,
      "learning_rate": 9.570204582755659e-06,
      "loss": 0.37026286125183105,
      "num_input_tokens_seen": 203345472,
      "step": 3950,
      "train_runtime": 24369.5002,
      "train_tokens_per_second": 8344.261
    },
    {
      "epoch": 0.2662941018044708,
      "grad_norm": 0.8112942297867278,
      "learning_rate": 9.569131445905178e-06,
      "loss": 0.39295547008514403,
      "num_input_tokens_seen": 203600728,
      "step": 3955,
      "train_runtime": 24404.7966,
      "train_tokens_per_second": 8342.652
    },
    {
      "epoch": 0.2666307568004309,
      "grad_norm": 6.301216651638736,
      "learning_rate": 9.568057031311659e-06,
      "loss": 0.4065236568450928,
      "num_input_tokens_seen": 203850344,
      "step": 3960,
      "train_runtime": 24435.7512,
      "train_tokens_per_second": 8342.299
    },
    {
      "epoch": 0.26696741179639105,
      "grad_norm": 0.9420906947457367,
      "learning_rate": 9.566981339275557e-06,
      "loss": 0.4255084037780762,
      "num_input_tokens_seen": 204093936,
      "step": 3965,
      "train_runtime": 24468.8198,
      "train_tokens_per_second": 8340.98
    },
    {
      "epoch": 0.2673040667923512,
      "grad_norm": 0.7157303968941934,
      "learning_rate": 9.565904370097687e-06,
      "loss": 0.3954904079437256,
      "num_input_tokens_seen": 204313824,
      "step": 3970,
      "train_runtime": 24495.4906,
      "train_tokens_per_second": 8340.875
    },
    {
      "epoch": 0.26764072178831133,
      "grad_norm": 0.6789304097360346,
      "learning_rate": 9.564826124079219e-06,
      "loss": 0.38766753673553467,
      "num_input_tokens_seen": 204559872,
      "step": 3975,
      "train_runtime": 24525.9659,
      "train_tokens_per_second": 8340.543
    },
    {
      "epoch": 0.26797737678427147,
      "grad_norm": 0.7520240263818112,
      "learning_rate": 9.563746601521681e-06,
      "loss": 0.367041802406311,
      "num_input_tokens_seen": 204824896,
      "step": 3980,
      "train_runtime": 24555.7972,
      "train_tokens_per_second": 8341.203
    },
    {
      "epoch": 0.2683140317802316,
      "grad_norm": 0.8693442551367947,
      "learning_rate": 9.562665802726961e-06,
      "loss": 0.37211105823516843,
      "num_input_tokens_seen": 205083360,
      "step": 3985,
      "train_runtime": 24585.3004,
      "train_tokens_per_second": 8341.706
    },
    {
      "epoch": 0.26865068677619175,
      "grad_norm": 0.7711682763930895,
      "learning_rate": 9.561583727997297e-06,
      "loss": 0.39172868728637694,
      "num_input_tokens_seen": 205331976,
      "step": 3990,
      "train_runtime": 24614.1226,
      "train_tokens_per_second": 8342.039
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.7504678434879254,
      "learning_rate": 9.560500377635288e-06,
      "loss": 0.4008033752441406,
      "num_input_tokens_seen": 205595600,
      "step": 3995,
      "train_runtime": 24643.7629,
      "train_tokens_per_second": 8342.703
    },
    {
      "epoch": 0.26932399676811203,
      "grad_norm": 0.7913517397418455,
      "learning_rate": 9.559415751943893e-06,
      "loss": 0.40540571212768556,
      "num_input_tokens_seen": 205852400,
      "step": 4000,
      "train_runtime": 24672.208,
      "train_tokens_per_second": 8343.493
    },
    {
      "epoch": 0.26966065176407217,
      "grad_norm": 0.6513701617943163,
      "learning_rate": 9.558329851226422e-06,
      "loss": 0.3956051588058472,
      "num_input_tokens_seen": 206126568,
      "step": 4005,
      "train_runtime": 24707.8648,
      "train_tokens_per_second": 8342.549
    },
    {
      "epoch": 0.2699973067600323,
      "grad_norm": 0.8809236208485097,
      "learning_rate": 9.557242675786542e-06,
      "loss": 0.41253128051757815,
      "num_input_tokens_seen": 206389000,
      "step": 4010,
      "train_runtime": 24743.9116,
      "train_tokens_per_second": 8341.001
    },
    {
      "epoch": 0.27033396175599245,
      "grad_norm": 0.6752439625372504,
      "learning_rate": 9.55615422592828e-06,
      "loss": 0.4170052528381348,
      "num_input_tokens_seen": 206663856,
      "step": 4015,
      "train_runtime": 24778.1592,
      "train_tokens_per_second": 8340.565
    },
    {
      "epoch": 0.2706706167519526,
      "grad_norm": 0.7254819731469194,
      "learning_rate": 9.555064501956018e-06,
      "loss": 0.4035297393798828,
      "num_input_tokens_seen": 206927168,
      "step": 4020,
      "train_runtime": 24806.6838,
      "train_tokens_per_second": 8341.589
    },
    {
      "epoch": 0.2710072717479127,
      "grad_norm": 0.729536397927028,
      "learning_rate": 9.553973504174493e-06,
      "loss": 0.3743337869644165,
      "num_input_tokens_seen": 207197008,
      "step": 4025,
      "train_runtime": 24836.8566,
      "train_tokens_per_second": 8342.32
    },
    {
      "epoch": 0.27134392674387287,
      "grad_norm": 0.9944731116444012,
      "learning_rate": 9.5528812328888e-06,
      "loss": 0.40724620819091795,
      "num_input_tokens_seen": 207429960,
      "step": 4030,
      "train_runtime": 24867.1503,
      "train_tokens_per_second": 8341.525
    },
    {
      "epoch": 0.271680581739833,
      "grad_norm": 0.711380818751653,
      "learning_rate": 9.551787688404386e-06,
      "loss": 0.37790932655334475,
      "num_input_tokens_seen": 207682536,
      "step": 4035,
      "train_runtime": 24895.7678,
      "train_tokens_per_second": 8342.082
    },
    {
      "epoch": 0.27201723673579314,
      "grad_norm": 0.7205947435258002,
      "learning_rate": 9.550692871027061e-06,
      "loss": 0.37463831901550293,
      "num_input_tokens_seen": 207939608,
      "step": 4040,
      "train_runtime": 24927.1754,
      "train_tokens_per_second": 8341.884
    },
    {
      "epoch": 0.2723538917317533,
      "grad_norm": 0.8125539159558312,
      "learning_rate": 9.549596781062984e-06,
      "loss": 0.37658646106719973,
      "num_input_tokens_seen": 208204768,
      "step": 4045,
      "train_runtime": 24955.341,
      "train_tokens_per_second": 8343.094
    },
    {
      "epoch": 0.2726905467277134,
      "grad_norm": 0.8288623584572793,
      "learning_rate": 9.548499418818677e-06,
      "loss": 0.3846138000488281,
      "num_input_tokens_seen": 208433640,
      "step": 4050,
      "train_runtime": 24983.6741,
      "train_tokens_per_second": 8342.794
    },
    {
      "epoch": 0.27302720172367356,
      "grad_norm": 0.8076150915439154,
      "learning_rate": 9.547400784601011e-06,
      "loss": 0.4360748291015625,
      "num_input_tokens_seen": 208680936,
      "step": 4055,
      "train_runtime": 25014.3264,
      "train_tokens_per_second": 8342.457
    },
    {
      "epoch": 0.2733638567196337,
      "grad_norm": 0.7834420920805673,
      "learning_rate": 9.546300878717216e-06,
      "loss": 0.4010347843170166,
      "num_input_tokens_seen": 208961928,
      "step": 4060,
      "train_runtime": 25045.6554,
      "train_tokens_per_second": 8343.241
    },
    {
      "epoch": 0.27370051171559384,
      "grad_norm": 0.7394826019868475,
      "learning_rate": 9.545199701474877e-06,
      "loss": 0.39485416412353513,
      "num_input_tokens_seen": 209209832,
      "step": 4065,
      "train_runtime": 25072.0742,
      "train_tokens_per_second": 8344.337
    },
    {
      "epoch": 0.274037166711554,
      "grad_norm": 0.8004556560043369,
      "learning_rate": 9.544097253181935e-06,
      "loss": 0.36254172325134276,
      "num_input_tokens_seen": 209452944,
      "step": 4070,
      "train_runtime": 25103.5587,
      "train_tokens_per_second": 8343.556
    },
    {
      "epoch": 0.2743738217075141,
      "grad_norm": 0.6625225498311411,
      "learning_rate": 9.542993534146687e-06,
      "loss": 0.398806095123291,
      "num_input_tokens_seen": 209697400,
      "step": 4075,
      "train_runtime": 25134.4091,
      "train_tokens_per_second": 8343.041
    },
    {
      "epoch": 0.27471047670347426,
      "grad_norm": 0.8054681280373299,
      "learning_rate": 9.541888544677784e-06,
      "loss": 0.39751219749450684,
      "num_input_tokens_seen": 209954200,
      "step": 4080,
      "train_runtime": 25166.7369,
      "train_tokens_per_second": 8342.528
    },
    {
      "epoch": 0.2750471316994344,
      "grad_norm": 0.7373775845398405,
      "learning_rate": 9.540782285084232e-06,
      "loss": 0.3591721773147583,
      "num_input_tokens_seen": 210202392,
      "step": 4085,
      "train_runtime": 25196.7251,
      "train_tokens_per_second": 8342.449
    },
    {
      "epoch": 0.27538378669539454,
      "grad_norm": 0.6765044143269571,
      "learning_rate": 9.539674755675392e-06,
      "loss": 0.38759493827819824,
      "num_input_tokens_seen": 210455416,
      "step": 4090,
      "train_runtime": 25228.481,
      "train_tokens_per_second": 8341.977
    },
    {
      "epoch": 0.2757204416913547,
      "grad_norm": 0.6399540181865827,
      "learning_rate": 9.538565956760983e-06,
      "loss": 0.34577934741973876,
      "num_input_tokens_seen": 210735816,
      "step": 4095,
      "train_runtime": 25259.6461,
      "train_tokens_per_second": 8342.786
    },
    {
      "epoch": 0.2760570966873148,
      "grad_norm": 0.6167253378268269,
      "learning_rate": 9.537455888651078e-06,
      "loss": 0.3741654634475708,
      "num_input_tokens_seen": 211010616,
      "step": 4100,
      "train_runtime": 25290.5185,
      "train_tokens_per_second": 8343.467
    },
    {
      "epoch": 0.27639375168327496,
      "grad_norm": 0.7013473590994608,
      "learning_rate": 9.536344551656103e-06,
      "loss": 0.41167469024658204,
      "num_input_tokens_seen": 211277568,
      "step": 4105,
      "train_runtime": 25324.5306,
      "train_tokens_per_second": 8342.803
    },
    {
      "epoch": 0.2767304066792351,
      "grad_norm": 0.6876393516673509,
      "learning_rate": 9.535231946086838e-06,
      "loss": 0.3873252868652344,
      "num_input_tokens_seen": 211540592,
      "step": 4110,
      "train_runtime": 25359.2432,
      "train_tokens_per_second": 8341.755
    },
    {
      "epoch": 0.27706706167519524,
      "grad_norm": 0.698897295935192,
      "learning_rate": 9.534118072254421e-06,
      "loss": 0.4023136615753174,
      "num_input_tokens_seen": 211799024,
      "step": 4115,
      "train_runtime": 25391.769,
      "train_tokens_per_second": 8341.247
    },
    {
      "epoch": 0.2774037166711554,
      "grad_norm": 0.6910699574663766,
      "learning_rate": 9.533002930470345e-06,
      "loss": 0.3745990753173828,
      "num_input_tokens_seen": 212045832,
      "step": 4120,
      "train_runtime": 25422.4364,
      "train_tokens_per_second": 8340.893
    },
    {
      "epoch": 0.2777403716671155,
      "grad_norm": 0.6487896914704242,
      "learning_rate": 9.531886521046452e-06,
      "loss": 0.3797455310821533,
      "num_input_tokens_seen": 212270064,
      "step": 4125,
      "train_runtime": 25452.2861,
      "train_tokens_per_second": 8339.921
    },
    {
      "epoch": 0.27807702666307565,
      "grad_norm": 0.8366687442039256,
      "learning_rate": 9.530768844294947e-06,
      "loss": 0.4085412502288818,
      "num_input_tokens_seen": 212533360,
      "step": 4130,
      "train_runtime": 25480.744,
      "train_tokens_per_second": 8340.94
    },
    {
      "epoch": 0.2784136816590358,
      "grad_norm": 0.8519285717088118,
      "learning_rate": 9.52964990052838e-06,
      "loss": 0.4124325752258301,
      "num_input_tokens_seen": 212802432,
      "step": 4135,
      "train_runtime": 25511.9141,
      "train_tokens_per_second": 8341.296
    },
    {
      "epoch": 0.27875033665499593,
      "grad_norm": 0.7352276844616217,
      "learning_rate": 9.528529690059663e-06,
      "loss": 0.3804769515991211,
      "num_input_tokens_seen": 213062920,
      "step": 4140,
      "train_runtime": 25543.2402,
      "train_tokens_per_second": 8341.264
    },
    {
      "epoch": 0.2790869916509561,
      "grad_norm": 0.7887118493297387,
      "learning_rate": 9.52740821320206e-06,
      "loss": 0.3981710195541382,
      "num_input_tokens_seen": 213321736,
      "step": 4145,
      "train_runtime": 25580.2653,
      "train_tokens_per_second": 8339.309
    },
    {
      "epoch": 0.2794236466469162,
      "grad_norm": 0.6171291651887711,
      "learning_rate": 9.526285470269185e-06,
      "loss": 0.41275882720947266,
      "num_input_tokens_seen": 213599296,
      "step": 4150,
      "train_runtime": 25612.124,
      "train_tokens_per_second": 8339.773
    },
    {
      "epoch": 0.2797603016428764,
      "grad_norm": 0.7157718016949131,
      "learning_rate": 9.525161461575012e-06,
      "loss": 0.40100951194763185,
      "num_input_tokens_seen": 213832240,
      "step": 4155,
      "train_runtime": 25640.4425,
      "train_tokens_per_second": 8339.647
    },
    {
      "epoch": 0.28009695663883655,
      "grad_norm": 0.7000630009679055,
      "learning_rate": 9.524036187433867e-06,
      "loss": 0.3759552001953125,
      "num_input_tokens_seen": 214079352,
      "step": 4160,
      "train_runtime": 25669.9885,
      "train_tokens_per_second": 8339.675
    },
    {
      "epoch": 0.2804336116347967,
      "grad_norm": 0.7364667371913711,
      "learning_rate": 9.522909648160427e-06,
      "loss": 0.3883460760116577,
      "num_input_tokens_seen": 214341344,
      "step": 4165,
      "train_runtime": 25701.7985,
      "train_tokens_per_second": 8339.547
    },
    {
      "epoch": 0.2807702666307568,
      "grad_norm": 0.7875076472171954,
      "learning_rate": 9.521781844069727e-06,
      "loss": 0.3881689548492432,
      "num_input_tokens_seen": 214603192,
      "step": 4170,
      "train_runtime": 25732.6598,
      "train_tokens_per_second": 8339.721
    },
    {
      "epoch": 0.28110692162671697,
      "grad_norm": 0.7386344079672109,
      "learning_rate": 9.520652775477154e-06,
      "loss": 0.37189648151397703,
      "num_input_tokens_seen": 214852488,
      "step": 4175,
      "train_runtime": 25764.9416,
      "train_tokens_per_second": 8338.947
    },
    {
      "epoch": 0.2814435766226771,
      "grad_norm": 0.5753689919733803,
      "learning_rate": 9.519522442698447e-06,
      "loss": 0.3522214651107788,
      "num_input_tokens_seen": 215104536,
      "step": 4180,
      "train_runtime": 25795.5519,
      "train_tokens_per_second": 8338.823
    },
    {
      "epoch": 0.28178023161863724,
      "grad_norm": 0.820880541584056,
      "learning_rate": 9.518390846049699e-06,
      "loss": 0.3984562397003174,
      "num_input_tokens_seen": 215339504,
      "step": 4185,
      "train_runtime": 25824.4227,
      "train_tokens_per_second": 8338.599
    },
    {
      "epoch": 0.2821168866145974,
      "grad_norm": 0.7066712924584547,
      "learning_rate": 9.517257985847361e-06,
      "loss": 0.41177897453308104,
      "num_input_tokens_seen": 215578384,
      "step": 4190,
      "train_runtime": 25855.9567,
      "train_tokens_per_second": 8337.668
    },
    {
      "epoch": 0.2824535416105575,
      "grad_norm": 0.6392449468355408,
      "learning_rate": 9.516123862408232e-06,
      "loss": 0.42540297508239744,
      "num_input_tokens_seen": 215838272,
      "step": 4195,
      "train_runtime": 25884.8238,
      "train_tokens_per_second": 8338.41
    },
    {
      "epoch": 0.28279019660651766,
      "grad_norm": 0.6678136878738598,
      "learning_rate": 9.514988476049466e-06,
      "loss": 0.36865365505218506,
      "num_input_tokens_seen": 216081440,
      "step": 4200,
      "train_runtime": 25915.3449,
      "train_tokens_per_second": 8337.973
    },
    {
      "epoch": 0.2831268516024778,
      "grad_norm": 0.5876426680708081,
      "learning_rate": 9.513851827088567e-06,
      "loss": 0.3833805561065674,
      "num_input_tokens_seen": 216319712,
      "step": 4205,
      "train_runtime": 25943.8753,
      "train_tokens_per_second": 8337.988
    },
    {
      "epoch": 0.28346350659843794,
      "grad_norm": 0.9933335722160103,
      "learning_rate": 9.512713915843402e-06,
      "loss": 0.37570858001708984,
      "num_input_tokens_seen": 216600008,
      "step": 4210,
      "train_runtime": 25976.9847,
      "train_tokens_per_second": 8338.151
    },
    {
      "epoch": 0.2838001615943981,
      "grad_norm": 0.6898227364172558,
      "learning_rate": 9.511574742632177e-06,
      "loss": 0.37197585105895997,
      "num_input_tokens_seen": 216842896,
      "step": 4215,
      "train_runtime": 26008.5079,
      "train_tokens_per_second": 8337.383
    },
    {
      "epoch": 0.2841368165903582,
      "grad_norm": 0.6760414566635102,
      "learning_rate": 9.510434307773464e-06,
      "loss": 0.44431452751159667,
      "num_input_tokens_seen": 217109376,
      "step": 4220,
      "train_runtime": 26039.2332,
      "train_tokens_per_second": 8337.779
    },
    {
      "epoch": 0.28447347158631836,
      "grad_norm": 0.6187155189305044,
      "learning_rate": 9.509292611586179e-06,
      "loss": 0.34885530471801757,
      "num_input_tokens_seen": 217315776,
      "step": 4225,
      "train_runtime": 26068.5086,
      "train_tokens_per_second": 8336.333
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.7141878618973462,
      "learning_rate": 9.508149654389592e-06,
      "loss": 0.4107461929321289,
      "num_input_tokens_seen": 217566424,
      "step": 4230,
      "train_runtime": 26102.3554,
      "train_tokens_per_second": 8335.126
    },
    {
      "epoch": 0.28514678157823864,
      "grad_norm": 0.572441758122331,
      "learning_rate": 9.507005436503331e-06,
      "loss": 0.3577643871307373,
      "num_input_tokens_seen": 217848000,
      "step": 4235,
      "train_runtime": 26140.1716,
      "train_tokens_per_second": 8333.84
    },
    {
      "epoch": 0.2854834365741988,
      "grad_norm": 0.8692374959181597,
      "learning_rate": 9.505859958247373e-06,
      "loss": 0.3937828540802002,
      "num_input_tokens_seen": 218098664,
      "step": 4240,
      "train_runtime": 26172.3396,
      "train_tokens_per_second": 8333.174
    },
    {
      "epoch": 0.2858200915701589,
      "grad_norm": 0.5965196175476413,
      "learning_rate": 9.504713219942045e-06,
      "loss": 0.3750722885131836,
      "num_input_tokens_seen": 218356480,
      "step": 4245,
      "train_runtime": 26201.2043,
      "train_tokens_per_second": 8333.834
    },
    {
      "epoch": 0.28615674656611906,
      "grad_norm": 0.6952473347507232,
      "learning_rate": 9.503565221908029e-06,
      "loss": 0.4015070915222168,
      "num_input_tokens_seen": 218624800,
      "step": 4250,
      "train_runtime": 26233.8422,
      "train_tokens_per_second": 8333.693
    },
    {
      "epoch": 0.2864934015620792,
      "grad_norm": 0.7497657759856564,
      "learning_rate": 9.50241596446636e-06,
      "loss": 0.4091782569885254,
      "num_input_tokens_seen": 218901440,
      "step": 4255,
      "train_runtime": 26263.7762,
      "train_tokens_per_second": 8334.728
    },
    {
      "epoch": 0.28683005655803934,
      "grad_norm": 0.8095209358057489,
      "learning_rate": 9.501265447938423e-06,
      "loss": 0.3998479604721069,
      "num_input_tokens_seen": 219141968,
      "step": 4260,
      "train_runtime": 26294.1672,
      "train_tokens_per_second": 8334.243
    },
    {
      "epoch": 0.2871667115539995,
      "grad_norm": 0.8011810441010684,
      "learning_rate": 9.500113672645958e-06,
      "loss": 0.3934728384017944,
      "num_input_tokens_seen": 219396624,
      "step": 4265,
      "train_runtime": 26323.569,
      "train_tokens_per_second": 8334.608
    },
    {
      "epoch": 0.2875033665499596,
      "grad_norm": 0.8462608522947167,
      "learning_rate": 9.498960638911054e-06,
      "loss": 0.37127046585083007,
      "num_input_tokens_seen": 219627464,
      "step": 4270,
      "train_runtime": 26353.7814,
      "train_tokens_per_second": 8333.812
    },
    {
      "epoch": 0.28784002154591976,
      "grad_norm": 0.6976360607813245,
      "learning_rate": 9.497806347056153e-06,
      "loss": 0.38576641082763674,
      "num_input_tokens_seen": 219882184,
      "step": 4275,
      "train_runtime": 26385.3814,
      "train_tokens_per_second": 8333.485
    },
    {
      "epoch": 0.2881766765418799,
      "grad_norm": 0.69800452178568,
      "learning_rate": 9.49665079740405e-06,
      "loss": 0.4180449485778809,
      "num_input_tokens_seen": 220146496,
      "step": 4280,
      "train_runtime": 26415.1456,
      "train_tokens_per_second": 8334.101
    },
    {
      "epoch": 0.28851333153784003,
      "grad_norm": 0.866267678237035,
      "learning_rate": 9.495493990277889e-06,
      "loss": 0.4000094413757324,
      "num_input_tokens_seen": 220393120,
      "step": 4285,
      "train_runtime": 26446.3277,
      "train_tokens_per_second": 8333.6
    },
    {
      "epoch": 0.2888499865338002,
      "grad_norm": 0.6715024469077907,
      "learning_rate": 9.49433592600117e-06,
      "loss": 0.3954871416091919,
      "num_input_tokens_seen": 220650184,
      "step": 4290,
      "train_runtime": 26477.1691,
      "train_tokens_per_second": 8333.602
    },
    {
      "epoch": 0.2891866415297603,
      "grad_norm": 0.7080465656881908,
      "learning_rate": 9.49317660489774e-06,
      "loss": 0.38829166889190675,
      "num_input_tokens_seen": 220925320,
      "step": 4295,
      "train_runtime": 26509.1664,
      "train_tokens_per_second": 8333.922
    },
    {
      "epoch": 0.28952329652572045,
      "grad_norm": 0.6653622931725003,
      "learning_rate": 9.4920160272918e-06,
      "loss": 0.3876731157302856,
      "num_input_tokens_seen": 221203224,
      "step": 4300,
      "train_runtime": 26539.9355,
      "train_tokens_per_second": 8334.731
    },
    {
      "epoch": 0.2898599515216806,
      "grad_norm": 0.7916065720642245,
      "learning_rate": 9.490854193507904e-06,
      "loss": 0.39028306007385255,
      "num_input_tokens_seen": 221456208,
      "step": 4305,
      "train_runtime": 26570.2064,
      "train_tokens_per_second": 8334.757
    },
    {
      "epoch": 0.29019660651764073,
      "grad_norm": 0.6915879138797075,
      "learning_rate": 9.489691103870951e-06,
      "loss": 0.40920300483703614,
      "num_input_tokens_seen": 221735136,
      "step": 4310,
      "train_runtime": 26603.9595,
      "train_tokens_per_second": 8334.667
    },
    {
      "epoch": 0.29053326151360087,
      "grad_norm": 0.65366333925972,
      "learning_rate": 9.488526758706198e-06,
      "loss": 0.3853466033935547,
      "num_input_tokens_seen": 222003328,
      "step": 4315,
      "train_runtime": 26633.6048,
      "train_tokens_per_second": 8335.459
    },
    {
      "epoch": 0.290869916509561,
      "grad_norm": 0.9236168247418115,
      "learning_rate": 9.48736115833925e-06,
      "loss": 0.40279550552368165,
      "num_input_tokens_seen": 222244144,
      "step": 4320,
      "train_runtime": 26666.8946,
      "train_tokens_per_second": 8334.084
    },
    {
      "epoch": 0.29120657150552115,
      "grad_norm": 0.6456287763916503,
      "learning_rate": 9.486194303096062e-06,
      "loss": 0.3805595636367798,
      "num_input_tokens_seen": 222515048,
      "step": 4325,
      "train_runtime": 26696.495,
      "train_tokens_per_second": 8334.991
    },
    {
      "epoch": 0.2915432265014813,
      "grad_norm": 0.555197022657934,
      "learning_rate": 9.485026193302945e-06,
      "loss": 0.3725703239440918,
      "num_input_tokens_seen": 222786656,
      "step": 4330,
      "train_runtime": 26725.5193,
      "train_tokens_per_second": 8336.102
    },
    {
      "epoch": 0.29187988149744143,
      "grad_norm": 0.6050077403851439,
      "learning_rate": 9.483856829286552e-06,
      "loss": 0.42488908767700195,
      "num_input_tokens_seen": 223051840,
      "step": 4335,
      "train_runtime": 26758.8464,
      "train_tokens_per_second": 8335.63
    },
    {
      "epoch": 0.29221653649340157,
      "grad_norm": 1.2393862155810138,
      "learning_rate": 9.482686211373896e-06,
      "loss": 0.4004028797149658,
      "num_input_tokens_seen": 223318136,
      "step": 4340,
      "train_runtime": 26791.1969,
      "train_tokens_per_second": 8335.504
    },
    {
      "epoch": 0.2925531914893617,
      "grad_norm": 0.8257089321213631,
      "learning_rate": 9.481514339892335e-06,
      "loss": 0.4182788848876953,
      "num_input_tokens_seen": 223595576,
      "step": 4345,
      "train_runtime": 26822.5696,
      "train_tokens_per_second": 8336.098
    },
    {
      "epoch": 0.29288984648532185,
      "grad_norm": 0.6930749724788373,
      "learning_rate": 9.48034121516958e-06,
      "loss": 0.4136636734008789,
      "num_input_tokens_seen": 223844376,
      "step": 4350,
      "train_runtime": 26855.3039,
      "train_tokens_per_second": 8335.202
    },
    {
      "epoch": 0.293226501481282,
      "grad_norm": 0.5936548397412114,
      "learning_rate": 9.47916683753369e-06,
      "loss": 0.38527374267578124,
      "num_input_tokens_seen": 224096800,
      "step": 4355,
      "train_runtime": 26882.3375,
      "train_tokens_per_second": 8336.21
    },
    {
      "epoch": 0.2935631564772421,
      "grad_norm": 0.6524758185768371,
      "learning_rate": 9.477991207313077e-06,
      "loss": 0.39638352394104004,
      "num_input_tokens_seen": 224375232,
      "step": 4360,
      "train_runtime": 26913.9554,
      "train_tokens_per_second": 8336.762
    },
    {
      "epoch": 0.29389981147320227,
      "grad_norm": 0.6696688847758661,
      "learning_rate": 9.476814324836504e-06,
      "loss": 0.37292141914367677,
      "num_input_tokens_seen": 224650360,
      "step": 4365,
      "train_runtime": 26946.8996,
      "train_tokens_per_second": 8336.779
    },
    {
      "epoch": 0.2942364664691624,
      "grad_norm": 0.7208510791514862,
      "learning_rate": 9.475636190433078e-06,
      "loss": 0.44341106414794923,
      "num_input_tokens_seen": 224937648,
      "step": 4370,
      "train_runtime": 26980.9086,
      "train_tokens_per_second": 8336.919
    },
    {
      "epoch": 0.29457312146512254,
      "grad_norm": 0.6611946217814506,
      "learning_rate": 9.474456804432264e-06,
      "loss": 0.39295039176940916,
      "num_input_tokens_seen": 225217712,
      "step": 4375,
      "train_runtime": 27014.0478,
      "train_tokens_per_second": 8337.059
    },
    {
      "epoch": 0.2949097764610827,
      "grad_norm": 0.6671818660368126,
      "learning_rate": 9.473276167163872e-06,
      "loss": 0.40974793434143064,
      "num_input_tokens_seen": 225496128,
      "step": 4380,
      "train_runtime": 27046.6043,
      "train_tokens_per_second": 8337.318
    },
    {
      "epoch": 0.2952464314570428,
      "grad_norm": 0.6144271533314608,
      "learning_rate": 9.472094278958066e-06,
      "loss": 0.39631237983703616,
      "num_input_tokens_seen": 225775368,
      "step": 4385,
      "train_runtime": 27077.7379,
      "train_tokens_per_second": 8338.044
    },
    {
      "epoch": 0.29558308645300296,
      "grad_norm": 0.671656122216855,
      "learning_rate": 9.470911140145353e-06,
      "loss": 0.3836498260498047,
      "num_input_tokens_seen": 226020480,
      "step": 4390,
      "train_runtime": 27106.6966,
      "train_tokens_per_second": 8338.179
    },
    {
      "epoch": 0.2959197414489631,
      "grad_norm": 0.8363340732543063,
      "learning_rate": 9.469726751056599e-06,
      "loss": 0.3979063034057617,
      "num_input_tokens_seen": 226278016,
      "step": 4395,
      "train_runtime": 27135.0181,
      "train_tokens_per_second": 8338.967
    },
    {
      "epoch": 0.29625639644492324,
      "grad_norm": 0.7535293275974959,
      "learning_rate": 9.468541112023009e-06,
      "loss": 0.4077214241027832,
      "num_input_tokens_seen": 226517536,
      "step": 4400,
      "train_runtime": 27163.2076,
      "train_tokens_per_second": 8339.131
    },
    {
      "epoch": 0.2965930514408834,
      "grad_norm": 1.0141756583325503,
      "learning_rate": 9.467354223376149e-06,
      "loss": 0.40714001655578613,
      "num_input_tokens_seen": 226729200,
      "step": 4405,
      "train_runtime": 27189.9435,
      "train_tokens_per_second": 8338.715
    },
    {
      "epoch": 0.2969297064368435,
      "grad_norm": 0.6753703830750982,
      "learning_rate": 9.466166085447923e-06,
      "loss": 0.3891974925994873,
      "num_input_tokens_seen": 226976280,
      "step": 4410,
      "train_runtime": 27220.1582,
      "train_tokens_per_second": 8338.536
    },
    {
      "epoch": 0.29726636143280366,
      "grad_norm": 0.792639968993046,
      "learning_rate": 9.464976698570595e-06,
      "loss": 0.3855931282043457,
      "num_input_tokens_seen": 227196504,
      "step": 4415,
      "train_runtime": 27253.1153,
      "train_tokens_per_second": 8336.533
    },
    {
      "epoch": 0.2976030164287638,
      "grad_norm": 0.7999972920595089,
      "learning_rate": 9.463786063076769e-06,
      "loss": 0.40854344367980955,
      "num_input_tokens_seen": 227462128,
      "step": 4420,
      "train_runtime": 27281.8111,
      "train_tokens_per_second": 8337.501
    },
    {
      "epoch": 0.29793967142472394,
      "grad_norm": 0.8199522519848642,
      "learning_rate": 9.462594179299408e-06,
      "loss": 0.3372624158859253,
      "num_input_tokens_seen": 227712168,
      "step": 4425,
      "train_runtime": 27311.2007,
      "train_tokens_per_second": 8337.684
    },
    {
      "epoch": 0.2982763264206841,
      "grad_norm": 0.6367459006346817,
      "learning_rate": 9.461401047571811e-06,
      "loss": 0.37390589714050293,
      "num_input_tokens_seen": 227989944,
      "step": 4430,
      "train_runtime": 27343.1881,
      "train_tokens_per_second": 8338.089
    },
    {
      "epoch": 0.2986129814166442,
      "grad_norm": 0.6607751726085161,
      "learning_rate": 9.460206668227639e-06,
      "loss": 0.4041748046875,
      "num_input_tokens_seen": 228257520,
      "step": 4435,
      "train_runtime": 27373.9644,
      "train_tokens_per_second": 8338.49
    },
    {
      "epoch": 0.29894963641260436,
      "grad_norm": 0.6245237380136043,
      "learning_rate": 9.459011041600895e-06,
      "loss": 0.3765630006790161,
      "num_input_tokens_seen": 228529080,
      "step": 4440,
      "train_runtime": 27406.4984,
      "train_tokens_per_second": 8338.5
    },
    {
      "epoch": 0.2992862914085645,
      "grad_norm": 0.7515421939804843,
      "learning_rate": 9.457814168025932e-06,
      "loss": 0.37979490756988527,
      "num_input_tokens_seen": 228793336,
      "step": 4445,
      "train_runtime": 27438.1096,
      "train_tokens_per_second": 8338.524
    },
    {
      "epoch": 0.29962294640452464,
      "grad_norm": 0.760986120286953,
      "learning_rate": 9.456616047837452e-06,
      "loss": 0.4283762454986572,
      "num_input_tokens_seen": 229014144,
      "step": 4450,
      "train_runtime": 27466.6219,
      "train_tokens_per_second": 8337.907
    },
    {
      "epoch": 0.2999596014004848,
      "grad_norm": 0.6223600574737844,
      "learning_rate": 9.455416681370506e-06,
      "loss": 0.4226823806762695,
      "num_input_tokens_seen": 229285584,
      "step": 4455,
      "train_runtime": 27496.8686,
      "train_tokens_per_second": 8338.607
    },
    {
      "epoch": 0.3002962563964449,
      "grad_norm": 0.8223625940918688,
      "learning_rate": 9.45421606896049e-06,
      "loss": 0.40646953582763673,
      "num_input_tokens_seen": 229552816,
      "step": 4460,
      "train_runtime": 27526.2201,
      "train_tokens_per_second": 8339.424
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.6720473576005771,
      "learning_rate": 9.453014210943155e-06,
      "loss": 0.39110302925109863,
      "num_input_tokens_seen": 229807520,
      "step": 4465,
      "train_runtime": 27554.0716,
      "train_tokens_per_second": 8340.238
    },
    {
      "epoch": 0.3009695663883652,
      "grad_norm": 0.6680715292938918,
      "learning_rate": 9.451811107654596e-06,
      "loss": 0.386443567276001,
      "num_input_tokens_seen": 230066584,
      "step": 4470,
      "train_runtime": 27584.4384,
      "train_tokens_per_second": 8340.448
    },
    {
      "epoch": 0.30130622138432533,
      "grad_norm": 0.6470378525581002,
      "learning_rate": 9.450606759431255e-06,
      "loss": 0.3894406080245972,
      "num_input_tokens_seen": 230343608,
      "step": 4475,
      "train_runtime": 27614.7401,
      "train_tokens_per_second": 8341.328
    },
    {
      "epoch": 0.3016428763802855,
      "grad_norm": 0.8367211329824955,
      "learning_rate": 9.449401166609928e-06,
      "loss": 0.4272471904754639,
      "num_input_tokens_seen": 230592552,
      "step": 4480,
      "train_runtime": 27642.0289,
      "train_tokens_per_second": 8342.099
    },
    {
      "epoch": 0.3019795313762456,
      "grad_norm": 0.949587400564466,
      "learning_rate": 9.448194329527752e-06,
      "loss": 0.3923869848251343,
      "num_input_tokens_seen": 230848184,
      "step": 4485,
      "train_runtime": 27671.9295,
      "train_tokens_per_second": 8342.323
    },
    {
      "epoch": 0.30231618637220575,
      "grad_norm": 0.6115084948003859,
      "learning_rate": 9.446986248522216e-06,
      "loss": 0.37474827766418456,
      "num_input_tokens_seen": 231113968,
      "step": 4490,
      "train_runtime": 27703.7768,
      "train_tokens_per_second": 8342.327
    },
    {
      "epoch": 0.3026528413681659,
      "grad_norm": 0.7803696057752092,
      "learning_rate": 9.445776923931157e-06,
      "loss": 0.36249709129333496,
      "num_input_tokens_seen": 231371848,
      "step": 4495,
      "train_runtime": 27734.5801,
      "train_tokens_per_second": 8342.36
    },
    {
      "epoch": 0.30298949636412603,
      "grad_norm": 0.6726473136494312,
      "learning_rate": 9.444566356092754e-06,
      "loss": 0.3772896766662598,
      "num_input_tokens_seen": 231652432,
      "step": 4500,
      "train_runtime": 27766.2477,
      "train_tokens_per_second": 8342.951
    },
    {
      "epoch": 0.30332615136008617,
      "grad_norm": 0.7413072101254118,
      "learning_rate": 9.443354545345545e-06,
      "loss": 0.3506791114807129,
      "num_input_tokens_seen": 231915816,
      "step": 4505,
      "train_runtime": 27801.2196,
      "train_tokens_per_second": 8341.93
    },
    {
      "epoch": 0.3036628063560463,
      "grad_norm": 0.6802833268563316,
      "learning_rate": 9.442141492028406e-06,
      "loss": 0.38966143131256104,
      "num_input_tokens_seen": 232167472,
      "step": 4510,
      "train_runtime": 27831.1983,
      "train_tokens_per_second": 8341.986
    },
    {
      "epoch": 0.30399946135200645,
      "grad_norm": 0.7604984222061342,
      "learning_rate": 9.440927196480563e-06,
      "loss": 0.38597960472106935,
      "num_input_tokens_seen": 232443640,
      "step": 4515,
      "train_runtime": 27866.9882,
      "train_tokens_per_second": 8341.183
    },
    {
      "epoch": 0.3043361163479666,
      "grad_norm": 0.626755747381137,
      "learning_rate": 9.439711659041593e-06,
      "loss": 0.3944221019744873,
      "num_input_tokens_seen": 232689752,
      "step": 4520,
      "train_runtime": 27901.007,
      "train_tokens_per_second": 8339.833
    },
    {
      "epoch": 0.30467277134392673,
      "grad_norm": 0.8312854454573649,
      "learning_rate": 9.438494880051413e-06,
      "loss": 0.39769792556762695,
      "num_input_tokens_seen": 232943776,
      "step": 4525,
      "train_runtime": 27932.1363,
      "train_tokens_per_second": 8339.633
    },
    {
      "epoch": 0.30500942633988687,
      "grad_norm": 0.6708775821034578,
      "learning_rate": 9.437276859850293e-06,
      "loss": 0.3947306632995605,
      "num_input_tokens_seen": 233205528,
      "step": 4530,
      "train_runtime": 27965.6125,
      "train_tokens_per_second": 8339.01
    },
    {
      "epoch": 0.305346081335847,
      "grad_norm": 0.6450363709739184,
      "learning_rate": 9.436057598778848e-06,
      "loss": 0.4011570930480957,
      "num_input_tokens_seen": 233470232,
      "step": 4535,
      "train_runtime": 27995.7453,
      "train_tokens_per_second": 8339.49
    },
    {
      "epoch": 0.30568273633180715,
      "grad_norm": 0.691133139379876,
      "learning_rate": 9.434837097178043e-06,
      "loss": 0.4221809864044189,
      "num_input_tokens_seen": 233741488,
      "step": 4540,
      "train_runtime": 28023.7125,
      "train_tokens_per_second": 8340.847
    },
    {
      "epoch": 0.3060193913277673,
      "grad_norm": 0.7688156865357128,
      "learning_rate": 9.433615355389183e-06,
      "loss": 0.4105339527130127,
      "num_input_tokens_seen": 233978248,
      "step": 4545,
      "train_runtime": 28053.9253,
      "train_tokens_per_second": 8340.303
    },
    {
      "epoch": 0.3063560463237274,
      "grad_norm": 0.6637436524234368,
      "learning_rate": 9.432392373753926e-06,
      "loss": 0.3822654962539673,
      "num_input_tokens_seen": 234260104,
      "step": 4550,
      "train_runtime": 28088.0762,
      "train_tokens_per_second": 8340.198
    },
    {
      "epoch": 0.30669270131968757,
      "grad_norm": 0.7454377268434984,
      "learning_rate": 9.431168152614278e-06,
      "loss": 0.38636274337768556,
      "num_input_tokens_seen": 234511816,
      "step": 4555,
      "train_runtime": 28118.5499,
      "train_tokens_per_second": 8340.111
    },
    {
      "epoch": 0.3070293563156477,
      "grad_norm": 0.6963587872589723,
      "learning_rate": 9.429942692312585e-06,
      "loss": 0.34262690544128416,
      "num_input_tokens_seen": 234783840,
      "step": 4560,
      "train_runtime": 28147.9729,
      "train_tokens_per_second": 8341.057
    },
    {
      "epoch": 0.30736601131160785,
      "grad_norm": 0.7410260312763024,
      "learning_rate": 9.428715993191546e-06,
      "loss": 0.3756542682647705,
      "num_input_tokens_seen": 235052936,
      "step": 4565,
      "train_runtime": 28174.9879,
      "train_tokens_per_second": 8342.61
    },
    {
      "epoch": 0.307702666307568,
      "grad_norm": 0.9057929802831579,
      "learning_rate": 9.427488055594199e-06,
      "loss": 0.3860651969909668,
      "num_input_tokens_seen": 235287520,
      "step": 4570,
      "train_runtime": 28202.1438,
      "train_tokens_per_second": 8342.895
    },
    {
      "epoch": 0.3080393213035281,
      "grad_norm": 0.6799483910633294,
      "learning_rate": 9.426258879863937e-06,
      "loss": 0.4013540744781494,
      "num_input_tokens_seen": 235581488,
      "step": 4575,
      "train_runtime": 28238.9488,
      "train_tokens_per_second": 8342.431
    },
    {
      "epoch": 0.30837597629948826,
      "grad_norm": 0.6218217948027893,
      "learning_rate": 9.425028466344494e-06,
      "loss": 0.35372772216796877,
      "num_input_tokens_seen": 235842264,
      "step": 4580,
      "train_runtime": 28272.0442,
      "train_tokens_per_second": 8341.889
    },
    {
      "epoch": 0.3087126312954484,
      "grad_norm": 0.7338564199985705,
      "learning_rate": 9.42379681537995e-06,
      "loss": 0.4014106750488281,
      "num_input_tokens_seen": 236092856,
      "step": 4585,
      "train_runtime": 28303.1061,
      "train_tokens_per_second": 8341.588
    },
    {
      "epoch": 0.30904928629140854,
      "grad_norm": 0.7663736028608944,
      "learning_rate": 9.422563927314732e-06,
      "loss": 0.39305777549743653,
      "num_input_tokens_seen": 236358592,
      "step": 4590,
      "train_runtime": 28333.3699,
      "train_tokens_per_second": 8342.057
    },
    {
      "epoch": 0.3093859412873687,
      "grad_norm": 0.6691380655811521,
      "learning_rate": 9.421329802493615e-06,
      "loss": 0.3737236976623535,
      "num_input_tokens_seen": 236618776,
      "step": 4595,
      "train_runtime": 28363.6207,
      "train_tokens_per_second": 8342.333
    },
    {
      "epoch": 0.3097225962833288,
      "grad_norm": 0.7252448161884978,
      "learning_rate": 9.420094441261717e-06,
      "loss": 0.3912565469741821,
      "num_input_tokens_seen": 236886704,
      "step": 4600,
      "train_runtime": 28389.6788,
      "train_tokens_per_second": 8344.114
    },
    {
      "epoch": 0.31005925127928896,
      "grad_norm": 1.2243578861529671,
      "learning_rate": 9.418857843964506e-06,
      "loss": 0.3722475290298462,
      "num_input_tokens_seen": 237135856,
      "step": 4605,
      "train_runtime": 28421.9232,
      "train_tokens_per_second": 8343.413
    },
    {
      "epoch": 0.3103959062752491,
      "grad_norm": 0.6418099335605021,
      "learning_rate": 9.417620010947786e-06,
      "loss": 0.40509729385375975,
      "num_input_tokens_seen": 237387984,
      "step": 4610,
      "train_runtime": 28447.2249,
      "train_tokens_per_second": 8344.856
    },
    {
      "epoch": 0.31073256127120924,
      "grad_norm": 0.778027148778391,
      "learning_rate": 9.416380942557719e-06,
      "loss": 0.37390871047973634,
      "num_input_tokens_seen": 237643416,
      "step": 4615,
      "train_runtime": 28482.9299,
      "train_tokens_per_second": 8343.363
    },
    {
      "epoch": 0.3110692162671694,
      "grad_norm": 0.6615680595434565,
      "learning_rate": 9.415140639140803e-06,
      "loss": 0.3932481288909912,
      "num_input_tokens_seen": 237881120,
      "step": 4620,
      "train_runtime": 28511.4971,
      "train_tokens_per_second": 8343.34
    },
    {
      "epoch": 0.3114058712631295,
      "grad_norm": 0.6805100599448501,
      "learning_rate": 9.413899101043887e-06,
      "loss": 0.42349720001220703,
      "num_input_tokens_seen": 238138024,
      "step": 4625,
      "train_runtime": 28545.2069,
      "train_tokens_per_second": 8342.487
    },
    {
      "epoch": 0.31174252625908966,
      "grad_norm": 0.8809058628005058,
      "learning_rate": 9.412656328614162e-06,
      "loss": 0.4077462673187256,
      "num_input_tokens_seen": 238363888,
      "step": 4630,
      "train_runtime": 28569.5388,
      "train_tokens_per_second": 8343.288
    },
    {
      "epoch": 0.3120791812550498,
      "grad_norm": 0.905842473597433,
      "learning_rate": 9.411412322199165e-06,
      "loss": 0.3960545063018799,
      "num_input_tokens_seen": 238646848,
      "step": 4635,
      "train_runtime": 28603.3238,
      "train_tokens_per_second": 8343.326
    },
    {
      "epoch": 0.31241583625100994,
      "grad_norm": 0.696567820301245,
      "learning_rate": 9.41016708214678e-06,
      "loss": 0.3858480453491211,
      "num_input_tokens_seen": 238868984,
      "step": 4640,
      "train_runtime": 28630.9371,
      "train_tokens_per_second": 8343.038
    },
    {
      "epoch": 0.31275249124697013,
      "grad_norm": 0.7810449749526278,
      "learning_rate": 9.408920608805233e-06,
      "loss": 0.3591245174407959,
      "num_input_tokens_seen": 239118136,
      "step": 4645,
      "train_runtime": 28659.6534,
      "train_tokens_per_second": 8343.372
    },
    {
      "epoch": 0.31308914624293027,
      "grad_norm": 0.6679093591315874,
      "learning_rate": 9.407672902523097e-06,
      "loss": 0.41380739212036133,
      "num_input_tokens_seen": 239378096,
      "step": 4650,
      "train_runtime": 28688.689,
      "train_tokens_per_second": 8343.989
    },
    {
      "epoch": 0.3134258012388904,
      "grad_norm": 0.8299777520921762,
      "learning_rate": 9.406423963649287e-06,
      "loss": 0.38919084072113036,
      "num_input_tokens_seen": 239600592,
      "step": 4655,
      "train_runtime": 28719.7062,
      "train_tokens_per_second": 8342.724
    },
    {
      "epoch": 0.31376245623485055,
      "grad_norm": 0.6425649985043917,
      "learning_rate": 9.405173792533069e-06,
      "loss": 0.3709535837173462,
      "num_input_tokens_seen": 239858392,
      "step": 4660,
      "train_runtime": 28750.5004,
      "train_tokens_per_second": 8342.755
    },
    {
      "epoch": 0.3140991112308107,
      "grad_norm": 0.7447702831099754,
      "learning_rate": 9.403922389524045e-06,
      "loss": 0.3843052387237549,
      "num_input_tokens_seen": 240114680,
      "step": 4665,
      "train_runtime": 28780.1214,
      "train_tokens_per_second": 8343.074
    },
    {
      "epoch": 0.31443576622677083,
      "grad_norm": 0.6428288980253829,
      "learning_rate": 9.40266975497217e-06,
      "loss": 0.4082589626312256,
      "num_input_tokens_seen": 240377072,
      "step": 4670,
      "train_runtime": 28806.8183,
      "train_tokens_per_second": 8344.451
    },
    {
      "epoch": 0.31477242122273097,
      "grad_norm": 0.6781946523138537,
      "learning_rate": 9.401415889227736e-06,
      "loss": 0.3962148904800415,
      "num_input_tokens_seen": 240659032,
      "step": 4675,
      "train_runtime": 28841.052,
      "train_tokens_per_second": 8344.322
    },
    {
      "epoch": 0.3151090762186911,
      "grad_norm": 0.5833446436664899,
      "learning_rate": 9.400160792641385e-06,
      "loss": 0.40200071334838866,
      "num_input_tokens_seen": 240936360,
      "step": 4680,
      "train_runtime": 28871.4463,
      "train_tokens_per_second": 8345.143
    },
    {
      "epoch": 0.31544573121465125,
      "grad_norm": 0.6290602025706544,
      "learning_rate": 9.398904465564097e-06,
      "loss": 0.3811753749847412,
      "num_input_tokens_seen": 241197160,
      "step": 4685,
      "train_runtime": 28907.0755,
      "train_tokens_per_second": 8343.88
    },
    {
      "epoch": 0.3157823862106114,
      "grad_norm": 0.6876139985575127,
      "learning_rate": 9.397646908347203e-06,
      "loss": 0.3670145034790039,
      "num_input_tokens_seen": 241448856,
      "step": 4690,
      "train_runtime": 28937.8456,
      "train_tokens_per_second": 8343.705
    },
    {
      "epoch": 0.3161190412065715,
      "grad_norm": 0.7833412983494945,
      "learning_rate": 9.396388121342375e-06,
      "loss": 0.36520750522613527,
      "num_input_tokens_seen": 241715304,
      "step": 4695,
      "train_runtime": 28968.3374,
      "train_tokens_per_second": 8344.121
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.7267846168249992,
      "learning_rate": 9.395128104901628e-06,
      "loss": 0.371771240234375,
      "num_input_tokens_seen": 241968896,
      "step": 4700,
      "train_runtime": 28996.9292,
      "train_tokens_per_second": 8344.639
    },
    {
      "epoch": 0.3167923511984918,
      "grad_norm": 0.6386269526363199,
      "learning_rate": 9.393866859377321e-06,
      "loss": 0.4120338439941406,
      "num_input_tokens_seen": 242240264,
      "step": 4705,
      "train_runtime": 29030.1381,
      "train_tokens_per_second": 8344.441
    },
    {
      "epoch": 0.31712900619445195,
      "grad_norm": 0.5949915747998211,
      "learning_rate": 9.392604385122157e-06,
      "loss": 0.38928020000457764,
      "num_input_tokens_seen": 242530592,
      "step": 4710,
      "train_runtime": 29064.355,
      "train_tokens_per_second": 8344.606
    },
    {
      "epoch": 0.3174656611904121,
      "grad_norm": 0.8009564082580207,
      "learning_rate": 9.391340682489185e-06,
      "loss": 0.3815764427185059,
      "num_input_tokens_seen": 242774824,
      "step": 4715,
      "train_runtime": 29095.5261,
      "train_tokens_per_second": 8344.06
    },
    {
      "epoch": 0.3178023161863722,
      "grad_norm": 0.6643277255911512,
      "learning_rate": 9.390075751831794e-06,
      "loss": 0.3853254556655884,
      "num_input_tokens_seen": 243048336,
      "step": 4720,
      "train_runtime": 29128.2046,
      "train_tokens_per_second": 8344.089
    },
    {
      "epoch": 0.31813897118233236,
      "grad_norm": 0.7420636003096855,
      "learning_rate": 9.388809593503718e-06,
      "loss": 0.3952138900756836,
      "num_input_tokens_seen": 243322592,
      "step": 4725,
      "train_runtime": 29159.6461,
      "train_tokens_per_second": 8344.497
    },
    {
      "epoch": 0.3184756261782925,
      "grad_norm": 0.7391363150456457,
      "learning_rate": 9.387542207859034e-06,
      "loss": 0.4226287841796875,
      "num_input_tokens_seen": 243587896,
      "step": 4730,
      "train_runtime": 29195.103,
      "train_tokens_per_second": 8343.45
    },
    {
      "epoch": 0.31881228117425264,
      "grad_norm": 0.6174239784347786,
      "learning_rate": 9.386273595252161e-06,
      "loss": 0.38719840049743653,
      "num_input_tokens_seen": 243861712,
      "step": 4735,
      "train_runtime": 29231.761,
      "train_tokens_per_second": 8342.354
    },
    {
      "epoch": 0.3191489361702128,
      "grad_norm": 0.7984178567073159,
      "learning_rate": 9.385003756037865e-06,
      "loss": 0.40962018966674807,
      "num_input_tokens_seen": 244118664,
      "step": 4740,
      "train_runtime": 29264.2591,
      "train_tokens_per_second": 8341.871
    },
    {
      "epoch": 0.3194855911661729,
      "grad_norm": 0.6640727890127178,
      "learning_rate": 9.383732690571253e-06,
      "loss": 0.3974289417266846,
      "num_input_tokens_seen": 244419072,
      "step": 4745,
      "train_runtime": 29300.3592,
      "train_tokens_per_second": 8341.846
    },
    {
      "epoch": 0.31982224616213306,
      "grad_norm": 0.695273634592087,
      "learning_rate": 9.382460399207769e-06,
      "loss": 0.39008047580718996,
      "num_input_tokens_seen": 244691768,
      "step": 4750,
      "train_runtime": 29332.8814,
      "train_tokens_per_second": 8341.893
    },
    {
      "epoch": 0.3201589011580932,
      "grad_norm": 0.7166410430146035,
      "learning_rate": 9.381186882303212e-06,
      "loss": 0.3712315559387207,
      "num_input_tokens_seen": 244923336,
      "step": 4755,
      "train_runtime": 29360.4319,
      "train_tokens_per_second": 8341.953
    },
    {
      "epoch": 0.32049555615405334,
      "grad_norm": 0.725764480401337,
      "learning_rate": 9.379912140213713e-06,
      "loss": 0.3775763988494873,
      "num_input_tokens_seen": 245174136,
      "step": 4760,
      "train_runtime": 29394.4063,
      "train_tokens_per_second": 8340.843
    },
    {
      "epoch": 0.3208322111500135,
      "grad_norm": 0.6625923530300807,
      "learning_rate": 9.37863617329575e-06,
      "loss": 0.36340746879577634,
      "num_input_tokens_seen": 245404920,
      "step": 4765,
      "train_runtime": 29423.0863,
      "train_tokens_per_second": 8340.557
    },
    {
      "epoch": 0.3211688661459736,
      "grad_norm": 0.7329330464823938,
      "learning_rate": 9.377358981906145e-06,
      "loss": 0.37799692153930664,
      "num_input_tokens_seen": 245668920,
      "step": 4770,
      "train_runtime": 29454.7836,
      "train_tokens_per_second": 8340.544
    },
    {
      "epoch": 0.32150552114193376,
      "grad_norm": 0.6498700173871859,
      "learning_rate": 9.376080566402059e-06,
      "loss": 0.39759120941162107,
      "num_input_tokens_seen": 245938208,
      "step": 4775,
      "train_runtime": 29485.012,
      "train_tokens_per_second": 8341.126
    },
    {
      "epoch": 0.3218421761378939,
      "grad_norm": 0.5970716265147497,
      "learning_rate": 9.374800927140994e-06,
      "loss": 0.31545438766479494,
      "num_input_tokens_seen": 246218448,
      "step": 4780,
      "train_runtime": 29513.5042,
      "train_tokens_per_second": 8342.569
    },
    {
      "epoch": 0.32217883113385404,
      "grad_norm": 0.6952441959576537,
      "learning_rate": 9.373520064480804e-06,
      "loss": 0.3953546524047852,
      "num_input_tokens_seen": 246479848,
      "step": 4785,
      "train_runtime": 29545.2463,
      "train_tokens_per_second": 8342.454
    },
    {
      "epoch": 0.3225154861298142,
      "grad_norm": 0.6794077113050575,
      "learning_rate": 9.372237978779672e-06,
      "loss": 0.3922430992126465,
      "num_input_tokens_seen": 246756984,
      "step": 4790,
      "train_runtime": 29575.5784,
      "train_tokens_per_second": 8343.268
    },
    {
      "epoch": 0.3228521411257743,
      "grad_norm": 0.6821464733590188,
      "learning_rate": 9.37095467039613e-06,
      "loss": 0.38663859367370607,
      "num_input_tokens_seen": 246988072,
      "step": 4795,
      "train_runtime": 29604.2888,
      "train_tokens_per_second": 8342.983
    },
    {
      "epoch": 0.32318879612173446,
      "grad_norm": 0.7247671373762176,
      "learning_rate": 9.369670139689056e-06,
      "loss": 0.37569193840026854,
      "num_input_tokens_seen": 247234992,
      "step": 4800,
      "train_runtime": 29630.3289,
      "train_tokens_per_second": 8343.984
    },
    {
      "epoch": 0.3235254511176946,
      "grad_norm": 0.678816610486564,
      "learning_rate": 9.368384387017659e-06,
      "loss": 0.37594757080078123,
      "num_input_tokens_seen": 247466344,
      "step": 4805,
      "train_runtime": 29662.7407,
      "train_tokens_per_second": 8342.666
    },
    {
      "epoch": 0.32386210611365474,
      "grad_norm": 0.6634323764014873,
      "learning_rate": 9.367097412741497e-06,
      "loss": 0.37766077518463137,
      "num_input_tokens_seen": 247724176,
      "step": 4810,
      "train_runtime": 29694.6667,
      "train_tokens_per_second": 8342.379
    },
    {
      "epoch": 0.3241987611096149,
      "grad_norm": 0.7601577020945164,
      "learning_rate": 9.36580921722047e-06,
      "loss": 0.398902440071106,
      "num_input_tokens_seen": 247984416,
      "step": 4815,
      "train_runtime": 29725.4342,
      "train_tokens_per_second": 8342.499
    },
    {
      "epoch": 0.324535416105575,
      "grad_norm": 0.6515995385318409,
      "learning_rate": 9.364519800814818e-06,
      "loss": 0.36098508834838866,
      "num_input_tokens_seen": 248247816,
      "step": 4820,
      "train_runtime": 29760.2566,
      "train_tokens_per_second": 8341.589
    },
    {
      "epoch": 0.32487207110153515,
      "grad_norm": 0.6310360763538257,
      "learning_rate": 9.36322916388512e-06,
      "loss": 0.3763983488082886,
      "num_input_tokens_seen": 248521536,
      "step": 4825,
      "train_runtime": 29795.5669,
      "train_tokens_per_second": 8340.89
    },
    {
      "epoch": 0.3252087260974953,
      "grad_norm": 0.6604569314535613,
      "learning_rate": 9.3619373067923e-06,
      "loss": 0.35733652114868164,
      "num_input_tokens_seen": 248794328,
      "step": 4830,
      "train_runtime": 29826.7228,
      "train_tokens_per_second": 8341.323
    },
    {
      "epoch": 0.32554538109345543,
      "grad_norm": 0.7164876779639466,
      "learning_rate": 9.360644229897622e-06,
      "loss": 0.3922275543212891,
      "num_input_tokens_seen": 249036368,
      "step": 4835,
      "train_runtime": 29855.7131,
      "train_tokens_per_second": 8341.33
    },
    {
      "epoch": 0.32588203608941557,
      "grad_norm": 0.7662705433132204,
      "learning_rate": 9.35934993356269e-06,
      "loss": 0.39902236461639407,
      "num_input_tokens_seen": 249291672,
      "step": 4840,
      "train_runtime": 29887.2853,
      "train_tokens_per_second": 8341.061
    },
    {
      "epoch": 0.3262186910853757,
      "grad_norm": 0.754651232196062,
      "learning_rate": 9.35805441814945e-06,
      "loss": 0.36565456390380857,
      "num_input_tokens_seen": 249556144,
      "step": 4845,
      "train_runtime": 29919.0104,
      "train_tokens_per_second": 8341.056
    },
    {
      "epoch": 0.32655534608133585,
      "grad_norm": 0.6895244192497568,
      "learning_rate": 9.356757684020188e-06,
      "loss": 0.3988660335540771,
      "num_input_tokens_seen": 249790360,
      "step": 4850,
      "train_runtime": 29952.0464,
      "train_tokens_per_second": 8339.676
    },
    {
      "epoch": 0.326892001077296,
      "grad_norm": 0.6799094978182154,
      "learning_rate": 9.355459731537533e-06,
      "loss": 0.37767231464385986,
      "num_input_tokens_seen": 250043536,
      "step": 4855,
      "train_runtime": 29982.6679,
      "train_tokens_per_second": 8339.603
    },
    {
      "epoch": 0.32722865607325613,
      "grad_norm": 0.6522749110604426,
      "learning_rate": 9.354160561064451e-06,
      "loss": 0.40085625648498535,
      "num_input_tokens_seen": 250313648,
      "step": 4860,
      "train_runtime": 30012.7208,
      "train_tokens_per_second": 8340.252
    },
    {
      "epoch": 0.32756531106921627,
      "grad_norm": 0.6935196372662568,
      "learning_rate": 9.352860172964254e-06,
      "loss": 0.4024970054626465,
      "num_input_tokens_seen": 250584928,
      "step": 4865,
      "train_runtime": 30042.1402,
      "train_tokens_per_second": 8341.114
    },
    {
      "epoch": 0.3279019660651764,
      "grad_norm": 0.6415802724048942,
      "learning_rate": 9.35155856760059e-06,
      "loss": 0.3614009380340576,
      "num_input_tokens_seen": 250861528,
      "step": 4870,
      "train_runtime": 30071.1033,
      "train_tokens_per_second": 8342.279
    },
    {
      "epoch": 0.32823862106113655,
      "grad_norm": 0.6708285013754662,
      "learning_rate": 9.35025574533745e-06,
      "loss": 0.3933092594146729,
      "num_input_tokens_seen": 251129528,
      "step": 4875,
      "train_runtime": 30104.5721,
      "train_tokens_per_second": 8341.907
    },
    {
      "epoch": 0.3285752760570967,
      "grad_norm": 0.6757723336805131,
      "learning_rate": 9.34895170653916e-06,
      "loss": 0.3645960330963135,
      "num_input_tokens_seen": 251391512,
      "step": 4880,
      "train_runtime": 30133.9733,
      "train_tokens_per_second": 8342.462
    },
    {
      "epoch": 0.3289119310530568,
      "grad_norm": 0.6247567765301895,
      "learning_rate": 9.347646451570394e-06,
      "loss": 0.3791051864624023,
      "num_input_tokens_seen": 251663936,
      "step": 4885,
      "train_runtime": 30170.2381,
      "train_tokens_per_second": 8341.463
    },
    {
      "epoch": 0.32924858604901697,
      "grad_norm": 0.6429194970717276,
      "learning_rate": 9.346339980796162e-06,
      "loss": 0.4167685031890869,
      "num_input_tokens_seen": 251963920,
      "step": 4890,
      "train_runtime": 30197.2606,
      "train_tokens_per_second": 8343.933
    },
    {
      "epoch": 0.3295852410449771,
      "grad_norm": 0.6869624573685872,
      "learning_rate": 9.345032294581813e-06,
      "loss": 0.38817539215087893,
      "num_input_tokens_seen": 252230472,
      "step": 4895,
      "train_runtime": 30230.7072,
      "train_tokens_per_second": 8343.519
    },
    {
      "epoch": 0.32992189604093725,
      "grad_norm": 0.6251308099148689,
      "learning_rate": 9.343723393293038e-06,
      "loss": 0.3838017463684082,
      "num_input_tokens_seen": 252477232,
      "step": 4900,
      "train_runtime": 30262.7568,
      "train_tokens_per_second": 8342.837
    },
    {
      "epoch": 0.3302585510368974,
      "grad_norm": 0.7569443127406553,
      "learning_rate": 9.342413277295869e-06,
      "loss": 0.3855480670928955,
      "num_input_tokens_seen": 252743144,
      "step": 4905,
      "train_runtime": 30295.101,
      "train_tokens_per_second": 8342.707
    },
    {
      "epoch": 0.3305952060328575,
      "grad_norm": 0.6093258239806219,
      "learning_rate": 9.341101946956672e-06,
      "loss": 0.38565993309020996,
      "num_input_tokens_seen": 252998048,
      "step": 4910,
      "train_runtime": 30324.3664,
      "train_tokens_per_second": 8343.061
    },
    {
      "epoch": 0.33093186102881766,
      "grad_norm": 0.6708342211981654,
      "learning_rate": 9.33978940264216e-06,
      "loss": 0.39599270820617677,
      "num_input_tokens_seen": 253224280,
      "step": 4915,
      "train_runtime": 30353.2568,
      "train_tokens_per_second": 8342.574
    },
    {
      "epoch": 0.3312685160247778,
      "grad_norm": 0.8472093343379826,
      "learning_rate": 9.338475644719377e-06,
      "loss": 0.37974135875701903,
      "num_input_tokens_seen": 253500064,
      "step": 4920,
      "train_runtime": 30384.329,
      "train_tokens_per_second": 8343.119
    },
    {
      "epoch": 0.33160517102073794,
      "grad_norm": 0.6885510524822153,
      "learning_rate": 9.337160673555719e-06,
      "loss": 0.38587629795074463,
      "num_input_tokens_seen": 253773472,
      "step": 4925,
      "train_runtime": 30411.7434,
      "train_tokens_per_second": 8344.588
    },
    {
      "epoch": 0.3319418260166981,
      "grad_norm": 0.7880324301150814,
      "learning_rate": 9.335844489518905e-06,
      "loss": 0.38245816230773927,
      "num_input_tokens_seen": 254027448,
      "step": 4930,
      "train_runtime": 30436.8342,
      "train_tokens_per_second": 8346.054
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.7275671184141373,
      "learning_rate": 9.334527092977008e-06,
      "loss": 0.3899287462234497,
      "num_input_tokens_seen": 254254600,
      "step": 4935,
      "train_runtime": 30464.6041,
      "train_tokens_per_second": 8345.902
    },
    {
      "epoch": 0.33261513600861836,
      "grad_norm": 0.9201634141894535,
      "learning_rate": 9.33320848429843e-06,
      "loss": 0.37147979736328124,
      "num_input_tokens_seen": 254544432,
      "step": 4940,
      "train_runtime": 30496.5846,
      "train_tokens_per_second": 8346.654
    },
    {
      "epoch": 0.3329517910045785,
      "grad_norm": 0.746075594887803,
      "learning_rate": 9.331888663851917e-06,
      "loss": 0.40908017158508303,
      "num_input_tokens_seen": 254813872,
      "step": 4945,
      "train_runtime": 30530.793,
      "train_tokens_per_second": 8346.127
    },
    {
      "epoch": 0.33328844600053864,
      "grad_norm": 0.6070705100595527,
      "learning_rate": 9.330567632006552e-06,
      "loss": 0.3875648260116577,
      "num_input_tokens_seen": 255071744,
      "step": 4950,
      "train_runtime": 30562.6247,
      "train_tokens_per_second": 8345.872
    },
    {
      "epoch": 0.3336251009964988,
      "grad_norm": 0.6732209274689368,
      "learning_rate": 9.329245389131759e-06,
      "loss": 0.3916248083114624,
      "num_input_tokens_seen": 255334072,
      "step": 4955,
      "train_runtime": 30593.8773,
      "train_tokens_per_second": 8345.921
    },
    {
      "epoch": 0.3339617559924589,
      "grad_norm": 0.6569546585786775,
      "learning_rate": 9.327921935597298e-06,
      "loss": 0.3677666187286377,
      "num_input_tokens_seen": 255574152,
      "step": 4960,
      "train_runtime": 30623.9096,
      "train_tokens_per_second": 8345.576
    },
    {
      "epoch": 0.33429841098841906,
      "grad_norm": 0.8722167338534346,
      "learning_rate": 9.326597271773267e-06,
      "loss": 0.39566693305969236,
      "num_input_tokens_seen": 255795728,
      "step": 4965,
      "train_runtime": 30654.4264,
      "train_tokens_per_second": 8344.496
    },
    {
      "epoch": 0.3346350659843792,
      "grad_norm": 0.7054131549918609,
      "learning_rate": 9.325271398030107e-06,
      "loss": 0.394504714012146,
      "num_input_tokens_seen": 256046536,
      "step": 4970,
      "train_runtime": 30682.8246,
      "train_tokens_per_second": 8344.947
    },
    {
      "epoch": 0.33497172098033934,
      "grad_norm": 0.7693720615485758,
      "learning_rate": 9.323944314738591e-06,
      "loss": 0.3982818603515625,
      "num_input_tokens_seen": 256301440,
      "step": 4975,
      "train_runtime": 30716.8067,
      "train_tokens_per_second": 8344.013
    },
    {
      "epoch": 0.3353083759762995,
      "grad_norm": 0.6655865396161899,
      "learning_rate": 9.322616022269838e-06,
      "loss": 0.37552011013031006,
      "num_input_tokens_seen": 256540744,
      "step": 4980,
      "train_runtime": 30741.2605,
      "train_tokens_per_second": 8345.16
    },
    {
      "epoch": 0.3356450309722596,
      "grad_norm": 0.874587408803514,
      "learning_rate": 9.321286520995297e-06,
      "loss": 0.3915414810180664,
      "num_input_tokens_seen": 256788376,
      "step": 4985,
      "train_runtime": 30774.1912,
      "train_tokens_per_second": 8344.277
    },
    {
      "epoch": 0.33598168596821976,
      "grad_norm": 0.7001930686721994,
      "learning_rate": 9.319955811286758e-06,
      "loss": 0.3776984930038452,
      "num_input_tokens_seen": 257036840,
      "step": 4990,
      "train_runtime": 30804.1063,
      "train_tokens_per_second": 8344.239
    },
    {
      "epoch": 0.3363183409641799,
      "grad_norm": 0.6339961947313105,
      "learning_rate": 9.318623893516354e-06,
      "loss": 0.38117218017578125,
      "num_input_tokens_seen": 257276520,
      "step": 4995,
      "train_runtime": 30833.9009,
      "train_tokens_per_second": 8343.95
    },
    {
      "epoch": 0.33665499596014004,
      "grad_norm": 0.6977093973729822,
      "learning_rate": 9.317290768056548e-06,
      "loss": 0.38987605571746825,
      "num_input_tokens_seen": 257511696,
      "step": 5000,
      "train_runtime": 30868.3169,
      "train_tokens_per_second": 8342.266
    },
    {
      "epoch": 0.3369916509561002,
      "grad_norm": 0.7021580634145813,
      "learning_rate": 9.315956435280147e-06,
      "loss": 0.39318969249725344,
      "num_input_tokens_seen": 257742400,
      "step": 5005,
      "train_runtime": 30902.9388,
      "train_tokens_per_second": 8340.385
    },
    {
      "epoch": 0.3373283059520603,
      "grad_norm": 0.6644578865950429,
      "learning_rate": 9.314620895560292e-06,
      "loss": 0.37369112968444823,
      "num_input_tokens_seen": 258006688,
      "step": 5010,
      "train_runtime": 30932.6925,
      "train_tokens_per_second": 8340.906
    },
    {
      "epoch": 0.33766496094802045,
      "grad_norm": 0.7828575706292819,
      "learning_rate": 9.313284149270459e-06,
      "loss": 0.41199846267700196,
      "num_input_tokens_seen": 258257176,
      "step": 5015,
      "train_runtime": 30964.1385,
      "train_tokens_per_second": 8340.525
    },
    {
      "epoch": 0.3380016159439806,
      "grad_norm": 0.7095301330098833,
      "learning_rate": 9.311946196784469e-06,
      "loss": 0.38463726043701174,
      "num_input_tokens_seen": 258509568,
      "step": 5020,
      "train_runtime": 30993.9482,
      "train_tokens_per_second": 8340.647
    },
    {
      "epoch": 0.33833827093994073,
      "grad_norm": 0.7132941416403792,
      "learning_rate": 9.310607038476476e-06,
      "loss": 0.3948333740234375,
      "num_input_tokens_seen": 258750848,
      "step": 5025,
      "train_runtime": 31023.4697,
      "train_tokens_per_second": 8340.487
    },
    {
      "epoch": 0.3386749259359009,
      "grad_norm": 0.707412430665757,
      "learning_rate": 9.30926667472097e-06,
      "loss": 0.3799560308456421,
      "num_input_tokens_seen": 259003232,
      "step": 5030,
      "train_runtime": 31053.7628,
      "train_tokens_per_second": 8340.478
    },
    {
      "epoch": 0.339011580931861,
      "grad_norm": 0.6987268133322435,
      "learning_rate": 9.307925105892779e-06,
      "loss": 0.4023834228515625,
      "num_input_tokens_seen": 259280960,
      "step": 5035,
      "train_runtime": 31085.549,
      "train_tokens_per_second": 8340.884
    },
    {
      "epoch": 0.33934823592782115,
      "grad_norm": 0.7522632906334726,
      "learning_rate": 9.30658233236707e-06,
      "loss": 0.36839320659637453,
      "num_input_tokens_seen": 259558272,
      "step": 5040,
      "train_runtime": 31116.6594,
      "train_tokens_per_second": 8341.457
    },
    {
      "epoch": 0.3396848909237813,
      "grad_norm": 0.7762470559434034,
      "learning_rate": 9.30523835451934e-06,
      "loss": 0.3748141050338745,
      "num_input_tokens_seen": 259811144,
      "step": 5045,
      "train_runtime": 31146.7856,
      "train_tokens_per_second": 8341.507
    },
    {
      "epoch": 0.34002154591974143,
      "grad_norm": 0.6806977871152003,
      "learning_rate": 9.303893172725437e-06,
      "loss": 0.4100802421569824,
      "num_input_tokens_seen": 260095000,
      "step": 5050,
      "train_runtime": 31181.035,
      "train_tokens_per_second": 8341.449
    },
    {
      "epoch": 0.34035820091570157,
      "grad_norm": 0.6448025918044382,
      "learning_rate": 9.302546787361533e-06,
      "loss": 0.38289475440979004,
      "num_input_tokens_seen": 260364424,
      "step": 5055,
      "train_runtime": 31209.6466,
      "train_tokens_per_second": 8342.434
    },
    {
      "epoch": 0.3406948559116617,
      "grad_norm": 0.6729938636180878,
      "learning_rate": 9.301199198804139e-06,
      "loss": 0.38710482120513917,
      "num_input_tokens_seen": 260617408,
      "step": 5060,
      "train_runtime": 31238.3962,
      "train_tokens_per_second": 8342.855
    },
    {
      "epoch": 0.34103151090762185,
      "grad_norm": 0.6978778327863823,
      "learning_rate": 9.299850407430102e-06,
      "loss": 0.4132237434387207,
      "num_input_tokens_seen": 260881632,
      "step": 5065,
      "train_runtime": 31265.7408,
      "train_tokens_per_second": 8344.009
    },
    {
      "epoch": 0.341368165903582,
      "grad_norm": 0.7664892016682658,
      "learning_rate": 9.298500413616613e-06,
      "loss": 0.3847651481628418,
      "num_input_tokens_seen": 261145328,
      "step": 5070,
      "train_runtime": 31297.559,
      "train_tokens_per_second": 8343.952
    },
    {
      "epoch": 0.3417048208995421,
      "grad_norm": 0.6975896316618444,
      "learning_rate": 9.297149217741188e-06,
      "loss": 0.3649940252304077,
      "num_input_tokens_seen": 261408856,
      "step": 5075,
      "train_runtime": 31332.8446,
      "train_tokens_per_second": 8342.966
    },
    {
      "epoch": 0.34204147589550227,
      "grad_norm": 0.6976851805977479,
      "learning_rate": 9.295796820181688e-06,
      "loss": 0.4122201919555664,
      "num_input_tokens_seen": 261674696,
      "step": 5080,
      "train_runtime": 31365.8878,
      "train_tokens_per_second": 8342.652
    },
    {
      "epoch": 0.3423781308914624,
      "grad_norm": 0.7404302123781178,
      "learning_rate": 9.294443221316305e-06,
      "loss": 0.38733491897583006,
      "num_input_tokens_seen": 261914096,
      "step": 5085,
      "train_runtime": 31396.0899,
      "train_tokens_per_second": 8342.252
    },
    {
      "epoch": 0.34271478588742255,
      "grad_norm": 0.6717996411875251,
      "learning_rate": 9.293088421523569e-06,
      "loss": 0.41142735481262205,
      "num_input_tokens_seen": 262158736,
      "step": 5090,
      "train_runtime": 31426.8358,
      "train_tokens_per_second": 8341.875
    },
    {
      "epoch": 0.3430514408833827,
      "grad_norm": 0.785695151364784,
      "learning_rate": 9.291732421182346e-06,
      "loss": 0.37002854347229003,
      "num_input_tokens_seen": 262395176,
      "step": 5095,
      "train_runtime": 31455.0455,
      "train_tokens_per_second": 8341.911
    },
    {
      "epoch": 0.3433880958793428,
      "grad_norm": 0.7304521415635425,
      "learning_rate": 9.290375220671837e-06,
      "loss": 0.38475751876831055,
      "num_input_tokens_seen": 262646952,
      "step": 5100,
      "train_runtime": 31484.5003,
      "train_tokens_per_second": 8342.103
    },
    {
      "epoch": 0.34372475087530296,
      "grad_norm": 0.7037378987028152,
      "learning_rate": 9.28901682037158e-06,
      "loss": 0.350843620300293,
      "num_input_tokens_seen": 262908968,
      "step": 5105,
      "train_runtime": 31518.2521,
      "train_tokens_per_second": 8341.483
    },
    {
      "epoch": 0.3440614058712631,
      "grad_norm": 0.6352479821961141,
      "learning_rate": 9.287657220661442e-06,
      "loss": 0.3983508586883545,
      "num_input_tokens_seen": 263168776,
      "step": 5110,
      "train_runtime": 31551.3654,
      "train_tokens_per_second": 8340.963
    },
    {
      "epoch": 0.34439806086722324,
      "grad_norm": 0.8540504231100057,
      "learning_rate": 9.286296421921636e-06,
      "loss": 0.3633908987045288,
      "num_input_tokens_seen": 263422264,
      "step": 5115,
      "train_runtime": 31582.3151,
      "train_tokens_per_second": 8340.816
    },
    {
      "epoch": 0.3447347158631834,
      "grad_norm": 0.8739816719965146,
      "learning_rate": 9.284934424532706e-06,
      "loss": 0.4369649887084961,
      "num_input_tokens_seen": 263655216,
      "step": 5120,
      "train_runtime": 31609.3271,
      "train_tokens_per_second": 8341.058
    },
    {
      "epoch": 0.3450713708591435,
      "grad_norm": 0.6016068722983827,
      "learning_rate": 9.283571228875525e-06,
      "loss": 0.34830656051635744,
      "num_input_tokens_seen": 263889768,
      "step": 5125,
      "train_runtime": 31641.4236,
      "train_tokens_per_second": 8340.009
    },
    {
      "epoch": 0.3454080258551037,
      "grad_norm": 0.6380146558077968,
      "learning_rate": 9.282206835331311e-06,
      "loss": 0.4099086284637451,
      "num_input_tokens_seen": 264163032,
      "step": 5130,
      "train_runtime": 31676.4906,
      "train_tokens_per_second": 8339.403
    },
    {
      "epoch": 0.34574468085106386,
      "grad_norm": 0.644720556310869,
      "learning_rate": 9.28084124428161e-06,
      "loss": 0.35596251487731934,
      "num_input_tokens_seen": 264410424,
      "step": 5135,
      "train_runtime": 31707.3273,
      "train_tokens_per_second": 8339.095
    },
    {
      "epoch": 0.346081335847024,
      "grad_norm": 0.6786726172430774,
      "learning_rate": 9.279474456108305e-06,
      "loss": 0.37679691314697267,
      "num_input_tokens_seen": 264697944,
      "step": 5140,
      "train_runtime": 31740.4097,
      "train_tokens_per_second": 8339.462
    },
    {
      "epoch": 0.34641799084298414,
      "grad_norm": 0.7317242601476821,
      "learning_rate": 9.278106471193615e-06,
      "loss": 0.3602623701095581,
      "num_input_tokens_seen": 264952800,
      "step": 5145,
      "train_runtime": 31768.3249,
      "train_tokens_per_second": 8340.156
    },
    {
      "epoch": 0.3467546458389443,
      "grad_norm": 0.7334034750053007,
      "learning_rate": 9.276737289920093e-06,
      "loss": 0.40178818702697755,
      "num_input_tokens_seen": 265207528,
      "step": 5150,
      "train_runtime": 31797.3457,
      "train_tokens_per_second": 8340.556
    },
    {
      "epoch": 0.3470913008349044,
      "grad_norm": 0.685065582093836,
      "learning_rate": 9.275366912670624e-06,
      "loss": 0.38415164947509767,
      "num_input_tokens_seen": 265460064,
      "step": 5155,
      "train_runtime": 31829.1274,
      "train_tokens_per_second": 8340.162
    },
    {
      "epoch": 0.34742795583086455,
      "grad_norm": 0.6869559162387655,
      "learning_rate": 9.273995339828432e-06,
      "loss": 0.3770427227020264,
      "num_input_tokens_seen": 265719888,
      "step": 5160,
      "train_runtime": 31860.8788,
      "train_tokens_per_second": 8340.005
    },
    {
      "epoch": 0.3477646108268247,
      "grad_norm": 0.9029558591542939,
      "learning_rate": 9.272622571777072e-06,
      "loss": 0.42333145141601564,
      "num_input_tokens_seen": 265983728,
      "step": 5165,
      "train_runtime": 31895.3161,
      "train_tokens_per_second": 8339.272
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.6566946188988227,
      "learning_rate": 9.271248608900434e-06,
      "loss": 0.3609964370727539,
      "num_input_tokens_seen": 266253736,
      "step": 5170,
      "train_runtime": 31926.1899,
      "train_tokens_per_second": 8339.665
    },
    {
      "epoch": 0.348437920818745,
      "grad_norm": 0.7689095427580899,
      "learning_rate": 9.269873451582741e-06,
      "loss": 0.39542255401611326,
      "num_input_tokens_seen": 266503616,
      "step": 5175,
      "train_runtime": 31956.7402,
      "train_tokens_per_second": 8339.512
    },
    {
      "epoch": 0.3487745758147051,
      "grad_norm": 0.7101987946799821,
      "learning_rate": 9.268497100208556e-06,
      "loss": 0.3946433782577515,
      "num_input_tokens_seen": 266744552,
      "step": 5180,
      "train_runtime": 31985.558,
      "train_tokens_per_second": 8339.531
    },
    {
      "epoch": 0.34911123081066525,
      "grad_norm": 0.7072367959647308,
      "learning_rate": 9.267119555162765e-06,
      "loss": 0.405084228515625,
      "num_input_tokens_seen": 267022160,
      "step": 5185,
      "train_runtime": 32017.3535,
      "train_tokens_per_second": 8339.92
    },
    {
      "epoch": 0.3494478858066254,
      "grad_norm": 0.6683854255194558,
      "learning_rate": 9.2657408168306e-06,
      "loss": 0.3779524564743042,
      "num_input_tokens_seen": 267285968,
      "step": 5190,
      "train_runtime": 32051.2288,
      "train_tokens_per_second": 8339.336
    },
    {
      "epoch": 0.34978454080258553,
      "grad_norm": 0.7011717171770939,
      "learning_rate": 9.264360885597617e-06,
      "loss": 0.37627859115600587,
      "num_input_tokens_seen": 267537408,
      "step": 5195,
      "train_runtime": 32079.3678,
      "train_tokens_per_second": 8339.859
    },
    {
      "epoch": 0.35012119579854567,
      "grad_norm": 0.7989305855863438,
      "learning_rate": 9.262979761849709e-06,
      "loss": 0.39826669692993166,
      "num_input_tokens_seen": 267794568,
      "step": 5200,
      "train_runtime": 32110.8733,
      "train_tokens_per_second": 8339.685
    },
    {
      "epoch": 0.3504578507945058,
      "grad_norm": 0.7241760428235579,
      "learning_rate": 9.261597445973106e-06,
      "loss": 0.3966125726699829,
      "num_input_tokens_seen": 268069296,
      "step": 5205,
      "train_runtime": 32139.3625,
      "train_tokens_per_second": 8340.84
    },
    {
      "epoch": 0.35079450579046595,
      "grad_norm": 0.7314207706935543,
      "learning_rate": 9.260213938354365e-06,
      "loss": 0.37574062347412107,
      "num_input_tokens_seen": 268312152,
      "step": 5210,
      "train_runtime": 32168.7823,
      "train_tokens_per_second": 8340.762
    },
    {
      "epoch": 0.3511311607864261,
      "grad_norm": 0.7184305342584864,
      "learning_rate": 9.25882923938038e-06,
      "loss": 0.39300949573516847,
      "num_input_tokens_seen": 268563216,
      "step": 5215,
      "train_runtime": 32196.7945,
      "train_tokens_per_second": 8341.303
    },
    {
      "epoch": 0.35146781578238623,
      "grad_norm": 0.7537921787040668,
      "learning_rate": 9.257443349438382e-06,
      "loss": 0.4221164703369141,
      "num_input_tokens_seen": 268835856,
      "step": 5220,
      "train_runtime": 32229.0244,
      "train_tokens_per_second": 8341.421
    },
    {
      "epoch": 0.35180447077834637,
      "grad_norm": 0.6870081236456057,
      "learning_rate": 9.256056268915924e-06,
      "loss": 0.38805551528930665,
      "num_input_tokens_seen": 269099048,
      "step": 5225,
      "train_runtime": 32259.3241,
      "train_tokens_per_second": 8341.745
    },
    {
      "epoch": 0.3521411257743065,
      "grad_norm": 0.6163234102610917,
      "learning_rate": 9.254667998200906e-06,
      "loss": 0.41520013809204104,
      "num_input_tokens_seen": 269356928,
      "step": 5230,
      "train_runtime": 32289.2202,
      "train_tokens_per_second": 8342.008
    },
    {
      "epoch": 0.35247778077026665,
      "grad_norm": 0.5671990357605662,
      "learning_rate": 9.253278537681548e-06,
      "loss": 0.40186185836791993,
      "num_input_tokens_seen": 269652400,
      "step": 5235,
      "train_runtime": 32317.5322,
      "train_tokens_per_second": 8343.843
    },
    {
      "epoch": 0.3528144357662268,
      "grad_norm": 0.7167682623635654,
      "learning_rate": 9.251887887746407e-06,
      "loss": 0.3919726133346558,
      "num_input_tokens_seen": 269917000,
      "step": 5240,
      "train_runtime": 32348.3646,
      "train_tokens_per_second": 8344.069
    },
    {
      "epoch": 0.3531510907621869,
      "grad_norm": 0.7357784274944127,
      "learning_rate": 9.25049604878438e-06,
      "loss": 0.3737499237060547,
      "num_input_tokens_seen": 270176704,
      "step": 5245,
      "train_runtime": 32378.4532,
      "train_tokens_per_second": 8344.336
    },
    {
      "epoch": 0.35348774575814707,
      "grad_norm": 0.7725304824441264,
      "learning_rate": 9.249103021184684e-06,
      "loss": 0.3933647871017456,
      "num_input_tokens_seen": 270421784,
      "step": 5250,
      "train_runtime": 32409.9683,
      "train_tokens_per_second": 8343.784
    },
    {
      "epoch": 0.3538244007541072,
      "grad_norm": 0.6810829158677065,
      "learning_rate": 9.24770880533688e-06,
      "loss": 0.4267380714416504,
      "num_input_tokens_seen": 270695752,
      "step": 5255,
      "train_runtime": 32441.7059,
      "train_tokens_per_second": 8344.067
    },
    {
      "epoch": 0.35416105575006734,
      "grad_norm": 1.0264330150940675,
      "learning_rate": 9.24631340163085e-06,
      "loss": 0.40450377464294435,
      "num_input_tokens_seen": 270906672,
      "step": 5260,
      "train_runtime": 32468.176,
      "train_tokens_per_second": 8343.76
    },
    {
      "epoch": 0.3544977107460275,
      "grad_norm": 0.6011496969750987,
      "learning_rate": 9.244916810456822e-06,
      "loss": 0.35047264099121095,
      "num_input_tokens_seen": 271185576,
      "step": 5265,
      "train_runtime": 32499.6479,
      "train_tokens_per_second": 8344.262
    },
    {
      "epoch": 0.3548343657419876,
      "grad_norm": 0.8252426411231506,
      "learning_rate": 9.24351903220534e-06,
      "loss": 0.39516458511352537,
      "num_input_tokens_seen": 271443472,
      "step": 5270,
      "train_runtime": 32531.5212,
      "train_tokens_per_second": 8344.014
    },
    {
      "epoch": 0.35517102073794776,
      "grad_norm": 0.7682525581134951,
      "learning_rate": 9.242120067267295e-06,
      "loss": 0.3620415925979614,
      "num_input_tokens_seen": 271702392,
      "step": 5275,
      "train_runtime": 32561.9976,
      "train_tokens_per_second": 8344.156
    },
    {
      "epoch": 0.3555076757339079,
      "grad_norm": 0.7281950904533832,
      "learning_rate": 9.2407199160339e-06,
      "loss": 0.38768293857574465,
      "num_input_tokens_seen": 271949424,
      "step": 5280,
      "train_runtime": 32594.8137,
      "train_tokens_per_second": 8343.334
    },
    {
      "epoch": 0.35584433072986804,
      "grad_norm": 0.6956573911373298,
      "learning_rate": 9.239318578896701e-06,
      "loss": 0.3809488296508789,
      "num_input_tokens_seen": 272201296,
      "step": 5285,
      "train_runtime": 32623.7703,
      "train_tokens_per_second": 8343.649
    },
    {
      "epoch": 0.3561809857258282,
      "grad_norm": 0.8236413692364504,
      "learning_rate": 9.237916056247583e-06,
      "loss": 0.39374918937683107,
      "num_input_tokens_seen": 272454520,
      "step": 5290,
      "train_runtime": 32654.9229,
      "train_tokens_per_second": 8343.444
    },
    {
      "epoch": 0.3565176407217883,
      "grad_norm": 0.7006445950116531,
      "learning_rate": 9.23651234847875e-06,
      "loss": 0.4058708190917969,
      "num_input_tokens_seen": 272708696,
      "step": 5295,
      "train_runtime": 32683.1981,
      "train_tokens_per_second": 8344.003
    },
    {
      "epoch": 0.35685429571774846,
      "grad_norm": 0.8041140366278605,
      "learning_rate": 9.23510745598275e-06,
      "loss": 0.38667347431182864,
      "num_input_tokens_seen": 272943544,
      "step": 5300,
      "train_runtime": 32714.071,
      "train_tokens_per_second": 8343.307
    },
    {
      "epoch": 0.3571909507137086,
      "grad_norm": 0.5522680237166616,
      "learning_rate": 9.233701379152456e-06,
      "loss": 0.37278237342834475,
      "num_input_tokens_seen": 273200064,
      "step": 5305,
      "train_runtime": 32747.2173,
      "train_tokens_per_second": 8342.696
    },
    {
      "epoch": 0.35752760570966874,
      "grad_norm": 0.7459077924316849,
      "learning_rate": 9.232294118381069e-06,
      "loss": 0.39784467220306396,
      "num_input_tokens_seen": 273466696,
      "step": 5310,
      "train_runtime": 32779.7188,
      "train_tokens_per_second": 8342.558
    },
    {
      "epoch": 0.3578642607056289,
      "grad_norm": 0.6489005876119589,
      "learning_rate": 9.230885674062129e-06,
      "loss": 0.3600412130355835,
      "num_input_tokens_seen": 273725768,
      "step": 5315,
      "train_runtime": 32810.2613,
      "train_tokens_per_second": 8342.688
    },
    {
      "epoch": 0.358200915701589,
      "grad_norm": 0.8279461644065005,
      "learning_rate": 9.2294760465895e-06,
      "loss": 0.4089359283447266,
      "num_input_tokens_seen": 273999816,
      "step": 5320,
      "train_runtime": 32841.2175,
      "train_tokens_per_second": 8343.169
    },
    {
      "epoch": 0.35853757069754916,
      "grad_norm": 0.7018599260418275,
      "learning_rate": 9.228065236357383e-06,
      "loss": 0.37081127166748046,
      "num_input_tokens_seen": 274273200,
      "step": 5325,
      "train_runtime": 32871.1957,
      "train_tokens_per_second": 8343.877
    },
    {
      "epoch": 0.3588742256935093,
      "grad_norm": 0.5649950046278993,
      "learning_rate": 9.226653243760304e-06,
      "loss": 0.3695035934448242,
      "num_input_tokens_seen": 274537064,
      "step": 5330,
      "train_runtime": 32903.878,
      "train_tokens_per_second": 8343.608
    },
    {
      "epoch": 0.35921088068946944,
      "grad_norm": 0.6413263935963638,
      "learning_rate": 9.225240069193124e-06,
      "loss": 0.4130889892578125,
      "num_input_tokens_seen": 274783136,
      "step": 5335,
      "train_runtime": 32933.2625,
      "train_tokens_per_second": 8343.635
    },
    {
      "epoch": 0.3595475356854296,
      "grad_norm": 0.7476701248348856,
      "learning_rate": 9.223825713051034e-06,
      "loss": 0.38680415153503417,
      "num_input_tokens_seen": 275040776,
      "step": 5340,
      "train_runtime": 32962.6317,
      "train_tokens_per_second": 8344.017
    },
    {
      "epoch": 0.3598841906813897,
      "grad_norm": 0.7262231170402146,
      "learning_rate": 9.222410175729549e-06,
      "loss": 0.38715109825134275,
      "num_input_tokens_seen": 275291848,
      "step": 5345,
      "train_runtime": 32994.3386,
      "train_tokens_per_second": 8343.609
    },
    {
      "epoch": 0.36022084567734985,
      "grad_norm": 0.6746617609986636,
      "learning_rate": 9.220993457624523e-06,
      "loss": 0.3688520908355713,
      "num_input_tokens_seen": 275531080,
      "step": 5350,
      "train_runtime": 33022.2448,
      "train_tokens_per_second": 8343.802
    },
    {
      "epoch": 0.36055750067331,
      "grad_norm": 0.7780945673585687,
      "learning_rate": 9.219575559132136e-06,
      "loss": 0.4017179012298584,
      "num_input_tokens_seen": 275791088,
      "step": 5355,
      "train_runtime": 33052.8049,
      "train_tokens_per_second": 8343.954
    },
    {
      "epoch": 0.36089415566927013,
      "grad_norm": 0.7308536344901853,
      "learning_rate": 9.2181564806489e-06,
      "loss": 0.3544175148010254,
      "num_input_tokens_seen": 276029688,
      "step": 5360,
      "train_runtime": 33083.3205,
      "train_tokens_per_second": 8343.47
    },
    {
      "epoch": 0.3612308106652303,
      "grad_norm": 0.5738375744598871,
      "learning_rate": 9.216736222571655e-06,
      "loss": 0.35495424270629883,
      "num_input_tokens_seen": 276292920,
      "step": 5365,
      "train_runtime": 33115.4605,
      "train_tokens_per_second": 8343.321
    },
    {
      "epoch": 0.3615674656611904,
      "grad_norm": 0.7036960404514623,
      "learning_rate": 9.215314785297572e-06,
      "loss": 0.42777557373046876,
      "num_input_tokens_seen": 276524904,
      "step": 5370,
      "train_runtime": 33141.45,
      "train_tokens_per_second": 8343.778
    },
    {
      "epoch": 0.36190412065715055,
      "grad_norm": 0.7418934241011746,
      "learning_rate": 9.21389216922415e-06,
      "loss": 0.3700996160507202,
      "num_input_tokens_seen": 276792168,
      "step": 5375,
      "train_runtime": 33172.174,
      "train_tokens_per_second": 8344.107
    },
    {
      "epoch": 0.3622407756531107,
      "grad_norm": 0.7200938238904515,
      "learning_rate": 9.212468374749222e-06,
      "loss": 0.42069168090820314,
      "num_input_tokens_seen": 277053872,
      "step": 5380,
      "train_runtime": 33205.1746,
      "train_tokens_per_second": 8343.696
    },
    {
      "epoch": 0.36257743064907083,
      "grad_norm": 0.6632122707434638,
      "learning_rate": 9.211043402270942e-06,
      "loss": 0.3830923795700073,
      "num_input_tokens_seen": 277306760,
      "step": 5385,
      "train_runtime": 33237.5108,
      "train_tokens_per_second": 8343.187
    },
    {
      "epoch": 0.36291408564503097,
      "grad_norm": 0.7108466025062012,
      "learning_rate": 9.209617252187806e-06,
      "loss": 0.37527804374694823,
      "num_input_tokens_seen": 277560352,
      "step": 5390,
      "train_runtime": 33268.4147,
      "train_tokens_per_second": 8343.059
    },
    {
      "epoch": 0.3632507406409911,
      "grad_norm": 0.8819895356118513,
      "learning_rate": 9.208189924898627e-06,
      "loss": 0.3487596035003662,
      "num_input_tokens_seen": 277795576,
      "step": 5395,
      "train_runtime": 33301.2375,
      "train_tokens_per_second": 8341.899
    },
    {
      "epoch": 0.36358739563695125,
      "grad_norm": 0.7767042516361655,
      "learning_rate": 9.206761420802554e-06,
      "loss": 0.4145012855529785,
      "num_input_tokens_seen": 278052728,
      "step": 5400,
      "train_runtime": 33333.1635,
      "train_tokens_per_second": 8341.624
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.6592251619657873,
      "learning_rate": 9.205331740299065e-06,
      "loss": 0.38005058765411376,
      "num_input_tokens_seen": 278287200,
      "step": 5405,
      "train_runtime": 33364.5451,
      "train_tokens_per_second": 8340.806
    },
    {
      "epoch": 0.36426070562887153,
      "grad_norm": 0.6704684206662174,
      "learning_rate": 9.203900883787967e-06,
      "loss": 0.41158604621887207,
      "num_input_tokens_seen": 278561656,
      "step": 5410,
      "train_runtime": 33400.4076,
      "train_tokens_per_second": 8340.068
    },
    {
      "epoch": 0.36459736062483167,
      "grad_norm": 0.7550658430076185,
      "learning_rate": 9.202468851669388e-06,
      "loss": 0.41437244415283203,
      "num_input_tokens_seen": 278807480,
      "step": 5415,
      "train_runtime": 33432.3764,
      "train_tokens_per_second": 8339.445
    },
    {
      "epoch": 0.3649340156207918,
      "grad_norm": 0.6382651373223271,
      "learning_rate": 9.201035644343797e-06,
      "loss": 0.3871955156326294,
      "num_input_tokens_seen": 279055192,
      "step": 5420,
      "train_runtime": 33463.8106,
      "train_tokens_per_second": 8339.014
    },
    {
      "epoch": 0.36527067061675195,
      "grad_norm": 0.7309882802254288,
      "learning_rate": 9.199601262211985e-06,
      "loss": 0.3819345712661743,
      "num_input_tokens_seen": 279288608,
      "step": 5425,
      "train_runtime": 33495.0628,
      "train_tokens_per_second": 8338.202
    },
    {
      "epoch": 0.3656073256127121,
      "grad_norm": 0.7012982576636724,
      "learning_rate": 9.19816570567507e-06,
      "loss": 0.3912306308746338,
      "num_input_tokens_seen": 279556696,
      "step": 5430,
      "train_runtime": 33526.4509,
      "train_tokens_per_second": 8338.392
    },
    {
      "epoch": 0.3659439806086722,
      "grad_norm": 0.7825466914732441,
      "learning_rate": 9.196728975134503e-06,
      "loss": 0.41402344703674315,
      "num_input_tokens_seen": 279792568,
      "step": 5435,
      "train_runtime": 33554.8766,
      "train_tokens_per_second": 8338.358
    },
    {
      "epoch": 0.36628063560463237,
      "grad_norm": 0.9155436426216613,
      "learning_rate": 9.195291070992061e-06,
      "loss": 0.38502039909362795,
      "num_input_tokens_seen": 280030704,
      "step": 5440,
      "train_runtime": 33584.0709,
      "train_tokens_per_second": 8338.2
    },
    {
      "epoch": 0.3666172906005925,
      "grad_norm": 0.6231808963834045,
      "learning_rate": 9.193851993649849e-06,
      "loss": 0.3621049404144287,
      "num_input_tokens_seen": 280293600,
      "step": 5445,
      "train_runtime": 33617.3917,
      "train_tokens_per_second": 8337.756
    },
    {
      "epoch": 0.36695394559655264,
      "grad_norm": 0.6576932627717301,
      "learning_rate": 9.1924117435103e-06,
      "loss": 0.3884142875671387,
      "num_input_tokens_seen": 280536128,
      "step": 5450,
      "train_runtime": 33647.1757,
      "train_tokens_per_second": 8337.583
    },
    {
      "epoch": 0.3672906005925128,
      "grad_norm": 0.7879700800000913,
      "learning_rate": 9.190970320976176e-06,
      "loss": 0.3699147701263428,
      "num_input_tokens_seen": 280799784,
      "step": 5455,
      "train_runtime": 33681.5816,
      "train_tokens_per_second": 8336.894
    },
    {
      "epoch": 0.3676272555884729,
      "grad_norm": 0.8409210207194304,
      "learning_rate": 9.189527726450565e-06,
      "loss": 0.4274420261383057,
      "num_input_tokens_seen": 281017120,
      "step": 5460,
      "train_runtime": 33711.5871,
      "train_tokens_per_second": 8335.921
    },
    {
      "epoch": 0.36796391058443306,
      "grad_norm": 0.6489655674813081,
      "learning_rate": 9.188083960336885e-06,
      "loss": 0.376585865020752,
      "num_input_tokens_seen": 281275224,
      "step": 5465,
      "train_runtime": 33745.6496,
      "train_tokens_per_second": 8335.155
    },
    {
      "epoch": 0.3683005655803932,
      "grad_norm": 0.7743051999310251,
      "learning_rate": 9.186639023038879e-06,
      "loss": 0.3980702877044678,
      "num_input_tokens_seen": 281540232,
      "step": 5470,
      "train_runtime": 33781.15,
      "train_tokens_per_second": 8334.241
    },
    {
      "epoch": 0.36863722057635334,
      "grad_norm": 0.7195438819586357,
      "learning_rate": 9.185192914960622e-06,
      "loss": 0.42609291076660155,
      "num_input_tokens_seen": 281795808,
      "step": 5475,
      "train_runtime": 33811.1631,
      "train_tokens_per_second": 8334.401
    },
    {
      "epoch": 0.3689738755723135,
      "grad_norm": 0.6125799153313906,
      "learning_rate": 9.183745636506511e-06,
      "loss": 0.3784531593322754,
      "num_input_tokens_seen": 282036288,
      "step": 5480,
      "train_runtime": 33842.6774,
      "train_tokens_per_second": 8333.746
    },
    {
      "epoch": 0.3693105305682736,
      "grad_norm": 0.5680666445664335,
      "learning_rate": 9.182297188081274e-06,
      "loss": 0.3766319990158081,
      "num_input_tokens_seen": 282321672,
      "step": 5485,
      "train_runtime": 33877.7555,
      "train_tokens_per_second": 8333.541
    },
    {
      "epoch": 0.36964718556423376,
      "grad_norm": 0.6847853877553535,
      "learning_rate": 9.180847570089964e-06,
      "loss": 0.3799124240875244,
      "num_input_tokens_seen": 282538608,
      "step": 5490,
      "train_runtime": 33905.2647,
      "train_tokens_per_second": 8333.178
    },
    {
      "epoch": 0.3699838405601939,
      "grad_norm": 0.6550570362488892,
      "learning_rate": 9.179396782937965e-06,
      "loss": 0.4135440349578857,
      "num_input_tokens_seen": 282824112,
      "step": 5495,
      "train_runtime": 33939.0758,
      "train_tokens_per_second": 8333.289
    },
    {
      "epoch": 0.37032049555615404,
      "grad_norm": 0.712124529592423,
      "learning_rate": 9.17794482703098e-06,
      "loss": 0.40084037780761717,
      "num_input_tokens_seen": 283082208,
      "step": 5500,
      "train_runtime": 33967.6574,
      "train_tokens_per_second": 8333.875
    },
    {
      "epoch": 0.3706571505521142,
      "grad_norm": 0.7675766456794723,
      "learning_rate": 9.176491702775049e-06,
      "loss": 0.3980029344558716,
      "num_input_tokens_seen": 283313304,
      "step": 5505,
      "train_runtime": 33995.8525,
      "train_tokens_per_second": 8333.761
    },
    {
      "epoch": 0.3709938055480743,
      "grad_norm": 0.6590811473219225,
      "learning_rate": 9.17503741057653e-06,
      "loss": 0.41051573753356935,
      "num_input_tokens_seen": 283577752,
      "step": 5510,
      "train_runtime": 34027.1053,
      "train_tokens_per_second": 8333.878
    },
    {
      "epoch": 0.37133046054403446,
      "grad_norm": 0.6842022900550432,
      "learning_rate": 9.173581950842111e-06,
      "loss": 0.3885434865951538,
      "num_input_tokens_seen": 283832376,
      "step": 5515,
      "train_runtime": 34058.509,
      "train_tokens_per_second": 8333.67
    },
    {
      "epoch": 0.3716671155399946,
      "grad_norm": 0.6903669695629125,
      "learning_rate": 9.172125323978811e-06,
      "loss": 0.38984169960021975,
      "num_input_tokens_seen": 284095192,
      "step": 5520,
      "train_runtime": 34088.2633,
      "train_tokens_per_second": 8334.106
    },
    {
      "epoch": 0.37200377053595474,
      "grad_norm": 0.6881989755065171,
      "learning_rate": 9.170667530393966e-06,
      "loss": 0.40131635665893556,
      "num_input_tokens_seen": 284374120,
      "step": 5525,
      "train_runtime": 34118.9981,
      "train_tokens_per_second": 8334.773
    },
    {
      "epoch": 0.3723404255319149,
      "grad_norm": 0.6574217624871604,
      "learning_rate": 9.169208570495247e-06,
      "loss": 0.35828752517700196,
      "num_input_tokens_seen": 284639160,
      "step": 5530,
      "train_runtime": 34149.2683,
      "train_tokens_per_second": 8335.147
    },
    {
      "epoch": 0.372677080527875,
      "grad_norm": 0.6993966707635625,
      "learning_rate": 9.167748444690646e-06,
      "loss": 0.38294119834899903,
      "num_input_tokens_seen": 284867624,
      "step": 5535,
      "train_runtime": 34176.5999,
      "train_tokens_per_second": 8335.166
    },
    {
      "epoch": 0.37301373552383515,
      "grad_norm": 0.797255225870366,
      "learning_rate": 9.16628715338848e-06,
      "loss": 0.38102545738220217,
      "num_input_tokens_seen": 285102592,
      "step": 5540,
      "train_runtime": 34205.0007,
      "train_tokens_per_second": 8335.114
    },
    {
      "epoch": 0.3733503905197953,
      "grad_norm": 0.6968420590979973,
      "learning_rate": 9.164824696997398e-06,
      "loss": 0.3673738479614258,
      "num_input_tokens_seen": 285346448,
      "step": 5545,
      "train_runtime": 34230.2306,
      "train_tokens_per_second": 8336.095
    },
    {
      "epoch": 0.37368704551575543,
      "grad_norm": 0.6887495271263051,
      "learning_rate": 9.163361075926371e-06,
      "loss": 0.4070948600769043,
      "num_input_tokens_seen": 285610816,
      "step": 5550,
      "train_runtime": 34260.2652,
      "train_tokens_per_second": 8336.503
    },
    {
      "epoch": 0.3740237005117156,
      "grad_norm": 0.6678833232767103,
      "learning_rate": 9.161896290584695e-06,
      "loss": 0.37053966522216797,
      "num_input_tokens_seen": 285869176,
      "step": 5555,
      "train_runtime": 34287.7596,
      "train_tokens_per_second": 8337.354
    },
    {
      "epoch": 0.3743603555076757,
      "grad_norm": 0.6782501852542312,
      "learning_rate": 9.160430341381991e-06,
      "loss": 0.3848405838012695,
      "num_input_tokens_seen": 286110512,
      "step": 5560,
      "train_runtime": 34316.2601,
      "train_tokens_per_second": 8337.462
    },
    {
      "epoch": 0.37469701050363585,
      "grad_norm": 0.7150761359520936,
      "learning_rate": 9.15896322872821e-06,
      "loss": 0.38939437866210935,
      "num_input_tokens_seen": 286372192,
      "step": 5565,
      "train_runtime": 34352.4327,
      "train_tokens_per_second": 8336.3
    },
    {
      "epoch": 0.375033665499596,
      "grad_norm": 0.7685837858213957,
      "learning_rate": 9.157494953033625e-06,
      "loss": 0.4125998497009277,
      "num_input_tokens_seen": 286650320,
      "step": 5570,
      "train_runtime": 34384.2365,
      "train_tokens_per_second": 8336.678
    },
    {
      "epoch": 0.37537032049555613,
      "grad_norm": 0.7276561749443938,
      "learning_rate": 9.156025514708832e-06,
      "loss": 0.38674585819244384,
      "num_input_tokens_seen": 286926176,
      "step": 5575,
      "train_runtime": 34417.0797,
      "train_tokens_per_second": 8336.738
    },
    {
      "epoch": 0.37570697549151627,
      "grad_norm": 0.6680066164441092,
      "learning_rate": 9.154554914164757e-06,
      "loss": 0.37763247489929197,
      "num_input_tokens_seen": 287210504,
      "step": 5580,
      "train_runtime": 34450.1435,
      "train_tokens_per_second": 8336.99
    },
    {
      "epoch": 0.3760436304874764,
      "grad_norm": 0.6396222613176532,
      "learning_rate": 9.153083151812649e-06,
      "loss": 0.37025539875030516,
      "num_input_tokens_seen": 287474944,
      "step": 5585,
      "train_runtime": 34481.1882,
      "train_tokens_per_second": 8337.153
    },
    {
      "epoch": 0.37638028548343655,
      "grad_norm": 0.7533857986966961,
      "learning_rate": 9.151610228064078e-06,
      "loss": 0.4086757183074951,
      "num_input_tokens_seen": 287733128,
      "step": 5590,
      "train_runtime": 34512.8387,
      "train_tokens_per_second": 8336.988
    },
    {
      "epoch": 0.3767169404793967,
      "grad_norm": 0.662408891900271,
      "learning_rate": 9.150136143330949e-06,
      "loss": 0.40886383056640624,
      "num_input_tokens_seen": 287987632,
      "step": 5595,
      "train_runtime": 34542.4609,
      "train_tokens_per_second": 8337.207
    },
    {
      "epoch": 0.37705359547535683,
      "grad_norm": 0.5965313412542876,
      "learning_rate": 9.14866089802548e-06,
      "loss": 0.3727555513381958,
      "num_input_tokens_seen": 288250400,
      "step": 5600,
      "train_runtime": 34577.4404,
      "train_tokens_per_second": 8336.372
    },
    {
      "epoch": 0.37739025047131697,
      "grad_norm": 0.7363199868057265,
      "learning_rate": 9.147184492560219e-06,
      "loss": 0.3568639039993286,
      "num_input_tokens_seen": 288494032,
      "step": 5605,
      "train_runtime": 34605.2872,
      "train_tokens_per_second": 8336.704
    },
    {
      "epoch": 0.3777269054672771,
      "grad_norm": 0.6922618753389043,
      "learning_rate": 9.14570692734804e-06,
      "loss": 0.3710559606552124,
      "num_input_tokens_seen": 288765056,
      "step": 5610,
      "train_runtime": 34634.3241,
      "train_tokens_per_second": 8337.54
    },
    {
      "epoch": 0.37806356046323725,
      "grad_norm": 0.6550346264120643,
      "learning_rate": 9.144228202802137e-06,
      "loss": 0.40007901191711426,
      "num_input_tokens_seen": 289039544,
      "step": 5615,
      "train_runtime": 34665.3449,
      "train_tokens_per_second": 8337.997
    },
    {
      "epoch": 0.37840021545919744,
      "grad_norm": 0.6883960296439051,
      "learning_rate": 9.142748319336033e-06,
      "loss": 0.38826305866241456,
      "num_input_tokens_seen": 289296664,
      "step": 5620,
      "train_runtime": 34695.7927,
      "train_tokens_per_second": 8338.091
    },
    {
      "epoch": 0.3787368704551576,
      "grad_norm": 0.6965370042794611,
      "learning_rate": 9.141267277363573e-06,
      "loss": 0.3761886119842529,
      "num_input_tokens_seen": 289524224,
      "step": 5625,
      "train_runtime": 34724.1965,
      "train_tokens_per_second": 8337.824
    },
    {
      "epoch": 0.3790735254511177,
      "grad_norm": 0.7191557520840252,
      "learning_rate": 9.139785077298921e-06,
      "loss": 0.36728496551513673,
      "num_input_tokens_seen": 289786552,
      "step": 5630,
      "train_runtime": 34757.1039,
      "train_tokens_per_second": 8337.477
    },
    {
      "epoch": 0.37941018044707786,
      "grad_norm": 0.6546433273291378,
      "learning_rate": 9.138301719556577e-06,
      "loss": 0.37585928440093996,
      "num_input_tokens_seen": 290039424,
      "step": 5635,
      "train_runtime": 34787.974,
      "train_tokens_per_second": 8337.347
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.6295725682449458,
      "learning_rate": 9.136817204551352e-06,
      "loss": 0.39262502193450927,
      "num_input_tokens_seen": 290311912,
      "step": 5640,
      "train_runtime": 34817.5597,
      "train_tokens_per_second": 8338.089
    },
    {
      "epoch": 0.38008349043899814,
      "grad_norm": 0.7074367724174596,
      "learning_rate": 9.135331532698385e-06,
      "loss": 0.3833719491958618,
      "num_input_tokens_seen": 290555968,
      "step": 5645,
      "train_runtime": 34846.9301,
      "train_tokens_per_second": 8338.065
    },
    {
      "epoch": 0.3804201454349583,
      "grad_norm": 0.7084623316587687,
      "learning_rate": 9.133844704413144e-06,
      "loss": 0.3692053556442261,
      "num_input_tokens_seen": 290802312,
      "step": 5650,
      "train_runtime": 34879.1963,
      "train_tokens_per_second": 8337.414
    },
    {
      "epoch": 0.3807568004309184,
      "grad_norm": 0.7648579831326876,
      "learning_rate": 9.132356720111413e-06,
      "loss": 0.38519768714904784,
      "num_input_tokens_seen": 291051584,
      "step": 5655,
      "train_runtime": 34910.3799,
      "train_tokens_per_second": 8337.107
    },
    {
      "epoch": 0.38109345542687856,
      "grad_norm": 0.7513410600664612,
      "learning_rate": 9.1308675802093e-06,
      "loss": 0.3839890480041504,
      "num_input_tokens_seen": 291303840,
      "step": 5660,
      "train_runtime": 34940.6325,
      "train_tokens_per_second": 8337.108
    },
    {
      "epoch": 0.3814301104228387,
      "grad_norm": 0.6544498489124774,
      "learning_rate": 9.129377285123241e-06,
      "loss": 0.38111891746521,
      "num_input_tokens_seen": 291553856,
      "step": 5665,
      "train_runtime": 34970.6423,
      "train_tokens_per_second": 8337.103
    },
    {
      "epoch": 0.38176676541879884,
      "grad_norm": 0.7297604800501273,
      "learning_rate": 9.127885835269996e-06,
      "loss": 0.40241031646728515,
      "num_input_tokens_seen": 291825400,
      "step": 5670,
      "train_runtime": 35001.4238,
      "train_tokens_per_second": 8337.529
    },
    {
      "epoch": 0.382103420414759,
      "grad_norm": 0.6472342357384739,
      "learning_rate": 9.126393231066636e-06,
      "loss": 0.3789690017700195,
      "num_input_tokens_seen": 292069296,
      "step": 5675,
      "train_runtime": 35032.9786,
      "train_tokens_per_second": 8336.982
    },
    {
      "epoch": 0.3824400754107191,
      "grad_norm": 0.7125000104110083,
      "learning_rate": 9.124899472930568e-06,
      "loss": 0.4007974624633789,
      "num_input_tokens_seen": 292330392,
      "step": 5680,
      "train_runtime": 35060.9801,
      "train_tokens_per_second": 8337.77
    },
    {
      "epoch": 0.38277673040667926,
      "grad_norm": 0.7222046538689625,
      "learning_rate": 9.123404561279517e-06,
      "loss": 0.4040034294128418,
      "num_input_tokens_seen": 292563248,
      "step": 5685,
      "train_runtime": 35087.1923,
      "train_tokens_per_second": 8338.178
    },
    {
      "epoch": 0.3831133854026394,
      "grad_norm": 0.6829701117533266,
      "learning_rate": 9.121908496531527e-06,
      "loss": 0.3856212615966797,
      "num_input_tokens_seen": 292839944,
      "step": 5690,
      "train_runtime": 35117.1426,
      "train_tokens_per_second": 8338.946
    },
    {
      "epoch": 0.38345004039859953,
      "grad_norm": 0.7435484970178073,
      "learning_rate": 9.120411279104971e-06,
      "loss": 0.38770408630371095,
      "num_input_tokens_seen": 293095056,
      "step": 5695,
      "train_runtime": 35143.682,
      "train_tokens_per_second": 8339.907
    },
    {
      "epoch": 0.3837866953945597,
      "grad_norm": 0.6819452531040607,
      "learning_rate": 9.118912909418538e-06,
      "loss": 0.37919111251831056,
      "num_input_tokens_seen": 293349848,
      "step": 5700,
      "train_runtime": 35173.8115,
      "train_tokens_per_second": 8340.007
    },
    {
      "epoch": 0.3841233503905198,
      "grad_norm": 0.7045901557344795,
      "learning_rate": 9.117413387891246e-06,
      "loss": 0.4140965938568115,
      "num_input_tokens_seen": 293608464,
      "step": 5705,
      "train_runtime": 35203.7243,
      "train_tokens_per_second": 8340.267
    },
    {
      "epoch": 0.38446000538647995,
      "grad_norm": 0.666552135255612,
      "learning_rate": 9.11591271494243e-06,
      "loss": 0.3777653217315674,
      "num_input_tokens_seen": 293864040,
      "step": 5710,
      "train_runtime": 35233.975,
      "train_tokens_per_second": 8340.36
    },
    {
      "epoch": 0.3847966603824401,
      "grad_norm": 0.6935350991169743,
      "learning_rate": 9.114410890991748e-06,
      "loss": 0.39169037342071533,
      "num_input_tokens_seen": 294114624,
      "step": 5715,
      "train_runtime": 35261.9011,
      "train_tokens_per_second": 8340.861
    },
    {
      "epoch": 0.38513331537840023,
      "grad_norm": 0.7348516414781409,
      "learning_rate": 9.112907916459177e-06,
      "loss": 0.37096438407897947,
      "num_input_tokens_seen": 294326000,
      "step": 5720,
      "train_runtime": 35290.5579,
      "train_tokens_per_second": 8340.078
    },
    {
      "epoch": 0.38546997037436037,
      "grad_norm": 0.673704559164417,
      "learning_rate": 9.111403791765025e-06,
      "loss": 0.37876081466674805,
      "num_input_tokens_seen": 294596536,
      "step": 5725,
      "train_runtime": 35323.0372,
      "train_tokens_per_second": 8340.068
    },
    {
      "epoch": 0.3858066253703205,
      "grad_norm": 0.6431595925319997,
      "learning_rate": 9.109898517329914e-06,
      "loss": 0.3627560377120972,
      "num_input_tokens_seen": 294881240,
      "step": 5730,
      "train_runtime": 35355.2933,
      "train_tokens_per_second": 8340.512
    },
    {
      "epoch": 0.38614328036628065,
      "grad_norm": 0.6395752149501135,
      "learning_rate": 9.108392093574785e-06,
      "loss": 0.3859114646911621,
      "num_input_tokens_seen": 295144808,
      "step": 5735,
      "train_runtime": 35382.7093,
      "train_tokens_per_second": 8341.498
    },
    {
      "epoch": 0.3864799353622408,
      "grad_norm": 0.7159553210928845,
      "learning_rate": 9.10688452092091e-06,
      "loss": 0.3605621576309204,
      "num_input_tokens_seen": 295403576,
      "step": 5740,
      "train_runtime": 35416.1794,
      "train_tokens_per_second": 8340.922
    },
    {
      "epoch": 0.38681659035820093,
      "grad_norm": 0.8603626035208404,
      "learning_rate": 9.105375799789874e-06,
      "loss": 0.387003755569458,
      "num_input_tokens_seen": 295645520,
      "step": 5745,
      "train_runtime": 35448.624,
      "train_tokens_per_second": 8340.113
    },
    {
      "epoch": 0.38715324535416107,
      "grad_norm": 0.7238281596360825,
      "learning_rate": 9.103865930603586e-06,
      "loss": 0.3665450572967529,
      "num_input_tokens_seen": 295876304,
      "step": 5750,
      "train_runtime": 35476.2839,
      "train_tokens_per_second": 8340.115
    },
    {
      "epoch": 0.3874899003501212,
      "grad_norm": 0.695952817062049,
      "learning_rate": 9.102354913784279e-06,
      "loss": 0.38305020332336426,
      "num_input_tokens_seen": 296132264,
      "step": 5755,
      "train_runtime": 35508.691,
      "train_tokens_per_second": 8339.712
    },
    {
      "epoch": 0.38782655534608135,
      "grad_norm": 0.6931679026062046,
      "learning_rate": 9.100842749754499e-06,
      "loss": 0.35318965911865235,
      "num_input_tokens_seen": 296396496,
      "step": 5760,
      "train_runtime": 35537.2373,
      "train_tokens_per_second": 8340.448
    },
    {
      "epoch": 0.3881632103420415,
      "grad_norm": 0.7099273326275661,
      "learning_rate": 9.099329438937122e-06,
      "loss": 0.40328054428100585,
      "num_input_tokens_seen": 296639472,
      "step": 5765,
      "train_runtime": 35571.1889,
      "train_tokens_per_second": 8339.318
    },
    {
      "epoch": 0.3884998653380016,
      "grad_norm": 0.5870152153666054,
      "learning_rate": 9.09781498175534e-06,
      "loss": 0.34922189712524415,
      "num_input_tokens_seen": 296891648,
      "step": 5770,
      "train_runtime": 35604.4028,
      "train_tokens_per_second": 8338.622
    },
    {
      "epoch": 0.38883652033396177,
      "grad_norm": 0.5994892843506954,
      "learning_rate": 9.096299378632666e-06,
      "loss": 0.3705258369445801,
      "num_input_tokens_seen": 297159392,
      "step": 5775,
      "train_runtime": 35635.1453,
      "train_tokens_per_second": 8338.941
    },
    {
      "epoch": 0.3891731753299219,
      "grad_norm": 0.6475668524884396,
      "learning_rate": 9.09478262999293e-06,
      "loss": 0.40535435676574705,
      "num_input_tokens_seen": 297440232,
      "step": 5780,
      "train_runtime": 35666.6168,
      "train_tokens_per_second": 8339.457
    },
    {
      "epoch": 0.38950983032588204,
      "grad_norm": 0.7750854253195375,
      "learning_rate": 9.09326473626029e-06,
      "loss": 0.3818108558654785,
      "num_input_tokens_seen": 297690672,
      "step": 5785,
      "train_runtime": 35694.1995,
      "train_tokens_per_second": 8340.029
    },
    {
      "epoch": 0.3898464853218422,
      "grad_norm": 0.6469968275432019,
      "learning_rate": 9.091745697859218e-06,
      "loss": 0.40662569999694825,
      "num_input_tokens_seen": 297965192,
      "step": 5790,
      "train_runtime": 35728.0222,
      "train_tokens_per_second": 8339.818
    },
    {
      "epoch": 0.3901831403178023,
      "grad_norm": 0.6857514306587776,
      "learning_rate": 9.090225515214512e-06,
      "loss": 0.40064148902893065,
      "num_input_tokens_seen": 298209272,
      "step": 5795,
      "train_runtime": 35756.6126,
      "train_tokens_per_second": 8339.975
    },
    {
      "epoch": 0.39051979531376246,
      "grad_norm": 0.8683966365512411,
      "learning_rate": 9.088704188751281e-06,
      "loss": 0.4058115482330322,
      "num_input_tokens_seen": 298453840,
      "step": 5800,
      "train_runtime": 35788.1906,
      "train_tokens_per_second": 8339.45
    },
    {
      "epoch": 0.3908564503097226,
      "grad_norm": 0.773051619092523,
      "learning_rate": 9.087181718894963e-06,
      "loss": 0.39449689388275144,
      "num_input_tokens_seen": 298703408,
      "step": 5805,
      "train_runtime": 35816.0695,
      "train_tokens_per_second": 8339.927
    },
    {
      "epoch": 0.39119310530568274,
      "grad_norm": 0.6358547384731529,
      "learning_rate": 9.085658106071309e-06,
      "loss": 0.38839995861053467,
      "num_input_tokens_seen": 298961272,
      "step": 5810,
      "train_runtime": 35846.5035,
      "train_tokens_per_second": 8340.04
    },
    {
      "epoch": 0.3915297603016429,
      "grad_norm": 0.6234081124930522,
      "learning_rate": 9.084133350706394e-06,
      "loss": 0.372831392288208,
      "num_input_tokens_seen": 299209056,
      "step": 5815,
      "train_runtime": 35873.3025,
      "train_tokens_per_second": 8340.717
    },
    {
      "epoch": 0.391866415297603,
      "grad_norm": 0.7249173847722905,
      "learning_rate": 9.082607453226613e-06,
      "loss": 0.3870790719985962,
      "num_input_tokens_seen": 299479760,
      "step": 5820,
      "train_runtime": 35902.7999,
      "train_tokens_per_second": 8341.404
    },
    {
      "epoch": 0.39220307029356316,
      "grad_norm": 0.5935993272339335,
      "learning_rate": 9.081080414058674e-06,
      "loss": 0.3978002071380615,
      "num_input_tokens_seen": 299732528,
      "step": 5825,
      "train_runtime": 35933.8603,
      "train_tokens_per_second": 8341.228
    },
    {
      "epoch": 0.3925397252895233,
      "grad_norm": 0.7008589551134432,
      "learning_rate": 9.079552233629612e-06,
      "loss": 0.36853115558624266,
      "num_input_tokens_seen": 299970832,
      "step": 5830,
      "train_runtime": 35962.4187,
      "train_tokens_per_second": 8341.231
    },
    {
      "epoch": 0.39287638028548344,
      "grad_norm": 0.7262640683442739,
      "learning_rate": 9.078022912366776e-06,
      "loss": 0.3802750587463379,
      "num_input_tokens_seen": 300235512,
      "step": 5835,
      "train_runtime": 35993.2772,
      "train_tokens_per_second": 8341.433
    },
    {
      "epoch": 0.3932130352814436,
      "grad_norm": 0.6884556228672779,
      "learning_rate": 9.076492450697835e-06,
      "loss": 0.38486638069152834,
      "num_input_tokens_seen": 300515784,
      "step": 5840,
      "train_runtime": 36023.5156,
      "train_tokens_per_second": 8342.211
    },
    {
      "epoch": 0.3935496902774037,
      "grad_norm": 0.7562678725402927,
      "learning_rate": 9.074960849050782e-06,
      "loss": 0.3845879793167114,
      "num_input_tokens_seen": 300777680,
      "step": 5845,
      "train_runtime": 36053.8663,
      "train_tokens_per_second": 8342.453
    },
    {
      "epoch": 0.39388634527336386,
      "grad_norm": 0.6831260130288721,
      "learning_rate": 9.07342810785392e-06,
      "loss": 0.3706568717956543,
      "num_input_tokens_seen": 301012312,
      "step": 5850,
      "train_runtime": 36081.0758,
      "train_tokens_per_second": 8342.665
    },
    {
      "epoch": 0.394223000269324,
      "grad_norm": 0.6943181174161223,
      "learning_rate": 9.071894227535878e-06,
      "loss": 0.37863473892211913,
      "num_input_tokens_seen": 301266880,
      "step": 5855,
      "train_runtime": 36112.0789,
      "train_tokens_per_second": 8342.552
    },
    {
      "epoch": 0.39455965526528414,
      "grad_norm": 0.74784620031604,
      "learning_rate": 9.070359208525598e-06,
      "loss": 0.40072102546691896,
      "num_input_tokens_seen": 301541736,
      "step": 5860,
      "train_runtime": 36143.6634,
      "train_tokens_per_second": 8342.866
    },
    {
      "epoch": 0.3948963102612443,
      "grad_norm": 0.8039389990161986,
      "learning_rate": 9.068823051252346e-06,
      "loss": 0.3883746385574341,
      "num_input_tokens_seen": 301799520,
      "step": 5865,
      "train_runtime": 36175.7599,
      "train_tokens_per_second": 8342.59
    },
    {
      "epoch": 0.3952329652572044,
      "grad_norm": 0.6975712883907496,
      "learning_rate": 9.067285756145702e-06,
      "loss": 0.3945502758026123,
      "num_input_tokens_seen": 302084136,
      "step": 5870,
      "train_runtime": 36204.3287,
      "train_tokens_per_second": 8343.868
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.7639114893622858,
      "learning_rate": 9.065747323635565e-06,
      "loss": 0.3211208820343018,
      "num_input_tokens_seen": 302285624,
      "step": 5875,
      "train_runtime": 36233.2401,
      "train_tokens_per_second": 8342.771
    },
    {
      "epoch": 0.3959062752491247,
      "grad_norm": 0.7097253835972193,
      "learning_rate": 9.064207754152154e-06,
      "loss": 0.39483156204223635,
      "num_input_tokens_seen": 302536456,
      "step": 5880,
      "train_runtime": 36259.5643,
      "train_tokens_per_second": 8343.632
    },
    {
      "epoch": 0.39624293024508483,
      "grad_norm": 0.6664164593477332,
      "learning_rate": 9.062667048126005e-06,
      "loss": 0.3697096824645996,
      "num_input_tokens_seen": 302804608,
      "step": 5885,
      "train_runtime": 36293.954,
      "train_tokens_per_second": 8343.114
    },
    {
      "epoch": 0.396579585241045,
      "grad_norm": 0.8327276743035087,
      "learning_rate": 9.061125205987971e-06,
      "loss": 0.41008601188659666,
      "num_input_tokens_seen": 303050480,
      "step": 5890,
      "train_runtime": 36326.3566,
      "train_tokens_per_second": 8342.441
    },
    {
      "epoch": 0.3969162402370051,
      "grad_norm": 0.784796722381948,
      "learning_rate": 9.059582228169222e-06,
      "loss": 0.3697114229202271,
      "num_input_tokens_seen": 303293632,
      "step": 5895,
      "train_runtime": 36355.447,
      "train_tokens_per_second": 8342.454
    },
    {
      "epoch": 0.39725289523296525,
      "grad_norm": 0.6172000755576919,
      "learning_rate": 9.058038115101248e-06,
      "loss": 0.4010317802429199,
      "num_input_tokens_seen": 303576888,
      "step": 5900,
      "train_runtime": 36386.4765,
      "train_tokens_per_second": 8343.124
    },
    {
      "epoch": 0.3975895502289254,
      "grad_norm": 0.6604090762522393,
      "learning_rate": 9.056492867215857e-06,
      "loss": 0.3461556911468506,
      "num_input_tokens_seen": 303811544,
      "step": 5905,
      "train_runtime": 36417.1542,
      "train_tokens_per_second": 8342.539
    },
    {
      "epoch": 0.39792620522488553,
      "grad_norm": 0.605526981955514,
      "learning_rate": 9.05494648494517e-06,
      "loss": 0.3832465887069702,
      "num_input_tokens_seen": 304055320,
      "step": 5910,
      "train_runtime": 36443.6476,
      "train_tokens_per_second": 8343.164
    },
    {
      "epoch": 0.39826286022084567,
      "grad_norm": 0.6325146102111908,
      "learning_rate": 9.053398968721629e-06,
      "loss": 0.3738247394561768,
      "num_input_tokens_seen": 304328320,
      "step": 5915,
      "train_runtime": 36476.7397,
      "train_tokens_per_second": 8343.079
    },
    {
      "epoch": 0.3985995152168058,
      "grad_norm": 0.6826930249378539,
      "learning_rate": 9.05185031897799e-06,
      "loss": 0.42293500900268555,
      "num_input_tokens_seen": 304581192,
      "step": 5920,
      "train_runtime": 36510.1138,
      "train_tokens_per_second": 8342.379
    },
    {
      "epoch": 0.39893617021276595,
      "grad_norm": 0.7080033749136879,
      "learning_rate": 9.050300536147331e-06,
      "loss": 0.4086330413818359,
      "num_input_tokens_seen": 304874000,
      "step": 5925,
      "train_runtime": 36543.615,
      "train_tokens_per_second": 8342.743
    },
    {
      "epoch": 0.3992728252087261,
      "grad_norm": 0.7337807790046751,
      "learning_rate": 9.048749620663044e-06,
      "loss": 0.34280033111572267,
      "num_input_tokens_seen": 305157088,
      "step": 5930,
      "train_runtime": 36575.1863,
      "train_tokens_per_second": 8343.282
    },
    {
      "epoch": 0.39960948020468623,
      "grad_norm": 0.65106000579047,
      "learning_rate": 9.047197572958834e-06,
      "loss": 0.36876654624938965,
      "num_input_tokens_seen": 305417784,
      "step": 5935,
      "train_runtime": 36604.5514,
      "train_tokens_per_second": 8343.711
    },
    {
      "epoch": 0.39994613520064637,
      "grad_norm": 0.692431208562081,
      "learning_rate": 9.04564439346873e-06,
      "loss": 0.38390161991119387,
      "num_input_tokens_seen": 305663680,
      "step": 5940,
      "train_runtime": 36634.0602,
      "train_tokens_per_second": 8343.702
    },
    {
      "epoch": 0.4002827901966065,
      "grad_norm": 0.7037521705577573,
      "learning_rate": 9.044090082627071e-06,
      "loss": 0.3786059617996216,
      "num_input_tokens_seen": 305913720,
      "step": 5945,
      "train_runtime": 36661.5591,
      "train_tokens_per_second": 8344.264
    },
    {
      "epoch": 0.40061944519256665,
      "grad_norm": 0.66886189496791,
      "learning_rate": 9.042534640868514e-06,
      "loss": 0.3942221403121948,
      "num_input_tokens_seen": 306159008,
      "step": 5950,
      "train_runtime": 36690.5685,
      "train_tokens_per_second": 8344.352
    },
    {
      "epoch": 0.4009561001885268,
      "grad_norm": 0.6019034926798098,
      "learning_rate": 9.04097806862804e-06,
      "loss": 0.387723708152771,
      "num_input_tokens_seen": 306429056,
      "step": 5955,
      "train_runtime": 36727.2262,
      "train_tokens_per_second": 8343.376
    },
    {
      "epoch": 0.4012927551844869,
      "grad_norm": 0.6352664686524633,
      "learning_rate": 9.03942036634093e-06,
      "loss": 0.39069309234619143,
      "num_input_tokens_seen": 306681440,
      "step": 5960,
      "train_runtime": 36759.5703,
      "train_tokens_per_second": 8342.901
    },
    {
      "epoch": 0.40162941018044707,
      "grad_norm": 0.6806234281475803,
      "learning_rate": 9.037861534442797e-06,
      "loss": 0.3634389162063599,
      "num_input_tokens_seen": 306945568,
      "step": 5965,
      "train_runtime": 36791.2378,
      "train_tokens_per_second": 8342.899
    },
    {
      "epoch": 0.4019660651764072,
      "grad_norm": 0.7614269876567773,
      "learning_rate": 9.036301573369563e-06,
      "loss": 0.391825270652771,
      "num_input_tokens_seen": 307213344,
      "step": 5970,
      "train_runtime": 36818.4687,
      "train_tokens_per_second": 8344.001
    },
    {
      "epoch": 0.40230272017236735,
      "grad_norm": 0.7918213988505681,
      "learning_rate": 9.034740483557465e-06,
      "loss": 0.3753262758255005,
      "num_input_tokens_seen": 307468920,
      "step": 5975,
      "train_runtime": 36851.124,
      "train_tokens_per_second": 8343.543
    },
    {
      "epoch": 0.4026393751683275,
      "grad_norm": 0.65429194452683,
      "learning_rate": 9.033178265443055e-06,
      "loss": 0.39017202854156496,
      "num_input_tokens_seen": 307735304,
      "step": 5980,
      "train_runtime": 36883.806,
      "train_tokens_per_second": 8343.372
    },
    {
      "epoch": 0.4029760301642876,
      "grad_norm": 0.6838668078459911,
      "learning_rate": 9.031614919463206e-06,
      "loss": 0.37329261302947997,
      "num_input_tokens_seen": 308002680,
      "step": 5985,
      "train_runtime": 36916.2575,
      "train_tokens_per_second": 8343.28
    },
    {
      "epoch": 0.40331268516024776,
      "grad_norm": 0.619087097207014,
      "learning_rate": 9.030050446055099e-06,
      "loss": 0.35289888381958007,
      "num_input_tokens_seen": 308258504,
      "step": 5990,
      "train_runtime": 36948.0628,
      "train_tokens_per_second": 8343.022
    },
    {
      "epoch": 0.4036493401562079,
      "grad_norm": 0.5780062747824698,
      "learning_rate": 9.028484845656235e-06,
      "loss": 0.3856215000152588,
      "num_input_tokens_seen": 308546520,
      "step": 5995,
      "train_runtime": 36980.9863,
      "train_tokens_per_second": 8343.383
    },
    {
      "epoch": 0.40398599515216804,
      "grad_norm": 0.6543312017392544,
      "learning_rate": 9.026918118704431e-06,
      "loss": 0.4088435173034668,
      "num_input_tokens_seen": 308811808,
      "step": 6000,
      "train_runtime": 37008.2733,
      "train_tokens_per_second": 8344.399
    },
    {
      "epoch": 0.4043226501481282,
      "grad_norm": 0.6687392784522165,
      "learning_rate": 9.025350265637816e-06,
      "loss": 0.37173459529876707,
      "num_input_tokens_seen": 309095096,
      "step": 6005,
      "train_runtime": 37043.8864,
      "train_tokens_per_second": 8344.025
    },
    {
      "epoch": 0.4046593051440883,
      "grad_norm": 0.6784830796091781,
      "learning_rate": 9.023781286894834e-06,
      "loss": 0.3737105131149292,
      "num_input_tokens_seen": 309358656,
      "step": 6010,
      "train_runtime": 37075.9364,
      "train_tokens_per_second": 8343.92
    },
    {
      "epoch": 0.40499596014004846,
      "grad_norm": 0.680870583814063,
      "learning_rate": 9.022211182914247e-06,
      "loss": 0.3655592918395996,
      "num_input_tokens_seen": 309615680,
      "step": 6015,
      "train_runtime": 37107.734,
      "train_tokens_per_second": 8343.697
    },
    {
      "epoch": 0.4053326151360086,
      "grad_norm": 2.216145980598881,
      "learning_rate": 9.020639954135128e-06,
      "loss": 0.3781093120574951,
      "num_input_tokens_seen": 309864680,
      "step": 6020,
      "train_runtime": 37140.2368,
      "train_tokens_per_second": 8343.099
    },
    {
      "epoch": 0.40566927013196874,
      "grad_norm": 0.7399954990192545,
      "learning_rate": 9.019067600996867e-06,
      "loss": 0.3067362070083618,
      "num_input_tokens_seen": 310093768,
      "step": 6025,
      "train_runtime": 37172.4121,
      "train_tokens_per_second": 8342.041
    },
    {
      "epoch": 0.4060059251279289,
      "grad_norm": 0.6914646023886861,
      "learning_rate": 9.017494123939169e-06,
      "loss": 0.3744835376739502,
      "num_input_tokens_seen": 310344288,
      "step": 6030,
      "train_runtime": 37203.4267,
      "train_tokens_per_second": 8341.82
    },
    {
      "epoch": 0.406342580123889,
      "grad_norm": 0.6476882800007183,
      "learning_rate": 9.015919523402048e-06,
      "loss": 0.387648606300354,
      "num_input_tokens_seen": 310603128,
      "step": 6035,
      "train_runtime": 37230.2752,
      "train_tokens_per_second": 8342.757
    },
    {
      "epoch": 0.40667923511984916,
      "grad_norm": 0.8671748451981152,
      "learning_rate": 9.014343799825838e-06,
      "loss": 0.3815485954284668,
      "num_input_tokens_seen": 310847968,
      "step": 6040,
      "train_runtime": 37261.8139,
      "train_tokens_per_second": 8342.266
    },
    {
      "epoch": 0.4070158901158093,
      "grad_norm": 0.5832875468168266,
      "learning_rate": 9.012766953651186e-06,
      "loss": 0.418215274810791,
      "num_input_tokens_seen": 311118080,
      "step": 6045,
      "train_runtime": 37290.2805,
      "train_tokens_per_second": 8343.141
    },
    {
      "epoch": 0.40735254511176944,
      "grad_norm": 0.6091771862487693,
      "learning_rate": 9.01118898531905e-06,
      "loss": 0.3947951316833496,
      "num_input_tokens_seen": 311352744,
      "step": 6050,
      "train_runtime": 37320.6939,
      "train_tokens_per_second": 8342.63
    },
    {
      "epoch": 0.4076892001077296,
      "grad_norm": 0.6608870636389189,
      "learning_rate": 9.009609895270708e-06,
      "loss": 0.3585469722747803,
      "num_input_tokens_seen": 311617456,
      "step": 6055,
      "train_runtime": 37355.0071,
      "train_tokens_per_second": 8342.053
    },
    {
      "epoch": 0.4080258551036897,
      "grad_norm": 0.7166899474517496,
      "learning_rate": 9.008029683947743e-06,
      "loss": 0.4171760559082031,
      "num_input_tokens_seen": 311887216,
      "step": 6060,
      "train_runtime": 37387.7904,
      "train_tokens_per_second": 8341.954
    },
    {
      "epoch": 0.40836251009964986,
      "grad_norm": 0.6810717848634962,
      "learning_rate": 9.006448351792057e-06,
      "loss": 0.39299988746643066,
      "num_input_tokens_seen": 312138352,
      "step": 6065,
      "train_runtime": 37416.3389,
      "train_tokens_per_second": 8342.301
    },
    {
      "epoch": 0.40869916509561,
      "grad_norm": 0.6180053289200077,
      "learning_rate": 9.004865899245864e-06,
      "loss": 0.3885251522064209,
      "num_input_tokens_seen": 312385704,
      "step": 6070,
      "train_runtime": 37447.6015,
      "train_tokens_per_second": 8341.942
    },
    {
      "epoch": 0.40903582009157013,
      "grad_norm": 0.62368450580202,
      "learning_rate": 9.003282326751694e-06,
      "loss": 0.39011297225952146,
      "num_input_tokens_seen": 312646240,
      "step": 6075,
      "train_runtime": 37479.8384,
      "train_tokens_per_second": 8341.718
    },
    {
      "epoch": 0.4093724750875303,
      "grad_norm": 0.9352145846464222,
      "learning_rate": 9.001697634752387e-06,
      "loss": 0.4065091133117676,
      "num_input_tokens_seen": 312899400,
      "step": 6080,
      "train_runtime": 37511.2631,
      "train_tokens_per_second": 8341.479
    },
    {
      "epoch": 0.4097091300834904,
      "grad_norm": 0.6113167582968716,
      "learning_rate": 9.000111823691097e-06,
      "loss": 0.3742032527923584,
      "num_input_tokens_seen": 313163904,
      "step": 6085,
      "train_runtime": 37541.5426,
      "train_tokens_per_second": 8341.796
    },
    {
      "epoch": 0.41004578507945055,
      "grad_norm": 0.6464386691779086,
      "learning_rate": 8.99852489401129e-06,
      "loss": 0.3951690673828125,
      "num_input_tokens_seen": 313429440,
      "step": 6090,
      "train_runtime": 37575.2898,
      "train_tokens_per_second": 8341.371
    },
    {
      "epoch": 0.4103824400754107,
      "grad_norm": 0.6287690203634758,
      "learning_rate": 8.996936846156748e-06,
      "loss": 0.3555378198623657,
      "num_input_tokens_seen": 313688112,
      "step": 6095,
      "train_runtime": 37605.3349,
      "train_tokens_per_second": 8341.585
    },
    {
      "epoch": 0.41071909507137083,
      "grad_norm": 0.6916558548816881,
      "learning_rate": 8.995347680571563e-06,
      "loss": 0.3971059322357178,
      "num_input_tokens_seen": 313952936,
      "step": 6100,
      "train_runtime": 37635.3378,
      "train_tokens_per_second": 8341.972
    },
    {
      "epoch": 0.41105575006733097,
      "grad_norm": 0.5781543551676644,
      "learning_rate": 8.993757397700137e-06,
      "loss": 0.387666130065918,
      "num_input_tokens_seen": 314212368,
      "step": 6105,
      "train_runtime": 37663.4781,
      "train_tokens_per_second": 8342.627
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 0.6714235150109914,
      "learning_rate": 8.99216599798719e-06,
      "loss": 0.41055145263671877,
      "num_input_tokens_seen": 314499920,
      "step": 6110,
      "train_runtime": 37699.4024,
      "train_tokens_per_second": 8342.305
    },
    {
      "epoch": 0.4117290600592513,
      "grad_norm": 0.7783761933464339,
      "learning_rate": 8.990573481877753e-06,
      "loss": 0.35719239711761475,
      "num_input_tokens_seen": 314737632,
      "step": 6115,
      "train_runtime": 37729.4233,
      "train_tokens_per_second": 8341.968
    },
    {
      "epoch": 0.41206571505521145,
      "grad_norm": 0.6444211110024752,
      "learning_rate": 8.988979849817167e-06,
      "loss": 0.3659427404403687,
      "num_input_tokens_seen": 314958768,
      "step": 6120,
      "train_runtime": 37758.4091,
      "train_tokens_per_second": 8341.421
    },
    {
      "epoch": 0.4124023700511716,
      "grad_norm": 0.8689109581249207,
      "learning_rate": 8.987385102251085e-06,
      "loss": 0.3761105537414551,
      "num_input_tokens_seen": 315185720,
      "step": 6125,
      "train_runtime": 37792.6658,
      "train_tokens_per_second": 8339.865
    },
    {
      "epoch": 0.4127390250471317,
      "grad_norm": 0.6865857712305324,
      "learning_rate": 8.985789239625475e-06,
      "loss": 0.3820502281188965,
      "num_input_tokens_seen": 315437120,
      "step": 6130,
      "train_runtime": 37823.3774,
      "train_tokens_per_second": 8339.74
    },
    {
      "epoch": 0.41307568004309186,
      "grad_norm": 0.7376632616001159,
      "learning_rate": 8.984192262386613e-06,
      "loss": 0.35550570487976074,
      "num_input_tokens_seen": 315703840,
      "step": 6135,
      "train_runtime": 37856.6533,
      "train_tokens_per_second": 8339.455
    },
    {
      "epoch": 0.413412335039052,
      "grad_norm": 0.6937796041337648,
      "learning_rate": 8.98259417098109e-06,
      "loss": 0.39723641872406007,
      "num_input_tokens_seen": 315975080,
      "step": 6140,
      "train_runtime": 37889.4207,
      "train_tokens_per_second": 8339.401
    },
    {
      "epoch": 0.41374899003501214,
      "grad_norm": 0.7133546714763493,
      "learning_rate": 8.980994965855808e-06,
      "loss": 0.3785146713256836,
      "num_input_tokens_seen": 316233040,
      "step": 6145,
      "train_runtime": 37921.0532,
      "train_tokens_per_second": 8339.247
    },
    {
      "epoch": 0.4140856450309723,
      "grad_norm": 0.6509505246618501,
      "learning_rate": 8.979394647457976e-06,
      "loss": 0.378758978843689,
      "num_input_tokens_seen": 316490744,
      "step": 6150,
      "train_runtime": 37951.1932,
      "train_tokens_per_second": 8339.415
    },
    {
      "epoch": 0.4144223000269324,
      "grad_norm": 0.6700512225552622,
      "learning_rate": 8.97779321623512e-06,
      "loss": 0.340434193611145,
      "num_input_tokens_seen": 316730400,
      "step": 6155,
      "train_runtime": 37983.6871,
      "train_tokens_per_second": 8338.59
    },
    {
      "epoch": 0.41475895502289256,
      "grad_norm": 0.6263441071620043,
      "learning_rate": 8.976190672635077e-06,
      "loss": 0.3832817316055298,
      "num_input_tokens_seen": 316994120,
      "step": 6160,
      "train_runtime": 38016.1764,
      "train_tokens_per_second": 8338.401
    },
    {
      "epoch": 0.4150956100188527,
      "grad_norm": 0.6596366015052689,
      "learning_rate": 8.974587017105991e-06,
      "loss": 0.39680254459381104,
      "num_input_tokens_seen": 317249160,
      "step": 6165,
      "train_runtime": 38049.7258,
      "train_tokens_per_second": 8337.752
    },
    {
      "epoch": 0.41543226501481284,
      "grad_norm": 0.6483787351724974,
      "learning_rate": 8.97298225009632e-06,
      "loss": 0.3907045602798462,
      "num_input_tokens_seen": 317506408,
      "step": 6170,
      "train_runtime": 38080.3808,
      "train_tokens_per_second": 8337.795
    },
    {
      "epoch": 0.415768920010773,
      "grad_norm": 0.7067146092073416,
      "learning_rate": 8.971376372054829e-06,
      "loss": 0.38661699295043944,
      "num_input_tokens_seen": 317756920,
      "step": 6175,
      "train_runtime": 38111.1023,
      "train_tokens_per_second": 8337.647
    },
    {
      "epoch": 0.4161055750067331,
      "grad_norm": 0.6040162449775972,
      "learning_rate": 8.969769383430602e-06,
      "loss": 0.37240090370178225,
      "num_input_tokens_seen": 318023672,
      "step": 6180,
      "train_runtime": 38141.4114,
      "train_tokens_per_second": 8338.015
    },
    {
      "epoch": 0.41644223000269326,
      "grad_norm": 0.6076911559362083,
      "learning_rate": 8.968161284673027e-06,
      "loss": 0.37006478309631347,
      "num_input_tokens_seen": 318292696,
      "step": 6185,
      "train_runtime": 38174.6735,
      "train_tokens_per_second": 8337.797
    },
    {
      "epoch": 0.4167788849986534,
      "grad_norm": 0.7725161902678804,
      "learning_rate": 8.9665520762318e-06,
      "loss": 0.3689037561416626,
      "num_input_tokens_seen": 318534832,
      "step": 6190,
      "train_runtime": 38206.4232,
      "train_tokens_per_second": 8337.206
    },
    {
      "epoch": 0.41711553999461354,
      "grad_norm": 0.7152011007921621,
      "learning_rate": 8.964941758556934e-06,
      "loss": 0.3963191509246826,
      "num_input_tokens_seen": 318772584,
      "step": 6195,
      "train_runtime": 38239.8837,
      "train_tokens_per_second": 8336.128
    },
    {
      "epoch": 0.4174521949905737,
      "grad_norm": 0.6581109139516929,
      "learning_rate": 8.963330332098747e-06,
      "loss": 0.37104551792144774,
      "num_input_tokens_seen": 319020432,
      "step": 6200,
      "train_runtime": 38267.9011,
      "train_tokens_per_second": 8336.502
    },
    {
      "epoch": 0.4177888499865338,
      "grad_norm": 0.646130052702144,
      "learning_rate": 8.961717797307872e-06,
      "loss": 0.35717215538024905,
      "num_input_tokens_seen": 319282768,
      "step": 6205,
      "train_runtime": 38299.7645,
      "train_tokens_per_second": 8336.416
    },
    {
      "epoch": 0.41812550498249396,
      "grad_norm": 0.6296037570469866,
      "learning_rate": 8.960104154635248e-06,
      "loss": 0.39736108779907225,
      "num_input_tokens_seen": 319543440,
      "step": 6210,
      "train_runtime": 38332.2831,
      "train_tokens_per_second": 8336.144
    },
    {
      "epoch": 0.4184621599784541,
      "grad_norm": 0.7444364698558681,
      "learning_rate": 8.958489404532125e-06,
      "loss": 0.3933404922485352,
      "num_input_tokens_seen": 319788384,
      "step": 6215,
      "train_runtime": 38358.5425,
      "train_tokens_per_second": 8336.823
    },
    {
      "epoch": 0.41879881497441424,
      "grad_norm": 1.8184272627378035,
      "learning_rate": 8.956873547450062e-06,
      "loss": 0.39119007587432864,
      "num_input_tokens_seen": 320039584,
      "step": 6220,
      "train_runtime": 38388.9388,
      "train_tokens_per_second": 8336.766
    },
    {
      "epoch": 0.4191354699703744,
      "grad_norm": 0.7627437949362296,
      "learning_rate": 8.95525658384093e-06,
      "loss": 0.4120905876159668,
      "num_input_tokens_seen": 320295848,
      "step": 6225,
      "train_runtime": 38420.5962,
      "train_tokens_per_second": 8336.566
    },
    {
      "epoch": 0.4194721249663345,
      "grad_norm": 0.7166747221838897,
      "learning_rate": 8.953638514156908e-06,
      "loss": 0.37010622024536133,
      "num_input_tokens_seen": 320553528,
      "step": 6230,
      "train_runtime": 38451.7375,
      "train_tokens_per_second": 8336.516
    },
    {
      "epoch": 0.41980877996229465,
      "grad_norm": 0.6758451266598695,
      "learning_rate": 8.952019338850481e-06,
      "loss": 0.38879220485687255,
      "num_input_tokens_seen": 320804088,
      "step": 6235,
      "train_runtime": 38483.8157,
      "train_tokens_per_second": 8336.078
    },
    {
      "epoch": 0.4201454349582548,
      "grad_norm": 0.652097784930252,
      "learning_rate": 8.950399058374447e-06,
      "loss": 0.36224000453948973,
      "num_input_tokens_seen": 321076072,
      "step": 6240,
      "train_runtime": 38516.5818,
      "train_tokens_per_second": 8336.048
    },
    {
      "epoch": 0.42048208995421493,
      "grad_norm": 0.6687338162145592,
      "learning_rate": 8.948777673181913e-06,
      "loss": 0.38503413200378417,
      "num_input_tokens_seen": 321348472,
      "step": 6245,
      "train_runtime": 38549.9392,
      "train_tokens_per_second": 8335.901
    },
    {
      "epoch": 0.42081874495017507,
      "grad_norm": 0.6212840099633836,
      "learning_rate": 8.947155183726296e-06,
      "loss": 0.35733513832092284,
      "num_input_tokens_seen": 321615536,
      "step": 6250,
      "train_runtime": 38585.4904,
      "train_tokens_per_second": 8335.142
    },
    {
      "epoch": 0.4211553999461352,
      "grad_norm": 0.6402037511269458,
      "learning_rate": 8.94553159046132e-06,
      "loss": 0.36295013427734374,
      "num_input_tokens_seen": 321875168,
      "step": 6255,
      "train_runtime": 38618.0052,
      "train_tokens_per_second": 8334.847
    },
    {
      "epoch": 0.42149205494209535,
      "grad_norm": 0.5853702366212903,
      "learning_rate": 8.943906893841015e-06,
      "loss": 0.36299285888671873,
      "num_input_tokens_seen": 322135256,
      "step": 6260,
      "train_runtime": 38651.0927,
      "train_tokens_per_second": 8334.441
    },
    {
      "epoch": 0.4218287099380555,
      "grad_norm": 0.6270511833516261,
      "learning_rate": 8.942281094319721e-06,
      "loss": 0.39774355888366697,
      "num_input_tokens_seen": 322394120,
      "step": 6265,
      "train_runtime": 38677.7849,
      "train_tokens_per_second": 8335.382
    },
    {
      "epoch": 0.42216536493401563,
      "grad_norm": 0.672086582354334,
      "learning_rate": 8.940654192352092e-06,
      "loss": 0.40533857345581054,
      "num_input_tokens_seen": 322630432,
      "step": 6270,
      "train_runtime": 38705.9109,
      "train_tokens_per_second": 8335.431
    },
    {
      "epoch": 0.42250201992997577,
      "grad_norm": 0.7400360527394102,
      "learning_rate": 8.939026188393083e-06,
      "loss": 0.37662224769592284,
      "num_input_tokens_seen": 322879600,
      "step": 6275,
      "train_runtime": 38740.4645,
      "train_tokens_per_second": 8334.428
    },
    {
      "epoch": 0.4228386749259359,
      "grad_norm": 0.625424911325638,
      "learning_rate": 8.937397082897963e-06,
      "loss": 0.3835691213607788,
      "num_input_tokens_seen": 323153336,
      "step": 6280,
      "train_runtime": 38768.1402,
      "train_tokens_per_second": 8335.539
    },
    {
      "epoch": 0.42317532992189605,
      "grad_norm": 1.0674920449396366,
      "learning_rate": 8.935766876322303e-06,
      "loss": 0.40459022521972654,
      "num_input_tokens_seen": 323417128,
      "step": 6285,
      "train_runtime": 38801.2894,
      "train_tokens_per_second": 8335.216
    },
    {
      "epoch": 0.4235119849178562,
      "grad_norm": 0.6388921182932564,
      "learning_rate": 8.934135569121985e-06,
      "loss": 0.388218092918396,
      "num_input_tokens_seen": 323688240,
      "step": 6290,
      "train_runtime": 38834.7229,
      "train_tokens_per_second": 8335.021
    },
    {
      "epoch": 0.4238486399138163,
      "grad_norm": 0.5764109438352905,
      "learning_rate": 8.932503161753202e-06,
      "loss": 0.3775338172912598,
      "num_input_tokens_seen": 323956136,
      "step": 6295,
      "train_runtime": 38863.1607,
      "train_tokens_per_second": 8335.815
    },
    {
      "epoch": 0.42418529490977647,
      "grad_norm": 0.8035324670489796,
      "learning_rate": 8.930869654672449e-06,
      "loss": 0.38253326416015626,
      "num_input_tokens_seen": 324211152,
      "step": 6300,
      "train_runtime": 38893.3938,
      "train_tokens_per_second": 8335.893
    },
    {
      "epoch": 0.4245219499057366,
      "grad_norm": 0.9052904903222522,
      "learning_rate": 8.929235048336531e-06,
      "loss": 0.36686158180236816,
      "num_input_tokens_seen": 324475792,
      "step": 6305,
      "train_runtime": 38923.6084,
      "train_tokens_per_second": 8336.221
    },
    {
      "epoch": 0.42485860490169675,
      "grad_norm": 0.6877434724125321,
      "learning_rate": 8.927599343202563e-06,
      "loss": 0.39807443618774413,
      "num_input_tokens_seen": 324755152,
      "step": 6310,
      "train_runtime": 38956.309,
      "train_tokens_per_second": 8336.394
    },
    {
      "epoch": 0.4251952598976569,
      "grad_norm": 0.6573821578522724,
      "learning_rate": 8.92596253972796e-06,
      "loss": 0.3873600959777832,
      "num_input_tokens_seen": 325054568,
      "step": 6315,
      "train_runtime": 38987.7777,
      "train_tokens_per_second": 8337.345
    },
    {
      "epoch": 0.425531914893617,
      "grad_norm": 0.6250912719254439,
      "learning_rate": 8.924324638370454e-06,
      "loss": 0.40852699279785154,
      "num_input_tokens_seen": 325324656,
      "step": 6320,
      "train_runtime": 39017.2586,
      "train_tokens_per_second": 8337.968
    },
    {
      "epoch": 0.42586856988957716,
      "grad_norm": 0.560991365466394,
      "learning_rate": 8.922685639588076e-06,
      "loss": 0.3364936351776123,
      "num_input_tokens_seen": 325557888,
      "step": 6325,
      "train_runtime": 39048.4365,
      "train_tokens_per_second": 8337.284
    },
    {
      "epoch": 0.4262052248855373,
      "grad_norm": 0.7216721918205741,
      "learning_rate": 8.921045543839167e-06,
      "loss": 0.40285377502441405,
      "num_input_tokens_seen": 325827560,
      "step": 6330,
      "train_runtime": 39081.6858,
      "train_tokens_per_second": 8337.091
    },
    {
      "epoch": 0.42654187988149744,
      "grad_norm": 0.7552657726435108,
      "learning_rate": 8.919404351582376e-06,
      "loss": 0.37372000217437745,
      "num_input_tokens_seen": 326055456,
      "step": 6335,
      "train_runtime": 39113.3262,
      "train_tokens_per_second": 8336.173
    },
    {
      "epoch": 0.4268785348774576,
      "grad_norm": 0.6789521062724837,
      "learning_rate": 8.917762063276657e-06,
      "loss": 0.37269010543823244,
      "num_input_tokens_seen": 326306584,
      "step": 6340,
      "train_runtime": 39147.0421,
      "train_tokens_per_second": 8335.408
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.5944740456637408,
      "learning_rate": 8.916118679381268e-06,
      "loss": 0.3733722686767578,
      "num_input_tokens_seen": 326564176,
      "step": 6345,
      "train_runtime": 39178.1521,
      "train_tokens_per_second": 8335.364
    },
    {
      "epoch": 0.42755184486937786,
      "grad_norm": 1.5978117943714971,
      "learning_rate": 8.914474200355777e-06,
      "loss": 0.38065276145935056,
      "num_input_tokens_seen": 326804592,
      "step": 6350,
      "train_runtime": 39206.3586,
      "train_tokens_per_second": 8335.5
    },
    {
      "epoch": 0.427888499865338,
      "grad_norm": 0.7443766542035906,
      "learning_rate": 8.912828626660059e-06,
      "loss": 0.39615108966827395,
      "num_input_tokens_seen": 327077280,
      "step": 6355,
      "train_runtime": 39238.4118,
      "train_tokens_per_second": 8335.64
    },
    {
      "epoch": 0.42822515486129814,
      "grad_norm": 0.6796710023313592,
      "learning_rate": 8.911181958754294e-06,
      "loss": 0.40175204277038573,
      "num_input_tokens_seen": 327335176,
      "step": 6360,
      "train_runtime": 39268.2626,
      "train_tokens_per_second": 8335.871
    },
    {
      "epoch": 0.4285618098572583,
      "grad_norm": 0.5826628112580934,
      "learning_rate": 8.909534197098966e-06,
      "loss": 0.38506431579589845,
      "num_input_tokens_seen": 327609496,
      "step": 6365,
      "train_runtime": 39303.354,
      "train_tokens_per_second": 8335.408
    },
    {
      "epoch": 0.4288984648532184,
      "grad_norm": 0.6567869148234816,
      "learning_rate": 8.907885342154865e-06,
      "loss": 0.3835031986236572,
      "num_input_tokens_seen": 327879360,
      "step": 6370,
      "train_runtime": 39329.8014,
      "train_tokens_per_second": 8336.665
    },
    {
      "epoch": 0.42923511984917856,
      "grad_norm": 0.7089240932589337,
      "learning_rate": 8.906235394383089e-06,
      "loss": 0.3608614444732666,
      "num_input_tokens_seen": 328097144,
      "step": 6375,
      "train_runtime": 39357.7168,
      "train_tokens_per_second": 8336.285
    },
    {
      "epoch": 0.4295717748451387,
      "grad_norm": 0.6591443709080359,
      "learning_rate": 8.904584354245042e-06,
      "loss": 0.372680139541626,
      "num_input_tokens_seen": 328359904,
      "step": 6380,
      "train_runtime": 39390.9906,
      "train_tokens_per_second": 8335.914
    },
    {
      "epoch": 0.42990842984109884,
      "grad_norm": 0.8217923151117263,
      "learning_rate": 8.902932222202433e-06,
      "loss": 0.3867887735366821,
      "num_input_tokens_seen": 328634256,
      "step": 6385,
      "train_runtime": 39422.1368,
      "train_tokens_per_second": 8336.287
    },
    {
      "epoch": 0.430245084837059,
      "grad_norm": 0.7315948726594934,
      "learning_rate": 8.901278998717272e-06,
      "loss": 0.37132930755615234,
      "num_input_tokens_seen": 328909064,
      "step": 6390,
      "train_runtime": 39455.7923,
      "train_tokens_per_second": 8336.141
    },
    {
      "epoch": 0.4305817398330191,
      "grad_norm": 0.7466846307719773,
      "learning_rate": 8.899624684251878e-06,
      "loss": 0.3827688217163086,
      "num_input_tokens_seen": 329163600,
      "step": 6395,
      "train_runtime": 39484.02,
      "train_tokens_per_second": 8336.628
    },
    {
      "epoch": 0.43091839482897926,
      "grad_norm": 0.7745806903657445,
      "learning_rate": 8.897969279268877e-06,
      "loss": 0.3764952182769775,
      "num_input_tokens_seen": 329427264,
      "step": 6400,
      "train_runtime": 39512.561,
      "train_tokens_per_second": 8337.279
    },
    {
      "epoch": 0.4312550498249394,
      "grad_norm": 0.637918947441338,
      "learning_rate": 8.896312784231196e-06,
      "loss": 0.3653470754623413,
      "num_input_tokens_seen": 329680696,
      "step": 6405,
      "train_runtime": 39541.321,
      "train_tokens_per_second": 8337.625
    },
    {
      "epoch": 0.43159170482089954,
      "grad_norm": 0.7389524891415091,
      "learning_rate": 8.89465519960207e-06,
      "loss": 0.4129926681518555,
      "num_input_tokens_seen": 329969696,
      "step": 6410,
      "train_runtime": 39571.6041,
      "train_tokens_per_second": 8338.547
    },
    {
      "epoch": 0.4319283598168597,
      "grad_norm": 0.673576179721604,
      "learning_rate": 8.892996525845037e-06,
      "loss": 0.38866827487945554,
      "num_input_tokens_seen": 330226456,
      "step": 6415,
      "train_runtime": 39601.526,
      "train_tokens_per_second": 8338.731
    },
    {
      "epoch": 0.4322650148128198,
      "grad_norm": 0.819624556394507,
      "learning_rate": 8.89133676342394e-06,
      "loss": 0.3862994432449341,
      "num_input_tokens_seen": 330454904,
      "step": 6420,
      "train_runtime": 39632.7022,
      "train_tokens_per_second": 8337.935
    },
    {
      "epoch": 0.43260166980877995,
      "grad_norm": 0.7265334665918796,
      "learning_rate": 8.889675912802923e-06,
      "loss": 0.3791198253631592,
      "num_input_tokens_seen": 330698024,
      "step": 6425,
      "train_runtime": 39661.4602,
      "train_tokens_per_second": 8338.019
    },
    {
      "epoch": 0.4329383248047401,
      "grad_norm": 0.7425617095690129,
      "learning_rate": 8.888013974446443e-06,
      "loss": 0.4081270217895508,
      "num_input_tokens_seen": 330961336,
      "step": 6430,
      "train_runtime": 39694.103,
      "train_tokens_per_second": 8337.796
    },
    {
      "epoch": 0.43327497980070023,
      "grad_norm": 0.7260178185390229,
      "learning_rate": 8.886350948819253e-06,
      "loss": 0.3779928684234619,
      "num_input_tokens_seen": 331218448,
      "step": 6435,
      "train_runtime": 39724.5725,
      "train_tokens_per_second": 8337.873
    },
    {
      "epoch": 0.4336116347966604,
      "grad_norm": 0.6674136912091239,
      "learning_rate": 8.884686836386412e-06,
      "loss": 0.37410221099853513,
      "num_input_tokens_seen": 331472568,
      "step": 6440,
      "train_runtime": 39753.4041,
      "train_tokens_per_second": 8338.218
    },
    {
      "epoch": 0.4339482897926205,
      "grad_norm": 0.6363999095948231,
      "learning_rate": 8.883021637613286e-06,
      "loss": 0.3879027128219604,
      "num_input_tokens_seen": 331747024,
      "step": 6445,
      "train_runtime": 39786.672,
      "train_tokens_per_second": 8338.145
    },
    {
      "epoch": 0.43428494478858065,
      "grad_norm": 0.5519062843106157,
      "learning_rate": 8.881355352965537e-06,
      "loss": 0.37331104278564453,
      "num_input_tokens_seen": 332012432,
      "step": 6450,
      "train_runtime": 39813.2148,
      "train_tokens_per_second": 8339.252
    },
    {
      "epoch": 0.4346215997845408,
      "grad_norm": 0.6261140723172173,
      "learning_rate": 8.879687982909145e-06,
      "loss": 0.3628070592880249,
      "num_input_tokens_seen": 332249296,
      "step": 6455,
      "train_runtime": 39840.8736,
      "train_tokens_per_second": 8339.408
    },
    {
      "epoch": 0.43495825478050093,
      "grad_norm": 0.7590329621432812,
      "learning_rate": 8.878019527910378e-06,
      "loss": 0.3879849433898926,
      "num_input_tokens_seen": 332482664,
      "step": 6460,
      "train_runtime": 39871.0369,
      "train_tokens_per_second": 8338.952
    },
    {
      "epoch": 0.43529490977646107,
      "grad_norm": 0.7869194933709005,
      "learning_rate": 8.876349988435815e-06,
      "loss": 0.4378767490386963,
      "num_input_tokens_seen": 332738688,
      "step": 6465,
      "train_runtime": 39898.012,
      "train_tokens_per_second": 8339.731
    },
    {
      "epoch": 0.4356315647724212,
      "grad_norm": 0.5604178175087633,
      "learning_rate": 8.874679364952339e-06,
      "loss": 0.37811200618743895,
      "num_input_tokens_seen": 333024960,
      "step": 6470,
      "train_runtime": 39927.2691,
      "train_tokens_per_second": 8340.79
    },
    {
      "epoch": 0.43596821976838135,
      "grad_norm": 0.6508530556497134,
      "learning_rate": 8.873007657927135e-06,
      "loss": 0.3676558494567871,
      "num_input_tokens_seen": 333294936,
      "step": 6475,
      "train_runtime": 39956.4286,
      "train_tokens_per_second": 8341.46
    },
    {
      "epoch": 0.4363048747643415,
      "grad_norm": 0.6997403453148796,
      "learning_rate": 8.87133486782769e-06,
      "loss": 0.33975834846496583,
      "num_input_tokens_seen": 333538608,
      "step": 6480,
      "train_runtime": 39986.1317,
      "train_tokens_per_second": 8341.357
    },
    {
      "epoch": 0.4366415297603016,
      "grad_norm": 0.6489296834846972,
      "learning_rate": 8.869660995121792e-06,
      "loss": 0.38961267471313477,
      "num_input_tokens_seen": 333801776,
      "step": 6485,
      "train_runtime": 40015.7065,
      "train_tokens_per_second": 8341.769
    },
    {
      "epoch": 0.43697818475626177,
      "grad_norm": 0.6477450645918214,
      "learning_rate": 8.867986040277537e-06,
      "loss": 0.3917898416519165,
      "num_input_tokens_seen": 334034784,
      "step": 6490,
      "train_runtime": 40047.3462,
      "train_tokens_per_second": 8340.997
    },
    {
      "epoch": 0.4373148397522219,
      "grad_norm": 0.7456628070824932,
      "learning_rate": 8.86631000376332e-06,
      "loss": 0.40244040489196775,
      "num_input_tokens_seen": 334277952,
      "step": 6495,
      "train_runtime": 40077.7573,
      "train_tokens_per_second": 8340.735
    },
    {
      "epoch": 0.43765149474818205,
      "grad_norm": 0.696609653110475,
      "learning_rate": 8.864632886047841e-06,
      "loss": 0.356403112411499,
      "num_input_tokens_seen": 334527768,
      "step": 6500,
      "train_runtime": 40109.5401,
      "train_tokens_per_second": 8340.354
    },
    {
      "epoch": 0.4379881497441422,
      "grad_norm": 0.7357038662464382,
      "learning_rate": 8.862954687600097e-06,
      "loss": 0.41670641899108884,
      "num_input_tokens_seen": 334781720,
      "step": 6505,
      "train_runtime": 40142.5985,
      "train_tokens_per_second": 8339.812
    },
    {
      "epoch": 0.4383248047401023,
      "grad_norm": 0.7520534735408295,
      "learning_rate": 8.861275408889393e-06,
      "loss": 0.39401199817657473,
      "num_input_tokens_seen": 335019848,
      "step": 6510,
      "train_runtime": 40169.9477,
      "train_tokens_per_second": 8340.062
    },
    {
      "epoch": 0.43866145973606246,
      "grad_norm": 0.752605695336146,
      "learning_rate": 8.859595050385332e-06,
      "loss": 0.3983294486999512,
      "num_input_tokens_seen": 335276984,
      "step": 6515,
      "train_runtime": 40206.5444,
      "train_tokens_per_second": 8338.866
    },
    {
      "epoch": 0.4389981147320226,
      "grad_norm": 0.6694838725124834,
      "learning_rate": 8.857913612557825e-06,
      "loss": 0.3874492883682251,
      "num_input_tokens_seen": 335550640,
      "step": 6520,
      "train_runtime": 40240.6904,
      "train_tokens_per_second": 8338.591
    },
    {
      "epoch": 0.43933476972798274,
      "grad_norm": 0.7506880901530358,
      "learning_rate": 8.856231095877077e-06,
      "loss": 0.37218761444091797,
      "num_input_tokens_seen": 335801088,
      "step": 6525,
      "train_runtime": 40270.1741,
      "train_tokens_per_second": 8338.705
    },
    {
      "epoch": 0.4396714247239429,
      "grad_norm": 0.7889851990145172,
      "learning_rate": 8.8545475008136e-06,
      "loss": 0.3742119073867798,
      "num_input_tokens_seen": 336065896,
      "step": 6530,
      "train_runtime": 40303.7881,
      "train_tokens_per_second": 8338.32
    },
    {
      "epoch": 0.440008079719903,
      "grad_norm": 0.7291571737593224,
      "learning_rate": 8.852862827838207e-06,
      "loss": 0.3729068756103516,
      "num_input_tokens_seen": 336307520,
      "step": 6535,
      "train_runtime": 40332.651,
      "train_tokens_per_second": 8338.344
    },
    {
      "epoch": 0.44034473471586316,
      "grad_norm": 0.5849080816610807,
      "learning_rate": 8.851177077422008e-06,
      "loss": 0.3848399639129639,
      "num_input_tokens_seen": 336582224,
      "step": 6540,
      "train_runtime": 40361.9958,
      "train_tokens_per_second": 8339.088
    },
    {
      "epoch": 0.4406813897118233,
      "grad_norm": 0.6984488116866454,
      "learning_rate": 8.849490250036421e-06,
      "loss": 0.4144181728363037,
      "num_input_tokens_seen": 336843776,
      "step": 6545,
      "train_runtime": 40390.1274,
      "train_tokens_per_second": 8339.755
    },
    {
      "epoch": 0.44101804470778344,
      "grad_norm": 0.705103904477996,
      "learning_rate": 8.847802346153159e-06,
      "loss": 0.3825979709625244,
      "num_input_tokens_seen": 337100984,
      "step": 6550,
      "train_runtime": 40422.9129,
      "train_tokens_per_second": 8339.354
    },
    {
      "epoch": 0.4413546997037436,
      "grad_norm": 0.6692649451692987,
      "learning_rate": 8.846113366244241e-06,
      "loss": 0.37798285484313965,
      "num_input_tokens_seen": 337378952,
      "step": 6555,
      "train_runtime": 40453.0236,
      "train_tokens_per_second": 8340.018
    },
    {
      "epoch": 0.4416913546997037,
      "grad_norm": 0.6434625406467402,
      "learning_rate": 8.844423310781986e-06,
      "loss": 0.3706270933151245,
      "num_input_tokens_seen": 337630672,
      "step": 6560,
      "train_runtime": 40484.4104,
      "train_tokens_per_second": 8339.77
    },
    {
      "epoch": 0.44202800969566386,
      "grad_norm": 0.7887590913777981,
      "learning_rate": 8.842732180239011e-06,
      "loss": 0.4007411003112793,
      "num_input_tokens_seen": 337873112,
      "step": 6565,
      "train_runtime": 40518.4475,
      "train_tokens_per_second": 8338.748
    },
    {
      "epoch": 0.442364664691624,
      "grad_norm": 0.7180058750205456,
      "learning_rate": 8.841039975088234e-06,
      "loss": 0.366942310333252,
      "num_input_tokens_seen": 338128064,
      "step": 6570,
      "train_runtime": 40553.1596,
      "train_tokens_per_second": 8337.897
    },
    {
      "epoch": 0.44270131968758414,
      "grad_norm": 0.8113268909213658,
      "learning_rate": 8.839346695802878e-06,
      "loss": 0.4380204677581787,
      "num_input_tokens_seen": 338408136,
      "step": 6575,
      "train_runtime": 40583.3475,
      "train_tokens_per_second": 8338.596
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.6042144593855177,
      "learning_rate": 8.837652342856459e-06,
      "loss": 0.3795351028442383,
      "num_input_tokens_seen": 338683680,
      "step": 6580,
      "train_runtime": 40621.2307,
      "train_tokens_per_second": 8337.603
    },
    {
      "epoch": 0.4433746296795044,
      "grad_norm": 0.6495373562942688,
      "learning_rate": 8.835956916722803e-06,
      "loss": 0.3622132301330566,
      "num_input_tokens_seen": 338943040,
      "step": 6585,
      "train_runtime": 40650.1399,
      "train_tokens_per_second": 8338.053
    },
    {
      "epoch": 0.44371128467546456,
      "grad_norm": 0.6637591597437986,
      "learning_rate": 8.834260417876024e-06,
      "loss": 0.38673095703125,
      "num_input_tokens_seen": 339193272,
      "step": 6590,
      "train_runtime": 40680.5064,
      "train_tokens_per_second": 8337.981
    },
    {
      "epoch": 0.4440479396714247,
      "grad_norm": 0.6337161233858215,
      "learning_rate": 8.832562846790549e-06,
      "loss": 0.40678882598876953,
      "num_input_tokens_seen": 339466176,
      "step": 6595,
      "train_runtime": 40708.3444,
      "train_tokens_per_second": 8338.983
    },
    {
      "epoch": 0.4443845946673849,
      "grad_norm": 0.6587255641321944,
      "learning_rate": 8.830864203941092e-06,
      "loss": 0.4110139846801758,
      "num_input_tokens_seen": 339730448,
      "step": 6600,
      "train_runtime": 40743.5356,
      "train_tokens_per_second": 8338.266
    },
    {
      "epoch": 0.44472124966334503,
      "grad_norm": 0.7176572909133402,
      "learning_rate": 8.829164489802677e-06,
      "loss": 0.37000303268432616,
      "num_input_tokens_seen": 339968264,
      "step": 6605,
      "train_runtime": 40773.0534,
      "train_tokens_per_second": 8338.062
    },
    {
      "epoch": 0.44505790465930517,
      "grad_norm": 0.7494589478326577,
      "learning_rate": 8.827463704850622e-06,
      "loss": 0.37254767417907714,
      "num_input_tokens_seen": 340222024,
      "step": 6610,
      "train_runtime": 40804.4024,
      "train_tokens_per_second": 8337.875
    },
    {
      "epoch": 0.4453945596552653,
      "grad_norm": 0.7586854043632947,
      "learning_rate": 8.825761849560547e-06,
      "loss": 0.3751363754272461,
      "num_input_tokens_seen": 340473040,
      "step": 6615,
      "train_runtime": 40836.5653,
      "train_tokens_per_second": 8337.455
    },
    {
      "epoch": 0.44573121465122545,
      "grad_norm": 0.7070015805350031,
      "learning_rate": 8.824058924408371e-06,
      "loss": 0.38225038051605226,
      "num_input_tokens_seen": 340720880,
      "step": 6620,
      "train_runtime": 40867.9474,
      "train_tokens_per_second": 8337.118
    },
    {
      "epoch": 0.4460678696471856,
      "grad_norm": 0.6556456650483835,
      "learning_rate": 8.822354929870311e-06,
      "loss": 0.362436580657959,
      "num_input_tokens_seen": 340961280,
      "step": 6625,
      "train_runtime": 40897.8936,
      "train_tokens_per_second": 8336.891
    },
    {
      "epoch": 0.44640452464314573,
      "grad_norm": 0.5326117480641144,
      "learning_rate": 8.820649866422884e-06,
      "loss": 0.3660923957824707,
      "num_input_tokens_seen": 341241016,
      "step": 6630,
      "train_runtime": 40929.7584,
      "train_tokens_per_second": 8337.235
    },
    {
      "epoch": 0.44674117963910587,
      "grad_norm": 0.6081884756131984,
      "learning_rate": 8.818943734542905e-06,
      "loss": 0.37025818824768064,
      "num_input_tokens_seen": 341519424,
      "step": 6635,
      "train_runtime": 40963.3958,
      "train_tokens_per_second": 8337.185
    },
    {
      "epoch": 0.447077834635066,
      "grad_norm": 0.6828751162680429,
      "learning_rate": 8.817236534707486e-06,
      "loss": 0.37492172718048095,
      "num_input_tokens_seen": 341784152,
      "step": 6640,
      "train_runtime": 40993.258,
      "train_tokens_per_second": 8337.57
    },
    {
      "epoch": 0.44741448963102615,
      "grad_norm": 0.6901600578584551,
      "learning_rate": 8.815528267394045e-06,
      "loss": 0.3770854949951172,
      "num_input_tokens_seen": 342048112,
      "step": 6645,
      "train_runtime": 41024.1002,
      "train_tokens_per_second": 8337.736
    },
    {
      "epoch": 0.4477511446269863,
      "grad_norm": 0.6785907134036357,
      "learning_rate": 8.81381893308029e-06,
      "loss": 0.40700554847717285,
      "num_input_tokens_seen": 342317696,
      "step": 6650,
      "train_runtime": 41052.2191,
      "train_tokens_per_second": 8338.592
    },
    {
      "epoch": 0.4480877996229464,
      "grad_norm": 0.6117799382646536,
      "learning_rate": 8.81210853224423e-06,
      "loss": 0.3783367395401001,
      "num_input_tokens_seen": 342579096,
      "step": 6655,
      "train_runtime": 41083.5747,
      "train_tokens_per_second": 8338.59
    },
    {
      "epoch": 0.44842445461890656,
      "grad_norm": 0.7260587852958514,
      "learning_rate": 8.810397065364177e-06,
      "loss": 0.3719600200653076,
      "num_input_tokens_seen": 342802016,
      "step": 6660,
      "train_runtime": 41112.27,
      "train_tokens_per_second": 8338.192
    },
    {
      "epoch": 0.4487611096148667,
      "grad_norm": 0.6330462839578629,
      "learning_rate": 8.808684532918735e-06,
      "loss": 0.4003697395324707,
      "num_input_tokens_seen": 343046472,
      "step": 6665,
      "train_runtime": 41142.9879,
      "train_tokens_per_second": 8337.909
    },
    {
      "epoch": 0.44909776461082684,
      "grad_norm": 0.813497507098597,
      "learning_rate": 8.806970935386807e-06,
      "loss": 0.3686700344085693,
      "num_input_tokens_seen": 343270336,
      "step": 6670,
      "train_runtime": 41173.9695,
      "train_tokens_per_second": 8337.072
    },
    {
      "epoch": 0.449434419606787,
      "grad_norm": 0.7869692018663015,
      "learning_rate": 8.805256273247597e-06,
      "loss": 0.3816176414489746,
      "num_input_tokens_seen": 343504352,
      "step": 6675,
      "train_runtime": 41202.5799,
      "train_tokens_per_second": 8336.962
    },
    {
      "epoch": 0.4497710746027471,
      "grad_norm": 0.7119184458972349,
      "learning_rate": 8.803540546980605e-06,
      "loss": 0.380031156539917,
      "num_input_tokens_seen": 343747152,
      "step": 6680,
      "train_runtime": 41234.6472,
      "train_tokens_per_second": 8336.367
    },
    {
      "epoch": 0.45010772959870726,
      "grad_norm": 0.6264317067871467,
      "learning_rate": 8.801823757065628e-06,
      "loss": 0.38188602924346926,
      "num_input_tokens_seen": 343994032,
      "step": 6685,
      "train_runtime": 41266.7881,
      "train_tokens_per_second": 8335.857
    },
    {
      "epoch": 0.4504443845946674,
      "grad_norm": 0.6185568666743475,
      "learning_rate": 8.800105903982758e-06,
      "loss": 0.36027965545654295,
      "num_input_tokens_seen": 344248016,
      "step": 6690,
      "train_runtime": 41294.273,
      "train_tokens_per_second": 8336.459
    },
    {
      "epoch": 0.45078103959062754,
      "grad_norm": 0.8065924284939741,
      "learning_rate": 8.798386988212393e-06,
      "loss": 0.34608304500579834,
      "num_input_tokens_seen": 344473592,
      "step": 6695,
      "train_runtime": 41322.995,
      "train_tokens_per_second": 8336.124
    },
    {
      "epoch": 0.4511176945865877,
      "grad_norm": 0.600077930593094,
      "learning_rate": 8.796667010235216e-06,
      "loss": 0.40233430862426756,
      "num_input_tokens_seen": 344738544,
      "step": 6700,
      "train_runtime": 41352.5471,
      "train_tokens_per_second": 8336.573
    },
    {
      "epoch": 0.4514543495825478,
      "grad_norm": 0.6698912812222598,
      "learning_rate": 8.794945970532219e-06,
      "loss": 0.37916929721832277,
      "num_input_tokens_seen": 344999568,
      "step": 6705,
      "train_runtime": 41388.8938,
      "train_tokens_per_second": 8335.559
    },
    {
      "epoch": 0.45179100457850796,
      "grad_norm": 0.6861446550327248,
      "learning_rate": 8.79322386958468e-06,
      "loss": 0.41646504402160645,
      "num_input_tokens_seen": 345283728,
      "step": 6710,
      "train_runtime": 41418.7508,
      "train_tokens_per_second": 8336.411
    },
    {
      "epoch": 0.4521276595744681,
      "grad_norm": 0.7894500844260524,
      "learning_rate": 8.791500707874183e-06,
      "loss": 0.3558710336685181,
      "num_input_tokens_seen": 345533424,
      "step": 6715,
      "train_runtime": 41447.8271,
      "train_tokens_per_second": 8336.587
    },
    {
      "epoch": 0.45246431457042824,
      "grad_norm": 0.6449802389017947,
      "learning_rate": 8.789776485882601e-06,
      "loss": 0.39746367931365967,
      "num_input_tokens_seen": 345792928,
      "step": 6720,
      "train_runtime": 41480.4846,
      "train_tokens_per_second": 8336.28
    },
    {
      "epoch": 0.4528009695663884,
      "grad_norm": 0.7320374585704675,
      "learning_rate": 8.788051204092112e-06,
      "loss": 0.37083072662353517,
      "num_input_tokens_seen": 346059816,
      "step": 6725,
      "train_runtime": 41513.4294,
      "train_tokens_per_second": 8336.093
    },
    {
      "epoch": 0.4531376245623485,
      "grad_norm": 0.824724733299454,
      "learning_rate": 8.786324862985183e-06,
      "loss": 0.3817046403884888,
      "num_input_tokens_seen": 346313480,
      "step": 6730,
      "train_runtime": 41546.4052,
      "train_tokens_per_second": 8335.582
    },
    {
      "epoch": 0.45347427955830866,
      "grad_norm": 0.7613473010667632,
      "learning_rate": 8.78459746304458e-06,
      "loss": 0.3658550024032593,
      "num_input_tokens_seen": 346500840,
      "step": 6735,
      "train_runtime": 41569.8195,
      "train_tokens_per_second": 8335.394
    },
    {
      "epoch": 0.4538109345542688,
      "grad_norm": 0.7698646185341481,
      "learning_rate": 8.782869004753363e-06,
      "loss": 0.38352031707763673,
      "num_input_tokens_seen": 346755424,
      "step": 6740,
      "train_runtime": 41601.021,
      "train_tokens_per_second": 8335.262
    },
    {
      "epoch": 0.45414758955022894,
      "grad_norm": 0.6270075963241716,
      "learning_rate": 8.781139488594892e-06,
      "loss": 0.38442516326904297,
      "num_input_tokens_seen": 347030712,
      "step": 6745,
      "train_runtime": 41632.3134,
      "train_tokens_per_second": 8335.61
    },
    {
      "epoch": 0.4544842445461891,
      "grad_norm": 0.6261836736422722,
      "learning_rate": 8.779408915052821e-06,
      "loss": 0.346458888053894,
      "num_input_tokens_seen": 347274952,
      "step": 6750,
      "train_runtime": 41661.6909,
      "train_tokens_per_second": 8335.594
    },
    {
      "epoch": 0.4548208995421492,
      "grad_norm": 0.6466146327014876,
      "learning_rate": 8.777677284611096e-06,
      "loss": 0.35758161544799805,
      "num_input_tokens_seen": 347501416,
      "step": 6755,
      "train_runtime": 41690.1048,
      "train_tokens_per_second": 8335.345
    },
    {
      "epoch": 0.45515755453810935,
      "grad_norm": 0.7522207888180874,
      "learning_rate": 8.775944597753969e-06,
      "loss": 0.3777958869934082,
      "num_input_tokens_seen": 347730112,
      "step": 6760,
      "train_runtime": 41721.5154,
      "train_tokens_per_second": 8334.551
    },
    {
      "epoch": 0.4554942095340695,
      "grad_norm": 0.7030893959883115,
      "learning_rate": 8.774210854965972e-06,
      "loss": 0.40615367889404297,
      "num_input_tokens_seen": 347978296,
      "step": 6765,
      "train_runtime": 41752.1946,
      "train_tokens_per_second": 8334.371
    },
    {
      "epoch": 0.45583086453002963,
      "grad_norm": 0.709164772024966,
      "learning_rate": 8.772476056731946e-06,
      "loss": 0.349475622177124,
      "num_input_tokens_seen": 348243328,
      "step": 6770,
      "train_runtime": 41781.3436,
      "train_tokens_per_second": 8334.9
    },
    {
      "epoch": 0.4561675195259898,
      "grad_norm": 0.68375606008556,
      "learning_rate": 8.77074020353702e-06,
      "loss": 0.36742286682128905,
      "num_input_tokens_seen": 348478744,
      "step": 6775,
      "train_runtime": 41811.8689,
      "train_tokens_per_second": 8334.446
    },
    {
      "epoch": 0.4565041745219499,
      "grad_norm": 0.808633646505253,
      "learning_rate": 8.76900329586662e-06,
      "loss": 0.4133578300476074,
      "num_input_tokens_seen": 348737888,
      "step": 6780,
      "train_runtime": 41846.0131,
      "train_tokens_per_second": 8333.838
    },
    {
      "epoch": 0.45684082951791005,
      "grad_norm": 0.6793254850132089,
      "learning_rate": 8.767265334206467e-06,
      "loss": 0.37294807434082033,
      "num_input_tokens_seen": 348998904,
      "step": 6785,
      "train_runtime": 41873.0853,
      "train_tokens_per_second": 8334.683
    },
    {
      "epoch": 0.4571774845138702,
      "grad_norm": 0.6033274579749357,
      "learning_rate": 8.765526319042577e-06,
      "loss": 0.3662102460861206,
      "num_input_tokens_seen": 349273016,
      "step": 6790,
      "train_runtime": 41907.167,
      "train_tokens_per_second": 8334.446
    },
    {
      "epoch": 0.45751413950983033,
      "grad_norm": 0.7587084872116829,
      "learning_rate": 8.763786250861258e-06,
      "loss": 0.35621516704559325,
      "num_input_tokens_seen": 349533528,
      "step": 6795,
      "train_runtime": 41937.6939,
      "train_tokens_per_second": 8334.591
    },
    {
      "epoch": 0.45785079450579047,
      "grad_norm": 0.697261697972445,
      "learning_rate": 8.762045130149114e-06,
      "loss": 0.3794650316238403,
      "num_input_tokens_seen": 349804688,
      "step": 6800,
      "train_runtime": 41972.5767,
      "train_tokens_per_second": 8334.125
    },
    {
      "epoch": 0.4581874495017506,
      "grad_norm": 0.6169009486826897,
      "learning_rate": 8.76030295739305e-06,
      "loss": 0.3692784786224365,
      "num_input_tokens_seen": 350079056,
      "step": 6805,
      "train_runtime": 42002.0054,
      "train_tokens_per_second": 8334.818
    },
    {
      "epoch": 0.45852410449771075,
      "grad_norm": 0.6615072396125007,
      "learning_rate": 8.758559733080252e-06,
      "loss": 0.39188737869262696,
      "num_input_tokens_seen": 350348336,
      "step": 6810,
      "train_runtime": 42032.8127,
      "train_tokens_per_second": 8335.115
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.6556552055025652,
      "learning_rate": 8.75681545769821e-06,
      "loss": 0.3802807807922363,
      "num_input_tokens_seen": 350625968,
      "step": 6815,
      "train_runtime": 42068.086,
      "train_tokens_per_second": 8334.726
    },
    {
      "epoch": 0.45919741448963103,
      "grad_norm": 0.7481504322273546,
      "learning_rate": 8.755070131734706e-06,
      "loss": 0.36862850189208984,
      "num_input_tokens_seen": 350873000,
      "step": 6820,
      "train_runtime": 42095.0522,
      "train_tokens_per_second": 8335.255
    },
    {
      "epoch": 0.45953406948559117,
      "grad_norm": 0.7553106326580985,
      "learning_rate": 8.753323755677812e-06,
      "loss": 0.3972615242004395,
      "num_input_tokens_seen": 351116520,
      "step": 6825,
      "train_runtime": 42125.2004,
      "train_tokens_per_second": 8335.071
    },
    {
      "epoch": 0.4598707244815513,
      "grad_norm": 0.7144172673794534,
      "learning_rate": 8.7515763300159e-06,
      "loss": 0.33542044162750245,
      "num_input_tokens_seen": 351362072,
      "step": 6830,
      "train_runtime": 42153.6254,
      "train_tokens_per_second": 8335.275
    },
    {
      "epoch": 0.46020737947751145,
      "grad_norm": 0.6486466790981725,
      "learning_rate": 8.74982785523763e-06,
      "loss": 0.3910716772079468,
      "num_input_tokens_seen": 351631552,
      "step": 6835,
      "train_runtime": 42183.069,
      "train_tokens_per_second": 8335.846
    },
    {
      "epoch": 0.4605440344734716,
      "grad_norm": 0.6582651678431258,
      "learning_rate": 8.748078331831957e-06,
      "loss": 0.3935364007949829,
      "num_input_tokens_seen": 351900168,
      "step": 6840,
      "train_runtime": 42217.1011,
      "train_tokens_per_second": 8335.489
    },
    {
      "epoch": 0.4608806894694317,
      "grad_norm": 0.6289507966898064,
      "learning_rate": 8.746327760288129e-06,
      "loss": 0.3724035739898682,
      "num_input_tokens_seen": 352195872,
      "step": 6845,
      "train_runtime": 42250.2329,
      "train_tokens_per_second": 8335.951
    },
    {
      "epoch": 0.46121734446539187,
      "grad_norm": 0.8052923739594976,
      "learning_rate": 8.744576141095691e-06,
      "loss": 0.38388452529907224,
      "num_input_tokens_seen": 352429640,
      "step": 6850,
      "train_runtime": 42278.4945,
      "train_tokens_per_second": 8335.908
    },
    {
      "epoch": 0.461553999461352,
      "grad_norm": 1.2647066224244488,
      "learning_rate": 8.742823474744476e-06,
      "loss": 0.398072624206543,
      "num_input_tokens_seen": 352673944,
      "step": 6855,
      "train_runtime": 42307.7833,
      "train_tokens_per_second": 8335.912
    },
    {
      "epoch": 0.46189065445731214,
      "grad_norm": 0.7221450921143207,
      "learning_rate": 8.74106976172461e-06,
      "loss": 0.3944756031036377,
      "num_input_tokens_seen": 352915416,
      "step": 6860,
      "train_runtime": 42340.6722,
      "train_tokens_per_second": 8335.14
    },
    {
      "epoch": 0.4622273094532723,
      "grad_norm": 1.0294267107284536,
      "learning_rate": 8.739315002526513e-06,
      "loss": 0.3851569175720215,
      "num_input_tokens_seen": 353170736,
      "step": 6865,
      "train_runtime": 42366.0182,
      "train_tokens_per_second": 8336.18
    },
    {
      "epoch": 0.4625639644492324,
      "grad_norm": 0.6342810131406622,
      "learning_rate": 8.737559197640902e-06,
      "loss": 0.3668498516082764,
      "num_input_tokens_seen": 353407912,
      "step": 6870,
      "train_runtime": 42400.3283,
      "train_tokens_per_second": 8335.028
    },
    {
      "epoch": 0.46290061944519256,
      "grad_norm": 0.6378887222992509,
      "learning_rate": 8.735802347558778e-06,
      "loss": 0.4022820472717285,
      "num_input_tokens_seen": 353650048,
      "step": 6875,
      "train_runtime": 42429.783,
      "train_tokens_per_second": 8334.948
    },
    {
      "epoch": 0.4632372744411527,
      "grad_norm": 0.6932672123939787,
      "learning_rate": 8.734044452771442e-06,
      "loss": 0.38662161827087405,
      "num_input_tokens_seen": 353895984,
      "step": 6880,
      "train_runtime": 42458.3824,
      "train_tokens_per_second": 8335.126
    },
    {
      "epoch": 0.46357392943711284,
      "grad_norm": 0.7834042028537866,
      "learning_rate": 8.73228551377048e-06,
      "loss": 0.40018477439880373,
      "num_input_tokens_seen": 354155024,
      "step": 6885,
      "train_runtime": 42489.8698,
      "train_tokens_per_second": 8335.046
    },
    {
      "epoch": 0.463910584433073,
      "grad_norm": 0.6291703163020351,
      "learning_rate": 8.730525531047776e-06,
      "loss": 0.40701723098754883,
      "num_input_tokens_seen": 354406256,
      "step": 6890,
      "train_runtime": 42520.4704,
      "train_tokens_per_second": 8334.956
    },
    {
      "epoch": 0.4642472394290331,
      "grad_norm": 0.6689345610626298,
      "learning_rate": 8.728764505095503e-06,
      "loss": 0.4083317756652832,
      "num_input_tokens_seen": 354676184,
      "step": 6895,
      "train_runtime": 42552.3611,
      "train_tokens_per_second": 8335.053
    },
    {
      "epoch": 0.46458389442499326,
      "grad_norm": 0.5770732814038618,
      "learning_rate": 8.727002436406126e-06,
      "loss": 0.359537410736084,
      "num_input_tokens_seen": 354963992,
      "step": 6900,
      "train_runtime": 42583.4026,
      "train_tokens_per_second": 8335.736
    },
    {
      "epoch": 0.4649205494209534,
      "grad_norm": 0.6442853521469436,
      "learning_rate": 8.725239325472404e-06,
      "loss": 0.40368995666503904,
      "num_input_tokens_seen": 355219200,
      "step": 6905,
      "train_runtime": 42616.8322,
      "train_tokens_per_second": 8335.185
    },
    {
      "epoch": 0.46525720441691354,
      "grad_norm": 0.6955463225616947,
      "learning_rate": 8.723475172787381e-06,
      "loss": 0.3772170066833496,
      "num_input_tokens_seen": 355484768,
      "step": 6910,
      "train_runtime": 42644.9033,
      "train_tokens_per_second": 8335.926
    },
    {
      "epoch": 0.4655938594128737,
      "grad_norm": 0.8748280962359676,
      "learning_rate": 8.7217099788444e-06,
      "loss": 0.37456018924713136,
      "num_input_tokens_seen": 355734944,
      "step": 6915,
      "train_runtime": 42673.8761,
      "train_tokens_per_second": 8336.129
    },
    {
      "epoch": 0.4659305144088338,
      "grad_norm": 0.648597875348657,
      "learning_rate": 8.719943744137091e-06,
      "loss": 0.37297959327697755,
      "num_input_tokens_seen": 355993064,
      "step": 6920,
      "train_runtime": 42704.986,
      "train_tokens_per_second": 8336.101
    },
    {
      "epoch": 0.46626716940479396,
      "grad_norm": 0.6251806952968482,
      "learning_rate": 8.718176469159378e-06,
      "loss": 0.3834844589233398,
      "num_input_tokens_seen": 356290824,
      "step": 6925,
      "train_runtime": 42735.6437,
      "train_tokens_per_second": 8337.088
    },
    {
      "epoch": 0.4666038244007541,
      "grad_norm": 0.6541556929414207,
      "learning_rate": 8.716408154405469e-06,
      "loss": 0.36685054302215575,
      "num_input_tokens_seen": 356569296,
      "step": 6930,
      "train_runtime": 42769.2145,
      "train_tokens_per_second": 8337.055
    },
    {
      "epoch": 0.46694047939671424,
      "grad_norm": 0.8505321972898082,
      "learning_rate": 8.714638800369872e-06,
      "loss": 0.4088714599609375,
      "num_input_tokens_seen": 356808224,
      "step": 6935,
      "train_runtime": 42795.0502,
      "train_tokens_per_second": 8337.605
    },
    {
      "epoch": 0.4672771343926744,
      "grad_norm": 0.6760130635369467,
      "learning_rate": 8.712868407547378e-06,
      "loss": 0.40453472137451174,
      "num_input_tokens_seen": 357085280,
      "step": 6940,
      "train_runtime": 42829.5552,
      "train_tokens_per_second": 8337.357
    },
    {
      "epoch": 0.4676137893886345,
      "grad_norm": 0.6472106574730438,
      "learning_rate": 8.711096976433076e-06,
      "loss": 0.3836667060852051,
      "num_input_tokens_seen": 357343592,
      "step": 6945,
      "train_runtime": 42861.739,
      "train_tokens_per_second": 8337.123
    },
    {
      "epoch": 0.46795044438459465,
      "grad_norm": 0.5863176735286573,
      "learning_rate": 8.709324507522337e-06,
      "loss": 0.35967581272125243,
      "num_input_tokens_seen": 357607096,
      "step": 6950,
      "train_runtime": 42895.6897,
      "train_tokens_per_second": 8336.667
    },
    {
      "epoch": 0.4682870993805548,
      "grad_norm": 0.6255100457314802,
      "learning_rate": 8.707551001310828e-06,
      "loss": 0.37271409034729003,
      "num_input_tokens_seen": 357880296,
      "step": 6955,
      "train_runtime": 42932.138,
      "train_tokens_per_second": 8335.953
    },
    {
      "epoch": 0.46862375437651493,
      "grad_norm": 0.6458108749666884,
      "learning_rate": 8.705776458294503e-06,
      "loss": 0.384765625,
      "num_input_tokens_seen": 358118560,
      "step": 6960,
      "train_runtime": 42961.2622,
      "train_tokens_per_second": 8335.848
    },
    {
      "epoch": 0.4689604093724751,
      "grad_norm": 0.6664742928865478,
      "learning_rate": 8.704000878969608e-06,
      "loss": 0.3607028484344482,
      "num_input_tokens_seen": 358346184,
      "step": 6965,
      "train_runtime": 42991.1062,
      "train_tokens_per_second": 8335.356
    },
    {
      "epoch": 0.4692970643684352,
      "grad_norm": 0.6636827288885906,
      "learning_rate": 8.702224263832679e-06,
      "loss": 0.3645676612854004,
      "num_input_tokens_seen": 358612416,
      "step": 6970,
      "train_runtime": 43021.1985,
      "train_tokens_per_second": 8335.714
    },
    {
      "epoch": 0.46963371936439535,
      "grad_norm": 0.6635412479011019,
      "learning_rate": 8.700446613380542e-06,
      "loss": 0.4147918701171875,
      "num_input_tokens_seen": 358886320,
      "step": 6975,
      "train_runtime": 43052.1785,
      "train_tokens_per_second": 8336.078
    },
    {
      "epoch": 0.4699703743603555,
      "grad_norm": 0.6426454797930109,
      "learning_rate": 8.698667928110307e-06,
      "loss": 0.36252119541168215,
      "num_input_tokens_seen": 359139064,
      "step": 6980,
      "train_runtime": 43084.0529,
      "train_tokens_per_second": 8335.777
    },
    {
      "epoch": 0.47030702935631563,
      "grad_norm": 0.8595407213704154,
      "learning_rate": 8.696888208519381e-06,
      "loss": 0.3871385812759399,
      "num_input_tokens_seen": 359389984,
      "step": 6985,
      "train_runtime": 43114.072,
      "train_tokens_per_second": 8335.793
    },
    {
      "epoch": 0.47064368435227577,
      "grad_norm": 0.8792781429777266,
      "learning_rate": 8.695107455105454e-06,
      "loss": 0.3343928098678589,
      "num_input_tokens_seen": 359635688,
      "step": 6990,
      "train_runtime": 43145.0843,
      "train_tokens_per_second": 8335.496
    },
    {
      "epoch": 0.4709803393482359,
      "grad_norm": 0.6227052731150478,
      "learning_rate": 8.693325668366513e-06,
      "loss": 0.38829870223999025,
      "num_input_tokens_seen": 359891208,
      "step": 6995,
      "train_runtime": 43171.4898,
      "train_tokens_per_second": 8336.317
    },
    {
      "epoch": 0.47131699434419605,
      "grad_norm": 0.7596993950292009,
      "learning_rate": 8.691542848800825e-06,
      "loss": 0.3879296541213989,
      "num_input_tokens_seen": 360129632,
      "step": 7000,
      "train_runtime": 43198.7004,
      "train_tokens_per_second": 8336.585
    },
    {
      "epoch": 0.4716536493401562,
      "grad_norm": 0.7173372692724945,
      "learning_rate": 8.68975899690695e-06,
      "loss": 0.37775793075561526,
      "num_input_tokens_seen": 360396320,
      "step": 7005,
      "train_runtime": 43227.8831,
      "train_tokens_per_second": 8337.126
    },
    {
      "epoch": 0.47199030433611633,
      "grad_norm": 0.755711507527088,
      "learning_rate": 8.687974113183738e-06,
      "loss": 0.3902088403701782,
      "num_input_tokens_seen": 360673312,
      "step": 7010,
      "train_runtime": 43258.7331,
      "train_tokens_per_second": 8337.584
    },
    {
      "epoch": 0.47232695933207647,
      "grad_norm": 0.765132577314007,
      "learning_rate": 8.686188198130326e-06,
      "loss": 0.3805851459503174,
      "num_input_tokens_seen": 360940872,
      "step": 7015,
      "train_runtime": 43292.7718,
      "train_tokens_per_second": 8337.209
    },
    {
      "epoch": 0.4726636143280366,
      "grad_norm": 0.6369271784476171,
      "learning_rate": 8.684401252246138e-06,
      "loss": 0.3950699806213379,
      "num_input_tokens_seen": 361199656,
      "step": 7020,
      "train_runtime": 43325.1896,
      "train_tokens_per_second": 8336.943
    },
    {
      "epoch": 0.47300026932399675,
      "grad_norm": 0.7162935870097515,
      "learning_rate": 8.68261327603089e-06,
      "loss": 0.4119518280029297,
      "num_input_tokens_seen": 361438264,
      "step": 7025,
      "train_runtime": 43356.2846,
      "train_tokens_per_second": 8336.468
    },
    {
      "epoch": 0.4733369243199569,
      "grad_norm": 0.6514937507856972,
      "learning_rate": 8.68082426998458e-06,
      "loss": 0.4030905246734619,
      "num_input_tokens_seen": 361716112,
      "step": 7030,
      "train_runtime": 43384.8654,
      "train_tokens_per_second": 8337.38
    },
    {
      "epoch": 0.473673579315917,
      "grad_norm": 0.8004663937268449,
      "learning_rate": 8.6790342346075e-06,
      "loss": 0.35478484630584717,
      "num_input_tokens_seen": 361936624,
      "step": 7035,
      "train_runtime": 43413.4818,
      "train_tokens_per_second": 8336.964
    },
    {
      "epoch": 0.47401023431187717,
      "grad_norm": 0.6502643899529759,
      "learning_rate": 8.677243170400228e-06,
      "loss": 0.375750732421875,
      "num_input_tokens_seen": 362169304,
      "step": 7040,
      "train_runtime": 43443.3662,
      "train_tokens_per_second": 8336.585
    },
    {
      "epoch": 0.4743468893078373,
      "grad_norm": 0.698685118685828,
      "learning_rate": 8.675451077863632e-06,
      "loss": 0.40714235305786134,
      "num_input_tokens_seen": 362429984,
      "step": 7045,
      "train_runtime": 43477.214,
      "train_tokens_per_second": 8336.09
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.65153247613981,
      "learning_rate": 8.67365795749886e-06,
      "loss": 0.36104259490966795,
      "num_input_tokens_seen": 362678368,
      "step": 7050,
      "train_runtime": 43506.5518,
      "train_tokens_per_second": 8336.178
    },
    {
      "epoch": 0.4750201992997576,
      "grad_norm": 0.6433706861921489,
      "learning_rate": 8.671863809807355e-06,
      "loss": 0.3644951581954956,
      "num_input_tokens_seen": 362916048,
      "step": 7055,
      "train_runtime": 43538.3237,
      "train_tokens_per_second": 8335.554
    },
    {
      "epoch": 0.4753568542957177,
      "grad_norm": 0.587156408303983,
      "learning_rate": 8.670068635290844e-06,
      "loss": 0.36927003860473634,
      "num_input_tokens_seen": 363169776,
      "step": 7060,
      "train_runtime": 43565.7184,
      "train_tokens_per_second": 8336.137
    },
    {
      "epoch": 0.47569350929167786,
      "grad_norm": 3.800129935336739,
      "learning_rate": 8.668272434451343e-06,
      "loss": 0.3495290279388428,
      "num_input_tokens_seen": 363426680,
      "step": 7065,
      "train_runtime": 43596.4406,
      "train_tokens_per_second": 8336.155
    },
    {
      "epoch": 0.476030164287638,
      "grad_norm": 0.6732751435341199,
      "learning_rate": 8.666475207791154e-06,
      "loss": 0.3936310291290283,
      "num_input_tokens_seen": 363662904,
      "step": 7070,
      "train_runtime": 43629.2685,
      "train_tokens_per_second": 8335.297
    },
    {
      "epoch": 0.47636681928359814,
      "grad_norm": 0.6170377227038959,
      "learning_rate": 8.664676955812863e-06,
      "loss": 0.36451971530914307,
      "num_input_tokens_seen": 363925408,
      "step": 7075,
      "train_runtime": 43662.8985,
      "train_tokens_per_second": 8334.889
    },
    {
      "epoch": 0.4767034742795583,
      "grad_norm": 0.7097812855137908,
      "learning_rate": 8.662877679019348e-06,
      "loss": 0.37722671031951904,
      "num_input_tokens_seen": 364178992,
      "step": 7080,
      "train_runtime": 43696.2177,
      "train_tokens_per_second": 8334.337
    },
    {
      "epoch": 0.4770401292755185,
      "grad_norm": 0.6382715763143733,
      "learning_rate": 8.66107737791377e-06,
      "loss": 0.3828958749771118,
      "num_input_tokens_seen": 364464784,
      "step": 7085,
      "train_runtime": 43729.9016,
      "train_tokens_per_second": 8334.452
    },
    {
      "epoch": 0.4773767842714786,
      "grad_norm": 0.7891566250158142,
      "learning_rate": 8.659276052999574e-06,
      "loss": 0.3672476053237915,
      "num_input_tokens_seen": 364688632,
      "step": 7090,
      "train_runtime": 43760.8815,
      "train_tokens_per_second": 8333.667
    },
    {
      "epoch": 0.47771343926743876,
      "grad_norm": 0.5816966767272904,
      "learning_rate": 8.657473704780504e-06,
      "loss": 0.3891436576843262,
      "num_input_tokens_seen": 364935200,
      "step": 7095,
      "train_runtime": 43788.4661,
      "train_tokens_per_second": 8334.049
    },
    {
      "epoch": 0.4780500942633989,
      "grad_norm": 0.6552141471842836,
      "learning_rate": 8.65567033376057e-06,
      "loss": 0.37781486511230467,
      "num_input_tokens_seen": 365205656,
      "step": 7100,
      "train_runtime": 43816.0644,
      "train_tokens_per_second": 8334.972
    },
    {
      "epoch": 0.47838674925935903,
      "grad_norm": 0.5963672233416513,
      "learning_rate": 8.65386594044409e-06,
      "loss": 0.3822510242462158,
      "num_input_tokens_seen": 365472440,
      "step": 7105,
      "train_runtime": 43847.6458,
      "train_tokens_per_second": 8335.053
    },
    {
      "epoch": 0.4787234042553192,
      "grad_norm": 0.6626510145868704,
      "learning_rate": 8.652060525335647e-06,
      "loss": 0.357573413848877,
      "num_input_tokens_seen": 365715128,
      "step": 7110,
      "train_runtime": 43877.8818,
      "train_tokens_per_second": 8334.84
    },
    {
      "epoch": 0.4790600592512793,
      "grad_norm": 0.6264217637796095,
      "learning_rate": 8.650254088940125e-06,
      "loss": 0.37399978637695314,
      "num_input_tokens_seen": 365957208,
      "step": 7115,
      "train_runtime": 43905.8754,
      "train_tokens_per_second": 8335.04
    },
    {
      "epoch": 0.47939671424723945,
      "grad_norm": 0.7220979263164878,
      "learning_rate": 8.648446631762686e-06,
      "loss": 0.378862190246582,
      "num_input_tokens_seen": 366191112,
      "step": 7120,
      "train_runtime": 43937.7624,
      "train_tokens_per_second": 8334.314
    },
    {
      "epoch": 0.4797333692431996,
      "grad_norm": 0.617734412398038,
      "learning_rate": 8.646638154308781e-06,
      "loss": 0.35373744964599607,
      "num_input_tokens_seen": 366449600,
      "step": 7125,
      "train_runtime": 43964.8522,
      "train_tokens_per_second": 8335.058
    },
    {
      "epoch": 0.48007002423915973,
      "grad_norm": 0.662779018077331,
      "learning_rate": 8.644828657084144e-06,
      "loss": 0.34715306758880615,
      "num_input_tokens_seen": 366715336,
      "step": 7130,
      "train_runtime": 43996.1973,
      "train_tokens_per_second": 8335.16
    },
    {
      "epoch": 0.48040667923511987,
      "grad_norm": 0.6783578754395524,
      "learning_rate": 8.643018140594797e-06,
      "loss": 0.3588275909423828,
      "num_input_tokens_seen": 366974544,
      "step": 7135,
      "train_runtime": 44030.8621,
      "train_tokens_per_second": 8334.485
    },
    {
      "epoch": 0.48074333423108,
      "grad_norm": 0.701723499274208,
      "learning_rate": 8.641206605347042e-06,
      "loss": 0.3538167953491211,
      "num_input_tokens_seen": 367221944,
      "step": 7140,
      "train_runtime": 44061.0696,
      "train_tokens_per_second": 8334.386
    },
    {
      "epoch": 0.48107998922704015,
      "grad_norm": 0.7592060233439405,
      "learning_rate": 8.639394051847472e-06,
      "loss": 0.3654845952987671,
      "num_input_tokens_seen": 367478088,
      "step": 7145,
      "train_runtime": 44091.4134,
      "train_tokens_per_second": 8334.459
    },
    {
      "epoch": 0.4814166442230003,
      "grad_norm": 0.79714612515017,
      "learning_rate": 8.63758048060296e-06,
      "loss": 0.3883105754852295,
      "num_input_tokens_seen": 367727384,
      "step": 7150,
      "train_runtime": 44123.3439,
      "train_tokens_per_second": 8334.078
    },
    {
      "epoch": 0.48175329921896043,
      "grad_norm": 0.6553316953278224,
      "learning_rate": 8.635765892120666e-06,
      "loss": 0.35214505195617674,
      "num_input_tokens_seen": 368018272,
      "step": 7155,
      "train_runtime": 44158.3524,
      "train_tokens_per_second": 8334.058
    },
    {
      "epoch": 0.48208995421492057,
      "grad_norm": 0.8069646865453213,
      "learning_rate": 8.633950286908032e-06,
      "loss": 0.39485721588134765,
      "num_input_tokens_seen": 368265776,
      "step": 7160,
      "train_runtime": 44185.7769,
      "train_tokens_per_second": 8334.487
    },
    {
      "epoch": 0.4824266092108807,
      "grad_norm": 0.7317292165844845,
      "learning_rate": 8.63213366547279e-06,
      "loss": 0.4142289161682129,
      "num_input_tokens_seen": 368511848,
      "step": 7165,
      "train_runtime": 44215.968,
      "train_tokens_per_second": 8334.361
    },
    {
      "epoch": 0.48276326420684085,
      "grad_norm": 0.742700344048468,
      "learning_rate": 8.630316028322952e-06,
      "loss": 0.3538456678390503,
      "num_input_tokens_seen": 368739960,
      "step": 7170,
      "train_runtime": 44241.6831,
      "train_tokens_per_second": 8334.673
    },
    {
      "epoch": 0.483099919202801,
      "grad_norm": 0.7829087157397598,
      "learning_rate": 8.628497375966811e-06,
      "loss": 0.3596886396408081,
      "num_input_tokens_seen": 369015872,
      "step": 7175,
      "train_runtime": 44273.4389,
      "train_tokens_per_second": 8334.927
    },
    {
      "epoch": 0.4834365741987611,
      "grad_norm": 0.6636265527524763,
      "learning_rate": 8.62667770891295e-06,
      "loss": 0.41170992851257326,
      "num_input_tokens_seen": 369291288,
      "step": 7180,
      "train_runtime": 44304.4965,
      "train_tokens_per_second": 8335.3
    },
    {
      "epoch": 0.48377322919472127,
      "grad_norm": 0.6053737008594526,
      "learning_rate": 8.624857027670232e-06,
      "loss": 0.36624870300292967,
      "num_input_tokens_seen": 369556512,
      "step": 7185,
      "train_runtime": 44334.9949,
      "train_tokens_per_second": 8335.549
    },
    {
      "epoch": 0.4841098841906814,
      "grad_norm": 0.6983319237137858,
      "learning_rate": 8.623035332747804e-06,
      "loss": 0.41037893295288086,
      "num_input_tokens_seen": 369785480,
      "step": 7190,
      "train_runtime": 44361.2372,
      "train_tokens_per_second": 8335.779
    },
    {
      "epoch": 0.48444653918664154,
      "grad_norm": 0.7827375766579721,
      "learning_rate": 8.6212126246551e-06,
      "loss": 0.35644211769104006,
      "num_input_tokens_seen": 370022144,
      "step": 7195,
      "train_runtime": 44393.1221,
      "train_tokens_per_second": 8335.123
    },
    {
      "epoch": 0.4847831941826017,
      "grad_norm": 0.6921063587723316,
      "learning_rate": 8.619388903901833e-06,
      "loss": 0.39018726348876953,
      "num_input_tokens_seen": 370294032,
      "step": 7200,
      "train_runtime": 44424.4346,
      "train_tokens_per_second": 8335.368
    },
    {
      "epoch": 0.4851198491785618,
      "grad_norm": 0.6475624706366795,
      "learning_rate": 8.617564170997998e-06,
      "loss": 0.3809091091156006,
      "num_input_tokens_seen": 370570080,
      "step": 7205,
      "train_runtime": 44455.6992,
      "train_tokens_per_second": 8335.716
    },
    {
      "epoch": 0.48545650417452196,
      "grad_norm": 0.6454407286981777,
      "learning_rate": 8.61573842645388e-06,
      "loss": 0.3847665309906006,
      "num_input_tokens_seen": 370844472,
      "step": 7210,
      "train_runtime": 44486.3518,
      "train_tokens_per_second": 8336.14
    },
    {
      "epoch": 0.4857931591704821,
      "grad_norm": 0.6942739469824448,
      "learning_rate": 8.61391167078004e-06,
      "loss": 0.36765251159667967,
      "num_input_tokens_seen": 371114224,
      "step": 7215,
      "train_runtime": 44519.0325,
      "train_tokens_per_second": 8336.08
    },
    {
      "epoch": 0.48612981416644224,
      "grad_norm": 0.6394194513114062,
      "learning_rate": 8.612083904487324e-06,
      "loss": 0.4097172260284424,
      "num_input_tokens_seen": 371373072,
      "step": 7220,
      "train_runtime": 44549.2841,
      "train_tokens_per_second": 8336.23
    },
    {
      "epoch": 0.4864664691624024,
      "grad_norm": 0.6827892752778869,
      "learning_rate": 8.610255128086867e-06,
      "loss": 0.39255073070526125,
      "num_input_tokens_seen": 371617312,
      "step": 7225,
      "train_runtime": 44576.7518,
      "train_tokens_per_second": 8336.572
    },
    {
      "epoch": 0.4868031241583625,
      "grad_norm": 0.6605325118657461,
      "learning_rate": 8.60842534209007e-06,
      "loss": 0.3467274188995361,
      "num_input_tokens_seen": 371873256,
      "step": 7230,
      "train_runtime": 44609.3643,
      "train_tokens_per_second": 8336.215
    },
    {
      "epoch": 0.48713977915432266,
      "grad_norm": 0.6866202208139071,
      "learning_rate": 8.606594547008636e-06,
      "loss": 0.33203327655792236,
      "num_input_tokens_seen": 372142584,
      "step": 7235,
      "train_runtime": 44635.7627,
      "train_tokens_per_second": 8337.319
    },
    {
      "epoch": 0.4874764341502828,
      "grad_norm": 0.6337239469032425,
      "learning_rate": 8.604762743354536e-06,
      "loss": 0.39234261512756347,
      "num_input_tokens_seen": 372431424,
      "step": 7240,
      "train_runtime": 44667.8128,
      "train_tokens_per_second": 8337.803
    },
    {
      "epoch": 0.48781308914624294,
      "grad_norm": 0.7083836099191185,
      "learning_rate": 8.602929931640029e-06,
      "loss": 0.3903968095779419,
      "num_input_tokens_seen": 372663064,
      "step": 7245,
      "train_runtime": 44698.9111,
      "train_tokens_per_second": 8337.184
    },
    {
      "epoch": 0.4881497441422031,
      "grad_norm": 0.7310951143289504,
      "learning_rate": 8.601096112377659e-06,
      "loss": 0.3911155700683594,
      "num_input_tokens_seen": 372912512,
      "step": 7250,
      "train_runtime": 44731.2088,
      "train_tokens_per_second": 8336.741
    },
    {
      "epoch": 0.4884863991381632,
      "grad_norm": 0.5633688823845033,
      "learning_rate": 8.59926128608024e-06,
      "loss": 0.35543341636657716,
      "num_input_tokens_seen": 373175448,
      "step": 7255,
      "train_runtime": 44762.6015,
      "train_tokens_per_second": 8336.769
    },
    {
      "epoch": 0.48882305413412336,
      "grad_norm": 0.600157740066565,
      "learning_rate": 8.597425453260881e-06,
      "loss": 0.3741034984588623,
      "num_input_tokens_seen": 373430432,
      "step": 7260,
      "train_runtime": 44792.6216,
      "train_tokens_per_second": 8336.874
    },
    {
      "epoch": 0.4891597091300835,
      "grad_norm": 0.6999546068117545,
      "learning_rate": 8.595588614432967e-06,
      "loss": 0.3597210168838501,
      "num_input_tokens_seen": 373655312,
      "step": 7265,
      "train_runtime": 44822.5469,
      "train_tokens_per_second": 8336.325
    },
    {
      "epoch": 0.48949636412604364,
      "grad_norm": 0.6859814859521575,
      "learning_rate": 8.59375077011016e-06,
      "loss": 0.368304967880249,
      "num_input_tokens_seen": 373926208,
      "step": 7270,
      "train_runtime": 44852.7265,
      "train_tokens_per_second": 8336.755
    },
    {
      "epoch": 0.4898330191220038,
      "grad_norm": 0.6574310567264673,
      "learning_rate": 8.591911920806412e-06,
      "loss": 0.3834777593612671,
      "num_input_tokens_seen": 374207296,
      "step": 7275,
      "train_runtime": 44881.6684,
      "train_tokens_per_second": 8337.642
    },
    {
      "epoch": 0.4901696741179639,
      "grad_norm": 0.5797504423876698,
      "learning_rate": 8.590072067035947e-06,
      "loss": 0.3819756507873535,
      "num_input_tokens_seen": 374487344,
      "step": 7280,
      "train_runtime": 44910.9542,
      "train_tokens_per_second": 8338.441
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.651393290687219,
      "learning_rate": 8.588231209313278e-06,
      "loss": 0.3650646686553955,
      "num_input_tokens_seen": 374787976,
      "step": 7285,
      "train_runtime": 44946.6497,
      "train_tokens_per_second": 8338.508
    },
    {
      "epoch": 0.4908429841098842,
      "grad_norm": 0.7246084159320156,
      "learning_rate": 8.586389348153192e-06,
      "loss": 0.3797840356826782,
      "num_input_tokens_seen": 375054496,
      "step": 7290,
      "train_runtime": 44976.6706,
      "train_tokens_per_second": 8338.867
    },
    {
      "epoch": 0.49117963910584433,
      "grad_norm": 0.694683525851556,
      "learning_rate": 8.584546484070762e-06,
      "loss": 0.4127194404602051,
      "num_input_tokens_seen": 375289256,
      "step": 7295,
      "train_runtime": 45005.8188,
      "train_tokens_per_second": 8338.683
    },
    {
      "epoch": 0.4915162941018045,
      "grad_norm": 0.7510627590758512,
      "learning_rate": 8.582702617581338e-06,
      "loss": 0.3785467863082886,
      "num_input_tokens_seen": 375544304,
      "step": 7300,
      "train_runtime": 45039.4538,
      "train_tokens_per_second": 8338.119
    },
    {
      "epoch": 0.4918529490977646,
      "grad_norm": 0.6349984899321119,
      "learning_rate": 8.58085774920055e-06,
      "loss": 0.37995409965515137,
      "num_input_tokens_seen": 375784416,
      "step": 7305,
      "train_runtime": 45074.6121,
      "train_tokens_per_second": 8336.942
    },
    {
      "epoch": 0.49218960409372475,
      "grad_norm": 0.7113053562047571,
      "learning_rate": 8.579011879444313e-06,
      "loss": 0.399746036529541,
      "num_input_tokens_seen": 376048688,
      "step": 7310,
      "train_runtime": 45106.5159,
      "train_tokens_per_second": 8336.904
    },
    {
      "epoch": 0.4925262590896849,
      "grad_norm": 0.6136347127369812,
      "learning_rate": 8.577165008828817e-06,
      "loss": 0.34152324199676515,
      "num_input_tokens_seen": 376274184,
      "step": 7315,
      "train_runtime": 45135.5372,
      "train_tokens_per_second": 8336.539
    },
    {
      "epoch": 0.49286291408564503,
      "grad_norm": 0.6566649540139976,
      "learning_rate": 8.57531713787053e-06,
      "loss": 0.38567614555358887,
      "num_input_tokens_seen": 376517872,
      "step": 7320,
      "train_runtime": 45169.1331,
      "train_tokens_per_second": 8335.734
    },
    {
      "epoch": 0.49319956908160517,
      "grad_norm": 0.703782903012243,
      "learning_rate": 8.573468267086208e-06,
      "loss": 0.39664392471313475,
      "num_input_tokens_seen": 376780320,
      "step": 7325,
      "train_runtime": 45200.2341,
      "train_tokens_per_second": 8335.805
    },
    {
      "epoch": 0.4935362240775653,
      "grad_norm": 0.7190517311546092,
      "learning_rate": 8.57161839699288e-06,
      "loss": 0.35924072265625,
      "num_input_tokens_seen": 377032088,
      "step": 7330,
      "train_runtime": 45232.448,
      "train_tokens_per_second": 8335.434
    },
    {
      "epoch": 0.49387287907352545,
      "grad_norm": 0.7335953508177087,
      "learning_rate": 8.569767528107857e-06,
      "loss": 0.37210080623626707,
      "num_input_tokens_seen": 377289192,
      "step": 7335,
      "train_runtime": 45261.6938,
      "train_tokens_per_second": 8335.729
    },
    {
      "epoch": 0.4942095340694856,
      "grad_norm": 0.5663040157443942,
      "learning_rate": 8.567915660948727e-06,
      "loss": 0.36352565288543703,
      "num_input_tokens_seen": 377558592,
      "step": 7340,
      "train_runtime": 45292.8939,
      "train_tokens_per_second": 8335.934
    },
    {
      "epoch": 0.49454618906544573,
      "grad_norm": 0.652104188975199,
      "learning_rate": 8.566062796033358e-06,
      "loss": 0.39132418632507326,
      "num_input_tokens_seen": 377807096,
      "step": 7345,
      "train_runtime": 45318.0867,
      "train_tokens_per_second": 8336.784
    },
    {
      "epoch": 0.49488284406140587,
      "grad_norm": 0.6367473383952845,
      "learning_rate": 8.564208933879903e-06,
      "loss": 0.3632572412490845,
      "num_input_tokens_seen": 378051672,
      "step": 7350,
      "train_runtime": 45348.5413,
      "train_tokens_per_second": 8336.578
    },
    {
      "epoch": 0.495219499057366,
      "grad_norm": 0.6459140751440573,
      "learning_rate": 8.562354075006782e-06,
      "loss": 0.3530298709869385,
      "num_input_tokens_seen": 378309808,
      "step": 7355,
      "train_runtime": 45377.2456,
      "train_tokens_per_second": 8336.994
    },
    {
      "epoch": 0.49555615405332615,
      "grad_norm": 0.5786974587477695,
      "learning_rate": 8.560498219932706e-06,
      "loss": 0.4171254634857178,
      "num_input_tokens_seen": 378577936,
      "step": 7360,
      "train_runtime": 45411.1916,
      "train_tokens_per_second": 8336.666
    },
    {
      "epoch": 0.4958928090492863,
      "grad_norm": 0.6782851800012041,
      "learning_rate": 8.558641369176654e-06,
      "loss": 0.3340815544128418,
      "num_input_tokens_seen": 378810400,
      "step": 7365,
      "train_runtime": 45442.1013,
      "train_tokens_per_second": 8336.111
    },
    {
      "epoch": 0.4962294640452464,
      "grad_norm": 0.6961640468763698,
      "learning_rate": 8.556783523257892e-06,
      "loss": 0.39092001914978025,
      "num_input_tokens_seen": 379049144,
      "step": 7370,
      "train_runtime": 45470.9897,
      "train_tokens_per_second": 8336.065
    },
    {
      "epoch": 0.49656611904120657,
      "grad_norm": 0.650177303195184,
      "learning_rate": 8.554924682695959e-06,
      "loss": 0.35159897804260254,
      "num_input_tokens_seen": 379290760,
      "step": 7375,
      "train_runtime": 45503.2043,
      "train_tokens_per_second": 8335.474
    },
    {
      "epoch": 0.4969027740371667,
      "grad_norm": 0.6996380881649468,
      "learning_rate": 8.553064848010677e-06,
      "loss": 0.37937912940979,
      "num_input_tokens_seen": 379535160,
      "step": 7380,
      "train_runtime": 45532.4035,
      "train_tokens_per_second": 8335.496
    },
    {
      "epoch": 0.49723942903312685,
      "grad_norm": 0.7311103668017179,
      "learning_rate": 8.551204019722137e-06,
      "loss": 0.37068610191345214,
      "num_input_tokens_seen": 379785960,
      "step": 7385,
      "train_runtime": 45557.2364,
      "train_tokens_per_second": 8336.457
    },
    {
      "epoch": 0.497576084029087,
      "grad_norm": 0.662318665405562,
      "learning_rate": 8.549342198350718e-06,
      "loss": 0.39316701889038086,
      "num_input_tokens_seen": 380043776,
      "step": 7390,
      "train_runtime": 45586.8133,
      "train_tokens_per_second": 8336.704
    },
    {
      "epoch": 0.4979127390250471,
      "grad_norm": 0.6434119566157755,
      "learning_rate": 8.54747938441707e-06,
      "loss": 0.35640218257904055,
      "num_input_tokens_seen": 380304040,
      "step": 7395,
      "train_runtime": 45614.2071,
      "train_tokens_per_second": 8337.403
    },
    {
      "epoch": 0.49824939402100726,
      "grad_norm": 0.6023259421672535,
      "learning_rate": 8.545615578442126e-06,
      "loss": 0.34729092121124266,
      "num_input_tokens_seen": 380554216,
      "step": 7400,
      "train_runtime": 45643.3992,
      "train_tokens_per_second": 8337.552
    },
    {
      "epoch": 0.4985860490169674,
      "grad_norm": 0.7145600260063468,
      "learning_rate": 8.54375078094709e-06,
      "loss": 0.36800098419189453,
      "num_input_tokens_seen": 380813904,
      "step": 7405,
      "train_runtime": 45672.2487,
      "train_tokens_per_second": 8337.971
    },
    {
      "epoch": 0.49892270401292754,
      "grad_norm": 0.5703411527498095,
      "learning_rate": 8.541884992453449e-06,
      "loss": 0.3637763738632202,
      "num_input_tokens_seen": 381085696,
      "step": 7410,
      "train_runtime": 45704.9933,
      "train_tokens_per_second": 8337.944
    },
    {
      "epoch": 0.4992593590088877,
      "grad_norm": 0.6201747984127052,
      "learning_rate": 8.540018213482966e-06,
      "loss": 0.36603221893310545,
      "num_input_tokens_seen": 381364424,
      "step": 7415,
      "train_runtime": 45737.1703,
      "train_tokens_per_second": 8338.173
    },
    {
      "epoch": 0.4995960140048478,
      "grad_norm": 0.6999738801249897,
      "learning_rate": 8.538150444557676e-06,
      "loss": 0.3900843381881714,
      "num_input_tokens_seen": 381600224,
      "step": 7420,
      "train_runtime": 45767.1463,
      "train_tokens_per_second": 8337.864
    },
    {
      "epoch": 0.49993266900080796,
      "grad_norm": 0.7310176806205884,
      "learning_rate": 8.536281686199896e-06,
      "loss": 0.38136217594146726,
      "num_input_tokens_seen": 381864088,
      "step": 7425,
      "train_runtime": 45795.9518,
      "train_tokens_per_second": 8338.381
    },
    {
      "epoch": 0.5002693239967682,
      "grad_norm": 0.6354380932699795,
      "learning_rate": 8.534411938932218e-06,
      "loss": 0.365736198425293,
      "num_input_tokens_seen": 382146176,
      "step": 7430,
      "train_runtime": 45824.9193,
      "train_tokens_per_second": 8339.266
    },
    {
      "epoch": 0.5006059789927283,
      "grad_norm": 0.9648955376258153,
      "learning_rate": 8.532541203277515e-06,
      "loss": 0.35866608619689944,
      "num_input_tokens_seen": 382401792,
      "step": 7435,
      "train_runtime": 45855.7549,
      "train_tokens_per_second": 8339.232
    },
    {
      "epoch": 0.5009426339886884,
      "grad_norm": 0.7578587282304725,
      "learning_rate": 8.530669479758926e-06,
      "loss": 0.3694236516952515,
      "num_input_tokens_seen": 382655288,
      "step": 7440,
      "train_runtime": 45883.1973,
      "train_tokens_per_second": 8339.769
    },
    {
      "epoch": 0.5012792889846486,
      "grad_norm": 0.6731682463369684,
      "learning_rate": 8.528796768899878e-06,
      "loss": 0.3731534004211426,
      "num_input_tokens_seen": 382894136,
      "step": 7445,
      "train_runtime": 45911.67,
      "train_tokens_per_second": 8339.8
    },
    {
      "epoch": 0.5016159439806087,
      "grad_norm": 0.6369235198739442,
      "learning_rate": 8.526923071224064e-06,
      "loss": 0.37650442123413086,
      "num_input_tokens_seen": 383175320,
      "step": 7450,
      "train_runtime": 45942.9732,
      "train_tokens_per_second": 8340.238
    },
    {
      "epoch": 0.5019525989765689,
      "grad_norm": 0.6611738406908919,
      "learning_rate": 8.525048387255461e-06,
      "loss": 0.3844642400741577,
      "num_input_tokens_seen": 383442624,
      "step": 7455,
      "train_runtime": 45971.8105,
      "train_tokens_per_second": 8340.821
    },
    {
      "epoch": 0.502289253972529,
      "grad_norm": 0.627617870302508,
      "learning_rate": 8.523172717518315e-06,
      "loss": 0.35893726348876953,
      "num_input_tokens_seen": 383688264,
      "step": 7460,
      "train_runtime": 46004.4474,
      "train_tokens_per_second": 8340.243
    },
    {
      "epoch": 0.5026259089684891,
      "grad_norm": 0.6047572654172919,
      "learning_rate": 8.521296062537156e-06,
      "loss": 0.3743009090423584,
      "num_input_tokens_seen": 383937280,
      "step": 7465,
      "train_runtime": 46035.4246,
      "train_tokens_per_second": 8340.04
    },
    {
      "epoch": 0.5029625639644493,
      "grad_norm": 0.7416501605131082,
      "learning_rate": 8.51941842283678e-06,
      "loss": 0.35652761459350585,
      "num_input_tokens_seen": 384188560,
      "step": 7470,
      "train_runtime": 46065.8292,
      "train_tokens_per_second": 8339.99
    },
    {
      "epoch": 0.5032992189604094,
      "grad_norm": 0.6447035424091886,
      "learning_rate": 8.517539798942265e-06,
      "loss": 0.3547863483428955,
      "num_input_tokens_seen": 384484344,
      "step": 7475,
      "train_runtime": 46093.5835,
      "train_tokens_per_second": 8341.385
    },
    {
      "epoch": 0.5036358739563696,
      "grad_norm": 0.5962421868883881,
      "learning_rate": 8.515660191378962e-06,
      "loss": 0.38811917304992677,
      "num_input_tokens_seen": 384744992,
      "step": 7480,
      "train_runtime": 46123.5969,
      "train_tokens_per_second": 8341.609
    },
    {
      "epoch": 0.5039725289523297,
      "grad_norm": 0.6147317133174536,
      "learning_rate": 8.513779600672495e-06,
      "loss": 0.35041046142578125,
      "num_input_tokens_seen": 385030040,
      "step": 7485,
      "train_runtime": 46159.6587,
      "train_tokens_per_second": 8341.267
    },
    {
      "epoch": 0.5043091839482898,
      "grad_norm": 0.6467805722666736,
      "learning_rate": 8.511898027348771e-06,
      "loss": 0.3693578243255615,
      "num_input_tokens_seen": 385269376,
      "step": 7490,
      "train_runtime": 46193.5052,
      "train_tokens_per_second": 8340.336
    },
    {
      "epoch": 0.50464583894425,
      "grad_norm": 0.6224611873935373,
      "learning_rate": 8.51001547193396e-06,
      "loss": 0.365088963508606,
      "num_input_tokens_seen": 385533344,
      "step": 7495,
      "train_runtime": 46225.458,
      "train_tokens_per_second": 8340.282
    },
    {
      "epoch": 0.5049824939402101,
      "grad_norm": 0.7539702456010547,
      "learning_rate": 8.508131934954515e-06,
      "loss": 0.36192262172698975,
      "num_input_tokens_seen": 385765000,
      "step": 7500,
      "train_runtime": 46251.1415,
      "train_tokens_per_second": 8340.659
    },
    {
      "epoch": 0.5053191489361702,
      "grad_norm": 0.6383047591326676,
      "learning_rate": 8.50624741693716e-06,
      "loss": 0.388753342628479,
      "num_input_tokens_seen": 386006280,
      "step": 7505,
      "train_runtime": 46284.3196,
      "train_tokens_per_second": 8339.893
    },
    {
      "epoch": 0.5056558039321304,
      "grad_norm": 0.731014529332956,
      "learning_rate": 8.5043619184089e-06,
      "loss": 0.406252384185791,
      "num_input_tokens_seen": 386273832,
      "step": 7510,
      "train_runtime": 46316.9101,
      "train_tokens_per_second": 8339.801
    },
    {
      "epoch": 0.5059924589280905,
      "grad_norm": 0.6856227421957123,
      "learning_rate": 8.502475439897e-06,
      "loss": 0.39432151317596437,
      "num_input_tokens_seen": 386550216,
      "step": 7515,
      "train_runtime": 46348.1639,
      "train_tokens_per_second": 8340.141
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.5548701453529146,
      "learning_rate": 8.500587981929013e-06,
      "loss": 0.38578619956970217,
      "num_input_tokens_seen": 386805176,
      "step": 7520,
      "train_runtime": 46375.4695,
      "train_tokens_per_second": 8340.728
    },
    {
      "epoch": 0.5066657689200108,
      "grad_norm": 0.6612666454142859,
      "learning_rate": 8.498699545032758e-06,
      "loss": 0.3705803871154785,
      "num_input_tokens_seen": 387090048,
      "step": 7525,
      "train_runtime": 46405.3792,
      "train_tokens_per_second": 8341.491
    },
    {
      "epoch": 0.507002423915971,
      "grad_norm": 0.65319765289502,
      "learning_rate": 8.496810129736334e-06,
      "loss": 0.37907302379608154,
      "num_input_tokens_seen": 387334768,
      "step": 7530,
      "train_runtime": 46435.7129,
      "train_tokens_per_second": 8341.312
    },
    {
      "epoch": 0.5073390789119311,
      "grad_norm": 0.7221470445683696,
      "learning_rate": 8.494919736568105e-06,
      "loss": 0.3734575748443604,
      "num_input_tokens_seen": 387571736,
      "step": 7535,
      "train_runtime": 46465.4243,
      "train_tokens_per_second": 8341.078
    },
    {
      "epoch": 0.5076757339078912,
      "grad_norm": 0.7425313851606345,
      "learning_rate": 8.49302836605672e-06,
      "loss": 0.3676410436630249,
      "num_input_tokens_seen": 387843064,
      "step": 7540,
      "train_runtime": 46494.885,
      "train_tokens_per_second": 8341.629
    },
    {
      "epoch": 0.5080123889038514,
      "grad_norm": 0.6437184660649315,
      "learning_rate": 8.491136018731088e-06,
      "loss": 0.35385403633117674,
      "num_input_tokens_seen": 388109640,
      "step": 7545,
      "train_runtime": 46529.0168,
      "train_tokens_per_second": 8341.239
    },
    {
      "epoch": 0.5083490438998115,
      "grad_norm": 0.7328397310040713,
      "learning_rate": 8.4892426951204e-06,
      "loss": 0.39201955795288085,
      "num_input_tokens_seen": 388387920,
      "step": 7550,
      "train_runtime": 46560.0868,
      "train_tokens_per_second": 8341.649
    },
    {
      "epoch": 0.5086856988957716,
      "grad_norm": 0.564818255571357,
      "learning_rate": 8.48734839575412e-06,
      "loss": 0.3611902236938477,
      "num_input_tokens_seen": 388671496,
      "step": 7555,
      "train_runtime": 46590.4209,
      "train_tokens_per_second": 8342.305
    },
    {
      "epoch": 0.5090223538917318,
      "grad_norm": 0.6137825471478212,
      "learning_rate": 8.485453121161983e-06,
      "loss": 0.35877237319946287,
      "num_input_tokens_seen": 388913384,
      "step": 7560,
      "train_runtime": 46621.4403,
      "train_tokens_per_second": 8341.943
    },
    {
      "epoch": 0.5093590088876919,
      "grad_norm": 0.7848932893956279,
      "learning_rate": 8.483556871873993e-06,
      "loss": 0.409788990020752,
      "num_input_tokens_seen": 389169640,
      "step": 7565,
      "train_runtime": 46655.558,
      "train_tokens_per_second": 8341.335
    },
    {
      "epoch": 0.5096956638836521,
      "grad_norm": 0.6912776431065939,
      "learning_rate": 8.481659648420433e-06,
      "loss": 0.37595205307006835,
      "num_input_tokens_seen": 389424448,
      "step": 7570,
      "train_runtime": 46686.2213,
      "train_tokens_per_second": 8341.314
    },
    {
      "epoch": 0.5100323188796122,
      "grad_norm": 0.7799498147069889,
      "learning_rate": 8.479761451331855e-06,
      "loss": 0.37680509090423586,
      "num_input_tokens_seen": 389634776,
      "step": 7575,
      "train_runtime": 46716.2684,
      "train_tokens_per_second": 8340.452
    },
    {
      "epoch": 0.5103689738755723,
      "grad_norm": 0.7343766559297956,
      "learning_rate": 8.477862281139082e-06,
      "loss": 0.3510934352874756,
      "num_input_tokens_seen": 389880720,
      "step": 7580,
      "train_runtime": 46748.4229,
      "train_tokens_per_second": 8339.976
    },
    {
      "epoch": 0.5107056288715325,
      "grad_norm": 0.5984562213830932,
      "learning_rate": 8.475962138373212e-06,
      "loss": 0.39998836517333985,
      "num_input_tokens_seen": 390119256,
      "step": 7585,
      "train_runtime": 46777.1164,
      "train_tokens_per_second": 8339.959
    },
    {
      "epoch": 0.5110422838674926,
      "grad_norm": 0.6021438084641919,
      "learning_rate": 8.474061023565614e-06,
      "loss": 0.37810866832733153,
      "num_input_tokens_seen": 390392200,
      "step": 7590,
      "train_runtime": 46807.7118,
      "train_tokens_per_second": 8340.339
    },
    {
      "epoch": 0.5113789388634528,
      "grad_norm": 0.7355085683751631,
      "learning_rate": 8.472158937247931e-06,
      "loss": 0.4002213478088379,
      "num_input_tokens_seen": 390643184,
      "step": 7595,
      "train_runtime": 46838.8211,
      "train_tokens_per_second": 8340.158
    },
    {
      "epoch": 0.5117155938594129,
      "grad_norm": 0.6096720830520197,
      "learning_rate": 8.47025587995207e-06,
      "loss": 0.37855312824249265,
      "num_input_tokens_seen": 390897592,
      "step": 7600,
      "train_runtime": 46872.428,
      "train_tokens_per_second": 8339.606
    },
    {
      "epoch": 0.512052248855373,
      "grad_norm": 0.5920900712702607,
      "learning_rate": 8.46835185221022e-06,
      "loss": 0.393257212638855,
      "num_input_tokens_seen": 391145368,
      "step": 7605,
      "train_runtime": 46898.0048,
      "train_tokens_per_second": 8340.341
    },
    {
      "epoch": 0.5123889038513332,
      "grad_norm": 0.7985486218610968,
      "learning_rate": 8.466446854554835e-06,
      "loss": 0.35846242904663084,
      "num_input_tokens_seen": 391393792,
      "step": 7610,
      "train_runtime": 46925.3983,
      "train_tokens_per_second": 8340.767
    },
    {
      "epoch": 0.5127255588472933,
      "grad_norm": 0.6905848866022356,
      "learning_rate": 8.464540887518638e-06,
      "loss": 0.3713674068450928,
      "num_input_tokens_seen": 391624728,
      "step": 7615,
      "train_runtime": 46952.7516,
      "train_tokens_per_second": 8340.826
    },
    {
      "epoch": 0.5130622138432535,
      "grad_norm": 0.7833230674715093,
      "learning_rate": 8.462633951634631e-06,
      "loss": 0.3607809066772461,
      "num_input_tokens_seen": 391870120,
      "step": 7620,
      "train_runtime": 46979.9431,
      "train_tokens_per_second": 8341.222
    },
    {
      "epoch": 0.5133988688392136,
      "grad_norm": 0.6492307425565776,
      "learning_rate": 8.46072604743608e-06,
      "loss": 0.3519000768661499,
      "num_input_tokens_seen": 392115856,
      "step": 7625,
      "train_runtime": 47009.8585,
      "train_tokens_per_second": 8341.141
    },
    {
      "epoch": 0.5137355238351737,
      "grad_norm": 0.8449883996039287,
      "learning_rate": 8.458817175456528e-06,
      "loss": 0.3830984115600586,
      "num_input_tokens_seen": 392365448,
      "step": 7630,
      "train_runtime": 47041.2489,
      "train_tokens_per_second": 8340.881
    },
    {
      "epoch": 0.5140721788311339,
      "grad_norm": 0.6981611679260012,
      "learning_rate": 8.45690733622978e-06,
      "loss": 0.3569530963897705,
      "num_input_tokens_seen": 392624352,
      "step": 7635,
      "train_runtime": 47071.4471,
      "train_tokens_per_second": 8341.03
    },
    {
      "epoch": 0.514408833827094,
      "grad_norm": 0.647905627904883,
      "learning_rate": 8.45499653028992e-06,
      "loss": 0.3651439666748047,
      "num_input_tokens_seen": 392886520,
      "step": 7640,
      "train_runtime": 47099.4989,
      "train_tokens_per_second": 8341.628
    },
    {
      "epoch": 0.5147454888230542,
      "grad_norm": 0.6355845728349334,
      "learning_rate": 8.453084758171297e-06,
      "loss": 0.38056604862213134,
      "num_input_tokens_seen": 393140056,
      "step": 7645,
      "train_runtime": 47127.8828,
      "train_tokens_per_second": 8341.984
    },
    {
      "epoch": 0.5150821438190143,
      "grad_norm": 0.6328638410046357,
      "learning_rate": 8.451172020408532e-06,
      "loss": 0.39309396743774416,
      "num_input_tokens_seen": 393408792,
      "step": 7650,
      "train_runtime": 47156.1421,
      "train_tokens_per_second": 8342.684
    },
    {
      "epoch": 0.5154187988149744,
      "grad_norm": 0.6248714557320574,
      "learning_rate": 8.449258317536517e-06,
      "loss": 0.3513326644897461,
      "num_input_tokens_seen": 393654552,
      "step": 7655,
      "train_runtime": 47190.4731,
      "train_tokens_per_second": 8341.823
    },
    {
      "epoch": 0.5157554538109346,
      "grad_norm": 0.7048099721263664,
      "learning_rate": 8.447343650090412e-06,
      "loss": 0.3510714530944824,
      "num_input_tokens_seen": 393912632,
      "step": 7660,
      "train_runtime": 47221.491,
      "train_tokens_per_second": 8341.808
    },
    {
      "epoch": 0.5160921088068947,
      "grad_norm": 0.6024937808214127,
      "learning_rate": 8.445428018605647e-06,
      "loss": 0.3957812309265137,
      "num_input_tokens_seen": 394177952,
      "step": 7665,
      "train_runtime": 47252.423,
      "train_tokens_per_second": 8341.963
    },
    {
      "epoch": 0.5164287638028549,
      "grad_norm": 0.6272781822698165,
      "learning_rate": 8.443511423617924e-06,
      "loss": 0.34112522602081297,
      "num_input_tokens_seen": 394427296,
      "step": 7670,
      "train_runtime": 47282.1975,
      "train_tokens_per_second": 8341.983
    },
    {
      "epoch": 0.516765418798815,
      "grad_norm": 0.5852525198475,
      "learning_rate": 8.441593865663213e-06,
      "loss": 0.3890520572662354,
      "num_input_tokens_seen": 394690208,
      "step": 7675,
      "train_runtime": 47310.4419,
      "train_tokens_per_second": 8342.56
    },
    {
      "epoch": 0.5171020737947751,
      "grad_norm": 0.6235578612467396,
      "learning_rate": 8.439675345277751e-06,
      "loss": 0.3883882999420166,
      "num_input_tokens_seen": 394935760,
      "step": 7680,
      "train_runtime": 47339.3475,
      "train_tokens_per_second": 8342.653
    },
    {
      "epoch": 0.5174387287907353,
      "grad_norm": 0.572447783953436,
      "learning_rate": 8.437755862998047e-06,
      "loss": 0.35449814796447754,
      "num_input_tokens_seen": 395193280,
      "step": 7685,
      "train_runtime": 47369.2388,
      "train_tokens_per_second": 8342.825
    },
    {
      "epoch": 0.5177753837866954,
      "grad_norm": 0.5769265566849234,
      "learning_rate": 8.435835419360878e-06,
      "loss": 0.38056881427764894,
      "num_input_tokens_seen": 395481400,
      "step": 7690,
      "train_runtime": 47405.0995,
      "train_tokens_per_second": 8342.592
    },
    {
      "epoch": 0.5181120387826555,
      "grad_norm": 0.6617518043091616,
      "learning_rate": 8.433914014903291e-06,
      "loss": 0.4177420139312744,
      "num_input_tokens_seen": 395737320,
      "step": 7695,
      "train_runtime": 47433.5404,
      "train_tokens_per_second": 8342.985
    },
    {
      "epoch": 0.5184486937786157,
      "grad_norm": 0.6867247814955132,
      "learning_rate": 8.431991650162599e-06,
      "loss": 0.381895637512207,
      "num_input_tokens_seen": 395989928,
      "step": 7700,
      "train_runtime": 47464.5381,
      "train_tokens_per_second": 8342.859
    },
    {
      "epoch": 0.5187853487745758,
      "grad_norm": 0.59042812470442,
      "learning_rate": 8.430068325676386e-06,
      "loss": 0.3787964105606079,
      "num_input_tokens_seen": 396245504,
      "step": 7705,
      "train_runtime": 47496.8065,
      "train_tokens_per_second": 8342.571
    },
    {
      "epoch": 0.519122003770536,
      "grad_norm": 0.6251395451579591,
      "learning_rate": 8.428144041982502e-06,
      "loss": 0.38366260528564455,
      "num_input_tokens_seen": 396497544,
      "step": 7710,
      "train_runtime": 47525.4424,
      "train_tokens_per_second": 8342.848
    },
    {
      "epoch": 0.5194586587664961,
      "grad_norm": 0.5845638599583094,
      "learning_rate": 8.42621879961907e-06,
      "loss": 0.37007317543029783,
      "num_input_tokens_seen": 396746408,
      "step": 7715,
      "train_runtime": 47556.994,
      "train_tokens_per_second": 8342.546
    },
    {
      "epoch": 0.5197953137624562,
      "grad_norm": 0.6936543911756424,
      "learning_rate": 8.424292599124476e-06,
      "loss": 0.3416330337524414,
      "num_input_tokens_seen": 397018504,
      "step": 7720,
      "train_runtime": 47585.5049,
      "train_tokens_per_second": 8343.266
    },
    {
      "epoch": 0.5201319687584164,
      "grad_norm": 0.9171300125753244,
      "learning_rate": 8.422365441037374e-06,
      "loss": 0.3919394493103027,
      "num_input_tokens_seen": 397303224,
      "step": 7725,
      "train_runtime": 47612.286,
      "train_tokens_per_second": 8344.553
    },
    {
      "epoch": 0.5204686237543765,
      "grad_norm": 0.6536633953055617,
      "learning_rate": 8.420437325896692e-06,
      "loss": 0.3955144643783569,
      "num_input_tokens_seen": 397580688,
      "step": 7730,
      "train_runtime": 47646.3851,
      "train_tokens_per_second": 8344.404
    },
    {
      "epoch": 0.5208052787503367,
      "grad_norm": 0.6631249220386531,
      "learning_rate": 8.418508254241617e-06,
      "loss": 0.3783026933670044,
      "num_input_tokens_seen": 397833288,
      "step": 7735,
      "train_runtime": 47677.9929,
      "train_tokens_per_second": 8344.17
    },
    {
      "epoch": 0.5211419337462968,
      "grad_norm": 0.7853563909813415,
      "learning_rate": 8.416578226611612e-06,
      "loss": 0.3487244129180908,
      "num_input_tokens_seen": 398073496,
      "step": 7740,
      "train_runtime": 47707.7569,
      "train_tokens_per_second": 8343.999
    },
    {
      "epoch": 0.5214785887422569,
      "grad_norm": 0.628453591261931,
      "learning_rate": 8.414647243546398e-06,
      "loss": 0.36919527053833007,
      "num_input_tokens_seen": 398332896,
      "step": 7745,
      "train_runtime": 47736.3198,
      "train_tokens_per_second": 8344.441
    },
    {
      "epoch": 0.5218152437382171,
      "grad_norm": 0.6217429447219092,
      "learning_rate": 8.412715305585973e-06,
      "loss": 0.3533369541168213,
      "num_input_tokens_seen": 398615368,
      "step": 7750,
      "train_runtime": 47767.331,
      "train_tokens_per_second": 8344.937
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.654906886965613,
      "learning_rate": 8.410782413270594e-06,
      "loss": 0.3846121311187744,
      "num_input_tokens_seen": 398882248,
      "step": 7755,
      "train_runtime": 47796.4034,
      "train_tokens_per_second": 8345.445
    },
    {
      "epoch": 0.5224885537301374,
      "grad_norm": 0.8865291679892634,
      "learning_rate": 8.408848567140787e-06,
      "loss": 0.3649510145187378,
      "num_input_tokens_seen": 399088576,
      "step": 7760,
      "train_runtime": 47821.9646,
      "train_tokens_per_second": 8345.299
    },
    {
      "epoch": 0.5228252087260975,
      "grad_norm": 0.6418574944058792,
      "learning_rate": 8.406913767737353e-06,
      "loss": 0.35866289138793944,
      "num_input_tokens_seen": 399341504,
      "step": 7765,
      "train_runtime": 47854.5916,
      "train_tokens_per_second": 8344.894
    },
    {
      "epoch": 0.5231618637220576,
      "grad_norm": 0.667300439471738,
      "learning_rate": 8.404978015601344e-06,
      "loss": 0.3917475700378418,
      "num_input_tokens_seen": 399590744,
      "step": 7770,
      "train_runtime": 47884.9963,
      "train_tokens_per_second": 8344.801
    },
    {
      "epoch": 0.5234985187180178,
      "grad_norm": 0.686940478540385,
      "learning_rate": 8.403041311274091e-06,
      "loss": 0.3780186653137207,
      "num_input_tokens_seen": 399856984,
      "step": 7775,
      "train_runtime": 47915.5633,
      "train_tokens_per_second": 8345.034
    },
    {
      "epoch": 0.5238351737139779,
      "grad_norm": 0.7571389671155001,
      "learning_rate": 8.401103655297188e-06,
      "loss": 0.37732892036437987,
      "num_input_tokens_seen": 400103296,
      "step": 7780,
      "train_runtime": 47945.6487,
      "train_tokens_per_second": 8344.934
    },
    {
      "epoch": 0.5241718287099381,
      "grad_norm": 0.6584152291053113,
      "learning_rate": 8.399165048212494e-06,
      "loss": 0.35012667179107665,
      "num_input_tokens_seen": 400366056,
      "step": 7785,
      "train_runtime": 47977.1445,
      "train_tokens_per_second": 8344.933
    },
    {
      "epoch": 0.5245084837058982,
      "grad_norm": 0.6010353788830914,
      "learning_rate": 8.397225490562131e-06,
      "loss": 0.37979326248168943,
      "num_input_tokens_seen": 400642920,
      "step": 7790,
      "train_runtime": 48005.0682,
      "train_tokens_per_second": 8345.846
    },
    {
      "epoch": 0.5248451387018583,
      "grad_norm": 0.7039792937171293,
      "learning_rate": 8.395284982888494e-06,
      "loss": 0.3541805982589722,
      "num_input_tokens_seen": 400902152,
      "step": 7795,
      "train_runtime": 48037.5243,
      "train_tokens_per_second": 8345.604
    },
    {
      "epoch": 0.5251817936978185,
      "grad_norm": 0.6222244227830678,
      "learning_rate": 8.393343525734239e-06,
      "loss": 0.38395464420318604,
      "num_input_tokens_seen": 401161184,
      "step": 7800,
      "train_runtime": 48067.0034,
      "train_tokens_per_second": 8345.875
    },
    {
      "epoch": 0.5255184486937786,
      "grad_norm": 0.6367182849929544,
      "learning_rate": 8.391401119642286e-06,
      "loss": 0.40142011642456055,
      "num_input_tokens_seen": 401381752,
      "step": 7805,
      "train_runtime": 48095.059,
      "train_tokens_per_second": 8345.592
    },
    {
      "epoch": 0.5258551036897388,
      "grad_norm": 0.6449859089671959,
      "learning_rate": 8.389457765155824e-06,
      "loss": 0.401448917388916,
      "num_input_tokens_seen": 401600072,
      "step": 7810,
      "train_runtime": 48121.7203,
      "train_tokens_per_second": 8345.505
    },
    {
      "epoch": 0.5261917586856989,
      "grad_norm": 0.7136512663179694,
      "learning_rate": 8.387513462818309e-06,
      "loss": 0.3598064422607422,
      "num_input_tokens_seen": 401858904,
      "step": 7815,
      "train_runtime": 48154.1888,
      "train_tokens_per_second": 8345.253
    },
    {
      "epoch": 0.526528413681659,
      "grad_norm": 0.6333516295597791,
      "learning_rate": 8.385568213173453e-06,
      "loss": 0.36281332969665525,
      "num_input_tokens_seen": 402106256,
      "step": 7820,
      "train_runtime": 48183.326,
      "train_tokens_per_second": 8345.34
    },
    {
      "epoch": 0.5268650686776192,
      "grad_norm": 0.5997603648779148,
      "learning_rate": 8.383622016765242e-06,
      "loss": 0.3611438751220703,
      "num_input_tokens_seen": 402356008,
      "step": 7825,
      "train_runtime": 48214.1521,
      "train_tokens_per_second": 8345.185
    },
    {
      "epoch": 0.5272017236735793,
      "grad_norm": 0.7660886558270676,
      "learning_rate": 8.381674874137926e-06,
      "loss": 0.39214377403259276,
      "num_input_tokens_seen": 402597696,
      "step": 7830,
      "train_runtime": 48246.0842,
      "train_tokens_per_second": 8344.671
    },
    {
      "epoch": 0.5275383786695395,
      "grad_norm": 0.715898870910737,
      "learning_rate": 8.379726785836013e-06,
      "loss": 0.3602346420288086,
      "num_input_tokens_seen": 402832600,
      "step": 7835,
      "train_runtime": 48273.9498,
      "train_tokens_per_second": 8344.72
    },
    {
      "epoch": 0.5278750336654996,
      "grad_norm": 0.6746795817692277,
      "learning_rate": 8.377777752404283e-06,
      "loss": 0.3501392602920532,
      "num_input_tokens_seen": 403106192,
      "step": 7840,
      "train_runtime": 48302.1861,
      "train_tokens_per_second": 8345.506
    },
    {
      "epoch": 0.5282116886614597,
      "grad_norm": 0.8291891576554807,
      "learning_rate": 8.375827774387774e-06,
      "loss": 0.34889845848083495,
      "num_input_tokens_seen": 403322664,
      "step": 7845,
      "train_runtime": 48329.3604,
      "train_tokens_per_second": 8345.293
    },
    {
      "epoch": 0.5285483436574199,
      "grad_norm": 0.587604698068178,
      "learning_rate": 8.373876852331793e-06,
      "loss": 0.35321383476257323,
      "num_input_tokens_seen": 403587536,
      "step": 7850,
      "train_runtime": 48360.02,
      "train_tokens_per_second": 8345.479
    },
    {
      "epoch": 0.52888499865338,
      "grad_norm": 0.6424912629032309,
      "learning_rate": 8.371924986781909e-06,
      "loss": 0.36687531471252444,
      "num_input_tokens_seen": 403850536,
      "step": 7855,
      "train_runtime": 48388.3045,
      "train_tokens_per_second": 8346.036
    },
    {
      "epoch": 0.5292216536493402,
      "grad_norm": 0.6987297171676459,
      "learning_rate": 8.369972178283955e-06,
      "loss": 0.35620162487030027,
      "num_input_tokens_seen": 404127520,
      "step": 7860,
      "train_runtime": 48417.1672,
      "train_tokens_per_second": 8346.782
    },
    {
      "epoch": 0.5295583086453003,
      "grad_norm": 0.6433897987366292,
      "learning_rate": 8.368018427384027e-06,
      "loss": 0.37619752883911134,
      "num_input_tokens_seen": 404379000,
      "step": 7865,
      "train_runtime": 48449.1225,
      "train_tokens_per_second": 8346.467
    },
    {
      "epoch": 0.5298949636412604,
      "grad_norm": 0.8802888437318735,
      "learning_rate": 8.366063734628485e-06,
      "loss": 0.37157843112945554,
      "num_input_tokens_seen": 404605920,
      "step": 7870,
      "train_runtime": 48480.1634,
      "train_tokens_per_second": 8345.804
    },
    {
      "epoch": 0.5302316186372206,
      "grad_norm": 0.7490288912254802,
      "learning_rate": 8.364108100563955e-06,
      "loss": 0.3841567039489746,
      "num_input_tokens_seen": 404861304,
      "step": 7875,
      "train_runtime": 48512.6824,
      "train_tokens_per_second": 8345.473
    },
    {
      "epoch": 0.5305682736331807,
      "grad_norm": 0.7238156645687599,
      "learning_rate": 8.362151525737321e-06,
      "loss": 0.41603918075561525,
      "num_input_tokens_seen": 405120440,
      "step": 7880,
      "train_runtime": 48542.5199,
      "train_tokens_per_second": 8345.682
    },
    {
      "epoch": 0.5309049286291408,
      "grad_norm": 0.706996738355739,
      "learning_rate": 8.360194010695733e-06,
      "loss": 0.35365705490112304,
      "num_input_tokens_seen": 405375984,
      "step": 7885,
      "train_runtime": 48575.0602,
      "train_tokens_per_second": 8345.352
    },
    {
      "epoch": 0.531241583625101,
      "grad_norm": 0.6297336408289489,
      "learning_rate": 8.358235555986607e-06,
      "loss": 0.3887474536895752,
      "num_input_tokens_seen": 405654368,
      "step": 7890,
      "train_runtime": 48607.4467,
      "train_tokens_per_second": 8345.519
    },
    {
      "epoch": 0.5315782386210611,
      "grad_norm": 0.6443323794519081,
      "learning_rate": 8.356276162157613e-06,
      "loss": 0.3496539831161499,
      "num_input_tokens_seen": 405886632,
      "step": 7895,
      "train_runtime": 48633.3711,
      "train_tokens_per_second": 8345.846
    },
    {
      "epoch": 0.5319148936170213,
      "grad_norm": 0.6849945846407484,
      "learning_rate": 8.354315829756696e-06,
      "loss": 0.3793375253677368,
      "num_input_tokens_seen": 406157840,
      "step": 7900,
      "train_runtime": 48667.1008,
      "train_tokens_per_second": 8345.635
    },
    {
      "epoch": 0.5322515486129814,
      "grad_norm": 0.6672917509232463,
      "learning_rate": 8.352354559332053e-06,
      "loss": 0.3634167671203613,
      "num_input_tokens_seen": 406420272,
      "step": 7905,
      "train_runtime": 48699.1833,
      "train_tokens_per_second": 8345.525
    },
    {
      "epoch": 0.5325882036089415,
      "grad_norm": 0.7236311850624043,
      "learning_rate": 8.350392351432146e-06,
      "loss": 0.3700665235519409,
      "num_input_tokens_seen": 406683832,
      "step": 7910,
      "train_runtime": 48732.8428,
      "train_tokens_per_second": 8345.169
    },
    {
      "epoch": 0.5329248586049017,
      "grad_norm": 0.5962247174812687,
      "learning_rate": 8.348429206605702e-06,
      "loss": 0.34813461303710935,
      "num_input_tokens_seen": 406965064,
      "step": 7915,
      "train_runtime": 48765.7912,
      "train_tokens_per_second": 8345.298
    },
    {
      "epoch": 0.5332615136008618,
      "grad_norm": 0.7332990152133011,
      "learning_rate": 8.346465125401706e-06,
      "loss": 0.4080685615539551,
      "num_input_tokens_seen": 407210520,
      "step": 7920,
      "train_runtime": 48799.7834,
      "train_tokens_per_second": 8344.515
    },
    {
      "epoch": 0.533598168596822,
      "grad_norm": 0.6688241775761121,
      "learning_rate": 8.344500108369412e-06,
      "loss": 0.3635000705718994,
      "num_input_tokens_seen": 407440008,
      "step": 7925,
      "train_runtime": 48828.5535,
      "train_tokens_per_second": 8344.298
    },
    {
      "epoch": 0.5339348235927821,
      "grad_norm": 0.6200846299679823,
      "learning_rate": 8.342534156058326e-06,
      "loss": 0.38973512649536135,
      "num_input_tokens_seen": 407698024,
      "step": 7930,
      "train_runtime": 48863.5803,
      "train_tokens_per_second": 8343.597
    },
    {
      "epoch": 0.5342714785887422,
      "grad_norm": 0.6723231441459591,
      "learning_rate": 8.34056726901822e-06,
      "loss": 0.4003415584564209,
      "num_input_tokens_seen": 407939616,
      "step": 7935,
      "train_runtime": 48895.5774,
      "train_tokens_per_second": 8343.078
    },
    {
      "epoch": 0.5346081335847024,
      "grad_norm": 0.7169231831373829,
      "learning_rate": 8.338599447799132e-06,
      "loss": 0.3740688800811768,
      "num_input_tokens_seen": 408183368,
      "step": 7940,
      "train_runtime": 48925.4134,
      "train_tokens_per_second": 8342.972
    },
    {
      "epoch": 0.5349447885806625,
      "grad_norm": 0.8277647438446416,
      "learning_rate": 8.336630692951353e-06,
      "loss": 0.36888928413391114,
      "num_input_tokens_seen": 408442264,
      "step": 7945,
      "train_runtime": 48958.7083,
      "train_tokens_per_second": 8342.587
    },
    {
      "epoch": 0.5352814435766227,
      "grad_norm": 0.7730860803783264,
      "learning_rate": 8.334661005025441e-06,
      "loss": 0.3906532287597656,
      "num_input_tokens_seen": 408698208,
      "step": 7950,
      "train_runtime": 48991.1963,
      "train_tokens_per_second": 8342.279
    },
    {
      "epoch": 0.5356180985725828,
      "grad_norm": 0.6695592457997913,
      "learning_rate": 8.332690384572213e-06,
      "loss": 0.40138988494873046,
      "num_input_tokens_seen": 408956416,
      "step": 7955,
      "train_runtime": 49024.0868,
      "train_tokens_per_second": 8341.949
    },
    {
      "epoch": 0.5359547535685429,
      "grad_norm": 0.6718941499793176,
      "learning_rate": 8.330718832142746e-06,
      "loss": 0.3946081161499023,
      "num_input_tokens_seen": 409228952,
      "step": 7960,
      "train_runtime": 49055.3668,
      "train_tokens_per_second": 8342.185
    },
    {
      "epoch": 0.5362914085645031,
      "grad_norm": 0.5891297811281302,
      "learning_rate": 8.328746348288379e-06,
      "loss": 0.3772918462753296,
      "num_input_tokens_seen": 409462128,
      "step": 7965,
      "train_runtime": 49086.0172,
      "train_tokens_per_second": 8341.726
    },
    {
      "epoch": 0.5366280635604632,
      "grad_norm": 0.6090595114406026,
      "learning_rate": 8.32677293356071e-06,
      "loss": 0.40877981185913087,
      "num_input_tokens_seen": 409746632,
      "step": 7970,
      "train_runtime": 49118.5681,
      "train_tokens_per_second": 8341.991
    },
    {
      "epoch": 0.5369647185564234,
      "grad_norm": 0.7679205293053646,
      "learning_rate": 8.324798588511598e-06,
      "loss": 0.3693244457244873,
      "num_input_tokens_seen": 409969464,
      "step": 7975,
      "train_runtime": 49144.8974,
      "train_tokens_per_second": 8342.056
    },
    {
      "epoch": 0.5373013735523835,
      "grad_norm": 0.7426790559102371,
      "learning_rate": 8.322823313693162e-06,
      "loss": 0.3832805633544922,
      "num_input_tokens_seen": 410226120,
      "step": 7980,
      "train_runtime": 49172.2996,
      "train_tokens_per_second": 8342.626
    },
    {
      "epoch": 0.5376380285483436,
      "grad_norm": 0.6919423675613927,
      "learning_rate": 8.320847109657782e-06,
      "loss": 0.36797146797180175,
      "num_input_tokens_seen": 410454792,
      "step": 7985,
      "train_runtime": 49202.8452,
      "train_tokens_per_second": 8342.095
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.7055490109238616,
      "learning_rate": 8.318869976958099e-06,
      "loss": 0.3302833080291748,
      "num_input_tokens_seen": 410704296,
      "step": 7990,
      "train_runtime": 49231.9122,
      "train_tokens_per_second": 8342.237
    },
    {
      "epoch": 0.5383113385402639,
      "grad_norm": 0.7081400475239026,
      "learning_rate": 8.316891916147007e-06,
      "loss": 0.3948878526687622,
      "num_input_tokens_seen": 410951232,
      "step": 7995,
      "train_runtime": 49265.0127,
      "train_tokens_per_second": 8341.645
    },
    {
      "epoch": 0.5386479935362241,
      "grad_norm": 0.6873684146031335,
      "learning_rate": 8.314912927777671e-06,
      "loss": 0.38301289081573486,
      "num_input_tokens_seen": 411212096,
      "step": 8000,
      "train_runtime": 49295.1345,
      "train_tokens_per_second": 8341.839
    },
    {
      "epoch": 0.5389846485321842,
      "grad_norm": 0.6216846909389204,
      "learning_rate": 8.312933012403503e-06,
      "loss": 0.37086408138275145,
      "num_input_tokens_seen": 411449936,
      "step": 8005,
      "train_runtime": 49324.7467,
      "train_tokens_per_second": 8341.653
    },
    {
      "epoch": 0.5393213035281443,
      "grad_norm": 0.6063313636813977,
      "learning_rate": 8.310952170578184e-06,
      "loss": 0.3812229156494141,
      "num_input_tokens_seen": 411696304,
      "step": 8010,
      "train_runtime": 49357.6583,
      "train_tokens_per_second": 8341.083
    },
    {
      "epoch": 0.5396579585241045,
      "grad_norm": 0.6799091731342173,
      "learning_rate": 8.308970402855647e-06,
      "loss": 0.3397177219390869,
      "num_input_tokens_seen": 411956512,
      "step": 8015,
      "train_runtime": 49386.3252,
      "train_tokens_per_second": 8341.51
    },
    {
      "epoch": 0.5399946135200646,
      "grad_norm": 0.6578328126086065,
      "learning_rate": 8.306987709790087e-06,
      "loss": 0.36124091148376464,
      "num_input_tokens_seen": 412214416,
      "step": 8020,
      "train_runtime": 49413.9429,
      "train_tokens_per_second": 8342.067
    },
    {
      "epoch": 0.5403312685160248,
      "grad_norm": 0.7162686686406264,
      "learning_rate": 8.305004091935962e-06,
      "loss": 0.3660966634750366,
      "num_input_tokens_seen": 412477344,
      "step": 8025,
      "train_runtime": 49444.9076,
      "train_tokens_per_second": 8342.16
    },
    {
      "epoch": 0.5406679235119849,
      "grad_norm": 0.6766470500339168,
      "learning_rate": 8.303019549847979e-06,
      "loss": 0.3955738067626953,
      "num_input_tokens_seen": 412732096,
      "step": 8030,
      "train_runtime": 49476.3523,
      "train_tokens_per_second": 8342.007
    },
    {
      "epoch": 0.541004578507945,
      "grad_norm": 0.6587476966021505,
      "learning_rate": 8.301034084081114e-06,
      "loss": 0.3581346035003662,
      "num_input_tokens_seen": 413008592,
      "step": 8035,
      "train_runtime": 49502.5771,
      "train_tokens_per_second": 8343.174
    },
    {
      "epoch": 0.5413412335039052,
      "grad_norm": 0.7708486821129813,
      "learning_rate": 8.299047695190592e-06,
      "loss": 0.3998836040496826,
      "num_input_tokens_seen": 413244080,
      "step": 8040,
      "train_runtime": 49528.9467,
      "train_tokens_per_second": 8343.486
    },
    {
      "epoch": 0.5416778884998653,
      "grad_norm": 0.6650894513191724,
      "learning_rate": 8.297060383731903e-06,
      "loss": 0.38129894733428954,
      "num_input_tokens_seen": 413500816,
      "step": 8045,
      "train_runtime": 49557.0442,
      "train_tokens_per_second": 8343.936
    },
    {
      "epoch": 0.5420145434958255,
      "grad_norm": 0.7252636579326407,
      "learning_rate": 8.29507215026079e-06,
      "loss": 0.3981478691101074,
      "num_input_tokens_seen": 413765536,
      "step": 8050,
      "train_runtime": 49586.7676,
      "train_tokens_per_second": 8344.273
    },
    {
      "epoch": 0.5423511984917856,
      "grad_norm": 0.6753457413500926,
      "learning_rate": 8.293082995333257e-06,
      "loss": 0.37377443313598635,
      "num_input_tokens_seen": 414036984,
      "step": 8055,
      "train_runtime": 49620.2427,
      "train_tokens_per_second": 8344.114
    },
    {
      "epoch": 0.5426878534877457,
      "grad_norm": 0.7760679111589714,
      "learning_rate": 8.291092919505568e-06,
      "loss": 0.36783013343811033,
      "num_input_tokens_seen": 414286928,
      "step": 8060,
      "train_runtime": 49646.7342,
      "train_tokens_per_second": 8344.696
    },
    {
      "epoch": 0.5430245084837059,
      "grad_norm": 0.8403133455863316,
      "learning_rate": 8.289101923334237e-06,
      "loss": 0.37348787784576415,
      "num_input_tokens_seen": 414555320,
      "step": 8065,
      "train_runtime": 49677.6571,
      "train_tokens_per_second": 8344.905
    },
    {
      "epoch": 0.543361163479666,
      "grad_norm": 0.7876653381583609,
      "learning_rate": 8.28711000737604e-06,
      "loss": 0.3777840375900269,
      "num_input_tokens_seen": 414798024,
      "step": 8070,
      "train_runtime": 49703.0928,
      "train_tokens_per_second": 8345.517
    },
    {
      "epoch": 0.5436978184756261,
      "grad_norm": 0.7201998106670032,
      "learning_rate": 8.285117172188013e-06,
      "loss": 0.38455877304077146,
      "num_input_tokens_seen": 415056096,
      "step": 8075,
      "train_runtime": 49736.8526,
      "train_tokens_per_second": 8345.041
    },
    {
      "epoch": 0.5440344734715863,
      "grad_norm": 0.7570525547005276,
      "learning_rate": 8.283123418327443e-06,
      "loss": 0.38727493286132814,
      "num_input_tokens_seen": 415324056,
      "step": 8080,
      "train_runtime": 49767.4995,
      "train_tokens_per_second": 8345.287
    },
    {
      "epoch": 0.5443711284675464,
      "grad_norm": 0.6069115031337261,
      "learning_rate": 8.281128746351878e-06,
      "loss": 0.347227954864502,
      "num_input_tokens_seen": 415602888,
      "step": 8085,
      "train_runtime": 49799.0576,
      "train_tokens_per_second": 8345.597
    },
    {
      "epoch": 0.5447077834635066,
      "grad_norm": 0.6082565987846037,
      "learning_rate": 8.279133156819123e-06,
      "loss": 0.39322333335876464,
      "num_input_tokens_seen": 415839528,
      "step": 8090,
      "train_runtime": 49828.7369,
      "train_tokens_per_second": 8345.376
    },
    {
      "epoch": 0.5450444384594667,
      "grad_norm": 0.6292301521078955,
      "learning_rate": 8.277136650287237e-06,
      "loss": 0.37183260917663574,
      "num_input_tokens_seen": 416109936,
      "step": 8095,
      "train_runtime": 49858.5443,
      "train_tokens_per_second": 8345.81
    },
    {
      "epoch": 0.5453810934554268,
      "grad_norm": 0.7442878212387799,
      "learning_rate": 8.275139227314537e-06,
      "loss": 0.36976823806762693,
      "num_input_tokens_seen": 416342808,
      "step": 8100,
      "train_runtime": 49885.3303,
      "train_tokens_per_second": 8345.997
    },
    {
      "epoch": 0.545717748451387,
      "grad_norm": 0.6779951056743466,
      "learning_rate": 8.273140888459594e-06,
      "loss": 0.41582889556884767,
      "num_input_tokens_seen": 416634832,
      "step": 8105,
      "train_runtime": 49920.3482,
      "train_tokens_per_second": 8345.992
    },
    {
      "epoch": 0.5460544034473471,
      "grad_norm": 0.7554126433760312,
      "learning_rate": 8.27114163428124e-06,
      "loss": 0.38335189819335935,
      "num_input_tokens_seen": 416895488,
      "step": 8110,
      "train_runtime": 49949.5875,
      "train_tokens_per_second": 8346.325
    },
    {
      "epoch": 0.5463910584433073,
      "grad_norm": 0.6782743185494216,
      "learning_rate": 8.26914146533856e-06,
      "loss": 0.37192554473876954,
      "num_input_tokens_seen": 417173112,
      "step": 8115,
      "train_runtime": 49980.4087,
      "train_tokens_per_second": 8346.733
    },
    {
      "epoch": 0.5467277134392674,
      "grad_norm": 0.6867036652881389,
      "learning_rate": 8.267140382190892e-06,
      "loss": 0.38272862434387206,
      "num_input_tokens_seen": 417421072,
      "step": 8120,
      "train_runtime": 50011.025,
      "train_tokens_per_second": 8346.581
    },
    {
      "epoch": 0.5470643684352275,
      "grad_norm": 0.6216834947779298,
      "learning_rate": 8.265138385397835e-06,
      "loss": 0.36701393127441406,
      "num_input_tokens_seen": 417669832,
      "step": 8125,
      "train_runtime": 50040.9753,
      "train_tokens_per_second": 8346.557
    },
    {
      "epoch": 0.5474010234311877,
      "grad_norm": 0.5446735907483835,
      "learning_rate": 8.263135475519241e-06,
      "loss": 0.33863718509674073,
      "num_input_tokens_seen": 417956576,
      "step": 8130,
      "train_runtime": 50079.2952,
      "train_tokens_per_second": 8345.896
    },
    {
      "epoch": 0.5477376784271478,
      "grad_norm": 0.6451268200392675,
      "learning_rate": 8.261131653115216e-06,
      "loss": 0.41192045211791994,
      "num_input_tokens_seen": 418235632,
      "step": 8135,
      "train_runtime": 50110.7166,
      "train_tokens_per_second": 8346.231
    },
    {
      "epoch": 0.548074333423108,
      "grad_norm": 0.711076301689756,
      "learning_rate": 8.259126918746122e-06,
      "loss": 0.3474350214004517,
      "num_input_tokens_seen": 418502008,
      "step": 8140,
      "train_runtime": 50140.8431,
      "train_tokens_per_second": 8346.529
    },
    {
      "epoch": 0.5484109884190681,
      "grad_norm": 0.6043168429902003,
      "learning_rate": 8.25712127297258e-06,
      "loss": 0.3674919843673706,
      "num_input_tokens_seen": 418756192,
      "step": 8145,
      "train_runtime": 50169.2919,
      "train_tokens_per_second": 8346.863
    },
    {
      "epoch": 0.5487476434150282,
      "grad_norm": 0.6232233860134962,
      "learning_rate": 8.255114716355459e-06,
      "loss": 0.4102166175842285,
      "num_input_tokens_seen": 419044560,
      "step": 8150,
      "train_runtime": 50198.0119,
      "train_tokens_per_second": 8347.832
    },
    {
      "epoch": 0.5490842984109884,
      "grad_norm": 0.5277624639135421,
      "learning_rate": 8.253107249455887e-06,
      "loss": 0.37326438426971437,
      "num_input_tokens_seen": 419311768,
      "step": 8155,
      "train_runtime": 50231.1091,
      "train_tokens_per_second": 8347.651
    },
    {
      "epoch": 0.5494209534069485,
      "grad_norm": 0.7416592652759905,
      "learning_rate": 8.251098872835247e-06,
      "loss": 0.3720337629318237,
      "num_input_tokens_seen": 419566840,
      "step": 8160,
      "train_runtime": 50260.5832,
      "train_tokens_per_second": 8347.831
    },
    {
      "epoch": 0.5497576084029087,
      "grad_norm": 0.6157764019580073,
      "learning_rate": 8.249089587055175e-06,
      "loss": 0.35301406383514405,
      "num_input_tokens_seen": 419809848,
      "step": 8165,
      "train_runtime": 50292.4831,
      "train_tokens_per_second": 8347.368
    },
    {
      "epoch": 0.5500942633988688,
      "grad_norm": 0.5564121889406249,
      "learning_rate": 8.247079392677558e-06,
      "loss": 0.3630223274230957,
      "num_input_tokens_seen": 420068776,
      "step": 8170,
      "train_runtime": 50324.4888,
      "train_tokens_per_second": 8347.204
    },
    {
      "epoch": 0.5504309183948289,
      "grad_norm": 0.6336229214414649,
      "learning_rate": 8.245068290264544e-06,
      "loss": 0.3596118688583374,
      "num_input_tokens_seen": 420328984,
      "step": 8175,
      "train_runtime": 50361.5286,
      "train_tokens_per_second": 8346.232
    },
    {
      "epoch": 0.5507675733907891,
      "grad_norm": 0.6124613000983918,
      "learning_rate": 8.243056280378531e-06,
      "loss": 0.3729127883911133,
      "num_input_tokens_seen": 420580432,
      "step": 8180,
      "train_runtime": 50397.3253,
      "train_tokens_per_second": 8345.293
    },
    {
      "epoch": 0.5511042283867492,
      "grad_norm": 0.624982194765141,
      "learning_rate": 8.241043363582171e-06,
      "loss": 0.36170690059661864,
      "num_input_tokens_seen": 420844600,
      "step": 8185,
      "train_runtime": 50430.4251,
      "train_tokens_per_second": 8345.054
    },
    {
      "epoch": 0.5514408833827094,
      "grad_norm": 0.6281829446528555,
      "learning_rate": 8.239029540438369e-06,
      "loss": 0.34545536041259767,
      "num_input_tokens_seen": 421090576,
      "step": 8190,
      "train_runtime": 50460.1538,
      "train_tokens_per_second": 8345.012
    },
    {
      "epoch": 0.5517775383786695,
      "grad_norm": 0.6691506257032837,
      "learning_rate": 8.237014811510284e-06,
      "loss": 0.354870343208313,
      "num_input_tokens_seen": 421350904,
      "step": 8195,
      "train_runtime": 50494.0181,
      "train_tokens_per_second": 8344.571
    },
    {
      "epoch": 0.5521141933746296,
      "grad_norm": 0.5747048623183061,
      "learning_rate": 8.234999177361329e-06,
      "loss": 0.3635202407836914,
      "num_input_tokens_seen": 421622568,
      "step": 8200,
      "train_runtime": 50523.1276,
      "train_tokens_per_second": 8345.14
    },
    {
      "epoch": 0.5524508483705898,
      "grad_norm": 0.6565385115191261,
      "learning_rate": 8.232982638555169e-06,
      "loss": 0.39693551063537597,
      "num_input_tokens_seen": 421893304,
      "step": 8205,
      "train_runtime": 50559.159,
      "train_tokens_per_second": 8344.548
    },
    {
      "epoch": 0.5527875033665499,
      "grad_norm": 0.6542385868523066,
      "learning_rate": 8.230965195655724e-06,
      "loss": 0.3799601554870605,
      "num_input_tokens_seen": 422173952,
      "step": 8210,
      "train_runtime": 50590.0803,
      "train_tokens_per_second": 8344.995
    },
    {
      "epoch": 0.55312415836251,
      "grad_norm": 0.6708474035486557,
      "learning_rate": 8.228946849227167e-06,
      "loss": 0.39617133140563965,
      "num_input_tokens_seen": 422455384,
      "step": 8215,
      "train_runtime": 50619.6783,
      "train_tokens_per_second": 8345.675
    },
    {
      "epoch": 0.5534608133584702,
      "grad_norm": 0.5613646194226947,
      "learning_rate": 8.226927599833916e-06,
      "loss": 0.37253365516662595,
      "num_input_tokens_seen": 422729056,
      "step": 8220,
      "train_runtime": 50654.6527,
      "train_tokens_per_second": 8345.315
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 0.5910604782801822,
      "learning_rate": 8.224907448040655e-06,
      "loss": 0.36936910152435304,
      "num_input_tokens_seen": 422999072,
      "step": 8225,
      "train_runtime": 50689.1508,
      "train_tokens_per_second": 8344.963
    },
    {
      "epoch": 0.5541341233503905,
      "grad_norm": 0.7494987691556876,
      "learning_rate": 8.222886394412309e-06,
      "loss": 0.40926084518432615,
      "num_input_tokens_seen": 423262624,
      "step": 8230,
      "train_runtime": 50724.2537,
      "train_tokens_per_second": 8344.383
    },
    {
      "epoch": 0.5544707783463506,
      "grad_norm": 0.5696496425783679,
      "learning_rate": 8.220864439514057e-06,
      "loss": 0.373108434677124,
      "num_input_tokens_seen": 423533208,
      "step": 8235,
      "train_runtime": 50754.567,
      "train_tokens_per_second": 8344.731
    },
    {
      "epoch": 0.5548074333423108,
      "grad_norm": 0.572387597852841,
      "learning_rate": 8.218841583911338e-06,
      "loss": 0.3593713045120239,
      "num_input_tokens_seen": 423773336,
      "step": 8240,
      "train_runtime": 50785.2547,
      "train_tokens_per_second": 8344.417
    },
    {
      "epoch": 0.5551440883382709,
      "grad_norm": 0.6559638624620918,
      "learning_rate": 8.216817828169834e-06,
      "loss": 0.40231051445007326,
      "num_input_tokens_seen": 424029792,
      "step": 8245,
      "train_runtime": 50815.4523,
      "train_tokens_per_second": 8344.505
    },
    {
      "epoch": 0.555480743334231,
      "grad_norm": 0.7806026455279574,
      "learning_rate": 8.214793172855481e-06,
      "loss": 0.36078591346740724,
      "num_input_tokens_seen": 424273208,
      "step": 8250,
      "train_runtime": 50842.2562,
      "train_tokens_per_second": 8344.893
    },
    {
      "epoch": 0.5558173983301912,
      "grad_norm": 0.625489470254559,
      "learning_rate": 8.21276761853447e-06,
      "loss": 0.3587883472442627,
      "num_input_tokens_seen": 424526896,
      "step": 8255,
      "train_runtime": 50877.2881,
      "train_tokens_per_second": 8344.134
    },
    {
      "epoch": 0.5561540533261513,
      "grad_norm": 0.731454975733539,
      "learning_rate": 8.210741165773238e-06,
      "loss": 0.3734226942062378,
      "num_input_tokens_seen": 424813328,
      "step": 8260,
      "train_runtime": 50906.0558,
      "train_tokens_per_second": 8345.045
    },
    {
      "epoch": 0.5564907083221114,
      "grad_norm": 0.6951700711858968,
      "learning_rate": 8.208713815138476e-06,
      "loss": 0.3747250556945801,
      "num_input_tokens_seen": 425088448,
      "step": 8265,
      "train_runtime": 50939.0206,
      "train_tokens_per_second": 8345.046
    },
    {
      "epoch": 0.5568273633180716,
      "grad_norm": 0.5841096340424691,
      "learning_rate": 8.20668556719713e-06,
      "loss": 0.34850254058837893,
      "num_input_tokens_seen": 425341696,
      "step": 8270,
      "train_runtime": 50964.7522,
      "train_tokens_per_second": 8345.801
    },
    {
      "epoch": 0.5571640183140317,
      "grad_norm": 0.694517657507754,
      "learning_rate": 8.204656422516388e-06,
      "loss": 0.33744144439697266,
      "num_input_tokens_seen": 425596768,
      "step": 8275,
      "train_runtime": 50994.8784,
      "train_tokens_per_second": 8345.873
    },
    {
      "epoch": 0.5575006733099919,
      "grad_norm": 0.7188253425776368,
      "learning_rate": 8.2026263816637e-06,
      "loss": 0.37355542182922363,
      "num_input_tokens_seen": 425839056,
      "step": 8280,
      "train_runtime": 51026.38,
      "train_tokens_per_second": 8345.469
    },
    {
      "epoch": 0.557837328305952,
      "grad_norm": 0.6178529417154219,
      "learning_rate": 8.200595445206753e-06,
      "loss": 0.3600980520248413,
      "num_input_tokens_seen": 426097608,
      "step": 8285,
      "train_runtime": 51057.7998,
      "train_tokens_per_second": 8345.397
    },
    {
      "epoch": 0.5581739833019121,
      "grad_norm": 0.8365491910069358,
      "learning_rate": 8.198563613713497e-06,
      "loss": 0.37148323059082033,
      "num_input_tokens_seen": 426334272,
      "step": 8290,
      "train_runtime": 51091.02,
      "train_tokens_per_second": 8344.603
    },
    {
      "epoch": 0.5585106382978723,
      "grad_norm": 0.6345345154854286,
      "learning_rate": 8.196530887752127e-06,
      "loss": 0.3624414443969727,
      "num_input_tokens_seen": 426598096,
      "step": 8295,
      "train_runtime": 51121.877,
      "train_tokens_per_second": 8344.727
    },
    {
      "epoch": 0.5588472932938324,
      "grad_norm": 0.719501723615319,
      "learning_rate": 8.194497267891086e-06,
      "loss": 0.37614076137542723,
      "num_input_tokens_seen": 426864440,
      "step": 8300,
      "train_runtime": 51151.9678,
      "train_tokens_per_second": 8345.025
    },
    {
      "epoch": 0.5591839482897926,
      "grad_norm": 0.6810481984336719,
      "learning_rate": 8.19246275469907e-06,
      "loss": 0.3779413938522339,
      "num_input_tokens_seen": 427128720,
      "step": 8305,
      "train_runtime": 51182.0643,
      "train_tokens_per_second": 8345.281
    },
    {
      "epoch": 0.5595206032857528,
      "grad_norm": 0.556355417952852,
      "learning_rate": 8.190427348745027e-06,
      "loss": 0.3560215473175049,
      "num_input_tokens_seen": 427403768,
      "step": 8310,
      "train_runtime": 51215.8118,
      "train_tokens_per_second": 8345.153
    },
    {
      "epoch": 0.559857258281713,
      "grad_norm": 0.562675525488597,
      "learning_rate": 8.188391050598146e-06,
      "loss": 0.34751291275024415,
      "num_input_tokens_seen": 427684560,
      "step": 8315,
      "train_runtime": 51247.2679,
      "train_tokens_per_second": 8345.509
    },
    {
      "epoch": 0.5601939132776731,
      "grad_norm": 0.7758817531898146,
      "learning_rate": 8.186353860827878e-06,
      "loss": 0.35610010623931887,
      "num_input_tokens_seen": 427929664,
      "step": 8320,
      "train_runtime": 51276.7881,
      "train_tokens_per_second": 8345.485
    },
    {
      "epoch": 0.5605305682736332,
      "grad_norm": 0.6390894314161747,
      "learning_rate": 8.184315780003911e-06,
      "loss": 0.365526008605957,
      "num_input_tokens_seen": 428173864,
      "step": 8325,
      "train_runtime": 51305.2981,
      "train_tokens_per_second": 8345.607
    },
    {
      "epoch": 0.5608672232695934,
      "grad_norm": 0.6021695467170687,
      "learning_rate": 8.182276808696193e-06,
      "loss": 0.3962390661239624,
      "num_input_tokens_seen": 428413376,
      "step": 8330,
      "train_runtime": 51336.4002,
      "train_tokens_per_second": 8345.217
    },
    {
      "epoch": 0.5612038782655535,
      "grad_norm": 0.5448671996434705,
      "learning_rate": 8.18023694747491e-06,
      "loss": 0.37192468643188475,
      "num_input_tokens_seen": 428679568,
      "step": 8335,
      "train_runtime": 51368.6773,
      "train_tokens_per_second": 8345.155
    },
    {
      "epoch": 0.5615405332615137,
      "grad_norm": 0.664717183563732,
      "learning_rate": 8.178196196910508e-06,
      "loss": 0.3603662490844727,
      "num_input_tokens_seen": 428950896,
      "step": 8340,
      "train_runtime": 51402.4308,
      "train_tokens_per_second": 8344.954
    },
    {
      "epoch": 0.5618771882574738,
      "grad_norm": 0.672247050920209,
      "learning_rate": 8.176154557573672e-06,
      "loss": 0.39259250164031984,
      "num_input_tokens_seen": 429230488,
      "step": 8345,
      "train_runtime": 51435.3337,
      "train_tokens_per_second": 8345.051
    },
    {
      "epoch": 0.5622138432534339,
      "grad_norm": 0.6662024688387073,
      "learning_rate": 8.174112030035344e-06,
      "loss": 0.35803871154785155,
      "num_input_tokens_seen": 429519832,
      "step": 8350,
      "train_runtime": 51470.1489,
      "train_tokens_per_second": 8345.028
    },
    {
      "epoch": 0.5625504982493941,
      "grad_norm": 0.6407893591820863,
      "learning_rate": 8.172068614866707e-06,
      "loss": 0.37200055122375486,
      "num_input_tokens_seen": 429769504,
      "step": 8355,
      "train_runtime": 51499.6057,
      "train_tokens_per_second": 8345.103
    },
    {
      "epoch": 0.5628871532453542,
      "grad_norm": 0.6485325883226092,
      "learning_rate": 8.170024312639197e-06,
      "loss": 0.39232349395751953,
      "num_input_tokens_seen": 430048568,
      "step": 8360,
      "train_runtime": 51529.7164,
      "train_tokens_per_second": 8345.642
    },
    {
      "epoch": 0.5632238082413143,
      "grad_norm": 0.6027445843658583,
      "learning_rate": 8.167979123924496e-06,
      "loss": 0.3706133127212524,
      "num_input_tokens_seen": 430311936,
      "step": 8365,
      "train_runtime": 51562.4204,
      "train_tokens_per_second": 8345.456
    },
    {
      "epoch": 0.5635604632372745,
      "grad_norm": 0.6393446607170127,
      "learning_rate": 8.165933049294537e-06,
      "loss": 0.3980983018875122,
      "num_input_tokens_seen": 430571376,
      "step": 8370,
      "train_runtime": 51592.2786,
      "train_tokens_per_second": 8345.655
    },
    {
      "epoch": 0.5638971182332346,
      "grad_norm": 0.5897424830474153,
      "learning_rate": 8.163886089321493e-06,
      "loss": 0.35391712188720703,
      "num_input_tokens_seen": 430843472,
      "step": 8375,
      "train_runtime": 51623.7262,
      "train_tokens_per_second": 8345.842
    },
    {
      "epoch": 0.5642337732291948,
      "grad_norm": 0.6468559257488488,
      "learning_rate": 8.161838244577795e-06,
      "loss": 0.35244431495666506,
      "num_input_tokens_seen": 431090768,
      "step": 8380,
      "train_runtime": 51655.5521,
      "train_tokens_per_second": 8345.488
    },
    {
      "epoch": 0.5645704282251549,
      "grad_norm": 0.628037782173836,
      "learning_rate": 8.159789515636114e-06,
      "loss": 0.35705389976501467,
      "num_input_tokens_seen": 431349480,
      "step": 8385,
      "train_runtime": 51687.2407,
      "train_tokens_per_second": 8345.376
    },
    {
      "epoch": 0.564907083221115,
      "grad_norm": 0.7152485221789631,
      "learning_rate": 8.15773990306937e-06,
      "loss": 0.3814382076263428,
      "num_input_tokens_seen": 431577744,
      "step": 8390,
      "train_runtime": 51718.3791,
      "train_tokens_per_second": 8344.765
    },
    {
      "epoch": 0.5652437382170752,
      "grad_norm": 1.165703430876709,
      "learning_rate": 8.155689407450732e-06,
      "loss": 0.37211730480194094,
      "num_input_tokens_seen": 431841160,
      "step": 8395,
      "train_runtime": 51750.4969,
      "train_tokens_per_second": 8344.677
    },
    {
      "epoch": 0.5655803932130353,
      "grad_norm": 0.7043636114910979,
      "learning_rate": 8.153638029353615e-06,
      "loss": 0.359681224822998,
      "num_input_tokens_seen": 432075200,
      "step": 8400,
      "train_runtime": 51776.062,
      "train_tokens_per_second": 8345.077
    },
    {
      "epoch": 0.5659170482089955,
      "grad_norm": 0.7326969537242338,
      "learning_rate": 8.151585769351678e-06,
      "loss": 0.39964632987976073,
      "num_input_tokens_seen": 432349032,
      "step": 8405,
      "train_runtime": 51804.8214,
      "train_tokens_per_second": 8345.73
    },
    {
      "epoch": 0.5662537032049556,
      "grad_norm": 0.7057252793956194,
      "learning_rate": 8.149532628018833e-06,
      "loss": 0.37884762287139895,
      "num_input_tokens_seen": 432596192,
      "step": 8410,
      "train_runtime": 51834.0833,
      "train_tokens_per_second": 8345.786
    },
    {
      "epoch": 0.5665903582009157,
      "grad_norm": 0.6716127016674683,
      "learning_rate": 8.147478605929228e-06,
      "loss": 0.3540342330932617,
      "num_input_tokens_seen": 432877792,
      "step": 8415,
      "train_runtime": 51869.5346,
      "train_tokens_per_second": 8345.511
    },
    {
      "epoch": 0.5669270131968759,
      "grad_norm": 0.6927393998020893,
      "learning_rate": 8.145423703657269e-06,
      "loss": 0.35318799018859864,
      "num_input_tokens_seen": 433143360,
      "step": 8420,
      "train_runtime": 51900.3582,
      "train_tokens_per_second": 8345.672
    },
    {
      "epoch": 0.567263668192836,
      "grad_norm": 0.5816888862087545,
      "learning_rate": 8.143367921777601e-06,
      "loss": 0.378446364402771,
      "num_input_tokens_seen": 433407960,
      "step": 8425,
      "train_runtime": 51931.8217,
      "train_tokens_per_second": 8345.711
    },
    {
      "epoch": 0.5676003231887962,
      "grad_norm": 0.60075721345595,
      "learning_rate": 8.141311260865117e-06,
      "loss": 0.40478291511535647,
      "num_input_tokens_seen": 433689472,
      "step": 8430,
      "train_runtime": 51964.5139,
      "train_tokens_per_second": 8345.878
    },
    {
      "epoch": 0.5679369781847563,
      "grad_norm": 0.6061182108830047,
      "learning_rate": 8.139253721494956e-06,
      "loss": 0.40162343978881837,
      "num_input_tokens_seen": 433938032,
      "step": 8435,
      "train_runtime": 51993.585,
      "train_tokens_per_second": 8345.992
    },
    {
      "epoch": 0.5682736331807164,
      "grad_norm": 0.6293946018431832,
      "learning_rate": 8.137195304242502e-06,
      "loss": 0.36252174377441404,
      "num_input_tokens_seen": 434196480,
      "step": 8440,
      "train_runtime": 52022.8849,
      "train_tokens_per_second": 8346.259
    },
    {
      "epoch": 0.5686102881766766,
      "grad_norm": 0.6111297345340696,
      "learning_rate": 8.135136009683387e-06,
      "loss": 0.3576167821884155,
      "num_input_tokens_seen": 434460520,
      "step": 8445,
      "train_runtime": 52053.3284,
      "train_tokens_per_second": 8346.45
    },
    {
      "epoch": 0.5689469431726367,
      "grad_norm": 0.6220114038871504,
      "learning_rate": 8.13307583839348e-06,
      "loss": 0.37273344993591306,
      "num_input_tokens_seen": 434723232,
      "step": 8450,
      "train_runtime": 52083.1581,
      "train_tokens_per_second": 8346.714
    },
    {
      "epoch": 0.5692835981685969,
      "grad_norm": 0.6558010514970117,
      "learning_rate": 8.131014790948908e-06,
      "loss": 0.39136567115783694,
      "num_input_tokens_seen": 434972088,
      "step": 8455,
      "train_runtime": 52112.8942,
      "train_tokens_per_second": 8346.727
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.7792999455012976,
      "learning_rate": 8.128952867926035e-06,
      "loss": 0.4125369548797607,
      "num_input_tokens_seen": 435235912,
      "step": 8460,
      "train_runtime": 52144.8018,
      "train_tokens_per_second": 8346.679
    },
    {
      "epoch": 0.5699569081605171,
      "grad_norm": 0.6133754497117616,
      "learning_rate": 8.126890069901468e-06,
      "loss": 0.38258388042449953,
      "num_input_tokens_seen": 435501552,
      "step": 8465,
      "train_runtime": 52176.6693,
      "train_tokens_per_second": 8346.672
    },
    {
      "epoch": 0.5702935631564773,
      "grad_norm": 0.597531991571466,
      "learning_rate": 8.124826397452064e-06,
      "loss": 0.3756241321563721,
      "num_input_tokens_seen": 435776632,
      "step": 8470,
      "train_runtime": 52208.154,
      "train_tokens_per_second": 8346.908
    },
    {
      "epoch": 0.5706302181524374,
      "grad_norm": 0.6566582125572921,
      "learning_rate": 8.122761851154923e-06,
      "loss": 0.36561896800994875,
      "num_input_tokens_seen": 436037264,
      "step": 8475,
      "train_runtime": 52243.5198,
      "train_tokens_per_second": 8346.246
    },
    {
      "epoch": 0.5709668731483976,
      "grad_norm": 0.7219277462916167,
      "learning_rate": 8.120696431587387e-06,
      "loss": 0.37534446716308595,
      "num_input_tokens_seen": 436291736,
      "step": 8480,
      "train_runtime": 52272.6885,
      "train_tokens_per_second": 8346.457
    },
    {
      "epoch": 0.5713035281443577,
      "grad_norm": 0.5916066519450789,
      "learning_rate": 8.118630139327046e-06,
      "loss": 0.35287330150604246,
      "num_input_tokens_seen": 436544168,
      "step": 8485,
      "train_runtime": 52305.8505,
      "train_tokens_per_second": 8345.991
    },
    {
      "epoch": 0.5716401831403178,
      "grad_norm": 0.546528345991408,
      "learning_rate": 8.116562974951731e-06,
      "loss": 0.39151749610900877,
      "num_input_tokens_seen": 436820968,
      "step": 8490,
      "train_runtime": 52337.0771,
      "train_tokens_per_second": 8346.3
    },
    {
      "epoch": 0.571976838136278,
      "grad_norm": 0.5793836876596657,
      "learning_rate": 8.11449493903952e-06,
      "loss": 0.3794196128845215,
      "num_input_tokens_seen": 437095104,
      "step": 8495,
      "train_runtime": 52370.7103,
      "train_tokens_per_second": 8346.175
    },
    {
      "epoch": 0.5723134931322381,
      "grad_norm": 0.6319061461364247,
      "learning_rate": 8.112426032168727e-06,
      "loss": 0.3823410987854004,
      "num_input_tokens_seen": 437349544,
      "step": 8500,
      "train_runtime": 52401.2454,
      "train_tokens_per_second": 8346.167
    },
    {
      "epoch": 0.5726501481281983,
      "grad_norm": 0.6162313898104138,
      "learning_rate": 8.11035625491792e-06,
      "loss": 0.3530158996582031,
      "num_input_tokens_seen": 437595752,
      "step": 8505,
      "train_runtime": 52428.9239,
      "train_tokens_per_second": 8346.457
    },
    {
      "epoch": 0.5729868031241584,
      "grad_norm": 0.6754240587782449,
      "learning_rate": 8.108285607865907e-06,
      "loss": 0.36786556243896484,
      "num_input_tokens_seen": 437848736,
      "step": 8510,
      "train_runtime": 52459.0615,
      "train_tokens_per_second": 8346.484
    },
    {
      "epoch": 0.5733234581201185,
      "grad_norm": 0.6944928618624996,
      "learning_rate": 8.106214091591731e-06,
      "loss": 0.3586641550064087,
      "num_input_tokens_seen": 438099048,
      "step": 8515,
      "train_runtime": 52489.3903,
      "train_tokens_per_second": 8346.43
    },
    {
      "epoch": 0.5736601131160787,
      "grad_norm": 0.6736665964808883,
      "learning_rate": 8.104141706674693e-06,
      "loss": 0.33544948101043703,
      "num_input_tokens_seen": 438366432,
      "step": 8520,
      "train_runtime": 52519.7868,
      "train_tokens_per_second": 8346.691
    },
    {
      "epoch": 0.5739967681120388,
      "grad_norm": 0.7119629975030851,
      "learning_rate": 8.102068453694324e-06,
      "loss": 0.3905355930328369,
      "num_input_tokens_seen": 438626088,
      "step": 8525,
      "train_runtime": 52550.6512,
      "train_tokens_per_second": 8346.73
    },
    {
      "epoch": 0.574333423107999,
      "grad_norm": 0.6338851962482754,
      "learning_rate": 8.099994333230404e-06,
      "loss": 0.37893812656402587,
      "num_input_tokens_seen": 438872192,
      "step": 8530,
      "train_runtime": 52584.5731,
      "train_tokens_per_second": 8346.026
    },
    {
      "epoch": 0.5746700781039591,
      "grad_norm": 1.0358723851773401,
      "learning_rate": 8.097919345862952e-06,
      "loss": 0.3809210777282715,
      "num_input_tokens_seen": 439111808,
      "step": 8535,
      "train_runtime": 52613.3893,
      "train_tokens_per_second": 8346.009
    },
    {
      "epoch": 0.5750067330999192,
      "grad_norm": 0.7204937350959398,
      "learning_rate": 8.095843492172235e-06,
      "loss": 0.3746265172958374,
      "num_input_tokens_seen": 439368232,
      "step": 8540,
      "train_runtime": 52640.5164,
      "train_tokens_per_second": 8346.579
    },
    {
      "epoch": 0.5753433880958794,
      "grad_norm": 0.7268012581990693,
      "learning_rate": 8.093766772738758e-06,
      "loss": 0.3204023361206055,
      "num_input_tokens_seen": 439615056,
      "step": 8545,
      "train_runtime": 52673.0659,
      "train_tokens_per_second": 8346.107
    },
    {
      "epoch": 0.5756800430918395,
      "grad_norm": 0.758629063325311,
      "learning_rate": 8.091689188143267e-06,
      "loss": 0.3806816339492798,
      "num_input_tokens_seen": 439856696,
      "step": 8550,
      "train_runtime": 52703.0804,
      "train_tokens_per_second": 8345.939
    },
    {
      "epoch": 0.5760166980877997,
      "grad_norm": 0.7625653418449093,
      "learning_rate": 8.089610738966754e-06,
      "loss": 0.37187392711639405,
      "num_input_tokens_seen": 440083128,
      "step": 8555,
      "train_runtime": 52731.0075,
      "train_tokens_per_second": 8345.813
    },
    {
      "epoch": 0.5763533530837598,
      "grad_norm": 0.7757997280946609,
      "learning_rate": 8.08753142579045e-06,
      "loss": 0.39926915168762206,
      "num_input_tokens_seen": 440343504,
      "step": 8560,
      "train_runtime": 52764.9171,
      "train_tokens_per_second": 8345.384
    },
    {
      "epoch": 0.5766900080797199,
      "grad_norm": 0.6975991476665848,
      "learning_rate": 8.085451249195828e-06,
      "loss": 0.3727346181869507,
      "num_input_tokens_seen": 440621016,
      "step": 8565,
      "train_runtime": 52797.9323,
      "train_tokens_per_second": 8345.422
    },
    {
      "epoch": 0.5770266630756801,
      "grad_norm": 0.6626869344738391,
      "learning_rate": 8.083370209764603e-06,
      "loss": 0.384766960144043,
      "num_input_tokens_seen": 440875984,
      "step": 8570,
      "train_runtime": 52828.6769,
      "train_tokens_per_second": 8345.391
    },
    {
      "epoch": 0.5773633180716402,
      "grad_norm": 0.5866775649079198,
      "learning_rate": 8.081288308078731e-06,
      "loss": 0.3703623294830322,
      "num_input_tokens_seen": 441166744,
      "step": 8575,
      "train_runtime": 52862.144,
      "train_tokens_per_second": 8345.608
    },
    {
      "epoch": 0.5776999730676003,
      "grad_norm": 0.6063497327955069,
      "learning_rate": 8.07920554472041e-06,
      "loss": 0.3540030479431152,
      "num_input_tokens_seen": 441426960,
      "step": 8580,
      "train_runtime": 52892.1721,
      "train_tokens_per_second": 8345.79
    },
    {
      "epoch": 0.5780366280635605,
      "grad_norm": 0.7338797404042379,
      "learning_rate": 8.077121920272075e-06,
      "loss": 0.38886799812316897,
      "num_input_tokens_seen": 441703496,
      "step": 8585,
      "train_runtime": 52925.9722,
      "train_tokens_per_second": 8345.685
    },
    {
      "epoch": 0.5783732830595206,
      "grad_norm": 0.6125703483504972,
      "learning_rate": 8.07503743531641e-06,
      "loss": 0.3385774612426758,
      "num_input_tokens_seen": 441945872,
      "step": 8590,
      "train_runtime": 52954.2421,
      "train_tokens_per_second": 8345.807
    },
    {
      "epoch": 0.5787099380554808,
      "grad_norm": 0.6490724759538308,
      "learning_rate": 8.072952090436328e-06,
      "loss": 0.3477421760559082,
      "num_input_tokens_seen": 442227760,
      "step": 8595,
      "train_runtime": 52985.578,
      "train_tokens_per_second": 8346.191
    },
    {
      "epoch": 0.5790465930514409,
      "grad_norm": 0.7210582283837331,
      "learning_rate": 8.070865886214996e-06,
      "loss": 0.3642889976501465,
      "num_input_tokens_seen": 442496064,
      "step": 8600,
      "train_runtime": 53016.3049,
      "train_tokens_per_second": 8346.415
    },
    {
      "epoch": 0.579383248047401,
      "grad_norm": 0.6839942493075155,
      "learning_rate": 8.06877882323581e-06,
      "loss": 0.38209397792816163,
      "num_input_tokens_seen": 442769112,
      "step": 8605,
      "train_runtime": 53045.6257,
      "train_tokens_per_second": 8346.949
    },
    {
      "epoch": 0.5797199030433612,
      "grad_norm": 0.6233147474525101,
      "learning_rate": 8.066690902082413e-06,
      "loss": 0.36062116622924806,
      "num_input_tokens_seen": 443039320,
      "step": 8610,
      "train_runtime": 53076.196,
      "train_tokens_per_second": 8347.232
    },
    {
      "epoch": 0.5800565580393213,
      "grad_norm": 0.6255073982338176,
      "learning_rate": 8.064602123338683e-06,
      "loss": 0.4078712463378906,
      "num_input_tokens_seen": 443330056,
      "step": 8615,
      "train_runtime": 53106.6931,
      "train_tokens_per_second": 8347.913
    },
    {
      "epoch": 0.5803932130352815,
      "grad_norm": 0.6813323743631753,
      "learning_rate": 8.06251248758874e-06,
      "loss": 0.37433183193206787,
      "num_input_tokens_seen": 443566000,
      "step": 8620,
      "train_runtime": 53134.8379,
      "train_tokens_per_second": 8347.932
    },
    {
      "epoch": 0.5807298680312416,
      "grad_norm": 0.7467990251385693,
      "learning_rate": 8.060421995416949e-06,
      "loss": 0.41181111335754395,
      "num_input_tokens_seen": 443792064,
      "step": 8625,
      "train_runtime": 53164.3657,
      "train_tokens_per_second": 8347.547
    },
    {
      "epoch": 0.5810665230272017,
      "grad_norm": 0.7349356493342193,
      "learning_rate": 8.058330647407904e-06,
      "loss": 0.3679357051849365,
      "num_input_tokens_seen": 444043024,
      "step": 8630,
      "train_runtime": 53192.395,
      "train_tokens_per_second": 8347.867
    },
    {
      "epoch": 0.5814031780231619,
      "grad_norm": 0.7161916626419421,
      "learning_rate": 8.056238444146445e-06,
      "loss": 0.38098907470703125,
      "num_input_tokens_seen": 444303464,
      "step": 8635,
      "train_runtime": 53218.526,
      "train_tokens_per_second": 8348.662
    },
    {
      "epoch": 0.581739833019122,
      "grad_norm": 0.6838212156778245,
      "learning_rate": 8.054145386217652e-06,
      "loss": 0.34535808563232423,
      "num_input_tokens_seen": 444575640,
      "step": 8640,
      "train_runtime": 53244.5146,
      "train_tokens_per_second": 8349.698
    },
    {
      "epoch": 0.5820764880150822,
      "grad_norm": 0.8584050928222599,
      "learning_rate": 8.05205147420684e-06,
      "loss": 0.35669693946838377,
      "num_input_tokens_seen": 444830208,
      "step": 8645,
      "train_runtime": 53276.4647,
      "train_tokens_per_second": 8349.469
    },
    {
      "epoch": 0.5824131430110423,
      "grad_norm": 0.7634519343265359,
      "learning_rate": 8.049956708699567e-06,
      "loss": 0.41454405784606935,
      "num_input_tokens_seen": 445099608,
      "step": 8650,
      "train_runtime": 53304.6135,
      "train_tokens_per_second": 8350.114
    },
    {
      "epoch": 0.5827497980070024,
      "grad_norm": 0.6604557986256854,
      "learning_rate": 8.047861090281623e-06,
      "loss": 0.3907956600189209,
      "num_input_tokens_seen": 445351192,
      "step": 8655,
      "train_runtime": 53334.9097,
      "train_tokens_per_second": 8350.088
    },
    {
      "epoch": 0.5830864530029626,
      "grad_norm": 0.6128443194099943,
      "learning_rate": 8.045764619539047e-06,
      "loss": 0.35973873138427737,
      "num_input_tokens_seen": 445614976,
      "step": 8660,
      "train_runtime": 53363.9213,
      "train_tokens_per_second": 8350.492
    },
    {
      "epoch": 0.5834231079989227,
      "grad_norm": 0.7781140634588403,
      "learning_rate": 8.043667297058106e-06,
      "loss": 0.3556912660598755,
      "num_input_tokens_seen": 445871792,
      "step": 8665,
      "train_runtime": 53390.1141,
      "train_tokens_per_second": 8351.205
    },
    {
      "epoch": 0.5837597629948829,
      "grad_norm": 0.5868726998668037,
      "learning_rate": 8.04156912342531e-06,
      "loss": 0.38425097465515134,
      "num_input_tokens_seen": 446146896,
      "step": 8670,
      "train_runtime": 53420.6735,
      "train_tokens_per_second": 8351.578
    },
    {
      "epoch": 0.584096417990843,
      "grad_norm": 0.7361270051641159,
      "learning_rate": 8.03947009922741e-06,
      "loss": 0.36755008697509767,
      "num_input_tokens_seen": 446409088,
      "step": 8675,
      "train_runtime": 53449.0218,
      "train_tokens_per_second": 8352.053
    },
    {
      "epoch": 0.5844330729868031,
      "grad_norm": 0.6583650411115769,
      "learning_rate": 8.037370225051387e-06,
      "loss": 0.4112565994262695,
      "num_input_tokens_seen": 446656776,
      "step": 8680,
      "train_runtime": 53479.422,
      "train_tokens_per_second": 8351.937
    },
    {
      "epoch": 0.5847697279827633,
      "grad_norm": 0.6236478614516069,
      "learning_rate": 8.035269501484467e-06,
      "loss": 0.3739811420440674,
      "num_input_tokens_seen": 446930448,
      "step": 8685,
      "train_runtime": 53517.033,
      "train_tokens_per_second": 8351.181
    },
    {
      "epoch": 0.5851063829787234,
      "grad_norm": 0.654142257716481,
      "learning_rate": 8.03316792911411e-06,
      "loss": 0.4086054801940918,
      "num_input_tokens_seen": 447190456,
      "step": 8690,
      "train_runtime": 53546.0294,
      "train_tokens_per_second": 8351.515
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.5990267827700653,
      "learning_rate": 8.031065508528011e-06,
      "loss": 0.34234137535095216,
      "num_input_tokens_seen": 447470088,
      "step": 8695,
      "train_runtime": 53574.0911,
      "train_tokens_per_second": 8352.36
    },
    {
      "epoch": 0.5857796929706437,
      "grad_norm": 0.5974272378102219,
      "learning_rate": 8.028962240314111e-06,
      "loss": 0.35796351432800294,
      "num_input_tokens_seen": 447737040,
      "step": 8700,
      "train_runtime": 53608.0868,
      "train_tokens_per_second": 8352.043
    },
    {
      "epoch": 0.5861163479666038,
      "grad_norm": 0.6437767366095397,
      "learning_rate": 8.026858125060577e-06,
      "loss": 0.39219210147857664,
      "num_input_tokens_seen": 447983248,
      "step": 8705,
      "train_runtime": 53636.8525,
      "train_tokens_per_second": 8352.154
    },
    {
      "epoch": 0.586453002962564,
      "grad_norm": 0.679468766631222,
      "learning_rate": 8.024753163355821e-06,
      "loss": 0.4106016159057617,
      "num_input_tokens_seen": 448247080,
      "step": 8710,
      "train_runtime": 53668.8312,
      "train_tokens_per_second": 8352.093
    },
    {
      "epoch": 0.5867896579585241,
      "grad_norm": 1.0571881017148985,
      "learning_rate": 8.022647355788489e-06,
      "loss": 0.367664098739624,
      "num_input_tokens_seen": 448504584,
      "step": 8715,
      "train_runtime": 53697.3841,
      "train_tokens_per_second": 8352.448
    },
    {
      "epoch": 0.5871263129544843,
      "grad_norm": 0.648290603320209,
      "learning_rate": 8.02054070294746e-06,
      "loss": 0.3927702188491821,
      "num_input_tokens_seen": 448754512,
      "step": 8720,
      "train_runtime": 53730.2052,
      "train_tokens_per_second": 8351.997
    },
    {
      "epoch": 0.5874629679504444,
      "grad_norm": 0.6609678665517676,
      "learning_rate": 8.018433205421859e-06,
      "loss": 0.3544990062713623,
      "num_input_tokens_seen": 449036616,
      "step": 8725,
      "train_runtime": 53756.761,
      "train_tokens_per_second": 8353.119
    },
    {
      "epoch": 0.5877996229464045,
      "grad_norm": 0.6614086148095739,
      "learning_rate": 8.016324863801036e-06,
      "loss": 0.3813911199569702,
      "num_input_tokens_seen": 449305776,
      "step": 8730,
      "train_runtime": 53790.9495,
      "train_tokens_per_second": 8352.814
    },
    {
      "epoch": 0.5881362779423647,
      "grad_norm": 0.6030798345932101,
      "learning_rate": 8.014215678674581e-06,
      "loss": 0.4030319690704346,
      "num_input_tokens_seen": 449568480,
      "step": 8735,
      "train_runtime": 53817.175,
      "train_tokens_per_second": 8353.625
    },
    {
      "epoch": 0.5884729329383248,
      "grad_norm": 0.5783912514040819,
      "learning_rate": 8.012105650632325e-06,
      "loss": 0.3609976291656494,
      "num_input_tokens_seen": 449815824,
      "step": 8740,
      "train_runtime": 53846.1273,
      "train_tokens_per_second": 8353.727
    },
    {
      "epoch": 0.588809587934285,
      "grad_norm": 0.6591283360586161,
      "learning_rate": 8.009994780264328e-06,
      "loss": 0.3852669715881348,
      "num_input_tokens_seen": 450036544,
      "step": 8745,
      "train_runtime": 53872.7784,
      "train_tokens_per_second": 8353.691
    },
    {
      "epoch": 0.5891462429302451,
      "grad_norm": 0.6974331402646584,
      "learning_rate": 8.00788306816089e-06,
      "loss": 0.3739362955093384,
      "num_input_tokens_seen": 450294928,
      "step": 8750,
      "train_runtime": 53906.098,
      "train_tokens_per_second": 8353.321
    },
    {
      "epoch": 0.5894828979262052,
      "grad_norm": 0.7307998997425771,
      "learning_rate": 8.005770514912543e-06,
      "loss": 0.38839290142059324,
      "num_input_tokens_seen": 450551624,
      "step": 8755,
      "train_runtime": 53933.8624,
      "train_tokens_per_second": 8353.78
    },
    {
      "epoch": 0.5898195529221654,
      "grad_norm": 0.6720530199854122,
      "learning_rate": 8.003657121110055e-06,
      "loss": 0.3778205394744873,
      "num_input_tokens_seen": 450826696,
      "step": 8760,
      "train_runtime": 53965.4675,
      "train_tokens_per_second": 8353.985
    },
    {
      "epoch": 0.5901562079181255,
      "grad_norm": 0.5899838155255056,
      "learning_rate": 8.001542887344432e-06,
      "loss": 0.3586574077606201,
      "num_input_tokens_seen": 451093472,
      "step": 8765,
      "train_runtime": 53998.2985,
      "train_tokens_per_second": 8353.846
    },
    {
      "epoch": 0.5904928629140856,
      "grad_norm": 0.6210418267066049,
      "learning_rate": 7.999427814206911e-06,
      "loss": 0.3756500244140625,
      "num_input_tokens_seen": 451357480,
      "step": 8770,
      "train_runtime": 54030.8969,
      "train_tokens_per_second": 8353.692
    },
    {
      "epoch": 0.5908295179100458,
      "grad_norm": 0.6957993623731414,
      "learning_rate": 7.997311902288968e-06,
      "loss": 0.3453741312026978,
      "num_input_tokens_seen": 451621432,
      "step": 8775,
      "train_runtime": 54059.6851,
      "train_tokens_per_second": 8354.126
    },
    {
      "epoch": 0.5911661729060059,
      "grad_norm": 0.8289861070482021,
      "learning_rate": 7.995195152182307e-06,
      "loss": 0.3358358383178711,
      "num_input_tokens_seen": 451855904,
      "step": 8780,
      "train_runtime": 54085.101,
      "train_tokens_per_second": 8354.536
    },
    {
      "epoch": 0.5915028279019661,
      "grad_norm": 0.7098611826248911,
      "learning_rate": 7.993077564478875e-06,
      "loss": 0.3639656066894531,
      "num_input_tokens_seen": 452131192,
      "step": 8785,
      "train_runtime": 54117.2796,
      "train_tokens_per_second": 8354.655
    },
    {
      "epoch": 0.5918394828979262,
      "grad_norm": 0.7005552030882023,
      "learning_rate": 7.990959139770844e-06,
      "loss": 0.397585391998291,
      "num_input_tokens_seen": 452401696,
      "step": 8790,
      "train_runtime": 54149.7304,
      "train_tokens_per_second": 8354.644
    },
    {
      "epoch": 0.5921761378938863,
      "grad_norm": 0.6542571754996472,
      "learning_rate": 7.988839878650628e-06,
      "loss": 0.3831373453140259,
      "num_input_tokens_seen": 452671248,
      "step": 8795,
      "train_runtime": 54180.7394,
      "train_tokens_per_second": 8354.837
    },
    {
      "epoch": 0.5925127928898465,
      "grad_norm": 0.6656953571327513,
      "learning_rate": 7.986719781710872e-06,
      "loss": 0.390764856338501,
      "num_input_tokens_seen": 452933776,
      "step": 8800,
      "train_runtime": 54210.5142,
      "train_tokens_per_second": 8355.091
    },
    {
      "epoch": 0.5928494478858066,
      "grad_norm": 0.6903151877760226,
      "learning_rate": 7.984598849544452e-06,
      "loss": 0.3716871738433838,
      "num_input_tokens_seen": 453184208,
      "step": 8805,
      "train_runtime": 54241.8401,
      "train_tokens_per_second": 8354.883
    },
    {
      "epoch": 0.5931861028817668,
      "grad_norm": 0.7244982407324604,
      "learning_rate": 7.98247708274448e-06,
      "loss": 0.3817361116409302,
      "num_input_tokens_seen": 453431312,
      "step": 8810,
      "train_runtime": 54270.522,
      "train_tokens_per_second": 8355.02
    },
    {
      "epoch": 0.5935227578777269,
      "grad_norm": 0.6676862903264942,
      "learning_rate": 7.980354481904303e-06,
      "loss": 0.4046356201171875,
      "num_input_tokens_seen": 453697552,
      "step": 8815,
      "train_runtime": 54301.179,
      "train_tokens_per_second": 8355.206
    },
    {
      "epoch": 0.593859412873687,
      "grad_norm": 0.7014127862195861,
      "learning_rate": 7.978231047617498e-06,
      "loss": 0.39393370151519774,
      "num_input_tokens_seen": 453952536,
      "step": 8820,
      "train_runtime": 54328.4748,
      "train_tokens_per_second": 8355.702
    },
    {
      "epoch": 0.5941960678696472,
      "grad_norm": 0.6481979476873357,
      "learning_rate": 7.976106780477878e-06,
      "loss": 0.39094018936157227,
      "num_input_tokens_seen": 454216672,
      "step": 8825,
      "train_runtime": 54360.5113,
      "train_tokens_per_second": 8355.636
    },
    {
      "epoch": 0.5945327228656073,
      "grad_norm": 0.7077470240707667,
      "learning_rate": 7.973981681079487e-06,
      "loss": 0.36669490337371824,
      "num_input_tokens_seen": 454446184,
      "step": 8830,
      "train_runtime": 54389.8914,
      "train_tokens_per_second": 8355.343
    },
    {
      "epoch": 0.5948693778615675,
      "grad_norm": 0.7019556025101246,
      "learning_rate": 7.971855750016603e-06,
      "loss": 0.36378610134124756,
      "num_input_tokens_seen": 454705544,
      "step": 8835,
      "train_runtime": 54422.0543,
      "train_tokens_per_second": 8355.171
    },
    {
      "epoch": 0.5952060328575276,
      "grad_norm": 0.6959249886502754,
      "learning_rate": 7.969728987883734e-06,
      "loss": 0.37087416648864746,
      "num_input_tokens_seen": 454965968,
      "step": 8840,
      "train_runtime": 54452.6315,
      "train_tokens_per_second": 8355.261
    },
    {
      "epoch": 0.5955426878534877,
      "grad_norm": 0.7503111898313857,
      "learning_rate": 7.967601395275624e-06,
      "loss": 0.375437593460083,
      "num_input_tokens_seen": 455216784,
      "step": 8845,
      "train_runtime": 54481.8555,
      "train_tokens_per_second": 8355.383
    },
    {
      "epoch": 0.5958793428494479,
      "grad_norm": 0.6324587511950892,
      "learning_rate": 7.965472972787247e-06,
      "loss": 0.35130293369293214,
      "num_input_tokens_seen": 455481056,
      "step": 8850,
      "train_runtime": 54515.1473,
      "train_tokens_per_second": 8355.128
    },
    {
      "epoch": 0.596215997845408,
      "grad_norm": 0.6369859478166993,
      "learning_rate": 7.963343721013808e-06,
      "loss": 0.36407558917999266,
      "num_input_tokens_seen": 455772152,
      "step": 8855,
      "train_runtime": 54546.3251,
      "train_tokens_per_second": 8355.689
    },
    {
      "epoch": 0.5965526528413682,
      "grad_norm": 0.708286734097933,
      "learning_rate": 7.961213640550748e-06,
      "loss": 0.3833615303039551,
      "num_input_tokens_seen": 456028856,
      "step": 8860,
      "train_runtime": 54573.8908,
      "train_tokens_per_second": 8356.173
    },
    {
      "epoch": 0.5968893078373283,
      "grad_norm": 0.7073558298719889,
      "learning_rate": 7.959082731993736e-06,
      "loss": 0.35933399200439453,
      "num_input_tokens_seen": 456288392,
      "step": 8865,
      "train_runtime": 54605.2235,
      "train_tokens_per_second": 8356.131
    },
    {
      "epoch": 0.5972259628332884,
      "grad_norm": 0.7379102444814177,
      "learning_rate": 7.956950995938673e-06,
      "loss": 0.3599724531173706,
      "num_input_tokens_seen": 456522744,
      "step": 8870,
      "train_runtime": 54636.2479,
      "train_tokens_per_second": 8355.675
    },
    {
      "epoch": 0.5975626178292486,
      "grad_norm": 0.8171911995639747,
      "learning_rate": 7.954818432981694e-06,
      "loss": 0.3825263500213623,
      "num_input_tokens_seen": 456748896,
      "step": 8875,
      "train_runtime": 54664.4079,
      "train_tokens_per_second": 8355.508
    },
    {
      "epoch": 0.5978992728252087,
      "grad_norm": 0.7135446283858827,
      "learning_rate": 7.952685043719162e-06,
      "loss": 0.3582231760025024,
      "num_input_tokens_seen": 457014144,
      "step": 8880,
      "train_runtime": 54694.6031,
      "train_tokens_per_second": 8355.745
    },
    {
      "epoch": 0.5982359278211689,
      "grad_norm": 0.7168812998255878,
      "learning_rate": 7.950550828747672e-06,
      "loss": 0.3767875671386719,
      "num_input_tokens_seen": 457272048,
      "step": 8885,
      "train_runtime": 54725.6967,
      "train_tokens_per_second": 8355.71
    },
    {
      "epoch": 0.598572582817129,
      "grad_norm": 0.598834198461999,
      "learning_rate": 7.948415788664052e-06,
      "loss": 0.3991069316864014,
      "num_input_tokens_seen": 457547608,
      "step": 8890,
      "train_runtime": 54758.7428,
      "train_tokens_per_second": 8355.7
    },
    {
      "epoch": 0.5989092378130891,
      "grad_norm": 0.8538206120084046,
      "learning_rate": 7.946279924065359e-06,
      "loss": 0.37126684188842773,
      "num_input_tokens_seen": 457815000,
      "step": 8895,
      "train_runtime": 54789.0342,
      "train_tokens_per_second": 8355.96
    },
    {
      "epoch": 0.5992458928090493,
      "grad_norm": 0.7022278091658763,
      "learning_rate": 7.94414323554888e-06,
      "loss": 0.3550544261932373,
      "num_input_tokens_seen": 458103232,
      "step": 8900,
      "train_runtime": 54818.9204,
      "train_tokens_per_second": 8356.663
    },
    {
      "epoch": 0.5995825478050094,
      "grad_norm": 0.6209223690931961,
      "learning_rate": 7.942005723712135e-06,
      "loss": 0.3623520374298096,
      "num_input_tokens_seen": 458387832,
      "step": 8905,
      "train_runtime": 54851.6525,
      "train_tokens_per_second": 8356.865
    },
    {
      "epoch": 0.5999192028009696,
      "grad_norm": 0.6541450910932609,
      "learning_rate": 7.939867389152868e-06,
      "loss": 0.35341243743896483,
      "num_input_tokens_seen": 458648160,
      "step": 8910,
      "train_runtime": 54884.9028,
      "train_tokens_per_second": 8356.545
    },
    {
      "epoch": 0.6002558577969297,
      "grad_norm": 0.8328501347852177,
      "learning_rate": 7.937728232469062e-06,
      "loss": 0.3771503925323486,
      "num_input_tokens_seen": 458891744,
      "step": 8915,
      "train_runtime": 54913.6701,
      "train_tokens_per_second": 8356.603
    },
    {
      "epoch": 0.6005925127928898,
      "grad_norm": 0.6801145567615978,
      "learning_rate": 7.935588254258921e-06,
      "loss": 0.4053555965423584,
      "num_input_tokens_seen": 459133872,
      "step": 8920,
      "train_runtime": 54944.0219,
      "train_tokens_per_second": 8356.394
    },
    {
      "epoch": 0.60092916778885,
      "grad_norm": 0.5941350294201105,
      "learning_rate": 7.933447455120889e-06,
      "loss": 0.3501303195953369,
      "num_input_tokens_seen": 459379392,
      "step": 8925,
      "train_runtime": 54972.0563,
      "train_tokens_per_second": 8356.598
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.859777304087846,
      "learning_rate": 7.931305835653629e-06,
      "loss": 0.3409982204437256,
      "num_input_tokens_seen": 459652008,
      "step": 8930,
      "train_runtime": 55006.3397,
      "train_tokens_per_second": 8356.346
    },
    {
      "epoch": 0.6016024777807703,
      "grad_norm": 0.7470743014409066,
      "learning_rate": 7.929163396456038e-06,
      "loss": 0.3885748624801636,
      "num_input_tokens_seen": 459892736,
      "step": 8935,
      "train_runtime": 55035.9167,
      "train_tokens_per_second": 8356.229
    },
    {
      "epoch": 0.6019391327767304,
      "grad_norm": 0.7081910313578139,
      "learning_rate": 7.927020138127245e-06,
      "loss": 0.3569533348083496,
      "num_input_tokens_seen": 460153792,
      "step": 8940,
      "train_runtime": 55066.2751,
      "train_tokens_per_second": 8356.363
    },
    {
      "epoch": 0.6022757877726905,
      "grad_norm": 0.5867650822258289,
      "learning_rate": 7.924876061266603e-06,
      "loss": 0.33462550640106203,
      "num_input_tokens_seen": 460438576,
      "step": 8945,
      "train_runtime": 55100.2434,
      "train_tokens_per_second": 8356.38
    },
    {
      "epoch": 0.6026124427686507,
      "grad_norm": 0.7844785207464715,
      "learning_rate": 7.922731166473698e-06,
      "loss": 0.39972786903381347,
      "num_input_tokens_seen": 460689944,
      "step": 8950,
      "train_runtime": 55134.9592,
      "train_tokens_per_second": 8355.678
    },
    {
      "epoch": 0.6029490977646108,
      "grad_norm": 0.6832842609871558,
      "learning_rate": 7.920585454348341e-06,
      "loss": 0.34897356033325194,
      "num_input_tokens_seen": 460939024,
      "step": 8955,
      "train_runtime": 55166.4493,
      "train_tokens_per_second": 8355.423
    },
    {
      "epoch": 0.603285752760571,
      "grad_norm": 1.1724496491650054,
      "learning_rate": 7.918438925490578e-06,
      "loss": 0.34812901020050047,
      "num_input_tokens_seen": 461208064,
      "step": 8960,
      "train_runtime": 55196.6003,
      "train_tokens_per_second": 8355.733
    },
    {
      "epoch": 0.6036224077565311,
      "grad_norm": 0.7944213449828361,
      "learning_rate": 7.916291580500672e-06,
      "loss": 0.3546807050704956,
      "num_input_tokens_seen": 461455536,
      "step": 8965,
      "train_runtime": 55230.4785,
      "train_tokens_per_second": 8355.089
    },
    {
      "epoch": 0.6039590627524912,
      "grad_norm": 0.6211312463795816,
      "learning_rate": 7.914143419979127e-06,
      "loss": 0.35325026512145996,
      "num_input_tokens_seen": 461723456,
      "step": 8970,
      "train_runtime": 55257.3435,
      "train_tokens_per_second": 8355.875
    },
    {
      "epoch": 0.6042957177484514,
      "grad_norm": 0.5985651500934066,
      "learning_rate": 7.911994444526667e-06,
      "loss": 0.329397177696228,
      "num_input_tokens_seen": 461988640,
      "step": 8975,
      "train_runtime": 55287.3733,
      "train_tokens_per_second": 8356.133
    },
    {
      "epoch": 0.6046323727444115,
      "grad_norm": 0.6804822740291245,
      "learning_rate": 7.909844654744246e-06,
      "loss": 0.3479720115661621,
      "num_input_tokens_seen": 462244240,
      "step": 8980,
      "train_runtime": 55318.9277,
      "train_tokens_per_second": 8355.987
    },
    {
      "epoch": 0.6049690277403716,
      "grad_norm": 0.6728507985098182,
      "learning_rate": 7.907694051233044e-06,
      "loss": 0.3725043058395386,
      "num_input_tokens_seen": 462499896,
      "step": 8985,
      "train_runtime": 55350.6229,
      "train_tokens_per_second": 8355.821
    },
    {
      "epoch": 0.6053056827363318,
      "grad_norm": 0.5234536714017051,
      "learning_rate": 7.905542634594476e-06,
      "loss": 0.38381028175354004,
      "num_input_tokens_seen": 462774904,
      "step": 8990,
      "train_runtime": 55383.4049,
      "train_tokens_per_second": 8355.841
    },
    {
      "epoch": 0.6056423377322919,
      "grad_norm": 0.7137318527239892,
      "learning_rate": 7.903390405430174e-06,
      "loss": 0.3634381055831909,
      "num_input_tokens_seen": 463041720,
      "step": 8995,
      "train_runtime": 55416.6281,
      "train_tokens_per_second": 8355.646
    },
    {
      "epoch": 0.6059789927282521,
      "grad_norm": 0.6840245808591199,
      "learning_rate": 7.901237364342003e-06,
      "loss": 0.35892040729522706,
      "num_input_tokens_seen": 463282784,
      "step": 9000,
      "train_runtime": 55447.5676,
      "train_tokens_per_second": 8355.331
    },
    {
      "epoch": 0.6063156477242122,
      "grad_norm": 0.5900456170245686,
      "learning_rate": 7.899083511932053e-06,
      "loss": 0.33178348541259767,
      "num_input_tokens_seen": 463533656,
      "step": 9005,
      "train_runtime": 55477.4624,
      "train_tokens_per_second": 8355.351
    },
    {
      "epoch": 0.6066523027201723,
      "grad_norm": 0.6168375921730539,
      "learning_rate": 7.896928848802645e-06,
      "loss": 0.36444332599639895,
      "num_input_tokens_seen": 463801592,
      "step": 9010,
      "train_runtime": 55510.6328,
      "train_tokens_per_second": 8355.185
    },
    {
      "epoch": 0.6069889577161325,
      "grad_norm": 0.7134960476271138,
      "learning_rate": 7.894773375556322e-06,
      "loss": 0.38894174098968504,
      "num_input_tokens_seen": 464017456,
      "step": 9015,
      "train_runtime": 55542.2583,
      "train_tokens_per_second": 8354.314
    },
    {
      "epoch": 0.6073256127120926,
      "grad_norm": 0.7235315588718757,
      "learning_rate": 7.892617092795855e-06,
      "loss": 0.3951845407485962,
      "num_input_tokens_seen": 464286816,
      "step": 9020,
      "train_runtime": 55574.756,
      "train_tokens_per_second": 8354.275
    },
    {
      "epoch": 0.6076622677080528,
      "grad_norm": 0.7017135059583793,
      "learning_rate": 7.890460001124242e-06,
      "loss": 0.37053399085998534,
      "num_input_tokens_seen": 464535336,
      "step": 9025,
      "train_runtime": 55607.5324,
      "train_tokens_per_second": 8353.82
    },
    {
      "epoch": 0.6079989227040129,
      "grad_norm": 0.8090472520867731,
      "learning_rate": 7.888302101144707e-06,
      "loss": 0.3741230726242065,
      "num_input_tokens_seen": 464776712,
      "step": 9030,
      "train_runtime": 55639.5517,
      "train_tokens_per_second": 8353.351
    },
    {
      "epoch": 0.608335577699973,
      "grad_norm": 0.6415211848871963,
      "learning_rate": 7.886143393460699e-06,
      "loss": 0.39028186798095704,
      "num_input_tokens_seen": 465060008,
      "step": 9035,
      "train_runtime": 55674.8877,
      "train_tokens_per_second": 8353.138
    },
    {
      "epoch": 0.6086722326959332,
      "grad_norm": 0.6107184174027389,
      "learning_rate": 7.883983878675895e-06,
      "loss": 0.35847771167755127,
      "num_input_tokens_seen": 465336184,
      "step": 9040,
      "train_runtime": 55705.067,
      "train_tokens_per_second": 8353.57
    },
    {
      "epoch": 0.6090088876918933,
      "grad_norm": 0.66760217147693,
      "learning_rate": 7.881823557394194e-06,
      "loss": 0.3705006122589111,
      "num_input_tokens_seen": 465602256,
      "step": 9045,
      "train_runtime": 55738.3738,
      "train_tokens_per_second": 8353.352
    },
    {
      "epoch": 0.6093455426878535,
      "grad_norm": 0.6956223331372131,
      "learning_rate": 7.879662430219728e-06,
      "loss": 0.35218517780303954,
      "num_input_tokens_seen": 465824040,
      "step": 9050,
      "train_runtime": 55766.3358,
      "train_tokens_per_second": 8353.141
    },
    {
      "epoch": 0.6096821976838136,
      "grad_norm": 0.7728638778879421,
      "learning_rate": 7.877500497756843e-06,
      "loss": 0.3936324596405029,
      "num_input_tokens_seen": 466073560,
      "step": 9055,
      "train_runtime": 55796.4367,
      "train_tokens_per_second": 8353.106
    },
    {
      "epoch": 0.6100188526797737,
      "grad_norm": 0.7592125884173825,
      "learning_rate": 7.87533776061012e-06,
      "loss": 0.3833607196807861,
      "num_input_tokens_seen": 466338624,
      "step": 9060,
      "train_runtime": 55829.7803,
      "train_tokens_per_second": 8352.865
    },
    {
      "epoch": 0.6103555076757339,
      "grad_norm": 0.646505969897749,
      "learning_rate": 7.873174219384362e-06,
      "loss": 0.38436715602874755,
      "num_input_tokens_seen": 466588008,
      "step": 9065,
      "train_runtime": 55860.4792,
      "train_tokens_per_second": 8352.739
    },
    {
      "epoch": 0.610692162671694,
      "grad_norm": 0.6694688508965312,
      "learning_rate": 7.871009874684595e-06,
      "loss": 0.39240255355834963,
      "num_input_tokens_seen": 466868800,
      "step": 9070,
      "train_runtime": 55893.1644,
      "train_tokens_per_second": 8352.878
    },
    {
      "epoch": 0.6110288176676542,
      "grad_norm": 0.6730577138798975,
      "learning_rate": 7.868844727116072e-06,
      "loss": 0.3798365116119385,
      "num_input_tokens_seen": 467136888,
      "step": 9075,
      "train_runtime": 55924.6422,
      "train_tokens_per_second": 8352.971
    },
    {
      "epoch": 0.6113654726636143,
      "grad_norm": 0.5922088669708675,
      "learning_rate": 7.866678777284267e-06,
      "loss": 0.35559816360473634,
      "num_input_tokens_seen": 467412912,
      "step": 9080,
      "train_runtime": 55957.9512,
      "train_tokens_per_second": 8352.931
    },
    {
      "epoch": 0.6117021276595744,
      "grad_norm": 0.6337634899606663,
      "learning_rate": 7.864512025794886e-06,
      "loss": 0.36407415866851806,
      "num_input_tokens_seen": 467670000,
      "step": 9085,
      "train_runtime": 55990.0004,
      "train_tokens_per_second": 8352.742
    },
    {
      "epoch": 0.6120387826555346,
      "grad_norm": 0.6251546779597406,
      "learning_rate": 7.86234447325385e-06,
      "loss": 0.3587490081787109,
      "num_input_tokens_seen": 467917792,
      "step": 9090,
      "train_runtime": 56020.8377,
      "train_tokens_per_second": 8352.567
    },
    {
      "epoch": 0.6123754376514947,
      "grad_norm": 0.6210415624879883,
      "learning_rate": 7.860176120267309e-06,
      "loss": 0.41840667724609376,
      "num_input_tokens_seen": 468205024,
      "step": 9095,
      "train_runtime": 56051.6383,
      "train_tokens_per_second": 8353.102
    },
    {
      "epoch": 0.6127120926474549,
      "grad_norm": 0.6978192031598606,
      "learning_rate": 7.858006967441637e-06,
      "loss": 0.35935587882995607,
      "num_input_tokens_seen": 468453520,
      "step": 9100,
      "train_runtime": 56085.5271,
      "train_tokens_per_second": 8352.485
    },
    {
      "epoch": 0.613048747643415,
      "grad_norm": 0.7322997402988699,
      "learning_rate": 7.85583701538343e-06,
      "loss": 0.38502931594848633,
      "num_input_tokens_seen": 468731272,
      "step": 9105,
      "train_runtime": 56114.5246,
      "train_tokens_per_second": 8353.118
    },
    {
      "epoch": 0.6133854026393751,
      "grad_norm": 0.721520115112052,
      "learning_rate": 7.853666264699506e-06,
      "loss": 0.36195964813232423,
      "num_input_tokens_seen": 468964496,
      "step": 9110,
      "train_runtime": 56146.6566,
      "train_tokens_per_second": 8352.492
    },
    {
      "epoch": 0.6137220576353353,
      "grad_norm": 0.6108460778677096,
      "learning_rate": 7.851494715996914e-06,
      "loss": 0.40300955772399905,
      "num_input_tokens_seen": 469208344,
      "step": 9115,
      "train_runtime": 56176.765,
      "train_tokens_per_second": 8352.356
    },
    {
      "epoch": 0.6140587126312954,
      "grad_norm": 0.6774126847087498,
      "learning_rate": 7.849322369882915e-06,
      "loss": 0.3705450057983398,
      "num_input_tokens_seen": 469453344,
      "step": 9120,
      "train_runtime": 56204.9197,
      "train_tokens_per_second": 8352.531
    },
    {
      "epoch": 0.6143953676272556,
      "grad_norm": 0.5598914670981792,
      "learning_rate": 7.847149226965003e-06,
      "loss": 0.3466792106628418,
      "num_input_tokens_seen": 469702472,
      "step": 9125,
      "train_runtime": 56232.9568,
      "train_tokens_per_second": 8352.797
    },
    {
      "epoch": 0.6147320226232157,
      "grad_norm": 0.6743457833013923,
      "learning_rate": 7.844975287850886e-06,
      "loss": 0.3458111763000488,
      "num_input_tokens_seen": 469963648,
      "step": 9130,
      "train_runtime": 56262.3264,
      "train_tokens_per_second": 8353.079
    },
    {
      "epoch": 0.6150686776191758,
      "grad_norm": 0.6210234682387259,
      "learning_rate": 7.842800553148506e-06,
      "loss": 0.39066529273986816,
      "num_input_tokens_seen": 470215872,
      "step": 9135,
      "train_runtime": 56293.9719,
      "train_tokens_per_second": 8352.864
    },
    {
      "epoch": 0.615405332615136,
      "grad_norm": 0.684268372694223,
      "learning_rate": 7.840625023466014e-06,
      "loss": 0.3716911315917969,
      "num_input_tokens_seen": 470480440,
      "step": 9140,
      "train_runtime": 56325.6509,
      "train_tokens_per_second": 8352.863
    },
    {
      "epoch": 0.6157419876110961,
      "grad_norm": 0.6120057309822424,
      "learning_rate": 7.838448699411792e-06,
      "loss": 0.347078800201416,
      "num_input_tokens_seen": 470738264,
      "step": 9145,
      "train_runtime": 56356.3964,
      "train_tokens_per_second": 8352.881
    },
    {
      "epoch": 0.6160786426070562,
      "grad_norm": 0.881017777286897,
      "learning_rate": 7.836271581594442e-06,
      "loss": 0.40745067596435547,
      "num_input_tokens_seen": 471006064,
      "step": 9150,
      "train_runtime": 56387.9211,
      "train_tokens_per_second": 8352.96
    },
    {
      "epoch": 0.6164152976030164,
      "grad_norm": 0.7105626539898199,
      "learning_rate": 7.834093670622789e-06,
      "loss": 0.3916271924972534,
      "num_input_tokens_seen": 471258672,
      "step": 9155,
      "train_runtime": 56416.4581,
      "train_tokens_per_second": 8353.213
    },
    {
      "epoch": 0.6167519525989765,
      "grad_norm": 0.5998943599695964,
      "learning_rate": 7.831914967105878e-06,
      "loss": 0.3943720817565918,
      "num_input_tokens_seen": 471537112,
      "step": 9160,
      "train_runtime": 56449.0792,
      "train_tokens_per_second": 8353.318
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.6628151953313433,
      "learning_rate": 7.829735471652978e-06,
      "loss": 0.348478889465332,
      "num_input_tokens_seen": 471792512,
      "step": 9165,
      "train_runtime": 56481.4143,
      "train_tokens_per_second": 8353.058
    },
    {
      "epoch": 0.6174252625908968,
      "grad_norm": 0.733355858411925,
      "learning_rate": 7.827555184873575e-06,
      "loss": 0.380437970161438,
      "num_input_tokens_seen": 472061960,
      "step": 9170,
      "train_runtime": 56514.4209,
      "train_tokens_per_second": 8352.947
    },
    {
      "epoch": 0.617761917586857,
      "grad_norm": 0.5392636735405187,
      "learning_rate": 7.825374107377381e-06,
      "loss": 0.3488961935043335,
      "num_input_tokens_seen": 472304832,
      "step": 9175,
      "train_runtime": 56545.2033,
      "train_tokens_per_second": 8352.695
    },
    {
      "epoch": 0.6180985725828171,
      "grad_norm": 0.7535327741255947,
      "learning_rate": 7.823192239774327e-06,
      "loss": 0.3772963285446167,
      "num_input_tokens_seen": 472552176,
      "step": 9180,
      "train_runtime": 56575.1837,
      "train_tokens_per_second": 8352.641
    },
    {
      "epoch": 0.6184352275787772,
      "grad_norm": 0.6506186743168164,
      "learning_rate": 7.821009582674567e-06,
      "loss": 0.3875298261642456,
      "num_input_tokens_seen": 472805720,
      "step": 9185,
      "train_runtime": 56606.3478,
      "train_tokens_per_second": 8352.521
    },
    {
      "epoch": 0.6187718825747374,
      "grad_norm": 0.7104540063837557,
      "learning_rate": 7.818826136688473e-06,
      "loss": 0.4013193130493164,
      "num_input_tokens_seen": 473078552,
      "step": 9190,
      "train_runtime": 56641.3416,
      "train_tokens_per_second": 8352.178
    },
    {
      "epoch": 0.6191085375706975,
      "grad_norm": 0.6784447743403043,
      "learning_rate": 7.816641902426638e-06,
      "loss": 0.38733530044555664,
      "num_input_tokens_seen": 473355232,
      "step": 9195,
      "train_runtime": 56675.0738,
      "train_tokens_per_second": 8352.088
    },
    {
      "epoch": 0.6194451925666576,
      "grad_norm": 0.6515975975468496,
      "learning_rate": 7.814456880499877e-06,
      "loss": 0.3875675916671753,
      "num_input_tokens_seen": 473610416,
      "step": 9200,
      "train_runtime": 56707.8395,
      "train_tokens_per_second": 8351.763
    },
    {
      "epoch": 0.6197818475626178,
      "grad_norm": 0.6537753776555876,
      "learning_rate": 7.812271071519222e-06,
      "loss": 0.39838006496429446,
      "num_input_tokens_seen": 473886152,
      "step": 9205,
      "train_runtime": 56740.6711,
      "train_tokens_per_second": 8351.79
    },
    {
      "epoch": 0.6201185025585779,
      "grad_norm": 0.7022189338409397,
      "learning_rate": 7.810084476095932e-06,
      "loss": 0.34320945739746095,
      "num_input_tokens_seen": 474144152,
      "step": 9210,
      "train_runtime": 56768.8512,
      "train_tokens_per_second": 8352.189
    },
    {
      "epoch": 0.6204551575545381,
      "grad_norm": 0.6255038980641974,
      "learning_rate": 7.807897094841476e-06,
      "loss": 0.35811076164245603,
      "num_input_tokens_seen": 474392504,
      "step": 9215,
      "train_runtime": 56796.3749,
      "train_tokens_per_second": 8352.514
    },
    {
      "epoch": 0.6207918125504982,
      "grad_norm": 0.652849292271752,
      "learning_rate": 7.805708928367555e-06,
      "loss": 0.3622297286987305,
      "num_input_tokens_seen": 474638768,
      "step": 9220,
      "train_runtime": 56827.3721,
      "train_tokens_per_second": 8352.291
    },
    {
      "epoch": 0.6211284675464583,
      "grad_norm": 0.6596564100699845,
      "learning_rate": 7.803519977286075e-06,
      "loss": 0.3956700086593628,
      "num_input_tokens_seen": 474882592,
      "step": 9225,
      "train_runtime": 56858.246,
      "train_tokens_per_second": 8352.044
    },
    {
      "epoch": 0.6214651225424185,
      "grad_norm": 0.6401819242047786,
      "learning_rate": 7.801330242209176e-06,
      "loss": 0.36794538497924806,
      "num_input_tokens_seen": 475123296,
      "step": 9230,
      "train_runtime": 56887.0024,
      "train_tokens_per_second": 8352.054
    },
    {
      "epoch": 0.6218017775383786,
      "grad_norm": 0.6555564045815278,
      "learning_rate": 7.799139723749207e-06,
      "loss": 0.38188967704772947,
      "num_input_tokens_seen": 475373376,
      "step": 9235,
      "train_runtime": 56914.3596,
      "train_tokens_per_second": 8352.433
    },
    {
      "epoch": 0.6221384325343388,
      "grad_norm": 0.6665905966485616,
      "learning_rate": 7.79694842251874e-06,
      "loss": 0.38590593338012696,
      "num_input_tokens_seen": 475644200,
      "step": 9240,
      "train_runtime": 56945.8435,
      "train_tokens_per_second": 8352.571
    },
    {
      "epoch": 0.6224750875302989,
      "grad_norm": 0.6005090595720074,
      "learning_rate": 7.794756339130565e-06,
      "loss": 0.36689636707305906,
      "num_input_tokens_seen": 475895168,
      "step": 9245,
      "train_runtime": 56976.777,
      "train_tokens_per_second": 8352.441
    },
    {
      "epoch": 0.622811742526259,
      "grad_norm": 0.5992637073779827,
      "learning_rate": 7.792563474197692e-06,
      "loss": 0.3804807662963867,
      "num_input_tokens_seen": 476145656,
      "step": 9250,
      "train_runtime": 57005.4655,
      "train_tokens_per_second": 8352.632
    },
    {
      "epoch": 0.6231483975222192,
      "grad_norm": 0.5986088967111153,
      "learning_rate": 7.79036982833335e-06,
      "loss": 0.39028100967407225,
      "num_input_tokens_seen": 476434448,
      "step": 9255,
      "train_runtime": 57037.6385,
      "train_tokens_per_second": 8352.983
    },
    {
      "epoch": 0.6234850525181793,
      "grad_norm": 0.7639042863550422,
      "learning_rate": 7.78817540215098e-06,
      "loss": 0.3685218095779419,
      "num_input_tokens_seen": 476683448,
      "step": 9260,
      "train_runtime": 57067.0012,
      "train_tokens_per_second": 8353.049
    },
    {
      "epoch": 0.6238217075141395,
      "grad_norm": 0.5919511541837735,
      "learning_rate": 7.78598019626425e-06,
      "loss": 0.36835460662841796,
      "num_input_tokens_seen": 476976568,
      "step": 9265,
      "train_runtime": 57099.5328,
      "train_tokens_per_second": 8353.423
    },
    {
      "epoch": 0.6241583625100996,
      "grad_norm": 0.7350737080376272,
      "learning_rate": 7.783784211287044e-06,
      "loss": 0.38853912353515624,
      "num_input_tokens_seen": 477210784,
      "step": 9270,
      "train_runtime": 57125.6979,
      "train_tokens_per_second": 8353.697
    },
    {
      "epoch": 0.6244950175060597,
      "grad_norm": 0.7579032731059617,
      "learning_rate": 7.78158744783346e-06,
      "loss": 0.3929029703140259,
      "num_input_tokens_seen": 477476240,
      "step": 9275,
      "train_runtime": 57157.0551,
      "train_tokens_per_second": 8353.759
    },
    {
      "epoch": 0.6248316725020199,
      "grad_norm": 0.6357908336344827,
      "learning_rate": 7.779389906517813e-06,
      "loss": 0.3563102722167969,
      "num_input_tokens_seen": 477731936,
      "step": 9280,
      "train_runtime": 57186.8846,
      "train_tokens_per_second": 8353.872
    },
    {
      "epoch": 0.6251683274979801,
      "grad_norm": 0.7049245664132862,
      "learning_rate": 7.777191587954645e-06,
      "loss": 0.3624164342880249,
      "num_input_tokens_seen": 477977584,
      "step": 9285,
      "train_runtime": 57216.2332,
      "train_tokens_per_second": 8353.881
    },
    {
      "epoch": 0.6255049824939403,
      "grad_norm": 0.7947066541452582,
      "learning_rate": 7.774992492758704e-06,
      "loss": 0.3606182336807251,
      "num_input_tokens_seen": 478215424,
      "step": 9290,
      "train_runtime": 57249.7251,
      "train_tokens_per_second": 8353.148
    },
    {
      "epoch": 0.6258416374899004,
      "grad_norm": 0.5996493579891897,
      "learning_rate": 7.77279262154496e-06,
      "loss": 0.3781902313232422,
      "num_input_tokens_seen": 478485448,
      "step": 9295,
      "train_runtime": 57280.855,
      "train_tokens_per_second": 8353.322
    },
    {
      "epoch": 0.6261782924858605,
      "grad_norm": 0.6669329912627631,
      "learning_rate": 7.770591974928604e-06,
      "loss": 0.3921350002288818,
      "num_input_tokens_seen": 478767552,
      "step": 9300,
      "train_runtime": 57311.3229,
      "train_tokens_per_second": 8353.804
    },
    {
      "epoch": 0.6265149474818207,
      "grad_norm": 0.6229717695828636,
      "learning_rate": 7.768390553525034e-06,
      "loss": 0.36296546459198,
      "num_input_tokens_seen": 479037456,
      "step": 9305,
      "train_runtime": 57338.5292,
      "train_tokens_per_second": 8354.547
    },
    {
      "epoch": 0.6268516024777808,
      "grad_norm": 0.6096635756864313,
      "learning_rate": 7.766188357949875e-06,
      "loss": 0.3690591335296631,
      "num_input_tokens_seen": 479294160,
      "step": 9310,
      "train_runtime": 57374.3059,
      "train_tokens_per_second": 8353.812
    },
    {
      "epoch": 0.627188257473741,
      "grad_norm": 0.7282007544993873,
      "learning_rate": 7.763985388818963e-06,
      "loss": 0.362040638923645,
      "num_input_tokens_seen": 479550488,
      "step": 9315,
      "train_runtime": 57409.1013,
      "train_tokens_per_second": 8353.214
    },
    {
      "epoch": 0.6275249124697011,
      "grad_norm": 0.8940127668513822,
      "learning_rate": 7.76178164674835e-06,
      "loss": 0.32610206604003905,
      "num_input_tokens_seen": 479804392,
      "step": 9320,
      "train_runtime": 57436.9428,
      "train_tokens_per_second": 8353.585
    },
    {
      "epoch": 0.6278615674656612,
      "grad_norm": 0.6492979106005826,
      "learning_rate": 7.759577132354305e-06,
      "loss": 0.36661763191223146,
      "num_input_tokens_seen": 480080440,
      "step": 9325,
      "train_runtime": 57470.6639,
      "train_tokens_per_second": 8353.487
    },
    {
      "epoch": 0.6281982224616214,
      "grad_norm": 0.6748469836633618,
      "learning_rate": 7.757371846253315e-06,
      "loss": 0.3543208360671997,
      "num_input_tokens_seen": 480324664,
      "step": 9330,
      "train_runtime": 57501.8511,
      "train_tokens_per_second": 8353.204
    },
    {
      "epoch": 0.6285348774575815,
      "grad_norm": 0.6369567753693987,
      "learning_rate": 7.755165789062083e-06,
      "loss": 0.3824007511138916,
      "num_input_tokens_seen": 480587088,
      "step": 9335,
      "train_runtime": 57535.2327,
      "train_tokens_per_second": 8352.918
    },
    {
      "epoch": 0.6288715324535417,
      "grad_norm": 0.6367996136184724,
      "learning_rate": 7.752958961397522e-06,
      "loss": 0.37522172927856445,
      "num_input_tokens_seen": 480881088,
      "step": 9340,
      "train_runtime": 57568.8781,
      "train_tokens_per_second": 8353.143
    },
    {
      "epoch": 0.6292081874495018,
      "grad_norm": 0.6799580986095594,
      "learning_rate": 7.750751363876765e-06,
      "loss": 0.37868332862854004,
      "num_input_tokens_seen": 481112512,
      "step": 9345,
      "train_runtime": 57598.8614,
      "train_tokens_per_second": 8352.813
    },
    {
      "epoch": 0.6295448424454619,
      "grad_norm": 0.6505715712654361,
      "learning_rate": 7.748542997117162e-06,
      "loss": 0.38275895118713377,
      "num_input_tokens_seen": 481379256,
      "step": 9350,
      "train_runtime": 57632.9735,
      "train_tokens_per_second": 8352.497
    },
    {
      "epoch": 0.6298814974414221,
      "grad_norm": 0.6493009901973241,
      "learning_rate": 7.746333861736275e-06,
      "loss": 0.3616541862487793,
      "num_input_tokens_seen": 481630968,
      "step": 9355,
      "train_runtime": 57660.8659,
      "train_tokens_per_second": 8352.822
    },
    {
      "epoch": 0.6302181524373822,
      "grad_norm": 0.6268043320740317,
      "learning_rate": 7.74412395835188e-06,
      "loss": 0.36249632835388185,
      "num_input_tokens_seen": 481858752,
      "step": 9360,
      "train_runtime": 57690.9411,
      "train_tokens_per_second": 8352.416
    },
    {
      "epoch": 0.6305548074333424,
      "grad_norm": 0.6068224579252816,
      "learning_rate": 7.74191328758197e-06,
      "loss": 0.3719178199768066,
      "num_input_tokens_seen": 482097208,
      "step": 9365,
      "train_runtime": 57721.2053,
      "train_tokens_per_second": 8352.168
    },
    {
      "epoch": 0.6308914624293025,
      "grad_norm": 0.6665264001058948,
      "learning_rate": 7.739701850044752e-06,
      "loss": 0.40233311653137205,
      "num_input_tokens_seen": 482368440,
      "step": 9370,
      "train_runtime": 57756.076,
      "train_tokens_per_second": 8351.822
    },
    {
      "epoch": 0.6312281174252626,
      "grad_norm": 0.8070853847793725,
      "learning_rate": 7.73748964635865e-06,
      "loss": 0.37592430114746095,
      "num_input_tokens_seen": 482632168,
      "step": 9375,
      "train_runtime": 57789.2249,
      "train_tokens_per_second": 8351.594
    },
    {
      "epoch": 0.6315647724212228,
      "grad_norm": 0.5758521868451719,
      "learning_rate": 7.735276677142297e-06,
      "loss": 0.34364781379699705,
      "num_input_tokens_seen": 482872080,
      "step": 9380,
      "train_runtime": 57819.2845,
      "train_tokens_per_second": 8351.402
    },
    {
      "epoch": 0.6319014274171829,
      "grad_norm": 0.6691039067841061,
      "learning_rate": 7.733062943014542e-06,
      "loss": 0.39099917411804197,
      "num_input_tokens_seen": 483143008,
      "step": 9385,
      "train_runtime": 57851.4382,
      "train_tokens_per_second": 8351.443
    },
    {
      "epoch": 0.632238082413143,
      "grad_norm": 0.5999688673563748,
      "learning_rate": 7.730848444594452e-06,
      "loss": 0.35693955421447754,
      "num_input_tokens_seen": 483420552,
      "step": 9390,
      "train_runtime": 57883.6155,
      "train_tokens_per_second": 8351.596
    },
    {
      "epoch": 0.6325747374091032,
      "grad_norm": 0.5654884771291862,
      "learning_rate": 7.7286331825013e-06,
      "loss": 0.35752553939819337,
      "num_input_tokens_seen": 483683336,
      "step": 9395,
      "train_runtime": 57914.8034,
      "train_tokens_per_second": 8351.636
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.8010951935863622,
      "learning_rate": 7.72641715735458e-06,
      "loss": 0.3652601957321167,
      "num_input_tokens_seen": 483910536,
      "step": 9400,
      "train_runtime": 57945.198,
      "train_tokens_per_second": 8351.176
    },
    {
      "epoch": 0.6332480474010235,
      "grad_norm": 0.5937130692880013,
      "learning_rate": 7.724200369773998e-06,
      "loss": 0.3849045276641846,
      "num_input_tokens_seen": 484185984,
      "step": 9405,
      "train_runtime": 57974.9989,
      "train_tokens_per_second": 8351.634
    },
    {
      "epoch": 0.6335847023969836,
      "grad_norm": 0.5943231777958589,
      "learning_rate": 7.721982820379467e-06,
      "loss": 0.36729583740234373,
      "num_input_tokens_seen": 484461464,
      "step": 9410,
      "train_runtime": 58006.0341,
      "train_tokens_per_second": 8351.915
    },
    {
      "epoch": 0.6339213573929438,
      "grad_norm": 0.7051720392099555,
      "learning_rate": 7.719764509791123e-06,
      "loss": 0.37789714336395264,
      "num_input_tokens_seen": 484677672,
      "step": 9415,
      "train_runtime": 58031.7488,
      "train_tokens_per_second": 8351.94
    },
    {
      "epoch": 0.6342580123889039,
      "grad_norm": 0.5910113509441429,
      "learning_rate": 7.717545438629303e-06,
      "loss": 0.41567516326904297,
      "num_input_tokens_seen": 484970544,
      "step": 9420,
      "train_runtime": 58061.293,
      "train_tokens_per_second": 8352.734
    },
    {
      "epoch": 0.634594667384864,
      "grad_norm": 0.6504047921194402,
      "learning_rate": 7.71532560751457e-06,
      "loss": 0.3653862476348877,
      "num_input_tokens_seen": 485229216,
      "step": 9425,
      "train_runtime": 58093.4605,
      "train_tokens_per_second": 8352.562
    },
    {
      "epoch": 0.6349313223808242,
      "grad_norm": 0.7080012843311778,
      "learning_rate": 7.713105017067686e-06,
      "loss": 0.3677579402923584,
      "num_input_tokens_seen": 485516160,
      "step": 9430,
      "train_runtime": 58123.7499,
      "train_tokens_per_second": 8353.146
    },
    {
      "epoch": 0.6352679773767843,
      "grad_norm": 0.863069704878791,
      "learning_rate": 7.710883667909639e-06,
      "loss": 0.3738752603530884,
      "num_input_tokens_seen": 485775216,
      "step": 9435,
      "train_runtime": 58152.9629,
      "train_tokens_per_second": 8353.404
    },
    {
      "epoch": 0.6356046323727444,
      "grad_norm": 0.557727495971737,
      "learning_rate": 7.708661560661617e-06,
      "loss": 0.38840508460998535,
      "num_input_tokens_seen": 486044784,
      "step": 9440,
      "train_runtime": 58185.0504,
      "train_tokens_per_second": 8353.431
    },
    {
      "epoch": 0.6359412873687046,
      "grad_norm": 0.6589530298761598,
      "learning_rate": 7.706438695945026e-06,
      "loss": 0.38015787601470946,
      "num_input_tokens_seen": 486300496,
      "step": 9445,
      "train_runtime": 58215.7346,
      "train_tokens_per_second": 8353.42
    },
    {
      "epoch": 0.6362779423646647,
      "grad_norm": 0.6493316812303702,
      "learning_rate": 7.704215074381484e-06,
      "loss": 0.3960240364074707,
      "num_input_tokens_seen": 486550528,
      "step": 9450,
      "train_runtime": 58246.0313,
      "train_tokens_per_second": 8353.368
    },
    {
      "epoch": 0.6366145973606249,
      "grad_norm": 0.6410242375179785,
      "learning_rate": 7.70199069659282e-06,
      "loss": 0.37774684429168703,
      "num_input_tokens_seen": 486803048,
      "step": 9455,
      "train_runtime": 58275.1306,
      "train_tokens_per_second": 8353.53
    },
    {
      "epoch": 0.636951252356585,
      "grad_norm": 0.6722443037324238,
      "learning_rate": 7.699765563201075e-06,
      "loss": 0.3903862714767456,
      "num_input_tokens_seen": 487075112,
      "step": 9460,
      "train_runtime": 58304.7729,
      "train_tokens_per_second": 8353.949
    },
    {
      "epoch": 0.6372879073525451,
      "grad_norm": 0.5689344048802987,
      "learning_rate": 7.697539674828498e-06,
      "loss": 0.36056272983551024,
      "num_input_tokens_seen": 487355760,
      "step": 9465,
      "train_runtime": 58339.1259,
      "train_tokens_per_second": 8353.841
    },
    {
      "epoch": 0.6376245623485053,
      "grad_norm": 0.6326002348502635,
      "learning_rate": 7.695313032097554e-06,
      "loss": 0.3481855869293213,
      "num_input_tokens_seen": 487613944,
      "step": 9470,
      "train_runtime": 58371.5589,
      "train_tokens_per_second": 8353.622
    },
    {
      "epoch": 0.6379612173444654,
      "grad_norm": 0.6742337980513478,
      "learning_rate": 7.693085635630913e-06,
      "loss": 0.3629417657852173,
      "num_input_tokens_seen": 487844128,
      "step": 9475,
      "train_runtime": 58399.9184,
      "train_tokens_per_second": 8353.507
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 0.7072928523440035,
      "learning_rate": 7.690857486051463e-06,
      "loss": 0.36119534969329836,
      "num_input_tokens_seen": 488112152,
      "step": 9480,
      "train_runtime": 58431.1406,
      "train_tokens_per_second": 8353.63
    },
    {
      "epoch": 0.6386345273363857,
      "grad_norm": 0.6162134843415107,
      "learning_rate": 7.688628583982298e-06,
      "loss": 0.35998642444610596,
      "num_input_tokens_seen": 488356832,
      "step": 9485,
      "train_runtime": 58459.8729,
      "train_tokens_per_second": 8353.71
    },
    {
      "epoch": 0.6389711823323458,
      "grad_norm": 0.6079693772498871,
      "learning_rate": 7.68639893004672e-06,
      "loss": 0.34569473266601564,
      "num_input_tokens_seen": 488626352,
      "step": 9490,
      "train_runtime": 58492.6862,
      "train_tokens_per_second": 8353.632
    },
    {
      "epoch": 0.639307837328306,
      "grad_norm": 0.6356548920879567,
      "learning_rate": 7.684168524868253e-06,
      "loss": 0.34182009696960447,
      "num_input_tokens_seen": 488901048,
      "step": 9495,
      "train_runtime": 58521.9501,
      "train_tokens_per_second": 8354.148
    },
    {
      "epoch": 0.6396444923242661,
      "grad_norm": 0.56820436875648,
      "learning_rate": 7.681937369070612e-06,
      "loss": 0.3589000225067139,
      "num_input_tokens_seen": 489154920,
      "step": 9500,
      "train_runtime": 58554.6983,
      "train_tokens_per_second": 8353.812
    },
    {
      "epoch": 0.6399811473202263,
      "grad_norm": 0.6928842620819191,
      "learning_rate": 7.67970546327774e-06,
      "loss": 0.37607102394104003,
      "num_input_tokens_seen": 489405920,
      "step": 9505,
      "train_runtime": 58587.7215,
      "train_tokens_per_second": 8353.387
    },
    {
      "epoch": 0.6403178023161864,
      "grad_norm": 0.6429962432085401,
      "learning_rate": 7.67747280811378e-06,
      "loss": 0.3567336559295654,
      "num_input_tokens_seen": 489656280,
      "step": 9510,
      "train_runtime": 58622.225,
      "train_tokens_per_second": 8352.741
    },
    {
      "epoch": 0.6406544573121465,
      "grad_norm": 0.690195450609345,
      "learning_rate": 7.675239404203088e-06,
      "loss": 0.35400426387786865,
      "num_input_tokens_seen": 489900848,
      "step": 9515,
      "train_runtime": 58651.953,
      "train_tokens_per_second": 8352.677
    },
    {
      "epoch": 0.6409911123081067,
      "grad_norm": 0.648984337769998,
      "learning_rate": 7.673005252170226e-06,
      "loss": 0.3632839202880859,
      "num_input_tokens_seen": 490161624,
      "step": 9520,
      "train_runtime": 58679.6619,
      "train_tokens_per_second": 8353.177
    },
    {
      "epoch": 0.6413277673040668,
      "grad_norm": 0.7130447645254481,
      "learning_rate": 7.670770352639968e-06,
      "loss": 0.36068427562713623,
      "num_input_tokens_seen": 490385224,
      "step": 9525,
      "train_runtime": 58709.7297,
      "train_tokens_per_second": 8352.708
    },
    {
      "epoch": 0.641664422300027,
      "grad_norm": 0.6434549780592344,
      "learning_rate": 7.668534706237299e-06,
      "loss": 0.3696594715118408,
      "num_input_tokens_seen": 490657728,
      "step": 9530,
      "train_runtime": 58742.762,
      "train_tokens_per_second": 8352.65
    },
    {
      "epoch": 0.6420010772959871,
      "grad_norm": 0.6425722422318173,
      "learning_rate": 7.666298313587409e-06,
      "loss": 0.4100175380706787,
      "num_input_tokens_seen": 490916112,
      "step": 9535,
      "train_runtime": 58776.9836,
      "train_tokens_per_second": 8352.183
    },
    {
      "epoch": 0.6423377322919472,
      "grad_norm": 0.756301664879849,
      "learning_rate": 7.664061175315696e-06,
      "loss": 0.3730136871337891,
      "num_input_tokens_seen": 491139096,
      "step": 9540,
      "train_runtime": 58803.4128,
      "train_tokens_per_second": 8352.221
    },
    {
      "epoch": 0.6426743872879074,
      "grad_norm": 0.6746866590998646,
      "learning_rate": 7.661823292047771e-06,
      "loss": 0.37233595848083495,
      "num_input_tokens_seen": 491397976,
      "step": 9545,
      "train_runtime": 58836.1045,
      "train_tokens_per_second": 8351.98
    },
    {
      "epoch": 0.6430110422838675,
      "grad_norm": 0.6674643602474986,
      "learning_rate": 7.659584664409452e-06,
      "loss": 0.38108053207397463,
      "num_input_tokens_seen": 491663968,
      "step": 9550,
      "train_runtime": 58867.0845,
      "train_tokens_per_second": 8352.103
    },
    {
      "epoch": 0.6433476972798277,
      "grad_norm": 0.7678581664664346,
      "learning_rate": 7.65734529302676e-06,
      "loss": 0.3330428123474121,
      "num_input_tokens_seen": 491904336,
      "step": 9555,
      "train_runtime": 58894.9641,
      "train_tokens_per_second": 8352.231
    },
    {
      "epoch": 0.6436843522757878,
      "grad_norm": 0.6455310190565688,
      "learning_rate": 7.655105178525932e-06,
      "loss": 0.3573255777359009,
      "num_input_tokens_seen": 492162256,
      "step": 9560,
      "train_runtime": 58924.023,
      "train_tokens_per_second": 8352.489
    },
    {
      "epoch": 0.6440210072717479,
      "grad_norm": 0.6658938526973475,
      "learning_rate": 7.652864321533406e-06,
      "loss": 0.3666593790054321,
      "num_input_tokens_seen": 492404816,
      "step": 9565,
      "train_runtime": 58952.0612,
      "train_tokens_per_second": 8352.631
    },
    {
      "epoch": 0.6443576622677081,
      "grad_norm": 0.6182397992548424,
      "learning_rate": 7.650622722675833e-06,
      "loss": 0.3860648155212402,
      "num_input_tokens_seen": 492670552,
      "step": 9570,
      "train_runtime": 58989.3443,
      "train_tokens_per_second": 8351.857
    },
    {
      "epoch": 0.6446943172636682,
      "grad_norm": 0.6798550411791584,
      "learning_rate": 7.648380382580067e-06,
      "loss": 0.389545202255249,
      "num_input_tokens_seen": 492914688,
      "step": 9575,
      "train_runtime": 59017.217,
      "train_tokens_per_second": 8352.049
    },
    {
      "epoch": 0.6450309722596284,
      "grad_norm": 0.705029986697118,
      "learning_rate": 7.646137301873173e-06,
      "loss": 0.3787783145904541,
      "num_input_tokens_seen": 493157376,
      "step": 9580,
      "train_runtime": 59046.3661,
      "train_tokens_per_second": 8352.036
    },
    {
      "epoch": 0.6453676272555885,
      "grad_norm": 0.7856712133289206,
      "learning_rate": 7.643893481182421e-06,
      "loss": 0.3429187059402466,
      "num_input_tokens_seen": 493413664,
      "step": 9585,
      "train_runtime": 59076.6971,
      "train_tokens_per_second": 8352.086
    },
    {
      "epoch": 0.6457042822515486,
      "grad_norm": 0.6952234772354412,
      "learning_rate": 7.641648921135287e-06,
      "loss": 0.3933689594268799,
      "num_input_tokens_seen": 493683448,
      "step": 9590,
      "train_runtime": 59104.4817,
      "train_tokens_per_second": 8352.724
    },
    {
      "epoch": 0.6460409372475088,
      "grad_norm": 0.8459314960606993,
      "learning_rate": 7.639403622359455e-06,
      "loss": 0.36925559043884276,
      "num_input_tokens_seen": 493933360,
      "step": 9595,
      "train_runtime": 59134.3351,
      "train_tokens_per_second": 8352.734
    },
    {
      "epoch": 0.6463775922434689,
      "grad_norm": 0.6358321185759126,
      "learning_rate": 7.637157585482817e-06,
      "loss": 0.3600678682327271,
      "num_input_tokens_seen": 494183968,
      "step": 9600,
      "train_runtime": 59166.1022,
      "train_tokens_per_second": 8352.485
    },
    {
      "epoch": 0.646714247239429,
      "grad_norm": 0.6500857468071418,
      "learning_rate": 7.63491081113347e-06,
      "loss": 0.3893348217010498,
      "num_input_tokens_seen": 494448552,
      "step": 9605,
      "train_runtime": 59196.794,
      "train_tokens_per_second": 8352.624
    },
    {
      "epoch": 0.6470509022353892,
      "grad_norm": 0.725415663778094,
      "learning_rate": 7.632663299939718e-06,
      "loss": 0.39886410236358644,
      "num_input_tokens_seen": 494712800,
      "step": 9610,
      "train_runtime": 59224.62,
      "train_tokens_per_second": 8353.161
    },
    {
      "epoch": 0.6473875572313493,
      "grad_norm": 0.7325837802862458,
      "learning_rate": 7.630415052530066e-06,
      "loss": 0.36340017318725587,
      "num_input_tokens_seen": 494966528,
      "step": 9615,
      "train_runtime": 59253.8896,
      "train_tokens_per_second": 8353.317
    },
    {
      "epoch": 0.6477242122273095,
      "grad_norm": 0.7005357668134669,
      "learning_rate": 7.628166069533235e-06,
      "loss": 0.3714491844177246,
      "num_input_tokens_seen": 495204776,
      "step": 9620,
      "train_runtime": 59282.126,
      "train_tokens_per_second": 8353.357
    },
    {
      "epoch": 0.6480608672232696,
      "grad_norm": 0.7594384802200328,
      "learning_rate": 7.625916351578141e-06,
      "loss": 0.4033350944519043,
      "num_input_tokens_seen": 495467440,
      "step": 9625,
      "train_runtime": 59307.8494,
      "train_tokens_per_second": 8354.163
    },
    {
      "epoch": 0.6483975222192297,
      "grad_norm": 0.618730047668511,
      "learning_rate": 7.623665899293914e-06,
      "loss": 0.3543778657913208,
      "num_input_tokens_seen": 495721480,
      "step": 9630,
      "train_runtime": 59336.7722,
      "train_tokens_per_second": 8354.372
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.6693378659357849,
      "learning_rate": 7.621414713309885e-06,
      "loss": 0.34654998779296875,
      "num_input_tokens_seen": 495978976,
      "step": 9635,
      "train_runtime": 59364.9955,
      "train_tokens_per_second": 8354.738
    },
    {
      "epoch": 0.64907083221115,
      "grad_norm": 0.6368097348644669,
      "learning_rate": 7.61916279425559e-06,
      "loss": 0.3752583980560303,
      "num_input_tokens_seen": 496244552,
      "step": 9640,
      "train_runtime": 59397.5634,
      "train_tokens_per_second": 8354.628
    },
    {
      "epoch": 0.6494074872071102,
      "grad_norm": 0.6485356456516718,
      "learning_rate": 7.6169101427607716e-06,
      "loss": 0.3737299919128418,
      "num_input_tokens_seen": 496526328,
      "step": 9645,
      "train_runtime": 59433.6681,
      "train_tokens_per_second": 8354.294
    },
    {
      "epoch": 0.6497441422030703,
      "grad_norm": 0.8348524664368608,
      "learning_rate": 7.614656759455375e-06,
      "loss": 0.3717931270599365,
      "num_input_tokens_seen": 496780216,
      "step": 9650,
      "train_runtime": 59467.9882,
      "train_tokens_per_second": 8353.742
    },
    {
      "epoch": 0.6500807971990304,
      "grad_norm": 0.7879133426889509,
      "learning_rate": 7.612402644969555e-06,
      "loss": 0.3779712438583374,
      "num_input_tokens_seen": 497018280,
      "step": 9655,
      "train_runtime": 59498.1918,
      "train_tokens_per_second": 8353.502
    },
    {
      "epoch": 0.6504174521949906,
      "grad_norm": 0.7165736093490771,
      "learning_rate": 7.6101477999336646e-06,
      "loss": 0.3735002279281616,
      "num_input_tokens_seen": 497267088,
      "step": 9660,
      "train_runtime": 59526.5225,
      "train_tokens_per_second": 8353.706
    },
    {
      "epoch": 0.6507541071909507,
      "grad_norm": 0.8441931223229201,
      "learning_rate": 7.607892224978266e-06,
      "loss": 0.3758198022842407,
      "num_input_tokens_seen": 497511968,
      "step": 9665,
      "train_runtime": 59557.8221,
      "train_tokens_per_second": 8353.428
    },
    {
      "epoch": 0.6510907621869109,
      "grad_norm": 0.6419795123474877,
      "learning_rate": 7.605635920734122e-06,
      "loss": 0.3466759204864502,
      "num_input_tokens_seen": 497763648,
      "step": 9670,
      "train_runtime": 59588.3525,
      "train_tokens_per_second": 8353.372
    },
    {
      "epoch": 0.651427417182871,
      "grad_norm": 0.6507274323107922,
      "learning_rate": 7.603378887832202e-06,
      "loss": 0.3853426933288574,
      "num_input_tokens_seen": 498002840,
      "step": 9675,
      "train_runtime": 59617.4501,
      "train_tokens_per_second": 8353.307
    },
    {
      "epoch": 0.6517640721788311,
      "grad_norm": 0.6086151643917757,
      "learning_rate": 7.601121126903676e-06,
      "loss": 0.3639655113220215,
      "num_input_tokens_seen": 498261600,
      "step": 9680,
      "train_runtime": 59649.5392,
      "train_tokens_per_second": 8353.151
    },
    {
      "epoch": 0.6521007271747913,
      "grad_norm": 0.6394917625359494,
      "learning_rate": 7.598862638579922e-06,
      "loss": 0.3716209173202515,
      "num_input_tokens_seen": 498549824,
      "step": 9685,
      "train_runtime": 59677.9337,
      "train_tokens_per_second": 8354.006
    },
    {
      "epoch": 0.6524373821707514,
      "grad_norm": 0.6722237060575774,
      "learning_rate": 7.5966034234925205e-06,
      "loss": 0.3665966749191284,
      "num_input_tokens_seen": 498828600,
      "step": 9690,
      "train_runtime": 59709.0109,
      "train_tokens_per_second": 8354.327
    },
    {
      "epoch": 0.6527740371667116,
      "grad_norm": 0.6508366265874906,
      "learning_rate": 7.59434348227325e-06,
      "loss": 0.4182003021240234,
      "num_input_tokens_seen": 499104168,
      "step": 9695,
      "train_runtime": 59740.1378,
      "train_tokens_per_second": 8354.587
    },
    {
      "epoch": 0.6531106921626717,
      "grad_norm": 0.7225803601409393,
      "learning_rate": 7.592082815554098e-06,
      "loss": 0.39524097442626954,
      "num_input_tokens_seen": 499374576,
      "step": 9700,
      "train_runtime": 59774.6544,
      "train_tokens_per_second": 8354.286
    },
    {
      "epoch": 0.6534473471586318,
      "grad_norm": 0.829748429230273,
      "learning_rate": 7.58982142396725e-06,
      "loss": 0.41488080024719237,
      "num_input_tokens_seen": 499582672,
      "step": 9705,
      "train_runtime": 59802.0071,
      "train_tokens_per_second": 8353.945
    },
    {
      "epoch": 0.653784002154592,
      "grad_norm": 0.7777813749701188,
      "learning_rate": 7.5875593081451e-06,
      "loss": 0.3701316356658936,
      "num_input_tokens_seen": 499847568,
      "step": 9710,
      "train_runtime": 59833.0167,
      "train_tokens_per_second": 8354.043
    },
    {
      "epoch": 0.6541206571505521,
      "grad_norm": 0.6529171544466738,
      "learning_rate": 7.585296468720243e-06,
      "loss": 0.35884201526641846,
      "num_input_tokens_seen": 500109336,
      "step": 9715,
      "train_runtime": 59861.685,
      "train_tokens_per_second": 8354.415
    },
    {
      "epoch": 0.6544573121465123,
      "grad_norm": 0.6119762685029182,
      "learning_rate": 7.5830329063254705e-06,
      "loss": 0.3759385347366333,
      "num_input_tokens_seen": 500366424,
      "step": 9720,
      "train_runtime": 59889.0273,
      "train_tokens_per_second": 8354.893
    },
    {
      "epoch": 0.6547939671424724,
      "grad_norm": 0.5963892677148679,
      "learning_rate": 7.5807686215937835e-06,
      "loss": 0.3801990270614624,
      "num_input_tokens_seen": 500609016,
      "step": 9725,
      "train_runtime": 59921.6362,
      "train_tokens_per_second": 8354.395
    },
    {
      "epoch": 0.6551306221384325,
      "grad_norm": 0.6754674097904906,
      "learning_rate": 7.578503615158379e-06,
      "loss": 0.34029135704040525,
      "num_input_tokens_seen": 500848800,
      "step": 9730,
      "train_runtime": 59951.5008,
      "train_tokens_per_second": 8354.233
    },
    {
      "epoch": 0.6554672771343927,
      "grad_norm": 0.5861556199879157,
      "learning_rate": 7.576237887652662e-06,
      "loss": 0.38237526416778567,
      "num_input_tokens_seen": 501120232,
      "step": 9735,
      "train_runtime": 59981.9966,
      "train_tokens_per_second": 8354.511
    },
    {
      "epoch": 0.6558039321303528,
      "grad_norm": 0.6626141557975646,
      "learning_rate": 7.573971439710235e-06,
      "loss": 0.33830549716949465,
      "num_input_tokens_seen": 501368904,
      "step": 9740,
      "train_runtime": 60011.1543,
      "train_tokens_per_second": 8354.595
    },
    {
      "epoch": 0.656140587126313,
      "grad_norm": 1.1335602425450946,
      "learning_rate": 7.571704271964902e-06,
      "loss": 0.3575699806213379,
      "num_input_tokens_seen": 501623160,
      "step": 9745,
      "train_runtime": 60039.8479,
      "train_tokens_per_second": 8354.837
    },
    {
      "epoch": 0.6564772421222731,
      "grad_norm": 0.7025111889895794,
      "learning_rate": 7.56943638505067e-06,
      "loss": 0.35959420204162595,
      "num_input_tokens_seen": 501891840,
      "step": 9750,
      "train_runtime": 60074.0033,
      "train_tokens_per_second": 8354.56
    },
    {
      "epoch": 0.6568138971182332,
      "grad_norm": 0.6531499401347184,
      "learning_rate": 7.567167779601746e-06,
      "loss": 0.3765146255493164,
      "num_input_tokens_seen": 502162752,
      "step": 9755,
      "train_runtime": 60107.1595,
      "train_tokens_per_second": 8354.458
    },
    {
      "epoch": 0.6571505521141934,
      "grad_norm": 0.6774634711227019,
      "learning_rate": 7.56489845625254e-06,
      "loss": 0.3987301826477051,
      "num_input_tokens_seen": 502416584,
      "step": 9760,
      "train_runtime": 60136.2527,
      "train_tokens_per_second": 8354.637
    },
    {
      "epoch": 0.6574872071101535,
      "grad_norm": 0.7612228183730559,
      "learning_rate": 7.562628415637659e-06,
      "loss": 0.3636819124221802,
      "num_input_tokens_seen": 502659312,
      "step": 9765,
      "train_runtime": 60165.9702,
      "train_tokens_per_second": 8354.545
    },
    {
      "epoch": 0.6578238621061137,
      "grad_norm": 0.6504262269855076,
      "learning_rate": 7.560357658391913e-06,
      "loss": 0.4025406837463379,
      "num_input_tokens_seen": 502928832,
      "step": 9770,
      "train_runtime": 60195.0051,
      "train_tokens_per_second": 8354.993
    },
    {
      "epoch": 0.6581605171020738,
      "grad_norm": 0.6899713572726605,
      "learning_rate": 7.558086185150311e-06,
      "loss": 0.3768664836883545,
      "num_input_tokens_seen": 503174520,
      "step": 9775,
      "train_runtime": 60226.7958,
      "train_tokens_per_second": 8354.662
    },
    {
      "epoch": 0.6584971720980339,
      "grad_norm": 0.6588225212042513,
      "learning_rate": 7.555813996548065e-06,
      "loss": 0.3853621959686279,
      "num_input_tokens_seen": 503440936,
      "step": 9780,
      "train_runtime": 60254.9543,
      "train_tokens_per_second": 8355.179
    },
    {
      "epoch": 0.6588338270939941,
      "grad_norm": 0.7117516827530522,
      "learning_rate": 7.553541093220586e-06,
      "loss": 0.3825618028640747,
      "num_input_tokens_seen": 503689696,
      "step": 9785,
      "train_runtime": 60285.5488,
      "train_tokens_per_second": 8355.065
    },
    {
      "epoch": 0.6591704820899542,
      "grad_norm": 0.6094916755968732,
      "learning_rate": 7.551267475803481e-06,
      "loss": 0.3678004026412964,
      "num_input_tokens_seen": 503957856,
      "step": 9790,
      "train_runtime": 60318.4002,
      "train_tokens_per_second": 8354.961
    },
    {
      "epoch": 0.6595071370859144,
      "grad_norm": 0.6879604976359944,
      "learning_rate": 7.5489931449325635e-06,
      "loss": 0.3611264228820801,
      "num_input_tokens_seen": 504208160,
      "step": 9795,
      "train_runtime": 60347.6548,
      "train_tokens_per_second": 8355.058
    },
    {
      "epoch": 0.6598437920818745,
      "grad_norm": 0.6772981294653642,
      "learning_rate": 7.546718101243842e-06,
      "loss": 0.37874517440795896,
      "num_input_tokens_seen": 504458024,
      "step": 9800,
      "train_runtime": 60379.075,
      "train_tokens_per_second": 8354.848
    },
    {
      "epoch": 0.6601804470778346,
      "grad_norm": 0.7569227911459748,
      "learning_rate": 7.544442345373522e-06,
      "loss": 0.3710306644439697,
      "num_input_tokens_seen": 504699480,
      "step": 9805,
      "train_runtime": 60407.5615,
      "train_tokens_per_second": 8354.906
    },
    {
      "epoch": 0.6605171020737948,
      "grad_norm": 0.6846868820245947,
      "learning_rate": 7.542165877958014e-06,
      "loss": 0.3689296722412109,
      "num_input_tokens_seen": 504960816,
      "step": 9810,
      "train_runtime": 60435.7034,
      "train_tokens_per_second": 8355.339
    },
    {
      "epoch": 0.6608537570697549,
      "grad_norm": 0.7438837382271939,
      "learning_rate": 7.539888699633926e-06,
      "loss": 0.37978827953338623,
      "num_input_tokens_seen": 505227936,
      "step": 9815,
      "train_runtime": 60470.017,
      "train_tokens_per_second": 8355.016
    },
    {
      "epoch": 0.661190412065715,
      "grad_norm": 0.7532802318630736,
      "learning_rate": 7.5376108110380606e-06,
      "loss": 0.34192771911621095,
      "num_input_tokens_seen": 505493576,
      "step": 9820,
      "train_runtime": 60499.5488,
      "train_tokens_per_second": 8355.328
    },
    {
      "epoch": 0.6615270670616752,
      "grad_norm": 0.671161121509805,
      "learning_rate": 7.535332212807426e-06,
      "loss": 0.3597960710525513,
      "num_input_tokens_seen": 505729864,
      "step": 9825,
      "train_runtime": 60527.0817,
      "train_tokens_per_second": 8355.431
    },
    {
      "epoch": 0.6618637220576353,
      "grad_norm": 0.6596864900775731,
      "learning_rate": 7.53305290557922e-06,
      "loss": 0.3942714691162109,
      "num_input_tokens_seen": 506002320,
      "step": 9830,
      "train_runtime": 60556.1946,
      "train_tokens_per_second": 8355.913
    },
    {
      "epoch": 0.6622003770535955,
      "grad_norm": 0.6156217039112897,
      "learning_rate": 7.530772889990848e-06,
      "loss": 0.37280371189117434,
      "num_input_tokens_seen": 506288336,
      "step": 9835,
      "train_runtime": 60589.5683,
      "train_tokens_per_second": 8356.031
    },
    {
      "epoch": 0.6625370320495556,
      "grad_norm": 0.5193542545616894,
      "learning_rate": 7.528492166679907e-06,
      "loss": 0.3398905754089355,
      "num_input_tokens_seen": 506560728,
      "step": 9840,
      "train_runtime": 60618.5621,
      "train_tokens_per_second": 8356.528
    },
    {
      "epoch": 0.6628736870455157,
      "grad_norm": 0.5774089241904171,
      "learning_rate": 7.526210736284194e-06,
      "loss": 0.37888374328613283,
      "num_input_tokens_seen": 506798352,
      "step": 9845,
      "train_runtime": 60649.6588,
      "train_tokens_per_second": 8356.162
    },
    {
      "epoch": 0.6632103420414759,
      "grad_norm": 0.6580794593938282,
      "learning_rate": 7.523928599441706e-06,
      "loss": 0.3781915187835693,
      "num_input_tokens_seen": 507069000,
      "step": 9850,
      "train_runtime": 60677.8915,
      "train_tokens_per_second": 8356.734
    },
    {
      "epoch": 0.663546997037436,
      "grad_norm": 0.6705296344097114,
      "learning_rate": 7.5216457567906295e-06,
      "loss": 0.37583074569702146,
      "num_input_tokens_seen": 507342600,
      "step": 9855,
      "train_runtime": 60707.8046,
      "train_tokens_per_second": 8357.123
    },
    {
      "epoch": 0.6638836520333962,
      "grad_norm": 0.7088142382702919,
      "learning_rate": 7.519362208969362e-06,
      "loss": 0.35044431686401367,
      "num_input_tokens_seen": 507605400,
      "step": 9860,
      "train_runtime": 60739.4922,
      "train_tokens_per_second": 8357.09
    },
    {
      "epoch": 0.6642203070293563,
      "grad_norm": 0.6122593456467742,
      "learning_rate": 7.5170779566164855e-06,
      "loss": 0.3731574535369873,
      "num_input_tokens_seen": 507862520,
      "step": 9865,
      "train_runtime": 60769.0049,
      "train_tokens_per_second": 8357.262
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.672576584228978,
      "learning_rate": 7.514793000370783e-06,
      "loss": 0.37289838790893554,
      "num_input_tokens_seen": 508105712,
      "step": 9870,
      "train_runtime": 60796.963,
      "train_tokens_per_second": 8357.419
    },
    {
      "epoch": 0.6648936170212766,
      "grad_norm": 0.6994753619449952,
      "learning_rate": 7.512507340871239e-06,
      "loss": 0.3597423553466797,
      "num_input_tokens_seen": 508349800,
      "step": 9875,
      "train_runtime": 60829.3729,
      "train_tokens_per_second": 8356.979
    },
    {
      "epoch": 0.6652302720172367,
      "grad_norm": 1.2457896462319678,
      "learning_rate": 7.510220978757028e-06,
      "loss": 0.3826481819152832,
      "num_input_tokens_seen": 508604464,
      "step": 9880,
      "train_runtime": 60856.4425,
      "train_tokens_per_second": 8357.447
    },
    {
      "epoch": 0.6655669270131969,
      "grad_norm": 0.7772326997566493,
      "learning_rate": 7.507933914667524e-06,
      "loss": 0.42517242431640623,
      "num_input_tokens_seen": 508873512,
      "step": 9885,
      "train_runtime": 60883.463,
      "train_tokens_per_second": 8358.156
    },
    {
      "epoch": 0.665903582009157,
      "grad_norm": 0.6758271384816631,
      "learning_rate": 7.505646149242299e-06,
      "loss": 0.3624406814575195,
      "num_input_tokens_seen": 509135472,
      "step": 9890,
      "train_runtime": 60912.5447,
      "train_tokens_per_second": 8358.467
    },
    {
      "epoch": 0.6662402370051171,
      "grad_norm": 0.6492229970540753,
      "learning_rate": 7.503357683121119e-06,
      "loss": 0.33897809982299804,
      "num_input_tokens_seen": 509392688,
      "step": 9895,
      "train_runtime": 60943.3055,
      "train_tokens_per_second": 8358.468
    },
    {
      "epoch": 0.6665768920010773,
      "grad_norm": 0.6024714520429473,
      "learning_rate": 7.501068516943944e-06,
      "loss": 0.312999963760376,
      "num_input_tokens_seen": 509648104,
      "step": 9900,
      "train_runtime": 60972.291,
      "train_tokens_per_second": 8358.684
    },
    {
      "epoch": 0.6669135469970374,
      "grad_norm": 0.6377245789652327,
      "learning_rate": 7.498778651350934e-06,
      "loss": 0.38643977642059324,
      "num_input_tokens_seen": 509912976,
      "step": 9905,
      "train_runtime": 61003.844,
      "train_tokens_per_second": 8358.702
    },
    {
      "epoch": 0.6672502019929976,
      "grad_norm": 0.6310817483518336,
      "learning_rate": 7.496488086982443e-06,
      "loss": 0.3563528060913086,
      "num_input_tokens_seen": 510177080,
      "step": 9910,
      "train_runtime": 61035.7946,
      "train_tokens_per_second": 8358.654
    },
    {
      "epoch": 0.6675868569889577,
      "grad_norm": 0.7233486354140175,
      "learning_rate": 7.494196824479019e-06,
      "loss": 0.36625871658325193,
      "num_input_tokens_seen": 510426880,
      "step": 9915,
      "train_runtime": 61068.4322,
      "train_tokens_per_second": 8358.277
    },
    {
      "epoch": 0.6679235119849178,
      "grad_norm": 0.603190932760509,
      "learning_rate": 7.491904864481406e-06,
      "loss": 0.36791064739227297,
      "num_input_tokens_seen": 510682912,
      "step": 9920,
      "train_runtime": 61094.5706,
      "train_tokens_per_second": 8358.892
    },
    {
      "epoch": 0.668260166980878,
      "grad_norm": 0.5272956187464529,
      "learning_rate": 7.489612207630543e-06,
      "loss": 0.3290459632873535,
      "num_input_tokens_seen": 510974424,
      "step": 9925,
      "train_runtime": 61125.2549,
      "train_tokens_per_second": 8359.465
    },
    {
      "epoch": 0.6685968219768381,
      "grad_norm": 0.6649183389374722,
      "learning_rate": 7.487318854567566e-06,
      "loss": 0.36833934783935546,
      "num_input_tokens_seen": 511212288,
      "step": 9930,
      "train_runtime": 61156.6666,
      "train_tokens_per_second": 8359.061
    },
    {
      "epoch": 0.6689334769727983,
      "grad_norm": 0.6637125644494949,
      "learning_rate": 7.485024805933803e-06,
      "loss": 0.35806262493133545,
      "num_input_tokens_seen": 511449872,
      "step": 9935,
      "train_runtime": 61187.907,
      "train_tokens_per_second": 8358.676
    },
    {
      "epoch": 0.6692701319687584,
      "grad_norm": 0.5901823726175713,
      "learning_rate": 7.482730062370776e-06,
      "loss": 0.3465218782424927,
      "num_input_tokens_seen": 511708736,
      "step": 9940,
      "train_runtime": 61214.0633,
      "train_tokens_per_second": 8359.333
    },
    {
      "epoch": 0.6696067869647185,
      "grad_norm": 0.712245081493509,
      "learning_rate": 7.480434624520205e-06,
      "loss": 0.3538621425628662,
      "num_input_tokens_seen": 511997152,
      "step": 9945,
      "train_runtime": 61245.4886,
      "train_tokens_per_second": 8359.753
    },
    {
      "epoch": 0.6699434419606787,
      "grad_norm": 0.5881344430210503,
      "learning_rate": 7.478138493024001e-06,
      "loss": 0.3783281326293945,
      "num_input_tokens_seen": 512274680,
      "step": 9950,
      "train_runtime": 61278.2957,
      "train_tokens_per_second": 8359.806
    },
    {
      "epoch": 0.6702800969566388,
      "grad_norm": 0.5840143308456538,
      "learning_rate": 7.475841668524268e-06,
      "loss": 0.3452629089355469,
      "num_input_tokens_seen": 512520672,
      "step": 9955,
      "train_runtime": 61309.4313,
      "train_tokens_per_second": 8359.573
    },
    {
      "epoch": 0.670616751952599,
      "grad_norm": 0.697264697514654,
      "learning_rate": 7.473544151663309e-06,
      "loss": 0.39688963890075685,
      "num_input_tokens_seen": 512783648,
      "step": 9960,
      "train_runtime": 61343.5618,
      "train_tokens_per_second": 8359.209
    },
    {
      "epoch": 0.6709534069485591,
      "grad_norm": 0.6493241204282691,
      "learning_rate": 7.471245943083615e-06,
      "loss": 0.36169986724853515,
      "num_input_tokens_seen": 513049568,
      "step": 9965,
      "train_runtime": 61373.7197,
      "train_tokens_per_second": 8359.434
    },
    {
      "epoch": 0.6712900619445192,
      "grad_norm": 0.6560739783025263,
      "learning_rate": 7.468947043427873e-06,
      "loss": 0.3433393001556396,
      "num_input_tokens_seen": 513307792,
      "step": 9970,
      "train_runtime": 61408.2479,
      "train_tokens_per_second": 8358.939
    },
    {
      "epoch": 0.6716267169404794,
      "grad_norm": 0.7807689068804154,
      "learning_rate": 7.466647453338965e-06,
      "loss": 0.36836695671081543,
      "num_input_tokens_seen": 513545624,
      "step": 9975,
      "train_runtime": 61436.1571,
      "train_tokens_per_second": 8359.013
    },
    {
      "epoch": 0.6719633719364395,
      "grad_norm": 0.5644728522931524,
      "learning_rate": 7.464347173459958e-06,
      "loss": 0.36575515270233155,
      "num_input_tokens_seen": 513834384,
      "step": 9980,
      "train_runtime": 61468.1559,
      "train_tokens_per_second": 8359.359
    },
    {
      "epoch": 0.6723000269323997,
      "grad_norm": 0.6294687153464186,
      "learning_rate": 7.462046204434126e-06,
      "loss": 0.34125614166259766,
      "num_input_tokens_seen": 514126112,
      "step": 9985,
      "train_runtime": 61502.9837,
      "train_tokens_per_second": 8359.369
    },
    {
      "epoch": 0.6726366819283598,
      "grad_norm": 0.6875858629576062,
      "learning_rate": 7.459744546904922e-06,
      "loss": 0.3762150764465332,
      "num_input_tokens_seen": 514386176,
      "step": 9990,
      "train_runtime": 61532.195,
      "train_tokens_per_second": 8359.627
    },
    {
      "epoch": 0.6729733369243199,
      "grad_norm": 0.6662310186881174,
      "learning_rate": 7.457442201516001e-06,
      "loss": 0.34650206565856934,
      "num_input_tokens_seen": 514623072,
      "step": 9995,
      "train_runtime": 61563.3621,
      "train_tokens_per_second": 8359.242
    },
    {
      "epoch": 0.6733099919202801,
      "grad_norm": 0.6807304443149015,
      "learning_rate": 7.455139168911203e-06,
      "loss": 0.3724966526031494,
      "num_input_tokens_seen": 514857096,
      "step": 10000,
      "train_runtime": 61593.2092,
      "train_tokens_per_second": 8358.991
    },
    {
      "epoch": 0.6736466469162402,
      "grad_norm": 0.7261782810052465,
      "learning_rate": 7.452835449734564e-06,
      "loss": 0.3619387626647949,
      "num_input_tokens_seen": 515118032,
      "step": 10005,
      "train_runtime": 61627.4265,
      "train_tokens_per_second": 8358.584
    },
    {
      "epoch": 0.6739833019122003,
      "grad_norm": 0.7097859489271743,
      "learning_rate": 7.450531044630315e-06,
      "loss": 0.3362336874008179,
      "num_input_tokens_seen": 515376896,
      "step": 10010,
      "train_runtime": 61658.2813,
      "train_tokens_per_second": 8358.6
    },
    {
      "epoch": 0.6743199569081605,
      "grad_norm": 0.6604156542251722,
      "learning_rate": 7.448225954242874e-06,
      "loss": 0.37511167526245115,
      "num_input_tokens_seen": 515652000,
      "step": 10015,
      "train_runtime": 61689.1196,
      "train_tokens_per_second": 8358.881
    },
    {
      "epoch": 0.6746566119041206,
      "grad_norm": 0.7003834568518906,
      "learning_rate": 7.44592017921685e-06,
      "loss": 0.36841545104980467,
      "num_input_tokens_seen": 515919544,
      "step": 10020,
      "train_runtime": 61721.4312,
      "train_tokens_per_second": 8358.84
    },
    {
      "epoch": 0.6749932669000808,
      "grad_norm": 0.611508423385469,
      "learning_rate": 7.443613720197048e-06,
      "loss": 0.38396248817443845,
      "num_input_tokens_seen": 516173536,
      "step": 10025,
      "train_runtime": 61752.4159,
      "train_tokens_per_second": 8358.759
    },
    {
      "epoch": 0.6753299218960409,
      "grad_norm": 0.7534250212081179,
      "learning_rate": 7.44130657782846e-06,
      "loss": 0.35410850048065184,
      "num_input_tokens_seen": 516405528,
      "step": 10030,
      "train_runtime": 61784.1449,
      "train_tokens_per_second": 8358.221
    },
    {
      "epoch": 0.675666576892001,
      "grad_norm": 0.6483349057523536,
      "learning_rate": 7.438998752756274e-06,
      "loss": 0.3710841178894043,
      "num_input_tokens_seen": 516659688,
      "step": 10035,
      "train_runtime": 61815.7154,
      "train_tokens_per_second": 8358.064
    },
    {
      "epoch": 0.6760032318879612,
      "grad_norm": 0.6610601223387931,
      "learning_rate": 7.436690245625864e-06,
      "loss": 0.36260688304901123,
      "num_input_tokens_seen": 516911352,
      "step": 10040,
      "train_runtime": 61845.7417,
      "train_tokens_per_second": 8358.075
    },
    {
      "epoch": 0.6763398868839213,
      "grad_norm": 0.8300369098143285,
      "learning_rate": 7.434381057082797e-06,
      "loss": 0.3682398319244385,
      "num_input_tokens_seen": 517148656,
      "step": 10045,
      "train_runtime": 61880.387,
      "train_tokens_per_second": 8357.23
    },
    {
      "epoch": 0.6766765418798815,
      "grad_norm": 0.6598488287423642,
      "learning_rate": 7.4320711877728305e-06,
      "loss": 0.3936509847640991,
      "num_input_tokens_seen": 517411552,
      "step": 10050,
      "train_runtime": 61911.492,
      "train_tokens_per_second": 8357.278
    },
    {
      "epoch": 0.6770131968758416,
      "grad_norm": 0.6998912158966115,
      "learning_rate": 7.429760638341909e-06,
      "loss": 0.34402642250061033,
      "num_input_tokens_seen": 517640560,
      "step": 10055,
      "train_runtime": 61937.1116,
      "train_tokens_per_second": 8357.519
    },
    {
      "epoch": 0.6773498518718017,
      "grad_norm": 0.6360524792219343,
      "learning_rate": 7.427449409436176e-06,
      "loss": 0.41823463439941405,
      "num_input_tokens_seen": 517900200,
      "step": 10060,
      "train_runtime": 61971.5059,
      "train_tokens_per_second": 8357.07
    },
    {
      "epoch": 0.6776865068677619,
      "grad_norm": 0.6182243002806643,
      "learning_rate": 7.425137501701955e-06,
      "loss": 0.34025278091430666,
      "num_input_tokens_seen": 518178968,
      "step": 10065,
      "train_runtime": 61998.8272,
      "train_tokens_per_second": 8357.883
    },
    {
      "epoch": 0.678023161863722,
      "grad_norm": 0.7529089322649867,
      "learning_rate": 7.422824915785765e-06,
      "loss": 0.35983307361602784,
      "num_input_tokens_seen": 518440352,
      "step": 10070,
      "train_runtime": 62029.0925,
      "train_tokens_per_second": 8358.019
    },
    {
      "epoch": 0.6783598168596822,
      "grad_norm": 0.6886918036315997,
      "learning_rate": 7.420511652334313e-06,
      "loss": 0.38345751762390134,
      "num_input_tokens_seen": 518673128,
      "step": 10075,
      "train_runtime": 62056.1056,
      "train_tokens_per_second": 8358.132
    },
    {
      "epoch": 0.6786964718556423,
      "grad_norm": 0.5703884721276026,
      "learning_rate": 7.418197711994495e-06,
      "loss": 0.3478293180465698,
      "num_input_tokens_seen": 518919688,
      "step": 10080,
      "train_runtime": 62084.9201,
      "train_tokens_per_second": 8358.224
    },
    {
      "epoch": 0.6790331268516024,
      "grad_norm": 0.6987048970954628,
      "learning_rate": 7.4158830954133984e-06,
      "loss": 0.36269614696502683,
      "num_input_tokens_seen": 519195024,
      "step": 10085,
      "train_runtime": 62116.8078,
      "train_tokens_per_second": 8358.366
    },
    {
      "epoch": 0.6793697818475626,
      "grad_norm": 0.6287667841061234,
      "learning_rate": 7.413567803238298e-06,
      "loss": 0.3595279693603516,
      "num_input_tokens_seen": 519452384,
      "step": 10090,
      "train_runtime": 62146.8924,
      "train_tokens_per_second": 8358.461
    },
    {
      "epoch": 0.6797064368435227,
      "grad_norm": 0.5937134129061175,
      "learning_rate": 7.411251836116658e-06,
      "loss": 0.3638514518737793,
      "num_input_tokens_seen": 519709976,
      "step": 10095,
      "train_runtime": 62177.8236,
      "train_tokens_per_second": 8358.446
    },
    {
      "epoch": 0.6800430918394829,
      "grad_norm": 0.6203136609129742,
      "learning_rate": 7.408935194696132e-06,
      "loss": 0.3751094818115234,
      "num_input_tokens_seen": 519962880,
      "step": 10100,
      "train_runtime": 62206.153,
      "train_tokens_per_second": 8358.705
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.5822101323080232,
      "learning_rate": 7.406617879624558e-06,
      "loss": 0.37724153995513915,
      "num_input_tokens_seen": 520231792,
      "step": 10105,
      "train_runtime": 62235.7611,
      "train_tokens_per_second": 8359.049
    },
    {
      "epoch": 0.6807164018314031,
      "grad_norm": 0.6278629623199532,
      "learning_rate": 7.40429989154997e-06,
      "loss": 0.3510415077209473,
      "num_input_tokens_seen": 520474928,
      "step": 10110,
      "train_runtime": 62264.3035,
      "train_tokens_per_second": 8359.122
    },
    {
      "epoch": 0.6810530568273633,
      "grad_norm": 0.5927604778106832,
      "learning_rate": 7.4019812311205826e-06,
      "loss": 0.3851910591125488,
      "num_input_tokens_seen": 520750032,
      "step": 10115,
      "train_runtime": 62294.9839,
      "train_tokens_per_second": 8359.422
    },
    {
      "epoch": 0.6813897118233234,
      "grad_norm": 0.5931276903857637,
      "learning_rate": 7.399661898984805e-06,
      "loss": 0.38865909576416013,
      "num_input_tokens_seen": 521047872,
      "step": 10120,
      "train_runtime": 62327.3252,
      "train_tokens_per_second": 8359.863
    },
    {
      "epoch": 0.6817263668192836,
      "grad_norm": 0.7611681728177291,
      "learning_rate": 7.397341895791228e-06,
      "loss": 0.3676868677139282,
      "num_input_tokens_seen": 521288424,
      "step": 10125,
      "train_runtime": 62357.2937,
      "train_tokens_per_second": 8359.702
    },
    {
      "epoch": 0.6820630218152437,
      "grad_norm": 0.7327077615758436,
      "learning_rate": 7.395021222188634e-06,
      "loss": 0.3669306516647339,
      "num_input_tokens_seen": 521566248,
      "step": 10130,
      "train_runtime": 62385.3724,
      "train_tokens_per_second": 8360.393
    },
    {
      "epoch": 0.6823996768112038,
      "grad_norm": 0.8459828394523544,
      "learning_rate": 7.392699878825993e-06,
      "loss": 0.4055652141571045,
      "num_input_tokens_seen": 521799368,
      "step": 10135,
      "train_runtime": 62415.7675,
      "train_tokens_per_second": 8360.057
    },
    {
      "epoch": 0.682736331807164,
      "grad_norm": 0.6324502287913406,
      "learning_rate": 7.390377866352462e-06,
      "loss": 0.36202435493469237,
      "num_input_tokens_seen": 522043880,
      "step": 10140,
      "train_runtime": 62448.4197,
      "train_tokens_per_second": 8359.601
    },
    {
      "epoch": 0.6830729868031241,
      "grad_norm": 0.7208481658045764,
      "learning_rate": 7.388055185417382e-06,
      "loss": 0.3662777662277222,
      "num_input_tokens_seen": 522317344,
      "step": 10145,
      "train_runtime": 62476.1391,
      "train_tokens_per_second": 8360.269
    },
    {
      "epoch": 0.6834096417990843,
      "grad_norm": 0.6340926954672919,
      "learning_rate": 7.3857318366702844e-06,
      "loss": 0.3444392204284668,
      "num_input_tokens_seen": 522580880,
      "step": 10150,
      "train_runtime": 62506.2151,
      "train_tokens_per_second": 8360.463
    },
    {
      "epoch": 0.6837462967950444,
      "grad_norm": 0.6444872730261649,
      "learning_rate": 7.383407820760885e-06,
      "loss": 0.37565701007843016,
      "num_input_tokens_seen": 522848792,
      "step": 10155,
      "train_runtime": 62537.5569,
      "train_tokens_per_second": 8360.557
    },
    {
      "epoch": 0.6840829517910045,
      "grad_norm": 0.6732383801663383,
      "learning_rate": 7.381083138339089e-06,
      "loss": 0.38091886043548584,
      "num_input_tokens_seen": 523099480,
      "step": 10160,
      "train_runtime": 62568.3065,
      "train_tokens_per_second": 8360.455
    },
    {
      "epoch": 0.6844196067869647,
      "grad_norm": 0.6547004626328791,
      "learning_rate": 7.378757790054988e-06,
      "loss": 0.3959184169769287,
      "num_input_tokens_seen": 523376776,
      "step": 10165,
      "train_runtime": 62602.3161,
      "train_tokens_per_second": 8360.342
    },
    {
      "epoch": 0.6847562617829248,
      "grad_norm": 0.7551067923000124,
      "learning_rate": 7.376431776558855e-06,
      "loss": 0.38664751052856444,
      "num_input_tokens_seen": 523638216,
      "step": 10170,
      "train_runtime": 62630.3448,
      "train_tokens_per_second": 8360.775
    },
    {
      "epoch": 0.685092916778885,
      "grad_norm": 0.6519755120283638,
      "learning_rate": 7.3741050985011546e-06,
      "loss": 0.34725251197814944,
      "num_input_tokens_seen": 523905568,
      "step": 10175,
      "train_runtime": 62662.6154,
      "train_tokens_per_second": 8360.736
    },
    {
      "epoch": 0.6854295717748451,
      "grad_norm": 0.6477280632767095,
      "learning_rate": 7.371777756532531e-06,
      "loss": 0.339398717880249,
      "num_input_tokens_seen": 524179816,
      "step": 10180,
      "train_runtime": 62699.4493,
      "train_tokens_per_second": 8360.198
    },
    {
      "epoch": 0.6857662267708052,
      "grad_norm": 0.6913457340071535,
      "learning_rate": 7.369449751303823e-06,
      "loss": 0.3820047855377197,
      "num_input_tokens_seen": 524442016,
      "step": 10185,
      "train_runtime": 62732.2913,
      "train_tokens_per_second": 8360.001
    },
    {
      "epoch": 0.6861028817667654,
      "grad_norm": 0.5708322886272178,
      "learning_rate": 7.367121083466048e-06,
      "loss": 0.35157318115234376,
      "num_input_tokens_seen": 524709600,
      "step": 10190,
      "train_runtime": 62763.2969,
      "train_tokens_per_second": 8360.134
    },
    {
      "epoch": 0.6864395367627255,
      "grad_norm": 0.774783728883351,
      "learning_rate": 7.364791753670409e-06,
      "loss": 0.37432193756103516,
      "num_input_tokens_seen": 524986256,
      "step": 10195,
      "train_runtime": 62794.7384,
      "train_tokens_per_second": 8360.354
    },
    {
      "epoch": 0.6867761917586857,
      "grad_norm": 0.6340533281986117,
      "learning_rate": 7.362461762568298e-06,
      "loss": 0.3750950336456299,
      "num_input_tokens_seen": 525251640,
      "step": 10200,
      "train_runtime": 62828.7978,
      "train_tokens_per_second": 8360.046
    },
    {
      "epoch": 0.6871128467546458,
      "grad_norm": 0.5703524719439249,
      "learning_rate": 7.360131110811287e-06,
      "loss": 0.3432144641876221,
      "num_input_tokens_seen": 525514432,
      "step": 10205,
      "train_runtime": 62858.1526,
      "train_tokens_per_second": 8360.323
    },
    {
      "epoch": 0.6874495017506059,
      "grad_norm": 0.6089138225210531,
      "learning_rate": 7.357799799051138e-06,
      "loss": 0.3621131181716919,
      "num_input_tokens_seen": 525761368,
      "step": 10210,
      "train_runtime": 62886.6918,
      "train_tokens_per_second": 8360.455
    },
    {
      "epoch": 0.6877861567465661,
      "grad_norm": 0.7244108147748454,
      "learning_rate": 7.355467827939794e-06,
      "loss": 0.372883939743042,
      "num_input_tokens_seen": 526003312,
      "step": 10215,
      "train_runtime": 62916.0249,
      "train_tokens_per_second": 8360.403
    },
    {
      "epoch": 0.6881228117425262,
      "grad_norm": 0.712234073098645,
      "learning_rate": 7.353135198129382e-06,
      "loss": 0.38681726455688475,
      "num_input_tokens_seen": 526250936,
      "step": 10220,
      "train_runtime": 62946.8547,
      "train_tokens_per_second": 8360.242
    },
    {
      "epoch": 0.6884594667384863,
      "grad_norm": 0.6573638226205807,
      "learning_rate": 7.350801910272216e-06,
      "loss": 0.38265199661254884,
      "num_input_tokens_seen": 526504512,
      "step": 10225,
      "train_runtime": 62975.6735,
      "train_tokens_per_second": 8360.443
    },
    {
      "epoch": 0.6887961217344465,
      "grad_norm": 0.7003859302060406,
      "learning_rate": 7.348467965020791e-06,
      "loss": 0.3415750741958618,
      "num_input_tokens_seen": 526745448,
      "step": 10230,
      "train_runtime": 63009.0763,
      "train_tokens_per_second": 8359.834
    },
    {
      "epoch": 0.6891327767304066,
      "grad_norm": 0.7829026697278321,
      "learning_rate": 7.3461333630277885e-06,
      "loss": 0.3599990367889404,
      "num_input_tokens_seen": 526961440,
      "step": 10235,
      "train_runtime": 63037.8872,
      "train_tokens_per_second": 8359.44
    },
    {
      "epoch": 0.6894694317263668,
      "grad_norm": 0.6477850196718318,
      "learning_rate": 7.343798104946074e-06,
      "loss": 0.35834789276123047,
      "num_input_tokens_seen": 527238008,
      "step": 10240,
      "train_runtime": 63070.91,
      "train_tokens_per_second": 8359.448
    },
    {
      "epoch": 0.6898060867223269,
      "grad_norm": 0.6197646729186858,
      "learning_rate": 7.341462191428692e-06,
      "loss": 0.37279281616210935,
      "num_input_tokens_seen": 527492488,
      "step": 10245,
      "train_runtime": 63103.9824,
      "train_tokens_per_second": 8359.1
    },
    {
      "epoch": 0.690142741718287,
      "grad_norm": 0.7502932822164534,
      "learning_rate": 7.339125623128874e-06,
      "loss": 0.3844339847564697,
      "num_input_tokens_seen": 527743000,
      "step": 10250,
      "train_runtime": 63132.2465,
      "train_tokens_per_second": 8359.326
    },
    {
      "epoch": 0.6904793967142472,
      "grad_norm": 0.5844466970724519,
      "learning_rate": 7.336788400700035e-06,
      "loss": 0.36040849685668946,
      "num_input_tokens_seen": 527988312,
      "step": 10255,
      "train_runtime": 63161.0886,
      "train_tokens_per_second": 8359.392
    },
    {
      "epoch": 0.6908160517102074,
      "grad_norm": 0.6939459073788363,
      "learning_rate": 7.334450524795771e-06,
      "loss": 0.3987321376800537,
      "num_input_tokens_seen": 528256720,
      "step": 10260,
      "train_runtime": 63189.1811,
      "train_tokens_per_second": 8359.923
    },
    {
      "epoch": 0.6911527067061676,
      "grad_norm": 0.6676255657966393,
      "learning_rate": 7.332111996069863e-06,
      "loss": 0.37420053482055665,
      "num_input_tokens_seen": 528513368,
      "step": 10265,
      "train_runtime": 63221.7734,
      "train_tokens_per_second": 8359.673
    },
    {
      "epoch": 0.6914893617021277,
      "grad_norm": 0.774041628739047,
      "learning_rate": 7.32977281517627e-06,
      "loss": 0.39228382110595705,
      "num_input_tokens_seen": 528747312,
      "step": 10270,
      "train_runtime": 63249.8887,
      "train_tokens_per_second": 8359.656
    },
    {
      "epoch": 0.6918260166980879,
      "grad_norm": 0.673864004650902,
      "learning_rate": 7.3274329827691406e-06,
      "loss": 0.3804271697998047,
      "num_input_tokens_seen": 528990192,
      "step": 10275,
      "train_runtime": 63283.6718,
      "train_tokens_per_second": 8359.031
    },
    {
      "epoch": 0.692162671694048,
      "grad_norm": 0.5914179939303112,
      "learning_rate": 7.325092499502796e-06,
      "loss": 0.3587521553039551,
      "num_input_tokens_seen": 529266784,
      "step": 10280,
      "train_runtime": 63315.511,
      "train_tokens_per_second": 8359.196
    },
    {
      "epoch": 0.6924993266900081,
      "grad_norm": 0.6061401479924983,
      "learning_rate": 7.322751366031751e-06,
      "loss": 0.3610358715057373,
      "num_input_tokens_seen": 529530984,
      "step": 10285,
      "train_runtime": 63345.2094,
      "train_tokens_per_second": 8359.448
    },
    {
      "epoch": 0.6928359816859683,
      "grad_norm": 0.6927541117853206,
      "learning_rate": 7.320409583010691e-06,
      "loss": 0.34702515602111816,
      "num_input_tokens_seen": 529793712,
      "step": 10290,
      "train_runtime": 63376.9934,
      "train_tokens_per_second": 8359.401
    },
    {
      "epoch": 0.6931726366819284,
      "grad_norm": 0.6559962980896995,
      "learning_rate": 7.318067151094493e-06,
      "loss": 0.3980299472808838,
      "num_input_tokens_seen": 530059632,
      "step": 10295,
      "train_runtime": 63407.9851,
      "train_tokens_per_second": 8359.509
    },
    {
      "epoch": 0.6935092916778886,
      "grad_norm": 0.9211067328645717,
      "learning_rate": 7.3157240709382046e-06,
      "loss": 0.3738506078720093,
      "num_input_tokens_seen": 530346320,
      "step": 10300,
      "train_runtime": 63441.3281,
      "train_tokens_per_second": 8359.635
    },
    {
      "epoch": 0.6938459466738487,
      "grad_norm": 0.6199045513179957,
      "learning_rate": 7.313380343197066e-06,
      "loss": 0.3570572853088379,
      "num_input_tokens_seen": 530630672,
      "step": 10305,
      "train_runtime": 63473.8455,
      "train_tokens_per_second": 8359.832
    },
    {
      "epoch": 0.6941826016698088,
      "grad_norm": 0.7224100069507878,
      "learning_rate": 7.311035968526492e-06,
      "loss": 0.37464725971221924,
      "num_input_tokens_seen": 530882192,
      "step": 10310,
      "train_runtime": 63506.549,
      "train_tokens_per_second": 8359.487
    },
    {
      "epoch": 0.694519256665769,
      "grad_norm": 0.6188025953178931,
      "learning_rate": 7.308690947582078e-06,
      "loss": 0.36537339687347414,
      "num_input_tokens_seen": 531128184,
      "step": 10315,
      "train_runtime": 63541.0014,
      "train_tokens_per_second": 8358.826
    },
    {
      "epoch": 0.6948559116617291,
      "grad_norm": 0.6213593663359893,
      "learning_rate": 7.306345281019602e-06,
      "loss": 0.3712362766265869,
      "num_input_tokens_seen": 531380400,
      "step": 10320,
      "train_runtime": 63571.9862,
      "train_tokens_per_second": 8358.719
    },
    {
      "epoch": 0.6951925666576892,
      "grad_norm": 0.7133682634695573,
      "learning_rate": 7.303998969495023e-06,
      "loss": 0.4040266990661621,
      "num_input_tokens_seen": 531637408,
      "step": 10325,
      "train_runtime": 63603.9242,
      "train_tokens_per_second": 8358.563
    },
    {
      "epoch": 0.6955292216536494,
      "grad_norm": 0.5895221558813948,
      "learning_rate": 7.301652013664479e-06,
      "loss": 0.395336651802063,
      "num_input_tokens_seen": 531914936,
      "step": 10330,
      "train_runtime": 63633.4469,
      "train_tokens_per_second": 8359.046
    },
    {
      "epoch": 0.6958658766496095,
      "grad_norm": 0.6863080501969525,
      "learning_rate": 7.299304414184288e-06,
      "loss": 0.3964931726455688,
      "num_input_tokens_seen": 532167584,
      "step": 10335,
      "train_runtime": 63661.3643,
      "train_tokens_per_second": 8359.349
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.5889321876274938,
      "learning_rate": 7.296956171710951e-06,
      "loss": 0.37654886245727537,
      "num_input_tokens_seen": 532446512,
      "step": 10340,
      "train_runtime": 63695.9251,
      "train_tokens_per_second": 8359.193
    },
    {
      "epoch": 0.6965391866415298,
      "grad_norm": 0.7560587122611903,
      "learning_rate": 7.294607286901146e-06,
      "loss": 0.3865635395050049,
      "num_input_tokens_seen": 532712656,
      "step": 10345,
      "train_runtime": 63724.8047,
      "train_tokens_per_second": 8359.581
    },
    {
      "epoch": 0.69687584163749,
      "grad_norm": 0.6824400640619078,
      "learning_rate": 7.292257760411726e-06,
      "loss": 0.37608747482299804,
      "num_input_tokens_seen": 532969360,
      "step": 10350,
      "train_runtime": 63754.3481,
      "train_tokens_per_second": 8359.734
    },
    {
      "epoch": 0.6972124966334501,
      "grad_norm": 0.6703064312886625,
      "learning_rate": 7.289907592899737e-06,
      "loss": 0.37522106170654296,
      "num_input_tokens_seen": 533196456,
      "step": 10355,
      "train_runtime": 63782.9944,
      "train_tokens_per_second": 8359.539
    },
    {
      "epoch": 0.6975491516294102,
      "grad_norm": 0.6430905950559078,
      "learning_rate": 7.287556785022389e-06,
      "loss": 0.38999171257019044,
      "num_input_tokens_seen": 533457920,
      "step": 10360,
      "train_runtime": 63815.7938,
      "train_tokens_per_second": 8359.34
    },
    {
      "epoch": 0.6978858066253704,
      "grad_norm": 0.6936191829242747,
      "learning_rate": 7.285205337437082e-06,
      "loss": 0.367938232421875,
      "num_input_tokens_seen": 533704608,
      "step": 10365,
      "train_runtime": 63843.7076,
      "train_tokens_per_second": 8359.549
    },
    {
      "epoch": 0.6982224616213305,
      "grad_norm": 0.6998414906104722,
      "learning_rate": 7.2828532508013895e-06,
      "loss": 0.3975669384002686,
      "num_input_tokens_seen": 533973024,
      "step": 10370,
      "train_runtime": 63877.8758,
      "train_tokens_per_second": 8359.28
    },
    {
      "epoch": 0.6985591166172906,
      "grad_norm": 0.6696284278944991,
      "learning_rate": 7.2805005257730625e-06,
      "loss": 0.37301805019378664,
      "num_input_tokens_seen": 534235392,
      "step": 10375,
      "train_runtime": 63908.0638,
      "train_tokens_per_second": 8359.436
    },
    {
      "epoch": 0.6988957716132508,
      "grad_norm": 0.5973485168138651,
      "learning_rate": 7.278147163010037e-06,
      "loss": 0.3508817195892334,
      "num_input_tokens_seen": 534508152,
      "step": 10380,
      "train_runtime": 63933.6723,
      "train_tokens_per_second": 8360.354
    },
    {
      "epoch": 0.6992324266092109,
      "grad_norm": 0.704675141671274,
      "learning_rate": 7.275793163170423e-06,
      "loss": 0.3626185417175293,
      "num_input_tokens_seen": 534784000,
      "step": 10385,
      "train_runtime": 63961.9392,
      "train_tokens_per_second": 8360.972
    },
    {
      "epoch": 0.6995690816051711,
      "grad_norm": 0.6534007112266147,
      "learning_rate": 7.273438526912506e-06,
      "loss": 0.39207441806793214,
      "num_input_tokens_seen": 535026296,
      "step": 10390,
      "train_runtime": 63997.6177,
      "train_tokens_per_second": 8360.097
    },
    {
      "epoch": 0.6999057366011312,
      "grad_norm": 0.6171243722548265,
      "learning_rate": 7.271083254894757e-06,
      "loss": 0.3568553447723389,
      "num_input_tokens_seen": 535279872,
      "step": 10395,
      "train_runtime": 64025.2619,
      "train_tokens_per_second": 8360.448
    },
    {
      "epoch": 0.7002423915970913,
      "grad_norm": 0.6350476158294553,
      "learning_rate": 7.268727347775815e-06,
      "loss": 0.3706211090087891,
      "num_input_tokens_seen": 535548528,
      "step": 10400,
      "train_runtime": 64056.9802,
      "train_tokens_per_second": 8360.502
    },
    {
      "epoch": 0.7005790465930515,
      "grad_norm": 0.6275478618786464,
      "learning_rate": 7.266370806214506e-06,
      "loss": 0.32037744522094724,
      "num_input_tokens_seen": 535799344,
      "step": 10405,
      "train_runtime": 64086.9571,
      "train_tokens_per_second": 8360.505
    },
    {
      "epoch": 0.7009157015890116,
      "grad_norm": 0.6283483480718476,
      "learning_rate": 7.264013630869827e-06,
      "loss": 0.3732919692993164,
      "num_input_tokens_seen": 536042864,
      "step": 10410,
      "train_runtime": 64121.8156,
      "train_tokens_per_second": 8359.758
    },
    {
      "epoch": 0.7012523565849718,
      "grad_norm": 0.6588764602352277,
      "learning_rate": 7.261655822400957e-06,
      "loss": 0.39177331924438474,
      "num_input_tokens_seen": 536315032,
      "step": 10415,
      "train_runtime": 64152.3431,
      "train_tokens_per_second": 8360.023
    },
    {
      "epoch": 0.7015890115809319,
      "grad_norm": 0.6353347318173787,
      "learning_rate": 7.259297381467248e-06,
      "loss": 0.3607137203216553,
      "num_input_tokens_seen": 536570608,
      "step": 10420,
      "train_runtime": 64183.6309,
      "train_tokens_per_second": 8359.929
    },
    {
      "epoch": 0.701925666576892,
      "grad_norm": 0.6257798154629359,
      "learning_rate": 7.2569383087282285e-06,
      "loss": 0.34423327445983887,
      "num_input_tokens_seen": 536781480,
      "step": 10425,
      "train_runtime": 64215.0367,
      "train_tokens_per_second": 8359.124
    },
    {
      "epoch": 0.7022623215728522,
      "grad_norm": 0.5807819066093601,
      "learning_rate": 7.25457860484361e-06,
      "loss": 0.3928612947463989,
      "num_input_tokens_seen": 537039648,
      "step": 10430,
      "train_runtime": 64247.0154,
      "train_tokens_per_second": 8358.982
    },
    {
      "epoch": 0.7025989765688123,
      "grad_norm": 0.6520551416408563,
      "learning_rate": 7.252218270473274e-06,
      "loss": 0.37448644638061523,
      "num_input_tokens_seen": 537290576,
      "step": 10435,
      "train_runtime": 64276.3549,
      "train_tokens_per_second": 8359.07
    },
    {
      "epoch": 0.7029356315647725,
      "grad_norm": 0.9483267963785021,
      "learning_rate": 7.249857306277279e-06,
      "loss": 0.3515018939971924,
      "num_input_tokens_seen": 537549560,
      "step": 10440,
      "train_runtime": 64306.7784,
      "train_tokens_per_second": 8359.143
    },
    {
      "epoch": 0.7032722865607326,
      "grad_norm": 0.8437986018895731,
      "learning_rate": 7.247495712915864e-06,
      "loss": 0.35799880027770997,
      "num_input_tokens_seen": 537813320,
      "step": 10445,
      "train_runtime": 64336.6482,
      "train_tokens_per_second": 8359.362
    },
    {
      "epoch": 0.7036089415566927,
      "grad_norm": 0.6402586826746676,
      "learning_rate": 7.245133491049439e-06,
      "loss": 0.3657562255859375,
      "num_input_tokens_seen": 538068552,
      "step": 10450,
      "train_runtime": 64366.8297,
      "train_tokens_per_second": 8359.407
    },
    {
      "epoch": 0.7039455965526529,
      "grad_norm": 0.7519485616441237,
      "learning_rate": 7.242770641338592e-06,
      "loss": 0.4163292407989502,
      "num_input_tokens_seen": 538331736,
      "step": 10455,
      "train_runtime": 64396.535,
      "train_tokens_per_second": 8359.638
    },
    {
      "epoch": 0.704282251548613,
      "grad_norm": 0.6099969735935115,
      "learning_rate": 7.240407164444088e-06,
      "loss": 0.361607027053833,
      "num_input_tokens_seen": 538574904,
      "step": 10460,
      "train_runtime": 64428.1352,
      "train_tokens_per_second": 8359.312
    },
    {
      "epoch": 0.7046189065445732,
      "grad_norm": 0.7292323086589414,
      "learning_rate": 7.238043061026862e-06,
      "loss": 0.33149693012237547,
      "num_input_tokens_seen": 538803856,
      "step": 10465,
      "train_runtime": 64455.6057,
      "train_tokens_per_second": 8359.302
    },
    {
      "epoch": 0.7049555615405333,
      "grad_norm": 0.6290159796387504,
      "learning_rate": 7.235678331748034e-06,
      "loss": 0.34483966827392576,
      "num_input_tokens_seen": 539028600,
      "step": 10470,
      "train_runtime": 64483.3848,
      "train_tokens_per_second": 8359.186
    },
    {
      "epoch": 0.7052922165364934,
      "grad_norm": 0.698588421444116,
      "learning_rate": 7.233312977268887e-06,
      "loss": 0.3922948122024536,
      "num_input_tokens_seen": 539300064,
      "step": 10475,
      "train_runtime": 64516.6221,
      "train_tokens_per_second": 8359.087
    },
    {
      "epoch": 0.7056288715324536,
      "grad_norm": 0.6772738740973412,
      "learning_rate": 7.230946998250889e-06,
      "loss": 0.3537127017974854,
      "num_input_tokens_seen": 539536312,
      "step": 10480,
      "train_runtime": 64545.7585,
      "train_tokens_per_second": 8358.974
    },
    {
      "epoch": 0.7059655265284137,
      "grad_norm": 0.6716747783705238,
      "learning_rate": 7.228580395355676e-06,
      "loss": 0.33184945583343506,
      "num_input_tokens_seen": 539777608,
      "step": 10485,
      "train_runtime": 64576.8765,
      "train_tokens_per_second": 8358.683
    },
    {
      "epoch": 0.7063021815243739,
      "grad_norm": 0.6592318540788714,
      "learning_rate": 7.226213169245062e-06,
      "loss": 0.3718853950500488,
      "num_input_tokens_seen": 540031104,
      "step": 10490,
      "train_runtime": 64606.6013,
      "train_tokens_per_second": 8358.76
    },
    {
      "epoch": 0.706638836520334,
      "grad_norm": 0.6900796665164551,
      "learning_rate": 7.223845320581035e-06,
      "loss": 0.36636569499969485,
      "num_input_tokens_seen": 540273360,
      "step": 10495,
      "train_runtime": 64633.471,
      "train_tokens_per_second": 8359.034
    },
    {
      "epoch": 0.7069754915162941,
      "grad_norm": 0.675424076328075,
      "learning_rate": 7.221476850025754e-06,
      "loss": 0.3648430347442627,
      "num_input_tokens_seen": 540549984,
      "step": 10500,
      "train_runtime": 64665.3606,
      "train_tokens_per_second": 8359.189
    },
    {
      "epoch": 0.7073121465122543,
      "grad_norm": 0.710393538947191,
      "learning_rate": 7.219107758241557e-06,
      "loss": 0.3732144832611084,
      "num_input_tokens_seen": 540814376,
      "step": 10505,
      "train_runtime": 64695.8919,
      "train_tokens_per_second": 8359.331
    },
    {
      "epoch": 0.7076488015082144,
      "grad_norm": 0.6757095411626399,
      "learning_rate": 7.2167380458909505e-06,
      "loss": 0.38040938377380373,
      "num_input_tokens_seen": 541086744,
      "step": 10510,
      "train_runtime": 64724.7639,
      "train_tokens_per_second": 8359.81
    },
    {
      "epoch": 0.7079854565041745,
      "grad_norm": 0.6366780140900492,
      "learning_rate": 7.214367713636619e-06,
      "loss": 0.37742507457733154,
      "num_input_tokens_seen": 541352920,
      "step": 10515,
      "train_runtime": 64752.4175,
      "train_tokens_per_second": 8360.351
    },
    {
      "epoch": 0.7083221115001347,
      "grad_norm": 0.6017340769769842,
      "learning_rate": 7.211996762141418e-06,
      "loss": 0.366849684715271,
      "num_input_tokens_seen": 541616472,
      "step": 10520,
      "train_runtime": 64783.0607,
      "train_tokens_per_second": 8360.464
    },
    {
      "epoch": 0.7086587664960948,
      "grad_norm": 0.707078099132898,
      "learning_rate": 7.209625192068374e-06,
      "loss": 0.3828901290893555,
      "num_input_tokens_seen": 541882304,
      "step": 10525,
      "train_runtime": 64813.5966,
      "train_tokens_per_second": 8360.627
    },
    {
      "epoch": 0.708995421492055,
      "grad_norm": 0.681531468979634,
      "learning_rate": 7.207253004080692e-06,
      "loss": 0.379799485206604,
      "num_input_tokens_seen": 542157432,
      "step": 10530,
      "train_runtime": 64844.1564,
      "train_tokens_per_second": 8360.93
    },
    {
      "epoch": 0.7093320764880151,
      "grad_norm": 0.8604919608045126,
      "learning_rate": 7.204880198841746e-06,
      "loss": 0.33632080554962157,
      "num_input_tokens_seen": 542418112,
      "step": 10535,
      "train_runtime": 64876.4628,
      "train_tokens_per_second": 8360.784
    },
    {
      "epoch": 0.7096687314839752,
      "grad_norm": 0.7554325942539468,
      "learning_rate": 7.2025067770150835e-06,
      "loss": 0.38117547035217286,
      "num_input_tokens_seen": 542689416,
      "step": 10540,
      "train_runtime": 64907.1948,
      "train_tokens_per_second": 8361.006
    },
    {
      "epoch": 0.7100053864799354,
      "grad_norm": 0.6950078268895109,
      "learning_rate": 7.2001327392644235e-06,
      "loss": 0.35231990814208985,
      "num_input_tokens_seen": 542976040,
      "step": 10545,
      "train_runtime": 64941.9951,
      "train_tokens_per_second": 8360.939
    },
    {
      "epoch": 0.7103420414758955,
      "grad_norm": 0.6943783411449824,
      "learning_rate": 7.197758086253659e-06,
      "loss": 0.40018653869628906,
      "num_input_tokens_seen": 543233968,
      "step": 10550,
      "train_runtime": 64972.0463,
      "train_tokens_per_second": 8361.041
    },
    {
      "epoch": 0.7106786964718557,
      "grad_norm": 0.6531777498103042,
      "learning_rate": 7.195382818646854e-06,
      "loss": 0.3890966892242432,
      "num_input_tokens_seen": 543474464,
      "step": 10555,
      "train_runtime": 64999.7501,
      "train_tokens_per_second": 8361.178
    },
    {
      "epoch": 0.7110153514678158,
      "grad_norm": 0.7171904485669206,
      "learning_rate": 7.1930069371082445e-06,
      "loss": 0.3551745891571045,
      "num_input_tokens_seen": 543697200,
      "step": 10560,
      "train_runtime": 65026.9139,
      "train_tokens_per_second": 8361.11
    },
    {
      "epoch": 0.7113520064637759,
      "grad_norm": 0.5806416941966891,
      "learning_rate": 7.190630442302238e-06,
      "loss": 0.3742053508758545,
      "num_input_tokens_seen": 543962296,
      "step": 10565,
      "train_runtime": 65059.6188,
      "train_tokens_per_second": 8360.982
    },
    {
      "epoch": 0.7116886614597361,
      "grad_norm": 0.6147406582240583,
      "learning_rate": 7.188253334893414e-06,
      "loss": 0.392845892906189,
      "num_input_tokens_seen": 544227552,
      "step": 10570,
      "train_runtime": 65088.9149,
      "train_tokens_per_second": 8361.294
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.6936674554634547,
      "learning_rate": 7.18587561554652e-06,
      "loss": 0.3721402645111084,
      "num_input_tokens_seen": 544475864,
      "step": 10575,
      "train_runtime": 65117.2515,
      "train_tokens_per_second": 8361.469
    },
    {
      "epoch": 0.7123619714516564,
      "grad_norm": 0.5132646939974334,
      "learning_rate": 7.183497284926483e-06,
      "loss": 0.35319502353668214,
      "num_input_tokens_seen": 544754640,
      "step": 10580,
      "train_runtime": 65150.1839,
      "train_tokens_per_second": 8361.521
    },
    {
      "epoch": 0.7126986264476165,
      "grad_norm": 0.7037043543135655,
      "learning_rate": 7.181118343698392e-06,
      "loss": 0.35889883041381837,
      "num_input_tokens_seen": 544991960,
      "step": 10585,
      "train_runtime": 65183.2848,
      "train_tokens_per_second": 8360.916
    },
    {
      "epoch": 0.7130352814435766,
      "grad_norm": 0.6789725087012636,
      "learning_rate": 7.178738792527512e-06,
      "loss": 0.38425796031951903,
      "num_input_tokens_seen": 545246576,
      "step": 10590,
      "train_runtime": 65214.9009,
      "train_tokens_per_second": 8360.767
    },
    {
      "epoch": 0.7133719364395368,
      "grad_norm": 0.6795474228124995,
      "learning_rate": 7.176358632079277e-06,
      "loss": 0.3629317760467529,
      "num_input_tokens_seen": 545501792,
      "step": 10595,
      "train_runtime": 65242.444,
      "train_tokens_per_second": 8361.149
    },
    {
      "epoch": 0.7137085914354969,
      "grad_norm": 0.5806729038959728,
      "learning_rate": 7.173977863019288e-06,
      "loss": 0.3603970050811768,
      "num_input_tokens_seen": 545761472,
      "step": 10600,
      "train_runtime": 65276.4143,
      "train_tokens_per_second": 8360.776
    },
    {
      "epoch": 0.7140452464314571,
      "grad_norm": 0.6203348901897491,
      "learning_rate": 7.171596486013323e-06,
      "loss": 0.3318899393081665,
      "num_input_tokens_seen": 546045552,
      "step": 10605,
      "train_runtime": 65308.2507,
      "train_tokens_per_second": 8361.05
    },
    {
      "epoch": 0.7143819014274172,
      "grad_norm": 0.6239861462971974,
      "learning_rate": 7.169214501727326e-06,
      "loss": 0.365164589881897,
      "num_input_tokens_seen": 546283672,
      "step": 10610,
      "train_runtime": 65334.2743,
      "train_tokens_per_second": 8361.364
    },
    {
      "epoch": 0.7147185564233773,
      "grad_norm": 0.6421159372538758,
      "learning_rate": 7.16683191082741e-06,
      "loss": 0.3660881519317627,
      "num_input_tokens_seen": 546542376,
      "step": 10615,
      "train_runtime": 65365.3444,
      "train_tokens_per_second": 8361.348
    },
    {
      "epoch": 0.7150552114193375,
      "grad_norm": 1.0345077327794605,
      "learning_rate": 7.16444871397986e-06,
      "loss": 0.3708617687225342,
      "num_input_tokens_seen": 546774400,
      "step": 10620,
      "train_runtime": 65395.5523,
      "train_tokens_per_second": 8361.033
    },
    {
      "epoch": 0.7153918664152976,
      "grad_norm": 0.8207765439808106,
      "learning_rate": 7.162064911851129e-06,
      "loss": 0.3636161327362061,
      "num_input_tokens_seen": 547013592,
      "step": 10625,
      "train_runtime": 65427.9806,
      "train_tokens_per_second": 8360.545
    },
    {
      "epoch": 0.7157285214112578,
      "grad_norm": 0.6733854191186601,
      "learning_rate": 7.15968050510784e-06,
      "loss": 0.3960456371307373,
      "num_input_tokens_seen": 547258960,
      "step": 10630,
      "train_runtime": 65457.7735,
      "train_tokens_per_second": 8360.488
    },
    {
      "epoch": 0.7160651764072179,
      "grad_norm": 0.7484872875902258,
      "learning_rate": 7.1572954944167834e-06,
      "loss": 0.3557002544403076,
      "num_input_tokens_seen": 547508264,
      "step": 10635,
      "train_runtime": 65487.217,
      "train_tokens_per_second": 8360.536
    },
    {
      "epoch": 0.716401831403178,
      "grad_norm": 0.5769963780782476,
      "learning_rate": 7.154909880444922e-06,
      "loss": 0.3925412893295288,
      "num_input_tokens_seen": 547772320,
      "step": 10640,
      "train_runtime": 65520.3951,
      "train_tokens_per_second": 8360.333
    },
    {
      "epoch": 0.7167384863991382,
      "grad_norm": 0.6522106464806976,
      "learning_rate": 7.152523663859384e-06,
      "loss": 0.36931040287017824,
      "num_input_tokens_seen": 548016800,
      "step": 10645,
      "train_runtime": 65549.8359,
      "train_tokens_per_second": 8360.308
    },
    {
      "epoch": 0.7170751413950983,
      "grad_norm": 0.7347716392017497,
      "learning_rate": 7.150136845327466e-06,
      "loss": 0.35829548835754393,
      "num_input_tokens_seen": 548252360,
      "step": 10650,
      "train_runtime": 65581.4653,
      "train_tokens_per_second": 8359.867
    },
    {
      "epoch": 0.7174117963910585,
      "grad_norm": 0.6408475325917141,
      "learning_rate": 7.147749425516637e-06,
      "loss": 0.37630043029785154,
      "num_input_tokens_seen": 548524520,
      "step": 10655,
      "train_runtime": 65616.2077,
      "train_tokens_per_second": 8359.589
    },
    {
      "epoch": 0.7177484513870186,
      "grad_norm": 0.59477771143305,
      "learning_rate": 7.145361405094529e-06,
      "loss": 0.3549040794372559,
      "num_input_tokens_seen": 548802984,
      "step": 10660,
      "train_runtime": 65651.3718,
      "train_tokens_per_second": 8359.353
    },
    {
      "epoch": 0.7180851063829787,
      "grad_norm": 0.6091669945806664,
      "learning_rate": 7.142972784728945e-06,
      "loss": 0.3378560543060303,
      "num_input_tokens_seen": 549043080,
      "step": 10665,
      "train_runtime": 65680.0346,
      "train_tokens_per_second": 8359.36
    },
    {
      "epoch": 0.7184217613789389,
      "grad_norm": 0.7259187890550948,
      "learning_rate": 7.140583565087856e-06,
      "loss": 0.3466786861419678,
      "num_input_tokens_seen": 549293360,
      "step": 10670,
      "train_runtime": 65714.5454,
      "train_tokens_per_second": 8358.779
    },
    {
      "epoch": 0.718758416374899,
      "grad_norm": 0.6466144939412135,
      "learning_rate": 7.138193746839399e-06,
      "loss": 0.3845326900482178,
      "num_input_tokens_seen": 549575256,
      "step": 10675,
      "train_runtime": 65747.1228,
      "train_tokens_per_second": 8358.925
    },
    {
      "epoch": 0.7190950713708592,
      "grad_norm": 0.6452547733344938,
      "learning_rate": 7.135803330651878e-06,
      "loss": 0.3682973384857178,
      "num_input_tokens_seen": 549857512,
      "step": 10680,
      "train_runtime": 65777.7172,
      "train_tokens_per_second": 8359.328
    },
    {
      "epoch": 0.7194317263668193,
      "grad_norm": 0.5933723833205323,
      "learning_rate": 7.133412317193766e-06,
      "loss": 0.3521927833557129,
      "num_input_tokens_seen": 550120344,
      "step": 10685,
      "train_runtime": 65806.5534,
      "train_tokens_per_second": 8359.659
    },
    {
      "epoch": 0.7197683813627794,
      "grad_norm": 0.6762457408358471,
      "learning_rate": 7.131020707133703e-06,
      "loss": 0.3691756248474121,
      "num_input_tokens_seen": 550363744,
      "step": 10690,
      "train_runtime": 65836.3258,
      "train_tokens_per_second": 8359.576
    },
    {
      "epoch": 0.7201050363587396,
      "grad_norm": 0.6832939003510723,
      "learning_rate": 7.128628501140495e-06,
      "loss": 0.3945106029510498,
      "num_input_tokens_seen": 550605560,
      "step": 10695,
      "train_runtime": 65867.6449,
      "train_tokens_per_second": 8359.272
    },
    {
      "epoch": 0.7204416913546997,
      "grad_norm": 0.6181329512018822,
      "learning_rate": 7.126235699883115e-06,
      "loss": 0.36069045066833494,
      "num_input_tokens_seen": 550822968,
      "step": 10700,
      "train_runtime": 65896.7719,
      "train_tokens_per_second": 8358.876
    },
    {
      "epoch": 0.7207783463506598,
      "grad_norm": 0.5785770650128453,
      "learning_rate": 7.1238423040306995e-06,
      "loss": 0.3613856315612793,
      "num_input_tokens_seen": 551074864,
      "step": 10705,
      "train_runtime": 65927.6264,
      "train_tokens_per_second": 8358.785
    },
    {
      "epoch": 0.72111500134662,
      "grad_norm": 0.56575080807301,
      "learning_rate": 7.121448314252557e-06,
      "loss": 0.33440916538238524,
      "num_input_tokens_seen": 551358872,
      "step": 10710,
      "train_runtime": 65956.7385,
      "train_tokens_per_second": 8359.402
    },
    {
      "epoch": 0.7214516563425801,
      "grad_norm": 0.6797106250922746,
      "learning_rate": 7.1190537312181586e-06,
      "loss": 0.35784177780151366,
      "num_input_tokens_seen": 551597776,
      "step": 10715,
      "train_runtime": 65988.8777,
      "train_tokens_per_second": 8358.951
    },
    {
      "epoch": 0.7217883113385403,
      "grad_norm": 0.70322779994017,
      "learning_rate": 7.116658555597141e-06,
      "loss": 0.3487898349761963,
      "num_input_tokens_seen": 551809904,
      "step": 10720,
      "train_runtime": 66019.1845,
      "train_tokens_per_second": 8358.327
    },
    {
      "epoch": 0.7221249663345004,
      "grad_norm": 0.6779714010332858,
      "learning_rate": 7.114262788059308e-06,
      "loss": 0.35342695713043215,
      "num_input_tokens_seen": 552067688,
      "step": 10725,
      "train_runtime": 66055.0,
      "train_tokens_per_second": 8357.697
    },
    {
      "epoch": 0.7224616213304605,
      "grad_norm": 0.6732200230854545,
      "learning_rate": 7.111866429274627e-06,
      "loss": 0.371171760559082,
      "num_input_tokens_seen": 552327184,
      "step": 10730,
      "train_runtime": 66084.9261,
      "train_tokens_per_second": 8357.839
    },
    {
      "epoch": 0.7227982763264207,
      "grad_norm": 0.6729897521593498,
      "learning_rate": 7.1094694799132356e-06,
      "loss": 0.368449330329895,
      "num_input_tokens_seen": 552591184,
      "step": 10735,
      "train_runtime": 66115.983,
      "train_tokens_per_second": 8357.906
    },
    {
      "epoch": 0.7231349313223808,
      "grad_norm": 0.6073606290202231,
      "learning_rate": 7.10707194064543e-06,
      "loss": 0.3535797119140625,
      "num_input_tokens_seen": 552834664,
      "step": 10740,
      "train_runtime": 66144.8633,
      "train_tokens_per_second": 8357.938
    },
    {
      "epoch": 0.723471586318341,
      "grad_norm": 0.6952961641634569,
      "learning_rate": 7.104673812141676e-06,
      "loss": 0.4039806365966797,
      "num_input_tokens_seen": 553103048,
      "step": 10745,
      "train_runtime": 66175.3971,
      "train_tokens_per_second": 8358.137
    },
    {
      "epoch": 0.7238082413143011,
      "grad_norm": 0.6843606089781188,
      "learning_rate": 7.102275095072601e-06,
      "loss": 0.3898960590362549,
      "num_input_tokens_seen": 553367640,
      "step": 10750,
      "train_runtime": 66208.0019,
      "train_tokens_per_second": 8358.018
    },
    {
      "epoch": 0.7241448963102612,
      "grad_norm": 0.676480142384188,
      "learning_rate": 7.099875790108998e-06,
      "loss": 0.36470861434936525,
      "num_input_tokens_seen": 553628960,
      "step": 10755,
      "train_runtime": 66236.0457,
      "train_tokens_per_second": 8358.424
    },
    {
      "epoch": 0.7244815513062214,
      "grad_norm": 0.61964871671152,
      "learning_rate": 7.097475897921827e-06,
      "loss": 0.3644301414489746,
      "num_input_tokens_seen": 553871696,
      "step": 10760,
      "train_runtime": 66261.0055,
      "train_tokens_per_second": 8358.939
    },
    {
      "epoch": 0.7248182063021815,
      "grad_norm": 0.7010979762110501,
      "learning_rate": 7.09507541918221e-06,
      "loss": 0.39059603214263916,
      "num_input_tokens_seen": 554121800,
      "step": 10765,
      "train_runtime": 66290.0779,
      "train_tokens_per_second": 8359.046
    },
    {
      "epoch": 0.7251548612981417,
      "grad_norm": 0.5510169303592481,
      "learning_rate": 7.092674354561433e-06,
      "loss": 0.3730147838592529,
      "num_input_tokens_seen": 554390096,
      "step": 10770,
      "train_runtime": 66319.4234,
      "train_tokens_per_second": 8359.393
    },
    {
      "epoch": 0.7254915162941018,
      "grad_norm": 0.6523434654842081,
      "learning_rate": 7.090272704730945e-06,
      "loss": 0.3369171619415283,
      "num_input_tokens_seen": 554624632,
      "step": 10775,
      "train_runtime": 66344.9061,
      "train_tokens_per_second": 8359.717
    },
    {
      "epoch": 0.7258281712900619,
      "grad_norm": 0.6188929747166261,
      "learning_rate": 7.08787047036236e-06,
      "loss": 0.34374656677246096,
      "num_input_tokens_seen": 554913312,
      "step": 10780,
      "train_runtime": 66376.6977,
      "train_tokens_per_second": 8360.062
    },
    {
      "epoch": 0.7261648262860221,
      "grad_norm": 0.7176897919761767,
      "learning_rate": 7.085467652127457e-06,
      "loss": 0.3726343631744385,
      "num_input_tokens_seen": 555184264,
      "step": 10785,
      "train_runtime": 66409.322,
      "train_tokens_per_second": 8360.035
    },
    {
      "epoch": 0.7265014812819822,
      "grad_norm": 0.5686912763222023,
      "learning_rate": 7.083064250698175e-06,
      "loss": 0.34858338832855223,
      "num_input_tokens_seen": 555459240,
      "step": 10790,
      "train_runtime": 66437.9226,
      "train_tokens_per_second": 8360.575
    },
    {
      "epoch": 0.7268381362779424,
      "grad_norm": 0.621809522759438,
      "learning_rate": 7.0806602667466175e-06,
      "loss": 0.36077589988708497,
      "num_input_tokens_seen": 555722344,
      "step": 10795,
      "train_runtime": 66469.7029,
      "train_tokens_per_second": 8360.536
    },
    {
      "epoch": 0.7271747912739025,
      "grad_norm": 0.5826057407257402,
      "learning_rate": 7.078255700945051e-06,
      "loss": 0.361968469619751,
      "num_input_tokens_seen": 555971200,
      "step": 10800,
      "train_runtime": 66499.6771,
      "train_tokens_per_second": 8360.51
    },
    {
      "epoch": 0.7275114462698626,
      "grad_norm": 0.6816806678210849,
      "learning_rate": 7.075850553965904e-06,
      "loss": 0.36708884239196776,
      "num_input_tokens_seen": 556208712,
      "step": 10805,
      "train_runtime": 66531.5391,
      "train_tokens_per_second": 8360.076
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.6468685805553185,
      "learning_rate": 7.07344482648177e-06,
      "loss": 0.36844027042388916,
      "num_input_tokens_seen": 556486952,
      "step": 10810,
      "train_runtime": 66562.7063,
      "train_tokens_per_second": 8360.341
    },
    {
      "epoch": 0.7281847562617829,
      "grad_norm": 0.8454964515399893,
      "learning_rate": 7.071038519165402e-06,
      "loss": 0.3865149736404419,
      "num_input_tokens_seen": 556734048,
      "step": 10815,
      "train_runtime": 66594.9766,
      "train_tokens_per_second": 8360.001
    },
    {
      "epoch": 0.7285214112577431,
      "grad_norm": 0.6592300878485745,
      "learning_rate": 7.068631632689716e-06,
      "loss": 0.38084075450897215,
      "num_input_tokens_seen": 557001968,
      "step": 10820,
      "train_runtime": 66627.053,
      "train_tokens_per_second": 8359.997
    },
    {
      "epoch": 0.7288580662537032,
      "grad_norm": 0.7258671001758901,
      "learning_rate": 7.066224167727791e-06,
      "loss": 0.37113070487976074,
      "num_input_tokens_seen": 557260192,
      "step": 10825,
      "train_runtime": 66658.836,
      "train_tokens_per_second": 8359.885
    },
    {
      "epoch": 0.7291947212496633,
      "grad_norm": 0.717199189270288,
      "learning_rate": 7.063816124952865e-06,
      "loss": 0.3641695022583008,
      "num_input_tokens_seen": 557512968,
      "step": 10830,
      "train_runtime": 66690.4643,
      "train_tokens_per_second": 8359.71
    },
    {
      "epoch": 0.7295313762456235,
      "grad_norm": 0.5308254205210227,
      "learning_rate": 7.061407505038341e-06,
      "loss": 0.35202085971832275,
      "num_input_tokens_seen": 557769928,
      "step": 10835,
      "train_runtime": 66721.1967,
      "train_tokens_per_second": 8359.711
    },
    {
      "epoch": 0.7298680312415836,
      "grad_norm": 0.5645644873512006,
      "learning_rate": 7.058998308657782e-06,
      "loss": 0.3844161033630371,
      "num_input_tokens_seen": 558051560,
      "step": 10840,
      "train_runtime": 66753.9916,
      "train_tokens_per_second": 8359.823
    },
    {
      "epoch": 0.7302046862375438,
      "grad_norm": 0.5644620965654115,
      "learning_rate": 7.056588536484912e-06,
      "loss": 0.34688332080841067,
      "num_input_tokens_seen": 558303208,
      "step": 10845,
      "train_runtime": 66786.1766,
      "train_tokens_per_second": 8359.562
    },
    {
      "epoch": 0.7305413412335039,
      "grad_norm": 0.6258307959856234,
      "learning_rate": 7.0541781891936145e-06,
      "loss": 0.34302966594696044,
      "num_input_tokens_seen": 558562584,
      "step": 10850,
      "train_runtime": 66819.4669,
      "train_tokens_per_second": 8359.279
    },
    {
      "epoch": 0.730877996229464,
      "grad_norm": 0.6563929908504171,
      "learning_rate": 7.0517672674579364e-06,
      "loss": 0.3676895618438721,
      "num_input_tokens_seen": 558825248,
      "step": 10855,
      "train_runtime": 66850.0152,
      "train_tokens_per_second": 8359.389
    },
    {
      "epoch": 0.7312146512254242,
      "grad_norm": 0.6428213445126294,
      "learning_rate": 7.049355771952087e-06,
      "loss": 0.34910609722137453,
      "num_input_tokens_seen": 559069848,
      "step": 10860,
      "train_runtime": 66880.6887,
      "train_tokens_per_second": 8359.212
    },
    {
      "epoch": 0.7315513062213843,
      "grad_norm": 0.6484794631604845,
      "learning_rate": 7.046943703350429e-06,
      "loss": 0.35908257961273193,
      "num_input_tokens_seen": 559327080,
      "step": 10865,
      "train_runtime": 66909.3528,
      "train_tokens_per_second": 8359.475
    },
    {
      "epoch": 0.7318879612173445,
      "grad_norm": 0.5376403840201005,
      "learning_rate": 7.044531062327492e-06,
      "loss": 0.3474988222122192,
      "num_input_tokens_seen": 559584296,
      "step": 10870,
      "train_runtime": 66938.6826,
      "train_tokens_per_second": 8359.655
    },
    {
      "epoch": 0.7322246162133046,
      "grad_norm": 0.590266032139924,
      "learning_rate": 7.042117849557963e-06,
      "loss": 0.36552603244781495,
      "num_input_tokens_seen": 559835736,
      "step": 10875,
      "train_runtime": 66970.2267,
      "train_tokens_per_second": 8359.472
    },
    {
      "epoch": 0.7325612712092647,
      "grad_norm": 0.8181001279913807,
      "learning_rate": 7.0397040657166885e-06,
      "loss": 0.36750819683074953,
      "num_input_tokens_seen": 560121184,
      "step": 10880,
      "train_runtime": 67001.129,
      "train_tokens_per_second": 8359.877
    },
    {
      "epoch": 0.7328979262052249,
      "grad_norm": 0.5767080984974863,
      "learning_rate": 7.037289711478676e-06,
      "loss": 0.3435854196548462,
      "num_input_tokens_seen": 560378680,
      "step": 10885,
      "train_runtime": 67033.3816,
      "train_tokens_per_second": 8359.696
    },
    {
      "epoch": 0.733234581201185,
      "grad_norm": 0.627304943021799,
      "learning_rate": 7.034874787519095e-06,
      "loss": 0.37999701499938965,
      "num_input_tokens_seen": 560644824,
      "step": 10890,
      "train_runtime": 67063.2622,
      "train_tokens_per_second": 8359.94
    },
    {
      "epoch": 0.7335712361971451,
      "grad_norm": 0.6168823657454228,
      "learning_rate": 7.0324592945132655e-06,
      "loss": 0.3995428323745728,
      "num_input_tokens_seen": 560918368,
      "step": 10895,
      "train_runtime": 67092.8915,
      "train_tokens_per_second": 8360.325
    },
    {
      "epoch": 0.7339078911931053,
      "grad_norm": 0.5959479926926255,
      "learning_rate": 7.0300432331366764e-06,
      "loss": 0.37626304626464846,
      "num_input_tokens_seen": 561172024,
      "step": 10900,
      "train_runtime": 67126.3108,
      "train_tokens_per_second": 8359.941
    },
    {
      "epoch": 0.7342445461890654,
      "grad_norm": 0.6105393515952683,
      "learning_rate": 7.02762660406497e-06,
      "loss": 0.37584078311920166,
      "num_input_tokens_seen": 561431008,
      "step": 10905,
      "train_runtime": 67158.2533,
      "train_tokens_per_second": 8359.821
    },
    {
      "epoch": 0.7345812011850256,
      "grad_norm": 0.5636442797302571,
      "learning_rate": 7.02520940797395e-06,
      "loss": 0.37336685657501223,
      "num_input_tokens_seen": 561671272,
      "step": 10910,
      "train_runtime": 67187.3256,
      "train_tokens_per_second": 8359.78
    },
    {
      "epoch": 0.7349178561809857,
      "grad_norm": 0.558975535512041,
      "learning_rate": 7.022791645539576e-06,
      "loss": 0.36878304481506347,
      "num_input_tokens_seen": 561912024,
      "step": 10915,
      "train_runtime": 67219.8165,
      "train_tokens_per_second": 8359.321
    },
    {
      "epoch": 0.7352545111769458,
      "grad_norm": 0.6012905783458087,
      "learning_rate": 7.02037331743797e-06,
      "loss": 0.39553213119506836,
      "num_input_tokens_seen": 562177672,
      "step": 10920,
      "train_runtime": 67250.0393,
      "train_tokens_per_second": 8359.514
    },
    {
      "epoch": 0.735591166172906,
      "grad_norm": 0.6025874932486689,
      "learning_rate": 7.0179544243454076e-06,
      "loss": 0.34576849937438964,
      "num_input_tokens_seen": 562450656,
      "step": 10925,
      "train_runtime": 67281.6167,
      "train_tokens_per_second": 8359.648
    },
    {
      "epoch": 0.7359278211688661,
      "grad_norm": 0.7253638249630419,
      "learning_rate": 7.0155349669383245e-06,
      "loss": 0.37188146114349363,
      "num_input_tokens_seen": 562718624,
      "step": 10930,
      "train_runtime": 67309.566,
      "train_tokens_per_second": 8360.158
    },
    {
      "epoch": 0.7362644761648263,
      "grad_norm": 0.6002993645325481,
      "learning_rate": 7.013114945893316e-06,
      "loss": 0.352215051651001,
      "num_input_tokens_seen": 562974752,
      "step": 10935,
      "train_runtime": 67340.9874,
      "train_tokens_per_second": 8360.061
    },
    {
      "epoch": 0.7366011311607864,
      "grad_norm": 0.6140749293613379,
      "learning_rate": 7.010694361887131e-06,
      "loss": 0.3519169330596924,
      "num_input_tokens_seen": 563233384,
      "step": 10940,
      "train_runtime": 67370.3409,
      "train_tokens_per_second": 8360.257
    },
    {
      "epoch": 0.7369377861567465,
      "grad_norm": 0.5786265327527237,
      "learning_rate": 7.008273215596681e-06,
      "loss": 0.37504940032958983,
      "num_input_tokens_seen": 563491928,
      "step": 10945,
      "train_runtime": 67405.8128,
      "train_tokens_per_second": 8359.693
    },
    {
      "epoch": 0.7372744411527067,
      "grad_norm": 0.6771444722010472,
      "learning_rate": 7.005851507699029e-06,
      "loss": 0.36923789978027344,
      "num_input_tokens_seen": 563766720,
      "step": 10950,
      "train_runtime": 67439.5498,
      "train_tokens_per_second": 8359.586
    },
    {
      "epoch": 0.7376110961486668,
      "grad_norm": 0.6218258068813654,
      "learning_rate": 7.003429238871399e-06,
      "loss": 0.3622065305709839,
      "num_input_tokens_seen": 564024600,
      "step": 10955,
      "train_runtime": 67468.7547,
      "train_tokens_per_second": 8359.79
    },
    {
      "epoch": 0.737947751144627,
      "grad_norm": 0.6130678874390081,
      "learning_rate": 7.001006409791171e-06,
      "loss": 0.39206857681274415,
      "num_input_tokens_seen": 564278384,
      "step": 10960,
      "train_runtime": 67497.4252,
      "train_tokens_per_second": 8359.999
    },
    {
      "epoch": 0.7382844061405871,
      "grad_norm": 0.6781983902568675,
      "learning_rate": 6.998583021135882e-06,
      "loss": 0.3724649906158447,
      "num_input_tokens_seen": 564540408,
      "step": 10965,
      "train_runtime": 67528.6891,
      "train_tokens_per_second": 8360.008
    },
    {
      "epoch": 0.7386210611365472,
      "grad_norm": 0.7120245841991889,
      "learning_rate": 6.996159073583224e-06,
      "loss": 0.36175124645233153,
      "num_input_tokens_seen": 564804840,
      "step": 10970,
      "train_runtime": 67556.6696,
      "train_tokens_per_second": 8360.46
    },
    {
      "epoch": 0.7389577161325074,
      "grad_norm": 0.5962535634091419,
      "learning_rate": 6.993734567811046e-06,
      "loss": 0.33404741287231443,
      "num_input_tokens_seen": 565049784,
      "step": 10975,
      "train_runtime": 67587.872,
      "train_tokens_per_second": 8360.225
    },
    {
      "epoch": 0.7392943711284675,
      "grad_norm": 0.5749413252763911,
      "learning_rate": 6.991309504497352e-06,
      "loss": 0.3726051330566406,
      "num_input_tokens_seen": 565314896,
      "step": 10980,
      "train_runtime": 67620.6408,
      "train_tokens_per_second": 8360.094
    },
    {
      "epoch": 0.7396310261244277,
      "grad_norm": 0.6670353297175254,
      "learning_rate": 6.988883884320307e-06,
      "loss": 0.3317781209945679,
      "num_input_tokens_seen": 565555544,
      "step": 10985,
      "train_runtime": 67650.4227,
      "train_tokens_per_second": 8359.971
    },
    {
      "epoch": 0.7399676811203878,
      "grad_norm": 0.6330642172028483,
      "learning_rate": 6.9864577079582254e-06,
      "loss": 0.38596382141113283,
      "num_input_tokens_seen": 565828376,
      "step": 10990,
      "train_runtime": 67679.5753,
      "train_tokens_per_second": 8360.401
    },
    {
      "epoch": 0.7403043361163479,
      "grad_norm": 0.684687883678918,
      "learning_rate": 6.984030976089581e-06,
      "loss": 0.3484159231185913,
      "num_input_tokens_seen": 566091664,
      "step": 10995,
      "train_runtime": 67708.2662,
      "train_tokens_per_second": 8360.747
    },
    {
      "epoch": 0.7406409911123081,
      "grad_norm": 0.6981559074616794,
      "learning_rate": 6.981603689392999e-06,
      "loss": 0.36209945678710936,
      "num_input_tokens_seen": 566332632,
      "step": 11000,
      "train_runtime": 67734.602,
      "train_tokens_per_second": 8361.054
    },
    {
      "epoch": 0.7409776461082682,
      "grad_norm": 0.6590154684823427,
      "learning_rate": 6.979175848547263e-06,
      "loss": 0.35679724216461184,
      "num_input_tokens_seen": 566584016,
      "step": 11005,
      "train_runtime": 67766.7542,
      "train_tokens_per_second": 8360.796
    },
    {
      "epoch": 0.7413143011042284,
      "grad_norm": 0.7503185695969601,
      "learning_rate": 6.976747454231316e-06,
      "loss": 0.3636085033416748,
      "num_input_tokens_seen": 566804816,
      "step": 11010,
      "train_runtime": 67797.167,
      "train_tokens_per_second": 8360.302
    },
    {
      "epoch": 0.7416509561001885,
      "grad_norm": 0.6363283527175061,
      "learning_rate": 6.974318507124245e-06,
      "loss": 0.3717265367507935,
      "num_input_tokens_seen": 567056384,
      "step": 11015,
      "train_runtime": 67828.8781,
      "train_tokens_per_second": 8360.103
    },
    {
      "epoch": 0.7419876110961486,
      "grad_norm": 0.8481965533382619,
      "learning_rate": 6.971889007905299e-06,
      "loss": 0.375919246673584,
      "num_input_tokens_seen": 567307432,
      "step": 11020,
      "train_runtime": 67859.0346,
      "train_tokens_per_second": 8360.087
    },
    {
      "epoch": 0.7423242660921088,
      "grad_norm": 0.6302235594821192,
      "learning_rate": 6.9694589572538795e-06,
      "loss": 0.40106749534606934,
      "num_input_tokens_seen": 567572760,
      "step": 11025,
      "train_runtime": 67888.653,
      "train_tokens_per_second": 8360.348
    },
    {
      "epoch": 0.7426609210880689,
      "grad_norm": 0.6310705242314048,
      "learning_rate": 6.9670283558495436e-06,
      "loss": 0.3647524118423462,
      "num_input_tokens_seen": 567838392,
      "step": 11030,
      "train_runtime": 67921.7204,
      "train_tokens_per_second": 8360.189
    },
    {
      "epoch": 0.742997576084029,
      "grad_norm": 0.5811352961198177,
      "learning_rate": 6.964597204371999e-06,
      "loss": 0.3671701431274414,
      "num_input_tokens_seen": 568097744,
      "step": 11035,
      "train_runtime": 67952.076,
      "train_tokens_per_second": 8360.271
    },
    {
      "epoch": 0.7433342310799892,
      "grad_norm": 0.6835735987222828,
      "learning_rate": 6.962165503501111e-06,
      "loss": 0.34223482608795164,
      "num_input_tokens_seen": 568351376,
      "step": 11040,
      "train_runtime": 67984.8269,
      "train_tokens_per_second": 8359.974
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.6031224443855633,
      "learning_rate": 6.959733253916899e-06,
      "loss": 0.381845760345459,
      "num_input_tokens_seen": 568620760,
      "step": 11045,
      "train_runtime": 68016.8684,
      "train_tokens_per_second": 8359.996
    },
    {
      "epoch": 0.7440075410719095,
      "grad_norm": 0.6431416821829419,
      "learning_rate": 6.95730045629953e-06,
      "loss": 0.35871272087097167,
      "num_input_tokens_seen": 568847008,
      "step": 11050,
      "train_runtime": 68043.3152,
      "train_tokens_per_second": 8360.072
    },
    {
      "epoch": 0.7443441960678696,
      "grad_norm": 0.726149115614154,
      "learning_rate": 6.954867111329329e-06,
      "loss": 0.402852725982666,
      "num_input_tokens_seen": 569122552,
      "step": 11055,
      "train_runtime": 68071.4652,
      "train_tokens_per_second": 8360.663
    },
    {
      "epoch": 0.7446808510638298,
      "grad_norm": 0.6772971893940513,
      "learning_rate": 6.952433219686775e-06,
      "loss": 0.3570889711380005,
      "num_input_tokens_seen": 569388296,
      "step": 11060,
      "train_runtime": 68099.7242,
      "train_tokens_per_second": 8361.095
    },
    {
      "epoch": 0.7450175060597899,
      "grad_norm": 0.5543073991834603,
      "learning_rate": 6.949998782052497e-06,
      "loss": 0.38263845443725586,
      "num_input_tokens_seen": 569652000,
      "step": 11065,
      "train_runtime": 68128.4944,
      "train_tokens_per_second": 8361.435
    },
    {
      "epoch": 0.74535416105575,
      "grad_norm": 0.674696821575928,
      "learning_rate": 6.947563799107277e-06,
      "loss": 0.3533199310302734,
      "num_input_tokens_seen": 569903080,
      "step": 11070,
      "train_runtime": 68158.3687,
      "train_tokens_per_second": 8361.454
    },
    {
      "epoch": 0.7456908160517102,
      "grad_norm": 0.636581819140668,
      "learning_rate": 6.945128271532052e-06,
      "loss": 0.35483999252319337,
      "num_input_tokens_seen": 570191200,
      "step": 11075,
      "train_runtime": 68188.9212,
      "train_tokens_per_second": 8361.933
    },
    {
      "epoch": 0.7460274710476703,
      "grad_norm": 0.6866493439065086,
      "learning_rate": 6.942692200007908e-06,
      "loss": 0.3714379549026489,
      "num_input_tokens_seen": 570441008,
      "step": 11080,
      "train_runtime": 68218.7624,
      "train_tokens_per_second": 8361.937
    },
    {
      "epoch": 0.7463641260436304,
      "grad_norm": 0.6704944103959332,
      "learning_rate": 6.9402555852160845e-06,
      "loss": 0.3725388526916504,
      "num_input_tokens_seen": 570710864,
      "step": 11085,
      "train_runtime": 68252.7189,
      "train_tokens_per_second": 8361.731
    },
    {
      "epoch": 0.7467007810395906,
      "grad_norm": 0.5967925217729835,
      "learning_rate": 6.937818427837975e-06,
      "loss": 0.35423455238342283,
      "num_input_tokens_seen": 570974128,
      "step": 11090,
      "train_runtime": 68280.8265,
      "train_tokens_per_second": 8362.144
    },
    {
      "epoch": 0.7470374360355507,
      "grad_norm": 0.6769701248476228,
      "learning_rate": 6.93538072855512e-06,
      "loss": 0.38665955066680907,
      "num_input_tokens_seen": 571240720,
      "step": 11095,
      "train_runtime": 68312.5779,
      "train_tokens_per_second": 8362.16
    },
    {
      "epoch": 0.7473740910315109,
      "grad_norm": 0.5893642357121762,
      "learning_rate": 6.932942488049217e-06,
      "loss": 0.32885358333587644,
      "num_input_tokens_seen": 571496480,
      "step": 11100,
      "train_runtime": 68342.1627,
      "train_tokens_per_second": 8362.283
    },
    {
      "epoch": 0.747710746027471,
      "grad_norm": 0.7056534969097168,
      "learning_rate": 6.930503707002108e-06,
      "loss": 0.37997496128082275,
      "num_input_tokens_seen": 571750680,
      "step": 11105,
      "train_runtime": 68374.3606,
      "train_tokens_per_second": 8362.063
    },
    {
      "epoch": 0.7480474010234311,
      "grad_norm": 0.7188059210464478,
      "learning_rate": 6.928064386095796e-06,
      "loss": 0.36277825832366944,
      "num_input_tokens_seen": 572009920,
      "step": 11110,
      "train_runtime": 68405.3613,
      "train_tokens_per_second": 8362.063
    },
    {
      "epoch": 0.7483840560193913,
      "grad_norm": 0.6901999726336663,
      "learning_rate": 6.9256245260124264e-06,
      "loss": 0.37076606750488283,
      "num_input_tokens_seen": 572285480,
      "step": 11115,
      "train_runtime": 68436.912,
      "train_tokens_per_second": 8362.234
    },
    {
      "epoch": 0.7487207110153514,
      "grad_norm": 0.6381967802181029,
      "learning_rate": 6.923184127434298e-06,
      "loss": 0.3691317081451416,
      "num_input_tokens_seen": 572521080,
      "step": 11120,
      "train_runtime": 68467.1419,
      "train_tokens_per_second": 8361.983
    },
    {
      "epoch": 0.7490573660113116,
      "grad_norm": 0.7720754242199899,
      "learning_rate": 6.920743191043861e-06,
      "loss": 0.35338869094848635,
      "num_input_tokens_seen": 572760952,
      "step": 11125,
      "train_runtime": 68495.928,
      "train_tokens_per_second": 8361.971
    },
    {
      "epoch": 0.7493940210072717,
      "grad_norm": 0.5528365103713967,
      "learning_rate": 6.918301717523714e-06,
      "loss": 0.336551833152771,
      "num_input_tokens_seen": 573015600,
      "step": 11130,
      "train_runtime": 68528.7223,
      "train_tokens_per_second": 8361.685
    },
    {
      "epoch": 0.7497306760032318,
      "grad_norm": 0.6692299005413853,
      "learning_rate": 6.915859707556611e-06,
      "loss": 0.3856123208999634,
      "num_input_tokens_seen": 573274320,
      "step": 11135,
      "train_runtime": 68559.471,
      "train_tokens_per_second": 8361.709
    },
    {
      "epoch": 0.750067330999192,
      "grad_norm": 0.664619126831313,
      "learning_rate": 6.913417161825449e-06,
      "loss": 0.4029876708984375,
      "num_input_tokens_seen": 573526800,
      "step": 11140,
      "train_runtime": 68587.2861,
      "train_tokens_per_second": 8361.999
    },
    {
      "epoch": 0.7504039859951521,
      "grad_norm": 0.644397121576548,
      "learning_rate": 6.91097408101328e-06,
      "loss": 0.37030735015869143,
      "num_input_tokens_seen": 573775432,
      "step": 11145,
      "train_runtime": 68617.0428,
      "train_tokens_per_second": 8361.996
    },
    {
      "epoch": 0.7507406409911123,
      "grad_norm": 0.7277260695653828,
      "learning_rate": 6.908530465803302e-06,
      "loss": 0.3577216386795044,
      "num_input_tokens_seen": 574026152,
      "step": 11150,
      "train_runtime": 68644.2767,
      "train_tokens_per_second": 8362.331
    },
    {
      "epoch": 0.7510772959870724,
      "grad_norm": 0.8324739417621774,
      "learning_rate": 6.906086316878865e-06,
      "loss": 0.3659271240234375,
      "num_input_tokens_seen": 574255168,
      "step": 11155,
      "train_runtime": 68674.9133,
      "train_tokens_per_second": 8361.935
    },
    {
      "epoch": 0.7514139509830325,
      "grad_norm": 0.5894696304219622,
      "learning_rate": 6.903641634923468e-06,
      "loss": 0.3377214431762695,
      "num_input_tokens_seen": 574509368,
      "step": 11160,
      "train_runtime": 68707.3028,
      "train_tokens_per_second": 8361.693
    },
    {
      "epoch": 0.7517506059789927,
      "grad_norm": 0.5724588823146445,
      "learning_rate": 6.9011964206207585e-06,
      "loss": 0.3798540115356445,
      "num_input_tokens_seen": 574754648,
      "step": 11165,
      "train_runtime": 68734.8862,
      "train_tokens_per_second": 8361.906
    },
    {
      "epoch": 0.7520872609749528,
      "grad_norm": 0.706769349878928,
      "learning_rate": 6.898750674654531e-06,
      "loss": 0.3486687183380127,
      "num_input_tokens_seen": 575023600,
      "step": 11170,
      "train_runtime": 68762.8883,
      "train_tokens_per_second": 8362.412
    },
    {
      "epoch": 0.752423915970913,
      "grad_norm": 0.5758710703657236,
      "learning_rate": 6.896304397708733e-06,
      "loss": 0.34326648712158203,
      "num_input_tokens_seen": 575251432,
      "step": 11175,
      "train_runtime": 68794.5116,
      "train_tokens_per_second": 8361.88
    },
    {
      "epoch": 0.7527605709668731,
      "grad_norm": 0.8070761461907264,
      "learning_rate": 6.893857590467456e-06,
      "loss": 0.35099196434020996,
      "num_input_tokens_seen": 575516136,
      "step": 11180,
      "train_runtime": 68822.7445,
      "train_tokens_per_second": 8362.296
    },
    {
      "epoch": 0.7530972259628332,
      "grad_norm": 0.6828069100388466,
      "learning_rate": 6.891410253614944e-06,
      "loss": 0.34926600456237794,
      "num_input_tokens_seen": 575747512,
      "step": 11185,
      "train_runtime": 68851.9607,
      "train_tokens_per_second": 8362.108
    },
    {
      "epoch": 0.7534338809587934,
      "grad_norm": 0.6425063946458784,
      "learning_rate": 6.888962387835586e-06,
      "loss": 0.377002477645874,
      "num_input_tokens_seen": 576030584,
      "step": 11190,
      "train_runtime": 68884.0501,
      "train_tokens_per_second": 8362.322
    },
    {
      "epoch": 0.7537705359547535,
      "grad_norm": 0.7100904420118469,
      "learning_rate": 6.886513993813919e-06,
      "loss": 0.3777203321456909,
      "num_input_tokens_seen": 576273592,
      "step": 11195,
      "train_runtime": 68915.0724,
      "train_tokens_per_second": 8362.084
    },
    {
      "epoch": 0.7541071909507137,
      "grad_norm": 0.8771748096172317,
      "learning_rate": 6.88406507223463e-06,
      "loss": 0.3521225929260254,
      "num_input_tokens_seen": 576548000,
      "step": 11200,
      "train_runtime": 68946.8427,
      "train_tokens_per_second": 8362.21
    },
    {
      "epoch": 0.7544438459466738,
      "grad_norm": 0.6783235258094383,
      "learning_rate": 6.881615623782551e-06,
      "loss": 0.3556192398071289,
      "num_input_tokens_seen": 576797048,
      "step": 11205,
      "train_runtime": 68976.2344,
      "train_tokens_per_second": 8362.258
    },
    {
      "epoch": 0.7547805009426339,
      "grad_norm": 0.6753758764517608,
      "learning_rate": 6.879165649142663e-06,
      "loss": 0.35202369689941404,
      "num_input_tokens_seen": 577024880,
      "step": 11210,
      "train_runtime": 69002.9472,
      "train_tokens_per_second": 8362.322
    },
    {
      "epoch": 0.7551171559385941,
      "grad_norm": 0.5785350512953161,
      "learning_rate": 6.876715149000093e-06,
      "loss": 0.3562492847442627,
      "num_input_tokens_seen": 577264888,
      "step": 11215,
      "train_runtime": 69031.5666,
      "train_tokens_per_second": 8362.332
    },
    {
      "epoch": 0.7554538109345542,
      "grad_norm": 0.5927404000097077,
      "learning_rate": 6.874264124040117e-06,
      "loss": 0.3591607570648193,
      "num_input_tokens_seen": 577544832,
      "step": 11220,
      "train_runtime": 69062.0673,
      "train_tokens_per_second": 8362.693
    },
    {
      "epoch": 0.7557904659305144,
      "grad_norm": 0.6351489841234415,
      "learning_rate": 6.871812574948154e-06,
      "loss": 0.3713825225830078,
      "num_input_tokens_seen": 577813936,
      "step": 11225,
      "train_runtime": 69093.1361,
      "train_tokens_per_second": 8362.827
    },
    {
      "epoch": 0.7561271209264745,
      "grad_norm": 0.6974427777982476,
      "learning_rate": 6.869360502409773e-06,
      "loss": 0.37527856826782224,
      "num_input_tokens_seen": 578071232,
      "step": 11230,
      "train_runtime": 69127.3443,
      "train_tokens_per_second": 8362.411
    },
    {
      "epoch": 0.7564637759224346,
      "grad_norm": 0.6680503342972961,
      "learning_rate": 6.866907907110689e-06,
      "loss": 0.34317874908447266,
      "num_input_tokens_seen": 578305608,
      "step": 11235,
      "train_runtime": 69154.033,
      "train_tokens_per_second": 8362.572
    },
    {
      "epoch": 0.7568004309183949,
      "grad_norm": 0.5328856987183879,
      "learning_rate": 6.864454789736763e-06,
      "loss": 0.33777642250061035,
      "num_input_tokens_seen": 578591872,
      "step": 11240,
      "train_runtime": 69183.9771,
      "train_tokens_per_second": 8363.091
    },
    {
      "epoch": 0.757137085914355,
      "grad_norm": 0.611949372070363,
      "learning_rate": 6.8620011509740005e-06,
      "loss": 0.3582913398742676,
      "num_input_tokens_seen": 578839624,
      "step": 11245,
      "train_runtime": 69212.565,
      "train_tokens_per_second": 8363.216
    },
    {
      "epoch": 0.7574737409103152,
      "grad_norm": 0.6055766245079415,
      "learning_rate": 6.859546991508552e-06,
      "loss": 0.3471322536468506,
      "num_input_tokens_seen": 579100136,
      "step": 11250,
      "train_runtime": 69244.2263,
      "train_tokens_per_second": 8363.154
    },
    {
      "epoch": 0.7578103959062753,
      "grad_norm": 0.6634498648327516,
      "learning_rate": 6.857092312026716e-06,
      "loss": 0.37633867263793946,
      "num_input_tokens_seen": 579382008,
      "step": 11255,
      "train_runtime": 69270.9587,
      "train_tokens_per_second": 8363.996
    },
    {
      "epoch": 0.7581470509022354,
      "grad_norm": 0.6665684574326222,
      "learning_rate": 6.8546371132149395e-06,
      "loss": 0.38732476234436036,
      "num_input_tokens_seen": 579641808,
      "step": 11260,
      "train_runtime": 69301.4989,
      "train_tokens_per_second": 8364.059
    },
    {
      "epoch": 0.7584837058981956,
      "grad_norm": 0.6353413817814473,
      "learning_rate": 6.852181395759807e-06,
      "loss": 0.37858726978302004,
      "num_input_tokens_seen": 579917784,
      "step": 11265,
      "train_runtime": 69333.1628,
      "train_tokens_per_second": 8364.219
    },
    {
      "epoch": 0.7588203608941557,
      "grad_norm": 0.5073625499645082,
      "learning_rate": 6.849725160348054e-06,
      "loss": 0.34138805866241456,
      "num_input_tokens_seen": 580150312,
      "step": 11270,
      "train_runtime": 69359.4497,
      "train_tokens_per_second": 8364.402
    },
    {
      "epoch": 0.7591570158901159,
      "grad_norm": 0.639879920808025,
      "learning_rate": 6.847268407666557e-06,
      "loss": 0.3464106559753418,
      "num_input_tokens_seen": 580412232,
      "step": 11275,
      "train_runtime": 69387.4384,
      "train_tokens_per_second": 8364.803
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.7780199545163043,
      "learning_rate": 6.844811138402341e-06,
      "loss": 0.3703222513198853,
      "num_input_tokens_seen": 580669176,
      "step": 11280,
      "train_runtime": 69420.0173,
      "train_tokens_per_second": 8364.578
    },
    {
      "epoch": 0.7598303258820361,
      "grad_norm": 0.666545804479237,
      "learning_rate": 6.842353353242574e-06,
      "loss": 0.35576980113983153,
      "num_input_tokens_seen": 580908656,
      "step": 11285,
      "train_runtime": 69445.8263,
      "train_tokens_per_second": 8364.918
    },
    {
      "epoch": 0.7601669808779963,
      "grad_norm": 0.7570249552205051,
      "learning_rate": 6.839895052874565e-06,
      "loss": 0.3743896484375,
      "num_input_tokens_seen": 581180112,
      "step": 11290,
      "train_runtime": 69476.2079,
      "train_tokens_per_second": 8365.167
    },
    {
      "epoch": 0.7605036358739564,
      "grad_norm": 0.7950690512708716,
      "learning_rate": 6.837436237985772e-06,
      "loss": 0.3548635959625244,
      "num_input_tokens_seen": 581411840,
      "step": 11295,
      "train_runtime": 69502.6098,
      "train_tokens_per_second": 8365.324
    },
    {
      "epoch": 0.7608402908699166,
      "grad_norm": 0.8168982810858051,
      "learning_rate": 6.834976909263794e-06,
      "loss": 0.3814142942428589,
      "num_input_tokens_seen": 581657136,
      "step": 11300,
      "train_runtime": 69535.388,
      "train_tokens_per_second": 8364.908
    },
    {
      "epoch": 0.7611769458658767,
      "grad_norm": 0.672538581832751,
      "learning_rate": 6.832517067396374e-06,
      "loss": 0.3600905179977417,
      "num_input_tokens_seen": 581930528,
      "step": 11305,
      "train_runtime": 69569.8869,
      "train_tokens_per_second": 8364.69
    },
    {
      "epoch": 0.7615136008618368,
      "grad_norm": 0.6186863670020163,
      "learning_rate": 6.8300567130714e-06,
      "loss": 0.38466434478759765,
      "num_input_tokens_seen": 582206784,
      "step": 11310,
      "train_runtime": 69600.9508,
      "train_tokens_per_second": 8364.926
    },
    {
      "epoch": 0.761850255857797,
      "grad_norm": 0.6210975268179282,
      "learning_rate": 6.8275958469769035e-06,
      "loss": 0.3704275608062744,
      "num_input_tokens_seen": 582481296,
      "step": 11315,
      "train_runtime": 69632.4705,
      "train_tokens_per_second": 8365.082
    },
    {
      "epoch": 0.7621869108537571,
      "grad_norm": 0.6478266131795628,
      "learning_rate": 6.825134469801055e-06,
      "loss": 0.35091581344604494,
      "num_input_tokens_seen": 582766496,
      "step": 11320,
      "train_runtime": 69662.5165,
      "train_tokens_per_second": 8365.568
    },
    {
      "epoch": 0.7625235658497173,
      "grad_norm": 0.6486569228997133,
      "learning_rate": 6.8226725822321715e-06,
      "loss": 0.3927011966705322,
      "num_input_tokens_seen": 583011376,
      "step": 11325,
      "train_runtime": 69697.9173,
      "train_tokens_per_second": 8364.832
    },
    {
      "epoch": 0.7628602208456774,
      "grad_norm": 0.6850860351303039,
      "learning_rate": 6.820210184958712e-06,
      "loss": 0.37799103260040284,
      "num_input_tokens_seen": 583276688,
      "step": 11330,
      "train_runtime": 69728.8754,
      "train_tokens_per_second": 8364.923
    },
    {
      "epoch": 0.7631968758416375,
      "grad_norm": 0.6920457785297098,
      "learning_rate": 6.81774727866928e-06,
      "loss": 0.37792081832885743,
      "num_input_tokens_seen": 583514192,
      "step": 11335,
      "train_runtime": 69762.2532,
      "train_tokens_per_second": 8364.325
    },
    {
      "epoch": 0.7635335308375977,
      "grad_norm": 0.7234269786021804,
      "learning_rate": 6.815283864052617e-06,
      "loss": 0.3670658111572266,
      "num_input_tokens_seen": 583771400,
      "step": 11340,
      "train_runtime": 69793.1408,
      "train_tokens_per_second": 8364.309
    },
    {
      "epoch": 0.7638701858335578,
      "grad_norm": 0.7351183510555562,
      "learning_rate": 6.812819941797611e-06,
      "loss": 0.37823798656463625,
      "num_input_tokens_seen": 584005344,
      "step": 11345,
      "train_runtime": 69822.3131,
      "train_tokens_per_second": 8364.165
    },
    {
      "epoch": 0.764206840829518,
      "grad_norm": 0.6464141684558143,
      "learning_rate": 6.810355512593288e-06,
      "loss": 0.38045520782470704,
      "num_input_tokens_seen": 584277752,
      "step": 11350,
      "train_runtime": 69855.1749,
      "train_tokens_per_second": 8364.13
    },
    {
      "epoch": 0.7645434958254781,
      "grad_norm": 0.5887469046161972,
      "learning_rate": 6.807890577128817e-06,
      "loss": 0.33489766120910647,
      "num_input_tokens_seen": 584525896,
      "step": 11355,
      "train_runtime": 69883.6997,
      "train_tokens_per_second": 8364.267
    },
    {
      "epoch": 0.7648801508214382,
      "grad_norm": 0.7202993385056916,
      "learning_rate": 6.805425136093513e-06,
      "loss": 0.3314122915267944,
      "num_input_tokens_seen": 584792840,
      "step": 11360,
      "train_runtime": 69913.4492,
      "train_tokens_per_second": 8364.526
    },
    {
      "epoch": 0.7652168058173984,
      "grad_norm": 0.6935939435281171,
      "learning_rate": 6.802959190176827e-06,
      "loss": 0.37778308391571047,
      "num_input_tokens_seen": 585028960,
      "step": 11365,
      "train_runtime": 69944.548,
      "train_tokens_per_second": 8364.182
    },
    {
      "epoch": 0.7655534608133585,
      "grad_norm": 0.6695051482186574,
      "learning_rate": 6.800492740068352e-06,
      "loss": 0.3565650701522827,
      "num_input_tokens_seen": 585272952,
      "step": 11370,
      "train_runtime": 69977.7108,
      "train_tokens_per_second": 8363.705
    },
    {
      "epoch": 0.7658901158093187,
      "grad_norm": 0.7315476103881744,
      "learning_rate": 6.798025786457825e-06,
      "loss": 0.35672051906585694,
      "num_input_tokens_seen": 585538312,
      "step": 11375,
      "train_runtime": 70007.0582,
      "train_tokens_per_second": 8363.99
    },
    {
      "epoch": 0.7662267708052788,
      "grad_norm": 0.5438466974024655,
      "learning_rate": 6.795558330035118e-06,
      "loss": 0.3592084884643555,
      "num_input_tokens_seen": 585798240,
      "step": 11380,
      "train_runtime": 70039.1239,
      "train_tokens_per_second": 8363.872
    },
    {
      "epoch": 0.7665634258012389,
      "grad_norm": 0.5250159378293867,
      "learning_rate": 6.793090371490251e-06,
      "loss": 0.3769026517868042,
      "num_input_tokens_seen": 586071064,
      "step": 11385,
      "train_runtime": 70070.4323,
      "train_tokens_per_second": 8364.028
    },
    {
      "epoch": 0.7669000807971991,
      "grad_norm": 0.5602235807679967,
      "learning_rate": 6.790621911513378e-06,
      "loss": 0.34200620651245117,
      "num_input_tokens_seen": 586344080,
      "step": 11390,
      "train_runtime": 70096.5018,
      "train_tokens_per_second": 8364.812
    },
    {
      "epoch": 0.7672367357931592,
      "grad_norm": 0.6042846004620436,
      "learning_rate": 6.788152950794799e-06,
      "loss": 0.3849551916122437,
      "num_input_tokens_seen": 586604400,
      "step": 11395,
      "train_runtime": 70127.7056,
      "train_tokens_per_second": 8364.802
    },
    {
      "epoch": 0.7675733907891193,
      "grad_norm": 0.7113269321893222,
      "learning_rate": 6.7856834900249495e-06,
      "loss": 0.38363258838653563,
      "num_input_tokens_seen": 586854344,
      "step": 11400,
      "train_runtime": 70158.6095,
      "train_tokens_per_second": 8364.68
    },
    {
      "epoch": 0.7679100457850795,
      "grad_norm": 0.7046236714964426,
      "learning_rate": 6.783213529894405e-06,
      "loss": 0.3427259922027588,
      "num_input_tokens_seen": 587104592,
      "step": 11405,
      "train_runtime": 70190.2229,
      "train_tokens_per_second": 8364.478
    },
    {
      "epoch": 0.7682467007810396,
      "grad_norm": 0.6580098124954733,
      "learning_rate": 6.780743071093883e-06,
      "loss": 0.33823044300079347,
      "num_input_tokens_seen": 587368952,
      "step": 11410,
      "train_runtime": 70221.4521,
      "train_tokens_per_second": 8364.523
    },
    {
      "epoch": 0.7685833557769998,
      "grad_norm": 0.6317843087325136,
      "learning_rate": 6.77827211431424e-06,
      "loss": 0.35743021965026855,
      "num_input_tokens_seen": 587641696,
      "step": 11415,
      "train_runtime": 70253.9137,
      "train_tokens_per_second": 8364.54
    },
    {
      "epoch": 0.7689200107729599,
      "grad_norm": 0.6767685994131393,
      "learning_rate": 6.775800660246471e-06,
      "loss": 0.3646695613861084,
      "num_input_tokens_seen": 587873568,
      "step": 11420,
      "train_runtime": 70284.0188,
      "train_tokens_per_second": 8364.257
    },
    {
      "epoch": 0.76925666576892,
      "grad_norm": 0.5937118645537143,
      "learning_rate": 6.77332870958171e-06,
      "loss": 0.3489880084991455,
      "num_input_tokens_seen": 588150584,
      "step": 11425,
      "train_runtime": 70315.7095,
      "train_tokens_per_second": 8364.427
    },
    {
      "epoch": 0.7695933207648802,
      "grad_norm": 0.6394066977972965,
      "learning_rate": 6.770856263011229e-06,
      "loss": 0.36341185569763185,
      "num_input_tokens_seen": 588393496,
      "step": 11430,
      "train_runtime": 70349.9978,
      "train_tokens_per_second": 8363.803
    },
    {
      "epoch": 0.7699299757608403,
      "grad_norm": 0.7041001299872311,
      "learning_rate": 6.768383321226442e-06,
      "loss": 0.3821458339691162,
      "num_input_tokens_seen": 588630728,
      "step": 11435,
      "train_runtime": 70380.4926,
      "train_tokens_per_second": 8363.549
    },
    {
      "epoch": 0.7702666307568005,
      "grad_norm": 0.6753887555969543,
      "learning_rate": 6.765909884918898e-06,
      "loss": 0.3527059078216553,
      "num_input_tokens_seen": 588877576,
      "step": 11440,
      "train_runtime": 70412.1765,
      "train_tokens_per_second": 8363.292
    },
    {
      "epoch": 0.7706032857527606,
      "grad_norm": 0.5493872744618454,
      "learning_rate": 6.763435954780286e-06,
      "loss": 0.37638809680938723,
      "num_input_tokens_seen": 589142064,
      "step": 11445,
      "train_runtime": 70443.7198,
      "train_tokens_per_second": 8363.301
    },
    {
      "epoch": 0.7709399407487207,
      "grad_norm": 0.6365265177135113,
      "learning_rate": 6.7609615315024325e-06,
      "loss": 0.35949316024780276,
      "num_input_tokens_seen": 589390256,
      "step": 11450,
      "train_runtime": 70474.4844,
      "train_tokens_per_second": 8363.172
    },
    {
      "epoch": 0.7712765957446809,
      "grad_norm": 0.7347479031581244,
      "learning_rate": 6.758486615777301e-06,
      "loss": 0.37460434436798096,
      "num_input_tokens_seen": 589641352,
      "step": 11455,
      "train_runtime": 70504.8971,
      "train_tokens_per_second": 8363.126
    },
    {
      "epoch": 0.771613250740641,
      "grad_norm": 0.8093432752704821,
      "learning_rate": 6.756011208296995e-06,
      "loss": 0.35287790298461913,
      "num_input_tokens_seen": 589894768,
      "step": 11460,
      "train_runtime": 70535.0349,
      "train_tokens_per_second": 8363.146
    },
    {
      "epoch": 0.7719499057366012,
      "grad_norm": 0.6703581341905691,
      "learning_rate": 6.753535309753755e-06,
      "loss": 0.36010894775390623,
      "num_input_tokens_seen": 590124952,
      "step": 11465,
      "train_runtime": 70563.0752,
      "train_tokens_per_second": 8363.084
    },
    {
      "epoch": 0.7722865607325613,
      "grad_norm": 0.5792236825393171,
      "learning_rate": 6.7510589208399565e-06,
      "loss": 0.37015910148620607,
      "num_input_tokens_seen": 590381584,
      "step": 11470,
      "train_runtime": 70592.6171,
      "train_tokens_per_second": 8363.22
    },
    {
      "epoch": 0.7726232157285214,
      "grad_norm": 0.6833944270800848,
      "learning_rate": 6.748582042248115e-06,
      "loss": 0.3562369108200073,
      "num_input_tokens_seen": 590630424,
      "step": 11475,
      "train_runtime": 70623.5998,
      "train_tokens_per_second": 8363.074
    },
    {
      "epoch": 0.7729598707244816,
      "grad_norm": 0.6751057774917895,
      "learning_rate": 6.746104674670882e-06,
      "loss": 0.3763397455215454,
      "num_input_tokens_seen": 590901672,
      "step": 11480,
      "train_runtime": 70653.9339,
      "train_tokens_per_second": 8363.323
    },
    {
      "epoch": 0.7732965257204417,
      "grad_norm": 0.6302841316988732,
      "learning_rate": 6.743626818801042e-06,
      "loss": 0.38190221786499023,
      "num_input_tokens_seen": 591159552,
      "step": 11485,
      "train_runtime": 70683.7724,
      "train_tokens_per_second": 8363.441
    },
    {
      "epoch": 0.7736331807164019,
      "grad_norm": 0.6323524635187835,
      "learning_rate": 6.741148475331525e-06,
      "loss": 0.37079148292541503,
      "num_input_tokens_seen": 591416520,
      "step": 11490,
      "train_runtime": 70718.0254,
      "train_tokens_per_second": 8363.024
    },
    {
      "epoch": 0.773969835712362,
      "grad_norm": 0.6203522181269473,
      "learning_rate": 6.738669644955388e-06,
      "loss": 0.34959964752197265,
      "num_input_tokens_seen": 591666992,
      "step": 11495,
      "train_runtime": 70748.9619,
      "train_tokens_per_second": 8362.907
    },
    {
      "epoch": 0.7743064907083221,
      "grad_norm": 0.6996613658636461,
      "learning_rate": 6.736190328365828e-06,
      "loss": 0.35959064960479736,
      "num_input_tokens_seen": 591927664,
      "step": 11500,
      "train_runtime": 70780.8961,
      "train_tokens_per_second": 8362.817
    },
    {
      "epoch": 0.7746431457042823,
      "grad_norm": 0.6599695605894135,
      "learning_rate": 6.73371052625618e-06,
      "loss": 0.36336143016815187,
      "num_input_tokens_seen": 592186728,
      "step": 11505,
      "train_runtime": 70809.7514,
      "train_tokens_per_second": 8363.067
    },
    {
      "epoch": 0.7749798007002424,
      "grad_norm": 0.6102568529791547,
      "learning_rate": 6.731230239319912e-06,
      "loss": 0.36689319610595705,
      "num_input_tokens_seen": 592443960,
      "step": 11510,
      "train_runtime": 70840.1373,
      "train_tokens_per_second": 8363.111
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.5305964644424537,
      "learning_rate": 6.728749468250627e-06,
      "loss": 0.38024487495422366,
      "num_input_tokens_seen": 592730472,
      "step": 11515,
      "train_runtime": 70870.717,
      "train_tokens_per_second": 8363.546
    },
    {
      "epoch": 0.7756531106921627,
      "grad_norm": 0.8370532842653917,
      "learning_rate": 6.726268213742066e-06,
      "loss": 0.3682197093963623,
      "num_input_tokens_seen": 592963896,
      "step": 11520,
      "train_runtime": 70899.3501,
      "train_tokens_per_second": 8363.46
    },
    {
      "epoch": 0.7759897656881228,
      "grad_norm": 0.6507831458221607,
      "learning_rate": 6.723786476488105e-06,
      "loss": 0.3702432870864868,
      "num_input_tokens_seen": 593237288,
      "step": 11525,
      "train_runtime": 70929.6786,
      "train_tokens_per_second": 8363.739
    },
    {
      "epoch": 0.776326420684083,
      "grad_norm": 0.6598361487117648,
      "learning_rate": 6.72130425718275e-06,
      "loss": 0.3565581560134888,
      "num_input_tokens_seen": 593486144,
      "step": 11530,
      "train_runtime": 70963.7968,
      "train_tokens_per_second": 8363.224
    },
    {
      "epoch": 0.7766630756800431,
      "grad_norm": 0.6816579029418405,
      "learning_rate": 6.718821556520151e-06,
      "loss": 0.37020533084869384,
      "num_input_tokens_seen": 593740768,
      "step": 11535,
      "train_runtime": 70995.6477,
      "train_tokens_per_second": 8363.059
    },
    {
      "epoch": 0.7769997306760033,
      "grad_norm": 0.6344726464837355,
      "learning_rate": 6.716338375194583e-06,
      "loss": 0.3551889419555664,
      "num_input_tokens_seen": 594006272,
      "step": 11540,
      "train_runtime": 71030.184,
      "train_tokens_per_second": 8362.73
    },
    {
      "epoch": 0.7773363856719634,
      "grad_norm": 0.6262765029824597,
      "learning_rate": 6.713854713900463e-06,
      "loss": 0.38368451595306396,
      "num_input_tokens_seen": 594268016,
      "step": 11545,
      "train_runtime": 71065.439,
      "train_tokens_per_second": 8362.265
    },
    {
      "epoch": 0.7776730406679235,
      "grad_norm": 0.681353224135269,
      "learning_rate": 6.7113705733323365e-06,
      "loss": 0.38232948780059817,
      "num_input_tokens_seen": 594503176,
      "step": 11550,
      "train_runtime": 71094.7829,
      "train_tokens_per_second": 8362.121
    },
    {
      "epoch": 0.7780096956638837,
      "grad_norm": 0.6615310844870296,
      "learning_rate": 6.708885954184885e-06,
      "loss": 0.36517086029052737,
      "num_input_tokens_seen": 594764280,
      "step": 11555,
      "train_runtime": 71124.071,
      "train_tokens_per_second": 8362.349
    },
    {
      "epoch": 0.7783463506598438,
      "grad_norm": 0.6751177161989232,
      "learning_rate": 6.7064008571529275e-06,
      "loss": 0.343701171875,
      "num_input_tokens_seen": 595027728,
      "step": 11560,
      "train_runtime": 71154.687,
      "train_tokens_per_second": 8362.453
    },
    {
      "epoch": 0.778683005655804,
      "grad_norm": 0.5639596664122049,
      "learning_rate": 6.703915282931411e-06,
      "loss": 0.36808929443359373,
      "num_input_tokens_seen": 595287952,
      "step": 11565,
      "train_runtime": 71182.1116,
      "train_tokens_per_second": 8362.887
    },
    {
      "epoch": 0.7790196606517641,
      "grad_norm": 0.6722439873697635,
      "learning_rate": 6.701429232215418e-06,
      "loss": 0.37281434535980223,
      "num_input_tokens_seen": 595514808,
      "step": 11570,
      "train_runtime": 71211.3994,
      "train_tokens_per_second": 8362.633
    },
    {
      "epoch": 0.7793563156477242,
      "grad_norm": 0.6861171847804377,
      "learning_rate": 6.6989427057001665e-06,
      "loss": 0.37390387058258057,
      "num_input_tokens_seen": 595777880,
      "step": 11575,
      "train_runtime": 71240.227,
      "train_tokens_per_second": 8362.942
    },
    {
      "epoch": 0.7796929706436844,
      "grad_norm": 0.6261688297308431,
      "learning_rate": 6.696455704081006e-06,
      "loss": 0.3776299238204956,
      "num_input_tokens_seen": 596039976,
      "step": 11580,
      "train_runtime": 71267.5618,
      "train_tokens_per_second": 8363.412
    },
    {
      "epoch": 0.7800296256396445,
      "grad_norm": 0.6315961538646053,
      "learning_rate": 6.693968228053416e-06,
      "loss": 0.38794381618499757,
      "num_input_tokens_seen": 596295720,
      "step": 11585,
      "train_runtime": 71300.5032,
      "train_tokens_per_second": 8363.135
    },
    {
      "epoch": 0.7803662806356046,
      "grad_norm": 0.6039899589680036,
      "learning_rate": 6.691480278313013e-06,
      "loss": 0.3584732532501221,
      "num_input_tokens_seen": 596544096,
      "step": 11590,
      "train_runtime": 71329.0949,
      "train_tokens_per_second": 8363.265
    },
    {
      "epoch": 0.7807029356315648,
      "grad_norm": 0.624290125200159,
      "learning_rate": 6.688991855555545e-06,
      "loss": 0.364134407043457,
      "num_input_tokens_seen": 596798280,
      "step": 11595,
      "train_runtime": 71355.6546,
      "train_tokens_per_second": 8363.714
    },
    {
      "epoch": 0.7810395906275249,
      "grad_norm": 0.5768656403138972,
      "learning_rate": 6.686502960476888e-06,
      "loss": 0.34170868396759035,
      "num_input_tokens_seen": 597082200,
      "step": 11600,
      "train_runtime": 71387.6006,
      "train_tokens_per_second": 8363.948
    },
    {
      "epoch": 0.7813762456234851,
      "grad_norm": 0.5889362509176054,
      "learning_rate": 6.684013593773058e-06,
      "loss": 0.32776219844818116,
      "num_input_tokens_seen": 597373256,
      "step": 11605,
      "train_runtime": 71421.1663,
      "train_tokens_per_second": 8364.093
    },
    {
      "epoch": 0.7817129006194452,
      "grad_norm": 0.6598890244070201,
      "learning_rate": 6.6815237561401955e-06,
      "loss": 0.32468628883361816,
      "num_input_tokens_seen": 597598000,
      "step": 11610,
      "train_runtime": 71449.0608,
      "train_tokens_per_second": 8363.973
    },
    {
      "epoch": 0.7820495556154053,
      "grad_norm": 0.6825915585933323,
      "learning_rate": 6.679033448274577e-06,
      "loss": 0.34545397758483887,
      "num_input_tokens_seen": 597845440,
      "step": 11615,
      "train_runtime": 71479.3406,
      "train_tokens_per_second": 8363.891
    },
    {
      "epoch": 0.7823862106113655,
      "grad_norm": 0.5419676800116091,
      "learning_rate": 6.676542670872609e-06,
      "loss": 0.36843137741088866,
      "num_input_tokens_seen": 598091480,
      "step": 11620,
      "train_runtime": 71511.8633,
      "train_tokens_per_second": 8363.528
    },
    {
      "epoch": 0.7827228656073256,
      "grad_norm": 0.5506457012698742,
      "learning_rate": 6.674051424630829e-06,
      "loss": 0.35717794895172117,
      "num_input_tokens_seen": 598367128,
      "step": 11625,
      "train_runtime": 71545.7904,
      "train_tokens_per_second": 8363.415
    },
    {
      "epoch": 0.7830595206032858,
      "grad_norm": 0.594571599890044,
      "learning_rate": 6.671559710245905e-06,
      "loss": 0.3690334320068359,
      "num_input_tokens_seen": 598634584,
      "step": 11630,
      "train_runtime": 71577.111,
      "train_tokens_per_second": 8363.492
    },
    {
      "epoch": 0.7833961755992459,
      "grad_norm": 0.6305620963700591,
      "learning_rate": 6.669067528414639e-06,
      "loss": 0.35454535484313965,
      "num_input_tokens_seen": 598897352,
      "step": 11635,
      "train_runtime": 71608.0782,
      "train_tokens_per_second": 8363.545
    },
    {
      "epoch": 0.783732830595206,
      "grad_norm": 0.6431792047344889,
      "learning_rate": 6.6665748798339606e-06,
      "loss": 0.3796380043029785,
      "num_input_tokens_seen": 599172296,
      "step": 11640,
      "train_runtime": 71637.8217,
      "train_tokens_per_second": 8363.91
    },
    {
      "epoch": 0.7840694855911662,
      "grad_norm": 0.6975182791911516,
      "learning_rate": 6.664081765200932e-06,
      "loss": 0.380670166015625,
      "num_input_tokens_seen": 599405888,
      "step": 11645,
      "train_runtime": 71668.0386,
      "train_tokens_per_second": 8363.643
    },
    {
      "epoch": 0.7844061405871263,
      "grad_norm": 0.7819132288259559,
      "learning_rate": 6.661588185212745e-06,
      "loss": 0.3757169723510742,
      "num_input_tokens_seen": 599655096,
      "step": 11650,
      "train_runtime": 71702.7615,
      "train_tokens_per_second": 8363.068
    },
    {
      "epoch": 0.7847427955830865,
      "grad_norm": 0.6222587073275907,
      "learning_rate": 6.65909414056672e-06,
      "loss": 0.36081752777099607,
      "num_input_tokens_seen": 599900216,
      "step": 11655,
      "train_runtime": 71729.633,
      "train_tokens_per_second": 8363.353
    },
    {
      "epoch": 0.7850794505790466,
      "grad_norm": 0.7227036389887456,
      "learning_rate": 6.6565996319603075e-06,
      "loss": 0.36037287712097166,
      "num_input_tokens_seen": 600152424,
      "step": 11660,
      "train_runtime": 71756.9615,
      "train_tokens_per_second": 8363.682
    },
    {
      "epoch": 0.7854161055750067,
      "grad_norm": 0.6371382706238344,
      "learning_rate": 6.6541046600910915e-06,
      "loss": 0.35015716552734377,
      "num_input_tokens_seen": 600394680,
      "step": 11665,
      "train_runtime": 71788.059,
      "train_tokens_per_second": 8363.434
    },
    {
      "epoch": 0.7857527605709669,
      "grad_norm": 0.6251106869504647,
      "learning_rate": 6.6516092256567824e-06,
      "loss": 0.3528814554214478,
      "num_input_tokens_seen": 600670688,
      "step": 11670,
      "train_runtime": 71819.5595,
      "train_tokens_per_second": 8363.609
    },
    {
      "epoch": 0.786089415566927,
      "grad_norm": 0.7162829117287818,
      "learning_rate": 6.64911332935522e-06,
      "loss": 0.38129196166992185,
      "num_input_tokens_seen": 600935952,
      "step": 11675,
      "train_runtime": 71851.1064,
      "train_tokens_per_second": 8363.628
    },
    {
      "epoch": 0.7864260705628872,
      "grad_norm": 0.6371912870743663,
      "learning_rate": 6.646616971884373e-06,
      "loss": 0.35258212089538576,
      "num_input_tokens_seen": 601200208,
      "step": 11680,
      "train_runtime": 71881.3311,
      "train_tokens_per_second": 8363.788
    },
    {
      "epoch": 0.7867627255588473,
      "grad_norm": 0.579993616537064,
      "learning_rate": 6.64412015394234e-06,
      "loss": 0.3527322292327881,
      "num_input_tokens_seen": 601441032,
      "step": 11685,
      "train_runtime": 71906.703,
      "train_tokens_per_second": 8364.186
    },
    {
      "epoch": 0.7870993805548074,
      "grad_norm": 0.6773218707669468,
      "learning_rate": 6.641622876227349e-06,
      "loss": 0.36829674243927,
      "num_input_tokens_seen": 601681408,
      "step": 11690,
      "train_runtime": 71935.4065,
      "train_tokens_per_second": 8364.19
    },
    {
      "epoch": 0.7874360355507676,
      "grad_norm": 0.6907279381245569,
      "learning_rate": 6.639125139437756e-06,
      "loss": 0.3589077711105347,
      "num_input_tokens_seen": 601920192,
      "step": 11695,
      "train_runtime": 71967.1233,
      "train_tokens_per_second": 8363.822
    },
    {
      "epoch": 0.7877726905467277,
      "grad_norm": 0.7529604564498005,
      "learning_rate": 6.6366269442720446e-06,
      "loss": 0.3614177942276001,
      "num_input_tokens_seen": 602151400,
      "step": 11700,
      "train_runtime": 71996.0798,
      "train_tokens_per_second": 8363.669
    },
    {
      "epoch": 0.7881093455426879,
      "grad_norm": 0.5961343283735043,
      "learning_rate": 6.634128291428827e-06,
      "loss": 0.39605035781860354,
      "num_input_tokens_seen": 602398512,
      "step": 11705,
      "train_runtime": 72025.1468,
      "train_tokens_per_second": 8363.725
    },
    {
      "epoch": 0.788446000538648,
      "grad_norm": 0.7118039907674051,
      "learning_rate": 6.631629181606842e-06,
      "loss": 0.37867534160614014,
      "num_input_tokens_seen": 602652880,
      "step": 11710,
      "train_runtime": 72052.9437,
      "train_tokens_per_second": 8364.029
    },
    {
      "epoch": 0.7887826555346081,
      "grad_norm": 0.5716596481731494,
      "learning_rate": 6.629129615504961e-06,
      "loss": 0.3602000713348389,
      "num_input_tokens_seen": 602917120,
      "step": 11715,
      "train_runtime": 72083.6951,
      "train_tokens_per_second": 8364.126
    },
    {
      "epoch": 0.7891193105305683,
      "grad_norm": 0.6161033554551477,
      "learning_rate": 6.626629593822178e-06,
      "loss": 0.33930392265319825,
      "num_input_tokens_seen": 603172400,
      "step": 11720,
      "train_runtime": 72115.6217,
      "train_tokens_per_second": 8363.963
    },
    {
      "epoch": 0.7894559655265284,
      "grad_norm": 0.6335075719558505,
      "learning_rate": 6.624129117257616e-06,
      "loss": 0.37834715843200684,
      "num_input_tokens_seen": 603430984,
      "step": 11725,
      "train_runtime": 72148.06,
      "train_tokens_per_second": 8363.787
    },
    {
      "epoch": 0.7897926205224886,
      "grad_norm": 0.6561717568026328,
      "learning_rate": 6.621628186510524e-06,
      "loss": 0.3702746629714966,
      "num_input_tokens_seen": 603685312,
      "step": 11730,
      "train_runtime": 72175.9179,
      "train_tokens_per_second": 8364.082
    },
    {
      "epoch": 0.7901292755184487,
      "grad_norm": 0.6242478754577795,
      "learning_rate": 6.619126802280281e-06,
      "loss": 0.3457314491271973,
      "num_input_tokens_seen": 603924888,
      "step": 11735,
      "train_runtime": 72207.0679,
      "train_tokens_per_second": 8363.792
    },
    {
      "epoch": 0.7904659305144088,
      "grad_norm": 0.6197779615766892,
      "learning_rate": 6.616624965266392e-06,
      "loss": 0.3620455741882324,
      "num_input_tokens_seen": 604186968,
      "step": 11740,
      "train_runtime": 72234.8445,
      "train_tokens_per_second": 8364.204
    },
    {
      "epoch": 0.790802585510369,
      "grad_norm": 0.6154780660237097,
      "learning_rate": 6.614122676168486e-06,
      "loss": 0.3843680381774902,
      "num_input_tokens_seen": 604435184,
      "step": 11745,
      "train_runtime": 72267.0449,
      "train_tokens_per_second": 8363.912
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.7843798372726337,
      "learning_rate": 6.611619935686321e-06,
      "loss": 0.3424374103546143,
      "num_input_tokens_seen": 604669056,
      "step": 11750,
      "train_runtime": 72300.0494,
      "train_tokens_per_second": 8363.328
    },
    {
      "epoch": 0.7914758955022893,
      "grad_norm": 0.6383322759232312,
      "learning_rate": 6.609116744519779e-06,
      "loss": 0.38023412227630615,
      "num_input_tokens_seen": 604936192,
      "step": 11755,
      "train_runtime": 72331.1439,
      "train_tokens_per_second": 8363.426
    },
    {
      "epoch": 0.7918125504982494,
      "grad_norm": 0.7541199673136507,
      "learning_rate": 6.60661310336887e-06,
      "loss": 0.335788631439209,
      "num_input_tokens_seen": 605168128,
      "step": 11760,
      "train_runtime": 72365.2924,
      "train_tokens_per_second": 8362.685
    },
    {
      "epoch": 0.7921492054942095,
      "grad_norm": 0.6660112419120152,
      "learning_rate": 6.604109012933732e-06,
      "loss": 0.3695520877838135,
      "num_input_tokens_seen": 605404896,
      "step": 11765,
      "train_runtime": 72394.6166,
      "train_tokens_per_second": 8362.568
    },
    {
      "epoch": 0.7924858604901697,
      "grad_norm": 0.926787392214373,
      "learning_rate": 6.601604473914622e-06,
      "loss": 0.3874047756195068,
      "num_input_tokens_seen": 605660984,
      "step": 11770,
      "train_runtime": 72426.1185,
      "train_tokens_per_second": 8362.466
    },
    {
      "epoch": 0.7928225154861298,
      "grad_norm": 0.5920309784521345,
      "learning_rate": 6.599099487011928e-06,
      "loss": 0.36225459575653074,
      "num_input_tokens_seen": 605928264,
      "step": 11775,
      "train_runtime": 72456.6493,
      "train_tokens_per_second": 8362.632
    },
    {
      "epoch": 0.79315917048209,
      "grad_norm": 0.6593155672035383,
      "learning_rate": 6.596594052926163e-06,
      "loss": 0.3749274253845215,
      "num_input_tokens_seen": 606199656,
      "step": 11780,
      "train_runtime": 72488.2911,
      "train_tokens_per_second": 8362.725
    },
    {
      "epoch": 0.7934958254780501,
      "grad_norm": 0.6402963904440483,
      "learning_rate": 6.5940881723579595e-06,
      "loss": 0.3697853088378906,
      "num_input_tokens_seen": 606455904,
      "step": 11785,
      "train_runtime": 72520.2558,
      "train_tokens_per_second": 8362.573
    },
    {
      "epoch": 0.7938324804740102,
      "grad_norm": 0.6533907377583371,
      "learning_rate": 6.591581846008085e-06,
      "loss": 0.34089322090148927,
      "num_input_tokens_seen": 606715432,
      "step": 11790,
      "train_runtime": 72554.3107,
      "train_tokens_per_second": 8362.224
    },
    {
      "epoch": 0.7941691354699704,
      "grad_norm": 0.6544487125063826,
      "learning_rate": 6.589075074577421e-06,
      "loss": 0.3590712070465088,
      "num_input_tokens_seen": 606968168,
      "step": 11795,
      "train_runtime": 72588.0028,
      "train_tokens_per_second": 8361.825
    },
    {
      "epoch": 0.7945057904659305,
      "grad_norm": 0.6216895750478085,
      "learning_rate": 6.5865678587669805e-06,
      "loss": 0.34460859298706054,
      "num_input_tokens_seen": 607243440,
      "step": 11800,
      "train_runtime": 72622.4895,
      "train_tokens_per_second": 8361.645
    },
    {
      "epoch": 0.7948424454618906,
      "grad_norm": 0.794012701227758,
      "learning_rate": 6.584060199277897e-06,
      "loss": 0.37389097213745115,
      "num_input_tokens_seen": 607497680,
      "step": 11805,
      "train_runtime": 72649.997,
      "train_tokens_per_second": 8361.978
    },
    {
      "epoch": 0.7951791004578508,
      "grad_norm": 0.6466405634366306,
      "learning_rate": 6.581552096811428e-06,
      "loss": 0.36005773544311526,
      "num_input_tokens_seen": 607747928,
      "step": 11810,
      "train_runtime": 72677.4484,
      "train_tokens_per_second": 8362.263
    },
    {
      "epoch": 0.7955157554538109,
      "grad_norm": 0.6561067701702801,
      "learning_rate": 6.57904355206896e-06,
      "loss": 0.36092994213104246,
      "num_input_tokens_seen": 607992136,
      "step": 11815,
      "train_runtime": 72706.535,
      "train_tokens_per_second": 8362.276
    },
    {
      "epoch": 0.7958524104497711,
      "grad_norm": 0.6345037354937736,
      "learning_rate": 6.576534565751997e-06,
      "loss": 0.35169095993041993,
      "num_input_tokens_seen": 608247872,
      "step": 11820,
      "train_runtime": 72740.5426,
      "train_tokens_per_second": 8361.883
    },
    {
      "epoch": 0.7961890654457312,
      "grad_norm": 0.6313629353081792,
      "learning_rate": 6.574025138562172e-06,
      "loss": 0.37473814487457274,
      "num_input_tokens_seen": 608484248,
      "step": 11825,
      "train_runtime": 72767.8856,
      "train_tokens_per_second": 8361.989
    },
    {
      "epoch": 0.7965257204416913,
      "grad_norm": 0.5379506937748278,
      "learning_rate": 6.571515271201234e-06,
      "loss": 0.33207390308380125,
      "num_input_tokens_seen": 608751176,
      "step": 11830,
      "train_runtime": 72798.8616,
      "train_tokens_per_second": 8362.097
    },
    {
      "epoch": 0.7968623754376515,
      "grad_norm": 0.5398611124724312,
      "learning_rate": 6.569004964371062e-06,
      "loss": 0.37789390087127683,
      "num_input_tokens_seen": 609025472,
      "step": 11835,
      "train_runtime": 72835.0765,
      "train_tokens_per_second": 8361.706
    },
    {
      "epoch": 0.7971990304336116,
      "grad_norm": 0.5903709866812008,
      "learning_rate": 6.566494218773656e-06,
      "loss": 0.3555665254592896,
      "num_input_tokens_seen": 609290440,
      "step": 11840,
      "train_runtime": 72866.3472,
      "train_tokens_per_second": 8361.754
    },
    {
      "epoch": 0.7975356854295718,
      "grad_norm": 0.6659453737879756,
      "learning_rate": 6.563983035111136e-06,
      "loss": 0.3606618642807007,
      "num_input_tokens_seen": 609562248,
      "step": 11845,
      "train_runtime": 72899.0693,
      "train_tokens_per_second": 8361.729
    },
    {
      "epoch": 0.7978723404255319,
      "grad_norm": 0.6286354341005861,
      "learning_rate": 6.561471414085748e-06,
      "loss": 0.3657133340835571,
      "num_input_tokens_seen": 609813792,
      "step": 11850,
      "train_runtime": 72929.2042,
      "train_tokens_per_second": 8361.723
    },
    {
      "epoch": 0.798208995421492,
      "grad_norm": 0.6219047786534081,
      "learning_rate": 6.558959356399859e-06,
      "loss": 0.37337608337402345,
      "num_input_tokens_seen": 610061800,
      "step": 11855,
      "train_runtime": 72958.7348,
      "train_tokens_per_second": 8361.738
    },
    {
      "epoch": 0.7985456504174522,
      "grad_norm": 0.5818249091170927,
      "learning_rate": 6.5564468627559544e-06,
      "loss": 0.349498987197876,
      "num_input_tokens_seen": 610297896,
      "step": 11860,
      "train_runtime": 72985.4741,
      "train_tokens_per_second": 8361.909
    },
    {
      "epoch": 0.7988823054134123,
      "grad_norm": 0.6259807680549376,
      "learning_rate": 6.5539339338566495e-06,
      "loss": 0.383138632774353,
      "num_input_tokens_seen": 610545216,
      "step": 11865,
      "train_runtime": 73016.3571,
      "train_tokens_per_second": 8361.759
    },
    {
      "epoch": 0.7992189604093725,
      "grad_norm": 0.8394887055736733,
      "learning_rate": 6.5514205704046754e-06,
      "loss": 0.3412919044494629,
      "num_input_tokens_seen": 610798880,
      "step": 11870,
      "train_runtime": 73046.3008,
      "train_tokens_per_second": 8361.804
    },
    {
      "epoch": 0.7995556154053326,
      "grad_norm": 0.627468881368645,
      "learning_rate": 6.548906773102886e-06,
      "loss": 0.32927989959716797,
      "num_input_tokens_seen": 611046288,
      "step": 11875,
      "train_runtime": 73075.8112,
      "train_tokens_per_second": 8361.813
    },
    {
      "epoch": 0.7998922704012927,
      "grad_norm": 0.5800446092481228,
      "learning_rate": 6.546392542654257e-06,
      "loss": 0.3554494619369507,
      "num_input_tokens_seen": 611307240,
      "step": 11880,
      "train_runtime": 73104.5066,
      "train_tokens_per_second": 8362.101
    },
    {
      "epoch": 0.8002289253972529,
      "grad_norm": 0.6321518744520843,
      "learning_rate": 6.543877879761882e-06,
      "loss": 0.38935065269470215,
      "num_input_tokens_seen": 611563400,
      "step": 11885,
      "train_runtime": 73140.7051,
      "train_tokens_per_second": 8361.464
    },
    {
      "epoch": 0.800565580393213,
      "grad_norm": 0.5567566057772835,
      "learning_rate": 6.5413627851289825e-06,
      "loss": 0.3465577125549316,
      "num_input_tokens_seen": 611832504,
      "step": 11890,
      "train_runtime": 73169.8869,
      "train_tokens_per_second": 8361.807
    },
    {
      "epoch": 0.8009022353891732,
      "grad_norm": 0.5927830993510849,
      "learning_rate": 6.538847259458896e-06,
      "loss": 0.3431912660598755,
      "num_input_tokens_seen": 612107648,
      "step": 11895,
      "train_runtime": 73201.9152,
      "train_tokens_per_second": 8361.908
    },
    {
      "epoch": 0.8012388903851333,
      "grad_norm": 0.6572683205004337,
      "learning_rate": 6.5363313034550805e-06,
      "loss": 0.337173318862915,
      "num_input_tokens_seen": 612347360,
      "step": 11900,
      "train_runtime": 73228.9268,
      "train_tokens_per_second": 8362.097
    },
    {
      "epoch": 0.8015755453810934,
      "grad_norm": 0.6128971843412115,
      "learning_rate": 6.533814917821114e-06,
      "loss": 0.35640878677368165,
      "num_input_tokens_seen": 612617696,
      "step": 11905,
      "train_runtime": 73256.5108,
      "train_tokens_per_second": 8362.638
    },
    {
      "epoch": 0.8019122003770536,
      "grad_norm": 0.5783362286051961,
      "learning_rate": 6.531298103260696e-06,
      "loss": 0.34691033363342283,
      "num_input_tokens_seen": 612859752,
      "step": 11910,
      "train_runtime": 73287.5566,
      "train_tokens_per_second": 8362.398
    },
    {
      "epoch": 0.8022488553730137,
      "grad_norm": 0.6026914020642675,
      "learning_rate": 6.528780860477648e-06,
      "loss": 0.3713871955871582,
      "num_input_tokens_seen": 613153432,
      "step": 11915,
      "train_runtime": 73316.3088,
      "train_tokens_per_second": 8363.125
    },
    {
      "epoch": 0.8025855103689739,
      "grad_norm": 0.5411242158030674,
      "learning_rate": 6.526263190175907e-06,
      "loss": 0.32230100631713865,
      "num_input_tokens_seen": 613436248,
      "step": 11920,
      "train_runtime": 73347.8991,
      "train_tokens_per_second": 8363.379
    },
    {
      "epoch": 0.802922165364934,
      "grad_norm": 0.5642148044656746,
      "learning_rate": 6.523745093059531e-06,
      "loss": 0.34615445137023926,
      "num_input_tokens_seen": 613699672,
      "step": 11925,
      "train_runtime": 73379.3171,
      "train_tokens_per_second": 8363.388
    },
    {
      "epoch": 0.8032588203608941,
      "grad_norm": 0.7284016066952003,
      "learning_rate": 6.521226569832699e-06,
      "loss": 0.36312649250030515,
      "num_input_tokens_seen": 613965600,
      "step": 11930,
      "train_runtime": 73410.9605,
      "train_tokens_per_second": 8363.405
    },
    {
      "epoch": 0.8035954753568543,
      "grad_norm": 0.6463551810092698,
      "learning_rate": 6.518707621199707e-06,
      "loss": 0.34615445137023926,
      "num_input_tokens_seen": 614213992,
      "step": 11935,
      "train_runtime": 73442.3567,
      "train_tokens_per_second": 8363.212
    },
    {
      "epoch": 0.8039321303528144,
      "grad_norm": 0.7655968084814359,
      "learning_rate": 6.516188247864972e-06,
      "loss": 0.36913399696350097,
      "num_input_tokens_seen": 614481696,
      "step": 11940,
      "train_runtime": 73476.279,
      "train_tokens_per_second": 8362.994
    },
    {
      "epoch": 0.8042687853487746,
      "grad_norm": 0.653100467292912,
      "learning_rate": 6.513668450533028e-06,
      "loss": 0.3527453660964966,
      "num_input_tokens_seen": 614753288,
      "step": 11945,
      "train_runtime": 73506.0159,
      "train_tokens_per_second": 8363.306
    },
    {
      "epoch": 0.8046054403447347,
      "grad_norm": 0.6880134265231777,
      "learning_rate": 6.511148229908531e-06,
      "loss": 0.36933221817016604,
      "num_input_tokens_seen": 615023208,
      "step": 11950,
      "train_runtime": 73536.7483,
      "train_tokens_per_second": 8363.481
    },
    {
      "epoch": 0.8049420953406948,
      "grad_norm": 0.781444540960237,
      "learning_rate": 6.508627586696248e-06,
      "loss": 0.3562591552734375,
      "num_input_tokens_seen": 615273488,
      "step": 11955,
      "train_runtime": 73567.0875,
      "train_tokens_per_second": 8363.434
    },
    {
      "epoch": 0.805278750336655,
      "grad_norm": 0.6255970948904388,
      "learning_rate": 6.5061065216010695e-06,
      "loss": 0.36382040977478025,
      "num_input_tokens_seen": 615527392,
      "step": 11960,
      "train_runtime": 73595.8371,
      "train_tokens_per_second": 8363.617
    },
    {
      "epoch": 0.8056154053326151,
      "grad_norm": 0.6102914588585403,
      "learning_rate": 6.5035850353280075e-06,
      "loss": 0.326744532585144,
      "num_input_tokens_seen": 615782744,
      "step": 11965,
      "train_runtime": 73625.1993,
      "train_tokens_per_second": 8363.75
    },
    {
      "epoch": 0.8059520603285752,
      "grad_norm": 0.703953065578744,
      "learning_rate": 6.501063128582183e-06,
      "loss": 0.3581666946411133,
      "num_input_tokens_seen": 616018016,
      "step": 11970,
      "train_runtime": 73651.3604,
      "train_tokens_per_second": 8363.973
    },
    {
      "epoch": 0.8062887153245354,
      "grad_norm": 0.7041356064350588,
      "learning_rate": 6.498540802068843e-06,
      "loss": 0.36394343376159666,
      "num_input_tokens_seen": 616280728,
      "step": 11975,
      "train_runtime": 73687.3907,
      "train_tokens_per_second": 8363.449
    },
    {
      "epoch": 0.8066253703204955,
      "grad_norm": 0.6705855653084377,
      "learning_rate": 6.4960180564933454e-06,
      "loss": 0.35794620513916015,
      "num_input_tokens_seen": 616530432,
      "step": 11980,
      "train_runtime": 73717.3344,
      "train_tokens_per_second": 8363.439
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.73414284510246,
      "learning_rate": 6.493494892561167e-06,
      "loss": 0.3539396286010742,
      "num_input_tokens_seen": 616782176,
      "step": 11985,
      "train_runtime": 73746.9813,
      "train_tokens_per_second": 8363.49
    },
    {
      "epoch": 0.8072986803124158,
      "grad_norm": 0.6965524955860983,
      "learning_rate": 6.490971310977907e-06,
      "loss": 0.3723776340484619,
      "num_input_tokens_seen": 617046752,
      "step": 11990,
      "train_runtime": 73779.7171,
      "train_tokens_per_second": 8363.366
    },
    {
      "epoch": 0.807635335308376,
      "grad_norm": 0.6665329937517741,
      "learning_rate": 6.488447312449273e-06,
      "loss": 0.37535507678985597,
      "num_input_tokens_seen": 617318088,
      "step": 11995,
      "train_runtime": 73812.5186,
      "train_tokens_per_second": 8363.325
    },
    {
      "epoch": 0.8079719903043361,
      "grad_norm": 0.644838708203876,
      "learning_rate": 6.485922897681095e-06,
      "loss": 0.37404546737670896,
      "num_input_tokens_seen": 617584656,
      "step": 12000,
      "train_runtime": 73840.4666,
      "train_tokens_per_second": 8363.77
    },
    {
      "epoch": 0.8083086453002962,
      "grad_norm": 0.5922247376222913,
      "learning_rate": 6.483398067379316e-06,
      "loss": 0.3915335893630981,
      "num_input_tokens_seen": 617874600,
      "step": 12005,
      "train_runtime": 73868.5225,
      "train_tokens_per_second": 8364.518
    },
    {
      "epoch": 0.8086453002962564,
      "grad_norm": 0.6804929558154248,
      "learning_rate": 6.480872822249998e-06,
      "loss": 0.3466752529144287,
      "num_input_tokens_seen": 618146808,
      "step": 12010,
      "train_runtime": 73903.2705,
      "train_tokens_per_second": 8364.269
    },
    {
      "epoch": 0.8089819552922165,
      "grad_norm": 0.7017532637597111,
      "learning_rate": 6.478347162999318e-06,
      "loss": 0.37797369956970217,
      "num_input_tokens_seen": 618383528,
      "step": 12015,
      "train_runtime": 73933.7464,
      "train_tokens_per_second": 8364.023
    },
    {
      "epoch": 0.8093186102881766,
      "grad_norm": 0.5909525753090459,
      "learning_rate": 6.475821090333568e-06,
      "loss": 0.3601799964904785,
      "num_input_tokens_seen": 618631448,
      "step": 12020,
      "train_runtime": 73961.8316,
      "train_tokens_per_second": 8364.199
    },
    {
      "epoch": 0.8096552652841368,
      "grad_norm": 0.7346729586754288,
      "learning_rate": 6.473294604959156e-06,
      "loss": 0.3805622339248657,
      "num_input_tokens_seen": 618882200,
      "step": 12025,
      "train_runtime": 73991.9814,
      "train_tokens_per_second": 8364.179
    },
    {
      "epoch": 0.8099919202800969,
      "grad_norm": 0.6282251500741876,
      "learning_rate": 6.470767707582605e-06,
      "loss": 0.3504250288009644,
      "num_input_tokens_seen": 619137744,
      "step": 12030,
      "train_runtime": 74024.6459,
      "train_tokens_per_second": 8363.941
    },
    {
      "epoch": 0.8103285752760571,
      "grad_norm": 0.7342277914780007,
      "learning_rate": 6.468240398910555e-06,
      "loss": 0.35168018341064455,
      "num_input_tokens_seen": 619357872,
      "step": 12035,
      "train_runtime": 74056.971,
      "train_tokens_per_second": 8363.262
    },
    {
      "epoch": 0.8106652302720172,
      "grad_norm": 0.6970962276789383,
      "learning_rate": 6.4657126796497595e-06,
      "loss": 0.3463627576828003,
      "num_input_tokens_seen": 619628488,
      "step": 12040,
      "train_runtime": 74091.9392,
      "train_tokens_per_second": 8362.968
    },
    {
      "epoch": 0.8110018852679773,
      "grad_norm": 0.7300516947524532,
      "learning_rate": 6.463184550507087e-06,
      "loss": 0.3676155567169189,
      "num_input_tokens_seen": 619868088,
      "step": 12045,
      "train_runtime": 74121.2185,
      "train_tokens_per_second": 8362.897
    },
    {
      "epoch": 0.8113385402639375,
      "grad_norm": 0.5648714312726304,
      "learning_rate": 6.46065601218952e-06,
      "loss": 0.32375755310058596,
      "num_input_tokens_seen": 620132552,
      "step": 12050,
      "train_runtime": 74147.0986,
      "train_tokens_per_second": 8363.544
    },
    {
      "epoch": 0.8116751952598976,
      "grad_norm": 0.5567142646046704,
      "learning_rate": 6.458127065404156e-06,
      "loss": 0.3801559925079346,
      "num_input_tokens_seen": 620387448,
      "step": 12055,
      "train_runtime": 74182.2107,
      "train_tokens_per_second": 8363.022
    },
    {
      "epoch": 0.8120118502558578,
      "grad_norm": 0.615323285680182,
      "learning_rate": 6.45559771085821e-06,
      "loss": 0.35763485431671144,
      "num_input_tokens_seen": 620660520,
      "step": 12060,
      "train_runtime": 74218.3862,
      "train_tokens_per_second": 8362.625
    },
    {
      "epoch": 0.8123485052518179,
      "grad_norm": 0.6813613817342248,
      "learning_rate": 6.453067949259005e-06,
      "loss": 0.34861345291137696,
      "num_input_tokens_seen": 620924560,
      "step": 12065,
      "train_runtime": 74250.9142,
      "train_tokens_per_second": 8362.517
    },
    {
      "epoch": 0.812685160247778,
      "grad_norm": 0.6096222599130287,
      "learning_rate": 6.450537781313982e-06,
      "loss": 0.34368329048156737,
      "num_input_tokens_seen": 621167168,
      "step": 12070,
      "train_runtime": 74279.5859,
      "train_tokens_per_second": 8362.556
    },
    {
      "epoch": 0.8130218152437382,
      "grad_norm": 0.6484658376199598,
      "learning_rate": 6.448007207730693e-06,
      "loss": 0.3604177951812744,
      "num_input_tokens_seen": 621430560,
      "step": 12075,
      "train_runtime": 74308.1268,
      "train_tokens_per_second": 8362.888
    },
    {
      "epoch": 0.8133584702396983,
      "grad_norm": 0.6357033863512115,
      "learning_rate": 6.445476229216805e-06,
      "loss": 0.3513907194137573,
      "num_input_tokens_seen": 621717288,
      "step": 12080,
      "train_runtime": 74342.7337,
      "train_tokens_per_second": 8362.852
    },
    {
      "epoch": 0.8136951252356585,
      "grad_norm": 0.7079997173746673,
      "learning_rate": 6.442944846480099e-06,
      "loss": 0.3805551528930664,
      "num_input_tokens_seen": 621958264,
      "step": 12085,
      "train_runtime": 74369.8839,
      "train_tokens_per_second": 8363.039
    },
    {
      "epoch": 0.8140317802316186,
      "grad_norm": 0.615306525780034,
      "learning_rate": 6.44041306022847e-06,
      "loss": 0.37595841884613035,
      "num_input_tokens_seen": 622236760,
      "step": 12090,
      "train_runtime": 74401.1219,
      "train_tokens_per_second": 8363.271
    },
    {
      "epoch": 0.8143684352275787,
      "grad_norm": 0.6768220528715624,
      "learning_rate": 6.43788087116992e-06,
      "loss": 0.38039298057556153,
      "num_input_tokens_seen": 622522456,
      "step": 12095,
      "train_runtime": 74429.1512,
      "train_tokens_per_second": 8363.96
    },
    {
      "epoch": 0.8147050902235389,
      "grad_norm": 0.7036456859801881,
      "learning_rate": 6.43534828001257e-06,
      "loss": 0.36355636119842527,
      "num_input_tokens_seen": 622807448,
      "step": 12100,
      "train_runtime": 74462.054,
      "train_tokens_per_second": 8364.092
    },
    {
      "epoch": 0.815041745219499,
      "grad_norm": 0.6971050999386554,
      "learning_rate": 6.43281528746465e-06,
      "loss": 0.3548264503479004,
      "num_input_tokens_seen": 623050432,
      "step": 12105,
      "train_runtime": 74495.8778,
      "train_tokens_per_second": 8363.556
    },
    {
      "epoch": 0.8153784002154592,
      "grad_norm": 0.5678278646594566,
      "learning_rate": 6.430281894234504e-06,
      "loss": 0.3692523241043091,
      "num_input_tokens_seen": 623306432,
      "step": 12110,
      "train_runtime": 74529.4749,
      "train_tokens_per_second": 8363.22
    },
    {
      "epoch": 0.8157150552114193,
      "grad_norm": 0.6110486020705574,
      "learning_rate": 6.427748101030587e-06,
      "loss": 0.3649147510528564,
      "num_input_tokens_seen": 623562984,
      "step": 12115,
      "train_runtime": 74559.8123,
      "train_tokens_per_second": 8363.258
    },
    {
      "epoch": 0.8160517102073794,
      "grad_norm": 0.5830379724779516,
      "learning_rate": 6.425213908561468e-06,
      "loss": 0.3635368824005127,
      "num_input_tokens_seen": 623826984,
      "step": 12120,
      "train_runtime": 74593.6888,
      "train_tokens_per_second": 8363.0
    },
    {
      "epoch": 0.8163883652033396,
      "grad_norm": 0.6647778723804982,
      "learning_rate": 6.422679317535823e-06,
      "loss": 0.36971220970153806,
      "num_input_tokens_seen": 624073888,
      "step": 12125,
      "train_runtime": 74626.0683,
      "train_tokens_per_second": 8362.679
    },
    {
      "epoch": 0.8167250201992997,
      "grad_norm": 0.6346888113488587,
      "learning_rate": 6.420144328662441e-06,
      "loss": 0.3967355966567993,
      "num_input_tokens_seen": 624347472,
      "step": 12130,
      "train_runtime": 74659.2544,
      "train_tokens_per_second": 8362.627
    },
    {
      "epoch": 0.8170616751952599,
      "grad_norm": 0.6025755310429041,
      "learning_rate": 6.417608942650229e-06,
      "loss": 0.37267906665802003,
      "num_input_tokens_seen": 624571936,
      "step": 12135,
      "train_runtime": 74691.8128,
      "train_tokens_per_second": 8361.987
    },
    {
      "epoch": 0.81739833019122,
      "grad_norm": 0.6318235048449532,
      "learning_rate": 6.415073160208196e-06,
      "loss": 0.3535665988922119,
      "num_input_tokens_seen": 624866584,
      "step": 12140,
      "train_runtime": 74725.0006,
      "train_tokens_per_second": 8362.216
    },
    {
      "epoch": 0.8177349851871801,
      "grad_norm": 0.8452228914489716,
      "learning_rate": 6.412536982045465e-06,
      "loss": 0.3732632637023926,
      "num_input_tokens_seen": 625090568,
      "step": 12145,
      "train_runtime": 74755.6162,
      "train_tokens_per_second": 8361.787
    },
    {
      "epoch": 0.8180716401831403,
      "grad_norm": 0.6608790990648166,
      "learning_rate": 6.410000408871273e-06,
      "loss": 0.38166027069091796,
      "num_input_tokens_seen": 625344320,
      "step": 12150,
      "train_runtime": 74787.2511,
      "train_tokens_per_second": 8361.643
    },
    {
      "epoch": 0.8184082951791004,
      "grad_norm": 0.6663859225736061,
      "learning_rate": 6.407463441394961e-06,
      "loss": 0.36986069679260253,
      "num_input_tokens_seen": 625616696,
      "step": 12155,
      "train_runtime": 74819.9873,
      "train_tokens_per_second": 8361.625
    },
    {
      "epoch": 0.8187449501750605,
      "grad_norm": 0.7629872932628746,
      "learning_rate": 6.4049260803259874e-06,
      "loss": 0.3785872459411621,
      "num_input_tokens_seen": 625877184,
      "step": 12160,
      "train_runtime": 74852.0468,
      "train_tokens_per_second": 8361.524
    },
    {
      "epoch": 0.8190816051710207,
      "grad_norm": 0.7298069521812959,
      "learning_rate": 6.402388326373914e-06,
      "loss": 0.3452877998352051,
      "num_input_tokens_seen": 626144680,
      "step": 12165,
      "train_runtime": 74878.2106,
      "train_tokens_per_second": 8362.175
    },
    {
      "epoch": 0.8194182601669808,
      "grad_norm": 0.6223797092857295,
      "learning_rate": 6.399850180248418e-06,
      "loss": 0.39779784679412844,
      "num_input_tokens_seen": 626427080,
      "step": 12170,
      "train_runtime": 74909.6334,
      "train_tokens_per_second": 8362.437
    },
    {
      "epoch": 0.819754915162941,
      "grad_norm": 0.6587844908132269,
      "learning_rate": 6.397311642659283e-06,
      "loss": 0.36663734912872314,
      "num_input_tokens_seen": 626680048,
      "step": 12175,
      "train_runtime": 74940.7454,
      "train_tokens_per_second": 8362.341
    },
    {
      "epoch": 0.8200915701589011,
      "grad_norm": 0.7494877788379269,
      "learning_rate": 6.394772714316402e-06,
      "loss": 0.3785989284515381,
      "num_input_tokens_seen": 626965512,
      "step": 12180,
      "train_runtime": 74969.5336,
      "train_tokens_per_second": 8362.937
    },
    {
      "epoch": 0.8204282251548612,
      "grad_norm": 0.6008951626584362,
      "learning_rate": 6.392233395929779e-06,
      "loss": 0.3390796184539795,
      "num_input_tokens_seen": 627216472,
      "step": 12185,
      "train_runtime": 74999.5074,
      "train_tokens_per_second": 8362.941
    },
    {
      "epoch": 0.8207648801508214,
      "grad_norm": 0.553464402885762,
      "learning_rate": 6.389693688209526e-06,
      "loss": 0.35080652236938475,
      "num_input_tokens_seen": 627465768,
      "step": 12190,
      "train_runtime": 75031.5996,
      "train_tokens_per_second": 8362.687
    },
    {
      "epoch": 0.8211015351467815,
      "grad_norm": 0.6041295400668112,
      "learning_rate": 6.387153591865867e-06,
      "loss": 0.35462350845336915,
      "num_input_tokens_seen": 627745528,
      "step": 12195,
      "train_runtime": 75062.5878,
      "train_tokens_per_second": 8362.961
    },
    {
      "epoch": 0.8214381901427417,
      "grad_norm": 0.7044242203002145,
      "learning_rate": 6.384613107609127e-06,
      "loss": 0.3859482526779175,
      "num_input_tokens_seen": 627998504,
      "step": 12200,
      "train_runtime": 75094.2007,
      "train_tokens_per_second": 8362.81
    },
    {
      "epoch": 0.8217748451387018,
      "grad_norm": 0.5173919257260599,
      "learning_rate": 6.382072236149747e-06,
      "loss": 0.35764479637145996,
      "num_input_tokens_seen": 628276624,
      "step": 12205,
      "train_runtime": 75123.6442,
      "train_tokens_per_second": 8363.234
    },
    {
      "epoch": 0.8221115001346619,
      "grad_norm": 0.7782577911739671,
      "learning_rate": 6.379530978198273e-06,
      "loss": 0.33992533683776854,
      "num_input_tokens_seen": 628517016,
      "step": 12210,
      "train_runtime": 75151.7864,
      "train_tokens_per_second": 8363.301
    },
    {
      "epoch": 0.8224481551306222,
      "grad_norm": 0.5800530698545718,
      "learning_rate": 6.376989334465361e-06,
      "loss": 0.3338246583938599,
      "num_input_tokens_seen": 628785376,
      "step": 12215,
      "train_runtime": 75183.9299,
      "train_tokens_per_second": 8363.295
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.5741237168052419,
      "learning_rate": 6.374447305661772e-06,
      "loss": 0.3520357131958008,
      "num_input_tokens_seen": 629067784,
      "step": 12220,
      "train_runtime": 75214.711,
      "train_tokens_per_second": 8363.627
    },
    {
      "epoch": 0.8231214651225425,
      "grad_norm": 0.6467750734389769,
      "learning_rate": 6.3719048924983776e-06,
      "loss": 0.3513361930847168,
      "num_input_tokens_seen": 629305016,
      "step": 12225,
      "train_runtime": 75245.2355,
      "train_tokens_per_second": 8363.387
    },
    {
      "epoch": 0.8234581201185026,
      "grad_norm": 0.6356934445258292,
      "learning_rate": 6.369362095686152e-06,
      "loss": 0.370013165473938,
      "num_input_tokens_seen": 629568976,
      "step": 12230,
      "train_runtime": 75274.5612,
      "train_tokens_per_second": 8363.635
    },
    {
      "epoch": 0.8237947751144628,
      "grad_norm": 0.6204560155137953,
      "learning_rate": 6.366818915936185e-06,
      "loss": 0.3420825481414795,
      "num_input_tokens_seen": 629828296,
      "step": 12235,
      "train_runtime": 75308.4632,
      "train_tokens_per_second": 8363.314
    },
    {
      "epoch": 0.8241314301104229,
      "grad_norm": 0.6478063131058364,
      "learning_rate": 6.364275353959667e-06,
      "loss": 0.3677258014678955,
      "num_input_tokens_seen": 630071272,
      "step": 12240,
      "train_runtime": 75337.1285,
      "train_tokens_per_second": 8363.357
    },
    {
      "epoch": 0.824468085106383,
      "grad_norm": 0.6397373180697052,
      "learning_rate": 6.3617314104678966e-06,
      "loss": 0.35830488204956057,
      "num_input_tokens_seen": 630338168,
      "step": 12245,
      "train_runtime": 75368.9058,
      "train_tokens_per_second": 8363.372
    },
    {
      "epoch": 0.8248047401023432,
      "grad_norm": 0.6957829299800613,
      "learning_rate": 6.359187086172278e-06,
      "loss": 0.379609489440918,
      "num_input_tokens_seen": 630569760,
      "step": 12250,
      "train_runtime": 75398.7848,
      "train_tokens_per_second": 8363.129
    },
    {
      "epoch": 0.8251413950983033,
      "grad_norm": 0.6336149932282042,
      "learning_rate": 6.356642381784326e-06,
      "loss": 0.37221903800964357,
      "num_input_tokens_seen": 630822752,
      "step": 12255,
      "train_runtime": 75428.2739,
      "train_tokens_per_second": 8363.213
    },
    {
      "epoch": 0.8254780500942634,
      "grad_norm": 0.8276086382349395,
      "learning_rate": 6.354097298015658e-06,
      "loss": 0.3461689710617065,
      "num_input_tokens_seen": 631040672,
      "step": 12260,
      "train_runtime": 75455.3956,
      "train_tokens_per_second": 8363.095
    },
    {
      "epoch": 0.8258147050902236,
      "grad_norm": 0.6202529196152856,
      "learning_rate": 6.3515518355779985e-06,
      "loss": 0.3908064365386963,
      "num_input_tokens_seen": 631274200,
      "step": 12265,
      "train_runtime": 75487.5161,
      "train_tokens_per_second": 8362.63
    },
    {
      "epoch": 0.8261513600861837,
      "grad_norm": 0.5183883824614285,
      "learning_rate": 6.349005995183177e-06,
      "loss": 0.3363948822021484,
      "num_input_tokens_seen": 631552656,
      "step": 12270,
      "train_runtime": 75519.0764,
      "train_tokens_per_second": 8362.823
    },
    {
      "epoch": 0.8264880150821439,
      "grad_norm": 0.6044695824746603,
      "learning_rate": 6.346459777543131e-06,
      "loss": 0.3592496395111084,
      "num_input_tokens_seen": 631797984,
      "step": 12275,
      "train_runtime": 75552.3675,
      "train_tokens_per_second": 8362.385
    },
    {
      "epoch": 0.826824670078104,
      "grad_norm": 0.6327961463253701,
      "learning_rate": 6.343913183369902e-06,
      "loss": 0.3626004695892334,
      "num_input_tokens_seen": 632029832,
      "step": 12280,
      "train_runtime": 75580.5473,
      "train_tokens_per_second": 8362.335
    },
    {
      "epoch": 0.8271613250740641,
      "grad_norm": 0.6173283531358659,
      "learning_rate": 6.341366213375637e-06,
      "loss": 0.34705319404602053,
      "num_input_tokens_seen": 632303536,
      "step": 12285,
      "train_runtime": 75611.0269,
      "train_tokens_per_second": 8362.584
    },
    {
      "epoch": 0.8274979800700243,
      "grad_norm": 0.6221716594804573,
      "learning_rate": 6.338818868272587e-06,
      "loss": 0.3721654415130615,
      "num_input_tokens_seen": 632544624,
      "step": 12290,
      "train_runtime": 75642.6986,
      "train_tokens_per_second": 8362.269
    },
    {
      "epoch": 0.8278346350659844,
      "grad_norm": 1.0181356014113025,
      "learning_rate": 6.336271148773111e-06,
      "loss": 0.3567339420318604,
      "num_input_tokens_seen": 632829264,
      "step": 12295,
      "train_runtime": 75673.8235,
      "train_tokens_per_second": 8362.591
    },
    {
      "epoch": 0.8281712900619446,
      "grad_norm": 0.5733310727420761,
      "learning_rate": 6.333723055589669e-06,
      "loss": 0.39951472282409667,
      "num_input_tokens_seen": 633087976,
      "step": 12300,
      "train_runtime": 75706.4965,
      "train_tokens_per_second": 8362.4
    },
    {
      "epoch": 0.8285079450579047,
      "grad_norm": 0.654891258994734,
      "learning_rate": 6.331174589434826e-06,
      "loss": 0.3736900329589844,
      "num_input_tokens_seen": 633342256,
      "step": 12305,
      "train_runtime": 75740.629,
      "train_tokens_per_second": 8361.988
    },
    {
      "epoch": 0.8288446000538648,
      "grad_norm": 0.6721667907150285,
      "learning_rate": 6.328625751021254e-06,
      "loss": 0.3438502311706543,
      "num_input_tokens_seen": 633571608,
      "step": 12310,
      "train_runtime": 75770.5865,
      "train_tokens_per_second": 8361.709
    },
    {
      "epoch": 0.829181255049825,
      "grad_norm": 0.6497612917590312,
      "learning_rate": 6.326076541061729e-06,
      "loss": 0.33492312431335447,
      "num_input_tokens_seen": 633821696,
      "step": 12315,
      "train_runtime": 75799.6763,
      "train_tokens_per_second": 8361.8
    },
    {
      "epoch": 0.8295179100457851,
      "grad_norm": 0.6494959902129418,
      "learning_rate": 6.323526960269127e-06,
      "loss": 0.3877936601638794,
      "num_input_tokens_seen": 634075944,
      "step": 12320,
      "train_runtime": 75835.2781,
      "train_tokens_per_second": 8361.227
    },
    {
      "epoch": 0.8298545650417453,
      "grad_norm": 0.6631926130000393,
      "learning_rate": 6.3209770093564315e-06,
      "loss": 0.35556161403656006,
      "num_input_tokens_seen": 634330168,
      "step": 12325,
      "train_runtime": 75866.6007,
      "train_tokens_per_second": 8361.125
    },
    {
      "epoch": 0.8301912200377054,
      "grad_norm": 0.7416387552982835,
      "learning_rate": 6.318426689036727e-06,
      "loss": 0.34862594604492186,
      "num_input_tokens_seen": 634570864,
      "step": 12330,
      "train_runtime": 75896.3396,
      "train_tokens_per_second": 8361.021
    },
    {
      "epoch": 0.8305278750336655,
      "grad_norm": 0.6588160689065086,
      "learning_rate": 6.315876000023202e-06,
      "loss": 0.38042070865631106,
      "num_input_tokens_seen": 634818568,
      "step": 12335,
      "train_runtime": 75930.5692,
      "train_tokens_per_second": 8360.514
    },
    {
      "epoch": 0.8308645300296257,
      "grad_norm": 0.6528677435312131,
      "learning_rate": 6.313324943029151e-06,
      "loss": 0.3695486783981323,
      "num_input_tokens_seen": 635056888,
      "step": 12340,
      "train_runtime": 75958.4375,
      "train_tokens_per_second": 8360.584
    },
    {
      "epoch": 0.8312011850255858,
      "grad_norm": 0.6041876484343911,
      "learning_rate": 6.310773518767967e-06,
      "loss": 0.3562894821166992,
      "num_input_tokens_seen": 635318080,
      "step": 12345,
      "train_runtime": 75990.2099,
      "train_tokens_per_second": 8360.525
    },
    {
      "epoch": 0.831537840021546,
      "grad_norm": 0.6326206162954919,
      "learning_rate": 6.30822172795315e-06,
      "loss": 0.35457358360290525,
      "num_input_tokens_seen": 635564264,
      "step": 12350,
      "train_runtime": 76018.101,
      "train_tokens_per_second": 8360.696
    },
    {
      "epoch": 0.8318744950175061,
      "grad_norm": 0.7474072205624553,
      "learning_rate": 6.3056695712982965e-06,
      "loss": 0.3938751697540283,
      "num_input_tokens_seen": 635845928,
      "step": 12355,
      "train_runtime": 76050.9321,
      "train_tokens_per_second": 8360.791
    },
    {
      "epoch": 0.8322111500134662,
      "grad_norm": 0.6132886853671778,
      "learning_rate": 6.303117049517111e-06,
      "loss": 0.3324575901031494,
      "num_input_tokens_seen": 636095744,
      "step": 12360,
      "train_runtime": 76081.2541,
      "train_tokens_per_second": 8360.742
    },
    {
      "epoch": 0.8325478050094264,
      "grad_norm": 0.6873731560909131,
      "learning_rate": 6.300564163323398e-06,
      "loss": 0.33866682052612307,
      "num_input_tokens_seen": 636349744,
      "step": 12365,
      "train_runtime": 76107.1501,
      "train_tokens_per_second": 8361.235
    },
    {
      "epoch": 0.8328844600053865,
      "grad_norm": 0.6547934815418962,
      "learning_rate": 6.298010913431065e-06,
      "loss": 0.345685338973999,
      "num_input_tokens_seen": 636605344,
      "step": 12370,
      "train_runtime": 76134.0425,
      "train_tokens_per_second": 8361.639
    },
    {
      "epoch": 0.8332211150013467,
      "grad_norm": 0.6283064238212986,
      "learning_rate": 6.295457300554119e-06,
      "loss": 0.34445626735687257,
      "num_input_tokens_seen": 636883856,
      "step": 12375,
      "train_runtime": 76165.5871,
      "train_tokens_per_second": 8361.832
    },
    {
      "epoch": 0.8335577699973068,
      "grad_norm": 0.7835082299841355,
      "learning_rate": 6.292903325406668e-06,
      "loss": 0.3782132625579834,
      "num_input_tokens_seen": 637109240,
      "step": 12380,
      "train_runtime": 76192.9254,
      "train_tokens_per_second": 8361.79
    },
    {
      "epoch": 0.8338944249932669,
      "grad_norm": 0.6886847209653963,
      "learning_rate": 6.2903489887029255e-06,
      "loss": 0.34002296924591063,
      "num_input_tokens_seen": 637368888,
      "step": 12385,
      "train_runtime": 76226.8125,
      "train_tokens_per_second": 8361.479
    },
    {
      "epoch": 0.8342310799892271,
      "grad_norm": 0.6000241298883137,
      "learning_rate": 6.287794291157204e-06,
      "loss": 0.36582255363464355,
      "num_input_tokens_seen": 637633896,
      "step": 12390,
      "train_runtime": 76258.0879,
      "train_tokens_per_second": 8361.525
    },
    {
      "epoch": 0.8345677349851872,
      "grad_norm": 0.7826177971188982,
      "learning_rate": 6.285239233483915e-06,
      "loss": 0.3497668027877808,
      "num_input_tokens_seen": 637901512,
      "step": 12395,
      "train_runtime": 76290.786,
      "train_tokens_per_second": 8361.449
    },
    {
      "epoch": 0.8349043899811474,
      "grad_norm": 0.5479271117918173,
      "learning_rate": 6.2826838163975724e-06,
      "loss": 0.3635211229324341,
      "num_input_tokens_seen": 638164784,
      "step": 12400,
      "train_runtime": 76318.1437,
      "train_tokens_per_second": 8361.901
    },
    {
      "epoch": 0.8352410449771075,
      "grad_norm": 0.6915759739641838,
      "learning_rate": 6.2801280406127905e-06,
      "loss": 0.3517460823059082,
      "num_input_tokens_seen": 638427552,
      "step": 12405,
      "train_runtime": 76353.7125,
      "train_tokens_per_second": 8361.447
    },
    {
      "epoch": 0.8355776999730676,
      "grad_norm": 0.6306252940917826,
      "learning_rate": 6.2775719068442845e-06,
      "loss": 0.35538997650146487,
      "num_input_tokens_seen": 638693080,
      "step": 12410,
      "train_runtime": 76385.1335,
      "train_tokens_per_second": 8361.484
    },
    {
      "epoch": 0.8359143549690278,
      "grad_norm": 0.8393975335922432,
      "learning_rate": 6.275015415806869e-06,
      "loss": 0.3388391971588135,
      "num_input_tokens_seen": 638923200,
      "step": 12415,
      "train_runtime": 76413.938,
      "train_tokens_per_second": 8361.344
    },
    {
      "epoch": 0.8362510099649879,
      "grad_norm": 0.6421124497850839,
      "learning_rate": 6.272458568215458e-06,
      "loss": 0.3739107608795166,
      "num_input_tokens_seen": 639165352,
      "step": 12420,
      "train_runtime": 76443.6507,
      "train_tokens_per_second": 8361.261
    },
    {
      "epoch": 0.836587664960948,
      "grad_norm": 0.6711576310230444,
      "learning_rate": 6.269901364785066e-06,
      "loss": 0.3682185173034668,
      "num_input_tokens_seen": 639438672,
      "step": 12425,
      "train_runtime": 76478.0475,
      "train_tokens_per_second": 8361.075
    },
    {
      "epoch": 0.8369243199569082,
      "grad_norm": 0.6116178394372694,
      "learning_rate": 6.2673438062308055e-06,
      "loss": 0.360455846786499,
      "num_input_tokens_seen": 639733768,
      "step": 12430,
      "train_runtime": 76508.2038,
      "train_tokens_per_second": 8361.636
    },
    {
      "epoch": 0.8372609749528683,
      "grad_norm": 0.6795262601047627,
      "learning_rate": 6.264785893267892e-06,
      "loss": 0.3431110382080078,
      "num_input_tokens_seen": 639980752,
      "step": 12435,
      "train_runtime": 76540.1247,
      "train_tokens_per_second": 8361.376
    },
    {
      "epoch": 0.8375976299488285,
      "grad_norm": 0.616051265767818,
      "learning_rate": 6.2622276266116366e-06,
      "loss": 0.3456950426101685,
      "num_input_tokens_seen": 640249872,
      "step": 12440,
      "train_runtime": 76574.801,
      "train_tokens_per_second": 8361.104
    },
    {
      "epoch": 0.8379342849447886,
      "grad_norm": 0.6512714810209196,
      "learning_rate": 6.25966900697745e-06,
      "loss": 0.37560200691223145,
      "num_input_tokens_seen": 640512936,
      "step": 12445,
      "train_runtime": 76604.984,
      "train_tokens_per_second": 8361.244
    },
    {
      "epoch": 0.8382709399407487,
      "grad_norm": 0.5875227507829216,
      "learning_rate": 6.257110035080843e-06,
      "loss": 0.3412102937698364,
      "num_input_tokens_seen": 640796744,
      "step": 12450,
      "train_runtime": 76635.1159,
      "train_tokens_per_second": 8361.659
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.5947960091372919,
      "learning_rate": 6.254550711637422e-06,
      "loss": 0.40421328544616697,
      "num_input_tokens_seen": 641043664,
      "step": 12455,
      "train_runtime": 76670.0675,
      "train_tokens_per_second": 8361.068
    },
    {
      "epoch": 0.838944249932669,
      "grad_norm": 0.6284958936460209,
      "learning_rate": 6.251991037362897e-06,
      "loss": 0.36992995738983153,
      "num_input_tokens_seen": 641289584,
      "step": 12460,
      "train_runtime": 76697.6243,
      "train_tokens_per_second": 8361.271
    },
    {
      "epoch": 0.8392809049286292,
      "grad_norm": 0.6964253095647055,
      "learning_rate": 6.249431012973068e-06,
      "loss": 0.36882972717285156,
      "num_input_tokens_seen": 641524808,
      "step": 12465,
      "train_runtime": 76726.5061,
      "train_tokens_per_second": 8361.189
    },
    {
      "epoch": 0.8396175599245893,
      "grad_norm": 0.6410032914590382,
      "learning_rate": 6.246870639183843e-06,
      "loss": 0.36358282566070554,
      "num_input_tokens_seen": 641778136,
      "step": 12470,
      "train_runtime": 76755.6524,
      "train_tokens_per_second": 8361.314
    },
    {
      "epoch": 0.8399542149205494,
      "grad_norm": 0.6318060380384801,
      "learning_rate": 6.2443099167112185e-06,
      "loss": 0.3874217987060547,
      "num_input_tokens_seen": 642045376,
      "step": 12475,
      "train_runtime": 76788.2894,
      "train_tokens_per_second": 8361.241
    },
    {
      "epoch": 0.8402908699165096,
      "grad_norm": 0.5609471744391404,
      "learning_rate": 6.241748846271293e-06,
      "loss": 0.3635607481002808,
      "num_input_tokens_seen": 642323688,
      "step": 12480,
      "train_runtime": 76821.7422,
      "train_tokens_per_second": 8361.223
    },
    {
      "epoch": 0.8406275249124697,
      "grad_norm": 0.7656166979973988,
      "learning_rate": 6.239187428580263e-06,
      "loss": 0.3794610023498535,
      "num_input_tokens_seen": 642563968,
      "step": 12485,
      "train_runtime": 76855.8711,
      "train_tokens_per_second": 8360.636
    },
    {
      "epoch": 0.8409641799084299,
      "grad_norm": 0.687934976639073,
      "learning_rate": 6.236625664354421e-06,
      "loss": 0.3719752073287964,
      "num_input_tokens_seen": 642836632,
      "step": 12490,
      "train_runtime": 76889.4173,
      "train_tokens_per_second": 8360.535
    },
    {
      "epoch": 0.84130083490439,
      "grad_norm": 0.780513985504746,
      "learning_rate": 6.2340635543101545e-06,
      "loss": 0.3576955795288086,
      "num_input_tokens_seen": 643060376,
      "step": 12495,
      "train_runtime": 76912.7611,
      "train_tokens_per_second": 8360.906
    },
    {
      "epoch": 0.8416374899003501,
      "grad_norm": 0.5989607873569913,
      "learning_rate": 6.23150109916395e-06,
      "loss": 0.341719913482666,
      "num_input_tokens_seen": 643330808,
      "step": 12500,
      "train_runtime": 76945.4847,
      "train_tokens_per_second": 8360.865
    },
    {
      "epoch": 0.8419741448963103,
      "grad_norm": 0.5614987272633878,
      "learning_rate": 6.22893829963239e-06,
      "loss": 0.3854731798171997,
      "num_input_tokens_seen": 643586392,
      "step": 12505,
      "train_runtime": 76973.3419,
      "train_tokens_per_second": 8361.16
    },
    {
      "epoch": 0.8423107998922704,
      "grad_norm": 0.6406348122931387,
      "learning_rate": 6.226375156432153e-06,
      "loss": 0.36084690093994143,
      "num_input_tokens_seen": 643871696,
      "step": 12510,
      "train_runtime": 77006.842,
      "train_tokens_per_second": 8361.227
    },
    {
      "epoch": 0.8426474548882306,
      "grad_norm": 0.6336082778953244,
      "learning_rate": 6.2238116702800145e-06,
      "loss": 0.35051660537719725,
      "num_input_tokens_seen": 644130360,
      "step": 12515,
      "train_runtime": 77036.0599,
      "train_tokens_per_second": 8361.414
    },
    {
      "epoch": 0.8429841098841907,
      "grad_norm": 0.6701161726262003,
      "learning_rate": 6.221247841892846e-06,
      "loss": 0.3795124292373657,
      "num_input_tokens_seen": 644366384,
      "step": 12520,
      "train_runtime": 77071.1222,
      "train_tokens_per_second": 8360.672
    },
    {
      "epoch": 0.8433207648801508,
      "grad_norm": 0.6041071572139411,
      "learning_rate": 6.218683671987611e-06,
      "loss": 0.3613196611404419,
      "num_input_tokens_seen": 644652160,
      "step": 12525,
      "train_runtime": 77105.0404,
      "train_tokens_per_second": 8360.701
    },
    {
      "epoch": 0.843657419876111,
      "grad_norm": 0.8145714788511211,
      "learning_rate": 6.216119161281373e-06,
      "loss": 0.37427239418029784,
      "num_input_tokens_seen": 644922096,
      "step": 12530,
      "train_runtime": 77136.9549,
      "train_tokens_per_second": 8360.741
    },
    {
      "epoch": 0.8439940748720711,
      "grad_norm": 0.6283399620042003,
      "learning_rate": 6.213554310491291e-06,
      "loss": 0.36890759468078616,
      "num_input_tokens_seen": 645160624,
      "step": 12535,
      "train_runtime": 77166.0465,
      "train_tokens_per_second": 8360.68
    },
    {
      "epoch": 0.8443307298680313,
      "grad_norm": 0.6271937472598833,
      "learning_rate": 6.210989120334616e-06,
      "loss": 0.3567379951477051,
      "num_input_tokens_seen": 645422624,
      "step": 12540,
      "train_runtime": 77200.2812,
      "train_tokens_per_second": 8360.366
    },
    {
      "epoch": 0.8446673848639914,
      "grad_norm": 0.6199987685138942,
      "learning_rate": 6.208423591528694e-06,
      "loss": 0.3777190685272217,
      "num_input_tokens_seen": 645665048,
      "step": 12545,
      "train_runtime": 77230.3883,
      "train_tokens_per_second": 8360.246
    },
    {
      "epoch": 0.8450040398599515,
      "grad_norm": 0.5817466589876813,
      "learning_rate": 6.205857724790969e-06,
      "loss": 0.35425164699554446,
      "num_input_tokens_seen": 645921232,
      "step": 12550,
      "train_runtime": 77263.6784,
      "train_tokens_per_second": 8359.96
    },
    {
      "epoch": 0.8453406948559117,
      "grad_norm": 0.7150366115550992,
      "learning_rate": 6.203291520838974e-06,
      "loss": 0.3935417175292969,
      "num_input_tokens_seen": 646206632,
      "step": 12555,
      "train_runtime": 77299.1626,
      "train_tokens_per_second": 8359.814
    },
    {
      "epoch": 0.8456773498518718,
      "grad_norm": 0.603607144980781,
      "learning_rate": 6.200724980390344e-06,
      "loss": 0.3399515151977539,
      "num_input_tokens_seen": 646431048,
      "step": 12560,
      "train_runtime": 77324.9465,
      "train_tokens_per_second": 8359.929
    },
    {
      "epoch": 0.846014004847832,
      "grad_norm": 0.6182743600469527,
      "learning_rate": 6.198158104162804e-06,
      "loss": 0.3219467639923096,
      "num_input_tokens_seen": 646682784,
      "step": 12565,
      "train_runtime": 77356.8679,
      "train_tokens_per_second": 8359.733
    },
    {
      "epoch": 0.8463506598437921,
      "grad_norm": 0.558893639201229,
      "learning_rate": 6.1955908928741685e-06,
      "loss": 0.36861453056335447,
      "num_input_tokens_seen": 646976424,
      "step": 12570,
      "train_runtime": 77383.4198,
      "train_tokens_per_second": 8360.659
    },
    {
      "epoch": 0.8466873148397522,
      "grad_norm": 0.6258044453996725,
      "learning_rate": 6.193023347242353e-06,
      "loss": 0.37322173118591306,
      "num_input_tokens_seen": 647252352,
      "step": 12575,
      "train_runtime": 77414.3916,
      "train_tokens_per_second": 8360.879
    },
    {
      "epoch": 0.8470239698357124,
      "grad_norm": 0.7181338979643858,
      "learning_rate": 6.190455467985361e-06,
      "loss": 0.3541293144226074,
      "num_input_tokens_seen": 647513144,
      "step": 12580,
      "train_runtime": 77442.6806,
      "train_tokens_per_second": 8361.192
    },
    {
      "epoch": 0.8473606248316725,
      "grad_norm": 0.8212488391910221,
      "learning_rate": 6.187887255821295e-06,
      "loss": 0.3890493392944336,
      "num_input_tokens_seen": 647774112,
      "step": 12585,
      "train_runtime": 77475.6824,
      "train_tokens_per_second": 8360.999
    },
    {
      "epoch": 0.8476972798276327,
      "grad_norm": 0.6585555768137688,
      "learning_rate": 6.185318711468345e-06,
      "loss": 0.3407646417617798,
      "num_input_tokens_seen": 648005944,
      "step": 12590,
      "train_runtime": 77507.9894,
      "train_tokens_per_second": 8360.505
    },
    {
      "epoch": 0.8480339348235928,
      "grad_norm": 0.6626460933197773,
      "learning_rate": 6.182749835644799e-06,
      "loss": 0.35547258853912356,
      "num_input_tokens_seen": 648259688,
      "step": 12595,
      "train_runtime": 77535.9552,
      "train_tokens_per_second": 8360.762
    },
    {
      "epoch": 0.8483705898195529,
      "grad_norm": 0.5792401135145839,
      "learning_rate": 6.180180629069031e-06,
      "loss": 0.3503227949142456,
      "num_input_tokens_seen": 648538040,
      "step": 12600,
      "train_runtime": 77565.6285,
      "train_tokens_per_second": 8361.152
    },
    {
      "epoch": 0.8487072448155131,
      "grad_norm": 1.1508711444858506,
      "learning_rate": 6.177611092459514e-06,
      "loss": 0.3341677665710449,
      "num_input_tokens_seen": 648790168,
      "step": 12605,
      "train_runtime": 77599.4993,
      "train_tokens_per_second": 8360.752
    },
    {
      "epoch": 0.8490438998114732,
      "grad_norm": 0.7194508827818901,
      "learning_rate": 6.175041226534809e-06,
      "loss": 0.3595294952392578,
      "num_input_tokens_seen": 649049176,
      "step": 12610,
      "train_runtime": 77634.2864,
      "train_tokens_per_second": 8360.342
    },
    {
      "epoch": 0.8493805548074334,
      "grad_norm": 0.5669170336253566,
      "learning_rate": 6.172471032013574e-06,
      "loss": 0.3258140802383423,
      "num_input_tokens_seen": 649309736,
      "step": 12615,
      "train_runtime": 77661.937,
      "train_tokens_per_second": 8360.72
    },
    {
      "epoch": 0.8497172098033935,
      "grad_norm": 0.6039734976396948,
      "learning_rate": 6.169900509614553e-06,
      "loss": 0.3576544761657715,
      "num_input_tokens_seen": 649573208,
      "step": 12620,
      "train_runtime": 77691.5545,
      "train_tokens_per_second": 8360.924
    },
    {
      "epoch": 0.8500538647993536,
      "grad_norm": 0.6972802461152998,
      "learning_rate": 6.167329660056585e-06,
      "loss": 0.3595376968383789,
      "num_input_tokens_seen": 649818688,
      "step": 12625,
      "train_runtime": 77724.9768,
      "train_tokens_per_second": 8360.487
    },
    {
      "epoch": 0.8503905197953138,
      "grad_norm": 0.5513324884680253,
      "learning_rate": 6.1647584840586e-06,
      "loss": 0.36918976306915285,
      "num_input_tokens_seen": 650082776,
      "step": 12630,
      "train_runtime": 77753.6782,
      "train_tokens_per_second": 8360.798
    },
    {
      "epoch": 0.8507271747912739,
      "grad_norm": 0.5889374166142537,
      "learning_rate": 6.1621869823396175e-06,
      "loss": 0.369230842590332,
      "num_input_tokens_seen": 650360488,
      "step": 12635,
      "train_runtime": 77787.8369,
      "train_tokens_per_second": 8360.696
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 0.6125412644478788,
      "learning_rate": 6.159615155618752e-06,
      "loss": 0.34952850341796876,
      "num_input_tokens_seen": 650630616,
      "step": 12640,
      "train_runtime": 77822.2731,
      "train_tokens_per_second": 8360.468
    },
    {
      "epoch": 0.8514004847831942,
      "grad_norm": 0.6740068368095228,
      "learning_rate": 6.157043004615207e-06,
      "loss": 0.3778586149215698,
      "num_input_tokens_seen": 650920256,
      "step": 12645,
      "train_runtime": 77854.1416,
      "train_tokens_per_second": 8360.766
    },
    {
      "epoch": 0.8517371397791543,
      "grad_norm": 0.7026411346236852,
      "learning_rate": 6.1544705300482755e-06,
      "loss": 0.3591524362564087,
      "num_input_tokens_seen": 651196304,
      "step": 12650,
      "train_runtime": 77886.1848,
      "train_tokens_per_second": 8360.87
    },
    {
      "epoch": 0.8520737947751145,
      "grad_norm": 0.5838174166122856,
      "learning_rate": 6.151897732637341e-06,
      "loss": 0.331775164604187,
      "num_input_tokens_seen": 651436048,
      "step": 12655,
      "train_runtime": 77914.8798,
      "train_tokens_per_second": 8360.868
    },
    {
      "epoch": 0.8524104497710746,
      "grad_norm": 0.7745537338453315,
      "learning_rate": 6.149324613101877e-06,
      "loss": 0.36592836380004884,
      "num_input_tokens_seen": 651695120,
      "step": 12660,
      "train_runtime": 77945.3758,
      "train_tokens_per_second": 8360.921
    },
    {
      "epoch": 0.8527471047670347,
      "grad_norm": 0.5502832174607998,
      "learning_rate": 6.146751172161453e-06,
      "loss": 0.3179174423217773,
      "num_input_tokens_seen": 651963408,
      "step": 12665,
      "train_runtime": 77978.7235,
      "train_tokens_per_second": 8360.786
    },
    {
      "epoch": 0.8530837597629949,
      "grad_norm": 0.6502473998184339,
      "learning_rate": 6.144177410535719e-06,
      "loss": 0.3575987100601196,
      "num_input_tokens_seen": 652232384,
      "step": 12670,
      "train_runtime": 78011.8986,
      "train_tokens_per_second": 8360.678
    },
    {
      "epoch": 0.853420414758955,
      "grad_norm": 0.6598655387058964,
      "learning_rate": 6.14160332894442e-06,
      "loss": 0.35810146331787107,
      "num_input_tokens_seen": 652483856,
      "step": 12675,
      "train_runtime": 78043.2231,
      "train_tokens_per_second": 8360.545
    },
    {
      "epoch": 0.8537570697549152,
      "grad_norm": 0.6715380890316313,
      "learning_rate": 6.139028928107391e-06,
      "loss": 0.35846521854400637,
      "num_input_tokens_seen": 652734344,
      "step": 12680,
      "train_runtime": 78079.2763,
      "train_tokens_per_second": 8359.892
    },
    {
      "epoch": 0.8540937247508753,
      "grad_norm": 0.6776927205947632,
      "learning_rate": 6.136454208744553e-06,
      "loss": 0.3340064287185669,
      "num_input_tokens_seen": 652988672,
      "step": 12685,
      "train_runtime": 78110.7221,
      "train_tokens_per_second": 8359.783
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.6055663028263393,
      "learning_rate": 6.133879171575921e-06,
      "loss": 0.36334891319274903,
      "num_input_tokens_seen": 653248072,
      "step": 12690,
      "train_runtime": 78139.0814,
      "train_tokens_per_second": 8360.068
    },
    {
      "epoch": 0.8547670347427956,
      "grad_norm": 0.636096889847177,
      "learning_rate": 6.131303817321595e-06,
      "loss": 0.3868595600128174,
      "num_input_tokens_seen": 653523608,
      "step": 12695,
      "train_runtime": 78168.1042,
      "train_tokens_per_second": 8360.489
    },
    {
      "epoch": 0.8551036897387557,
      "grad_norm": 0.6312143318095289,
      "learning_rate": 6.128728146701763e-06,
      "loss": 0.3899166345596313,
      "num_input_tokens_seen": 653813120,
      "step": 12700,
      "train_runtime": 78195.7281,
      "train_tokens_per_second": 8361.238
    },
    {
      "epoch": 0.8554403447347159,
      "grad_norm": 0.6034737496612217,
      "learning_rate": 6.126152160436705e-06,
      "loss": 0.3433213233947754,
      "num_input_tokens_seen": 654065880,
      "step": 12705,
      "train_runtime": 78225.3934,
      "train_tokens_per_second": 8361.299
    },
    {
      "epoch": 0.855776999730676,
      "grad_norm": 0.8220707988739069,
      "learning_rate": 6.123575859246784e-06,
      "loss": 0.37622890472412107,
      "num_input_tokens_seen": 654302336,
      "step": 12710,
      "train_runtime": 78255.0804,
      "train_tokens_per_second": 8361.148
    },
    {
      "epoch": 0.8561136547266361,
      "grad_norm": 0.6684712245730813,
      "learning_rate": 6.120999243852459e-06,
      "loss": 0.34521443843841554,
      "num_input_tokens_seen": 654546312,
      "step": 12715,
      "train_runtime": 78287.0512,
      "train_tokens_per_second": 8360.85
    },
    {
      "epoch": 0.8564503097225963,
      "grad_norm": 0.6861176815085666,
      "learning_rate": 6.118422314974269e-06,
      "loss": 0.37166242599487304,
      "num_input_tokens_seen": 654800120,
      "step": 12720,
      "train_runtime": 78318.063,
      "train_tokens_per_second": 8360.78
    },
    {
      "epoch": 0.8567869647185564,
      "grad_norm": 0.6537787093604126,
      "learning_rate": 6.115845073332845e-06,
      "loss": 0.3552509546279907,
      "num_input_tokens_seen": 655071952,
      "step": 12725,
      "train_runtime": 78348.5514,
      "train_tokens_per_second": 8360.996
    },
    {
      "epoch": 0.8571236197145166,
      "grad_norm": 0.7111333863861787,
      "learning_rate": 6.113267519648905e-06,
      "loss": 0.38918347358703614,
      "num_input_tokens_seen": 655358664,
      "step": 12730,
      "train_runtime": 78380.9702,
      "train_tokens_per_second": 8361.196
    },
    {
      "epoch": 0.8574602747104767,
      "grad_norm": 0.6346950934868341,
      "learning_rate": 6.110689654643253e-06,
      "loss": 0.36266469955444336,
      "num_input_tokens_seen": 655597656,
      "step": 12735,
      "train_runtime": 78409.1005,
      "train_tokens_per_second": 8361.244
    },
    {
      "epoch": 0.8577969297064368,
      "grad_norm": 0.642650615556372,
      "learning_rate": 6.1081114790367805e-06,
      "loss": 0.37669308185577394,
      "num_input_tokens_seen": 655843872,
      "step": 12740,
      "train_runtime": 78440.6986,
      "train_tokens_per_second": 8361.015
    },
    {
      "epoch": 0.858133584702397,
      "grad_norm": 0.7019113529065354,
      "learning_rate": 6.105532993550467e-06,
      "loss": 0.3631018877029419,
      "num_input_tokens_seen": 656106456,
      "step": 12745,
      "train_runtime": 78472.549,
      "train_tokens_per_second": 8360.968
    },
    {
      "epoch": 0.8584702396983571,
      "grad_norm": 0.5828494654947747,
      "learning_rate": 6.1029541989053765e-06,
      "loss": 0.35289137363433837,
      "num_input_tokens_seen": 656355136,
      "step": 12750,
      "train_runtime": 78503.0333,
      "train_tokens_per_second": 8360.889
    },
    {
      "epoch": 0.8588068946943173,
      "grad_norm": 0.5947666098210919,
      "learning_rate": 6.100375095822661e-06,
      "loss": 0.3587950229644775,
      "num_input_tokens_seen": 656611432,
      "step": 12755,
      "train_runtime": 78536.4244,
      "train_tokens_per_second": 8360.597
    },
    {
      "epoch": 0.8591435496902774,
      "grad_norm": 0.6810615424520319,
      "learning_rate": 6.097795685023558e-06,
      "loss": 0.36793832778930663,
      "num_input_tokens_seen": 656874320,
      "step": 12760,
      "train_runtime": 78568.8968,
      "train_tokens_per_second": 8360.488
    },
    {
      "epoch": 0.8594802046862375,
      "grad_norm": 0.6416840503294701,
      "learning_rate": 6.095215967229396e-06,
      "loss": 0.32996644973754885,
      "num_input_tokens_seen": 657117632,
      "step": 12765,
      "train_runtime": 78597.4466,
      "train_tokens_per_second": 8360.547
    },
    {
      "epoch": 0.8598168596821977,
      "grad_norm": 0.6428123587539508,
      "learning_rate": 6.092635943161578e-06,
      "loss": 0.3776623487472534,
      "num_input_tokens_seen": 657398544,
      "step": 12770,
      "train_runtime": 78626.5373,
      "train_tokens_per_second": 8361.026
    },
    {
      "epoch": 0.8601535146781578,
      "grad_norm": 0.6988722851576044,
      "learning_rate": 6.090055613541603e-06,
      "loss": 0.3423515558242798,
      "num_input_tokens_seen": 657663208,
      "step": 12775,
      "train_runtime": 78654.9333,
      "train_tokens_per_second": 8361.373
    },
    {
      "epoch": 0.860490169674118,
      "grad_norm": 0.673280275923216,
      "learning_rate": 6.087474979091052e-06,
      "loss": 0.3563959836959839,
      "num_input_tokens_seen": 657887064,
      "step": 12780,
      "train_runtime": 78686.1118,
      "train_tokens_per_second": 8360.904
    },
    {
      "epoch": 0.8608268246700781,
      "grad_norm": 0.6042496556512292,
      "learning_rate": 6.084894040531591e-06,
      "loss": 0.3455789804458618,
      "num_input_tokens_seen": 658157520,
      "step": 12785,
      "train_runtime": 78716.0953,
      "train_tokens_per_second": 8361.156
    },
    {
      "epoch": 0.8611634796660382,
      "grad_norm": 0.6424499492836285,
      "learning_rate": 6.0823127985849705e-06,
      "loss": 0.3525089263916016,
      "num_input_tokens_seen": 658391760,
      "step": 12790,
      "train_runtime": 78746.3735,
      "train_tokens_per_second": 8360.915
    },
    {
      "epoch": 0.8615001346619984,
      "grad_norm": 0.6878199597183813,
      "learning_rate": 6.079731253973028e-06,
      "loss": 0.35139970779418944,
      "num_input_tokens_seen": 658639992,
      "step": 12795,
      "train_runtime": 78779.7298,
      "train_tokens_per_second": 8360.526
    },
    {
      "epoch": 0.8618367896579585,
      "grad_norm": 0.6603818375952162,
      "learning_rate": 6.077149407417683e-06,
      "loss": 0.34958500862121583,
      "num_input_tokens_seen": 658886464,
      "step": 12800,
      "train_runtime": 78812.5414,
      "train_tokens_per_second": 8360.173
    },
    {
      "epoch": 0.8621734446539187,
      "grad_norm": 0.6560318418978957,
      "learning_rate": 6.07456725964094e-06,
      "loss": 0.38059568405151367,
      "num_input_tokens_seen": 659127192,
      "step": 12805,
      "train_runtime": 78841.0487,
      "train_tokens_per_second": 8360.203
    },
    {
      "epoch": 0.8625100996498788,
      "grad_norm": 0.6165060778558217,
      "learning_rate": 6.07198481136489e-06,
      "loss": 0.3488000392913818,
      "num_input_tokens_seen": 659370808,
      "step": 12810,
      "train_runtime": 78872.9039,
      "train_tokens_per_second": 8359.915
    },
    {
      "epoch": 0.8628467546458389,
      "grad_norm": 0.6690467260171085,
      "learning_rate": 6.069402063311706e-06,
      "loss": 0.3449772596359253,
      "num_input_tokens_seen": 659606696,
      "step": 12815,
      "train_runtime": 78902.3683,
      "train_tokens_per_second": 8359.783
    },
    {
      "epoch": 0.8631834096417991,
      "grad_norm": 0.6559098760806569,
      "learning_rate": 6.066819016203646e-06,
      "loss": 0.3570109367370605,
      "num_input_tokens_seen": 659856328,
      "step": 12820,
      "train_runtime": 78933.5443,
      "train_tokens_per_second": 8359.644
    },
    {
      "epoch": 0.8635200646377592,
      "grad_norm": 0.9119171182949001,
      "learning_rate": 6.064235670763048e-06,
      "loss": 0.36870219707489016,
      "num_input_tokens_seen": 660119408,
      "step": 12825,
      "train_runtime": 78964.0222,
      "train_tokens_per_second": 8359.749
    },
    {
      "epoch": 0.8638567196337193,
      "grad_norm": 0.6125582687146782,
      "learning_rate": 6.06165202771234e-06,
      "loss": 0.3674616813659668,
      "num_input_tokens_seen": 660379544,
      "step": 12830,
      "train_runtime": 78998.9914,
      "train_tokens_per_second": 8359.341
    },
    {
      "epoch": 0.8641933746296795,
      "grad_norm": 0.6752284037326365,
      "learning_rate": 6.059068087774026e-06,
      "loss": 0.3512548923492432,
      "num_input_tokens_seen": 660599808,
      "step": 12835,
      "train_runtime": 79025.3542,
      "train_tokens_per_second": 8359.34
    },
    {
      "epoch": 0.8645300296256396,
      "grad_norm": 0.5700016918891491,
      "learning_rate": 6.0564838516707005e-06,
      "loss": 0.3870248794555664,
      "num_input_tokens_seen": 660876752,
      "step": 12840,
      "train_runtime": 79058.2814,
      "train_tokens_per_second": 8359.361
    },
    {
      "epoch": 0.8648666846215998,
      "grad_norm": 0.5542608794481988,
      "learning_rate": 6.053899320125033e-06,
      "loss": 0.3414546012878418,
      "num_input_tokens_seen": 661139520,
      "step": 12845,
      "train_runtime": 79088.3946,
      "train_tokens_per_second": 8359.501
    },
    {
      "epoch": 0.8652033396175599,
      "grad_norm": 0.6261068314476096,
      "learning_rate": 6.051314493859782e-06,
      "loss": 0.35768375396728513,
      "num_input_tokens_seen": 661412872,
      "step": 12850,
      "train_runtime": 79119.6681,
      "train_tokens_per_second": 8359.652
    },
    {
      "epoch": 0.86553999461352,
      "grad_norm": 0.6734145010117373,
      "learning_rate": 6.048729373597786e-06,
      "loss": 0.3505547285079956,
      "num_input_tokens_seen": 661669656,
      "step": 12855,
      "train_runtime": 79148.6494,
      "train_tokens_per_second": 8359.835
    },
    {
      "epoch": 0.8658766496094802,
      "grad_norm": 0.5372331922394887,
      "learning_rate": 6.046143960061963e-06,
      "loss": 0.3773908853530884,
      "num_input_tokens_seen": 661945624,
      "step": 12860,
      "train_runtime": 79181.274,
      "train_tokens_per_second": 8359.876
    },
    {
      "epoch": 0.8662133046054403,
      "grad_norm": 0.6301849450684814,
      "learning_rate": 6.043558253975319e-06,
      "loss": 0.38001585006713867,
      "num_input_tokens_seen": 662227128,
      "step": 12865,
      "train_runtime": 79210.9734,
      "train_tokens_per_second": 8360.295
    },
    {
      "epoch": 0.8665499596014005,
      "grad_norm": 0.6343517765677632,
      "learning_rate": 6.040972256060936e-06,
      "loss": 0.3400850534439087,
      "num_input_tokens_seen": 662467672,
      "step": 12870,
      "train_runtime": 79239.9167,
      "train_tokens_per_second": 8360.277
    },
    {
      "epoch": 0.8668866145973606,
      "grad_norm": 0.7899362470882324,
      "learning_rate": 6.038385967041982e-06,
      "loss": 0.3567202091217041,
      "num_input_tokens_seen": 662696192,
      "step": 12875,
      "train_runtime": 79273.0156,
      "train_tokens_per_second": 8359.669
    },
    {
      "epoch": 0.8672232695933207,
      "grad_norm": 0.6058949097916906,
      "learning_rate": 6.035799387641703e-06,
      "loss": 0.36471047401428225,
      "num_input_tokens_seen": 662960728,
      "step": 12880,
      "train_runtime": 79304.0331,
      "train_tokens_per_second": 8359.735
    },
    {
      "epoch": 0.8675599245892809,
      "grad_norm": 0.6659547579172281,
      "learning_rate": 6.033212518583427e-06,
      "loss": 0.3678407669067383,
      "num_input_tokens_seen": 663201304,
      "step": 12885,
      "train_runtime": 79335.4358,
      "train_tokens_per_second": 8359.459
    },
    {
      "epoch": 0.867896579585241,
      "grad_norm": 0.6727319021353564,
      "learning_rate": 6.030625360590567e-06,
      "loss": 0.34593958854675294,
      "num_input_tokens_seen": 663471488,
      "step": 12890,
      "train_runtime": 79369.8563,
      "train_tokens_per_second": 8359.238
    },
    {
      "epoch": 0.8682332345812012,
      "grad_norm": 0.5508787623704847,
      "learning_rate": 6.028037914386609e-06,
      "loss": 0.3570049524307251,
      "num_input_tokens_seen": 663731104,
      "step": 12895,
      "train_runtime": 79400.7683,
      "train_tokens_per_second": 8359.253
    },
    {
      "epoch": 0.8685698895771613,
      "grad_norm": 0.6413263416987328,
      "learning_rate": 6.025450180695128e-06,
      "loss": 0.3772146224975586,
      "num_input_tokens_seen": 663993920,
      "step": 12900,
      "train_runtime": 79431.8112,
      "train_tokens_per_second": 8359.295
    },
    {
      "epoch": 0.8689065445731214,
      "grad_norm": 0.7330655588085616,
      "learning_rate": 6.022862160239774e-06,
      "loss": 0.3618407964706421,
      "num_input_tokens_seen": 664233152,
      "step": 12905,
      "train_runtime": 79462.9991,
      "train_tokens_per_second": 8359.024
    },
    {
      "epoch": 0.8692431995690816,
      "grad_norm": 0.5477016222314597,
      "learning_rate": 6.020273853744277e-06,
      "loss": 0.35313663482666013,
      "num_input_tokens_seen": 664498976,
      "step": 12910,
      "train_runtime": 79493.3836,
      "train_tokens_per_second": 8359.173
    },
    {
      "epoch": 0.8695798545650417,
      "grad_norm": 0.5569301783399404,
      "learning_rate": 6.017685261932452e-06,
      "loss": 0.34578800201416016,
      "num_input_tokens_seen": 664738192,
      "step": 12915,
      "train_runtime": 79525.7776,
      "train_tokens_per_second": 8358.776
    },
    {
      "epoch": 0.8699165095610019,
      "grad_norm": 0.6763002985094875,
      "learning_rate": 6.015096385528189e-06,
      "loss": 0.3654573917388916,
      "num_input_tokens_seen": 664982760,
      "step": 12920,
      "train_runtime": 79558.6348,
      "train_tokens_per_second": 8358.398
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.616631114161216,
      "learning_rate": 6.012507225255457e-06,
      "loss": 0.3833828687667847,
      "num_input_tokens_seen": 665218408,
      "step": 12925,
      "train_runtime": 79585.2351,
      "train_tokens_per_second": 8358.566
    },
    {
      "epoch": 0.8705898195529221,
      "grad_norm": 0.733101908031184,
      "learning_rate": 6.009917781838309e-06,
      "loss": 0.3590146780014038,
      "num_input_tokens_seen": 665472304,
      "step": 12930,
      "train_runtime": 79616.3453,
      "train_tokens_per_second": 8358.488
    },
    {
      "epoch": 0.8709264745488823,
      "grad_norm": 0.6278281471357632,
      "learning_rate": 6.007328056000873e-06,
      "loss": 0.3749275207519531,
      "num_input_tokens_seen": 665742776,
      "step": 12935,
      "train_runtime": 79648.9242,
      "train_tokens_per_second": 8358.465
    },
    {
      "epoch": 0.8712631295448424,
      "grad_norm": 0.7802250877628468,
      "learning_rate": 6.004738048467359e-06,
      "loss": 0.3681314706802368,
      "num_input_tokens_seen": 666007800,
      "step": 12940,
      "train_runtime": 79684.2858,
      "train_tokens_per_second": 8358.082
    },
    {
      "epoch": 0.8715997845408026,
      "grad_norm": 0.6493274106300405,
      "learning_rate": 6.002147759962054e-06,
      "loss": 0.3732911109924316,
      "num_input_tokens_seen": 666267128,
      "step": 12945,
      "train_runtime": 79714.6893,
      "train_tokens_per_second": 8358.147
    },
    {
      "epoch": 0.8719364395367627,
      "grad_norm": 0.648246272114855,
      "learning_rate": 5.999557191209323e-06,
      "loss": 0.3878971576690674,
      "num_input_tokens_seen": 666548464,
      "step": 12950,
      "train_runtime": 79741.7922,
      "train_tokens_per_second": 8358.835
    },
    {
      "epoch": 0.8722730945327228,
      "grad_norm": 0.7414918623968676,
      "learning_rate": 5.996966342933611e-06,
      "loss": 0.3335369825363159,
      "num_input_tokens_seen": 666802152,
      "step": 12955,
      "train_runtime": 79774.2285,
      "train_tokens_per_second": 8358.616
    },
    {
      "epoch": 0.872609749528683,
      "grad_norm": 0.8903123266257061,
      "learning_rate": 5.99437521585944e-06,
      "loss": 0.36038334369659425,
      "num_input_tokens_seen": 667047320,
      "step": 12960,
      "train_runtime": 79803.3572,
      "train_tokens_per_second": 8358.637
    },
    {
      "epoch": 0.8729464045246431,
      "grad_norm": 0.5522375949829542,
      "learning_rate": 5.99178381071141e-06,
      "loss": 0.37600126266479494,
      "num_input_tokens_seen": 667308888,
      "step": 12965,
      "train_runtime": 79830.1957,
      "train_tokens_per_second": 8359.104
    },
    {
      "epoch": 0.8732830595206033,
      "grad_norm": 0.6285315512148892,
      "learning_rate": 5.989192128214201e-06,
      "loss": 0.35153708457946775,
      "num_input_tokens_seen": 667570632,
      "step": 12970,
      "train_runtime": 79860.8921,
      "train_tokens_per_second": 8359.168
    },
    {
      "epoch": 0.8736197145165634,
      "grad_norm": 0.6211720897845763,
      "learning_rate": 5.986600169092568e-06,
      "loss": 0.34699037075042727,
      "num_input_tokens_seen": 667827376,
      "step": 12975,
      "train_runtime": 79891.7846,
      "train_tokens_per_second": 8359.15
    },
    {
      "epoch": 0.8739563695125235,
      "grad_norm": 0.5921000347503103,
      "learning_rate": 5.984007934071343e-06,
      "loss": 0.3637322187423706,
      "num_input_tokens_seen": 668104488,
      "step": 12980,
      "train_runtime": 79923.8586,
      "train_tokens_per_second": 8359.262
    },
    {
      "epoch": 0.8742930245084837,
      "grad_norm": 0.6441117299237299,
      "learning_rate": 5.981415423875436e-06,
      "loss": 0.35071163177490233,
      "num_input_tokens_seen": 668370536,
      "step": 12985,
      "train_runtime": 79956.1668,
      "train_tokens_per_second": 8359.212
    },
    {
      "epoch": 0.8746296795044438,
      "grad_norm": 0.6240763316385677,
      "learning_rate": 5.978822639229836e-06,
      "loss": 0.3455819606781006,
      "num_input_tokens_seen": 668607456,
      "step": 12990,
      "train_runtime": 79987.3933,
      "train_tokens_per_second": 8358.91
    },
    {
      "epoch": 0.874966334500404,
      "grad_norm": 0.6306984348524409,
      "learning_rate": 5.976229580859607e-06,
      "loss": 0.37087626457214357,
      "num_input_tokens_seen": 668886864,
      "step": 12995,
      "train_runtime": 80020.839,
      "train_tokens_per_second": 8358.908
    },
    {
      "epoch": 0.8753029894963641,
      "grad_norm": 0.602709172300639,
      "learning_rate": 5.9736362494898895e-06,
      "loss": 0.3443007230758667,
      "num_input_tokens_seen": 669147096,
      "step": 13000,
      "train_runtime": 80053.4667,
      "train_tokens_per_second": 8358.752
    },
    {
      "epoch": 0.8756396444923242,
      "grad_norm": 0.5810050418133745,
      "learning_rate": 5.971042645845897e-06,
      "loss": 0.36269965171813967,
      "num_input_tokens_seen": 669394448,
      "step": 13005,
      "train_runtime": 80086.656,
      "train_tokens_per_second": 8358.377
    },
    {
      "epoch": 0.8759762994882844,
      "grad_norm": 0.6104899173495202,
      "learning_rate": 5.968448770652926e-06,
      "loss": 0.3556150674819946,
      "num_input_tokens_seen": 669683776,
      "step": 13010,
      "train_runtime": 80121.5614,
      "train_tokens_per_second": 8358.347
    },
    {
      "epoch": 0.8763129544842445,
      "grad_norm": 0.6146643601076629,
      "learning_rate": 5.965854624636345e-06,
      "loss": 0.3771155834197998,
      "num_input_tokens_seen": 669937000,
      "step": 13015,
      "train_runtime": 80153.2906,
      "train_tokens_per_second": 8358.197
    },
    {
      "epoch": 0.8766496094802047,
      "grad_norm": 0.6221659356213541,
      "learning_rate": 5.963260208521598e-06,
      "loss": 0.3618452072143555,
      "num_input_tokens_seen": 670188288,
      "step": 13020,
      "train_runtime": 80184.0792,
      "train_tokens_per_second": 8358.122
    },
    {
      "epoch": 0.8769862644761648,
      "grad_norm": 0.5980915087213089,
      "learning_rate": 5.960665523034205e-06,
      "loss": 0.3761177062988281,
      "num_input_tokens_seen": 670438024,
      "step": 13025,
      "train_runtime": 80211.755,
      "train_tokens_per_second": 8358.351
    },
    {
      "epoch": 0.8773229194721249,
      "grad_norm": 0.6293263560441469,
      "learning_rate": 5.958070568899762e-06,
      "loss": 0.31571502685546876,
      "num_input_tokens_seen": 670695088,
      "step": 13030,
      "train_runtime": 80240.8708,
      "train_tokens_per_second": 8358.522
    },
    {
      "epoch": 0.8776595744680851,
      "grad_norm": 0.613386498526115,
      "learning_rate": 5.955475346843939e-06,
      "loss": 0.34776930809020995,
      "num_input_tokens_seen": 670953656,
      "step": 13035,
      "train_runtime": 80274.8324,
      "train_tokens_per_second": 8358.207
    },
    {
      "epoch": 0.8779962294640452,
      "grad_norm": 0.6127088088711503,
      "learning_rate": 5.952879857592482e-06,
      "loss": 0.32555651664733887,
      "num_input_tokens_seen": 671194960,
      "step": 13040,
      "train_runtime": 80306.9985,
      "train_tokens_per_second": 8357.864
    },
    {
      "epoch": 0.8783328844600053,
      "grad_norm": 0.5715958214100246,
      "learning_rate": 5.950284101871212e-06,
      "loss": 0.3715256690979004,
      "num_input_tokens_seen": 671436320,
      "step": 13045,
      "train_runtime": 80339.2051,
      "train_tokens_per_second": 8357.518
    },
    {
      "epoch": 0.8786695394559655,
      "grad_norm": 0.7604549248117403,
      "learning_rate": 5.947688080406023e-06,
      "loss": 0.34642324447631834,
      "num_input_tokens_seen": 671680048,
      "step": 13050,
      "train_runtime": 80369.3029,
      "train_tokens_per_second": 8357.42
    },
    {
      "epoch": 0.8790061944519256,
      "grad_norm": 0.5749773549246331,
      "learning_rate": 5.945091793922885e-06,
      "loss": 0.36411921977996825,
      "num_input_tokens_seen": 671934712,
      "step": 13055,
      "train_runtime": 80400.5911,
      "train_tokens_per_second": 8357.335
    },
    {
      "epoch": 0.8793428494478858,
      "grad_norm": 0.5404499122193702,
      "learning_rate": 5.942495243147839e-06,
      "loss": 0.3814393520355225,
      "num_input_tokens_seen": 672205768,
      "step": 13060,
      "train_runtime": 80435.4124,
      "train_tokens_per_second": 8357.087
    },
    {
      "epoch": 0.8796795044438459,
      "grad_norm": 0.6178171126765887,
      "learning_rate": 5.939898428807004e-06,
      "loss": 0.3885827541351318,
      "num_input_tokens_seen": 672458880,
      "step": 13065,
      "train_runtime": 80464.3805,
      "train_tokens_per_second": 8357.224
    },
    {
      "epoch": 0.880016159439806,
      "grad_norm": 0.7703917702526525,
      "learning_rate": 5.937301351626571e-06,
      "loss": 0.3797243595123291,
      "num_input_tokens_seen": 672698960,
      "step": 13070,
      "train_runtime": 80497.6525,
      "train_tokens_per_second": 8356.753
    },
    {
      "epoch": 0.8803528144357662,
      "grad_norm": 0.5466466759879106,
      "learning_rate": 5.934704012332803e-06,
      "loss": 0.35700135231018065,
      "num_input_tokens_seen": 672978832,
      "step": 13075,
      "train_runtime": 80531.2056,
      "train_tokens_per_second": 8356.746
    },
    {
      "epoch": 0.8806894694317263,
      "grad_norm": 0.663173906833156,
      "learning_rate": 5.93210641165204e-06,
      "loss": 0.39308757781982423,
      "num_input_tokens_seen": 673237072,
      "step": 13080,
      "train_runtime": 80562.6954,
      "train_tokens_per_second": 8356.685
    },
    {
      "epoch": 0.8810261244276865,
      "grad_norm": 0.6740342019780106,
      "learning_rate": 5.92950855031069e-06,
      "loss": 0.35691306591033933,
      "num_input_tokens_seen": 673498128,
      "step": 13085,
      "train_runtime": 80594.494,
      "train_tokens_per_second": 8356.627
    },
    {
      "epoch": 0.8813627794236466,
      "grad_norm": 0.7541494572586337,
      "learning_rate": 5.926910429035239e-06,
      "loss": 0.34788856506347654,
      "num_input_tokens_seen": 673754072,
      "step": 13090,
      "train_runtime": 80621.7871,
      "train_tokens_per_second": 8356.973
    },
    {
      "epoch": 0.8816994344196067,
      "grad_norm": 0.5912874354630311,
      "learning_rate": 5.924312048552241e-06,
      "loss": 0.3523898124694824,
      "num_input_tokens_seen": 674026600,
      "step": 13095,
      "train_runtime": 80655.9802,
      "train_tokens_per_second": 8356.809
    },
    {
      "epoch": 0.8820360894155669,
      "grad_norm": 0.5912435598521816,
      "learning_rate": 5.9217134095883265e-06,
      "loss": 0.35706634521484376,
      "num_input_tokens_seen": 674271616,
      "step": 13100,
      "train_runtime": 80688.1017,
      "train_tokens_per_second": 8356.519
    },
    {
      "epoch": 0.882372744411527,
      "grad_norm": 0.6676697241534663,
      "learning_rate": 5.919114512870197e-06,
      "loss": 0.35798444747924807,
      "num_input_tokens_seen": 674543728,
      "step": 13105,
      "train_runtime": 80717.5636,
      "train_tokens_per_second": 8356.84
    },
    {
      "epoch": 0.8827093994074872,
      "grad_norm": 0.631702906041094,
      "learning_rate": 5.916515359124623e-06,
      "loss": 0.34296715259552,
      "num_input_tokens_seen": 674771968,
      "step": 13110,
      "train_runtime": 80748.623,
      "train_tokens_per_second": 8356.452
    },
    {
      "epoch": 0.8830460544034473,
      "grad_norm": 0.6229605313080638,
      "learning_rate": 5.913915949078453e-06,
      "loss": 0.35898048877716066,
      "num_input_tokens_seen": 675033176,
      "step": 13115,
      "train_runtime": 80778.7869,
      "train_tokens_per_second": 8356.565
    },
    {
      "epoch": 0.8833827093994074,
      "grad_norm": 0.7181691637246672,
      "learning_rate": 5.911316283458601e-06,
      "loss": 0.35623142719268797,
      "num_input_tokens_seen": 675263168,
      "step": 13120,
      "train_runtime": 80807.6039,
      "train_tokens_per_second": 8356.431
    },
    {
      "epoch": 0.8837193643953676,
      "grad_norm": 0.6335555101527864,
      "learning_rate": 5.908716362992057e-06,
      "loss": 0.35456418991088867,
      "num_input_tokens_seen": 675510080,
      "step": 13125,
      "train_runtime": 80836.0382,
      "train_tokens_per_second": 8356.546
    },
    {
      "epoch": 0.8840560193913277,
      "grad_norm": 0.5883535057275991,
      "learning_rate": 5.90611618840588e-06,
      "loss": 0.3699319362640381,
      "num_input_tokens_seen": 675764960,
      "step": 13130,
      "train_runtime": 80869.9224,
      "train_tokens_per_second": 8356.196
    },
    {
      "epoch": 0.8843926743872879,
      "grad_norm": 0.6705370519749358,
      "learning_rate": 5.903515760427198e-06,
      "loss": 0.3926721096038818,
      "num_input_tokens_seen": 676031952,
      "step": 13135,
      "train_runtime": 80898.3238,
      "train_tokens_per_second": 8356.563
    },
    {
      "epoch": 0.884729329383248,
      "grad_norm": 0.6028790310520268,
      "learning_rate": 5.900915079783218e-06,
      "loss": 0.3491120576858521,
      "num_input_tokens_seen": 676322408,
      "step": 13140,
      "train_runtime": 80930.0175,
      "train_tokens_per_second": 8356.88
    },
    {
      "epoch": 0.8850659843792081,
      "grad_norm": 0.6636768547007416,
      "learning_rate": 5.898314147201207e-06,
      "loss": 0.3428701400756836,
      "num_input_tokens_seen": 676568376,
      "step": 13145,
      "train_runtime": 80960.2862,
      "train_tokens_per_second": 8356.793
    },
    {
      "epoch": 0.8854026393751683,
      "grad_norm": 0.658617460559585,
      "learning_rate": 5.895712963408511e-06,
      "loss": 0.3859280824661255,
      "num_input_tokens_seen": 676806248,
      "step": 13150,
      "train_runtime": 80994.9232,
      "train_tokens_per_second": 8356.156
    },
    {
      "epoch": 0.8857392943711284,
      "grad_norm": 0.6912933384558435,
      "learning_rate": 5.893111529132539e-06,
      "loss": 0.368697452545166,
      "num_input_tokens_seen": 677048160,
      "step": 13155,
      "train_runtime": 81026.0385,
      "train_tokens_per_second": 8355.933
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.6475137048551329,
      "learning_rate": 5.8905098451007755e-06,
      "loss": 0.3717859983444214,
      "num_input_tokens_seen": 677282944,
      "step": 13160,
      "train_runtime": 81053.5525,
      "train_tokens_per_second": 8355.993
    },
    {
      "epoch": 0.8864126043630487,
      "grad_norm": 0.6055045131655981,
      "learning_rate": 5.887907912040774e-06,
      "loss": 0.37036609649658203,
      "num_input_tokens_seen": 677551912,
      "step": 13165,
      "train_runtime": 81086.6847,
      "train_tokens_per_second": 8355.896
    },
    {
      "epoch": 0.8867492593590088,
      "grad_norm": 0.5755981580554987,
      "learning_rate": 5.885305730680158e-06,
      "loss": 0.37028448581695556,
      "num_input_tokens_seen": 677832160,
      "step": 13170,
      "train_runtime": 81116.0161,
      "train_tokens_per_second": 8356.33
    },
    {
      "epoch": 0.887085914354969,
      "grad_norm": 0.6577782079013551,
      "learning_rate": 5.882703301746616e-06,
      "loss": 0.3542908191680908,
      "num_input_tokens_seen": 678078200,
      "step": 13175,
      "train_runtime": 81148.3506,
      "train_tokens_per_second": 8356.032
    },
    {
      "epoch": 0.8874225693509291,
      "grad_norm": 0.6446898364479628,
      "learning_rate": 5.88010062596791e-06,
      "loss": 0.3630987644195557,
      "num_input_tokens_seen": 678316240,
      "step": 13180,
      "train_runtime": 81179.0278,
      "train_tokens_per_second": 8355.806
    },
    {
      "epoch": 0.8877592243468893,
      "grad_norm": 0.6112383613306287,
      "learning_rate": 5.877497704071871e-06,
      "loss": 0.33677263259887696,
      "num_input_tokens_seen": 678602976,
      "step": 13185,
      "train_runtime": 81209.8483,
      "train_tokens_per_second": 8356.166
    },
    {
      "epoch": 0.8880958793428494,
      "grad_norm": 0.6994302117687429,
      "learning_rate": 5.874894536786396e-06,
      "loss": 0.33505859375,
      "num_input_tokens_seen": 678852520,
      "step": 13190,
      "train_runtime": 81241.0464,
      "train_tokens_per_second": 8356.029
    },
    {
      "epoch": 0.8884325343388096,
      "grad_norm": 0.6821480824613465,
      "learning_rate": 5.8722911248394555e-06,
      "loss": 0.33796401023864747,
      "num_input_tokens_seen": 679076728,
      "step": 13195,
      "train_runtime": 81271.9701,
      "train_tokens_per_second": 8355.608
    },
    {
      "epoch": 0.8887691893347698,
      "grad_norm": 0.6432445191850359,
      "learning_rate": 5.8696874689590824e-06,
      "loss": 0.3454821825027466,
      "num_input_tokens_seen": 679327096,
      "step": 13200,
      "train_runtime": 81304.7703,
      "train_tokens_per_second": 8355.317
    },
    {
      "epoch": 0.8891058443307299,
      "grad_norm": 0.6483590674380476,
      "learning_rate": 5.867083569873383e-06,
      "loss": 0.35104620456695557,
      "num_input_tokens_seen": 679594072,
      "step": 13205,
      "train_runtime": 81337.1174,
      "train_tokens_per_second": 8355.276
    },
    {
      "epoch": 0.8894424993266901,
      "grad_norm": 0.6494957001899682,
      "learning_rate": 5.8644794283105255e-06,
      "loss": 0.35859639644622804,
      "num_input_tokens_seen": 679860896,
      "step": 13210,
      "train_runtime": 81368.1531,
      "train_tokens_per_second": 8355.368
    },
    {
      "epoch": 0.8897791543226502,
      "grad_norm": 0.6500607169201391,
      "learning_rate": 5.861875044998755e-06,
      "loss": 0.40003271102905275,
      "num_input_tokens_seen": 680137096,
      "step": 13215,
      "train_runtime": 81403.6216,
      "train_tokens_per_second": 8355.121
    },
    {
      "epoch": 0.8901158093186103,
      "grad_norm": 0.6499120615207054,
      "learning_rate": 5.859270420666377e-06,
      "loss": 0.376279616355896,
      "num_input_tokens_seen": 680392280,
      "step": 13220,
      "train_runtime": 81431.102,
      "train_tokens_per_second": 8355.435
    },
    {
      "epoch": 0.8904524643145705,
      "grad_norm": 0.6943584686829152,
      "learning_rate": 5.856665556041764e-06,
      "loss": 0.32646162509918214,
      "num_input_tokens_seen": 680631360,
      "step": 13225,
      "train_runtime": 81461.4195,
      "train_tokens_per_second": 8355.26
    },
    {
      "epoch": 0.8907891193105306,
      "grad_norm": 0.6225818777936154,
      "learning_rate": 5.8540604518533605e-06,
      "loss": 0.37882609367370607,
      "num_input_tokens_seen": 680899840,
      "step": 13230,
      "train_runtime": 81492.9353,
      "train_tokens_per_second": 8355.324
    },
    {
      "epoch": 0.8911257743064908,
      "grad_norm": 0.665543897547396,
      "learning_rate": 5.851455108829675e-06,
      "loss": 0.33943424224853513,
      "num_input_tokens_seen": 681132536,
      "step": 13235,
      "train_runtime": 81524.9002,
      "train_tokens_per_second": 8354.902
    },
    {
      "epoch": 0.8914624293024509,
      "grad_norm": 0.5861185902234483,
      "learning_rate": 5.848849527699283e-06,
      "loss": 0.3572589874267578,
      "num_input_tokens_seen": 681396864,
      "step": 13240,
      "train_runtime": 81553.2124,
      "train_tokens_per_second": 8355.242
    },
    {
      "epoch": 0.891799084298411,
      "grad_norm": 0.6737200437349476,
      "learning_rate": 5.846243709190827e-06,
      "loss": 0.3545804500579834,
      "num_input_tokens_seen": 681649296,
      "step": 13245,
      "train_runtime": 81585.1224,
      "train_tokens_per_second": 8355.069
    },
    {
      "epoch": 0.8921357392943712,
      "grad_norm": 1.0914195362437003,
      "learning_rate": 5.843637654033019e-06,
      "loss": 0.3703150749206543,
      "num_input_tokens_seen": 681869008,
      "step": 13250,
      "train_runtime": 81612.9288,
      "train_tokens_per_second": 8354.914
    },
    {
      "epoch": 0.8924723942903313,
      "grad_norm": 0.5879180227469825,
      "learning_rate": 5.841031362954629e-06,
      "loss": 0.38281750679016113,
      "num_input_tokens_seen": 682143400,
      "step": 13255,
      "train_runtime": 81642.1065,
      "train_tokens_per_second": 8355.289
    },
    {
      "epoch": 0.8928090492862915,
      "grad_norm": 0.5959332066857956,
      "learning_rate": 5.8384248366845e-06,
      "loss": 0.3569962024688721,
      "num_input_tokens_seen": 682414160,
      "step": 13260,
      "train_runtime": 81674.0819,
      "train_tokens_per_second": 8355.333
    },
    {
      "epoch": 0.8931457042822516,
      "grad_norm": 0.5527677255712462,
      "learning_rate": 5.8358180759515396e-06,
      "loss": 0.3633003234863281,
      "num_input_tokens_seen": 682668024,
      "step": 13265,
      "train_runtime": 81703.5144,
      "train_tokens_per_second": 8355.43
    },
    {
      "epoch": 0.8934823592782117,
      "grad_norm": 0.6444113735574559,
      "learning_rate": 5.833211081484719e-06,
      "loss": 0.3606697082519531,
      "num_input_tokens_seen": 682927520,
      "step": 13270,
      "train_runtime": 81736.3803,
      "train_tokens_per_second": 8355.245
    },
    {
      "epoch": 0.8938190142741719,
      "grad_norm": 0.5567127561256134,
      "learning_rate": 5.830603854013077e-06,
      "loss": 0.36896984577178954,
      "num_input_tokens_seen": 683182672,
      "step": 13275,
      "train_runtime": 81764.5955,
      "train_tokens_per_second": 8355.483
    },
    {
      "epoch": 0.894155669270132,
      "grad_norm": 0.8056776213724244,
      "learning_rate": 5.827996394265716e-06,
      "loss": 0.3482972145080566,
      "num_input_tokens_seen": 683424056,
      "step": 13280,
      "train_runtime": 81794.1322,
      "train_tokens_per_second": 8355.417
    },
    {
      "epoch": 0.8944923242660922,
      "grad_norm": 0.6090327950035972,
      "learning_rate": 5.825388702971802e-06,
      "loss": 0.3394944429397583,
      "num_input_tokens_seen": 683683928,
      "step": 13285,
      "train_runtime": 81822.2872,
      "train_tokens_per_second": 8355.718
    },
    {
      "epoch": 0.8948289792620523,
      "grad_norm": 0.6526084353001578,
      "learning_rate": 5.822780780860571e-06,
      "loss": 0.3509685754776001,
      "num_input_tokens_seen": 683941624,
      "step": 13290,
      "train_runtime": 81853.8843,
      "train_tokens_per_second": 8355.64
    },
    {
      "epoch": 0.8951656342580124,
      "grad_norm": 0.5811923982870922,
      "learning_rate": 5.820172628661317e-06,
      "loss": 0.35237894058227537,
      "num_input_tokens_seen": 684226304,
      "step": 13295,
      "train_runtime": 81885.8159,
      "train_tokens_per_second": 8355.859
    },
    {
      "epoch": 0.8955022892539726,
      "grad_norm": 0.6585470828152886,
      "learning_rate": 5.8175642471034045e-06,
      "loss": 0.34798598289489746,
      "num_input_tokens_seen": 684492784,
      "step": 13300,
      "train_runtime": 81920.6442,
      "train_tokens_per_second": 8355.559
    },
    {
      "epoch": 0.8958389442499327,
      "grad_norm": 0.6650646725548803,
      "learning_rate": 5.8149556369162565e-06,
      "loss": 0.36943299770355226,
      "num_input_tokens_seen": 684734440,
      "step": 13305,
      "train_runtime": 81952.5442,
      "train_tokens_per_second": 8355.255
    },
    {
      "epoch": 0.8961755992458929,
      "grad_norm": 0.6229412195257346,
      "learning_rate": 5.812346798829361e-06,
      "loss": 0.37810468673706055,
      "num_input_tokens_seen": 685009440,
      "step": 13310,
      "train_runtime": 81982.2544,
      "train_tokens_per_second": 8355.582
    },
    {
      "epoch": 0.896512254241853,
      "grad_norm": 0.7880550877016411,
      "learning_rate": 5.809737733572276e-06,
      "loss": 0.35051565170288085,
      "num_input_tokens_seen": 685257632,
      "step": 13315,
      "train_runtime": 82013.9311,
      "train_tokens_per_second": 8355.381
    },
    {
      "epoch": 0.8968489092378131,
      "grad_norm": 0.7544075045178883,
      "learning_rate": 5.8071284418746166e-06,
      "loss": 0.35827059745788575,
      "num_input_tokens_seen": 685520944,
      "step": 13320,
      "train_runtime": 82041.8955,
      "train_tokens_per_second": 8355.742
    },
    {
      "epoch": 0.8971855642337733,
      "grad_norm": 0.6158205759677424,
      "learning_rate": 5.80451892446606e-06,
      "loss": 0.365170955657959,
      "num_input_tokens_seen": 685789984,
      "step": 13325,
      "train_runtime": 82070.7912,
      "train_tokens_per_second": 8356.079
    },
    {
      "epoch": 0.8975222192297334,
      "grad_norm": 0.594625704206593,
      "learning_rate": 5.801909182076354e-06,
      "loss": 0.3513838052749634,
      "num_input_tokens_seen": 686065736,
      "step": 13330,
      "train_runtime": 82098.3286,
      "train_tokens_per_second": 8356.635
    },
    {
      "epoch": 0.8978588742256935,
      "grad_norm": 0.7117501001920108,
      "learning_rate": 5.7992992154353e-06,
      "loss": 0.39045875072479247,
      "num_input_tokens_seen": 686322048,
      "step": 13335,
      "train_runtime": 82134.1686,
      "train_tokens_per_second": 8356.109
    },
    {
      "epoch": 0.8981955292216537,
      "grad_norm": 0.6356143322903928,
      "learning_rate": 5.79668902527277e-06,
      "loss": 0.3708018779754639,
      "num_input_tokens_seen": 686579264,
      "step": 13340,
      "train_runtime": 82166.281,
      "train_tokens_per_second": 8355.973
    },
    {
      "epoch": 0.8985321842176138,
      "grad_norm": 0.6428724221233346,
      "learning_rate": 5.794078612318695e-06,
      "loss": 0.354721212387085,
      "num_input_tokens_seen": 686831480,
      "step": 13345,
      "train_runtime": 82196.675,
      "train_tokens_per_second": 8355.952
    },
    {
      "epoch": 0.898868839213574,
      "grad_norm": 0.5158126897386706,
      "learning_rate": 5.791467977303068e-06,
      "loss": 0.33667314052581787,
      "num_input_tokens_seen": 687089584,
      "step": 13350,
      "train_runtime": 82224.5811,
      "train_tokens_per_second": 8356.255
    },
    {
      "epoch": 0.8992054942095341,
      "grad_norm": 0.7420776549273923,
      "learning_rate": 5.788857120955945e-06,
      "loss": 0.37637383937835694,
      "num_input_tokens_seen": 687348160,
      "step": 13355,
      "train_runtime": 82256.6194,
      "train_tokens_per_second": 8356.144
    },
    {
      "epoch": 0.8995421492054942,
      "grad_norm": 0.6210120967842535,
      "learning_rate": 5.786246044007443e-06,
      "loss": 0.33445785045623777,
      "num_input_tokens_seen": 687578280,
      "step": 13360,
      "train_runtime": 82286.6847,
      "train_tokens_per_second": 8355.887
    },
    {
      "epoch": 0.8998788042014544,
      "grad_norm": 0.6349388280569066,
      "learning_rate": 5.783634747187743e-06,
      "loss": 0.3746604919433594,
      "num_input_tokens_seen": 687858280,
      "step": 13365,
      "train_runtime": 82319.6388,
      "train_tokens_per_second": 8355.944
    },
    {
      "epoch": 0.9002154591974145,
      "grad_norm": 0.6151404464794595,
      "learning_rate": 5.781023231227084e-06,
      "loss": 0.3644608974456787,
      "num_input_tokens_seen": 688106112,
      "step": 13370,
      "train_runtime": 82351.5095,
      "train_tokens_per_second": 8355.719
    },
    {
      "epoch": 0.9005521141933747,
      "grad_norm": 0.5685263665461737,
      "learning_rate": 5.778411496855769e-06,
      "loss": 0.36220531463623046,
      "num_input_tokens_seen": 688364248,
      "step": 13375,
      "train_runtime": 82385.4933,
      "train_tokens_per_second": 8355.406
    },
    {
      "epoch": 0.9008887691893348,
      "grad_norm": 0.6380932196290142,
      "learning_rate": 5.775799544804161e-06,
      "loss": 0.372160530090332,
      "num_input_tokens_seen": 688647384,
      "step": 13380,
      "train_runtime": 82418.4742,
      "train_tokens_per_second": 8355.498
    },
    {
      "epoch": 0.9012254241852949,
      "grad_norm": 0.6208987806785653,
      "learning_rate": 5.7731873758026845e-06,
      "loss": 0.35392146110534667,
      "num_input_tokens_seen": 688901168,
      "step": 13385,
      "train_runtime": 82450.7685,
      "train_tokens_per_second": 8355.303
    },
    {
      "epoch": 0.9015620791812551,
      "grad_norm": 0.707785088020397,
      "learning_rate": 5.770574990581823e-06,
      "loss": 0.35057864189147947,
      "num_input_tokens_seen": 689153560,
      "step": 13390,
      "train_runtime": 82481.706,
      "train_tokens_per_second": 8355.229
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.5959845241318221,
      "learning_rate": 5.767962389872124e-06,
      "loss": 0.33239409923553465,
      "num_input_tokens_seen": 689412648,
      "step": 13395,
      "train_runtime": 82509.2896,
      "train_tokens_per_second": 8355.576
    },
    {
      "epoch": 0.9022353891731754,
      "grad_norm": 0.696687241545238,
      "learning_rate": 5.765349574404191e-06,
      "loss": 0.3477863073348999,
      "num_input_tokens_seen": 689650976,
      "step": 13400,
      "train_runtime": 82540.4479,
      "train_tokens_per_second": 8355.309
    },
    {
      "epoch": 0.9025720441691355,
      "grad_norm": 0.6440139481246556,
      "learning_rate": 5.7627365449086905e-06,
      "loss": 0.35566258430480957,
      "num_input_tokens_seen": 689874352,
      "step": 13405,
      "train_runtime": 82566.4951,
      "train_tokens_per_second": 8355.379
    },
    {
      "epoch": 0.9029086991650956,
      "grad_norm": 0.6671425407082485,
      "learning_rate": 5.760123302116345e-06,
      "loss": 0.379826021194458,
      "num_input_tokens_seen": 690147568,
      "step": 13410,
      "train_runtime": 82594.6595,
      "train_tokens_per_second": 8355.838
    },
    {
      "epoch": 0.9032453541610558,
      "grad_norm": 0.6608855712042792,
      "learning_rate": 5.757509846757945e-06,
      "loss": 0.36635336875915525,
      "num_input_tokens_seen": 690418464,
      "step": 13415,
      "train_runtime": 82624.8134,
      "train_tokens_per_second": 8356.067
    },
    {
      "epoch": 0.9035820091570159,
      "grad_norm": 0.7053568526225722,
      "learning_rate": 5.75489617956433e-06,
      "loss": 0.38388853073120116,
      "num_input_tokens_seen": 690668920,
      "step": 13420,
      "train_runtime": 82654.5075,
      "train_tokens_per_second": 8356.095
    },
    {
      "epoch": 0.9039186641529761,
      "grad_norm": 0.6243151158567132,
      "learning_rate": 5.752282301266406e-06,
      "loss": 0.3743810892105103,
      "num_input_tokens_seen": 690932576,
      "step": 13425,
      "train_runtime": 82685.0139,
      "train_tokens_per_second": 8356.201
    },
    {
      "epoch": 0.9042553191489362,
      "grad_norm": 0.6021351960416654,
      "learning_rate": 5.749668212595136e-06,
      "loss": 0.35646777153015136,
      "num_input_tokens_seen": 691187264,
      "step": 13430,
      "train_runtime": 82711.9794,
      "train_tokens_per_second": 8356.556
    },
    {
      "epoch": 0.9045919741448963,
      "grad_norm": 0.6240661837569327,
      "learning_rate": 5.747053914281539e-06,
      "loss": 0.3667402744293213,
      "num_input_tokens_seen": 691421200,
      "step": 13435,
      "train_runtime": 82741.585,
      "train_tokens_per_second": 8356.393
    },
    {
      "epoch": 0.9049286291408565,
      "grad_norm": 0.5925245305090393,
      "learning_rate": 5.744439407056699e-06,
      "loss": 0.34764633178710935,
      "num_input_tokens_seen": 691665296,
      "step": 13440,
      "train_runtime": 82768.7871,
      "train_tokens_per_second": 8356.596
    },
    {
      "epoch": 0.9052652841368166,
      "grad_norm": 0.5421059411878397,
      "learning_rate": 5.741824691651752e-06,
      "loss": 0.3322668313980103,
      "num_input_tokens_seen": 691932304,
      "step": 13445,
      "train_runtime": 82799.5433,
      "train_tokens_per_second": 8356.716
    },
    {
      "epoch": 0.9056019391327768,
      "grad_norm": 0.740526296961368,
      "learning_rate": 5.739209768797896e-06,
      "loss": 0.3581199884414673,
      "num_input_tokens_seen": 692180344,
      "step": 13450,
      "train_runtime": 82829.7312,
      "train_tokens_per_second": 8356.665
    },
    {
      "epoch": 0.9059385941287369,
      "grad_norm": 0.6487192288650974,
      "learning_rate": 5.736594639226385e-06,
      "loss": 0.32801964282989504,
      "num_input_tokens_seen": 692446416,
      "step": 13455,
      "train_runtime": 82860.6819,
      "train_tokens_per_second": 8356.755
    },
    {
      "epoch": 0.906275249124697,
      "grad_norm": 0.5947696974588971,
      "learning_rate": 5.7339793036685306e-06,
      "loss": 0.3690043926239014,
      "num_input_tokens_seen": 692715696,
      "step": 13460,
      "train_runtime": 82890.1014,
      "train_tokens_per_second": 8357.038
    },
    {
      "epoch": 0.9066119041206572,
      "grad_norm": 0.6523864636214958,
      "learning_rate": 5.731363762855706e-06,
      "loss": 0.3398882389068604,
      "num_input_tokens_seen": 692959352,
      "step": 13465,
      "train_runtime": 82923.3728,
      "train_tokens_per_second": 8356.623
    },
    {
      "epoch": 0.9069485591166173,
      "grad_norm": 0.6635409196494506,
      "learning_rate": 5.728748017519337e-06,
      "loss": 0.38260316848754883,
      "num_input_tokens_seen": 693213360,
      "step": 13470,
      "train_runtime": 82954.5257,
      "train_tokens_per_second": 8356.547
    },
    {
      "epoch": 0.9072852141125775,
      "grad_norm": 0.5907922782885596,
      "learning_rate": 5.726132068390908e-06,
      "loss": 0.3552544116973877,
      "num_input_tokens_seen": 693450264,
      "step": 13475,
      "train_runtime": 82983.7019,
      "train_tokens_per_second": 8356.463
    },
    {
      "epoch": 0.9076218691085376,
      "grad_norm": 0.6629022196453397,
      "learning_rate": 5.723515916201962e-06,
      "loss": 0.39786829948425295,
      "num_input_tokens_seen": 693727488,
      "step": 13480,
      "train_runtime": 83010.8243,
      "train_tokens_per_second": 8357.073
    },
    {
      "epoch": 0.9079585241044977,
      "grad_norm": 0.5788025340218351,
      "learning_rate": 5.720899561684098e-06,
      "loss": 0.3507582664489746,
      "num_input_tokens_seen": 693986168,
      "step": 13485,
      "train_runtime": 83044.3382,
      "train_tokens_per_second": 8356.815
    },
    {
      "epoch": 0.9082951791004579,
      "grad_norm": 0.746085931288479,
      "learning_rate": 5.71828300556897e-06,
      "loss": 0.3621519565582275,
      "num_input_tokens_seen": 694255736,
      "step": 13490,
      "train_runtime": 83070.91,
      "train_tokens_per_second": 8357.387
    },
    {
      "epoch": 0.908631834096418,
      "grad_norm": 0.5557121070047347,
      "learning_rate": 5.715666248588289e-06,
      "loss": 0.3518456220626831,
      "num_input_tokens_seen": 694525896,
      "step": 13495,
      "train_runtime": 83100.748,
      "train_tokens_per_second": 8357.637
    },
    {
      "epoch": 0.9089684890923782,
      "grad_norm": 0.6553426569653834,
      "learning_rate": 5.713049291473825e-06,
      "loss": 0.37739195823669436,
      "num_input_tokens_seen": 694795800,
      "step": 13500,
      "train_runtime": 83130.9514,
      "train_tokens_per_second": 8357.847
    },
    {
      "epoch": 0.9093051440883383,
      "grad_norm": 0.722991303847563,
      "learning_rate": 5.710432134957399e-06,
      "loss": 0.3340196371078491,
      "num_input_tokens_seen": 695015632,
      "step": 13505,
      "train_runtime": 83159.6075,
      "train_tokens_per_second": 8357.611
    },
    {
      "epoch": 0.9096417990842984,
      "grad_norm": 0.6788133989345068,
      "learning_rate": 5.707814779770892e-06,
      "loss": 0.3379532814025879,
      "num_input_tokens_seen": 695255144,
      "step": 13510,
      "train_runtime": 83188.411,
      "train_tokens_per_second": 8357.596
    },
    {
      "epoch": 0.9099784540802586,
      "grad_norm": 0.6417574257323145,
      "learning_rate": 5.7051972266462395e-06,
      "loss": 0.35926411151885984,
      "num_input_tokens_seen": 695521560,
      "step": 13515,
      "train_runtime": 83214.3644,
      "train_tokens_per_second": 8358.191
    },
    {
      "epoch": 0.9103151090762187,
      "grad_norm": 0.779000274628635,
      "learning_rate": 5.702579476315431e-06,
      "loss": 0.33089332580566405,
      "num_input_tokens_seen": 695772488,
      "step": 13520,
      "train_runtime": 83244.0512,
      "train_tokens_per_second": 8358.225
    },
    {
      "epoch": 0.9106517640721788,
      "grad_norm": 0.6634564232237014,
      "learning_rate": 5.699961529510513e-06,
      "loss": 0.3610454797744751,
      "num_input_tokens_seen": 696004376,
      "step": 13525,
      "train_runtime": 83274.2637,
      "train_tokens_per_second": 8357.977
    },
    {
      "epoch": 0.910988419068139,
      "grad_norm": 0.6535047092031706,
      "learning_rate": 5.697343386963586e-06,
      "loss": 0.36852355003356935,
      "num_input_tokens_seen": 696254344,
      "step": 13530,
      "train_runtime": 83303.5783,
      "train_tokens_per_second": 8358.036
    },
    {
      "epoch": 0.9113250740640991,
      "grad_norm": 0.5762681709443156,
      "learning_rate": 5.694725049406801e-06,
      "loss": 0.36067972183227537,
      "num_input_tokens_seen": 696503704,
      "step": 13535,
      "train_runtime": 83333.1131,
      "train_tokens_per_second": 8358.067
    },
    {
      "epoch": 0.9116617290600593,
      "grad_norm": 0.5909215534500072,
      "learning_rate": 5.692106517572373e-06,
      "loss": 0.37316346168518066,
      "num_input_tokens_seen": 696757112,
      "step": 13540,
      "train_runtime": 83360.8059,
      "train_tokens_per_second": 8358.33
    },
    {
      "epoch": 0.9119983840560194,
      "grad_norm": 0.6945554074478593,
      "learning_rate": 5.689487792192566e-06,
      "loss": 0.4025864124298096,
      "num_input_tokens_seen": 697011192,
      "step": 13545,
      "train_runtime": 83394.617,
      "train_tokens_per_second": 8357.988
    },
    {
      "epoch": 0.9123350390519795,
      "grad_norm": 0.6588805103529365,
      "learning_rate": 5.686868873999696e-06,
      "loss": 0.3683311939239502,
      "num_input_tokens_seen": 697255816,
      "step": 13550,
      "train_runtime": 83427.8269,
      "train_tokens_per_second": 8357.593
    },
    {
      "epoch": 0.9126716940479397,
      "grad_norm": 0.5290167765378848,
      "learning_rate": 5.684249763726135e-06,
      "loss": 0.3581395626068115,
      "num_input_tokens_seen": 697523568,
      "step": 13555,
      "train_runtime": 83461.2615,
      "train_tokens_per_second": 8357.453
    },
    {
      "epoch": 0.9130083490438998,
      "grad_norm": 0.6550305377261901,
      "learning_rate": 5.681630462104308e-06,
      "loss": 0.34157938957214357,
      "num_input_tokens_seen": 697751992,
      "step": 13560,
      "train_runtime": 83491.0929,
      "train_tokens_per_second": 8357.203
    },
    {
      "epoch": 0.91334500403986,
      "grad_norm": 0.6963928585434378,
      "learning_rate": 5.679010969866697e-06,
      "loss": 0.35550198554992674,
      "num_input_tokens_seen": 697999824,
      "step": 13565,
      "train_runtime": 83519.2134,
      "train_tokens_per_second": 8357.356
    },
    {
      "epoch": 0.9136816590358201,
      "grad_norm": 0.5939214263038702,
      "learning_rate": 5.676391287745835e-06,
      "loss": 0.3482653141021729,
      "num_input_tokens_seen": 698296448,
      "step": 13570,
      "train_runtime": 83553.4858,
      "train_tokens_per_second": 8357.478
    },
    {
      "epoch": 0.9140183140317802,
      "grad_norm": 0.6272391774602417,
      "learning_rate": 5.673771416474305e-06,
      "loss": 0.3388658046722412,
      "num_input_tokens_seen": 698554480,
      "step": 13575,
      "train_runtime": 83586.6225,
      "train_tokens_per_second": 8357.252
    },
    {
      "epoch": 0.9143549690277404,
      "grad_norm": 0.6515459624450703,
      "learning_rate": 5.671151356784747e-06,
      "loss": 0.3396682500839233,
      "num_input_tokens_seen": 698800456,
      "step": 13580,
      "train_runtime": 83620.7232,
      "train_tokens_per_second": 8356.786
    },
    {
      "epoch": 0.9146916240237005,
      "grad_norm": 0.5536824577234217,
      "learning_rate": 5.668531109409851e-06,
      "loss": 0.37335591316223143,
      "num_input_tokens_seen": 699074480,
      "step": 13585,
      "train_runtime": 83653.5705,
      "train_tokens_per_second": 8356.78
    },
    {
      "epoch": 0.9150282790196607,
      "grad_norm": 0.7046621523450236,
      "learning_rate": 5.665910675082362e-06,
      "loss": 0.3552588224411011,
      "num_input_tokens_seen": 699329680,
      "step": 13590,
      "train_runtime": 83684.7479,
      "train_tokens_per_second": 8356.716
    },
    {
      "epoch": 0.9153649340156208,
      "grad_norm": 0.6173463986080291,
      "learning_rate": 5.663290054535078e-06,
      "loss": 0.33337738513946535,
      "num_input_tokens_seen": 699579224,
      "step": 13595,
      "train_runtime": 83714.9472,
      "train_tokens_per_second": 8356.682
    },
    {
      "epoch": 0.9157015890115809,
      "grad_norm": 0.7795234892695109,
      "learning_rate": 5.660669248500846e-06,
      "loss": 0.38270201683044436,
      "num_input_tokens_seen": 699841232,
      "step": 13600,
      "train_runtime": 83745.4644,
      "train_tokens_per_second": 8356.766
    },
    {
      "epoch": 0.9160382440075411,
      "grad_norm": 0.6314114125213657,
      "learning_rate": 5.658048257712563e-06,
      "loss": 0.3439387321472168,
      "num_input_tokens_seen": 700094408,
      "step": 13605,
      "train_runtime": 83774.897,
      "train_tokens_per_second": 8356.852
    },
    {
      "epoch": 0.9163748990035012,
      "grad_norm": 0.5711464586290839,
      "learning_rate": 5.655427082903184e-06,
      "loss": 0.3569528579711914,
      "num_input_tokens_seen": 700350872,
      "step": 13610,
      "train_runtime": 83807.5052,
      "train_tokens_per_second": 8356.661
    },
    {
      "epoch": 0.9167115539994614,
      "grad_norm": 0.7333933219341897,
      "learning_rate": 5.652805724805711e-06,
      "loss": 0.3550983428955078,
      "num_input_tokens_seen": 700611976,
      "step": 13615,
      "train_runtime": 83836.5213,
      "train_tokens_per_second": 8356.883
    },
    {
      "epoch": 0.9170482089954215,
      "grad_norm": 0.7723623466088162,
      "learning_rate": 5.650184184153199e-06,
      "loss": 0.3645094156265259,
      "num_input_tokens_seen": 700875248,
      "step": 13620,
      "train_runtime": 83870.9285,
      "train_tokens_per_second": 8356.593
    },
    {
      "epoch": 0.9173848639913816,
      "grad_norm": 0.6165251705305488,
      "learning_rate": 5.647562461678754e-06,
      "loss": 0.39477858543395994,
      "num_input_tokens_seen": 701147976,
      "step": 13625,
      "train_runtime": 83901.1786,
      "train_tokens_per_second": 8356.831
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.6025639690856172,
      "learning_rate": 5.64494055811553e-06,
      "loss": 0.3476600408554077,
      "num_input_tokens_seen": 701396512,
      "step": 13630,
      "train_runtime": 83933.8281,
      "train_tokens_per_second": 8356.541
    },
    {
      "epoch": 0.9180581739833019,
      "grad_norm": 0.6553005158678815,
      "learning_rate": 5.642318474196737e-06,
      "loss": 0.3417848587036133,
      "num_input_tokens_seen": 701666272,
      "step": 13635,
      "train_runtime": 83964.1746,
      "train_tokens_per_second": 8356.734
    },
    {
      "epoch": 0.9183948289792621,
      "grad_norm": 0.6242627728951042,
      "learning_rate": 5.63969621065563e-06,
      "loss": 0.3576439142227173,
      "num_input_tokens_seen": 701919520,
      "step": 13640,
      "train_runtime": 83991.5972,
      "train_tokens_per_second": 8357.021
    },
    {
      "epoch": 0.9187314839752222,
      "grad_norm": 0.5848449554463806,
      "learning_rate": 5.637073768225517e-06,
      "loss": 0.3263015031814575,
      "num_input_tokens_seen": 702161576,
      "step": 13645,
      "train_runtime": 84022.4733,
      "train_tokens_per_second": 8356.831
    },
    {
      "epoch": 0.9190681389711823,
      "grad_norm": 0.6088435036534938,
      "learning_rate": 5.634451147639758e-06,
      "loss": 0.3304743766784668,
      "num_input_tokens_seen": 702388928,
      "step": 13650,
      "train_runtime": 84055.1651,
      "train_tokens_per_second": 8356.285
    },
    {
      "epoch": 0.9194047939671425,
      "grad_norm": 0.5915969485516099,
      "learning_rate": 5.631828349631757e-06,
      "loss": 0.3419053554534912,
      "num_input_tokens_seen": 702652152,
      "step": 13655,
      "train_runtime": 84089.1261,
      "train_tokens_per_second": 8356.041
    },
    {
      "epoch": 0.9197414489631026,
      "grad_norm": 0.6356227188332457,
      "learning_rate": 5.629205374934974e-06,
      "loss": 0.3411503076553345,
      "num_input_tokens_seen": 702900320,
      "step": 13660,
      "train_runtime": 84116.3829,
      "train_tokens_per_second": 8356.283
    },
    {
      "epoch": 0.9200781039590628,
      "grad_norm": 0.6518606012520866,
      "learning_rate": 5.626582224282913e-06,
      "loss": 0.36418373584747316,
      "num_input_tokens_seen": 703163728,
      "step": 13665,
      "train_runtime": 84147.157,
      "train_tokens_per_second": 8356.358
    },
    {
      "epoch": 0.9204147589550229,
      "grad_norm": 0.632723644516242,
      "learning_rate": 5.623958898409132e-06,
      "loss": 0.39714875221252444,
      "num_input_tokens_seen": 703436080,
      "step": 13670,
      "train_runtime": 84175.6672,
      "train_tokens_per_second": 8356.763
    },
    {
      "epoch": 0.920751413950983,
      "grad_norm": 0.6622887952631968,
      "learning_rate": 5.621335398047234e-06,
      "loss": 0.38545827865600585,
      "num_input_tokens_seen": 703687368,
      "step": 13675,
      "train_runtime": 84204.3348,
      "train_tokens_per_second": 8356.902
    },
    {
      "epoch": 0.9210880689469432,
      "grad_norm": 0.6210162187706881,
      "learning_rate": 5.618711723930874e-06,
      "loss": 0.3702589750289917,
      "num_input_tokens_seen": 703953616,
      "step": 13680,
      "train_runtime": 84234.836,
      "train_tokens_per_second": 8357.037
    },
    {
      "epoch": 0.9214247239429033,
      "grad_norm": 0.5999035588244432,
      "learning_rate": 5.616087876793753e-06,
      "loss": 0.3752654314041138,
      "num_input_tokens_seen": 704234024,
      "step": 13685,
      "train_runtime": 84265.8786,
      "train_tokens_per_second": 8357.286
    },
    {
      "epoch": 0.9217613789388635,
      "grad_norm": 0.6004637591287154,
      "learning_rate": 5.613463857369621e-06,
      "loss": 0.3636165618896484,
      "num_input_tokens_seen": 704496664,
      "step": 13690,
      "train_runtime": 84293.8234,
      "train_tokens_per_second": 8357.631
    },
    {
      "epoch": 0.9220980339348236,
      "grad_norm": 0.6382231377612089,
      "learning_rate": 5.610839666392278e-06,
      "loss": 0.3471952438354492,
      "num_input_tokens_seen": 704748744,
      "step": 13695,
      "train_runtime": 84324.7937,
      "train_tokens_per_second": 8357.551
    },
    {
      "epoch": 0.9224346889307837,
      "grad_norm": 0.699422635241344,
      "learning_rate": 5.608215304595571e-06,
      "loss": 0.3674611568450928,
      "num_input_tokens_seen": 704992512,
      "step": 13700,
      "train_runtime": 84351.2193,
      "train_tokens_per_second": 8357.822
    },
    {
      "epoch": 0.9227713439267439,
      "grad_norm": 0.7104146285520443,
      "learning_rate": 5.605590772713393e-06,
      "loss": 0.3484257698059082,
      "num_input_tokens_seen": 705206872,
      "step": 13705,
      "train_runtime": 84378.6607,
      "train_tokens_per_second": 8357.645
    },
    {
      "epoch": 0.923107998922704,
      "grad_norm": 0.7231077469044681,
      "learning_rate": 5.602966071479687e-06,
      "loss": 0.3604652643203735,
      "num_input_tokens_seen": 705448368,
      "step": 13710,
      "train_runtime": 84405.8101,
      "train_tokens_per_second": 8357.818
    },
    {
      "epoch": 0.9234446539186641,
      "grad_norm": 0.5686733053226621,
      "learning_rate": 5.600341201628439e-06,
      "loss": 0.3624136924743652,
      "num_input_tokens_seen": 705720936,
      "step": 13715,
      "train_runtime": 84439.3213,
      "train_tokens_per_second": 8357.729
    },
    {
      "epoch": 0.9237813089146243,
      "grad_norm": 0.5671753288844125,
      "learning_rate": 5.597716163893692e-06,
      "loss": 0.30830399990081786,
      "num_input_tokens_seen": 705991872,
      "step": 13720,
      "train_runtime": 84466.935,
      "train_tokens_per_second": 8358.204
    },
    {
      "epoch": 0.9241179639105844,
      "grad_norm": 0.6987340921636698,
      "learning_rate": 5.595090959009525e-06,
      "loss": 0.3380585670471191,
      "num_input_tokens_seen": 706252928,
      "step": 13725,
      "train_runtime": 84497.1116,
      "train_tokens_per_second": 8358.309
    },
    {
      "epoch": 0.9244546189065446,
      "grad_norm": 0.5518380365386968,
      "learning_rate": 5.592465587710068e-06,
      "loss": 0.3300445556640625,
      "num_input_tokens_seen": 706542064,
      "step": 13730,
      "train_runtime": 84526.3481,
      "train_tokens_per_second": 8358.838
    },
    {
      "epoch": 0.9247912739025047,
      "grad_norm": 0.526886577380852,
      "learning_rate": 5.589840050729498e-06,
      "loss": 0.3172940969467163,
      "num_input_tokens_seen": 706807632,
      "step": 13735,
      "train_runtime": 84556.5101,
      "train_tokens_per_second": 8358.997
    },
    {
      "epoch": 0.9251279288984648,
      "grad_norm": 0.7553012755217439,
      "learning_rate": 5.587214348802039e-06,
      "loss": 0.3852890729904175,
      "num_input_tokens_seen": 707050864,
      "step": 13740,
      "train_runtime": 84585.3871,
      "train_tokens_per_second": 8359.019
    },
    {
      "epoch": 0.925464583894425,
      "grad_norm": 0.6106811763772959,
      "learning_rate": 5.5845884826619615e-06,
      "loss": 0.3656655788421631,
      "num_input_tokens_seen": 707327168,
      "step": 13745,
      "train_runtime": 84613.9632,
      "train_tokens_per_second": 8359.462
    },
    {
      "epoch": 0.9258012388903851,
      "grad_norm": 0.6066104500687765,
      "learning_rate": 5.5819624530435775e-06,
      "loss": 0.3613640308380127,
      "num_input_tokens_seen": 707565848,
      "step": 13750,
      "train_runtime": 84642.3841,
      "train_tokens_per_second": 8359.474
    },
    {
      "epoch": 0.9261378938863453,
      "grad_norm": 0.6844312631167492,
      "learning_rate": 5.57933626068125e-06,
      "loss": 0.39635381698608396,
      "num_input_tokens_seen": 707823008,
      "step": 13755,
      "train_runtime": 84670.177,
      "train_tokens_per_second": 8359.768
    },
    {
      "epoch": 0.9264745488823054,
      "grad_norm": 0.6199377450988277,
      "learning_rate": 5.5767099063093835e-06,
      "loss": 0.35148770809173585,
      "num_input_tokens_seen": 708097936,
      "step": 13760,
      "train_runtime": 84703.1277,
      "train_tokens_per_second": 8359.761
    },
    {
      "epoch": 0.9268112038782655,
      "grad_norm": 0.7733153861535556,
      "learning_rate": 5.574083390662431e-06,
      "loss": 0.3512766361236572,
      "num_input_tokens_seen": 708376952,
      "step": 13765,
      "train_runtime": 84734.7587,
      "train_tokens_per_second": 8359.934
    },
    {
      "epoch": 0.9271478588742257,
      "grad_norm": 0.764985223027613,
      "learning_rate": 5.57145671447489e-06,
      "loss": 0.36069037914276125,
      "num_input_tokens_seen": 708621064,
      "step": 13770,
      "train_runtime": 84764.3221,
      "train_tokens_per_second": 8359.898
    },
    {
      "epoch": 0.9274845138701858,
      "grad_norm": 0.6310638950523032,
      "learning_rate": 5.5688298784813e-06,
      "loss": 0.3971395015716553,
      "num_input_tokens_seen": 708883248,
      "step": 13775,
      "train_runtime": 84796.9674,
      "train_tokens_per_second": 8359.771
    },
    {
      "epoch": 0.927821168866146,
      "grad_norm": 0.7856414787684248,
      "learning_rate": 5.566202883416249e-06,
      "loss": 0.3961209535598755,
      "num_input_tokens_seen": 709159384,
      "step": 13780,
      "train_runtime": 84829.051,
      "train_tokens_per_second": 8359.865
    },
    {
      "epoch": 0.9281578238621061,
      "grad_norm": 0.718893804336572,
      "learning_rate": 5.563575730014366e-06,
      "loss": 0.38128087520599363,
      "num_input_tokens_seen": 709413040,
      "step": 13785,
      "train_runtime": 84858.6002,
      "train_tokens_per_second": 8359.943
    },
    {
      "epoch": 0.9284944788580662,
      "grad_norm": 0.6966228567915038,
      "learning_rate": 5.560948419010328e-06,
      "loss": 0.3499451160430908,
      "num_input_tokens_seen": 709675136,
      "step": 13790,
      "train_runtime": 84884.1929,
      "train_tokens_per_second": 8360.51
    },
    {
      "epoch": 0.9288311338540264,
      "grad_norm": 0.6849527239142031,
      "learning_rate": 5.558320951138853e-06,
      "loss": 0.3309880256652832,
      "num_input_tokens_seen": 709910344,
      "step": 13795,
      "train_runtime": 84913.7124,
      "train_tokens_per_second": 8360.373
    },
    {
      "epoch": 0.9291677888499865,
      "grad_norm": 0.714273984373423,
      "learning_rate": 5.5556933271347046e-06,
      "loss": 0.36818068027496337,
      "num_input_tokens_seen": 710146792,
      "step": 13800,
      "train_runtime": 84942.1196,
      "train_tokens_per_second": 8360.361
    },
    {
      "epoch": 0.9295044438459467,
      "grad_norm": 0.6333850850025499,
      "learning_rate": 5.55306554773269e-06,
      "loss": 0.37214179039001466,
      "num_input_tokens_seen": 710385832,
      "step": 13805,
      "train_runtime": 84975.9981,
      "train_tokens_per_second": 8359.841
    },
    {
      "epoch": 0.9298410988419068,
      "grad_norm": 0.8756047021008705,
      "learning_rate": 5.550437613667658e-06,
      "loss": 0.3601330995559692,
      "num_input_tokens_seen": 710598968,
      "step": 13810,
      "train_runtime": 85004.2904,
      "train_tokens_per_second": 8359.566
    },
    {
      "epoch": 0.9301777538378669,
      "grad_norm": 0.6286365662330439,
      "learning_rate": 5.547809525674501e-06,
      "loss": 0.35212481021881104,
      "num_input_tokens_seen": 710839120,
      "step": 13815,
      "train_runtime": 85034.6593,
      "train_tokens_per_second": 8359.405
    },
    {
      "epoch": 0.9305144088338271,
      "grad_norm": 0.7137853369611423,
      "learning_rate": 5.545181284488159e-06,
      "loss": 0.38329458236694336,
      "num_input_tokens_seen": 711113640,
      "step": 13820,
      "train_runtime": 85069.0755,
      "train_tokens_per_second": 8359.25
    },
    {
      "epoch": 0.9308510638297872,
      "grad_norm": 0.72269540061033,
      "learning_rate": 5.5425528908436085e-06,
      "loss": 0.3534586429595947,
      "num_input_tokens_seen": 711345304,
      "step": 13825,
      "train_runtime": 85099.5845,
      "train_tokens_per_second": 8358.975
    },
    {
      "epoch": 0.9311877188257474,
      "grad_norm": 0.62853715109335,
      "learning_rate": 5.539924345475873e-06,
      "loss": 0.3750478744506836,
      "num_input_tokens_seen": 711573904,
      "step": 13830,
      "train_runtime": 85126.3628,
      "train_tokens_per_second": 8359.031
    },
    {
      "epoch": 0.9315243738217075,
      "grad_norm": 0.675859664863836,
      "learning_rate": 5.5372956491200145e-06,
      "loss": 0.3734222412109375,
      "num_input_tokens_seen": 711829104,
      "step": 13835,
      "train_runtime": 85153.7778,
      "train_tokens_per_second": 8359.337
    },
    {
      "epoch": 0.9318610288176676,
      "grad_norm": 0.9003956511243558,
      "learning_rate": 5.534666802511143e-06,
      "loss": 0.3576240539550781,
      "num_input_tokens_seen": 712058352,
      "step": 13840,
      "train_runtime": 85182.1994,
      "train_tokens_per_second": 8359.239
    },
    {
      "epoch": 0.9321976838136278,
      "grad_norm": 0.7246618138958477,
      "learning_rate": 5.532037806384404e-06,
      "loss": 0.37378973960876466,
      "num_input_tokens_seen": 712334984,
      "step": 13845,
      "train_runtime": 85211.1407,
      "train_tokens_per_second": 8359.646
    },
    {
      "epoch": 0.9325343388095879,
      "grad_norm": 0.6723459129364898,
      "learning_rate": 5.529408661474989e-06,
      "loss": 0.3412851572036743,
      "num_input_tokens_seen": 712555848,
      "step": 13850,
      "train_runtime": 85239.7405,
      "train_tokens_per_second": 8359.432
    },
    {
      "epoch": 0.932870993805548,
      "grad_norm": 0.5815520598384242,
      "learning_rate": 5.526779368518129e-06,
      "loss": 0.3767396926879883,
      "num_input_tokens_seen": 712827264,
      "step": 13855,
      "train_runtime": 85270.6559,
      "train_tokens_per_second": 8359.585
    },
    {
      "epoch": 0.9332076488015082,
      "grad_norm": 0.7027072687773963,
      "learning_rate": 5.5241499282491e-06,
      "loss": 0.35355525016784667,
      "num_input_tokens_seen": 713056376,
      "step": 13860,
      "train_runtime": 85305.4827,
      "train_tokens_per_second": 8358.858
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.6457704729993184,
      "learning_rate": 5.5215203414032135e-06,
      "loss": 0.3357942819595337,
      "num_input_tokens_seen": 713318160,
      "step": 13865,
      "train_runtime": 85335.5073,
      "train_tokens_per_second": 8358.984
    },
    {
      "epoch": 0.9338809587934285,
      "grad_norm": 0.6437526893817489,
      "learning_rate": 5.518890608715828e-06,
      "loss": 0.3578556299209595,
      "num_input_tokens_seen": 713564248,
      "step": 13870,
      "train_runtime": 85365.2378,
      "train_tokens_per_second": 8358.956
    },
    {
      "epoch": 0.9342176137893886,
      "grad_norm": 0.7543089310790887,
      "learning_rate": 5.516260730922337e-06,
      "loss": 0.37209558486938477,
      "num_input_tokens_seen": 713800176,
      "step": 13875,
      "train_runtime": 85395.2615,
      "train_tokens_per_second": 8358.78
    },
    {
      "epoch": 0.9345542687853488,
      "grad_norm": 0.5642489668785322,
      "learning_rate": 5.51363070875818e-06,
      "loss": 0.3516117811203003,
      "num_input_tokens_seen": 714055360,
      "step": 13880,
      "train_runtime": 85421.247,
      "train_tokens_per_second": 8359.224
    },
    {
      "epoch": 0.9348909237813089,
      "grad_norm": 0.6946055988242056,
      "learning_rate": 5.511000542958832e-06,
      "loss": 0.37625324726104736,
      "num_input_tokens_seen": 714302328,
      "step": 13885,
      "train_runtime": 85452.2699,
      "train_tokens_per_second": 8359.08
    },
    {
      "epoch": 0.935227578777269,
      "grad_norm": 0.7022100296429534,
      "learning_rate": 5.508370234259812e-06,
      "loss": 0.35982165336608884,
      "num_input_tokens_seen": 714542968,
      "step": 13890,
      "train_runtime": 85482.1463,
      "train_tokens_per_second": 8358.973
    },
    {
      "epoch": 0.9355642337732292,
      "grad_norm": 0.6237046487275945,
      "learning_rate": 5.505739783396678e-06,
      "loss": 0.4107686519622803,
      "num_input_tokens_seen": 714787688,
      "step": 13895,
      "train_runtime": 85510.0151,
      "train_tokens_per_second": 8359.111
    },
    {
      "epoch": 0.9359008887691893,
      "grad_norm": 0.687076769634906,
      "learning_rate": 5.503109191105026e-06,
      "loss": 0.36827161312103274,
      "num_input_tokens_seen": 715032760,
      "step": 13900,
      "train_runtime": 85541.491,
      "train_tokens_per_second": 8358.9
    },
    {
      "epoch": 0.9362375437651494,
      "grad_norm": 0.6087965544943061,
      "learning_rate": 5.500478458120493e-06,
      "loss": 0.36850454807281496,
      "num_input_tokens_seen": 715309080,
      "step": 13905,
      "train_runtime": 85572.6857,
      "train_tokens_per_second": 8359.082
    },
    {
      "epoch": 0.9365741987611096,
      "grad_norm": 0.5892906087628781,
      "learning_rate": 5.497847585178754e-06,
      "loss": 0.37303524017333983,
      "num_input_tokens_seen": 715570752,
      "step": 13910,
      "train_runtime": 85606.7957,
      "train_tokens_per_second": 8358.808
    },
    {
      "epoch": 0.9369108537570697,
      "grad_norm": 0.6449938548418706,
      "learning_rate": 5.495216573015526e-06,
      "loss": 0.33706130981445315,
      "num_input_tokens_seen": 715808296,
      "step": 13915,
      "train_runtime": 85635.8257,
      "train_tokens_per_second": 8358.748
    },
    {
      "epoch": 0.9372475087530299,
      "grad_norm": 0.6977441607931472,
      "learning_rate": 5.492585422366562e-06,
      "loss": 0.33948483467102053,
      "num_input_tokens_seen": 716053496,
      "step": 13920,
      "train_runtime": 85663.9856,
      "train_tokens_per_second": 8358.863
    },
    {
      "epoch": 0.93758416374899,
      "grad_norm": 0.6058794751047483,
      "learning_rate": 5.489954133967657e-06,
      "loss": 0.3295562505722046,
      "num_input_tokens_seen": 716292144,
      "step": 13925,
      "train_runtime": 85698.3552,
      "train_tokens_per_second": 8358.295
    },
    {
      "epoch": 0.9379208187449501,
      "grad_norm": 0.5877859108055806,
      "learning_rate": 5.48732270855464e-06,
      "loss": 0.33592839241027833,
      "num_input_tokens_seen": 716564576,
      "step": 13930,
      "train_runtime": 85728.7301,
      "train_tokens_per_second": 8358.511
    },
    {
      "epoch": 0.9382574737409103,
      "grad_norm": 0.7041710234125325,
      "learning_rate": 5.48469114686338e-06,
      "loss": 0.34801621437072755,
      "num_input_tokens_seen": 716800040,
      "step": 13935,
      "train_runtime": 85759.2845,
      "train_tokens_per_second": 8358.279
    },
    {
      "epoch": 0.9385941287368704,
      "grad_norm": 0.6084909212103816,
      "learning_rate": 5.482059449629788e-06,
      "loss": 0.34277198314666746,
      "num_input_tokens_seen": 717068648,
      "step": 13940,
      "train_runtime": 85790.9697,
      "train_tokens_per_second": 8358.323
    },
    {
      "epoch": 0.9389307837328306,
      "grad_norm": 0.5836880594854189,
      "learning_rate": 5.479427617589808e-06,
      "loss": 0.3264045715332031,
      "num_input_tokens_seen": 717327592,
      "step": 13945,
      "train_runtime": 85821.7448,
      "train_tokens_per_second": 8358.343
    },
    {
      "epoch": 0.9392674387287907,
      "grad_norm": 0.540846766704371,
      "learning_rate": 5.4767956514794226e-06,
      "loss": 0.36751489639282225,
      "num_input_tokens_seen": 717606424,
      "step": 13950,
      "train_runtime": 85854.6964,
      "train_tokens_per_second": 8358.383
    },
    {
      "epoch": 0.9396040937247508,
      "grad_norm": 0.7938836563548223,
      "learning_rate": 5.474163552034655e-06,
      "loss": 0.34305806159973146,
      "num_input_tokens_seen": 717846576,
      "step": 13955,
      "train_runtime": 85886.7482,
      "train_tokens_per_second": 8358.06
    },
    {
      "epoch": 0.939940748720711,
      "grad_norm": 0.7648841591211527,
      "learning_rate": 5.4715313199915595e-06,
      "loss": 0.33446311950683594,
      "num_input_tokens_seen": 718086824,
      "step": 13960,
      "train_runtime": 85918.3467,
      "train_tokens_per_second": 8357.782
    },
    {
      "epoch": 0.9402774037166711,
      "grad_norm": 0.7314906059693517,
      "learning_rate": 5.468898956086236e-06,
      "loss": 0.3605980157852173,
      "num_input_tokens_seen": 718358320,
      "step": 13965,
      "train_runtime": 85946.7812,
      "train_tokens_per_second": 8358.176
    },
    {
      "epoch": 0.9406140587126313,
      "grad_norm": 0.61336491462741,
      "learning_rate": 5.466266461054815e-06,
      "loss": 0.362291145324707,
      "num_input_tokens_seen": 718622888,
      "step": 13970,
      "train_runtime": 85975.3795,
      "train_tokens_per_second": 8358.473
    },
    {
      "epoch": 0.9409507137085914,
      "grad_norm": 0.7433814417394373,
      "learning_rate": 5.463633835633464e-06,
      "loss": 0.3795844554901123,
      "num_input_tokens_seen": 718903208,
      "step": 13975,
      "train_runtime": 86008.8841,
      "train_tokens_per_second": 8358.476
    },
    {
      "epoch": 0.9412873687045515,
      "grad_norm": 0.5995027897843211,
      "learning_rate": 5.46100108055839e-06,
      "loss": 0.35777459144592283,
      "num_input_tokens_seen": 719178416,
      "step": 13980,
      "train_runtime": 86040.1744,
      "train_tokens_per_second": 8358.635
    },
    {
      "epoch": 0.9416240237005117,
      "grad_norm": 0.6169309461635175,
      "learning_rate": 5.458368196565833e-06,
      "loss": 0.33479578495025636,
      "num_input_tokens_seen": 719436792,
      "step": 13985,
      "train_runtime": 86069.0303,
      "train_tokens_per_second": 8358.835
    },
    {
      "epoch": 0.9419606786964718,
      "grad_norm": 0.5849935631956745,
      "learning_rate": 5.455735184392073e-06,
      "loss": 0.3792692184448242,
      "num_input_tokens_seen": 719711336,
      "step": 13990,
      "train_runtime": 86098.782,
      "train_tokens_per_second": 8359.135
    },
    {
      "epoch": 0.942297333692432,
      "grad_norm": 0.7023143590154773,
      "learning_rate": 5.4531020447734204e-06,
      "loss": 0.3533689022064209,
      "num_input_tokens_seen": 719964304,
      "step": 13995,
      "train_runtime": 86128.6326,
      "train_tokens_per_second": 8359.175
    },
    {
      "epoch": 0.9426339886883921,
      "grad_norm": 0.6277999988326576,
      "learning_rate": 5.450468778446226e-06,
      "loss": 0.35790510177612306,
      "num_input_tokens_seen": 720238944,
      "step": 14000,
      "train_runtime": 86156.8195,
      "train_tokens_per_second": 8359.628
    },
    {
      "epoch": 0.9429706436843522,
      "grad_norm": 0.7210742478980305,
      "learning_rate": 5.447835386146875e-06,
      "loss": 0.35814385414123534,
      "num_input_tokens_seen": 720516784,
      "step": 14005,
      "train_runtime": 86184.5909,
      "train_tokens_per_second": 8360.158
    },
    {
      "epoch": 0.9433072986803124,
      "grad_norm": 0.6616121074006898,
      "learning_rate": 5.445201868611784e-06,
      "loss": 0.3976274013519287,
      "num_input_tokens_seen": 720760616,
      "step": 14010,
      "train_runtime": 86213.6128,
      "train_tokens_per_second": 8360.172
    },
    {
      "epoch": 0.9436439536762725,
      "grad_norm": 0.5983596102367572,
      "learning_rate": 5.442568226577413e-06,
      "loss": 0.3653991222381592,
      "num_input_tokens_seen": 721018392,
      "step": 14015,
      "train_runtime": 86245.6548,
      "train_tokens_per_second": 8360.055
    },
    {
      "epoch": 0.9439806086722327,
      "grad_norm": 0.6085005187180216,
      "learning_rate": 5.439934460780247e-06,
      "loss": 0.3373982906341553,
      "num_input_tokens_seen": 721287424,
      "step": 14020,
      "train_runtime": 86276.081,
      "train_tokens_per_second": 8360.225
    },
    {
      "epoch": 0.9443172636681928,
      "grad_norm": 0.6284948877098328,
      "learning_rate": 5.437300571956813e-06,
      "loss": 0.36469502449035646,
      "num_input_tokens_seen": 721519432,
      "step": 14025,
      "train_runtime": 86302.0729,
      "train_tokens_per_second": 8360.395
    },
    {
      "epoch": 0.9446539186641529,
      "grad_norm": 0.574735887821787,
      "learning_rate": 5.434666560843667e-06,
      "loss": 0.36296539306640624,
      "num_input_tokens_seen": 721764936,
      "step": 14030,
      "train_runtime": 86329.2713,
      "train_tokens_per_second": 8360.605
    },
    {
      "epoch": 0.9449905736601131,
      "grad_norm": 0.6334548211225146,
      "learning_rate": 5.432032428177405e-06,
      "loss": 0.36661243438720703,
      "num_input_tokens_seen": 722031296,
      "step": 14035,
      "train_runtime": 86364.8221,
      "train_tokens_per_second": 8360.248
    },
    {
      "epoch": 0.9453272286560732,
      "grad_norm": 0.6571459423961725,
      "learning_rate": 5.429398174694651e-06,
      "loss": 0.35982475280761717,
      "num_input_tokens_seen": 722316400,
      "step": 14040,
      "train_runtime": 86398.5782,
      "train_tokens_per_second": 8360.281
    },
    {
      "epoch": 0.9456638836520334,
      "grad_norm": 0.6247176010698106,
      "learning_rate": 5.426763801132067e-06,
      "loss": 0.35411183834075927,
      "num_input_tokens_seen": 722528640,
      "step": 14045,
      "train_runtime": 86431.0963,
      "train_tokens_per_second": 8359.591
    },
    {
      "epoch": 0.9460005386479935,
      "grad_norm": 0.7083514915609673,
      "learning_rate": 5.424129308226347e-06,
      "loss": 0.3723160743713379,
      "num_input_tokens_seen": 722768080,
      "step": 14050,
      "train_runtime": 86458.1405,
      "train_tokens_per_second": 8359.746
    },
    {
      "epoch": 0.9463371936439536,
      "grad_norm": 0.6202174100328659,
      "learning_rate": 5.421494696714219e-06,
      "loss": 0.36843457221984866,
      "num_input_tokens_seen": 723060920,
      "step": 14055,
      "train_runtime": 86486.4596,
      "train_tokens_per_second": 8360.394
    },
    {
      "epoch": 0.9466738486399138,
      "grad_norm": 0.5997943768449644,
      "learning_rate": 5.418859967332442e-06,
      "loss": 0.3628130674362183,
      "num_input_tokens_seen": 723332240,
      "step": 14060,
      "train_runtime": 86519.0018,
      "train_tokens_per_second": 8360.386
    },
    {
      "epoch": 0.9470105036358739,
      "grad_norm": 0.8757422937471547,
      "learning_rate": 5.416225120817811e-06,
      "loss": 0.36904425621032716,
      "num_input_tokens_seen": 723551640,
      "step": 14065,
      "train_runtime": 86548.5863,
      "train_tokens_per_second": 8360.063
    },
    {
      "epoch": 0.947347158631834,
      "grad_norm": 0.5604317822082541,
      "learning_rate": 5.413590157907153e-06,
      "loss": 0.34093663692474363,
      "num_input_tokens_seen": 723850192,
      "step": 14070,
      "train_runtime": 86581.0313,
      "train_tokens_per_second": 8360.378
    },
    {
      "epoch": 0.9476838136277942,
      "grad_norm": 0.5868582010527802,
      "learning_rate": 5.4109550793373255e-06,
      "loss": 0.3498985767364502,
      "num_input_tokens_seen": 724107672,
      "step": 14075,
      "train_runtime": 86611.1563,
      "train_tokens_per_second": 8360.443
    },
    {
      "epoch": 0.9480204686237543,
      "grad_norm": 0.5571920822326955,
      "learning_rate": 5.40831988584522e-06,
      "loss": 0.36889612674713135,
      "num_input_tokens_seen": 724382616,
      "step": 14080,
      "train_runtime": 86642.2426,
      "train_tokens_per_second": 8360.617
    },
    {
      "epoch": 0.9483571236197145,
      "grad_norm": 0.6349056586621602,
      "learning_rate": 5.40568457816776e-06,
      "loss": 0.3716681957244873,
      "num_input_tokens_seen": 724656136,
      "step": 14085,
      "train_runtime": 86674.6601,
      "train_tokens_per_second": 8360.646
    },
    {
      "epoch": 0.9486937786156746,
      "grad_norm": 0.543020645491933,
      "learning_rate": 5.403049157041903e-06,
      "loss": 0.3487908363342285,
      "num_input_tokens_seen": 724917560,
      "step": 14090,
      "train_runtime": 86706.1614,
      "train_tokens_per_second": 8360.623
    },
    {
      "epoch": 0.9490304336116347,
      "grad_norm": 0.6874745360187002,
      "learning_rate": 5.400413623204635e-06,
      "loss": 0.34180307388305664,
      "num_input_tokens_seen": 725168424,
      "step": 14095,
      "train_runtime": 86733.0998,
      "train_tokens_per_second": 8360.919
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.6516504316653042,
      "learning_rate": 5.397777977392973e-06,
      "loss": 0.3835340976715088,
      "num_input_tokens_seen": 725417928,
      "step": 14100,
      "train_runtime": 86764.4317,
      "train_tokens_per_second": 8360.775
    },
    {
      "epoch": 0.949703743603555,
      "grad_norm": 0.6370008024888683,
      "learning_rate": 5.395142220343967e-06,
      "loss": 0.3734624147415161,
      "num_input_tokens_seen": 725673992,
      "step": 14105,
      "train_runtime": 86798.9615,
      "train_tokens_per_second": 8360.399
    },
    {
      "epoch": 0.9500403985995152,
      "grad_norm": 0.6366939272355171,
      "learning_rate": 5.3925063527947e-06,
      "loss": 0.3412612438201904,
      "num_input_tokens_seen": 725917504,
      "step": 14110,
      "train_runtime": 86827.7909,
      "train_tokens_per_second": 8360.428
    },
    {
      "epoch": 0.9503770535954753,
      "grad_norm": 0.7579134758733556,
      "learning_rate": 5.389870375482283e-06,
      "loss": 0.3830456495285034,
      "num_input_tokens_seen": 726145272,
      "step": 14115,
      "train_runtime": 86856.0046,
      "train_tokens_per_second": 8360.335
    },
    {
      "epoch": 0.9507137085914354,
      "grad_norm": 0.5958805602713254,
      "learning_rate": 5.387234289143859e-06,
      "loss": 0.32933456897735597,
      "num_input_tokens_seen": 726400240,
      "step": 14120,
      "train_runtime": 86890.0082,
      "train_tokens_per_second": 8359.997
    },
    {
      "epoch": 0.9510503635873956,
      "grad_norm": 0.6368490001569849,
      "learning_rate": 5.384598094516601e-06,
      "loss": 0.36543726921081543,
      "num_input_tokens_seen": 726663616,
      "step": 14125,
      "train_runtime": 86921.5657,
      "train_tokens_per_second": 8359.992
    },
    {
      "epoch": 0.9513870185833557,
      "grad_norm": 0.6251902688319988,
      "learning_rate": 5.381961792337712e-06,
      "loss": 0.3773162841796875,
      "num_input_tokens_seen": 726937488,
      "step": 14130,
      "train_runtime": 86949.4797,
      "train_tokens_per_second": 8360.458
    },
    {
      "epoch": 0.9517236735793159,
      "grad_norm": 0.529620916813458,
      "learning_rate": 5.3793253833444245e-06,
      "loss": 0.32413268089294434,
      "num_input_tokens_seen": 727217080,
      "step": 14135,
      "train_runtime": 86981.7654,
      "train_tokens_per_second": 8360.569
    },
    {
      "epoch": 0.952060328575276,
      "grad_norm": 0.5911820009936902,
      "learning_rate": 5.3766888682740045e-06,
      "loss": 0.3413226127624512,
      "num_input_tokens_seen": 727486352,
      "step": 14140,
      "train_runtime": 87011.1518,
      "train_tokens_per_second": 8360.84
    },
    {
      "epoch": 0.9523969835712361,
      "grad_norm": 0.6409359778541225,
      "learning_rate": 5.374052247863745e-06,
      "loss": 0.36683523654937744,
      "num_input_tokens_seen": 727738784,
      "step": 14145,
      "train_runtime": 87044.6622,
      "train_tokens_per_second": 8360.522
    },
    {
      "epoch": 0.9527336385671963,
      "grad_norm": 0.8312829198369863,
      "learning_rate": 5.371415522850966e-06,
      "loss": 0.3453713893890381,
      "num_input_tokens_seen": 727989488,
      "step": 14150,
      "train_runtime": 87074.1796,
      "train_tokens_per_second": 8360.567
    },
    {
      "epoch": 0.9530702935631564,
      "grad_norm": 0.6423465996208764,
      "learning_rate": 5.368778693973019e-06,
      "loss": 0.33897373676300047,
      "num_input_tokens_seen": 728239472,
      "step": 14155,
      "train_runtime": 87103.3094,
      "train_tokens_per_second": 8360.641
    },
    {
      "epoch": 0.9534069485591166,
      "grad_norm": 0.7918282852884608,
      "learning_rate": 5.366141761967287e-06,
      "loss": 0.3669609069824219,
      "num_input_tokens_seen": 728494240,
      "step": 14160,
      "train_runtime": 87133.1968,
      "train_tokens_per_second": 8360.697
    },
    {
      "epoch": 0.9537436035550767,
      "grad_norm": 0.7811001229225133,
      "learning_rate": 5.363504727571179e-06,
      "loss": 0.3758628606796265,
      "num_input_tokens_seen": 728750816,
      "step": 14165,
      "train_runtime": 87169.1989,
      "train_tokens_per_second": 8360.187
    },
    {
      "epoch": 0.954080258551037,
      "grad_norm": 0.6881444183489839,
      "learning_rate": 5.360867591522133e-06,
      "loss": 0.41786885261535645,
      "num_input_tokens_seen": 729026344,
      "step": 14170,
      "train_runtime": 87196.6827,
      "train_tokens_per_second": 8360.712
    },
    {
      "epoch": 0.9544169135469971,
      "grad_norm": 0.6466902780796403,
      "learning_rate": 5.358230354557617e-06,
      "loss": 0.3468768119812012,
      "num_input_tokens_seen": 729291064,
      "step": 14175,
      "train_runtime": 87228.5801,
      "train_tokens_per_second": 8360.689
    },
    {
      "epoch": 0.9547535685429572,
      "grad_norm": 0.741146075980091,
      "learning_rate": 5.355593017415123e-06,
      "loss": 0.3932826995849609,
      "num_input_tokens_seen": 729568600,
      "step": 14180,
      "train_runtime": 87258.0802,
      "train_tokens_per_second": 8361.043
    },
    {
      "epoch": 0.9550902235389174,
      "grad_norm": 0.6489795026132003,
      "learning_rate": 5.352955580832176e-06,
      "loss": 0.36314241886138915,
      "num_input_tokens_seen": 729833944,
      "step": 14185,
      "train_runtime": 87291.5111,
      "train_tokens_per_second": 8360.881
    },
    {
      "epoch": 0.9554268785348775,
      "grad_norm": 0.5487351901589599,
      "learning_rate": 5.350318045546326e-06,
      "loss": 0.3662402153015137,
      "num_input_tokens_seen": 730104176,
      "step": 14190,
      "train_runtime": 87319.0674,
      "train_tokens_per_second": 8361.337
    },
    {
      "epoch": 0.9557635335308377,
      "grad_norm": 0.7469056186744565,
      "learning_rate": 5.347680412295152e-06,
      "loss": 0.36307051181793215,
      "num_input_tokens_seen": 730355336,
      "step": 14195,
      "train_runtime": 87348.111,
      "train_tokens_per_second": 8361.433
    },
    {
      "epoch": 0.9561001885267978,
      "grad_norm": 0.8200625015801231,
      "learning_rate": 5.34504268181626e-06,
      "loss": 0.3784991979598999,
      "num_input_tokens_seen": 730588160,
      "step": 14200,
      "train_runtime": 87373.3333,
      "train_tokens_per_second": 8361.684
    },
    {
      "epoch": 0.9564368435227579,
      "grad_norm": 0.7703156171761787,
      "learning_rate": 5.342404854847282e-06,
      "loss": 0.35217292308807374,
      "num_input_tokens_seen": 730812424,
      "step": 14205,
      "train_runtime": 87401.5762,
      "train_tokens_per_second": 8361.547
    },
    {
      "epoch": 0.9567734985187181,
      "grad_norm": 0.5900923562798783,
      "learning_rate": 5.3397669321258776e-06,
      "loss": 0.35503149032592773,
      "num_input_tokens_seen": 731084576,
      "step": 14210,
      "train_runtime": 87432.2144,
      "train_tokens_per_second": 8361.73
    },
    {
      "epoch": 0.9571101535146782,
      "grad_norm": 0.6059012636347123,
      "learning_rate": 5.337128914389734e-06,
      "loss": 0.37281012535095215,
      "num_input_tokens_seen": 731360632,
      "step": 14215,
      "train_runtime": 87462.828,
      "train_tokens_per_second": 8361.96
    },
    {
      "epoch": 0.9574468085106383,
      "grad_norm": 0.6052290753783702,
      "learning_rate": 5.334490802376565e-06,
      "loss": 0.35882718563079835,
      "num_input_tokens_seen": 731614528,
      "step": 14220,
      "train_runtime": 87494.6482,
      "train_tokens_per_second": 8361.82
    },
    {
      "epoch": 0.9577834635065985,
      "grad_norm": 1.0021853095980249,
      "learning_rate": 5.33185259682411e-06,
      "loss": 0.38712215423583984,
      "num_input_tokens_seen": 731879728,
      "step": 14225,
      "train_runtime": 87527.2179,
      "train_tokens_per_second": 8361.739
    },
    {
      "epoch": 0.9581201185025586,
      "grad_norm": 0.6258285931480688,
      "learning_rate": 5.329214298470134e-06,
      "loss": 0.38699243068695066,
      "num_input_tokens_seen": 732140648,
      "step": 14230,
      "train_runtime": 87556.2694,
      "train_tokens_per_second": 8361.944
    },
    {
      "epoch": 0.9584567734985188,
      "grad_norm": 0.6035219068310738,
      "learning_rate": 5.326575908052428e-06,
      "loss": 0.3572932243347168,
      "num_input_tokens_seen": 732415056,
      "step": 14235,
      "train_runtime": 87586.2315,
      "train_tokens_per_second": 8362.217
    },
    {
      "epoch": 0.9587934284944789,
      "grad_norm": 0.760452601237789,
      "learning_rate": 5.323937426308813e-06,
      "loss": 0.4015679359436035,
      "num_input_tokens_seen": 732682840,
      "step": 14240,
      "train_runtime": 87617.6486,
      "train_tokens_per_second": 8362.275
    },
    {
      "epoch": 0.959130083490439,
      "grad_norm": 0.6245059183783409,
      "learning_rate": 5.321298853977128e-06,
      "loss": 0.34355151653289795,
      "num_input_tokens_seen": 732982096,
      "step": 14245,
      "train_runtime": 87652.8017,
      "train_tokens_per_second": 8362.335
    },
    {
      "epoch": 0.9594667384863992,
      "grad_norm": 0.5598390813095574,
      "learning_rate": 5.318660191795244e-06,
      "loss": 0.3299668788909912,
      "num_input_tokens_seen": 733249112,
      "step": 14250,
      "train_runtime": 87683.3976,
      "train_tokens_per_second": 8362.462
    },
    {
      "epoch": 0.9598033934823593,
      "grad_norm": 0.6285031747741826,
      "learning_rate": 5.316021440501053e-06,
      "loss": 0.3323509216308594,
      "num_input_tokens_seen": 733494280,
      "step": 14255,
      "train_runtime": 87717.4864,
      "train_tokens_per_second": 8362.008
    },
    {
      "epoch": 0.9601400484783195,
      "grad_norm": 0.6053642715267822,
      "learning_rate": 5.313382600832474e-06,
      "loss": 0.3179682493209839,
      "num_input_tokens_seen": 733751568,
      "step": 14260,
      "train_runtime": 87744.5273,
      "train_tokens_per_second": 8362.363
    },
    {
      "epoch": 0.9604767034742796,
      "grad_norm": 0.5914927140142677,
      "learning_rate": 5.310743673527451e-06,
      "loss": 0.3719813823699951,
      "num_input_tokens_seen": 734029832,
      "step": 14265,
      "train_runtime": 87776.5852,
      "train_tokens_per_second": 8362.479
    },
    {
      "epoch": 0.9608133584702397,
      "grad_norm": 0.5883631687596198,
      "learning_rate": 5.3081046593239496e-06,
      "loss": 0.3751855373382568,
      "num_input_tokens_seen": 734291384,
      "step": 14270,
      "train_runtime": 87809.7661,
      "train_tokens_per_second": 8362.297
    },
    {
      "epoch": 0.9611500134661999,
      "grad_norm": 0.6427857183989779,
      "learning_rate": 5.305465558959964e-06,
      "loss": 0.36624219417572024,
      "num_input_tokens_seen": 734536112,
      "step": 14275,
      "train_runtime": 87839.3834,
      "train_tokens_per_second": 8362.264
    },
    {
      "epoch": 0.96148666846216,
      "grad_norm": 0.6113274606828261,
      "learning_rate": 5.302826373173506e-06,
      "loss": 0.34711906909942625,
      "num_input_tokens_seen": 734808512,
      "step": 14280,
      "train_runtime": 87875.3038,
      "train_tokens_per_second": 8361.946
    },
    {
      "epoch": 0.9618233234581202,
      "grad_norm": 0.6430902827857716,
      "learning_rate": 5.3001871027026194e-06,
      "loss": 0.3507866382598877,
      "num_input_tokens_seen": 735046224,
      "step": 14285,
      "train_runtime": 87904.9722,
      "train_tokens_per_second": 8361.828
    },
    {
      "epoch": 0.9621599784540803,
      "grad_norm": 0.5913404062989938,
      "learning_rate": 5.2975477482853685e-06,
      "loss": 0.3438394546508789,
      "num_input_tokens_seen": 735314352,
      "step": 14290,
      "train_runtime": 87935.8242,
      "train_tokens_per_second": 8361.943
    },
    {
      "epoch": 0.9624966334500404,
      "grad_norm": 0.6630694231570018,
      "learning_rate": 5.294908310659838e-06,
      "loss": 0.3502235412597656,
      "num_input_tokens_seen": 735591168,
      "step": 14295,
      "train_runtime": 87967.0441,
      "train_tokens_per_second": 8362.122
    },
    {
      "epoch": 0.9628332884460006,
      "grad_norm": 0.6996964521243003,
      "learning_rate": 5.292268790564138e-06,
      "loss": 0.33524651527404786,
      "num_input_tokens_seen": 735840040,
      "step": 14300,
      "train_runtime": 87998.5021,
      "train_tokens_per_second": 8361.961
    },
    {
      "epoch": 0.9631699434419607,
      "grad_norm": 0.6569820410902291,
      "learning_rate": 5.289629188736403e-06,
      "loss": 0.3850408554077148,
      "num_input_tokens_seen": 736090512,
      "step": 14305,
      "train_runtime": 88028.2693,
      "train_tokens_per_second": 8361.979
    },
    {
      "epoch": 0.9635065984379209,
      "grad_norm": 0.6526470344334132,
      "learning_rate": 5.286989505914788e-06,
      "loss": 0.3760151147842407,
      "num_input_tokens_seen": 736373752,
      "step": 14310,
      "train_runtime": 88056.2856,
      "train_tokens_per_second": 8362.535
    },
    {
      "epoch": 0.963843253433881,
      "grad_norm": 0.6933414143944796,
      "learning_rate": 5.284349742837475e-06,
      "loss": 0.3672949314117432,
      "num_input_tokens_seen": 736648592,
      "step": 14315,
      "train_runtime": 88089.9115,
      "train_tokens_per_second": 8362.463
    },
    {
      "epoch": 0.9641799084298411,
      "grad_norm": 0.7487858712936426,
      "learning_rate": 5.281709900242662e-06,
      "loss": 0.35797595977783203,
      "num_input_tokens_seen": 736910840,
      "step": 14320,
      "train_runtime": 88118.7266,
      "train_tokens_per_second": 8362.704
    },
    {
      "epoch": 0.9645165634258013,
      "grad_norm": 0.7650827733447019,
      "learning_rate": 5.279069978868574e-06,
      "loss": 0.3477782249450684,
      "num_input_tokens_seen": 737173344,
      "step": 14325,
      "train_runtime": 88151.3332,
      "train_tokens_per_second": 8362.589
    },
    {
      "epoch": 0.9648532184217614,
      "grad_norm": 0.6367357115166513,
      "learning_rate": 5.276429979453455e-06,
      "loss": 0.3619274854660034,
      "num_input_tokens_seen": 737431032,
      "step": 14330,
      "train_runtime": 88183.3002,
      "train_tokens_per_second": 8362.479
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.576357115575759,
      "learning_rate": 5.273789902735575e-06,
      "loss": 0.3567034244537354,
      "num_input_tokens_seen": 737694728,
      "step": 14335,
      "train_runtime": 88210.5816,
      "train_tokens_per_second": 8362.882
    },
    {
      "epoch": 0.9655265284136817,
      "grad_norm": 0.6473147540295975,
      "learning_rate": 5.27114974945322e-06,
      "loss": 0.33646454811096194,
      "num_input_tokens_seen": 737939136,
      "step": 14340,
      "train_runtime": 88243.7851,
      "train_tokens_per_second": 8362.505
    },
    {
      "epoch": 0.9658631834096418,
      "grad_norm": 0.6355379740381388,
      "learning_rate": 5.2685095203447046e-06,
      "loss": 0.35042057037353513,
      "num_input_tokens_seen": 738196352,
      "step": 14345,
      "train_runtime": 88275.2604,
      "train_tokens_per_second": 8362.438
    },
    {
      "epoch": 0.966199838405602,
      "grad_norm": 0.5955511218868927,
      "learning_rate": 5.265869216148357e-06,
      "loss": 0.3428061485290527,
      "num_input_tokens_seen": 738458440,
      "step": 14350,
      "train_runtime": 88303.839,
      "train_tokens_per_second": 8362.699
    },
    {
      "epoch": 0.9665364934015621,
      "grad_norm": 0.62574068382919,
      "learning_rate": 5.263228837602532e-06,
      "loss": 0.408629035949707,
      "num_input_tokens_seen": 738720392,
      "step": 14355,
      "train_runtime": 88336.0284,
      "train_tokens_per_second": 8362.617
    },
    {
      "epoch": 0.9668731483975223,
      "grad_norm": 0.6892173591059155,
      "learning_rate": 5.2605883854456e-06,
      "loss": 0.3529597043991089,
      "num_input_tokens_seen": 738975472,
      "step": 14360,
      "train_runtime": 88364.6851,
      "train_tokens_per_second": 8362.792
    },
    {
      "epoch": 0.9672098033934824,
      "grad_norm": 0.5501760577305538,
      "learning_rate": 5.257947860415957e-06,
      "loss": 0.3290332078933716,
      "num_input_tokens_seen": 739253888,
      "step": 14365,
      "train_runtime": 88396.4496,
      "train_tokens_per_second": 8362.936
    },
    {
      "epoch": 0.9675464583894425,
      "grad_norm": 0.6954511434519749,
      "learning_rate": 5.255307263252019e-06,
      "loss": 0.35056846141815184,
      "num_input_tokens_seen": 739535304,
      "step": 14370,
      "train_runtime": 88428.412,
      "train_tokens_per_second": 8363.096
    },
    {
      "epoch": 0.9678831133854027,
      "grad_norm": 0.5986790159846712,
      "learning_rate": 5.2526665946922185e-06,
      "loss": 0.36737747192382814,
      "num_input_tokens_seen": 739795160,
      "step": 14375,
      "train_runtime": 88457.9532,
      "train_tokens_per_second": 8363.241
    },
    {
      "epoch": 0.9682197683813628,
      "grad_norm": 0.5711589258694139,
      "learning_rate": 5.25002585547501e-06,
      "loss": 0.34208879470825193,
      "num_input_tokens_seen": 740060960,
      "step": 14380,
      "train_runtime": 88490.2956,
      "train_tokens_per_second": 8363.188
    },
    {
      "epoch": 0.968556423377323,
      "grad_norm": 0.6617538302283817,
      "learning_rate": 5.2473850463388675e-06,
      "loss": 0.3895174264907837,
      "num_input_tokens_seen": 740323992,
      "step": 14385,
      "train_runtime": 88519.8059,
      "train_tokens_per_second": 8363.371
    },
    {
      "epoch": 0.9688930783732831,
      "grad_norm": 0.58873523773307,
      "learning_rate": 5.244744168022288e-06,
      "loss": 0.35665168762207033,
      "num_input_tokens_seen": 740583328,
      "step": 14390,
      "train_runtime": 88547.8761,
      "train_tokens_per_second": 8363.649
    },
    {
      "epoch": 0.9692297333692432,
      "grad_norm": 0.687296731797204,
      "learning_rate": 5.24210322126378e-06,
      "loss": 0.34968135356903074,
      "num_input_tokens_seen": 740868904,
      "step": 14395,
      "train_runtime": 88581.4957,
      "train_tokens_per_second": 8363.698
    },
    {
      "epoch": 0.9695663883652034,
      "grad_norm": 0.5907313149254284,
      "learning_rate": 5.23946220680188e-06,
      "loss": 0.35584030151367185,
      "num_input_tokens_seen": 741157160,
      "step": 14400,
      "train_runtime": 88618.7738,
      "train_tokens_per_second": 8363.433
    },
    {
      "epoch": 0.9699030433611635,
      "grad_norm": 0.6465651002155556,
      "learning_rate": 5.236821125375136e-06,
      "loss": 0.351446533203125,
      "num_input_tokens_seen": 741424192,
      "step": 14405,
      "train_runtime": 88650.5268,
      "train_tokens_per_second": 8363.449
    },
    {
      "epoch": 0.9702396983571236,
      "grad_norm": 0.5472166159912781,
      "learning_rate": 5.234179977722119e-06,
      "loss": 0.31578528881073,
      "num_input_tokens_seen": 741719056,
      "step": 14410,
      "train_runtime": 88687.8221,
      "train_tokens_per_second": 8363.257
    },
    {
      "epoch": 0.9705763533530838,
      "grad_norm": 0.5546413346852146,
      "learning_rate": 5.23153876458142e-06,
      "loss": 0.35041203498840334,
      "num_input_tokens_seen": 741961000,
      "step": 14415,
      "train_runtime": 88719.2436,
      "train_tokens_per_second": 8363.022
    },
    {
      "epoch": 0.9709130083490439,
      "grad_norm": 0.6474009495862352,
      "learning_rate": 5.2288974866916435e-06,
      "loss": 0.3520882844924927,
      "num_input_tokens_seen": 742186208,
      "step": 14420,
      "train_runtime": 88752.3956,
      "train_tokens_per_second": 8362.436
    },
    {
      "epoch": 0.9712496633450041,
      "grad_norm": 0.601668234170292,
      "learning_rate": 5.226256144791414e-06,
      "loss": 0.3570445775985718,
      "num_input_tokens_seen": 742426024,
      "step": 14425,
      "train_runtime": 88783.9134,
      "train_tokens_per_second": 8362.168
    },
    {
      "epoch": 0.9715863183409642,
      "grad_norm": 0.6403460989752152,
      "learning_rate": 5.223614739619375e-06,
      "loss": 0.36797733306884767,
      "num_input_tokens_seen": 742658696,
      "step": 14430,
      "train_runtime": 88810.9412,
      "train_tokens_per_second": 8362.243
    },
    {
      "epoch": 0.9719229733369243,
      "grad_norm": 0.5955233306232816,
      "learning_rate": 5.2209732719141855e-06,
      "loss": 0.3318456172943115,
      "num_input_tokens_seen": 742900528,
      "step": 14435,
      "train_runtime": 88842.7448,
      "train_tokens_per_second": 8361.972
    },
    {
      "epoch": 0.9722596283328845,
      "grad_norm": 0.5584249761353324,
      "learning_rate": 5.218331742414528e-06,
      "loss": 0.34693429470062254,
      "num_input_tokens_seen": 743185552,
      "step": 14440,
      "train_runtime": 88875.411,
      "train_tokens_per_second": 8362.105
    },
    {
      "epoch": 0.9725962833288446,
      "grad_norm": 0.5920376825373118,
      "learning_rate": 5.215690151859094e-06,
      "loss": 0.3266302108764648,
      "num_input_tokens_seen": 743422800,
      "step": 14445,
      "train_runtime": 88903.8314,
      "train_tokens_per_second": 8362.101
    },
    {
      "epoch": 0.9729329383248048,
      "grad_norm": 0.6443995456709581,
      "learning_rate": 5.213048500986597e-06,
      "loss": 0.36126294136047366,
      "num_input_tokens_seen": 743669904,
      "step": 14450,
      "train_runtime": 88937.2042,
      "train_tokens_per_second": 8361.741
    },
    {
      "epoch": 0.9732695933207649,
      "grad_norm": 0.7267624526453857,
      "learning_rate": 5.210406790535766e-06,
      "loss": 0.3521712779998779,
      "num_input_tokens_seen": 743918096,
      "step": 14455,
      "train_runtime": 88965.9142,
      "train_tokens_per_second": 8361.833
    },
    {
      "epoch": 0.973606248316725,
      "grad_norm": 0.6127641805123557,
      "learning_rate": 5.207765021245346e-06,
      "loss": 0.35202717781066895,
      "num_input_tokens_seen": 744164832,
      "step": 14460,
      "train_runtime": 88994.7981,
      "train_tokens_per_second": 8361.891
    },
    {
      "epoch": 0.9739429033126852,
      "grad_norm": 0.645537050130955,
      "learning_rate": 5.2051231938541015e-06,
      "loss": 0.36264159679412844,
      "num_input_tokens_seen": 744401424,
      "step": 14465,
      "train_runtime": 89022.2616,
      "train_tokens_per_second": 8361.969
    },
    {
      "epoch": 0.9742795583086453,
      "grad_norm": 0.6806433439643508,
      "learning_rate": 5.202481309100808e-06,
      "loss": 0.3530406951904297,
      "num_input_tokens_seen": 744634232,
      "step": 14470,
      "train_runtime": 89052.9072,
      "train_tokens_per_second": 8361.706
    },
    {
      "epoch": 0.9746162133046055,
      "grad_norm": 1.1360714228856998,
      "learning_rate": 5.199839367724266e-06,
      "loss": 0.3529019355773926,
      "num_input_tokens_seen": 744906680,
      "step": 14475,
      "train_runtime": 89083.7213,
      "train_tokens_per_second": 8361.872
    },
    {
      "epoch": 0.9749528683005656,
      "grad_norm": 0.6608403740410719,
      "learning_rate": 5.197197370463278e-06,
      "loss": 0.32339093685150144,
      "num_input_tokens_seen": 745184704,
      "step": 14480,
      "train_runtime": 89114.7339,
      "train_tokens_per_second": 8362.082
    },
    {
      "epoch": 0.9752895232965257,
      "grad_norm": 0.5799338074572421,
      "learning_rate": 5.194555318056675e-06,
      "loss": 0.3731882333755493,
      "num_input_tokens_seen": 745473600,
      "step": 14485,
      "train_runtime": 89145.0643,
      "train_tokens_per_second": 8362.478
    },
    {
      "epoch": 0.9756261782924859,
      "grad_norm": 0.7800057093064784,
      "learning_rate": 5.191913211243299e-06,
      "loss": 0.36658635139465334,
      "num_input_tokens_seen": 745736688,
      "step": 14490,
      "train_runtime": 89178.7465,
      "train_tokens_per_second": 8362.269
    },
    {
      "epoch": 0.975962833288446,
      "grad_norm": 0.6563491973832075,
      "learning_rate": 5.1892710507620055e-06,
      "loss": 0.36483492851257326,
      "num_input_tokens_seen": 746004288,
      "step": 14495,
      "train_runtime": 89209.2758,
      "train_tokens_per_second": 8362.407
    },
    {
      "epoch": 0.9762994882844062,
      "grad_norm": 0.628663992301826,
      "learning_rate": 5.186628837351666e-06,
      "loss": 0.3439019680023193,
      "num_input_tokens_seen": 746282832,
      "step": 14500,
      "train_runtime": 89241.5739,
      "train_tokens_per_second": 8362.502
    },
    {
      "epoch": 0.9766361432803663,
      "grad_norm": 0.5984484263151677,
      "learning_rate": 5.183986571751167e-06,
      "loss": 0.3592672348022461,
      "num_input_tokens_seen": 746537296,
      "step": 14505,
      "train_runtime": 89272.5539,
      "train_tokens_per_second": 8362.45
    },
    {
      "epoch": 0.9769727982763264,
      "grad_norm": 0.6165789902645931,
      "learning_rate": 5.181344254699408e-06,
      "loss": 0.37874755859375,
      "num_input_tokens_seen": 746792608,
      "step": 14510,
      "train_runtime": 89304.257,
      "train_tokens_per_second": 8362.341
    },
    {
      "epoch": 0.9773094532722866,
      "grad_norm": 0.540056323599675,
      "learning_rate": 5.178701886935309e-06,
      "loss": 0.3450993537902832,
      "num_input_tokens_seen": 747045280,
      "step": 14515,
      "train_runtime": 89334.9227,
      "train_tokens_per_second": 8362.298
    },
    {
      "epoch": 0.9776461082682467,
      "grad_norm": 0.5919400091069926,
      "learning_rate": 5.176059469197796e-06,
      "loss": 0.3634330749511719,
      "num_input_tokens_seen": 747311296,
      "step": 14520,
      "train_runtime": 89368.532,
      "train_tokens_per_second": 8362.13
    },
    {
      "epoch": 0.9779827632642069,
      "grad_norm": 0.6919317566849034,
      "learning_rate": 5.173417002225815e-06,
      "loss": 0.34366180896759035,
      "num_input_tokens_seen": 747570408,
      "step": 14525,
      "train_runtime": 89395.9356,
      "train_tokens_per_second": 8362.465
    },
    {
      "epoch": 0.978319418260167,
      "grad_norm": 0.6723946904390986,
      "learning_rate": 5.170774486758322e-06,
      "loss": 0.3595534086227417,
      "num_input_tokens_seen": 747797280,
      "step": 14530,
      "train_runtime": 89421.9595,
      "train_tokens_per_second": 8362.569
    },
    {
      "epoch": 0.9786560732561271,
      "grad_norm": 0.6418375733853215,
      "learning_rate": 5.168131923534287e-06,
      "loss": 0.3850479364395142,
      "num_input_tokens_seen": 748069144,
      "step": 14535,
      "train_runtime": 89453.2736,
      "train_tokens_per_second": 8362.68
    },
    {
      "epoch": 0.9789927282520873,
      "grad_norm": 0.6071518605261178,
      "learning_rate": 5.165489313292696e-06,
      "loss": 0.3487050533294678,
      "num_input_tokens_seen": 748311472,
      "step": 14540,
      "train_runtime": 89485.9102,
      "train_tokens_per_second": 8362.338
    },
    {
      "epoch": 0.9793293832480474,
      "grad_norm": 0.6626189417646114,
      "learning_rate": 5.162846656772548e-06,
      "loss": 0.3451840400695801,
      "num_input_tokens_seen": 748551560,
      "step": 14545,
      "train_runtime": 89520.3163,
      "train_tokens_per_second": 8361.806
    },
    {
      "epoch": 0.9796660382440076,
      "grad_norm": 0.6463291770965629,
      "learning_rate": 5.16020395471285e-06,
      "loss": 0.3357709407806396,
      "num_input_tokens_seen": 748813384,
      "step": 14550,
      "train_runtime": 89549.2981,
      "train_tokens_per_second": 8362.024
    },
    {
      "epoch": 0.9800026932399677,
      "grad_norm": 1.0417105225372099,
      "learning_rate": 5.157561207852628e-06,
      "loss": 0.3749570608139038,
      "num_input_tokens_seen": 749061352,
      "step": 14555,
      "train_runtime": 89576.9344,
      "train_tokens_per_second": 8362.212
    },
    {
      "epoch": 0.9803393482359278,
      "grad_norm": 0.700988877194742,
      "learning_rate": 5.154918416930915e-06,
      "loss": 0.342649507522583,
      "num_input_tokens_seen": 749327200,
      "step": 14560,
      "train_runtime": 89605.9422,
      "train_tokens_per_second": 8362.472
    },
    {
      "epoch": 0.980676003231888,
      "grad_norm": 0.6050950753181012,
      "learning_rate": 5.152275582686761e-06,
      "loss": 0.37602529525756834,
      "num_input_tokens_seen": 749582488,
      "step": 14565,
      "train_runtime": 89641.8757,
      "train_tokens_per_second": 8361.968
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.6043880880196205,
      "learning_rate": 5.149632705859227e-06,
      "loss": 0.36406569480895995,
      "num_input_tokens_seen": 749848456,
      "step": 14570,
      "train_runtime": 89672.6974,
      "train_tokens_per_second": 8362.06
    },
    {
      "epoch": 0.9813493132238083,
      "grad_norm": 0.6107858541565223,
      "learning_rate": 5.146989787187383e-06,
      "loss": 0.3663525104522705,
      "num_input_tokens_seen": 750139688,
      "step": 14575,
      "train_runtime": 89701.8013,
      "train_tokens_per_second": 8362.593
    },
    {
      "epoch": 0.9816859682197684,
      "grad_norm": 0.5793255349162568,
      "learning_rate": 5.144346827410314e-06,
      "loss": 0.35837926864624026,
      "num_input_tokens_seen": 750418968,
      "step": 14580,
      "train_runtime": 89729.644,
      "train_tokens_per_second": 8363.111
    },
    {
      "epoch": 0.9820226232157285,
      "grad_norm": 0.7725196351462432,
      "learning_rate": 5.141703827267117e-06,
      "loss": 0.3677496910095215,
      "num_input_tokens_seen": 750689672,
      "step": 14585,
      "train_runtime": 89762.7368,
      "train_tokens_per_second": 8363.043
    },
    {
      "epoch": 0.9823592782116887,
      "grad_norm": 0.5602852832783709,
      "learning_rate": 5.139060787496893e-06,
      "loss": 0.36788253784179686,
      "num_input_tokens_seen": 750950984,
      "step": 14590,
      "train_runtime": 89796.2108,
      "train_tokens_per_second": 8362.836
    },
    {
      "epoch": 0.9826959332076488,
      "grad_norm": 0.663202974632322,
      "learning_rate": 5.136417708838766e-06,
      "loss": 0.36667675971984864,
      "num_input_tokens_seen": 751174008,
      "step": 14595,
      "train_runtime": 89823.9435,
      "train_tokens_per_second": 8362.737
    },
    {
      "epoch": 0.983032588203609,
      "grad_norm": 0.7215162040150019,
      "learning_rate": 5.13377459203186e-06,
      "loss": 0.3619992733001709,
      "num_input_tokens_seen": 751415784,
      "step": 14600,
      "train_runtime": 89854.8465,
      "train_tokens_per_second": 8362.552
    },
    {
      "epoch": 0.9833692431995691,
      "grad_norm": 0.5916735829111606,
      "learning_rate": 5.131131437815318e-06,
      "loss": 0.3326805830001831,
      "num_input_tokens_seen": 751670512,
      "step": 14605,
      "train_runtime": 89887.2978,
      "train_tokens_per_second": 8362.366
    },
    {
      "epoch": 0.9837058981955292,
      "grad_norm": 0.6580944822007222,
      "learning_rate": 5.128488246928287e-06,
      "loss": 0.3637146234512329,
      "num_input_tokens_seen": 751952016,
      "step": 14610,
      "train_runtime": 89915.0076,
      "train_tokens_per_second": 8362.92
    },
    {
      "epoch": 0.9840425531914894,
      "grad_norm": 0.6269937936499203,
      "learning_rate": 5.125845020109925e-06,
      "loss": 0.36372296810150145,
      "num_input_tokens_seen": 752198920,
      "step": 14615,
      "train_runtime": 89945.6415,
      "train_tokens_per_second": 8362.817
    },
    {
      "epoch": 0.9843792081874495,
      "grad_norm": 0.7539813088839651,
      "learning_rate": 5.123201758099406e-06,
      "loss": 0.35615763664245603,
      "num_input_tokens_seen": 752437992,
      "step": 14620,
      "train_runtime": 89976.59,
      "train_tokens_per_second": 8362.597
    },
    {
      "epoch": 0.9847158631834096,
      "grad_norm": 0.6482627432196469,
      "learning_rate": 5.12055846163591e-06,
      "loss": 0.36782193183898926,
      "num_input_tokens_seen": 752688064,
      "step": 14625,
      "train_runtime": 90008.8188,
      "train_tokens_per_second": 8362.381
    },
    {
      "epoch": 0.9850525181793698,
      "grad_norm": 0.6197282673328127,
      "learning_rate": 5.117915131458623e-06,
      "loss": 0.3248273849487305,
      "num_input_tokens_seen": 752949144,
      "step": 14630,
      "train_runtime": 90039.4664,
      "train_tokens_per_second": 8362.435
    },
    {
      "epoch": 0.9853891731753299,
      "grad_norm": 0.6767873378883582,
      "learning_rate": 5.115271768306745e-06,
      "loss": 0.34547929763793944,
      "num_input_tokens_seen": 753202432,
      "step": 14635,
      "train_runtime": 90072.3407,
      "train_tokens_per_second": 8362.195
    },
    {
      "epoch": 0.9857258281712901,
      "grad_norm": 0.7477016981809171,
      "learning_rate": 5.112628372919485e-06,
      "loss": 0.3570772647857666,
      "num_input_tokens_seen": 753456512,
      "step": 14640,
      "train_runtime": 90100.8523,
      "train_tokens_per_second": 8362.368
    },
    {
      "epoch": 0.9860624831672502,
      "grad_norm": 0.6430091963259881,
      "learning_rate": 5.1099849460360595e-06,
      "loss": 0.3487353563308716,
      "num_input_tokens_seen": 753684648,
      "step": 14645,
      "train_runtime": 90128.7521,
      "train_tokens_per_second": 8362.311
    },
    {
      "epoch": 0.9863991381632103,
      "grad_norm": 0.6146746772590236,
      "learning_rate": 5.107341488395695e-06,
      "loss": 0.3252723217010498,
      "num_input_tokens_seen": 753940256,
      "step": 14650,
      "train_runtime": 90157.8583,
      "train_tokens_per_second": 8362.446
    },
    {
      "epoch": 0.9867357931591705,
      "grad_norm": 0.5741507831451229,
      "learning_rate": 5.104698000737626e-06,
      "loss": 0.37476043701171874,
      "num_input_tokens_seen": 754199752,
      "step": 14655,
      "train_runtime": 90190.8699,
      "train_tokens_per_second": 8362.263
    },
    {
      "epoch": 0.9870724481551306,
      "grad_norm": 0.672267227807992,
      "learning_rate": 5.102054483801094e-06,
      "loss": 0.382643985748291,
      "num_input_tokens_seen": 754473816,
      "step": 14660,
      "train_runtime": 90216.6599,
      "train_tokens_per_second": 8362.91
    },
    {
      "epoch": 0.9874091031510908,
      "grad_norm": 0.5928519664216304,
      "learning_rate": 5.099410938325351e-06,
      "loss": 0.3637427091598511,
      "num_input_tokens_seen": 754738880,
      "step": 14665,
      "train_runtime": 90245.6427,
      "train_tokens_per_second": 8363.161
    },
    {
      "epoch": 0.9877457581470509,
      "grad_norm": 0.7404637562742247,
      "learning_rate": 5.096767365049657e-06,
      "loss": 0.34601993560791017,
      "num_input_tokens_seen": 754970080,
      "step": 14670,
      "train_runtime": 90277.2059,
      "train_tokens_per_second": 8362.798
    },
    {
      "epoch": 0.988082413143011,
      "grad_norm": 0.5856229791724301,
      "learning_rate": 5.094123764713277e-06,
      "loss": 0.3318432569503784,
      "num_input_tokens_seen": 755249912,
      "step": 14675,
      "train_runtime": 90313.2216,
      "train_tokens_per_second": 8362.562
    },
    {
      "epoch": 0.9884190681389712,
      "grad_norm": 0.5856170948690114,
      "learning_rate": 5.091480138055487e-06,
      "loss": 0.36562409400939944,
      "num_input_tokens_seen": 755526744,
      "step": 14680,
      "train_runtime": 90346.5773,
      "train_tokens_per_second": 8362.539
    },
    {
      "epoch": 0.9887557231349313,
      "grad_norm": 0.6567953853707483,
      "learning_rate": 5.088836485815568e-06,
      "loss": 0.3500864505767822,
      "num_input_tokens_seen": 755750064,
      "step": 14685,
      "train_runtime": 90376.127,
      "train_tokens_per_second": 8362.275
    },
    {
      "epoch": 0.9890923781308915,
      "grad_norm": 0.6648340395777387,
      "learning_rate": 5.086192808732809e-06,
      "loss": 0.3539912462234497,
      "num_input_tokens_seen": 755981768,
      "step": 14690,
      "train_runtime": 90405.0171,
      "train_tokens_per_second": 8362.166
    },
    {
      "epoch": 0.9894290331268516,
      "grad_norm": 0.6899622824092835,
      "learning_rate": 5.083549107546505e-06,
      "loss": 0.36251192092895507,
      "num_input_tokens_seen": 756223472,
      "step": 14695,
      "train_runtime": 90433.0647,
      "train_tokens_per_second": 8362.245
    },
    {
      "epoch": 0.9897656881228117,
      "grad_norm": 0.6474276293021785,
      "learning_rate": 5.080905382995961e-06,
      "loss": 0.3799863815307617,
      "num_input_tokens_seen": 756495648,
      "step": 14700,
      "train_runtime": 90465.3872,
      "train_tokens_per_second": 8362.266
    },
    {
      "epoch": 0.9901023431187719,
      "grad_norm": 0.6218443029262584,
      "learning_rate": 5.0782616358204835e-06,
      "loss": 0.36276276111602784,
      "num_input_tokens_seen": 756742168,
      "step": 14705,
      "train_runtime": 90496.3354,
      "train_tokens_per_second": 8362.131
    },
    {
      "epoch": 0.990438998114732,
      "grad_norm": 0.7107829374971519,
      "learning_rate": 5.075617866759389e-06,
      "loss": 0.3320207118988037,
      "num_input_tokens_seen": 756990488,
      "step": 14710,
      "train_runtime": 90522.5699,
      "train_tokens_per_second": 8362.45
    },
    {
      "epoch": 0.9907756531106922,
      "grad_norm": 0.6974327050208646,
      "learning_rate": 5.0729740765519995e-06,
      "loss": 0.3975100517272949,
      "num_input_tokens_seen": 757242728,
      "step": 14715,
      "train_runtime": 90553.6659,
      "train_tokens_per_second": 8362.364
    },
    {
      "epoch": 0.9911123081066523,
      "grad_norm": 0.611984968965646,
      "learning_rate": 5.070330265937643e-06,
      "loss": 0.3454824209213257,
      "num_input_tokens_seen": 757497144,
      "step": 14720,
      "train_runtime": 90582.4639,
      "train_tokens_per_second": 8362.514
    },
    {
      "epoch": 0.9914489631026124,
      "grad_norm": 0.6442486150231649,
      "learning_rate": 5.067686435655651e-06,
      "loss": 0.38298299312591555,
      "num_input_tokens_seen": 757768680,
      "step": 14725,
      "train_runtime": 90615.2826,
      "train_tokens_per_second": 8362.482
    },
    {
      "epoch": 0.9917856180985726,
      "grad_norm": 0.6348703031902034,
      "learning_rate": 5.065042586445363e-06,
      "loss": 0.38900146484375,
      "num_input_tokens_seen": 758015080,
      "step": 14730,
      "train_runtime": 90645.4686,
      "train_tokens_per_second": 8362.416
    },
    {
      "epoch": 0.9921222730945327,
      "grad_norm": 0.6582675436514699,
      "learning_rate": 5.0623987190461245e-06,
      "loss": 0.305422043800354,
      "num_input_tokens_seen": 758253336,
      "step": 14735,
      "train_runtime": 90676.563,
      "train_tokens_per_second": 8362.176
    },
    {
      "epoch": 0.9924589280904929,
      "grad_norm": 0.6880543081352709,
      "learning_rate": 5.059754834197281e-06,
      "loss": 0.3703434467315674,
      "num_input_tokens_seen": 758493128,
      "step": 14740,
      "train_runtime": 90710.5008,
      "train_tokens_per_second": 8361.69
    },
    {
      "epoch": 0.992795583086453,
      "grad_norm": 0.6682124746717244,
      "learning_rate": 5.05711093263819e-06,
      "loss": 0.3130594968795776,
      "num_input_tokens_seen": 758741760,
      "step": 14745,
      "train_runtime": 90739.3177,
      "train_tokens_per_second": 8361.775
    },
    {
      "epoch": 0.9931322380824131,
      "grad_norm": 0.6849999604258528,
      "learning_rate": 5.054467015108208e-06,
      "loss": 0.36989803314208985,
      "num_input_tokens_seen": 759025152,
      "step": 14750,
      "train_runtime": 90768.9888,
      "train_tokens_per_second": 8362.164
    },
    {
      "epoch": 0.9934688930783733,
      "grad_norm": 0.7990832944839223,
      "learning_rate": 5.051823082346701e-06,
      "loss": 0.35957190990447996,
      "num_input_tokens_seen": 759286688,
      "step": 14755,
      "train_runtime": 90802.74,
      "train_tokens_per_second": 8361.936
    },
    {
      "epoch": 0.9938055480743334,
      "grad_norm": 0.6589347055874071,
      "learning_rate": 5.049179135093033e-06,
      "loss": 0.36083264350891114,
      "num_input_tokens_seen": 759571472,
      "step": 14760,
      "train_runtime": 90829.999,
      "train_tokens_per_second": 8362.562
    },
    {
      "epoch": 0.9941422030702936,
      "grad_norm": 0.659623974325216,
      "learning_rate": 5.046535174086575e-06,
      "loss": 0.3653231143951416,
      "num_input_tokens_seen": 759809256,
      "step": 14765,
      "train_runtime": 90857.0493,
      "train_tokens_per_second": 8362.689
    },
    {
      "epoch": 0.9944788580662537,
      "grad_norm": 0.674700228042783,
      "learning_rate": 5.043891200066706e-06,
      "loss": 0.3586362361907959,
      "num_input_tokens_seen": 760064672,
      "step": 14770,
      "train_runtime": 90889.1117,
      "train_tokens_per_second": 8362.549
    },
    {
      "epoch": 0.9948155130622138,
      "grad_norm": 0.5945504083432771,
      "learning_rate": 5.041247213772802e-06,
      "loss": 0.36374197006225584,
      "num_input_tokens_seen": 760347904,
      "step": 14775,
      "train_runtime": 90919.1486,
      "train_tokens_per_second": 8362.902
    },
    {
      "epoch": 0.995152168058174,
      "grad_norm": 0.7608599518831826,
      "learning_rate": 5.038603215944247e-06,
      "loss": 0.3569845676422119,
      "num_input_tokens_seen": 760586408,
      "step": 14780,
      "train_runtime": 90953.2014,
      "train_tokens_per_second": 8362.393
    },
    {
      "epoch": 0.9954888230541341,
      "grad_norm": 0.639936802278777,
      "learning_rate": 5.0359592073204254e-06,
      "loss": 0.34589252471923826,
      "num_input_tokens_seen": 760842136,
      "step": 14785,
      "train_runtime": 90983.7275,
      "train_tokens_per_second": 8362.398
    },
    {
      "epoch": 0.9958254780500942,
      "grad_norm": 0.7083271073380089,
      "learning_rate": 5.033315188640725e-06,
      "loss": 0.3430428981781006,
      "num_input_tokens_seen": 761110600,
      "step": 14790,
      "train_runtime": 91013.6452,
      "train_tokens_per_second": 8362.599
    },
    {
      "epoch": 0.9961621330460544,
      "grad_norm": 0.5334955039539556,
      "learning_rate": 5.03067116064454e-06,
      "loss": 0.3328507900238037,
      "num_input_tokens_seen": 761370096,
      "step": 14795,
      "train_runtime": 91039.6612,
      "train_tokens_per_second": 8363.059
    },
    {
      "epoch": 0.9964987880420145,
      "grad_norm": 0.524148643404444,
      "learning_rate": 5.028027124071262e-06,
      "loss": 0.3636810779571533,
      "num_input_tokens_seen": 761606896,
      "step": 14800,
      "train_runtime": 91064.9382,
      "train_tokens_per_second": 8363.338
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.5251853592393707,
      "learning_rate": 5.025383079660289e-06,
      "loss": 0.32831132411956787,
      "num_input_tokens_seen": 761879608,
      "step": 14805,
      "train_runtime": 91095.4949,
      "train_tokens_per_second": 8363.527
    },
    {
      "epoch": 0.9971720980339348,
      "grad_norm": 0.6379306266340367,
      "learning_rate": 5.0227390281510166e-06,
      "loss": 0.362441086769104,
      "num_input_tokens_seen": 762140368,
      "step": 14810,
      "train_runtime": 91126.7037,
      "train_tokens_per_second": 8363.524
    },
    {
      "epoch": 0.997508753029895,
      "grad_norm": 0.6239000118868194,
      "learning_rate": 5.020094970282849e-06,
      "loss": 0.3600430488586426,
      "num_input_tokens_seen": 762377424,
      "step": 14815,
      "train_runtime": 91156.2599,
      "train_tokens_per_second": 8363.413
    },
    {
      "epoch": 0.9978454080258551,
      "grad_norm": 0.6606610473154998,
      "learning_rate": 5.017450906795187e-06,
      "loss": 0.3697696924209595,
      "num_input_tokens_seen": 762648888,
      "step": 14820,
      "train_runtime": 91190.198,
      "train_tokens_per_second": 8363.277
    },
    {
      "epoch": 0.9981820630218152,
      "grad_norm": 0.7303227337453166,
      "learning_rate": 5.014806838427434e-06,
      "loss": 0.3719185352325439,
      "num_input_tokens_seen": 762875464,
      "step": 14825,
      "train_runtime": 91222.2438,
      "train_tokens_per_second": 8362.823
    },
    {
      "epoch": 0.9985187180177754,
      "grad_norm": 0.6112202405088385,
      "learning_rate": 5.0121627659189964e-06,
      "loss": 0.343916130065918,
      "num_input_tokens_seen": 763155456,
      "step": 14830,
      "train_runtime": 91250.849,
      "train_tokens_per_second": 8363.27
    },
    {
      "epoch": 0.9988553730137355,
      "grad_norm": 0.5672140973077854,
      "learning_rate": 5.009518690009281e-06,
      "loss": 0.3465181112289429,
      "num_input_tokens_seen": 763398848,
      "step": 14835,
      "train_runtime": 91282.4298,
      "train_tokens_per_second": 8363.043
    },
    {
      "epoch": 0.9991920280096956,
      "grad_norm": 0.6467179062631327,
      "learning_rate": 5.006874611437692e-06,
      "loss": 0.3693327665328979,
      "num_input_tokens_seen": 763662592,
      "step": 14840,
      "train_runtime": 91311.1425,
      "train_tokens_per_second": 8363.301
    },
    {
      "epoch": 0.9995286830056558,
      "grad_norm": 0.5924378704896538,
      "learning_rate": 5.004230530943641e-06,
      "loss": 0.3489140272140503,
      "num_input_tokens_seen": 763901144,
      "step": 14845,
      "train_runtime": 91337.1477,
      "train_tokens_per_second": 8363.532
    },
    {
      "epoch": 0.9998653380016159,
      "grad_norm": 0.6253377088271591,
      "learning_rate": 5.001586449266536e-06,
      "loss": 0.31801815032958985,
      "num_input_tokens_seen": 764167184,
      "step": 14850,
      "train_runtime": 91366.2546,
      "train_tokens_per_second": 8363.779
    },
    {
      "epoch": 1.000201992997576,
      "grad_norm": 0.5278000491225869,
      "learning_rate": 4.998942367145787e-06,
      "loss": 0.3537947177886963,
      "num_input_tokens_seen": 764415920,
      "step": 14855,
      "train_runtime": 91548.6356,
      "train_tokens_per_second": 8349.834
    },
    {
      "epoch": 1.0005386479935363,
      "grad_norm": 0.5988935511967799,
      "learning_rate": 4.996298285320798e-06,
      "loss": 0.3004462242126465,
      "num_input_tokens_seen": 764659432,
      "step": 14860,
      "train_runtime": 91578.8075,
      "train_tokens_per_second": 8349.742
    },
    {
      "epoch": 1.0008753029894963,
      "grad_norm": 0.7045294075526491,
      "learning_rate": 4.993654204530983e-06,
      "loss": 0.29091382026672363,
      "num_input_tokens_seen": 764926112,
      "step": 14865,
      "train_runtime": 91611.9217,
      "train_tokens_per_second": 8349.635
    },
    {
      "epoch": 1.0012119579854566,
      "grad_norm": 0.7129376818795925,
      "learning_rate": 4.99101012551575e-06,
      "loss": 0.29558095932006834,
      "num_input_tokens_seen": 765136600,
      "step": 14870,
      "train_runtime": 91640.0892,
      "train_tokens_per_second": 8349.366
    },
    {
      "epoch": 1.0015486129814166,
      "grad_norm": 0.6478838846215902,
      "learning_rate": 4.9883660490145075e-06,
      "loss": 0.300225567817688,
      "num_input_tokens_seen": 765415616,
      "step": 14875,
      "train_runtime": 91669.3806,
      "train_tokens_per_second": 8349.741
    },
    {
      "epoch": 1.0018852679773769,
      "grad_norm": 0.6204352812580606,
      "learning_rate": 4.985721975766661e-06,
      "loss": 0.32076153755187986,
      "num_input_tokens_seen": 765676760,
      "step": 14880,
      "train_runtime": 91700.8321,
      "train_tokens_per_second": 8349.725
    },
    {
      "epoch": 1.002221922973337,
      "grad_norm": 0.676470695644764,
      "learning_rate": 4.983077906511619e-06,
      "loss": 0.3248899459838867,
      "num_input_tokens_seen": 765900856,
      "step": 14885,
      "train_runtime": 91729.1845,
      "train_tokens_per_second": 8349.588
    },
    {
      "epoch": 1.0025585779692971,
      "grad_norm": 0.5986537593266208,
      "learning_rate": 4.980433841988786e-06,
      "loss": 0.3039379596710205,
      "num_input_tokens_seen": 766163560,
      "step": 14890,
      "train_runtime": 91757.7698,
      "train_tokens_per_second": 8349.849
    },
    {
      "epoch": 1.0028952329652572,
      "grad_norm": 0.6129681187802878,
      "learning_rate": 4.977789782937568e-06,
      "loss": 0.3060157775878906,
      "num_input_tokens_seen": 766437864,
      "step": 14895,
      "train_runtime": 91788.1286,
      "train_tokens_per_second": 8350.076
    },
    {
      "epoch": 1.0032318879612174,
      "grad_norm": 0.6300176061061673,
      "learning_rate": 4.975145730097364e-06,
      "loss": 0.31664488315582273,
      "num_input_tokens_seen": 766680816,
      "step": 14900,
      "train_runtime": 91815.9417,
      "train_tokens_per_second": 8350.193
    },
    {
      "epoch": 1.0035685429571775,
      "grad_norm": 0.6423021830575799,
      "learning_rate": 4.97250168420758e-06,
      "loss": 0.31142263412475585,
      "num_input_tokens_seen": 766925248,
      "step": 14905,
      "train_runtime": 91849.7045,
      "train_tokens_per_second": 8349.785
    },
    {
      "epoch": 1.0039051979531377,
      "grad_norm": 0.615350828008354,
      "learning_rate": 4.96985764600761e-06,
      "loss": 0.3118563175201416,
      "num_input_tokens_seen": 767187816,
      "step": 14910,
      "train_runtime": 91878.793,
      "train_tokens_per_second": 8349.999
    },
    {
      "epoch": 1.0042418529490977,
      "grad_norm": 0.6232151546297223,
      "learning_rate": 4.967213616236855e-06,
      "loss": 0.32876875400543215,
      "num_input_tokens_seen": 767435152,
      "step": 14915,
      "train_runtime": 91909.3338,
      "train_tokens_per_second": 8349.915
    },
    {
      "epoch": 1.004578507945058,
      "grad_norm": 0.6494546609461149,
      "learning_rate": 4.964569595634706e-06,
      "loss": 0.314912748336792,
      "num_input_tokens_seen": 767691560,
      "step": 14920,
      "train_runtime": 91937.9356,
      "train_tokens_per_second": 8350.107
    },
    {
      "epoch": 1.004915162941018,
      "grad_norm": 0.6266535575614761,
      "learning_rate": 4.961925584940559e-06,
      "loss": 0.3316687345504761,
      "num_input_tokens_seen": 767954896,
      "step": 14925,
      "train_runtime": 91968.8312,
      "train_tokens_per_second": 8350.165
    },
    {
      "epoch": 1.0052518179369783,
      "grad_norm": 0.7185997978342075,
      "learning_rate": 4.959281584893799e-06,
      "loss": 0.2799243927001953,
      "num_input_tokens_seen": 768208432,
      "step": 14930,
      "train_runtime": 91997.664,
      "train_tokens_per_second": 8350.304
    },
    {
      "epoch": 1.0055884729329383,
      "grad_norm": 0.7148756303281492,
      "learning_rate": 4.956637596233814e-06,
      "loss": 0.33184566497802737,
      "num_input_tokens_seen": 768436808,
      "step": 14935,
      "train_runtime": 92026.8937,
      "train_tokens_per_second": 8350.133
    },
    {
      "epoch": 1.0059251279288985,
      "grad_norm": 0.6691804532424843,
      "learning_rate": 4.953993619699986e-06,
      "loss": 0.28480091094970705,
      "num_input_tokens_seen": 768698960,
      "step": 14940,
      "train_runtime": 92054.9802,
      "train_tokens_per_second": 8350.433
    },
    {
      "epoch": 1.0062617829248586,
      "grad_norm": 0.6430020992985699,
      "learning_rate": 4.951349656031698e-06,
      "loss": 0.30924439430236816,
      "num_input_tokens_seen": 768939736,
      "step": 14945,
      "train_runtime": 92082.3986,
      "train_tokens_per_second": 8350.562
    },
    {
      "epoch": 1.0065984379208188,
      "grad_norm": 0.6328475673733119,
      "learning_rate": 4.948705705968322e-06,
      "loss": 0.31339619159698484,
      "num_input_tokens_seen": 769203640,
      "step": 14950,
      "train_runtime": 92112.5979,
      "train_tokens_per_second": 8350.689
    },
    {
      "epoch": 1.0069350929167789,
      "grad_norm": 0.6188406280926582,
      "learning_rate": 4.946061770249233e-06,
      "loss": 0.3146347522735596,
      "num_input_tokens_seen": 769453240,
      "step": 14955,
      "train_runtime": 92147.8073,
      "train_tokens_per_second": 8350.207
    },
    {
      "epoch": 1.007271747912739,
      "grad_norm": 0.6895118481586512,
      "learning_rate": 4.9434178496137955e-06,
      "loss": 0.3305472135543823,
      "num_input_tokens_seen": 769708304,
      "step": 14960,
      "train_runtime": 92180.2748,
      "train_tokens_per_second": 8350.033
    },
    {
      "epoch": 1.0076084029086991,
      "grad_norm": 0.5187130268232959,
      "learning_rate": 4.940773944801378e-06,
      "loss": 0.31608099937438966,
      "num_input_tokens_seen": 769968488,
      "step": 14965,
      "train_runtime": 92212.0008,
      "train_tokens_per_second": 8349.981
    },
    {
      "epoch": 1.0079450579046594,
      "grad_norm": 0.7492400560256901,
      "learning_rate": 4.938130056551337e-06,
      "loss": 0.32681200504302976,
      "num_input_tokens_seen": 770210912,
      "step": 14970,
      "train_runtime": 92244.501,
      "train_tokens_per_second": 8349.667
    },
    {
      "epoch": 1.0082817129006194,
      "grad_norm": 0.672958389930208,
      "learning_rate": 4.9354861856030305e-06,
      "loss": 0.32323408126831055,
      "num_input_tokens_seen": 770473832,
      "step": 14975,
      "train_runtime": 92274.6024,
      "train_tokens_per_second": 8349.793
    },
    {
      "epoch": 1.0086183678965797,
      "grad_norm": 0.6975823870137224,
      "learning_rate": 4.932842332695803e-06,
      "loss": 0.2915213108062744,
      "num_input_tokens_seen": 770712848,
      "step": 14980,
      "train_runtime": 92308.743,
      "train_tokens_per_second": 8349.294
    },
    {
      "epoch": 1.0089550228925397,
      "grad_norm": 0.5757540195793606,
      "learning_rate": 4.9301984985690055e-06,
      "loss": 0.31422791481018064,
      "num_input_tokens_seen": 770992112,
      "step": 14985,
      "train_runtime": 92345.1337,
      "train_tokens_per_second": 8349.028
    },
    {
      "epoch": 1.0092916778885,
      "grad_norm": 0.7840028070342749,
      "learning_rate": 4.927554683961974e-06,
      "loss": 0.3297339916229248,
      "num_input_tokens_seen": 771236064,
      "step": 14990,
      "train_runtime": 92374.6012,
      "train_tokens_per_second": 8349.006
    },
    {
      "epoch": 1.00962833288446,
      "grad_norm": 0.6226143001977579,
      "learning_rate": 4.924910889614046e-06,
      "loss": 0.30892233848571776,
      "num_input_tokens_seen": 771499712,
      "step": 14995,
      "train_runtime": 92411.6495,
      "train_tokens_per_second": 8348.511
    },
    {
      "epoch": 1.0099649878804202,
      "grad_norm": 0.6381048003277241,
      "learning_rate": 4.922267116264546e-06,
      "loss": 0.33637826442718505,
      "num_input_tokens_seen": 771748408,
      "step": 15000,
      "train_runtime": 92441.4138,
      "train_tokens_per_second": 8348.514
    },
    {
      "epoch": 1.0103016428763802,
      "grad_norm": 0.7134208956599921,
      "learning_rate": 4.919623364652801e-06,
      "loss": 0.30307559967041015,
      "num_input_tokens_seen": 772004640,
      "step": 15005,
      "train_runtime": 92469.0997,
      "train_tokens_per_second": 8348.785
    },
    {
      "epoch": 1.0106382978723405,
      "grad_norm": 0.693294063707901,
      "learning_rate": 4.916979635518124e-06,
      "loss": 0.3179266929626465,
      "num_input_tokens_seen": 772238336,
      "step": 15010,
      "train_runtime": 92497.2112,
      "train_tokens_per_second": 8348.774
    },
    {
      "epoch": 1.0109749528683005,
      "grad_norm": 0.6269857829293229,
      "learning_rate": 4.914335929599828e-06,
      "loss": 0.3171516418457031,
      "num_input_tokens_seen": 772510824,
      "step": 15015,
      "train_runtime": 92524.7113,
      "train_tokens_per_second": 8349.238
    },
    {
      "epoch": 1.0113116078642608,
      "grad_norm": 0.69837604370234,
      "learning_rate": 4.9116922476372155e-06,
      "loss": 0.3218095779418945,
      "num_input_tokens_seen": 772768904,
      "step": 15020,
      "train_runtime": 92551.6633,
      "train_tokens_per_second": 8349.595
    },
    {
      "epoch": 1.0116482628602208,
      "grad_norm": 0.7113731397586375,
      "learning_rate": 4.909048590369586e-06,
      "loss": 0.3203019380569458,
      "num_input_tokens_seen": 773037752,
      "step": 15025,
      "train_runtime": 92583.7903,
      "train_tokens_per_second": 8349.601
    },
    {
      "epoch": 1.011984917856181,
      "grad_norm": 0.6595239440962553,
      "learning_rate": 4.906404958536227e-06,
      "loss": 0.316718316078186,
      "num_input_tokens_seen": 773291920,
      "step": 15030,
      "train_runtime": 92613.6809,
      "train_tokens_per_second": 8349.651
    },
    {
      "epoch": 1.012321572852141,
      "grad_norm": 0.7048969791515814,
      "learning_rate": 4.903761352876422e-06,
      "loss": 0.3176868915557861,
      "num_input_tokens_seen": 773500008,
      "step": 15035,
      "train_runtime": 92641.0177,
      "train_tokens_per_second": 8349.433
    },
    {
      "epoch": 1.0126582278481013,
      "grad_norm": 0.7326156896405658,
      "learning_rate": 4.9011177741294485e-06,
      "loss": 0.3358032464981079,
      "num_input_tokens_seen": 773745480,
      "step": 15040,
      "train_runtime": 92672.6457,
      "train_tokens_per_second": 8349.233
    },
    {
      "epoch": 1.0129948828440614,
      "grad_norm": 0.570130743214808,
      "learning_rate": 4.898474223034576e-06,
      "loss": 0.3094195365905762,
      "num_input_tokens_seen": 774012608,
      "step": 15045,
      "train_runtime": 92706.2828,
      "train_tokens_per_second": 8349.085
    },
    {
      "epoch": 1.0133315378400216,
      "grad_norm": 0.6266426153448574,
      "learning_rate": 4.895830700331061e-06,
      "loss": 0.3052509307861328,
      "num_input_tokens_seen": 774250320,
      "step": 15050,
      "train_runtime": 92737.1183,
      "train_tokens_per_second": 8348.872
    },
    {
      "epoch": 1.0136681928359816,
      "grad_norm": 0.6502038260616333,
      "learning_rate": 4.893187206758158e-06,
      "loss": 0.30704851150512696,
      "num_input_tokens_seen": 774537848,
      "step": 15055,
      "train_runtime": 92767.8123,
      "train_tokens_per_second": 8349.209
    },
    {
      "epoch": 1.014004847831942,
      "grad_norm": 1.3596754078085211,
      "learning_rate": 4.890543743055115e-06,
      "loss": 0.30096116065979006,
      "num_input_tokens_seen": 774795176,
      "step": 15060,
      "train_runtime": 92796.195,
      "train_tokens_per_second": 8349.428
    },
    {
      "epoch": 1.014341502827902,
      "grad_norm": 0.6017542732281932,
      "learning_rate": 4.887900309961163e-06,
      "loss": 0.3151562690734863,
      "num_input_tokens_seen": 775060640,
      "step": 15065,
      "train_runtime": 92828.134,
      "train_tokens_per_second": 8349.415
    },
    {
      "epoch": 1.0146781578238622,
      "grad_norm": 0.6242942090036027,
      "learning_rate": 4.885256908215532e-06,
      "loss": 0.3258845567703247,
      "num_input_tokens_seen": 775341008,
      "step": 15070,
      "train_runtime": 92858.8255,
      "train_tokens_per_second": 8349.675
    },
    {
      "epoch": 1.0150148128198222,
      "grad_norm": 0.7289311827651086,
      "learning_rate": 4.88261353855744e-06,
      "loss": 0.29760816097259524,
      "num_input_tokens_seen": 775608448,
      "step": 15075,
      "train_runtime": 92889.3041,
      "train_tokens_per_second": 8349.814
    },
    {
      "epoch": 1.0153514678157824,
      "grad_norm": 0.6058566931627722,
      "learning_rate": 4.8799702017261e-06,
      "loss": 0.2893998622894287,
      "num_input_tokens_seen": 775860488,
      "step": 15080,
      "train_runtime": 92924.5677,
      "train_tokens_per_second": 8349.358
    },
    {
      "epoch": 1.0156881228117425,
      "grad_norm": 0.7572033664265153,
      "learning_rate": 4.877326898460709e-06,
      "loss": 0.31896772384643557,
      "num_input_tokens_seen": 776104392,
      "step": 15085,
      "train_runtime": 92953.4685,
      "train_tokens_per_second": 8349.386
    },
    {
      "epoch": 1.0160247778077027,
      "grad_norm": 0.6304288878247143,
      "learning_rate": 4.87468362950046e-06,
      "loss": 0.2908316135406494,
      "num_input_tokens_seen": 776372584,
      "step": 15090,
      "train_runtime": 92985.4515,
      "train_tokens_per_second": 8349.398
    },
    {
      "epoch": 1.0163614328036628,
      "grad_norm": 0.6544430246694305,
      "learning_rate": 4.872040395584533e-06,
      "loss": 0.3058916091918945,
      "num_input_tokens_seen": 776609904,
      "step": 15095,
      "train_runtime": 93020.0501,
      "train_tokens_per_second": 8348.844
    },
    {
      "epoch": 1.016698087799623,
      "grad_norm": 0.7006409861399933,
      "learning_rate": 4.869397197452104e-06,
      "loss": 0.2816466331481934,
      "num_input_tokens_seen": 776867600,
      "step": 15100,
      "train_runtime": 93050.798,
      "train_tokens_per_second": 8348.855
    },
    {
      "epoch": 1.017034742795583,
      "grad_norm": 0.6947893500039443,
      "learning_rate": 4.866754035842329e-06,
      "loss": 0.29557826519012453,
      "num_input_tokens_seen": 777130304,
      "step": 15105,
      "train_runtime": 93080.2997,
      "train_tokens_per_second": 8349.031
    },
    {
      "epoch": 1.0173713977915433,
      "grad_norm": 0.6290890295052495,
      "learning_rate": 4.8641109114943666e-06,
      "loss": 0.2872951030731201,
      "num_input_tokens_seen": 777389584,
      "step": 15110,
      "train_runtime": 93109.6602,
      "train_tokens_per_second": 8349.183
    },
    {
      "epoch": 1.0177080527875033,
      "grad_norm": 0.6061230366904564,
      "learning_rate": 4.861467825147352e-06,
      "loss": 0.3256107807159424,
      "num_input_tokens_seen": 777660448,
      "step": 15115,
      "train_runtime": 93135.6663,
      "train_tokens_per_second": 8349.76
    },
    {
      "epoch": 1.0180447077834636,
      "grad_norm": 0.6063993917573989,
      "learning_rate": 4.858824777540419e-06,
      "loss": 0.2960064888000488,
      "num_input_tokens_seen": 777912040,
      "step": 15120,
      "train_runtime": 93165.1237,
      "train_tokens_per_second": 8349.82
    },
    {
      "epoch": 1.0183813627794236,
      "grad_norm": 0.6418021750838492,
      "learning_rate": 4.856181769412686e-06,
      "loss": 0.29643323421478274,
      "num_input_tokens_seen": 778146784,
      "step": 15125,
      "train_runtime": 93196.7556,
      "train_tokens_per_second": 8349.505
    },
    {
      "epoch": 1.0187180177753838,
      "grad_norm": 0.7585980485372937,
      "learning_rate": 4.8535388015032644e-06,
      "loss": 0.3028522729873657,
      "num_input_tokens_seen": 778391208,
      "step": 15130,
      "train_runtime": 93224.287,
      "train_tokens_per_second": 8349.661
    },
    {
      "epoch": 1.0190546727713439,
      "grad_norm": 0.5664580053742905,
      "learning_rate": 4.850895874551248e-06,
      "loss": 0.3149651288986206,
      "num_input_tokens_seen": 778662104,
      "step": 15135,
      "train_runtime": 93255.4505,
      "train_tokens_per_second": 8349.776
    },
    {
      "epoch": 1.0193913277673041,
      "grad_norm": 0.6467036774259785,
      "learning_rate": 4.848252989295726e-06,
      "loss": 0.3287036895751953,
      "num_input_tokens_seen": 778894928,
      "step": 15140,
      "train_runtime": 93281.2316,
      "train_tokens_per_second": 8349.964
    },
    {
      "epoch": 1.0197279827632642,
      "grad_norm": 0.633569900278401,
      "learning_rate": 4.845610146475771e-06,
      "loss": 0.29628787040710447,
      "num_input_tokens_seen": 779160192,
      "step": 15145,
      "train_runtime": 93315.7432,
      "train_tokens_per_second": 8349.719
    },
    {
      "epoch": 1.0200646377592244,
      "grad_norm": 0.6850562283777203,
      "learning_rate": 4.842967346830447e-06,
      "loss": 0.29238238334655764,
      "num_input_tokens_seen": 779412496,
      "step": 15150,
      "train_runtime": 93345.3378,
      "train_tokens_per_second": 8349.774
    },
    {
      "epoch": 1.0204012927551844,
      "grad_norm": 0.6917886235508459,
      "learning_rate": 4.840324591098803e-06,
      "loss": 0.3131434917449951,
      "num_input_tokens_seen": 779639776,
      "step": 15155,
      "train_runtime": 93373.3486,
      "train_tokens_per_second": 8349.704
    },
    {
      "epoch": 1.0207379477511447,
      "grad_norm": 0.7789561243635031,
      "learning_rate": 4.83768188001988e-06,
      "loss": 0.3021040439605713,
      "num_input_tokens_seen": 779868248,
      "step": 15160,
      "train_runtime": 93403.8776,
      "train_tokens_per_second": 8349.42
    },
    {
      "epoch": 1.0210746027471047,
      "grad_norm": 0.726656143768006,
      "learning_rate": 4.835039214332699e-06,
      "loss": 0.3078037738800049,
      "num_input_tokens_seen": 780094416,
      "step": 15165,
      "train_runtime": 93436.655,
      "train_tokens_per_second": 8348.912
    },
    {
      "epoch": 1.021411257743065,
      "grad_norm": 0.6485429718299148,
      "learning_rate": 4.832396594776277e-06,
      "loss": 0.30429506301879883,
      "num_input_tokens_seen": 780354008,
      "step": 15170,
      "train_runtime": 93467.6393,
      "train_tokens_per_second": 8348.922
    },
    {
      "epoch": 1.021747912739025,
      "grad_norm": 0.7360768016303965,
      "learning_rate": 4.82975402208961e-06,
      "loss": 0.28963651657104494,
      "num_input_tokens_seen": 780614944,
      "step": 15175,
      "train_runtime": 93494.2117,
      "train_tokens_per_second": 8349.34
    },
    {
      "epoch": 1.0220845677349852,
      "grad_norm": 0.6695696379500725,
      "learning_rate": 4.827111497011692e-06,
      "loss": 0.2888345241546631,
      "num_input_tokens_seen": 780908704,
      "step": 15180,
      "train_runtime": 93525.1365,
      "train_tokens_per_second": 8349.72
    },
    {
      "epoch": 1.0224212227309453,
      "grad_norm": 0.7833775621664594,
      "learning_rate": 4.82446902028149e-06,
      "loss": 0.3134031295776367,
      "num_input_tokens_seen": 781174784,
      "step": 15185,
      "train_runtime": 93555.3995,
      "train_tokens_per_second": 8349.863
    },
    {
      "epoch": 1.0227578777269055,
      "grad_norm": 0.670752810236938,
      "learning_rate": 4.821826592637968e-06,
      "loss": 0.3243523120880127,
      "num_input_tokens_seen": 781438208,
      "step": 15190,
      "train_runtime": 93583.4596,
      "train_tokens_per_second": 8350.174
    },
    {
      "epoch": 1.0230945327228655,
      "grad_norm": 0.6076943013213648,
      "learning_rate": 4.8191842148200694e-06,
      "loss": 0.29744279384613037,
      "num_input_tokens_seen": 781713608,
      "step": 15195,
      "train_runtime": 93613.6649,
      "train_tokens_per_second": 8350.422
    },
    {
      "epoch": 1.0234311877188258,
      "grad_norm": 0.8405548175384202,
      "learning_rate": 4.8165418875667315e-06,
      "loss": 0.3002565145492554,
      "num_input_tokens_seen": 781982928,
      "step": 15200,
      "train_runtime": 93641.5988,
      "train_tokens_per_second": 8350.807
    },
    {
      "epoch": 1.0237678427147858,
      "grad_norm": 0.5661179717115511,
      "learning_rate": 4.813899611616867e-06,
      "loss": 0.32268385887145995,
      "num_input_tokens_seen": 782247728,
      "step": 15205,
      "train_runtime": 93669.5983,
      "train_tokens_per_second": 8351.138
    },
    {
      "epoch": 1.024104497710746,
      "grad_norm": 0.7343043957094861,
      "learning_rate": 4.8112573877093855e-06,
      "loss": 0.30162479877471926,
      "num_input_tokens_seen": 782532768,
      "step": 15210,
      "train_runtime": 93699.9702,
      "train_tokens_per_second": 8351.473
    },
    {
      "epoch": 1.024441152706706,
      "grad_norm": 0.6280210131273671,
      "learning_rate": 4.808615216583171e-06,
      "loss": 0.31537761688232424,
      "num_input_tokens_seen": 782791304,
      "step": 15215,
      "train_runtime": 93731.9789,
      "train_tokens_per_second": 8351.379
    },
    {
      "epoch": 1.0247778077026664,
      "grad_norm": 0.7178320975426257,
      "learning_rate": 4.805973098977102e-06,
      "loss": 0.3200103282928467,
      "num_input_tokens_seen": 783034328,
      "step": 15220,
      "train_runtime": 93762.1816,
      "train_tokens_per_second": 8351.281
    },
    {
      "epoch": 1.0251144626986264,
      "grad_norm": 0.6608909994390402,
      "learning_rate": 4.803331035630037e-06,
      "loss": 0.32470428943634033,
      "num_input_tokens_seen": 783305824,
      "step": 15225,
      "train_runtime": 93793.6356,
      "train_tokens_per_second": 8351.375
    },
    {
      "epoch": 1.0254511176945866,
      "grad_norm": 0.5760486855455214,
      "learning_rate": 4.800689027280822e-06,
      "loss": 0.2915608882904053,
      "num_input_tokens_seen": 783563232,
      "step": 15230,
      "train_runtime": 93828.1492,
      "train_tokens_per_second": 8351.046
    },
    {
      "epoch": 1.0257877726905467,
      "grad_norm": 0.7052035160794182,
      "learning_rate": 4.798047074668283e-06,
      "loss": 0.2981018781661987,
      "num_input_tokens_seen": 783815120,
      "step": 15235,
      "train_runtime": 93858.6414,
      "train_tokens_per_second": 8351.017
    },
    {
      "epoch": 1.026124427686507,
      "grad_norm": 0.611529196293041,
      "learning_rate": 4.795405178531236e-06,
      "loss": 0.2836286544799805,
      "num_input_tokens_seen": 784064184,
      "step": 15240,
      "train_runtime": 93884.8303,
      "train_tokens_per_second": 8351.34
    },
    {
      "epoch": 1.026461082682467,
      "grad_norm": 0.6787538736606665,
      "learning_rate": 4.792763339608478e-06,
      "loss": 0.30867910385131836,
      "num_input_tokens_seen": 784301808,
      "step": 15245,
      "train_runtime": 93911.7429,
      "train_tokens_per_second": 8351.477
    },
    {
      "epoch": 1.0267977376784272,
      "grad_norm": 0.6977268497341499,
      "learning_rate": 4.790121558638793e-06,
      "loss": 0.2761770486831665,
      "num_input_tokens_seen": 784546112,
      "step": 15250,
      "train_runtime": 93943.8774,
      "train_tokens_per_second": 8351.221
    },
    {
      "epoch": 1.0271343926743872,
      "grad_norm": 0.5681659349652446,
      "learning_rate": 4.787479836360944e-06,
      "loss": 0.3260411500930786,
      "num_input_tokens_seen": 784806032,
      "step": 15255,
      "train_runtime": 93974.9633,
      "train_tokens_per_second": 8351.225
    },
    {
      "epoch": 1.0274710476703475,
      "grad_norm": 0.641979553707752,
      "learning_rate": 4.784838173513682e-06,
      "loss": 0.3096033573150635,
      "num_input_tokens_seen": 785065960,
      "step": 15260,
      "train_runtime": 94008.0494,
      "train_tokens_per_second": 8351.05
    },
    {
      "epoch": 1.0278077026663075,
      "grad_norm": 0.8025053696188786,
      "learning_rate": 4.782196570835737e-06,
      "loss": 0.3049262046813965,
      "num_input_tokens_seen": 785329432,
      "step": 15265,
      "train_runtime": 94040.6942,
      "train_tokens_per_second": 8350.953
    },
    {
      "epoch": 1.0281443576622677,
      "grad_norm": 0.7992750170041185,
      "learning_rate": 4.779555029065827e-06,
      "loss": 0.2899137020111084,
      "num_input_tokens_seen": 785557712,
      "step": 15270,
      "train_runtime": 94069.2432,
      "train_tokens_per_second": 8350.845
    },
    {
      "epoch": 1.0284810126582278,
      "grad_norm": 0.6649636078368687,
      "learning_rate": 4.776913548942649e-06,
      "loss": 0.31001992225646974,
      "num_input_tokens_seen": 785827256,
      "step": 15275,
      "train_runtime": 94099.6422,
      "train_tokens_per_second": 8351.012
    },
    {
      "epoch": 1.028817667654188,
      "grad_norm": 0.6708645827507029,
      "learning_rate": 4.774272131204889e-06,
      "loss": 0.32271533012390136,
      "num_input_tokens_seen": 786085392,
      "step": 15280,
      "train_runtime": 94129.6709,
      "train_tokens_per_second": 8351.09
    },
    {
      "epoch": 1.029154322650148,
      "grad_norm": 0.6751908667727994,
      "learning_rate": 4.7716307765912045e-06,
      "loss": 0.31547017097473146,
      "num_input_tokens_seen": 786340080,
      "step": 15285,
      "train_runtime": 94158.7221,
      "train_tokens_per_second": 8351.219
    },
    {
      "epoch": 1.0294909776461083,
      "grad_norm": 0.53098020881673,
      "learning_rate": 4.768989485840246e-06,
      "loss": 0.3139127016067505,
      "num_input_tokens_seen": 786623576,
      "step": 15290,
      "train_runtime": 94187.4907,
      "train_tokens_per_second": 8351.678
    },
    {
      "epoch": 1.0298276326420683,
      "grad_norm": 0.6550538546282838,
      "learning_rate": 4.766348259690641e-06,
      "loss": 0.28084578514099123,
      "num_input_tokens_seen": 786904040,
      "step": 15295,
      "train_runtime": 94220.3907,
      "train_tokens_per_second": 8351.738
    },
    {
      "epoch": 1.0301642876380286,
      "grad_norm": 0.7178197741659091,
      "learning_rate": 4.763707098881002e-06,
      "loss": 0.30819058418273926,
      "num_input_tokens_seen": 787159080,
      "step": 15300,
      "train_runtime": 94250.6189,
      "train_tokens_per_second": 8351.766
    },
    {
      "epoch": 1.0305009426339886,
      "grad_norm": 0.7249252621587535,
      "learning_rate": 4.761066004149917e-06,
      "loss": 0.3271448612213135,
      "num_input_tokens_seen": 787396088,
      "step": 15305,
      "train_runtime": 94280.406,
      "train_tokens_per_second": 8351.641
    },
    {
      "epoch": 1.0308375976299489,
      "grad_norm": 0.7038034023494721,
      "learning_rate": 4.758424976235964e-06,
      "loss": 0.3069882869720459,
      "num_input_tokens_seen": 787658056,
      "step": 15310,
      "train_runtime": 94310.2689,
      "train_tokens_per_second": 8351.774
    },
    {
      "epoch": 1.031174252625909,
      "grad_norm": 0.6506482566937491,
      "learning_rate": 4.755784015877694e-06,
      "loss": 0.3205685615539551,
      "num_input_tokens_seen": 787952560,
      "step": 15315,
      "train_runtime": 94338.861,
      "train_tokens_per_second": 8352.365
    },
    {
      "epoch": 1.0315109076218691,
      "grad_norm": 0.6736718291271461,
      "learning_rate": 4.753143123813646e-06,
      "loss": 0.3196721076965332,
      "num_input_tokens_seen": 788209392,
      "step": 15320,
      "train_runtime": 94371.4944,
      "train_tokens_per_second": 8352.198
    },
    {
      "epoch": 1.0318475626178292,
      "grad_norm": 0.7968501118224294,
      "learning_rate": 4.7505023007823365e-06,
      "loss": 0.3078810214996338,
      "num_input_tokens_seen": 788465488,
      "step": 15325,
      "train_runtime": 94402.4261,
      "train_tokens_per_second": 8352.174
    },
    {
      "epoch": 1.0321842176137894,
      "grad_norm": 0.6068589422954311,
      "learning_rate": 4.747861547522264e-06,
      "loss": 0.28024132251739503,
      "num_input_tokens_seen": 788722848,
      "step": 15330,
      "train_runtime": 94433.0859,
      "train_tokens_per_second": 8352.188
    },
    {
      "epoch": 1.0325208726097495,
      "grad_norm": 0.6080389461052023,
      "learning_rate": 4.745220864771904e-06,
      "loss": 0.29949126243591306,
      "num_input_tokens_seen": 788963432,
      "step": 15335,
      "train_runtime": 94466.8949,
      "train_tokens_per_second": 8351.745
    },
    {
      "epoch": 1.0328575276057097,
      "grad_norm": 2.010386123766531,
      "learning_rate": 4.742580253269718e-06,
      "loss": 0.3167204618453979,
      "num_input_tokens_seen": 789192128,
      "step": 15340,
      "train_runtime": 94493.9579,
      "train_tokens_per_second": 8351.773
    },
    {
      "epoch": 1.0331941826016697,
      "grad_norm": 0.8251401412230449,
      "learning_rate": 4.739939713754141e-06,
      "loss": 0.2982097387313843,
      "num_input_tokens_seen": 789428792,
      "step": 15345,
      "train_runtime": 94522.1298,
      "train_tokens_per_second": 8351.788
    },
    {
      "epoch": 1.03353083759763,
      "grad_norm": 0.7364326475309304,
      "learning_rate": 4.737299246963597e-06,
      "loss": 0.30305795669555663,
      "num_input_tokens_seen": 789663104,
      "step": 15350,
      "train_runtime": 94554.359,
      "train_tokens_per_second": 8351.419
    },
    {
      "epoch": 1.03386749259359,
      "grad_norm": 0.5877296366252737,
      "learning_rate": 4.734658853636479e-06,
      "loss": 0.30371832847595215,
      "num_input_tokens_seen": 789940656,
      "step": 15355,
      "train_runtime": 94585.5486,
      "train_tokens_per_second": 8351.6
    },
    {
      "epoch": 1.0342041475895503,
      "grad_norm": 0.7106774642355852,
      "learning_rate": 4.732018534511167e-06,
      "loss": 0.3331176280975342,
      "num_input_tokens_seen": 790201064,
      "step": 15360,
      "train_runtime": 94618.7141,
      "train_tokens_per_second": 8351.425
    },
    {
      "epoch": 1.0345408025855103,
      "grad_norm": 0.5816842783767852,
      "learning_rate": 4.729378290326017e-06,
      "loss": 0.28656654357910155,
      "num_input_tokens_seen": 790448184,
      "step": 15365,
      "train_runtime": 94647.6991,
      "train_tokens_per_second": 8351.478
    },
    {
      "epoch": 1.0348774575814705,
      "grad_norm": 0.6101788544439645,
      "learning_rate": 4.726738121819365e-06,
      "loss": 0.3154762744903564,
      "num_input_tokens_seen": 790711176,
      "step": 15370,
      "train_runtime": 94680.2434,
      "train_tokens_per_second": 8351.385
    },
    {
      "epoch": 1.0352141125774306,
      "grad_norm": 0.7226304542055705,
      "learning_rate": 4.7240980297295255e-06,
      "loss": 0.27692904472351076,
      "num_input_tokens_seen": 790953272,
      "step": 15375,
      "train_runtime": 94712.4272,
      "train_tokens_per_second": 8351.103
    },
    {
      "epoch": 1.0355507675733908,
      "grad_norm": 0.5264383835148649,
      "learning_rate": 4.721458014794795e-06,
      "loss": 0.3037320613861084,
      "num_input_tokens_seen": 791228544,
      "step": 15380,
      "train_runtime": 94746.7864,
      "train_tokens_per_second": 8350.98
    },
    {
      "epoch": 1.0358874225693508,
      "grad_norm": 0.6923664083560249,
      "learning_rate": 4.718818077753439e-06,
      "loss": 0.3228918075561523,
      "num_input_tokens_seen": 791511368,
      "step": 15385,
      "train_runtime": 94778.8052,
      "train_tokens_per_second": 8351.143
    },
    {
      "epoch": 1.036224077565311,
      "grad_norm": 0.6221417647522788,
      "learning_rate": 4.716178219343712e-06,
      "loss": 0.30662922859191893,
      "num_input_tokens_seen": 791771016,
      "step": 15390,
      "train_runtime": 94807.8476,
      "train_tokens_per_second": 8351.324
    },
    {
      "epoch": 1.0365607325612711,
      "grad_norm": 0.9053307847772525,
      "learning_rate": 4.713538440303841e-06,
      "loss": 0.315276575088501,
      "num_input_tokens_seen": 791995784,
      "step": 15395,
      "train_runtime": 94836.7252,
      "train_tokens_per_second": 8351.151
    },
    {
      "epoch": 1.0368973875572314,
      "grad_norm": 0.6611566752692288,
      "learning_rate": 4.7108987413720335e-06,
      "loss": 0.31512951850891113,
      "num_input_tokens_seen": 792240832,
      "step": 15400,
      "train_runtime": 94866.5999,
      "train_tokens_per_second": 8351.104
    },
    {
      "epoch": 1.0372340425531914,
      "grad_norm": 0.8229324173696816,
      "learning_rate": 4.708259123286469e-06,
      "loss": 0.3153153657913208,
      "num_input_tokens_seen": 792489528,
      "step": 15405,
      "train_runtime": 94898.8256,
      "train_tokens_per_second": 8350.889
    },
    {
      "epoch": 1.0375706975491517,
      "grad_norm": 0.6493358868318335,
      "learning_rate": 4.7056195867853105e-06,
      "loss": 0.30100345611572266,
      "num_input_tokens_seen": 792757976,
      "step": 15410,
      "train_runtime": 94929.7537,
      "train_tokens_per_second": 8350.996
    },
    {
      "epoch": 1.0379073525451117,
      "grad_norm": 0.66328843454707,
      "learning_rate": 4.702980132606696e-06,
      "loss": 0.2830946445465088,
      "num_input_tokens_seen": 793036808,
      "step": 15415,
      "train_runtime": 94956.8336,
      "train_tokens_per_second": 8351.551
    },
    {
      "epoch": 1.038244007541072,
      "grad_norm": 0.7228542173872321,
      "learning_rate": 4.700340761488741e-06,
      "loss": 0.29207863807678225,
      "num_input_tokens_seen": 793299432,
      "step": 15420,
      "train_runtime": 94988.121,
      "train_tokens_per_second": 8351.565
    },
    {
      "epoch": 1.038580662537032,
      "grad_norm": 0.6336382230132335,
      "learning_rate": 4.6977014741695345e-06,
      "loss": 0.28584580421447753,
      "num_input_tokens_seen": 793560800,
      "step": 15425,
      "train_runtime": 95016.1344,
      "train_tokens_per_second": 8351.853
    },
    {
      "epoch": 1.0389173175329922,
      "grad_norm": 0.6428487909543191,
      "learning_rate": 4.69506227138715e-06,
      "loss": 0.3088440179824829,
      "num_input_tokens_seen": 793844264,
      "step": 15430,
      "train_runtime": 95050.388,
      "train_tokens_per_second": 8351.826
    },
    {
      "epoch": 1.0392539725289522,
      "grad_norm": 0.6743983935842776,
      "learning_rate": 4.6924231538796265e-06,
      "loss": 0.3357588768005371,
      "num_input_tokens_seen": 794116000,
      "step": 15435,
      "train_runtime": 95082.043,
      "train_tokens_per_second": 8351.903
    },
    {
      "epoch": 1.0395906275249125,
      "grad_norm": 0.6906432898720729,
      "learning_rate": 4.6897841223849885e-06,
      "loss": 0.31035866737365725,
      "num_input_tokens_seen": 794362872,
      "step": 15440,
      "train_runtime": 95116.0981,
      "train_tokens_per_second": 8351.508
    },
    {
      "epoch": 1.0399272825208725,
      "grad_norm": 0.6865393280160691,
      "learning_rate": 4.68714517764123e-06,
      "loss": 0.2838796377182007,
      "num_input_tokens_seen": 794618368,
      "step": 15445,
      "train_runtime": 95148.3103,
      "train_tokens_per_second": 8351.366
    },
    {
      "epoch": 1.0402639375168328,
      "grad_norm": 0.5728253023467373,
      "learning_rate": 4.684506320386327e-06,
      "loss": 0.29650976657867434,
      "num_input_tokens_seen": 794849656,
      "step": 15450,
      "train_runtime": 95175.0167,
      "train_tokens_per_second": 8351.453
    },
    {
      "epoch": 1.0406005925127928,
      "grad_norm": 2.604582967192607,
      "learning_rate": 4.6818675513582235e-06,
      "loss": 0.285698676109314,
      "num_input_tokens_seen": 795094344,
      "step": 15455,
      "train_runtime": 95201.6083,
      "train_tokens_per_second": 8351.69
    },
    {
      "epoch": 1.040937247508753,
      "grad_norm": 0.6850403529100854,
      "learning_rate": 4.679228871294846e-06,
      "loss": 0.2871422290802002,
      "num_input_tokens_seen": 795357784,
      "step": 15460,
      "train_runtime": 95230.9303,
      "train_tokens_per_second": 8351.885
    },
    {
      "epoch": 1.041273902504713,
      "grad_norm": 0.6281388552908435,
      "learning_rate": 4.67659028093409e-06,
      "loss": 0.3151197910308838,
      "num_input_tokens_seen": 795617216,
      "step": 15465,
      "train_runtime": 95266.4885,
      "train_tokens_per_second": 8351.491
    },
    {
      "epoch": 1.0416105575006733,
      "grad_norm": 0.6543680875641293,
      "learning_rate": 4.673951781013834e-06,
      "loss": 0.3227803707122803,
      "num_input_tokens_seen": 795867392,
      "step": 15470,
      "train_runtime": 95297.7256,
      "train_tokens_per_second": 8351.379
    },
    {
      "epoch": 1.0419472124966334,
      "grad_norm": 0.5524681755481217,
      "learning_rate": 4.67131337227192e-06,
      "loss": 0.31559076309204104,
      "num_input_tokens_seen": 796140600,
      "step": 15475,
      "train_runtime": 95330.0394,
      "train_tokens_per_second": 8351.414
    },
    {
      "epoch": 1.0422838674925936,
      "grad_norm": 0.6503026750387748,
      "learning_rate": 4.668675055446175e-06,
      "loss": 0.33051512241363523,
      "num_input_tokens_seen": 796391792,
      "step": 15480,
      "train_runtime": 95363.1791,
      "train_tokens_per_second": 8351.146
    },
    {
      "epoch": 1.0426205224885536,
      "grad_norm": 0.7569955524087405,
      "learning_rate": 4.666036831274392e-06,
      "loss": 0.27715728282928465,
      "num_input_tokens_seen": 796643928,
      "step": 15485,
      "train_runtime": 95391.9699,
      "train_tokens_per_second": 8351.268
    },
    {
      "epoch": 1.0429571774845139,
      "grad_norm": 0.6662971526520505,
      "learning_rate": 4.663398700494346e-06,
      "loss": 0.30719428062438964,
      "num_input_tokens_seen": 796890992,
      "step": 15490,
      "train_runtime": 95419.2165,
      "train_tokens_per_second": 8351.473
    },
    {
      "epoch": 1.043293832480474,
      "grad_norm": 0.6364190941153873,
      "learning_rate": 4.66076066384378e-06,
      "loss": 0.3221767425537109,
      "num_input_tokens_seen": 797161232,
      "step": 15495,
      "train_runtime": 95448.6675,
      "train_tokens_per_second": 8351.727
    },
    {
      "epoch": 1.0436304874764342,
      "grad_norm": 0.6137679942502171,
      "learning_rate": 4.658122722060413e-06,
      "loss": 0.3077971935272217,
      "num_input_tokens_seen": 797429408,
      "step": 15500,
      "train_runtime": 95481.4254,
      "train_tokens_per_second": 8351.671
    },
    {
      "epoch": 1.0439671424723942,
      "grad_norm": 0.6918469849332286,
      "learning_rate": 4.655484875881935e-06,
      "loss": 0.3127863645553589,
      "num_input_tokens_seen": 797657408,
      "step": 15505,
      "train_runtime": 95514.7542,
      "train_tokens_per_second": 8351.143
    },
    {
      "epoch": 1.0443037974683544,
      "grad_norm": 0.6055163112684561,
      "learning_rate": 4.652847126046014e-06,
      "loss": 0.3275908470153809,
      "num_input_tokens_seen": 797878904,
      "step": 15510,
      "train_runtime": 95541.5188,
      "train_tokens_per_second": 8351.122
    },
    {
      "epoch": 1.0446404524643145,
      "grad_norm": 0.6194689951560682,
      "learning_rate": 4.650209473290286e-06,
      "loss": 0.3311581611633301,
      "num_input_tokens_seen": 798148776,
      "step": 15515,
      "train_runtime": 95572.5528,
      "train_tokens_per_second": 8351.234
    },
    {
      "epoch": 1.0449771074602747,
      "grad_norm": 0.6959131184747207,
      "learning_rate": 4.647571918352366e-06,
      "loss": 0.3128347396850586,
      "num_input_tokens_seen": 798399752,
      "step": 15520,
      "train_runtime": 95604.0382,
      "train_tokens_per_second": 8351.109
    },
    {
      "epoch": 1.0453137624562348,
      "grad_norm": 0.664235922896763,
      "learning_rate": 4.644934461969833e-06,
      "loss": 0.301772928237915,
      "num_input_tokens_seen": 798654672,
      "step": 15525,
      "train_runtime": 95634.5304,
      "train_tokens_per_second": 8351.112
    },
    {
      "epoch": 1.045650417452195,
      "grad_norm": 0.5793033655642059,
      "learning_rate": 4.642297104880247e-06,
      "loss": 0.2943442344665527,
      "num_input_tokens_seen": 798925232,
      "step": 15530,
      "train_runtime": 95669.8935,
      "train_tokens_per_second": 8350.853
    },
    {
      "epoch": 1.045987072448155,
      "grad_norm": 0.7433716375306815,
      "learning_rate": 4.639659847821133e-06,
      "loss": 0.2991571187973022,
      "num_input_tokens_seen": 799186136,
      "step": 15535,
      "train_runtime": 95699.8778,
      "train_tokens_per_second": 8350.963
    },
    {
      "epoch": 1.0463237274441153,
      "grad_norm": 0.6378533847115975,
      "learning_rate": 4.6370226915299945e-06,
      "loss": 0.29785001277923584,
      "num_input_tokens_seen": 799449488,
      "step": 15540,
      "train_runtime": 95730.8222,
      "train_tokens_per_second": 8351.015
    },
    {
      "epoch": 1.0466603824400753,
      "grad_norm": 0.5949632781225688,
      "learning_rate": 4.634385636744302e-06,
      "loss": 0.3093514919281006,
      "num_input_tokens_seen": 799713088,
      "step": 15545,
      "train_runtime": 95761.9891,
      "train_tokens_per_second": 8351.049
    },
    {
      "epoch": 1.0469970374360356,
      "grad_norm": 0.7084729097146005,
      "learning_rate": 4.631748684201503e-06,
      "loss": 0.32886857986450196,
      "num_input_tokens_seen": 799988984,
      "step": 15550,
      "train_runtime": 95794.926,
      "train_tokens_per_second": 8351.058
    },
    {
      "epoch": 1.0473336924319956,
      "grad_norm": 0.6503427188076387,
      "learning_rate": 4.629111834639008e-06,
      "loss": 0.3032204627990723,
      "num_input_tokens_seen": 800265296,
      "step": 15555,
      "train_runtime": 95823.963,
      "train_tokens_per_second": 8351.411
    },
    {
      "epoch": 1.0476703474279558,
      "grad_norm": 0.6132870756834028,
      "learning_rate": 4.626475088794206e-06,
      "loss": 0.29300580024719236,
      "num_input_tokens_seen": 800538408,
      "step": 15560,
      "train_runtime": 95853.806,
      "train_tokens_per_second": 8351.66
    },
    {
      "epoch": 1.0480070024239159,
      "grad_norm": 0.5988025791632424,
      "learning_rate": 4.623838447404454e-06,
      "loss": 0.31331744194030764,
      "num_input_tokens_seen": 800814712,
      "step": 15565,
      "train_runtime": 95885.4855,
      "train_tokens_per_second": 8351.782
    },
    {
      "epoch": 1.0483436574198761,
      "grad_norm": 0.563082910933,
      "learning_rate": 4.621201911207082e-06,
      "loss": 0.2929849147796631,
      "num_input_tokens_seen": 801065040,
      "step": 15570,
      "train_runtime": 95915.1146,
      "train_tokens_per_second": 8351.812
    },
    {
      "epoch": 1.0486803124158361,
      "grad_norm": 0.6911048699614594,
      "learning_rate": 4.618565480939387e-06,
      "loss": 0.32099735736846924,
      "num_input_tokens_seen": 801341728,
      "step": 15575,
      "train_runtime": 95950.3092,
      "train_tokens_per_second": 8351.633
    },
    {
      "epoch": 1.0490169674117964,
      "grad_norm": 0.704652955098657,
      "learning_rate": 4.6159291573386395e-06,
      "loss": 0.31405038833618165,
      "num_input_tokens_seen": 801602728,
      "step": 15580,
      "train_runtime": 95981.298,
      "train_tokens_per_second": 8351.655
    },
    {
      "epoch": 1.0493536224077564,
      "grad_norm": 0.6963777015214433,
      "learning_rate": 4.613292941142077e-06,
      "loss": 0.30476818084716795,
      "num_input_tokens_seen": 801856456,
      "step": 15585,
      "train_runtime": 96014.0707,
      "train_tokens_per_second": 8351.447
    },
    {
      "epoch": 1.0496902774037167,
      "grad_norm": 0.6495890766139463,
      "learning_rate": 4.610656833086912e-06,
      "loss": 0.31113150119781496,
      "num_input_tokens_seen": 802101080,
      "step": 15590,
      "train_runtime": 96044.3291,
      "train_tokens_per_second": 8351.363
    },
    {
      "epoch": 1.0500269323996767,
      "grad_norm": 0.7331127934995962,
      "learning_rate": 4.60802083391032e-06,
      "loss": 0.34190595149993896,
      "num_input_tokens_seen": 802371672,
      "step": 15595,
      "train_runtime": 96073.6423,
      "train_tokens_per_second": 8351.632
    },
    {
      "epoch": 1.050363587395637,
      "grad_norm": 0.8070981942889235,
      "learning_rate": 4.605384944349454e-06,
      "loss": 0.29348604679107665,
      "num_input_tokens_seen": 802616424,
      "step": 15600,
      "train_runtime": 96100.0068,
      "train_tokens_per_second": 8351.887
    },
    {
      "epoch": 1.050700242391597,
      "grad_norm": 0.7422329233299905,
      "learning_rate": 4.602749165141429e-06,
      "loss": 0.2912889003753662,
      "num_input_tokens_seen": 802847160,
      "step": 15605,
      "train_runtime": 96131.7273,
      "train_tokens_per_second": 8351.532
    },
    {
      "epoch": 1.0510368973875572,
      "grad_norm": 0.6763115520977678,
      "learning_rate": 4.600113497023333e-06,
      "loss": 0.31812753677368166,
      "num_input_tokens_seen": 803109456,
      "step": 15610,
      "train_runtime": 96163.9397,
      "train_tokens_per_second": 8351.462
    },
    {
      "epoch": 1.0513735523835175,
      "grad_norm": 0.7499693070490822,
      "learning_rate": 4.597477940732222e-06,
      "loss": 0.30219709873199463,
      "num_input_tokens_seen": 803340736,
      "step": 15615,
      "train_runtime": 96193.0741,
      "train_tokens_per_second": 8351.337
    },
    {
      "epoch": 1.0517102073794775,
      "grad_norm": 0.732750119615819,
      "learning_rate": 4.594842497005124e-06,
      "loss": 0.2944667816162109,
      "num_input_tokens_seen": 803618960,
      "step": 15620,
      "train_runtime": 96225.6872,
      "train_tokens_per_second": 8351.397
    },
    {
      "epoch": 1.0520468623754375,
      "grad_norm": 0.6186045318811447,
      "learning_rate": 4.5922071665790264e-06,
      "loss": 0.31853930950164794,
      "num_input_tokens_seen": 803869712,
      "step": 15625,
      "train_runtime": 96260.6781,
      "train_tokens_per_second": 8350.967
    },
    {
      "epoch": 1.0523835173713978,
      "grad_norm": 0.5807448883085781,
      "learning_rate": 4.589571950190897e-06,
      "loss": 0.30025424957275393,
      "num_input_tokens_seen": 804134384,
      "step": 15630,
      "train_runtime": 96293.9331,
      "train_tokens_per_second": 8350.831
    },
    {
      "epoch": 1.052720172367358,
      "grad_norm": 0.6106168830876476,
      "learning_rate": 4.586936848577661e-06,
      "loss": 0.3019110202789307,
      "num_input_tokens_seen": 804405840,
      "step": 15635,
      "train_runtime": 96324.8914,
      "train_tokens_per_second": 8350.965
    },
    {
      "epoch": 1.053056827363318,
      "grad_norm": 0.6515009069198882,
      "learning_rate": 4.584301862476219e-06,
      "loss": 0.31414499282836916,
      "num_input_tokens_seen": 804669432,
      "step": 15640,
      "train_runtime": 96355.8507,
      "train_tokens_per_second": 8351.018
    },
    {
      "epoch": 1.053393482359278,
      "grad_norm": 0.603638334182503,
      "learning_rate": 4.581666992623435e-06,
      "loss": 0.28463091850280764,
      "num_input_tokens_seen": 804943584,
      "step": 15645,
      "train_runtime": 96384.8038,
      "train_tokens_per_second": 8351.354
    },
    {
      "epoch": 1.0537301373552383,
      "grad_norm": 0.6562130509710659,
      "learning_rate": 4.579032239756144e-06,
      "loss": 0.3115992546081543,
      "num_input_tokens_seen": 805216312,
      "step": 15650,
      "train_runtime": 96418.4533,
      "train_tokens_per_second": 8351.268
    },
    {
      "epoch": 1.0540667923511986,
      "grad_norm": 0.5824329892200311,
      "learning_rate": 4.576397604611143e-06,
      "loss": 0.285434627532959,
      "num_input_tokens_seen": 805480400,
      "step": 15655,
      "train_runtime": 96449.783,
      "train_tokens_per_second": 8351.293
    },
    {
      "epoch": 1.0544034473471586,
      "grad_norm": 0.6841043172182464,
      "learning_rate": 4.573763087925202e-06,
      "loss": 0.3071301460266113,
      "num_input_tokens_seen": 805744488,
      "step": 15660,
      "train_runtime": 96480.6989,
      "train_tokens_per_second": 8351.354
    },
    {
      "epoch": 1.0547401023431189,
      "grad_norm": 0.6429284774520287,
      "learning_rate": 4.571128690435053e-06,
      "loss": 0.31865506172180175,
      "num_input_tokens_seen": 805997504,
      "step": 15665,
      "train_runtime": 96512.7382,
      "train_tokens_per_second": 8351.203
    },
    {
      "epoch": 1.055076757339079,
      "grad_norm": 0.7381136711633498,
      "learning_rate": 4.568494412877401e-06,
      "loss": 0.30289087295532224,
      "num_input_tokens_seen": 806238984,
      "step": 15670,
      "train_runtime": 96548.3507,
      "train_tokens_per_second": 8350.624
    },
    {
      "epoch": 1.0554134123350392,
      "grad_norm": 0.6703426355363162,
      "learning_rate": 4.565860255988908e-06,
      "loss": 0.31318044662475586,
      "num_input_tokens_seen": 806516464,
      "step": 15675,
      "train_runtime": 96578.0145,
      "train_tokens_per_second": 8350.932
    },
    {
      "epoch": 1.0557500673309992,
      "grad_norm": 0.6345403718189958,
      "learning_rate": 4.563226220506212e-06,
      "loss": 0.3183109283447266,
      "num_input_tokens_seen": 806811288,
      "step": 15680,
      "train_runtime": 96613.6823,
      "train_tokens_per_second": 8350.901
    },
    {
      "epoch": 1.0560867223269594,
      "grad_norm": 0.6286372888860233,
      "learning_rate": 4.5605923071659085e-06,
      "loss": 0.30764615535736084,
      "num_input_tokens_seen": 807057096,
      "step": 15685,
      "train_runtime": 96643.0988,
      "train_tokens_per_second": 8350.903
    },
    {
      "epoch": 1.0564233773229195,
      "grad_norm": 0.6677183126674162,
      "learning_rate": 4.557958516704566e-06,
      "loss": 0.2955521583557129,
      "num_input_tokens_seen": 807307312,
      "step": 15690,
      "train_runtime": 96671.4798,
      "train_tokens_per_second": 8351.039
    },
    {
      "epoch": 1.0567600323188797,
      "grad_norm": 0.726495580165145,
      "learning_rate": 4.5553248498587144e-06,
      "loss": 0.30329504013061526,
      "num_input_tokens_seen": 807562832,
      "step": 15695,
      "train_runtime": 96701.2356,
      "train_tokens_per_second": 8351.112
    },
    {
      "epoch": 1.0570966873148397,
      "grad_norm": 0.8006645385576787,
      "learning_rate": 4.552691307364853e-06,
      "loss": 0.3124422550201416,
      "num_input_tokens_seen": 807798152,
      "step": 15700,
      "train_runtime": 96732.7828,
      "train_tokens_per_second": 8350.821
    },
    {
      "epoch": 1.0574333423108,
      "grad_norm": 0.6777673782527114,
      "learning_rate": 4.5500578899594375e-06,
      "loss": 0.33155772686004636,
      "num_input_tokens_seen": 808055008,
      "step": 15705,
      "train_runtime": 96766.4361,
      "train_tokens_per_second": 8350.571
    },
    {
      "epoch": 1.05776999730676,
      "grad_norm": 0.6580600264629202,
      "learning_rate": 4.5474245983789e-06,
      "loss": 0.31063170433044435,
      "num_input_tokens_seen": 808310520,
      "step": 15710,
      "train_runtime": 96797.2808,
      "train_tokens_per_second": 8350.55
    },
    {
      "epoch": 1.0581066523027203,
      "grad_norm": 0.6204302733213267,
      "learning_rate": 4.54479143335963e-06,
      "loss": 0.2965527057647705,
      "num_input_tokens_seen": 808557520,
      "step": 15715,
      "train_runtime": 96826.7249,
      "train_tokens_per_second": 8350.561
    },
    {
      "epoch": 1.0584433072986803,
      "grad_norm": 0.6902064329648425,
      "learning_rate": 4.5421583956379855e-06,
      "loss": 0.3253329753875732,
      "num_input_tokens_seen": 808800784,
      "step": 15720,
      "train_runtime": 96857.2134,
      "train_tokens_per_second": 8350.444
    },
    {
      "epoch": 1.0587799622946406,
      "grad_norm": 0.6053749036752375,
      "learning_rate": 4.539525485950284e-06,
      "loss": 0.2887387752532959,
      "num_input_tokens_seen": 809071832,
      "step": 15725,
      "train_runtime": 96887.2331,
      "train_tokens_per_second": 8350.655
    },
    {
      "epoch": 1.0591166172906006,
      "grad_norm": 0.6904294178561866,
      "learning_rate": 4.536892705032815e-06,
      "loss": 0.32090206146240235,
      "num_input_tokens_seen": 809353488,
      "step": 15730,
      "train_runtime": 96922.3474,
      "train_tokens_per_second": 8350.535
    },
    {
      "epoch": 1.0594532722865608,
      "grad_norm": 0.6594756234053325,
      "learning_rate": 4.534260053621821e-06,
      "loss": 0.2997774839401245,
      "num_input_tokens_seen": 809604520,
      "step": 15735,
      "train_runtime": 96949.1623,
      "train_tokens_per_second": 8350.815
    },
    {
      "epoch": 1.0597899272825209,
      "grad_norm": 0.7391185645869478,
      "learning_rate": 4.531627532453519e-06,
      "loss": 0.2769421100616455,
      "num_input_tokens_seen": 809844112,
      "step": 15740,
      "train_runtime": 96977.9157,
      "train_tokens_per_second": 8350.81
    },
    {
      "epoch": 1.0601265822784811,
      "grad_norm": 0.607411254432396,
      "learning_rate": 4.5289951422640826e-06,
      "loss": 0.31239068508148193,
      "num_input_tokens_seen": 810076400,
      "step": 15745,
      "train_runtime": 97006.1349,
      "train_tokens_per_second": 8350.775
    },
    {
      "epoch": 1.0604632372744411,
      "grad_norm": 0.6551549986182215,
      "learning_rate": 4.526362883789655e-06,
      "loss": 0.2952704429626465,
      "num_input_tokens_seen": 810336984,
      "step": 15750,
      "train_runtime": 97034.6489,
      "train_tokens_per_second": 8351.007
    },
    {
      "epoch": 1.0607998922704014,
      "grad_norm": 0.9540504769298384,
      "learning_rate": 4.5237307577663345e-06,
      "loss": 0.3223764419555664,
      "num_input_tokens_seen": 810601072,
      "step": 15755,
      "train_runtime": 97068.4935,
      "train_tokens_per_second": 8350.815
    },
    {
      "epoch": 1.0611365472663614,
      "grad_norm": 0.5920943799195969,
      "learning_rate": 4.52109876493019e-06,
      "loss": 0.3064307451248169,
      "num_input_tokens_seen": 810852048,
      "step": 15760,
      "train_runtime": 97100.5164,
      "train_tokens_per_second": 8350.646
    },
    {
      "epoch": 1.0614732022623217,
      "grad_norm": 0.6332480884221939,
      "learning_rate": 4.518466906017249e-06,
      "loss": 0.3272072792053223,
      "num_input_tokens_seen": 811096992,
      "step": 15765,
      "train_runtime": 97126.9942,
      "train_tokens_per_second": 8350.892
    },
    {
      "epoch": 1.0618098572582817,
      "grad_norm": 0.6757567059177355,
      "learning_rate": 4.515835181763503e-06,
      "loss": 0.3022955894470215,
      "num_input_tokens_seen": 811360152,
      "step": 15770,
      "train_runtime": 97155.899,
      "train_tokens_per_second": 8351.116
    },
    {
      "epoch": 1.062146512254242,
      "grad_norm": 0.7297902213618249,
      "learning_rate": 4.513203592904903e-06,
      "loss": 0.30367996692657473,
      "num_input_tokens_seen": 811619776,
      "step": 15775,
      "train_runtime": 97183.5228,
      "train_tokens_per_second": 8351.413
    },
    {
      "epoch": 1.062483167250202,
      "grad_norm": 0.6072795221021702,
      "learning_rate": 4.510572140177368e-06,
      "loss": 0.31002278327941896,
      "num_input_tokens_seen": 811880232,
      "step": 15780,
      "train_runtime": 97212.6315,
      "train_tokens_per_second": 8351.592
    },
    {
      "epoch": 1.0628198222461622,
      "grad_norm": 0.6844325011011434,
      "learning_rate": 4.507940824316771e-06,
      "loss": 0.31415681838989257,
      "num_input_tokens_seen": 812127664,
      "step": 15785,
      "train_runtime": 97247.4072,
      "train_tokens_per_second": 8351.15
    },
    {
      "epoch": 1.0631564772421223,
      "grad_norm": 0.7402784372628899,
      "learning_rate": 4.505309646058954e-06,
      "loss": 0.30674521923065184,
      "num_input_tokens_seen": 812367672,
      "step": 15790,
      "train_runtime": 97274.414,
      "train_tokens_per_second": 8351.299
    },
    {
      "epoch": 1.0634931322380825,
      "grad_norm": 0.8329157665652496,
      "learning_rate": 4.502678606139716e-06,
      "loss": 0.29350240230560304,
      "num_input_tokens_seen": 812612024,
      "step": 15795,
      "train_runtime": 97307.764,
      "train_tokens_per_second": 8350.947
    },
    {
      "epoch": 1.0638297872340425,
      "grad_norm": 0.6194243993515839,
      "learning_rate": 4.500047705294823e-06,
      "loss": 0.28825223445892334,
      "num_input_tokens_seen": 812876888,
      "step": 15800,
      "train_runtime": 97341.8042,
      "train_tokens_per_second": 8350.748
    },
    {
      "epoch": 1.0641664422300028,
      "grad_norm": 0.6442677791036808,
      "learning_rate": 4.497416944259993e-06,
      "loss": 0.3108521461486816,
      "num_input_tokens_seen": 813142944,
      "step": 15805,
      "train_runtime": 97373.7855,
      "train_tokens_per_second": 8350.738
    },
    {
      "epoch": 1.0645030972259628,
      "grad_norm": 0.6882252374404035,
      "learning_rate": 4.494786323770912e-06,
      "loss": 0.3220226526260376,
      "num_input_tokens_seen": 813375120,
      "step": 15810,
      "train_runtime": 97402.7915,
      "train_tokens_per_second": 8350.635
    },
    {
      "epoch": 1.064839752221923,
      "grad_norm": 0.5642652695878719,
      "learning_rate": 4.492155844563224e-06,
      "loss": 0.306705379486084,
      "num_input_tokens_seen": 813649232,
      "step": 15815,
      "train_runtime": 97434.2391,
      "train_tokens_per_second": 8350.753
    },
    {
      "epoch": 1.065176407217883,
      "grad_norm": 0.6523321638182542,
      "learning_rate": 4.489525507372537e-06,
      "loss": 0.3220046043395996,
      "num_input_tokens_seen": 813916704,
      "step": 15820,
      "train_runtime": 97468.1544,
      "train_tokens_per_second": 8350.591
    },
    {
      "epoch": 1.0655130622138433,
      "grad_norm": 0.6863169276054384,
      "learning_rate": 4.486895312934411e-06,
      "loss": 0.305257773399353,
      "num_input_tokens_seen": 814164760,
      "step": 15825,
      "train_runtime": 97494.9628,
      "train_tokens_per_second": 8350.839
    },
    {
      "epoch": 1.0658497172098034,
      "grad_norm": 0.6220388273732657,
      "learning_rate": 4.484265261984378e-06,
      "loss": 0.3422244548797607,
      "num_input_tokens_seen": 814419896,
      "step": 15830,
      "train_runtime": 97520.2152,
      "train_tokens_per_second": 8351.293
    },
    {
      "epoch": 1.0661863722057636,
      "grad_norm": 0.5977503769637414,
      "learning_rate": 4.4816353552579165e-06,
      "loss": 0.29603071212768556,
      "num_input_tokens_seen": 814671296,
      "step": 15835,
      "train_runtime": 97550.6161,
      "train_tokens_per_second": 8351.268
    },
    {
      "epoch": 1.0665230272017237,
      "grad_norm": 0.7429184089514365,
      "learning_rate": 4.479005593490477e-06,
      "loss": 0.31331372261047363,
      "num_input_tokens_seen": 814920952,
      "step": 15840,
      "train_runtime": 97579.4132,
      "train_tokens_per_second": 8351.361
    },
    {
      "epoch": 1.066859682197684,
      "grad_norm": 0.6298450076245277,
      "learning_rate": 4.476375977417461e-06,
      "loss": 0.27030391693115235,
      "num_input_tokens_seen": 815184144,
      "step": 15845,
      "train_runtime": 97608.6026,
      "train_tokens_per_second": 8351.56
    },
    {
      "epoch": 1.067196337193644,
      "grad_norm": 0.822405276795839,
      "learning_rate": 4.473746507774235e-06,
      "loss": 0.2967180013656616,
      "num_input_tokens_seen": 815463000,
      "step": 15850,
      "train_runtime": 97636.3535,
      "train_tokens_per_second": 8352.043
    },
    {
      "epoch": 1.0675329921896042,
      "grad_norm": 0.5841219737760913,
      "learning_rate": 4.471117185296117e-06,
      "loss": 0.2960835933685303,
      "num_input_tokens_seen": 815717272,
      "step": 15855,
      "train_runtime": 97665.4472,
      "train_tokens_per_second": 8352.158
    },
    {
      "epoch": 1.0678696471855642,
      "grad_norm": 0.6247367427473405,
      "learning_rate": 4.468488010718394e-06,
      "loss": 0.2904985427856445,
      "num_input_tokens_seen": 815952648,
      "step": 15860,
      "train_runtime": 97696.4701,
      "train_tokens_per_second": 8351.915
    },
    {
      "epoch": 1.0682063021815245,
      "grad_norm": 0.6824053561293733,
      "learning_rate": 4.465858984776302e-06,
      "loss": 0.31074156761169436,
      "num_input_tokens_seen": 816206312,
      "step": 15865,
      "train_runtime": 97728.5358,
      "train_tokens_per_second": 8351.771
    },
    {
      "epoch": 1.0685429571774845,
      "grad_norm": 0.7722560873605057,
      "learning_rate": 4.463230108205044e-06,
      "loss": 0.30661611557006835,
      "num_input_tokens_seen": 816426208,
      "step": 15870,
      "train_runtime": 97756.9717,
      "train_tokens_per_second": 8351.591
    },
    {
      "epoch": 1.0688796121734447,
      "grad_norm": 0.6305922940268565,
      "learning_rate": 4.460601381739772e-06,
      "loss": 0.3136931896209717,
      "num_input_tokens_seen": 816694624,
      "step": 15875,
      "train_runtime": 97784.4536,
      "train_tokens_per_second": 8351.988
    },
    {
      "epoch": 1.0692162671694048,
      "grad_norm": 0.7565672527601736,
      "learning_rate": 4.457972806115607e-06,
      "loss": 0.3402997970581055,
      "num_input_tokens_seen": 816941776,
      "step": 15880,
      "train_runtime": 97813.5699,
      "train_tokens_per_second": 8352.029
    },
    {
      "epoch": 1.069552922165365,
      "grad_norm": 0.5809738133542407,
      "learning_rate": 4.455344382067616e-06,
      "loss": 0.3103238582611084,
      "num_input_tokens_seen": 817206152,
      "step": 15885,
      "train_runtime": 97844.9195,
      "train_tokens_per_second": 8352.055
    },
    {
      "epoch": 1.069889577161325,
      "grad_norm": 0.5782918551450515,
      "learning_rate": 4.452716110330832e-06,
      "loss": 0.3041639804840088,
      "num_input_tokens_seen": 817468704,
      "step": 15890,
      "train_runtime": 97873.8197,
      "train_tokens_per_second": 8352.271
    },
    {
      "epoch": 1.0702262321572853,
      "grad_norm": 0.7245373371115373,
      "learning_rate": 4.450087991640242e-06,
      "loss": 0.3205798864364624,
      "num_input_tokens_seen": 817709800,
      "step": 15895,
      "train_runtime": 97901.6796,
      "train_tokens_per_second": 8352.357
    },
    {
      "epoch": 1.0705628871532453,
      "grad_norm": 0.6437165452006176,
      "learning_rate": 4.447460026730794e-06,
      "loss": 0.3166641712188721,
      "num_input_tokens_seen": 817950784,
      "step": 15900,
      "train_runtime": 97930.836,
      "train_tokens_per_second": 8352.331
    },
    {
      "epoch": 1.0708995421492056,
      "grad_norm": 0.6317724533169219,
      "learning_rate": 4.444832216337384e-06,
      "loss": 0.29520449638366697,
      "num_input_tokens_seen": 818212544,
      "step": 15905,
      "train_runtime": 97958.6296,
      "train_tokens_per_second": 8352.634
    },
    {
      "epoch": 1.0712361971451656,
      "grad_norm": 0.8211433033816183,
      "learning_rate": 4.442204561194877e-06,
      "loss": 0.30286417007446287,
      "num_input_tokens_seen": 818473952,
      "step": 15910,
      "train_runtime": 97987.6722,
      "train_tokens_per_second": 8352.826
    },
    {
      "epoch": 1.0715728521411259,
      "grad_norm": 0.6185412990585721,
      "learning_rate": 4.439577062038085e-06,
      "loss": 0.2956798553466797,
      "num_input_tokens_seen": 818741568,
      "step": 15915,
      "train_runtime": 98022.4465,
      "train_tokens_per_second": 8352.593
    },
    {
      "epoch": 1.0719095071370859,
      "grad_norm": 0.6493471003627679,
      "learning_rate": 4.436949719601782e-06,
      "loss": 0.33915271759033205,
      "num_input_tokens_seen": 819008768,
      "step": 15920,
      "train_runtime": 98050.5413,
      "train_tokens_per_second": 8352.924
    },
    {
      "epoch": 1.0722461621330461,
      "grad_norm": 0.8142018414036551,
      "learning_rate": 4.434322534620692e-06,
      "loss": 0.3197824478149414,
      "num_input_tokens_seen": 819252936,
      "step": 15925,
      "train_runtime": 98076.0623,
      "train_tokens_per_second": 8353.241
    },
    {
      "epoch": 1.0725828171290062,
      "grad_norm": 0.6251361915066007,
      "learning_rate": 4.431695507829504e-06,
      "loss": 0.28504817485809325,
      "num_input_tokens_seen": 819516104,
      "step": 15930,
      "train_runtime": 98106.7714,
      "train_tokens_per_second": 8353.308
    },
    {
      "epoch": 1.0729194721249664,
      "grad_norm": 0.6991682369550378,
      "learning_rate": 4.429068639962853e-06,
      "loss": 0.31056933403015136,
      "num_input_tokens_seen": 819777984,
      "step": 15935,
      "train_runtime": 98135.9889,
      "train_tokens_per_second": 8353.49
    },
    {
      "epoch": 1.0732561271209264,
      "grad_norm": 0.702241420442548,
      "learning_rate": 4.426441931755337e-06,
      "loss": 0.3028721809387207,
      "num_input_tokens_seen": 820042392,
      "step": 15940,
      "train_runtime": 98168.046,
      "train_tokens_per_second": 8353.455
    },
    {
      "epoch": 1.0735927821168867,
      "grad_norm": 0.5713690378712825,
      "learning_rate": 4.4238153839415045e-06,
      "loss": 0.32488455772399905,
      "num_input_tokens_seen": 820321856,
      "step": 15945,
      "train_runtime": 98198.7572,
      "train_tokens_per_second": 8353.689
    },
    {
      "epoch": 1.0739294371128467,
      "grad_norm": 0.7342310571631137,
      "learning_rate": 4.421188997255865e-06,
      "loss": 0.33701553344726565,
      "num_input_tokens_seen": 820578784,
      "step": 15950,
      "train_runtime": 98226.2656,
      "train_tokens_per_second": 8353.965
    },
    {
      "epoch": 1.074266092108807,
      "grad_norm": 0.7040041225222362,
      "learning_rate": 4.418562772432874e-06,
      "loss": 0.3020047187805176,
      "num_input_tokens_seen": 820841416,
      "step": 15955,
      "train_runtime": 98255.0864,
      "train_tokens_per_second": 8354.188
    },
    {
      "epoch": 1.074602747104767,
      "grad_norm": 0.6473711784548206,
      "learning_rate": 4.415936710206952e-06,
      "loss": 0.31671957969665526,
      "num_input_tokens_seen": 821133328,
      "step": 15960,
      "train_runtime": 98289.801,
      "train_tokens_per_second": 8354.207
    },
    {
      "epoch": 1.0749394021007272,
      "grad_norm": 0.6843065317613914,
      "learning_rate": 4.413310811312465e-06,
      "loss": 0.30799479484558107,
      "num_input_tokens_seen": 821387944,
      "step": 15965,
      "train_runtime": 98322.7513,
      "train_tokens_per_second": 8353.997
    },
    {
      "epoch": 1.0752760570966873,
      "grad_norm": 0.7018572590507703,
      "learning_rate": 4.41068507648374e-06,
      "loss": 0.28204917907714844,
      "num_input_tokens_seen": 821650920,
      "step": 15970,
      "train_runtime": 98353.3749,
      "train_tokens_per_second": 8354.069
    },
    {
      "epoch": 1.0756127120926475,
      "grad_norm": 0.5769386720063828,
      "learning_rate": 4.408059506455053e-06,
      "loss": 0.29321515560150146,
      "num_input_tokens_seen": 821946104,
      "step": 15975,
      "train_runtime": 98381.4918,
      "train_tokens_per_second": 8354.682
    },
    {
      "epoch": 1.0759493670886076,
      "grad_norm": 0.6068614643069487,
      "learning_rate": 4.40543410196064e-06,
      "loss": 0.3099503517150879,
      "num_input_tokens_seen": 822207536,
      "step": 15980,
      "train_runtime": 98412.195,
      "train_tokens_per_second": 8354.732
    },
    {
      "epoch": 1.0762860220845678,
      "grad_norm": 0.8020392795451264,
      "learning_rate": 4.402808863734681e-06,
      "loss": 0.318156361579895,
      "num_input_tokens_seen": 822454576,
      "step": 15985,
      "train_runtime": 98444.0984,
      "train_tokens_per_second": 8354.534
    },
    {
      "epoch": 1.0766226770805278,
      "grad_norm": 0.6148821398978891,
      "learning_rate": 4.40018379251132e-06,
      "loss": 0.31072587966918946,
      "num_input_tokens_seen": 822700568,
      "step": 15990,
      "train_runtime": 98474.132,
      "train_tokens_per_second": 8354.484
    },
    {
      "epoch": 1.076959332076488,
      "grad_norm": 0.787725813801612,
      "learning_rate": 4.397558889024649e-06,
      "loss": 0.3489112615585327,
      "num_input_tokens_seen": 822927376,
      "step": 15995,
      "train_runtime": 98502.7104,
      "train_tokens_per_second": 8354.363
    },
    {
      "epoch": 1.0772959870724481,
      "grad_norm": 0.5347062568932408,
      "learning_rate": 4.394934154008715e-06,
      "loss": 0.309527587890625,
      "num_input_tokens_seen": 823191152,
      "step": 16000,
      "train_runtime": 98532.5679,
      "train_tokens_per_second": 8354.508
    },
    {
      "epoch": 1.0776326420684084,
      "grad_norm": 0.7104648242100337,
      "learning_rate": 4.392309588197514e-06,
      "loss": 0.3081902742385864,
      "num_input_tokens_seen": 823463912,
      "step": 16005,
      "train_runtime": 98561.701,
      "train_tokens_per_second": 8354.806
    },
    {
      "epoch": 1.0779692970643684,
      "grad_norm": 0.6179395531035086,
      "learning_rate": 4.389685192324999e-06,
      "loss": 0.2975563287734985,
      "num_input_tokens_seen": 823716536,
      "step": 16010,
      "train_runtime": 98592.3196,
      "train_tokens_per_second": 8354.774
    },
    {
      "epoch": 1.0783059520603286,
      "grad_norm": 0.6423212390958619,
      "learning_rate": 4.387060967125073e-06,
      "loss": 0.3238647937774658,
      "num_input_tokens_seen": 823959952,
      "step": 16015,
      "train_runtime": 98621.8072,
      "train_tokens_per_second": 8354.744
    },
    {
      "epoch": 1.0786426070562887,
      "grad_norm": 0.7630131879112164,
      "learning_rate": 4.384436913331596e-06,
      "loss": 0.3374919891357422,
      "num_input_tokens_seen": 824220376,
      "step": 16020,
      "train_runtime": 98648.7756,
      "train_tokens_per_second": 8355.1
    },
    {
      "epoch": 1.078979262052249,
      "grad_norm": 0.6179730599153226,
      "learning_rate": 4.38181303167837e-06,
      "loss": 0.2995894432067871,
      "num_input_tokens_seen": 824472336,
      "step": 16025,
      "train_runtime": 98679.9335,
      "train_tokens_per_second": 8355.015
    },
    {
      "epoch": 1.079315917048209,
      "grad_norm": 0.6337326404809014,
      "learning_rate": 4.379189322899159e-06,
      "loss": 0.3033179759979248,
      "num_input_tokens_seen": 824712728,
      "step": 16030,
      "train_runtime": 98708.6614,
      "train_tokens_per_second": 8355.019
    },
    {
      "epoch": 1.0796525720441692,
      "grad_norm": 1.0694158776527598,
      "learning_rate": 4.376565787727676e-06,
      "loss": 0.30269851684570315,
      "num_input_tokens_seen": 824984688,
      "step": 16035,
      "train_runtime": 98738.9722,
      "train_tokens_per_second": 8355.208
    },
    {
      "epoch": 1.0799892270401292,
      "grad_norm": 0.6507166916053683,
      "learning_rate": 4.373942426897581e-06,
      "loss": 0.3006168842315674,
      "num_input_tokens_seen": 825214728,
      "step": 16040,
      "train_runtime": 98767.5291,
      "train_tokens_per_second": 8355.122
    },
    {
      "epoch": 1.0803258820360895,
      "grad_norm": 0.7551861244513868,
      "learning_rate": 4.37131924114249e-06,
      "loss": 0.32126874923706056,
      "num_input_tokens_seen": 825441536,
      "step": 16045,
      "train_runtime": 98794.9673,
      "train_tokens_per_second": 8355.097
    },
    {
      "epoch": 1.0806625370320495,
      "grad_norm": 0.6771006518408688,
      "learning_rate": 4.368696231195968e-06,
      "loss": 0.3098970651626587,
      "num_input_tokens_seen": 825703528,
      "step": 16050,
      "train_runtime": 98826.1717,
      "train_tokens_per_second": 8355.11
    },
    {
      "epoch": 1.0809991920280098,
      "grad_norm": 0.752467399538668,
      "learning_rate": 4.366073397791534e-06,
      "loss": 0.2963526964187622,
      "num_input_tokens_seen": 825951648,
      "step": 16055,
      "train_runtime": 98855.2579,
      "train_tokens_per_second": 8355.162
    },
    {
      "epoch": 1.0813358470239698,
      "grad_norm": 0.6233817656005872,
      "learning_rate": 4.363450741662651e-06,
      "loss": 0.2817226886749268,
      "num_input_tokens_seen": 826206984,
      "step": 16060,
      "train_runtime": 98884.8713,
      "train_tokens_per_second": 8355.242
    },
    {
      "epoch": 1.08167250201993,
      "grad_norm": 0.6548068746479812,
      "learning_rate": 4.360828263542739e-06,
      "loss": 0.28666985034942627,
      "num_input_tokens_seen": 826476528,
      "step": 16065,
      "train_runtime": 98915.7089,
      "train_tokens_per_second": 8355.362
    },
    {
      "epoch": 1.08200915701589,
      "grad_norm": 0.7125352154245406,
      "learning_rate": 4.358205964165164e-06,
      "loss": 0.31947157382965086,
      "num_input_tokens_seen": 826734312,
      "step": 16070,
      "train_runtime": 98945.5882,
      "train_tokens_per_second": 8355.444
    },
    {
      "epoch": 1.0823458120118503,
      "grad_norm": 0.7211445520303865,
      "learning_rate": 4.355583844263247e-06,
      "loss": 0.32896206378936765,
      "num_input_tokens_seen": 826993512,
      "step": 16075,
      "train_runtime": 98976.3381,
      "train_tokens_per_second": 8355.467
    },
    {
      "epoch": 1.0826824670078103,
      "grad_norm": 0.8187355062010029,
      "learning_rate": 4.352961904570251e-06,
      "loss": 0.31351509094238283,
      "num_input_tokens_seen": 827249520,
      "step": 16080,
      "train_runtime": 99008.1473,
      "train_tokens_per_second": 8355.368
    },
    {
      "epoch": 1.0830191220037706,
      "grad_norm": 0.6879110974936716,
      "learning_rate": 4.350340145819397e-06,
      "loss": 0.3158433437347412,
      "num_input_tokens_seen": 827500624,
      "step": 16085,
      "train_runtime": 99039.3649,
      "train_tokens_per_second": 8355.27
    },
    {
      "epoch": 1.0833557769997306,
      "grad_norm": 0.659944779578145,
      "learning_rate": 4.347718568743847e-06,
      "loss": 0.3170581340789795,
      "num_input_tokens_seen": 827754200,
      "step": 16090,
      "train_runtime": 99072.3642,
      "train_tokens_per_second": 8355.046
    },
    {
      "epoch": 1.0836924319956909,
      "grad_norm": 0.6337664950267237,
      "learning_rate": 4.34509717407672e-06,
      "loss": 0.3210428237915039,
      "num_input_tokens_seen": 828030808,
      "step": 16095,
      "train_runtime": 99103.8224,
      "train_tokens_per_second": 8355.185
    },
    {
      "epoch": 1.084029086991651,
      "grad_norm": 0.762008048331662,
      "learning_rate": 4.342475962551081e-06,
      "loss": 0.3140867710113525,
      "num_input_tokens_seen": 828294944,
      "step": 16100,
      "train_runtime": 99134.6484,
      "train_tokens_per_second": 8355.252
    },
    {
      "epoch": 1.0843657419876112,
      "grad_norm": 0.6576531270708682,
      "learning_rate": 4.339854934899943e-06,
      "loss": 0.294956636428833,
      "num_input_tokens_seen": 828540312,
      "step": 16105,
      "train_runtime": 99165.4375,
      "train_tokens_per_second": 8355.132
    },
    {
      "epoch": 1.0847023969835712,
      "grad_norm": 0.5319474549895239,
      "learning_rate": 4.337234091856265e-06,
      "loss": 0.27748429775238037,
      "num_input_tokens_seen": 828812624,
      "step": 16110,
      "train_runtime": 99195.894,
      "train_tokens_per_second": 8355.312
    },
    {
      "epoch": 1.0850390519795314,
      "grad_norm": 0.7042615033850165,
      "learning_rate": 4.334613434152961e-06,
      "loss": 0.3048213481903076,
      "num_input_tokens_seen": 829050848,
      "step": 16115,
      "train_runtime": 99227.9981,
      "train_tokens_per_second": 8355.009
    },
    {
      "epoch": 1.0853757069754915,
      "grad_norm": 0.7183731275690736,
      "learning_rate": 4.331992962522888e-06,
      "loss": 0.3379840612411499,
      "num_input_tokens_seen": 829318536,
      "step": 16120,
      "train_runtime": 99257.6464,
      "train_tokens_per_second": 8355.211
    },
    {
      "epoch": 1.0857123619714517,
      "grad_norm": 0.6699782737338129,
      "learning_rate": 4.329372677698854e-06,
      "loss": 0.3062744140625,
      "num_input_tokens_seen": 829598488,
      "step": 16125,
      "train_runtime": 99285.6008,
      "train_tokens_per_second": 8355.678
    },
    {
      "epoch": 1.0860490169674117,
      "grad_norm": 0.6330541669084931,
      "learning_rate": 4.3267525804136114e-06,
      "loss": 0.3056508541107178,
      "num_input_tokens_seen": 829862296,
      "step": 16130,
      "train_runtime": 99320.5054,
      "train_tokens_per_second": 8355.397
    },
    {
      "epoch": 1.086385671963372,
      "grad_norm": 0.7453286499577733,
      "learning_rate": 4.324132671399864e-06,
      "loss": 0.315975284576416,
      "num_input_tokens_seen": 830088376,
      "step": 16135,
      "train_runtime": 99349.5256,
      "train_tokens_per_second": 8355.232
    },
    {
      "epoch": 1.086722326959332,
      "grad_norm": 0.650126537118071,
      "learning_rate": 4.321512951390258e-06,
      "loss": 0.32438490390777586,
      "num_input_tokens_seen": 830351128,
      "step": 16140,
      "train_runtime": 99377.5771,
      "train_tokens_per_second": 8355.518
    },
    {
      "epoch": 1.0870589819552923,
      "grad_norm": 0.77085612513116,
      "learning_rate": 4.318893421117393e-06,
      "loss": 0.3009478092193604,
      "num_input_tokens_seen": 830617384,
      "step": 16145,
      "train_runtime": 99406.7055,
      "train_tokens_per_second": 8355.748
    },
    {
      "epoch": 1.0873956369512523,
      "grad_norm": 0.7129286979703636,
      "learning_rate": 4.31627408131381e-06,
      "loss": 0.2962451934814453,
      "num_input_tokens_seen": 830843352,
      "step": 16150,
      "train_runtime": 99436.4451,
      "train_tokens_per_second": 8355.521
    },
    {
      "epoch": 1.0877322919472125,
      "grad_norm": 0.6233857845264886,
      "learning_rate": 4.313654932712001e-06,
      "loss": 0.28501286506652834,
      "num_input_tokens_seen": 831100672,
      "step": 16155,
      "train_runtime": 99467.1708,
      "train_tokens_per_second": 8355.527
    },
    {
      "epoch": 1.0880689469431726,
      "grad_norm": 0.672506002549499,
      "learning_rate": 4.3110359760444e-06,
      "loss": 0.3002964735031128,
      "num_input_tokens_seen": 831380704,
      "step": 16160,
      "train_runtime": 99501.831,
      "train_tokens_per_second": 8355.431
    },
    {
      "epoch": 1.0884056019391328,
      "grad_norm": 0.6580075391687634,
      "learning_rate": 4.308417212043392e-06,
      "loss": 0.28516302108764646,
      "num_input_tokens_seen": 831624784,
      "step": 16165,
      "train_runtime": 99534.0203,
      "train_tokens_per_second": 8355.181
    },
    {
      "epoch": 1.0887422569350929,
      "grad_norm": 0.6598779751014826,
      "learning_rate": 4.305798641441304e-06,
      "loss": 0.29999685287475586,
      "num_input_tokens_seen": 831889464,
      "step": 16170,
      "train_runtime": 99564.8803,
      "train_tokens_per_second": 8355.25
    },
    {
      "epoch": 1.089078911931053,
      "grad_norm": 0.6060413828942235,
      "learning_rate": 4.303180264970416e-06,
      "loss": 0.2979151725769043,
      "num_input_tokens_seen": 832134976,
      "step": 16175,
      "train_runtime": 99598.065,
      "train_tokens_per_second": 8354.931
    },
    {
      "epoch": 1.0894155669270131,
      "grad_norm": 0.6350326800227432,
      "learning_rate": 4.3005620833629415e-06,
      "loss": 0.27677078247070314,
      "num_input_tokens_seen": 832379328,
      "step": 16180,
      "train_runtime": 99626.0767,
      "train_tokens_per_second": 8355.035
    },
    {
      "epoch": 1.0897522219229734,
      "grad_norm": 0.7435592644638447,
      "learning_rate": 4.297944097351053e-06,
      "loss": 0.3168666362762451,
      "num_input_tokens_seen": 832606192,
      "step": 16185,
      "train_runtime": 99654.848,
      "train_tokens_per_second": 8354.899
    },
    {
      "epoch": 1.0900888769189334,
      "grad_norm": 0.6505648259019212,
      "learning_rate": 4.295326307666856e-06,
      "loss": 0.31599721908569334,
      "num_input_tokens_seen": 832881136,
      "step": 16190,
      "train_runtime": 99688.8642,
      "train_tokens_per_second": 8354.806
    },
    {
      "epoch": 1.0904255319148937,
      "grad_norm": 0.8286863493832625,
      "learning_rate": 4.292708715042411e-06,
      "loss": 0.3045308589935303,
      "num_input_tokens_seen": 833137144,
      "step": 16195,
      "train_runtime": 99717.7372,
      "train_tokens_per_second": 8354.954
    },
    {
      "epoch": 1.0907621869108537,
      "grad_norm": 0.7069465733210872,
      "learning_rate": 4.290091320209718e-06,
      "loss": 0.3311828851699829,
      "num_input_tokens_seen": 833413320,
      "step": 16200,
      "train_runtime": 99748.3468,
      "train_tokens_per_second": 8355.159
    },
    {
      "epoch": 1.091098841906814,
      "grad_norm": 0.6568966713232546,
      "learning_rate": 4.287474123900726e-06,
      "loss": 0.3059386253356934,
      "num_input_tokens_seen": 833647480,
      "step": 16205,
      "train_runtime": 99775.0984,
      "train_tokens_per_second": 8355.266
    },
    {
      "epoch": 1.091435496902774,
      "grad_norm": 0.7064265144783768,
      "learning_rate": 4.284857126847321e-06,
      "loss": 0.308731746673584,
      "num_input_tokens_seen": 833900512,
      "step": 16210,
      "train_runtime": 99804.5666,
      "train_tokens_per_second": 8355.334
    },
    {
      "epoch": 1.0917721518987342,
      "grad_norm": 0.6212354191409861,
      "learning_rate": 4.28224032978134e-06,
      "loss": 0.31665606498718263,
      "num_input_tokens_seen": 834175128,
      "step": 16215,
      "train_runtime": 99838.3178,
      "train_tokens_per_second": 8355.26
    },
    {
      "epoch": 1.0921088068946943,
      "grad_norm": 0.5150327318250116,
      "learning_rate": 4.279623733434561e-06,
      "loss": 0.29777135848999026,
      "num_input_tokens_seen": 834434152,
      "step": 16220,
      "train_runtime": 99866.5916,
      "train_tokens_per_second": 8355.488
    },
    {
      "epoch": 1.0924454618906545,
      "grad_norm": 0.6219736264731023,
      "learning_rate": 4.277007338538711e-06,
      "loss": 0.27379517555236815,
      "num_input_tokens_seen": 834717152,
      "step": 16225,
      "train_runtime": 99899.1819,
      "train_tokens_per_second": 8355.595
    },
    {
      "epoch": 1.0927821168866145,
      "grad_norm": 0.5098697199187242,
      "learning_rate": 4.27439114582545e-06,
      "loss": 0.27320327758789065,
      "num_input_tokens_seen": 834965080,
      "step": 16230,
      "train_runtime": 99929.6822,
      "train_tokens_per_second": 8355.526
    },
    {
      "epoch": 1.0931187718825748,
      "grad_norm": 0.6970196880808259,
      "learning_rate": 4.271775156026395e-06,
      "loss": 0.3258587121963501,
      "num_input_tokens_seen": 835186128,
      "step": 16235,
      "train_runtime": 99956.9163,
      "train_tokens_per_second": 8355.461
    },
    {
      "epoch": 1.0934554268785348,
      "grad_norm": 0.623934002693791,
      "learning_rate": 4.269159369873093e-06,
      "loss": 0.3008262634277344,
      "num_input_tokens_seen": 835426864,
      "step": 16240,
      "train_runtime": 99989.3113,
      "train_tokens_per_second": 8355.162
    },
    {
      "epoch": 1.093792081874495,
      "grad_norm": 0.6127854938000932,
      "learning_rate": 4.266543788097043e-06,
      "loss": 0.29367814064025877,
      "num_input_tokens_seen": 835701040,
      "step": 16245,
      "train_runtime": 100021.168,
      "train_tokens_per_second": 8355.242
    },
    {
      "epoch": 1.094128736870455,
      "grad_norm": 0.7312990980119649,
      "learning_rate": 4.263928411429683e-06,
      "loss": 0.3308006763458252,
      "num_input_tokens_seen": 835951800,
      "step": 16250,
      "train_runtime": 100051.696,
      "train_tokens_per_second": 8355.199
    },
    {
      "epoch": 1.0944653918664153,
      "grad_norm": 0.6121910727011586,
      "learning_rate": 4.261313240602398e-06,
      "loss": 0.31551346778869627,
      "num_input_tokens_seen": 836204184,
      "step": 16255,
      "train_runtime": 100084.1472,
      "train_tokens_per_second": 8355.011
    },
    {
      "epoch": 1.0948020468623754,
      "grad_norm": 0.6262799403063705,
      "learning_rate": 4.2586982763465064e-06,
      "loss": 0.3067952871322632,
      "num_input_tokens_seen": 836481392,
      "step": 16260,
      "train_runtime": 100114.5907,
      "train_tokens_per_second": 8355.24
    },
    {
      "epoch": 1.0951387018583356,
      "grad_norm": 0.6211312855295074,
      "learning_rate": 4.256083519393281e-06,
      "loss": 0.3015486001968384,
      "num_input_tokens_seen": 836740568,
      "step": 16265,
      "train_runtime": 100143.0303,
      "train_tokens_per_second": 8355.455
    },
    {
      "epoch": 1.0954753568542956,
      "grad_norm": 0.6012898210368819,
      "learning_rate": 4.253468970473925e-06,
      "loss": 0.3237235307693481,
      "num_input_tokens_seen": 837018288,
      "step": 16270,
      "train_runtime": 100174.1369,
      "train_tokens_per_second": 8355.633
    },
    {
      "epoch": 1.095812011850256,
      "grad_norm": 0.618442328475905,
      "learning_rate": 4.250854630319593e-06,
      "loss": 0.3361919403076172,
      "num_input_tokens_seen": 837278120,
      "step": 16275,
      "train_runtime": 100206.2196,
      "train_tokens_per_second": 8355.55
    },
    {
      "epoch": 1.096148666846216,
      "grad_norm": 0.6395248194220398,
      "learning_rate": 4.248240499661373e-06,
      "loss": 0.31615161895751953,
      "num_input_tokens_seen": 837521376,
      "step": 16280,
      "train_runtime": 100233.7743,
      "train_tokens_per_second": 8355.68
    },
    {
      "epoch": 1.0964853218421762,
      "grad_norm": 0.6599387180862631,
      "learning_rate": 4.245626579230302e-06,
      "loss": 0.293076229095459,
      "num_input_tokens_seen": 837789944,
      "step": 16285,
      "train_runtime": 100266.911,
      "train_tokens_per_second": 8355.597
    },
    {
      "epoch": 1.0968219768381362,
      "grad_norm": 0.7152045821298734,
      "learning_rate": 4.243012869757351e-06,
      "loss": 0.3020111083984375,
      "num_input_tokens_seen": 838071792,
      "step": 16290,
      "train_runtime": 100300.5372,
      "train_tokens_per_second": 8355.606
    },
    {
      "epoch": 1.0971586318340965,
      "grad_norm": 0.6507997820639266,
      "learning_rate": 4.240399371973438e-06,
      "loss": 0.30306663513183596,
      "num_input_tokens_seen": 838337896,
      "step": 16295,
      "train_runtime": 100333.2931,
      "train_tokens_per_second": 8355.531
    },
    {
      "epoch": 1.0974952868300565,
      "grad_norm": 0.6673865571914459,
      "learning_rate": 4.2377860866094165e-06,
      "loss": 0.3142717361450195,
      "num_input_tokens_seen": 838606608,
      "step": 16300,
      "train_runtime": 100363.1436,
      "train_tokens_per_second": 8355.723
    },
    {
      "epoch": 1.0978319418260167,
      "grad_norm": 0.7023912193813008,
      "learning_rate": 4.235173014396088e-06,
      "loss": 0.31052603721618655,
      "num_input_tokens_seen": 838869224,
      "step": 16305,
      "train_runtime": 100392.2122,
      "train_tokens_per_second": 8355.919
    },
    {
      "epoch": 1.0981685968219768,
      "grad_norm": 0.6157763404913839,
      "learning_rate": 4.2325601560641854e-06,
      "loss": 0.3173912525177002,
      "num_input_tokens_seen": 839141920,
      "step": 16310,
      "train_runtime": 100426.0856,
      "train_tokens_per_second": 8355.816
    },
    {
      "epoch": 1.098505251817937,
      "grad_norm": 0.6620756314632117,
      "learning_rate": 4.229947512344389e-06,
      "loss": 0.3076141357421875,
      "num_input_tokens_seen": 839394160,
      "step": 16315,
      "train_runtime": 100453.3881,
      "train_tokens_per_second": 8356.056
    },
    {
      "epoch": 1.098841906813897,
      "grad_norm": 0.6445728869402848,
      "learning_rate": 4.227335083967314e-06,
      "loss": 0.2993314743041992,
      "num_input_tokens_seen": 839652640,
      "step": 16320,
      "train_runtime": 100485.4606,
      "train_tokens_per_second": 8355.962
    },
    {
      "epoch": 1.0991785618098573,
      "grad_norm": 0.5640863434472387,
      "learning_rate": 4.224722871663522e-06,
      "loss": 0.30032930374145506,
      "num_input_tokens_seen": 839921584,
      "step": 16325,
      "train_runtime": 100519.5474,
      "train_tokens_per_second": 8355.803
    },
    {
      "epoch": 1.0995152168058173,
      "grad_norm": 0.6795870547362438,
      "learning_rate": 4.222110876163503e-06,
      "loss": 0.3285876750946045,
      "num_input_tokens_seen": 840185048,
      "step": 16330,
      "train_runtime": 100552.5362,
      "train_tokens_per_second": 8355.682
    },
    {
      "epoch": 1.0998518718017776,
      "grad_norm": 0.6556088541344682,
      "learning_rate": 4.2194990981977e-06,
      "loss": 0.2812621593475342,
      "num_input_tokens_seen": 840468208,
      "step": 16335,
      "train_runtime": 100583.3,
      "train_tokens_per_second": 8355.942
    },
    {
      "epoch": 1.1001885267977376,
      "grad_norm": 0.6294321230271787,
      "learning_rate": 4.216887538496482e-06,
      "loss": 0.2955659866333008,
      "num_input_tokens_seen": 840719600,
      "step": 16340,
      "train_runtime": 100611.7538,
      "train_tokens_per_second": 8356.077
    },
    {
      "epoch": 1.1005251817936978,
      "grad_norm": 0.6503555986634307,
      "learning_rate": 4.2142761977901695e-06,
      "loss": 0.35608773231506347,
      "num_input_tokens_seen": 840970496,
      "step": 16345,
      "train_runtime": 100644.8588,
      "train_tokens_per_second": 8355.822
    },
    {
      "epoch": 1.1008618367896579,
      "grad_norm": 0.6242222688268579,
      "learning_rate": 4.21166507680901e-06,
      "loss": 0.3124131917953491,
      "num_input_tokens_seen": 841223088,
      "step": 16350,
      "train_runtime": 100676.5031,
      "train_tokens_per_second": 8355.704
    },
    {
      "epoch": 1.1011984917856181,
      "grad_norm": 1.2687306911267917,
      "learning_rate": 4.209054176283202e-06,
      "loss": 0.3060481071472168,
      "num_input_tokens_seen": 841466376,
      "step": 16355,
      "train_runtime": 100705.2119,
      "train_tokens_per_second": 8355.738
    },
    {
      "epoch": 1.1015351467815782,
      "grad_norm": 0.6332302976617876,
      "learning_rate": 4.206443496942868e-06,
      "loss": 0.3210864305496216,
      "num_input_tokens_seen": 841719192,
      "step": 16360,
      "train_runtime": 100733.9969,
      "train_tokens_per_second": 8355.86
    },
    {
      "epoch": 1.1018718017775384,
      "grad_norm": 0.7303905410480417,
      "learning_rate": 4.203833039518081e-06,
      "loss": 0.2985135793685913,
      "num_input_tokens_seen": 841964696,
      "step": 16365,
      "train_runtime": 100765.7406,
      "train_tokens_per_second": 8355.664
    },
    {
      "epoch": 1.1022084567734984,
      "grad_norm": 0.6318332442374439,
      "learning_rate": 4.201222804738845e-06,
      "loss": 0.3015625,
      "num_input_tokens_seen": 842250960,
      "step": 16370,
      "train_runtime": 100795.8519,
      "train_tokens_per_second": 8356.008
    },
    {
      "epoch": 1.1025451117694587,
      "grad_norm": 0.6683067816978097,
      "learning_rate": 4.198612793335107e-06,
      "loss": 0.2900685787200928,
      "num_input_tokens_seen": 842512312,
      "step": 16375,
      "train_runtime": 100824.0114,
      "train_tokens_per_second": 8356.267
    },
    {
      "epoch": 1.1028817667654187,
      "grad_norm": 0.5904485576234717,
      "learning_rate": 4.196003006036743e-06,
      "loss": 0.31495821475982666,
      "num_input_tokens_seen": 842760776,
      "step": 16380,
      "train_runtime": 100857.4715,
      "train_tokens_per_second": 8355.958
    },
    {
      "epoch": 1.103218421761379,
      "grad_norm": 0.6343962950354208,
      "learning_rate": 4.193393443573576e-06,
      "loss": 0.305485200881958,
      "num_input_tokens_seen": 843035416,
      "step": 16385,
      "train_runtime": 100886.4151,
      "train_tokens_per_second": 8356.283
    },
    {
      "epoch": 1.103555076757339,
      "grad_norm": 0.6702562032048817,
      "learning_rate": 4.19078410667536e-06,
      "loss": 0.33074121475219725,
      "num_input_tokens_seen": 843284744,
      "step": 16390,
      "train_runtime": 100918.2779,
      "train_tokens_per_second": 8356.115
    },
    {
      "epoch": 1.1038917317532992,
      "grad_norm": 0.6725211285040439,
      "learning_rate": 4.188174996071788e-06,
      "loss": 0.31855006217956544,
      "num_input_tokens_seen": 843511208,
      "step": 16395,
      "train_runtime": 100948.803,
      "train_tokens_per_second": 8355.832
    },
    {
      "epoch": 1.1042283867492593,
      "grad_norm": 0.6869772398644177,
      "learning_rate": 4.18556611249249e-06,
      "loss": 0.30276265144348147,
      "num_input_tokens_seen": 843770648,
      "step": 16400,
      "train_runtime": 100981.3717,
      "train_tokens_per_second": 8355.706
    },
    {
      "epoch": 1.1045650417452195,
      "grad_norm": 0.8075758307117828,
      "learning_rate": 4.182957456667032e-06,
      "loss": 0.32773098945617674,
      "num_input_tokens_seen": 844026296,
      "step": 16405,
      "train_runtime": 101009.8188,
      "train_tokens_per_second": 8355.884
    },
    {
      "epoch": 1.1049016967411796,
      "grad_norm": 0.7182886860958013,
      "learning_rate": 4.1803490293249145e-06,
      "loss": 0.3246901035308838,
      "num_input_tokens_seen": 844287512,
      "step": 16410,
      "train_runtime": 101037.3833,
      "train_tokens_per_second": 8356.189
    },
    {
      "epoch": 1.1052383517371398,
      "grad_norm": 0.7464129096947015,
      "learning_rate": 4.1777408311955785e-06,
      "loss": 0.30339000225067136,
      "num_input_tokens_seen": 844526352,
      "step": 16415,
      "train_runtime": 101067.982,
      "train_tokens_per_second": 8356.023
    },
    {
      "epoch": 1.1055750067330998,
      "grad_norm": 0.6523691787478235,
      "learning_rate": 4.175132863008395e-06,
      "loss": 0.336089563369751,
      "num_input_tokens_seen": 844790512,
      "step": 16420,
      "train_runtime": 101098.9366,
      "train_tokens_per_second": 8356.077
    },
    {
      "epoch": 1.10591166172906,
      "grad_norm": 0.8397678041658462,
      "learning_rate": 4.17252512549268e-06,
      "loss": 0.30681347846984863,
      "num_input_tokens_seen": 845014624,
      "step": 16425,
      "train_runtime": 101129.6981,
      "train_tokens_per_second": 8355.751
    },
    {
      "epoch": 1.10624831672502,
      "grad_norm": 0.6657614392790722,
      "learning_rate": 4.169917619377671e-06,
      "loss": 0.3078404664993286,
      "num_input_tokens_seen": 845246600,
      "step": 16430,
      "train_runtime": 101159.1267,
      "train_tokens_per_second": 8355.614
    },
    {
      "epoch": 1.1065849717209804,
      "grad_norm": 0.6965886475909844,
      "learning_rate": 4.167310345392557e-06,
      "loss": 0.3397346496582031,
      "num_input_tokens_seen": 845500904,
      "step": 16435,
      "train_runtime": 101194.3705,
      "train_tokens_per_second": 8355.217
    },
    {
      "epoch": 1.1069216267169404,
      "grad_norm": 0.5782147362736061,
      "learning_rate": 4.1647033042664455e-06,
      "loss": 0.3291481971740723,
      "num_input_tokens_seen": 845749664,
      "step": 16440,
      "train_runtime": 101220.6234,
      "train_tokens_per_second": 8355.507
    },
    {
      "epoch": 1.1072582817129006,
      "grad_norm": 0.6636459003058068,
      "learning_rate": 4.162096496728394e-06,
      "loss": 0.2982253313064575,
      "num_input_tokens_seen": 846004424,
      "step": 16445,
      "train_runtime": 101251.3114,
      "train_tokens_per_second": 8355.491
    },
    {
      "epoch": 1.1075949367088607,
      "grad_norm": 0.591892685379499,
      "learning_rate": 4.1594899235073835e-06,
      "loss": 0.2939328670501709,
      "num_input_tokens_seen": 846273128,
      "step": 16450,
      "train_runtime": 101279.7621,
      "train_tokens_per_second": 8355.797
    },
    {
      "epoch": 1.107931591704821,
      "grad_norm": 0.5947888400364327,
      "learning_rate": 4.156883585332338e-06,
      "loss": 0.3273641586303711,
      "num_input_tokens_seen": 846543144,
      "step": 16455,
      "train_runtime": 101309.3005,
      "train_tokens_per_second": 8356.026
    },
    {
      "epoch": 1.108268246700781,
      "grad_norm": 0.665973045990279,
      "learning_rate": 4.154277482932107e-06,
      "loss": 0.3275979280471802,
      "num_input_tokens_seen": 846787744,
      "step": 16460,
      "train_runtime": 101339.6495,
      "train_tokens_per_second": 8355.937
    },
    {
      "epoch": 1.1086049016967412,
      "grad_norm": 0.6246058536379457,
      "learning_rate": 4.151671617035482e-06,
      "loss": 0.30124380588531496,
      "num_input_tokens_seen": 847038496,
      "step": 16465,
      "train_runtime": 101370.4833,
      "train_tokens_per_second": 8355.869
    },
    {
      "epoch": 1.1089415566927012,
      "grad_norm": 0.6733598887996348,
      "learning_rate": 4.149065988371184e-06,
      "loss": 0.2985043525695801,
      "num_input_tokens_seen": 847303304,
      "step": 16470,
      "train_runtime": 101403.1128,
      "train_tokens_per_second": 8355.792
    },
    {
      "epoch": 1.1092782116886615,
      "grad_norm": 0.6389267998077245,
      "learning_rate": 4.14646059766787e-06,
      "loss": 0.3076008319854736,
      "num_input_tokens_seen": 847551112,
      "step": 16475,
      "train_runtime": 101432.6604,
      "train_tokens_per_second": 8355.801
    },
    {
      "epoch": 1.1096148666846215,
      "grad_norm": 0.5942833260764463,
      "learning_rate": 4.143855445654127e-06,
      "loss": 0.2926583766937256,
      "num_input_tokens_seen": 847824144,
      "step": 16480,
      "train_runtime": 101464.6427,
      "train_tokens_per_second": 8355.858
    },
    {
      "epoch": 1.1099515216805818,
      "grad_norm": 0.6950204327978016,
      "learning_rate": 4.14125053305848e-06,
      "loss": 0.2904445886611938,
      "num_input_tokens_seen": 848058464,
      "step": 16485,
      "train_runtime": 101493.5644,
      "train_tokens_per_second": 8355.786
    },
    {
      "epoch": 1.1102881766765418,
      "grad_norm": 0.585102752325167,
      "learning_rate": 4.1386458606093805e-06,
      "loss": 0.30385332107543944,
      "num_input_tokens_seen": 848321648,
      "step": 16490,
      "train_runtime": 101524.1159,
      "train_tokens_per_second": 8355.863
    },
    {
      "epoch": 1.110624831672502,
      "grad_norm": 0.6915941911528798,
      "learning_rate": 4.1360414290352214e-06,
      "loss": 0.3327202796936035,
      "num_input_tokens_seen": 848598528,
      "step": 16495,
      "train_runtime": 101556.5177,
      "train_tokens_per_second": 8355.924
    },
    {
      "epoch": 1.110961486668462,
      "grad_norm": 0.6730912127610442,
      "learning_rate": 4.1334372390643205e-06,
      "loss": 0.3025339126586914,
      "num_input_tokens_seen": 848867424,
      "step": 16500,
      "train_runtime": 101584.9159,
      "train_tokens_per_second": 8356.235
    },
    {
      "epoch": 1.1112981416644223,
      "grad_norm": 0.5480459802531522,
      "learning_rate": 4.130833291424934e-06,
      "loss": 0.31500208377838135,
      "num_input_tokens_seen": 849132144,
      "step": 16505,
      "train_runtime": 101616.1148,
      "train_tokens_per_second": 8356.274
    },
    {
      "epoch": 1.1116347966603823,
      "grad_norm": 0.6721068108290534,
      "learning_rate": 4.128229586845244e-06,
      "loss": 0.30233724117279054,
      "num_input_tokens_seen": 849394104,
      "step": 16510,
      "train_runtime": 101642.6156,
      "train_tokens_per_second": 8356.673
    },
    {
      "epoch": 1.1119714516563426,
      "grad_norm": 0.5835791695618432,
      "learning_rate": 4.125626126053372e-06,
      "loss": 0.2942690372467041,
      "num_input_tokens_seen": 849666648,
      "step": 16515,
      "train_runtime": 101673.4159,
      "train_tokens_per_second": 8356.822
    },
    {
      "epoch": 1.1123081066523026,
      "grad_norm": 0.6447735991443272,
      "learning_rate": 4.123022909777364e-06,
      "loss": 0.3135091543197632,
      "num_input_tokens_seen": 849942728,
      "step": 16520,
      "train_runtime": 101702.0681,
      "train_tokens_per_second": 8357.182
    },
    {
      "epoch": 1.1126447616482629,
      "grad_norm": 0.7774133334210737,
      "learning_rate": 4.120419938745205e-06,
      "loss": 0.3070396423339844,
      "num_input_tokens_seen": 850217456,
      "step": 16525,
      "train_runtime": 101735.1075,
      "train_tokens_per_second": 8357.169
    },
    {
      "epoch": 1.112981416644223,
      "grad_norm": 0.7618952262824671,
      "learning_rate": 4.1178172136848035e-06,
      "loss": 0.29273672103881837,
      "num_input_tokens_seen": 850454888,
      "step": 16530,
      "train_runtime": 101764.1579,
      "train_tokens_per_second": 8357.116
    },
    {
      "epoch": 1.1133180716401831,
      "grad_norm": 0.7068670791459627,
      "learning_rate": 4.115214735324007e-06,
      "loss": 0.2861247301101685,
      "num_input_tokens_seen": 850739368,
      "step": 16535,
      "train_runtime": 101797.9267,
      "train_tokens_per_second": 8357.138
    },
    {
      "epoch": 1.1136547266361432,
      "grad_norm": 0.5333438834439996,
      "learning_rate": 4.112612504390586e-06,
      "loss": 0.269451379776001,
      "num_input_tokens_seen": 851006432,
      "step": 16540,
      "train_runtime": 101825.7735,
      "train_tokens_per_second": 8357.476
    },
    {
      "epoch": 1.1139913816321034,
      "grad_norm": 0.5923599149598823,
      "learning_rate": 4.11001052161225e-06,
      "loss": 0.3202071189880371,
      "num_input_tokens_seen": 851260200,
      "step": 16545,
      "train_runtime": 101857.551,
      "train_tokens_per_second": 8357.36
    },
    {
      "epoch": 1.1143280366280637,
      "grad_norm": 0.7072071509110393,
      "learning_rate": 4.107408787716632e-06,
      "loss": 0.30836460590362547,
      "num_input_tokens_seen": 851518248,
      "step": 16550,
      "train_runtime": 101889.8203,
      "train_tokens_per_second": 8357.246
    },
    {
      "epoch": 1.1146646916240237,
      "grad_norm": 0.6630331308992137,
      "learning_rate": 4.104807303431303e-06,
      "loss": 0.30173659324645996,
      "num_input_tokens_seen": 851770600,
      "step": 16555,
      "train_runtime": 101919.7846,
      "train_tokens_per_second": 8357.265
    },
    {
      "epoch": 1.1150013466199837,
      "grad_norm": 0.7150374894524848,
      "learning_rate": 4.102206069483754e-06,
      "loss": 0.32788848876953125,
      "num_input_tokens_seen": 852020568,
      "step": 16560,
      "train_runtime": 101949.6282,
      "train_tokens_per_second": 8357.27
    },
    {
      "epoch": 1.115338001615944,
      "grad_norm": 0.6787738155967814,
      "learning_rate": 4.099605086601417e-06,
      "loss": 0.2890244722366333,
      "num_input_tokens_seen": 852244592,
      "step": 16565,
      "train_runtime": 101980.2432,
      "train_tokens_per_second": 8356.958
    },
    {
      "epoch": 1.1156746566119042,
      "grad_norm": 0.6603856520445882,
      "learning_rate": 4.097004355511645e-06,
      "loss": 0.28168582916259766,
      "num_input_tokens_seen": 852506288,
      "step": 16570,
      "train_runtime": 102010.9047,
      "train_tokens_per_second": 8357.011
    },
    {
      "epoch": 1.1160113116078643,
      "grad_norm": 0.8466373914261651,
      "learning_rate": 4.094403876941728e-06,
      "loss": 0.2696842670440674,
      "num_input_tokens_seen": 852764976,
      "step": 16575,
      "train_runtime": 102040.8344,
      "train_tokens_per_second": 8357.095
    },
    {
      "epoch": 1.1163479666038243,
      "grad_norm": 0.6416820051673234,
      "learning_rate": 4.0918036516188755e-06,
      "loss": 0.3106646776199341,
      "num_input_tokens_seen": 852990456,
      "step": 16580,
      "train_runtime": 102071.2321,
      "train_tokens_per_second": 8356.816
    },
    {
      "epoch": 1.1166846215997845,
      "grad_norm": 0.6232943055788859,
      "learning_rate": 4.0892036802702395e-06,
      "loss": 0.3074917793273926,
      "num_input_tokens_seen": 853249304,
      "step": 16585,
      "train_runtime": 102098.158,
      "train_tokens_per_second": 8357.147
    },
    {
      "epoch": 1.1170212765957448,
      "grad_norm": 0.6868181276893939,
      "learning_rate": 4.086603963622887e-06,
      "loss": 0.2935677766799927,
      "num_input_tokens_seen": 853505968,
      "step": 16590,
      "train_runtime": 102129.6838,
      "train_tokens_per_second": 8357.08
    },
    {
      "epoch": 1.1173579315917048,
      "grad_norm": 0.6715687615293335,
      "learning_rate": 4.084004502403825e-06,
      "loss": 0.3107813835144043,
      "num_input_tokens_seen": 853767120,
      "step": 16595,
      "train_runtime": 102162.5542,
      "train_tokens_per_second": 8356.948
    },
    {
      "epoch": 1.1176945865876649,
      "grad_norm": 0.5755481318443784,
      "learning_rate": 4.081405297339982e-06,
      "loss": 0.30759921073913576,
      "num_input_tokens_seen": 854049880,
      "step": 16600,
      "train_runtime": 102195.2273,
      "train_tokens_per_second": 8357.043
    },
    {
      "epoch": 1.118031241583625,
      "grad_norm": 0.5945732377806127,
      "learning_rate": 4.0788063491582205e-06,
      "loss": 0.30746939182281496,
      "num_input_tokens_seen": 854288280,
      "step": 16605,
      "train_runtime": 102229.3299,
      "train_tokens_per_second": 8356.587
    },
    {
      "epoch": 1.1183678965795854,
      "grad_norm": 0.7984418393566359,
      "learning_rate": 4.076207658585323e-06,
      "loss": 0.30744004249572754,
      "num_input_tokens_seen": 854555704,
      "step": 16610,
      "train_runtime": 102261.2696,
      "train_tokens_per_second": 8356.592
    },
    {
      "epoch": 1.1187045515755454,
      "grad_norm": 0.5807436278587919,
      "learning_rate": 4.07360922634801e-06,
      "loss": 0.30896430015563964,
      "num_input_tokens_seen": 854824008,
      "step": 16615,
      "train_runtime": 102291.6499,
      "train_tokens_per_second": 8356.733
    },
    {
      "epoch": 1.1190412065715054,
      "grad_norm": 0.7151463555061205,
      "learning_rate": 4.071011053172922e-06,
      "loss": 0.30258519649505616,
      "num_input_tokens_seen": 855052008,
      "step": 16620,
      "train_runtime": 102321.4161,
      "train_tokens_per_second": 8356.53
    },
    {
      "epoch": 1.1193778615674657,
      "grad_norm": 0.6868634181348422,
      "learning_rate": 4.068413139786631e-06,
      "loss": 0.29320287704467773,
      "num_input_tokens_seen": 855304704,
      "step": 16625,
      "train_runtime": 102353.728,
      "train_tokens_per_second": 8356.361
    },
    {
      "epoch": 1.119714516563426,
      "grad_norm": 0.7544111318990142,
      "learning_rate": 4.065815486915635e-06,
      "loss": 0.30193209648132324,
      "num_input_tokens_seen": 855574912,
      "step": 16630,
      "train_runtime": 102386.1873,
      "train_tokens_per_second": 8356.351
    },
    {
      "epoch": 1.120051171559386,
      "grad_norm": 0.6464741260390477,
      "learning_rate": 4.063218095286359e-06,
      "loss": 0.31113808155059813,
      "num_input_tokens_seen": 855853744,
      "step": 16635,
      "train_runtime": 102415.5856,
      "train_tokens_per_second": 8356.675
    },
    {
      "epoch": 1.120387826555346,
      "grad_norm": 0.6629813144242627,
      "learning_rate": 4.0606209656251555e-06,
      "loss": 0.302203369140625,
      "num_input_tokens_seen": 856120440,
      "step": 16640,
      "train_runtime": 102446.3907,
      "train_tokens_per_second": 8356.765
    },
    {
      "epoch": 1.1207244815513062,
      "grad_norm": 0.6698328991470311,
      "learning_rate": 4.058024098658304e-06,
      "loss": 0.31984121799468995,
      "num_input_tokens_seen": 856385616,
      "step": 16645,
      "train_runtime": 102478.1102,
      "train_tokens_per_second": 8356.766
    },
    {
      "epoch": 1.1210611365472665,
      "grad_norm": 0.6700775303489109,
      "learning_rate": 4.055427495112008e-06,
      "loss": 0.33400535583496094,
      "num_input_tokens_seen": 856646616,
      "step": 16650,
      "train_runtime": 102507.6242,
      "train_tokens_per_second": 8356.906
    },
    {
      "epoch": 1.1213977915432265,
      "grad_norm": 0.6643913430065304,
      "learning_rate": 4.052831155712403e-06,
      "loss": 0.2927570104598999,
      "num_input_tokens_seen": 856905184,
      "step": 16655,
      "train_runtime": 102535.6994,
      "train_tokens_per_second": 8357.14
    },
    {
      "epoch": 1.1217344465391865,
      "grad_norm": 0.5931971822089357,
      "learning_rate": 4.050235081185544e-06,
      "loss": 0.2802257537841797,
      "num_input_tokens_seen": 857165176,
      "step": 16660,
      "train_runtime": 102564.8261,
      "train_tokens_per_second": 8357.302
    },
    {
      "epoch": 1.1220711015351468,
      "grad_norm": 0.6938874345723243,
      "learning_rate": 4.047639272257415e-06,
      "loss": 0.3122962474822998,
      "num_input_tokens_seen": 857425872,
      "step": 16665,
      "train_runtime": 102598.7751,
      "train_tokens_per_second": 8357.077
    },
    {
      "epoch": 1.122407756531107,
      "grad_norm": 1.3596397003921774,
      "learning_rate": 4.045043729653927e-06,
      "loss": 0.2781905889511108,
      "num_input_tokens_seen": 857694848,
      "step": 16670,
      "train_runtime": 102629.6924,
      "train_tokens_per_second": 8357.18
    },
    {
      "epoch": 1.122744411527067,
      "grad_norm": 0.6370073443045681,
      "learning_rate": 4.0424484541009175e-06,
      "loss": 0.3133222579956055,
      "num_input_tokens_seen": 857969512,
      "step": 16675,
      "train_runtime": 102662.3405,
      "train_tokens_per_second": 8357.198
    },
    {
      "epoch": 1.1230810665230273,
      "grad_norm": 0.7301205068238853,
      "learning_rate": 4.039853446324141e-06,
      "loss": 0.28787984848022463,
      "num_input_tokens_seen": 858208936,
      "step": 16680,
      "train_runtime": 102693.2398,
      "train_tokens_per_second": 8357.015
    },
    {
      "epoch": 1.1234177215189873,
      "grad_norm": 0.7455630458947646,
      "learning_rate": 4.037258707049289e-06,
      "loss": 0.29384377002716067,
      "num_input_tokens_seen": 858455144,
      "step": 16685,
      "train_runtime": 102723.8181,
      "train_tokens_per_second": 8356.924
    },
    {
      "epoch": 1.1237543765149476,
      "grad_norm": 0.5889459397926163,
      "learning_rate": 4.034664237001969e-06,
      "loss": 0.3010182619094849,
      "num_input_tokens_seen": 858704176,
      "step": 16690,
      "train_runtime": 102753.9531,
      "train_tokens_per_second": 8356.897
    },
    {
      "epoch": 1.1240910315109076,
      "grad_norm": 0.7040713380892127,
      "learning_rate": 4.032070036907716e-06,
      "loss": 0.31245949268341067,
      "num_input_tokens_seen": 858948032,
      "step": 16695,
      "train_runtime": 102785.0121,
      "train_tokens_per_second": 8356.744
    },
    {
      "epoch": 1.1244276865068679,
      "grad_norm": 0.6805682988249598,
      "learning_rate": 4.029476107491991e-06,
      "loss": 0.3175481081008911,
      "num_input_tokens_seen": 859205544,
      "step": 16700,
      "train_runtime": 102813.8226,
      "train_tokens_per_second": 8356.907
    },
    {
      "epoch": 1.124764341502828,
      "grad_norm": 0.6579298119514798,
      "learning_rate": 4.0268824494801784e-06,
      "loss": 0.30815463066101073,
      "num_input_tokens_seen": 859469984,
      "step": 16705,
      "train_runtime": 102843.7665,
      "train_tokens_per_second": 8357.045
    },
    {
      "epoch": 1.1251009964987881,
      "grad_norm": 0.6435975062607976,
      "learning_rate": 4.024289063597585e-06,
      "loss": 0.3092358589172363,
      "num_input_tokens_seen": 859695880,
      "step": 16710,
      "train_runtime": 102870.7074,
      "train_tokens_per_second": 8357.052
    },
    {
      "epoch": 1.1254376514947482,
      "grad_norm": 0.9486265404984269,
      "learning_rate": 4.021695950569443e-06,
      "loss": 0.2914973497390747,
      "num_input_tokens_seen": 859968328,
      "step": 16715,
      "train_runtime": 102900.0038,
      "train_tokens_per_second": 8357.321
    },
    {
      "epoch": 1.1257743064907082,
      "grad_norm": 0.7281107075268011,
      "learning_rate": 4.01910311112091e-06,
      "loss": 0.31534428596496583,
      "num_input_tokens_seen": 860222400,
      "step": 16720,
      "train_runtime": 102932.1177,
      "train_tokens_per_second": 8357.182
    },
    {
      "epoch": 1.1261109614866684,
      "grad_norm": 0.6345269625123698,
      "learning_rate": 4.016510545977064e-06,
      "loss": 0.30355381965637207,
      "num_input_tokens_seen": 860487672,
      "step": 16725,
      "train_runtime": 102963.9041,
      "train_tokens_per_second": 8357.178
    },
    {
      "epoch": 1.1264476164826287,
      "grad_norm": 0.784908706274172,
      "learning_rate": 4.013918255862908e-06,
      "loss": 0.3155029773712158,
      "num_input_tokens_seen": 860717488,
      "step": 16730,
      "train_runtime": 102997.2319,
      "train_tokens_per_second": 8356.705
    },
    {
      "epoch": 1.1267842714785887,
      "grad_norm": 0.7229268100086466,
      "learning_rate": 4.011326241503369e-06,
      "loss": 0.2955645561218262,
      "num_input_tokens_seen": 860957640,
      "step": 16735,
      "train_runtime": 103029.4317,
      "train_tokens_per_second": 8356.424
    },
    {
      "epoch": 1.127120926474549,
      "grad_norm": 0.7028340771785913,
      "learning_rate": 4.00873450362329e-06,
      "loss": 0.32819485664367676,
      "num_input_tokens_seen": 861213752,
      "step": 16740,
      "train_runtime": 103065.403,
      "train_tokens_per_second": 8355.993
    },
    {
      "epoch": 1.127457581470509,
      "grad_norm": 0.6642220763815289,
      "learning_rate": 4.0061430429474476e-06,
      "loss": 0.3117096185684204,
      "num_input_tokens_seen": 861447640,
      "step": 16745,
      "train_runtime": 103098.7758,
      "train_tokens_per_second": 8355.556
    },
    {
      "epoch": 1.1277942364664693,
      "grad_norm": 0.6494085717148491,
      "learning_rate": 4.003551860200534e-06,
      "loss": 0.31165523529052735,
      "num_input_tokens_seen": 861680128,
      "step": 16750,
      "train_runtime": 103132.7522,
      "train_tokens_per_second": 8355.058
    },
    {
      "epoch": 1.1281308914624293,
      "grad_norm": 0.6568902423268487,
      "learning_rate": 4.000960956107167e-06,
      "loss": 0.320721435546875,
      "num_input_tokens_seen": 861922304,
      "step": 16755,
      "train_runtime": 103160.9547,
      "train_tokens_per_second": 8355.121
    },
    {
      "epoch": 1.1284675464583895,
      "grad_norm": 0.6234923232116176,
      "learning_rate": 3.998370331391881e-06,
      "loss": 0.2836894512176514,
      "num_input_tokens_seen": 862151488,
      "step": 16760,
      "train_runtime": 103189.7446,
      "train_tokens_per_second": 8355.011
    },
    {
      "epoch": 1.1288042014543496,
      "grad_norm": 0.7842739078439134,
      "learning_rate": 3.995779986779139e-06,
      "loss": 0.30156965255737306,
      "num_input_tokens_seen": 862378768,
      "step": 16765,
      "train_runtime": 103218.0031,
      "train_tokens_per_second": 8354.926
    },
    {
      "epoch": 1.1291408564503098,
      "grad_norm": 0.6480849728368644,
      "learning_rate": 3.9931899229933204e-06,
      "loss": 0.3299871444702148,
      "num_input_tokens_seen": 862645184,
      "step": 16770,
      "train_runtime": 103245.5333,
      "train_tokens_per_second": 8355.278
    },
    {
      "epoch": 1.1294775114462698,
      "grad_norm": 0.6416987955820219,
      "learning_rate": 3.990600140758731e-06,
      "loss": 0.3031825065612793,
      "num_input_tokens_seen": 862918072,
      "step": 16775,
      "train_runtime": 103275.5884,
      "train_tokens_per_second": 8355.489
    },
    {
      "epoch": 1.12981416644223,
      "grad_norm": 0.5472629459513358,
      "learning_rate": 3.988010640799591e-06,
      "loss": 0.28463385105133054,
      "num_input_tokens_seen": 863172528,
      "step": 16780,
      "train_runtime": 103308.1478,
      "train_tokens_per_second": 8355.319
    },
    {
      "epoch": 1.1301508214381901,
      "grad_norm": 0.8074127964549361,
      "learning_rate": 3.985421423840051e-06,
      "loss": 0.33401663303375245,
      "num_input_tokens_seen": 863422616,
      "step": 16785,
      "train_runtime": 103341.3869,
      "train_tokens_per_second": 8355.052
    },
    {
      "epoch": 1.1304874764341504,
      "grad_norm": 0.6178680068046836,
      "learning_rate": 3.982832490604173e-06,
      "loss": 0.2883934020996094,
      "num_input_tokens_seen": 863689112,
      "step": 16790,
      "train_runtime": 103370.0446,
      "train_tokens_per_second": 8355.313
    },
    {
      "epoch": 1.1308241314301104,
      "grad_norm": 0.7074732179919915,
      "learning_rate": 3.9802438418159465e-06,
      "loss": 0.3236823081970215,
      "num_input_tokens_seen": 863986496,
      "step": 16795,
      "train_runtime": 103403.5199,
      "train_tokens_per_second": 8355.484
    },
    {
      "epoch": 1.1311607864260707,
      "grad_norm": 0.7027931563332835,
      "learning_rate": 3.977655478199275e-06,
      "loss": 0.317358136177063,
      "num_input_tokens_seen": 864255944,
      "step": 16800,
      "train_runtime": 103434.6292,
      "train_tokens_per_second": 8355.576
    },
    {
      "epoch": 1.1314974414220307,
      "grad_norm": 0.7488900177830008,
      "learning_rate": 3.975067400477993e-06,
      "loss": 0.33366944789886477,
      "num_input_tokens_seen": 864486072,
      "step": 16805,
      "train_runtime": 103461.2321,
      "train_tokens_per_second": 8355.652
    },
    {
      "epoch": 1.131834096417991,
      "grad_norm": 0.6642281892631939,
      "learning_rate": 3.972479609375842e-06,
      "loss": 0.3194951295852661,
      "num_input_tokens_seen": 864763728,
      "step": 16810,
      "train_runtime": 103491.2527,
      "train_tokens_per_second": 8355.911
    },
    {
      "epoch": 1.132170751413951,
      "grad_norm": 0.6795507984780602,
      "learning_rate": 3.969892105616492e-06,
      "loss": 0.3067826509475708,
      "num_input_tokens_seen": 865029656,
      "step": 16815,
      "train_runtime": 103521.0546,
      "train_tokens_per_second": 8356.075
    },
    {
      "epoch": 1.1325074064099112,
      "grad_norm": 0.6654103361282868,
      "learning_rate": 3.967304889923529e-06,
      "loss": 0.2940589666366577,
      "num_input_tokens_seen": 865279488,
      "step": 16820,
      "train_runtime": 103547.3399,
      "train_tokens_per_second": 8356.366
    },
    {
      "epoch": 1.1328440614058712,
      "grad_norm": 0.7243421134732356,
      "learning_rate": 3.964717963020463e-06,
      "loss": 0.31104493141174316,
      "num_input_tokens_seen": 865541352,
      "step": 16825,
      "train_runtime": 103579.7202,
      "train_tokens_per_second": 8356.282
    },
    {
      "epoch": 1.1331807164018315,
      "grad_norm": 0.6669677508154379,
      "learning_rate": 3.962131325630715e-06,
      "loss": 0.2953691005706787,
      "num_input_tokens_seen": 865797056,
      "step": 16830,
      "train_runtime": 103610.0225,
      "train_tokens_per_second": 8356.306
    },
    {
      "epoch": 1.1335173713977915,
      "grad_norm": 0.7564798926426155,
      "learning_rate": 3.959544978477634e-06,
      "loss": 0.336243200302124,
      "num_input_tokens_seen": 866043088,
      "step": 16835,
      "train_runtime": 103642.8167,
      "train_tokens_per_second": 8356.036
    },
    {
      "epoch": 1.1338540263937518,
      "grad_norm": 0.6110168677281218,
      "learning_rate": 3.9569589222844805e-06,
      "loss": 0.30751473903656007,
      "num_input_tokens_seen": 866282144,
      "step": 16840,
      "train_runtime": 103673.3117,
      "train_tokens_per_second": 8355.884
    },
    {
      "epoch": 1.1341906813897118,
      "grad_norm": 0.6500893130043437,
      "learning_rate": 3.954373157774439e-06,
      "loss": 0.30432472229003904,
      "num_input_tokens_seen": 866552896,
      "step": 16845,
      "train_runtime": 103703.0289,
      "train_tokens_per_second": 8356.1
    },
    {
      "epoch": 1.134527336385672,
      "grad_norm": 0.7085496508790573,
      "learning_rate": 3.951787685670609e-06,
      "loss": 0.31612162590026854,
      "num_input_tokens_seen": 866806400,
      "step": 16850,
      "train_runtime": 103737.6948,
      "train_tokens_per_second": 8355.752
    },
    {
      "epoch": 1.134863991381632,
      "grad_norm": 0.6918391922174646,
      "learning_rate": 3.949202506696012e-06,
      "loss": 0.3126729726791382,
      "num_input_tokens_seen": 867057232,
      "step": 16855,
      "train_runtime": 103769.4126,
      "train_tokens_per_second": 8355.615
    },
    {
      "epoch": 1.1352006463775923,
      "grad_norm": 0.5714223234438583,
      "learning_rate": 3.946617621573581e-06,
      "loss": 0.30724587440490725,
      "num_input_tokens_seen": 867340016,
      "step": 16860,
      "train_runtime": 103803.1823,
      "train_tokens_per_second": 8355.621
    },
    {
      "epoch": 1.1355373013735524,
      "grad_norm": 0.7397899669432535,
      "learning_rate": 3.944033031026175e-06,
      "loss": 0.28271188735961916,
      "num_input_tokens_seen": 867563336,
      "step": 16865,
      "train_runtime": 103831.5605,
      "train_tokens_per_second": 8355.488
    },
    {
      "epoch": 1.1358739563695126,
      "grad_norm": 0.6602808679727027,
      "learning_rate": 3.941448735776563e-06,
      "loss": 0.29501805305480955,
      "num_input_tokens_seen": 867836512,
      "step": 16870,
      "train_runtime": 103859.7523,
      "train_tokens_per_second": 8355.85
    },
    {
      "epoch": 1.1362106113654726,
      "grad_norm": 0.5925545393420967,
      "learning_rate": 3.938864736547439e-06,
      "loss": 0.3131247043609619,
      "num_input_tokens_seen": 868114296,
      "step": 16875,
      "train_runtime": 103890.8742,
      "train_tokens_per_second": 8356.021
    },
    {
      "epoch": 1.1365472663614329,
      "grad_norm": 0.7435790980947489,
      "learning_rate": 3.936281034061405e-06,
      "loss": 0.2967257499694824,
      "num_input_tokens_seen": 868369984,
      "step": 16880,
      "train_runtime": 103918.7799,
      "train_tokens_per_second": 8356.237
    },
    {
      "epoch": 1.136883921357393,
      "grad_norm": 0.6386582701979503,
      "learning_rate": 3.93369762904099e-06,
      "loss": 0.31355862617492675,
      "num_input_tokens_seen": 868604856,
      "step": 16885,
      "train_runtime": 103950.1779,
      "train_tokens_per_second": 8355.973
    },
    {
      "epoch": 1.1372205763533532,
      "grad_norm": 0.6913966656736273,
      "learning_rate": 3.931114522208631e-06,
      "loss": 0.3171189785003662,
      "num_input_tokens_seen": 868832984,
      "step": 16890,
      "train_runtime": 103976.5011,
      "train_tokens_per_second": 8356.051
    },
    {
      "epoch": 1.1375572313493132,
      "grad_norm": 0.6646203076518454,
      "learning_rate": 3.928531714286689e-06,
      "loss": 0.3087935447692871,
      "num_input_tokens_seen": 869099304,
      "step": 16895,
      "train_runtime": 104007.7428,
      "train_tokens_per_second": 8356.102
    },
    {
      "epoch": 1.1378938863452734,
      "grad_norm": 0.7916760871924293,
      "learning_rate": 3.925949205997434e-06,
      "loss": 0.3291207551956177,
      "num_input_tokens_seen": 869331080,
      "step": 16900,
      "train_runtime": 104034.2972,
      "train_tokens_per_second": 8356.197
    },
    {
      "epoch": 1.1382305413412335,
      "grad_norm": 0.7422982519232997,
      "learning_rate": 3.923366998063062e-06,
      "loss": 0.2871209383010864,
      "num_input_tokens_seen": 869579584,
      "step": 16905,
      "train_runtime": 104066.0577,
      "train_tokens_per_second": 8356.035
    },
    {
      "epoch": 1.1385671963371937,
      "grad_norm": 0.5484138276192869,
      "learning_rate": 3.920785091205674e-06,
      "loss": 0.2841703176498413,
      "num_input_tokens_seen": 869863360,
      "step": 16910,
      "train_runtime": 104100.4205,
      "train_tokens_per_second": 8356.002
    },
    {
      "epoch": 1.1389038513331537,
      "grad_norm": 0.6959066677230432,
      "learning_rate": 3.918203486147294e-06,
      "loss": 0.34468488693237304,
      "num_input_tokens_seen": 870112336,
      "step": 16915,
      "train_runtime": 104131.7253,
      "train_tokens_per_second": 8355.881
    },
    {
      "epoch": 1.139240506329114,
      "grad_norm": 0.6948086597549358,
      "learning_rate": 3.9156221836098594e-06,
      "loss": 0.3188822269439697,
      "num_input_tokens_seen": 870393416,
      "step": 16920,
      "train_runtime": 104160.8919,
      "train_tokens_per_second": 8356.24
    },
    {
      "epoch": 1.139577161325074,
      "grad_norm": 0.6494490374191518,
      "learning_rate": 3.913041184315224e-06,
      "loss": 0.3146040439605713,
      "num_input_tokens_seen": 870639048,
      "step": 16925,
      "train_runtime": 104192.4096,
      "train_tokens_per_second": 8356.07
    },
    {
      "epoch": 1.1399138163210343,
      "grad_norm": 0.667020973531052,
      "learning_rate": 3.910460488985154e-06,
      "loss": 0.31246581077575686,
      "num_input_tokens_seen": 870876720,
      "step": 16930,
      "train_runtime": 104221.4898,
      "train_tokens_per_second": 8356.019
    },
    {
      "epoch": 1.1402504713169943,
      "grad_norm": 0.650516640950316,
      "learning_rate": 3.9078800983413355e-06,
      "loss": 0.3025388240814209,
      "num_input_tokens_seen": 871148712,
      "step": 16935,
      "train_runtime": 104253.6616,
      "train_tokens_per_second": 8356.049
    },
    {
      "epoch": 1.1405871263129546,
      "grad_norm": 0.6281895385559993,
      "learning_rate": 3.905300013105365e-06,
      "loss": 0.2823931694030762,
      "num_input_tokens_seen": 871416544,
      "step": 16940,
      "train_runtime": 104283.1234,
      "train_tokens_per_second": 8356.257
    },
    {
      "epoch": 1.1409237813089146,
      "grad_norm": 0.6698283320293591,
      "learning_rate": 3.902720233998754e-06,
      "loss": 0.30697948932647706,
      "num_input_tokens_seen": 871671440,
      "step": 16945,
      "train_runtime": 104314.9175,
      "train_tokens_per_second": 8356.153
    },
    {
      "epoch": 1.1412604363048748,
      "grad_norm": 0.5489959160820703,
      "learning_rate": 3.90014076174293e-06,
      "loss": 0.31130781173706057,
      "num_input_tokens_seen": 871946584,
      "step": 16950,
      "train_runtime": 104343.7818,
      "train_tokens_per_second": 8356.479
    },
    {
      "epoch": 1.1415970913008349,
      "grad_norm": 0.5944262760638814,
      "learning_rate": 3.897561597059237e-06,
      "loss": 0.2916730403900146,
      "num_input_tokens_seen": 872221664,
      "step": 16955,
      "train_runtime": 104370.6688,
      "train_tokens_per_second": 8356.962
    },
    {
      "epoch": 1.1419337462967951,
      "grad_norm": 0.6502159521949479,
      "learning_rate": 3.894982740668927e-06,
      "loss": 0.30435729026794434,
      "num_input_tokens_seen": 872462624,
      "step": 16960,
      "train_runtime": 104404.6402,
      "train_tokens_per_second": 8356.55
    },
    {
      "epoch": 1.1422704012927551,
      "grad_norm": 0.6002263215958288,
      "learning_rate": 3.89240419329317e-06,
      "loss": 0.32088651657104494,
      "num_input_tokens_seen": 872755080,
      "step": 16965,
      "train_runtime": 104435.4875,
      "train_tokens_per_second": 8356.882
    },
    {
      "epoch": 1.1426070562887154,
      "grad_norm": 0.593978712260147,
      "learning_rate": 3.889825955653046e-06,
      "loss": 0.3060433387756348,
      "num_input_tokens_seen": 873025792,
      "step": 16970,
      "train_runtime": 104462.8395,
      "train_tokens_per_second": 8357.286
    },
    {
      "epoch": 1.1429437112846754,
      "grad_norm": 0.5888642133664785,
      "learning_rate": 3.887248028469558e-06,
      "loss": 0.27785696983337405,
      "num_input_tokens_seen": 873307432,
      "step": 16975,
      "train_runtime": 104491.8892,
      "train_tokens_per_second": 8357.658
    },
    {
      "epoch": 1.1432803662806357,
      "grad_norm": 0.5559333055950306,
      "learning_rate": 3.884670412463607e-06,
      "loss": 0.2932558059692383,
      "num_input_tokens_seen": 873555768,
      "step": 16980,
      "train_runtime": 104522.6905,
      "train_tokens_per_second": 8357.571
    },
    {
      "epoch": 1.1436170212765957,
      "grad_norm": 0.6871794847630817,
      "learning_rate": 3.8820931083560194e-06,
      "loss": 0.3167438983917236,
      "num_input_tokens_seen": 873791512,
      "step": 16985,
      "train_runtime": 104548.2634,
      "train_tokens_per_second": 8357.781
    },
    {
      "epoch": 1.143953676272556,
      "grad_norm": 0.5984519143600413,
      "learning_rate": 3.879516116867531e-06,
      "loss": 0.29274654388427734,
      "num_input_tokens_seen": 874042816,
      "step": 16990,
      "train_runtime": 104585.1442,
      "train_tokens_per_second": 8357.237
    },
    {
      "epoch": 1.144290331268516,
      "grad_norm": 0.7263983290598066,
      "learning_rate": 3.876939438718786e-06,
      "loss": 0.3093990087509155,
      "num_input_tokens_seen": 874282544,
      "step": 16995,
      "train_runtime": 104615.9862,
      "train_tokens_per_second": 8357.064
    },
    {
      "epoch": 1.1446269862644762,
      "grad_norm": 0.591618717626823,
      "learning_rate": 3.874363074630345e-06,
      "loss": 0.29619851112365725,
      "num_input_tokens_seen": 874523944,
      "step": 17000,
      "train_runtime": 104649.6326,
      "train_tokens_per_second": 8356.684
    },
    {
      "epoch": 1.1449636412604363,
      "grad_norm": 0.7253934823007644,
      "learning_rate": 3.871787025322681e-06,
      "loss": 0.31036360263824464,
      "num_input_tokens_seen": 874794648,
      "step": 17005,
      "train_runtime": 104680.7345,
      "train_tokens_per_second": 8356.787
    },
    {
      "epoch": 1.1453002962563965,
      "grad_norm": 0.7628185984406366,
      "learning_rate": 3.86921129151618e-06,
      "loss": 0.30371053218841554,
      "num_input_tokens_seen": 875047656,
      "step": 17010,
      "train_runtime": 104715.4224,
      "train_tokens_per_second": 8356.435
    },
    {
      "epoch": 1.1456369512523565,
      "grad_norm": 0.6704364165340144,
      "learning_rate": 3.866635873931133e-06,
      "loss": 0.32897803783416746,
      "num_input_tokens_seen": 875292928,
      "step": 17015,
      "train_runtime": 104746.0595,
      "train_tokens_per_second": 8356.333
    },
    {
      "epoch": 1.1459736062483168,
      "grad_norm": 0.6840333999320259,
      "learning_rate": 3.864060773287751e-06,
      "loss": 0.313144850730896,
      "num_input_tokens_seen": 875560664,
      "step": 17020,
      "train_runtime": 104781.0006,
      "train_tokens_per_second": 8356.101
    },
    {
      "epoch": 1.1463102612442768,
      "grad_norm": 0.6184192616141311,
      "learning_rate": 3.861485990306151e-06,
      "loss": 0.3094632625579834,
      "num_input_tokens_seen": 875818288,
      "step": 17025,
      "train_runtime": 104808.247,
      "train_tokens_per_second": 8356.387
    },
    {
      "epoch": 1.146646916240237,
      "grad_norm": 0.6442864589026236,
      "learning_rate": 3.858911525706365e-06,
      "loss": 0.29442310333251953,
      "num_input_tokens_seen": 876097192,
      "step": 17030,
      "train_runtime": 104840.3945,
      "train_tokens_per_second": 8356.485
    },
    {
      "epoch": 1.146983571236197,
      "grad_norm": 0.6831757728562,
      "learning_rate": 3.85633738020833e-06,
      "loss": 0.33914618492126464,
      "num_input_tokens_seen": 876385656,
      "step": 17035,
      "train_runtime": 104867.8141,
      "train_tokens_per_second": 8357.051
    },
    {
      "epoch": 1.1473202262321573,
      "grad_norm": 0.69212545449256,
      "learning_rate": 3.853763554531902e-06,
      "loss": 0.32245583534240724,
      "num_input_tokens_seen": 876614400,
      "step": 17040,
      "train_runtime": 104895.486,
      "train_tokens_per_second": 8357.027
    },
    {
      "epoch": 1.1476568812281174,
      "grad_norm": 0.609036815109161,
      "learning_rate": 3.851190049396839e-06,
      "loss": 0.2844985008239746,
      "num_input_tokens_seen": 876873016,
      "step": 17045,
      "train_runtime": 104928.4707,
      "train_tokens_per_second": 8356.865
    },
    {
      "epoch": 1.1479935362240776,
      "grad_norm": 0.5885364381426806,
      "learning_rate": 3.848616865522815e-06,
      "loss": 0.3054988384246826,
      "num_input_tokens_seen": 877137536,
      "step": 17050,
      "train_runtime": 104958.8816,
      "train_tokens_per_second": 8356.963
    },
    {
      "epoch": 1.1483301912200377,
      "grad_norm": 0.655744425926811,
      "learning_rate": 3.846044003629414e-06,
      "loss": 0.2989304304122925,
      "num_input_tokens_seen": 877397424,
      "step": 17055,
      "train_runtime": 104987.9082,
      "train_tokens_per_second": 8357.128
    },
    {
      "epoch": 1.148666846215998,
      "grad_norm": 0.6156546249883895,
      "learning_rate": 3.843471464436128e-06,
      "loss": 0.2828042984008789,
      "num_input_tokens_seen": 877629272,
      "step": 17060,
      "train_runtime": 105015.3117,
      "train_tokens_per_second": 8357.155
    },
    {
      "epoch": 1.149003501211958,
      "grad_norm": 0.6802363430087818,
      "learning_rate": 3.840899248662358e-06,
      "loss": 0.3308852672576904,
      "num_input_tokens_seen": 877872872,
      "step": 17065,
      "train_runtime": 105049.6997,
      "train_tokens_per_second": 8356.739
    },
    {
      "epoch": 1.1493401562079182,
      "grad_norm": 0.5976390080552436,
      "learning_rate": 3.838327357027419e-06,
      "loss": 0.3099637985229492,
      "num_input_tokens_seen": 878129752,
      "step": 17070,
      "train_runtime": 105077.9725,
      "train_tokens_per_second": 8356.935
    },
    {
      "epoch": 1.1496768112038782,
      "grad_norm": 0.6435928684016471,
      "learning_rate": 3.835755790250528e-06,
      "loss": 0.30886523723602294,
      "num_input_tokens_seen": 878393728,
      "step": 17075,
      "train_runtime": 105108.3119,
      "train_tokens_per_second": 8357.034
    },
    {
      "epoch": 1.1500134661998385,
      "grad_norm": 0.5827325614714502,
      "learning_rate": 3.83318454905082e-06,
      "loss": 0.29399499893188474,
      "num_input_tokens_seen": 878645352,
      "step": 17080,
      "train_runtime": 105138.1593,
      "train_tokens_per_second": 8357.055
    },
    {
      "epoch": 1.1503501211957985,
      "grad_norm": 0.6727532231389034,
      "learning_rate": 3.830613634147331e-06,
      "loss": 0.3152384996414185,
      "num_input_tokens_seen": 878902632,
      "step": 17085,
      "train_runtime": 105169.5963,
      "train_tokens_per_second": 8357.003
    },
    {
      "epoch": 1.1506867761917587,
      "grad_norm": 0.6053368601370069,
      "learning_rate": 3.82804304625901e-06,
      "loss": 0.3099790573120117,
      "num_input_tokens_seen": 879190264,
      "step": 17090,
      "train_runtime": 105202.488,
      "train_tokens_per_second": 8357.124
    },
    {
      "epoch": 1.1510234311877188,
      "grad_norm": 0.6888479518635179,
      "learning_rate": 3.825472786104715e-06,
      "loss": 0.2965691089630127,
      "num_input_tokens_seen": 879468472,
      "step": 17095,
      "train_runtime": 105238.3121,
      "train_tokens_per_second": 8356.923
    },
    {
      "epoch": 1.151360086183679,
      "grad_norm": 0.8068324914827761,
      "learning_rate": 3.8229028544032095e-06,
      "loss": 0.31577157974243164,
      "num_input_tokens_seen": 879694432,
      "step": 17100,
      "train_runtime": 105270.3618,
      "train_tokens_per_second": 8356.525
    },
    {
      "epoch": 1.151696741179639,
      "grad_norm": 0.8214807951011464,
      "learning_rate": 3.8203332518731666e-06,
      "loss": 0.3195110559463501,
      "num_input_tokens_seen": 879962888,
      "step": 17105,
      "train_runtime": 105300.7119,
      "train_tokens_per_second": 8356.666
    },
    {
      "epoch": 1.1520333961755993,
      "grad_norm": 0.548850801433862,
      "learning_rate": 3.8177639792331695e-06,
      "loss": 0.2933152198791504,
      "num_input_tokens_seen": 880232768,
      "step": 17110,
      "train_runtime": 105331.5774,
      "train_tokens_per_second": 8356.78
    },
    {
      "epoch": 1.1523700511715593,
      "grad_norm": 0.7897373865941482,
      "learning_rate": 3.815195037201704e-06,
      "loss": 0.341682505607605,
      "num_input_tokens_seen": 880471456,
      "step": 17115,
      "train_runtime": 105357.5207,
      "train_tokens_per_second": 8356.987
    },
    {
      "epoch": 1.1527067061675196,
      "grad_norm": 0.6455756598935148,
      "learning_rate": 3.8126264264971694e-06,
      "loss": 0.3241943120956421,
      "num_input_tokens_seen": 880731784,
      "step": 17120,
      "train_runtime": 105387.2251,
      "train_tokens_per_second": 8357.102
    },
    {
      "epoch": 1.1530433611634796,
      "grad_norm": 0.694613792227836,
      "learning_rate": 3.8100581478378673e-06,
      "loss": 0.30988268852233886,
      "num_input_tokens_seen": 880983016,
      "step": 17125,
      "train_runtime": 105422.2858,
      "train_tokens_per_second": 8356.706
    },
    {
      "epoch": 1.1533800161594399,
      "grad_norm": 0.7381870814088827,
      "learning_rate": 3.807490201942012e-06,
      "loss": 0.3057543277740479,
      "num_input_tokens_seen": 881248600,
      "step": 17130,
      "train_runtime": 105456.5022,
      "train_tokens_per_second": 8356.513
    },
    {
      "epoch": 1.1537166711553999,
      "grad_norm": 0.693247255943654,
      "learning_rate": 3.804922589527717e-06,
      "loss": 0.29403057098388674,
      "num_input_tokens_seen": 881504952,
      "step": 17135,
      "train_runtime": 105485.2836,
      "train_tokens_per_second": 8356.663
    },
    {
      "epoch": 1.1540533261513601,
      "grad_norm": 0.6009906827454817,
      "learning_rate": 3.8023553113130096e-06,
      "loss": 0.3182498931884766,
      "num_input_tokens_seen": 881789248,
      "step": 17140,
      "train_runtime": 105519.0647,
      "train_tokens_per_second": 8356.682
    },
    {
      "epoch": 1.1543899811473202,
      "grad_norm": 0.6238271653417137,
      "learning_rate": 3.799788368015821e-06,
      "loss": 0.27831790447235105,
      "num_input_tokens_seen": 882050328,
      "step": 17145,
      "train_runtime": 105545.8709,
      "train_tokens_per_second": 8357.033
    },
    {
      "epoch": 1.1547266361432804,
      "grad_norm": 0.7187802324314922,
      "learning_rate": 3.7972217603539873e-06,
      "loss": 0.3025989532470703,
      "num_input_tokens_seen": 882293056,
      "step": 17150,
      "train_runtime": 105575.845,
      "train_tokens_per_second": 8356.959
    },
    {
      "epoch": 1.1550632911392404,
      "grad_norm": 0.7498164474969755,
      "learning_rate": 3.7946554890452524e-06,
      "loss": 0.3129431962966919,
      "num_input_tokens_seen": 882569632,
      "step": 17155,
      "train_runtime": 105604.8396,
      "train_tokens_per_second": 8357.284
    },
    {
      "epoch": 1.1553999461352007,
      "grad_norm": 0.6735483106589689,
      "learning_rate": 3.792089554807269e-06,
      "loss": 0.3003907442092896,
      "num_input_tokens_seen": 882837760,
      "step": 17160,
      "train_runtime": 105632.8598,
      "train_tokens_per_second": 8357.605
    },
    {
      "epoch": 1.1557366011311607,
      "grad_norm": 0.7496721644714822,
      "learning_rate": 3.789523958357587e-06,
      "loss": 0.31588494777679443,
      "num_input_tokens_seen": 883076184,
      "step": 17165,
      "train_runtime": 105663.4149,
      "train_tokens_per_second": 8357.445
    },
    {
      "epoch": 1.156073256127121,
      "grad_norm": 0.6145600972554757,
      "learning_rate": 3.7869587004136714e-06,
      "loss": 0.27930805683135984,
      "num_input_tokens_seen": 883353056,
      "step": 17170,
      "train_runtime": 105697.2284,
      "train_tokens_per_second": 8357.391
    },
    {
      "epoch": 1.156409911123081,
      "grad_norm": 0.6585477866443202,
      "learning_rate": 3.7843937816928864e-06,
      "loss": 0.3084217071533203,
      "num_input_tokens_seen": 883625544,
      "step": 17175,
      "train_runtime": 105726.4498,
      "train_tokens_per_second": 8357.658
    },
    {
      "epoch": 1.1567465661190413,
      "grad_norm": 0.7300518975272322,
      "learning_rate": 3.781829202912506e-06,
      "loss": 0.3192928314208984,
      "num_input_tokens_seen": 883882496,
      "step": 17180,
      "train_runtime": 105756.3311,
      "train_tokens_per_second": 8357.727
    },
    {
      "epoch": 1.1570832211150013,
      "grad_norm": 0.6336384004015022,
      "learning_rate": 3.7792649647897023e-06,
      "loss": 0.3120211124420166,
      "num_input_tokens_seen": 884154792,
      "step": 17185,
      "train_runtime": 105787.546,
      "train_tokens_per_second": 8357.834
    },
    {
      "epoch": 1.1574198761109615,
      "grad_norm": 0.6466410870493421,
      "learning_rate": 3.7767010680415595e-06,
      "loss": 0.2996551752090454,
      "num_input_tokens_seen": 884401432,
      "step": 17190,
      "train_runtime": 105816.0897,
      "train_tokens_per_second": 8357.911
    },
    {
      "epoch": 1.1577565311069216,
      "grad_norm": 0.7541401830885729,
      "learning_rate": 3.774137513385061e-06,
      "loss": 0.2956112861633301,
      "num_input_tokens_seen": 884647136,
      "step": 17195,
      "train_runtime": 105846.6008,
      "train_tokens_per_second": 8357.823
    },
    {
      "epoch": 1.1580931861028818,
      "grad_norm": 0.6157972660874018,
      "learning_rate": 3.7715743015371005e-06,
      "loss": 0.31571102142333984,
      "num_input_tokens_seen": 884919216,
      "step": 17200,
      "train_runtime": 105878.1885,
      "train_tokens_per_second": 8357.899
    },
    {
      "epoch": 1.1584298410988418,
      "grad_norm": 0.6295744337615994,
      "learning_rate": 3.769011433214467e-06,
      "loss": 0.28450348377227785,
      "num_input_tokens_seen": 885173736,
      "step": 17205,
      "train_runtime": 105910.4434,
      "train_tokens_per_second": 8357.757
    },
    {
      "epoch": 1.158766496094802,
      "grad_norm": 0.6494366879717509,
      "learning_rate": 3.766448909133863e-06,
      "loss": 0.3168362855911255,
      "num_input_tokens_seen": 885459392,
      "step": 17210,
      "train_runtime": 105941.841,
      "train_tokens_per_second": 8357.976
    },
    {
      "epoch": 1.1591031510907621,
      "grad_norm": 0.7031313270290526,
      "learning_rate": 3.7638867300118864e-06,
      "loss": 0.28278446197509766,
      "num_input_tokens_seen": 885707376,
      "step": 17215,
      "train_runtime": 105977.1074,
      "train_tokens_per_second": 8357.535
    },
    {
      "epoch": 1.1594398060867224,
      "grad_norm": 0.8835556193950778,
      "learning_rate": 3.7613248965650447e-06,
      "loss": 0.2978202342987061,
      "num_input_tokens_seen": 885982280,
      "step": 17220,
      "train_runtime": 106007.1592,
      "train_tokens_per_second": 8357.759
    },
    {
      "epoch": 1.1597764610826824,
      "grad_norm": 0.6319147388648432,
      "learning_rate": 3.758763409509745e-06,
      "loss": 0.29429779052734373,
      "num_input_tokens_seen": 886251240,
      "step": 17225,
      "train_runtime": 106036.9409,
      "train_tokens_per_second": 8357.948
    },
    {
      "epoch": 1.1601131160786426,
      "grad_norm": 0.7278088642178367,
      "learning_rate": 3.7562022695623013e-06,
      "loss": 0.31409993171691897,
      "num_input_tokens_seen": 886505096,
      "step": 17230,
      "train_runtime": 106069.0049,
      "train_tokens_per_second": 8357.815
    },
    {
      "epoch": 1.1604497710746027,
      "grad_norm": 0.6034775307255161,
      "learning_rate": 3.7536414774389252e-06,
      "loss": 0.312406063079834,
      "num_input_tokens_seen": 886774968,
      "step": 17235,
      "train_runtime": 106100.1593,
      "train_tokens_per_second": 8357.904
    },
    {
      "epoch": 1.160786426070563,
      "grad_norm": 0.6716976688153422,
      "learning_rate": 3.7510810338557363e-06,
      "loss": 0.30361332893371584,
      "num_input_tokens_seen": 887002240,
      "step": 17240,
      "train_runtime": 106125.6102,
      "train_tokens_per_second": 8358.041
    },
    {
      "epoch": 1.161123081066523,
      "grad_norm": 0.6205892684514043,
      "learning_rate": 3.748520939528753e-06,
      "loss": 0.30615642070770266,
      "num_input_tokens_seen": 887256864,
      "step": 17245,
      "train_runtime": 106156.5412,
      "train_tokens_per_second": 8358.005
    },
    {
      "epoch": 1.1614597360624832,
      "grad_norm": 0.5952860787848324,
      "learning_rate": 3.7459611951739007e-06,
      "loss": 0.29608104228973386,
      "num_input_tokens_seen": 887526256,
      "step": 17250,
      "train_runtime": 106186.7001,
      "train_tokens_per_second": 8358.168
    },
    {
      "epoch": 1.1617963910584432,
      "grad_norm": 0.6475725641758359,
      "learning_rate": 3.7434018015069984e-06,
      "loss": 0.31608555316925047,
      "num_input_tokens_seen": 887790992,
      "step": 17255,
      "train_runtime": 106215.0848,
      "train_tokens_per_second": 8358.427
    },
    {
      "epoch": 1.1621330460544035,
      "grad_norm": 0.681384216955638,
      "learning_rate": 3.740842759243778e-06,
      "loss": 0.3070371627807617,
      "num_input_tokens_seen": 888048632,
      "step": 17260,
      "train_runtime": 106242.7411,
      "train_tokens_per_second": 8358.676
    },
    {
      "epoch": 1.1624697010503635,
      "grad_norm": 0.5906588540166086,
      "learning_rate": 3.738284069099862e-06,
      "loss": 0.2860537052154541,
      "num_input_tokens_seen": 888320784,
      "step": 17265,
      "train_runtime": 106274.2425,
      "train_tokens_per_second": 8358.759
    },
    {
      "epoch": 1.1628063560463238,
      "grad_norm": 0.8230519926104491,
      "learning_rate": 3.735725731790785e-06,
      "loss": 0.30562572479248046,
      "num_input_tokens_seen": 888585184,
      "step": 17270,
      "train_runtime": 106308.7225,
      "train_tokens_per_second": 8358.535
    },
    {
      "epoch": 1.1631430110422838,
      "grad_norm": 0.6501726067118748,
      "learning_rate": 3.733167748031974e-06,
      "loss": 0.3070846080780029,
      "num_input_tokens_seen": 888846952,
      "step": 17275,
      "train_runtime": 106341.6997,
      "train_tokens_per_second": 8358.405
    },
    {
      "epoch": 1.163479666038244,
      "grad_norm": 0.6491224600852812,
      "learning_rate": 3.7306101185387646e-06,
      "loss": 0.3076375961303711,
      "num_input_tokens_seen": 889111304,
      "step": 17280,
      "train_runtime": 106372.1556,
      "train_tokens_per_second": 8358.497
    },
    {
      "epoch": 1.163816321034204,
      "grad_norm": 0.6926289144231863,
      "learning_rate": 3.728052844026386e-06,
      "loss": 0.30945303440093996,
      "num_input_tokens_seen": 889366640,
      "step": 17285,
      "train_runtime": 106405.6989,
      "train_tokens_per_second": 8358.261
    },
    {
      "epoch": 1.1641529760301643,
      "grad_norm": 0.7010163904580071,
      "learning_rate": 3.7254959252099745e-06,
      "loss": 0.3006307125091553,
      "num_input_tokens_seen": 889607640,
      "step": 17290,
      "train_runtime": 106436.6555,
      "train_tokens_per_second": 8358.095
    },
    {
      "epoch": 1.1644896310261243,
      "grad_norm": 0.7365723648211843,
      "learning_rate": 3.7229393628045617e-06,
      "loss": 0.2995476484298706,
      "num_input_tokens_seen": 889869888,
      "step": 17295,
      "train_runtime": 106467.6263,
      "train_tokens_per_second": 8358.126
    },
    {
      "epoch": 1.1648262860220846,
      "grad_norm": 0.6615889569774944,
      "learning_rate": 3.720383157525087e-06,
      "loss": 0.2873203992843628,
      "num_input_tokens_seen": 890121824,
      "step": 17300,
      "train_runtime": 106500.2582,
      "train_tokens_per_second": 8357.931
    },
    {
      "epoch": 1.1651629410180446,
      "grad_norm": 0.6781146767516973,
      "learning_rate": 3.717827310086379e-06,
      "loss": 0.30064854621887205,
      "num_input_tokens_seen": 890375088,
      "step": 17305,
      "train_runtime": 106534.6077,
      "train_tokens_per_second": 8357.614
    },
    {
      "epoch": 1.1654995960140049,
      "grad_norm": 0.7233842376156325,
      "learning_rate": 3.7152718212031778e-06,
      "loss": 0.3064406871795654,
      "num_input_tokens_seen": 890614856,
      "step": 17310,
      "train_runtime": 106561.7045,
      "train_tokens_per_second": 8357.738
    },
    {
      "epoch": 1.165836251009965,
      "grad_norm": 0.6800131154136905,
      "learning_rate": 3.7127166915901123e-06,
      "loss": 0.3312011957168579,
      "num_input_tokens_seen": 890855304,
      "step": 17315,
      "train_runtime": 106596.4337,
      "train_tokens_per_second": 8357.271
    },
    {
      "epoch": 1.1661729060059252,
      "grad_norm": 0.6826348818357022,
      "learning_rate": 3.710161921961721e-06,
      "loss": 0.3120392084121704,
      "num_input_tokens_seen": 891113200,
      "step": 17320,
      "train_runtime": 106624.5379,
      "train_tokens_per_second": 8357.487
    },
    {
      "epoch": 1.1665095610018852,
      "grad_norm": 0.6705289686225374,
      "learning_rate": 3.707607513032434e-06,
      "loss": 0.2946956157684326,
      "num_input_tokens_seen": 891359744,
      "step": 17325,
      "train_runtime": 106657.1286,
      "train_tokens_per_second": 8357.245
    },
    {
      "epoch": 1.1668462159978454,
      "grad_norm": 0.6908991565834035,
      "learning_rate": 3.705053465516587e-06,
      "loss": 0.2907521724700928,
      "num_input_tokens_seen": 891616320,
      "step": 17330,
      "train_runtime": 106689.7675,
      "train_tokens_per_second": 8357.093
    },
    {
      "epoch": 1.1671828709938055,
      "grad_norm": 0.6728744609889177,
      "learning_rate": 3.702499780128407e-06,
      "loss": 0.2937815189361572,
      "num_input_tokens_seen": 891848136,
      "step": 17335,
      "train_runtime": 106719.4217,
      "train_tokens_per_second": 8356.943
    },
    {
      "epoch": 1.1675195259897657,
      "grad_norm": 0.6149184196978748,
      "learning_rate": 3.6999464575820266e-06,
      "loss": 0.3108397960662842,
      "num_input_tokens_seen": 892109216,
      "step": 17340,
      "train_runtime": 106749.035,
      "train_tokens_per_second": 8357.071
    },
    {
      "epoch": 1.1678561809857257,
      "grad_norm": 0.612736667282439,
      "learning_rate": 3.697393498591473e-06,
      "loss": 0.3043938636779785,
      "num_input_tokens_seen": 892364928,
      "step": 17345,
      "train_runtime": 106781.3183,
      "train_tokens_per_second": 8356.939
    },
    {
      "epoch": 1.168192835981686,
      "grad_norm": 0.6812382853567459,
      "learning_rate": 3.694840903870676e-06,
      "loss": 0.32756943702697755,
      "num_input_tokens_seen": 892625152,
      "step": 17350,
      "train_runtime": 106811.5766,
      "train_tokens_per_second": 8357.008
    },
    {
      "epoch": 1.168529490977646,
      "grad_norm": 0.6610571252665851,
      "learning_rate": 3.6922886741334564e-06,
      "loss": 0.32044053077697754,
      "num_input_tokens_seen": 892880704,
      "step": 17355,
      "train_runtime": 106845.0796,
      "train_tokens_per_second": 8356.779
    },
    {
      "epoch": 1.1688661459736063,
      "grad_norm": 0.6195905866762739,
      "learning_rate": 3.689736810093541e-06,
      "loss": 0.2980691432952881,
      "num_input_tokens_seen": 893157360,
      "step": 17360,
      "train_runtime": 106875.6839,
      "train_tokens_per_second": 8356.974
    },
    {
      "epoch": 1.1692028009695663,
      "grad_norm": 0.7392615814371957,
      "learning_rate": 3.6871853124645464e-06,
      "loss": 0.3228400230407715,
      "num_input_tokens_seen": 893420920,
      "step": 17365,
      "train_runtime": 106904.5127,
      "train_tokens_per_second": 8357.186
    },
    {
      "epoch": 1.1695394559655266,
      "grad_norm": 0.595724537340978,
      "learning_rate": 3.684634181959994e-06,
      "loss": 0.2952144622802734,
      "num_input_tokens_seen": 893707088,
      "step": 17370,
      "train_runtime": 106938.6959,
      "train_tokens_per_second": 8357.191
    },
    {
      "epoch": 1.1698761109614866,
      "grad_norm": 0.8489407422831275,
      "learning_rate": 3.6820834192932974e-06,
      "loss": 0.31411545276641845,
      "num_input_tokens_seen": 893962288,
      "step": 17375,
      "train_runtime": 106968.874,
      "train_tokens_per_second": 8357.219
    },
    {
      "epoch": 1.1702127659574468,
      "grad_norm": 0.6132194017163738,
      "learning_rate": 3.679533025177772e-06,
      "loss": 0.3329578399658203,
      "num_input_tokens_seen": 894255232,
      "step": 17380,
      "train_runtime": 107001.9142,
      "train_tokens_per_second": 8357.376
    },
    {
      "epoch": 1.1705494209534069,
      "grad_norm": 0.6980800093904247,
      "learning_rate": 3.6769830003266234e-06,
      "loss": 0.32050509452819825,
      "num_input_tokens_seen": 894524800,
      "step": 17385,
      "train_runtime": 107033.7433,
      "train_tokens_per_second": 8357.409
    },
    {
      "epoch": 1.1708860759493671,
      "grad_norm": 0.6359702261168011,
      "learning_rate": 3.6744333454529607e-06,
      "loss": 0.31168227195739745,
      "num_input_tokens_seen": 894803632,
      "step": 17390,
      "train_runtime": 107065.0093,
      "train_tokens_per_second": 8357.573
    },
    {
      "epoch": 1.1712227309453271,
      "grad_norm": 0.6457011217899132,
      "learning_rate": 3.6718840612697847e-06,
      "loss": 0.3085618495941162,
      "num_input_tokens_seen": 895078736,
      "step": 17395,
      "train_runtime": 107098.4744,
      "train_tokens_per_second": 8357.53
    },
    {
      "epoch": 1.1715593859412874,
      "grad_norm": 0.6377436529115635,
      "learning_rate": 3.6693351484899986e-06,
      "loss": 0.29672846794128416,
      "num_input_tokens_seen": 895330472,
      "step": 17400,
      "train_runtime": 107127.867,
      "train_tokens_per_second": 8357.587
    },
    {
      "epoch": 1.1718960409372474,
      "grad_norm": 0.8929972768987774,
      "learning_rate": 3.6667866078263937e-06,
      "loss": 0.3002431392669678,
      "num_input_tokens_seen": 895581016,
      "step": 17405,
      "train_runtime": 107157.2562,
      "train_tokens_per_second": 8357.633
    },
    {
      "epoch": 1.1722326959332077,
      "grad_norm": 0.6254823602812576,
      "learning_rate": 3.664238439991663e-06,
      "loss": 0.3329931259155273,
      "num_input_tokens_seen": 895844016,
      "step": 17410,
      "train_runtime": 107186.8079,
      "train_tokens_per_second": 8357.782
    },
    {
      "epoch": 1.1725693509291677,
      "grad_norm": 0.6790236352663005,
      "learning_rate": 3.6616906456983925e-06,
      "loss": 0.3145754814147949,
      "num_input_tokens_seen": 896105632,
      "step": 17415,
      "train_runtime": 107216.8351,
      "train_tokens_per_second": 8357.882
    },
    {
      "epoch": 1.172906005925128,
      "grad_norm": 0.7263934084804098,
      "learning_rate": 3.659143225659065e-06,
      "loss": 0.3272714138031006,
      "num_input_tokens_seen": 896366032,
      "step": 17420,
      "train_runtime": 107247.7689,
      "train_tokens_per_second": 8357.899
    },
    {
      "epoch": 1.173242660921088,
      "grad_norm": 0.6014501507927307,
      "learning_rate": 3.6565961805860595e-06,
      "loss": 0.3017958879470825,
      "num_input_tokens_seen": 896642728,
      "step": 17425,
      "train_runtime": 107277.226,
      "train_tokens_per_second": 8358.183
    },
    {
      "epoch": 1.1735793159170482,
      "grad_norm": 0.6332386472355778,
      "learning_rate": 3.654049511191649e-06,
      "loss": 0.30283441543579104,
      "num_input_tokens_seen": 896881816,
      "step": 17430,
      "train_runtime": 107308.3184,
      "train_tokens_per_second": 8357.99
    },
    {
      "epoch": 1.1739159709130083,
      "grad_norm": 0.7004051062741918,
      "learning_rate": 3.6515032181879995e-06,
      "loss": 0.3129396438598633,
      "num_input_tokens_seen": 897127800,
      "step": 17435,
      "train_runtime": 107339.4427,
      "train_tokens_per_second": 8357.858
    },
    {
      "epoch": 1.1742526259089685,
      "grad_norm": 0.5963933297311941,
      "learning_rate": 3.6489573022871755e-06,
      "loss": 0.2858293056488037,
      "num_input_tokens_seen": 897393048,
      "step": 17440,
      "train_runtime": 107371.2422,
      "train_tokens_per_second": 8357.853
    },
    {
      "epoch": 1.1745892809049285,
      "grad_norm": 0.5984318766861859,
      "learning_rate": 3.6464117642011333e-06,
      "loss": 0.3054333209991455,
      "num_input_tokens_seen": 897634992,
      "step": 17445,
      "train_runtime": 107401.0463,
      "train_tokens_per_second": 8357.786
    },
    {
      "epoch": 1.1749259359008888,
      "grad_norm": 0.6935803506640997,
      "learning_rate": 3.643866604641726e-06,
      "loss": 0.30799026489257814,
      "num_input_tokens_seen": 897922672,
      "step": 17450,
      "train_runtime": 107430.6458,
      "train_tokens_per_second": 8358.161
    },
    {
      "epoch": 1.1752625908968488,
      "grad_norm": 0.6216460243938668,
      "learning_rate": 3.641321824320697e-06,
      "loss": 0.2967031240463257,
      "num_input_tokens_seen": 898176712,
      "step": 17455,
      "train_runtime": 107459.9823,
      "train_tokens_per_second": 8358.244
    },
    {
      "epoch": 1.175599245892809,
      "grad_norm": 0.6095991782920825,
      "learning_rate": 3.6387774239496893e-06,
      "loss": 0.28469176292419435,
      "num_input_tokens_seen": 898461472,
      "step": 17460,
      "train_runtime": 107495.0882,
      "train_tokens_per_second": 8358.163
    },
    {
      "epoch": 1.175935900888769,
      "grad_norm": 0.6545094627667176,
      "learning_rate": 3.6362334042402326e-06,
      "loss": 0.30136406421661377,
      "num_input_tokens_seen": 898711848,
      "step": 17465,
      "train_runtime": 107524.0861,
      "train_tokens_per_second": 8358.237
    },
    {
      "epoch": 1.1762725558847293,
      "grad_norm": 0.6028345891926392,
      "learning_rate": 3.6336897659037563e-06,
      "loss": 0.28113360404968263,
      "num_input_tokens_seen": 898965800,
      "step": 17470,
      "train_runtime": 107556.5477,
      "train_tokens_per_second": 8358.076
    },
    {
      "epoch": 1.1766092108806894,
      "grad_norm": 0.6232474251773212,
      "learning_rate": 3.6311465096515787e-06,
      "loss": 0.2892298698425293,
      "num_input_tokens_seen": 899236104,
      "step": 17475,
      "train_runtime": 107589.0351,
      "train_tokens_per_second": 8358.065
    },
    {
      "epoch": 1.1769458658766496,
      "grad_norm": 0.651213191199993,
      "learning_rate": 3.6286036361949175e-06,
      "loss": 0.26990952491760256,
      "num_input_tokens_seen": 899498968,
      "step": 17480,
      "train_runtime": 107619.3013,
      "train_tokens_per_second": 8358.157
    },
    {
      "epoch": 1.1772825208726099,
      "grad_norm": 0.591868604352822,
      "learning_rate": 3.6260611462448736e-06,
      "loss": 0.256899356842041,
      "num_input_tokens_seen": 899760992,
      "step": 17485,
      "train_runtime": 107649.0166,
      "train_tokens_per_second": 8358.283
    },
    {
      "epoch": 1.17761917586857,
      "grad_norm": 0.8056078000175037,
      "learning_rate": 3.623519040512451e-06,
      "loss": 0.33391318321228025,
      "num_input_tokens_seen": 899997264,
      "step": 17490,
      "train_runtime": 107678.6015,
      "train_tokens_per_second": 8358.181
    },
    {
      "epoch": 1.17795583086453,
      "grad_norm": 0.5708324562428944,
      "learning_rate": 3.6209773197085373e-06,
      "loss": 0.27828378677368165,
      "num_input_tokens_seen": 900257504,
      "step": 17495,
      "train_runtime": 107711.5978,
      "train_tokens_per_second": 8358.037
    },
    {
      "epoch": 1.1782924858604902,
      "grad_norm": 0.6562786501366338,
      "learning_rate": 3.6184359845439214e-06,
      "loss": 0.3139003276824951,
      "num_input_tokens_seen": 900529856,
      "step": 17500,
      "train_runtime": 107746.6936,
      "train_tokens_per_second": 8357.842
    },
    {
      "epoch": 1.1786291408564504,
      "grad_norm": 0.7018848449457978,
      "learning_rate": 3.615895035729275e-06,
      "loss": 0.3316058874130249,
      "num_input_tokens_seen": 900798024,
      "step": 17505,
      "train_runtime": 107774.4502,
      "train_tokens_per_second": 8358.178
    },
    {
      "epoch": 1.1789657958524105,
      "grad_norm": 0.5933983628245367,
      "learning_rate": 3.6133544739751703e-06,
      "loss": 0.29215245246887206,
      "num_input_tokens_seen": 901046680,
      "step": 17510,
      "train_runtime": 107806.6192,
      "train_tokens_per_second": 8357.99
    },
    {
      "epoch": 1.1793024508483705,
      "grad_norm": 0.6947782982274862,
      "learning_rate": 3.6108142999920633e-06,
      "loss": 0.3105462074279785,
      "num_input_tokens_seen": 901298696,
      "step": 17515,
      "train_runtime": 107835.0723,
      "train_tokens_per_second": 8358.122
    },
    {
      "epoch": 1.1796391058443307,
      "grad_norm": 0.6294325517172377,
      "learning_rate": 3.608274514490308e-06,
      "loss": 0.3064380168914795,
      "num_input_tokens_seen": 901557080,
      "step": 17520,
      "train_runtime": 107864.5779,
      "train_tokens_per_second": 8358.231
    },
    {
      "epoch": 1.179975760840291,
      "grad_norm": 0.7284163513837525,
      "learning_rate": 3.605735118180146e-06,
      "loss": 0.2888055324554443,
      "num_input_tokens_seen": 901802280,
      "step": 17525,
      "train_runtime": 107894.6568,
      "train_tokens_per_second": 8358.174
    },
    {
      "epoch": 1.180312415836251,
      "grad_norm": 0.6377215897923488,
      "learning_rate": 3.6031961117717144e-06,
      "loss": 0.2864884853363037,
      "num_input_tokens_seen": 902042752,
      "step": 17530,
      "train_runtime": 107923.6931,
      "train_tokens_per_second": 8358.153
    },
    {
      "epoch": 1.180649070832211,
      "grad_norm": 0.5801913873209914,
      "learning_rate": 3.600657495975034e-06,
      "loss": 0.29247317314147947,
      "num_input_tokens_seen": 902316752,
      "step": 17535,
      "train_runtime": 107954.2966,
      "train_tokens_per_second": 8358.322
    },
    {
      "epoch": 1.1809857258281713,
      "grad_norm": 0.6039455854754936,
      "learning_rate": 3.598119271500024e-06,
      "loss": 0.2927104473114014,
      "num_input_tokens_seen": 902571104,
      "step": 17540,
      "train_runtime": 107985.3307,
      "train_tokens_per_second": 8358.275
    },
    {
      "epoch": 1.1813223808241315,
      "grad_norm": 0.670737427105334,
      "learning_rate": 3.5955814390564882e-06,
      "loss": 0.30347309112548826,
      "num_input_tokens_seen": 902832224,
      "step": 17545,
      "train_runtime": 108019.2053,
      "train_tokens_per_second": 8358.071
    },
    {
      "epoch": 1.1816590358200916,
      "grad_norm": 0.6691128214199659,
      "learning_rate": 3.5930439993541265e-06,
      "loss": 0.3122389316558838,
      "num_input_tokens_seen": 903078704,
      "step": 17550,
      "train_runtime": 108051.5011,
      "train_tokens_per_second": 8357.854
    },
    {
      "epoch": 1.1819956908160516,
      "grad_norm": 0.7038432150766367,
      "learning_rate": 3.590506953102522e-06,
      "loss": 0.3280792236328125,
      "num_input_tokens_seen": 903337240,
      "step": 17555,
      "train_runtime": 108083.743,
      "train_tokens_per_second": 8357.753
    },
    {
      "epoch": 1.1823323458120119,
      "grad_norm": 0.680678468611471,
      "learning_rate": 3.5879703010111554e-06,
      "loss": 0.31617274284362795,
      "num_input_tokens_seen": 903576560,
      "step": 17560,
      "train_runtime": 108115.1237,
      "train_tokens_per_second": 8357.541
    },
    {
      "epoch": 1.182669000807972,
      "grad_norm": 0.7540320693959238,
      "learning_rate": 3.5854340437893898e-06,
      "loss": 0.2980616092681885,
      "num_input_tokens_seen": 903835160,
      "step": 17565,
      "train_runtime": 108144.9969,
      "train_tokens_per_second": 8357.623
    },
    {
      "epoch": 1.1830056558039321,
      "grad_norm": 0.6278263155310905,
      "learning_rate": 3.5828981821464838e-06,
      "loss": 0.315705680847168,
      "num_input_tokens_seen": 904088752,
      "step": 17570,
      "train_runtime": 108170.4479,
      "train_tokens_per_second": 8358.001
    },
    {
      "epoch": 1.1833423107998922,
      "grad_norm": 0.8138169828028491,
      "learning_rate": 3.580362716791582e-06,
      "loss": 0.3137835502624512,
      "num_input_tokens_seen": 904381088,
      "step": 17575,
      "train_runtime": 108196.7138,
      "train_tokens_per_second": 8358.674
    },
    {
      "epoch": 1.1836789657958524,
      "grad_norm": 0.6621880005692178,
      "learning_rate": 3.5778276484337205e-06,
      "loss": 0.33135275840759276,
      "num_input_tokens_seen": 904640912,
      "step": 17580,
      "train_runtime": 108224.9634,
      "train_tokens_per_second": 8358.893
    },
    {
      "epoch": 1.1840156207918127,
      "grad_norm": 0.584836368775387,
      "learning_rate": 3.575292977781821e-06,
      "loss": 0.2994485378265381,
      "num_input_tokens_seen": 904886192,
      "step": 17585,
      "train_runtime": 108255.8995,
      "train_tokens_per_second": 8358.77
    },
    {
      "epoch": 1.1843522757877727,
      "grad_norm": 0.7076429907046901,
      "learning_rate": 3.5727587055446976e-06,
      "loss": 0.3045213222503662,
      "num_input_tokens_seen": 905161512,
      "step": 17590,
      "train_runtime": 108285.0438,
      "train_tokens_per_second": 8359.063
    },
    {
      "epoch": 1.1846889307837327,
      "grad_norm": 0.5727640907793196,
      "learning_rate": 3.570224832431051e-06,
      "loss": 0.312225341796875,
      "num_input_tokens_seen": 905441328,
      "step": 17595,
      "train_runtime": 108312.8729,
      "train_tokens_per_second": 8359.499
    },
    {
      "epoch": 1.185025585779693,
      "grad_norm": 0.7585003993055028,
      "learning_rate": 3.5676913591494715e-06,
      "loss": 0.30545864105224607,
      "num_input_tokens_seen": 905704280,
      "step": 17600,
      "train_runtime": 108340.9156,
      "train_tokens_per_second": 8359.762
    },
    {
      "epoch": 1.1853622407756532,
      "grad_norm": 0.659218519348306,
      "learning_rate": 3.5651582864084345e-06,
      "loss": 0.3256580352783203,
      "num_input_tokens_seen": 905955480,
      "step": 17605,
      "train_runtime": 108371.9062,
      "train_tokens_per_second": 8359.689
    },
    {
      "epoch": 1.1856988957716132,
      "grad_norm": 0.6632137744483709,
      "learning_rate": 3.5626256149163096e-06,
      "loss": 0.3006434917449951,
      "num_input_tokens_seen": 906213536,
      "step": 17610,
      "train_runtime": 108403.9844,
      "train_tokens_per_second": 8359.596
    },
    {
      "epoch": 1.1860355507675733,
      "grad_norm": 0.6559579351289216,
      "learning_rate": 3.5600933453813464e-06,
      "loss": 0.3051601886749268,
      "num_input_tokens_seen": 906458944,
      "step": 17615,
      "train_runtime": 108434.9928,
      "train_tokens_per_second": 8359.469
    },
    {
      "epoch": 1.1863722057635335,
      "grad_norm": 0.6494718524606543,
      "learning_rate": 3.5575614785116886e-06,
      "loss": 0.286408805847168,
      "num_input_tokens_seen": 906696928,
      "step": 17620,
      "train_runtime": 108467.2016,
      "train_tokens_per_second": 8359.181
    },
    {
      "epoch": 1.1867088607594938,
      "grad_norm": 0.7914755717214959,
      "learning_rate": 3.5550300150153618e-06,
      "loss": 0.3059598684310913,
      "num_input_tokens_seen": 906967624,
      "step": 17625,
      "train_runtime": 108499.0846,
      "train_tokens_per_second": 8359.219
    },
    {
      "epoch": 1.1870455157554538,
      "grad_norm": 0.6520222570524081,
      "learning_rate": 3.5524989556002863e-06,
      "loss": 0.3035943269729614,
      "num_input_tokens_seen": 907220448,
      "step": 17630,
      "train_runtime": 108531.4787,
      "train_tokens_per_second": 8359.054
    },
    {
      "epoch": 1.1873821707514138,
      "grad_norm": 0.6350804033489559,
      "learning_rate": 3.5499683009742604e-06,
      "loss": 0.3052224636077881,
      "num_input_tokens_seen": 907454024,
      "step": 17635,
      "train_runtime": 108560.4445,
      "train_tokens_per_second": 8358.975
    },
    {
      "epoch": 1.187718825747374,
      "grad_norm": 0.6541887140263067,
      "learning_rate": 3.5474380518449757e-06,
      "loss": 0.2898722171783447,
      "num_input_tokens_seen": 907726184,
      "step": 17640,
      "train_runtime": 108590.1642,
      "train_tokens_per_second": 8359.193
    },
    {
      "epoch": 1.1880554807433343,
      "grad_norm": 0.8552360252533019,
      "learning_rate": 3.5449082089200072e-06,
      "loss": 0.3070579290390015,
      "num_input_tokens_seen": 907982856,
      "step": 17645,
      "train_runtime": 108618.9996,
      "train_tokens_per_second": 8359.337
    },
    {
      "epoch": 1.1883921357392944,
      "grad_norm": 0.5650194210924182,
      "learning_rate": 3.5423787729068206e-06,
      "loss": 0.31574201583862305,
      "num_input_tokens_seen": 908255936,
      "step": 17650,
      "train_runtime": 108653.308,
      "train_tokens_per_second": 8359.211
    },
    {
      "epoch": 1.1887287907352544,
      "grad_norm": 0.6596526016037457,
      "learning_rate": 3.539849744512759e-06,
      "loss": 0.2935312271118164,
      "num_input_tokens_seen": 908491944,
      "step": 17655,
      "train_runtime": 108681.7848,
      "train_tokens_per_second": 8359.192
    },
    {
      "epoch": 1.1890654457312146,
      "grad_norm": 0.6692171178279086,
      "learning_rate": 3.5373211244450633e-06,
      "loss": 0.2952733516693115,
      "num_input_tokens_seen": 908755296,
      "step": 17660,
      "train_runtime": 108712.0794,
      "train_tokens_per_second": 8359.285
    },
    {
      "epoch": 1.189402100727175,
      "grad_norm": 0.6418049066271427,
      "learning_rate": 3.5347929134108483e-06,
      "loss": 0.30740468502044677,
      "num_input_tokens_seen": 909007120,
      "step": 17665,
      "train_runtime": 108739.3064,
      "train_tokens_per_second": 8359.508
    },
    {
      "epoch": 1.189738755723135,
      "grad_norm": 0.7573163730495946,
      "learning_rate": 3.532265112117125e-06,
      "loss": 0.32855868339538574,
      "num_input_tokens_seen": 909277368,
      "step": 17670,
      "train_runtime": 108768.8105,
      "train_tokens_per_second": 8359.725
    },
    {
      "epoch": 1.190075410719095,
      "grad_norm": 0.5688288230638641,
      "learning_rate": 3.5297377212707805e-06,
      "loss": 0.26004083156585694,
      "num_input_tokens_seen": 909526896,
      "step": 17675,
      "train_runtime": 108796.1096,
      "train_tokens_per_second": 8359.921
    },
    {
      "epoch": 1.1904120657150552,
      "grad_norm": 0.766137262897706,
      "learning_rate": 3.527210741578596e-06,
      "loss": 0.3046592712402344,
      "num_input_tokens_seen": 909795144,
      "step": 17680,
      "train_runtime": 108829.1568,
      "train_tokens_per_second": 8359.847
    },
    {
      "epoch": 1.1907487207110155,
      "grad_norm": 0.6619703521081188,
      "learning_rate": 3.5246841737472304e-06,
      "loss": 0.3188671588897705,
      "num_input_tokens_seen": 910031120,
      "step": 17685,
      "train_runtime": 108859.2788,
      "train_tokens_per_second": 8359.702
    },
    {
      "epoch": 1.1910853757069755,
      "grad_norm": 0.6330019078844673,
      "learning_rate": 3.522158018483231e-06,
      "loss": 0.29268274307250974,
      "num_input_tokens_seen": 910300464,
      "step": 17690,
      "train_runtime": 108888.0435,
      "train_tokens_per_second": 8359.967
    },
    {
      "epoch": 1.1914220307029355,
      "grad_norm": 0.6366796403488081,
      "learning_rate": 3.519632276493029e-06,
      "loss": 0.306062650680542,
      "num_input_tokens_seen": 910546184,
      "step": 17695,
      "train_runtime": 108917.4396,
      "train_tokens_per_second": 8359.967
    },
    {
      "epoch": 1.1917586856988958,
      "grad_norm": 0.6172210345213752,
      "learning_rate": 3.5171069484829417e-06,
      "loss": 0.3183453559875488,
      "num_input_tokens_seen": 910800488,
      "step": 17700,
      "train_runtime": 108947.798,
      "train_tokens_per_second": 8359.972
    },
    {
      "epoch": 1.192095340694856,
      "grad_norm": 0.7398563763803442,
      "learning_rate": 3.5145820351591663e-06,
      "loss": 0.29471845626831056,
      "num_input_tokens_seen": 911038040,
      "step": 17705,
      "train_runtime": 108978.4129,
      "train_tokens_per_second": 8359.803
    },
    {
      "epoch": 1.192431995690816,
      "grad_norm": 0.9310108694045112,
      "learning_rate": 3.51205753722779e-06,
      "loss": 0.31592864990234376,
      "num_input_tokens_seen": 911324720,
      "step": 17710,
      "train_runtime": 109008.5072,
      "train_tokens_per_second": 8360.125
    },
    {
      "epoch": 1.1927686506867763,
      "grad_norm": 0.7238291288615214,
      "learning_rate": 3.5095334553947757e-06,
      "loss": 0.3373671054840088,
      "num_input_tokens_seen": 911608448,
      "step": 17715,
      "train_runtime": 109043.8204,
      "train_tokens_per_second": 8360.019
    },
    {
      "epoch": 1.1931053056827363,
      "grad_norm": 0.5907626052366977,
      "learning_rate": 3.50700979036598e-06,
      "loss": 0.29745752811431886,
      "num_input_tokens_seen": 911861624,
      "step": 17720,
      "train_runtime": 109071.2788,
      "train_tokens_per_second": 8360.236
    },
    {
      "epoch": 1.1934419606786966,
      "grad_norm": 0.6521100780922855,
      "learning_rate": 3.5044865428471343e-06,
      "loss": 0.2840247631072998,
      "num_input_tokens_seen": 912105928,
      "step": 17725,
      "train_runtime": 109099.4595,
      "train_tokens_per_second": 8360.316
    },
    {
      "epoch": 1.1937786156746566,
      "grad_norm": 0.6837134010009283,
      "learning_rate": 3.50196371354386e-06,
      "loss": 0.3068502187728882,
      "num_input_tokens_seen": 912362696,
      "step": 17730,
      "train_runtime": 109131.6198,
      "train_tokens_per_second": 8360.205
    },
    {
      "epoch": 1.1941152706706168,
      "grad_norm": 0.6742536093847684,
      "learning_rate": 3.499441303161655e-06,
      "loss": 0.3386415481567383,
      "num_input_tokens_seen": 912640320,
      "step": 17735,
      "train_runtime": 109162.3869,
      "train_tokens_per_second": 8360.392
    },
    {
      "epoch": 1.1944519256665769,
      "grad_norm": 0.6478513627543618,
      "learning_rate": 3.4969193124059042e-06,
      "loss": 0.305743932723999,
      "num_input_tokens_seen": 912917632,
      "step": 17740,
      "train_runtime": 109196.0981,
      "train_tokens_per_second": 8360.35
    },
    {
      "epoch": 1.1947885806625371,
      "grad_norm": 0.5600754424255185,
      "learning_rate": 3.4943977419818748e-06,
      "loss": 0.2972008228302002,
      "num_input_tokens_seen": 913202696,
      "step": 17745,
      "train_runtime": 109224.2497,
      "train_tokens_per_second": 8360.805
    },
    {
      "epoch": 1.1951252356584972,
      "grad_norm": 0.7358925333253512,
      "learning_rate": 3.4918765925947173e-06,
      "loss": 0.29028513431549074,
      "num_input_tokens_seen": 913464576,
      "step": 17750,
      "train_runtime": 109256.4979,
      "train_tokens_per_second": 8360.735
    },
    {
      "epoch": 1.1954618906544574,
      "grad_norm": 0.7223458362732181,
      "learning_rate": 3.4893558649494595e-06,
      "loss": 0.32371978759765624,
      "num_input_tokens_seen": 913734064,
      "step": 17755,
      "train_runtime": 109291.5325,
      "train_tokens_per_second": 8360.52
    },
    {
      "epoch": 1.1957985456504174,
      "grad_norm": 0.6113124115290576,
      "learning_rate": 3.4868355597510194e-06,
      "loss": 0.2947253704071045,
      "num_input_tokens_seen": 913981552,
      "step": 17760,
      "train_runtime": 109322.9503,
      "train_tokens_per_second": 8360.381
    },
    {
      "epoch": 1.1961352006463777,
      "grad_norm": 0.7106784691055003,
      "learning_rate": 3.4843156777041887e-06,
      "loss": 0.2838107109069824,
      "num_input_tokens_seen": 914245864,
      "step": 17765,
      "train_runtime": 109356.0738,
      "train_tokens_per_second": 8360.266
    },
    {
      "epoch": 1.1964718556423377,
      "grad_norm": 0.7198384528886875,
      "learning_rate": 3.4817962195136457e-06,
      "loss": 0.30479955673217773,
      "num_input_tokens_seen": 914502944,
      "step": 17770,
      "train_runtime": 109381.8717,
      "train_tokens_per_second": 8360.644
    },
    {
      "epoch": 1.196808510638298,
      "grad_norm": 0.7120328741180209,
      "learning_rate": 3.479277185883947e-06,
      "loss": 0.30356736183166505,
      "num_input_tokens_seen": 914757168,
      "step": 17775,
      "train_runtime": 109412.7148,
      "train_tokens_per_second": 8360.611
    },
    {
      "epoch": 1.197145165634258,
      "grad_norm": 0.6494303322317979,
      "learning_rate": 3.476758577519537e-06,
      "loss": 0.289382791519165,
      "num_input_tokens_seen": 915026536,
      "step": 17780,
      "train_runtime": 109446.1826,
      "train_tokens_per_second": 8360.516
    },
    {
      "epoch": 1.1974818206302182,
      "grad_norm": 0.7209427901269357,
      "learning_rate": 3.4742403951247318e-06,
      "loss": 0.3056520938873291,
      "num_input_tokens_seen": 915290232,
      "step": 17785,
      "train_runtime": 109472.6045,
      "train_tokens_per_second": 8360.907
    },
    {
      "epoch": 1.1978184756261783,
      "grad_norm": 0.6484945753672452,
      "learning_rate": 3.471722639403735e-06,
      "loss": 0.2945056200027466,
      "num_input_tokens_seen": 915548264,
      "step": 17790,
      "train_runtime": 109500.4284,
      "train_tokens_per_second": 8361.139
    },
    {
      "epoch": 1.1981551306221385,
      "grad_norm": 0.618348197447593,
      "learning_rate": 3.4692053110606277e-06,
      "loss": 0.296243953704834,
      "num_input_tokens_seen": 915768736,
      "step": 17795,
      "train_runtime": 109531.1463,
      "train_tokens_per_second": 8360.807
    },
    {
      "epoch": 1.1984917856180985,
      "grad_norm": 0.6192461031846919,
      "learning_rate": 3.4666884107993766e-06,
      "loss": 0.29240467548370364,
      "num_input_tokens_seen": 916008312,
      "step": 17800,
      "train_runtime": 109561.7087,
      "train_tokens_per_second": 8360.661
    },
    {
      "epoch": 1.1988284406140588,
      "grad_norm": 0.6796452257178552,
      "learning_rate": 3.464171939323819e-06,
      "loss": 0.29559690952301027,
      "num_input_tokens_seen": 916244952,
      "step": 17805,
      "train_runtime": 109592.3767,
      "train_tokens_per_second": 8360.481
    },
    {
      "epoch": 1.1991650956100188,
      "grad_norm": 0.6435753332184259,
      "learning_rate": 3.4616558973376848e-06,
      "loss": 0.30375447273254397,
      "num_input_tokens_seen": 916501640,
      "step": 17810,
      "train_runtime": 109617.8323,
      "train_tokens_per_second": 8360.881
    },
    {
      "epoch": 1.199501750605979,
      "grad_norm": 0.686545210275044,
      "learning_rate": 3.4591402855445697e-06,
      "loss": 0.30944070816040037,
      "num_input_tokens_seen": 916766864,
      "step": 17815,
      "train_runtime": 109650.3923,
      "train_tokens_per_second": 8360.817
    },
    {
      "epoch": 1.199838405601939,
      "grad_norm": 0.6302817617686535,
      "learning_rate": 3.4566251046479625e-06,
      "loss": 0.3125229120254517,
      "num_input_tokens_seen": 917017296,
      "step": 17820,
      "train_runtime": 109680.9544,
      "train_tokens_per_second": 8360.771
    },
    {
      "epoch": 1.2001750605978994,
      "grad_norm": 0.7632478734992566,
      "learning_rate": 3.454110355351221e-06,
      "loss": 0.32356953620910645,
      "num_input_tokens_seen": 917279752,
      "step": 17825,
      "train_runtime": 109712.0151,
      "train_tokens_per_second": 8360.796
    },
    {
      "epoch": 1.2005117155938594,
      "grad_norm": 0.6233719843505926,
      "learning_rate": 3.451596038357593e-06,
      "loss": 0.30062987804412844,
      "num_input_tokens_seen": 917560016,
      "step": 17830,
      "train_runtime": 109742.5452,
      "train_tokens_per_second": 8361.024
    },
    {
      "epoch": 1.2008483705898196,
      "grad_norm": 0.7332736073312713,
      "learning_rate": 3.4490821543701924e-06,
      "loss": 0.33019819259643557,
      "num_input_tokens_seen": 917823424,
      "step": 17835,
      "train_runtime": 109770.9723,
      "train_tokens_per_second": 8361.258
    },
    {
      "epoch": 1.2011850255857797,
      "grad_norm": 0.6339193969886071,
      "learning_rate": 3.4465687040920227e-06,
      "loss": 0.30566437244415284,
      "num_input_tokens_seen": 918089328,
      "step": 17840,
      "train_runtime": 109803.5451,
      "train_tokens_per_second": 8361.199
    },
    {
      "epoch": 1.20152168058174,
      "grad_norm": 0.6335696248494531,
      "learning_rate": 3.44405568822596e-06,
      "loss": 0.2890420436859131,
      "num_input_tokens_seen": 918339664,
      "step": 17845,
      "train_runtime": 109837.1011,
      "train_tokens_per_second": 8360.924
    },
    {
      "epoch": 1.2018583355777,
      "grad_norm": 0.7717956655839819,
      "learning_rate": 3.441543107474765e-06,
      "loss": 0.29753594398498534,
      "num_input_tokens_seen": 918582600,
      "step": 17850,
      "train_runtime": 109868.2555,
      "train_tokens_per_second": 8360.764
    },
    {
      "epoch": 1.2021949905736602,
      "grad_norm": 0.6066020364451304,
      "learning_rate": 3.439030962541069e-06,
      "loss": 0.2866447925567627,
      "num_input_tokens_seen": 918835496,
      "step": 17855,
      "train_runtime": 109896.2493,
      "train_tokens_per_second": 8360.936
    },
    {
      "epoch": 1.2025316455696202,
      "grad_norm": 0.633256533396687,
      "learning_rate": 3.4365192541273885e-06,
      "loss": 0.29886136054992674,
      "num_input_tokens_seen": 919067400,
      "step": 17860,
      "train_runtime": 109925.4256,
      "train_tokens_per_second": 8360.826
    },
    {
      "epoch": 1.2028683005655805,
      "grad_norm": 0.6486377023624809,
      "learning_rate": 3.4340079829361108e-06,
      "loss": 0.31514968872070315,
      "num_input_tokens_seen": 919303168,
      "step": 17865,
      "train_runtime": 109956.7035,
      "train_tokens_per_second": 8360.592
    },
    {
      "epoch": 1.2032049555615405,
      "grad_norm": 0.5634125551447571,
      "learning_rate": 3.4314971496695078e-06,
      "loss": 0.3128427743911743,
      "num_input_tokens_seen": 919580952,
      "step": 17870,
      "train_runtime": 109992.2053,
      "train_tokens_per_second": 8360.419
    },
    {
      "epoch": 1.2035416105575008,
      "grad_norm": 0.6350600896341285,
      "learning_rate": 3.428986755029724e-06,
      "loss": 0.2886244297027588,
      "num_input_tokens_seen": 919841040,
      "step": 17875,
      "train_runtime": 110026.7687,
      "train_tokens_per_second": 8360.157
    },
    {
      "epoch": 1.2038782655534608,
      "grad_norm": 0.61193729042251,
      "learning_rate": 3.4264767997187874e-06,
      "loss": 0.2990025758743286,
      "num_input_tokens_seen": 920129176,
      "step": 17880,
      "train_runtime": 110055.8025,
      "train_tokens_per_second": 8360.569
    },
    {
      "epoch": 1.204214920549421,
      "grad_norm": 0.6672345499525344,
      "learning_rate": 3.423967284438594e-06,
      "loss": 0.31085565090179446,
      "num_input_tokens_seen": 920388984,
      "step": 17885,
      "train_runtime": 110088.725,
      "train_tokens_per_second": 8360.429
    },
    {
      "epoch": 1.204551575545381,
      "grad_norm": 0.5547310428136645,
      "learning_rate": 3.421458209890925e-06,
      "loss": 0.29582488536834717,
      "num_input_tokens_seen": 920671600,
      "step": 17890,
      "train_runtime": 110121.4693,
      "train_tokens_per_second": 8360.51
    },
    {
      "epoch": 1.2048882305413413,
      "grad_norm": 0.601579720525666,
      "learning_rate": 3.418949576777433e-06,
      "loss": 0.2837364912033081,
      "num_input_tokens_seen": 920951568,
      "step": 17895,
      "train_runtime": 110153.5446,
      "train_tokens_per_second": 8360.617
    },
    {
      "epoch": 1.2052248855373013,
      "grad_norm": 0.7768773766061078,
      "learning_rate": 3.4164413857996513e-06,
      "loss": 0.2914083003997803,
      "num_input_tokens_seen": 921198984,
      "step": 17900,
      "train_runtime": 110184.6734,
      "train_tokens_per_second": 8360.5
    },
    {
      "epoch": 1.2055615405332616,
      "grad_norm": 0.5828659297323998,
      "learning_rate": 3.4139336376589853e-06,
      "loss": 0.28088438510894775,
      "num_input_tokens_seen": 921472400,
      "step": 17905,
      "train_runtime": 110215.2214,
      "train_tokens_per_second": 8360.664
    },
    {
      "epoch": 1.2058981955292216,
      "grad_norm": 0.8504071656716345,
      "learning_rate": 3.411426333056722e-06,
      "loss": 0.30802462100982664,
      "num_input_tokens_seen": 921707688,
      "step": 17910,
      "train_runtime": 110247.9499,
      "train_tokens_per_second": 8360.316
    },
    {
      "epoch": 1.2062348505251819,
      "grad_norm": 0.6850431944082769,
      "learning_rate": 3.408919472694017e-06,
      "loss": 0.3306396961212158,
      "num_input_tokens_seen": 921958928,
      "step": 17915,
      "train_runtime": 110275.4713,
      "train_tokens_per_second": 8360.508
    },
    {
      "epoch": 1.206571505521142,
      "grad_norm": 0.6003141008463565,
      "learning_rate": 3.406413057271909e-06,
      "loss": 0.2970249176025391,
      "num_input_tokens_seen": 922215616,
      "step": 17920,
      "train_runtime": 110311.1259,
      "train_tokens_per_second": 8360.132
    },
    {
      "epoch": 1.2069081605171021,
      "grad_norm": 0.6923337698503513,
      "learning_rate": 3.403907087491306e-06,
      "loss": 0.29839510917663575,
      "num_input_tokens_seen": 922478760,
      "step": 17925,
      "train_runtime": 110340.851,
      "train_tokens_per_second": 8360.265
    },
    {
      "epoch": 1.2072448155130622,
      "grad_norm": 0.6621220284900677,
      "learning_rate": 3.4014015640529996e-06,
      "loss": 0.3103820323944092,
      "num_input_tokens_seen": 922736344,
      "step": 17930,
      "train_runtime": 110374.796,
      "train_tokens_per_second": 8360.028
    },
    {
      "epoch": 1.2075814705090224,
      "grad_norm": 0.6680035074334569,
      "learning_rate": 3.398896487657646e-06,
      "loss": 0.31158339977264404,
      "num_input_tokens_seen": 922994744,
      "step": 17935,
      "train_runtime": 110405.3783,
      "train_tokens_per_second": 8360.052
    },
    {
      "epoch": 1.2079181255049825,
      "grad_norm": 0.69645497304019,
      "learning_rate": 3.396391859005785e-06,
      "loss": 0.29932029247283937,
      "num_input_tokens_seen": 923266328,
      "step": 17940,
      "train_runtime": 110440.2574,
      "train_tokens_per_second": 8359.871
    },
    {
      "epoch": 1.2082547805009427,
      "grad_norm": 0.6180974662201534,
      "learning_rate": 3.393887678797826e-06,
      "loss": 0.3020059823989868,
      "num_input_tokens_seen": 923498184,
      "step": 17945,
      "train_runtime": 110467.3294,
      "train_tokens_per_second": 8359.921
    },
    {
      "epoch": 1.2085914354969027,
      "grad_norm": 0.7078681795579722,
      "learning_rate": 3.3913839477340575e-06,
      "loss": 0.2952367067337036,
      "num_input_tokens_seen": 923746336,
      "step": 17950,
      "train_runtime": 110495.5115,
      "train_tokens_per_second": 8360.035
    },
    {
      "epoch": 1.208928090492863,
      "grad_norm": 0.7264876015861517,
      "learning_rate": 3.388880666514637e-06,
      "loss": 0.28288543224334717,
      "num_input_tokens_seen": 924014600,
      "step": 17955,
      "train_runtime": 110527.0975,
      "train_tokens_per_second": 8360.073
    },
    {
      "epoch": 1.209264745488823,
      "grad_norm": 0.5910437102381817,
      "learning_rate": 3.3863778358396033e-06,
      "loss": 0.31498055458068847,
      "num_input_tokens_seen": 924264192,
      "step": 17960,
      "train_runtime": 110560.9838,
      "train_tokens_per_second": 8359.768
    },
    {
      "epoch": 1.2096014004847833,
      "grad_norm": 0.605986580230836,
      "learning_rate": 3.3838754564088592e-06,
      "loss": 0.2993941307067871,
      "num_input_tokens_seen": 924528848,
      "step": 17965,
      "train_runtime": 110596.5907,
      "train_tokens_per_second": 8359.47
    },
    {
      "epoch": 1.2099380554807433,
      "grad_norm": 0.5917787096247646,
      "learning_rate": 3.3813735289221923e-06,
      "loss": 0.2809807300567627,
      "num_input_tokens_seen": 924790352,
      "step": 17970,
      "train_runtime": 110627.3164,
      "train_tokens_per_second": 8359.512
    },
    {
      "epoch": 1.2102747104767035,
      "grad_norm": 0.5558368733061633,
      "learning_rate": 3.378872054079255e-06,
      "loss": 0.29021282196044923,
      "num_input_tokens_seen": 925051192,
      "step": 17975,
      "train_runtime": 110660.7586,
      "train_tokens_per_second": 8359.343
    },
    {
      "epoch": 1.2106113654726636,
      "grad_norm": 0.705830523866984,
      "learning_rate": 3.3763710325795784e-06,
      "loss": 0.3304572343826294,
      "num_input_tokens_seen": 925290240,
      "step": 17980,
      "train_runtime": 110691.4091,
      "train_tokens_per_second": 8359.187
    },
    {
      "epoch": 1.2109480204686238,
      "grad_norm": 0.6078447367689197,
      "learning_rate": 3.3738704651225673e-06,
      "loss": 0.3038602828979492,
      "num_input_tokens_seen": 925567440,
      "step": 17985,
      "train_runtime": 110722.4791,
      "train_tokens_per_second": 8359.345
    },
    {
      "epoch": 1.2112846754645838,
      "grad_norm": 0.6646182291991405,
      "learning_rate": 3.3713703524074936e-06,
      "loss": 0.2717027187347412,
      "num_input_tokens_seen": 925810800,
      "step": 17990,
      "train_runtime": 110750.7459,
      "train_tokens_per_second": 8359.409
    },
    {
      "epoch": 1.211621330460544,
      "grad_norm": 0.6832674664355688,
      "learning_rate": 3.3688706951335082e-06,
      "loss": 0.30864458084106444,
      "num_input_tokens_seen": 926046000,
      "step": 17995,
      "train_runtime": 110783.472,
      "train_tokens_per_second": 8359.063
    },
    {
      "epoch": 1.2119579854565041,
      "grad_norm": 0.6003448315272588,
      "learning_rate": 3.3663714939996305e-06,
      "loss": 0.3056600093841553,
      "num_input_tokens_seen": 926339968,
      "step": 18000,
      "train_runtime": 110818.1281,
      "train_tokens_per_second": 8359.101
    },
    {
      "epoch": 1.2122946404524644,
      "grad_norm": 0.676504770848526,
      "learning_rate": 3.3638727497047575e-06,
      "loss": 0.29108219146728515,
      "num_input_tokens_seen": 926578776,
      "step": 18005,
      "train_runtime": 110848.0384,
      "train_tokens_per_second": 8359.0
    },
    {
      "epoch": 1.2126312954484244,
      "grad_norm": 0.6673069624956047,
      "learning_rate": 3.3613744629476507e-06,
      "loss": 0.29642107486724856,
      "num_input_tokens_seen": 926822592,
      "step": 18010,
      "train_runtime": 110879.4015,
      "train_tokens_per_second": 8358.835
    },
    {
      "epoch": 1.2129679504443847,
      "grad_norm": 0.6876527240529698,
      "learning_rate": 3.358876634426952e-06,
      "loss": 0.2782243490219116,
      "num_input_tokens_seen": 927096128,
      "step": 18015,
      "train_runtime": 110914.997,
      "train_tokens_per_second": 8358.618
    },
    {
      "epoch": 1.2133046054403447,
      "grad_norm": 0.6229740090159124,
      "learning_rate": 3.3563792648411676e-06,
      "loss": 0.313086724281311,
      "num_input_tokens_seen": 927332480,
      "step": 18020,
      "train_runtime": 110945.5134,
      "train_tokens_per_second": 8358.45
    },
    {
      "epoch": 1.213641260436305,
      "grad_norm": 0.6204490139079647,
      "learning_rate": 3.353882354888681e-06,
      "loss": 0.2888501167297363,
      "num_input_tokens_seen": 927601416,
      "step": 18025,
      "train_runtime": 110977.139,
      "train_tokens_per_second": 8358.491
    },
    {
      "epoch": 1.213977915432265,
      "grad_norm": 0.6372245524079381,
      "learning_rate": 3.3513859052677444e-06,
      "loss": 0.2762479305267334,
      "num_input_tokens_seen": 927835872,
      "step": 18030,
      "train_runtime": 111003.1077,
      "train_tokens_per_second": 8358.648
    },
    {
      "epoch": 1.2143145704282252,
      "grad_norm": 0.7223668754427582,
      "learning_rate": 3.3488899166764842e-06,
      "loss": 0.30127754211425783,
      "num_input_tokens_seen": 928111104,
      "step": 18035,
      "train_runtime": 111031.5668,
      "train_tokens_per_second": 8358.984
    },
    {
      "epoch": 1.2146512254241852,
      "grad_norm": 0.6888100387517707,
      "learning_rate": 3.346394389812892e-06,
      "loss": 0.2941565990447998,
      "num_input_tokens_seen": 928377480,
      "step": 18040,
      "train_runtime": 111064.7131,
      "train_tokens_per_second": 8358.888
    },
    {
      "epoch": 1.2149878804201455,
      "grad_norm": 0.5957457514591588,
      "learning_rate": 3.3438993253748365e-06,
      "loss": 0.2881103754043579,
      "num_input_tokens_seen": 928614152,
      "step": 18045,
      "train_runtime": 111092.559,
      "train_tokens_per_second": 8358.923
    },
    {
      "epoch": 1.2153245354161055,
      "grad_norm": 0.6520122751535721,
      "learning_rate": 3.3414047240600526e-06,
      "loss": 0.2996713638305664,
      "num_input_tokens_seen": 928866288,
      "step": 18050,
      "train_runtime": 111122.4018,
      "train_tokens_per_second": 8358.947
    },
    {
      "epoch": 1.2156611904120658,
      "grad_norm": 0.7206978147535007,
      "learning_rate": 3.338910586566151e-06,
      "loss": 0.2935003757476807,
      "num_input_tokens_seen": 929127080,
      "step": 18055,
      "train_runtime": 111149.2273,
      "train_tokens_per_second": 8359.276
    },
    {
      "epoch": 1.2159978454080258,
      "grad_norm": 0.6455111720923492,
      "learning_rate": 3.336416913590605e-06,
      "loss": 0.2972100257873535,
      "num_input_tokens_seen": 929366616,
      "step": 18060,
      "train_runtime": 111181.4138,
      "train_tokens_per_second": 8359.011
    },
    {
      "epoch": 1.216334500403986,
      "grad_norm": 0.6445851775421455,
      "learning_rate": 3.333923705830766e-06,
      "loss": 0.3194100856781006,
      "num_input_tokens_seen": 929630904,
      "step": 18065,
      "train_runtime": 111213.7913,
      "train_tokens_per_second": 8358.953
    },
    {
      "epoch": 1.216671155399946,
      "grad_norm": 0.6513221848565514,
      "learning_rate": 3.3314309639838484e-06,
      "loss": 0.3287325382232666,
      "num_input_tokens_seen": 929899232,
      "step": 18070,
      "train_runtime": 111243.9699,
      "train_tokens_per_second": 8359.098
    },
    {
      "epoch": 1.2170078103959063,
      "grad_norm": 0.6432842986922573,
      "learning_rate": 3.328938688746942e-06,
      "loss": 0.2986437797546387,
      "num_input_tokens_seen": 930169784,
      "step": 18075,
      "train_runtime": 111274.6557,
      "train_tokens_per_second": 8359.224
    },
    {
      "epoch": 1.2173444653918664,
      "grad_norm": 0.6581643893305507,
      "learning_rate": 3.3264468808170012e-06,
      "loss": 0.34569289684295657,
      "num_input_tokens_seen": 930445640,
      "step": 18080,
      "train_runtime": 111304.9138,
      "train_tokens_per_second": 8359.43
    },
    {
      "epoch": 1.2176811203878266,
      "grad_norm": 0.6585789961059643,
      "learning_rate": 3.3239555408908562e-06,
      "loss": 0.32164907455444336,
      "num_input_tokens_seen": 930662152,
      "step": 18085,
      "train_runtime": 111332.8938,
      "train_tokens_per_second": 8359.274
    },
    {
      "epoch": 1.2180177753837866,
      "grad_norm": 0.6638295396499506,
      "learning_rate": 3.3214646696651974e-06,
      "loss": 0.29711949825286865,
      "num_input_tokens_seen": 930902824,
      "step": 18090,
      "train_runtime": 111364.9453,
      "train_tokens_per_second": 8359.029
    },
    {
      "epoch": 1.2183544303797469,
      "grad_norm": 0.631896146634691,
      "learning_rate": 3.3189742678365923e-06,
      "loss": 0.31453118324279783,
      "num_input_tokens_seen": 931174856,
      "step": 18095,
      "train_runtime": 111394.6802,
      "train_tokens_per_second": 8359.24
    },
    {
      "epoch": 1.218691085375707,
      "grad_norm": 0.646876490649323,
      "learning_rate": 3.316484336101472e-06,
      "loss": 0.2915815353393555,
      "num_input_tokens_seen": 931413416,
      "step": 18100,
      "train_runtime": 111426.5714,
      "train_tokens_per_second": 8358.988
    },
    {
      "epoch": 1.2190277403716672,
      "grad_norm": 0.7439975417210088,
      "learning_rate": 3.313994875156141e-06,
      "loss": 0.32032041549682616,
      "num_input_tokens_seen": 931666584,
      "step": 18105,
      "train_runtime": 111460.7903,
      "train_tokens_per_second": 8358.694
    },
    {
      "epoch": 1.2193643953676272,
      "grad_norm": 0.6490013009599328,
      "learning_rate": 3.3115058856967643e-06,
      "loss": 0.3188755035400391,
      "num_input_tokens_seen": 931933112,
      "step": 18110,
      "train_runtime": 111492.678,
      "train_tokens_per_second": 8358.693
    },
    {
      "epoch": 1.2197010503635874,
      "grad_norm": 0.7574255415929944,
      "learning_rate": 3.3090173684193845e-06,
      "loss": 0.27324326038360597,
      "num_input_tokens_seen": 932179680,
      "step": 18115,
      "train_runtime": 111520.5508,
      "train_tokens_per_second": 8358.815
    },
    {
      "epoch": 1.2200377053595475,
      "grad_norm": 0.6197830946521603,
      "learning_rate": 3.3065293240199037e-06,
      "loss": 0.3170691967010498,
      "num_input_tokens_seen": 932439744,
      "step": 18120,
      "train_runtime": 111549.8357,
      "train_tokens_per_second": 8358.952
    },
    {
      "epoch": 1.2203743603555077,
      "grad_norm": 0.7615325899585683,
      "learning_rate": 3.3040417531940982e-06,
      "loss": 0.32556819915771484,
      "num_input_tokens_seen": 932702040,
      "step": 18125,
      "train_runtime": 111577.0644,
      "train_tokens_per_second": 8359.263
    },
    {
      "epoch": 1.2207110153514678,
      "grad_norm": 0.6209966879764605,
      "learning_rate": 3.301554656637607e-06,
      "loss": 0.2885535478591919,
      "num_input_tokens_seen": 932956488,
      "step": 18130,
      "train_runtime": 111605.2733,
      "train_tokens_per_second": 8359.43
    },
    {
      "epoch": 1.221047670347428,
      "grad_norm": 0.5980356628601349,
      "learning_rate": 3.2990680350459427e-06,
      "loss": 0.3009043216705322,
      "num_input_tokens_seen": 933196392,
      "step": 18135,
      "train_runtime": 111639.1543,
      "train_tokens_per_second": 8359.042
    },
    {
      "epoch": 1.221384325343388,
      "grad_norm": 0.659953648782314,
      "learning_rate": 3.2965818891144763e-06,
      "loss": 0.3159477710723877,
      "num_input_tokens_seen": 933462504,
      "step": 18140,
      "train_runtime": 111667.462,
      "train_tokens_per_second": 8359.306
    },
    {
      "epoch": 1.2217209803393483,
      "grad_norm": 0.6524580883363426,
      "learning_rate": 3.294096219538454e-06,
      "loss": 0.3084556102752686,
      "num_input_tokens_seen": 933740232,
      "step": 18145,
      "train_runtime": 111697.1878,
      "train_tokens_per_second": 8359.568
    },
    {
      "epoch": 1.2220576353353083,
      "grad_norm": 0.6754214655826412,
      "learning_rate": 3.291611027012983e-06,
      "loss": 0.3128241539001465,
      "num_input_tokens_seen": 934001864,
      "step": 18150,
      "train_runtime": 111725.5625,
      "train_tokens_per_second": 8359.787
    },
    {
      "epoch": 1.2223942903312686,
      "grad_norm": 0.7402317379291143,
      "learning_rate": 3.289126312233043e-06,
      "loss": 0.294179630279541,
      "num_input_tokens_seen": 934274432,
      "step": 18155,
      "train_runtime": 111758.7218,
      "train_tokens_per_second": 8359.745
    },
    {
      "epoch": 1.2227309453272286,
      "grad_norm": 0.6164750997927926,
      "learning_rate": 3.2866420758934714e-06,
      "loss": 0.30426487922668455,
      "num_input_tokens_seen": 934543344,
      "step": 18160,
      "train_runtime": 111795.013,
      "train_tokens_per_second": 8359.437
    },
    {
      "epoch": 1.2230676003231888,
      "grad_norm": 0.6003661910982957,
      "learning_rate": 3.2841583186889826e-06,
      "loss": 0.3126729726791382,
      "num_input_tokens_seen": 934824224,
      "step": 18165,
      "train_runtime": 111824.0486,
      "train_tokens_per_second": 8359.778
    },
    {
      "epoch": 1.2234042553191489,
      "grad_norm": 0.7004135821137168,
      "learning_rate": 3.281675041314146e-06,
      "loss": 0.2962830066680908,
      "num_input_tokens_seen": 935071336,
      "step": 18170,
      "train_runtime": 111852.5926,
      "train_tokens_per_second": 8359.854
    },
    {
      "epoch": 1.2237409103151091,
      "grad_norm": 0.6335219131229692,
      "learning_rate": 3.279192244463406e-06,
      "loss": 0.2944872140884399,
      "num_input_tokens_seen": 935315832,
      "step": 18175,
      "train_runtime": 111881.8472,
      "train_tokens_per_second": 8359.853
    },
    {
      "epoch": 1.2240775653110691,
      "grad_norm": 0.6490675622133055,
      "learning_rate": 3.2767099288310667e-06,
      "loss": 0.3050816059112549,
      "num_input_tokens_seen": 935574264,
      "step": 18180,
      "train_runtime": 111910.7138,
      "train_tokens_per_second": 8360.006
    },
    {
      "epoch": 1.2244142203070294,
      "grad_norm": 0.584894779860017,
      "learning_rate": 3.2742280951113025e-06,
      "loss": 0.2938250541687012,
      "num_input_tokens_seen": 935854368,
      "step": 18185,
      "train_runtime": 111939.2493,
      "train_tokens_per_second": 8360.377
    },
    {
      "epoch": 1.2247508753029894,
      "grad_norm": 0.6777156094920629,
      "learning_rate": 3.2717467439981464e-06,
      "loss": 0.31197314262390136,
      "num_input_tokens_seen": 936098840,
      "step": 18190,
      "train_runtime": 111969.8078,
      "train_tokens_per_second": 8360.279
    },
    {
      "epoch": 1.2250875302989497,
      "grad_norm": 0.7208839146516077,
      "learning_rate": 3.2692658761855033e-06,
      "loss": 0.271687650680542,
      "num_input_tokens_seen": 936333936,
      "step": 18195,
      "train_runtime": 111999.7932,
      "train_tokens_per_second": 8360.14
    },
    {
      "epoch": 1.2254241852949097,
      "grad_norm": 0.6345899286920527,
      "learning_rate": 3.266785492367138e-06,
      "loss": 0.2827309608459473,
      "num_input_tokens_seen": 936600488,
      "step": 18200,
      "train_runtime": 112030.291,
      "train_tokens_per_second": 8360.243
    },
    {
      "epoch": 1.22576084029087,
      "grad_norm": 0.6474482048342068,
      "learning_rate": 3.2643055932366852e-06,
      "loss": 0.2858921527862549,
      "num_input_tokens_seen": 936844112,
      "step": 18205,
      "train_runtime": 112060.0828,
      "train_tokens_per_second": 8360.195
    },
    {
      "epoch": 1.22609749528683,
      "grad_norm": 0.6265471215123033,
      "learning_rate": 3.2618261794876365e-06,
      "loss": 0.31128382682800293,
      "num_input_tokens_seen": 937127464,
      "step": 18210,
      "train_runtime": 112091.4711,
      "train_tokens_per_second": 8360.382
    },
    {
      "epoch": 1.2264341502827902,
      "grad_norm": 0.6507452891501014,
      "learning_rate": 3.259347251813356e-06,
      "loss": 0.29832887649536133,
      "num_input_tokens_seen": 937387848,
      "step": 18215,
      "train_runtime": 112121.1918,
      "train_tokens_per_second": 8360.488
    },
    {
      "epoch": 1.2267708052787503,
      "grad_norm": 0.6432466072376616,
      "learning_rate": 3.256868810907064e-06,
      "loss": 0.3160618782043457,
      "num_input_tokens_seen": 937663752,
      "step": 18220,
      "train_runtime": 112156.5976,
      "train_tokens_per_second": 8360.308
    },
    {
      "epoch": 1.2271074602747105,
      "grad_norm": 0.6535600952497245,
      "learning_rate": 3.2543908574618514e-06,
      "loss": 0.32107720375061033,
      "num_input_tokens_seen": 937917984,
      "step": 18225,
      "train_runtime": 112187.3767,
      "train_tokens_per_second": 8360.281
    },
    {
      "epoch": 1.2274441152706705,
      "grad_norm": 0.5872029148142969,
      "learning_rate": 3.251913392170668e-06,
      "loss": 0.2937935829162598,
      "num_input_tokens_seen": 938161472,
      "step": 18230,
      "train_runtime": 112215.48,
      "train_tokens_per_second": 8360.357
    },
    {
      "epoch": 1.2277807702666308,
      "grad_norm": 0.6121316694618761,
      "learning_rate": 3.249436415726333e-06,
      "loss": 0.29731974601745603,
      "num_input_tokens_seen": 938434216,
      "step": 18235,
      "train_runtime": 112247.2909,
      "train_tokens_per_second": 8360.418
    },
    {
      "epoch": 1.2281174252625908,
      "grad_norm": 0.7079249930293646,
      "learning_rate": 3.24695992882152e-06,
      "loss": 0.3068737506866455,
      "num_input_tokens_seen": 938676584,
      "step": 18240,
      "train_runtime": 112279.5408,
      "train_tokens_per_second": 8360.175
    },
    {
      "epoch": 1.228454080258551,
      "grad_norm": 0.6211268675449834,
      "learning_rate": 3.244483932148773e-06,
      "loss": 0.29485907554626467,
      "num_input_tokens_seen": 938938272,
      "step": 18245,
      "train_runtime": 112308.928,
      "train_tokens_per_second": 8360.317
    },
    {
      "epoch": 1.228790735254511,
      "grad_norm": 0.5882601015813599,
      "learning_rate": 3.2420084264004966e-06,
      "loss": 0.3042133808135986,
      "num_input_tokens_seen": 939198360,
      "step": 18250,
      "train_runtime": 112337.2433,
      "train_tokens_per_second": 8360.525
    },
    {
      "epoch": 1.2291273902504714,
      "grad_norm": 0.6469167475932236,
      "learning_rate": 3.2395334122689594e-06,
      "loss": 0.30333480834960935,
      "num_input_tokens_seen": 939449784,
      "step": 18255,
      "train_runtime": 112368.5857,
      "train_tokens_per_second": 8360.431
    },
    {
      "epoch": 1.2294640452464314,
      "grad_norm": 0.6583376805296486,
      "learning_rate": 3.2370588904462873e-06,
      "loss": 0.31109232902526857,
      "num_input_tokens_seen": 939716752,
      "step": 18260,
      "train_runtime": 112399.518,
      "train_tokens_per_second": 8360.505
    },
    {
      "epoch": 1.2298007002423916,
      "grad_norm": 0.6952496288601652,
      "learning_rate": 3.2345848616244775e-06,
      "loss": 0.305296802520752,
      "num_input_tokens_seen": 939943832,
      "step": 18265,
      "train_runtime": 112426.6655,
      "train_tokens_per_second": 8360.506
    },
    {
      "epoch": 1.2301373552383517,
      "grad_norm": 0.7078185666646475,
      "learning_rate": 3.2321113264953797e-06,
      "loss": 0.30515754222869873,
      "num_input_tokens_seen": 940207824,
      "step": 18270,
      "train_runtime": 112462.3622,
      "train_tokens_per_second": 8360.2
    },
    {
      "epoch": 1.230474010234312,
      "grad_norm": 0.798282230137791,
      "learning_rate": 3.2296382857507124e-06,
      "loss": 0.2854955196380615,
      "num_input_tokens_seen": 940451272,
      "step": 18275,
      "train_runtime": 112498.4387,
      "train_tokens_per_second": 8359.683
    },
    {
      "epoch": 1.230810665230272,
      "grad_norm": 0.6567094725476114,
      "learning_rate": 3.227165740082053e-06,
      "loss": 0.3149935960769653,
      "num_input_tokens_seen": 940683896,
      "step": 18280,
      "train_runtime": 112524.8248,
      "train_tokens_per_second": 8359.79
    },
    {
      "epoch": 1.2311473202262322,
      "grad_norm": 0.6410597543792673,
      "learning_rate": 3.2246936901808424e-06,
      "loss": 0.29353866577148435,
      "num_input_tokens_seen": 940934320,
      "step": 18285,
      "train_runtime": 112554.4521,
      "train_tokens_per_second": 8359.814
    },
    {
      "epoch": 1.2314839752221922,
      "grad_norm": 0.6939228161241674,
      "learning_rate": 3.2222221367383786e-06,
      "loss": 0.3212655782699585,
      "num_input_tokens_seen": 941189680,
      "step": 18290,
      "train_runtime": 112585.7151,
      "train_tokens_per_second": 8359.761
    },
    {
      "epoch": 1.2318206302181525,
      "grad_norm": 0.6685122421447941,
      "learning_rate": 3.2197510804458265e-06,
      "loss": 0.30394325256347654,
      "num_input_tokens_seen": 941431944,
      "step": 18295,
      "train_runtime": 112612.9058,
      "train_tokens_per_second": 8359.894
    },
    {
      "epoch": 1.2321572852141125,
      "grad_norm": 0.6127419869861721,
      "learning_rate": 3.217280521994207e-06,
      "loss": 0.30586729049682615,
      "num_input_tokens_seen": 941707536,
      "step": 18300,
      "train_runtime": 112642.6203,
      "train_tokens_per_second": 8360.135
    },
    {
      "epoch": 1.2324939402100727,
      "grad_norm": 0.7504085159558997,
      "learning_rate": 3.214810462074407e-06,
      "loss": 0.2951645851135254,
      "num_input_tokens_seen": 941970136,
      "step": 18305,
      "train_runtime": 112675.3673,
      "train_tokens_per_second": 8360.036
    },
    {
      "epoch": 1.2328305952060328,
      "grad_norm": 0.6513291171710486,
      "learning_rate": 3.2123409013771666e-06,
      "loss": 0.2999980926513672,
      "num_input_tokens_seen": 942227488,
      "step": 18310,
      "train_runtime": 112709.3823,
      "train_tokens_per_second": 8359.796
    },
    {
      "epoch": 1.233167250201993,
      "grad_norm": 0.5891160087214137,
      "learning_rate": 3.209871840593095e-06,
      "loss": 0.3140019178390503,
      "num_input_tokens_seen": 942496008,
      "step": 18315,
      "train_runtime": 112743.5766,
      "train_tokens_per_second": 8359.643
    },
    {
      "epoch": 1.233503905197953,
      "grad_norm": 0.6331166218356793,
      "learning_rate": 3.207403280412652e-06,
      "loss": 0.2968922138214111,
      "num_input_tokens_seen": 942755184,
      "step": 18320,
      "train_runtime": 112777.4448,
      "train_tokens_per_second": 8359.43
    },
    {
      "epoch": 1.2338405601939133,
      "grad_norm": 0.6220156573968736,
      "learning_rate": 3.2049352215261677e-06,
      "loss": 0.287923526763916,
      "num_input_tokens_seen": 943030288,
      "step": 18325,
      "train_runtime": 112811.1074,
      "train_tokens_per_second": 8359.374
    },
    {
      "epoch": 1.2341772151898733,
      "grad_norm": 0.6616543770509772,
      "learning_rate": 3.2024676646238222e-06,
      "loss": 0.29592883586883545,
      "num_input_tokens_seen": 943290928,
      "step": 18330,
      "train_runtime": 112843.4699,
      "train_tokens_per_second": 8359.287
    },
    {
      "epoch": 1.2345138701858336,
      "grad_norm": 0.6727085075534636,
      "learning_rate": 3.2000006103956653e-06,
      "loss": 0.30621800422668455,
      "num_input_tokens_seen": 943555576,
      "step": 18335,
      "train_runtime": 112870.9305,
      "train_tokens_per_second": 8359.598
    },
    {
      "epoch": 1.2348505251817936,
      "grad_norm": 0.7223363524270422,
      "learning_rate": 3.1975340595315956e-06,
      "loss": 0.2802441596984863,
      "num_input_tokens_seen": 943797448,
      "step": 18340,
      "train_runtime": 112901.8171,
      "train_tokens_per_second": 8359.453
    },
    {
      "epoch": 1.2351871801777539,
      "grad_norm": 0.6294467213632026,
      "learning_rate": 3.1950680127213796e-06,
      "loss": 0.30479798316955564,
      "num_input_tokens_seen": 944059736,
      "step": 18345,
      "train_runtime": 112931.676,
      "train_tokens_per_second": 8359.565
    },
    {
      "epoch": 1.235523835173714,
      "grad_norm": 0.6361149131446852,
      "learning_rate": 3.192602470654638e-06,
      "loss": 0.31455469131469727,
      "num_input_tokens_seen": 944306208,
      "step": 18350,
      "train_runtime": 112964.3904,
      "train_tokens_per_second": 8359.326
    },
    {
      "epoch": 1.2358604901696741,
      "grad_norm": 0.7191784990901213,
      "learning_rate": 3.190137434020853e-06,
      "loss": 0.2854630470275879,
      "num_input_tokens_seen": 944559480,
      "step": 18355,
      "train_runtime": 112995.3532,
      "train_tokens_per_second": 8359.277
    },
    {
      "epoch": 1.2361971451656342,
      "grad_norm": 0.789928781629323,
      "learning_rate": 3.1876729035093616e-06,
      "loss": 0.2906643390655518,
      "num_input_tokens_seen": 944827520,
      "step": 18360,
      "train_runtime": 113025.46,
      "train_tokens_per_second": 8359.422
    },
    {
      "epoch": 1.2365338001615944,
      "grad_norm": 0.702563024690507,
      "learning_rate": 3.185208879809366e-06,
      "loss": 0.30158684253692625,
      "num_input_tokens_seen": 945093480,
      "step": 18365,
      "train_runtime": 113057.1329,
      "train_tokens_per_second": 8359.433
    },
    {
      "epoch": 1.2368704551575544,
      "grad_norm": 0.6696502131638835,
      "learning_rate": 3.1827453636099183e-06,
      "loss": 0.3003855228424072,
      "num_input_tokens_seen": 945379280,
      "step": 18370,
      "train_runtime": 113089.045,
      "train_tokens_per_second": 8359.601
    },
    {
      "epoch": 1.2372071101535147,
      "grad_norm": 0.8116160685110233,
      "learning_rate": 3.1802823555999356e-06,
      "loss": 0.2935532569885254,
      "num_input_tokens_seen": 945653656,
      "step": 18375,
      "train_runtime": 113120.1346,
      "train_tokens_per_second": 8359.729
    },
    {
      "epoch": 1.2375437651494747,
      "grad_norm": 0.6442068129706245,
      "learning_rate": 3.1778198564681885e-06,
      "loss": 0.3098647117614746,
      "num_input_tokens_seen": 945923016,
      "step": 18380,
      "train_runtime": 113153.5448,
      "train_tokens_per_second": 8359.641
    },
    {
      "epoch": 1.237880420145435,
      "grad_norm": 0.6060970654001319,
      "learning_rate": 3.17535786690331e-06,
      "loss": 0.3165290355682373,
      "num_input_tokens_seen": 946178088,
      "step": 18385,
      "train_runtime": 113184.4141,
      "train_tokens_per_second": 8359.615
    },
    {
      "epoch": 1.238217075141395,
      "grad_norm": 0.6501929703160678,
      "learning_rate": 3.172896387593784e-06,
      "loss": 0.3022042989730835,
      "num_input_tokens_seen": 946452416,
      "step": 18390,
      "train_runtime": 113214.737,
      "train_tokens_per_second": 8359.799
    },
    {
      "epoch": 1.2385537301373553,
      "grad_norm": 0.7250743021881255,
      "learning_rate": 3.1704354192279573e-06,
      "loss": 0.30698471069335936,
      "num_input_tokens_seen": 946702568,
      "step": 18395,
      "train_runtime": 113241.9369,
      "train_tokens_per_second": 8360.0
    },
    {
      "epoch": 1.2388903851333153,
      "grad_norm": 0.6920646302661121,
      "learning_rate": 3.1679749624940316e-06,
      "loss": 0.306855583190918,
      "num_input_tokens_seen": 946974600,
      "step": 18400,
      "train_runtime": 113272.4684,
      "train_tokens_per_second": 8360.148
    },
    {
      "epoch": 1.2392270401292755,
      "grad_norm": 0.6729364271093754,
      "learning_rate": 3.165515018080067e-06,
      "loss": 0.30208549499511717,
      "num_input_tokens_seen": 947234264,
      "step": 18405,
      "train_runtime": 113305.5945,
      "train_tokens_per_second": 8359.996
    },
    {
      "epoch": 1.2395636951252356,
      "grad_norm": 0.6623334411533914,
      "learning_rate": 3.163055586673976e-06,
      "loss": 0.28679752349853516,
      "num_input_tokens_seen": 947499712,
      "step": 18410,
      "train_runtime": 113339.0752,
      "train_tokens_per_second": 8359.868
    },
    {
      "epoch": 1.2399003501211958,
      "grad_norm": 1.546137945007892,
      "learning_rate": 3.1605966689635348e-06,
      "loss": 0.3073254585266113,
      "num_input_tokens_seen": 947779624,
      "step": 18415,
      "train_runtime": 113370.9579,
      "train_tokens_per_second": 8359.986
    },
    {
      "epoch": 1.2402370051171558,
      "grad_norm": 0.7283860052669471,
      "learning_rate": 3.1581382656363666e-06,
      "loss": 0.2918886184692383,
      "num_input_tokens_seen": 948030200,
      "step": 18420,
      "train_runtime": 113400.2423,
      "train_tokens_per_second": 8360.037
    },
    {
      "epoch": 1.240573660113116,
      "grad_norm": 0.648642040376813,
      "learning_rate": 3.1556803773799616e-06,
      "loss": 0.3260128259658813,
      "num_input_tokens_seen": 948292704,
      "step": 18425,
      "train_runtime": 113435.756,
      "train_tokens_per_second": 8359.734
    },
    {
      "epoch": 1.2409103151090761,
      "grad_norm": 0.6894021097574702,
      "learning_rate": 3.1532230048816564e-06,
      "loss": 0.3005608320236206,
      "num_input_tokens_seen": 948550816,
      "step": 18430,
      "train_runtime": 113465.0169,
      "train_tokens_per_second": 8359.853
    },
    {
      "epoch": 1.2412469701050364,
      "grad_norm": 0.6224480436898541,
      "learning_rate": 3.150766148828651e-06,
      "loss": 0.2982094049453735,
      "num_input_tokens_seen": 948816208,
      "step": 18435,
      "train_runtime": 113492.754,
      "train_tokens_per_second": 8360.148
    },
    {
      "epoch": 1.2415836251009964,
      "grad_norm": 0.723465788987981,
      "learning_rate": 3.1483098099079934e-06,
      "loss": 0.282586669921875,
      "num_input_tokens_seen": 949065456,
      "step": 18440,
      "train_runtime": 113523.0036,
      "train_tokens_per_second": 8360.116
    },
    {
      "epoch": 1.2419202800969567,
      "grad_norm": 0.5935488997418149,
      "learning_rate": 3.1458539888065937e-06,
      "loss": 0.2950422763824463,
      "num_input_tokens_seen": 949299816,
      "step": 18445,
      "train_runtime": 113551.902,
      "train_tokens_per_second": 8360.052
    },
    {
      "epoch": 1.2422569350929167,
      "grad_norm": 0.689343875737127,
      "learning_rate": 3.143398686211212e-06,
      "loss": 0.2848012685775757,
      "num_input_tokens_seen": 949532328,
      "step": 18450,
      "train_runtime": 113581.2966,
      "train_tokens_per_second": 8359.936
    },
    {
      "epoch": 1.242593590088877,
      "grad_norm": 0.5968698495747523,
      "learning_rate": 3.1409439028084706e-06,
      "loss": 0.3004512548446655,
      "num_input_tokens_seen": 949797344,
      "step": 18455,
      "train_runtime": 113612.5171,
      "train_tokens_per_second": 8359.971
    },
    {
      "epoch": 1.2429302450848372,
      "grad_norm": 0.6848318573288349,
      "learning_rate": 3.1384896392848363e-06,
      "loss": 0.28642032146453855,
      "num_input_tokens_seen": 950040256,
      "step": 18460,
      "train_runtime": 113644.1335,
      "train_tokens_per_second": 8359.783
    },
    {
      "epoch": 1.2432669000807972,
      "grad_norm": 0.6764666105065046,
      "learning_rate": 3.1360358963266404e-06,
      "loss": 0.32849347591400146,
      "num_input_tokens_seen": 950321240,
      "step": 18465,
      "train_runtime": 113677.7484,
      "train_tokens_per_second": 8359.782
    },
    {
      "epoch": 1.2436035550767572,
      "grad_norm": 0.6865328979474214,
      "learning_rate": 3.1335826746200594e-06,
      "loss": 0.30597572326660155,
      "num_input_tokens_seen": 950534768,
      "step": 18470,
      "train_runtime": 113703.6548,
      "train_tokens_per_second": 8359.756
    },
    {
      "epoch": 1.2439402100727175,
      "grad_norm": 0.7021974459977733,
      "learning_rate": 3.1311299748511335e-06,
      "loss": 0.3221548080444336,
      "num_input_tokens_seen": 950790008,
      "step": 18475,
      "train_runtime": 113734.4714,
      "train_tokens_per_second": 8359.735
    },
    {
      "epoch": 1.2442768650686777,
      "grad_norm": 0.6703235767646125,
      "learning_rate": 3.1286777977057504e-06,
      "loss": 0.28579187393188477,
      "num_input_tokens_seen": 951025872,
      "step": 18480,
      "train_runtime": 113769.9579,
      "train_tokens_per_second": 8359.2
    },
    {
      "epoch": 1.2446135200646378,
      "grad_norm": 0.670193807268741,
      "learning_rate": 3.126226143869655e-06,
      "loss": 0.3239318609237671,
      "num_input_tokens_seen": 951269040,
      "step": 18485,
      "train_runtime": 113801.3425,
      "train_tokens_per_second": 8359.032
    },
    {
      "epoch": 1.2449501750605978,
      "grad_norm": 0.6520196557535212,
      "learning_rate": 3.1237750140284424e-06,
      "loss": 0.27232682704925537,
      "num_input_tokens_seen": 951511656,
      "step": 18490,
      "train_runtime": 113835.3355,
      "train_tokens_per_second": 8358.667
    },
    {
      "epoch": 1.245286830056558,
      "grad_norm": 0.6279942282030285,
      "learning_rate": 3.1213244088675647e-06,
      "loss": 0.3052088260650635,
      "num_input_tokens_seen": 951763240,
      "step": 18495,
      "train_runtime": 113863.5411,
      "train_tokens_per_second": 8358.806
    },
    {
      "epoch": 1.2456234850525183,
      "grad_norm": 0.667818129969126,
      "learning_rate": 3.1188743290723246e-06,
      "loss": 0.3171834468841553,
      "num_input_tokens_seen": 952018096,
      "step": 18500,
      "train_runtime": 113894.0883,
      "train_tokens_per_second": 8358.802
    },
    {
      "epoch": 1.2459601400484783,
      "grad_norm": 0.7289622644896387,
      "learning_rate": 3.116424775327882e-06,
      "loss": 0.2964337825775146,
      "num_input_tokens_seen": 952271856,
      "step": 18505,
      "train_runtime": 113920.6023,
      "train_tokens_per_second": 8359.084
    },
    {
      "epoch": 1.2462967950444384,
      "grad_norm": 0.6078869429952752,
      "learning_rate": 3.1139757483192423e-06,
      "loss": 0.3054831027984619,
      "num_input_tokens_seen": 952545976,
      "step": 18510,
      "train_runtime": 113953.1108,
      "train_tokens_per_second": 8359.105
    },
    {
      "epoch": 1.2466334500403986,
      "grad_norm": 0.7382420167397549,
      "learning_rate": 3.1115272487312727e-06,
      "loss": 0.2921756267547607,
      "num_input_tokens_seen": 952795712,
      "step": 18515,
      "train_runtime": 113985.7117,
      "train_tokens_per_second": 8358.905
    },
    {
      "epoch": 1.2469701050363589,
      "grad_norm": 0.7810629201629543,
      "learning_rate": 3.109079277248684e-06,
      "loss": 0.29273257255554197,
      "num_input_tokens_seen": 953035720,
      "step": 18520,
      "train_runtime": 114016.3818,
      "train_tokens_per_second": 8358.761
    },
    {
      "epoch": 1.2473067600323189,
      "grad_norm": 0.5770538653572754,
      "learning_rate": 3.1066318345560463e-06,
      "loss": 0.32892603874206544,
      "num_input_tokens_seen": 953299168,
      "step": 18525,
      "train_runtime": 114047.6669,
      "train_tokens_per_second": 8358.778
    },
    {
      "epoch": 1.247643415028279,
      "grad_norm": 0.6515467071740286,
      "learning_rate": 3.1041849213377782e-06,
      "loss": 0.3114762783050537,
      "num_input_tokens_seen": 953586280,
      "step": 18530,
      "train_runtime": 114082.4611,
      "train_tokens_per_second": 8358.746
    },
    {
      "epoch": 1.2479800700242392,
      "grad_norm": 0.6586671445315933,
      "learning_rate": 3.101738538278154e-06,
      "loss": 0.3032058238983154,
      "num_input_tokens_seen": 953850872,
      "step": 18535,
      "train_runtime": 114111.2996,
      "train_tokens_per_second": 8358.952
    },
    {
      "epoch": 1.2483167250201994,
      "grad_norm": 0.73830240313759,
      "learning_rate": 3.0992926860612914e-06,
      "loss": 0.3054978370666504,
      "num_input_tokens_seen": 954096880,
      "step": 18540,
      "train_runtime": 114141.179,
      "train_tokens_per_second": 8358.919
    },
    {
      "epoch": 1.2486533800161594,
      "grad_norm": 0.6982939525988804,
      "learning_rate": 3.0968473653711693e-06,
      "loss": 0.3063164472579956,
      "num_input_tokens_seen": 954380096,
      "step": 18545,
      "train_runtime": 114175.5902,
      "train_tokens_per_second": 8358.88
    },
    {
      "epoch": 1.2489900350121195,
      "grad_norm": 0.7164470811692033,
      "learning_rate": 3.094402576891613e-06,
      "loss": 0.3234349250793457,
      "num_input_tokens_seen": 954636704,
      "step": 18550,
      "train_runtime": 114209.338,
      "train_tokens_per_second": 8358.657
    },
    {
      "epoch": 1.2493266900080797,
      "grad_norm": 0.5649864729099529,
      "learning_rate": 3.0919583213063004e-06,
      "loss": 0.3134467601776123,
      "num_input_tokens_seen": 954917496,
      "step": 18555,
      "train_runtime": 114245.625,
      "train_tokens_per_second": 8358.46
    },
    {
      "epoch": 1.24966334500404,
      "grad_norm": 0.6725170172348172,
      "learning_rate": 3.0895145992987574e-06,
      "loss": 0.3051428318023682,
      "num_input_tokens_seen": 955196296,
      "step": 18560,
      "train_runtime": 114276.1336,
      "train_tokens_per_second": 8358.668
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.7303544944109375,
      "learning_rate": 3.087071411552366e-06,
      "loss": 0.2929638385772705,
      "num_input_tokens_seen": 955448312,
      "step": 18565,
      "train_runtime": 114305.8353,
      "train_tokens_per_second": 8358.701
    },
    {
      "epoch": 1.25033665499596,
      "grad_norm": 0.6348547711197915,
      "learning_rate": 3.084628758750353e-06,
      "loss": 0.3208238840103149,
      "num_input_tokens_seen": 955695136,
      "step": 18570,
      "train_runtime": 114337.2495,
      "train_tokens_per_second": 8358.563
    },
    {
      "epoch": 1.2506733099919203,
      "grad_norm": 0.8339843012916267,
      "learning_rate": 3.082186641575801e-06,
      "loss": 0.3317065715789795,
      "num_input_tokens_seen": 955947240,
      "step": 18575,
      "train_runtime": 114365.6615,
      "train_tokens_per_second": 8358.691
    },
    {
      "epoch": 1.2510099649878805,
      "grad_norm": 0.7333186042591734,
      "learning_rate": 3.0797450607116384e-06,
      "loss": 0.3068979740142822,
      "num_input_tokens_seen": 956195000,
      "step": 18580,
      "train_runtime": 114394.7657,
      "train_tokens_per_second": 8358.73
    },
    {
      "epoch": 1.2513466199838406,
      "grad_norm": 0.6188315369519921,
      "learning_rate": 3.0773040168406475e-06,
      "loss": 0.3139481067657471,
      "num_input_tokens_seen": 956477632,
      "step": 18585,
      "train_runtime": 114427.758,
      "train_tokens_per_second": 8358.79
    },
    {
      "epoch": 1.2516832749798006,
      "grad_norm": 0.687385910224288,
      "learning_rate": 3.074863510645456e-06,
      "loss": 0.3119869947433472,
      "num_input_tokens_seen": 956738464,
      "step": 18590,
      "train_runtime": 114463.4572,
      "train_tokens_per_second": 8358.462
    },
    {
      "epoch": 1.2520199299757608,
      "grad_norm": 0.6200157399846669,
      "learning_rate": 3.072423542808546e-06,
      "loss": 0.28959572315216064,
      "num_input_tokens_seen": 956994872,
      "step": 18595,
      "train_runtime": 114489.7561,
      "train_tokens_per_second": 8358.782
    },
    {
      "epoch": 1.252356584971721,
      "grad_norm": 0.6388458388326855,
      "learning_rate": 3.0699841140122443e-06,
      "loss": 0.2989231586456299,
      "num_input_tokens_seen": 957249488,
      "step": 18600,
      "train_runtime": 114519.9517,
      "train_tokens_per_second": 8358.801
    },
    {
      "epoch": 1.2526932399676811,
      "grad_norm": 0.7091786565691167,
      "learning_rate": 3.0675452249387327e-06,
      "loss": 0.2964563131332397,
      "num_input_tokens_seen": 957518520,
      "step": 18605,
      "train_runtime": 114548.4371,
      "train_tokens_per_second": 8359.071
    },
    {
      "epoch": 1.2530298949636411,
      "grad_norm": 0.6771355428796569,
      "learning_rate": 3.0651068762700354e-06,
      "loss": 0.3373867511749268,
      "num_input_tokens_seen": 957761120,
      "step": 18610,
      "train_runtime": 114578.2171,
      "train_tokens_per_second": 8359.016
    },
    {
      "epoch": 1.2533665499596014,
      "grad_norm": 0.6502573405128462,
      "learning_rate": 3.0626690686880313e-06,
      "loss": 0.3185829401016235,
      "num_input_tokens_seen": 958050488,
      "step": 18615,
      "train_runtime": 114614.6539,
      "train_tokens_per_second": 8358.883
    },
    {
      "epoch": 1.2537032049555616,
      "grad_norm": 0.6304179154732418,
      "learning_rate": 3.060231802874443e-06,
      "loss": 0.30222296714782715,
      "num_input_tokens_seen": 958316176,
      "step": 18620,
      "train_runtime": 114642.2594,
      "train_tokens_per_second": 8359.188
    },
    {
      "epoch": 1.2540398599515217,
      "grad_norm": 0.668487141117124,
      "learning_rate": 3.0577950795108462e-06,
      "loss": 0.3091724395751953,
      "num_input_tokens_seen": 958576424,
      "step": 18625,
      "train_runtime": 114672.8302,
      "train_tokens_per_second": 8359.229
    },
    {
      "epoch": 1.2543765149474817,
      "grad_norm": 0.729627613527525,
      "learning_rate": 3.055358899278662e-06,
      "loss": 0.29625868797302246,
      "num_input_tokens_seen": 958808264,
      "step": 18630,
      "train_runtime": 114708.1065,
      "train_tokens_per_second": 8358.679
    },
    {
      "epoch": 1.254713169943442,
      "grad_norm": 0.6180858467217152,
      "learning_rate": 3.052923262859162e-06,
      "loss": 0.2930983304977417,
      "num_input_tokens_seen": 959089336,
      "step": 18635,
      "train_runtime": 114736.6225,
      "train_tokens_per_second": 8359.051
    },
    {
      "epoch": 1.2550498249394022,
      "grad_norm": 0.6328394797194425,
      "learning_rate": 3.0504881709334603e-06,
      "loss": 0.299950098991394,
      "num_input_tokens_seen": 959341056,
      "step": 18640,
      "train_runtime": 114768.19,
      "train_tokens_per_second": 8358.946
    },
    {
      "epoch": 1.2553864799353622,
      "grad_norm": 0.6642241934696849,
      "learning_rate": 3.0480536241825263e-06,
      "loss": 0.28252859115600587,
      "num_input_tokens_seen": 959592536,
      "step": 18645,
      "train_runtime": 114796.3051,
      "train_tokens_per_second": 8359.089
    },
    {
      "epoch": 1.2557231349313223,
      "grad_norm": 0.6683640572411111,
      "learning_rate": 3.0456196232871715e-06,
      "loss": 0.30548121929168703,
      "num_input_tokens_seen": 959835672,
      "step": 18650,
      "train_runtime": 114823.8407,
      "train_tokens_per_second": 8359.202
    },
    {
      "epoch": 1.2560597899272825,
      "grad_norm": 0.7630324436695216,
      "learning_rate": 3.043186168928059e-06,
      "loss": 0.30261590480804446,
      "num_input_tokens_seen": 960056272,
      "step": 18655,
      "train_runtime": 114855.1297,
      "train_tokens_per_second": 8358.845
    },
    {
      "epoch": 1.2563964449232428,
      "grad_norm": 0.7418429000471426,
      "learning_rate": 3.040753261785692e-06,
      "loss": 0.278994083404541,
      "num_input_tokens_seen": 960290456,
      "step": 18660,
      "train_runtime": 114884.5476,
      "train_tokens_per_second": 8358.743
    },
    {
      "epoch": 1.2567330999192028,
      "grad_norm": 0.7698892027836927,
      "learning_rate": 3.038320902540429e-06,
      "loss": 0.285690450668335,
      "num_input_tokens_seen": 960533624,
      "step": 18665,
      "train_runtime": 114913.3422,
      "train_tokens_per_second": 8358.765
    },
    {
      "epoch": 1.2570697549151628,
      "grad_norm": 0.6476298958528087,
      "learning_rate": 3.0358890918724706e-06,
      "loss": 0.30154948234558104,
      "num_input_tokens_seen": 960801048,
      "step": 18670,
      "train_runtime": 114946.6346,
      "train_tokens_per_second": 8358.671
    },
    {
      "epoch": 1.257406409911123,
      "grad_norm": 0.7562029668972012,
      "learning_rate": 3.0334578304618646e-06,
      "loss": 0.31045963764190676,
      "num_input_tokens_seen": 961032368,
      "step": 18675,
      "train_runtime": 114975.4298,
      "train_tokens_per_second": 8358.589
    },
    {
      "epoch": 1.2577430649070833,
      "grad_norm": 0.6315678845534666,
      "learning_rate": 3.0310271189885037e-06,
      "loss": 0.30917465686798096,
      "num_input_tokens_seen": 961285584,
      "step": 18680,
      "train_runtime": 115003.7065,
      "train_tokens_per_second": 8358.736
    },
    {
      "epoch": 1.2580797199030433,
      "grad_norm": 1.5017580458858695,
      "learning_rate": 3.0285969581321328e-06,
      "loss": 0.29650652408599854,
      "num_input_tokens_seen": 961539488,
      "step": 18685,
      "train_runtime": 115034.1006,
      "train_tokens_per_second": 8358.734
    },
    {
      "epoch": 1.2584163748990034,
      "grad_norm": 0.6679269635079449,
      "learning_rate": 3.0261673485723343e-06,
      "loss": 0.34806137084960936,
      "num_input_tokens_seen": 961781416,
      "step": 18690,
      "train_runtime": 115068.0721,
      "train_tokens_per_second": 8358.369
    },
    {
      "epoch": 1.2587530298949636,
      "grad_norm": 0.6174304173376781,
      "learning_rate": 3.023738290988544e-06,
      "loss": 0.2779110908508301,
      "num_input_tokens_seen": 962042424,
      "step": 18695,
      "train_runtime": 115102.3898,
      "train_tokens_per_second": 8358.145
    },
    {
      "epoch": 1.2590896848909239,
      "grad_norm": 0.7581978004967864,
      "learning_rate": 3.021309786060037e-06,
      "loss": 0.30838067531585694,
      "num_input_tokens_seen": 962274800,
      "step": 18700,
      "train_runtime": 115130.7465,
      "train_tokens_per_second": 8358.104
    },
    {
      "epoch": 1.259426339886884,
      "grad_norm": 0.5734301181345368,
      "learning_rate": 3.018881834465941e-06,
      "loss": 0.28553023338317873,
      "num_input_tokens_seen": 962521112,
      "step": 18705,
      "train_runtime": 115160.6061,
      "train_tokens_per_second": 8358.076
    },
    {
      "epoch": 1.259762994882844,
      "grad_norm": 0.6552022459780517,
      "learning_rate": 3.016454436885221e-06,
      "loss": 0.3150408983230591,
      "num_input_tokens_seen": 962776528,
      "step": 18710,
      "train_runtime": 115193.9088,
      "train_tokens_per_second": 8357.877
    },
    {
      "epoch": 1.2600996498788042,
      "grad_norm": 0.6550870333720717,
      "learning_rate": 3.0140275939966934e-06,
      "loss": 0.30576119422912595,
      "num_input_tokens_seen": 963029376,
      "step": 18715,
      "train_runtime": 115224.3057,
      "train_tokens_per_second": 8357.867
    },
    {
      "epoch": 1.2604363048747644,
      "grad_norm": 0.6384828281409937,
      "learning_rate": 3.0116013064790166e-06,
      "loss": 0.348641562461853,
      "num_input_tokens_seen": 963323920,
      "step": 18720,
      "train_runtime": 115254.4865,
      "train_tokens_per_second": 8358.234
    },
    {
      "epoch": 1.2607729598707245,
      "grad_norm": 0.7212303870288262,
      "learning_rate": 3.0091755750106937e-06,
      "loss": 0.29804325103759766,
      "num_input_tokens_seen": 963582600,
      "step": 18725,
      "train_runtime": 115285.6737,
      "train_tokens_per_second": 8358.216
    },
    {
      "epoch": 1.2611096148666845,
      "grad_norm": 0.5967036014710859,
      "learning_rate": 3.006750400270072e-06,
      "loss": 0.28386673927307127,
      "num_input_tokens_seen": 963835184,
      "step": 18730,
      "train_runtime": 115315.8102,
      "train_tokens_per_second": 8358.222
    },
    {
      "epoch": 1.2614462698626447,
      "grad_norm": 0.586999543747888,
      "learning_rate": 3.0043257829353466e-06,
      "loss": 0.2827768802642822,
      "num_input_tokens_seen": 964111736,
      "step": 18735,
      "train_runtime": 115345.3336,
      "train_tokens_per_second": 8358.481
    },
    {
      "epoch": 1.261782924858605,
      "grad_norm": 0.662621561152035,
      "learning_rate": 3.0019017236845504e-06,
      "loss": 0.3197850227355957,
      "num_input_tokens_seen": 964369280,
      "step": 18740,
      "train_runtime": 115376.2476,
      "train_tokens_per_second": 8358.473
    },
    {
      "epoch": 1.262119579854565,
      "grad_norm": 0.7128186013795516,
      "learning_rate": 2.9994782231955673e-06,
      "loss": 0.3055188417434692,
      "num_input_tokens_seen": 964628128,
      "step": 18745,
      "train_runtime": 115408.5763,
      "train_tokens_per_second": 8358.375
    },
    {
      "epoch": 1.262456234850525,
      "grad_norm": 0.6288993924577417,
      "learning_rate": 2.9970552821461185e-06,
      "loss": 0.2980309247970581,
      "num_input_tokens_seen": 964870120,
      "step": 18750,
      "train_runtime": 115439.2227,
      "train_tokens_per_second": 8358.252
    },
    {
      "epoch": 1.2627928898464853,
      "grad_norm": 0.7694419873253359,
      "learning_rate": 2.994632901213775e-06,
      "loss": 0.2910336494445801,
      "num_input_tokens_seen": 965116368,
      "step": 18755,
      "train_runtime": 115468.9711,
      "train_tokens_per_second": 8358.231
    },
    {
      "epoch": 1.2631295448424456,
      "grad_norm": 0.7194882352562777,
      "learning_rate": 2.9922110810759443e-06,
      "loss": 0.31582036018371584,
      "num_input_tokens_seen": 965350608,
      "step": 18760,
      "train_runtime": 115497.4932,
      "train_tokens_per_second": 8358.195
    },
    {
      "epoch": 1.2634661998384056,
      "grad_norm": 0.6115841694323227,
      "learning_rate": 2.989789822409883e-06,
      "loss": 0.30611820220947267,
      "num_input_tokens_seen": 965624648,
      "step": 18765,
      "train_runtime": 115529.5513,
      "train_tokens_per_second": 8358.248
    },
    {
      "epoch": 1.2638028548343656,
      "grad_norm": 0.6191238900004944,
      "learning_rate": 2.9873691258926864e-06,
      "loss": 0.3253757953643799,
      "num_input_tokens_seen": 965927928,
      "step": 18770,
      "train_runtime": 115561.3323,
      "train_tokens_per_second": 8358.574
    },
    {
      "epoch": 1.2641395098303259,
      "grad_norm": 0.6249288837072322,
      "learning_rate": 2.9849489922012985e-06,
      "loss": 0.30664787292480467,
      "num_input_tokens_seen": 966160384,
      "step": 18775,
      "train_runtime": 115592.9056,
      "train_tokens_per_second": 8358.302
    },
    {
      "epoch": 1.264476164826286,
      "grad_norm": 0.640463089539323,
      "learning_rate": 2.982529422012498e-06,
      "loss": 0.32817587852478025,
      "num_input_tokens_seen": 966444048,
      "step": 18780,
      "train_runtime": 115623.8075,
      "train_tokens_per_second": 8358.521
    },
    {
      "epoch": 1.2648128198222461,
      "grad_norm": 0.6995393005602683,
      "learning_rate": 2.9801104160029127e-06,
      "loss": 0.2688779830932617,
      "num_input_tokens_seen": 966674096,
      "step": 18785,
      "train_runtime": 115654.406,
      "train_tokens_per_second": 8358.299
    },
    {
      "epoch": 1.2651494748182062,
      "grad_norm": 0.6837463072694947,
      "learning_rate": 2.9776919748490063e-06,
      "loss": 0.32268438339233396,
      "num_input_tokens_seen": 966932336,
      "step": 18790,
      "train_runtime": 115682.0813,
      "train_tokens_per_second": 8358.532
    },
    {
      "epoch": 1.2654861298141664,
      "grad_norm": 0.7368554861738245,
      "learning_rate": 2.975274099227092e-06,
      "loss": 0.3244370460510254,
      "num_input_tokens_seen": 967167656,
      "step": 18795,
      "train_runtime": 115711.8969,
      "train_tokens_per_second": 8358.412
    },
    {
      "epoch": 1.2658227848101267,
      "grad_norm": 0.8263175223848045,
      "learning_rate": 2.9728567898133183e-06,
      "loss": 0.30799107551574706,
      "num_input_tokens_seen": 967428560,
      "step": 18800,
      "train_runtime": 115741.3408,
      "train_tokens_per_second": 8358.539
    },
    {
      "epoch": 1.2661594398060867,
      "grad_norm": 0.6650198851966721,
      "learning_rate": 2.9704400472836816e-06,
      "loss": 0.3003499507904053,
      "num_input_tokens_seen": 967682936,
      "step": 18805,
      "train_runtime": 115774.5598,
      "train_tokens_per_second": 8358.338
    },
    {
      "epoch": 1.266496094802047,
      "grad_norm": 0.6750397864853851,
      "learning_rate": 2.9680238723140108e-06,
      "loss": 0.32263615131378176,
      "num_input_tokens_seen": 967943400,
      "step": 18810,
      "train_runtime": 115800.2747,
      "train_tokens_per_second": 8358.731
    },
    {
      "epoch": 1.266832749798007,
      "grad_norm": 0.6428381951394027,
      "learning_rate": 2.9656082655799866e-06,
      "loss": 0.32791891098022463,
      "num_input_tokens_seen": 968198552,
      "step": 18815,
      "train_runtime": 115828.1138,
      "train_tokens_per_second": 8358.925
    },
    {
      "epoch": 1.2671694047939672,
      "grad_norm": 1.0924912705234273,
      "learning_rate": 2.9631932277571225e-06,
      "loss": 0.3148000717163086,
      "num_input_tokens_seen": 968475064,
      "step": 18820,
      "train_runtime": 115857.2379,
      "train_tokens_per_second": 8359.211
    },
    {
      "epoch": 1.2675060597899273,
      "grad_norm": 0.6337481121893204,
      "learning_rate": 2.9607787595207794e-06,
      "loss": 0.2861155986785889,
      "num_input_tokens_seen": 968730048,
      "step": 18825,
      "train_runtime": 115887.4366,
      "train_tokens_per_second": 8359.233
    },
    {
      "epoch": 1.2678427147858875,
      "grad_norm": 0.5436615143431183,
      "learning_rate": 2.9583648615461514e-06,
      "loss": 0.2969028949737549,
      "num_input_tokens_seen": 968977568,
      "step": 18830,
      "train_runtime": 115919.1874,
      "train_tokens_per_second": 8359.078
    },
    {
      "epoch": 1.2681793697818475,
      "grad_norm": 0.6299312020203105,
      "learning_rate": 2.955951534508281e-06,
      "loss": 0.3313584327697754,
      "num_input_tokens_seen": 969223840,
      "step": 18835,
      "train_runtime": 115951.5267,
      "train_tokens_per_second": 8358.871
    },
    {
      "epoch": 1.2685160247778078,
      "grad_norm": 0.6171914136147806,
      "learning_rate": 2.9535387790820452e-06,
      "loss": 0.31123526096343995,
      "num_input_tokens_seen": 969502096,
      "step": 18840,
      "train_runtime": 115985.2981,
      "train_tokens_per_second": 8358.836
    },
    {
      "epoch": 1.2688526797737678,
      "grad_norm": 0.6376060502527366,
      "learning_rate": 2.951126595942165e-06,
      "loss": 0.2920564889907837,
      "num_input_tokens_seen": 969777040,
      "step": 18845,
      "train_runtime": 116017.6168,
      "train_tokens_per_second": 8358.877
    },
    {
      "epoch": 1.269189334769728,
      "grad_norm": 0.7072477973900015,
      "learning_rate": 2.948714985763197e-06,
      "loss": 0.2797840118408203,
      "num_input_tokens_seen": 970007896,
      "step": 18850,
      "train_runtime": 116050.2956,
      "train_tokens_per_second": 8358.513
    },
    {
      "epoch": 1.269525989765688,
      "grad_norm": 0.6361644309082884,
      "learning_rate": 2.9463039492195445e-06,
      "loss": 0.3075706005096436,
      "num_input_tokens_seen": 970291120,
      "step": 18855,
      "train_runtime": 116084.1081,
      "train_tokens_per_second": 8358.518
    },
    {
      "epoch": 1.2698626447616483,
      "grad_norm": 0.576420364010783,
      "learning_rate": 2.9438934869854417e-06,
      "loss": 0.2924260139465332,
      "num_input_tokens_seen": 970548480,
      "step": 18860,
      "train_runtime": 116111.5282,
      "train_tokens_per_second": 8358.761
    },
    {
      "epoch": 1.2701992997576084,
      "grad_norm": 0.6871175376264038,
      "learning_rate": 2.9414835997349705e-06,
      "loss": 0.28102641105651854,
      "num_input_tokens_seen": 970799528,
      "step": 18865,
      "train_runtime": 116142.6997,
      "train_tokens_per_second": 8358.679
    },
    {
      "epoch": 1.2705359547535686,
      "grad_norm": 0.5929877812675389,
      "learning_rate": 2.939074288142045e-06,
      "loss": 0.32398509979248047,
      "num_input_tokens_seen": 971077344,
      "step": 18870,
      "train_runtime": 116172.5751,
      "train_tokens_per_second": 8358.921
    },
    {
      "epoch": 1.2708726097495286,
      "grad_norm": 0.6147399530082972,
      "learning_rate": 2.9366655528804257e-06,
      "loss": 0.2902873992919922,
      "num_input_tokens_seen": 971333064,
      "step": 18875,
      "train_runtime": 116201.1837,
      "train_tokens_per_second": 8359.063
    },
    {
      "epoch": 1.271209264745489,
      "grad_norm": 0.6279400825357576,
      "learning_rate": 2.934257394623702e-06,
      "loss": 0.2958557605743408,
      "num_input_tokens_seen": 971601224,
      "step": 18880,
      "train_runtime": 116234.9592,
      "train_tokens_per_second": 8358.942
    },
    {
      "epoch": 1.271545919741449,
      "grad_norm": 0.6248425062276546,
      "learning_rate": 2.9318498140453134e-06,
      "loss": 0.31405913829803467,
      "num_input_tokens_seen": 971860656,
      "step": 18885,
      "train_runtime": 116263.8688,
      "train_tokens_per_second": 8359.094
    },
    {
      "epoch": 1.2718825747374092,
      "grad_norm": 0.7459205349977819,
      "learning_rate": 2.929442811818527e-06,
      "loss": 0.3079646110534668,
      "num_input_tokens_seen": 972094520,
      "step": 18890,
      "train_runtime": 116292.5849,
      "train_tokens_per_second": 8359.041
    },
    {
      "epoch": 1.2722192297333692,
      "grad_norm": 0.6956055819821743,
      "learning_rate": 2.927036388616457e-06,
      "loss": 0.2970764398574829,
      "num_input_tokens_seen": 972330088,
      "step": 18895,
      "train_runtime": 116321.9938,
      "train_tokens_per_second": 8358.953
    },
    {
      "epoch": 1.2725558847293295,
      "grad_norm": 0.762570617775703,
      "learning_rate": 2.9246305451120494e-06,
      "loss": 0.32904183864593506,
      "num_input_tokens_seen": 972575400,
      "step": 18900,
      "train_runtime": 116350.307,
      "train_tokens_per_second": 8359.027
    },
    {
      "epoch": 1.2728925397252895,
      "grad_norm": 0.682138275260431,
      "learning_rate": 2.922225281978095e-06,
      "loss": 0.2944671154022217,
      "num_input_tokens_seen": 972817896,
      "step": 18905,
      "train_runtime": 116382.45,
      "train_tokens_per_second": 8358.802
    },
    {
      "epoch": 1.2732291947212497,
      "grad_norm": 0.635444122494314,
      "learning_rate": 2.919820599887212e-06,
      "loss": 0.3074968338012695,
      "num_input_tokens_seen": 973076776,
      "step": 18910,
      "train_runtime": 116410.9348,
      "train_tokens_per_second": 8358.981
    },
    {
      "epoch": 1.2735658497172098,
      "grad_norm": 0.6987862498147768,
      "learning_rate": 2.917416499511866e-06,
      "loss": 0.30770838260650635,
      "num_input_tokens_seen": 973305480,
      "step": 18915,
      "train_runtime": 116442.5253,
      "train_tokens_per_second": 8358.677
    },
    {
      "epoch": 1.27390250471317,
      "grad_norm": 0.6326784722587996,
      "learning_rate": 2.9150129815243553e-06,
      "loss": 0.27744617462158205,
      "num_input_tokens_seen": 973559264,
      "step": 18920,
      "train_runtime": 116469.2639,
      "train_tokens_per_second": 8358.937
    },
    {
      "epoch": 1.27423915970913,
      "grad_norm": 0.7421199630638027,
      "learning_rate": 2.912610046596816e-06,
      "loss": 0.30734972953796386,
      "num_input_tokens_seen": 973800904,
      "step": 18925,
      "train_runtime": 116500.5777,
      "train_tokens_per_second": 8358.765
    },
    {
      "epoch": 1.2745758147050903,
      "grad_norm": 0.6000305932237072,
      "learning_rate": 2.91020769540122e-06,
      "loss": 0.31457953453063964,
      "num_input_tokens_seen": 974024784,
      "step": 18930,
      "train_runtime": 116529.58,
      "train_tokens_per_second": 8358.605
    },
    {
      "epoch": 1.2749124697010503,
      "grad_norm": 0.6062639640805295,
      "learning_rate": 2.9078059286093795e-06,
      "loss": 0.29296417236328126,
      "num_input_tokens_seen": 974253688,
      "step": 18935,
      "train_runtime": 116560.6128,
      "train_tokens_per_second": 8358.344
    },
    {
      "epoch": 1.2752491246970106,
      "grad_norm": 0.520348904750659,
      "learning_rate": 2.905404746892937e-06,
      "loss": 0.2825160026550293,
      "num_input_tokens_seen": 974514712,
      "step": 18940,
      "train_runtime": 116593.5779,
      "train_tokens_per_second": 8358.219
    },
    {
      "epoch": 1.2755857796929706,
      "grad_norm": 0.6429156462257553,
      "learning_rate": 2.903004150923382e-06,
      "loss": 0.3014225959777832,
      "num_input_tokens_seen": 974775632,
      "step": 18945,
      "train_runtime": 116625.2318,
      "train_tokens_per_second": 8358.188
    },
    {
      "epoch": 1.2759224346889309,
      "grad_norm": 0.6667694994423854,
      "learning_rate": 2.9006041413720252e-06,
      "loss": 0.3288427829742432,
      "num_input_tokens_seen": 975031616,
      "step": 18950,
      "train_runtime": 116656.9281,
      "train_tokens_per_second": 8358.112
    },
    {
      "epoch": 1.2762590896848909,
      "grad_norm": 0.6089511768942755,
      "learning_rate": 2.8982047189100276e-06,
      "loss": 0.29043145179748536,
      "num_input_tokens_seen": 975308568,
      "step": 18955,
      "train_runtime": 116689.0708,
      "train_tokens_per_second": 8358.183
    },
    {
      "epoch": 1.2765957446808511,
      "grad_norm": 0.6992781935470396,
      "learning_rate": 2.895805884208378e-06,
      "loss": 0.2972787618637085,
      "num_input_tokens_seen": 975545976,
      "step": 18960,
      "train_runtime": 116718.5601,
      "train_tokens_per_second": 8358.105
    },
    {
      "epoch": 1.2769323996768112,
      "grad_norm": 0.7305106111031626,
      "learning_rate": 2.8934076379379016e-06,
      "loss": 0.3120333909988403,
      "num_input_tokens_seen": 975813680,
      "step": 18965,
      "train_runtime": 116751.6443,
      "train_tokens_per_second": 8358.029
    },
    {
      "epoch": 1.2772690546727714,
      "grad_norm": 0.6081098339695414,
      "learning_rate": 2.891009980769266e-06,
      "loss": 0.29751224517822267,
      "num_input_tokens_seen": 976077688,
      "step": 18970,
      "train_runtime": 116780.8175,
      "train_tokens_per_second": 8358.202
    },
    {
      "epoch": 1.2776057096687314,
      "grad_norm": 0.6899333933222006,
      "learning_rate": 2.88861291337296e-06,
      "loss": 0.2934125900268555,
      "num_input_tokens_seen": 976342416,
      "step": 18975,
      "train_runtime": 116810.7486,
      "train_tokens_per_second": 8358.327
    },
    {
      "epoch": 1.2779423646646917,
      "grad_norm": 0.673952472034726,
      "learning_rate": 2.8862164364193236e-06,
      "loss": 0.29152803421020507,
      "num_input_tokens_seen": 976596992,
      "step": 18980,
      "train_runtime": 116839.6786,
      "train_tokens_per_second": 8358.436
    },
    {
      "epoch": 1.2782790196606517,
      "grad_norm": 0.6143271560887512,
      "learning_rate": 2.8838205505785168e-06,
      "loss": 0.3096057653427124,
      "num_input_tokens_seen": 976861408,
      "step": 18985,
      "train_runtime": 116869.9091,
      "train_tokens_per_second": 8358.537
    },
    {
      "epoch": 1.278615674656612,
      "grad_norm": 0.6865477069120016,
      "learning_rate": 2.881425256520547e-06,
      "loss": 0.2975276231765747,
      "num_input_tokens_seen": 977123808,
      "step": 18990,
      "train_runtime": 116904.5174,
      "train_tokens_per_second": 8358.307
    },
    {
      "epoch": 1.278952329652572,
      "grad_norm": 0.5820282455412795,
      "learning_rate": 2.8790305549152487e-06,
      "loss": 0.25682644844055175,
      "num_input_tokens_seen": 977392480,
      "step": 18995,
      "train_runtime": 116937.1969,
      "train_tokens_per_second": 8358.268
    },
    {
      "epoch": 1.2792889846485322,
      "grad_norm": 0.6570155874236377,
      "learning_rate": 2.876636446432292e-06,
      "loss": 0.31268982887268065,
      "num_input_tokens_seen": 977648960,
      "step": 19000,
      "train_runtime": 116968.9301,
      "train_tokens_per_second": 8358.194
    },
    {
      "epoch": 1.2796256396444923,
      "grad_norm": 0.6471598573123991,
      "learning_rate": 2.8742429317411826e-06,
      "loss": 0.3017135143280029,
      "num_input_tokens_seen": 977920336,
      "step": 19005,
      "train_runtime": 117001.8599,
      "train_tokens_per_second": 8358.161
    },
    {
      "epoch": 1.2799622946404525,
      "grad_norm": 0.6977558885676273,
      "learning_rate": 2.8718500115112585e-06,
      "loss": 0.3094404458999634,
      "num_input_tokens_seen": 978166448,
      "step": 19010,
      "train_runtime": 117038.174,
      "train_tokens_per_second": 8357.67
    },
    {
      "epoch": 1.2802989496364126,
      "grad_norm": 0.6491851588560112,
      "learning_rate": 2.869457686411691e-06,
      "loss": 0.3197046756744385,
      "num_input_tokens_seen": 978403128,
      "step": 19015,
      "train_runtime": 117066.9275,
      "train_tokens_per_second": 8357.639
    },
    {
      "epoch": 1.2806356046323728,
      "grad_norm": 0.673482065432389,
      "learning_rate": 2.8670659571114912e-06,
      "loss": 0.30575602054595946,
      "num_input_tokens_seen": 978677752,
      "step": 19020,
      "train_runtime": 117098.3942,
      "train_tokens_per_second": 8357.738
    },
    {
      "epoch": 1.2809722596283328,
      "grad_norm": 0.5642820615393821,
      "learning_rate": 2.864674824279491e-06,
      "loss": 0.2879039764404297,
      "num_input_tokens_seen": 978944832,
      "step": 19025,
      "train_runtime": 117128.9646,
      "train_tokens_per_second": 8357.837
    },
    {
      "epoch": 1.281308914624293,
      "grad_norm": 0.6105714379782573,
      "learning_rate": 2.862284288584371e-06,
      "loss": 0.28074357509613035,
      "num_input_tokens_seen": 979225112,
      "step": 19030,
      "train_runtime": 117161.2401,
      "train_tokens_per_second": 8357.927
    },
    {
      "epoch": 1.2816455696202531,
      "grad_norm": 0.7145378825499431,
      "learning_rate": 2.8598943506946297e-06,
      "loss": 0.30459573268890383,
      "num_input_tokens_seen": 979480640,
      "step": 19035,
      "train_runtime": 117195.5214,
      "train_tokens_per_second": 8357.663
    },
    {
      "epoch": 1.2819822246162134,
      "grad_norm": 0.6428910408870873,
      "learning_rate": 2.85750501127861e-06,
      "loss": 0.27977454662323,
      "num_input_tokens_seen": 979726456,
      "step": 19040,
      "train_runtime": 117225.6332,
      "train_tokens_per_second": 8357.613
    },
    {
      "epoch": 1.2823188796121734,
      "grad_norm": 0.6203471632932999,
      "learning_rate": 2.8551162710044827e-06,
      "loss": 0.29895853996276855,
      "num_input_tokens_seen": 979976272,
      "step": 19045,
      "train_runtime": 117257.189,
      "train_tokens_per_second": 8357.494
    },
    {
      "epoch": 1.2826555346081336,
      "grad_norm": 0.6601702836091164,
      "learning_rate": 2.8527281305402503e-06,
      "loss": 0.32022643089294434,
      "num_input_tokens_seen": 980233368,
      "step": 19050,
      "train_runtime": 117291.0125,
      "train_tokens_per_second": 8357.276
    },
    {
      "epoch": 1.2829921896040937,
      "grad_norm": 0.7382190492558451,
      "learning_rate": 2.8503405905537496e-06,
      "loss": 0.2962524175643921,
      "num_input_tokens_seen": 980483440,
      "step": 19055,
      "train_runtime": 117316.619,
      "train_tokens_per_second": 8357.584
    },
    {
      "epoch": 1.283328844600054,
      "grad_norm": 0.5966689721847331,
      "learning_rate": 2.847953651712647e-06,
      "loss": 0.30246973037719727,
      "num_input_tokens_seen": 980765648,
      "step": 19060,
      "train_runtime": 117353.1325,
      "train_tokens_per_second": 8357.388
    },
    {
      "epoch": 1.283665499596014,
      "grad_norm": 0.6118635181867711,
      "learning_rate": 2.8455673146844426e-06,
      "loss": 0.28995609283447266,
      "num_input_tokens_seen": 981000416,
      "step": 19065,
      "train_runtime": 117381.7431,
      "train_tokens_per_second": 8357.351
    },
    {
      "epoch": 1.2840021545919742,
      "grad_norm": 0.6465878782285631,
      "learning_rate": 2.8431815801364717e-06,
      "loss": 0.31263012886047364,
      "num_input_tokens_seen": 981255888,
      "step": 19070,
      "train_runtime": 117410.7944,
      "train_tokens_per_second": 8357.459
    },
    {
      "epoch": 1.2843388095879342,
      "grad_norm": 0.7263430108285459,
      "learning_rate": 2.8407964487358906e-06,
      "loss": 0.31687126159667967,
      "num_input_tokens_seen": 981506752,
      "step": 19075,
      "train_runtime": 117445.9551,
      "train_tokens_per_second": 8357.093
    },
    {
      "epoch": 1.2846754645838945,
      "grad_norm": 0.6925386129735001,
      "learning_rate": 2.838411921149702e-06,
      "loss": 0.304840612411499,
      "num_input_tokens_seen": 981745224,
      "step": 19080,
      "train_runtime": 117475.7257,
      "train_tokens_per_second": 8357.005
    },
    {
      "epoch": 1.2850121195798545,
      "grad_norm": 0.6321300827828227,
      "learning_rate": 2.836027998044722e-06,
      "loss": 0.2880533695220947,
      "num_input_tokens_seen": 981991144,
      "step": 19085,
      "train_runtime": 117501.5471,
      "train_tokens_per_second": 8357.261
    },
    {
      "epoch": 1.2853487745758148,
      "grad_norm": 0.6850136334122487,
      "learning_rate": 2.8336446800876143e-06,
      "loss": 0.3066259384155273,
      "num_input_tokens_seen": 982221456,
      "step": 19090,
      "train_runtime": 117532.078,
      "train_tokens_per_second": 8357.05
    },
    {
      "epoch": 1.2856854295717748,
      "grad_norm": 0.8101642453253817,
      "learning_rate": 2.8312619679448637e-06,
      "loss": 0.2940936326980591,
      "num_input_tokens_seen": 982483192,
      "step": 19095,
      "train_runtime": 117559.2533,
      "train_tokens_per_second": 8357.345
    },
    {
      "epoch": 1.286022084567735,
      "grad_norm": 0.6769730922302745,
      "learning_rate": 2.8288798622827884e-06,
      "loss": 0.31775078773498533,
      "num_input_tokens_seen": 982700896,
      "step": 19100,
      "train_runtime": 117587.833,
      "train_tokens_per_second": 8357.165
    },
    {
      "epoch": 1.286358739563695,
      "grad_norm": 0.8532360888023992,
      "learning_rate": 2.826498363767536e-06,
      "loss": 0.31343779563903806,
      "num_input_tokens_seen": 982924264,
      "step": 19105,
      "train_runtime": 117616.377,
      "train_tokens_per_second": 8357.036
    },
    {
      "epoch": 1.2866953945596553,
      "grad_norm": 0.7498311077592399,
      "learning_rate": 2.8241174730650855e-06,
      "loss": 0.31284489631652834,
      "num_input_tokens_seen": 983199472,
      "step": 19110,
      "train_runtime": 117647.4786,
      "train_tokens_per_second": 8357.166
    },
    {
      "epoch": 1.2870320495556153,
      "grad_norm": 0.6822164508006636,
      "learning_rate": 2.821737190841244e-06,
      "loss": 0.2972156047821045,
      "num_input_tokens_seen": 983455856,
      "step": 19115,
      "train_runtime": 117681.6369,
      "train_tokens_per_second": 8356.919
    },
    {
      "epoch": 1.2873687045515756,
      "grad_norm": 0.6165988724968101,
      "learning_rate": 2.8193575177616537e-06,
      "loss": 0.2881195068359375,
      "num_input_tokens_seen": 983713392,
      "step": 19120,
      "train_runtime": 117715.4632,
      "train_tokens_per_second": 8356.705
    },
    {
      "epoch": 1.2877053595475356,
      "grad_norm": 0.6621432980398686,
      "learning_rate": 2.816978454491777e-06,
      "loss": 0.2784890651702881,
      "num_input_tokens_seen": 983981616,
      "step": 19125,
      "train_runtime": 117747.1249,
      "train_tokens_per_second": 8356.736
    },
    {
      "epoch": 1.2880420145434959,
      "grad_norm": 0.6662292545497164,
      "learning_rate": 2.814600001696919e-06,
      "loss": 0.31694707870483396,
      "num_input_tokens_seen": 984238344,
      "step": 19130,
      "train_runtime": 117778.835,
      "train_tokens_per_second": 8356.666
    },
    {
      "epoch": 1.288378669539456,
      "grad_norm": 0.5879152365069383,
      "learning_rate": 2.812222160042196e-06,
      "loss": 0.30867748260498046,
      "num_input_tokens_seen": 984508520,
      "step": 19135,
      "train_runtime": 117806.8377,
      "train_tokens_per_second": 8356.973
    },
    {
      "epoch": 1.2887153245354162,
      "grad_norm": 0.668309883808289,
      "learning_rate": 2.809844930192573e-06,
      "loss": 0.31049792766571044,
      "num_input_tokens_seen": 984753016,
      "step": 19140,
      "train_runtime": 117836.4566,
      "train_tokens_per_second": 8356.947
    },
    {
      "epoch": 1.2890519795313762,
      "grad_norm": 0.6970060901774897,
      "learning_rate": 2.807468312812831e-06,
      "loss": 0.3239492654800415,
      "num_input_tokens_seen": 985015504,
      "step": 19145,
      "train_runtime": 117866.6165,
      "train_tokens_per_second": 8357.036
    },
    {
      "epoch": 1.2893886345273364,
      "grad_norm": 0.5818147879400498,
      "learning_rate": 2.805092308567584e-06,
      "loss": 0.2860324144363403,
      "num_input_tokens_seen": 985287512,
      "step": 19150,
      "train_runtime": 117899.1601,
      "train_tokens_per_second": 8357.036
    },
    {
      "epoch": 1.2897252895232965,
      "grad_norm": 0.6442690296288094,
      "learning_rate": 2.802716918121273e-06,
      "loss": 0.29507813453674314,
      "num_input_tokens_seen": 985544288,
      "step": 19155,
      "train_runtime": 117926.842,
      "train_tokens_per_second": 8357.252
    },
    {
      "epoch": 1.2900619445192567,
      "grad_norm": 0.6236217869956391,
      "learning_rate": 2.8003421421381704e-06,
      "loss": 0.29388861656188964,
      "num_input_tokens_seen": 985824632,
      "step": 19160,
      "train_runtime": 117959.2179,
      "train_tokens_per_second": 8357.334
    },
    {
      "epoch": 1.2903985995152167,
      "grad_norm": 0.72230255461879,
      "learning_rate": 2.797967981282371e-06,
      "loss": 0.32345309257507326,
      "num_input_tokens_seen": 986075416,
      "step": 19165,
      "train_runtime": 117988.853,
      "train_tokens_per_second": 8357.361
    },
    {
      "epoch": 1.290735254511177,
      "grad_norm": 0.6204021386066758,
      "learning_rate": 2.795594436217808e-06,
      "loss": 0.28948545455932617,
      "num_input_tokens_seen": 986330088,
      "step": 19170,
      "train_runtime": 118023.5855,
      "train_tokens_per_second": 8357.059
    },
    {
      "epoch": 1.291071909507137,
      "grad_norm": 0.5758355112834527,
      "learning_rate": 2.7932215076082277e-06,
      "loss": 0.29665560722351075,
      "num_input_tokens_seen": 986573240,
      "step": 19175,
      "train_runtime": 118051.6459,
      "train_tokens_per_second": 8357.132
    },
    {
      "epoch": 1.2914085645030973,
      "grad_norm": 0.5680949777505496,
      "learning_rate": 2.790849196117219e-06,
      "loss": 0.3050487518310547,
      "num_input_tokens_seen": 986847288,
      "step": 19180,
      "train_runtime": 118081.9686,
      "train_tokens_per_second": 8357.307
    },
    {
      "epoch": 1.2917452194990573,
      "grad_norm": 0.6133835491636843,
      "learning_rate": 2.7884775024081856e-06,
      "loss": 0.2952148914337158,
      "num_input_tokens_seen": 987099696,
      "step": 19185,
      "train_runtime": 118111.1662,
      "train_tokens_per_second": 8357.378
    },
    {
      "epoch": 1.2920818744950175,
      "grad_norm": 0.6879901260589167,
      "learning_rate": 2.786106427144367e-06,
      "loss": 0.2914983034133911,
      "num_input_tokens_seen": 987357128,
      "step": 19190,
      "train_runtime": 118146.9736,
      "train_tokens_per_second": 8357.024
    },
    {
      "epoch": 1.2924185294909776,
      "grad_norm": 0.7234167988529141,
      "learning_rate": 2.783735970988828e-06,
      "loss": 0.27358303070068357,
      "num_input_tokens_seen": 987608608,
      "step": 19195,
      "train_runtime": 118179.3068,
      "train_tokens_per_second": 8356.866
    },
    {
      "epoch": 1.2927551844869378,
      "grad_norm": 0.6249886325539862,
      "learning_rate": 2.7813661346044566e-06,
      "loss": 0.30453245639801024,
      "num_input_tokens_seen": 987857424,
      "step": 19200,
      "train_runtime": 118215.2578,
      "train_tokens_per_second": 8356.429
    },
    {
      "epoch": 1.2930918394828979,
      "grad_norm": 0.7470808605285114,
      "learning_rate": 2.7789969186539713e-06,
      "loss": 0.30964767932891846,
      "num_input_tokens_seen": 988087528,
      "step": 19205,
      "train_runtime": 118241.1197,
      "train_tokens_per_second": 8356.547
    },
    {
      "epoch": 1.293428494478858,
      "grad_norm": 0.7026190085528019,
      "learning_rate": 2.776628323799915e-06,
      "loss": 0.3154508113861084,
      "num_input_tokens_seen": 988344880,
      "step": 19210,
      "train_runtime": 118270.0473,
      "train_tokens_per_second": 8356.68
    },
    {
      "epoch": 1.2937651494748181,
      "grad_norm": 0.7254134796701002,
      "learning_rate": 2.774260350704657e-06,
      "loss": 0.3222126245498657,
      "num_input_tokens_seen": 988600432,
      "step": 19215,
      "train_runtime": 118299.0953,
      "train_tokens_per_second": 8356.788
    },
    {
      "epoch": 1.2941018044707784,
      "grad_norm": 0.6820085556447519,
      "learning_rate": 2.771893000030398e-06,
      "loss": 0.3145334243774414,
      "num_input_tokens_seen": 988853776,
      "step": 19220,
      "train_runtime": 118333.436,
      "train_tokens_per_second": 8356.504
    },
    {
      "epoch": 1.2944384594667384,
      "grad_norm": 0.6256594268134094,
      "learning_rate": 2.7695262724391526e-06,
      "loss": 0.28402533531188967,
      "num_input_tokens_seen": 989103216,
      "step": 19225,
      "train_runtime": 118359.8422,
      "train_tokens_per_second": 8356.747
    },
    {
      "epoch": 1.2947751144626987,
      "grad_norm": 0.6837357076653562,
      "learning_rate": 2.767160168592777e-06,
      "loss": 0.33648340702056884,
      "num_input_tokens_seen": 989368056,
      "step": 19230,
      "train_runtime": 118390.9004,
      "train_tokens_per_second": 8356.791
    },
    {
      "epoch": 1.2951117694586587,
      "grad_norm": 0.6564620490404995,
      "learning_rate": 2.7647946891529355e-06,
      "loss": 0.3208168506622314,
      "num_input_tokens_seen": 989598456,
      "step": 19235,
      "train_runtime": 118419.2522,
      "train_tokens_per_second": 8356.736
    },
    {
      "epoch": 1.295448424454619,
      "grad_norm": 0.6493074987182532,
      "learning_rate": 2.7624298347811342e-06,
      "loss": 0.3055669069290161,
      "num_input_tokens_seen": 989825808,
      "step": 19240,
      "train_runtime": 118452.0843,
      "train_tokens_per_second": 8356.339
    },
    {
      "epoch": 1.295785079450579,
      "grad_norm": 0.6823857793111066,
      "learning_rate": 2.7600656061386945e-06,
      "loss": 0.3116645812988281,
      "num_input_tokens_seen": 990107072,
      "step": 19245,
      "train_runtime": 118484.2984,
      "train_tokens_per_second": 8356.441
    },
    {
      "epoch": 1.2961217344465392,
      "grad_norm": 0.647498268750319,
      "learning_rate": 2.757702003886765e-06,
      "loss": 0.28616149425506593,
      "num_input_tokens_seen": 990371504,
      "step": 19250,
      "train_runtime": 118515.3811,
      "train_tokens_per_second": 8356.481
    },
    {
      "epoch": 1.2964583894424992,
      "grad_norm": 0.7517616171612741,
      "learning_rate": 2.7553390286863205e-06,
      "loss": 0.2865654945373535,
      "num_input_tokens_seen": 990607496,
      "step": 19255,
      "train_runtime": 118543.2685,
      "train_tokens_per_second": 8356.506
    },
    {
      "epoch": 1.2967950444384595,
      "grad_norm": 0.6775582833326768,
      "learning_rate": 2.752976681198159e-06,
      "loss": 0.301178503036499,
      "num_input_tokens_seen": 990879256,
      "step": 19260,
      "train_runtime": 118575.5816,
      "train_tokens_per_second": 8356.52
    },
    {
      "epoch": 1.2971316994344195,
      "grad_norm": 0.7298537188010265,
      "learning_rate": 2.750614962082901e-06,
      "loss": 0.28229522705078125,
      "num_input_tokens_seen": 991135432,
      "step": 19265,
      "train_runtime": 118605.4988,
      "train_tokens_per_second": 8356.572
    },
    {
      "epoch": 1.2974683544303798,
      "grad_norm": 0.6326491610433641,
      "learning_rate": 2.748253872001e-06,
      "loss": 0.30921053886413574,
      "num_input_tokens_seen": 991398688,
      "step": 19270,
      "train_runtime": 118634.2023,
      "train_tokens_per_second": 8356.77
    },
    {
      "epoch": 1.2978050094263398,
      "grad_norm": 0.7521673825859909,
      "learning_rate": 2.74589341161272e-06,
      "loss": 0.29870266914367677,
      "num_input_tokens_seen": 991619392,
      "step": 19275,
      "train_runtime": 118665.535,
      "train_tokens_per_second": 8356.423
    },
    {
      "epoch": 1.2981416644223,
      "grad_norm": 0.7240697698369928,
      "learning_rate": 2.743533581578163e-06,
      "loss": 0.30806636810302734,
      "num_input_tokens_seen": 991861928,
      "step": 19280,
      "train_runtime": 118696.8634,
      "train_tokens_per_second": 8356.261
    },
    {
      "epoch": 1.29847831941826,
      "grad_norm": 0.6702630066931201,
      "learning_rate": 2.7411743825572407e-06,
      "loss": 0.3455987453460693,
      "num_input_tokens_seen": 992147144,
      "step": 19285,
      "train_runtime": 118730.3688,
      "train_tokens_per_second": 8356.305
    },
    {
      "epoch": 1.2988149744142203,
      "grad_norm": 0.5750050982813291,
      "learning_rate": 2.7388158152097015e-06,
      "loss": 0.3006103038787842,
      "num_input_tokens_seen": 992426104,
      "step": 19290,
      "train_runtime": 118759.2808,
      "train_tokens_per_second": 8356.619
    },
    {
      "epoch": 1.2991516294101804,
      "grad_norm": 0.6048075222102912,
      "learning_rate": 2.736457880195109e-06,
      "loss": 0.3133401870727539,
      "num_input_tokens_seen": 992668968,
      "step": 19295,
      "train_runtime": 118790.6752,
      "train_tokens_per_second": 8356.455
    },
    {
      "epoch": 1.2994882844061406,
      "grad_norm": 0.6725738827393627,
      "learning_rate": 2.7341005781728523e-06,
      "loss": 0.3334936618804932,
      "num_input_tokens_seen": 992934328,
      "step": 19300,
      "train_runtime": 118822.3268,
      "train_tokens_per_second": 8356.463
    },
    {
      "epoch": 1.2998249394021006,
      "grad_norm": 0.6623206309421017,
      "learning_rate": 2.7317439098021444e-06,
      "loss": 0.31380429267883303,
      "num_input_tokens_seen": 993192296,
      "step": 19305,
      "train_runtime": 118853.591,
      "train_tokens_per_second": 8356.435
    },
    {
      "epoch": 1.300161594398061,
      "grad_norm": 0.7176620777271376,
      "learning_rate": 2.7293878757420185e-06,
      "loss": 0.3134188413619995,
      "num_input_tokens_seen": 993440528,
      "step": 19310,
      "train_runtime": 118884.773,
      "train_tokens_per_second": 8356.331
    },
    {
      "epoch": 1.3004982493940211,
      "grad_norm": 0.6200238344922845,
      "learning_rate": 2.7270324766513306e-06,
      "loss": 0.3019558429718018,
      "num_input_tokens_seen": 993725016,
      "step": 19315,
      "train_runtime": 118918.193,
      "train_tokens_per_second": 8356.375
    },
    {
      "epoch": 1.3008349043899812,
      "grad_norm": 0.6554903562285915,
      "learning_rate": 2.724677713188767e-06,
      "loss": 0.30286622047424316,
      "num_input_tokens_seen": 993952616,
      "step": 19320,
      "train_runtime": 118945.8392,
      "train_tokens_per_second": 8356.346
    },
    {
      "epoch": 1.3011715593859412,
      "grad_norm": 0.676608743398712,
      "learning_rate": 2.722323586012821e-06,
      "loss": 0.3134955406188965,
      "num_input_tokens_seen": 994198688,
      "step": 19325,
      "train_runtime": 118977.5211,
      "train_tokens_per_second": 8356.189
    },
    {
      "epoch": 1.3015082143819015,
      "grad_norm": 0.7397686381572275,
      "learning_rate": 2.7199700957818252e-06,
      "loss": 0.29197895526885986,
      "num_input_tokens_seen": 994455944,
      "step": 19330,
      "train_runtime": 119005.3115,
      "train_tokens_per_second": 8356.4
    },
    {
      "epoch": 1.3018448693778617,
      "grad_norm": 0.8095738445022868,
      "learning_rate": 2.717617243153918e-06,
      "loss": 0.2962174892425537,
      "num_input_tokens_seen": 994693104,
      "step": 19335,
      "train_runtime": 119035.0953,
      "train_tokens_per_second": 8356.301
    },
    {
      "epoch": 1.3021815243738217,
      "grad_norm": 0.6936079586720052,
      "learning_rate": 2.715265028787073e-06,
      "loss": 0.31534805297851565,
      "num_input_tokens_seen": 994934824,
      "step": 19340,
      "train_runtime": 119066.243,
      "train_tokens_per_second": 8356.145
    },
    {
      "epoch": 1.3025181793697818,
      "grad_norm": 0.7667507794134386,
      "learning_rate": 2.7129134533390765e-06,
      "loss": 0.33009459972381594,
      "num_input_tokens_seen": 995215112,
      "step": 19345,
      "train_runtime": 119099.2839,
      "train_tokens_per_second": 8356.18
    },
    {
      "epoch": 1.302854834365742,
      "grad_norm": 0.649247270557763,
      "learning_rate": 2.7105625174675404e-06,
      "loss": 0.28858771324157717,
      "num_input_tokens_seen": 995491176,
      "step": 19350,
      "train_runtime": 119129.1765,
      "train_tokens_per_second": 8356.401
    },
    {
      "epoch": 1.3031914893617023,
      "grad_norm": 0.6731169967091668,
      "learning_rate": 2.7082122218298957e-06,
      "loss": 0.2886997699737549,
      "num_input_tokens_seen": 995744776,
      "step": 19355,
      "train_runtime": 119156.5502,
      "train_tokens_per_second": 8356.61
    },
    {
      "epoch": 1.3035281443576623,
      "grad_norm": 0.6808450990929112,
      "learning_rate": 2.7058625670833947e-06,
      "loss": 0.3042797088623047,
      "num_input_tokens_seen": 996011192,
      "step": 19360,
      "train_runtime": 119188.4774,
      "train_tokens_per_second": 8356.606
    },
    {
      "epoch": 1.3038647993536223,
      "grad_norm": 0.6247622587186782,
      "learning_rate": 2.70351355388511e-06,
      "loss": 0.29733991622924805,
      "num_input_tokens_seen": 996268328,
      "step": 19365,
      "train_runtime": 119216.158,
      "train_tokens_per_second": 8356.823
    },
    {
      "epoch": 1.3042014543495826,
      "grad_norm": 0.6807647687841835,
      "learning_rate": 2.7011651828919407e-06,
      "loss": 0.3124992847442627,
      "num_input_tokens_seen": 996540664,
      "step": 19370,
      "train_runtime": 119244.3897,
      "train_tokens_per_second": 8357.128
    },
    {
      "epoch": 1.3045381093455428,
      "grad_norm": 0.6274046614570492,
      "learning_rate": 2.698817454760593e-06,
      "loss": 0.30438709259033203,
      "num_input_tokens_seen": 996810264,
      "step": 19375,
      "train_runtime": 119278.1495,
      "train_tokens_per_second": 8357.023
    },
    {
      "epoch": 1.3048747643415028,
      "grad_norm": 0.615394552609364,
      "learning_rate": 2.696470370147611e-06,
      "loss": 0.31673617362976075,
      "num_input_tokens_seen": 997070664,
      "step": 19380,
      "train_runtime": 119310.4261,
      "train_tokens_per_second": 8356.945
    },
    {
      "epoch": 1.3052114193374629,
      "grad_norm": 0.6880109474416504,
      "learning_rate": 2.69412392970934e-06,
      "loss": 0.32934131622314455,
      "num_input_tokens_seen": 997331848,
      "step": 19385,
      "train_runtime": 119345.8004,
      "train_tokens_per_second": 8356.656
    },
    {
      "epoch": 1.3055480743334231,
      "grad_norm": 0.592347607846234,
      "learning_rate": 2.6917781341019615e-06,
      "loss": 0.29868111610412595,
      "num_input_tokens_seen": 997589376,
      "step": 19390,
      "train_runtime": 119373.9914,
      "train_tokens_per_second": 8356.84
    },
    {
      "epoch": 1.3058847293293834,
      "grad_norm": 0.6929448316162584,
      "learning_rate": 2.689432983981467e-06,
      "loss": 0.29558582305908204,
      "num_input_tokens_seen": 997842080,
      "step": 19395,
      "train_runtime": 119404.5625,
      "train_tokens_per_second": 8356.817
    },
    {
      "epoch": 1.3062213843253434,
      "grad_norm": 0.6713159273765152,
      "learning_rate": 2.6870884800036713e-06,
      "loss": 0.30768465995788574,
      "num_input_tokens_seen": 998087504,
      "step": 19400,
      "train_runtime": 119435.5124,
      "train_tokens_per_second": 8356.706
    },
    {
      "epoch": 1.3065580393213034,
      "grad_norm": 0.7420024228328281,
      "learning_rate": 2.684744622824207e-06,
      "loss": 0.3182835102081299,
      "num_input_tokens_seen": 998328168,
      "step": 19405,
      "train_runtime": 119463.8894,
      "train_tokens_per_second": 8356.736
    },
    {
      "epoch": 1.3068946943172637,
      "grad_norm": 0.6897953751308821,
      "learning_rate": 2.6824014130985255e-06,
      "loss": 0.30908761024475095,
      "num_input_tokens_seen": 998584384,
      "step": 19410,
      "train_runtime": 119499.7797,
      "train_tokens_per_second": 8356.37
    },
    {
      "epoch": 1.307231349313224,
      "grad_norm": 0.6329770504278176,
      "learning_rate": 2.680058851481896e-06,
      "loss": 0.2973031759262085,
      "num_input_tokens_seen": 998848384,
      "step": 19415,
      "train_runtime": 119530.2494,
      "train_tokens_per_second": 8356.449
    },
    {
      "epoch": 1.307568004309184,
      "grad_norm": 0.7921819062996943,
      "learning_rate": 2.6777169386294154e-06,
      "loss": 0.3180563926696777,
      "num_input_tokens_seen": 999107280,
      "step": 19420,
      "train_runtime": 119560.6492,
      "train_tokens_per_second": 8356.489
    },
    {
      "epoch": 1.307904659305144,
      "grad_norm": 0.6403605215835085,
      "learning_rate": 2.6753756751959846e-06,
      "loss": 0.29218928813934325,
      "num_input_tokens_seen": 999385872,
      "step": 19425,
      "train_runtime": 119588.9707,
      "train_tokens_per_second": 8356.84
    },
    {
      "epoch": 1.3082413143011042,
      "grad_norm": 0.5815131748356693,
      "learning_rate": 2.6730350618363367e-06,
      "loss": 0.2993615627288818,
      "num_input_tokens_seen": 999651144,
      "step": 19430,
      "train_runtime": 119624.4018,
      "train_tokens_per_second": 8356.582
    },
    {
      "epoch": 1.3085779692970645,
      "grad_norm": 0.6177363450708162,
      "learning_rate": 2.6706950992050097e-06,
      "loss": 0.3005399703979492,
      "num_input_tokens_seen": 999920976,
      "step": 19435,
      "train_runtime": 119657.9618,
      "train_tokens_per_second": 8356.493
    },
    {
      "epoch": 1.3089146242930245,
      "grad_norm": 0.6783208598933199,
      "learning_rate": 2.6683557879563726e-06,
      "loss": 0.32118470668792726,
      "num_input_tokens_seen": 1000188680,
      "step": 19440,
      "train_runtime": 119691.7627,
      "train_tokens_per_second": 8356.37
    },
    {
      "epoch": 1.3092512792889845,
      "grad_norm": 0.6780072787882727,
      "learning_rate": 2.6660171287446045e-06,
      "loss": 0.3425604820251465,
      "num_input_tokens_seen": 1000465280,
      "step": 19445,
      "train_runtime": 119720.5169,
      "train_tokens_per_second": 8356.674
    },
    {
      "epoch": 1.3095879342849448,
      "grad_norm": 0.589389102702863,
      "learning_rate": 2.6636791222237034e-06,
      "loss": 0.31085567474365233,
      "num_input_tokens_seen": 1000723600,
      "step": 19450,
      "train_runtime": 119755.5139,
      "train_tokens_per_second": 8356.389
    },
    {
      "epoch": 1.309924589280905,
      "grad_norm": 0.6440215383541844,
      "learning_rate": 2.6613417690474864e-06,
      "loss": 0.3136088132858276,
      "num_input_tokens_seen": 1000998064,
      "step": 19455,
      "train_runtime": 119787.6401,
      "train_tokens_per_second": 8356.439
    },
    {
      "epoch": 1.310261244276865,
      "grad_norm": 0.5800586685436736,
      "learning_rate": 2.6590050698695856e-06,
      "loss": 0.30886077880859375,
      "num_input_tokens_seen": 1001258768,
      "step": 19460,
      "train_runtime": 119821.7322,
      "train_tokens_per_second": 8356.237
    },
    {
      "epoch": 1.310597899272825,
      "grad_norm": 0.644019521819607,
      "learning_rate": 2.6566690253434508e-06,
      "loss": 0.31041090488433837,
      "num_input_tokens_seen": 1001541240,
      "step": 19465,
      "train_runtime": 119854.3654,
      "train_tokens_per_second": 8356.318
    },
    {
      "epoch": 1.3109345542687854,
      "grad_norm": 0.682433195726019,
      "learning_rate": 2.654333636122354e-06,
      "loss": 0.2880502939224243,
      "num_input_tokens_seen": 1001810984,
      "step": 19470,
      "train_runtime": 119884.2783,
      "train_tokens_per_second": 8356.483
    },
    {
      "epoch": 1.3112712092647456,
      "grad_norm": 0.6964754341215903,
      "learning_rate": 2.6519989028593728e-06,
      "loss": 0.3222483158111572,
      "num_input_tokens_seen": 1002089304,
      "step": 19475,
      "train_runtime": 119916.9039,
      "train_tokens_per_second": 8356.531
    },
    {
      "epoch": 1.3116078642607056,
      "grad_norm": 0.7128359990469624,
      "learning_rate": 2.649664826207414e-06,
      "loss": 0.3114827394485474,
      "num_input_tokens_seen": 1002337576,
      "step": 19480,
      "train_runtime": 119949.1743,
      "train_tokens_per_second": 8356.352
    },
    {
      "epoch": 1.3119445192566657,
      "grad_norm": 0.712887291435525,
      "learning_rate": 2.647331406819189e-06,
      "loss": 0.3266038417816162,
      "num_input_tokens_seen": 1002585952,
      "step": 19485,
      "train_runtime": 119980.4859,
      "train_tokens_per_second": 8356.242
    },
    {
      "epoch": 1.312281174252626,
      "grad_norm": 0.6232668968141699,
      "learning_rate": 2.644998645347236e-06,
      "loss": 0.2959095001220703,
      "num_input_tokens_seen": 1002827944,
      "step": 19490,
      "train_runtime": 120014.117,
      "train_tokens_per_second": 8355.917
    },
    {
      "epoch": 1.3126178292485862,
      "grad_norm": 0.6475840950549915,
      "learning_rate": 2.6426665424439025e-06,
      "loss": 0.30884976387023927,
      "num_input_tokens_seen": 1003090648,
      "step": 19495,
      "train_runtime": 120042.5864,
      "train_tokens_per_second": 8356.123
    },
    {
      "epoch": 1.3129544842445462,
      "grad_norm": 0.7016394213543232,
      "learning_rate": 2.6403350987613526e-06,
      "loss": 0.31875271797180177,
      "num_input_tokens_seen": 1003356280,
      "step": 19500,
      "train_runtime": 120076.6656,
      "train_tokens_per_second": 8355.964
    },
    {
      "epoch": 1.3132911392405062,
      "grad_norm": 0.6842468817532963,
      "learning_rate": 2.6380043149515676e-06,
      "loss": 0.29985475540161133,
      "num_input_tokens_seen": 1003611456,
      "step": 19505,
      "train_runtime": 120109.2582,
      "train_tokens_per_second": 8355.821
    },
    {
      "epoch": 1.3136277942364665,
      "grad_norm": 0.6778969582621874,
      "learning_rate": 2.6356741916663453e-06,
      "loss": 0.31459214687347414,
      "num_input_tokens_seen": 1003874248,
      "step": 19510,
      "train_runtime": 120134.6892,
      "train_tokens_per_second": 8356.24
    },
    {
      "epoch": 1.3139644492324267,
      "grad_norm": 0.7153785008495804,
      "learning_rate": 2.6333447295572933e-06,
      "loss": 0.310893177986145,
      "num_input_tokens_seen": 1004139408,
      "step": 19515,
      "train_runtime": 120166.7323,
      "train_tokens_per_second": 8356.218
    },
    {
      "epoch": 1.3143011042283868,
      "grad_norm": 0.6427454408927988,
      "learning_rate": 2.631015929275845e-06,
      "loss": 0.3071840047836304,
      "num_input_tokens_seen": 1004381296,
      "step": 19520,
      "train_runtime": 120192.6598,
      "train_tokens_per_second": 8356.428
    },
    {
      "epoch": 1.3146377592243468,
      "grad_norm": 0.7297145498030481,
      "learning_rate": 2.628687791473234e-06,
      "loss": 0.29586875438690186,
      "num_input_tokens_seen": 1004667000,
      "step": 19525,
      "train_runtime": 120226.4835,
      "train_tokens_per_second": 8356.453
    },
    {
      "epoch": 1.314974414220307,
      "grad_norm": 0.7719324825633266,
      "learning_rate": 2.6263603168005224e-06,
      "loss": 0.2903263568878174,
      "num_input_tokens_seen": 1004916768,
      "step": 19530,
      "train_runtime": 120251.8161,
      "train_tokens_per_second": 8356.77
    },
    {
      "epoch": 1.3153110692162673,
      "grad_norm": 0.5997586976075788,
      "learning_rate": 2.6240335059085797e-06,
      "loss": 0.30548701286315916,
      "num_input_tokens_seen": 1005179456,
      "step": 19535,
      "train_runtime": 120285.8791,
      "train_tokens_per_second": 8356.587
    },
    {
      "epoch": 1.3156477242122273,
      "grad_norm": 0.6583159478843866,
      "learning_rate": 2.62170735944809e-06,
      "loss": 0.3155561685562134,
      "num_input_tokens_seen": 1005447312,
      "step": 19540,
      "train_runtime": 120322.6579,
      "train_tokens_per_second": 8356.259
    },
    {
      "epoch": 1.3159843792081873,
      "grad_norm": 0.6047863342412795,
      "learning_rate": 2.6193818780695537e-06,
      "loss": 0.29338994026184084,
      "num_input_tokens_seen": 1005726784,
      "step": 19545,
      "train_runtime": 120354.98,
      "train_tokens_per_second": 8356.337
    },
    {
      "epoch": 1.3163210342041476,
      "grad_norm": 0.6969856619440332,
      "learning_rate": 2.6170570624232837e-06,
      "loss": 0.3203129291534424,
      "num_input_tokens_seen": 1005938392,
      "step": 19550,
      "train_runtime": 120384.874,
      "train_tokens_per_second": 8356.02
    },
    {
      "epoch": 1.3166576892001078,
      "grad_norm": 0.7156789600264727,
      "learning_rate": 2.6147329131594078e-06,
      "loss": 0.27625370025634766,
      "num_input_tokens_seen": 1006209304,
      "step": 19555,
      "train_runtime": 120414.6022,
      "train_tokens_per_second": 8356.207
    },
    {
      "epoch": 1.3169943441960679,
      "grad_norm": 0.596579722239996,
      "learning_rate": 2.612409430927867e-06,
      "loss": 0.2753139495849609,
      "num_input_tokens_seen": 1006465104,
      "step": 19560,
      "train_runtime": 120441.9491,
      "train_tokens_per_second": 8356.433
    },
    {
      "epoch": 1.317330999192028,
      "grad_norm": 0.750469136376467,
      "learning_rate": 2.610086616378413e-06,
      "loss": 0.28801689147949217,
      "num_input_tokens_seen": 1006715456,
      "step": 19565,
      "train_runtime": 120469.8385,
      "train_tokens_per_second": 8356.577
    },
    {
      "epoch": 1.3176676541879881,
      "grad_norm": 0.5619028382099868,
      "learning_rate": 2.60776447016062e-06,
      "loss": 0.31370973587036133,
      "num_input_tokens_seen": 1006971832,
      "step": 19570,
      "train_runtime": 120500.7306,
      "train_tokens_per_second": 8356.562
    },
    {
      "epoch": 1.3180043091839484,
      "grad_norm": 0.6476164691510053,
      "learning_rate": 2.6054429929238607e-06,
      "loss": 0.31538031101226804,
      "num_input_tokens_seen": 1007240664,
      "step": 19575,
      "train_runtime": 120532.956,
      "train_tokens_per_second": 8356.558
    },
    {
      "epoch": 1.3183409641799084,
      "grad_norm": 0.573713592658333,
      "learning_rate": 2.603122185317335e-06,
      "loss": 0.28681769371032717,
      "num_input_tokens_seen": 1007471848,
      "step": 19580,
      "train_runtime": 120559.4147,
      "train_tokens_per_second": 8356.642
    },
    {
      "epoch": 1.3186776191758685,
      "grad_norm": 0.6369265271870169,
      "learning_rate": 2.6008020479900465e-06,
      "loss": 0.2564453125,
      "num_input_tokens_seen": 1007722448,
      "step": 19585,
      "train_runtime": 120592.019,
      "train_tokens_per_second": 8356.461
    },
    {
      "epoch": 1.3190142741718287,
      "grad_norm": 0.5454208706650109,
      "learning_rate": 2.598482581590815e-06,
      "loss": 0.29861650466918943,
      "num_input_tokens_seen": 1008007240,
      "step": 19590,
      "train_runtime": 120623.7139,
      "train_tokens_per_second": 8356.626
    },
    {
      "epoch": 1.319350929167789,
      "grad_norm": 0.6115300762288272,
      "learning_rate": 2.5961637867682715e-06,
      "loss": 0.2846058368682861,
      "num_input_tokens_seen": 1008277976,
      "step": 19595,
      "train_runtime": 120654.5471,
      "train_tokens_per_second": 8356.734
    },
    {
      "epoch": 1.319687584163749,
      "grad_norm": 0.6396236428574612,
      "learning_rate": 2.5938456641708598e-06,
      "loss": 0.2912113666534424,
      "num_input_tokens_seen": 1008543088,
      "step": 19600,
      "train_runtime": 120685.7319,
      "train_tokens_per_second": 8356.772
    },
    {
      "epoch": 1.320024239159709,
      "grad_norm": 0.6727860985762553,
      "learning_rate": 2.591528214446836e-06,
      "loss": 0.3102696895599365,
      "num_input_tokens_seen": 1008825008,
      "step": 19605,
      "train_runtime": 120716.9019,
      "train_tokens_per_second": 8356.949
    },
    {
      "epoch": 1.3203608941556693,
      "grad_norm": 0.6264920169484289,
      "learning_rate": 2.5892114382442664e-06,
      "loss": 0.31895108222961427,
      "num_input_tokens_seen": 1009076792,
      "step": 19610,
      "train_runtime": 120747.3636,
      "train_tokens_per_second": 8356.926
    },
    {
      "epoch": 1.3206975491516295,
      "grad_norm": 0.5425160199128981,
      "learning_rate": 2.5868953362110287e-06,
      "loss": 0.2934732913970947,
      "num_input_tokens_seen": 1009353184,
      "step": 19615,
      "train_runtime": 120779.9823,
      "train_tokens_per_second": 8356.958
    },
    {
      "epoch": 1.3210342041475895,
      "grad_norm": 0.6680101683702643,
      "learning_rate": 2.5845799089948196e-06,
      "loss": 0.29140944480895997,
      "num_input_tokens_seen": 1009595792,
      "step": 19620,
      "train_runtime": 120809.7411,
      "train_tokens_per_second": 8356.907
    },
    {
      "epoch": 1.3213708591435496,
      "grad_norm": 0.6200698380267011,
      "learning_rate": 2.5822651572431323e-06,
      "loss": 0.2900580167770386,
      "num_input_tokens_seen": 1009871328,
      "step": 19625,
      "train_runtime": 120838.9431,
      "train_tokens_per_second": 8357.168
    },
    {
      "epoch": 1.3217075141395098,
      "grad_norm": 0.7548517779602734,
      "learning_rate": 2.5799510816032867e-06,
      "loss": 0.29996702671051023,
      "num_input_tokens_seen": 1010078080,
      "step": 19630,
      "train_runtime": 120868.8932,
      "train_tokens_per_second": 8356.808
    },
    {
      "epoch": 1.32204416913547,
      "grad_norm": 0.6291695045808948,
      "learning_rate": 2.577637682722404e-06,
      "loss": 0.3069286823272705,
      "num_input_tokens_seen": 1010324760,
      "step": 19635,
      "train_runtime": 120897.1756,
      "train_tokens_per_second": 8356.893
    },
    {
      "epoch": 1.32238082413143,
      "grad_norm": 0.8091177424993845,
      "learning_rate": 2.5753249612474195e-06,
      "loss": 0.3246387243270874,
      "num_input_tokens_seen": 1010566624,
      "step": 19640,
      "train_runtime": 120925.0604,
      "train_tokens_per_second": 8356.966
    },
    {
      "epoch": 1.3227174791273901,
      "grad_norm": 0.6348654394276988,
      "learning_rate": 2.5730129178250774e-06,
      "loss": 0.3150815010070801,
      "num_input_tokens_seen": 1010822688,
      "step": 19645,
      "train_runtime": 120956.1513,
      "train_tokens_per_second": 8356.935
    },
    {
      "epoch": 1.3230541341233504,
      "grad_norm": 0.672483798406663,
      "learning_rate": 2.570701553101933e-06,
      "loss": 0.30807952880859374,
      "num_input_tokens_seen": 1011068040,
      "step": 19650,
      "train_runtime": 120987.7726,
      "train_tokens_per_second": 8356.779
    },
    {
      "epoch": 1.3233907891193106,
      "grad_norm": 0.6732254372956349,
      "learning_rate": 2.568390867724354e-06,
      "loss": 0.3285548210144043,
      "num_input_tokens_seen": 1011336192,
      "step": 19655,
      "train_runtime": 121018.9997,
      "train_tokens_per_second": 8356.838
    },
    {
      "epoch": 1.3237274441152707,
      "grad_norm": 0.7241005772308821,
      "learning_rate": 2.5660808623385143e-06,
      "loss": 0.3077761888504028,
      "num_input_tokens_seen": 1011580976,
      "step": 19660,
      "train_runtime": 121048.7164,
      "train_tokens_per_second": 8356.809
    },
    {
      "epoch": 1.3240640991112307,
      "grad_norm": 0.6075311689792906,
      "learning_rate": 2.5637715375903984e-06,
      "loss": 0.2838373899459839,
      "num_input_tokens_seen": 1011834872,
      "step": 19665,
      "train_runtime": 121079.3834,
      "train_tokens_per_second": 8356.789
    },
    {
      "epoch": 1.324400754107191,
      "grad_norm": 0.7488223287151273,
      "learning_rate": 2.5614628941258075e-06,
      "loss": 0.2925628662109375,
      "num_input_tokens_seen": 1012076824,
      "step": 19670,
      "train_runtime": 121106.8388,
      "train_tokens_per_second": 8356.892
    },
    {
      "epoch": 1.3247374091031512,
      "grad_norm": 0.688867538030082,
      "learning_rate": 2.5591549325903388e-06,
      "loss": 0.324657940864563,
      "num_input_tokens_seen": 1012365840,
      "step": 19675,
      "train_runtime": 121140.6472,
      "train_tokens_per_second": 8356.946
    },
    {
      "epoch": 1.3250740640991112,
      "grad_norm": 0.6652611598163317,
      "learning_rate": 2.5568476536294105e-06,
      "loss": 0.31885063648223877,
      "num_input_tokens_seen": 1012637552,
      "step": 19680,
      "train_runtime": 121172.9323,
      "train_tokens_per_second": 8356.962
    },
    {
      "epoch": 1.3254107190950712,
      "grad_norm": 0.6374336741433521,
      "learning_rate": 2.5545410578882466e-06,
      "loss": 0.30245041847229004,
      "num_input_tokens_seen": 1012911480,
      "step": 19685,
      "train_runtime": 121207.5927,
      "train_tokens_per_second": 8356.832
    },
    {
      "epoch": 1.3257473740910315,
      "grad_norm": 0.7221258397861101,
      "learning_rate": 2.5522351460118776e-06,
      "loss": 0.3004828691482544,
      "num_input_tokens_seen": 1013162968,
      "step": 19690,
      "train_runtime": 121237.3234,
      "train_tokens_per_second": 8356.857
    },
    {
      "epoch": 1.3260840290869917,
      "grad_norm": 0.5602002611993842,
      "learning_rate": 2.549929918645144e-06,
      "loss": 0.3067863941192627,
      "num_input_tokens_seen": 1013440528,
      "step": 19695,
      "train_runtime": 121271.5854,
      "train_tokens_per_second": 8356.785
    },
    {
      "epoch": 1.3264206840829518,
      "grad_norm": 0.6782507829053125,
      "learning_rate": 2.5476253764326963e-06,
      "loss": 0.29280045032501223,
      "num_input_tokens_seen": 1013686000,
      "step": 19700,
      "train_runtime": 121302.2888,
      "train_tokens_per_second": 8356.693
    },
    {
      "epoch": 1.3267573390789118,
      "grad_norm": 0.6619450160000074,
      "learning_rate": 2.5453215200189927e-06,
      "loss": 0.2941920757293701,
      "num_input_tokens_seen": 1013932680,
      "step": 19705,
      "train_runtime": 121334.6008,
      "train_tokens_per_second": 8356.501
    },
    {
      "epoch": 1.327093994074872,
      "grad_norm": 0.6757662890745023,
      "learning_rate": 2.543018350048298e-06,
      "loss": 0.2949738740921021,
      "num_input_tokens_seen": 1014197432,
      "step": 19710,
      "train_runtime": 121363.5817,
      "train_tokens_per_second": 8356.687
    },
    {
      "epoch": 1.3274306490708323,
      "grad_norm": 0.7121998968805081,
      "learning_rate": 2.540715867164685e-06,
      "loss": 0.30037670135498046,
      "num_input_tokens_seen": 1014455104,
      "step": 19715,
      "train_runtime": 121393.2709,
      "train_tokens_per_second": 8356.766
    },
    {
      "epoch": 1.3277673040667923,
      "grad_norm": 0.7092696337549579,
      "learning_rate": 2.538414072012042e-06,
      "loss": 0.30840773582458497,
      "num_input_tokens_seen": 1014688768,
      "step": 19720,
      "train_runtime": 121421.6362,
      "train_tokens_per_second": 8356.738
    },
    {
      "epoch": 1.3281039590627524,
      "grad_norm": 0.7036668637499363,
      "learning_rate": 2.53611296523405e-06,
      "loss": 0.3049553394317627,
      "num_input_tokens_seen": 1014946768,
      "step": 19725,
      "train_runtime": 121451.638,
      "train_tokens_per_second": 8356.798
    },
    {
      "epoch": 1.3284406140587126,
      "grad_norm": 0.6216392746532222,
      "learning_rate": 2.5338125474742127e-06,
      "loss": 0.2998421907424927,
      "num_input_tokens_seen": 1015194720,
      "step": 19730,
      "train_runtime": 121480.1919,
      "train_tokens_per_second": 8356.875
    },
    {
      "epoch": 1.3287772690546729,
      "grad_norm": 0.6668253968620014,
      "learning_rate": 2.531512819375832e-06,
      "loss": 0.29500107765197753,
      "num_input_tokens_seen": 1015444216,
      "step": 19735,
      "train_runtime": 121511.8196,
      "train_tokens_per_second": 8356.753
    },
    {
      "epoch": 1.3291139240506329,
      "grad_norm": 0.6269395752666356,
      "learning_rate": 2.52921378158202e-06,
      "loss": 0.2917795658111572,
      "num_input_tokens_seen": 1015703800,
      "step": 19740,
      "train_runtime": 121544.8142,
      "train_tokens_per_second": 8356.62
    },
    {
      "epoch": 1.329450579046593,
      "grad_norm": 0.7543847610817661,
      "learning_rate": 2.5269154347356957e-06,
      "loss": 0.2789111614227295,
      "num_input_tokens_seen": 1015965184,
      "step": 19745,
      "train_runtime": 121573.947,
      "train_tokens_per_second": 8356.767
    },
    {
      "epoch": 1.3297872340425532,
      "grad_norm": 0.766292208268358,
      "learning_rate": 2.5246177794795834e-06,
      "loss": 0.3052054882049561,
      "num_input_tokens_seen": 1016185336,
      "step": 19750,
      "train_runtime": 121607.9015,
      "train_tokens_per_second": 8356.244
    },
    {
      "epoch": 1.3301238890385134,
      "grad_norm": 0.6094413315618321,
      "learning_rate": 2.5223208164562163e-06,
      "loss": 0.3031134605407715,
      "num_input_tokens_seen": 1016439088,
      "step": 19755,
      "train_runtime": 121638.1213,
      "train_tokens_per_second": 8356.254
    },
    {
      "epoch": 1.3304605440344734,
      "grad_norm": 0.6153110024372244,
      "learning_rate": 2.5200245463079315e-06,
      "loss": 0.31504011154174805,
      "num_input_tokens_seen": 1016693832,
      "step": 19760,
      "train_runtime": 121668.8185,
      "train_tokens_per_second": 8356.24
    },
    {
      "epoch": 1.3307971990304335,
      "grad_norm": 0.6764269925887849,
      "learning_rate": 2.5177289696768725e-06,
      "loss": 0.3188283443450928,
      "num_input_tokens_seen": 1016956800,
      "step": 19765,
      "train_runtime": 121705.2933,
      "train_tokens_per_second": 8355.896
    },
    {
      "epoch": 1.3311338540263937,
      "grad_norm": 0.6706351976659286,
      "learning_rate": 2.515434087204996e-06,
      "loss": 0.2973954439163208,
      "num_input_tokens_seen": 1017232336,
      "step": 19770,
      "train_runtime": 121739.4229,
      "train_tokens_per_second": 8355.817
    },
    {
      "epoch": 1.331470509022354,
      "grad_norm": 0.5464044858813836,
      "learning_rate": 2.51313989953405e-06,
      "loss": 0.2998525619506836,
      "num_input_tokens_seen": 1017512936,
      "step": 19775,
      "train_runtime": 121773.9951,
      "train_tokens_per_second": 8355.749
    },
    {
      "epoch": 1.331807164018314,
      "grad_norm": 0.5758055618660524,
      "learning_rate": 2.510846407305603e-06,
      "loss": 0.297439432144165,
      "num_input_tokens_seen": 1017775992,
      "step": 19780,
      "train_runtime": 121805.4301,
      "train_tokens_per_second": 8355.752
    },
    {
      "epoch": 1.332143819014274,
      "grad_norm": 0.6187215350597134,
      "learning_rate": 2.508553611161021e-06,
      "loss": 0.3005701541900635,
      "num_input_tokens_seen": 1018034768,
      "step": 19785,
      "train_runtime": 121837.2039,
      "train_tokens_per_second": 8355.697
    },
    {
      "epoch": 1.3324804740102343,
      "grad_norm": 0.6412042168494314,
      "learning_rate": 2.5062615117414773e-06,
      "loss": 0.30263481140136717,
      "num_input_tokens_seen": 1018311208,
      "step": 19790,
      "train_runtime": 121869.5039,
      "train_tokens_per_second": 8355.751
    },
    {
      "epoch": 1.3328171290061945,
      "grad_norm": 0.7193157661987096,
      "learning_rate": 2.50397010968795e-06,
      "loss": 0.28984713554382324,
      "num_input_tokens_seen": 1018566664,
      "step": 19795,
      "train_runtime": 121897.083,
      "train_tokens_per_second": 8355.956
    },
    {
      "epoch": 1.3331537840021546,
      "grad_norm": 0.6218130918016056,
      "learning_rate": 2.501679405641222e-06,
      "loss": 0.3209359169006348,
      "num_input_tokens_seen": 1018835520,
      "step": 19800,
      "train_runtime": 121931.9442,
      "train_tokens_per_second": 8355.772
    },
    {
      "epoch": 1.3334904389981148,
      "grad_norm": 0.6860979956956623,
      "learning_rate": 2.499389400241881e-06,
      "loss": 0.3041622877120972,
      "num_input_tokens_seen": 1019069008,
      "step": 19805,
      "train_runtime": 121960.7332,
      "train_tokens_per_second": 8355.714
    },
    {
      "epoch": 1.3338270939940748,
      "grad_norm": 0.5911751543508106,
      "learning_rate": 2.497100094130321e-06,
      "loss": 0.26626157760620117,
      "num_input_tokens_seen": 1019332256,
      "step": 19810,
      "train_runtime": 121995.9134,
      "train_tokens_per_second": 8355.462
    },
    {
      "epoch": 1.334163748990035,
      "grad_norm": 0.5725164107764497,
      "learning_rate": 2.494811487946737e-06,
      "loss": 0.2966012954711914,
      "num_input_tokens_seen": 1019607744,
      "step": 19815,
      "train_runtime": 122027.0817,
      "train_tokens_per_second": 8355.586
    },
    {
      "epoch": 1.3345004039859951,
      "grad_norm": 0.6760503176370376,
      "learning_rate": 2.492523582331135e-06,
      "loss": 0.298583984375,
      "num_input_tokens_seen": 1019869128,
      "step": 19820,
      "train_runtime": 122056.924,
      "train_tokens_per_second": 8355.684
    },
    {
      "epoch": 1.3348370589819554,
      "grad_norm": 0.743217628171435,
      "learning_rate": 2.4902363779233134e-06,
      "loss": 0.29581451416015625,
      "num_input_tokens_seen": 1020127056,
      "step": 19825,
      "train_runtime": 122087.278,
      "train_tokens_per_second": 8355.72
    },
    {
      "epoch": 1.3351737139779154,
      "grad_norm": 0.6945997693382276,
      "learning_rate": 2.4879498753628885e-06,
      "loss": 0.30193638801574707,
      "num_input_tokens_seen": 1020346384,
      "step": 19830,
      "train_runtime": 122117.4104,
      "train_tokens_per_second": 8355.454
    },
    {
      "epoch": 1.3355103689738757,
      "grad_norm": 0.559035017271517,
      "learning_rate": 2.4856640752892702e-06,
      "loss": 0.27355208396911623,
      "num_input_tokens_seen": 1020623544,
      "step": 19835,
      "train_runtime": 122151.1441,
      "train_tokens_per_second": 8355.415
    },
    {
      "epoch": 1.3358470239698357,
      "grad_norm": 0.6532305965431634,
      "learning_rate": 2.4833789783416757e-06,
      "loss": 0.27185511589050293,
      "num_input_tokens_seen": 1020885512,
      "step": 19840,
      "train_runtime": 122185.3352,
      "train_tokens_per_second": 8355.221
    },
    {
      "epoch": 1.336183678965796,
      "grad_norm": 0.6946567046253066,
      "learning_rate": 2.4810945851591245e-06,
      "loss": 0.31412625312805176,
      "num_input_tokens_seen": 1021125384,
      "step": 19845,
      "train_runtime": 122216.9931,
      "train_tokens_per_second": 8355.02
    },
    {
      "epoch": 1.336520333961756,
      "grad_norm": 0.6775582952958535,
      "learning_rate": 2.4788108963804406e-06,
      "loss": 0.3008406639099121,
      "num_input_tokens_seen": 1021376880,
      "step": 19850,
      "train_runtime": 122247.886,
      "train_tokens_per_second": 8354.966
    },
    {
      "epoch": 1.3368569889577162,
      "grad_norm": 0.654205857245267,
      "learning_rate": 2.4765279126442502e-06,
      "loss": 0.31642167568206786,
      "num_input_tokens_seen": 1021637192,
      "step": 19855,
      "train_runtime": 122280.397,
      "train_tokens_per_second": 8354.873
    },
    {
      "epoch": 1.3371936439536762,
      "grad_norm": 0.6343246018012364,
      "learning_rate": 2.4742456345889828e-06,
      "loss": 0.3070626974105835,
      "num_input_tokens_seen": 1021885488,
      "step": 19860,
      "train_runtime": 122311.6625,
      "train_tokens_per_second": 8354.767
    },
    {
      "epoch": 1.3375302989496365,
      "grad_norm": 0.7065580864038769,
      "learning_rate": 2.471964062852867e-06,
      "loss": 0.30606870651245116,
      "num_input_tokens_seen": 1022152176,
      "step": 19865,
      "train_runtime": 122342.3621,
      "train_tokens_per_second": 8354.851
    },
    {
      "epoch": 1.3378669539455965,
      "grad_norm": 0.5820797523318217,
      "learning_rate": 2.469683198073943e-06,
      "loss": 0.28144302368164065,
      "num_input_tokens_seen": 1022438816,
      "step": 19870,
      "train_runtime": 122371.6962,
      "train_tokens_per_second": 8355.19
    },
    {
      "epoch": 1.3382036089415568,
      "grad_norm": 0.8360706462934245,
      "learning_rate": 2.4674030408900414e-06,
      "loss": 0.27997355461120604,
      "num_input_tokens_seen": 1022655776,
      "step": 19875,
      "train_runtime": 122396.025,
      "train_tokens_per_second": 8355.302
    },
    {
      "epoch": 1.3385402639375168,
      "grad_norm": 0.732190001176413,
      "learning_rate": 2.465123591938805e-06,
      "loss": 0.307498574256897,
      "num_input_tokens_seen": 1022909232,
      "step": 19880,
      "train_runtime": 122426.3653,
      "train_tokens_per_second": 8355.302
    },
    {
      "epoch": 1.338876918933477,
      "grad_norm": 0.7546625331758499,
      "learning_rate": 2.462844851857673e-06,
      "loss": 0.3201789379119873,
      "num_input_tokens_seen": 1023139280,
      "step": 19885,
      "train_runtime": 122455.3349,
      "train_tokens_per_second": 8355.204
    },
    {
      "epoch": 1.339213573929437,
      "grad_norm": 0.6835169110162995,
      "learning_rate": 2.4605668212838878e-06,
      "loss": 0.2969949245452881,
      "num_input_tokens_seen": 1023400200,
      "step": 19890,
      "train_runtime": 122484.6398,
      "train_tokens_per_second": 8355.335
    },
    {
      "epoch": 1.3395502289253973,
      "grad_norm": 0.6726827366624085,
      "learning_rate": 2.458289500854493e-06,
      "loss": 0.3436745643615723,
      "num_input_tokens_seen": 1023666136,
      "step": 19895,
      "train_runtime": 122510.9128,
      "train_tokens_per_second": 8355.714
    },
    {
      "epoch": 1.3398868839213574,
      "grad_norm": 0.7050365996532545,
      "learning_rate": 2.4560128912063353e-06,
      "loss": 0.3135679721832275,
      "num_input_tokens_seen": 1023894400,
      "step": 19900,
      "train_runtime": 122535.6286,
      "train_tokens_per_second": 8355.891
    },
    {
      "epoch": 1.3402235389173176,
      "grad_norm": 0.6960981023110862,
      "learning_rate": 2.4537369929760596e-06,
      "loss": 0.3381784915924072,
      "num_input_tokens_seen": 1024171352,
      "step": 19905,
      "train_runtime": 122566.9134,
      "train_tokens_per_second": 8356.018
    },
    {
      "epoch": 1.3405601939132776,
      "grad_norm": 0.6907654041594551,
      "learning_rate": 2.4514618068001144e-06,
      "loss": 0.31571331024169924,
      "num_input_tokens_seen": 1024422080,
      "step": 19910,
      "train_runtime": 122599.5887,
      "train_tokens_per_second": 8355.836
    },
    {
      "epoch": 1.3408968489092379,
      "grad_norm": 0.6496888684620415,
      "learning_rate": 2.4491873333147463e-06,
      "loss": 0.29081087112426757,
      "num_input_tokens_seen": 1024671952,
      "step": 19915,
      "train_runtime": 122626.7819,
      "train_tokens_per_second": 8356.021
    },
    {
      "epoch": 1.341233503905198,
      "grad_norm": 0.738718078298942,
      "learning_rate": 2.4469135731560106e-06,
      "loss": 0.2896077632904053,
      "num_input_tokens_seen": 1024915000,
      "step": 19920,
      "train_runtime": 122656.3285,
      "train_tokens_per_second": 8355.99
    },
    {
      "epoch": 1.3415701589011582,
      "grad_norm": 0.7144240294504486,
      "learning_rate": 2.4446405269597494e-06,
      "loss": 0.315073561668396,
      "num_input_tokens_seen": 1025179224,
      "step": 19925,
      "train_runtime": 122687.6095,
      "train_tokens_per_second": 8356.013
    },
    {
      "epoch": 1.3419068138971182,
      "grad_norm": 0.5991716112854333,
      "learning_rate": 2.442368195361618e-06,
      "loss": 0.2739919662475586,
      "num_input_tokens_seen": 1025442648,
      "step": 19930,
      "train_runtime": 122719.2559,
      "train_tokens_per_second": 8356.004
    },
    {
      "epoch": 1.3422434688930784,
      "grad_norm": 0.6350400876633809,
      "learning_rate": 2.4400965789970647e-06,
      "loss": 0.2863043546676636,
      "num_input_tokens_seen": 1025683448,
      "step": 19935,
      "train_runtime": 122748.6654,
      "train_tokens_per_second": 8355.964
    },
    {
      "epoch": 1.3425801238890385,
      "grad_norm": 0.7359660916015447,
      "learning_rate": 2.437825678501339e-06,
      "loss": 0.3118752479553223,
      "num_input_tokens_seen": 1025948096,
      "step": 19940,
      "train_runtime": 122775.5262,
      "train_tokens_per_second": 8356.292
    },
    {
      "epoch": 1.3429167788849987,
      "grad_norm": 0.6095079438801473,
      "learning_rate": 2.4355554945094956e-06,
      "loss": 0.30511715412139895,
      "num_input_tokens_seen": 1026205824,
      "step": 19945,
      "train_runtime": 122807.038,
      "train_tokens_per_second": 8356.246
    },
    {
      "epoch": 1.3432534338809587,
      "grad_norm": 0.662796524364306,
      "learning_rate": 2.4332860276563763e-06,
      "loss": 0.2911284685134888,
      "num_input_tokens_seen": 1026462232,
      "step": 19950,
      "train_runtime": 122837.5206,
      "train_tokens_per_second": 8356.26
    },
    {
      "epoch": 1.343590088876919,
      "grad_norm": 0.5792707714282304,
      "learning_rate": 2.4310172785766396e-06,
      "loss": 0.30453689098358155,
      "num_input_tokens_seen": 1026713088,
      "step": 19955,
      "train_runtime": 122864.7021,
      "train_tokens_per_second": 8356.453
    },
    {
      "epoch": 1.343926743872879,
      "grad_norm": 0.7787437324931943,
      "learning_rate": 2.428749247904724e-06,
      "loss": 0.3315553665161133,
      "num_input_tokens_seen": 1026967912,
      "step": 19960,
      "train_runtime": 122896.6613,
      "train_tokens_per_second": 8356.353
    },
    {
      "epoch": 1.3442633988688393,
      "grad_norm": 0.641173890549149,
      "learning_rate": 2.4264819362748856e-06,
      "loss": 0.31028721332550047,
      "num_input_tokens_seen": 1027233032,
      "step": 19965,
      "train_runtime": 122927.3002,
      "train_tokens_per_second": 8356.427
    },
    {
      "epoch": 1.3446000538647993,
      "grad_norm": 0.707912278900383,
      "learning_rate": 2.4242153443211663e-06,
      "loss": 0.3248644590377808,
      "num_input_tokens_seen": 1027497048,
      "step": 19970,
      "train_runtime": 122958.7497,
      "train_tokens_per_second": 8356.437
    },
    {
      "epoch": 1.3449367088607596,
      "grad_norm": 0.6318321883108848,
      "learning_rate": 2.4219494726774124e-06,
      "loss": 0.29428920745849607,
      "num_input_tokens_seen": 1027765200,
      "step": 19975,
      "train_runtime": 122987.676,
      "train_tokens_per_second": 8356.652
    },
    {
      "epoch": 1.3452733638567196,
      "grad_norm": 0.6586235051325013,
      "learning_rate": 2.419684321977268e-06,
      "loss": 0.28247780799865724,
      "num_input_tokens_seen": 1028008104,
      "step": 19980,
      "train_runtime": 123015.0385,
      "train_tokens_per_second": 8356.768
    },
    {
      "epoch": 1.3456100188526798,
      "grad_norm": 0.6234038653869889,
      "learning_rate": 2.417419892854175e-06,
      "loss": 0.2880507469177246,
      "num_input_tokens_seen": 1028287888,
      "step": 19985,
      "train_runtime": 123042.2798,
      "train_tokens_per_second": 8357.191
    },
    {
      "epoch": 1.3459466738486399,
      "grad_norm": 0.8056522681471315,
      "learning_rate": 2.415156185941371e-06,
      "loss": 0.29158530235290525,
      "num_input_tokens_seen": 1028544160,
      "step": 19990,
      "train_runtime": 123067.2988,
      "train_tokens_per_second": 8357.575
    },
    {
      "epoch": 1.3462833288446001,
      "grad_norm": 0.6469773743255736,
      "learning_rate": 2.412893201871902e-06,
      "loss": 0.3076338768005371,
      "num_input_tokens_seen": 1028782600,
      "step": 19995,
      "train_runtime": 123098.6616,
      "train_tokens_per_second": 8357.383
    },
    {
      "epoch": 1.3466199838405601,
      "grad_norm": 0.7059132779715629,
      "learning_rate": 2.410630941278595e-06,
      "loss": 0.31584360599517824,
      "num_input_tokens_seen": 1029037496,
      "step": 20000,
      "train_runtime": 123130.706,
      "train_tokens_per_second": 8357.278
    },
    {
      "epoch": 1.3469566388365204,
      "grad_norm": 0.6745118762784275,
      "learning_rate": 2.408369404794092e-06,
      "loss": 0.3103919982910156,
      "num_input_tokens_seen": 1029292632,
      "step": 20005,
      "train_runtime": 123162.5095,
      "train_tokens_per_second": 8357.191
    },
    {
      "epoch": 1.3472932938324804,
      "grad_norm": 0.6599998781788835,
      "learning_rate": 2.4061085930508176e-06,
      "loss": 0.3092803001403809,
      "num_input_tokens_seen": 1029537384,
      "step": 20010,
      "train_runtime": 123193.9048,
      "train_tokens_per_second": 8357.048
    },
    {
      "epoch": 1.3476299488284407,
      "grad_norm": 0.7750289575230128,
      "learning_rate": 2.4038485066810058e-06,
      "loss": 0.3083575010299683,
      "num_input_tokens_seen": 1029785016,
      "step": 20015,
      "train_runtime": 123222.7499,
      "train_tokens_per_second": 8357.101
    },
    {
      "epoch": 1.3479666038244007,
      "grad_norm": 0.9468639587966495,
      "learning_rate": 2.40158914631668e-06,
      "loss": 0.27394053936004636,
      "num_input_tokens_seen": 1030026920,
      "step": 20020,
      "train_runtime": 123252.4376,
      "train_tokens_per_second": 8357.051
    },
    {
      "epoch": 1.348303258820361,
      "grad_norm": 0.6829687195801646,
      "learning_rate": 2.399330512589664e-06,
      "loss": 0.30153598785400393,
      "num_input_tokens_seen": 1030283336,
      "step": 20025,
      "train_runtime": 123284.558,
      "train_tokens_per_second": 8356.954
    },
    {
      "epoch": 1.348639913816321,
      "grad_norm": 0.7791853574971936,
      "learning_rate": 2.3970726061315775e-06,
      "loss": 0.30663743019104006,
      "num_input_tokens_seen": 1030538064,
      "step": 20030,
      "train_runtime": 123314.4385,
      "train_tokens_per_second": 8356.994
    },
    {
      "epoch": 1.3489765688122812,
      "grad_norm": 0.6507070872800583,
      "learning_rate": 2.394815427573836e-06,
      "loss": 0.30552842617034914,
      "num_input_tokens_seen": 1030804472,
      "step": 20035,
      "train_runtime": 123344.0102,
      "train_tokens_per_second": 8357.151
    },
    {
      "epoch": 1.3493132238082413,
      "grad_norm": 0.6025165010628415,
      "learning_rate": 2.3925589775476516e-06,
      "loss": 0.28455300331115724,
      "num_input_tokens_seen": 1031043400,
      "step": 20040,
      "train_runtime": 123376.811,
      "train_tokens_per_second": 8356.865
    },
    {
      "epoch": 1.3496498788042015,
      "grad_norm": 0.6514926653169254,
      "learning_rate": 2.3903032566840377e-06,
      "loss": 0.30849218368530273,
      "num_input_tokens_seen": 1031297296,
      "step": 20045,
      "train_runtime": 123407.8301,
      "train_tokens_per_second": 8356.822
    },
    {
      "epoch": 1.3499865338001615,
      "grad_norm": 0.6296407948557022,
      "learning_rate": 2.3880482656137926e-06,
      "loss": 0.30874104499816896,
      "num_input_tokens_seen": 1031575104,
      "step": 20050,
      "train_runtime": 123438.757,
      "train_tokens_per_second": 8356.979
    },
    {
      "epoch": 1.3503231887961218,
      "grad_norm": 0.5869701071989414,
      "learning_rate": 2.3857940049675254e-06,
      "loss": 0.2993377447128296,
      "num_input_tokens_seen": 1031833760,
      "step": 20055,
      "train_runtime": 123471.2499,
      "train_tokens_per_second": 8356.875
    },
    {
      "epoch": 1.3506598437920818,
      "grad_norm": 0.6436679884686825,
      "learning_rate": 2.383540475375625e-06,
      "loss": 0.2896669149398804,
      "num_input_tokens_seen": 1032091408,
      "step": 20060,
      "train_runtime": 123503.5685,
      "train_tokens_per_second": 8356.774
    },
    {
      "epoch": 1.350996498788042,
      "grad_norm": 0.6844995697508771,
      "learning_rate": 2.3812876774682886e-06,
      "loss": 0.29706151485443116,
      "num_input_tokens_seen": 1032371808,
      "step": 20065,
      "train_runtime": 123536.4895,
      "train_tokens_per_second": 8356.817
    },
    {
      "epoch": 1.351333153784002,
      "grad_norm": 0.646712297699358,
      "learning_rate": 2.379035611875503e-06,
      "loss": 0.3277004241943359,
      "num_input_tokens_seen": 1032659224,
      "step": 20070,
      "train_runtime": 123570.3788,
      "train_tokens_per_second": 8356.851
    },
    {
      "epoch": 1.3516698087799623,
      "grad_norm": 0.6791301453511425,
      "learning_rate": 2.376784279227051e-06,
      "loss": 0.2940059661865234,
      "num_input_tokens_seen": 1032932800,
      "step": 20075,
      "train_runtime": 123602.5906,
      "train_tokens_per_second": 8356.886
    },
    {
      "epoch": 1.3520064637759224,
      "grad_norm": 0.6395628244993978,
      "learning_rate": 2.3745336801525105e-06,
      "loss": 0.3013000011444092,
      "num_input_tokens_seen": 1033205680,
      "step": 20080,
      "train_runtime": 123636.8236,
      "train_tokens_per_second": 8356.78
    },
    {
      "epoch": 1.3523431187718826,
      "grad_norm": 0.6830809754284611,
      "learning_rate": 2.3722838152812543e-06,
      "loss": 0.3180394649505615,
      "num_input_tokens_seen": 1033450512,
      "step": 20085,
      "train_runtime": 123665.7706,
      "train_tokens_per_second": 8356.803
    },
    {
      "epoch": 1.3526797737678427,
      "grad_norm": 0.5456755130612314,
      "learning_rate": 2.3700346852424485e-06,
      "loss": 0.2990797758102417,
      "num_input_tokens_seen": 1033706712,
      "step": 20090,
      "train_runtime": 123696.3598,
      "train_tokens_per_second": 8356.808
    },
    {
      "epoch": 1.353016428763803,
      "grad_norm": 0.680846047442662,
      "learning_rate": 2.3677862906650594e-06,
      "loss": 0.32927467823028567,
      "num_input_tokens_seen": 1033955880,
      "step": 20095,
      "train_runtime": 123727.1592,
      "train_tokens_per_second": 8356.741
    },
    {
      "epoch": 1.353353083759763,
      "grad_norm": 0.7796202284295132,
      "learning_rate": 2.365538632177837e-06,
      "loss": 0.33127450942993164,
      "num_input_tokens_seen": 1034194896,
      "step": 20100,
      "train_runtime": 123754.4741,
      "train_tokens_per_second": 8356.828
    },
    {
      "epoch": 1.3536897387557232,
      "grad_norm": 0.6210641677246218,
      "learning_rate": 2.36329171040934e-06,
      "loss": 0.3047150135040283,
      "num_input_tokens_seen": 1034481272,
      "step": 20105,
      "train_runtime": 123787.2766,
      "train_tokens_per_second": 8356.927
    },
    {
      "epoch": 1.3540263937516832,
      "grad_norm": 0.7685414384530671,
      "learning_rate": 2.3610455259879045e-06,
      "loss": 0.2965384006500244,
      "num_input_tokens_seen": 1034725520,
      "step": 20110,
      "train_runtime": 123818.2183,
      "train_tokens_per_second": 8356.812
    },
    {
      "epoch": 1.3543630487476435,
      "grad_norm": 0.7575771226929834,
      "learning_rate": 2.358800079541674e-06,
      "loss": 0.3188737630844116,
      "num_input_tokens_seen": 1034969472,
      "step": 20115,
      "train_runtime": 123848.981,
      "train_tokens_per_second": 8356.706
    },
    {
      "epoch": 1.3546997037436035,
      "grad_norm": 0.6882029575843972,
      "learning_rate": 2.3565553716985794e-06,
      "loss": 0.32381980419158934,
      "num_input_tokens_seen": 1035243784,
      "step": 20120,
      "train_runtime": 123881.8336,
      "train_tokens_per_second": 8356.704
    },
    {
      "epoch": 1.3550363587395637,
      "grad_norm": 0.7266545810417295,
      "learning_rate": 2.3543114030863457e-06,
      "loss": 0.2857825756072998,
      "num_input_tokens_seen": 1035475872,
      "step": 20125,
      "train_runtime": 123910.5477,
      "train_tokens_per_second": 8356.64
    },
    {
      "epoch": 1.3553730137355238,
      "grad_norm": 0.655915576229465,
      "learning_rate": 2.352068174332491e-06,
      "loss": 0.2948249340057373,
      "num_input_tokens_seen": 1035704216,
      "step": 20130,
      "train_runtime": 123941.9464,
      "train_tokens_per_second": 8356.366
    },
    {
      "epoch": 1.355709668731484,
      "grad_norm": 0.569363809374053,
      "learning_rate": 2.349825686064329e-06,
      "loss": 0.27097353935241697,
      "num_input_tokens_seen": 1035964544,
      "step": 20135,
      "train_runtime": 123971.3519,
      "train_tokens_per_second": 8356.483
    },
    {
      "epoch": 1.356046323727444,
      "grad_norm": 0.6479370591483051,
      "learning_rate": 2.3475839389089605e-06,
      "loss": 0.3203598976135254,
      "num_input_tokens_seen": 1036249168,
      "step": 20140,
      "train_runtime": 124003.38,
      "train_tokens_per_second": 8356.62
    },
    {
      "epoch": 1.3563829787234043,
      "grad_norm": 0.6964799176507515,
      "learning_rate": 2.345342933493289e-06,
      "loss": 0.3017817497253418,
      "num_input_tokens_seen": 1036487560,
      "step": 20145,
      "train_runtime": 124030.7796,
      "train_tokens_per_second": 8356.696
    },
    {
      "epoch": 1.3567196337193643,
      "grad_norm": 0.6068571320580438,
      "learning_rate": 2.343102670443998e-06,
      "loss": 0.3065997838973999,
      "num_input_tokens_seen": 1036730928,
      "step": 20150,
      "train_runtime": 124061.8669,
      "train_tokens_per_second": 8356.564
    },
    {
      "epoch": 1.3570562887153246,
      "grad_norm": 0.5960939076836714,
      "learning_rate": 2.3408631503875763e-06,
      "loss": 0.31101717948913576,
      "num_input_tokens_seen": 1037002992,
      "step": 20155,
      "train_runtime": 124090.7781,
      "train_tokens_per_second": 8356.809
    },
    {
      "epoch": 1.3573929437112846,
      "grad_norm": 0.7145809398409917,
      "learning_rate": 2.338624373950291e-06,
      "loss": 0.26345434188842776,
      "num_input_tokens_seen": 1037271248,
      "step": 20160,
      "train_runtime": 124122.1331,
      "train_tokens_per_second": 8356.86
    },
    {
      "epoch": 1.3577295987072449,
      "grad_norm": 0.630593660205498,
      "learning_rate": 2.336386341758216e-06,
      "loss": 0.30486245155334474,
      "num_input_tokens_seen": 1037559824,
      "step": 20165,
      "train_runtime": 124151.2884,
      "train_tokens_per_second": 8357.222
    },
    {
      "epoch": 1.3580662537032049,
      "grad_norm": 0.7802000544191725,
      "learning_rate": 2.3341490544372063e-06,
      "loss": 0.3087857484817505,
      "num_input_tokens_seen": 1037816272,
      "step": 20170,
      "train_runtime": 124180.8329,
      "train_tokens_per_second": 8357.298
    },
    {
      "epoch": 1.3584029086991651,
      "grad_norm": 0.6390184224048322,
      "learning_rate": 2.3319125126129128e-06,
      "loss": 0.3040606498718262,
      "num_input_tokens_seen": 1038076672,
      "step": 20175,
      "train_runtime": 124214.7883,
      "train_tokens_per_second": 8357.11
    },
    {
      "epoch": 1.3587395636951252,
      "grad_norm": 0.6897422840886855,
      "learning_rate": 2.329676716910777e-06,
      "loss": 0.3234599828720093,
      "num_input_tokens_seen": 1038333752,
      "step": 20180,
      "train_runtime": 124245.6113,
      "train_tokens_per_second": 8357.106
    },
    {
      "epoch": 1.3590762186910854,
      "grad_norm": 0.6791543586217502,
      "learning_rate": 2.3274416679560325e-06,
      "loss": 0.3043846607208252,
      "num_input_tokens_seen": 1038602304,
      "step": 20185,
      "train_runtime": 124276.832,
      "train_tokens_per_second": 8357.168
    },
    {
      "epoch": 1.3594128736870454,
      "grad_norm": 0.711422969300409,
      "learning_rate": 2.3252073663737016e-06,
      "loss": 0.3216113090515137,
      "num_input_tokens_seen": 1038892504,
      "step": 20190,
      "train_runtime": 124307.1171,
      "train_tokens_per_second": 8357.466
    },
    {
      "epoch": 1.3597495286830057,
      "grad_norm": 0.7553638924216122,
      "learning_rate": 2.3229738127886054e-06,
      "loss": 0.31137871742248535,
      "num_input_tokens_seen": 1039124744,
      "step": 20195,
      "train_runtime": 124337.242,
      "train_tokens_per_second": 8357.309
    },
    {
      "epoch": 1.3600861836789657,
      "grad_norm": 0.586604207245732,
      "learning_rate": 2.3207410078253416e-06,
      "loss": 0.3081186056137085,
      "num_input_tokens_seen": 1039378912,
      "step": 20200,
      "train_runtime": 124368.872,
      "train_tokens_per_second": 8357.227
    },
    {
      "epoch": 1.360422838674926,
      "grad_norm": 0.6483407390073895,
      "learning_rate": 2.3185089521083154e-06,
      "loss": 0.2980147361755371,
      "num_input_tokens_seen": 1039619864,
      "step": 20205,
      "train_runtime": 124397.1692,
      "train_tokens_per_second": 8357.263
    },
    {
      "epoch": 1.360759493670886,
      "grad_norm": 0.6707145791158808,
      "learning_rate": 2.3162776462617067e-06,
      "loss": 0.30695669651031493,
      "num_input_tokens_seen": 1039872488,
      "step": 20210,
      "train_runtime": 124426.6525,
      "train_tokens_per_second": 8357.313
    },
    {
      "epoch": 1.3610961486668463,
      "grad_norm": 0.6743615899423291,
      "learning_rate": 2.314047090909498e-06,
      "loss": 0.29269530773162844,
      "num_input_tokens_seen": 1040130632,
      "step": 20215,
      "train_runtime": 124454.7547,
      "train_tokens_per_second": 8357.5
    },
    {
      "epoch": 1.3614328036628063,
      "grad_norm": 0.6798266699997294,
      "learning_rate": 2.3118172866754557e-06,
      "loss": 0.2917036533355713,
      "num_input_tokens_seen": 1040378448,
      "step": 20220,
      "train_runtime": 124485.5888,
      "train_tokens_per_second": 8357.421
    },
    {
      "epoch": 1.3617694586587665,
      "grad_norm": 0.7948866166972844,
      "learning_rate": 2.309588234183137e-06,
      "loss": 0.2935960292816162,
      "num_input_tokens_seen": 1040640072,
      "step": 20225,
      "train_runtime": 124517.468,
      "train_tokens_per_second": 8357.382
    },
    {
      "epoch": 1.3621061136547266,
      "grad_norm": 0.6416619572555756,
      "learning_rate": 2.3073599340558896e-06,
      "loss": 0.31913299560546876,
      "num_input_tokens_seen": 1040883312,
      "step": 20230,
      "train_runtime": 124547.6216,
      "train_tokens_per_second": 8357.312
    },
    {
      "epoch": 1.3624427686506868,
      "grad_norm": 0.6872414629151978,
      "learning_rate": 2.3051323869168506e-06,
      "loss": 0.3184078216552734,
      "num_input_tokens_seen": 1041161648,
      "step": 20235,
      "train_runtime": 124583.9215,
      "train_tokens_per_second": 8357.111
    },
    {
      "epoch": 1.3627794236466468,
      "grad_norm": 0.6060926122243196,
      "learning_rate": 2.3029055933889443e-06,
      "loss": 0.29862046241760254,
      "num_input_tokens_seen": 1041402152,
      "step": 20240,
      "train_runtime": 124613.645,
      "train_tokens_per_second": 8357.048
    },
    {
      "epoch": 1.363116078642607,
      "grad_norm": 0.5720983290207879,
      "learning_rate": 2.300679554094892e-06,
      "loss": 0.2781268835067749,
      "num_input_tokens_seen": 1041662256,
      "step": 20245,
      "train_runtime": 124645.3219,
      "train_tokens_per_second": 8357.01
    },
    {
      "epoch": 1.3634527336385671,
      "grad_norm": 0.6155020382402444,
      "learning_rate": 2.298454269657191e-06,
      "loss": 0.29693989753723143,
      "num_input_tokens_seen": 1041916608,
      "step": 20250,
      "train_runtime": 124674.2904,
      "train_tokens_per_second": 8357.109
    },
    {
      "epoch": 1.3637893886345274,
      "grad_norm": 0.6580989737577928,
      "learning_rate": 2.2962297406981432e-06,
      "loss": 0.31287531852722167,
      "num_input_tokens_seen": 1042171808,
      "step": 20255,
      "train_runtime": 124705.577,
      "train_tokens_per_second": 8357.059
    },
    {
      "epoch": 1.3641260436304874,
      "grad_norm": 0.6613358953030363,
      "learning_rate": 2.2940059678398223e-06,
      "loss": 0.2971006393432617,
      "num_input_tokens_seen": 1042418712,
      "step": 20260,
      "train_runtime": 124739.0943,
      "train_tokens_per_second": 8356.792
    },
    {
      "epoch": 1.3644626986264476,
      "grad_norm": 0.6720105825725601,
      "learning_rate": 2.291782951704106e-06,
      "loss": 0.30727190971374513,
      "num_input_tokens_seen": 1042670144,
      "step": 20265,
      "train_runtime": 124769.4889,
      "train_tokens_per_second": 8356.772
    },
    {
      "epoch": 1.3647993536224077,
      "grad_norm": 0.6722465941623115,
      "learning_rate": 2.289560692912651e-06,
      "loss": 0.3067721128463745,
      "num_input_tokens_seen": 1042915824,
      "step": 20270,
      "train_runtime": 124797.4333,
      "train_tokens_per_second": 8356.869
    },
    {
      "epoch": 1.365136008618368,
      "grad_norm": 0.6533622787579281,
      "learning_rate": 2.287339192086906e-06,
      "loss": 0.3068347930908203,
      "num_input_tokens_seen": 1043151704,
      "step": 20275,
      "train_runtime": 124829.5891,
      "train_tokens_per_second": 8356.606
    },
    {
      "epoch": 1.365472663614328,
      "grad_norm": 0.6632525201571032,
      "learning_rate": 2.2851184498481056e-06,
      "loss": 0.3028784036636353,
      "num_input_tokens_seen": 1043395696,
      "step": 20280,
      "train_runtime": 124859.7314,
      "train_tokens_per_second": 8356.543
    },
    {
      "epoch": 1.3658093186102882,
      "grad_norm": 0.712055056071256,
      "learning_rate": 2.2828984668172732e-06,
      "loss": 0.29187831878662107,
      "num_input_tokens_seen": 1043629416,
      "step": 20285,
      "train_runtime": 124889.7776,
      "train_tokens_per_second": 8356.404
    },
    {
      "epoch": 1.3661459736062485,
      "grad_norm": 0.6769254231077144,
      "learning_rate": 2.2806792436152185e-06,
      "loss": 0.30111083984375,
      "num_input_tokens_seen": 1043892736,
      "step": 20290,
      "train_runtime": 124919.3664,
      "train_tokens_per_second": 8356.532
    },
    {
      "epoch": 1.3664826286022085,
      "grad_norm": 0.7055914977439469,
      "learning_rate": 2.2784607808625463e-06,
      "loss": 0.3000594139099121,
      "num_input_tokens_seen": 1044134352,
      "step": 20295,
      "train_runtime": 124951.0291,
      "train_tokens_per_second": 8356.349
    },
    {
      "epoch": 1.3668192835981685,
      "grad_norm": 0.5988483165718044,
      "learning_rate": 2.276243079179634e-06,
      "loss": 0.30394251346588136,
      "num_input_tokens_seen": 1044397352,
      "step": 20300,
      "train_runtime": 124984.4331,
      "train_tokens_per_second": 8356.219
    },
    {
      "epoch": 1.3671559385941288,
      "grad_norm": 0.7386792190767895,
      "learning_rate": 2.2740261391866634e-06,
      "loss": 0.31372871398925783,
      "num_input_tokens_seen": 1044671480,
      "step": 20305,
      "train_runtime": 125012.8144,
      "train_tokens_per_second": 8356.515
    },
    {
      "epoch": 1.367492593590089,
      "grad_norm": 0.7574215705661316,
      "learning_rate": 2.2718099615035865e-06,
      "loss": 0.2739867687225342,
      "num_input_tokens_seen": 1044937720,
      "step": 20310,
      "train_runtime": 125045.0544,
      "train_tokens_per_second": 8356.49
    },
    {
      "epoch": 1.367829248586049,
      "grad_norm": 0.5901299209263264,
      "learning_rate": 2.2695945467501567e-06,
      "loss": 0.312354588508606,
      "num_input_tokens_seen": 1045201776,
      "step": 20315,
      "train_runtime": 125076.7191,
      "train_tokens_per_second": 8356.485
    },
    {
      "epoch": 1.368165903582009,
      "grad_norm": 0.6525961254746567,
      "learning_rate": 2.267379895545905e-06,
      "loss": 0.3005099296569824,
      "num_input_tokens_seen": 1045457280,
      "step": 20320,
      "train_runtime": 125108.0005,
      "train_tokens_per_second": 8356.438
    },
    {
      "epoch": 1.3685025585779693,
      "grad_norm": 0.5219279704155199,
      "learning_rate": 2.2651660085101513e-06,
      "loss": 0.31668288707733155,
      "num_input_tokens_seen": 1045730192,
      "step": 20325,
      "train_runtime": 125142.9345,
      "train_tokens_per_second": 8356.286
    },
    {
      "epoch": 1.3688392135739296,
      "grad_norm": 0.6703583586364402,
      "learning_rate": 2.262952886262003e-06,
      "loss": 0.30512497425079343,
      "num_input_tokens_seen": 1046002104,
      "step": 20330,
      "train_runtime": 125171.64,
      "train_tokens_per_second": 8356.542
    },
    {
      "epoch": 1.3691758685698896,
      "grad_norm": 0.595099166679219,
      "learning_rate": 2.260740529420352e-06,
      "loss": 0.2874576568603516,
      "num_input_tokens_seen": 1046255776,
      "step": 20335,
      "train_runtime": 125206.5253,
      "train_tokens_per_second": 8356.24
    },
    {
      "epoch": 1.3695125235658496,
      "grad_norm": 0.717611198982159,
      "learning_rate": 2.2585289386038754e-06,
      "loss": 0.2921773672103882,
      "num_input_tokens_seen": 1046540064,
      "step": 20340,
      "train_runtime": 125237.2564,
      "train_tokens_per_second": 8356.46
    },
    {
      "epoch": 1.3698491785618099,
      "grad_norm": 0.6466840100412323,
      "learning_rate": 2.256318114431043e-06,
      "loss": 0.33261291980743407,
      "num_input_tokens_seen": 1046829352,
      "step": 20345,
      "train_runtime": 125269.6516,
      "train_tokens_per_second": 8356.608
    },
    {
      "epoch": 1.3701858335577701,
      "grad_norm": 0.7013335039814285,
      "learning_rate": 2.2541080575200973e-06,
      "loss": 0.29471554756164553,
      "num_input_tokens_seen": 1047065952,
      "step": 20350,
      "train_runtime": 125300.0283,
      "train_tokens_per_second": 8356.47
    },
    {
      "epoch": 1.3705224885537302,
      "grad_norm": 0.6359190121532701,
      "learning_rate": 2.251898768489081e-06,
      "loss": 0.3120358943939209,
      "num_input_tokens_seen": 1047349096,
      "step": 20355,
      "train_runtime": 125333.8615,
      "train_tokens_per_second": 8356.474
    },
    {
      "epoch": 1.3708591435496902,
      "grad_norm": 0.8587845823323448,
      "learning_rate": 2.249690247955807e-06,
      "loss": 0.30320723056793214,
      "num_input_tokens_seen": 1047588744,
      "step": 20360,
      "train_runtime": 125366.2972,
      "train_tokens_per_second": 8356.223
    },
    {
      "epoch": 1.3711957985456504,
      "grad_norm": 0.6490175822251917,
      "learning_rate": 2.247482496537887e-06,
      "loss": 0.3226908206939697,
      "num_input_tokens_seen": 1047858224,
      "step": 20365,
      "train_runtime": 125397.3664,
      "train_tokens_per_second": 8356.302
    },
    {
      "epoch": 1.3715324535416107,
      "grad_norm": 0.7064370814611398,
      "learning_rate": 2.24527551485271e-06,
      "loss": 0.30781965255737304,
      "num_input_tokens_seen": 1048119472,
      "step": 20370,
      "train_runtime": 125426.368,
      "train_tokens_per_second": 8356.452
    },
    {
      "epoch": 1.3718691085375707,
      "grad_norm": 0.6572414394306653,
      "learning_rate": 2.243069303517451e-06,
      "loss": 0.3275381803512573,
      "num_input_tokens_seen": 1048377536,
      "step": 20375,
      "train_runtime": 125458.7928,
      "train_tokens_per_second": 8356.35
    },
    {
      "epoch": 1.3722057635335307,
      "grad_norm": 0.6885773622772247,
      "learning_rate": 2.2408638631490696e-06,
      "loss": 0.2958162784576416,
      "num_input_tokens_seen": 1048635304,
      "step": 20380,
      "train_runtime": 125489.7163,
      "train_tokens_per_second": 8356.345
    },
    {
      "epoch": 1.372542418529491,
      "grad_norm": 0.648178326932461,
      "learning_rate": 2.2386591943643117e-06,
      "loss": 0.2885350227355957,
      "num_input_tokens_seen": 1048904920,
      "step": 20385,
      "train_runtime": 125518.8828,
      "train_tokens_per_second": 8356.551
    },
    {
      "epoch": 1.3728790735254512,
      "grad_norm": 0.6382322147695644,
      "learning_rate": 2.2364552977797023e-06,
      "loss": 0.3082895755767822,
      "num_input_tokens_seen": 1049168752,
      "step": 20390,
      "train_runtime": 125550.8641,
      "train_tokens_per_second": 8356.524
    },
    {
      "epoch": 1.3732157285214113,
      "grad_norm": 0.806121699198174,
      "learning_rate": 2.23425217401156e-06,
      "loss": 0.31714444160461425,
      "num_input_tokens_seen": 1049426784,
      "step": 20395,
      "train_runtime": 125580.9393,
      "train_tokens_per_second": 8356.577
    },
    {
      "epoch": 1.3735523835173713,
      "grad_norm": 0.613972433506254,
      "learning_rate": 2.2320498236759746e-06,
      "loss": 0.28239006996154786,
      "num_input_tokens_seen": 1049696728,
      "step": 20400,
      "train_runtime": 125611.9263,
      "train_tokens_per_second": 8356.665
    },
    {
      "epoch": 1.3738890385133316,
      "grad_norm": 0.6356477099421,
      "learning_rate": 2.2298482473888335e-06,
      "loss": 0.30055975914001465,
      "num_input_tokens_seen": 1049943208,
      "step": 20405,
      "train_runtime": 125643.5558,
      "train_tokens_per_second": 8356.523
    },
    {
      "epoch": 1.3742256935092918,
      "grad_norm": 0.5912739734059913,
      "learning_rate": 2.227647445765792e-06,
      "loss": 0.31168451309204104,
      "num_input_tokens_seen": 1050205736,
      "step": 20410,
      "train_runtime": 125672.8701,
      "train_tokens_per_second": 8356.662
    },
    {
      "epoch": 1.3745623485052518,
      "grad_norm": 0.7327180617605101,
      "learning_rate": 2.225447419422305e-06,
      "loss": 0.28993678092956543,
      "num_input_tokens_seen": 1050451400,
      "step": 20415,
      "train_runtime": 125708.0361,
      "train_tokens_per_second": 8356.279
    },
    {
      "epoch": 1.3748990035012119,
      "grad_norm": 0.6242433550181666,
      "learning_rate": 2.223248168973599e-06,
      "loss": 0.29593021869659425,
      "num_input_tokens_seen": 1050719392,
      "step": 20420,
      "train_runtime": 125735.1483,
      "train_tokens_per_second": 8356.608
    },
    {
      "epoch": 1.375235658497172,
      "grad_norm": 0.6891206318068553,
      "learning_rate": 2.221049695034688e-06,
      "loss": 0.28548712730407716,
      "num_input_tokens_seen": 1050985328,
      "step": 20425,
      "train_runtime": 125770.9314,
      "train_tokens_per_second": 8356.345
    },
    {
      "epoch": 1.3755723134931324,
      "grad_norm": 0.7205558695448427,
      "learning_rate": 2.218851998220369e-06,
      "loss": 0.31765754222869874,
      "num_input_tokens_seen": 1051262224,
      "step": 20430,
      "train_runtime": 125799.8224,
      "train_tokens_per_second": 8356.627
    },
    {
      "epoch": 1.3759089684890924,
      "grad_norm": 0.6859552924053286,
      "learning_rate": 2.21665507914522e-06,
      "loss": 0.28239238262176514,
      "num_input_tokens_seen": 1051482008,
      "step": 20435,
      "train_runtime": 125827.7013,
      "train_tokens_per_second": 8356.522
    },
    {
      "epoch": 1.3762456234850524,
      "grad_norm": 0.6280327345164156,
      "learning_rate": 2.214458938423601e-06,
      "loss": 0.2960052967071533,
      "num_input_tokens_seen": 1051744720,
      "step": 20440,
      "train_runtime": 125860.683,
      "train_tokens_per_second": 8356.42
    },
    {
      "epoch": 1.3765822784810127,
      "grad_norm": 0.5559334150935812,
      "learning_rate": 2.212263576669661e-06,
      "loss": 0.2922870874404907,
      "num_input_tokens_seen": 1052009600,
      "step": 20445,
      "train_runtime": 125892.4273,
      "train_tokens_per_second": 8356.417
    },
    {
      "epoch": 1.376918933476973,
      "grad_norm": 0.8768908431754949,
      "learning_rate": 2.2100689944973182e-06,
      "loss": 0.3218863487243652,
      "num_input_tokens_seen": 1052270552,
      "step": 20450,
      "train_runtime": 125922.356,
      "train_tokens_per_second": 8356.503
    },
    {
      "epoch": 1.377255588472933,
      "grad_norm": 0.6328735949592746,
      "learning_rate": 2.2078751925202886e-06,
      "loss": 0.34476265907287595,
      "num_input_tokens_seen": 1052538432,
      "step": 20455,
      "train_runtime": 125955.7747,
      "train_tokens_per_second": 8356.413
    },
    {
      "epoch": 1.377592243468893,
      "grad_norm": 0.6264100346823674,
      "learning_rate": 2.205682171352054e-06,
      "loss": 0.30696663856506345,
      "num_input_tokens_seen": 1052802720,
      "step": 20460,
      "train_runtime": 125984.2203,
      "train_tokens_per_second": 8356.624
    },
    {
      "epoch": 1.3779288984648532,
      "grad_norm": 0.6790020462776697,
      "learning_rate": 2.2034899316058926e-06,
      "loss": 0.28307905197143557,
      "num_input_tokens_seen": 1053044536,
      "step": 20465,
      "train_runtime": 126012.2406,
      "train_tokens_per_second": 8356.684
    },
    {
      "epoch": 1.3782655534608135,
      "grad_norm": 0.7429484517300131,
      "learning_rate": 2.201298473894854e-06,
      "loss": 0.30101547241210935,
      "num_input_tokens_seen": 1053279808,
      "step": 20470,
      "train_runtime": 126044.365,
      "train_tokens_per_second": 8356.421
    },
    {
      "epoch": 1.3786022084567735,
      "grad_norm": 0.6265333900763055,
      "learning_rate": 2.1991077988317723e-06,
      "loss": 0.294628381729126,
      "num_input_tokens_seen": 1053559424,
      "step": 20475,
      "train_runtime": 126075.8929,
      "train_tokens_per_second": 8356.549
    },
    {
      "epoch": 1.3789388634527335,
      "grad_norm": 0.6034272518950323,
      "learning_rate": 2.1969179070292637e-06,
      "loss": 0.28416264057159424,
      "num_input_tokens_seen": 1053804080,
      "step": 20480,
      "train_runtime": 126107.7702,
      "train_tokens_per_second": 8356.377
    },
    {
      "epoch": 1.3792755184486938,
      "grad_norm": 0.6080275610013723,
      "learning_rate": 2.1947287990997236e-06,
      "loss": 0.29663825035095215,
      "num_input_tokens_seen": 1054061512,
      "step": 20485,
      "train_runtime": 126138.5847,
      "train_tokens_per_second": 8356.377
    },
    {
      "epoch": 1.379612173444654,
      "grad_norm": 0.7420718290050353,
      "learning_rate": 2.192540475655328e-06,
      "loss": 0.30174903869628905,
      "num_input_tokens_seen": 1054316584,
      "step": 20490,
      "train_runtime": 126168.3345,
      "train_tokens_per_second": 8356.428
    },
    {
      "epoch": 1.379948828440614,
      "grad_norm": 0.7231177582610299,
      "learning_rate": 2.1903529373080397e-06,
      "loss": 0.31407790184020995,
      "num_input_tokens_seen": 1054562048,
      "step": 20495,
      "train_runtime": 126194.4803,
      "train_tokens_per_second": 8356.642
    },
    {
      "epoch": 1.380285483436574,
      "grad_norm": 1.101233227462399,
      "learning_rate": 2.1881661846695895e-06,
      "loss": 0.2909216403961182,
      "num_input_tokens_seen": 1054785712,
      "step": 20500,
      "train_runtime": 126226.1105,
      "train_tokens_per_second": 8356.32
    },
    {
      "epoch": 1.3806221384325343,
      "grad_norm": 0.7774016360582718,
      "learning_rate": 2.185980218351503e-06,
      "loss": 0.32069370746612547,
      "num_input_tokens_seen": 1055043112,
      "step": 20505,
      "train_runtime": 126256.5858,
      "train_tokens_per_second": 8356.341
    },
    {
      "epoch": 1.3809587934284946,
      "grad_norm": 0.6412098197889331,
      "learning_rate": 2.1837950389650714e-06,
      "loss": 0.29931304454803465,
      "num_input_tokens_seen": 1055331784,
      "step": 20510,
      "train_runtime": 126285.1331,
      "train_tokens_per_second": 8356.738
    },
    {
      "epoch": 1.3812954484244546,
      "grad_norm": 0.5592825033721487,
      "learning_rate": 2.1816106471213787e-06,
      "loss": 0.31114993095397947,
      "num_input_tokens_seen": 1055615272,
      "step": 20515,
      "train_runtime": 126312.8662,
      "train_tokens_per_second": 8357.148
    },
    {
      "epoch": 1.3816321034204146,
      "grad_norm": 0.6554929413153016,
      "learning_rate": 2.179427043431282e-06,
      "loss": 0.32677254676818845,
      "num_input_tokens_seen": 1055887184,
      "step": 20520,
      "train_runtime": 126347.3835,
      "train_tokens_per_second": 8357.017
    },
    {
      "epoch": 1.381968758416375,
      "grad_norm": 0.6254013005947292,
      "learning_rate": 2.177244228505418e-06,
      "loss": 0.3227147340774536,
      "num_input_tokens_seen": 1056112136,
      "step": 20525,
      "train_runtime": 126376.3991,
      "train_tokens_per_second": 8356.878
    },
    {
      "epoch": 1.3823054134123351,
      "grad_norm": 0.7616895982950401,
      "learning_rate": 2.1750622029542047e-06,
      "loss": 0.28806285858154296,
      "num_input_tokens_seen": 1056390920,
      "step": 20530,
      "train_runtime": 126407.3805,
      "train_tokens_per_second": 8357.035
    },
    {
      "epoch": 1.3826420684082952,
      "grad_norm": 0.70293144626987,
      "learning_rate": 2.1728809673878374e-06,
      "loss": 0.3112710237503052,
      "num_input_tokens_seen": 1056622344,
      "step": 20535,
      "train_runtime": 126436.3969,
      "train_tokens_per_second": 8356.948
    },
    {
      "epoch": 1.3829787234042552,
      "grad_norm": 0.6438111938499254,
      "learning_rate": 2.1707005224162915e-06,
      "loss": 0.29096403121948244,
      "num_input_tokens_seen": 1056886016,
      "step": 20540,
      "train_runtime": 126467.7183,
      "train_tokens_per_second": 8356.963
    },
    {
      "epoch": 1.3833153784002155,
      "grad_norm": 0.628772407474854,
      "learning_rate": 2.1685208686493257e-06,
      "loss": 0.29936683177948,
      "num_input_tokens_seen": 1057113864,
      "step": 20545,
      "train_runtime": 126499.7725,
      "train_tokens_per_second": 8356.646
    },
    {
      "epoch": 1.3836520333961757,
      "grad_norm": 0.6720597281875523,
      "learning_rate": 2.166342006696467e-06,
      "loss": 0.30106339454650877,
      "num_input_tokens_seen": 1057375080,
      "step": 20550,
      "train_runtime": 126536.0998,
      "train_tokens_per_second": 8356.312
    },
    {
      "epoch": 1.3839886883921357,
      "grad_norm": 0.7374189931883665,
      "learning_rate": 2.1641639371670335e-06,
      "loss": 0.30194265842437745,
      "num_input_tokens_seen": 1057635176,
      "step": 20555,
      "train_runtime": 126568.1993,
      "train_tokens_per_second": 8356.247
    },
    {
      "epoch": 1.3843253433880958,
      "grad_norm": 0.8102176605159486,
      "learning_rate": 2.161986660670109e-06,
      "loss": 0.3201554298400879,
      "num_input_tokens_seen": 1057899280,
      "step": 20560,
      "train_runtime": 126601.96,
      "train_tokens_per_second": 8356.105
    },
    {
      "epoch": 1.384661998384056,
      "grad_norm": 0.535987374170239,
      "learning_rate": 2.159810177814566e-06,
      "loss": 0.2877042293548584,
      "num_input_tokens_seen": 1058141712,
      "step": 20565,
      "train_runtime": 126631.9755,
      "train_tokens_per_second": 8356.039
    },
    {
      "epoch": 1.3849986533800163,
      "grad_norm": 0.630766460107833,
      "learning_rate": 2.1576344892090507e-06,
      "loss": 0.3136085033416748,
      "num_input_tokens_seen": 1058391384,
      "step": 20570,
      "train_runtime": 126663.815,
      "train_tokens_per_second": 8355.91
    },
    {
      "epoch": 1.3853353083759763,
      "grad_norm": 0.7289085150996589,
      "learning_rate": 2.1554595954619866e-06,
      "loss": 0.2955606937408447,
      "num_input_tokens_seen": 1058647784,
      "step": 20575,
      "train_runtime": 126689.368,
      "train_tokens_per_second": 8356.248
    },
    {
      "epoch": 1.3856719633719363,
      "grad_norm": 0.6906887118555867,
      "learning_rate": 2.1532854971815763e-06,
      "loss": 0.30385465621948243,
      "num_input_tokens_seen": 1058885560,
      "step": 20580,
      "train_runtime": 126718.5807,
      "train_tokens_per_second": 8356.198
    },
    {
      "epoch": 1.3860086183678966,
      "grad_norm": 0.6474209019602356,
      "learning_rate": 2.151112194975799e-06,
      "loss": 0.28007428646087645,
      "num_input_tokens_seen": 1059140512,
      "step": 20585,
      "train_runtime": 126749.9889,
      "train_tokens_per_second": 8356.139
    },
    {
      "epoch": 1.3863452733638568,
      "grad_norm": 0.543472932977308,
      "learning_rate": 2.1489396894524106e-06,
      "loss": 0.27717204093933107,
      "num_input_tokens_seen": 1059404504,
      "step": 20590,
      "train_runtime": 126779.7153,
      "train_tokens_per_second": 8356.262
    },
    {
      "epoch": 1.3866819283598169,
      "grad_norm": 0.5807433605199038,
      "learning_rate": 2.1467679812189497e-06,
      "loss": 0.3013914585113525,
      "num_input_tokens_seen": 1059664312,
      "step": 20595,
      "train_runtime": 126812.4036,
      "train_tokens_per_second": 8356.157
    },
    {
      "epoch": 1.3870185833557769,
      "grad_norm": 0.7623124982099005,
      "learning_rate": 2.1445970708827213e-06,
      "loss": 0.3028479337692261,
      "num_input_tokens_seen": 1059887488,
      "step": 20600,
      "train_runtime": 126842.4751,
      "train_tokens_per_second": 8355.935
    },
    {
      "epoch": 1.3873552383517371,
      "grad_norm": 0.5655175133550693,
      "learning_rate": 2.14242695905082e-06,
      "loss": 0.27884960174560547,
      "num_input_tokens_seen": 1060158552,
      "step": 20605,
      "train_runtime": 126871.5403,
      "train_tokens_per_second": 8356.157
    },
    {
      "epoch": 1.3876918933476974,
      "grad_norm": 0.6866712255081071,
      "learning_rate": 2.1402576463301044e-06,
      "loss": 0.32149229049682615,
      "num_input_tokens_seen": 1060440616,
      "step": 20610,
      "train_runtime": 126903.3568,
      "train_tokens_per_second": 8356.285
    },
    {
      "epoch": 1.3880285483436574,
      "grad_norm": 0.7008572770180914,
      "learning_rate": 2.13808913332722e-06,
      "loss": 0.28740930557250977,
      "num_input_tokens_seen": 1060689368,
      "step": 20615,
      "train_runtime": 126934.1632,
      "train_tokens_per_second": 8356.217
    },
    {
      "epoch": 1.3883652033396174,
      "grad_norm": 0.7235429739795294,
      "learning_rate": 2.1359214206485845e-06,
      "loss": 0.31825432777404783,
      "num_input_tokens_seen": 1060953056,
      "step": 20620,
      "train_runtime": 126962.8626,
      "train_tokens_per_second": 8356.405
    },
    {
      "epoch": 1.3887018583355777,
      "grad_norm": 0.616819234807941,
      "learning_rate": 2.1337545089003905e-06,
      "loss": 0.261979341506958,
      "num_input_tokens_seen": 1061182632,
      "step": 20625,
      "train_runtime": 126988.0671,
      "train_tokens_per_second": 8356.554
    },
    {
      "epoch": 1.389038513331538,
      "grad_norm": 0.6391688964583764,
      "learning_rate": 2.131588398688608e-06,
      "loss": 0.3065532684326172,
      "num_input_tokens_seen": 1061442752,
      "step": 20630,
      "train_runtime": 127014.9718,
      "train_tokens_per_second": 8356.832
    },
    {
      "epoch": 1.389375168327498,
      "grad_norm": 0.6541038327891988,
      "learning_rate": 2.1294230906189833e-06,
      "loss": 0.2841207027435303,
      "num_input_tokens_seen": 1061681760,
      "step": 20635,
      "train_runtime": 127046.9686,
      "train_tokens_per_second": 8356.608
    },
    {
      "epoch": 1.389711823323458,
      "grad_norm": 0.7002981704234682,
      "learning_rate": 2.1272585852970358e-06,
      "loss": 0.2945998668670654,
      "num_input_tokens_seen": 1061921984,
      "step": 20640,
      "train_runtime": 127076.8154,
      "train_tokens_per_second": 8356.536
    },
    {
      "epoch": 1.3900484783194182,
      "grad_norm": 0.7894461743967995,
      "learning_rate": 2.125094883328068e-06,
      "loss": 0.2877465009689331,
      "num_input_tokens_seen": 1062144608,
      "step": 20645,
      "train_runtime": 127105.9847,
      "train_tokens_per_second": 8356.37
    },
    {
      "epoch": 1.3903851333153785,
      "grad_norm": 0.6828144949160467,
      "learning_rate": 2.1229319853171447e-06,
      "loss": 0.2707461595535278,
      "num_input_tokens_seen": 1062401128,
      "step": 20650,
      "train_runtime": 127138.3138,
      "train_tokens_per_second": 8356.263
    },
    {
      "epoch": 1.3907217883113385,
      "grad_norm": 0.6542406169804238,
      "learning_rate": 2.1207698918691217e-06,
      "loss": 0.3154615879058838,
      "num_input_tokens_seen": 1062656088,
      "step": 20655,
      "train_runtime": 127171.1725,
      "train_tokens_per_second": 8356.108
    },
    {
      "epoch": 1.3910584433072986,
      "grad_norm": 0.6054116721334164,
      "learning_rate": 2.1186086035886133e-06,
      "loss": 0.27914814949035643,
      "num_input_tokens_seen": 1062936136,
      "step": 20660,
      "train_runtime": 127205.6373,
      "train_tokens_per_second": 8356.046
    },
    {
      "epoch": 1.3913950983032588,
      "grad_norm": 0.6499422936046781,
      "learning_rate": 2.116448121080023e-06,
      "loss": 0.3020500183105469,
      "num_input_tokens_seen": 1063230440,
      "step": 20665,
      "train_runtime": 127236.181,
      "train_tokens_per_second": 8356.353
    },
    {
      "epoch": 1.391731753299219,
      "grad_norm": 0.7090337083459695,
      "learning_rate": 2.1142884449475205e-06,
      "loss": 0.3097307920455933,
      "num_input_tokens_seen": 1063483408,
      "step": 20670,
      "train_runtime": 127262.8407,
      "train_tokens_per_second": 8356.59
    },
    {
      "epoch": 1.392068408295179,
      "grad_norm": 0.7074354301311644,
      "learning_rate": 2.112129575795052e-06,
      "loss": 0.2985063076019287,
      "num_input_tokens_seen": 1063727368,
      "step": 20675,
      "train_runtime": 127293.2868,
      "train_tokens_per_second": 8356.508
    },
    {
      "epoch": 1.3924050632911391,
      "grad_norm": 0.5779495698598853,
      "learning_rate": 2.1099715142263394e-06,
      "loss": 0.27476089000701903,
      "num_input_tokens_seen": 1063971240,
      "step": 20680,
      "train_runtime": 127323.6255,
      "train_tokens_per_second": 8356.432
    },
    {
      "epoch": 1.3927417182870994,
      "grad_norm": 0.6269555361627216,
      "learning_rate": 2.1078142608448773e-06,
      "loss": 0.2756513595581055,
      "num_input_tokens_seen": 1064237760,
      "step": 20685,
      "train_runtime": 127350.7516,
      "train_tokens_per_second": 8356.745
    },
    {
      "epoch": 1.3930783732830596,
      "grad_norm": 0.707908172698073,
      "learning_rate": 2.1056578162539327e-06,
      "loss": 0.32199935913085936,
      "num_input_tokens_seen": 1064483448,
      "step": 20690,
      "train_runtime": 127379.7369,
      "train_tokens_per_second": 8356.772
    },
    {
      "epoch": 1.3934150282790196,
      "grad_norm": 0.8119924289275169,
      "learning_rate": 2.103502181056554e-06,
      "loss": 0.3065761089324951,
      "num_input_tokens_seen": 1064724568,
      "step": 20695,
      "train_runtime": 127410.1736,
      "train_tokens_per_second": 8356.668
    },
    {
      "epoch": 1.3937516832749797,
      "grad_norm": 0.6298362676096052,
      "learning_rate": 2.1013473558555503e-06,
      "loss": 0.29851341247558594,
      "num_input_tokens_seen": 1065005584,
      "step": 20700,
      "train_runtime": 127441.2248,
      "train_tokens_per_second": 8356.837
    },
    {
      "epoch": 1.39408833827094,
      "grad_norm": 0.6301055103710732,
      "learning_rate": 2.099193341253519e-06,
      "loss": 0.2922239303588867,
      "num_input_tokens_seen": 1065239840,
      "step": 20705,
      "train_runtime": 127472.2768,
      "train_tokens_per_second": 8356.639
    },
    {
      "epoch": 1.3944249932669002,
      "grad_norm": 0.6201238380690933,
      "learning_rate": 2.0970401378528167e-06,
      "loss": 0.29210991859436036,
      "num_input_tokens_seen": 1065468288,
      "step": 20710,
      "train_runtime": 127502.7219,
      "train_tokens_per_second": 8356.436
    },
    {
      "epoch": 1.3947616482628602,
      "grad_norm": 0.5972899223082472,
      "learning_rate": 2.0948877462555844e-06,
      "loss": 0.2766505241394043,
      "num_input_tokens_seen": 1065738648,
      "step": 20715,
      "train_runtime": 127530.833,
      "train_tokens_per_second": 8356.714
    },
    {
      "epoch": 1.3950983032588202,
      "grad_norm": 0.5966685804604998,
      "learning_rate": 2.09273616706373e-06,
      "loss": 0.32006402015686036,
      "num_input_tokens_seen": 1066028376,
      "step": 20720,
      "train_runtime": 127562.7644,
      "train_tokens_per_second": 8356.893
    },
    {
      "epoch": 1.3954349582547805,
      "grad_norm": 0.6911314910133048,
      "learning_rate": 2.090585400878936e-06,
      "loss": 0.3169880390167236,
      "num_input_tokens_seen": 1066277680,
      "step": 20725,
      "train_runtime": 127593.2784,
      "train_tokens_per_second": 8356.848
    },
    {
      "epoch": 1.3957716132507407,
      "grad_norm": 0.707256772082191,
      "learning_rate": 2.0884354483026576e-06,
      "loss": 0.30948314666748045,
      "num_input_tokens_seen": 1066544320,
      "step": 20730,
      "train_runtime": 127623.1507,
      "train_tokens_per_second": 8356.982
    },
    {
      "epoch": 1.3961082682467008,
      "grad_norm": 0.6135070224152639,
      "learning_rate": 2.0862863099361214e-06,
      "loss": 0.3114339828491211,
      "num_input_tokens_seen": 1066830424,
      "step": 20735,
      "train_runtime": 127658.0116,
      "train_tokens_per_second": 8356.941
    },
    {
      "epoch": 1.3964449232426608,
      "grad_norm": 0.7186834917218913,
      "learning_rate": 2.084137986380326e-06,
      "loss": 0.32537550926208497,
      "num_input_tokens_seen": 1067082296,
      "step": 20740,
      "train_runtime": 127692.1135,
      "train_tokens_per_second": 8356.681
    },
    {
      "epoch": 1.396781578238621,
      "grad_norm": 0.648631343618548,
      "learning_rate": 2.081990478236048e-06,
      "loss": 0.290079927444458,
      "num_input_tokens_seen": 1067317864,
      "step": 20745,
      "train_runtime": 127722.4,
      "train_tokens_per_second": 8356.544
    },
    {
      "epoch": 1.3971182332345813,
      "grad_norm": 0.6968913197216609,
      "learning_rate": 2.079843786103824e-06,
      "loss": 0.29417641162872316,
      "num_input_tokens_seen": 1067555136,
      "step": 20750,
      "train_runtime": 127754.4411,
      "train_tokens_per_second": 8356.305
    },
    {
      "epoch": 1.3974548882305413,
      "grad_norm": 0.682936484932301,
      "learning_rate": 2.0776979105839774e-06,
      "loss": 0.3058482646942139,
      "num_input_tokens_seen": 1067835808,
      "step": 20755,
      "train_runtime": 127785.9439,
      "train_tokens_per_second": 8356.442
    },
    {
      "epoch": 1.3977915432265013,
      "grad_norm": 0.6898594208105641,
      "learning_rate": 2.075552852276588e-06,
      "loss": 0.3294797420501709,
      "num_input_tokens_seen": 1068105768,
      "step": 20760,
      "train_runtime": 127814.3031,
      "train_tokens_per_second": 8356.7
    },
    {
      "epoch": 1.3981281982224616,
      "grad_norm": 0.7540153531009443,
      "learning_rate": 2.07340861178152e-06,
      "loss": 0.31100120544433596,
      "num_input_tokens_seen": 1068360704,
      "step": 20765,
      "train_runtime": 127845.2779,
      "train_tokens_per_second": 8356.669
    },
    {
      "epoch": 1.3984648532184218,
      "grad_norm": 0.6980598993910977,
      "learning_rate": 2.0712651896984013e-06,
      "loss": 0.3024468421936035,
      "num_input_tokens_seen": 1068630024,
      "step": 20770,
      "train_runtime": 127878.6341,
      "train_tokens_per_second": 8356.596
    },
    {
      "epoch": 1.3988015082143819,
      "grad_norm": 0.610829880549451,
      "learning_rate": 2.0691225866266335e-06,
      "loss": 0.30399270057678224,
      "num_input_tokens_seen": 1068874808,
      "step": 20775,
      "train_runtime": 127906.0982,
      "train_tokens_per_second": 8356.715
    },
    {
      "epoch": 1.3991381632103421,
      "grad_norm": 0.7422148534847935,
      "learning_rate": 2.066980803165388e-06,
      "loss": 0.30172908306121826,
      "num_input_tokens_seen": 1069151208,
      "step": 20780,
      "train_runtime": 127936.607,
      "train_tokens_per_second": 8356.883
    },
    {
      "epoch": 1.3994748182063022,
      "grad_norm": 0.6682205293106211,
      "learning_rate": 2.0648398399136088e-06,
      "loss": 0.30202603340148926,
      "num_input_tokens_seen": 1069415232,
      "step": 20785,
      "train_runtime": 127967.4543,
      "train_tokens_per_second": 8356.931
    },
    {
      "epoch": 1.3998114732022624,
      "grad_norm": 0.5883692660701106,
      "learning_rate": 2.062699697470007e-06,
      "loss": 0.287111234664917,
      "num_input_tokens_seen": 1069661536,
      "step": 20790,
      "train_runtime": 127999.3653,
      "train_tokens_per_second": 8356.772
    },
    {
      "epoch": 1.4001481281982224,
      "grad_norm": 0.5879651362803939,
      "learning_rate": 2.060560376433072e-06,
      "loss": 0.3184560060501099,
      "num_input_tokens_seen": 1069926744,
      "step": 20795,
      "train_runtime": 128029.8426,
      "train_tokens_per_second": 8356.854
    },
    {
      "epoch": 1.4004847831941827,
      "grad_norm": 0.5847038742492293,
      "learning_rate": 2.05842187740105e-06,
      "loss": 0.3074421167373657,
      "num_input_tokens_seen": 1070173608,
      "step": 20800,
      "train_runtime": 128061.0631,
      "train_tokens_per_second": 8356.745
    },
    {
      "epoch": 1.4008214381901427,
      "grad_norm": 0.7739068500898593,
      "learning_rate": 2.0562842009719745e-06,
      "loss": 0.27733232975006106,
      "num_input_tokens_seen": 1070376984,
      "step": 20805,
      "train_runtime": 128086.1031,
      "train_tokens_per_second": 8356.699
    },
    {
      "epoch": 1.401158093186103,
      "grad_norm": 0.688475196211515,
      "learning_rate": 2.0541473477436317e-06,
      "loss": 0.28714442253112793,
      "num_input_tokens_seen": 1070627208,
      "step": 20810,
      "train_runtime": 128116.2472,
      "train_tokens_per_second": 8356.686
    },
    {
      "epoch": 1.401494748182063,
      "grad_norm": 0.6104501870469582,
      "learning_rate": 2.052011318313591e-06,
      "loss": 0.2912898540496826,
      "num_input_tokens_seen": 1070893280,
      "step": 20815,
      "train_runtime": 128148.9772,
      "train_tokens_per_second": 8356.628
    },
    {
      "epoch": 1.4018314031780232,
      "grad_norm": 0.6260556422070833,
      "learning_rate": 2.049876113279185e-06,
      "loss": 0.32019524574279784,
      "num_input_tokens_seen": 1071146744,
      "step": 20820,
      "train_runtime": 128183.5934,
      "train_tokens_per_second": 8356.348
    },
    {
      "epoch": 1.4021680581739833,
      "grad_norm": 0.596186161142486,
      "learning_rate": 2.0477417332375165e-06,
      "loss": 0.28453288078308103,
      "num_input_tokens_seen": 1071422512,
      "step": 20825,
      "train_runtime": 128214.2604,
      "train_tokens_per_second": 8356.5
    },
    {
      "epoch": 1.4025047131699435,
      "grad_norm": 0.7095916150984886,
      "learning_rate": 2.045608178785458e-06,
      "loss": 0.29633607864379885,
      "num_input_tokens_seen": 1071678896,
      "step": 20830,
      "train_runtime": 128245.9092,
      "train_tokens_per_second": 8356.437
    },
    {
      "epoch": 1.4028413681659035,
      "grad_norm": 0.5557405248915998,
      "learning_rate": 2.043475450519652e-06,
      "loss": 0.2947739601135254,
      "num_input_tokens_seen": 1071936104,
      "step": 20835,
      "train_runtime": 128278.292,
      "train_tokens_per_second": 8356.333
    },
    {
      "epoch": 1.4031780231618638,
      "grad_norm": 0.6740692069110648,
      "learning_rate": 2.0413435490365064e-06,
      "loss": 0.3021059513092041,
      "num_input_tokens_seen": 1072184480,
      "step": 20840,
      "train_runtime": 128304.9257,
      "train_tokens_per_second": 8356.534
    },
    {
      "epoch": 1.4035146781578238,
      "grad_norm": 0.6225036049185637,
      "learning_rate": 2.0392124749322064e-06,
      "loss": 0.31174867153167723,
      "num_input_tokens_seen": 1072452152,
      "step": 20845,
      "train_runtime": 128332.2856,
      "train_tokens_per_second": 8356.838
    },
    {
      "epoch": 1.403851333153784,
      "grad_norm": 0.6997505321896831,
      "learning_rate": 2.037082228802693e-06,
      "loss": 0.28815364837646484,
      "num_input_tokens_seen": 1072722656,
      "step": 20850,
      "train_runtime": 128359.9937,
      "train_tokens_per_second": 8357.142
    },
    {
      "epoch": 1.404187988149744,
      "grad_norm": 0.627648095824335,
      "learning_rate": 2.03495281124369e-06,
      "loss": 0.2795103073120117,
      "num_input_tokens_seen": 1072979864,
      "step": 20855,
      "train_runtime": 128394.9001,
      "train_tokens_per_second": 8356.873
    },
    {
      "epoch": 1.4045246431457044,
      "grad_norm": 0.6236736576833406,
      "learning_rate": 2.0328242228506746e-06,
      "loss": 0.2842702865600586,
      "num_input_tokens_seen": 1073237952,
      "step": 20860,
      "train_runtime": 128422.6164,
      "train_tokens_per_second": 8357.079
    },
    {
      "epoch": 1.4048612981416644,
      "grad_norm": 0.6249433022780402,
      "learning_rate": 2.0306964642189053e-06,
      "loss": 0.30205984115600587,
      "num_input_tokens_seen": 1073488336,
      "step": 20865,
      "train_runtime": 128454.4995,
      "train_tokens_per_second": 8356.954
    },
    {
      "epoch": 1.4051979531376246,
      "grad_norm": 0.6385097754153668,
      "learning_rate": 2.028569535943402e-06,
      "loss": 0.29488072395324705,
      "num_input_tokens_seen": 1073730128,
      "step": 20870,
      "train_runtime": 128484.7386,
      "train_tokens_per_second": 8356.869
    },
    {
      "epoch": 1.4055346081335847,
      "grad_norm": 0.656217658554165,
      "learning_rate": 2.0264434386189524e-06,
      "loss": 0.29990372657775877,
      "num_input_tokens_seen": 1073991408,
      "step": 20875,
      "train_runtime": 128517.9547,
      "train_tokens_per_second": 8356.742
    },
    {
      "epoch": 1.405871263129545,
      "grad_norm": 0.6072721851611018,
      "learning_rate": 2.0243181728401136e-06,
      "loss": 0.2929666042327881,
      "num_input_tokens_seen": 1074256048,
      "step": 20880,
      "train_runtime": 128545.7525,
      "train_tokens_per_second": 8356.994
    },
    {
      "epoch": 1.406207918125505,
      "grad_norm": 0.6790328006045276,
      "learning_rate": 2.022193739201208e-06,
      "loss": 0.3138271331787109,
      "num_input_tokens_seen": 1074487592,
      "step": 20885,
      "train_runtime": 128575.9048,
      "train_tokens_per_second": 8356.835
    },
    {
      "epoch": 1.4065445731214652,
      "grad_norm": 0.6231859734722964,
      "learning_rate": 2.0200701382963267e-06,
      "loss": 0.28175153732299807,
      "num_input_tokens_seen": 1074739744,
      "step": 20890,
      "train_runtime": 128606.5308,
      "train_tokens_per_second": 8356.805
    },
    {
      "epoch": 1.4068812281174252,
      "grad_norm": 0.6483558522867493,
      "learning_rate": 2.0179473707193324e-06,
      "loss": 0.3042271614074707,
      "num_input_tokens_seen": 1075005904,
      "step": 20895,
      "train_runtime": 128636.256,
      "train_tokens_per_second": 8356.943
    },
    {
      "epoch": 1.4072178831133855,
      "grad_norm": 0.5979859560067431,
      "learning_rate": 2.0158254370638435e-06,
      "loss": 0.30184450149536135,
      "num_input_tokens_seen": 1075265504,
      "step": 20900,
      "train_runtime": 128672.1969,
      "train_tokens_per_second": 8356.627
    },
    {
      "epoch": 1.4075545381093455,
      "grad_norm": 0.564422636666687,
      "learning_rate": 2.0137043379232585e-06,
      "loss": 0.2984570741653442,
      "num_input_tokens_seen": 1075530400,
      "step": 20905,
      "train_runtime": 128703.3852,
      "train_tokens_per_second": 8356.66
    },
    {
      "epoch": 1.4078911931053057,
      "grad_norm": 0.5425607256560027,
      "learning_rate": 2.0115840738907293e-06,
      "loss": 0.28650531768798826,
      "num_input_tokens_seen": 1075767432,
      "step": 20910,
      "train_runtime": 128733.387,
      "train_tokens_per_second": 8356.553
    },
    {
      "epoch": 1.4082278481012658,
      "grad_norm": 0.738709954824736,
      "learning_rate": 2.0094646455591865e-06,
      "loss": 0.30636916160583494,
      "num_input_tokens_seen": 1076028432,
      "step": 20915,
      "train_runtime": 128760.8382,
      "train_tokens_per_second": 8356.799
    },
    {
      "epoch": 1.408564503097226,
      "grad_norm": 0.6474378662923264,
      "learning_rate": 2.007346053521319e-06,
      "loss": 0.3198882579803467,
      "num_input_tokens_seen": 1076288480,
      "step": 20920,
      "train_runtime": 128790.4894,
      "train_tokens_per_second": 8356.894
    },
    {
      "epoch": 1.408901158093186,
      "grad_norm": 0.6832053035281319,
      "learning_rate": 2.005228298369583e-06,
      "loss": 0.31244099140167236,
      "num_input_tokens_seen": 1076527304,
      "step": 20925,
      "train_runtime": 128819.9273,
      "train_tokens_per_second": 8356.838
    },
    {
      "epoch": 1.4092378130891463,
      "grad_norm": 0.7026081215396923,
      "learning_rate": 2.003111380696208e-06,
      "loss": 0.286331844329834,
      "num_input_tokens_seen": 1076786472,
      "step": 20930,
      "train_runtime": 128849.0904,
      "train_tokens_per_second": 8356.958
    },
    {
      "epoch": 1.4095744680851063,
      "grad_norm": 0.7282657410338225,
      "learning_rate": 2.0009953010931747e-06,
      "loss": 0.285398006439209,
      "num_input_tokens_seen": 1077034976,
      "step": 20935,
      "train_runtime": 128877.3539,
      "train_tokens_per_second": 8357.054
    },
    {
      "epoch": 1.4099111230810666,
      "grad_norm": 0.7567887232399207,
      "learning_rate": 1.998880060152244e-06,
      "loss": 0.3289190292358398,
      "num_input_tokens_seen": 1077305104,
      "step": 20940,
      "train_runtime": 128910.2418,
      "train_tokens_per_second": 8357.017
    },
    {
      "epoch": 1.4102477780770266,
      "grad_norm": 0.640784340931643,
      "learning_rate": 1.9967656584649354e-06,
      "loss": 0.3086037874221802,
      "num_input_tokens_seen": 1077566384,
      "step": 20945,
      "train_runtime": 128938.8888,
      "train_tokens_per_second": 8357.187
    },
    {
      "epoch": 1.4105844330729869,
      "grad_norm": 0.5964719299225596,
      "learning_rate": 1.994652096622533e-06,
      "loss": 0.32252964973449705,
      "num_input_tokens_seen": 1077821096,
      "step": 20950,
      "train_runtime": 128970.1108,
      "train_tokens_per_second": 8357.139
    },
    {
      "epoch": 1.410921088068947,
      "grad_norm": 0.5700632493478853,
      "learning_rate": 1.992539375216089e-06,
      "loss": 0.3176435470581055,
      "num_input_tokens_seen": 1078086136,
      "step": 20955,
      "train_runtime": 129001.0409,
      "train_tokens_per_second": 8357.189
    },
    {
      "epoch": 1.4112577430649071,
      "grad_norm": 0.6350965983156464,
      "learning_rate": 1.9904274948364184e-06,
      "loss": 0.3055450677871704,
      "num_input_tokens_seen": 1078347424,
      "step": 20960,
      "train_runtime": 129031.6501,
      "train_tokens_per_second": 8357.232
    },
    {
      "epoch": 1.4115943980608672,
      "grad_norm": 0.6867824235036502,
      "learning_rate": 1.988316456074102e-06,
      "loss": 0.3131978511810303,
      "num_input_tokens_seen": 1078619336,
      "step": 20965,
      "train_runtime": 129064.1325,
      "train_tokens_per_second": 8357.235
    },
    {
      "epoch": 1.4119310530568274,
      "grad_norm": 0.6113454432091829,
      "learning_rate": 1.9862062595194854e-06,
      "loss": 0.30844357013702395,
      "num_input_tokens_seen": 1078867360,
      "step": 20970,
      "train_runtime": 129095.5148,
      "train_tokens_per_second": 8357.125
    },
    {
      "epoch": 1.4122677080527875,
      "grad_norm": 0.7661154325809073,
      "learning_rate": 1.984096905762676e-06,
      "loss": 0.3022831201553345,
      "num_input_tokens_seen": 1079108672,
      "step": 20975,
      "train_runtime": 129124.6349,
      "train_tokens_per_second": 8357.109
    },
    {
      "epoch": 1.4126043630487477,
      "grad_norm": 0.7531573039960046,
      "learning_rate": 1.9819883953935533e-06,
      "loss": 0.3031305313110352,
      "num_input_tokens_seen": 1079387064,
      "step": 20980,
      "train_runtime": 129156.4366,
      "train_tokens_per_second": 8357.207
    },
    {
      "epoch": 1.4129410180447077,
      "grad_norm": 0.7047779628330395,
      "learning_rate": 1.9798807290017485e-06,
      "loss": 0.29775049686431887,
      "num_input_tokens_seen": 1079633712,
      "step": 20985,
      "train_runtime": 129186.0856,
      "train_tokens_per_second": 8357.198
    },
    {
      "epoch": 1.413277673040668,
      "grad_norm": 0.6316653914738319,
      "learning_rate": 1.9777739071766693e-06,
      "loss": 0.30454163551330565,
      "num_input_tokens_seen": 1079882600,
      "step": 20990,
      "train_runtime": 129218.7808,
      "train_tokens_per_second": 8357.01
    },
    {
      "epoch": 1.413614328036628,
      "grad_norm": 0.6060763815334613,
      "learning_rate": 1.97566793050748e-06,
      "loss": 0.2993695020675659,
      "num_input_tokens_seen": 1080123432,
      "step": 20995,
      "train_runtime": 129248.5878,
      "train_tokens_per_second": 8356.946
    },
    {
      "epoch": 1.4139509830325883,
      "grad_norm": 0.7236262703811732,
      "learning_rate": 1.9735627995831096e-06,
      "loss": 0.28127713203430177,
      "num_input_tokens_seen": 1080398136,
      "step": 21000,
      "train_runtime": 129276.4441,
      "train_tokens_per_second": 8357.27
    },
    {
      "epoch": 1.4142876380285483,
      "grad_norm": 0.6756731429270078,
      "learning_rate": 1.9714585149922523e-06,
      "loss": 0.31168427467346194,
      "num_input_tokens_seen": 1080667344,
      "step": 21005,
      "train_runtime": 129306.4246,
      "train_tokens_per_second": 8357.414
    },
    {
      "epoch": 1.4146242930245085,
      "grad_norm": 0.606384552795419,
      "learning_rate": 1.9693550773233634e-06,
      "loss": 0.2843266487121582,
      "num_input_tokens_seen": 1080939136,
      "step": 21010,
      "train_runtime": 129336.3498,
      "train_tokens_per_second": 8357.582
    },
    {
      "epoch": 1.4149609480204686,
      "grad_norm": 0.6776129172299469,
      "learning_rate": 1.967252487164663e-06,
      "loss": 0.281050968170166,
      "num_input_tokens_seen": 1081209000,
      "step": 21015,
      "train_runtime": 129370.0021,
      "train_tokens_per_second": 8357.494
    },
    {
      "epoch": 1.4152976030164288,
      "grad_norm": 0.753165275334734,
      "learning_rate": 1.9651507451041344e-06,
      "loss": 0.30775420665740966,
      "num_input_tokens_seen": 1081452088,
      "step": 21020,
      "train_runtime": 129399.8208,
      "train_tokens_per_second": 8357.447
    },
    {
      "epoch": 1.4156342580123888,
      "grad_norm": 0.5885933504641184,
      "learning_rate": 1.9630498517295204e-06,
      "loss": 0.3032418727874756,
      "num_input_tokens_seen": 1081736584,
      "step": 21025,
      "train_runtime": 129432.9183,
      "train_tokens_per_second": 8357.507
    },
    {
      "epoch": 1.415970913008349,
      "grad_norm": 0.6600097012848987,
      "learning_rate": 1.960949807628335e-06,
      "loss": 0.3026545763015747,
      "num_input_tokens_seen": 1081996832,
      "step": 21030,
      "train_runtime": 129460.917,
      "train_tokens_per_second": 8357.71
    },
    {
      "epoch": 1.4163075680043091,
      "grad_norm": 0.7072769954982299,
      "learning_rate": 1.9588506133878415e-06,
      "loss": 0.3372961521148682,
      "num_input_tokens_seen": 1082263192,
      "step": 21035,
      "train_runtime": 129493.6758,
      "train_tokens_per_second": 8357.653
    },
    {
      "epoch": 1.4166442230002694,
      "grad_norm": 0.6389836162209389,
      "learning_rate": 1.956752269595078e-06,
      "loss": 0.3274827003479004,
      "num_input_tokens_seen": 1082520696,
      "step": 21040,
      "train_runtime": 129527.7994,
      "train_tokens_per_second": 8357.439
    },
    {
      "epoch": 1.4169808779962294,
      "grad_norm": 0.6473158881138733,
      "learning_rate": 1.9546547768368384e-06,
      "loss": 0.2972825288772583,
      "num_input_tokens_seen": 1082801088,
      "step": 21045,
      "train_runtime": 129558.166,
      "train_tokens_per_second": 8357.644
    },
    {
      "epoch": 1.4173175329921897,
      "grad_norm": 0.7938316383096427,
      "learning_rate": 1.9525581356996793e-06,
      "loss": 0.2811481714248657,
      "num_input_tokens_seen": 1083047312,
      "step": 21050,
      "train_runtime": 129586.1788,
      "train_tokens_per_second": 8357.738
    },
    {
      "epoch": 1.4176541879881497,
      "grad_norm": 0.6887326393469879,
      "learning_rate": 1.95046234676992e-06,
      "loss": 0.3022762298583984,
      "num_input_tokens_seen": 1083333120,
      "step": 21055,
      "train_runtime": 129614.3818,
      "train_tokens_per_second": 8358.124
    },
    {
      "epoch": 1.41799084298411,
      "grad_norm": 0.803842443848654,
      "learning_rate": 1.9483674106336416e-06,
      "loss": 0.27750999927520753,
      "num_input_tokens_seen": 1083580264,
      "step": 21060,
      "train_runtime": 129642.5035,
      "train_tokens_per_second": 8358.218
    },
    {
      "epoch": 1.41832749798007,
      "grad_norm": 0.6198821289071458,
      "learning_rate": 1.9462733278766853e-06,
      "loss": 0.29550838470458984,
      "num_input_tokens_seen": 1083814912,
      "step": 21065,
      "train_runtime": 129675.1386,
      "train_tokens_per_second": 8357.924
    },
    {
      "epoch": 1.4186641529760302,
      "grad_norm": 0.6390212444625646,
      "learning_rate": 1.9441800990846553e-06,
      "loss": 0.2937769412994385,
      "num_input_tokens_seen": 1084096616,
      "step": 21070,
      "train_runtime": 129710.0383,
      "train_tokens_per_second": 8357.847
    },
    {
      "epoch": 1.4190008079719902,
      "grad_norm": 0.7061893216695715,
      "learning_rate": 1.942087724842915e-06,
      "loss": 0.3254528284072876,
      "num_input_tokens_seen": 1084341504,
      "step": 21075,
      "train_runtime": 129737.5607,
      "train_tokens_per_second": 8357.961
    },
    {
      "epoch": 1.4193374629679505,
      "grad_norm": 0.6430981970846072,
      "learning_rate": 1.9399962057365944e-06,
      "loss": 0.2945624113082886,
      "num_input_tokens_seen": 1084579016,
      "step": 21080,
      "train_runtime": 129768.4208,
      "train_tokens_per_second": 8357.804
    },
    {
      "epoch": 1.4196741179639105,
      "grad_norm": 0.6734021878503895,
      "learning_rate": 1.937905542350574e-06,
      "loss": 0.32351059913635255,
      "num_input_tokens_seen": 1084830688,
      "step": 21085,
      "train_runtime": 129798.7396,
      "train_tokens_per_second": 8357.791
    },
    {
      "epoch": 1.4200107729598708,
      "grad_norm": 1.102346086546148,
      "learning_rate": 1.935815735269506e-06,
      "loss": 0.2996234893798828,
      "num_input_tokens_seen": 1085092128,
      "step": 21090,
      "train_runtime": 129833.3453,
      "train_tokens_per_second": 8357.577
    },
    {
      "epoch": 1.4203474279558308,
      "grad_norm": 0.6726013388554297,
      "learning_rate": 1.933726785077797e-06,
      "loss": 0.323046350479126,
      "num_input_tokens_seen": 1085341760,
      "step": 21095,
      "train_runtime": 129861.7315,
      "train_tokens_per_second": 8357.672
    },
    {
      "epoch": 1.420684082951791,
      "grad_norm": 0.6618471112213669,
      "learning_rate": 1.9316386923596146e-06,
      "loss": 0.3098498582839966,
      "num_input_tokens_seen": 1085603120,
      "step": 21100,
      "train_runtime": 129892.2926,
      "train_tokens_per_second": 8357.718
    },
    {
      "epoch": 1.421020737947751,
      "grad_norm": 0.7287050854372379,
      "learning_rate": 1.929551457698888e-06,
      "loss": 0.31035666465759276,
      "num_input_tokens_seen": 1085876032,
      "step": 21105,
      "train_runtime": 129923.8044,
      "train_tokens_per_second": 8357.791
    },
    {
      "epoch": 1.4213573929437113,
      "grad_norm": 0.6112709115632613,
      "learning_rate": 1.9274650816793044e-06,
      "loss": 0.3067057132720947,
      "num_input_tokens_seen": 1086160256,
      "step": 21110,
      "train_runtime": 129955.9102,
      "train_tokens_per_second": 8357.913
    },
    {
      "epoch": 1.4216940479396714,
      "grad_norm": 0.6397788002780581,
      "learning_rate": 1.9253795648843136e-06,
      "loss": 0.3073048830032349,
      "num_input_tokens_seen": 1086448704,
      "step": 21115,
      "train_runtime": 129990.5526,
      "train_tokens_per_second": 8357.905
    },
    {
      "epoch": 1.4220307029356316,
      "grad_norm": 0.708558640675028,
      "learning_rate": 1.9232949078971236e-06,
      "loss": 0.305779242515564,
      "num_input_tokens_seen": 1086705544,
      "step": 21120,
      "train_runtime": 130023.2119,
      "train_tokens_per_second": 8357.781
    },
    {
      "epoch": 1.4223673579315916,
      "grad_norm": 0.5946328023542219,
      "learning_rate": 1.9212111113006998e-06,
      "loss": 0.2870795726776123,
      "num_input_tokens_seen": 1086980432,
      "step": 21125,
      "train_runtime": 130049.0809,
      "train_tokens_per_second": 8358.232
    },
    {
      "epoch": 1.4227040129275519,
      "grad_norm": 0.7595351032503853,
      "learning_rate": 1.919128175677775e-06,
      "loss": 0.2997849464416504,
      "num_input_tokens_seen": 1087214272,
      "step": 21130,
      "train_runtime": 130080.8246,
      "train_tokens_per_second": 8357.99
    },
    {
      "epoch": 1.423040667923512,
      "grad_norm": 0.6581124715692033,
      "learning_rate": 1.917046101610827e-06,
      "loss": 0.2966975212097168,
      "num_input_tokens_seen": 1087485880,
      "step": 21135,
      "train_runtime": 130115.6322,
      "train_tokens_per_second": 8357.842
    },
    {
      "epoch": 1.4233773229194722,
      "grad_norm": 0.6478464068291329,
      "learning_rate": 1.9149648896821087e-06,
      "loss": 0.3110191822052002,
      "num_input_tokens_seen": 1087725968,
      "step": 21140,
      "train_runtime": 130145.9911,
      "train_tokens_per_second": 8357.737
    },
    {
      "epoch": 1.4237139779154322,
      "grad_norm": 0.7064017072476337,
      "learning_rate": 1.912884540473621e-06,
      "loss": 0.29104998111724856,
      "num_input_tokens_seen": 1087962344,
      "step": 21145,
      "train_runtime": 130175.5757,
      "train_tokens_per_second": 8357.653
    },
    {
      "epoch": 1.4240506329113924,
      "grad_norm": 0.6823650617051144,
      "learning_rate": 1.9108050545671283e-06,
      "loss": 0.29151747226715086,
      "num_input_tokens_seen": 1088205872,
      "step": 21150,
      "train_runtime": 130202.0637,
      "train_tokens_per_second": 8357.824
    },
    {
      "epoch": 1.4243872879073525,
      "grad_norm": 0.7109425747454821,
      "learning_rate": 1.9087264325441507e-06,
      "loss": 0.26427202224731444,
      "num_input_tokens_seen": 1088445840,
      "step": 21155,
      "train_runtime": 130233.6196,
      "train_tokens_per_second": 8357.641
    },
    {
      "epoch": 1.4247239429033127,
      "grad_norm": 0.6096382619876911,
      "learning_rate": 1.9066486749859686e-06,
      "loss": 0.3048095226287842,
      "num_input_tokens_seen": 1088698112,
      "step": 21160,
      "train_runtime": 130261.7202,
      "train_tokens_per_second": 8357.775
    },
    {
      "epoch": 1.4250605978992728,
      "grad_norm": 0.5890457099894726,
      "learning_rate": 1.9045717824736209e-06,
      "loss": 0.3080806016921997,
      "num_input_tokens_seen": 1088968264,
      "step": 21165,
      "train_runtime": 130294.5658,
      "train_tokens_per_second": 8357.741
    },
    {
      "epoch": 1.425397252895233,
      "grad_norm": 0.7110544904813829,
      "learning_rate": 1.9024957555879037e-06,
      "loss": 0.3064640998840332,
      "num_input_tokens_seen": 1089210416,
      "step": 21170,
      "train_runtime": 130324.6196,
      "train_tokens_per_second": 8357.672
    },
    {
      "epoch": 1.425733907891193,
      "grad_norm": 0.7031326398367332,
      "learning_rate": 1.9004205949093695e-06,
      "loss": 0.31092081069946287,
      "num_input_tokens_seen": 1089474048,
      "step": 21175,
      "train_runtime": 130357.595,
      "train_tokens_per_second": 8357.58
    },
    {
      "epoch": 1.4260705628871533,
      "grad_norm": 0.5674729367580545,
      "learning_rate": 1.8983463010183355e-06,
      "loss": 0.3143479347229004,
      "num_input_tokens_seen": 1089727384,
      "step": 21180,
      "train_runtime": 130390.8724,
      "train_tokens_per_second": 8357.39
    },
    {
      "epoch": 1.4264072178831133,
      "grad_norm": 0.6425981180325138,
      "learning_rate": 1.8962728744948644e-06,
      "loss": 0.30100188255310056,
      "num_input_tokens_seen": 1089991872,
      "step": 21185,
      "train_runtime": 130422.6274,
      "train_tokens_per_second": 8357.383
    },
    {
      "epoch": 1.4267438728790736,
      "grad_norm": 0.6995396048246585,
      "learning_rate": 1.894200315918789e-06,
      "loss": 0.3035569429397583,
      "num_input_tokens_seen": 1090286360,
      "step": 21190,
      "train_runtime": 130452.5027,
      "train_tokens_per_second": 8357.727
    },
    {
      "epoch": 1.4270805278750336,
      "grad_norm": 0.7826443940024519,
      "learning_rate": 1.8921286258696913e-06,
      "loss": 0.29504225254058836,
      "num_input_tokens_seen": 1090517664,
      "step": 21195,
      "train_runtime": 130481.0192,
      "train_tokens_per_second": 8357.673
    },
    {
      "epoch": 1.4274171828709938,
      "grad_norm": 0.6364516806621332,
      "learning_rate": 1.890057804926913e-06,
      "loss": 0.2822399616241455,
      "num_input_tokens_seen": 1090779328,
      "step": 21200,
      "train_runtime": 130512.4941,
      "train_tokens_per_second": 8357.662
    },
    {
      "epoch": 1.4277538378669539,
      "grad_norm": 0.6042661270336602,
      "learning_rate": 1.887987853669553e-06,
      "loss": 0.2620558500289917,
      "num_input_tokens_seen": 1091043080,
      "step": 21205,
      "train_runtime": 130542.7904,
      "train_tokens_per_second": 8357.743
    },
    {
      "epoch": 1.4280904928629141,
      "grad_norm": 0.6842053943842599,
      "learning_rate": 1.8859187726764654e-06,
      "loss": 0.32583234310150144,
      "num_input_tokens_seen": 1091317744,
      "step": 21210,
      "train_runtime": 130574.9281,
      "train_tokens_per_second": 8357.789
    },
    {
      "epoch": 1.4284271478588741,
      "grad_norm": 0.7212542580045628,
      "learning_rate": 1.8838505625262638e-06,
      "loss": 0.29590868949890137,
      "num_input_tokens_seen": 1091586168,
      "step": 21215,
      "train_runtime": 130608.5936,
      "train_tokens_per_second": 8357.69
    },
    {
      "epoch": 1.4287638028548344,
      "grad_norm": 0.7300664932698939,
      "learning_rate": 1.8817832237973154e-06,
      "loss": 0.3039144515991211,
      "num_input_tokens_seen": 1091847896,
      "step": 21220,
      "train_runtime": 130636.3643,
      "train_tokens_per_second": 8357.917
    },
    {
      "epoch": 1.4291004578507944,
      "grad_norm": 0.6403335062631452,
      "learning_rate": 1.8797167570677439e-06,
      "loss": 0.2929843425750732,
      "num_input_tokens_seen": 1092097816,
      "step": 21225,
      "train_runtime": 130671.7,
      "train_tokens_per_second": 8357.57
    },
    {
      "epoch": 1.4294371128467547,
      "grad_norm": 0.874750774128768,
      "learning_rate": 1.8776511629154342e-06,
      "loss": 0.31774110794067384,
      "num_input_tokens_seen": 1092363208,
      "step": 21230,
      "train_runtime": 130704.196,
      "train_tokens_per_second": 8357.522
    },
    {
      "epoch": 1.4297737678427147,
      "grad_norm": 0.6090041130733985,
      "learning_rate": 1.8755864419180176e-06,
      "loss": 0.2990549564361572,
      "num_input_tokens_seen": 1092619928,
      "step": 21235,
      "train_runtime": 130733.6888,
      "train_tokens_per_second": 8357.6
    },
    {
      "epoch": 1.430110422838675,
      "grad_norm": 0.6354384829109294,
      "learning_rate": 1.8735225946528906e-06,
      "loss": 0.30085344314575196,
      "num_input_tokens_seen": 1092892840,
      "step": 21240,
      "train_runtime": 130765.2275,
      "train_tokens_per_second": 8357.672
    },
    {
      "epoch": 1.430447077834635,
      "grad_norm": 0.6485542561714097,
      "learning_rate": 1.8714596216972008e-06,
      "loss": 0.31369495391845703,
      "num_input_tokens_seen": 1093161488,
      "step": 21245,
      "train_runtime": 130792.5459,
      "train_tokens_per_second": 8357.98
    },
    {
      "epoch": 1.4307837328305952,
      "grad_norm": 0.6550250974633568,
      "learning_rate": 1.8693975236278512e-06,
      "loss": 0.30474085807800294,
      "num_input_tokens_seen": 1093404760,
      "step": 21250,
      "train_runtime": 130824.8878,
      "train_tokens_per_second": 8357.773
    },
    {
      "epoch": 1.4311203878265553,
      "grad_norm": 0.7361223887845512,
      "learning_rate": 1.8673363010215017e-06,
      "loss": 0.3206819772720337,
      "num_input_tokens_seen": 1093653040,
      "step": 21255,
      "train_runtime": 130854.5653,
      "train_tokens_per_second": 8357.775
    },
    {
      "epoch": 1.4314570428225155,
      "grad_norm": 0.650838796132633,
      "learning_rate": 1.8652759544545661e-06,
      "loss": 0.30043060779571534,
      "num_input_tokens_seen": 1093920864,
      "step": 21260,
      "train_runtime": 130881.4069,
      "train_tokens_per_second": 8358.107
    },
    {
      "epoch": 1.4317936978184758,
      "grad_norm": 0.644458090494655,
      "learning_rate": 1.863216484503214e-06,
      "loss": 0.30520968437194823,
      "num_input_tokens_seen": 1094178432,
      "step": 21265,
      "train_runtime": 130907.5229,
      "train_tokens_per_second": 8358.408
    },
    {
      "epoch": 1.4321303528144358,
      "grad_norm": 0.7083843302559016,
      "learning_rate": 1.8611578917433698e-06,
      "loss": 0.3214149475097656,
      "num_input_tokens_seen": 1094442528,
      "step": 21270,
      "train_runtime": 130939.282,
      "train_tokens_per_second": 8358.397
    },
    {
      "epoch": 1.4324670078103958,
      "grad_norm": 0.6599018845033143,
      "learning_rate": 1.85910017675071e-06,
      "loss": 0.31797022819519044,
      "num_input_tokens_seen": 1094689112,
      "step": 21275,
      "train_runtime": 130969.7257,
      "train_tokens_per_second": 8358.337
    },
    {
      "epoch": 1.432803662806356,
      "grad_norm": 0.6763352662731626,
      "learning_rate": 1.8570433401006737e-06,
      "loss": 0.26765966415405273,
      "num_input_tokens_seen": 1094963792,
      "step": 21280,
      "train_runtime": 131001.2259,
      "train_tokens_per_second": 8358.424
    },
    {
      "epoch": 1.4331403178023163,
      "grad_norm": 0.5603727527373569,
      "learning_rate": 1.8549873823684418e-06,
      "loss": 0.2703015089035034,
      "num_input_tokens_seen": 1095214896,
      "step": 21285,
      "train_runtime": 131029.1148,
      "train_tokens_per_second": 8358.561
    },
    {
      "epoch": 1.4334769727982763,
      "grad_norm": 0.6042509925582977,
      "learning_rate": 1.8529323041289616e-06,
      "loss": 0.33124818801879885,
      "num_input_tokens_seen": 1095492616,
      "step": 21290,
      "train_runtime": 131057.5883,
      "train_tokens_per_second": 8358.864
    },
    {
      "epoch": 1.4338136277942364,
      "grad_norm": 0.5619739363205349,
      "learning_rate": 1.850878105956927e-06,
      "loss": 0.2978641986846924,
      "num_input_tokens_seen": 1095777048,
      "step": 21295,
      "train_runtime": 131091.8155,
      "train_tokens_per_second": 8358.852
    },
    {
      "epoch": 1.4341502827901966,
      "grad_norm": 0.5963440261805192,
      "learning_rate": 1.8488247884267885e-06,
      "loss": 0.2861639499664307,
      "num_input_tokens_seen": 1096061144,
      "step": 21300,
      "train_runtime": 131123.2483,
      "train_tokens_per_second": 8359.015
    },
    {
      "epoch": 1.4344869377861569,
      "grad_norm": 0.6613986249677305,
      "learning_rate": 1.8467723521127495e-06,
      "loss": 0.2987335681915283,
      "num_input_tokens_seen": 1096296176,
      "step": 21305,
      "train_runtime": 131151.6776,
      "train_tokens_per_second": 8358.995
    },
    {
      "epoch": 1.434823592782117,
      "grad_norm": 0.6602627495329266,
      "learning_rate": 1.8447207975887671e-06,
      "loss": 0.28681154251098634,
      "num_input_tokens_seen": 1096548808,
      "step": 21310,
      "train_runtime": 131186.2658,
      "train_tokens_per_second": 8358.717
    },
    {
      "epoch": 1.435160247778077,
      "grad_norm": 0.7023288285284095,
      "learning_rate": 1.8426701254285522e-06,
      "loss": 0.31384806632995604,
      "num_input_tokens_seen": 1096810576,
      "step": 21315,
      "train_runtime": 131215.8468,
      "train_tokens_per_second": 8358.827
    },
    {
      "epoch": 1.4354969027740372,
      "grad_norm": 0.6436161173387435,
      "learning_rate": 1.840620336205569e-06,
      "loss": 0.3215021133422852,
      "num_input_tokens_seen": 1097074368,
      "step": 21320,
      "train_runtime": 131245.6798,
      "train_tokens_per_second": 8358.937
    },
    {
      "epoch": 1.4358335577699974,
      "grad_norm": 0.6177404558691388,
      "learning_rate": 1.8385714304930324e-06,
      "loss": 0.2807640552520752,
      "num_input_tokens_seen": 1097350960,
      "step": 21325,
      "train_runtime": 131281.964,
      "train_tokens_per_second": 8358.734
    },
    {
      "epoch": 1.4361702127659575,
      "grad_norm": 0.5882650417235131,
      "learning_rate": 1.8365234088639178e-06,
      "loss": 0.29798188209533694,
      "num_input_tokens_seen": 1097614320,
      "step": 21330,
      "train_runtime": 131308.7336,
      "train_tokens_per_second": 8359.035
    },
    {
      "epoch": 1.4365068677619175,
      "grad_norm": 0.6602024841203226,
      "learning_rate": 1.834476271890941e-06,
      "loss": 0.27694220542907716,
      "num_input_tokens_seen": 1097841536,
      "step": 21335,
      "train_runtime": 131339.8409,
      "train_tokens_per_second": 8358.785
    },
    {
      "epoch": 1.4368435227578777,
      "grad_norm": 0.6941661726279934,
      "learning_rate": 1.8324300201465823e-06,
      "loss": 0.28451759815216066,
      "num_input_tokens_seen": 1098116056,
      "step": 21340,
      "train_runtime": 131371.8937,
      "train_tokens_per_second": 8358.836
    },
    {
      "epoch": 1.437180177753838,
      "grad_norm": 0.8579169288644947,
      "learning_rate": 1.8303846542030674e-06,
      "loss": 0.31587610244750974,
      "num_input_tokens_seen": 1098368712,
      "step": 21345,
      "train_runtime": 131401.7468,
      "train_tokens_per_second": 8358.859
    },
    {
      "epoch": 1.437516832749798,
      "grad_norm": 0.5546002785691599,
      "learning_rate": 1.828340174632377e-06,
      "loss": 0.2807796001434326,
      "num_input_tokens_seen": 1098632120,
      "step": 21350,
      "train_runtime": 131437.5949,
      "train_tokens_per_second": 8358.584
    },
    {
      "epoch": 1.437853487745758,
      "grad_norm": 0.7340902986853459,
      "learning_rate": 1.8262965820062434e-06,
      "loss": 0.2966428756713867,
      "num_input_tokens_seen": 1098902808,
      "step": 21355,
      "train_runtime": 131469.0058,
      "train_tokens_per_second": 8358.645
    },
    {
      "epoch": 1.4381901427417183,
      "grad_norm": 0.6350122119201628,
      "learning_rate": 1.8242538768961499e-06,
      "loss": 0.2801060199737549,
      "num_input_tokens_seen": 1099161776,
      "step": 21360,
      "train_runtime": 131505.9357,
      "train_tokens_per_second": 8358.267
    },
    {
      "epoch": 1.4385267977376786,
      "grad_norm": 0.6912343741571223,
      "learning_rate": 1.8222120598733317e-06,
      "loss": 0.28026351928710935,
      "num_input_tokens_seen": 1099438448,
      "step": 21365,
      "train_runtime": 131534.621,
      "train_tokens_per_second": 8358.548
    },
    {
      "epoch": 1.4388634527336386,
      "grad_norm": 0.557387717733823,
      "learning_rate": 1.820171131508781e-06,
      "loss": 0.3202632427215576,
      "num_input_tokens_seen": 1099691960,
      "step": 21370,
      "train_runtime": 131571.8554,
      "train_tokens_per_second": 8358.109
    },
    {
      "epoch": 1.4392001077295986,
      "grad_norm": 0.6504054586688567,
      "learning_rate": 1.8181310923732304e-06,
      "loss": 0.27750577926635744,
      "num_input_tokens_seen": 1099942264,
      "step": 21375,
      "train_runtime": 131601.1899,
      "train_tokens_per_second": 8358.148
    },
    {
      "epoch": 1.4395367627255589,
      "grad_norm": 0.7122200419626099,
      "learning_rate": 1.8160919430371764e-06,
      "loss": 0.3200272560119629,
      "num_input_tokens_seen": 1100180136,
      "step": 21380,
      "train_runtime": 131631.128,
      "train_tokens_per_second": 8358.054
    },
    {
      "epoch": 1.439873417721519,
      "grad_norm": 0.6051706141703245,
      "learning_rate": 1.814053684070855e-06,
      "loss": 0.28946547508239745,
      "num_input_tokens_seen": 1100447528,
      "step": 21385,
      "train_runtime": 131659.8461,
      "train_tokens_per_second": 8358.262
    },
    {
      "epoch": 1.4402100727174791,
      "grad_norm": 0.6160563341295477,
      "learning_rate": 1.8120163160442633e-06,
      "loss": 0.27372188568115235,
      "num_input_tokens_seen": 1100713560,
      "step": 21390,
      "train_runtime": 131691.9583,
      "train_tokens_per_second": 8358.244
    },
    {
      "epoch": 1.4405467277134392,
      "grad_norm": 0.6279262581221682,
      "learning_rate": 1.8099798395271428e-06,
      "loss": 0.3096756458282471,
      "num_input_tokens_seen": 1100997152,
      "step": 21395,
      "train_runtime": 131725.517,
      "train_tokens_per_second": 8358.268
    },
    {
      "epoch": 1.4408833827093994,
      "grad_norm": 0.6792576036687424,
      "learning_rate": 1.807944255088988e-06,
      "loss": 0.28123018741607664,
      "num_input_tokens_seen": 1101256840,
      "step": 21400,
      "train_runtime": 131758.7584,
      "train_tokens_per_second": 8358.13
    },
    {
      "epoch": 1.4412200377053597,
      "grad_norm": 0.6232626304249771,
      "learning_rate": 1.8059095632990437e-06,
      "loss": 0.28386900424957273,
      "num_input_tokens_seen": 1101522616,
      "step": 21405,
      "train_runtime": 131788.5141,
      "train_tokens_per_second": 8358.26
    },
    {
      "epoch": 1.4415566927013197,
      "grad_norm": 0.6505246415648119,
      "learning_rate": 1.8038757647263045e-06,
      "loss": 0.31407151222229,
      "num_input_tokens_seen": 1101781984,
      "step": 21410,
      "train_runtime": 131822.6348,
      "train_tokens_per_second": 8358.064
    },
    {
      "epoch": 1.4418933476972797,
      "grad_norm": 0.6465375072895345,
      "learning_rate": 1.8018428599395143e-06,
      "loss": 0.28914899826049806,
      "num_input_tokens_seen": 1102055520,
      "step": 21415,
      "train_runtime": 131857.0656,
      "train_tokens_per_second": 8357.956
    },
    {
      "epoch": 1.44223000269324,
      "grad_norm": 0.8265605662080039,
      "learning_rate": 1.7998108495071743e-06,
      "loss": 0.2877967119216919,
      "num_input_tokens_seen": 1102315616,
      "step": 21420,
      "train_runtime": 131887.8837,
      "train_tokens_per_second": 8357.975
    },
    {
      "epoch": 1.4425666576892002,
      "grad_norm": 0.5959141857573635,
      "learning_rate": 1.797779733997521e-06,
      "loss": 0.29606573581695556,
      "num_input_tokens_seen": 1102567056,
      "step": 21425,
      "train_runtime": 131918.1764,
      "train_tokens_per_second": 8357.962
    },
    {
      "epoch": 1.4429033126851603,
      "grad_norm": 0.9752032172137891,
      "learning_rate": 1.7957495139785585e-06,
      "loss": 0.3197441577911377,
      "num_input_tokens_seen": 1102826760,
      "step": 21430,
      "train_runtime": 131951.1582,
      "train_tokens_per_second": 8357.841
    },
    {
      "epoch": 1.4432399676811203,
      "grad_norm": 0.7019079433869365,
      "learning_rate": 1.7937201900180228e-06,
      "loss": 0.29850196838378906,
      "num_input_tokens_seen": 1103054104,
      "step": 21435,
      "train_runtime": 131978.6604,
      "train_tokens_per_second": 8357.822
    },
    {
      "epoch": 1.4435766226770805,
      "grad_norm": 0.689427324124126,
      "learning_rate": 1.7916917626834134e-06,
      "loss": 0.33110685348510743,
      "num_input_tokens_seen": 1103318728,
      "step": 21440,
      "train_runtime": 132011.4174,
      "train_tokens_per_second": 8357.752
    },
    {
      "epoch": 1.4439132776730408,
      "grad_norm": 0.6466828903693559,
      "learning_rate": 1.7896642325419722e-06,
      "loss": 0.3111570835113525,
      "num_input_tokens_seen": 1103579520,
      "step": 21445,
      "train_runtime": 132042.3159,
      "train_tokens_per_second": 8357.772
    },
    {
      "epoch": 1.4442499326690008,
      "grad_norm": 0.7673582650185317,
      "learning_rate": 1.7876376001606915e-06,
      "loss": 0.2995492458343506,
      "num_input_tokens_seen": 1103823136,
      "step": 21450,
      "train_runtime": 132068.7103,
      "train_tokens_per_second": 8357.946
    },
    {
      "epoch": 1.4445865876649608,
      "grad_norm": 0.6589140259368582,
      "learning_rate": 1.7856118661063128e-06,
      "loss": 0.28098621368408205,
      "num_input_tokens_seen": 1104093824,
      "step": 21455,
      "train_runtime": 132100.6819,
      "train_tokens_per_second": 8357.972
    },
    {
      "epoch": 1.444923242660921,
      "grad_norm": 0.6360346059916626,
      "learning_rate": 1.7835870309453251e-06,
      "loss": 0.30194740295410155,
      "num_input_tokens_seen": 1104341656,
      "step": 21460,
      "train_runtime": 132128.7869,
      "train_tokens_per_second": 8358.07
    },
    {
      "epoch": 1.4452598976568813,
      "grad_norm": 0.7605703979828327,
      "learning_rate": 1.7815630952439667e-06,
      "loss": 0.30184597969055177,
      "num_input_tokens_seen": 1104584376,
      "step": 21465,
      "train_runtime": 132160.7177,
      "train_tokens_per_second": 8357.887
    },
    {
      "epoch": 1.4455965526528414,
      "grad_norm": 0.6227524554977869,
      "learning_rate": 1.7795400595682289e-06,
      "loss": 0.30388836860656737,
      "num_input_tokens_seen": 1104824480,
      "step": 21470,
      "train_runtime": 132192.3587,
      "train_tokens_per_second": 8357.703
    },
    {
      "epoch": 1.4459332076488014,
      "grad_norm": 0.6326347123016764,
      "learning_rate": 1.7775179244838408e-06,
      "loss": 0.3347285270690918,
      "num_input_tokens_seen": 1105074456,
      "step": 21475,
      "train_runtime": 132228.9633,
      "train_tokens_per_second": 8357.28
    },
    {
      "epoch": 1.4462698626447617,
      "grad_norm": 0.6700568328794966,
      "learning_rate": 1.7754966905562925e-06,
      "loss": 0.2945703983306885,
      "num_input_tokens_seen": 1105304688,
      "step": 21480,
      "train_runtime": 132256.8565,
      "train_tokens_per_second": 8357.258
    },
    {
      "epoch": 1.446606517640722,
      "grad_norm": 0.7419891967906372,
      "learning_rate": 1.7734763583508098e-06,
      "loss": 0.30434494018554686,
      "num_input_tokens_seen": 1105562344,
      "step": 21485,
      "train_runtime": 132290.6678,
      "train_tokens_per_second": 8357.07
    },
    {
      "epoch": 1.446943172636682,
      "grad_norm": 0.6640690512611899,
      "learning_rate": 1.7714569284323757e-06,
      "loss": 0.3147833824157715,
      "num_input_tokens_seen": 1105846032,
      "step": 21490,
      "train_runtime": 132323.5301,
      "train_tokens_per_second": 8357.138
    },
    {
      "epoch": 1.447279827632642,
      "grad_norm": 0.7394265379688365,
      "learning_rate": 1.7694384013657173e-06,
      "loss": 0.2976743221282959,
      "num_input_tokens_seen": 1106098504,
      "step": 21495,
      "train_runtime": 132353.6911,
      "train_tokens_per_second": 8357.141
    },
    {
      "epoch": 1.4476164826286022,
      "grad_norm": 0.642341907334742,
      "learning_rate": 1.767420777715308e-06,
      "loss": 0.3167498826980591,
      "num_input_tokens_seen": 1106377376,
      "step": 21500,
      "train_runtime": 132385.1149,
      "train_tokens_per_second": 8357.264
    },
    {
      "epoch": 1.4479531376245625,
      "grad_norm": 0.643683589527546,
      "learning_rate": 1.7654040580453702e-06,
      "loss": 0.2906998872756958,
      "num_input_tokens_seen": 1106653880,
      "step": 21505,
      "train_runtime": 132418.5674,
      "train_tokens_per_second": 8357.241
    },
    {
      "epoch": 1.4482897926205225,
      "grad_norm": 0.5551967223075788,
      "learning_rate": 1.7633882429198734e-06,
      "loss": 0.31531763076782227,
      "num_input_tokens_seen": 1106931728,
      "step": 21510,
      "train_runtime": 132450.2693,
      "train_tokens_per_second": 8357.338
    },
    {
      "epoch": 1.4486264476164825,
      "grad_norm": 0.7023031968372223,
      "learning_rate": 1.7613733329025318e-06,
      "loss": 0.3006064653396606,
      "num_input_tokens_seen": 1107205144,
      "step": 21515,
      "train_runtime": 132481.9879,
      "train_tokens_per_second": 8357.401
    },
    {
      "epoch": 1.4489631026124428,
      "grad_norm": 0.5884159351720257,
      "learning_rate": 1.759359328556814e-06,
      "loss": 0.3067363977432251,
      "num_input_tokens_seen": 1107473968,
      "step": 21520,
      "train_runtime": 132513.2377,
      "train_tokens_per_second": 8357.459
    },
    {
      "epoch": 1.449299757608403,
      "grad_norm": 0.672731963899376,
      "learning_rate": 1.7573462304459227e-06,
      "loss": 0.30321426391601564,
      "num_input_tokens_seen": 1107741000,
      "step": 21525,
      "train_runtime": 132543.9362,
      "train_tokens_per_second": 8357.538
    },
    {
      "epoch": 1.449636412604363,
      "grad_norm": 0.7091441796833302,
      "learning_rate": 1.7553340391328216e-06,
      "loss": 0.29336156845092776,
      "num_input_tokens_seen": 1107995008,
      "step": 21530,
      "train_runtime": 132574.2974,
      "train_tokens_per_second": 8357.54
    },
    {
      "epoch": 1.449973067600323,
      "grad_norm": 0.6173418617318239,
      "learning_rate": 1.7533227551802068e-06,
      "loss": 0.3004711866378784,
      "num_input_tokens_seen": 1108269408,
      "step": 21535,
      "train_runtime": 132605.1107,
      "train_tokens_per_second": 8357.667
    },
    {
      "epoch": 1.4503097225962833,
      "grad_norm": 0.6593514082355765,
      "learning_rate": 1.7513123791505321e-06,
      "loss": 0.32212080955505373,
      "num_input_tokens_seen": 1108503928,
      "step": 21540,
      "train_runtime": 132633.969,
      "train_tokens_per_second": 8357.617
    },
    {
      "epoch": 1.4506463775922436,
      "grad_norm": 0.604290442848206,
      "learning_rate": 1.7493029116059913e-06,
      "loss": 0.31188812255859377,
      "num_input_tokens_seen": 1108750016,
      "step": 21545,
      "train_runtime": 132665.2756,
      "train_tokens_per_second": 8357.5
    },
    {
      "epoch": 1.4509830325882036,
      "grad_norm": 0.6494852780535535,
      "learning_rate": 1.7472943531085252e-06,
      "loss": 0.29781789779663087,
      "num_input_tokens_seen": 1109016992,
      "step": 21550,
      "train_runtime": 132696.8584,
      "train_tokens_per_second": 8357.523
    },
    {
      "epoch": 1.4513196875841636,
      "grad_norm": 0.6680083024011922,
      "learning_rate": 1.7452867042198212e-06,
      "loss": 0.2774949073791504,
      "num_input_tokens_seen": 1109258872,
      "step": 21555,
      "train_runtime": 132728.085,
      "train_tokens_per_second": 8357.379
    },
    {
      "epoch": 1.4516563425801239,
      "grad_norm": 0.5644592553244968,
      "learning_rate": 1.7432799655013122e-06,
      "loss": 0.2818763732910156,
      "num_input_tokens_seen": 1109520864,
      "step": 21560,
      "train_runtime": 132759.9032,
      "train_tokens_per_second": 8357.349
    },
    {
      "epoch": 1.4519929975760841,
      "grad_norm": 0.7564845596028599,
      "learning_rate": 1.741274137514174e-06,
      "loss": 0.3203913450241089,
      "num_input_tokens_seen": 1109764424,
      "step": 21565,
      "train_runtime": 132788.0842,
      "train_tokens_per_second": 8357.41
    },
    {
      "epoch": 1.4523296525720442,
      "grad_norm": 0.6096956691288908,
      "learning_rate": 1.7392692208193351e-06,
      "loss": 0.2873049259185791,
      "num_input_tokens_seen": 1110016264,
      "step": 21570,
      "train_runtime": 132821.7899,
      "train_tokens_per_second": 8357.185
    },
    {
      "epoch": 1.4526663075680042,
      "grad_norm": 0.8323818483598184,
      "learning_rate": 1.7372652159774578e-06,
      "loss": 0.26852295398712156,
      "num_input_tokens_seen": 1110274256,
      "step": 21575,
      "train_runtime": 132853.9567,
      "train_tokens_per_second": 8357.103
    },
    {
      "epoch": 1.4530029625639644,
      "grad_norm": 0.6740738393326057,
      "learning_rate": 1.7352621235489625e-06,
      "loss": 0.29173738956451417,
      "num_input_tokens_seen": 1110528632,
      "step": 21580,
      "train_runtime": 132885.8449,
      "train_tokens_per_second": 8357.012
    },
    {
      "epoch": 1.4533396175599247,
      "grad_norm": 0.7487156118900867,
      "learning_rate": 1.733259944094e-06,
      "loss": 0.31221718788146974,
      "num_input_tokens_seen": 1110787832,
      "step": 21585,
      "train_runtime": 132915.5168,
      "train_tokens_per_second": 8357.097
    },
    {
      "epoch": 1.4536762725558847,
      "grad_norm": 0.776835762133548,
      "learning_rate": 1.7312586781724789e-06,
      "loss": 0.3010872364044189,
      "num_input_tokens_seen": 1111048816,
      "step": 21590,
      "train_runtime": 132946.7228,
      "train_tokens_per_second": 8357.098
    },
    {
      "epoch": 1.4540129275518447,
      "grad_norm": 0.6429627048295165,
      "learning_rate": 1.7292583263440454e-06,
      "loss": 0.3124294996261597,
      "num_input_tokens_seen": 1111299096,
      "step": 21595,
      "train_runtime": 132978.0711,
      "train_tokens_per_second": 8357.01
    },
    {
      "epoch": 1.454349582547805,
      "grad_norm": 0.6810128304287453,
      "learning_rate": 1.7272588891680914e-06,
      "loss": 0.28821558952331544,
      "num_input_tokens_seen": 1111530216,
      "step": 21600,
      "train_runtime": 133007.8723,
      "train_tokens_per_second": 8356.875
    },
    {
      "epoch": 1.4546862375437652,
      "grad_norm": 0.638780074329626,
      "learning_rate": 1.7252603672037527e-06,
      "loss": 0.2918674945831299,
      "num_input_tokens_seen": 1111782528,
      "step": 21605,
      "train_runtime": 133035.498,
      "train_tokens_per_second": 8357.037
    },
    {
      "epoch": 1.4550228925397253,
      "grad_norm": 0.6316198935439115,
      "learning_rate": 1.72326276100991e-06,
      "loss": 0.28084726333618165,
      "num_input_tokens_seen": 1112052968,
      "step": 21610,
      "train_runtime": 133070.0081,
      "train_tokens_per_second": 8356.902
    },
    {
      "epoch": 1.4553595475356853,
      "grad_norm": 0.6981509984213002,
      "learning_rate": 1.721266071145186e-06,
      "loss": 0.31381940841674805,
      "num_input_tokens_seen": 1112320752,
      "step": 21615,
      "train_runtime": 133097.3257,
      "train_tokens_per_second": 8357.198
    },
    {
      "epoch": 1.4556962025316456,
      "grad_norm": 0.6213059901188921,
      "learning_rate": 1.7192702981679531e-06,
      "loss": 0.3121327877044678,
      "num_input_tokens_seen": 1112603200,
      "step": 21620,
      "train_runtime": 133124.557,
      "train_tokens_per_second": 8357.611
    },
    {
      "epoch": 1.4560328575276058,
      "grad_norm": 0.6417522232574675,
      "learning_rate": 1.717275442636317e-06,
      "loss": 0.27664613723754883,
      "num_input_tokens_seen": 1112858832,
      "step": 21625,
      "train_runtime": 133156.4355,
      "train_tokens_per_second": 8357.529
    },
    {
      "epoch": 1.4563695125235658,
      "grad_norm": 0.6809535911708648,
      "learning_rate": 1.7152815051081384e-06,
      "loss": 0.31713180541992186,
      "num_input_tokens_seen": 1113118488,
      "step": 21630,
      "train_runtime": 133184.268,
      "train_tokens_per_second": 8357.733
    },
    {
      "epoch": 1.4567061675195259,
      "grad_norm": 0.5156761062856282,
      "learning_rate": 1.7132884861410093e-06,
      "loss": 0.3103309154510498,
      "num_input_tokens_seen": 1113366128,
      "step": 21635,
      "train_runtime": 133213.0608,
      "train_tokens_per_second": 8357.785
    },
    {
      "epoch": 1.4570428225154861,
      "grad_norm": 0.6180767098187174,
      "learning_rate": 1.7112963862922766e-06,
      "loss": 0.2871995449066162,
      "num_input_tokens_seen": 1113616104,
      "step": 21640,
      "train_runtime": 133239.3715,
      "train_tokens_per_second": 8358.011
    },
    {
      "epoch": 1.4573794775114464,
      "grad_norm": 0.6208414348115595,
      "learning_rate": 1.7093052061190224e-06,
      "loss": 0.3135540962219238,
      "num_input_tokens_seen": 1113896608,
      "step": 21645,
      "train_runtime": 133269.3169,
      "train_tokens_per_second": 8358.238
    },
    {
      "epoch": 1.4577161325074064,
      "grad_norm": 0.6762031266684863,
      "learning_rate": 1.7073149461780742e-06,
      "loss": 0.30667529106140134,
      "num_input_tokens_seen": 1114185960,
      "step": 21650,
      "train_runtime": 133304.7119,
      "train_tokens_per_second": 8358.189
    },
    {
      "epoch": 1.4580527875033664,
      "grad_norm": 0.7430403083516292,
      "learning_rate": 1.705325607026001e-06,
      "loss": 0.3093313217163086,
      "num_input_tokens_seen": 1114424624,
      "step": 21655,
      "train_runtime": 133335.9392,
      "train_tokens_per_second": 8358.021
    },
    {
      "epoch": 1.4583894424993267,
      "grad_norm": 0.6770901728239235,
      "learning_rate": 1.7033371892191164e-06,
      "loss": 0.30808780193328855,
      "num_input_tokens_seen": 1114694416,
      "step": 21660,
      "train_runtime": 133368.4121,
      "train_tokens_per_second": 8358.009
    },
    {
      "epoch": 1.458726097495287,
      "grad_norm": 0.7401452970352431,
      "learning_rate": 1.7013496933134726e-06,
      "loss": 0.3069624662399292,
      "num_input_tokens_seen": 1114933680,
      "step": 21665,
      "train_runtime": 133396.4397,
      "train_tokens_per_second": 8358.047
    },
    {
      "epoch": 1.459062752491247,
      "grad_norm": 0.6683752751329867,
      "learning_rate": 1.6993631198648724e-06,
      "loss": 0.32300219535827634,
      "num_input_tokens_seen": 1115210560,
      "step": 21670,
      "train_runtime": 133426.273,
      "train_tokens_per_second": 8358.253
    },
    {
      "epoch": 1.459399407487207,
      "grad_norm": 0.6248576121681623,
      "learning_rate": 1.697377469428847e-06,
      "loss": 0.2937227725982666,
      "num_input_tokens_seen": 1115476672,
      "step": 21675,
      "train_runtime": 133459.6535,
      "train_tokens_per_second": 8358.157
    },
    {
      "epoch": 1.4597360624831672,
      "grad_norm": 0.6093574926304915,
      "learning_rate": 1.6953927425606842e-06,
      "loss": 0.322925329208374,
      "num_input_tokens_seen": 1115730064,
      "step": 21680,
      "train_runtime": 133492.3358,
      "train_tokens_per_second": 8358.008
    },
    {
      "epoch": 1.4600727174791275,
      "grad_norm": 0.6201220048288818,
      "learning_rate": 1.6934089398154002e-06,
      "loss": 0.2789339542388916,
      "num_input_tokens_seen": 1115989000,
      "step": 21685,
      "train_runtime": 133520.8092,
      "train_tokens_per_second": 8358.165
    },
    {
      "epoch": 1.4604093724750875,
      "grad_norm": 0.6111632901577095,
      "learning_rate": 1.6914260617477634e-06,
      "loss": 0.2883721113204956,
      "num_input_tokens_seen": 1116269888,
      "step": 21690,
      "train_runtime": 133548.4861,
      "train_tokens_per_second": 8358.536
    },
    {
      "epoch": 1.4607460274710475,
      "grad_norm": 0.5768021756400129,
      "learning_rate": 1.6894441089122788e-06,
      "loss": 0.2706777811050415,
      "num_input_tokens_seen": 1116524312,
      "step": 21695,
      "train_runtime": 133578.9169,
      "train_tokens_per_second": 8358.537
    },
    {
      "epoch": 1.4610826824670078,
      "grad_norm": 0.7023480969703217,
      "learning_rate": 1.6874630818631926e-06,
      "loss": 0.31003992557525634,
      "num_input_tokens_seen": 1116777856,
      "step": 21700,
      "train_runtime": 133609.2706,
      "train_tokens_per_second": 8358.536
    },
    {
      "epoch": 1.461419337462968,
      "grad_norm": 0.7150237907890772,
      "learning_rate": 1.685482981154492e-06,
      "loss": 0.31594417095184324,
      "num_input_tokens_seen": 1117043272,
      "step": 21705,
      "train_runtime": 133636.0,
      "train_tokens_per_second": 8358.85
    },
    {
      "epoch": 1.461755992458928,
      "grad_norm": 0.6406091214460472,
      "learning_rate": 1.683503807339907e-06,
      "loss": 0.29458544254302976,
      "num_input_tokens_seen": 1117301776,
      "step": 21710,
      "train_runtime": 133668.8701,
      "train_tokens_per_second": 8358.728
    },
    {
      "epoch": 1.462092647454888,
      "grad_norm": 0.5929353571838196,
      "learning_rate": 1.6815255609729047e-06,
      "loss": 0.3022796154022217,
      "num_input_tokens_seen": 1117570800,
      "step": 21715,
      "train_runtime": 133700.1259,
      "train_tokens_per_second": 8358.786
    },
    {
      "epoch": 1.4624293024508483,
      "grad_norm": 0.753451751437872,
      "learning_rate": 1.6795482426067012e-06,
      "loss": 0.29369215965270995,
      "num_input_tokens_seen": 1117824992,
      "step": 21720,
      "train_runtime": 133730.1092,
      "train_tokens_per_second": 8358.813
    },
    {
      "epoch": 1.4627659574468086,
      "grad_norm": 0.6986441963844277,
      "learning_rate": 1.6775718527942403e-06,
      "loss": 0.28773884773254393,
      "num_input_tokens_seen": 1118046592,
      "step": 21725,
      "train_runtime": 133761.3461,
      "train_tokens_per_second": 8358.518
    },
    {
      "epoch": 1.4631026124427686,
      "grad_norm": 0.683799197241156,
      "learning_rate": 1.675596392088219e-06,
      "loss": 0.3110309600830078,
      "num_input_tokens_seen": 1118322784,
      "step": 21730,
      "train_runtime": 133792.4559,
      "train_tokens_per_second": 8358.639
    },
    {
      "epoch": 1.4634392674387287,
      "grad_norm": 0.5766543571947229,
      "learning_rate": 1.6736218610410633e-06,
      "loss": 0.285445237159729,
      "num_input_tokens_seen": 1118595880,
      "step": 21735,
      "train_runtime": 133824.9155,
      "train_tokens_per_second": 8358.652
    },
    {
      "epoch": 1.463775922434689,
      "grad_norm": 0.6544324223063842,
      "learning_rate": 1.6716482602049483e-06,
      "loss": 0.29945979118347166,
      "num_input_tokens_seen": 1118846960,
      "step": 21740,
      "train_runtime": 133855.9805,
      "train_tokens_per_second": 8358.588
    },
    {
      "epoch": 1.4641125774306492,
      "grad_norm": 0.6789699340563715,
      "learning_rate": 1.669675590131784e-06,
      "loss": 0.3022740602493286,
      "num_input_tokens_seen": 1119082536,
      "step": 21745,
      "train_runtime": 133884.0125,
      "train_tokens_per_second": 8358.597
    },
    {
      "epoch": 1.4644492324266092,
      "grad_norm": 0.7513769771282448,
      "learning_rate": 1.6677038513732214e-06,
      "loss": 0.30121960639953616,
      "num_input_tokens_seen": 1119345200,
      "step": 21750,
      "train_runtime": 133914.1548,
      "train_tokens_per_second": 8358.677
    },
    {
      "epoch": 1.4647858874225694,
      "grad_norm": 0.6862410059093264,
      "learning_rate": 1.6657330444806508e-06,
      "loss": 0.3175745725631714,
      "num_input_tokens_seen": 1119601560,
      "step": 21755,
      "train_runtime": 133943.5539,
      "train_tokens_per_second": 8358.757
    },
    {
      "epoch": 1.4651225424185295,
      "grad_norm": 0.62646264670656,
      "learning_rate": 1.6637631700052021e-06,
      "loss": 0.2980278253555298,
      "num_input_tokens_seen": 1119887992,
      "step": 21760,
      "train_runtime": 133976.6443,
      "train_tokens_per_second": 8358.83
    },
    {
      "epoch": 1.4654591974144897,
      "grad_norm": 0.6846120431312245,
      "learning_rate": 1.6617942284977428e-06,
      "loss": 0.29366674423217776,
      "num_input_tokens_seen": 1120180120,
      "step": 21765,
      "train_runtime": 134005.715,
      "train_tokens_per_second": 8359.197
    },
    {
      "epoch": 1.4657958524104497,
      "grad_norm": 0.7289715537530498,
      "learning_rate": 1.6598262205088862e-06,
      "loss": 0.3128157377243042,
      "num_input_tokens_seen": 1120452408,
      "step": 21770,
      "train_runtime": 134036.176,
      "train_tokens_per_second": 8359.328
    },
    {
      "epoch": 1.46613250740641,
      "grad_norm": 0.7007526795910939,
      "learning_rate": 1.6578591465889722e-06,
      "loss": 0.30434188842773435,
      "num_input_tokens_seen": 1120725464,
      "step": 21775,
      "train_runtime": 134070.648,
      "train_tokens_per_second": 8359.216
    },
    {
      "epoch": 1.46646916240237,
      "grad_norm": 0.7096527544769341,
      "learning_rate": 1.6558930072880942e-06,
      "loss": 0.3101084232330322,
      "num_input_tokens_seen": 1120994096,
      "step": 21780,
      "train_runtime": 134098.7656,
      "train_tokens_per_second": 8359.466
    },
    {
      "epoch": 1.4668058173983303,
      "grad_norm": 0.6827917901534096,
      "learning_rate": 1.6539278031560695e-06,
      "loss": 0.31211166381835936,
      "num_input_tokens_seen": 1121239176,
      "step": 21785,
      "train_runtime": 134129.582,
      "train_tokens_per_second": 8359.373
    },
    {
      "epoch": 1.4671424723942903,
      "grad_norm": 0.8197064064646364,
      "learning_rate": 1.6519635347424672e-06,
      "loss": 0.29163532257080077,
      "num_input_tokens_seen": 1121493088,
      "step": 21790,
      "train_runtime": 134155.3926,
      "train_tokens_per_second": 8359.657
    },
    {
      "epoch": 1.4674791273902505,
      "grad_norm": 0.6928214640160376,
      "learning_rate": 1.650000202596586e-06,
      "loss": 0.3295187473297119,
      "num_input_tokens_seen": 1121771176,
      "step": 21795,
      "train_runtime": 134184.7057,
      "train_tokens_per_second": 8359.903
    },
    {
      "epoch": 1.4678157823862106,
      "grad_norm": 0.6973196233564714,
      "learning_rate": 1.6480378072674659e-06,
      "loss": 0.30800912380218504,
      "num_input_tokens_seen": 1122040848,
      "step": 21800,
      "train_runtime": 134214.1032,
      "train_tokens_per_second": 8360.082
    },
    {
      "epoch": 1.4681524373821708,
      "grad_norm": 0.675849601991147,
      "learning_rate": 1.646076349303884e-06,
      "loss": 0.2947542667388916,
      "num_input_tokens_seen": 1122297744,
      "step": 21805,
      "train_runtime": 134242.3743,
      "train_tokens_per_second": 8360.235
    },
    {
      "epoch": 1.4684890923781309,
      "grad_norm": 0.7286966587452776,
      "learning_rate": 1.6441158292543557e-06,
      "loss": 0.29759912490844725,
      "num_input_tokens_seen": 1122540584,
      "step": 21810,
      "train_runtime": 134275.2013,
      "train_tokens_per_second": 8359.999
    },
    {
      "epoch": 1.468825747374091,
      "grad_norm": 0.6378531541504576,
      "learning_rate": 1.6421562476671339e-06,
      "loss": 0.3271633148193359,
      "num_input_tokens_seen": 1122801608,
      "step": 21815,
      "train_runtime": 134307.6435,
      "train_tokens_per_second": 8359.923
    },
    {
      "epoch": 1.4691624023700511,
      "grad_norm": 0.5822558741741543,
      "learning_rate": 1.640197605090213e-06,
      "loss": 0.2794620513916016,
      "num_input_tokens_seen": 1123061024,
      "step": 21820,
      "train_runtime": 134339.9225,
      "train_tokens_per_second": 8359.846
    },
    {
      "epoch": 1.4694990573660114,
      "grad_norm": 0.6718105602377253,
      "learning_rate": 1.6382399020713146e-06,
      "loss": 0.2918576717376709,
      "num_input_tokens_seen": 1123293320,
      "step": 21825,
      "train_runtime": 134372.1809,
      "train_tokens_per_second": 8359.568
    },
    {
      "epoch": 1.4698357123619714,
      "grad_norm": 0.8151754094781176,
      "learning_rate": 1.6362831391579115e-06,
      "loss": 0.29859471321105957,
      "num_input_tokens_seen": 1123540776,
      "step": 21830,
      "train_runtime": 134404.4027,
      "train_tokens_per_second": 8359.405
    },
    {
      "epoch": 1.4701723673579317,
      "grad_norm": 0.6560330433791852,
      "learning_rate": 1.634327316897199e-06,
      "loss": 0.3015228509902954,
      "num_input_tokens_seen": 1123786880,
      "step": 21835,
      "train_runtime": 134432.056,
      "train_tokens_per_second": 8359.516
    },
    {
      "epoch": 1.4705090223538917,
      "grad_norm": 0.651315044295407,
      "learning_rate": 1.6323724358361216e-06,
      "loss": 0.28355405330657957,
      "num_input_tokens_seen": 1124028768,
      "step": 21840,
      "train_runtime": 134461.4935,
      "train_tokens_per_second": 8359.484
    },
    {
      "epoch": 1.470845677349852,
      "grad_norm": 0.6888953822468835,
      "learning_rate": 1.6304184965213537e-06,
      "loss": 0.29400811195373533,
      "num_input_tokens_seen": 1124271432,
      "step": 21845,
      "train_runtime": 134487.4097,
      "train_tokens_per_second": 8359.678
    },
    {
      "epoch": 1.471182332345812,
      "grad_norm": 0.6807922554309767,
      "learning_rate": 1.6284654994993088e-06,
      "loss": 0.30544424057006836,
      "num_input_tokens_seen": 1124550048,
      "step": 21850,
      "train_runtime": 134516.7692,
      "train_tokens_per_second": 8359.925
    },
    {
      "epoch": 1.4715189873417722,
      "grad_norm": 0.696441508342031,
      "learning_rate": 1.6265134453161358e-06,
      "loss": 0.30437135696411133,
      "num_input_tokens_seen": 1124794792,
      "step": 21855,
      "train_runtime": 134548.1459,
      "train_tokens_per_second": 8359.794
    },
    {
      "epoch": 1.4718556423377323,
      "grad_norm": 0.7008552418492205,
      "learning_rate": 1.6245623345177203e-06,
      "loss": 0.3018461227416992,
      "num_input_tokens_seen": 1125026096,
      "step": 21860,
      "train_runtime": 134580.4824,
      "train_tokens_per_second": 8359.504
    },
    {
      "epoch": 1.4721922973336925,
      "grad_norm": 0.566444886645963,
      "learning_rate": 1.622612167649683e-06,
      "loss": 0.2781670093536377,
      "num_input_tokens_seen": 1125274968,
      "step": 21865,
      "train_runtime": 134610.454,
      "train_tokens_per_second": 8359.492
    },
    {
      "epoch": 1.4725289523296525,
      "grad_norm": 0.5340278243000767,
      "learning_rate": 1.6206629452573868e-06,
      "loss": 0.26104638576507566,
      "num_input_tokens_seen": 1125548568,
      "step": 21870,
      "train_runtime": 134642.8192,
      "train_tokens_per_second": 8359.514
    },
    {
      "epoch": 1.4728656073256128,
      "grad_norm": 0.6336792361999635,
      "learning_rate": 1.618714667885919e-06,
      "loss": 0.28565826416015627,
      "num_input_tokens_seen": 1125801336,
      "step": 21875,
      "train_runtime": 134675.4338,
      "train_tokens_per_second": 8359.367
    },
    {
      "epoch": 1.4732022623215728,
      "grad_norm": 0.6033098987231856,
      "learning_rate": 1.6167673360801151e-06,
      "loss": 0.29342403411865237,
      "num_input_tokens_seen": 1126080896,
      "step": 21880,
      "train_runtime": 134710.1918,
      "train_tokens_per_second": 8359.285
    },
    {
      "epoch": 1.473538917317533,
      "grad_norm": 0.7444874538757394,
      "learning_rate": 1.6148209503845336e-06,
      "loss": 0.3379021167755127,
      "num_input_tokens_seen": 1126330744,
      "step": 21885,
      "train_runtime": 134743.7513,
      "train_tokens_per_second": 8359.057
    },
    {
      "epoch": 1.473875572313493,
      "grad_norm": 0.6590175921057586,
      "learning_rate": 1.6128755113434808e-06,
      "loss": 0.28277461528778075,
      "num_input_tokens_seen": 1126584408,
      "step": 21890,
      "train_runtime": 134771.6405,
      "train_tokens_per_second": 8359.21
    },
    {
      "epoch": 1.4742122273094533,
      "grad_norm": 0.5984885282535287,
      "learning_rate": 1.6109310195009908e-06,
      "loss": 0.2949079036712646,
      "num_input_tokens_seen": 1126864440,
      "step": 21895,
      "train_runtime": 134802.7286,
      "train_tokens_per_second": 8359.359
    },
    {
      "epoch": 1.4745488823054134,
      "grad_norm": 0.7060265564671923,
      "learning_rate": 1.6089874754008323e-06,
      "loss": 0.2935539484024048,
      "num_input_tokens_seen": 1127105112,
      "step": 21900,
      "train_runtime": 134832.7976,
      "train_tokens_per_second": 8359.28
    },
    {
      "epoch": 1.4748855373013736,
      "grad_norm": 0.6959022472559766,
      "learning_rate": 1.6070448795865162e-06,
      "loss": 0.3041229248046875,
      "num_input_tokens_seen": 1127359872,
      "step": 21905,
      "train_runtime": 134866.3828,
      "train_tokens_per_second": 8359.087
    },
    {
      "epoch": 1.4752221922973336,
      "grad_norm": 0.635563647368659,
      "learning_rate": 1.605103232601277e-06,
      "loss": 0.2948932647705078,
      "num_input_tokens_seen": 1127611096,
      "step": 21910,
      "train_runtime": 134897.9019,
      "train_tokens_per_second": 8358.997
    },
    {
      "epoch": 1.475558847293294,
      "grad_norm": 0.5055916717376825,
      "learning_rate": 1.603162534988096e-06,
      "loss": 0.26693480014801024,
      "num_input_tokens_seen": 1127860160,
      "step": 21915,
      "train_runtime": 134928.5293,
      "train_tokens_per_second": 8358.945
    },
    {
      "epoch": 1.475895502289254,
      "grad_norm": 0.60371936044259,
      "learning_rate": 1.6012227872896791e-06,
      "loss": 0.28634166717529297,
      "num_input_tokens_seen": 1128114840,
      "step": 21920,
      "train_runtime": 134956.5042,
      "train_tokens_per_second": 8359.099
    },
    {
      "epoch": 1.4762321572852142,
      "grad_norm": 0.6154144406151074,
      "learning_rate": 1.599283990048472e-06,
      "loss": 0.3006128787994385,
      "num_input_tokens_seen": 1128359808,
      "step": 21925,
      "train_runtime": 134985.9705,
      "train_tokens_per_second": 8359.089
    },
    {
      "epoch": 1.4765688122811742,
      "grad_norm": 0.6861906838389773,
      "learning_rate": 1.5973461438066533e-06,
      "loss": 0.29517812728881837,
      "num_input_tokens_seen": 1128613232,
      "step": 21930,
      "train_runtime": 135016.3568,
      "train_tokens_per_second": 8359.085
    },
    {
      "epoch": 1.4769054672771345,
      "grad_norm": 0.7120921449728638,
      "learning_rate": 1.5954092491061351e-06,
      "loss": 0.29170522689819334,
      "num_input_tokens_seen": 1128858144,
      "step": 21935,
      "train_runtime": 135045.2695,
      "train_tokens_per_second": 8359.109
    },
    {
      "epoch": 1.4772421222730945,
      "grad_norm": 0.5997837178650863,
      "learning_rate": 1.5934733064885637e-06,
      "loss": 0.30538015365600585,
      "num_input_tokens_seen": 1129123584,
      "step": 21940,
      "train_runtime": 135076.2745,
      "train_tokens_per_second": 8359.156
    },
    {
      "epoch": 1.4775787772690547,
      "grad_norm": 0.6533201977204609,
      "learning_rate": 1.5915383164953203e-06,
      "loss": 0.30964365005493166,
      "num_input_tokens_seen": 1129362200,
      "step": 21945,
      "train_runtime": 135103.0862,
      "train_tokens_per_second": 8359.263
    },
    {
      "epoch": 1.4779154322650148,
      "grad_norm": 0.7393473270047337,
      "learning_rate": 1.5896042796675155e-06,
      "loss": 0.30026695728302,
      "num_input_tokens_seen": 1129602200,
      "step": 21950,
      "train_runtime": 135134.2656,
      "train_tokens_per_second": 8359.11
    },
    {
      "epoch": 1.478252087260975,
      "grad_norm": 0.7139130926529175,
      "learning_rate": 1.5876711965460023e-06,
      "loss": 0.30302140712738035,
      "num_input_tokens_seen": 1129862880,
      "step": 21955,
      "train_runtime": 135168.0224,
      "train_tokens_per_second": 8358.951
    },
    {
      "epoch": 1.478588742256935,
      "grad_norm": 0.6001835390418051,
      "learning_rate": 1.5857390676713546e-06,
      "loss": 0.29804272651672364,
      "num_input_tokens_seen": 1130100896,
      "step": 21960,
      "train_runtime": 135199.0149,
      "train_tokens_per_second": 8358.795
    },
    {
      "epoch": 1.4789253972528953,
      "grad_norm": 0.6378987163072389,
      "learning_rate": 1.5838078935838913e-06,
      "loss": 0.26572303771972655,
      "num_input_tokens_seen": 1130341952,
      "step": 21965,
      "train_runtime": 135227.3513,
      "train_tokens_per_second": 8358.826
    },
    {
      "epoch": 1.4792620522488553,
      "grad_norm": 0.6742579849624568,
      "learning_rate": 1.581877674823657e-06,
      "loss": 0.2932957410812378,
      "num_input_tokens_seen": 1130576384,
      "step": 21970,
      "train_runtime": 135257.8061,
      "train_tokens_per_second": 8358.678
    },
    {
      "epoch": 1.4795987072448156,
      "grad_norm": 0.6701137172998733,
      "learning_rate": 1.5799484119304308e-06,
      "loss": 0.2972468376159668,
      "num_input_tokens_seen": 1130834120,
      "step": 21975,
      "train_runtime": 135287.6285,
      "train_tokens_per_second": 8358.74
    },
    {
      "epoch": 1.4799353622407756,
      "grad_norm": 0.6248755736055632,
      "learning_rate": 1.578020105443726e-06,
      "loss": 0.28776533603668214,
      "num_input_tokens_seen": 1131094728,
      "step": 21980,
      "train_runtime": 135319.3992,
      "train_tokens_per_second": 8358.703
    },
    {
      "epoch": 1.4802720172367358,
      "grad_norm": 0.7427103568763174,
      "learning_rate": 1.576092755902787e-06,
      "loss": 0.30176410675048826,
      "num_input_tokens_seen": 1131348616,
      "step": 21985,
      "train_runtime": 135348.4956,
      "train_tokens_per_second": 8358.782
    },
    {
      "epoch": 1.4806086722326959,
      "grad_norm": 0.6273601348392186,
      "learning_rate": 1.5741663638465903e-06,
      "loss": 0.3162957191467285,
      "num_input_tokens_seen": 1131591568,
      "step": 21990,
      "train_runtime": 135380.705,
      "train_tokens_per_second": 8358.588
    },
    {
      "epoch": 1.4809453272286561,
      "grad_norm": 0.7449002337624812,
      "learning_rate": 1.5722409298138463e-06,
      "loss": 0.2865452766418457,
      "num_input_tokens_seen": 1131820216,
      "step": 21995,
      "train_runtime": 135414.2689,
      "train_tokens_per_second": 8358.205
    },
    {
      "epoch": 1.4812819822246162,
      "grad_norm": 0.7592922615809121,
      "learning_rate": 1.570316454342994e-06,
      "loss": 0.3187229156494141,
      "num_input_tokens_seen": 1132071920,
      "step": 22000,
      "train_runtime": 135446.2072,
      "train_tokens_per_second": 8358.092
    },
    {
      "epoch": 1.4816186372205764,
      "grad_norm": 0.6319825710427134,
      "learning_rate": 1.568392937972212e-06,
      "loss": 0.3112966775894165,
      "num_input_tokens_seen": 1132349864,
      "step": 22005,
      "train_runtime": 135479.0344,
      "train_tokens_per_second": 8358.119
    },
    {
      "epoch": 1.4819552922165364,
      "grad_norm": 0.622184816894414,
      "learning_rate": 1.5664703812393988e-06,
      "loss": 0.29169373512268065,
      "num_input_tokens_seen": 1132627184,
      "step": 22010,
      "train_runtime": 135509.2398,
      "train_tokens_per_second": 8358.302
    },
    {
      "epoch": 1.4822919472124967,
      "grad_norm": 0.6989696380463097,
      "learning_rate": 1.5645487846821967e-06,
      "loss": 0.2974118232727051,
      "num_input_tokens_seen": 1132885776,
      "step": 22015,
      "train_runtime": 135541.9932,
      "train_tokens_per_second": 8358.19
    },
    {
      "epoch": 1.4826286022084567,
      "grad_norm": 0.724986366530846,
      "learning_rate": 1.5626281488379724e-06,
      "loss": 0.3043977737426758,
      "num_input_tokens_seen": 1133147328,
      "step": 22020,
      "train_runtime": 135568.1884,
      "train_tokens_per_second": 8358.505
    },
    {
      "epoch": 1.482965257204417,
      "grad_norm": 0.6635690715126672,
      "learning_rate": 1.560708474243825e-06,
      "loss": 0.2851078510284424,
      "num_input_tokens_seen": 1133395216,
      "step": 22025,
      "train_runtime": 135596.006,
      "train_tokens_per_second": 8358.618
    },
    {
      "epoch": 1.483301912200377,
      "grad_norm": 0.7164037094642488,
      "learning_rate": 1.5587897614365854e-06,
      "loss": 0.3000472545623779,
      "num_input_tokens_seen": 1133684184,
      "step": 22030,
      "train_runtime": 135621.2124,
      "train_tokens_per_second": 8359.195
    },
    {
      "epoch": 1.4836385671963372,
      "grad_norm": 0.7424740225086487,
      "learning_rate": 1.5568720109528163e-06,
      "loss": 0.2932755947113037,
      "num_input_tokens_seen": 1133937352,
      "step": 22035,
      "train_runtime": 135650.3527,
      "train_tokens_per_second": 8359.266
    },
    {
      "epoch": 1.4839752221922973,
      "grad_norm": 0.7651509698668074,
      "learning_rate": 1.5549552233288095e-06,
      "loss": 0.3283426761627197,
      "num_input_tokens_seen": 1134195760,
      "step": 22040,
      "train_runtime": 135681.6526,
      "train_tokens_per_second": 8359.242
    },
    {
      "epoch": 1.4843118771882575,
      "grad_norm": 0.6132473153472802,
      "learning_rate": 1.5530393991005893e-06,
      "loss": 0.30064690113067627,
      "num_input_tokens_seen": 1134439928,
      "step": 22045,
      "train_runtime": 135710.3615,
      "train_tokens_per_second": 8359.273
    },
    {
      "epoch": 1.4846485321842176,
      "grad_norm": 0.7041710795549715,
      "learning_rate": 1.551124538803908e-06,
      "loss": 0.2949697971343994,
      "num_input_tokens_seen": 1134710216,
      "step": 22050,
      "train_runtime": 135742.0577,
      "train_tokens_per_second": 8359.312
    },
    {
      "epoch": 1.4849851871801778,
      "grad_norm": 0.6795619638301715,
      "learning_rate": 1.5492106429742547e-06,
      "loss": 0.30388355255126953,
      "num_input_tokens_seen": 1134974544,
      "step": 22055,
      "train_runtime": 135770.8275,
      "train_tokens_per_second": 8359.488
    },
    {
      "epoch": 1.4853218421761378,
      "grad_norm": 0.6692486484965178,
      "learning_rate": 1.5472977121468386e-06,
      "loss": 0.29910898208618164,
      "num_input_tokens_seen": 1135257872,
      "step": 22060,
      "train_runtime": 135802.7338,
      "train_tokens_per_second": 8359.61
    },
    {
      "epoch": 1.485658497172098,
      "grad_norm": 0.738541825549814,
      "learning_rate": 1.545385746856609e-06,
      "loss": 0.30290651321411133,
      "num_input_tokens_seen": 1135512136,
      "step": 22065,
      "train_runtime": 135835.6497,
      "train_tokens_per_second": 8359.456
    },
    {
      "epoch": 1.485995152168058,
      "grad_norm": 0.7479115831963229,
      "learning_rate": 1.5434747476382388e-06,
      "loss": 0.2726182460784912,
      "num_input_tokens_seen": 1135793456,
      "step": 22070,
      "train_runtime": 135863.2731,
      "train_tokens_per_second": 8359.827
    },
    {
      "epoch": 1.4863318071640184,
      "grad_norm": 0.723057693928631,
      "learning_rate": 1.5415647150261337e-06,
      "loss": 0.2920900821685791,
      "num_input_tokens_seen": 1136032920,
      "step": 22075,
      "train_runtime": 135893.9062,
      "train_tokens_per_second": 8359.705
    },
    {
      "epoch": 1.4866684621599784,
      "grad_norm": 0.6586509878103383,
      "learning_rate": 1.539655649554428e-06,
      "loss": 0.29311509132385255,
      "num_input_tokens_seen": 1136281800,
      "step": 22080,
      "train_runtime": 135928.9587,
      "train_tokens_per_second": 8359.38
    },
    {
      "epoch": 1.4870051171559386,
      "grad_norm": 0.6824665927635349,
      "learning_rate": 1.5377475517569856e-06,
      "loss": 0.2945427894592285,
      "num_input_tokens_seen": 1136517176,
      "step": 22085,
      "train_runtime": 135956.7664,
      "train_tokens_per_second": 8359.401
    },
    {
      "epoch": 1.4873417721518987,
      "grad_norm": 0.713635211061795,
      "learning_rate": 1.5358404221674e-06,
      "loss": 0.2975715160369873,
      "num_input_tokens_seen": 1136748064,
      "step": 22090,
      "train_runtime": 135985.2431,
      "train_tokens_per_second": 8359.349
    },
    {
      "epoch": 1.487678427147859,
      "grad_norm": 0.6955820474979858,
      "learning_rate": 1.533934261318994e-06,
      "loss": 0.30663821697235105,
      "num_input_tokens_seen": 1137011216,
      "step": 22095,
      "train_runtime": 136016.2844,
      "train_tokens_per_second": 8359.376
    },
    {
      "epoch": 1.488015082143819,
      "grad_norm": 0.6796553391075357,
      "learning_rate": 1.5320290697448177e-06,
      "loss": 0.28195343017578123,
      "num_input_tokens_seen": 1137258032,
      "step": 22100,
      "train_runtime": 136042.8269,
      "train_tokens_per_second": 8359.559
    },
    {
      "epoch": 1.4883517371397792,
      "grad_norm": 0.7111212249382538,
      "learning_rate": 1.530124847977657e-06,
      "loss": 0.3139911890029907,
      "num_input_tokens_seen": 1137499920,
      "step": 22105,
      "train_runtime": 136070.124,
      "train_tokens_per_second": 8359.66
    },
    {
      "epoch": 1.4886883921357392,
      "grad_norm": 0.6333256374591162,
      "learning_rate": 1.528221596550014e-06,
      "loss": 0.28489389419555666,
      "num_input_tokens_seen": 1137746896,
      "step": 22110,
      "train_runtime": 136097.3328,
      "train_tokens_per_second": 8359.803
    },
    {
      "epoch": 1.4890250471316995,
      "grad_norm": 0.7074782559068168,
      "learning_rate": 1.5263193159941331e-06,
      "loss": 0.2861085653305054,
      "num_input_tokens_seen": 1137996416,
      "step": 22115,
      "train_runtime": 136123.9873,
      "train_tokens_per_second": 8359.999
    },
    {
      "epoch": 1.4893617021276595,
      "grad_norm": 0.7531998566451017,
      "learning_rate": 1.5244180068419783e-06,
      "loss": 0.2770880699157715,
      "num_input_tokens_seen": 1138236808,
      "step": 22120,
      "train_runtime": 136152.3546,
      "train_tokens_per_second": 8360.023
    },
    {
      "epoch": 1.4896983571236198,
      "grad_norm": 0.7350795330945097,
      "learning_rate": 1.5225176696252453e-06,
      "loss": 0.32627182006835936,
      "num_input_tokens_seen": 1138465928,
      "step": 22125,
      "train_runtime": 136178.7091,
      "train_tokens_per_second": 8360.088
    },
    {
      "epoch": 1.4900350121195798,
      "grad_norm": 0.6582921999759891,
      "learning_rate": 1.5206183048753575e-06,
      "loss": 0.3021306276321411,
      "num_input_tokens_seen": 1138719264,
      "step": 22130,
      "train_runtime": 136208.5462,
      "train_tokens_per_second": 8360.116
    },
    {
      "epoch": 1.49037166711554,
      "grad_norm": 0.725103379104582,
      "learning_rate": 1.5187199131234653e-06,
      "loss": 0.3248953104019165,
      "num_input_tokens_seen": 1138995672,
      "step": 22135,
      "train_runtime": 136234.2412,
      "train_tokens_per_second": 8360.568
    },
    {
      "epoch": 1.4907083221115,
      "grad_norm": 0.7332931854167154,
      "learning_rate": 1.5168224949004484e-06,
      "loss": 0.30140366554260256,
      "num_input_tokens_seen": 1139248720,
      "step": 22140,
      "train_runtime": 136261.4156,
      "train_tokens_per_second": 8360.758
    },
    {
      "epoch": 1.4910449771074603,
      "grad_norm": 0.653426860026319,
      "learning_rate": 1.514926050736914e-06,
      "loss": 0.27651591300964357,
      "num_input_tokens_seen": 1139509608,
      "step": 22145,
      "train_runtime": 136296.7192,
      "train_tokens_per_second": 8360.507
    },
    {
      "epoch": 1.4913816321034203,
      "grad_norm": 0.7128516480137939,
      "learning_rate": 1.5130305811631945e-06,
      "loss": 0.3044422149658203,
      "num_input_tokens_seen": 1139775584,
      "step": 22150,
      "train_runtime": 136327.075,
      "train_tokens_per_second": 8360.596
    },
    {
      "epoch": 1.4917182870993806,
      "grad_norm": 0.6495679402500402,
      "learning_rate": 1.5111360867093571e-06,
      "loss": 0.3039050579071045,
      "num_input_tokens_seen": 1140038280,
      "step": 22155,
      "train_runtime": 136359.2479,
      "train_tokens_per_second": 8360.55
    },
    {
      "epoch": 1.4920549420953406,
      "grad_norm": 0.635847185566265,
      "learning_rate": 1.5092425679051848e-06,
      "loss": 0.2789382219314575,
      "num_input_tokens_seen": 1140296216,
      "step": 22160,
      "train_runtime": 136388.45,
      "train_tokens_per_second": 8360.651
    },
    {
      "epoch": 1.4923915970913009,
      "grad_norm": 0.5870563444418894,
      "learning_rate": 1.5073500252801988e-06,
      "loss": 0.3098041772842407,
      "num_input_tokens_seen": 1140568576,
      "step": 22165,
      "train_runtime": 136421.6201,
      "train_tokens_per_second": 8360.615
    },
    {
      "epoch": 1.492728252087261,
      "grad_norm": 0.6105165028439109,
      "learning_rate": 1.5054584593636412e-06,
      "loss": 0.2940333127975464,
      "num_input_tokens_seen": 1140853736,
      "step": 22170,
      "train_runtime": 136450.496,
      "train_tokens_per_second": 8360.935
    },
    {
      "epoch": 1.4930649070832211,
      "grad_norm": 0.7011436552727367,
      "learning_rate": 1.5035678706844814e-06,
      "loss": 0.31178696155548097,
      "num_input_tokens_seen": 1141104152,
      "step": 22175,
      "train_runtime": 136482.1355,
      "train_tokens_per_second": 8360.832
    },
    {
      "epoch": 1.4934015620791812,
      "grad_norm": 0.5938806461735502,
      "learning_rate": 1.501678259771418e-06,
      "loss": 0.3047629356384277,
      "num_input_tokens_seen": 1141352264,
      "step": 22180,
      "train_runtime": 136513.689,
      "train_tokens_per_second": 8360.717
    },
    {
      "epoch": 1.4937382170751414,
      "grad_norm": 0.6752512036262284,
      "learning_rate": 1.499789627152874e-06,
      "loss": 0.30664968490600586,
      "num_input_tokens_seen": 1141621288,
      "step": 22185,
      "train_runtime": 136547.1014,
      "train_tokens_per_second": 8360.641
    },
    {
      "epoch": 1.4940748720711015,
      "grad_norm": 0.6916291184618136,
      "learning_rate": 1.497901973356999e-06,
      "loss": 0.262829065322876,
      "num_input_tokens_seen": 1141894368,
      "step": 22190,
      "train_runtime": 136576.1323,
      "train_tokens_per_second": 8360.863
    },
    {
      "epoch": 1.4944115270670617,
      "grad_norm": 0.5789580808188955,
      "learning_rate": 1.4960152989116694e-06,
      "loss": 0.29398431777954104,
      "num_input_tokens_seen": 1142157304,
      "step": 22195,
      "train_runtime": 136611.2407,
      "train_tokens_per_second": 8360.639
    },
    {
      "epoch": 1.4947481820630217,
      "grad_norm": 0.6826180485448771,
      "learning_rate": 1.4941296043444869e-06,
      "loss": 0.314375638961792,
      "num_input_tokens_seen": 1142384704,
      "step": 22200,
      "train_runtime": 136643.9949,
      "train_tokens_per_second": 8360.299
    },
    {
      "epoch": 1.495084837058982,
      "grad_norm": 0.6007649716197117,
      "learning_rate": 1.4922448901827847e-06,
      "loss": 0.28628902435302733,
      "num_input_tokens_seen": 1142638776,
      "step": 22205,
      "train_runtime": 136674.9419,
      "train_tokens_per_second": 8360.265
    },
    {
      "epoch": 1.495421492054942,
      "grad_norm": 0.7936250483572991,
      "learning_rate": 1.4903611569536103e-06,
      "loss": 0.2737450122833252,
      "num_input_tokens_seen": 1142870216,
      "step": 22210,
      "train_runtime": 136703.2467,
      "train_tokens_per_second": 8360.227
    },
    {
      "epoch": 1.4957581470509023,
      "grad_norm": 0.6448916826688876,
      "learning_rate": 1.488478405183748e-06,
      "loss": 0.3046762466430664,
      "num_input_tokens_seen": 1143123808,
      "step": 22215,
      "train_runtime": 136735.9885,
      "train_tokens_per_second": 8360.08
    },
    {
      "epoch": 1.4960948020468623,
      "grad_norm": 0.7093695813606637,
      "learning_rate": 1.4865966353997024e-06,
      "loss": 0.3008828401565552,
      "num_input_tokens_seen": 1143404152,
      "step": 22220,
      "train_runtime": 136763.0035,
      "train_tokens_per_second": 8360.479
    },
    {
      "epoch": 1.4964314570428225,
      "grad_norm": 0.8281297189796107,
      "learning_rate": 1.4847158481277047e-06,
      "loss": 0.2860808610916138,
      "num_input_tokens_seen": 1143662864,
      "step": 22225,
      "train_runtime": 136793.131,
      "train_tokens_per_second": 8360.528
    },
    {
      "epoch": 1.4967681120387826,
      "grad_norm": 0.8095734916619246,
      "learning_rate": 1.48283604389371e-06,
      "loss": 0.28035707473754884,
      "num_input_tokens_seen": 1143891856,
      "step": 22230,
      "train_runtime": 136823.5444,
      "train_tokens_per_second": 8360.344
    },
    {
      "epoch": 1.4971047670347428,
      "grad_norm": 0.6389210935877347,
      "learning_rate": 1.480957223223401e-06,
      "loss": 0.2800656795501709,
      "num_input_tokens_seen": 1144155176,
      "step": 22235,
      "train_runtime": 136854.0644,
      "train_tokens_per_second": 8360.403
    },
    {
      "epoch": 1.497441422030703,
      "grad_norm": 0.6814080653727317,
      "learning_rate": 1.4790793866421826e-06,
      "loss": 0.3287003755569458,
      "num_input_tokens_seen": 1144404184,
      "step": 22240,
      "train_runtime": 136882.0188,
      "train_tokens_per_second": 8360.515
    },
    {
      "epoch": 1.497778077026663,
      "grad_norm": 0.7395344009387596,
      "learning_rate": 1.4772025346751862e-06,
      "loss": 0.30235815048217773,
      "num_input_tokens_seen": 1144669480,
      "step": 22245,
      "train_runtime": 136913.0469,
      "train_tokens_per_second": 8360.558
    },
    {
      "epoch": 1.4981147320226231,
      "grad_norm": 0.5791328526567672,
      "learning_rate": 1.475326667847266e-06,
      "loss": 0.3045390844345093,
      "num_input_tokens_seen": 1144931816,
      "step": 22250,
      "train_runtime": 136947.8464,
      "train_tokens_per_second": 8360.349
    },
    {
      "epoch": 1.4984513870185834,
      "grad_norm": 0.6557482059439009,
      "learning_rate": 1.4734517866830073e-06,
      "loss": 0.3077830791473389,
      "num_input_tokens_seen": 1145206744,
      "step": 22255,
      "train_runtime": 136978.2259,
      "train_tokens_per_second": 8360.502
    },
    {
      "epoch": 1.4987880420145436,
      "grad_norm": 0.6359778306296116,
      "learning_rate": 1.4715778917067081e-06,
      "loss": 0.3083902359008789,
      "num_input_tokens_seen": 1145439312,
      "step": 22260,
      "train_runtime": 137002.6665,
      "train_tokens_per_second": 8360.708
    },
    {
      "epoch": 1.4991246970105037,
      "grad_norm": 0.6011034708857709,
      "learning_rate": 1.4697049834424016e-06,
      "loss": 0.29170403480529783,
      "num_input_tokens_seen": 1145701648,
      "step": 22265,
      "train_runtime": 137033.3896,
      "train_tokens_per_second": 8360.748
    },
    {
      "epoch": 1.4994613520064637,
      "grad_norm": 0.7060946481986697,
      "learning_rate": 1.4678330624138387e-06,
      "loss": 0.28395633697509765,
      "num_input_tokens_seen": 1145951928,
      "step": 22270,
      "train_runtime": 137065.9321,
      "train_tokens_per_second": 8360.589
    },
    {
      "epoch": 1.499798007002424,
      "grad_norm": 0.6084311628493997,
      "learning_rate": 1.465962129144497e-06,
      "loss": 0.2933329105377197,
      "num_input_tokens_seen": 1146210752,
      "step": 22275,
      "train_runtime": 137097.8383,
      "train_tokens_per_second": 8360.531
    },
    {
      "epoch": 1.5001346619983842,
      "grad_norm": 0.6651078389996864,
      "learning_rate": 1.4640921841575766e-06,
      "loss": 0.314755916595459,
      "num_input_tokens_seen": 1146501016,
      "step": 22280,
      "train_runtime": 137126.6384,
      "train_tokens_per_second": 8360.892
    },
    {
      "epoch": 1.5004713169943442,
      "grad_norm": 0.615974027771379,
      "learning_rate": 1.4622232279760018e-06,
      "loss": 0.2965149641036987,
      "num_input_tokens_seen": 1146770576,
      "step": 22285,
      "train_runtime": 137156.1005,
      "train_tokens_per_second": 8361.061
    },
    {
      "epoch": 1.5008079719903042,
      "grad_norm": 0.7124037160640546,
      "learning_rate": 1.4603552611224202e-06,
      "loss": 0.3240455389022827,
      "num_input_tokens_seen": 1147001408,
      "step": 22290,
      "train_runtime": 137188.1344,
      "train_tokens_per_second": 8360.792
    },
    {
      "epoch": 1.5011446269862645,
      "grad_norm": 0.6410059344074623,
      "learning_rate": 1.4584882841192027e-06,
      "loss": 0.31077635288238525,
      "num_input_tokens_seen": 1147258520,
      "step": 22295,
      "train_runtime": 137220.1976,
      "train_tokens_per_second": 8360.712
    },
    {
      "epoch": 1.5014812819822247,
      "grad_norm": 0.622924163142642,
      "learning_rate": 1.456622297488442e-06,
      "loss": 0.28331871032714845,
      "num_input_tokens_seen": 1147501936,
      "step": 22300,
      "train_runtime": 137255.8712,
      "train_tokens_per_second": 8360.312
    },
    {
      "epoch": 1.5018179369781848,
      "grad_norm": 0.7691474065776647,
      "learning_rate": 1.4547573017519595e-06,
      "loss": 0.2967668533325195,
      "num_input_tokens_seen": 1147743616,
      "step": 22305,
      "train_runtime": 137282.923,
      "train_tokens_per_second": 8360.425
    },
    {
      "epoch": 1.5021545919741448,
      "grad_norm": 0.7024357910186878,
      "learning_rate": 1.4528932974312897e-06,
      "loss": 0.294305419921875,
      "num_input_tokens_seen": 1148009976,
      "step": 22310,
      "train_runtime": 137314.541,
      "train_tokens_per_second": 8360.44
    },
    {
      "epoch": 1.502491246970105,
      "grad_norm": 0.6866025699900984,
      "learning_rate": 1.4510302850477005e-06,
      "loss": 0.2981562614440918,
      "num_input_tokens_seen": 1148266664,
      "step": 22315,
      "train_runtime": 137343.251,
      "train_tokens_per_second": 8360.561
    },
    {
      "epoch": 1.5028279019660653,
      "grad_norm": 0.7232889464874734,
      "learning_rate": 1.4491682651221746e-06,
      "loss": 0.28380043506622316,
      "num_input_tokens_seen": 1148505776,
      "step": 22320,
      "train_runtime": 137369.2596,
      "train_tokens_per_second": 8360.719
    },
    {
      "epoch": 1.5031645569620253,
      "grad_norm": 0.6769213362288188,
      "learning_rate": 1.447307238175421e-06,
      "loss": 0.2925002336502075,
      "num_input_tokens_seen": 1148740496,
      "step": 22325,
      "train_runtime": 137403.0061,
      "train_tokens_per_second": 8360.374
    },
    {
      "epoch": 1.5035012119579854,
      "grad_norm": 0.6643046426245174,
      "learning_rate": 1.4454472047278688e-06,
      "loss": 0.2971970558166504,
      "num_input_tokens_seen": 1148984416,
      "step": 22330,
      "train_runtime": 137432.7028,
      "train_tokens_per_second": 8360.342
    },
    {
      "epoch": 1.5038378669539456,
      "grad_norm": 0.6622076211890006,
      "learning_rate": 1.443588165299672e-06,
      "loss": 0.27638468742370603,
      "num_input_tokens_seen": 1149251864,
      "step": 22335,
      "train_runtime": 137460.892,
      "train_tokens_per_second": 8360.573
    },
    {
      "epoch": 1.5041745219499059,
      "grad_norm": 0.7025423376864047,
      "learning_rate": 1.4417301204107037e-06,
      "loss": 0.30760965347290037,
      "num_input_tokens_seen": 1149518808,
      "step": 22340,
      "train_runtime": 137493.3243,
      "train_tokens_per_second": 8360.543
    },
    {
      "epoch": 1.504511176945866,
      "grad_norm": 0.6614139413846337,
      "learning_rate": 1.4398730705805607e-06,
      "loss": 0.2996813297271729,
      "num_input_tokens_seen": 1149781176,
      "step": 22345,
      "train_runtime": 137526.2204,
      "train_tokens_per_second": 8360.451
    },
    {
      "epoch": 1.504847831941826,
      "grad_norm": 0.573382456265421,
      "learning_rate": 1.43801701632856e-06,
      "loss": 0.30378174781799316,
      "num_input_tokens_seen": 1150053680,
      "step": 22350,
      "train_runtime": 137557.1939,
      "train_tokens_per_second": 8360.549
    },
    {
      "epoch": 1.5051844869377862,
      "grad_norm": 0.6301834184231963,
      "learning_rate": 1.436161958173745e-06,
      "loss": 0.32085657119750977,
      "num_input_tokens_seen": 1150320456,
      "step": 22355,
      "train_runtime": 137591.0898,
      "train_tokens_per_second": 8360.428
    },
    {
      "epoch": 1.5055211419337464,
      "grad_norm": 0.6953437534118194,
      "learning_rate": 1.4343078966348718e-06,
      "loss": 0.29875998497009276,
      "num_input_tokens_seen": 1150562536,
      "step": 22360,
      "train_runtime": 137621.9625,
      "train_tokens_per_second": 8360.312
    },
    {
      "epoch": 1.5058577969297064,
      "grad_norm": 0.6626805062546809,
      "learning_rate": 1.432454832230426e-06,
      "loss": 0.28462049961090086,
      "num_input_tokens_seen": 1150836568,
      "step": 22365,
      "train_runtime": 137653.7896,
      "train_tokens_per_second": 8360.37
    },
    {
      "epoch": 1.5061944519256665,
      "grad_norm": 0.5481692105725593,
      "learning_rate": 1.4306027654786103e-06,
      "loss": 0.27988805770874026,
      "num_input_tokens_seen": 1151103800,
      "step": 22370,
      "train_runtime": 137685.4853,
      "train_tokens_per_second": 8360.386
    },
    {
      "epoch": 1.5065311069216267,
      "grad_norm": 0.6488325991209066,
      "learning_rate": 1.4287516968973491e-06,
      "loss": 0.30617494583129884,
      "num_input_tokens_seen": 1151368800,
      "step": 22375,
      "train_runtime": 137717.4096,
      "train_tokens_per_second": 8360.372
    },
    {
      "epoch": 1.506867761917587,
      "grad_norm": 0.779603215355148,
      "learning_rate": 1.4269016270042884e-06,
      "loss": 0.3189589023590088,
      "num_input_tokens_seen": 1151618432,
      "step": 22380,
      "train_runtime": 137749.6141,
      "train_tokens_per_second": 8360.23
    },
    {
      "epoch": 1.507204416913547,
      "grad_norm": 0.849216422601139,
      "learning_rate": 1.425052556316794e-06,
      "loss": 0.2873330593109131,
      "num_input_tokens_seen": 1151873880,
      "step": 22385,
      "train_runtime": 137783.7723,
      "train_tokens_per_second": 8360.011
    },
    {
      "epoch": 1.507541071909507,
      "grad_norm": 0.7828295973766324,
      "learning_rate": 1.4232044853519522e-06,
      "loss": 0.32483787536621095,
      "num_input_tokens_seen": 1152111016,
      "step": 22390,
      "train_runtime": 137813.061,
      "train_tokens_per_second": 8359.955
    },
    {
      "epoch": 1.5078777269054673,
      "grad_norm": 0.5911290227784272,
      "learning_rate": 1.4213574146265708e-06,
      "loss": 0.2831369161605835,
      "num_input_tokens_seen": 1152357664,
      "step": 22395,
      "train_runtime": 137843.197,
      "train_tokens_per_second": 8359.917
    },
    {
      "epoch": 1.5082143819014275,
      "grad_norm": 0.6566278367083462,
      "learning_rate": 1.4195113446571752e-06,
      "loss": 0.32633228302001954,
      "num_input_tokens_seen": 1152629592,
      "step": 22400,
      "train_runtime": 137873.8263,
      "train_tokens_per_second": 8360.032
    },
    {
      "epoch": 1.5085510368973876,
      "grad_norm": 0.6248963935621498,
      "learning_rate": 1.417666275960018e-06,
      "loss": 0.2903961181640625,
      "num_input_tokens_seen": 1152915096,
      "step": 22405,
      "train_runtime": 137905.3818,
      "train_tokens_per_second": 8360.189
    },
    {
      "epoch": 1.5088876918933476,
      "grad_norm": 0.563650173950329,
      "learning_rate": 1.4158222090510605e-06,
      "loss": 0.2681227684020996,
      "num_input_tokens_seen": 1153179544,
      "step": 22410,
      "train_runtime": 137937.0708,
      "train_tokens_per_second": 8360.186
    },
    {
      "epoch": 1.5092243468893078,
      "grad_norm": 0.7252261117077876,
      "learning_rate": 1.4139791444459945e-06,
      "loss": 0.3145525693893433,
      "num_input_tokens_seen": 1153435144,
      "step": 22415,
      "train_runtime": 137970.6364,
      "train_tokens_per_second": 8360.005
    },
    {
      "epoch": 1.509561001885268,
      "grad_norm": 0.7129563287681342,
      "learning_rate": 1.4121370826602254e-06,
      "loss": 0.2963120937347412,
      "num_input_tokens_seen": 1153674736,
      "step": 22420,
      "train_runtime": 138000.5345,
      "train_tokens_per_second": 8359.929
    },
    {
      "epoch": 1.5098976568812281,
      "grad_norm": 0.6448615308467331,
      "learning_rate": 1.4102960242088803e-06,
      "loss": 0.30205645561218264,
      "num_input_tokens_seen": 1153910120,
      "step": 22425,
      "train_runtime": 138029.3858,
      "train_tokens_per_second": 8359.887
    },
    {
      "epoch": 1.5102343118771882,
      "grad_norm": 0.6366483868901622,
      "learning_rate": 1.4084559696068057e-06,
      "loss": 0.2837998867034912,
      "num_input_tokens_seen": 1154163136,
      "step": 22430,
      "train_runtime": 138059.056,
      "train_tokens_per_second": 8359.923
    },
    {
      "epoch": 1.5105709668731484,
      "grad_norm": 0.6610850639340312,
      "learning_rate": 1.406616919368567e-06,
      "loss": 0.30032052993774416,
      "num_input_tokens_seen": 1154407688,
      "step": 22435,
      "train_runtime": 138087.8738,
      "train_tokens_per_second": 8359.95
    },
    {
      "epoch": 1.5109076218691087,
      "grad_norm": 0.6406030054348781,
      "learning_rate": 1.4047788740084474e-06,
      "loss": 0.28491208553314207,
      "num_input_tokens_seen": 1154672208,
      "step": 22440,
      "train_runtime": 138119.3655,
      "train_tokens_per_second": 8359.959
    },
    {
      "epoch": 1.5112442768650687,
      "grad_norm": 0.5990151098258003,
      "learning_rate": 1.402941834040452e-06,
      "loss": 0.28939085006713866,
      "num_input_tokens_seen": 1154952400,
      "step": 22445,
      "train_runtime": 138152.8077,
      "train_tokens_per_second": 8359.963
    },
    {
      "epoch": 1.5115809318610287,
      "grad_norm": 0.6530502275473666,
      "learning_rate": 1.4011057999783006e-06,
      "loss": 0.28585023880004884,
      "num_input_tokens_seen": 1155194272,
      "step": 22450,
      "train_runtime": 138183.8978,
      "train_tokens_per_second": 8359.833
    },
    {
      "epoch": 1.511917586856989,
      "grad_norm": 0.6616276996440408,
      "learning_rate": 1.3992707723354398e-06,
      "loss": 0.3070800542831421,
      "num_input_tokens_seen": 1155455600,
      "step": 22455,
      "train_runtime": 138217.4521,
      "train_tokens_per_second": 8359.694
    },
    {
      "epoch": 1.5122542418529492,
      "grad_norm": 0.6523330023754762,
      "learning_rate": 1.397436751625022e-06,
      "loss": 0.2849736213684082,
      "num_input_tokens_seen": 1155689616,
      "step": 22460,
      "train_runtime": 138244.1178,
      "train_tokens_per_second": 8359.774
    },
    {
      "epoch": 1.5125908968489092,
      "grad_norm": 0.7570570139634323,
      "learning_rate": 1.3956037383599308e-06,
      "loss": 0.2897013187408447,
      "num_input_tokens_seen": 1155940624,
      "step": 22465,
      "train_runtime": 138275.1682,
      "train_tokens_per_second": 8359.712
    },
    {
      "epoch": 1.5129275518448693,
      "grad_norm": 0.7184082868546665,
      "learning_rate": 1.39377173305276e-06,
      "loss": 0.2764572620391846,
      "num_input_tokens_seen": 1156176328,
      "step": 22470,
      "train_runtime": 138306.3113,
      "train_tokens_per_second": 8359.534
    },
    {
      "epoch": 1.5132642068408295,
      "grad_norm": 0.6414566825179518,
      "learning_rate": 1.3919407362158254e-06,
      "loss": 0.3127893924713135,
      "num_input_tokens_seen": 1156465736,
      "step": 22475,
      "train_runtime": 138332.475,
      "train_tokens_per_second": 8360.045
    },
    {
      "epoch": 1.5136008618367898,
      "grad_norm": 0.749524241109927,
      "learning_rate": 1.3901107483611582e-06,
      "loss": 0.28145995140075686,
      "num_input_tokens_seen": 1156713992,
      "step": 22480,
      "train_runtime": 138359.4223,
      "train_tokens_per_second": 8360.211
    },
    {
      "epoch": 1.5139375168327498,
      "grad_norm": 0.6512494789683223,
      "learning_rate": 1.3882817700005091e-06,
      "loss": 0.296793794631958,
      "num_input_tokens_seen": 1156965968,
      "step": 22485,
      "train_runtime": 138391.0554,
      "train_tokens_per_second": 8360.121
    },
    {
      "epoch": 1.5142741718287098,
      "grad_norm": 0.7143875116024254,
      "learning_rate": 1.3864538016453466e-06,
      "loss": 0.3032499074935913,
      "num_input_tokens_seen": 1157188496,
      "step": 22490,
      "train_runtime": 138418.2691,
      "train_tokens_per_second": 8360.085
    },
    {
      "epoch": 1.51461082682467,
      "grad_norm": 0.7785330415877907,
      "learning_rate": 1.384626843806855e-06,
      "loss": 0.3033439159393311,
      "num_input_tokens_seen": 1157424896,
      "step": 22495,
      "train_runtime": 138448.7106,
      "train_tokens_per_second": 8359.954
    },
    {
      "epoch": 1.5149474818206303,
      "grad_norm": 0.6576054816970321,
      "learning_rate": 1.3828008969959367e-06,
      "loss": 0.3073155879974365,
      "num_input_tokens_seen": 1157701256,
      "step": 22500,
      "train_runtime": 138477.1591,
      "train_tokens_per_second": 8360.233
    },
    {
      "epoch": 1.5152841368165904,
      "grad_norm": 0.8000168117059241,
      "learning_rate": 1.3809759617232166e-06,
      "loss": 0.2859421491622925,
      "num_input_tokens_seen": 1157952968,
      "step": 22505,
      "train_runtime": 138508.6732,
      "train_tokens_per_second": 8360.148
    },
    {
      "epoch": 1.5156207918125504,
      "grad_norm": 0.6314647092145452,
      "learning_rate": 1.3791520384990247e-06,
      "loss": 0.28838419914245605,
      "num_input_tokens_seen": 1158206160,
      "step": 22510,
      "train_runtime": 138539.5992,
      "train_tokens_per_second": 8360.109
    },
    {
      "epoch": 1.5159574468085106,
      "grad_norm": 0.5893821783828139,
      "learning_rate": 1.3773291278334216e-06,
      "loss": 0.29623496532440186,
      "num_input_tokens_seen": 1158456376,
      "step": 22515,
      "train_runtime": 138572.662,
      "train_tokens_per_second": 8359.92
    },
    {
      "epoch": 1.5162941018044709,
      "grad_norm": 0.6391574614775891,
      "learning_rate": 1.3755072302361754e-06,
      "loss": 0.2954308032989502,
      "num_input_tokens_seen": 1158736480,
      "step": 22520,
      "train_runtime": 138603.3149,
      "train_tokens_per_second": 8360.092
    },
    {
      "epoch": 1.516630756800431,
      "grad_norm": 0.6917748795248135,
      "learning_rate": 1.3736863462167749e-06,
      "loss": 0.29259729385375977,
      "num_input_tokens_seen": 1158957928,
      "step": 22525,
      "train_runtime": 138635.7193,
      "train_tokens_per_second": 8359.735
    },
    {
      "epoch": 1.516967411796391,
      "grad_norm": 0.6546617245594852,
      "learning_rate": 1.3718664762844242e-06,
      "loss": 0.28532435894012453,
      "num_input_tokens_seen": 1159203296,
      "step": 22530,
      "train_runtime": 138662.5372,
      "train_tokens_per_second": 8359.888
    },
    {
      "epoch": 1.5173040667923512,
      "grad_norm": 0.6674103047036467,
      "learning_rate": 1.370047620948044e-06,
      "loss": 0.2980503082275391,
      "num_input_tokens_seen": 1159457104,
      "step": 22535,
      "train_runtime": 138691.0865,
      "train_tokens_per_second": 8359.997
    },
    {
      "epoch": 1.5176407217883114,
      "grad_norm": 0.6348873102594415,
      "learning_rate": 1.368229780716272e-06,
      "loss": 0.31056416034698486,
      "num_input_tokens_seen": 1159721112,
      "step": 22540,
      "train_runtime": 138724.1854,
      "train_tokens_per_second": 8359.906
    },
    {
      "epoch": 1.5179773767842715,
      "grad_norm": 0.7427102323047035,
      "learning_rate": 1.3664129560974604e-06,
      "loss": 0.29146828651428225,
      "num_input_tokens_seen": 1159980768,
      "step": 22545,
      "train_runtime": 138756.3546,
      "train_tokens_per_second": 8359.839
    },
    {
      "epoch": 1.5183140317802315,
      "grad_norm": 0.6857177573097472,
      "learning_rate": 1.3645971475996777e-06,
      "loss": 0.30243401527404784,
      "num_input_tokens_seen": 1160224952,
      "step": 22550,
      "train_runtime": 138788.6111,
      "train_tokens_per_second": 8359.655
    },
    {
      "epoch": 1.5186506867761917,
      "grad_norm": 0.8136535161289397,
      "learning_rate": 1.3627823557307135e-06,
      "loss": 0.3012071132659912,
      "num_input_tokens_seen": 1160508112,
      "step": 22555,
      "train_runtime": 138820.8343,
      "train_tokens_per_second": 8359.755
    },
    {
      "epoch": 1.518987341772152,
      "grad_norm": 0.6582558202699675,
      "learning_rate": 1.3609685809980616e-06,
      "loss": 0.2990730285644531,
      "num_input_tokens_seen": 1160773520,
      "step": 22560,
      "train_runtime": 138857.1392,
      "train_tokens_per_second": 8359.48
    },
    {
      "epoch": 1.519323996768112,
      "grad_norm": 0.6331840250919515,
      "learning_rate": 1.3591558239089443e-06,
      "loss": 0.27659432888031005,
      "num_input_tokens_seen": 1161027096,
      "step": 22565,
      "train_runtime": 138889.6008,
      "train_tokens_per_second": 8359.352
    },
    {
      "epoch": 1.519660651764072,
      "grad_norm": 0.7163908551654113,
      "learning_rate": 1.3573440849702902e-06,
      "loss": 0.31561081409454345,
      "num_input_tokens_seen": 1161291480,
      "step": 22570,
      "train_runtime": 138918.4908,
      "train_tokens_per_second": 8359.517
    },
    {
      "epoch": 1.5199973067600323,
      "grad_norm": 0.6573742143765177,
      "learning_rate": 1.3555333646887476e-06,
      "loss": 0.31214122772216796,
      "num_input_tokens_seen": 1161565600,
      "step": 22575,
      "train_runtime": 138954.8662,
      "train_tokens_per_second": 8359.301
    },
    {
      "epoch": 1.5203339617559926,
      "grad_norm": 0.6269101792891109,
      "learning_rate": 1.3537236635706779e-06,
      "loss": 0.2915287971496582,
      "num_input_tokens_seen": 1161839032,
      "step": 22580,
      "train_runtime": 138983.9367,
      "train_tokens_per_second": 8359.52
    },
    {
      "epoch": 1.5206706167519526,
      "grad_norm": 0.7302903167747169,
      "learning_rate": 1.351914982122159e-06,
      "loss": 0.2796888828277588,
      "num_input_tokens_seen": 1162107184,
      "step": 22585,
      "train_runtime": 139013.3945,
      "train_tokens_per_second": 8359.678
    },
    {
      "epoch": 1.5210072717479126,
      "grad_norm": 0.6731163613008256,
      "learning_rate": 1.3501073208489823e-06,
      "loss": 0.30295047760009763,
      "num_input_tokens_seen": 1162372616,
      "step": 22590,
      "train_runtime": 139045.7593,
      "train_tokens_per_second": 8359.641
    },
    {
      "epoch": 1.5213439267438729,
      "grad_norm": 0.6478627897094131,
      "learning_rate": 1.3483006802566546e-06,
      "loss": 0.2931240081787109,
      "num_input_tokens_seen": 1162638336,
      "step": 22595,
      "train_runtime": 139076.2089,
      "train_tokens_per_second": 8359.721
    },
    {
      "epoch": 1.5216805817398331,
      "grad_norm": 0.7001865423394553,
      "learning_rate": 1.3464950608503957e-06,
      "loss": 0.3085663318634033,
      "num_input_tokens_seen": 1162885880,
      "step": 22600,
      "train_runtime": 139108.0022,
      "train_tokens_per_second": 8359.59
    },
    {
      "epoch": 1.5220172367357931,
      "grad_norm": 0.7527750213572646,
      "learning_rate": 1.344690463135146e-06,
      "loss": 0.3028801441192627,
      "num_input_tokens_seen": 1163131000,
      "step": 22605,
      "train_runtime": 139137.5234,
      "train_tokens_per_second": 8359.578
    },
    {
      "epoch": 1.5223538917317532,
      "grad_norm": 0.6296911590075803,
      "learning_rate": 1.3428868876155493e-06,
      "loss": 0.32263808250427245,
      "num_input_tokens_seen": 1163372168,
      "step": 22610,
      "train_runtime": 139165.0936,
      "train_tokens_per_second": 8359.655
    },
    {
      "epoch": 1.5226905467277134,
      "grad_norm": 0.5872162020294657,
      "learning_rate": 1.3410843347959745e-06,
      "loss": 0.28458514213562014,
      "num_input_tokens_seen": 1163622136,
      "step": 22615,
      "train_runtime": 139193.6581,
      "train_tokens_per_second": 8359.735
    },
    {
      "epoch": 1.5230272017236737,
      "grad_norm": 0.6944840467931233,
      "learning_rate": 1.3392828051804979e-06,
      "loss": 0.29966304302215574,
      "num_input_tokens_seen": 1163886080,
      "step": 22620,
      "train_runtime": 139226.798,
      "train_tokens_per_second": 8359.641
    },
    {
      "epoch": 1.5233638567196337,
      "grad_norm": 0.6519517470349513,
      "learning_rate": 1.3374822992729114e-06,
      "loss": 0.3090278148651123,
      "num_input_tokens_seen": 1164138968,
      "step": 22625,
      "train_runtime": 139259.5612,
      "train_tokens_per_second": 8359.49
    },
    {
      "epoch": 1.5237005117155937,
      "grad_norm": 0.7383776345718224,
      "learning_rate": 1.3356828175767212e-06,
      "loss": 0.3016170263290405,
      "num_input_tokens_seen": 1164379704,
      "step": 22630,
      "train_runtime": 139286.6203,
      "train_tokens_per_second": 8359.595
    },
    {
      "epoch": 1.524037166711554,
      "grad_norm": 0.6571899118913771,
      "learning_rate": 1.3338843605951462e-06,
      "loss": 0.27048587799072266,
      "num_input_tokens_seen": 1164650304,
      "step": 22635,
      "train_runtime": 139318.1571,
      "train_tokens_per_second": 8359.645
    },
    {
      "epoch": 1.5243738217075142,
      "grad_norm": 0.6694753754284136,
      "learning_rate": 1.33208692883112e-06,
      "loss": 0.2947390079498291,
      "num_input_tokens_seen": 1164927232,
      "step": 22640,
      "train_runtime": 139351.0106,
      "train_tokens_per_second": 8359.661
    },
    {
      "epoch": 1.5247104767034743,
      "grad_norm": 0.5617331953245761,
      "learning_rate": 1.3302905227872876e-06,
      "loss": 0.28046383857727053,
      "num_input_tokens_seen": 1165219344,
      "step": 22645,
      "train_runtime": 139383.9078,
      "train_tokens_per_second": 8359.784
    },
    {
      "epoch": 1.5250471316994343,
      "grad_norm": 1.0260083267960638,
      "learning_rate": 1.328495142966007e-06,
      "loss": 0.3081449747085571,
      "num_input_tokens_seen": 1165462624,
      "step": 22650,
      "train_runtime": 139412.6343,
      "train_tokens_per_second": 8359.806
    },
    {
      "epoch": 1.5253837866953945,
      "grad_norm": 0.6472188427913277,
      "learning_rate": 1.3267007898693552e-06,
      "loss": 0.2968280792236328,
      "num_input_tokens_seen": 1165726464,
      "step": 22655,
      "train_runtime": 139445.8731,
      "train_tokens_per_second": 8359.706
    },
    {
      "epoch": 1.5257204416913548,
      "grad_norm": 0.6533912121704911,
      "learning_rate": 1.324907463999111e-06,
      "loss": 0.2974425792694092,
      "num_input_tokens_seen": 1165996152,
      "step": 22660,
      "train_runtime": 139476.3705,
      "train_tokens_per_second": 8359.811
    },
    {
      "epoch": 1.5260570966873148,
      "grad_norm": 0.6486096244215718,
      "learning_rate": 1.3231151658567776e-06,
      "loss": 0.29709796905517577,
      "num_input_tokens_seen": 1166242248,
      "step": 22665,
      "train_runtime": 139507.6408,
      "train_tokens_per_second": 8359.702
    },
    {
      "epoch": 1.5263937516832748,
      "grad_norm": 0.60866953911307,
      "learning_rate": 1.321323895943563e-06,
      "loss": 0.29802823066711426,
      "num_input_tokens_seen": 1166493488,
      "step": 22670,
      "train_runtime": 139537.1267,
      "train_tokens_per_second": 8359.736
    },
    {
      "epoch": 1.526730406679235,
      "grad_norm": 0.6131081445252852,
      "learning_rate": 1.3195336547603905e-06,
      "loss": 0.30552306175231936,
      "num_input_tokens_seen": 1166745936,
      "step": 22675,
      "train_runtime": 139567.3066,
      "train_tokens_per_second": 8359.737
    },
    {
      "epoch": 1.5270670616751953,
      "grad_norm": 0.6353516545798206,
      "learning_rate": 1.3177444428078951e-06,
      "loss": 0.2834911823272705,
      "num_input_tokens_seen": 1166988472,
      "step": 22680,
      "train_runtime": 139598.1405,
      "train_tokens_per_second": 8359.628
    },
    {
      "epoch": 1.5274037166711554,
      "grad_norm": 0.7245344590229508,
      "learning_rate": 1.3159562605864245e-06,
      "loss": 0.2919331073760986,
      "num_input_tokens_seen": 1167253920,
      "step": 22685,
      "train_runtime": 139628.5768,
      "train_tokens_per_second": 8359.706
    },
    {
      "epoch": 1.5277403716671154,
      "grad_norm": 0.7092051868605261,
      "learning_rate": 1.3141691085960385e-06,
      "loss": 0.2977416515350342,
      "num_input_tokens_seen": 1167540816,
      "step": 22690,
      "train_runtime": 139659.0578,
      "train_tokens_per_second": 8359.936
    },
    {
      "epoch": 1.5280770266630757,
      "grad_norm": 0.6716474438869504,
      "learning_rate": 1.3123829873365073e-06,
      "loss": 0.2779765844345093,
      "num_input_tokens_seen": 1167779112,
      "step": 22695,
      "train_runtime": 139689.8513,
      "train_tokens_per_second": 8359.799
    },
    {
      "epoch": 1.528413681659036,
      "grad_norm": 0.7463071889405585,
      "learning_rate": 1.3105978973073136e-06,
      "loss": 0.2825263023376465,
      "num_input_tokens_seen": 1168017904,
      "step": 22700,
      "train_runtime": 139718.89,
      "train_tokens_per_second": 8359.771
    },
    {
      "epoch": 1.528750336654996,
      "grad_norm": 0.6743765634741533,
      "learning_rate": 1.3088138390076565e-06,
      "loss": 0.3133948802947998,
      "num_input_tokens_seen": 1168290616,
      "step": 22705,
      "train_runtime": 139750.517,
      "train_tokens_per_second": 8359.83
    },
    {
      "epoch": 1.529086991650956,
      "grad_norm": 0.6162167262455808,
      "learning_rate": 1.3070308129364357e-06,
      "loss": 0.2877573251724243,
      "num_input_tokens_seen": 1168534024,
      "step": 22710,
      "train_runtime": 139779.8616,
      "train_tokens_per_second": 8359.817
    },
    {
      "epoch": 1.5294236466469162,
      "grad_norm": 0.7655407025617664,
      "learning_rate": 1.3052488195922736e-06,
      "loss": 0.282314395904541,
      "num_input_tokens_seen": 1168765944,
      "step": 22715,
      "train_runtime": 139809.3151,
      "train_tokens_per_second": 8359.714
    },
    {
      "epoch": 1.5297603016428765,
      "grad_norm": 0.6469061163207948,
      "learning_rate": 1.3034678594734972e-06,
      "loss": 0.2827145099639893,
      "num_input_tokens_seen": 1169023912,
      "step": 22720,
      "train_runtime": 139838.9376,
      "train_tokens_per_second": 8359.788
    },
    {
      "epoch": 1.5300969566388365,
      "grad_norm": 0.8451589936758108,
      "learning_rate": 1.3016879330781468e-06,
      "loss": 0.30217053890228274,
      "num_input_tokens_seen": 1169255624,
      "step": 22725,
      "train_runtime": 139868.0082,
      "train_tokens_per_second": 8359.707
    },
    {
      "epoch": 1.5304336116347965,
      "grad_norm": 0.6152732625861611,
      "learning_rate": 1.2999090409039728e-06,
      "loss": 0.29002938270568845,
      "num_input_tokens_seen": 1169517096,
      "step": 22730,
      "train_runtime": 139895.276,
      "train_tokens_per_second": 8359.947
    },
    {
      "epoch": 1.5307702666307568,
      "grad_norm": 0.7058685696214855,
      "learning_rate": 1.2981311834484367e-06,
      "loss": 0.30576837062835693,
      "num_input_tokens_seen": 1169767168,
      "step": 22735,
      "train_runtime": 139929.7516,
      "train_tokens_per_second": 8359.674
    },
    {
      "epoch": 1.531106921626717,
      "grad_norm": 0.7277214769197768,
      "learning_rate": 1.296354361208711e-06,
      "loss": 0.28068568706512453,
      "num_input_tokens_seen": 1170010272,
      "step": 22740,
      "train_runtime": 139960.8103,
      "train_tokens_per_second": 8359.556
    },
    {
      "epoch": 1.531443576622677,
      "grad_norm": 0.7035727091869138,
      "learning_rate": 1.294578574681678e-06,
      "loss": 0.3145493507385254,
      "num_input_tokens_seen": 1170248056,
      "step": 22745,
      "train_runtime": 139988.971,
      "train_tokens_per_second": 8359.573
    },
    {
      "epoch": 1.531780231618637,
      "grad_norm": 0.6371961519014206,
      "learning_rate": 1.292803824363929e-06,
      "loss": 0.28414483070373536,
      "num_input_tokens_seen": 1170514616,
      "step": 22750,
      "train_runtime": 140019.3369,
      "train_tokens_per_second": 8359.664
    },
    {
      "epoch": 1.5321168866145973,
      "grad_norm": 0.6633069104564601,
      "learning_rate": 1.2910301107517726e-06,
      "loss": 0.3204554319381714,
      "num_input_tokens_seen": 1170780416,
      "step": 22755,
      "train_runtime": 140052.9513,
      "train_tokens_per_second": 8359.555
    },
    {
      "epoch": 1.5324535416105576,
      "grad_norm": 0.7160150818866493,
      "learning_rate": 1.2892574343412156e-06,
      "loss": 0.3198665142059326,
      "num_input_tokens_seen": 1171050528,
      "step": 22760,
      "train_runtime": 140083.7824,
      "train_tokens_per_second": 8359.644
    },
    {
      "epoch": 1.5327901966065176,
      "grad_norm": 0.7040910277948873,
      "learning_rate": 1.2874857956279858e-06,
      "loss": 0.29843361377716066,
      "num_input_tokens_seen": 1171288616,
      "step": 22765,
      "train_runtime": 140114.5229,
      "train_tokens_per_second": 8359.509
    },
    {
      "epoch": 1.5331268516024776,
      "grad_norm": 0.7178361842954455,
      "learning_rate": 1.2857151951075147e-06,
      "loss": 0.3059422969818115,
      "num_input_tokens_seen": 1171556632,
      "step": 22770,
      "train_runtime": 140147.3656,
      "train_tokens_per_second": 8359.462
    },
    {
      "epoch": 1.5334635065984379,
      "grad_norm": 0.6011042336110684,
      "learning_rate": 1.2839456332749456e-06,
      "loss": 0.3003576993942261,
      "num_input_tokens_seen": 1171808888,
      "step": 22775,
      "train_runtime": 140175.2388,
      "train_tokens_per_second": 8359.6
    },
    {
      "epoch": 1.5338001615943981,
      "grad_norm": 0.8265468951138524,
      "learning_rate": 1.2821771106251308e-06,
      "loss": 0.2962388038635254,
      "num_input_tokens_seen": 1172063216,
      "step": 22780,
      "train_runtime": 140203.2801,
      "train_tokens_per_second": 8359.742
    },
    {
      "epoch": 1.5341368165903582,
      "grad_norm": 0.6084251950415271,
      "learning_rate": 1.2804096276526312e-06,
      "loss": 0.28495206832885744,
      "num_input_tokens_seen": 1172334832,
      "step": 22785,
      "train_runtime": 140232.6479,
      "train_tokens_per_second": 8359.928
    },
    {
      "epoch": 1.5344734715863182,
      "grad_norm": 0.7750202485567135,
      "learning_rate": 1.2786431848517184e-06,
      "loss": 0.28902473449707033,
      "num_input_tokens_seen": 1172563640,
      "step": 22790,
      "train_runtime": 140263.1559,
      "train_tokens_per_second": 8359.741
    },
    {
      "epoch": 1.5348101265822784,
      "grad_norm": 0.6767864644884974,
      "learning_rate": 1.2768777827163725e-06,
      "loss": 0.29035372734069825,
      "num_input_tokens_seen": 1172807024,
      "step": 22795,
      "train_runtime": 140291.4022,
      "train_tokens_per_second": 8359.793
    },
    {
      "epoch": 1.5351467815782387,
      "grad_norm": 0.6474523077128271,
      "learning_rate": 1.2751134217402811e-06,
      "loss": 0.2924686908721924,
      "num_input_tokens_seen": 1173043032,
      "step": 22800,
      "train_runtime": 140317.5316,
      "train_tokens_per_second": 8359.918
    },
    {
      "epoch": 1.5354834365741987,
      "grad_norm": 0.5686240435833865,
      "learning_rate": 1.2733501024168465e-06,
      "loss": 0.2848621368408203,
      "num_input_tokens_seen": 1173308760,
      "step": 22805,
      "train_runtime": 140348.5332,
      "train_tokens_per_second": 8359.965
    },
    {
      "epoch": 1.5358200915701588,
      "grad_norm": 0.7037932384029524,
      "learning_rate": 1.2715878252391695e-06,
      "loss": 0.29651594161987305,
      "num_input_tokens_seen": 1173544424,
      "step": 22810,
      "train_runtime": 140377.7275,
      "train_tokens_per_second": 8359.905
    },
    {
      "epoch": 1.536156746566119,
      "grad_norm": 0.6963338681139166,
      "learning_rate": 1.269826590700069e-06,
      "loss": 0.30859320163726806,
      "num_input_tokens_seen": 1173807568,
      "step": 22815,
      "train_runtime": 140412.2431,
      "train_tokens_per_second": 8359.724
    },
    {
      "epoch": 1.5364934015620793,
      "grad_norm": 0.5959535858187579,
      "learning_rate": 1.2680663992920684e-06,
      "loss": 0.28382391929626466,
      "num_input_tokens_seen": 1174075064,
      "step": 22820,
      "train_runtime": 140442.3837,
      "train_tokens_per_second": 8359.834
    },
    {
      "epoch": 1.5368300565580393,
      "grad_norm": 0.7628208322889882,
      "learning_rate": 1.2663072515073988e-06,
      "loss": 0.3021712303161621,
      "num_input_tokens_seen": 1174331488,
      "step": 22825,
      "train_runtime": 140471.7147,
      "train_tokens_per_second": 8359.914
    },
    {
      "epoch": 1.5371667115539993,
      "grad_norm": 0.7138075702525751,
      "learning_rate": 1.2645491478380002e-06,
      "loss": 0.30213727951049807,
      "num_input_tokens_seen": 1174582056,
      "step": 22830,
      "train_runtime": 140500.6421,
      "train_tokens_per_second": 8359.976
    },
    {
      "epoch": 1.5375033665499596,
      "grad_norm": 0.6980411719601997,
      "learning_rate": 1.2627920887755212e-06,
      "loss": 0.3025203704833984,
      "num_input_tokens_seen": 1174831400,
      "step": 22835,
      "train_runtime": 140531.5966,
      "train_tokens_per_second": 8359.909
    },
    {
      "epoch": 1.5378400215459198,
      "grad_norm": 0.6350360402756156,
      "learning_rate": 1.261036074811317e-06,
      "loss": 0.30528059005737307,
      "num_input_tokens_seen": 1175097712,
      "step": 22840,
      "train_runtime": 140565.6296,
      "train_tokens_per_second": 8359.78
    },
    {
      "epoch": 1.5381766765418798,
      "grad_norm": 0.6891928461482402,
      "learning_rate": 1.2592811064364524e-06,
      "loss": 0.29288086891174314,
      "num_input_tokens_seen": 1175358712,
      "step": 22845,
      "train_runtime": 140593.6492,
      "train_tokens_per_second": 8359.97
    },
    {
      "epoch": 1.5385133315378399,
      "grad_norm": 0.9440766994596648,
      "learning_rate": 1.2575271841416963e-06,
      "loss": 0.3165269374847412,
      "num_input_tokens_seen": 1175572808,
      "step": 22850,
      "train_runtime": 140624.0773,
      "train_tokens_per_second": 8359.684
    },
    {
      "epoch": 1.5388499865338001,
      "grad_norm": 0.6416561876733814,
      "learning_rate": 1.2557743084175327e-06,
      "loss": 0.27585878372192385,
      "num_input_tokens_seen": 1175827152,
      "step": 22855,
      "train_runtime": 140653.7974,
      "train_tokens_per_second": 8359.726
    },
    {
      "epoch": 1.5391866415297604,
      "grad_norm": 0.631883192724963,
      "learning_rate": 1.254022479754141e-06,
      "loss": 0.28116726875305176,
      "num_input_tokens_seen": 1176096760,
      "step": 22860,
      "train_runtime": 140686.9677,
      "train_tokens_per_second": 8359.671
    },
    {
      "epoch": 1.5395232965257204,
      "grad_norm": 0.642689212824122,
      "learning_rate": 1.252271698641419e-06,
      "loss": 0.2886949062347412,
      "num_input_tokens_seen": 1176350336,
      "step": 22865,
      "train_runtime": 140713.8871,
      "train_tokens_per_second": 8359.874
    },
    {
      "epoch": 1.5398599515216804,
      "grad_norm": 0.614487096485123,
      "learning_rate": 1.2505219655689659e-06,
      "loss": 0.29910759925842284,
      "num_input_tokens_seen": 1176632696,
      "step": 22870,
      "train_runtime": 140750.6502,
      "train_tokens_per_second": 8359.696
    },
    {
      "epoch": 1.5401966065176407,
      "grad_norm": 0.6500667493224104,
      "learning_rate": 1.2487732810260877e-06,
      "loss": 0.30614299774169923,
      "num_input_tokens_seen": 1176892384,
      "step": 22875,
      "train_runtime": 140779.2882,
      "train_tokens_per_second": 8359.84
    },
    {
      "epoch": 1.540533261513601,
      "grad_norm": 0.6826890780459333,
      "learning_rate": 1.2470256455018027e-06,
      "loss": 0.2769783020019531,
      "num_input_tokens_seen": 1177145656,
      "step": 22880,
      "train_runtime": 140808.7665,
      "train_tokens_per_second": 8359.889
    },
    {
      "epoch": 1.540869916509561,
      "grad_norm": 1.0632452252391085,
      "learning_rate": 1.2452790594848257e-06,
      "loss": 0.28171873092651367,
      "num_input_tokens_seen": 1177376968,
      "step": 22885,
      "train_runtime": 140836.9628,
      "train_tokens_per_second": 8359.858
    },
    {
      "epoch": 1.541206571505521,
      "grad_norm": 0.713822919473914,
      "learning_rate": 1.2435335234635892e-06,
      "loss": 0.32900609970092776,
      "num_input_tokens_seen": 1177639160,
      "step": 22890,
      "train_runtime": 140867.536,
      "train_tokens_per_second": 8359.905
    },
    {
      "epoch": 1.5415432265014812,
      "grad_norm": 0.7615907830964759,
      "learning_rate": 1.2417890379262221e-06,
      "loss": 0.3054653167724609,
      "num_input_tokens_seen": 1177885552,
      "step": 22895,
      "train_runtime": 140897.7699,
      "train_tokens_per_second": 8359.859
    },
    {
      "epoch": 1.5418798814974415,
      "grad_norm": 0.7673941413012852,
      "learning_rate": 1.2400456033605672e-06,
      "loss": 0.2679884433746338,
      "num_input_tokens_seen": 1178136696,
      "step": 22900,
      "train_runtime": 140927.5745,
      "train_tokens_per_second": 8359.874
    },
    {
      "epoch": 1.5422165364934015,
      "grad_norm": 0.6069540531302895,
      "learning_rate": 1.2383032202541706e-06,
      "loss": 0.30822124481201174,
      "num_input_tokens_seen": 1178400352,
      "step": 22905,
      "train_runtime": 140957.6736,
      "train_tokens_per_second": 8359.959
    },
    {
      "epoch": 1.5425531914893615,
      "grad_norm": 0.7035946953706492,
      "learning_rate": 1.2365618890942827e-06,
      "loss": 0.32069308757781984,
      "num_input_tokens_seen": 1178665616,
      "step": 22910,
      "train_runtime": 140989.1186,
      "train_tokens_per_second": 8359.976
    },
    {
      "epoch": 1.5428898464853218,
      "grad_norm": 0.5935358891461194,
      "learning_rate": 1.2348216103678618e-06,
      "loss": 0.3144097805023193,
      "num_input_tokens_seen": 1178928880,
      "step": 22915,
      "train_runtime": 141020.5706,
      "train_tokens_per_second": 8359.978
    },
    {
      "epoch": 1.543226501481282,
      "grad_norm": 0.7725470431995894,
      "learning_rate": 1.2330823845615713e-06,
      "loss": 0.28970847129821775,
      "num_input_tokens_seen": 1179156496,
      "step": 22920,
      "train_runtime": 141054.6862,
      "train_tokens_per_second": 8359.57
    },
    {
      "epoch": 1.543563156477242,
      "grad_norm": 0.6633043111003611,
      "learning_rate": 1.2313442121617785e-06,
      "loss": 0.296297025680542,
      "num_input_tokens_seen": 1179443656,
      "step": 22925,
      "train_runtime": 141081.8859,
      "train_tokens_per_second": 8359.994
    },
    {
      "epoch": 1.543899811473202,
      "grad_norm": 0.6458000275349427,
      "learning_rate": 1.229607093654563e-06,
      "loss": 0.2769587516784668,
      "num_input_tokens_seen": 1179726440,
      "step": 22930,
      "train_runtime": 141113.9667,
      "train_tokens_per_second": 8360.097
    },
    {
      "epoch": 1.5442364664691623,
      "grad_norm": 0.5744967911482114,
      "learning_rate": 1.2278710295256974e-06,
      "loss": 0.30580930709838866,
      "num_input_tokens_seen": 1180003080,
      "step": 22935,
      "train_runtime": 141148.3571,
      "train_tokens_per_second": 8360.02
    },
    {
      "epoch": 1.5445731214651226,
      "grad_norm": 0.6541469867517778,
      "learning_rate": 1.2261360202606725e-06,
      "loss": 0.3028374910354614,
      "num_input_tokens_seen": 1180278832,
      "step": 22940,
      "train_runtime": 141178.348,
      "train_tokens_per_second": 8360.197
    },
    {
      "epoch": 1.5449097764610826,
      "grad_norm": 0.6927627823919419,
      "learning_rate": 1.2244020663446731e-06,
      "loss": 0.3159763813018799,
      "num_input_tokens_seen": 1180515112,
      "step": 22945,
      "train_runtime": 141209.7818,
      "train_tokens_per_second": 8360.009
    },
    {
      "epoch": 1.5452464314570427,
      "grad_norm": 0.7033606546343739,
      "learning_rate": 1.222669168262598e-06,
      "loss": 0.29160988330841064,
      "num_input_tokens_seen": 1180779360,
      "step": 22950,
      "train_runtime": 141240.3163,
      "train_tokens_per_second": 8360.073
    },
    {
      "epoch": 1.545583086453003,
      "grad_norm": 0.5856637018193247,
      "learning_rate": 1.2209373264990448e-06,
      "loss": 0.2916308879852295,
      "num_input_tokens_seen": 1181045920,
      "step": 22955,
      "train_runtime": 141274.326,
      "train_tokens_per_second": 8359.947
    },
    {
      "epoch": 1.5459197414489632,
      "grad_norm": 0.7230919009592404,
      "learning_rate": 1.2192065415383176e-06,
      "loss": 0.30676984786987305,
      "num_input_tokens_seen": 1181297496,
      "step": 22960,
      "train_runtime": 141308.403,
      "train_tokens_per_second": 8359.712
    },
    {
      "epoch": 1.5462563964449232,
      "grad_norm": 0.6868194753798107,
      "learning_rate": 1.217476813864425e-06,
      "loss": 0.3111604928970337,
      "num_input_tokens_seen": 1181575432,
      "step": 22965,
      "train_runtime": 141338.0363,
      "train_tokens_per_second": 8359.925
    },
    {
      "epoch": 1.5465930514408834,
      "grad_norm": 0.6368084220125848,
      "learning_rate": 1.2157481439610801e-06,
      "loss": 0.2771847486495972,
      "num_input_tokens_seen": 1181836248,
      "step": 22970,
      "train_runtime": 141370.015,
      "train_tokens_per_second": 8359.879
    },
    {
      "epoch": 1.5469297064368437,
      "grad_norm": 0.6637656792339086,
      "learning_rate": 1.2140205323116976e-06,
      "loss": 0.2878734588623047,
      "num_input_tokens_seen": 1182071432,
      "step": 22975,
      "train_runtime": 141398.3046,
      "train_tokens_per_second": 8359.87
    },
    {
      "epoch": 1.5472663614328037,
      "grad_norm": 0.5214145241233855,
      "learning_rate": 1.2122939793994043e-06,
      "loss": 0.28056299686431885,
      "num_input_tokens_seen": 1182326088,
      "step": 22980,
      "train_runtime": 141427.7008,
      "train_tokens_per_second": 8359.933
    },
    {
      "epoch": 1.5476030164287637,
      "grad_norm": 0.6708724042579567,
      "learning_rate": 1.2105684857070182e-06,
      "loss": 0.28717889785766604,
      "num_input_tokens_seen": 1182586744,
      "step": 22985,
      "train_runtime": 141459.7391,
      "train_tokens_per_second": 8359.882
    },
    {
      "epoch": 1.547939671424724,
      "grad_norm": 0.6065240741300233,
      "learning_rate": 1.2088440517170729e-06,
      "loss": 0.3035863399505615,
      "num_input_tokens_seen": 1182863704,
      "step": 22990,
      "train_runtime": 141488.7381,
      "train_tokens_per_second": 8360.126
    },
    {
      "epoch": 1.5482763264206842,
      "grad_norm": 0.7908902402068876,
      "learning_rate": 1.2071206779117994e-06,
      "loss": 0.3017385482788086,
      "num_input_tokens_seen": 1183130216,
      "step": 22995,
      "train_runtime": 141519.498,
      "train_tokens_per_second": 8360.192
    },
    {
      "epoch": 1.5486129814166443,
      "grad_norm": 0.5481942857233729,
      "learning_rate": 1.2053983647731337e-06,
      "loss": 0.29206008911132814,
      "num_input_tokens_seen": 1183423520,
      "step": 23000,
      "train_runtime": 141551.6493,
      "train_tokens_per_second": 8360.365
    },
    {
      "epoch": 1.5489496364126043,
      "grad_norm": 0.7431647845598125,
      "learning_rate": 1.2036771127827152e-06,
      "loss": 0.2966840982437134,
      "num_input_tokens_seen": 1183665296,
      "step": 23005,
      "train_runtime": 141578.9908,
      "train_tokens_per_second": 8360.459
    },
    {
      "epoch": 1.5492862914085646,
      "grad_norm": 0.6577511610512886,
      "learning_rate": 1.2019569224218869e-06,
      "loss": 0.30553600788116453,
      "num_input_tokens_seen": 1183952624,
      "step": 23010,
      "train_runtime": 141613.6644,
      "train_tokens_per_second": 8360.441
    },
    {
      "epoch": 1.5496229464045248,
      "grad_norm": 0.6400915337622956,
      "learning_rate": 1.2002377941716936e-06,
      "loss": 0.27906370162963867,
      "num_input_tokens_seen": 1184205048,
      "step": 23015,
      "train_runtime": 141647.7548,
      "train_tokens_per_second": 8360.211
    },
    {
      "epoch": 1.5499596014004848,
      "grad_norm": 0.723090103183172,
      "learning_rate": 1.1985197285128853e-06,
      "loss": 0.30386335849761964,
      "num_input_tokens_seen": 1184435640,
      "step": 23020,
      "train_runtime": 141677.4333,
      "train_tokens_per_second": 8360.087
    },
    {
      "epoch": 1.5502962563964449,
      "grad_norm": 0.6987033435081649,
      "learning_rate": 1.1968027259259107e-06,
      "loss": 0.2917176723480225,
      "num_input_tokens_seen": 1184671264,
      "step": 23025,
      "train_runtime": 141706.8258,
      "train_tokens_per_second": 8360.016
    },
    {
      "epoch": 1.5506329113924051,
      "grad_norm": 0.6937165387575358,
      "learning_rate": 1.1950867868909293e-06,
      "loss": 0.2735679864883423,
      "num_input_tokens_seen": 1184915512,
      "step": 23030,
      "train_runtime": 141736.1877,
      "train_tokens_per_second": 8360.007
    },
    {
      "epoch": 1.5509695663883654,
      "grad_norm": 0.6845153138450224,
      "learning_rate": 1.1933719118877923e-06,
      "loss": 0.30825827121734617,
      "num_input_tokens_seen": 1185225296,
      "step": 23035,
      "train_runtime": 141770.2095,
      "train_tokens_per_second": 8360.186
    },
    {
      "epoch": 1.5513062213843254,
      "grad_norm": 0.6049426785969441,
      "learning_rate": 1.1916581013960632e-06,
      "loss": 0.275950288772583,
      "num_input_tokens_seen": 1185516616,
      "step": 23040,
      "train_runtime": 141800.816,
      "train_tokens_per_second": 8360.436
    },
    {
      "epoch": 1.5516428763802854,
      "grad_norm": 0.6830612082386875,
      "learning_rate": 1.189945355895002e-06,
      "loss": 0.3086043357849121,
      "num_input_tokens_seen": 1185764304,
      "step": 23045,
      "train_runtime": 141832.8515,
      "train_tokens_per_second": 8360.294
    },
    {
      "epoch": 1.5519795313762457,
      "grad_norm": 0.6365550490915899,
      "learning_rate": 1.1882336758635727e-06,
      "loss": 0.274019718170166,
      "num_input_tokens_seen": 1186048880,
      "step": 23050,
      "train_runtime": 141863.1899,
      "train_tokens_per_second": 8360.512
    },
    {
      "epoch": 1.552316186372206,
      "grad_norm": 0.6451667495858604,
      "learning_rate": 1.1865230617804412e-06,
      "loss": 0.3103852033615112,
      "num_input_tokens_seen": 1186345416,
      "step": 23055,
      "train_runtime": 141899.8056,
      "train_tokens_per_second": 8360.444
    },
    {
      "epoch": 1.552652841368166,
      "grad_norm": 0.6686553681351011,
      "learning_rate": 1.1848135141239753e-06,
      "loss": 0.30016701221466063,
      "num_input_tokens_seen": 1186600688,
      "step": 23060,
      "train_runtime": 141928.0764,
      "train_tokens_per_second": 8360.578
    },
    {
      "epoch": 1.552989496364126,
      "grad_norm": 0.7377064387088341,
      "learning_rate": 1.1831050333722438e-06,
      "loss": 0.3005824089050293,
      "num_input_tokens_seen": 1186860224,
      "step": 23065,
      "train_runtime": 141957.9572,
      "train_tokens_per_second": 8360.646
    },
    {
      "epoch": 1.5533261513600862,
      "grad_norm": 0.6189396532746998,
      "learning_rate": 1.1813976200030191e-06,
      "loss": 0.28497920036315916,
      "num_input_tokens_seen": 1187114848,
      "step": 23070,
      "train_runtime": 141990.4245,
      "train_tokens_per_second": 8360.527
    },
    {
      "epoch": 1.5536628063560465,
      "grad_norm": 0.6280378097867121,
      "learning_rate": 1.1796912744937717e-06,
      "loss": 0.3084261417388916,
      "num_input_tokens_seen": 1187379328,
      "step": 23075,
      "train_runtime": 142021.9412,
      "train_tokens_per_second": 8360.534
    },
    {
      "epoch": 1.5539994613520065,
      "grad_norm": 0.7423460561451716,
      "learning_rate": 1.1779859973216802e-06,
      "loss": 0.29273371696472167,
      "num_input_tokens_seen": 1187624368,
      "step": 23080,
      "train_runtime": 142051.6515,
      "train_tokens_per_second": 8360.511
    },
    {
      "epoch": 1.5543361163479665,
      "grad_norm": 0.8161547632206546,
      "learning_rate": 1.1762817889636142e-06,
      "loss": 0.3029959201812744,
      "num_input_tokens_seen": 1187846424,
      "step": 23085,
      "train_runtime": 142080.8482,
      "train_tokens_per_second": 8360.356
    },
    {
      "epoch": 1.5546727713439268,
      "grad_norm": 0.6816197135383731,
      "learning_rate": 1.1745786498961542e-06,
      "loss": 0.3192852735519409,
      "num_input_tokens_seen": 1188130728,
      "step": 23090,
      "train_runtime": 142112.1783,
      "train_tokens_per_second": 8360.513
    },
    {
      "epoch": 1.555009426339887,
      "grad_norm": 0.6610291312460334,
      "learning_rate": 1.1728765805955766e-06,
      "loss": 0.27566957473754883,
      "num_input_tokens_seen": 1188384032,
      "step": 23095,
      "train_runtime": 142143.186,
      "train_tokens_per_second": 8360.471
    },
    {
      "epoch": 1.555346081335847,
      "grad_norm": 0.724825008998817,
      "learning_rate": 1.1711755815378595e-06,
      "loss": 0.3177782535552979,
      "num_input_tokens_seen": 1188627848,
      "step": 23100,
      "train_runtime": 142179.5679,
      "train_tokens_per_second": 8360.047
    },
    {
      "epoch": 1.555682736331807,
      "grad_norm": 0.572008468166391,
      "learning_rate": 1.1694756531986818e-06,
      "loss": 0.2873601675033569,
      "num_input_tokens_seen": 1188891384,
      "step": 23105,
      "train_runtime": 142210.3781,
      "train_tokens_per_second": 8360.089
    },
    {
      "epoch": 1.5560193913277673,
      "grad_norm": 0.6894203741657663,
      "learning_rate": 1.1677767960534232e-06,
      "loss": 0.3037106990814209,
      "num_input_tokens_seen": 1189147768,
      "step": 23110,
      "train_runtime": 142243.2325,
      "train_tokens_per_second": 8359.96
    },
    {
      "epoch": 1.5563560463237276,
      "grad_norm": 0.632115033482747,
      "learning_rate": 1.166079010577163e-06,
      "loss": 0.3034500598907471,
      "num_input_tokens_seen": 1189392328,
      "step": 23115,
      "train_runtime": 142274.3874,
      "train_tokens_per_second": 8359.849
    },
    {
      "epoch": 1.5566927013196876,
      "grad_norm": 0.7806419507017245,
      "learning_rate": 1.1643822972446823e-06,
      "loss": 0.32990565299987795,
      "num_input_tokens_seen": 1189661904,
      "step": 23120,
      "train_runtime": 142305.9297,
      "train_tokens_per_second": 8359.89
    },
    {
      "epoch": 1.5570293563156477,
      "grad_norm": 0.581372979067898,
      "learning_rate": 1.1626866565304594e-06,
      "loss": 0.2879173755645752,
      "num_input_tokens_seen": 1189932152,
      "step": 23125,
      "train_runtime": 142335.8535,
      "train_tokens_per_second": 8360.031
    },
    {
      "epoch": 1.557366011311608,
      "grad_norm": 0.6351449182732297,
      "learning_rate": 1.16099208890868e-06,
      "loss": 0.29297752380371095,
      "num_input_tokens_seen": 1190195872,
      "step": 23130,
      "train_runtime": 142367.039,
      "train_tokens_per_second": 8360.052
    },
    {
      "epoch": 1.5577026663075682,
      "grad_norm": 0.6506986186729775,
      "learning_rate": 1.1592985948532187e-06,
      "loss": 0.2894761085510254,
      "num_input_tokens_seen": 1190440648,
      "step": 23135,
      "train_runtime": 142398.8236,
      "train_tokens_per_second": 8359.905
    },
    {
      "epoch": 1.5580393213035282,
      "grad_norm": 0.6103245127942409,
      "learning_rate": 1.1576061748376594e-06,
      "loss": 0.2879594564437866,
      "num_input_tokens_seen": 1190694176,
      "step": 23140,
      "train_runtime": 142433.3019,
      "train_tokens_per_second": 8359.661
    },
    {
      "epoch": 1.5583759762994882,
      "grad_norm": 0.646399668279499,
      "learning_rate": 1.1559148293352805e-06,
      "loss": 0.2879641056060791,
      "num_input_tokens_seen": 1190960248,
      "step": 23145,
      "train_runtime": 142468.7744,
      "train_tokens_per_second": 8359.448
    },
    {
      "epoch": 1.5587126312954485,
      "grad_norm": 0.6296818936443628,
      "learning_rate": 1.154224558819062e-06,
      "loss": 0.2925204038619995,
      "num_input_tokens_seen": 1191223032,
      "step": 23150,
      "train_runtime": 142502.1235,
      "train_tokens_per_second": 8359.335
    },
    {
      "epoch": 1.5590492862914087,
      "grad_norm": 0.6463382098551967,
      "learning_rate": 1.1525353637616821e-06,
      "loss": 0.30001111030578614,
      "num_input_tokens_seen": 1191483400,
      "step": 23155,
      "train_runtime": 142538.0554,
      "train_tokens_per_second": 8359.055
    },
    {
      "epoch": 1.5593859412873687,
      "grad_norm": 0.6363088412188602,
      "learning_rate": 1.1508472446355196e-06,
      "loss": 0.3032949924468994,
      "num_input_tokens_seen": 1191765184,
      "step": 23160,
      "train_runtime": 142565.7543,
      "train_tokens_per_second": 8359.407
    },
    {
      "epoch": 1.5597225962833288,
      "grad_norm": 0.6115432764457933,
      "learning_rate": 1.1491602019126501e-06,
      "loss": 0.30321083068847654,
      "num_input_tokens_seen": 1192053616,
      "step": 23165,
      "train_runtime": 142594.1988,
      "train_tokens_per_second": 8359.762
    },
    {
      "epoch": 1.560059251279289,
      "grad_norm": 0.7075340052509642,
      "learning_rate": 1.1474742360648515e-06,
      "loss": 0.3023722171783447,
      "num_input_tokens_seen": 1192318464,
      "step": 23170,
      "train_runtime": 142623.0255,
      "train_tokens_per_second": 8359.93
    },
    {
      "epoch": 1.5603959062752493,
      "grad_norm": 0.8104415343726629,
      "learning_rate": 1.1457893475635968e-06,
      "loss": 0.32656681537628174,
      "num_input_tokens_seen": 1192562656,
      "step": 23175,
      "train_runtime": 142652.4561,
      "train_tokens_per_second": 8359.917
    },
    {
      "epoch": 1.5607325612712093,
      "grad_norm": 0.584552472123191,
      "learning_rate": 1.144105536880063e-06,
      "loss": 0.272817063331604,
      "num_input_tokens_seen": 1192816336,
      "step": 23180,
      "train_runtime": 142684.3886,
      "train_tokens_per_second": 8359.824
    },
    {
      "epoch": 1.5610692162671693,
      "grad_norm": 0.6555881089780414,
      "learning_rate": 1.1424228044851176e-06,
      "loss": 0.281112813949585,
      "num_input_tokens_seen": 1193020576,
      "step": 23185,
      "train_runtime": 142709.4028,
      "train_tokens_per_second": 8359.79
    },
    {
      "epoch": 1.5614058712631296,
      "grad_norm": 0.6062957149626632,
      "learning_rate": 1.1407411508493355e-06,
      "loss": 0.30576157569885254,
      "num_input_tokens_seen": 1193261608,
      "step": 23190,
      "train_runtime": 142736.6621,
      "train_tokens_per_second": 8359.882
    },
    {
      "epoch": 1.5617425262590898,
      "grad_norm": 0.6759735960761273,
      "learning_rate": 1.1390605764429846e-06,
      "loss": 0.2658079624176025,
      "num_input_tokens_seen": 1193500248,
      "step": 23195,
      "train_runtime": 142767.4448,
      "train_tokens_per_second": 8359.751
    },
    {
      "epoch": 1.5620791812550499,
      "grad_norm": 0.6492994176812896,
      "learning_rate": 1.1373810817360314e-06,
      "loss": 0.29275538921356203,
      "num_input_tokens_seen": 1193775672,
      "step": 23200,
      "train_runtime": 142798.2048,
      "train_tokens_per_second": 8359.879
    },
    {
      "epoch": 1.5624158362510099,
      "grad_norm": 0.7556346988661108,
      "learning_rate": 1.135702667198142e-06,
      "loss": 0.31007847785949705,
      "num_input_tokens_seen": 1194052696,
      "step": 23205,
      "train_runtime": 142826.5451,
      "train_tokens_per_second": 8360.159
    },
    {
      "epoch": 1.5627524912469701,
      "grad_norm": 0.6918717087475715,
      "learning_rate": 1.134025333298679e-06,
      "loss": 0.25963749885559084,
      "num_input_tokens_seen": 1194282984,
      "step": 23210,
      "train_runtime": 142856.4502,
      "train_tokens_per_second": 8360.021
    },
    {
      "epoch": 1.5630891462429304,
      "grad_norm": 0.6177005390079087,
      "learning_rate": 1.1323490805067045e-06,
      "loss": 0.2747525453567505,
      "num_input_tokens_seen": 1194537336,
      "step": 23215,
      "train_runtime": 142887.906,
      "train_tokens_per_second": 8359.961
    },
    {
      "epoch": 1.5634258012388904,
      "grad_norm": 0.6814020576782962,
      "learning_rate": 1.1306739092909757e-06,
      "loss": 0.27852401733398435,
      "num_input_tokens_seen": 1194770712,
      "step": 23220,
      "train_runtime": 142921.4975,
      "train_tokens_per_second": 8359.629
    },
    {
      "epoch": 1.5637624562348504,
      "grad_norm": 0.6997389537665967,
      "learning_rate": 1.1289998201199493e-06,
      "loss": 0.29431188106536865,
      "num_input_tokens_seen": 1195041760,
      "step": 23225,
      "train_runtime": 142953.7031,
      "train_tokens_per_second": 8359.642
    },
    {
      "epoch": 1.5640991112308107,
      "grad_norm": 0.6686719979193267,
      "learning_rate": 1.1273268134617816e-06,
      "loss": 0.295851993560791,
      "num_input_tokens_seen": 1195309296,
      "step": 23230,
      "train_runtime": 142985.8149,
      "train_tokens_per_second": 8359.636
    },
    {
      "epoch": 1.564435766226771,
      "grad_norm": 0.5930119257597855,
      "learning_rate": 1.1256548897843189e-06,
      "loss": 0.3016411542892456,
      "num_input_tokens_seen": 1195589824,
      "step": 23235,
      "train_runtime": 143017.7037,
      "train_tokens_per_second": 8359.733
    },
    {
      "epoch": 1.564772421222731,
      "grad_norm": 0.6519286480678321,
      "learning_rate": 1.1239840495551136e-06,
      "loss": 0.29764366149902344,
      "num_input_tokens_seen": 1195850128,
      "step": 23240,
      "train_runtime": 143048.8283,
      "train_tokens_per_second": 8359.734
    },
    {
      "epoch": 1.565109076218691,
      "grad_norm": 0.6651384755857174,
      "learning_rate": 1.1223142932414087e-06,
      "loss": 0.3278395652770996,
      "num_input_tokens_seen": 1196117704,
      "step": 23245,
      "train_runtime": 143082.5556,
      "train_tokens_per_second": 8359.633
    },
    {
      "epoch": 1.5654457312146512,
      "grad_norm": 0.6045527039483317,
      "learning_rate": 1.120645621310147e-06,
      "loss": 0.28243656158447267,
      "num_input_tokens_seen": 1196372568,
      "step": 23250,
      "train_runtime": 143108.7669,
      "train_tokens_per_second": 8359.883
    },
    {
      "epoch": 1.5657823862106115,
      "grad_norm": 0.6563141657451216,
      "learning_rate": 1.1189780342279666e-06,
      "loss": 0.29192841053009033,
      "num_input_tokens_seen": 1196621456,
      "step": 23255,
      "train_runtime": 143138.0716,
      "train_tokens_per_second": 8359.91
    },
    {
      "epoch": 1.5661190412065715,
      "grad_norm": 0.6455511762904207,
      "learning_rate": 1.1173115324612033e-06,
      "loss": 0.3079479455947876,
      "num_input_tokens_seen": 1196854440,
      "step": 23260,
      "train_runtime": 143167.5784,
      "train_tokens_per_second": 8359.815
    },
    {
      "epoch": 1.5664556962025316,
      "grad_norm": 0.6795695754809344,
      "learning_rate": 1.1156461164758891e-06,
      "loss": 0.3266674757003784,
      "num_input_tokens_seen": 1197124368,
      "step": 23265,
      "train_runtime": 143197.2219,
      "train_tokens_per_second": 8359.969
    },
    {
      "epoch": 1.5667923511984918,
      "grad_norm": 0.5629942030418789,
      "learning_rate": 1.1139817867377522e-06,
      "loss": 0.25736141204833984,
      "num_input_tokens_seen": 1197396264,
      "step": 23270,
      "train_runtime": 143227.6853,
      "train_tokens_per_second": 8360.089
    },
    {
      "epoch": 1.567129006194452,
      "grad_norm": 0.6766158378594144,
      "learning_rate": 1.1123185437122153e-06,
      "loss": 0.2947237014770508,
      "num_input_tokens_seen": 1197659720,
      "step": 23275,
      "train_runtime": 143258.2462,
      "train_tokens_per_second": 8360.145
    },
    {
      "epoch": 1.567465661190412,
      "grad_norm": 0.6083361869511047,
      "learning_rate": 1.1106563878644038e-06,
      "loss": 0.30858473777770995,
      "num_input_tokens_seen": 1197932928,
      "step": 23280,
      "train_runtime": 143288.2142,
      "train_tokens_per_second": 8360.303
    },
    {
      "epoch": 1.5678023161863721,
      "grad_norm": 0.9126258906182185,
      "learning_rate": 1.1089953196591286e-06,
      "loss": 0.30234713554382325,
      "num_input_tokens_seen": 1198183528,
      "step": 23285,
      "train_runtime": 143319.5455,
      "train_tokens_per_second": 8360.224
    },
    {
      "epoch": 1.5681389711823324,
      "grad_norm": 0.6583754770024951,
      "learning_rate": 1.107335339560906e-06,
      "loss": 0.3016059398651123,
      "num_input_tokens_seen": 1198427344,
      "step": 23290,
      "train_runtime": 143348.431,
      "train_tokens_per_second": 8360.24
    },
    {
      "epoch": 1.5684756261782926,
      "grad_norm": 0.7042709313496065,
      "learning_rate": 1.1056764480339427e-06,
      "loss": 0.30268714427947996,
      "num_input_tokens_seen": 1198680544,
      "step": 23295,
      "train_runtime": 143383.3223,
      "train_tokens_per_second": 8359.972
    },
    {
      "epoch": 1.5688122811742526,
      "grad_norm": 0.5951551840488251,
      "learning_rate": 1.1040186455421425e-06,
      "loss": 0.32471671104431155,
      "num_input_tokens_seen": 1198967192,
      "step": 23300,
      "train_runtime": 143417.977,
      "train_tokens_per_second": 8359.951
    },
    {
      "epoch": 1.5691489361702127,
      "grad_norm": 0.7762190869721295,
      "learning_rate": 1.1023619325491052e-06,
      "loss": 0.30966644287109374,
      "num_input_tokens_seen": 1199216032,
      "step": 23305,
      "train_runtime": 143446.0181,
      "train_tokens_per_second": 8360.051
    },
    {
      "epoch": 1.569485591166173,
      "grad_norm": 0.5699275100029038,
      "learning_rate": 1.1007063095181248e-06,
      "loss": 0.2784873962402344,
      "num_input_tokens_seen": 1199462504,
      "step": 23310,
      "train_runtime": 143476.7527,
      "train_tokens_per_second": 8359.978
    },
    {
      "epoch": 1.5698222461621332,
      "grad_norm": 0.8813617667286745,
      "learning_rate": 1.0990517769121905e-06,
      "loss": 0.2906619071960449,
      "num_input_tokens_seen": 1199693504,
      "step": 23315,
      "train_runtime": 143503.7539,
      "train_tokens_per_second": 8360.015
    },
    {
      "epoch": 1.5701589011580932,
      "grad_norm": 0.649236219684033,
      "learning_rate": 1.0973983351939876e-06,
      "loss": 0.30652332305908203,
      "num_input_tokens_seen": 1199955584,
      "step": 23320,
      "train_runtime": 143533.9834,
      "train_tokens_per_second": 8360.08
    },
    {
      "epoch": 1.5704955561540532,
      "grad_norm": 0.6502748064161727,
      "learning_rate": 1.095745984825895e-06,
      "loss": 0.29300885200500487,
      "num_input_tokens_seen": 1200229392,
      "step": 23325,
      "train_runtime": 143568.702,
      "train_tokens_per_second": 8359.965
    },
    {
      "epoch": 1.5708322111500135,
      "grad_norm": 0.6948282270918872,
      "learning_rate": 1.0940947262699902e-06,
      "loss": 0.2954150676727295,
      "num_input_tokens_seen": 1200456128,
      "step": 23330,
      "train_runtime": 143599.8516,
      "train_tokens_per_second": 8359.731
    },
    {
      "epoch": 1.5711688661459737,
      "grad_norm": 0.6044159328164566,
      "learning_rate": 1.092444559988038e-06,
      "loss": 0.288616418838501,
      "num_input_tokens_seen": 1200713408,
      "step": 23335,
      "train_runtime": 143629.5433,
      "train_tokens_per_second": 8359.794
    },
    {
      "epoch": 1.5715055211419338,
      "grad_norm": 0.6170058421273449,
      "learning_rate": 1.0907954864415048e-06,
      "loss": 0.2937595367431641,
      "num_input_tokens_seen": 1200977856,
      "step": 23340,
      "train_runtime": 143658.9086,
      "train_tokens_per_second": 8359.926
    },
    {
      "epoch": 1.5718421761378938,
      "grad_norm": 0.7679418610772841,
      "learning_rate": 1.089147506091549e-06,
      "loss": 0.2810963153839111,
      "num_input_tokens_seen": 1201239656,
      "step": 23345,
      "train_runtime": 143688.2251,
      "train_tokens_per_second": 8360.042
    },
    {
      "epoch": 1.572178831133854,
      "grad_norm": 0.8196264488187123,
      "learning_rate": 1.0875006193990217e-06,
      "loss": 0.31924853324890134,
      "num_input_tokens_seen": 1201481120,
      "step": 23350,
      "train_runtime": 143720.287,
      "train_tokens_per_second": 8359.858
    },
    {
      "epoch": 1.5725154861298143,
      "grad_norm": 0.6845513931408063,
      "learning_rate": 1.0858548268244706e-06,
      "loss": 0.28528895378112795,
      "num_input_tokens_seen": 1201740456,
      "step": 23355,
      "train_runtime": 143748.792,
      "train_tokens_per_second": 8360.004
    },
    {
      "epoch": 1.5728521411257743,
      "grad_norm": 0.676486081890053,
      "learning_rate": 1.0842101288281359e-06,
      "loss": 0.28965229988098146,
      "num_input_tokens_seen": 1201997752,
      "step": 23360,
      "train_runtime": 143776.2852,
      "train_tokens_per_second": 8360.195
    },
    {
      "epoch": 1.5731887961217343,
      "grad_norm": 0.6248384907635289,
      "learning_rate": 1.082566525869952e-06,
      "loss": 0.2914091110229492,
      "num_input_tokens_seen": 1202272664,
      "step": 23365,
      "train_runtime": 143803.0548,
      "train_tokens_per_second": 8360.55
    },
    {
      "epoch": 1.5735254511176946,
      "grad_norm": 0.7879066584836453,
      "learning_rate": 1.0809240184095476e-06,
      "loss": 0.30899786949157715,
      "num_input_tokens_seen": 1202512880,
      "step": 23370,
      "train_runtime": 143831.7169,
      "train_tokens_per_second": 8360.554
    },
    {
      "epoch": 1.5738621061136548,
      "grad_norm": 0.6406333257306212,
      "learning_rate": 1.0792826069062429e-06,
      "loss": 0.26851141452789307,
      "num_input_tokens_seen": 1202762400,
      "step": 23375,
      "train_runtime": 143863.9769,
      "train_tokens_per_second": 8360.414
    },
    {
      "epoch": 1.5741987611096149,
      "grad_norm": 0.6688903146226225,
      "learning_rate": 1.0776422918190576e-06,
      "loss": 0.3093066692352295,
      "num_input_tokens_seen": 1203030400,
      "step": 23380,
      "train_runtime": 143896.8842,
      "train_tokens_per_second": 8360.364
    },
    {
      "epoch": 1.574535416105575,
      "grad_norm": 0.7143298615795994,
      "learning_rate": 1.0760030736066952e-06,
      "loss": 0.2826624155044556,
      "num_input_tokens_seen": 1203260200,
      "step": 23385,
      "train_runtime": 143927.8594,
      "train_tokens_per_second": 8360.162
    },
    {
      "epoch": 1.5748720711015352,
      "grad_norm": 0.6386960561355802,
      "learning_rate": 1.0743649527275619e-06,
      "loss": 0.31038103103637693,
      "num_input_tokens_seen": 1203527000,
      "step": 23390,
      "train_runtime": 143961.3697,
      "train_tokens_per_second": 8360.069
    },
    {
      "epoch": 1.5752087260974954,
      "grad_norm": 0.7395511412699992,
      "learning_rate": 1.0727279296397514e-06,
      "loss": 0.30342764854431153,
      "num_input_tokens_seen": 1203790664,
      "step": 23395,
      "train_runtime": 143993.6779,
      "train_tokens_per_second": 8360.024
    },
    {
      "epoch": 1.5755453810934554,
      "grad_norm": 0.7715601316351341,
      "learning_rate": 1.0710920048010526e-06,
      "loss": 0.3054181098937988,
      "num_input_tokens_seen": 1204054136,
      "step": 23400,
      "train_runtime": 144023.1311,
      "train_tokens_per_second": 8360.144
    },
    {
      "epoch": 1.5758820360894155,
      "grad_norm": 0.6133704767642179,
      "learning_rate": 1.0694571786689462e-06,
      "loss": 0.29744958877563477,
      "num_input_tokens_seen": 1204350208,
      "step": 23405,
      "train_runtime": 144055.1152,
      "train_tokens_per_second": 8360.343
    },
    {
      "epoch": 1.5762186910853757,
      "grad_norm": 0.6524835202980357,
      "learning_rate": 1.067823451700606e-06,
      "loss": 0.288527250289917,
      "num_input_tokens_seen": 1204622784,
      "step": 23410,
      "train_runtime": 144086.5107,
      "train_tokens_per_second": 8360.413
    },
    {
      "epoch": 1.576555346081336,
      "grad_norm": 0.6024572791351785,
      "learning_rate": 1.066190824352899e-06,
      "loss": 0.2972898006439209,
      "num_input_tokens_seen": 1204895912,
      "step": 23415,
      "train_runtime": 144114.8514,
      "train_tokens_per_second": 8360.664
    },
    {
      "epoch": 1.576892001077296,
      "grad_norm": 0.5932539958753045,
      "learning_rate": 1.0645592970823837e-06,
      "loss": 0.2690598487854004,
      "num_input_tokens_seen": 1205127216,
      "step": 23420,
      "train_runtime": 144146.5153,
      "train_tokens_per_second": 8360.433
    },
    {
      "epoch": 1.577228656073256,
      "grad_norm": 0.6136141103837925,
      "learning_rate": 1.0629288703453105e-06,
      "loss": 0.30337493419647216,
      "num_input_tokens_seen": 1205393600,
      "step": 23425,
      "train_runtime": 144178.2092,
      "train_tokens_per_second": 8360.442
    },
    {
      "epoch": 1.5775653110692163,
      "grad_norm": 0.83458075892431,
      "learning_rate": 1.061299544597627e-06,
      "loss": 0.29721770286560056,
      "num_input_tokens_seen": 1205636472,
      "step": 23430,
      "train_runtime": 144205.1456,
      "train_tokens_per_second": 8360.565
    },
    {
      "epoch": 1.5779019660651765,
      "grad_norm": 0.5905094419348519,
      "learning_rate": 1.0596713202949637e-06,
      "loss": 0.2961164712905884,
      "num_input_tokens_seen": 1205901744,
      "step": 23435,
      "train_runtime": 144232.1391,
      "train_tokens_per_second": 8360.839
    },
    {
      "epoch": 1.5782386210611365,
      "grad_norm": 0.6228808404823696,
      "learning_rate": 1.0580441978926527e-06,
      "loss": 0.3093919277191162,
      "num_input_tokens_seen": 1206141720,
      "step": 23440,
      "train_runtime": 144264.2871,
      "train_tokens_per_second": 8360.64
    },
    {
      "epoch": 1.5785752760570966,
      "grad_norm": 0.6642365329214925,
      "learning_rate": 1.056418177845711e-06,
      "loss": 0.28920223712921145,
      "num_input_tokens_seen": 1206372248,
      "step": 23445,
      "train_runtime": 144293.7192,
      "train_tokens_per_second": 8360.532
    },
    {
      "epoch": 1.5789119310530568,
      "grad_norm": 0.5942111558022745,
      "learning_rate": 1.0547932606088506e-06,
      "loss": 0.3007926940917969,
      "num_input_tokens_seen": 1206620000,
      "step": 23450,
      "train_runtime": 144321.2873,
      "train_tokens_per_second": 8360.652
    },
    {
      "epoch": 1.579248586049017,
      "grad_norm": 0.6422102037803108,
      "learning_rate": 1.053169446636475e-06,
      "loss": 0.29891166687011717,
      "num_input_tokens_seen": 1206871256,
      "step": 23455,
      "train_runtime": 144355.4032,
      "train_tokens_per_second": 8360.416
    },
    {
      "epoch": 1.579585241044977,
      "grad_norm": 0.7623544405142704,
      "learning_rate": 1.0515467363826782e-06,
      "loss": 0.3001178026199341,
      "num_input_tokens_seen": 1207154200,
      "step": 23460,
      "train_runtime": 144386.6948,
      "train_tokens_per_second": 8360.564
    },
    {
      "epoch": 1.5799218960409371,
      "grad_norm": 0.6530079817207814,
      "learning_rate": 1.0499251303012454e-06,
      "loss": 0.28798685073852537,
      "num_input_tokens_seen": 1207409488,
      "step": 23465,
      "train_runtime": 144422.7792,
      "train_tokens_per_second": 8360.243
    },
    {
      "epoch": 1.5802585510368974,
      "grad_norm": 0.6952289851231613,
      "learning_rate": 1.0483046288456544e-06,
      "loss": 0.30144472122192384,
      "num_input_tokens_seen": 1207654600,
      "step": 23470,
      "train_runtime": 144453.6301,
      "train_tokens_per_second": 8360.154
    },
    {
      "epoch": 1.5805952060328576,
      "grad_norm": 0.6764923081112497,
      "learning_rate": 1.0466852324690719e-06,
      "loss": 0.31884446144104006,
      "num_input_tokens_seen": 1207911048,
      "step": 23475,
      "train_runtime": 144484.4623,
      "train_tokens_per_second": 8360.145
    },
    {
      "epoch": 1.5809318610288177,
      "grad_norm": 0.7237701169155566,
      "learning_rate": 1.0450669416243596e-06,
      "loss": 0.29470338821411135,
      "num_input_tokens_seen": 1208180424,
      "step": 23480,
      "train_runtime": 144514.5964,
      "train_tokens_per_second": 8360.266
    },
    {
      "epoch": 1.5812685160247777,
      "grad_norm": 0.5392405444229407,
      "learning_rate": 1.0434497567640639e-06,
      "loss": 0.30573101043701173,
      "num_input_tokens_seen": 1208451360,
      "step": 23485,
      "train_runtime": 144542.7625,
      "train_tokens_per_second": 8360.511
    },
    {
      "epoch": 1.581605171020738,
      "grad_norm": 0.7339514795469303,
      "learning_rate": 1.0418336783404282e-06,
      "loss": 0.31420207023620605,
      "num_input_tokens_seen": 1208712408,
      "step": 23490,
      "train_runtime": 144577.7539,
      "train_tokens_per_second": 8360.293
    },
    {
      "epoch": 1.5819418260166982,
      "grad_norm": 0.6521132183821964,
      "learning_rate": 1.0402187068053825e-06,
      "loss": 0.28568623065948484,
      "num_input_tokens_seen": 1208971032,
      "step": 23495,
      "train_runtime": 144607.5336,
      "train_tokens_per_second": 8360.36
    },
    {
      "epoch": 1.5822784810126582,
      "grad_norm": 0.6360581559894941,
      "learning_rate": 1.0386048426105495e-06,
      "loss": 0.289516806602478,
      "num_input_tokens_seen": 1209232256,
      "step": 23500,
      "train_runtime": 144639.1637,
      "train_tokens_per_second": 8360.338
    },
    {
      "epoch": 1.5826151360086183,
      "grad_norm": 0.644058322110758,
      "learning_rate": 1.0369920862072396e-06,
      "loss": 0.2771589756011963,
      "num_input_tokens_seen": 1209472400,
      "step": 23505,
      "train_runtime": 144667.2371,
      "train_tokens_per_second": 8360.375
    },
    {
      "epoch": 1.5829517910045785,
      "grad_norm": 0.7431002275034424,
      "learning_rate": 1.0353804380464556e-06,
      "loss": 0.2816773891448975,
      "num_input_tokens_seen": 1209718344,
      "step": 23510,
      "train_runtime": 144696.9764,
      "train_tokens_per_second": 8360.357
    },
    {
      "epoch": 1.5832884460005388,
      "grad_norm": 0.7035085867109769,
      "learning_rate": 1.0337698985788903e-06,
      "loss": 0.28846981525421145,
      "num_input_tokens_seen": 1209987000,
      "step": 23515,
      "train_runtime": 144722.6845,
      "train_tokens_per_second": 8360.728
    },
    {
      "epoch": 1.5836251009964988,
      "grad_norm": 0.6483816374026294,
      "learning_rate": 1.0321604682549247e-06,
      "loss": 0.2702645301818848,
      "num_input_tokens_seen": 1210253288,
      "step": 23520,
      "train_runtime": 144752.5908,
      "train_tokens_per_second": 8360.84
    },
    {
      "epoch": 1.5839617559924588,
      "grad_norm": 0.6853542900560545,
      "learning_rate": 1.0305521475246311e-06,
      "loss": 0.31032357215881345,
      "num_input_tokens_seen": 1210502032,
      "step": 23525,
      "train_runtime": 144782.4723,
      "train_tokens_per_second": 8360.833
    },
    {
      "epoch": 1.584298410988419,
      "grad_norm": 0.7278333156602855,
      "learning_rate": 1.028944936837774e-06,
      "loss": 0.3138599872589111,
      "num_input_tokens_seen": 1210737816,
      "step": 23530,
      "train_runtime": 144812.7614,
      "train_tokens_per_second": 8360.712
    },
    {
      "epoch": 1.5846350659843793,
      "grad_norm": 0.6629194868070156,
      "learning_rate": 1.0273388366438003e-06,
      "loss": 0.299422287940979,
      "num_input_tokens_seen": 1210994136,
      "step": 23535,
      "train_runtime": 144844.2746,
      "train_tokens_per_second": 8360.663
    },
    {
      "epoch": 1.5849717209803393,
      "grad_norm": 0.6326292898098356,
      "learning_rate": 1.0257338473918538e-06,
      "loss": 0.2888387441635132,
      "num_input_tokens_seen": 1211257760,
      "step": 23540,
      "train_runtime": 144874.9683,
      "train_tokens_per_second": 8360.711
    },
    {
      "epoch": 1.5853083759762994,
      "grad_norm": 0.6190634443688546,
      "learning_rate": 1.0241299695307644e-06,
      "loss": 0.29530422687530516,
      "num_input_tokens_seen": 1211502472,
      "step": 23545,
      "train_runtime": 144905.8715,
      "train_tokens_per_second": 8360.617
    },
    {
      "epoch": 1.5856450309722596,
      "grad_norm": 0.5602222133692157,
      "learning_rate": 1.0225272035090506e-06,
      "loss": 0.28389706611633303,
      "num_input_tokens_seen": 1211782920,
      "step": 23550,
      "train_runtime": 144934.8764,
      "train_tokens_per_second": 8360.879
    },
    {
      "epoch": 1.5859816859682199,
      "grad_norm": 0.7243875219161162,
      "learning_rate": 1.0209255497749209e-06,
      "loss": 0.3025680065155029,
      "num_input_tokens_seen": 1212042080,
      "step": 23555,
      "train_runtime": 144968.0172,
      "train_tokens_per_second": 8360.755
    },
    {
      "epoch": 1.58631834096418,
      "grad_norm": 0.8336036106804539,
      "learning_rate": 1.0193250087762735e-06,
      "loss": 0.2802864074707031,
      "num_input_tokens_seen": 1212304488,
      "step": 23560,
      "train_runtime": 145000.5744,
      "train_tokens_per_second": 8360.687
    },
    {
      "epoch": 1.58665499596014,
      "grad_norm": 0.5440361622503938,
      "learning_rate": 1.0177255809606934e-06,
      "loss": 0.27771122455596925,
      "num_input_tokens_seen": 1212581312,
      "step": 23565,
      "train_runtime": 145029.0832,
      "train_tokens_per_second": 8360.953
    },
    {
      "epoch": 1.5869916509561002,
      "grad_norm": 0.713519418721198,
      "learning_rate": 1.0161272667754562e-06,
      "loss": 0.2999070882797241,
      "num_input_tokens_seen": 1212855608,
      "step": 23570,
      "train_runtime": 145059.2187,
      "train_tokens_per_second": 8361.107
    },
    {
      "epoch": 1.5873283059520604,
      "grad_norm": 0.5548812305866564,
      "learning_rate": 1.014530066667524e-06,
      "loss": 0.2997293949127197,
      "num_input_tokens_seen": 1213128152,
      "step": 23575,
      "train_runtime": 145088.5681,
      "train_tokens_per_second": 8361.294
    },
    {
      "epoch": 1.5876649609480205,
      "grad_norm": 0.6678296192597087,
      "learning_rate": 1.0129339810835526e-06,
      "loss": 0.2756863832473755,
      "num_input_tokens_seen": 1213360160,
      "step": 23580,
      "train_runtime": 145114.3074,
      "train_tokens_per_second": 8361.41
    },
    {
      "epoch": 1.5880016159439805,
      "grad_norm": 0.6701443157461248,
      "learning_rate": 1.0113390104698767e-06,
      "loss": 0.2857106447219849,
      "num_input_tokens_seen": 1213637192,
      "step": 23585,
      "train_runtime": 145141.2804,
      "train_tokens_per_second": 8361.764
    },
    {
      "epoch": 1.5883382709399407,
      "grad_norm": 0.6438125592847754,
      "learning_rate": 1.009745155272529e-06,
      "loss": 0.31817386150360105,
      "num_input_tokens_seen": 1213907136,
      "step": 23590,
      "train_runtime": 145176.3483,
      "train_tokens_per_second": 8361.604
    },
    {
      "epoch": 1.588674925935901,
      "grad_norm": 0.6633413404335571,
      "learning_rate": 1.0081524159372246e-06,
      "loss": 0.302812385559082,
      "num_input_tokens_seen": 1214156528,
      "step": 23595,
      "train_runtime": 145203.8954,
      "train_tokens_per_second": 8361.735
    },
    {
      "epoch": 1.589011580931861,
      "grad_norm": 0.6885658263056451,
      "learning_rate": 1.0065607929093685e-06,
      "loss": 0.2925971269607544,
      "num_input_tokens_seen": 1214395240,
      "step": 23600,
      "train_runtime": 145236.3857,
      "train_tokens_per_second": 8361.508
    },
    {
      "epoch": 1.589348235927821,
      "grad_norm": 0.6323618340875348,
      "learning_rate": 1.0049702866340521e-06,
      "loss": 0.26336758136749266,
      "num_input_tokens_seen": 1214660712,
      "step": 23605,
      "train_runtime": 145267.912,
      "train_tokens_per_second": 8361.521
    },
    {
      "epoch": 1.5896848909237813,
      "grad_norm": 0.6899535640200979,
      "learning_rate": 1.0033808975560556e-06,
      "loss": 0.28592615127563475,
      "num_input_tokens_seen": 1214919400,
      "step": 23610,
      "train_runtime": 145296.0633,
      "train_tokens_per_second": 8361.681
    },
    {
      "epoch": 1.5900215459197415,
      "grad_norm": 0.6384584035660178,
      "learning_rate": 1.0017926261198473e-06,
      "loss": 0.2882401466369629,
      "num_input_tokens_seen": 1215161360,
      "step": 23615,
      "train_runtime": 145323.0556,
      "train_tokens_per_second": 8361.793
    },
    {
      "epoch": 1.5903582009157016,
      "grad_norm": 0.8377479692130131,
      "learning_rate": 1.0002054727695814e-06,
      "loss": 0.2994795083999634,
      "num_input_tokens_seen": 1215402632,
      "step": 23620,
      "train_runtime": 145351.6187,
      "train_tokens_per_second": 8361.81
    },
    {
      "epoch": 1.5906948559116616,
      "grad_norm": 0.5815025739591071,
      "learning_rate": 9.986194379490993e-07,
      "loss": 0.29154505729675295,
      "num_input_tokens_seen": 1215652752,
      "step": 23625,
      "train_runtime": 145382.3933,
      "train_tokens_per_second": 8361.76
    },
    {
      "epoch": 1.5910315109076218,
      "grad_norm": 0.6678025833673056,
      "learning_rate": 9.970345221019345e-07,
      "loss": 0.3099416494369507,
      "num_input_tokens_seen": 1215899856,
      "step": 23630,
      "train_runtime": 145412.3453,
      "train_tokens_per_second": 8361.737
    },
    {
      "epoch": 1.591368165903582,
      "grad_norm": 0.824195588691417,
      "learning_rate": 9.954507256712987e-07,
      "loss": 0.2955456733703613,
      "num_input_tokens_seen": 1216155312,
      "step": 23635,
      "train_runtime": 145442.0961,
      "train_tokens_per_second": 8361.783
    },
    {
      "epoch": 1.5917048208995421,
      "grad_norm": 0.7431420474850635,
      "learning_rate": 9.938680491000991e-07,
      "loss": 0.2936047792434692,
      "num_input_tokens_seen": 1216392600,
      "step": 23640,
      "train_runtime": 145469.5262,
      "train_tokens_per_second": 8361.838
    },
    {
      "epoch": 1.5920414758955022,
      "grad_norm": 0.6707766447058537,
      "learning_rate": 9.92286492830925e-07,
      "loss": 0.2891110897064209,
      "num_input_tokens_seen": 1216660664,
      "step": 23645,
      "train_runtime": 145502.848,
      "train_tokens_per_second": 8361.765
    },
    {
      "epoch": 1.5923781308914624,
      "grad_norm": 0.648089382730063,
      "learning_rate": 9.907060573060535e-07,
      "loss": 0.3095808267593384,
      "num_input_tokens_seen": 1216939192,
      "step": 23650,
      "train_runtime": 145533.1479,
      "train_tokens_per_second": 8361.938
    },
    {
      "epoch": 1.5927147858874227,
      "grad_norm": 0.6380320629754798,
      "learning_rate": 9.891267429674484e-07,
      "loss": 0.3120683431625366,
      "num_input_tokens_seen": 1217182136,
      "step": 23655,
      "train_runtime": 145560.8961,
      "train_tokens_per_second": 8362.013
    },
    {
      "epoch": 1.5930514408833827,
      "grad_norm": 0.6330264193186063,
      "learning_rate": 9.8754855025676e-07,
      "loss": 0.2787421703338623,
      "num_input_tokens_seen": 1217434992,
      "step": 23660,
      "train_runtime": 145590.8162,
      "train_tokens_per_second": 8362.032
    },
    {
      "epoch": 1.5933880958793427,
      "grad_norm": 0.7103133190450351,
      "learning_rate": 9.85971479615324e-07,
      "loss": 0.28728713989257815,
      "num_input_tokens_seen": 1217713760,
      "step": 23665,
      "train_runtime": 145622.0204,
      "train_tokens_per_second": 8362.154
    },
    {
      "epoch": 1.593724750875303,
      "grad_norm": 0.5953273155878402,
      "learning_rate": 9.843955314841647e-07,
      "loss": 0.2867331027984619,
      "num_input_tokens_seen": 1217989864,
      "step": 23670,
      "train_runtime": 145655.3023,
      "train_tokens_per_second": 8362.139
    },
    {
      "epoch": 1.5940614058712632,
      "grad_norm": 0.6876706614645887,
      "learning_rate": 9.82820706303989e-07,
      "loss": 0.29907841682434083,
      "num_input_tokens_seen": 1218257720,
      "step": 23675,
      "train_runtime": 145689.4775,
      "train_tokens_per_second": 8362.016
    },
    {
      "epoch": 1.5943980608672232,
      "grad_norm": 0.7478138186158814,
      "learning_rate": 9.812470045151952e-07,
      "loss": 0.29465272426605227,
      "num_input_tokens_seen": 1218496656,
      "step": 23680,
      "train_runtime": 145719.2542,
      "train_tokens_per_second": 8361.947
    },
    {
      "epoch": 1.5947347158631833,
      "grad_norm": 0.6744272534596973,
      "learning_rate": 9.79674426557859e-07,
      "loss": 0.2935919761657715,
      "num_input_tokens_seen": 1218737136,
      "step": 23685,
      "train_runtime": 145747.035,
      "train_tokens_per_second": 8362.003
    },
    {
      "epoch": 1.5950713708591435,
      "grad_norm": 0.5834842503949419,
      "learning_rate": 9.781029728717513e-07,
      "loss": 0.2769655227661133,
      "num_input_tokens_seen": 1218985112,
      "step": 23690,
      "train_runtime": 145776.6463,
      "train_tokens_per_second": 8362.005
    },
    {
      "epoch": 1.5954080258551038,
      "grad_norm": 0.6678784800698959,
      "learning_rate": 9.765326438963218e-07,
      "loss": 0.3156560182571411,
      "num_input_tokens_seen": 1219237376,
      "step": 23695,
      "train_runtime": 145805.564,
      "train_tokens_per_second": 8362.077
    },
    {
      "epoch": 1.5957446808510638,
      "grad_norm": 0.7100237004046561,
      "learning_rate": 9.749634400707087e-07,
      "loss": 0.28241448402404784,
      "num_input_tokens_seen": 1219476600,
      "step": 23700,
      "train_runtime": 145833.7132,
      "train_tokens_per_second": 8362.103
    },
    {
      "epoch": 1.5960813358470238,
      "grad_norm": 0.6330056537954944,
      "learning_rate": 9.733953618337344e-07,
      "loss": 0.3010775089263916,
      "num_input_tokens_seen": 1219705112,
      "step": 23705,
      "train_runtime": 145865.9093,
      "train_tokens_per_second": 8361.824
    },
    {
      "epoch": 1.596417990842984,
      "grad_norm": 0.5913911327619016,
      "learning_rate": 9.718284096239077e-07,
      "loss": 0.30742506980895995,
      "num_input_tokens_seen": 1220003816,
      "step": 23710,
      "train_runtime": 145897.0747,
      "train_tokens_per_second": 8362.086
    },
    {
      "epoch": 1.5967546458389443,
      "grad_norm": 0.720690014517704,
      "learning_rate": 9.702625838794215e-07,
      "loss": 0.28258519172668456,
      "num_input_tokens_seen": 1220251040,
      "step": 23715,
      "train_runtime": 145931.5985,
      "train_tokens_per_second": 8361.801
    },
    {
      "epoch": 1.5970913008349044,
      "grad_norm": 0.6307921233023355,
      "learning_rate": 9.686978850381535e-07,
      "loss": 0.28132071495056155,
      "num_input_tokens_seen": 1220523056,
      "step": 23720,
      "train_runtime": 145960.0686,
      "train_tokens_per_second": 8362.034
    },
    {
      "epoch": 1.5974279558308644,
      "grad_norm": 0.5971936128262016,
      "learning_rate": 9.67134313537666e-07,
      "loss": 0.28850390911102297,
      "num_input_tokens_seen": 1220774984,
      "step": 23725,
      "train_runtime": 145991.6978,
      "train_tokens_per_second": 8361.948
    },
    {
      "epoch": 1.5977646108268246,
      "grad_norm": 0.6401787712685798,
      "learning_rate": 9.6557186981521e-07,
      "loss": 0.2918362617492676,
      "num_input_tokens_seen": 1221043480,
      "step": 23730,
      "train_runtime": 146023.1212,
      "train_tokens_per_second": 8361.987
    },
    {
      "epoch": 1.5981012658227849,
      "grad_norm": 0.6671128150184781,
      "learning_rate": 9.640105543077133e-07,
      "loss": 0.2828370094299316,
      "num_input_tokens_seen": 1221291088,
      "step": 23735,
      "train_runtime": 146051.1914,
      "train_tokens_per_second": 8362.075
    },
    {
      "epoch": 1.598437920818745,
      "grad_norm": 0.6035815682896537,
      "learning_rate": 9.624503674517972e-07,
      "loss": 0.3029410600662231,
      "num_input_tokens_seen": 1221539752,
      "step": 23740,
      "train_runtime": 146085.1037,
      "train_tokens_per_second": 8361.837
    },
    {
      "epoch": 1.598774575814705,
      "grad_norm": 0.5795016833106137,
      "learning_rate": 9.608913096837603e-07,
      "loss": 0.2747305393218994,
      "num_input_tokens_seen": 1221799600,
      "step": 23745,
      "train_runtime": 146117.8877,
      "train_tokens_per_second": 8361.739
    },
    {
      "epoch": 1.5991112308106652,
      "grad_norm": 0.7293764320134308,
      "learning_rate": 9.593333814395889e-07,
      "loss": 0.30777502059936523,
      "num_input_tokens_seen": 1222034488,
      "step": 23750,
      "train_runtime": 146149.0796,
      "train_tokens_per_second": 8361.561
    },
    {
      "epoch": 1.5994478858066254,
      "grad_norm": 0.5778462266155275,
      "learning_rate": 9.577765831549529e-07,
      "loss": 0.30039410591125487,
      "num_input_tokens_seen": 1222295736,
      "step": 23755,
      "train_runtime": 146180.5241,
      "train_tokens_per_second": 8361.55
    },
    {
      "epoch": 1.5997845408025855,
      "grad_norm": 0.6127373211559891,
      "learning_rate": 9.562209152652058e-07,
      "loss": 0.3062568664550781,
      "num_input_tokens_seen": 1222543184,
      "step": 23760,
      "train_runtime": 146211.6263,
      "train_tokens_per_second": 8361.464
    },
    {
      "epoch": 1.6001211957985455,
      "grad_norm": 0.7200216717304828,
      "learning_rate": 9.54666378205385e-07,
      "loss": 0.31950066089630125,
      "num_input_tokens_seen": 1222815344,
      "step": 23765,
      "train_runtime": 146239.4645,
      "train_tokens_per_second": 8361.733
    },
    {
      "epoch": 1.6004578507945058,
      "grad_norm": 0.6789385423787447,
      "learning_rate": 9.531129724102117e-07,
      "loss": 0.2981275081634521,
      "num_input_tokens_seen": 1223060840,
      "step": 23770,
      "train_runtime": 146272.9019,
      "train_tokens_per_second": 8361.5
    },
    {
      "epoch": 1.600794505790466,
      "grad_norm": 0.6971041335590995,
      "learning_rate": 9.515606983140896e-07,
      "loss": 0.3016944169998169,
      "num_input_tokens_seen": 1223320744,
      "step": 23775,
      "train_runtime": 146305.3144,
      "train_tokens_per_second": 8361.424
    },
    {
      "epoch": 1.601131160786426,
      "grad_norm": 0.6767868798568395,
      "learning_rate": 9.500095563511119e-07,
      "loss": 0.2934077024459839,
      "num_input_tokens_seen": 1223595280,
      "step": 23780,
      "train_runtime": 146334.3028,
      "train_tokens_per_second": 8361.644
    },
    {
      "epoch": 1.601467815782386,
      "grad_norm": 0.5978612545267852,
      "learning_rate": 9.484595469550445e-07,
      "loss": 0.2953498125076294,
      "num_input_tokens_seen": 1223863632,
      "step": 23785,
      "train_runtime": 146364.9215,
      "train_tokens_per_second": 8361.728
    },
    {
      "epoch": 1.6018044707783463,
      "grad_norm": 0.6220249786225521,
      "learning_rate": 9.469106705593462e-07,
      "loss": 0.2786733388900757,
      "num_input_tokens_seen": 1224091144,
      "step": 23790,
      "train_runtime": 146395.329,
      "train_tokens_per_second": 8361.545
    },
    {
      "epoch": 1.6021411257743066,
      "grad_norm": 0.6432345852994249,
      "learning_rate": 9.453629275971549e-07,
      "loss": 0.29642181396484374,
      "num_input_tokens_seen": 1224356400,
      "step": 23795,
      "train_runtime": 146424.6478,
      "train_tokens_per_second": 8361.682
    },
    {
      "epoch": 1.6024777807702666,
      "grad_norm": 0.5792357910123949,
      "learning_rate": 9.438163185012916e-07,
      "loss": 0.29220259189605713,
      "num_input_tokens_seen": 1224603976,
      "step": 23800,
      "train_runtime": 146455.1097,
      "train_tokens_per_second": 8361.634
    },
    {
      "epoch": 1.6028144357662266,
      "grad_norm": 0.6187052863013284,
      "learning_rate": 9.422708437042604e-07,
      "loss": 0.2868655204772949,
      "num_input_tokens_seen": 1224874864,
      "step": 23805,
      "train_runtime": 146484.0244,
      "train_tokens_per_second": 8361.832
    },
    {
      "epoch": 1.6031510907621869,
      "grad_norm": 0.6898313930803983,
      "learning_rate": 9.40726503638249e-07,
      "loss": 0.3110226631164551,
      "num_input_tokens_seen": 1225137840,
      "step": 23810,
      "train_runtime": 146515.1065,
      "train_tokens_per_second": 8361.853
    },
    {
      "epoch": 1.6034877457581471,
      "grad_norm": 0.6824168967070373,
      "learning_rate": 9.391832987351268e-07,
      "loss": 0.2913219451904297,
      "num_input_tokens_seen": 1225393664,
      "step": 23815,
      "train_runtime": 146546.9211,
      "train_tokens_per_second": 8361.784
    },
    {
      "epoch": 1.6038244007541071,
      "grad_norm": 0.6416622963337121,
      "learning_rate": 9.376412294264458e-07,
      "loss": 0.296143102645874,
      "num_input_tokens_seen": 1225656344,
      "step": 23820,
      "train_runtime": 146579.0881,
      "train_tokens_per_second": 8361.741
    },
    {
      "epoch": 1.6041610557500672,
      "grad_norm": 0.6491185518275011,
      "learning_rate": 9.361002961434401e-07,
      "loss": 0.2969661712646484,
      "num_input_tokens_seen": 1225896016,
      "step": 23825,
      "train_runtime": 146608.4237,
      "train_tokens_per_second": 8361.702
    },
    {
      "epoch": 1.6044977107460274,
      "grad_norm": 0.7071362161716219,
      "learning_rate": 9.345604993170299e-07,
      "loss": 0.3105741262435913,
      "num_input_tokens_seen": 1226172664,
      "step": 23830,
      "train_runtime": 146635.7828,
      "train_tokens_per_second": 8362.029
    },
    {
      "epoch": 1.6048343657419877,
      "grad_norm": 0.6587829784727106,
      "learning_rate": 9.330218393778101e-07,
      "loss": 0.2888102293014526,
      "num_input_tokens_seen": 1226420176,
      "step": 23835,
      "train_runtime": 146663.9906,
      "train_tokens_per_second": 8362.108
    },
    {
      "epoch": 1.6051710207379477,
      "grad_norm": 0.7097418159346436,
      "learning_rate": 9.314843167560656e-07,
      "loss": 0.28895087242126466,
      "num_input_tokens_seen": 1226683224,
      "step": 23840,
      "train_runtime": 146695.0572,
      "train_tokens_per_second": 8362.131
    },
    {
      "epoch": 1.6055076757339077,
      "grad_norm": 0.6150528265499648,
      "learning_rate": 9.299479318817578e-07,
      "loss": 0.2957641124725342,
      "num_input_tokens_seen": 1226934176,
      "step": 23845,
      "train_runtime": 146723.015,
      "train_tokens_per_second": 8362.248
    },
    {
      "epoch": 1.605844330729868,
      "grad_norm": 0.6727078976027892,
      "learning_rate": 9.284126851845316e-07,
      "loss": 0.28337299823760986,
      "num_input_tokens_seen": 1227217760,
      "step": 23850,
      "train_runtime": 146757.0359,
      "train_tokens_per_second": 8362.241
    },
    {
      "epoch": 1.6061809857258282,
      "grad_norm": 0.6859478632759055,
      "learning_rate": 9.268785770937172e-07,
      "loss": 0.2946218967437744,
      "num_input_tokens_seen": 1227461704,
      "step": 23855,
      "train_runtime": 146785.5534,
      "train_tokens_per_second": 8362.279
    },
    {
      "epoch": 1.6065176407217883,
      "grad_norm": 0.6877453225928777,
      "learning_rate": 9.253456080383178e-07,
      "loss": 0.30165188312530516,
      "num_input_tokens_seen": 1227724744,
      "step": 23860,
      "train_runtime": 146815.2334,
      "train_tokens_per_second": 8362.38
    },
    {
      "epoch": 1.6068542957177483,
      "grad_norm": 0.7250063832055329,
      "learning_rate": 9.238137784470286e-07,
      "loss": 0.32261099815368655,
      "num_input_tokens_seen": 1227966512,
      "step": 23865,
      "train_runtime": 146848.2607,
      "train_tokens_per_second": 8362.145
    },
    {
      "epoch": 1.6071909507137085,
      "grad_norm": 0.6899506612529626,
      "learning_rate": 9.222830887482153e-07,
      "loss": 0.26985774040222166,
      "num_input_tokens_seen": 1228231040,
      "step": 23870,
      "train_runtime": 146879.3885,
      "train_tokens_per_second": 8362.174
    },
    {
      "epoch": 1.6075276057096688,
      "grad_norm": 0.7555405981372472,
      "learning_rate": 9.207535393699351e-07,
      "loss": 0.30493524074554446,
      "num_input_tokens_seen": 1228469696,
      "step": 23875,
      "train_runtime": 146910.5583,
      "train_tokens_per_second": 8362.025
    },
    {
      "epoch": 1.6078642607056288,
      "grad_norm": 0.6569791089162996,
      "learning_rate": 9.192251307399197e-07,
      "loss": 0.29795596599578855,
      "num_input_tokens_seen": 1228729848,
      "step": 23880,
      "train_runtime": 146938.7106,
      "train_tokens_per_second": 8362.193
    },
    {
      "epoch": 1.6082009157015889,
      "grad_norm": 0.6514612154299279,
      "learning_rate": 9.176978632855842e-07,
      "loss": 0.2779392719268799,
      "num_input_tokens_seen": 1228977928,
      "step": 23885,
      "train_runtime": 146970.7223,
      "train_tokens_per_second": 8362.06
    },
    {
      "epoch": 1.608537570697549,
      "grad_norm": 0.6271218134397402,
      "learning_rate": 9.161717374340235e-07,
      "loss": 0.2775870323181152,
      "num_input_tokens_seen": 1229262528,
      "step": 23890,
      "train_runtime": 147003.493,
      "train_tokens_per_second": 8362.131
    },
    {
      "epoch": 1.6088742256935094,
      "grad_norm": 0.7648109827073816,
      "learning_rate": 9.14646753612014e-07,
      "loss": 0.29368042945861816,
      "num_input_tokens_seen": 1229537328,
      "step": 23895,
      "train_runtime": 147034.2189,
      "train_tokens_per_second": 8362.253
    },
    {
      "epoch": 1.6092108806894694,
      "grad_norm": 0.6443228120399127,
      "learning_rate": 9.131229122460112e-07,
      "loss": 0.2932413101196289,
      "num_input_tokens_seen": 1229822112,
      "step": 23900,
      "train_runtime": 147063.4031,
      "train_tokens_per_second": 8362.53
    },
    {
      "epoch": 1.6095475356854294,
      "grad_norm": 0.65649828824507,
      "learning_rate": 9.116002137621566e-07,
      "loss": 0.2983255863189697,
      "num_input_tokens_seen": 1230076888,
      "step": 23905,
      "train_runtime": 147091.5833,
      "train_tokens_per_second": 8362.66
    },
    {
      "epoch": 1.6098841906813897,
      "grad_norm": 0.8233882077307837,
      "learning_rate": 9.100786585862626e-07,
      "loss": 0.3283552646636963,
      "num_input_tokens_seen": 1230304792,
      "step": 23910,
      "train_runtime": 147121.5927,
      "train_tokens_per_second": 8362.503
    },
    {
      "epoch": 1.61022084567735,
      "grad_norm": 0.6656350125790199,
      "learning_rate": 9.085582471438326e-07,
      "loss": 0.2980748176574707,
      "num_input_tokens_seen": 1230558360,
      "step": 23915,
      "train_runtime": 147153.1002,
      "train_tokens_per_second": 8362.436
    },
    {
      "epoch": 1.61055750067331,
      "grad_norm": 0.895898321092663,
      "learning_rate": 9.070389798600394e-07,
      "loss": 0.3171079158782959,
      "num_input_tokens_seen": 1230815704,
      "step": 23920,
      "train_runtime": 147181.1625,
      "train_tokens_per_second": 8362.59
    },
    {
      "epoch": 1.61089415566927,
      "grad_norm": 0.7213360936445516,
      "learning_rate": 9.055208571597451e-07,
      "loss": 0.3079305410385132,
      "num_input_tokens_seen": 1231077096,
      "step": 23925,
      "train_runtime": 147208.5541,
      "train_tokens_per_second": 8362.81
    },
    {
      "epoch": 1.6112308106652302,
      "grad_norm": 0.6777266370607684,
      "learning_rate": 9.040038794674872e-07,
      "loss": 0.3125328540802002,
      "num_input_tokens_seen": 1231334568,
      "step": 23930,
      "train_runtime": 147235.6571,
      "train_tokens_per_second": 8363.019
    },
    {
      "epoch": 1.6115674656611905,
      "grad_norm": 0.6914477395573208,
      "learning_rate": 9.024880472074831e-07,
      "loss": 0.2857400894165039,
      "num_input_tokens_seen": 1231558616,
      "step": 23935,
      "train_runtime": 147265.2303,
      "train_tokens_per_second": 8362.861
    },
    {
      "epoch": 1.6119041206571505,
      "grad_norm": 0.5862451933031427,
      "learning_rate": 9.009733608036308e-07,
      "loss": 0.31400446891784667,
      "num_input_tokens_seen": 1231828192,
      "step": 23940,
      "train_runtime": 147297.8583,
      "train_tokens_per_second": 8362.838
    },
    {
      "epoch": 1.6122407756531107,
      "grad_norm": 0.6209592504922359,
      "learning_rate": 8.994598206795068e-07,
      "loss": 0.30269927978515626,
      "num_input_tokens_seen": 1232086088,
      "step": 23945,
      "train_runtime": 147330.777,
      "train_tokens_per_second": 8362.72
    },
    {
      "epoch": 1.612577430649071,
      "grad_norm": 0.6538549095443367,
      "learning_rate": 8.979474272583672e-07,
      "loss": 0.2936129331588745,
      "num_input_tokens_seen": 1232325232,
      "step": 23950,
      "train_runtime": 147362.4322,
      "train_tokens_per_second": 8362.547
    },
    {
      "epoch": 1.612914085645031,
      "grad_norm": 0.6782538346929446,
      "learning_rate": 8.964361809631517e-07,
      "loss": 0.3155693531036377,
      "num_input_tokens_seen": 1232600288,
      "step": 23955,
      "train_runtime": 147393.652,
      "train_tokens_per_second": 8362.642
    },
    {
      "epoch": 1.613250740640991,
      "grad_norm": 0.6848652093508729,
      "learning_rate": 8.949260822164707e-07,
      "loss": 0.3084084987640381,
      "num_input_tokens_seen": 1232871376,
      "step": 23960,
      "train_runtime": 147424.2909,
      "train_tokens_per_second": 8362.742
    },
    {
      "epoch": 1.6135873956369513,
      "grad_norm": 0.6277998924510451,
      "learning_rate": 8.934171314406226e-07,
      "loss": 0.29707396030426025,
      "num_input_tokens_seen": 1233146192,
      "step": 23965,
      "train_runtime": 147455.2866,
      "train_tokens_per_second": 8362.848
    },
    {
      "epoch": 1.6139240506329116,
      "grad_norm": 0.7466883429634857,
      "learning_rate": 8.919093290575765e-07,
      "loss": 0.29536330699920654,
      "num_input_tokens_seen": 1233381464,
      "step": 23970,
      "train_runtime": 147486.852,
      "train_tokens_per_second": 8362.654
    },
    {
      "epoch": 1.6142607056288716,
      "grad_norm": 0.7750805061531985,
      "learning_rate": 8.904026754889877e-07,
      "loss": 0.30999696254730225,
      "num_input_tokens_seen": 1233620784,
      "step": 23975,
      "train_runtime": 147517.4372,
      "train_tokens_per_second": 8362.542
    },
    {
      "epoch": 1.6145973606248316,
      "grad_norm": 0.7267991824633424,
      "learning_rate": 8.888971711561867e-07,
      "loss": 0.31446199417114257,
      "num_input_tokens_seen": 1233860928,
      "step": 23980,
      "train_runtime": 147548.0621,
      "train_tokens_per_second": 8362.434
    },
    {
      "epoch": 1.6149340156207919,
      "grad_norm": 0.7328197312686097,
      "learning_rate": 8.873928164801821e-07,
      "loss": 0.33335831165313723,
      "num_input_tokens_seen": 1234126480,
      "step": 23985,
      "train_runtime": 147578.7869,
      "train_tokens_per_second": 8362.492
    },
    {
      "epoch": 1.6152706706167521,
      "grad_norm": 0.6663101561225745,
      "learning_rate": 8.858896118816624e-07,
      "loss": 0.2909775972366333,
      "num_input_tokens_seen": 1234333992,
      "step": 23990,
      "train_runtime": 147610.4313,
      "train_tokens_per_second": 8362.105
    },
    {
      "epoch": 1.6156073256127121,
      "grad_norm": 0.6412106421557306,
      "learning_rate": 8.84387557780994e-07,
      "loss": 0.27634515762329104,
      "num_input_tokens_seen": 1234570688,
      "step": 23995,
      "train_runtime": 147643.1462,
      "train_tokens_per_second": 8361.856
    },
    {
      "epoch": 1.6159439806086722,
      "grad_norm": 0.6432797818286116,
      "learning_rate": 8.828866545982196e-07,
      "loss": 0.28171381950378416,
      "num_input_tokens_seen": 1234826744,
      "step": 24000,
      "train_runtime": 147674.845,
      "train_tokens_per_second": 8361.795
    },
    {
      "epoch": 1.6162806356046324,
      "grad_norm": 0.6198185740369195,
      "learning_rate": 8.813869027530664e-07,
      "loss": 0.29250049591064453,
      "num_input_tokens_seen": 1235078312,
      "step": 24005,
      "train_runtime": 147708.5707,
      "train_tokens_per_second": 8361.589
    },
    {
      "epoch": 1.6166172906005927,
      "grad_norm": 0.600600310211218,
      "learning_rate": 8.798883026649302e-07,
      "loss": 0.30993027687072755,
      "num_input_tokens_seen": 1235364912,
      "step": 24010,
      "train_runtime": 147744.2688,
      "train_tokens_per_second": 8361.508
    },
    {
      "epoch": 1.6169539455965527,
      "grad_norm": 0.5986496987589849,
      "learning_rate": 8.783908547528947e-07,
      "loss": 0.26591441631317136,
      "num_input_tokens_seen": 1235613664,
      "step": 24015,
      "train_runtime": 147778.8945,
      "train_tokens_per_second": 8361.232
    },
    {
      "epoch": 1.6172906005925127,
      "grad_norm": 0.633691718741667,
      "learning_rate": 8.768945594357109e-07,
      "loss": 0.29343485832214355,
      "num_input_tokens_seen": 1235866968,
      "step": 24020,
      "train_runtime": 147807.264,
      "train_tokens_per_second": 8361.341
    },
    {
      "epoch": 1.617627255588473,
      "grad_norm": 0.7708757062625551,
      "learning_rate": 8.753994171318175e-07,
      "loss": 0.3216904878616333,
      "num_input_tokens_seen": 1236106696,
      "step": 24025,
      "train_runtime": 147837.5268,
      "train_tokens_per_second": 8361.251
    },
    {
      "epoch": 1.6179639105844332,
      "grad_norm": 0.6306482144171498,
      "learning_rate": 8.739054282593245e-07,
      "loss": 0.30921273231506347,
      "num_input_tokens_seen": 1236375064,
      "step": 24030,
      "train_runtime": 147872.0186,
      "train_tokens_per_second": 8361.116
    },
    {
      "epoch": 1.6183005655803933,
      "grad_norm": 0.7298195677785214,
      "learning_rate": 8.724125932360217e-07,
      "loss": 0.31004810333251953,
      "num_input_tokens_seen": 1236627904,
      "step": 24035,
      "train_runtime": 147904.0681,
      "train_tokens_per_second": 8361.013
    },
    {
      "epoch": 1.6186372205763533,
      "grad_norm": 0.7026834494049439,
      "learning_rate": 8.70920912479375e-07,
      "loss": 0.31732585430145266,
      "num_input_tokens_seen": 1236881336,
      "step": 24040,
      "train_runtime": 147936.0765,
      "train_tokens_per_second": 8360.918
    },
    {
      "epoch": 1.6189738755723135,
      "grad_norm": 0.6694856205298427,
      "learning_rate": 8.694303864065285e-07,
      "loss": 0.2913722038269043,
      "num_input_tokens_seen": 1237155640,
      "step": 24045,
      "train_runtime": 147970.5741,
      "train_tokens_per_second": 8360.822
    },
    {
      "epoch": 1.6193105305682738,
      "grad_norm": 0.5723045295637822,
      "learning_rate": 8.679410154343015e-07,
      "loss": 0.2872626304626465,
      "num_input_tokens_seen": 1237418952,
      "step": 24050,
      "train_runtime": 148002.5185,
      "train_tokens_per_second": 8360.797
    },
    {
      "epoch": 1.6196471855642338,
      "grad_norm": 0.7673890718497243,
      "learning_rate": 8.664527999791961e-07,
      "loss": 0.32081654071807864,
      "num_input_tokens_seen": 1237678600,
      "step": 24055,
      "train_runtime": 148030.3192,
      "train_tokens_per_second": 8360.98
    },
    {
      "epoch": 1.6199838405601938,
      "grad_norm": 0.6944391540607289,
      "learning_rate": 8.649657404573819e-07,
      "loss": 0.2948997259140015,
      "num_input_tokens_seen": 1237946752,
      "step": 24060,
      "train_runtime": 148063.6647,
      "train_tokens_per_second": 8360.908
    },
    {
      "epoch": 1.620320495556154,
      "grad_norm": 0.7930468418655989,
      "learning_rate": 8.634798372847148e-07,
      "loss": 0.31767778396606444,
      "num_input_tokens_seen": 1238222576,
      "step": 24065,
      "train_runtime": 148096.1918,
      "train_tokens_per_second": 8360.935
    },
    {
      "epoch": 1.6206571505521143,
      "grad_norm": 0.7113586997944203,
      "learning_rate": 8.619950908767189e-07,
      "loss": 0.2730276107788086,
      "num_input_tokens_seen": 1238442728,
      "step": 24070,
      "train_runtime": 148126.3738,
      "train_tokens_per_second": 8360.717
    },
    {
      "epoch": 1.6209938055480744,
      "grad_norm": 0.5853673179727644,
      "learning_rate": 8.605115016486016e-07,
      "loss": 0.2884242534637451,
      "num_input_tokens_seen": 1238718728,
      "step": 24075,
      "train_runtime": 148161.0737,
      "train_tokens_per_second": 8360.622
    },
    {
      "epoch": 1.6213304605440344,
      "grad_norm": 0.770156076960694,
      "learning_rate": 8.590290700152426e-07,
      "loss": 0.28478455543518066,
      "num_input_tokens_seen": 1238966664,
      "step": 24080,
      "train_runtime": 148189.5737,
      "train_tokens_per_second": 8360.687
    },
    {
      "epoch": 1.6216671155399947,
      "grad_norm": 0.7080851072577723,
      "learning_rate": 8.575477963912005e-07,
      "loss": 0.2978858709335327,
      "num_input_tokens_seen": 1239230432,
      "step": 24085,
      "train_runtime": 148222.8915,
      "train_tokens_per_second": 8360.587
    },
    {
      "epoch": 1.622003770535955,
      "grad_norm": 0.8270168338808922,
      "learning_rate": 8.560676811907071e-07,
      "loss": 0.31110224723815916,
      "num_input_tokens_seen": 1239482288,
      "step": 24090,
      "train_runtime": 148251.1403,
      "train_tokens_per_second": 8360.693
    },
    {
      "epoch": 1.622340425531915,
      "grad_norm": 0.7551708176117605,
      "learning_rate": 8.54588724827673e-07,
      "loss": 0.28424944877624514,
      "num_input_tokens_seen": 1239703240,
      "step": 24095,
      "train_runtime": 148278.2276,
      "train_tokens_per_second": 8360.656
    },
    {
      "epoch": 1.622677080527875,
      "grad_norm": 0.6411194882317476,
      "learning_rate": 8.531109277156818e-07,
      "loss": 0.30069005489349365,
      "num_input_tokens_seen": 1239989800,
      "step": 24100,
      "train_runtime": 148309.9799,
      "train_tokens_per_second": 8360.798
    },
    {
      "epoch": 1.6230137355238352,
      "grad_norm": 0.5401260592072046,
      "learning_rate": 8.516342902679986e-07,
      "loss": 0.26456236839294434,
      "num_input_tokens_seen": 1240212320,
      "step": 24105,
      "train_runtime": 148340.7189,
      "train_tokens_per_second": 8360.566
    },
    {
      "epoch": 1.6233503905197955,
      "grad_norm": 0.6814378801459691,
      "learning_rate": 8.501588128975557e-07,
      "loss": 0.270048999786377,
      "num_input_tokens_seen": 1240475496,
      "step": 24110,
      "train_runtime": 148370.9373,
      "train_tokens_per_second": 8360.637
    },
    {
      "epoch": 1.6236870455157555,
      "grad_norm": 0.6221906608223347,
      "learning_rate": 8.486844960169704e-07,
      "loss": 0.2822227716445923,
      "num_input_tokens_seen": 1240725584,
      "step": 24115,
      "train_runtime": 148401.6602,
      "train_tokens_per_second": 8360.591
    },
    {
      "epoch": 1.6240237005117155,
      "grad_norm": 0.6693991401315648,
      "learning_rate": 8.472113400385257e-07,
      "loss": 0.28075566291809084,
      "num_input_tokens_seen": 1240986736,
      "step": 24120,
      "train_runtime": 148435.0241,
      "train_tokens_per_second": 8360.471
    },
    {
      "epoch": 1.6243603555076758,
      "grad_norm": 0.6625937045955483,
      "learning_rate": 8.457393453741886e-07,
      "loss": 0.30159735679626465,
      "num_input_tokens_seen": 1241227128,
      "step": 24125,
      "train_runtime": 148465.441,
      "train_tokens_per_second": 8360.377
    },
    {
      "epoch": 1.624697010503636,
      "grad_norm": 0.623257449308574,
      "learning_rate": 8.442685124355965e-07,
      "loss": 0.2810337543487549,
      "num_input_tokens_seen": 1241476016,
      "step": 24130,
      "train_runtime": 148498.2315,
      "train_tokens_per_second": 8360.207
    },
    {
      "epoch": 1.625033665499596,
      "grad_norm": 0.703311402728029,
      "learning_rate": 8.427988416340633e-07,
      "loss": 0.31512699127197263,
      "num_input_tokens_seen": 1241705304,
      "step": 24135,
      "train_runtime": 148532.9353,
      "train_tokens_per_second": 8359.798
    },
    {
      "epoch": 1.625370320495556,
      "grad_norm": 0.7568147507409431,
      "learning_rate": 8.413303333805777e-07,
      "loss": 0.2885798454284668,
      "num_input_tokens_seen": 1241950528,
      "step": 24140,
      "train_runtime": 148561.2648,
      "train_tokens_per_second": 8359.854
    },
    {
      "epoch": 1.6257069754915163,
      "grad_norm": 0.686443364123146,
      "learning_rate": 8.398629880858034e-07,
      "loss": 0.30902719497680664,
      "num_input_tokens_seen": 1242246288,
      "step": 24145,
      "train_runtime": 148596.2473,
      "train_tokens_per_second": 8359.877
    },
    {
      "epoch": 1.6260436304874766,
      "grad_norm": 0.6132069371552243,
      "learning_rate": 8.383968061600778e-07,
      "loss": 0.3079632043838501,
      "num_input_tokens_seen": 1242485880,
      "step": 24150,
      "train_runtime": 148630.8262,
      "train_tokens_per_second": 8359.544
    },
    {
      "epoch": 1.6263802854834366,
      "grad_norm": 0.5927851399308158,
      "learning_rate": 8.36931788013417e-07,
      "loss": 0.2837127923965454,
      "num_input_tokens_seen": 1242764264,
      "step": 24155,
      "train_runtime": 148660.7688,
      "train_tokens_per_second": 8359.733
    },
    {
      "epoch": 1.6267169404793966,
      "grad_norm": 0.6980783894403393,
      "learning_rate": 8.354679340555044e-07,
      "loss": 0.29796566963195803,
      "num_input_tokens_seen": 1243015280,
      "step": 24160,
      "train_runtime": 148691.8902,
      "train_tokens_per_second": 8359.671
    },
    {
      "epoch": 1.6270535954753569,
      "grad_norm": 0.6265215783905168,
      "learning_rate": 8.34005244695707e-07,
      "loss": 0.288420295715332,
      "num_input_tokens_seen": 1243263656,
      "step": 24165,
      "train_runtime": 148723.9122,
      "train_tokens_per_second": 8359.541
    },
    {
      "epoch": 1.6273902504713171,
      "grad_norm": 0.5925665730774862,
      "learning_rate": 8.325437203430558e-07,
      "loss": 0.2981341123580933,
      "num_input_tokens_seen": 1243513920,
      "step": 24170,
      "train_runtime": 148754.4878,
      "train_tokens_per_second": 8359.505
    },
    {
      "epoch": 1.6277269054672772,
      "grad_norm": 0.6174789408187711,
      "learning_rate": 8.310833614062652e-07,
      "loss": 0.29852986335754395,
      "num_input_tokens_seen": 1243762224,
      "step": 24175,
      "train_runtime": 148782.8109,
      "train_tokens_per_second": 8359.583
    },
    {
      "epoch": 1.6280635604632372,
      "grad_norm": 0.7718118288037109,
      "learning_rate": 8.296241682937189e-07,
      "loss": 0.3043161392211914,
      "num_input_tokens_seen": 1244022920,
      "step": 24180,
      "train_runtime": 148815.3554,
      "train_tokens_per_second": 8359.506
    },
    {
      "epoch": 1.6284002154591974,
      "grad_norm": 0.5749711150139978,
      "learning_rate": 8.281661414134761e-07,
      "loss": 0.29904799461364745,
      "num_input_tokens_seen": 1244273016,
      "step": 24185,
      "train_runtime": 148846.1269,
      "train_tokens_per_second": 8359.458
    },
    {
      "epoch": 1.6287368704551577,
      "grad_norm": 0.6859323356274802,
      "learning_rate": 8.267092811732686e-07,
      "loss": 0.30904836654663087,
      "num_input_tokens_seen": 1244534480,
      "step": 24190,
      "train_runtime": 148876.2474,
      "train_tokens_per_second": 8359.523
    },
    {
      "epoch": 1.6290735254511177,
      "grad_norm": 0.6411932034556799,
      "learning_rate": 8.252535879805029e-07,
      "loss": 0.28147592544555666,
      "num_input_tokens_seen": 1244793736,
      "step": 24195,
      "train_runtime": 148912.8068,
      "train_tokens_per_second": 8359.212
    },
    {
      "epoch": 1.6294101804470777,
      "grad_norm": 0.6821630725009705,
      "learning_rate": 8.237990622422576e-07,
      "loss": 0.2901399374008179,
      "num_input_tokens_seen": 1245055296,
      "step": 24200,
      "train_runtime": 148939.4519,
      "train_tokens_per_second": 8359.473
    },
    {
      "epoch": 1.629746835443038,
      "grad_norm": 0.6090322676438463,
      "learning_rate": 8.223457043652905e-07,
      "loss": 0.3023641347885132,
      "num_input_tokens_seen": 1245323536,
      "step": 24205,
      "train_runtime": 148973.3499,
      "train_tokens_per_second": 8359.371
    },
    {
      "epoch": 1.6300834904389983,
      "grad_norm": 0.8265689164030401,
      "learning_rate": 8.208935147560227e-07,
      "loss": 0.2932861804962158,
      "num_input_tokens_seen": 1245556536,
      "step": 24210,
      "train_runtime": 149007.7909,
      "train_tokens_per_second": 8359.003
    },
    {
      "epoch": 1.6304201454349583,
      "grad_norm": 0.648114347998774,
      "learning_rate": 8.194424938205597e-07,
      "loss": 0.3295659065246582,
      "num_input_tokens_seen": 1245840296,
      "step": 24215,
      "train_runtime": 149040.1045,
      "train_tokens_per_second": 8359.094
    },
    {
      "epoch": 1.6307568004309183,
      "grad_norm": 0.6982012117871018,
      "learning_rate": 8.179926419646705e-07,
      "loss": 0.2803459882736206,
      "num_input_tokens_seen": 1246087184,
      "step": 24220,
      "train_runtime": 149069.3233,
      "train_tokens_per_second": 8359.112
    },
    {
      "epoch": 1.6310934554268786,
      "grad_norm": 0.7083952080864633,
      "learning_rate": 8.165439595938047e-07,
      "loss": 0.31992335319519044,
      "num_input_tokens_seen": 1246310816,
      "step": 24225,
      "train_runtime": 149097.7327,
      "train_tokens_per_second": 8359.019
    },
    {
      "epoch": 1.6314301104228388,
      "grad_norm": 0.7073647017337159,
      "learning_rate": 8.150964471130801e-07,
      "loss": 0.28247556686401365,
      "num_input_tokens_seen": 1246589024,
      "step": 24230,
      "train_runtime": 149129.4667,
      "train_tokens_per_second": 8359.106
    },
    {
      "epoch": 1.6317667654187988,
      "grad_norm": 0.6358312473279518,
      "learning_rate": 8.136501049272899e-07,
      "loss": 0.2945717334747314,
      "num_input_tokens_seen": 1246840848,
      "step": 24235,
      "train_runtime": 149161.6007,
      "train_tokens_per_second": 8358.993
    },
    {
      "epoch": 1.6321034204147589,
      "grad_norm": 0.7109296670702442,
      "learning_rate": 8.122049334408983e-07,
      "loss": 0.29415187835693357,
      "num_input_tokens_seen": 1247085624,
      "step": 24240,
      "train_runtime": 149189.6603,
      "train_tokens_per_second": 8359.062
    },
    {
      "epoch": 1.6324400754107191,
      "grad_norm": 0.8871208553851592,
      "learning_rate": 8.107609330580429e-07,
      "loss": 0.2898645639419556,
      "num_input_tokens_seen": 1247332040,
      "step": 24245,
      "train_runtime": 149220.0115,
      "train_tokens_per_second": 8359.013
    },
    {
      "epoch": 1.6327767304066794,
      "grad_norm": 0.6484583593090408,
      "learning_rate": 8.093181041825332e-07,
      "loss": 0.30123889446258545,
      "num_input_tokens_seen": 1247601576,
      "step": 24250,
      "train_runtime": 149255.5368,
      "train_tokens_per_second": 8358.829
    },
    {
      "epoch": 1.6331133854026394,
      "grad_norm": 0.6935591801879868,
      "learning_rate": 8.078764472178552e-07,
      "loss": 0.29544141292572024,
      "num_input_tokens_seen": 1247847784,
      "step": 24255,
      "train_runtime": 149287.5092,
      "train_tokens_per_second": 8358.688
    },
    {
      "epoch": 1.6334500403985994,
      "grad_norm": 0.6075351533580892,
      "learning_rate": 8.06435962567158e-07,
      "loss": 0.327945613861084,
      "num_input_tokens_seen": 1248069528,
      "step": 24260,
      "train_runtime": 149321.2357,
      "train_tokens_per_second": 8358.286
    },
    {
      "epoch": 1.6337866953945597,
      "grad_norm": 0.6685047728832834,
      "learning_rate": 8.049966506332746e-07,
      "loss": 0.25960488319396974,
      "num_input_tokens_seen": 1248330288,
      "step": 24265,
      "train_runtime": 149350.7506,
      "train_tokens_per_second": 8358.38
    },
    {
      "epoch": 1.63412335039052,
      "grad_norm": 0.6469818700365366,
      "learning_rate": 8.035585118186984e-07,
      "loss": 0.2763599157333374,
      "num_input_tokens_seen": 1248574624,
      "step": 24270,
      "train_runtime": 149381.3648,
      "train_tokens_per_second": 8358.302
    },
    {
      "epoch": 1.63446000538648,
      "grad_norm": 0.6577350482119343,
      "learning_rate": 8.021215465256038e-07,
      "loss": 0.26905217170715334,
      "num_input_tokens_seen": 1248828616,
      "step": 24275,
      "train_runtime": 149410.8702,
      "train_tokens_per_second": 8358.352
    },
    {
      "epoch": 1.63479666038244,
      "grad_norm": 0.6734925693846262,
      "learning_rate": 8.006857551558328e-07,
      "loss": 0.2990940809249878,
      "num_input_tokens_seen": 1249072704,
      "step": 24280,
      "train_runtime": 149443.3363,
      "train_tokens_per_second": 8358.169
    },
    {
      "epoch": 1.6351333153784002,
      "grad_norm": 0.6282880589905192,
      "learning_rate": 7.992511381108997e-07,
      "loss": 0.3036353588104248,
      "num_input_tokens_seen": 1249370248,
      "step": 24285,
      "train_runtime": 149475.5576,
      "train_tokens_per_second": 8358.358
    },
    {
      "epoch": 1.6354699703743605,
      "grad_norm": 0.6414262043134169,
      "learning_rate": 7.978176957919909e-07,
      "loss": 0.2923716068267822,
      "num_input_tokens_seen": 1249614360,
      "step": 24290,
      "train_runtime": 149505.5323,
      "train_tokens_per_second": 8358.315
    },
    {
      "epoch": 1.6358066253703205,
      "grad_norm": 0.5868220189272834,
      "learning_rate": 7.963854285999634e-07,
      "loss": 0.2752955675125122,
      "num_input_tokens_seen": 1249868312,
      "step": 24295,
      "train_runtime": 149536.8898,
      "train_tokens_per_second": 8358.261
    },
    {
      "epoch": 1.6361432803662805,
      "grad_norm": 0.8129773527959091,
      "learning_rate": 7.949543369353452e-07,
      "loss": 0.29357328414916994,
      "num_input_tokens_seen": 1250092816,
      "step": 24300,
      "train_runtime": 149567.0559,
      "train_tokens_per_second": 8358.076
    },
    {
      "epoch": 1.6364799353622408,
      "grad_norm": 0.6530519314552922,
      "learning_rate": 7.9352442119834e-07,
      "loss": 0.3173997402191162,
      "num_input_tokens_seen": 1250337680,
      "step": 24305,
      "train_runtime": 149595.8598,
      "train_tokens_per_second": 8358.104
    },
    {
      "epoch": 1.636816590358201,
      "grad_norm": 0.664981931445702,
      "learning_rate": 7.920956817888148e-07,
      "loss": 0.3024116277694702,
      "num_input_tokens_seen": 1250598056,
      "step": 24310,
      "train_runtime": 149626.6423,
      "train_tokens_per_second": 8358.124
    },
    {
      "epoch": 1.637153245354161,
      "grad_norm": 0.8058707615510541,
      "learning_rate": 7.906681191063165e-07,
      "loss": 0.2993587017059326,
      "num_input_tokens_seen": 1250866440,
      "step": 24315,
      "train_runtime": 149655.1197,
      "train_tokens_per_second": 8358.327
    },
    {
      "epoch": 1.637489900350121,
      "grad_norm": 0.7155240670205786,
      "learning_rate": 7.892417335500541e-07,
      "loss": 0.3101596117019653,
      "num_input_tokens_seen": 1251117024,
      "step": 24320,
      "train_runtime": 149684.9817,
      "train_tokens_per_second": 8358.334
    },
    {
      "epoch": 1.6378265553460813,
      "grad_norm": 0.7450487563305583,
      "learning_rate": 7.878165255189146e-07,
      "loss": 0.3204202175140381,
      "num_input_tokens_seen": 1251354704,
      "step": 24325,
      "train_runtime": 149719.2931,
      "train_tokens_per_second": 8358.006
    },
    {
      "epoch": 1.6381632103420416,
      "grad_norm": 0.8193222271188244,
      "learning_rate": 7.863924954114522e-07,
      "loss": 0.28904759883880615,
      "num_input_tokens_seen": 1251597008,
      "step": 24330,
      "train_runtime": 149748.5381,
      "train_tokens_per_second": 8357.991
    },
    {
      "epoch": 1.6384998653380016,
      "grad_norm": 0.5758213253817959,
      "learning_rate": 7.849696436258919e-07,
      "loss": 0.2803504467010498,
      "num_input_tokens_seen": 1251844944,
      "step": 24335,
      "train_runtime": 149776.643,
      "train_tokens_per_second": 8358.079
    },
    {
      "epoch": 1.6388365203339617,
      "grad_norm": 0.6269059878517124,
      "learning_rate": 7.835479705601307e-07,
      "loss": 0.3026492357254028,
      "num_input_tokens_seen": 1252086824,
      "step": 24340,
      "train_runtime": 149807.2038,
      "train_tokens_per_second": 8357.988
    },
    {
      "epoch": 1.639173175329922,
      "grad_norm": 0.5662339213424912,
      "learning_rate": 7.821274766117337e-07,
      "loss": 0.299318790435791,
      "num_input_tokens_seen": 1252335720,
      "step": 24345,
      "train_runtime": 149835.9139,
      "train_tokens_per_second": 8358.048
    },
    {
      "epoch": 1.6395098303258822,
      "grad_norm": 0.7164572155565451,
      "learning_rate": 7.807081621779367e-07,
      "loss": 0.313852858543396,
      "num_input_tokens_seen": 1252600856,
      "step": 24350,
      "train_runtime": 149870.0131,
      "train_tokens_per_second": 8357.915
    },
    {
      "epoch": 1.6398464853218422,
      "grad_norm": 0.5828616124838266,
      "learning_rate": 7.792900276556509e-07,
      "loss": 0.30215229988098147,
      "num_input_tokens_seen": 1252868840,
      "step": 24355,
      "train_runtime": 149900.1659,
      "train_tokens_per_second": 8358.022
    },
    {
      "epoch": 1.6401831403178022,
      "grad_norm": 0.6465271266822389,
      "learning_rate": 7.778730734414469e-07,
      "loss": 0.3042720317840576,
      "num_input_tokens_seen": 1253119824,
      "step": 24360,
      "train_runtime": 149934.6862,
      "train_tokens_per_second": 8357.771
    },
    {
      "epoch": 1.6405197953137625,
      "grad_norm": 0.8192141167011209,
      "learning_rate": 7.764572999315772e-07,
      "loss": 0.29519200325012207,
      "num_input_tokens_seen": 1253389136,
      "step": 24365,
      "train_runtime": 149966.7371,
      "train_tokens_per_second": 8357.781
    },
    {
      "epoch": 1.6408564503097227,
      "grad_norm": 0.6847851242114603,
      "learning_rate": 7.750427075219536e-07,
      "loss": 0.30560519695281985,
      "num_input_tokens_seen": 1253622840,
      "step": 24370,
      "train_runtime": 149999.5027,
      "train_tokens_per_second": 8357.513
    },
    {
      "epoch": 1.6411931053056827,
      "grad_norm": 0.7640291513433882,
      "learning_rate": 7.736292966081655e-07,
      "loss": 0.2901760101318359,
      "num_input_tokens_seen": 1253846096,
      "step": 24375,
      "train_runtime": 150028.3782,
      "train_tokens_per_second": 8357.393
    },
    {
      "epoch": 1.6415297603016428,
      "grad_norm": 0.577920084801074,
      "learning_rate": 7.722170675854673e-07,
      "loss": 0.266463041305542,
      "num_input_tokens_seen": 1254111800,
      "step": 24380,
      "train_runtime": 150063.5784,
      "train_tokens_per_second": 8357.203
    },
    {
      "epoch": 1.641866415297603,
      "grad_norm": 0.6428575111159175,
      "learning_rate": 7.70806020848785e-07,
      "loss": 0.2797550678253174,
      "num_input_tokens_seen": 1254382760,
      "step": 24385,
      "train_runtime": 150092.3391,
      "train_tokens_per_second": 8357.407
    },
    {
      "epoch": 1.6422030702935633,
      "grad_norm": 0.6470982188763604,
      "learning_rate": 7.693961567927133e-07,
      "loss": 0.27973299026489257,
      "num_input_tokens_seen": 1254647544,
      "step": 24390,
      "train_runtime": 150123.4839,
      "train_tokens_per_second": 8357.437
    },
    {
      "epoch": 1.6425397252895233,
      "grad_norm": 0.6057733459798235,
      "learning_rate": 7.679874758115153e-07,
      "loss": 0.278795862197876,
      "num_input_tokens_seen": 1254917776,
      "step": 24395,
      "train_runtime": 150159.661,
      "train_tokens_per_second": 8357.223
    },
    {
      "epoch": 1.6428763802854833,
      "grad_norm": 0.6318874762557986,
      "learning_rate": 7.66579978299124e-07,
      "loss": 0.30783424377441404,
      "num_input_tokens_seen": 1255191032,
      "step": 24400,
      "train_runtime": 150187.4263,
      "train_tokens_per_second": 8357.497
    },
    {
      "epoch": 1.6432130352814436,
      "grad_norm": 0.7262228042418619,
      "learning_rate": 7.651736646491447e-07,
      "loss": 0.31876723766326903,
      "num_input_tokens_seen": 1255428632,
      "step": 24405,
      "train_runtime": 150216.8507,
      "train_tokens_per_second": 8357.442
    },
    {
      "epoch": 1.6435496902774038,
      "grad_norm": 0.5683289655771215,
      "learning_rate": 7.63768535254844e-07,
      "loss": 0.27198209762573244,
      "num_input_tokens_seen": 1255685544,
      "step": 24410,
      "train_runtime": 150250.5353,
      "train_tokens_per_second": 8357.278
    },
    {
      "epoch": 1.6438863452733639,
      "grad_norm": 0.7328704897010073,
      "learning_rate": 7.623645905091664e-07,
      "loss": 0.30343379974365237,
      "num_input_tokens_seen": 1255942104,
      "step": 24415,
      "train_runtime": 150279.6753,
      "train_tokens_per_second": 8357.365
    },
    {
      "epoch": 1.6442230002693239,
      "grad_norm": 0.7074989043414779,
      "learning_rate": 7.609618308047156e-07,
      "loss": 0.29104299545288087,
      "num_input_tokens_seen": 1256213984,
      "step": 24420,
      "train_runtime": 150309.8278,
      "train_tokens_per_second": 8357.497
    },
    {
      "epoch": 1.6445596552652841,
      "grad_norm": 0.6442770892331537,
      "learning_rate": 7.595602565337729e-07,
      "loss": 0.3005190134048462,
      "num_input_tokens_seen": 1256473440,
      "step": 24425,
      "train_runtime": 150341.9082,
      "train_tokens_per_second": 8357.44
    },
    {
      "epoch": 1.6448963102612444,
      "grad_norm": 0.7038205307072376,
      "learning_rate": 7.581598680882829e-07,
      "loss": 0.3200518131256104,
      "num_input_tokens_seen": 1256735992,
      "step": 24430,
      "train_runtime": 150374.4388,
      "train_tokens_per_second": 8357.378
    },
    {
      "epoch": 1.6452329652572044,
      "grad_norm": 0.7484611298703149,
      "learning_rate": 7.567606658598592e-07,
      "loss": 0.2992230415344238,
      "num_input_tokens_seen": 1256984952,
      "step": 24435,
      "train_runtime": 150404.7665,
      "train_tokens_per_second": 8357.348
    },
    {
      "epoch": 1.6455696202531644,
      "grad_norm": 0.6460482148836196,
      "learning_rate": 7.553626502397848e-07,
      "loss": 0.31764955520629884,
      "num_input_tokens_seen": 1257247264,
      "step": 24440,
      "train_runtime": 150433.8721,
      "train_tokens_per_second": 8357.475
    },
    {
      "epoch": 1.6459062752491247,
      "grad_norm": 0.641541624566756,
      "learning_rate": 7.539658216190099e-07,
      "loss": 0.27633473873138426,
      "num_input_tokens_seen": 1257507864,
      "step": 24445,
      "train_runtime": 150463.8797,
      "train_tokens_per_second": 8357.54
    },
    {
      "epoch": 1.646242930245085,
      "grad_norm": 0.7496081054124683,
      "learning_rate": 7.525701803881525e-07,
      "loss": 0.30477280616760255,
      "num_input_tokens_seen": 1257755608,
      "step": 24450,
      "train_runtime": 150496.8114,
      "train_tokens_per_second": 8357.357
    },
    {
      "epoch": 1.646579585241045,
      "grad_norm": 0.66987989051448,
      "learning_rate": 7.511757269375019e-07,
      "loss": 0.2896285057067871,
      "num_input_tokens_seen": 1257985704,
      "step": 24455,
      "train_runtime": 150531.4893,
      "train_tokens_per_second": 8356.96
    },
    {
      "epoch": 1.646916240237005,
      "grad_norm": 0.6592016139009291,
      "learning_rate": 7.497824616570087e-07,
      "loss": 0.31701107025146485,
      "num_input_tokens_seen": 1258218536,
      "step": 24460,
      "train_runtime": 150561.167,
      "train_tokens_per_second": 8356.86
    },
    {
      "epoch": 1.6472528952329653,
      "grad_norm": 0.6506573283527837,
      "learning_rate": 7.483903849362995e-07,
      "loss": 0.3016084671020508,
      "num_input_tokens_seen": 1258480232,
      "step": 24465,
      "train_runtime": 150594.3988,
      "train_tokens_per_second": 8356.753
    },
    {
      "epoch": 1.6475895502289255,
      "grad_norm": 0.6592264983093828,
      "learning_rate": 7.469994971646594e-07,
      "loss": 0.30023980140686035,
      "num_input_tokens_seen": 1258750864,
      "step": 24470,
      "train_runtime": 150623.6293,
      "train_tokens_per_second": 8356.928
    },
    {
      "epoch": 1.6479262052248855,
      "grad_norm": 0.5889017580433618,
      "learning_rate": 7.456097987310495e-07,
      "loss": 0.29707822799682615,
      "num_input_tokens_seen": 1258996112,
      "step": 24475,
      "train_runtime": 150655.2205,
      "train_tokens_per_second": 8356.804
    },
    {
      "epoch": 1.6482628602208456,
      "grad_norm": 0.8155842574893346,
      "learning_rate": 7.442212900240931e-07,
      "loss": 0.31392340660095214,
      "num_input_tokens_seen": 1259246512,
      "step": 24480,
      "train_runtime": 150684.4466,
      "train_tokens_per_second": 8356.845
    },
    {
      "epoch": 1.6485995152168058,
      "grad_norm": 0.6267655272595841,
      "learning_rate": 7.42833971432082e-07,
      "loss": 0.26655728816986085,
      "num_input_tokens_seen": 1259500168,
      "step": 24485,
      "train_runtime": 150714.4275,
      "train_tokens_per_second": 8356.865
    },
    {
      "epoch": 1.648936170212766,
      "grad_norm": 0.6710817220578564,
      "learning_rate": 7.414478433429756e-07,
      "loss": 0.31193540096282957,
      "num_input_tokens_seen": 1259769176,
      "step": 24490,
      "train_runtime": 150741.1065,
      "train_tokens_per_second": 8357.171
    },
    {
      "epoch": 1.649272825208726,
      "grad_norm": 0.6383261897818037,
      "learning_rate": 7.400629061444003e-07,
      "loss": 0.2899024248123169,
      "num_input_tokens_seen": 1260011392,
      "step": 24495,
      "train_runtime": 150770.4319,
      "train_tokens_per_second": 8357.152
    },
    {
      "epoch": 1.6496094802046861,
      "grad_norm": 0.676005360249914,
      "learning_rate": 7.386791602236482e-07,
      "loss": 0.3224844694137573,
      "num_input_tokens_seen": 1260294024,
      "step": 24500,
      "train_runtime": 150802.1603,
      "train_tokens_per_second": 8357.268
    },
    {
      "epoch": 1.6499461352006464,
      "grad_norm": 0.6726244138790403,
      "learning_rate": 7.372966059676834e-07,
      "loss": 0.34412670135498047,
      "num_input_tokens_seen": 1260535192,
      "step": 24505,
      "train_runtime": 150833.1765,
      "train_tokens_per_second": 8357.148
    },
    {
      "epoch": 1.6502827901966066,
      "grad_norm": 0.6094798970821154,
      "learning_rate": 7.359152437631273e-07,
      "loss": 0.32269465923309326,
      "num_input_tokens_seen": 1260808296,
      "step": 24510,
      "train_runtime": 150865.7301,
      "train_tokens_per_second": 8357.155
    },
    {
      "epoch": 1.6506194451925666,
      "grad_norm": 0.7566517466952112,
      "learning_rate": 7.345350739962781e-07,
      "loss": 0.28818683624267577,
      "num_input_tokens_seen": 1261073656,
      "step": 24515,
      "train_runtime": 150898.4829,
      "train_tokens_per_second": 8357.1
    },
    {
      "epoch": 1.6509561001885267,
      "grad_norm": 0.6770486149029457,
      "learning_rate": 7.331560970530921e-07,
      "loss": 0.2973813772201538,
      "num_input_tokens_seen": 1261330832,
      "step": 24520,
      "train_runtime": 150930.5218,
      "train_tokens_per_second": 8357.03
    },
    {
      "epoch": 1.651292755184487,
      "grad_norm": 0.6082305211516752,
      "learning_rate": 7.317783133191986e-07,
      "loss": 0.2943634271621704,
      "num_input_tokens_seen": 1261606056,
      "step": 24525,
      "train_runtime": 150960.5703,
      "train_tokens_per_second": 8357.189
    },
    {
      "epoch": 1.6516294101804472,
      "grad_norm": 0.718488850191112,
      "learning_rate": 7.304017231798893e-07,
      "loss": 0.2836528539657593,
      "num_input_tokens_seen": 1261878624,
      "step": 24530,
      "train_runtime": 150990.95,
      "train_tokens_per_second": 8357.313
    },
    {
      "epoch": 1.6519660651764072,
      "grad_norm": 0.7782957847089798,
      "learning_rate": 7.290263270201231e-07,
      "loss": 0.2961276531219482,
      "num_input_tokens_seen": 1262137080,
      "step": 24535,
      "train_runtime": 151020.9665,
      "train_tokens_per_second": 8357.363
    },
    {
      "epoch": 1.6523027201723672,
      "grad_norm": 0.6702771995717126,
      "learning_rate": 7.276521252245261e-07,
      "loss": 0.2887396812438965,
      "num_input_tokens_seen": 1262390960,
      "step": 24540,
      "train_runtime": 151049.7556,
      "train_tokens_per_second": 8357.451
    },
    {
      "epoch": 1.6526393751683275,
      "grad_norm": 0.6429827730220521,
      "learning_rate": 7.262791181773882e-07,
      "loss": 0.2910745143890381,
      "num_input_tokens_seen": 1262656816,
      "step": 24545,
      "train_runtime": 151082.6635,
      "train_tokens_per_second": 8357.39
    },
    {
      "epoch": 1.6529760301642877,
      "grad_norm": 0.816091044232686,
      "learning_rate": 7.24907306262666e-07,
      "loss": 0.30447936058044434,
      "num_input_tokens_seen": 1262894704,
      "step": 24550,
      "train_runtime": 151110.3032,
      "train_tokens_per_second": 8357.436
    },
    {
      "epoch": 1.6533126851602478,
      "grad_norm": 0.669738593594749,
      "learning_rate": 7.235366898639856e-07,
      "loss": 0.31218686103820803,
      "num_input_tokens_seen": 1263149576,
      "step": 24555,
      "train_runtime": 151140.8532,
      "train_tokens_per_second": 8357.433
    },
    {
      "epoch": 1.6536493401562078,
      "grad_norm": 0.7189464372003882,
      "learning_rate": 7.221672693646309e-07,
      "loss": 0.31140248775482177,
      "num_input_tokens_seen": 1263416472,
      "step": 24560,
      "train_runtime": 151175.1638,
      "train_tokens_per_second": 8357.302
    },
    {
      "epoch": 1.653985995152168,
      "grad_norm": 0.5835034213227324,
      "learning_rate": 7.207990451475594e-07,
      "loss": 0.3014613151550293,
      "num_input_tokens_seen": 1263705848,
      "step": 24565,
      "train_runtime": 151208.2757,
      "train_tokens_per_second": 8357.385
    },
    {
      "epoch": 1.6543226501481283,
      "grad_norm": 0.7522066513567441,
      "learning_rate": 7.194320175953901e-07,
      "loss": 0.30106658935546876,
      "num_input_tokens_seen": 1263944808,
      "step": 24570,
      "train_runtime": 151239.3305,
      "train_tokens_per_second": 8357.249
    },
    {
      "epoch": 1.6546593051440883,
      "grad_norm": 0.6576814475229851,
      "learning_rate": 7.180661870904066e-07,
      "loss": 0.27959983348846434,
      "num_input_tokens_seen": 1264206984,
      "step": 24575,
      "train_runtime": 151268.5759,
      "train_tokens_per_second": 8357.367
    },
    {
      "epoch": 1.6549959601400483,
      "grad_norm": 0.6039091449747372,
      "learning_rate": 7.167015540145599e-07,
      "loss": 0.2811901569366455,
      "num_input_tokens_seen": 1264437504,
      "step": 24580,
      "train_runtime": 151295.7432,
      "train_tokens_per_second": 8357.39
    },
    {
      "epoch": 1.6553326151360086,
      "grad_norm": 0.725813746968183,
      "learning_rate": 7.153381187494657e-07,
      "loss": 0.3058518648147583,
      "num_input_tokens_seen": 1264682304,
      "step": 24585,
      "train_runtime": 151323.2575,
      "train_tokens_per_second": 8357.488
    },
    {
      "epoch": 1.6556692701319689,
      "grad_norm": 0.7188805774270236,
      "learning_rate": 7.139758816764036e-07,
      "loss": 0.2913860082626343,
      "num_input_tokens_seen": 1264922416,
      "step": 24590,
      "train_runtime": 151350.8443,
      "train_tokens_per_second": 8357.551
    },
    {
      "epoch": 1.6560059251279289,
      "grad_norm": 0.5670189547698657,
      "learning_rate": 7.126148431763191e-07,
      "loss": 0.27695236206054685,
      "num_input_tokens_seen": 1265198560,
      "step": 24595,
      "train_runtime": 151381.4978,
      "train_tokens_per_second": 8357.683
    },
    {
      "epoch": 1.656342580123889,
      "grad_norm": 0.6400671585457508,
      "learning_rate": 7.11255003629821e-07,
      "loss": 0.29895381927490233,
      "num_input_tokens_seen": 1265474544,
      "step": 24600,
      "train_runtime": 151413.5372,
      "train_tokens_per_second": 8357.737
    },
    {
      "epoch": 1.6566792351198492,
      "grad_norm": 0.5753697256339022,
      "learning_rate": 7.098963634171874e-07,
      "loss": 0.29350881576538085,
      "num_input_tokens_seen": 1265755328,
      "step": 24605,
      "train_runtime": 151446.8899,
      "train_tokens_per_second": 8357.751
    },
    {
      "epoch": 1.6570158901158094,
      "grad_norm": 0.8758205314077223,
      "learning_rate": 7.085389229183537e-07,
      "loss": 0.30609936714172364,
      "num_input_tokens_seen": 1265973864,
      "step": 24610,
      "train_runtime": 151473.9519,
      "train_tokens_per_second": 8357.7
    },
    {
      "epoch": 1.6573525451117694,
      "grad_norm": 0.7325169174615904,
      "learning_rate": 7.071826825129264e-07,
      "loss": 0.29387166500091555,
      "num_input_tokens_seen": 1266216272,
      "step": 24615,
      "train_runtime": 151509.5156,
      "train_tokens_per_second": 8357.338
    },
    {
      "epoch": 1.6576892001077295,
      "grad_norm": 0.6436266929079452,
      "learning_rate": 7.058276425801735e-07,
      "loss": 0.309558367729187,
      "num_input_tokens_seen": 1266479568,
      "step": 24620,
      "train_runtime": 151540.6783,
      "train_tokens_per_second": 8357.357
    },
    {
      "epoch": 1.6580258551036897,
      "grad_norm": 0.6356683728338197,
      "learning_rate": 7.044738034990273e-07,
      "loss": 0.3014700412750244,
      "num_input_tokens_seen": 1266725056,
      "step": 24625,
      "train_runtime": 151570.5036,
      "train_tokens_per_second": 8357.332
    },
    {
      "epoch": 1.65836251009965,
      "grad_norm": 0.71432476376305,
      "learning_rate": 7.031211656480841e-07,
      "loss": 0.31534428596496583,
      "num_input_tokens_seen": 1266978744,
      "step": 24630,
      "train_runtime": 151604.7459,
      "train_tokens_per_second": 8357.118
    },
    {
      "epoch": 1.65869916509561,
      "grad_norm": 0.6496281665154438,
      "learning_rate": 7.017697294056058e-07,
      "loss": 0.2696672916412354,
      "num_input_tokens_seen": 1267231584,
      "step": 24635,
      "train_runtime": 151634.6402,
      "train_tokens_per_second": 8357.138
    },
    {
      "epoch": 1.65903582009157,
      "grad_norm": 0.6448398944688153,
      "learning_rate": 7.004194951495163e-07,
      "loss": 0.3157979488372803,
      "num_input_tokens_seen": 1267513640,
      "step": 24640,
      "train_runtime": 151665.0715,
      "train_tokens_per_second": 8357.321
    },
    {
      "epoch": 1.6593724750875303,
      "grad_norm": 0.6682465556689301,
      "learning_rate": 6.990704632574046e-07,
      "loss": 0.29840543270111086,
      "num_input_tokens_seen": 1267783352,
      "step": 24645,
      "train_runtime": 151702.4083,
      "train_tokens_per_second": 8357.042
    },
    {
      "epoch": 1.6597091300834905,
      "grad_norm": 0.668636994864507,
      "learning_rate": 6.977226341065219e-07,
      "loss": 0.2871983051300049,
      "num_input_tokens_seen": 1268041128,
      "step": 24650,
      "train_runtime": 151730.5354,
      "train_tokens_per_second": 8357.191
    },
    {
      "epoch": 1.6600457850794506,
      "grad_norm": 0.6347886259242137,
      "learning_rate": 6.96376008073788e-07,
      "loss": 0.30836474895477295,
      "num_input_tokens_seen": 1268276656,
      "step": 24655,
      "train_runtime": 151764.5074,
      "train_tokens_per_second": 8356.873
    },
    {
      "epoch": 1.6603824400754106,
      "grad_norm": 0.5776195361319348,
      "learning_rate": 6.950305855357781e-07,
      "loss": 0.27713427543640134,
      "num_input_tokens_seen": 1268523976,
      "step": 24660,
      "train_runtime": 151795.2703,
      "train_tokens_per_second": 8356.808
    },
    {
      "epoch": 1.6607190950713708,
      "grad_norm": 0.6189013495582839,
      "learning_rate": 6.936863668687383e-07,
      "loss": 0.29589390754699707,
      "num_input_tokens_seen": 1268777968,
      "step": 24665,
      "train_runtime": 151825.9003,
      "train_tokens_per_second": 8356.795
    },
    {
      "epoch": 1.661055750067331,
      "grad_norm": 0.6804728574356337,
      "learning_rate": 6.923433524485745e-07,
      "loss": 0.3222536563873291,
      "num_input_tokens_seen": 1269039344,
      "step": 24670,
      "train_runtime": 151857.9284,
      "train_tokens_per_second": 8356.754
    },
    {
      "epoch": 1.6613924050632911,
      "grad_norm": 0.626660101372101,
      "learning_rate": 6.910015426508554e-07,
      "loss": 0.30066187381744386,
      "num_input_tokens_seen": 1269308032,
      "step": 24675,
      "train_runtime": 151890.7376,
      "train_tokens_per_second": 8356.718
    },
    {
      "epoch": 1.6617290600592511,
      "grad_norm": 0.6323332045834307,
      "learning_rate": 6.896609378508152e-07,
      "loss": 0.3063807010650635,
      "num_input_tokens_seen": 1269538280,
      "step": 24680,
      "train_runtime": 151919.4631,
      "train_tokens_per_second": 8356.653
    },
    {
      "epoch": 1.6620657150552114,
      "grad_norm": 0.5733149765152857,
      "learning_rate": 6.883215384233488e-07,
      "loss": 0.28978352546691893,
      "num_input_tokens_seen": 1269820384,
      "step": 24685,
      "train_runtime": 151953.3163,
      "train_tokens_per_second": 8356.648
    },
    {
      "epoch": 1.6624023700511716,
      "grad_norm": 0.7042255080586403,
      "learning_rate": 6.869833447430152e-07,
      "loss": 0.2829467296600342,
      "num_input_tokens_seen": 1270044096,
      "step": 24690,
      "train_runtime": 151983.5673,
      "train_tokens_per_second": 8356.457
    },
    {
      "epoch": 1.6627390250471317,
      "grad_norm": 0.6472566158265216,
      "learning_rate": 6.856463571840361e-07,
      "loss": 0.3014203071594238,
      "num_input_tokens_seen": 1270286816,
      "step": 24695,
      "train_runtime": 152018.1632,
      "train_tokens_per_second": 8356.152
    },
    {
      "epoch": 1.6630756800430917,
      "grad_norm": 0.6837747919242632,
      "learning_rate": 6.843105761202945e-07,
      "loss": 0.3120609521865845,
      "num_input_tokens_seen": 1270559072,
      "step": 24700,
      "train_runtime": 152048.517,
      "train_tokens_per_second": 8356.274
    },
    {
      "epoch": 1.663412335039052,
      "grad_norm": 0.7016770519505987,
      "learning_rate": 6.829760019253412e-07,
      "loss": 0.31083495616912843,
      "num_input_tokens_seen": 1270808752,
      "step": 24705,
      "train_runtime": 152077.4113,
      "train_tokens_per_second": 8356.328
    },
    {
      "epoch": 1.6637489900350122,
      "grad_norm": 0.6259997512694349,
      "learning_rate": 6.816426349723804e-07,
      "loss": 0.2980978965759277,
      "num_input_tokens_seen": 1271065168,
      "step": 24710,
      "train_runtime": 152105.5466,
      "train_tokens_per_second": 8356.468
    },
    {
      "epoch": 1.6640856450309722,
      "grad_norm": 0.6179002921674936,
      "learning_rate": 6.803104756342877e-07,
      "loss": 0.27123332023620605,
      "num_input_tokens_seen": 1271335496,
      "step": 24715,
      "train_runtime": 152132.6404,
      "train_tokens_per_second": 8356.757
    },
    {
      "epoch": 1.6644223000269323,
      "grad_norm": 0.6802002344079286,
      "learning_rate": 6.789795242835962e-07,
      "loss": 0.26348793506622314,
      "num_input_tokens_seen": 1271584248,
      "step": 24720,
      "train_runtime": 152160.023,
      "train_tokens_per_second": 8356.888
    },
    {
      "epoch": 1.6647589550228925,
      "grad_norm": 0.707305493491447,
      "learning_rate": 6.776497812925015e-07,
      "loss": 0.30634593963623047,
      "num_input_tokens_seen": 1271821136,
      "step": 24725,
      "train_runtime": 152191.0281,
      "train_tokens_per_second": 8356.742
    },
    {
      "epoch": 1.6650956100188528,
      "grad_norm": 0.6730915586863759,
      "learning_rate": 6.763212470328628e-07,
      "loss": 0.3349052667617798,
      "num_input_tokens_seen": 1272084008,
      "step": 24730,
      "train_runtime": 152225.2639,
      "train_tokens_per_second": 8356.589
    },
    {
      "epoch": 1.6654322650148128,
      "grad_norm": 0.7169688049360834,
      "learning_rate": 6.749939218762003e-07,
      "loss": 0.2913651943206787,
      "num_input_tokens_seen": 1272335032,
      "step": 24735,
      "train_runtime": 152256.0373,
      "train_tokens_per_second": 8356.549
    },
    {
      "epoch": 1.6657689200107728,
      "grad_norm": 0.5842793353749551,
      "learning_rate": 6.736678061936958e-07,
      "loss": 0.3160650968551636,
      "num_input_tokens_seen": 1272606656,
      "step": 24740,
      "train_runtime": 152286.6124,
      "train_tokens_per_second": 8356.655
    },
    {
      "epoch": 1.666105575006733,
      "grad_norm": 0.6967752161453884,
      "learning_rate": 6.723429003561937e-07,
      "loss": 0.3107904434204102,
      "num_input_tokens_seen": 1272897288,
      "step": 24745,
      "train_runtime": 152318.4413,
      "train_tokens_per_second": 8356.817
    },
    {
      "epoch": 1.6664422300026933,
      "grad_norm": 0.7425013448494789,
      "learning_rate": 6.710192047341984e-07,
      "loss": 0.2908594846725464,
      "num_input_tokens_seen": 1273145112,
      "step": 24750,
      "train_runtime": 152348.9664,
      "train_tokens_per_second": 8356.769
    },
    {
      "epoch": 1.6667788849986533,
      "grad_norm": 0.6492093784404421,
      "learning_rate": 6.69696719697881e-07,
      "loss": 0.27853055000305177,
      "num_input_tokens_seen": 1273409040,
      "step": 24755,
      "train_runtime": 152384.6787,
      "train_tokens_per_second": 8356.542
    },
    {
      "epoch": 1.6671155399946134,
      "grad_norm": 0.6621713369599981,
      "learning_rate": 6.683754456170655e-07,
      "loss": 0.2867394685745239,
      "num_input_tokens_seen": 1273669136,
      "step": 24760,
      "train_runtime": 152413.0701,
      "train_tokens_per_second": 8356.692
    },
    {
      "epoch": 1.6674521949905736,
      "grad_norm": 0.7386844331279635,
      "learning_rate": 6.670553828612448e-07,
      "loss": 0.28548855781555177,
      "num_input_tokens_seen": 1273887048,
      "step": 24765,
      "train_runtime": 152441.4881,
      "train_tokens_per_second": 8356.564
    },
    {
      "epoch": 1.6677888499865339,
      "grad_norm": 0.5777141662567643,
      "learning_rate": 6.657365317995696e-07,
      "loss": 0.2929078578948975,
      "num_input_tokens_seen": 1274142408,
      "step": 24770,
      "train_runtime": 152475.679,
      "train_tokens_per_second": 8356.365
    },
    {
      "epoch": 1.668125504982494,
      "grad_norm": 0.5975245223287966,
      "learning_rate": 6.644188928008521e-07,
      "loss": 0.3170823097229004,
      "num_input_tokens_seen": 1274384544,
      "step": 24775,
      "train_runtime": 152505.1897,
      "train_tokens_per_second": 8356.336
    },
    {
      "epoch": 1.668462159978454,
      "grad_norm": 0.7285948947125273,
      "learning_rate": 6.631024662335667e-07,
      "loss": 0.31711106300354003,
      "num_input_tokens_seen": 1274630328,
      "step": 24780,
      "train_runtime": 152534.826,
      "train_tokens_per_second": 8356.323
    },
    {
      "epoch": 1.6687988149744142,
      "grad_norm": 0.6672870924745835,
      "learning_rate": 6.617872524658469e-07,
      "loss": 0.2980704069137573,
      "num_input_tokens_seen": 1274907056,
      "step": 24785,
      "train_runtime": 152570.5577,
      "train_tokens_per_second": 8356.18
    },
    {
      "epoch": 1.6691354699703744,
      "grad_norm": 0.6567870709945342,
      "learning_rate": 6.604732518654882e-07,
      "loss": 0.3302984476089478,
      "num_input_tokens_seen": 1275155064,
      "step": 24790,
      "train_runtime": 152599.4289,
      "train_tokens_per_second": 8356.224
    },
    {
      "epoch": 1.6694721249663345,
      "grad_norm": 0.7633681055133597,
      "learning_rate": 6.591604647999473e-07,
      "loss": 0.3283034324645996,
      "num_input_tokens_seen": 1275425120,
      "step": 24795,
      "train_runtime": 152629.2835,
      "train_tokens_per_second": 8356.359
    },
    {
      "epoch": 1.6698087799622945,
      "grad_norm": 0.6088419914153808,
      "learning_rate": 6.57848891636339e-07,
      "loss": 0.28100731372833254,
      "num_input_tokens_seen": 1275689352,
      "step": 24800,
      "train_runtime": 152664.6483,
      "train_tokens_per_second": 8356.154
    },
    {
      "epoch": 1.6701454349582547,
      "grad_norm": 0.6458245973388046,
      "learning_rate": 6.565385327414441e-07,
      "loss": 0.3123847246170044,
      "num_input_tokens_seen": 1275962496,
      "step": 24805,
      "train_runtime": 152695.1454,
      "train_tokens_per_second": 8356.274
    },
    {
      "epoch": 1.670482089954215,
      "grad_norm": 0.6706589427349469,
      "learning_rate": 6.552293884816968e-07,
      "loss": 0.28674731254577634,
      "num_input_tokens_seen": 1276225336,
      "step": 24810,
      "train_runtime": 152726.7744,
      "train_tokens_per_second": 8356.265
    },
    {
      "epoch": 1.670818744950175,
      "grad_norm": 0.5806515766601483,
      "learning_rate": 6.539214592231968e-07,
      "loss": 0.2943121910095215,
      "num_input_tokens_seen": 1276497752,
      "step": 24815,
      "train_runtime": 152757.1591,
      "train_tokens_per_second": 8356.386
    },
    {
      "epoch": 1.671155399946135,
      "grad_norm": 0.6279732216817283,
      "learning_rate": 6.526147453317028e-07,
      "loss": 0.26099610328674316,
      "num_input_tokens_seen": 1276743112,
      "step": 24820,
      "train_runtime": 152788.0306,
      "train_tokens_per_second": 8356.303
    },
    {
      "epoch": 1.6714920549420953,
      "grad_norm": 0.713842057511617,
      "learning_rate": 6.513092471726307e-07,
      "loss": 0.3016034126281738,
      "num_input_tokens_seen": 1277021752,
      "step": 24825,
      "train_runtime": 152819.277,
      "train_tokens_per_second": 8356.418
    },
    {
      "epoch": 1.6718287099380555,
      "grad_norm": 0.5795232595783595,
      "learning_rate": 6.500049651110635e-07,
      "loss": 0.2878747463226318,
      "num_input_tokens_seen": 1277289080,
      "step": 24830,
      "train_runtime": 152846.809,
      "train_tokens_per_second": 8356.662
    },
    {
      "epoch": 1.6721653649340156,
      "grad_norm": 0.6194546608670671,
      "learning_rate": 6.487018995117339e-07,
      "loss": 0.29423160552978517,
      "num_input_tokens_seen": 1277587704,
      "step": 24835,
      "train_runtime": 152879.3897,
      "train_tokens_per_second": 8356.834
    },
    {
      "epoch": 1.6725020199299756,
      "grad_norm": 0.6867036285654052,
      "learning_rate": 6.474000507390455e-07,
      "loss": 0.2912751197814941,
      "num_input_tokens_seen": 1277849368,
      "step": 24840,
      "train_runtime": 152911.0484,
      "train_tokens_per_second": 8356.815
    },
    {
      "epoch": 1.6728386749259359,
      "grad_norm": 0.6127641722083538,
      "learning_rate": 6.460994191570513e-07,
      "loss": 0.29958865642547605,
      "num_input_tokens_seen": 1278118344,
      "step": 24845,
      "train_runtime": 152942.5408,
      "train_tokens_per_second": 8356.853
    },
    {
      "epoch": 1.673175329921896,
      "grad_norm": 0.6647532649005168,
      "learning_rate": 6.448000051294717e-07,
      "loss": 0.2961106777191162,
      "num_input_tokens_seen": 1278375608,
      "step": 24850,
      "train_runtime": 152973.8044,
      "train_tokens_per_second": 8356.827
    },
    {
      "epoch": 1.6735119849178561,
      "grad_norm": 0.6222367778389418,
      "learning_rate": 6.435018090196832e-07,
      "loss": 0.29656505584716797,
      "num_input_tokens_seen": 1278654320,
      "step": 24855,
      "train_runtime": 153004.4355,
      "train_tokens_per_second": 8356.976
    },
    {
      "epoch": 1.6738486399138162,
      "grad_norm": 0.63639415200674,
      "learning_rate": 6.42204831190722e-07,
      "loss": 0.2926618576049805,
      "num_input_tokens_seen": 1278930792,
      "step": 24860,
      "train_runtime": 153032.2639,
      "train_tokens_per_second": 8357.262
    },
    {
      "epoch": 1.6741852949097764,
      "grad_norm": 0.8154123845184823,
      "learning_rate": 6.409090720052835e-07,
      "loss": 0.3029963493347168,
      "num_input_tokens_seen": 1279174144,
      "step": 24865,
      "train_runtime": 153064.1051,
      "train_tokens_per_second": 8357.114
    },
    {
      "epoch": 1.6745219499057367,
      "grad_norm": 0.6157771670665775,
      "learning_rate": 6.396145318257229e-07,
      "loss": 0.2720723390579224,
      "num_input_tokens_seen": 1279442328,
      "step": 24870,
      "train_runtime": 153091.3453,
      "train_tokens_per_second": 8357.379
    },
    {
      "epoch": 1.6748586049016967,
      "grad_norm": 0.6593009387640669,
      "learning_rate": 6.383212110140535e-07,
      "loss": 0.30841591358184817,
      "num_input_tokens_seen": 1279698984,
      "step": 24875,
      "train_runtime": 153124.8829,
      "train_tokens_per_second": 8357.224
    },
    {
      "epoch": 1.6751952598976567,
      "grad_norm": 0.9378673644796045,
      "learning_rate": 6.370291099319509e-07,
      "loss": 0.3143094539642334,
      "num_input_tokens_seen": 1279916840,
      "step": 24880,
      "train_runtime": 153151.3049,
      "train_tokens_per_second": 8357.205
    },
    {
      "epoch": 1.675531914893617,
      "grad_norm": 0.6232843468974941,
      "learning_rate": 6.357382289407433e-07,
      "loss": 0.29632785320281985,
      "num_input_tokens_seen": 1280198056,
      "step": 24885,
      "train_runtime": 153181.8574,
      "train_tokens_per_second": 8357.374
    },
    {
      "epoch": 1.6758685698895772,
      "grad_norm": 0.6630602534851162,
      "learning_rate": 6.344485684014251e-07,
      "loss": 0.3144750833511353,
      "num_input_tokens_seen": 1280444760,
      "step": 24890,
      "train_runtime": 153211.8869,
      "train_tokens_per_second": 8357.346
    },
    {
      "epoch": 1.6762052248855372,
      "grad_norm": 0.6705063604819603,
      "learning_rate": 6.33160128674642e-07,
      "loss": 0.3127013921737671,
      "num_input_tokens_seen": 1280685928,
      "step": 24895,
      "train_runtime": 153242.1103,
      "train_tokens_per_second": 8357.272
    },
    {
      "epoch": 1.6765418798814973,
      "grad_norm": 0.6244597392278398,
      "learning_rate": 6.318729101207055e-07,
      "loss": 0.2881596565246582,
      "num_input_tokens_seen": 1280944256,
      "step": 24900,
      "train_runtime": 153270.6234,
      "train_tokens_per_second": 8357.402
    },
    {
      "epoch": 1.6768785348774575,
      "grad_norm": 0.585788295007258,
      "learning_rate": 6.305869130995807e-07,
      "loss": 0.305736780166626,
      "num_input_tokens_seen": 1281208696,
      "step": 24905,
      "train_runtime": 153303.7613,
      "train_tokens_per_second": 8357.321
    },
    {
      "epoch": 1.6772151898734178,
      "grad_norm": 0.6980608908234077,
      "learning_rate": 6.293021379708924e-07,
      "loss": 0.29666292667388916,
      "num_input_tokens_seen": 1281477240,
      "step": 24910,
      "train_runtime": 153335.5012,
      "train_tokens_per_second": 8357.342
    },
    {
      "epoch": 1.6775518448693778,
      "grad_norm": 0.6990755198988304,
      "learning_rate": 6.280185850939241e-07,
      "loss": 0.3206045389175415,
      "num_input_tokens_seen": 1281720688,
      "step": 24915,
      "train_runtime": 153363.3738,
      "train_tokens_per_second": 8357.411
    },
    {
      "epoch": 1.677888499865338,
      "grad_norm": 0.6930671169281918,
      "learning_rate": 6.267362548276173e-07,
      "loss": 0.3038783073425293,
      "num_input_tokens_seen": 1281955664,
      "step": 24920,
      "train_runtime": 153397.0416,
      "train_tokens_per_second": 8357.108
    },
    {
      "epoch": 1.6782251548612983,
      "grad_norm": 0.8280050588663744,
      "learning_rate": 6.254551475305703e-07,
      "loss": 0.30324745178222656,
      "num_input_tokens_seen": 1282172696,
      "step": 24925,
      "train_runtime": 153427.9172,
      "train_tokens_per_second": 8356.841
    },
    {
      "epoch": 1.6785618098572583,
      "grad_norm": 0.6703552988718168,
      "learning_rate": 6.241752635610443e-07,
      "loss": 0.31419103145599364,
      "num_input_tokens_seen": 1282430960,
      "step": 24930,
      "train_runtime": 153460.7933,
      "train_tokens_per_second": 8356.734
    },
    {
      "epoch": 1.6788984648532184,
      "grad_norm": 0.7573231850891989,
      "learning_rate": 6.228966032769496e-07,
      "loss": 0.29470531940460204,
      "num_input_tokens_seen": 1282683808,
      "step": 24935,
      "train_runtime": 153490.5868,
      "train_tokens_per_second": 8356.759
    },
    {
      "epoch": 1.6792351198491786,
      "grad_norm": 0.6877388661705258,
      "learning_rate": 6.216191670358651e-07,
      "loss": 0.29354305267333985,
      "num_input_tokens_seen": 1282927816,
      "step": 24940,
      "train_runtime": 153519.2793,
      "train_tokens_per_second": 8356.786
    },
    {
      "epoch": 1.6795717748451389,
      "grad_norm": 0.6471956558081552,
      "learning_rate": 6.203429551950163e-07,
      "loss": 0.32773828506469727,
      "num_input_tokens_seen": 1283192704,
      "step": 24945,
      "train_runtime": 153546.6128,
      "train_tokens_per_second": 8357.024
    },
    {
      "epoch": 1.679908429841099,
      "grad_norm": 0.7121249188304748,
      "learning_rate": 6.19067968111296e-07,
      "loss": 0.2773040771484375,
      "num_input_tokens_seen": 1283431232,
      "step": 24950,
      "train_runtime": 153577.5583,
      "train_tokens_per_second": 8356.893
    },
    {
      "epoch": 1.680245084837059,
      "grad_norm": 0.5933012558983565,
      "learning_rate": 6.177942061412478e-07,
      "loss": 0.27400946617126465,
      "num_input_tokens_seen": 1283687672,
      "step": 24955,
      "train_runtime": 153612.34,
      "train_tokens_per_second": 8356.67
    },
    {
      "epoch": 1.6805817398330192,
      "grad_norm": 0.5803085864038622,
      "learning_rate": 6.16521669641077e-07,
      "loss": 0.27742390632629393,
      "num_input_tokens_seen": 1283965344,
      "step": 24960,
      "train_runtime": 153645.3576,
      "train_tokens_per_second": 8356.682
    },
    {
      "epoch": 1.6809183948289794,
      "grad_norm": 0.6282407289129438,
      "learning_rate": 6.152503589666426e-07,
      "loss": 0.3035838842391968,
      "num_input_tokens_seen": 1284240528,
      "step": 24965,
      "train_runtime": 153676.1295,
      "train_tokens_per_second": 8356.799
    },
    {
      "epoch": 1.6812550498249395,
      "grad_norm": 0.7040600596823373,
      "learning_rate": 6.139802744734636e-07,
      "loss": 0.2799068450927734,
      "num_input_tokens_seen": 1284505336,
      "step": 24970,
      "train_runtime": 153707.2904,
      "train_tokens_per_second": 8356.828
    },
    {
      "epoch": 1.6815917048208995,
      "grad_norm": 0.6946433117252738,
      "learning_rate": 6.127114165167142e-07,
      "loss": 0.27260446548461914,
      "num_input_tokens_seen": 1284739272,
      "step": 24975,
      "train_runtime": 153736.0579,
      "train_tokens_per_second": 8356.786
    },
    {
      "epoch": 1.6819283598168597,
      "grad_norm": 0.7538027151599112,
      "learning_rate": 6.11443785451229e-07,
      "loss": 0.3058010578155518,
      "num_input_tokens_seen": 1285006552,
      "step": 24980,
      "train_runtime": 153764.6511,
      "train_tokens_per_second": 8356.97
    },
    {
      "epoch": 1.68226501481282,
      "grad_norm": 0.6324182966666955,
      "learning_rate": 6.101773816314926e-07,
      "loss": 0.2823632717132568,
      "num_input_tokens_seen": 1285276400,
      "step": 24985,
      "train_runtime": 153796.4544,
      "train_tokens_per_second": 8356.996
    },
    {
      "epoch": 1.68260166980878,
      "grad_norm": 0.6894895487848052,
      "learning_rate": 6.089122054116558e-07,
      "loss": 0.3190539121627808,
      "num_input_tokens_seen": 1285553552,
      "step": 24990,
      "train_runtime": 153826.1006,
      "train_tokens_per_second": 8357.187
    },
    {
      "epoch": 1.68293832480474,
      "grad_norm": 0.6746137294599758,
      "learning_rate": 6.076482571455161e-07,
      "loss": 0.2867159366607666,
      "num_input_tokens_seen": 1285812632,
      "step": 24995,
      "train_runtime": 153857.5374,
      "train_tokens_per_second": 8357.164
    },
    {
      "epoch": 1.6832749798007003,
      "grad_norm": 0.7565605525094377,
      "learning_rate": 6.063855371865369e-07,
      "loss": 0.3001354217529297,
      "num_input_tokens_seen": 1286045704,
      "step": 25000,
      "train_runtime": 153889.103,
      "train_tokens_per_second": 8356.964
    },
    {
      "epoch": 1.6836116347966605,
      "grad_norm": 0.7787218384665338,
      "learning_rate": 6.051240458878316e-07,
      "loss": 0.33756093978881835,
      "num_input_tokens_seen": 1286303896,
      "step": 25005,
      "train_runtime": 153919.1414,
      "train_tokens_per_second": 8357.011
    },
    {
      "epoch": 1.6839482897926206,
      "grad_norm": 0.7194824380388133,
      "learning_rate": 6.038637836021721e-07,
      "loss": 0.3080334424972534,
      "num_input_tokens_seen": 1286558456,
      "step": 25010,
      "train_runtime": 153951.1219,
      "train_tokens_per_second": 8356.928
    },
    {
      "epoch": 1.6842849447885806,
      "grad_norm": 0.636855887860322,
      "learning_rate": 6.026047506819882e-07,
      "loss": 0.2978890657424927,
      "num_input_tokens_seen": 1286806232,
      "step": 25015,
      "train_runtime": 153983.4696,
      "train_tokens_per_second": 8356.782
    },
    {
      "epoch": 1.6846215997845408,
      "grad_norm": 0.7322756293495313,
      "learning_rate": 6.013469474793632e-07,
      "loss": 0.29526772499084475,
      "num_input_tokens_seen": 1287066240,
      "step": 25020,
      "train_runtime": 154011.3093,
      "train_tokens_per_second": 8356.959
    },
    {
      "epoch": 1.684958254780501,
      "grad_norm": 0.7396685114324698,
      "learning_rate": 6.000903743460368e-07,
      "loss": 0.29828224182128904,
      "num_input_tokens_seen": 1287318152,
      "step": 25025,
      "train_runtime": 154041.4809,
      "train_tokens_per_second": 8356.958
    },
    {
      "epoch": 1.6852949097764611,
      "grad_norm": 0.6365369085516817,
      "learning_rate": 5.98835031633409e-07,
      "loss": 0.29223194122314455,
      "num_input_tokens_seen": 1287577904,
      "step": 25030,
      "train_runtime": 154072.6348,
      "train_tokens_per_second": 8356.954
    },
    {
      "epoch": 1.6856315647724212,
      "grad_norm": 0.6371902787418645,
      "learning_rate": 5.97580919692528e-07,
      "loss": 0.3029315948486328,
      "num_input_tokens_seen": 1287868808,
      "step": 25035,
      "train_runtime": 154103.7483,
      "train_tokens_per_second": 8357.154
    },
    {
      "epoch": 1.6859682197683814,
      "grad_norm": 0.6407492850172966,
      "learning_rate": 5.963280388741072e-07,
      "loss": 0.2878507137298584,
      "num_input_tokens_seen": 1288142320,
      "step": 25040,
      "train_runtime": 154136.8124,
      "train_tokens_per_second": 8357.136
    },
    {
      "epoch": 1.6863048747643417,
      "grad_norm": 0.8319443648342152,
      "learning_rate": 5.950763895285055e-07,
      "loss": 0.30640416145324706,
      "num_input_tokens_seen": 1288372864,
      "step": 25045,
      "train_runtime": 154164.7336,
      "train_tokens_per_second": 8357.118
    },
    {
      "epoch": 1.6866415297603017,
      "grad_norm": 0.5853979169083778,
      "learning_rate": 5.938259720057465e-07,
      "loss": 0.30520217418670653,
      "num_input_tokens_seen": 1288652600,
      "step": 25050,
      "train_runtime": 154193.0006,
      "train_tokens_per_second": 8357.4
    },
    {
      "epoch": 1.6869781847562617,
      "grad_norm": 0.7144940705504398,
      "learning_rate": 5.925767866555043e-07,
      "loss": 0.2833669185638428,
      "num_input_tokens_seen": 1288900664,
      "step": 25055,
      "train_runtime": 154219.8957,
      "train_tokens_per_second": 8357.551
    },
    {
      "epoch": 1.687314839752222,
      "grad_norm": 0.7258395527672962,
      "learning_rate": 5.913288338271095e-07,
      "loss": 0.3092818737030029,
      "num_input_tokens_seen": 1289165264,
      "step": 25060,
      "train_runtime": 154248.5295,
      "train_tokens_per_second": 8357.715
    },
    {
      "epoch": 1.6876514947481822,
      "grad_norm": 0.5782448828320352,
      "learning_rate": 5.900821138695478e-07,
      "loss": 0.2968343734741211,
      "num_input_tokens_seen": 1289425584,
      "step": 25065,
      "train_runtime": 154279.9861,
      "train_tokens_per_second": 8357.698
    },
    {
      "epoch": 1.6879881497441422,
      "grad_norm": 0.6734948372967838,
      "learning_rate": 5.888366271314605e-07,
      "loss": 0.29281589984893797,
      "num_input_tokens_seen": 1289687216,
      "step": 25070,
      "train_runtime": 154310.3476,
      "train_tokens_per_second": 8357.749
    },
    {
      "epoch": 1.6883248047401023,
      "grad_norm": 0.5806080795252254,
      "learning_rate": 5.875923739611439e-07,
      "loss": 0.30453858375549314,
      "num_input_tokens_seen": 1289970456,
      "step": 25075,
      "train_runtime": 154338.6404,
      "train_tokens_per_second": 8358.052
    },
    {
      "epoch": 1.6886614597360625,
      "grad_norm": 0.6431862606665103,
      "learning_rate": 5.863493547065513e-07,
      "loss": 0.28416364192962645,
      "num_input_tokens_seen": 1290211048,
      "step": 25080,
      "train_runtime": 154370.4709,
      "train_tokens_per_second": 8357.888
    },
    {
      "epoch": 1.6889981147320228,
      "grad_norm": 0.6561303551295199,
      "learning_rate": 5.851075697152858e-07,
      "loss": 0.2987945318222046,
      "num_input_tokens_seen": 1290476520,
      "step": 25085,
      "train_runtime": 154400.8898,
      "train_tokens_per_second": 8357.96
    },
    {
      "epoch": 1.6893347697279828,
      "grad_norm": 0.6536530647034721,
      "learning_rate": 5.838670193346118e-07,
      "loss": 0.30833442211151124,
      "num_input_tokens_seen": 1290721944,
      "step": 25090,
      "train_runtime": 154433.7389,
      "train_tokens_per_second": 8357.772
    },
    {
      "epoch": 1.6896714247239428,
      "grad_norm": 0.6069576919496391,
      "learning_rate": 5.826277039114426e-07,
      "loss": 0.2854807376861572,
      "num_input_tokens_seen": 1290974304,
      "step": 25095,
      "train_runtime": 154462.2358,
      "train_tokens_per_second": 8357.864
    },
    {
      "epoch": 1.690008079719903,
      "grad_norm": 0.6000086380493743,
      "learning_rate": 5.813896237923505e-07,
      "loss": 0.32449994087219236,
      "num_input_tokens_seen": 1291229392,
      "step": 25100,
      "train_runtime": 154491.8026,
      "train_tokens_per_second": 8357.915
    },
    {
      "epoch": 1.6903447347158633,
      "grad_norm": 0.6273643507827594,
      "learning_rate": 5.801527793235606e-07,
      "loss": 0.2946127653121948,
      "num_input_tokens_seen": 1291512784,
      "step": 25105,
      "train_runtime": 154521.5806,
      "train_tokens_per_second": 8358.139
    },
    {
      "epoch": 1.6906813897118234,
      "grad_norm": 0.6540016592133578,
      "learning_rate": 5.789171708509516e-07,
      "loss": 0.2910575866699219,
      "num_input_tokens_seen": 1291733008,
      "step": 25110,
      "train_runtime": 154548.5924,
      "train_tokens_per_second": 8358.103
    },
    {
      "epoch": 1.6910180447077834,
      "grad_norm": 0.6407241178744663,
      "learning_rate": 5.776827987200584e-07,
      "loss": 0.2972250461578369,
      "num_input_tokens_seen": 1291991440,
      "step": 25115,
      "train_runtime": 154579.7383,
      "train_tokens_per_second": 8358.09
    },
    {
      "epoch": 1.6913546997037436,
      "grad_norm": 0.6257782244271743,
      "learning_rate": 5.764496632760691e-07,
      "loss": 0.2869912624359131,
      "num_input_tokens_seen": 1292235256,
      "step": 25120,
      "train_runtime": 154608.893,
      "train_tokens_per_second": 8358.091
    },
    {
      "epoch": 1.6916913546997039,
      "grad_norm": 0.6009010803890104,
      "learning_rate": 5.752177648638241e-07,
      "loss": 0.27775373458862307,
      "num_input_tokens_seen": 1292499416,
      "step": 25125,
      "train_runtime": 154645.162,
      "train_tokens_per_second": 8357.839
    },
    {
      "epoch": 1.692028009695664,
      "grad_norm": 0.7216408638562958,
      "learning_rate": 5.739871038278239e-07,
      "loss": 0.30296647548675537,
      "num_input_tokens_seen": 1292762080,
      "step": 25130,
      "train_runtime": 154676.5319,
      "train_tokens_per_second": 8357.842
    },
    {
      "epoch": 1.692364664691624,
      "grad_norm": 0.7591894530783491,
      "learning_rate": 5.727576805122142e-07,
      "loss": 0.2801056861877441,
      "num_input_tokens_seen": 1292996720,
      "step": 25135,
      "train_runtime": 154705.4518,
      "train_tokens_per_second": 8357.797
    },
    {
      "epoch": 1.6927013196875842,
      "grad_norm": 0.5999079129545056,
      "learning_rate": 5.715294952608041e-07,
      "loss": 0.3096039295196533,
      "num_input_tokens_seen": 1293284824,
      "step": 25140,
      "train_runtime": 154736.7016,
      "train_tokens_per_second": 8357.971
    },
    {
      "epoch": 1.6930379746835444,
      "grad_norm": 0.6482333214550756,
      "learning_rate": 5.703025484170471e-07,
      "loss": 0.30091152191162107,
      "num_input_tokens_seen": 1293567136,
      "step": 25145,
      "train_runtime": 154773.8451,
      "train_tokens_per_second": 8357.789
    },
    {
      "epoch": 1.6933746296795045,
      "grad_norm": 0.6025783342763544,
      "learning_rate": 5.690768403240587e-07,
      "loss": 0.31388542652130125,
      "num_input_tokens_seen": 1293831520,
      "step": 25150,
      "train_runtime": 154801.7838,
      "train_tokens_per_second": 8357.988
    },
    {
      "epoch": 1.6937112846754645,
      "grad_norm": 0.6027071897692966,
      "learning_rate": 5.67852371324602e-07,
      "loss": 0.3045590162277222,
      "num_input_tokens_seen": 1294105024,
      "step": 25155,
      "train_runtime": 154834.3493,
      "train_tokens_per_second": 8357.997
    },
    {
      "epoch": 1.6940479396714248,
      "grad_norm": 0.6538552625942754,
      "learning_rate": 5.66629141761097e-07,
      "loss": 0.29695720672607423,
      "num_input_tokens_seen": 1294392192,
      "step": 25160,
      "train_runtime": 154868.4835,
      "train_tokens_per_second": 8358.009
    },
    {
      "epoch": 1.694384594667385,
      "grad_norm": 0.6444162258141245,
      "learning_rate": 5.654071519756154e-07,
      "loss": 0.30961055755615235,
      "num_input_tokens_seen": 1294648200,
      "step": 25165,
      "train_runtime": 154899.1315,
      "train_tokens_per_second": 8358.008
    },
    {
      "epoch": 1.694721249663345,
      "grad_norm": 0.6982665209424952,
      "learning_rate": 5.641864023098831e-07,
      "loss": 0.3370542526245117,
      "num_input_tokens_seen": 1294901432,
      "step": 25170,
      "train_runtime": 154926.7055,
      "train_tokens_per_second": 8358.155
    },
    {
      "epoch": 1.695057904659305,
      "grad_norm": 0.6262299196983416,
      "learning_rate": 5.629668931052773e-07,
      "loss": 0.2924381732940674,
      "num_input_tokens_seen": 1295177944,
      "step": 25175,
      "train_runtime": 154956.5835,
      "train_tokens_per_second": 8358.328
    },
    {
      "epoch": 1.6953945596552653,
      "grad_norm": 0.8122871770053646,
      "learning_rate": 5.61748624702833e-07,
      "loss": 0.3111175537109375,
      "num_input_tokens_seen": 1295446168,
      "step": 25180,
      "train_runtime": 154988.7959,
      "train_tokens_per_second": 8358.321
    },
    {
      "epoch": 1.6957312146512256,
      "grad_norm": 0.6969033694193383,
      "learning_rate": 5.605315974432313e-07,
      "loss": 0.30250959396362304,
      "num_input_tokens_seen": 1295697384,
      "step": 25185,
      "train_runtime": 155018.017,
      "train_tokens_per_second": 8358.366
    },
    {
      "epoch": 1.6960678696471856,
      "grad_norm": 0.5898269042153695,
      "learning_rate": 5.593158116668146e-07,
      "loss": 0.2969617128372192,
      "num_input_tokens_seen": 1295982160,
      "step": 25190,
      "train_runtime": 155048.0269,
      "train_tokens_per_second": 8358.585
    },
    {
      "epoch": 1.6964045246431456,
      "grad_norm": 0.6848871667108906,
      "learning_rate": 5.581012677135683e-07,
      "loss": 0.30820045471191404,
      "num_input_tokens_seen": 1296220664,
      "step": 25195,
      "train_runtime": 155082.7272,
      "train_tokens_per_second": 8358.253
    },
    {
      "epoch": 1.6967411796391059,
      "grad_norm": 0.6154500201081657,
      "learning_rate": 5.56887965923139e-07,
      "loss": 0.30482871532440187,
      "num_input_tokens_seen": 1296495288,
      "step": 25200,
      "train_runtime": 155112.4202,
      "train_tokens_per_second": 8358.423
    },
    {
      "epoch": 1.6970778346350661,
      "grad_norm": 0.7169435751175564,
      "learning_rate": 5.556759066348227e-07,
      "loss": 0.2932065486907959,
      "num_input_tokens_seen": 1296757936,
      "step": 25205,
      "train_runtime": 155143.0841,
      "train_tokens_per_second": 8358.464
    },
    {
      "epoch": 1.6974144896310261,
      "grad_norm": 0.6985329715449524,
      "learning_rate": 5.54465090187567e-07,
      "loss": 0.2974205255508423,
      "num_input_tokens_seen": 1297007664,
      "step": 25210,
      "train_runtime": 155171.7806,
      "train_tokens_per_second": 8358.528
    },
    {
      "epoch": 1.6977511446269862,
      "grad_norm": 0.6712291802433147,
      "learning_rate": 5.532555169199727e-07,
      "loss": 0.2888886213302612,
      "num_input_tokens_seen": 1297268904,
      "step": 25215,
      "train_runtime": 155207.4821,
      "train_tokens_per_second": 8358.288
    },
    {
      "epoch": 1.6980877996229464,
      "grad_norm": 0.5975063585150914,
      "learning_rate": 5.520471871702937e-07,
      "loss": 0.28809595108032227,
      "num_input_tokens_seen": 1297534656,
      "step": 25220,
      "train_runtime": 155239.6581,
      "train_tokens_per_second": 8358.268
    },
    {
      "epoch": 1.6984244546189067,
      "grad_norm": 0.7565069192875064,
      "learning_rate": 5.508401012764341e-07,
      "loss": 0.29517331123352053,
      "num_input_tokens_seen": 1297801232,
      "step": 25225,
      "train_runtime": 155269.783,
      "train_tokens_per_second": 8358.363
    },
    {
      "epoch": 1.6987611096148667,
      "grad_norm": 0.6551417551090734,
      "learning_rate": 5.496342595759552e-07,
      "loss": 0.2842202425003052,
      "num_input_tokens_seen": 1298072216,
      "step": 25230,
      "train_runtime": 155302.5893,
      "train_tokens_per_second": 8358.342
    },
    {
      "epoch": 1.6990977646108267,
      "grad_norm": 0.6400244416990853,
      "learning_rate": 5.484296624060626e-07,
      "loss": 0.29554221630096433,
      "num_input_tokens_seen": 1298339888,
      "step": 25235,
      "train_runtime": 155332.8051,
      "train_tokens_per_second": 8358.44
    },
    {
      "epoch": 1.699434419606787,
      "grad_norm": 0.6279590178301934,
      "learning_rate": 5.472263101036212e-07,
      "loss": 0.3003653049468994,
      "num_input_tokens_seen": 1298603880,
      "step": 25240,
      "train_runtime": 155366.7718,
      "train_tokens_per_second": 8358.312
    },
    {
      "epoch": 1.6997710746027472,
      "grad_norm": 0.6717653169467266,
      "learning_rate": 5.460242030051421e-07,
      "loss": 0.30445232391357424,
      "num_input_tokens_seen": 1298857744,
      "step": 25245,
      "train_runtime": 155395.0753,
      "train_tokens_per_second": 8358.423
    },
    {
      "epoch": 1.7001077295987073,
      "grad_norm": 0.6351101686838609,
      "learning_rate": 5.448233414467924e-07,
      "loss": 0.3035956621170044,
      "num_input_tokens_seen": 1299143288,
      "step": 25250,
      "train_runtime": 155428.6048,
      "train_tokens_per_second": 8358.457
    },
    {
      "epoch": 1.7004443845946673,
      "grad_norm": 0.6471961373176617,
      "learning_rate": 5.43623725764389e-07,
      "loss": 0.3183321475982666,
      "num_input_tokens_seen": 1299396920,
      "step": 25255,
      "train_runtime": 155457.8978,
      "train_tokens_per_second": 8358.513
    },
    {
      "epoch": 1.7007810395906275,
      "grad_norm": 0.6897866747178667,
      "learning_rate": 5.424253562934001e-07,
      "loss": 0.30020053386688234,
      "num_input_tokens_seen": 1299661104,
      "step": 25260,
      "train_runtime": 155486.499,
      "train_tokens_per_second": 8358.675
    },
    {
      "epoch": 1.7011176945865878,
      "grad_norm": 0.6577159451121951,
      "learning_rate": 5.412282333689461e-07,
      "loss": 0.2879279613494873,
      "num_input_tokens_seen": 1299929648,
      "step": 25265,
      "train_runtime": 155516.2482,
      "train_tokens_per_second": 8358.803
    },
    {
      "epoch": 1.7014543495825478,
      "grad_norm": 0.6335745308129616,
      "learning_rate": 5.400323573257987e-07,
      "loss": 0.2918604612350464,
      "num_input_tokens_seen": 1300194680,
      "step": 25270,
      "train_runtime": 155549.9319,
      "train_tokens_per_second": 8358.697
    },
    {
      "epoch": 1.7017910045785078,
      "grad_norm": 0.602417149609731,
      "learning_rate": 5.388377284983803e-07,
      "loss": 0.2854419469833374,
      "num_input_tokens_seen": 1300460200,
      "step": 25275,
      "train_runtime": 155584.0457,
      "train_tokens_per_second": 8358.57
    },
    {
      "epoch": 1.702127659574468,
      "grad_norm": 0.6479356613172405,
      "learning_rate": 5.37644347220767e-07,
      "loss": 0.28288946151733396,
      "num_input_tokens_seen": 1300713688,
      "step": 25280,
      "train_runtime": 155613.6923,
      "train_tokens_per_second": 8358.607
    },
    {
      "epoch": 1.7024643145704283,
      "grad_norm": 0.7475429254619368,
      "learning_rate": 5.364522138266809e-07,
      "loss": 0.30722129344940186,
      "num_input_tokens_seen": 1300994640,
      "step": 25285,
      "train_runtime": 155645.5762,
      "train_tokens_per_second": 8358.7
    },
    {
      "epoch": 1.7028009695663884,
      "grad_norm": 0.6553953008008376,
      "learning_rate": 5.352613286495018e-07,
      "loss": 0.3007795810699463,
      "num_input_tokens_seen": 1301239672,
      "step": 25290,
      "train_runtime": 155675.562,
      "train_tokens_per_second": 8358.664
    },
    {
      "epoch": 1.7031376245623484,
      "grad_norm": 0.6924825419988966,
      "learning_rate": 5.340716920222527e-07,
      "loss": 0.2899085760116577,
      "num_input_tokens_seen": 1301492152,
      "step": 25295,
      "train_runtime": 155707.6192,
      "train_tokens_per_second": 8358.564
    },
    {
      "epoch": 1.7034742795583087,
      "grad_norm": 0.7028206331022429,
      "learning_rate": 5.32883304277616e-07,
      "loss": 0.3263566017150879,
      "num_input_tokens_seen": 1301732160,
      "step": 25300,
      "train_runtime": 155739.0099,
      "train_tokens_per_second": 8358.421
    },
    {
      "epoch": 1.703810934554269,
      "grad_norm": 0.6910432608072555,
      "learning_rate": 5.31696165747918e-07,
      "loss": 0.2739458799362183,
      "num_input_tokens_seen": 1301994888,
      "step": 25305,
      "train_runtime": 155773.9123,
      "train_tokens_per_second": 8358.234
    },
    {
      "epoch": 1.704147589550229,
      "grad_norm": 1.1519379167014636,
      "learning_rate": 5.305102767651393e-07,
      "loss": 0.2800089597702026,
      "num_input_tokens_seen": 1302267936,
      "step": 25310,
      "train_runtime": 155805.17,
      "train_tokens_per_second": 8358.31
    },
    {
      "epoch": 1.704484244546189,
      "grad_norm": 0.6460904887278585,
      "learning_rate": 5.293256376609091e-07,
      "loss": 0.29753289222717283,
      "num_input_tokens_seen": 1302530400,
      "step": 25315,
      "train_runtime": 155838.1044,
      "train_tokens_per_second": 8358.228
    },
    {
      "epoch": 1.7048208995421492,
      "grad_norm": 0.5924512231418766,
      "learning_rate": 5.281422487665089e-07,
      "loss": 0.27810642719268797,
      "num_input_tokens_seen": 1302808368,
      "step": 25320,
      "train_runtime": 155870.114,
      "train_tokens_per_second": 8358.295
    },
    {
      "epoch": 1.7051575545381095,
      "grad_norm": 0.8172620177570578,
      "learning_rate": 5.269601104128674e-07,
      "loss": 0.28679211139678956,
      "num_input_tokens_seen": 1303070080,
      "step": 25325,
      "train_runtime": 155901.6524,
      "train_tokens_per_second": 8358.283
    },
    {
      "epoch": 1.7054942095340695,
      "grad_norm": 0.6266649408709838,
      "learning_rate": 5.257792229305702e-07,
      "loss": 0.3052984714508057,
      "num_input_tokens_seen": 1303330720,
      "step": 25330,
      "train_runtime": 155932.5147,
      "train_tokens_per_second": 8358.3
    },
    {
      "epoch": 1.7058308645300295,
      "grad_norm": 0.831578000245705,
      "learning_rate": 5.245995866498438e-07,
      "loss": 0.2755401849746704,
      "num_input_tokens_seen": 1303562736,
      "step": 25335,
      "train_runtime": 155963.3847,
      "train_tokens_per_second": 8358.133
    },
    {
      "epoch": 1.7061675195259898,
      "grad_norm": 0.6657919020828802,
      "learning_rate": 5.234212019005741e-07,
      "loss": 0.29737756252288816,
      "num_input_tokens_seen": 1303802048,
      "step": 25340,
      "train_runtime": 155991.4928,
      "train_tokens_per_second": 8358.161
    },
    {
      "epoch": 1.70650417452195,
      "grad_norm": 0.6249343414087439,
      "learning_rate": 5.222440690122882e-07,
      "loss": 0.2909064531326294,
      "num_input_tokens_seen": 1304046688,
      "step": 25345,
      "train_runtime": 156024.8235,
      "train_tokens_per_second": 8357.944
    },
    {
      "epoch": 1.70684082951791,
      "grad_norm": 0.6439603052683572,
      "learning_rate": 5.210681883141716e-07,
      "loss": 0.27885620594024657,
      "num_input_tokens_seen": 1304303432,
      "step": 25350,
      "train_runtime": 156058.2664,
      "train_tokens_per_second": 8357.798
    },
    {
      "epoch": 1.70717748451387,
      "grad_norm": 0.657210126898185,
      "learning_rate": 5.19893560135053e-07,
      "loss": 0.3141366243362427,
      "num_input_tokens_seen": 1304571560,
      "step": 25355,
      "train_runtime": 156089.2884,
      "train_tokens_per_second": 8357.854
    },
    {
      "epoch": 1.7075141395098303,
      "grad_norm": 0.6645382062207716,
      "learning_rate": 5.187201848034146e-07,
      "loss": 0.29487786293029783,
      "num_input_tokens_seen": 1304834376,
      "step": 25360,
      "train_runtime": 156119.8863,
      "train_tokens_per_second": 8357.9
    },
    {
      "epoch": 1.7078507945057906,
      "grad_norm": 0.6583724376811803,
      "learning_rate": 5.175480626473873e-07,
      "loss": 0.29004855155944825,
      "num_input_tokens_seen": 1305077552,
      "step": 25365,
      "train_runtime": 156150.9944,
      "train_tokens_per_second": 8357.792
    },
    {
      "epoch": 1.7081874495017506,
      "grad_norm": 0.73322393319492,
      "learning_rate": 5.163771939947504e-07,
      "loss": 0.2993566274642944,
      "num_input_tokens_seen": 1305339792,
      "step": 25370,
      "train_runtime": 156178.7022,
      "train_tokens_per_second": 8357.988
    },
    {
      "epoch": 1.7085241044977106,
      "grad_norm": 0.5525270457086267,
      "learning_rate": 5.152075791729327e-07,
      "loss": 0.32078018188476565,
      "num_input_tokens_seen": 1305613432,
      "step": 25375,
      "train_runtime": 156213.4803,
      "train_tokens_per_second": 8357.879
    },
    {
      "epoch": 1.7088607594936709,
      "grad_norm": 0.6278229118589885,
      "learning_rate": 5.140392185090165e-07,
      "loss": 0.2948938846588135,
      "num_input_tokens_seen": 1305849096,
      "step": 25380,
      "train_runtime": 156241.0775,
      "train_tokens_per_second": 8357.911
    },
    {
      "epoch": 1.7091974144896311,
      "grad_norm": 0.7872382734167945,
      "learning_rate": 5.128721123297265e-07,
      "loss": 0.30551419258117674,
      "num_input_tokens_seen": 1306121512,
      "step": 25385,
      "train_runtime": 156274.9399,
      "train_tokens_per_second": 8357.844
    },
    {
      "epoch": 1.7095340694855912,
      "grad_norm": 0.6896832895707034,
      "learning_rate": 5.117062609614438e-07,
      "loss": 0.2921993017196655,
      "num_input_tokens_seen": 1306381856,
      "step": 25390,
      "train_runtime": 156306.8001,
      "train_tokens_per_second": 8357.806
    },
    {
      "epoch": 1.7098707244815512,
      "grad_norm": 0.6869864881809644,
      "learning_rate": 5.105416647301909e-07,
      "loss": 0.29342117309570315,
      "num_input_tokens_seen": 1306640648,
      "step": 25395,
      "train_runtime": 156337.9125,
      "train_tokens_per_second": 8357.798
    },
    {
      "epoch": 1.7102073794775114,
      "grad_norm": 0.7106316520699582,
      "learning_rate": 5.093783239616468e-07,
      "loss": 0.27153050899505615,
      "num_input_tokens_seen": 1306878712,
      "step": 25400,
      "train_runtime": 156366.3757,
      "train_tokens_per_second": 8357.799
    },
    {
      "epoch": 1.7105440344734717,
      "grad_norm": 0.826219683620497,
      "learning_rate": 5.082162389811352e-07,
      "loss": 0.2772348880767822,
      "num_input_tokens_seen": 1307147584,
      "step": 25405,
      "train_runtime": 156397.2935,
      "train_tokens_per_second": 8357.866
    },
    {
      "epoch": 1.7108806894694317,
      "grad_norm": 0.6621425586696408,
      "learning_rate": 5.070554101136288e-07,
      "loss": 0.29272890090942383,
      "num_input_tokens_seen": 1307389760,
      "step": 25410,
      "train_runtime": 156426.552,
      "train_tokens_per_second": 8357.851
    },
    {
      "epoch": 1.7112173444653918,
      "grad_norm": 0.5723935531722559,
      "learning_rate": 5.058958376837497e-07,
      "loss": 0.30031347274780273,
      "num_input_tokens_seen": 1307657392,
      "step": 25415,
      "train_runtime": 156457.7886,
      "train_tokens_per_second": 8357.893
    },
    {
      "epoch": 1.711553999461352,
      "grad_norm": 0.62493588066649,
      "learning_rate": 5.047375220157691e-07,
      "loss": 0.30083141326904295,
      "num_input_tokens_seen": 1307918608,
      "step": 25420,
      "train_runtime": 156487.0711,
      "train_tokens_per_second": 8357.998
    },
    {
      "epoch": 1.7118906544573123,
      "grad_norm": 0.782600261881126,
      "learning_rate": 5.035804634336056e-07,
      "loss": 0.31206326484680175,
      "num_input_tokens_seen": 1308169544,
      "step": 25425,
      "train_runtime": 156515.8422,
      "train_tokens_per_second": 8358.065
    },
    {
      "epoch": 1.7122273094532723,
      "grad_norm": 0.6996392538631012,
      "learning_rate": 5.024246622608292e-07,
      "loss": 0.2819211959838867,
      "num_input_tokens_seen": 1308456832,
      "step": 25430,
      "train_runtime": 156546.0172,
      "train_tokens_per_second": 8358.289
    },
    {
      "epoch": 1.7125639644492323,
      "grad_norm": 0.9729396675746177,
      "learning_rate": 5.01270118820652e-07,
      "loss": 0.3106332778930664,
      "num_input_tokens_seen": 1308728968,
      "step": 25435,
      "train_runtime": 156577.8348,
      "train_tokens_per_second": 8358.328
    },
    {
      "epoch": 1.7129006194451926,
      "grad_norm": 0.6655089371947147,
      "learning_rate": 5.001168334359429e-07,
      "loss": 0.30740971565246583,
      "num_input_tokens_seen": 1309004128,
      "step": 25440,
      "train_runtime": 156608.2382,
      "train_tokens_per_second": 8358.463
    },
    {
      "epoch": 1.7132372744411528,
      "grad_norm": 0.6007638641872738,
      "learning_rate": 4.989648064292102e-07,
      "loss": 0.301995062828064,
      "num_input_tokens_seen": 1309272536,
      "step": 25445,
      "train_runtime": 156640.2082,
      "train_tokens_per_second": 8358.47
    },
    {
      "epoch": 1.7135739294371128,
      "grad_norm": 0.6792347789593666,
      "learning_rate": 4.978140381226177e-07,
      "loss": 0.30781214237213134,
      "num_input_tokens_seen": 1309529328,
      "step": 25450,
      "train_runtime": 156677.2741,
      "train_tokens_per_second": 8358.132
    },
    {
      "epoch": 1.7139105844330729,
      "grad_norm": 0.5576940627587856,
      "learning_rate": 4.966645288379729e-07,
      "loss": 0.28174552917480467,
      "num_input_tokens_seen": 1309798248,
      "step": 25455,
      "train_runtime": 156707.7347,
      "train_tokens_per_second": 8358.223
    },
    {
      "epoch": 1.7142472394290331,
      "grad_norm": 0.7795958758228526,
      "learning_rate": 4.955162788967327e-07,
      "loss": 0.30873157978057864,
      "num_input_tokens_seen": 1310012104,
      "step": 25460,
      "train_runtime": 156741.9102,
      "train_tokens_per_second": 8357.765
    },
    {
      "epoch": 1.7145838944249934,
      "grad_norm": 0.6298424383095472,
      "learning_rate": 4.943692886200013e-07,
      "loss": 0.3071903228759766,
      "num_input_tokens_seen": 1310264672,
      "step": 25465,
      "train_runtime": 156776.7402,
      "train_tokens_per_second": 8357.52
    },
    {
      "epoch": 1.7149205494209534,
      "grad_norm": 0.6285997763474522,
      "learning_rate": 4.932235583285305e-07,
      "loss": 0.31211133003234864,
      "num_input_tokens_seen": 1310546224,
      "step": 25470,
      "train_runtime": 156808.3551,
      "train_tokens_per_second": 8357.63
    },
    {
      "epoch": 1.7152572044169134,
      "grad_norm": 0.7115829508400912,
      "learning_rate": 4.920790883427201e-07,
      "loss": 0.29173007011413576,
      "num_input_tokens_seen": 1310765720,
      "step": 25475,
      "train_runtime": 156835.7556,
      "train_tokens_per_second": 8357.569
    },
    {
      "epoch": 1.7155938594128737,
      "grad_norm": 0.7287581767843716,
      "learning_rate": 4.909358789826196e-07,
      "loss": 0.3067842960357666,
      "num_input_tokens_seen": 1310974416,
      "step": 25480,
      "train_runtime": 156865.1734,
      "train_tokens_per_second": 8357.333
    },
    {
      "epoch": 1.715930514408834,
      "grad_norm": 0.6532999769238355,
      "learning_rate": 4.897939305679206e-07,
      "loss": 0.30706071853637695,
      "num_input_tokens_seen": 1311257232,
      "step": 25485,
      "train_runtime": 156895.3284,
      "train_tokens_per_second": 8357.529
    },
    {
      "epoch": 1.716267169404794,
      "grad_norm": 0.6338798327163583,
      "learning_rate": 4.886532434179686e-07,
      "loss": 0.28668437004089353,
      "num_input_tokens_seen": 1311496824,
      "step": 25490,
      "train_runtime": 156926.8811,
      "train_tokens_per_second": 8357.375
    },
    {
      "epoch": 1.716603824400754,
      "grad_norm": 0.6431232650341788,
      "learning_rate": 4.875138178517497e-07,
      "loss": 0.2722508668899536,
      "num_input_tokens_seen": 1311757872,
      "step": 25495,
      "train_runtime": 156959.3206,
      "train_tokens_per_second": 8357.311
    },
    {
      "epoch": 1.7169404793967142,
      "grad_norm": 0.6810239618823012,
      "learning_rate": 4.863756541879028e-07,
      "loss": 0.318449068069458,
      "num_input_tokens_seen": 1312018464,
      "step": 25500,
      "train_runtime": 156991.1835,
      "train_tokens_per_second": 8357.275
    },
    {
      "epoch": 1.7172771343926745,
      "grad_norm": 0.6347756335953764,
      "learning_rate": 4.852387527447117e-07,
      "loss": 0.276993989944458,
      "num_input_tokens_seen": 1312300584,
      "step": 25505,
      "train_runtime": 157019.3302,
      "train_tokens_per_second": 8357.573
    },
    {
      "epoch": 1.7176137893886345,
      "grad_norm": 0.6660389534825137,
      "learning_rate": 4.84103113840107e-07,
      "loss": 0.29533069133758544,
      "num_input_tokens_seen": 1312562720,
      "step": 25510,
      "train_runtime": 157053.7632,
      "train_tokens_per_second": 8357.41
    },
    {
      "epoch": 1.7179504443845945,
      "grad_norm": 0.6371437811618621,
      "learning_rate": 4.829687377916664e-07,
      "loss": 0.2890852689743042,
      "num_input_tokens_seen": 1312824408,
      "step": 25515,
      "train_runtime": 157083.035,
      "train_tokens_per_second": 8357.519
    },
    {
      "epoch": 1.7182870993805548,
      "grad_norm": 0.7510638936249103,
      "learning_rate": 4.81835624916614e-07,
      "loss": 0.2845533609390259,
      "num_input_tokens_seen": 1313073152,
      "step": 25520,
      "train_runtime": 157112.9607,
      "train_tokens_per_second": 8357.51
    },
    {
      "epoch": 1.718623754376515,
      "grad_norm": 0.6688860639994233,
      "learning_rate": 4.807037755318211e-07,
      "loss": 0.2933858156204224,
      "num_input_tokens_seen": 1313355864,
      "step": 25525,
      "train_runtime": 157145.5681,
      "train_tokens_per_second": 8357.575
    },
    {
      "epoch": 1.718960409372475,
      "grad_norm": 0.5739320758744232,
      "learning_rate": 4.795731899538086e-07,
      "loss": 0.3091202020645142,
      "num_input_tokens_seen": 1313590512,
      "step": 25530,
      "train_runtime": 157178.2875,
      "train_tokens_per_second": 8357.328
    },
    {
      "epoch": 1.719297064368435,
      "grad_norm": 0.5723054178696603,
      "learning_rate": 4.78443868498737e-07,
      "loss": 0.28513364791870116,
      "num_input_tokens_seen": 1313844688,
      "step": 25535,
      "train_runtime": 157209.2906,
      "train_tokens_per_second": 8357.297
    },
    {
      "epoch": 1.7196337193643954,
      "grad_norm": 0.6224199267358416,
      "learning_rate": 4.773158114824206e-07,
      "loss": 0.284136700630188,
      "num_input_tokens_seen": 1314105880,
      "step": 25540,
      "train_runtime": 157239.895,
      "train_tokens_per_second": 8357.331
    },
    {
      "epoch": 1.7199703743603556,
      "grad_norm": 0.6715207183739592,
      "learning_rate": 4.7618901922031445e-07,
      "loss": 0.2867898464202881,
      "num_input_tokens_seen": 1314374736,
      "step": 25545,
      "train_runtime": 157271.1304,
      "train_tokens_per_second": 8357.381
    },
    {
      "epoch": 1.7203070293563156,
      "grad_norm": 0.6655488872205229,
      "learning_rate": 4.750634920275243e-07,
      "loss": 0.28437850475311277,
      "num_input_tokens_seen": 1314627440,
      "step": 25550,
      "train_runtime": 157298.8308,
      "train_tokens_per_second": 8357.516
    },
    {
      "epoch": 1.7206436843522757,
      "grad_norm": 0.6425874933393432,
      "learning_rate": 4.739392302188001e-07,
      "loss": 0.3136065244674683,
      "num_input_tokens_seen": 1314903632,
      "step": 25555,
      "train_runtime": 157329.8328,
      "train_tokens_per_second": 8357.624
    },
    {
      "epoch": 1.720980339348236,
      "grad_norm": 0.6601304036985154,
      "learning_rate": 4.728162341085368e-07,
      "loss": 0.30288076400756836,
      "num_input_tokens_seen": 1315156808,
      "step": 25560,
      "train_runtime": 157362.9094,
      "train_tokens_per_second": 8357.476
    },
    {
      "epoch": 1.7213169943441962,
      "grad_norm": 0.6022547802485434,
      "learning_rate": 4.7169450401077843e-07,
      "loss": 0.30704526901245116,
      "num_input_tokens_seen": 1315420456,
      "step": 25565,
      "train_runtime": 157396.8575,
      "train_tokens_per_second": 8357.349
    },
    {
      "epoch": 1.7216536493401562,
      "grad_norm": 0.6403858660145809,
      "learning_rate": 4.7057404023921174e-07,
      "loss": 0.3089473724365234,
      "num_input_tokens_seen": 1315696720,
      "step": 25570,
      "train_runtime": 157426.2592,
      "train_tokens_per_second": 8357.543
    },
    {
      "epoch": 1.7219903043361162,
      "grad_norm": 0.7343246539877568,
      "learning_rate": 4.694548431071705e-07,
      "loss": 0.3086869478225708,
      "num_input_tokens_seen": 1315946360,
      "step": 25575,
      "train_runtime": 157456.1639,
      "train_tokens_per_second": 8357.541
    },
    {
      "epoch": 1.7223269593320765,
      "grad_norm": 0.5864157534698733,
      "learning_rate": 4.683369129276377e-07,
      "loss": 0.296161150932312,
      "num_input_tokens_seen": 1316225776,
      "step": 25580,
      "train_runtime": 157487.3918,
      "train_tokens_per_second": 8357.658
    },
    {
      "epoch": 1.7226636143280367,
      "grad_norm": 0.5984536601680418,
      "learning_rate": 4.6722025001323444e-07,
      "loss": 0.2812647342681885,
      "num_input_tokens_seen": 1316440360,
      "step": 25585,
      "train_runtime": 157520.1452,
      "train_tokens_per_second": 8357.283
    },
    {
      "epoch": 1.7230002693239967,
      "grad_norm": 0.7562415701514201,
      "learning_rate": 4.6610485467623643e-07,
      "loss": 0.30520100593566896,
      "num_input_tokens_seen": 1316673320,
      "step": 25590,
      "train_runtime": 157549.8154,
      "train_tokens_per_second": 8357.187
    },
    {
      "epoch": 1.7233369243199568,
      "grad_norm": 0.6584866016600502,
      "learning_rate": 4.649907272285564e-07,
      "loss": 0.2941312313079834,
      "num_input_tokens_seen": 1316928816,
      "step": 25595,
      "train_runtime": 157580.3754,
      "train_tokens_per_second": 8357.188
    },
    {
      "epoch": 1.723673579315917,
      "grad_norm": 0.6656442676093736,
      "learning_rate": 4.638778679817596e-07,
      "loss": 0.29076247215270995,
      "num_input_tokens_seen": 1317201776,
      "step": 25600,
      "train_runtime": 157614.2514,
      "train_tokens_per_second": 8357.124
    },
    {
      "epoch": 1.7240102343118773,
      "grad_norm": 0.760826579054111,
      "learning_rate": 4.6276627724705204e-07,
      "loss": 0.3008571624755859,
      "num_input_tokens_seen": 1317451368,
      "step": 25605,
      "train_runtime": 157642.7711,
      "train_tokens_per_second": 8357.195
    },
    {
      "epoch": 1.7243468893078373,
      "grad_norm": 0.5663029024829883,
      "learning_rate": 4.616559553352873e-07,
      "loss": 0.2825214385986328,
      "num_input_tokens_seen": 1317713144,
      "step": 25610,
      "train_runtime": 157670.276,
      "train_tokens_per_second": 8357.397
    },
    {
      "epoch": 1.7246835443037973,
      "grad_norm": 0.6211198687357852,
      "learning_rate": 4.6054690255696256e-07,
      "loss": 0.3079988479614258,
      "num_input_tokens_seen": 1317978208,
      "step": 25615,
      "train_runtime": 157696.4388,
      "train_tokens_per_second": 8357.692
    },
    {
      "epoch": 1.7250201992997576,
      "grad_norm": 0.7089888476573826,
      "learning_rate": 4.594391192222214e-07,
      "loss": 0.29544806480407715,
      "num_input_tokens_seen": 1318213384,
      "step": 25620,
      "train_runtime": 157723.2441,
      "train_tokens_per_second": 8357.762
    },
    {
      "epoch": 1.7253568542957178,
      "grad_norm": 0.6704171706298899,
      "learning_rate": 4.583326056408505e-07,
      "loss": 0.2831497430801392,
      "num_input_tokens_seen": 1318459944,
      "step": 25625,
      "train_runtime": 157754.4767,
      "train_tokens_per_second": 8357.671
    },
    {
      "epoch": 1.7256935092916779,
      "grad_norm": 0.773449691157141,
      "learning_rate": 4.5722736212228613e-07,
      "loss": 0.302567195892334,
      "num_input_tokens_seen": 1318683160,
      "step": 25630,
      "train_runtime": 157780.5391,
      "train_tokens_per_second": 8357.705
    },
    {
      "epoch": 1.726030164287638,
      "grad_norm": 0.6551334740933409,
      "learning_rate": 4.5612338897560227e-07,
      "loss": 0.3076305627822876,
      "num_input_tokens_seen": 1318950240,
      "step": 25635,
      "train_runtime": 157811.7055,
      "train_tokens_per_second": 8357.747
    },
    {
      "epoch": 1.7263668192835981,
      "grad_norm": 0.7035323250666441,
      "learning_rate": 4.5502068650952527e-07,
      "loss": 0.3098435401916504,
      "num_input_tokens_seen": 1319204592,
      "step": 25640,
      "train_runtime": 157842.4742,
      "train_tokens_per_second": 8357.729
    },
    {
      "epoch": 1.7267034742795584,
      "grad_norm": 0.676116238611635,
      "learning_rate": 4.5391925503241894e-07,
      "loss": 0.30146336555480957,
      "num_input_tokens_seen": 1319478144,
      "step": 25645,
      "train_runtime": 157870.808,
      "train_tokens_per_second": 8357.962
    },
    {
      "epoch": 1.7270401292755184,
      "grad_norm": 0.7423332809538111,
      "learning_rate": 4.5281909485229704e-07,
      "loss": 0.2968470573425293,
      "num_input_tokens_seen": 1319727504,
      "step": 25650,
      "train_runtime": 157901.9295,
      "train_tokens_per_second": 8357.893
    },
    {
      "epoch": 1.7273767842714784,
      "grad_norm": 0.6425856001215274,
      "learning_rate": 4.51720206276815e-07,
      "loss": 0.25838408470153806,
      "num_input_tokens_seen": 1319978704,
      "step": 25655,
      "train_runtime": 157933.3466,
      "train_tokens_per_second": 8357.821
    },
    {
      "epoch": 1.7277134392674387,
      "grad_norm": 0.6494738638653509,
      "learning_rate": 4.506225896132743e-07,
      "loss": 0.2746278285980225,
      "num_input_tokens_seen": 1320228584,
      "step": 25660,
      "train_runtime": 157968.0889,
      "train_tokens_per_second": 8357.565
    },
    {
      "epoch": 1.728050094263399,
      "grad_norm": 0.6553455113799943,
      "learning_rate": 4.4952624516861944e-07,
      "loss": 0.29344930648803713,
      "num_input_tokens_seen": 1320473280,
      "step": 25665,
      "train_runtime": 157996.7609,
      "train_tokens_per_second": 8357.597
    },
    {
      "epoch": 1.728386749259359,
      "grad_norm": 0.6039977881488069,
      "learning_rate": 4.4843117324944005e-07,
      "loss": 0.2652871131896973,
      "num_input_tokens_seen": 1320772024,
      "step": 25670,
      "train_runtime": 158026.5459,
      "train_tokens_per_second": 8357.912
    },
    {
      "epoch": 1.728723404255319,
      "grad_norm": 0.7083737163313358,
      "learning_rate": 4.4733737416196745e-07,
      "loss": 0.29836416244506836,
      "num_input_tokens_seen": 1320978312,
      "step": 25675,
      "train_runtime": 158056.8691,
      "train_tokens_per_second": 8357.614
    },
    {
      "epoch": 1.7290600592512793,
      "grad_norm": 0.5778415436718939,
      "learning_rate": 4.4624484821208346e-07,
      "loss": 0.28352766036987304,
      "num_input_tokens_seen": 1321232968,
      "step": 25680,
      "train_runtime": 158088.8082,
      "train_tokens_per_second": 8357.536
    },
    {
      "epoch": 1.7293967142472395,
      "grad_norm": 0.6906342176270012,
      "learning_rate": 4.451535957053044e-07,
      "loss": 0.2656360149383545,
      "num_input_tokens_seen": 1321485648,
      "step": 25685,
      "train_runtime": 158118.1839,
      "train_tokens_per_second": 8357.582
    },
    {
      "epoch": 1.7297333692431995,
      "grad_norm": 0.702768003827997,
      "learning_rate": 4.440636169467999e-07,
      "loss": 0.29286558628082277,
      "num_input_tokens_seen": 1321731784,
      "step": 25690,
      "train_runtime": 158146.9484,
      "train_tokens_per_second": 8357.618
    },
    {
      "epoch": 1.7300700242391596,
      "grad_norm": 0.7395719764857467,
      "learning_rate": 4.429749122413757e-07,
      "loss": 0.28691964149475097,
      "num_input_tokens_seen": 1321985704,
      "step": 25695,
      "train_runtime": 158175.1548,
      "train_tokens_per_second": 8357.733
    },
    {
      "epoch": 1.7304066792351198,
      "grad_norm": 0.7156807561196736,
      "learning_rate": 4.418874818934865e-07,
      "loss": 0.2911983013153076,
      "num_input_tokens_seen": 1322247968,
      "step": 25700,
      "train_runtime": 158207.8583,
      "train_tokens_per_second": 8357.663
    },
    {
      "epoch": 1.73074333423108,
      "grad_norm": 0.759660902733546,
      "learning_rate": 4.4080132620722806e-07,
      "loss": 0.2999105930328369,
      "num_input_tokens_seen": 1322474688,
      "step": 25705,
      "train_runtime": 158236.7264,
      "train_tokens_per_second": 8357.571
    },
    {
      "epoch": 1.73107998922704,
      "grad_norm": 0.6515693455427687,
      "learning_rate": 4.3971644548634116e-07,
      "loss": 0.3027797222137451,
      "num_input_tokens_seen": 1322720096,
      "step": 25710,
      "train_runtime": 158269.2211,
      "train_tokens_per_second": 8357.406
    },
    {
      "epoch": 1.7314166442230001,
      "grad_norm": 0.8515381412421363,
      "learning_rate": 4.3863284003420826e-07,
      "loss": 0.314762020111084,
      "num_input_tokens_seen": 1322974584,
      "step": 25715,
      "train_runtime": 158302.6761,
      "train_tokens_per_second": 8357.247
    },
    {
      "epoch": 1.7317532992189604,
      "grad_norm": 0.6323774839958511,
      "learning_rate": 4.375505101538563e-07,
      "loss": 0.2825151920318604,
      "num_input_tokens_seen": 1323240832,
      "step": 25720,
      "train_runtime": 158330.3744,
      "train_tokens_per_second": 8357.467
    },
    {
      "epoch": 1.7320899542149206,
      "grad_norm": 0.6806260726778389,
      "learning_rate": 4.3646945614795487e-07,
      "loss": 0.31633195877075193,
      "num_input_tokens_seen": 1323512632,
      "step": 25725,
      "train_runtime": 158362.0829,
      "train_tokens_per_second": 8357.51
    },
    {
      "epoch": 1.7324266092108807,
      "grad_norm": 0.6382077585049454,
      "learning_rate": 4.3538967831881995e-07,
      "loss": 0.3103560209274292,
      "num_input_tokens_seen": 1323786552,
      "step": 25730,
      "train_runtime": 158398.7836,
      "train_tokens_per_second": 8357.303
    },
    {
      "epoch": 1.7327632642068407,
      "grad_norm": 0.671918477880032,
      "learning_rate": 4.343111769684039e-07,
      "loss": 0.295884108543396,
      "num_input_tokens_seen": 1324018840,
      "step": 25735,
      "train_runtime": 158428.4204,
      "train_tokens_per_second": 8357.205
    },
    {
      "epoch": 1.733099919202801,
      "grad_norm": 0.5862701494974917,
      "learning_rate": 4.332339523983103e-07,
      "loss": 0.27646489143371583,
      "num_input_tokens_seen": 1324288808,
      "step": 25740,
      "train_runtime": 158462.1058,
      "train_tokens_per_second": 8357.132
    },
    {
      "epoch": 1.7334365741987612,
      "grad_norm": 0.6420927772689585,
      "learning_rate": 4.321580049097773e-07,
      "loss": 0.2840766668319702,
      "num_input_tokens_seen": 1324562520,
      "step": 25745,
      "train_runtime": 158495.9926,
      "train_tokens_per_second": 8357.073
    },
    {
      "epoch": 1.7337732291947212,
      "grad_norm": 0.7373893330007768,
      "learning_rate": 4.310833348036941e-07,
      "loss": 0.27754693031311034,
      "num_input_tokens_seen": 1324808616,
      "step": 25750,
      "train_runtime": 158526.8552,
      "train_tokens_per_second": 8356.998
    },
    {
      "epoch": 1.7341098841906812,
      "grad_norm": 0.6275723484957894,
      "learning_rate": 4.300099423805865e-07,
      "loss": 0.3004399299621582,
      "num_input_tokens_seen": 1325040648,
      "step": 25755,
      "train_runtime": 158555.0734,
      "train_tokens_per_second": 8356.974
    },
    {
      "epoch": 1.7344465391866415,
      "grad_norm": 0.6393664346301751,
      "learning_rate": 4.289378279406264e-07,
      "loss": 0.30637011528015134,
      "num_input_tokens_seen": 1325311208,
      "step": 25760,
      "train_runtime": 158587.4122,
      "train_tokens_per_second": 8356.976
    },
    {
      "epoch": 1.7347831941826017,
      "grad_norm": 0.7685920611747342,
      "learning_rate": 4.2786699178362647e-07,
      "loss": 0.2978309869766235,
      "num_input_tokens_seen": 1325572656,
      "step": 25765,
      "train_runtime": 158616.3239,
      "train_tokens_per_second": 8357.101
    },
    {
      "epoch": 1.7351198491785618,
      "grad_norm": 0.6478123285345493,
      "learning_rate": 4.2679743420904254e-07,
      "loss": 0.3122110843658447,
      "num_input_tokens_seen": 1325844296,
      "step": 25770,
      "train_runtime": 158645.9465,
      "train_tokens_per_second": 8357.253
    },
    {
      "epoch": 1.7354565041745218,
      "grad_norm": 0.6805829079460136,
      "learning_rate": 4.257291555159726e-07,
      "loss": 0.25813138484954834,
      "num_input_tokens_seen": 1326088536,
      "step": 25775,
      "train_runtime": 158676.4538,
      "train_tokens_per_second": 8357.185
    },
    {
      "epoch": 1.735793159170482,
      "grad_norm": 0.5796258104235578,
      "learning_rate": 4.246621560031594e-07,
      "loss": 0.2914740562438965,
      "num_input_tokens_seen": 1326336288,
      "step": 25780,
      "train_runtime": 158704.2809,
      "train_tokens_per_second": 8357.281
    },
    {
      "epoch": 1.7361298141664423,
      "grad_norm": 0.6351225540838559,
      "learning_rate": 4.235964359689837e-07,
      "loss": 0.27538037300109863,
      "num_input_tokens_seen": 1326609976,
      "step": 25785,
      "train_runtime": 158733.7053,
      "train_tokens_per_second": 8357.456
    },
    {
      "epoch": 1.7364664691624023,
      "grad_norm": 0.6893069095435417,
      "learning_rate": 4.225319957114726e-07,
      "loss": 0.29334063529968263,
      "num_input_tokens_seen": 1326839432,
      "step": 25790,
      "train_runtime": 158760.7865,
      "train_tokens_per_second": 8357.476
    },
    {
      "epoch": 1.7368031241583624,
      "grad_norm": 0.674486289730744,
      "learning_rate": 4.214688355282909e-07,
      "loss": 0.2885859489440918,
      "num_input_tokens_seen": 1327088320,
      "step": 25795,
      "train_runtime": 158791.6923,
      "train_tokens_per_second": 8357.417
    },
    {
      "epoch": 1.7371397791543226,
      "grad_norm": 0.7362994210258667,
      "learning_rate": 4.204069557167506e-07,
      "loss": 0.2798640012741089,
      "num_input_tokens_seen": 1327338336,
      "step": 25800,
      "train_runtime": 158824.5678,
      "train_tokens_per_second": 8357.261
    },
    {
      "epoch": 1.7374764341502829,
      "grad_norm": 0.7958104213974944,
      "learning_rate": 4.1934635657380153e-07,
      "loss": 0.30298466682434083,
      "num_input_tokens_seen": 1327598744,
      "step": 25805,
      "train_runtime": 158855.2875,
      "train_tokens_per_second": 8357.284
    },
    {
      "epoch": 1.7378130891462429,
      "grad_norm": 0.6294091336966349,
      "learning_rate": 4.1828703839603637e-07,
      "loss": 0.2835930109024048,
      "num_input_tokens_seen": 1327877104,
      "step": 25810,
      "train_runtime": 158885.2756,
      "train_tokens_per_second": 8357.459
    },
    {
      "epoch": 1.738149744142203,
      "grad_norm": 0.7434950588703381,
      "learning_rate": 4.172290014796926e-07,
      "loss": 0.3295255184173584,
      "num_input_tokens_seen": 1328137168,
      "step": 25815,
      "train_runtime": 158913.0121,
      "train_tokens_per_second": 8357.636
    },
    {
      "epoch": 1.7384863991381632,
      "grad_norm": 0.7165963393336137,
      "learning_rate": 4.1617224612064353e-07,
      "loss": 0.2946991205215454,
      "num_input_tokens_seen": 1328378248,
      "step": 25820,
      "train_runtime": 158940.1072,
      "train_tokens_per_second": 8357.728
    },
    {
      "epoch": 1.7388230541341234,
      "grad_norm": 0.6007255426266559,
      "learning_rate": 4.151167726144101e-07,
      "loss": 0.28665313720703123,
      "num_input_tokens_seen": 1328664656,
      "step": 25825,
      "train_runtime": 158974.0757,
      "train_tokens_per_second": 8357.744
    },
    {
      "epoch": 1.7391597091300834,
      "grad_norm": 0.6625661520302867,
      "learning_rate": 4.1406258125615106e-07,
      "loss": 0.3052952289581299,
      "num_input_tokens_seen": 1328935136,
      "step": 25830,
      "train_runtime": 159004.3092,
      "train_tokens_per_second": 8357.856
    },
    {
      "epoch": 1.7394963641260435,
      "grad_norm": 0.6108525840558636,
      "learning_rate": 4.130096723406674e-07,
      "loss": 0.29311294555664064,
      "num_input_tokens_seen": 1329190776,
      "step": 25835,
      "train_runtime": 159039.1755,
      "train_tokens_per_second": 8357.631
    },
    {
      "epoch": 1.7398330191220037,
      "grad_norm": 0.7118298144269312,
      "learning_rate": 4.119580461624023e-07,
      "loss": 0.29893457889556885,
      "num_input_tokens_seen": 1329441448,
      "step": 25840,
      "train_runtime": 159073.7813,
      "train_tokens_per_second": 8357.389
    },
    {
      "epoch": 1.740169674117964,
      "grad_norm": 0.6178913201143932,
      "learning_rate": 4.109077030154396e-07,
      "loss": 0.2977069139480591,
      "num_input_tokens_seen": 1329700936,
      "step": 25845,
      "train_runtime": 159105.0027,
      "train_tokens_per_second": 8357.38
    },
    {
      "epoch": 1.740506329113924,
      "grad_norm": 0.7153496715260806,
      "learning_rate": 4.0985864319350366e-07,
      "loss": 0.298319149017334,
      "num_input_tokens_seen": 1329957864,
      "step": 25850,
      "train_runtime": 159133.6312,
      "train_tokens_per_second": 8357.491
    },
    {
      "epoch": 1.740842984109884,
      "grad_norm": 0.7499010228602133,
      "learning_rate": 4.0881086698996154e-07,
      "loss": 0.2970526933670044,
      "num_input_tokens_seen": 1330212128,
      "step": 25855,
      "train_runtime": 159166.0487,
      "train_tokens_per_second": 8357.386
    },
    {
      "epoch": 1.7411796391058443,
      "grad_norm": 0.5153937991338019,
      "learning_rate": 4.077643746978194e-07,
      "loss": 0.29752097129821775,
      "num_input_tokens_seen": 1330487424,
      "step": 25860,
      "train_runtime": 159196.9708,
      "train_tokens_per_second": 8357.492
    },
    {
      "epoch": 1.7415162941018045,
      "grad_norm": 0.5333314184106328,
      "learning_rate": 4.067191666097281e-07,
      "loss": 0.2894236326217651,
      "num_input_tokens_seen": 1330743096,
      "step": 25865,
      "train_runtime": 159227.8327,
      "train_tokens_per_second": 8357.478
    },
    {
      "epoch": 1.7418529490977646,
      "grad_norm": 0.6347719269272296,
      "learning_rate": 4.056752430179728e-07,
      "loss": 0.27895326614379884,
      "num_input_tokens_seen": 1331000912,
      "step": 25870,
      "train_runtime": 159257.6749,
      "train_tokens_per_second": 8357.531
    },
    {
      "epoch": 1.7421896040937246,
      "grad_norm": 0.6643251505840373,
      "learning_rate": 4.046326042144866e-07,
      "loss": 0.2896248817443848,
      "num_input_tokens_seen": 1331258568,
      "step": 25875,
      "train_runtime": 159289.3406,
      "train_tokens_per_second": 8357.487
    },
    {
      "epoch": 1.7425262590896848,
      "grad_norm": 0.6068873348777418,
      "learning_rate": 4.0359125049083956e-07,
      "loss": 0.2804113864898682,
      "num_input_tokens_seen": 1331517488,
      "step": 25880,
      "train_runtime": 159318.5198,
      "train_tokens_per_second": 8357.581
    },
    {
      "epoch": 1.742862914085645,
      "grad_norm": 0.6461057044150119,
      "learning_rate": 4.025511821382422e-07,
      "loss": 0.2966874837875366,
      "num_input_tokens_seen": 1331760504,
      "step": 25885,
      "train_runtime": 159346.4614,
      "train_tokens_per_second": 8357.641
    },
    {
      "epoch": 1.7431995690816051,
      "grad_norm": 0.6432559508708617,
      "learning_rate": 4.015123994475462e-07,
      "loss": 0.29219861030578614,
      "num_input_tokens_seen": 1331990608,
      "step": 25890,
      "train_runtime": 159377.6571,
      "train_tokens_per_second": 8357.449
    },
    {
      "epoch": 1.7435362240775654,
      "grad_norm": 0.7393370191630527,
      "learning_rate": 4.0047490270924417e-07,
      "loss": 0.29007601737976074,
      "num_input_tokens_seen": 1332291992,
      "step": 25895,
      "train_runtime": 159405.1599,
      "train_tokens_per_second": 8357.898
    },
    {
      "epoch": 1.7438728790735256,
      "grad_norm": 0.6349774288298972,
      "learning_rate": 3.994386922134691e-07,
      "loss": 0.29868102073669434,
      "num_input_tokens_seen": 1332560504,
      "step": 25900,
      "train_runtime": 159437.0688,
      "train_tokens_per_second": 8357.909
    },
    {
      "epoch": 1.7442095340694856,
      "grad_norm": 0.6196884596674892,
      "learning_rate": 3.984037682499936e-07,
      "loss": 0.281439995765686,
      "num_input_tokens_seen": 1332823528,
      "step": 25905,
      "train_runtime": 159467.081,
      "train_tokens_per_second": 8357.985
    },
    {
      "epoch": 1.7445461890654457,
      "grad_norm": 0.6171797271685329,
      "learning_rate": 3.973701311082301e-07,
      "loss": 0.31310460567474363,
      "num_input_tokens_seen": 1333070464,
      "step": 25910,
      "train_runtime": 159497.4116,
      "train_tokens_per_second": 8357.944
    },
    {
      "epoch": 1.744882844061406,
      "grad_norm": 0.8525036188274069,
      "learning_rate": 3.9633778107723454e-07,
      "loss": 0.29258713722229,
      "num_input_tokens_seen": 1333306032,
      "step": 25915,
      "train_runtime": 159527.688,
      "train_tokens_per_second": 8357.835
    },
    {
      "epoch": 1.7452194990573662,
      "grad_norm": 0.6774357587706222,
      "learning_rate": 3.953067184456966e-07,
      "loss": 0.31081037521362304,
      "num_input_tokens_seen": 1333553240,
      "step": 25920,
      "train_runtime": 159555.1971,
      "train_tokens_per_second": 8357.943
    },
    {
      "epoch": 1.7455561540533262,
      "grad_norm": 0.6872824240103099,
      "learning_rate": 3.9427694350195224e-07,
      "loss": 0.30258779525756835,
      "num_input_tokens_seen": 1333840472,
      "step": 25925,
      "train_runtime": 159583.7292,
      "train_tokens_per_second": 8358.249
    },
    {
      "epoch": 1.7458928090492862,
      "grad_norm": 0.6038876775990113,
      "learning_rate": 3.9324845653397446e-07,
      "loss": 0.30083909034729006,
      "num_input_tokens_seen": 1334107128,
      "step": 25930,
      "train_runtime": 159616.3816,
      "train_tokens_per_second": 8358.209
    },
    {
      "epoch": 1.7462294640452465,
      "grad_norm": 0.6502176493902306,
      "learning_rate": 3.922212578293755e-07,
      "loss": 0.27406558990478513,
      "num_input_tokens_seen": 1334348688,
      "step": 25935,
      "train_runtime": 159645.8537,
      "train_tokens_per_second": 8358.179
    },
    {
      "epoch": 1.7465661190412067,
      "grad_norm": 0.6242740166166935,
      "learning_rate": 3.911953476754088e-07,
      "loss": 0.2934547424316406,
      "num_input_tokens_seen": 1334619776,
      "step": 25940,
      "train_runtime": 159678.19,
      "train_tokens_per_second": 8358.185
    },
    {
      "epoch": 1.7469027740371668,
      "grad_norm": 0.760947127873944,
      "learning_rate": 3.9017072635896716e-07,
      "loss": 0.3102132797241211,
      "num_input_tokens_seen": 1334886792,
      "step": 25945,
      "train_runtime": 159710.8831,
      "train_tokens_per_second": 8358.145
    },
    {
      "epoch": 1.7472394290331268,
      "grad_norm": 0.6668003347450837,
      "learning_rate": 3.891473941665819e-07,
      "loss": 0.3055806875228882,
      "num_input_tokens_seen": 1335161648,
      "step": 25950,
      "train_runtime": 159743.8657,
      "train_tokens_per_second": 8358.14
    },
    {
      "epoch": 1.747576084029087,
      "grad_norm": 0.7780062077985637,
      "learning_rate": 3.881253513844246e-07,
      "loss": 0.3042929172515869,
      "num_input_tokens_seen": 1335401912,
      "step": 25955,
      "train_runtime": 159769.3511,
      "train_tokens_per_second": 8358.311
    },
    {
      "epoch": 1.7479127390250473,
      "grad_norm": 0.6581583636706244,
      "learning_rate": 3.871045982983063e-07,
      "loss": 0.28629069328308104,
      "num_input_tokens_seen": 1335656744,
      "step": 25960,
      "train_runtime": 159803.008,
      "train_tokens_per_second": 8358.145
    },
    {
      "epoch": 1.7482493940210073,
      "grad_norm": 0.672378725586984,
      "learning_rate": 3.8608513519367897e-07,
      "loss": 0.2794161319732666,
      "num_input_tokens_seen": 1335918576,
      "step": 25965,
      "train_runtime": 159837.766,
      "train_tokens_per_second": 8357.966
    },
    {
      "epoch": 1.7485860490169673,
      "grad_norm": 0.7423186305936402,
      "learning_rate": 3.8506696235562915e-07,
      "loss": 0.298835301399231,
      "num_input_tokens_seen": 1336166400,
      "step": 25970,
      "train_runtime": 159867.8893,
      "train_tokens_per_second": 8357.941
    },
    {
      "epoch": 1.7489227040129276,
      "grad_norm": 0.689435035843814,
      "learning_rate": 3.840500800688884e-07,
      "loss": 0.2886026859283447,
      "num_input_tokens_seen": 1336435352,
      "step": 25975,
      "train_runtime": 159898.019,
      "train_tokens_per_second": 8358.048
    },
    {
      "epoch": 1.7492593590088878,
      "grad_norm": 0.7230613985298435,
      "learning_rate": 3.8303448861782333e-07,
      "loss": 0.2952162504196167,
      "num_input_tokens_seen": 1336674920,
      "step": 25980,
      "train_runtime": 159928.3568,
      "train_tokens_per_second": 8357.961
    },
    {
      "epoch": 1.7495960140048479,
      "grad_norm": 0.7523500595490094,
      "learning_rate": 3.8202018828644105e-07,
      "loss": 0.28498640060424807,
      "num_input_tokens_seen": 1336923712,
      "step": 25985,
      "train_runtime": 159957.2648,
      "train_tokens_per_second": 8358.006
    },
    {
      "epoch": 1.749932669000808,
      "grad_norm": 0.7221552649784742,
      "learning_rate": 3.8100717935838804e-07,
      "loss": 0.3184696674346924,
      "num_input_tokens_seen": 1337170296,
      "step": 25990,
      "train_runtime": 159984.4018,
      "train_tokens_per_second": 8358.129
    },
    {
      "epoch": 1.7502693239967682,
      "grad_norm": 0.7130768027513776,
      "learning_rate": 3.799954621169477e-07,
      "loss": 0.28164589405059814,
      "num_input_tokens_seen": 1337431320,
      "step": 25995,
      "train_runtime": 160014.0799,
      "train_tokens_per_second": 8358.21
    },
    {
      "epoch": 1.7506059789927284,
      "grad_norm": 0.7423363104676663,
      "learning_rate": 3.7898503684504483e-07,
      "loss": 0.2978879690170288,
      "num_input_tokens_seen": 1337681392,
      "step": 26000,
      "train_runtime": 160047.1574,
      "train_tokens_per_second": 8358.045
    },
    {
      "epoch": 1.7509426339886884,
      "grad_norm": 0.635724260027889,
      "learning_rate": 3.7797590382524066e-07,
      "loss": 0.2901010036468506,
      "num_input_tokens_seen": 1337913456,
      "step": 26005,
      "train_runtime": 160078.77,
      "train_tokens_per_second": 8357.844
    },
    {
      "epoch": 1.7512792889846485,
      "grad_norm": 0.6220688821443228,
      "learning_rate": 3.76968063339736e-07,
      "loss": 0.29820914268493653,
      "num_input_tokens_seen": 1338202632,
      "step": 26010,
      "train_runtime": 160112.5379,
      "train_tokens_per_second": 8357.888
    },
    {
      "epoch": 1.7516159439806087,
      "grad_norm": 0.6614273045548646,
      "learning_rate": 3.759615156703722e-07,
      "loss": 0.2745553255081177,
      "num_input_tokens_seen": 1338453432,
      "step": 26015,
      "train_runtime": 160140.8085,
      "train_tokens_per_second": 8357.978
    },
    {
      "epoch": 1.751952598976569,
      "grad_norm": 0.6290129562712029,
      "learning_rate": 3.749562610986235e-07,
      "loss": 0.30714025497436526,
      "num_input_tokens_seen": 1338716112,
      "step": 26020,
      "train_runtime": 160167.5936,
      "train_tokens_per_second": 8358.221
    },
    {
      "epoch": 1.752289253972529,
      "grad_norm": 0.6577823315841561,
      "learning_rate": 3.739522999056094e-07,
      "loss": 0.277442479133606,
      "num_input_tokens_seen": 1338949136,
      "step": 26025,
      "train_runtime": 160201.2011,
      "train_tokens_per_second": 8357.922
    },
    {
      "epoch": 1.752625908968489,
      "grad_norm": 0.7375825472624327,
      "learning_rate": 3.7294963237208326e-07,
      "loss": 0.2933690071105957,
      "num_input_tokens_seen": 1339208744,
      "step": 26030,
      "train_runtime": 160227.0204,
      "train_tokens_per_second": 8358.195
    },
    {
      "epoch": 1.7529625639644493,
      "grad_norm": 0.7226864957850594,
      "learning_rate": 3.7194825877843786e-07,
      "loss": 0.28744142055511473,
      "num_input_tokens_seen": 1339468536,
      "step": 26035,
      "train_runtime": 160255.2625,
      "train_tokens_per_second": 8358.344
    },
    {
      "epoch": 1.7532992189604095,
      "grad_norm": 0.6428372078997174,
      "learning_rate": 3.7094817940470375e-07,
      "loss": 0.3043669700622559,
      "num_input_tokens_seen": 1339711568,
      "step": 26040,
      "train_runtime": 160285.5992,
      "train_tokens_per_second": 8358.278
    },
    {
      "epoch": 1.7536358739563696,
      "grad_norm": 0.6454554020316284,
      "learning_rate": 3.6994939453055044e-07,
      "loss": 0.2933466911315918,
      "num_input_tokens_seen": 1339980696,
      "step": 26045,
      "train_runtime": 160319.2745,
      "train_tokens_per_second": 8358.201
    },
    {
      "epoch": 1.7539725289523296,
      "grad_norm": 0.6983490098433837,
      "learning_rate": 3.6895190443528515e-07,
      "loss": 0.2914438009262085,
      "num_input_tokens_seen": 1340240880,
      "step": 26050,
      "train_runtime": 160346.9112,
      "train_tokens_per_second": 8358.383
    },
    {
      "epoch": 1.7543091839482898,
      "grad_norm": 0.6157056656737339,
      "learning_rate": 3.679557093978525e-07,
      "loss": 0.3060291290283203,
      "num_input_tokens_seen": 1340509872,
      "step": 26055,
      "train_runtime": 160374.6891,
      "train_tokens_per_second": 8358.612
    },
    {
      "epoch": 1.75464583894425,
      "grad_norm": 0.7149341006726245,
      "learning_rate": 3.669608096968341e-07,
      "loss": 0.28987536430358884,
      "num_input_tokens_seen": 1340742184,
      "step": 26060,
      "train_runtime": 160402.3079,
      "train_tokens_per_second": 8358.622
    },
    {
      "epoch": 1.75498249394021,
      "grad_norm": 0.6358804245303117,
      "learning_rate": 3.6596720561045295e-07,
      "loss": 0.29163172245025637,
      "num_input_tokens_seen": 1341019056,
      "step": 26065,
      "train_runtime": 160438.0094,
      "train_tokens_per_second": 8358.487
    },
    {
      "epoch": 1.7553191489361701,
      "grad_norm": 0.6298372942590204,
      "learning_rate": 3.6497489741656455e-07,
      "loss": 0.29502182006835936,
      "num_input_tokens_seen": 1341294912,
      "step": 26070,
      "train_runtime": 160466.155,
      "train_tokens_per_second": 8358.74
    },
    {
      "epoch": 1.7556558039321304,
      "grad_norm": 0.609302394078109,
      "learning_rate": 3.639838853926669e-07,
      "loss": 0.2910851001739502,
      "num_input_tokens_seen": 1341564776,
      "step": 26075,
      "train_runtime": 160501.3222,
      "train_tokens_per_second": 8358.59
    },
    {
      "epoch": 1.7559924589280906,
      "grad_norm": 0.632763515902012,
      "learning_rate": 3.629941698158923e-07,
      "loss": 0.30943498611450193,
      "num_input_tokens_seen": 1341822216,
      "step": 26080,
      "train_runtime": 160534.2643,
      "train_tokens_per_second": 8358.479
    },
    {
      "epoch": 1.7563291139240507,
      "grad_norm": 0.614162726041024,
      "learning_rate": 3.6200575096301207e-07,
      "loss": 0.28748140335083006,
      "num_input_tokens_seen": 1342077440,
      "step": 26085,
      "train_runtime": 160563.1973,
      "train_tokens_per_second": 8358.562
    },
    {
      "epoch": 1.7566657689200107,
      "grad_norm": 0.7686091209760204,
      "learning_rate": 3.61018629110434e-07,
      "loss": 0.31841301918029785,
      "num_input_tokens_seen": 1342353672,
      "step": 26090,
      "train_runtime": 160598.1646,
      "train_tokens_per_second": 8358.462
    },
    {
      "epoch": 1.757002423915971,
      "grad_norm": 0.5803069921493731,
      "learning_rate": 3.6003280453420386e-07,
      "loss": 0.27993144989013674,
      "num_input_tokens_seen": 1342616504,
      "step": 26095,
      "train_runtime": 160628.5304,
      "train_tokens_per_second": 8358.518
    },
    {
      "epoch": 1.7573390789119312,
      "grad_norm": 0.7066718748896264,
      "learning_rate": 3.59048277510004e-07,
      "loss": 0.2861770153045654,
      "num_input_tokens_seen": 1342874392,
      "step": 26100,
      "train_runtime": 160661.0838,
      "train_tokens_per_second": 8358.43
    },
    {
      "epoch": 1.7576757339078912,
      "grad_norm": 0.6826930598193696,
      "learning_rate": 3.5806504831315423e-07,
      "loss": 0.28505425453186034,
      "num_input_tokens_seen": 1343132480,
      "step": 26105,
      "train_runtime": 160690.6049,
      "train_tokens_per_second": 8358.5
    },
    {
      "epoch": 1.7580123889038513,
      "grad_norm": 0.6565966659136724,
      "learning_rate": 3.570831172186112e-07,
      "loss": 0.3115654945373535,
      "num_input_tokens_seen": 1343395856,
      "step": 26110,
      "train_runtime": 160718.8069,
      "train_tokens_per_second": 8358.672
    },
    {
      "epoch": 1.7583490438998115,
      "grad_norm": 0.6955866530465504,
      "learning_rate": 3.5610248450097084e-07,
      "loss": 0.30111353397369384,
      "num_input_tokens_seen": 1343646720,
      "step": 26115,
      "train_runtime": 160744.9624,
      "train_tokens_per_second": 8358.873
    },
    {
      "epoch": 1.7586856988957718,
      "grad_norm": 0.5826636846347298,
      "learning_rate": 3.551231504344604e-07,
      "loss": 0.27691121101379396,
      "num_input_tokens_seen": 1343910808,
      "step": 26120,
      "train_runtime": 160772.5994,
      "train_tokens_per_second": 8359.079
    },
    {
      "epoch": 1.7590223538917318,
      "grad_norm": 0.5943421143097994,
      "learning_rate": 3.5414511529295036e-07,
      "loss": 0.3096380949020386,
      "num_input_tokens_seen": 1344171824,
      "step": 26125,
      "train_runtime": 160807.1015,
      "train_tokens_per_second": 8358.908
    },
    {
      "epoch": 1.7593590088876918,
      "grad_norm": 0.7208633863296395,
      "learning_rate": 3.5316837934994465e-07,
      "loss": 0.31474230289459226,
      "num_input_tokens_seen": 1344437368,
      "step": 26130,
      "train_runtime": 160836.6229,
      "train_tokens_per_second": 8359.025
    },
    {
      "epoch": 1.759695663883652,
      "grad_norm": 0.6608315185374786,
      "learning_rate": 3.521929428785836e-07,
      "loss": 0.299714994430542,
      "num_input_tokens_seen": 1344713704,
      "step": 26135,
      "train_runtime": 160865.5468,
      "train_tokens_per_second": 8359.24
    },
    {
      "epoch": 1.7600323188796123,
      "grad_norm": 0.6771044873697255,
      "learning_rate": 3.5121880615164584e-07,
      "loss": 0.31691908836364746,
      "num_input_tokens_seen": 1344966792,
      "step": 26140,
      "train_runtime": 160894.425,
      "train_tokens_per_second": 8359.313
    },
    {
      "epoch": 1.7603689738755723,
      "grad_norm": 0.6313719827054307,
      "learning_rate": 3.5024596944154445e-07,
      "loss": 0.3013108253479004,
      "num_input_tokens_seen": 1345217760,
      "step": 26145,
      "train_runtime": 160926.7618,
      "train_tokens_per_second": 8359.192
    },
    {
      "epoch": 1.7607056288715324,
      "grad_norm": 0.7549681255252368,
      "learning_rate": 3.4927443302033127e-07,
      "loss": 0.27835588455200194,
      "num_input_tokens_seen": 1345479400,
      "step": 26150,
      "train_runtime": 160954.8125,
      "train_tokens_per_second": 8359.361
    },
    {
      "epoch": 1.7610422838674926,
      "grad_norm": 0.6833786727646106,
      "learning_rate": 3.4830419715969233e-07,
      "loss": 0.2756466865539551,
      "num_input_tokens_seen": 1345731608,
      "step": 26155,
      "train_runtime": 160984.1043,
      "train_tokens_per_second": 8359.407
    },
    {
      "epoch": 1.7613789388634529,
      "grad_norm": 0.7572691825130193,
      "learning_rate": 3.4733526213095114e-07,
      "loss": 0.29552474021911623,
      "num_input_tokens_seen": 1345983568,
      "step": 26160,
      "train_runtime": 161011.9583,
      "train_tokens_per_second": 8359.525
    },
    {
      "epoch": 1.761715593859413,
      "grad_norm": 0.6998814637279851,
      "learning_rate": 3.4636762820506876e-07,
      "loss": 0.2991434097290039,
      "num_input_tokens_seen": 1346237984,
      "step": 26165,
      "train_runtime": 161041.7791,
      "train_tokens_per_second": 8359.557
    },
    {
      "epoch": 1.762052248855373,
      "grad_norm": 0.8465319496253078,
      "learning_rate": 3.4540129565263867e-07,
      "loss": 0.30602202415466306,
      "num_input_tokens_seen": 1346482880,
      "step": 26170,
      "train_runtime": 161073.1055,
      "train_tokens_per_second": 8359.452
    },
    {
      "epoch": 1.7623889038513332,
      "grad_norm": 0.6757035810867636,
      "learning_rate": 3.444362647438942e-07,
      "loss": 0.29701688289642336,
      "num_input_tokens_seen": 1346753504,
      "step": 26175,
      "train_runtime": 161103.1717,
      "train_tokens_per_second": 8359.572
    },
    {
      "epoch": 1.7627255588472934,
      "grad_norm": 0.6457375646529487,
      "learning_rate": 3.434725357487029e-07,
      "loss": 0.30782666206359866,
      "num_input_tokens_seen": 1347026312,
      "step": 26180,
      "train_runtime": 161135.0485,
      "train_tokens_per_second": 8359.611
    },
    {
      "epoch": 1.7630622138432535,
      "grad_norm": 0.7533737492037593,
      "learning_rate": 3.4251010893656844e-07,
      "loss": 0.326169753074646,
      "num_input_tokens_seen": 1347287880,
      "step": 26185,
      "train_runtime": 161163.9742,
      "train_tokens_per_second": 8359.734
    },
    {
      "epoch": 1.7633988688392135,
      "grad_norm": 0.70727613066161,
      "learning_rate": 3.4154898457663065e-07,
      "loss": 0.29431443214416503,
      "num_input_tokens_seen": 1347532104,
      "step": 26190,
      "train_runtime": 161193.7457,
      "train_tokens_per_second": 8359.705
    },
    {
      "epoch": 1.7637355238351737,
      "grad_norm": 0.6774049056283339,
      "learning_rate": 3.40589162937664e-07,
      "loss": 0.2965696811676025,
      "num_input_tokens_seen": 1347751224,
      "step": 26195,
      "train_runtime": 161220.8005,
      "train_tokens_per_second": 8359.661
    },
    {
      "epoch": 1.764072178831134,
      "grad_norm": 0.6843391537925229,
      "learning_rate": 3.3963064428808034e-07,
      "loss": 0.27552683353424073,
      "num_input_tokens_seen": 1347988528,
      "step": 26200,
      "train_runtime": 161250.2481,
      "train_tokens_per_second": 8359.606
    },
    {
      "epoch": 1.764408833827094,
      "grad_norm": 0.6091266474018148,
      "learning_rate": 3.386734288959265e-07,
      "loss": 0.30062198638916016,
      "num_input_tokens_seen": 1348203096,
      "step": 26205,
      "train_runtime": 161275.4977,
      "train_tokens_per_second": 8359.628
    },
    {
      "epoch": 1.764745488823054,
      "grad_norm": 0.6650496967778534,
      "learning_rate": 3.377175170288838e-07,
      "loss": 0.30403034687042235,
      "num_input_tokens_seen": 1348469960,
      "step": 26210,
      "train_runtime": 161308.695,
      "train_tokens_per_second": 8359.562
    },
    {
      "epoch": 1.7650821438190143,
      "grad_norm": 0.6527643305299018,
      "learning_rate": 3.367629089542712e-07,
      "loss": 0.2928946018218994,
      "num_input_tokens_seen": 1348737264,
      "step": 26215,
      "train_runtime": 161339.2469,
      "train_tokens_per_second": 8359.635
    },
    {
      "epoch": 1.7654187988149745,
      "grad_norm": 0.7307395416886756,
      "learning_rate": 3.3580960493904024e-07,
      "loss": 0.3066680431365967,
      "num_input_tokens_seen": 1348990568,
      "step": 26220,
      "train_runtime": 161369.0036,
      "train_tokens_per_second": 8359.663
    },
    {
      "epoch": 1.7657554538109346,
      "grad_norm": 0.7083323429786569,
      "learning_rate": 3.3485760524978093e-07,
      "loss": 0.3145547866821289,
      "num_input_tokens_seen": 1349239216,
      "step": 26225,
      "train_runtime": 161397.9881,
      "train_tokens_per_second": 8359.703
    },
    {
      "epoch": 1.7660921088068946,
      "grad_norm": 0.714037254060508,
      "learning_rate": 3.3390691015271646e-07,
      "loss": 0.2972515106201172,
      "num_input_tokens_seen": 1349494768,
      "step": 26230,
      "train_runtime": 161429.5546,
      "train_tokens_per_second": 8359.651
    },
    {
      "epoch": 1.7664287638028549,
      "grad_norm": 0.7330955720075035,
      "learning_rate": 3.329575199137053e-07,
      "loss": 0.28320086002349854,
      "num_input_tokens_seen": 1349756624,
      "step": 26235,
      "train_runtime": 161459.7669,
      "train_tokens_per_second": 8359.709
    },
    {
      "epoch": 1.766765418798815,
      "grad_norm": 0.7402490246999724,
      "learning_rate": 3.3200943479824113e-07,
      "loss": 0.3181619167327881,
      "num_input_tokens_seen": 1350013000,
      "step": 26240,
      "train_runtime": 161490.5444,
      "train_tokens_per_second": 8359.703
    },
    {
      "epoch": 1.7671020737947751,
      "grad_norm": 0.5560532932211901,
      "learning_rate": 3.3106265507145406e-07,
      "loss": 0.25999748706817627,
      "num_input_tokens_seen": 1350270952,
      "step": 26245,
      "train_runtime": 161524.1111,
      "train_tokens_per_second": 8359.563
    },
    {
      "epoch": 1.7674387287907352,
      "grad_norm": 0.6617291993887443,
      "learning_rate": 3.3011718099810687e-07,
      "loss": 0.28427298069000245,
      "num_input_tokens_seen": 1350516464,
      "step": 26250,
      "train_runtime": 161553.1615,
      "train_tokens_per_second": 8359.579
    },
    {
      "epoch": 1.7677753837866954,
      "grad_norm": 0.6462073249148715,
      "learning_rate": 3.2917301284259904e-07,
      "loss": 0.2877047538757324,
      "num_input_tokens_seen": 1350792656,
      "step": 26255,
      "train_runtime": 161583.3314,
      "train_tokens_per_second": 8359.728
    },
    {
      "epoch": 1.7681120387826557,
      "grad_norm": 0.6238342836887955,
      "learning_rate": 3.2823015086896335e-07,
      "loss": 0.3080324172973633,
      "num_input_tokens_seen": 1351056256,
      "step": 26260,
      "train_runtime": 161616.0432,
      "train_tokens_per_second": 8359.667
    },
    {
      "epoch": 1.7684486937786157,
      "grad_norm": 0.6786572884022523,
      "learning_rate": 3.2728859534087e-07,
      "loss": 0.29061193466186525,
      "num_input_tokens_seen": 1351317424,
      "step": 26265,
      "train_runtime": 161643.9565,
      "train_tokens_per_second": 8359.839
    },
    {
      "epoch": 1.7687853487745757,
      "grad_norm": 0.7186532745028049,
      "learning_rate": 3.263483465216194e-07,
      "loss": 0.30810601711273194,
      "num_input_tokens_seen": 1351603016,
      "step": 26270,
      "train_runtime": 161679.6112,
      "train_tokens_per_second": 8359.762
    },
    {
      "epoch": 1.769122003770536,
      "grad_norm": 0.6226063503137219,
      "learning_rate": 3.2540940467415126e-07,
      "loss": 0.28761563301086424,
      "num_input_tokens_seen": 1351855744,
      "step": 26275,
      "train_runtime": 161712.5569,
      "train_tokens_per_second": 8359.621
    },
    {
      "epoch": 1.7694586587664962,
      "grad_norm": 0.616939056065546,
      "learning_rate": 3.2447177006103726e-07,
      "loss": 0.2969227313995361,
      "num_input_tokens_seen": 1352102528,
      "step": 26280,
      "train_runtime": 161743.4974,
      "train_tokens_per_second": 8359.548
    },
    {
      "epoch": 1.7697953137624562,
      "grad_norm": 0.6597648470942332,
      "learning_rate": 3.235354429444831e-07,
      "loss": 0.2906570672988892,
      "num_input_tokens_seen": 1352341648,
      "step": 26285,
      "train_runtime": 161769.3582,
      "train_tokens_per_second": 8359.69
    },
    {
      "epoch": 1.7701319687584163,
      "grad_norm": 0.7033296734242606,
      "learning_rate": 3.226004235863306e-07,
      "loss": 0.2992046356201172,
      "num_input_tokens_seen": 1352589112,
      "step": 26290,
      "train_runtime": 161798.2669,
      "train_tokens_per_second": 8359.726
    },
    {
      "epoch": 1.7704686237543765,
      "grad_norm": 0.7301168696319766,
      "learning_rate": 3.2166671224805437e-07,
      "loss": 0.2927102565765381,
      "num_input_tokens_seen": 1352862392,
      "step": 26295,
      "train_runtime": 161833.6674,
      "train_tokens_per_second": 8359.586
    },
    {
      "epoch": 1.7708052787503368,
      "grad_norm": 0.6865672683032009,
      "learning_rate": 3.207343091907633e-07,
      "loss": 0.288381290435791,
      "num_input_tokens_seen": 1353097552,
      "step": 26300,
      "train_runtime": 161868.2024,
      "train_tokens_per_second": 8359.255
    },
    {
      "epoch": 1.7711419337462968,
      "grad_norm": 0.7484313373401098,
      "learning_rate": 3.1980321467520226e-07,
      "loss": 0.2924871206283569,
      "num_input_tokens_seen": 1353323160,
      "step": 26305,
      "train_runtime": 161897.969,
      "train_tokens_per_second": 8359.111
    },
    {
      "epoch": 1.7714785887422568,
      "grad_norm": 0.7750237513177343,
      "learning_rate": 3.1887342896174735e-07,
      "loss": 0.30837421417236327,
      "num_input_tokens_seen": 1353546704,
      "step": 26310,
      "train_runtime": 161926.5018,
      "train_tokens_per_second": 8359.019
    },
    {
      "epoch": 1.771815243738217,
      "grad_norm": 0.6017880443372428,
      "learning_rate": 3.1794495231041276e-07,
      "loss": 0.2868243932723999,
      "num_input_tokens_seen": 1353805224,
      "step": 26315,
      "train_runtime": 161955.1198,
      "train_tokens_per_second": 8359.138
    },
    {
      "epoch": 1.7721518987341773,
      "grad_norm": 0.74483230427819,
      "learning_rate": 3.1701778498084035e-07,
      "loss": 0.31068806648254393,
      "num_input_tokens_seen": 1354038288,
      "step": 26320,
      "train_runtime": 161986.3984,
      "train_tokens_per_second": 8358.963
    },
    {
      "epoch": 1.7724885537301374,
      "grad_norm": 0.6619973570850624,
      "learning_rate": 3.160919272323121e-07,
      "loss": 0.31393444538116455,
      "num_input_tokens_seen": 1354309968,
      "step": 26325,
      "train_runtime": 162012.7417,
      "train_tokens_per_second": 8359.281
    },
    {
      "epoch": 1.7728252087260974,
      "grad_norm": 0.6418288018798243,
      "learning_rate": 3.1516737932374087e-07,
      "loss": 0.3026572227478027,
      "num_input_tokens_seen": 1354563576,
      "step": 26330,
      "train_runtime": 162041.6411,
      "train_tokens_per_second": 8359.355
    },
    {
      "epoch": 1.7731618637220576,
      "grad_norm": 0.609871000308674,
      "learning_rate": 3.142441415136727e-07,
      "loss": 0.2854583501815796,
      "num_input_tokens_seen": 1354851232,
      "step": 26335,
      "train_runtime": 162075.7264,
      "train_tokens_per_second": 8359.372
    },
    {
      "epoch": 1.773498518718018,
      "grad_norm": 0.647794847843074,
      "learning_rate": 3.133222140602893e-07,
      "loss": 0.28263752460479735,
      "num_input_tokens_seen": 1355085360,
      "step": 26340,
      "train_runtime": 162102.0491,
      "train_tokens_per_second": 8359.459
    },
    {
      "epoch": 1.773835173713978,
      "grad_norm": 0.6463326375230871,
      "learning_rate": 3.124015972214034e-07,
      "loss": 0.303174090385437,
      "num_input_tokens_seen": 1355355960,
      "step": 26345,
      "train_runtime": 162135.7812,
      "train_tokens_per_second": 8359.388
    },
    {
      "epoch": 1.774171828709938,
      "grad_norm": 0.7043469258666549,
      "learning_rate": 3.114822912544635e-07,
      "loss": 0.31141347885131837,
      "num_input_tokens_seen": 1355594096,
      "step": 26350,
      "train_runtime": 162167.7319,
      "train_tokens_per_second": 8359.21
    },
    {
      "epoch": 1.7745084837058982,
      "grad_norm": 0.6243547491252317,
      "learning_rate": 3.1056429641655007e-07,
      "loss": 0.28743271827697753,
      "num_input_tokens_seen": 1355826832,
      "step": 26355,
      "train_runtime": 162196.7593,
      "train_tokens_per_second": 8359.149
    },
    {
      "epoch": 1.7748451387018584,
      "grad_norm": 0.6981146736512425,
      "learning_rate": 3.0964761296437665e-07,
      "loss": 0.2846538066864014,
      "num_input_tokens_seen": 1356094152,
      "step": 26360,
      "train_runtime": 162230.5071,
      "train_tokens_per_second": 8359.058
    },
    {
      "epoch": 1.7751817936978185,
      "grad_norm": 0.7246966105720261,
      "learning_rate": 3.087322411542937e-07,
      "loss": 0.30587449073791506,
      "num_input_tokens_seen": 1356355104,
      "step": 26365,
      "train_runtime": 162261.7414,
      "train_tokens_per_second": 8359.057
    },
    {
      "epoch": 1.7755184486937785,
      "grad_norm": 0.6771860855755931,
      "learning_rate": 3.0781818124227867e-07,
      "loss": 0.27478020191192626,
      "num_input_tokens_seen": 1356600752,
      "step": 26370,
      "train_runtime": 162292.3196,
      "train_tokens_per_second": 8358.995
    },
    {
      "epoch": 1.7758551036897388,
      "grad_norm": 0.6768268804229209,
      "learning_rate": 3.0690543348394776e-07,
      "loss": 0.29021215438842773,
      "num_input_tokens_seen": 1356863824,
      "step": 26375,
      "train_runtime": 162325.0282,
      "train_tokens_per_second": 8358.932
    },
    {
      "epoch": 1.776191758685699,
      "grad_norm": 0.5904130049320521,
      "learning_rate": 3.059939981345467e-07,
      "loss": 0.2668785095214844,
      "num_input_tokens_seen": 1357130352,
      "step": 26380,
      "train_runtime": 162354.1694,
      "train_tokens_per_second": 8359.073
    },
    {
      "epoch": 1.776528413681659,
      "grad_norm": 0.6965135073528634,
      "learning_rate": 3.050838754489566e-07,
      "loss": 0.28300886154174804,
      "num_input_tokens_seen": 1357385304,
      "step": 26385,
      "train_runtime": 162380.0726,
      "train_tokens_per_second": 8359.31
    },
    {
      "epoch": 1.776865068677619,
      "grad_norm": 0.6516154874506989,
      "learning_rate": 3.041750656816894e-07,
      "loss": 0.2924177646636963,
      "num_input_tokens_seen": 1357637832,
      "step": 26390,
      "train_runtime": 162415.6855,
      "train_tokens_per_second": 8359.032
    },
    {
      "epoch": 1.7772017236735793,
      "grad_norm": 0.6979500319596583,
      "learning_rate": 3.0326756908689135e-07,
      "loss": 0.30799920558929444,
      "num_input_tokens_seen": 1357905008,
      "step": 26395,
      "train_runtime": 162447.5097,
      "train_tokens_per_second": 8359.039
    },
    {
      "epoch": 1.7775383786695396,
      "grad_norm": 0.5919015521771285,
      "learning_rate": 3.0236138591833995e-07,
      "loss": 0.26824440956115725,
      "num_input_tokens_seen": 1358172992,
      "step": 26400,
      "train_runtime": 162479.4077,
      "train_tokens_per_second": 8359.047
    },
    {
      "epoch": 1.7778750336654996,
      "grad_norm": 0.6701968006293861,
      "learning_rate": 3.0145651642944863e-07,
      "loss": 0.31132969856262205,
      "num_input_tokens_seen": 1358441208,
      "step": 26405,
      "train_runtime": 162510.4769,
      "train_tokens_per_second": 8359.099
    },
    {
      "epoch": 1.7782116886614596,
      "grad_norm": 0.783662506375726,
      "learning_rate": 3.005529608732588e-07,
      "loss": 0.274139928817749,
      "num_input_tokens_seen": 1358690648,
      "step": 26410,
      "train_runtime": 162543.7088,
      "train_tokens_per_second": 8358.925
    },
    {
      "epoch": 1.7785483436574199,
      "grad_norm": 0.642705316862753,
      "learning_rate": 2.996507195024495e-07,
      "loss": 0.3140422344207764,
      "num_input_tokens_seen": 1358965536,
      "step": 26415,
      "train_runtime": 162573.3205,
      "train_tokens_per_second": 8359.093
    },
    {
      "epoch": 1.7788849986533801,
      "grad_norm": 0.7355615148061111,
      "learning_rate": 2.9874979256932614e-07,
      "loss": 0.2955437898635864,
      "num_input_tokens_seen": 1359234696,
      "step": 26420,
      "train_runtime": 162603.582,
      "train_tokens_per_second": 8359.193
    },
    {
      "epoch": 1.7792216536493402,
      "grad_norm": 0.6801729146129405,
      "learning_rate": 2.9785018032583325e-07,
      "loss": 0.31080672740936277,
      "num_input_tokens_seen": 1359489264,
      "step": 26425,
      "train_runtime": 162636.8133,
      "train_tokens_per_second": 8359.05
    },
    {
      "epoch": 1.7795583086453002,
      "grad_norm": 0.713237018850876,
      "learning_rate": 2.969518830235435e-07,
      "loss": 0.3060844659805298,
      "num_input_tokens_seen": 1359753208,
      "step": 26430,
      "train_runtime": 162670.1856,
      "train_tokens_per_second": 8358.958
    },
    {
      "epoch": 1.7798949636412604,
      "grad_norm": 0.6853028451178209,
      "learning_rate": 2.960549009136626e-07,
      "loss": 0.30226118564605714,
      "num_input_tokens_seen": 1359998016,
      "step": 26435,
      "train_runtime": 162704.2448,
      "train_tokens_per_second": 8358.713
    },
    {
      "epoch": 1.7802316186372207,
      "grad_norm": 0.7000964570773769,
      "learning_rate": 2.951592342470289e-07,
      "loss": 0.32621045112609864,
      "num_input_tokens_seen": 1360278736,
      "step": 26440,
      "train_runtime": 162738.5324,
      "train_tokens_per_second": 8358.676
    },
    {
      "epoch": 1.7805682736331807,
      "grad_norm": 0.6116208269975089,
      "learning_rate": 2.942648832741124e-07,
      "loss": 0.2819814205169678,
      "num_input_tokens_seen": 1360510136,
      "step": 26445,
      "train_runtime": 162768.2726,
      "train_tokens_per_second": 8358.571
    },
    {
      "epoch": 1.7809049286291407,
      "grad_norm": 0.579691022178006,
      "learning_rate": 2.933718482450154e-07,
      "loss": 0.28756370544433596,
      "num_input_tokens_seen": 1360773448,
      "step": 26450,
      "train_runtime": 162796.4665,
      "train_tokens_per_second": 8358.741
    },
    {
      "epoch": 1.781241583625101,
      "grad_norm": 0.7023713646733455,
      "learning_rate": 2.9248012940947423e-07,
      "loss": 0.3132743835449219,
      "num_input_tokens_seen": 1361053968,
      "step": 26455,
      "train_runtime": 162825.9201,
      "train_tokens_per_second": 8358.951
    },
    {
      "epoch": 1.7815782386210612,
      "grad_norm": 0.6810203108350382,
      "learning_rate": 2.915897270168522e-07,
      "loss": 0.2973148345947266,
      "num_input_tokens_seen": 1361295048,
      "step": 26460,
      "train_runtime": 162852.9392,
      "train_tokens_per_second": 8359.045
    },
    {
      "epoch": 1.7819148936170213,
      "grad_norm": 0.7450756977222269,
      "learning_rate": 2.9070064131615004e-07,
      "loss": 0.3071122646331787,
      "num_input_tokens_seen": 1361532640,
      "step": 26465,
      "train_runtime": 162881.6903,
      "train_tokens_per_second": 8359.028
    },
    {
      "epoch": 1.7822515486129813,
      "grad_norm": 0.7172742541035068,
      "learning_rate": 2.8981287255599567e-07,
      "loss": 0.294301438331604,
      "num_input_tokens_seen": 1361742088,
      "step": 26470,
      "train_runtime": 162911.6064,
      "train_tokens_per_second": 8358.779
    },
    {
      "epoch": 1.7825882036089415,
      "grad_norm": 0.6719903872886102,
      "learning_rate": 2.8892642098465205e-07,
      "loss": 0.2995366811752319,
      "num_input_tokens_seen": 1361998752,
      "step": 26475,
      "train_runtime": 162942.6536,
      "train_tokens_per_second": 8358.761
    },
    {
      "epoch": 1.7829248586049018,
      "grad_norm": 0.6472984822272575,
      "learning_rate": 2.8804128685001263e-07,
      "loss": 0.2926197052001953,
      "num_input_tokens_seen": 1362275624,
      "step": 26480,
      "train_runtime": 162974.4772,
      "train_tokens_per_second": 8358.828
    },
    {
      "epoch": 1.7832615136008618,
      "grad_norm": 0.7700820344435674,
      "learning_rate": 2.8715747039960216e-07,
      "loss": 0.2760331153869629,
      "num_input_tokens_seen": 1362515848,
      "step": 26485,
      "train_runtime": 163003.2964,
      "train_tokens_per_second": 8358.824
    },
    {
      "epoch": 1.7835981685968219,
      "grad_norm": 0.7197819584875104,
      "learning_rate": 2.8627497188057684e-07,
      "loss": 0.29818997383117674,
      "num_input_tokens_seen": 1362770312,
      "step": 26490,
      "train_runtime": 163035.1143,
      "train_tokens_per_second": 8358.753
    },
    {
      "epoch": 1.783934823592782,
      "grad_norm": 0.660831789840294,
      "learning_rate": 2.853937915397248e-07,
      "loss": 0.32532808780670164,
      "num_input_tokens_seen": 1363021856,
      "step": 26495,
      "train_runtime": 163065.321,
      "train_tokens_per_second": 8358.748
    },
    {
      "epoch": 1.7842714785887424,
      "grad_norm": 0.696359885713326,
      "learning_rate": 2.845139296234639e-07,
      "loss": 0.2761091232299805,
      "num_input_tokens_seen": 1363298264,
      "step": 26500,
      "train_runtime": 163097.2907,
      "train_tokens_per_second": 8358.804
    },
    {
      "epoch": 1.7846081335847024,
      "grad_norm": 0.6508649050997728,
      "learning_rate": 2.836353863778479e-07,
      "loss": 0.2874312400817871,
      "num_input_tokens_seen": 1363553944,
      "step": 26505,
      "train_runtime": 163129.167,
      "train_tokens_per_second": 8358.738
    },
    {
      "epoch": 1.7849447885806624,
      "grad_norm": 0.5671582819902711,
      "learning_rate": 2.8275816204855534e-07,
      "loss": 0.2974745750427246,
      "num_input_tokens_seen": 1363795160,
      "step": 26510,
      "train_runtime": 163160.1038,
      "train_tokens_per_second": 8358.631
    },
    {
      "epoch": 1.7852814435766227,
      "grad_norm": 0.6765996026421778,
      "learning_rate": 2.818822568809015e-07,
      "loss": 0.317954421043396,
      "num_input_tokens_seen": 1364037880,
      "step": 26515,
      "train_runtime": 163191.8871,
      "train_tokens_per_second": 8358.491
    },
    {
      "epoch": 1.785618098572583,
      "grad_norm": 0.5601671218730068,
      "learning_rate": 2.810076711198284e-07,
      "loss": 0.28429155349731444,
      "num_input_tokens_seen": 1364287080,
      "step": 26520,
      "train_runtime": 163220.5637,
      "train_tokens_per_second": 8358.549
    },
    {
      "epoch": 1.785954753568543,
      "grad_norm": 0.6423177319961557,
      "learning_rate": 2.801344050099131e-07,
      "loss": 0.3109766721725464,
      "num_input_tokens_seen": 1364575888,
      "step": 26525,
      "train_runtime": 163254.4577,
      "train_tokens_per_second": 8358.583
    },
    {
      "epoch": 1.786291408564503,
      "grad_norm": 0.6401253078758091,
      "learning_rate": 2.7926245879536076e-07,
      "loss": 0.29339141845703126,
      "num_input_tokens_seen": 1364838800,
      "step": 26530,
      "train_runtime": 163285.1244,
      "train_tokens_per_second": 8358.623
    },
    {
      "epoch": 1.7866280635604632,
      "grad_norm": 0.6505546091532378,
      "learning_rate": 2.7839183272000857e-07,
      "loss": 0.28722617626190183,
      "num_input_tokens_seen": 1365125568,
      "step": 26535,
      "train_runtime": 163313.2563,
      "train_tokens_per_second": 8358.939
    },
    {
      "epoch": 1.7869647185564235,
      "grad_norm": 0.7451032875949253,
      "learning_rate": 2.775225270273241e-07,
      "loss": 0.3065593719482422,
      "num_input_tokens_seen": 1365403496,
      "step": 26540,
      "train_runtime": 163341.9199,
      "train_tokens_per_second": 8359.174
    },
    {
      "epoch": 1.7873013735523835,
      "grad_norm": 0.5491026442076314,
      "learning_rate": 2.7665454196040665e-07,
      "loss": 0.264717173576355,
      "num_input_tokens_seen": 1365678240,
      "step": 26545,
      "train_runtime": 163369.8107,
      "train_tokens_per_second": 8359.428
    },
    {
      "epoch": 1.7876380285483435,
      "grad_norm": 0.724656475117513,
      "learning_rate": 2.7578787776198325e-07,
      "loss": 0.3115692138671875,
      "num_input_tokens_seen": 1365920640,
      "step": 26550,
      "train_runtime": 163403.052,
      "train_tokens_per_second": 8359.211
    },
    {
      "epoch": 1.7879746835443038,
      "grad_norm": 0.5986272575492035,
      "learning_rate": 2.7492253467441775e-07,
      "loss": 0.2747222423553467,
      "num_input_tokens_seen": 1366183600,
      "step": 26555,
      "train_runtime": 163434.6776,
      "train_tokens_per_second": 8359.203
    },
    {
      "epoch": 1.788311338540264,
      "grad_norm": 0.6411277422851874,
      "learning_rate": 2.740585129396961e-07,
      "loss": 0.29883060455322263,
      "num_input_tokens_seen": 1366445328,
      "step": 26560,
      "train_runtime": 163466.2464,
      "train_tokens_per_second": 8359.189
    },
    {
      "epoch": 1.788647993536224,
      "grad_norm": 0.6694396488962485,
      "learning_rate": 2.731958127994433e-07,
      "loss": 0.279448127746582,
      "num_input_tokens_seen": 1366698352,
      "step": 26565,
      "train_runtime": 163493.6684,
      "train_tokens_per_second": 8359.335
    },
    {
      "epoch": 1.788984648532184,
      "grad_norm": 0.7346016268295213,
      "learning_rate": 2.7233443449490695e-07,
      "loss": 0.31168107986450194,
      "num_input_tokens_seen": 1366945440,
      "step": 26570,
      "train_runtime": 163525.9102,
      "train_tokens_per_second": 8359.198
    },
    {
      "epoch": 1.7893213035281443,
      "grad_norm": 0.7283329727837069,
      "learning_rate": 2.7147437826697153e-07,
      "loss": 0.29676721096038816,
      "num_input_tokens_seen": 1367195128,
      "step": 26575,
      "train_runtime": 163556.2214,
      "train_tokens_per_second": 8359.175
    },
    {
      "epoch": 1.7896579585241046,
      "grad_norm": 0.7328171664885637,
      "learning_rate": 2.706156443561481e-07,
      "loss": 0.30084004402160647,
      "num_input_tokens_seen": 1367476672,
      "step": 26580,
      "train_runtime": 163588.5168,
      "train_tokens_per_second": 8359.246
    },
    {
      "epoch": 1.7899946135200646,
      "grad_norm": 0.6158903080812466,
      "learning_rate": 2.6975823300257884e-07,
      "loss": 0.3031869888305664,
      "num_input_tokens_seen": 1367734040,
      "step": 26585,
      "train_runtime": 163617.4449,
      "train_tokens_per_second": 8359.341
    },
    {
      "epoch": 1.7903312685160246,
      "grad_norm": 0.648007095948897,
      "learning_rate": 2.6890214444603604e-07,
      "loss": 0.2958799362182617,
      "num_input_tokens_seen": 1368005264,
      "step": 26590,
      "train_runtime": 163651.0495,
      "train_tokens_per_second": 8359.282
    },
    {
      "epoch": 1.790667923511985,
      "grad_norm": 0.7707529456147812,
      "learning_rate": 2.6804737892592193e-07,
      "loss": 0.30699868202209474,
      "num_input_tokens_seen": 1368236952,
      "step": 26595,
      "train_runtime": 163681.0499,
      "train_tokens_per_second": 8359.165
    },
    {
      "epoch": 1.7910045785079451,
      "grad_norm": 0.9056712748503344,
      "learning_rate": 2.671939366812687e-07,
      "loss": 0.29925525188446045,
      "num_input_tokens_seen": 1368492968,
      "step": 26600,
      "train_runtime": 163710.332,
      "train_tokens_per_second": 8359.234
    },
    {
      "epoch": 1.7913412335039052,
      "grad_norm": 4.5643605988133915,
      "learning_rate": 2.66341817950741e-07,
      "loss": 0.3009171724319458,
      "num_input_tokens_seen": 1368783832,
      "step": 26605,
      "train_runtime": 163746.6241,
      "train_tokens_per_second": 8359.158
    },
    {
      "epoch": 1.7916778884998652,
      "grad_norm": 0.644492128945891,
      "learning_rate": 2.6549102297262764e-07,
      "loss": 0.29766478538513186,
      "num_input_tokens_seen": 1369043848,
      "step": 26610,
      "train_runtime": 163776.5922,
      "train_tokens_per_second": 8359.216
    },
    {
      "epoch": 1.7920145434958255,
      "grad_norm": 0.7917535642304331,
      "learning_rate": 2.6464155198485387e-07,
      "loss": 0.3020035266876221,
      "num_input_tokens_seen": 1369314728,
      "step": 26615,
      "train_runtime": 163808.9122,
      "train_tokens_per_second": 8359.22
    },
    {
      "epoch": 1.7923511984917857,
      "grad_norm": 0.6165261422594354,
      "learning_rate": 2.6379340522496853e-07,
      "loss": 0.29959874153137206,
      "num_input_tokens_seen": 1369581792,
      "step": 26620,
      "train_runtime": 163840.7857,
      "train_tokens_per_second": 8359.224
    },
    {
      "epoch": 1.7926878534877457,
      "grad_norm": 0.728964103540598,
      "learning_rate": 2.6294658293015584e-07,
      "loss": 0.3151498794555664,
      "num_input_tokens_seen": 1369850344,
      "step": 26625,
      "train_runtime": 163874.2031,
      "train_tokens_per_second": 8359.158
    },
    {
      "epoch": 1.7930245084837058,
      "grad_norm": 0.7175973098200038,
      "learning_rate": 2.621010853372258e-07,
      "loss": 0.32748396396636964,
      "num_input_tokens_seen": 1370087376,
      "step": 26630,
      "train_runtime": 163900.2093,
      "train_tokens_per_second": 8359.278
    },
    {
      "epoch": 1.793361163479666,
      "grad_norm": 0.8444120377016635,
      "learning_rate": 2.612569126826187e-07,
      "loss": 0.31508784294128417,
      "num_input_tokens_seen": 1370349088,
      "step": 26635,
      "train_runtime": 163932.5449,
      "train_tokens_per_second": 8359.225
    },
    {
      "epoch": 1.7936978184756263,
      "grad_norm": 0.6005965668858685,
      "learning_rate": 2.6041406520240563e-07,
      "loss": 0.31157310009002687,
      "num_input_tokens_seen": 1370632256,
      "step": 26640,
      "train_runtime": 163962.7862,
      "train_tokens_per_second": 8359.411
    },
    {
      "epoch": 1.7940344734715863,
      "grad_norm": 0.871677079693244,
      "learning_rate": 2.5957254313228595e-07,
      "loss": 0.3005971908569336,
      "num_input_tokens_seen": 1370868768,
      "step": 26645,
      "train_runtime": 163993.4463,
      "train_tokens_per_second": 8359.29
    },
    {
      "epoch": 1.7943711284675463,
      "grad_norm": 0.6160897195680374,
      "learning_rate": 2.5873234670758753e-07,
      "loss": 0.28684320449829104,
      "num_input_tokens_seen": 1371138432,
      "step": 26650,
      "train_runtime": 164025.1142,
      "train_tokens_per_second": 8359.32
    },
    {
      "epoch": 1.7947077834635066,
      "grad_norm": 0.6993907950241754,
      "learning_rate": 2.578934761632712e-07,
      "loss": 0.30126285552978516,
      "num_input_tokens_seen": 1371422032,
      "step": 26655,
      "train_runtime": 164059.5415,
      "train_tokens_per_second": 8359.295
    },
    {
      "epoch": 1.7950444384594668,
      "grad_norm": 0.6478421754531897,
      "learning_rate": 2.570559317339211e-07,
      "loss": 0.2930849552154541,
      "num_input_tokens_seen": 1371682552,
      "step": 26660,
      "train_runtime": 164094.1039,
      "train_tokens_per_second": 8359.121
    },
    {
      "epoch": 1.7953810934554268,
      "grad_norm": 0.6260584071131169,
      "learning_rate": 2.56219713653757e-07,
      "loss": 0.290463924407959,
      "num_input_tokens_seen": 1371950920,
      "step": 26665,
      "train_runtime": 164124.8901,
      "train_tokens_per_second": 8359.189
    },
    {
      "epoch": 1.7957177484513869,
      "grad_norm": 0.7585643335471508,
      "learning_rate": 2.5538482215662187e-07,
      "loss": 0.32201323509216306,
      "num_input_tokens_seen": 1372207576,
      "step": 26670,
      "train_runtime": 164155.7502,
      "train_tokens_per_second": 8359.181
    },
    {
      "epoch": 1.7960544034473471,
      "grad_norm": 0.6605595012498481,
      "learning_rate": 2.545512574759923e-07,
      "loss": 0.26940791606903075,
      "num_input_tokens_seen": 1372453432,
      "step": 26675,
      "train_runtime": 164185.7914,
      "train_tokens_per_second": 8359.149
    },
    {
      "epoch": 1.7963910584433074,
      "grad_norm": 0.6810211095368512,
      "learning_rate": 2.537190198449713e-07,
      "loss": 0.3050240993499756,
      "num_input_tokens_seen": 1372712296,
      "step": 26680,
      "train_runtime": 164218.9444,
      "train_tokens_per_second": 8359.037
    },
    {
      "epoch": 1.7967277134392674,
      "grad_norm": 0.6694711987910362,
      "learning_rate": 2.5288810949629207e-07,
      "loss": 0.29526543617248535,
      "num_input_tokens_seen": 1372966640,
      "step": 26685,
      "train_runtime": 164252.3932,
      "train_tokens_per_second": 8358.884
    },
    {
      "epoch": 1.7970643684352274,
      "grad_norm": 0.6533442737381525,
      "learning_rate": 2.5205852666231544e-07,
      "loss": 0.28032846450805665,
      "num_input_tokens_seen": 1373215840,
      "step": 26690,
      "train_runtime": 164281.2594,
      "train_tokens_per_second": 8358.932
    },
    {
      "epoch": 1.7974010234311877,
      "grad_norm": 0.7232758490811922,
      "learning_rate": 2.5123027157503144e-07,
      "loss": 0.28046555519104005,
      "num_input_tokens_seen": 1373440984,
      "step": 26695,
      "train_runtime": 164307.3097,
      "train_tokens_per_second": 8358.977
    },
    {
      "epoch": 1.797737678427148,
      "grad_norm": 0.6202957790087975,
      "learning_rate": 2.504033444660592e-07,
      "loss": 0.30008535385131835,
      "num_input_tokens_seen": 1373724040,
      "step": 26700,
      "train_runtime": 164340.6067,
      "train_tokens_per_second": 8359.006
    },
    {
      "epoch": 1.798074333423108,
      "grad_norm": 0.5774880359271494,
      "learning_rate": 2.4957774556664706e-07,
      "loss": 0.288060736656189,
      "num_input_tokens_seen": 1373980216,
      "step": 26705,
      "train_runtime": 164371.2911,
      "train_tokens_per_second": 8359.004
    },
    {
      "epoch": 1.798410988419068,
      "grad_norm": 0.7455754755744658,
      "learning_rate": 2.487534751076692e-07,
      "loss": 0.2858297348022461,
      "num_input_tokens_seen": 1374230128,
      "step": 26710,
      "train_runtime": 164400.4462,
      "train_tokens_per_second": 8359.041
    },
    {
      "epoch": 1.7987476434150282,
      "grad_norm": 0.6315800165213423,
      "learning_rate": 2.47930533319633e-07,
      "loss": 0.2767197608947754,
      "num_input_tokens_seen": 1374481120,
      "step": 26715,
      "train_runtime": 164430.2477,
      "train_tokens_per_second": 8359.053
    },
    {
      "epoch": 1.7990842984109885,
      "grad_norm": 0.6389397814398871,
      "learning_rate": 2.4710892043266823e-07,
      "loss": 0.3152724027633667,
      "num_input_tokens_seen": 1374754152,
      "step": 26720,
      "train_runtime": 164462.5686,
      "train_tokens_per_second": 8359.07
    },
    {
      "epoch": 1.7994209534069485,
      "grad_norm": 0.6330817246401353,
      "learning_rate": 2.462886366765388e-07,
      "loss": 0.28899197578430175,
      "num_input_tokens_seen": 1375006704,
      "step": 26725,
      "train_runtime": 164495.8041,
      "train_tokens_per_second": 8358.917
    },
    {
      "epoch": 1.7997576084029085,
      "grad_norm": 0.7031120804006123,
      "learning_rate": 2.454696822806341e-07,
      "loss": 0.30874583721160886,
      "num_input_tokens_seen": 1375241760,
      "step": 26730,
      "train_runtime": 164522.7735,
      "train_tokens_per_second": 8358.975
    },
    {
      "epoch": 1.8000942633988688,
      "grad_norm": 0.6206827132380031,
      "learning_rate": 2.446520574739708e-07,
      "loss": 0.28034272193908694,
      "num_input_tokens_seen": 1375511392,
      "step": 26735,
      "train_runtime": 164554.9921,
      "train_tokens_per_second": 8358.977
    },
    {
      "epoch": 1.800430918394829,
      "grad_norm": 0.8331894968965721,
      "learning_rate": 2.438357624851967e-07,
      "loss": 0.2741487979888916,
      "num_input_tokens_seen": 1375763040,
      "step": 26740,
      "train_runtime": 164584.6344,
      "train_tokens_per_second": 8359.0
    },
    {
      "epoch": 1.800767573390789,
      "grad_norm": 0.643849542726069,
      "learning_rate": 2.430207975425847e-07,
      "loss": 0.30466017723083494,
      "num_input_tokens_seen": 1376032472,
      "step": 26745,
      "train_runtime": 164620.7023,
      "train_tokens_per_second": 8358.806
    },
    {
      "epoch": 1.801104228386749,
      "grad_norm": 0.6666792504274854,
      "learning_rate": 2.4220716287403744e-07,
      "loss": 0.30779523849487306,
      "num_input_tokens_seen": 1376305128,
      "step": 26750,
      "train_runtime": 164652.232,
      "train_tokens_per_second": 8358.861
    },
    {
      "epoch": 1.8014408833827094,
      "grad_norm": 0.6478472032571472,
      "learning_rate": 2.413948587070869e-07,
      "loss": 0.29144694805145266,
      "num_input_tokens_seen": 1376502960,
      "step": 26755,
      "train_runtime": 164681.0702,
      "train_tokens_per_second": 8358.599
    },
    {
      "epoch": 1.8017775383786696,
      "grad_norm": 0.6572064508575829,
      "learning_rate": 2.405838852688891e-07,
      "loss": 0.3054184913635254,
      "num_input_tokens_seen": 1376780416,
      "step": 26760,
      "train_runtime": 164712.8037,
      "train_tokens_per_second": 8358.673
    },
    {
      "epoch": 1.8021141933746296,
      "grad_norm": 0.6978890676548193,
      "learning_rate": 2.397742427862315e-07,
      "loss": 0.2644207239151001,
      "num_input_tokens_seen": 1377062160,
      "step": 26765,
      "train_runtime": 164745.997,
      "train_tokens_per_second": 8358.699
    },
    {
      "epoch": 1.8024508483705897,
      "grad_norm": 0.5996805201412253,
      "learning_rate": 2.3896593148552695e-07,
      "loss": 0.29239442348480227,
      "num_input_tokens_seen": 1377316536,
      "step": 26770,
      "train_runtime": 164772.9522,
      "train_tokens_per_second": 8358.875
    },
    {
      "epoch": 1.80278750336655,
      "grad_norm": 0.5817281035638224,
      "learning_rate": 2.3815895159281844e-07,
      "loss": 0.2740891695022583,
      "num_input_tokens_seen": 1377578688,
      "step": 26775,
      "train_runtime": 164802.1439,
      "train_tokens_per_second": 8358.985
    },
    {
      "epoch": 1.8031241583625102,
      "grad_norm": 0.6385209309217973,
      "learning_rate": 2.3735330333377493e-07,
      "loss": 0.27954792976379395,
      "num_input_tokens_seen": 1377856800,
      "step": 26780,
      "train_runtime": 164837.1124,
      "train_tokens_per_second": 8358.899
    },
    {
      "epoch": 1.8034608133584702,
      "grad_norm": 0.6903350333739159,
      "learning_rate": 2.3654898693369232e-07,
      "loss": 0.299359130859375,
      "num_input_tokens_seen": 1378102224,
      "step": 26785,
      "train_runtime": 164868.2647,
      "train_tokens_per_second": 8358.808
    },
    {
      "epoch": 1.8037974683544302,
      "grad_norm": 0.6817812504667816,
      "learning_rate": 2.3574600261749735e-07,
      "loss": 0.3022841691970825,
      "num_input_tokens_seen": 1378350904,
      "step": 26790,
      "train_runtime": 164902.9334,
      "train_tokens_per_second": 8358.559
    },
    {
      "epoch": 1.8041341233503905,
      "grad_norm": 0.6749330511951313,
      "learning_rate": 2.3494435060973875e-07,
      "loss": 0.31226515769958496,
      "num_input_tokens_seen": 1378624248,
      "step": 26795,
      "train_runtime": 164935.338,
      "train_tokens_per_second": 8358.574
    },
    {
      "epoch": 1.8044707783463507,
      "grad_norm": 0.7392381807499524,
      "learning_rate": 2.3414403113459937e-07,
      "loss": 0.28716259002685546,
      "num_input_tokens_seen": 1378886048,
      "step": 26800,
      "train_runtime": 164967.0214,
      "train_tokens_per_second": 8358.556
    },
    {
      "epoch": 1.8048074333423108,
      "grad_norm": 0.657683853126936,
      "learning_rate": 2.3334504441588413e-07,
      "loss": 0.27746422290802003,
      "num_input_tokens_seen": 1379135952,
      "step": 26805,
      "train_runtime": 164998.642,
      "train_tokens_per_second": 8358.469
    },
    {
      "epoch": 1.8051440883382708,
      "grad_norm": 0.6642940447371937,
      "learning_rate": 2.3254739067702704e-07,
      "loss": 0.2895087480545044,
      "num_input_tokens_seen": 1379403888,
      "step": 26810,
      "train_runtime": 165029.4464,
      "train_tokens_per_second": 8358.532
    },
    {
      "epoch": 1.805480743334231,
      "grad_norm": 0.7028182683398114,
      "learning_rate": 2.3175107014109077e-07,
      "loss": 0.27565274238586424,
      "num_input_tokens_seen": 1379671640,
      "step": 26815,
      "train_runtime": 165057.1623,
      "train_tokens_per_second": 8358.751
    },
    {
      "epoch": 1.8058173983301913,
      "grad_norm": 0.6670112118703009,
      "learning_rate": 2.3095608303076223e-07,
      "loss": 0.3078090906143188,
      "num_input_tokens_seen": 1379915112,
      "step": 26820,
      "train_runtime": 165086.2303,
      "train_tokens_per_second": 8358.754
    },
    {
      "epoch": 1.8061540533261513,
      "grad_norm": 0.6515052127604407,
      "learning_rate": 2.3016242956835798e-07,
      "loss": 0.3135065078735352,
      "num_input_tokens_seen": 1380187528,
      "step": 26825,
      "train_runtime": 165119.1677,
      "train_tokens_per_second": 8358.736
    },
    {
      "epoch": 1.8064907083221113,
      "grad_norm": 0.6509731835562325,
      "learning_rate": 2.2937010997581999e-07,
      "loss": 0.28349905014038085,
      "num_input_tokens_seen": 1380457640,
      "step": 26830,
      "train_runtime": 165148.995,
      "train_tokens_per_second": 8358.862
    },
    {
      "epoch": 1.8068273633180716,
      "grad_norm": 0.8465728612972613,
      "learning_rate": 2.2857912447471818e-07,
      "loss": 0.2916706085205078,
      "num_input_tokens_seen": 1380701992,
      "step": 26835,
      "train_runtime": 165177.8972,
      "train_tokens_per_second": 8358.879
    },
    {
      "epoch": 1.8071640183140318,
      "grad_norm": 0.6445220344656567,
      "learning_rate": 2.2778947328625012e-07,
      "loss": 0.2780611991882324,
      "num_input_tokens_seen": 1380935232,
      "step": 26840,
      "train_runtime": 165206.2866,
      "train_tokens_per_second": 8358.854
    },
    {
      "epoch": 1.8075006733099919,
      "grad_norm": 0.853304596510962,
      "learning_rate": 2.2700115663123745e-07,
      "loss": 0.32016854286193847,
      "num_input_tokens_seen": 1381197752,
      "step": 26845,
      "train_runtime": 165236.0519,
      "train_tokens_per_second": 8358.937
    },
    {
      "epoch": 1.807837328305952,
      "grad_norm": 0.6460968961925075,
      "learning_rate": 2.2621417473013164e-07,
      "loss": 0.2796643257141113,
      "num_input_tokens_seen": 1381453336,
      "step": 26850,
      "train_runtime": 165265.8499,
      "train_tokens_per_second": 8358.976
    },
    {
      "epoch": 1.8081739833019121,
      "grad_norm": 0.7369275946833516,
      "learning_rate": 2.254285278030094e-07,
      "loss": 0.3094628095626831,
      "num_input_tokens_seen": 1381731640,
      "step": 26855,
      "train_runtime": 165297.5318,
      "train_tokens_per_second": 8359.058
    },
    {
      "epoch": 1.8085106382978724,
      "grad_norm": 0.7077324047195571,
      "learning_rate": 2.24644216069575e-07,
      "loss": 0.29753885269165037,
      "num_input_tokens_seen": 1381974112,
      "step": 26860,
      "train_runtime": 165326.0355,
      "train_tokens_per_second": 8359.083
    },
    {
      "epoch": 1.8088472932938324,
      "grad_norm": 0.5726977503576736,
      "learning_rate": 2.238612397491574e-07,
      "loss": 0.2651700973510742,
      "num_input_tokens_seen": 1382242440,
      "step": 26865,
      "train_runtime": 165360.0796,
      "train_tokens_per_second": 8358.985
    },
    {
      "epoch": 1.8091839482897925,
      "grad_norm": 0.6486558678319373,
      "learning_rate": 2.230795990607143e-07,
      "loss": 0.29917302131652834,
      "num_input_tokens_seen": 1382503800,
      "step": 26870,
      "train_runtime": 165392.528,
      "train_tokens_per_second": 8358.925
    },
    {
      "epoch": 1.809520603285753,
      "grad_norm": 0.6680403921960552,
      "learning_rate": 2.2229929422282915e-07,
      "loss": 0.26502370834350586,
      "num_input_tokens_seen": 1382756368,
      "step": 26875,
      "train_runtime": 165423.0049,
      "train_tokens_per_second": 8358.912
    },
    {
      "epoch": 1.809857258281713,
      "grad_norm": 0.6948890426670201,
      "learning_rate": 2.215203254537107e-07,
      "loss": 0.3094813823699951,
      "num_input_tokens_seen": 1383012800,
      "step": 26880,
      "train_runtime": 165460.084,
      "train_tokens_per_second": 8358.589
    },
    {
      "epoch": 1.810193913277673,
      "grad_norm": 0.6770743278653019,
      "learning_rate": 2.2074269297119588e-07,
      "loss": 0.30295090675354003,
      "num_input_tokens_seen": 1383284912,
      "step": 26885,
      "train_runtime": 165491.9359,
      "train_tokens_per_second": 8358.624
    },
    {
      "epoch": 1.8105305682736332,
      "grad_norm": 0.7240749097966586,
      "learning_rate": 2.199663969927479e-07,
      "loss": 0.292708683013916,
      "num_input_tokens_seen": 1383529776,
      "step": 26890,
      "train_runtime": 165522.2062,
      "train_tokens_per_second": 8358.575
    },
    {
      "epoch": 1.8108672232695935,
      "grad_norm": 0.6519548034636742,
      "learning_rate": 2.1919143773545316e-07,
      "loss": 0.29218297004699706,
      "num_input_tokens_seen": 1383796576,
      "step": 26895,
      "train_runtime": 165556.9158,
      "train_tokens_per_second": 8358.434
    },
    {
      "epoch": 1.8112038782655535,
      "grad_norm": 0.7450983315320532,
      "learning_rate": 2.184178154160288e-07,
      "loss": 0.2884012460708618,
      "num_input_tokens_seen": 1384053936,
      "step": 26900,
      "train_runtime": 165588.8159,
      "train_tokens_per_second": 8358.378
    },
    {
      "epoch": 1.8115405332615135,
      "grad_norm": 0.6882137737150592,
      "learning_rate": 2.1764553025081514e-07,
      "loss": 0.2890641212463379,
      "num_input_tokens_seen": 1384313944,
      "step": 26905,
      "train_runtime": 165622.3071,
      "train_tokens_per_second": 8358.258
    },
    {
      "epoch": 1.8118771882574738,
      "grad_norm": 0.7831007398305819,
      "learning_rate": 2.168745824557794e-07,
      "loss": 0.28020777702331545,
      "num_input_tokens_seen": 1384561960,
      "step": 26910,
      "train_runtime": 165651.1278,
      "train_tokens_per_second": 8358.301
    },
    {
      "epoch": 1.812213843253434,
      "grad_norm": 0.7933353119802672,
      "learning_rate": 2.161049722465136e-07,
      "loss": 0.27577311992645265,
      "num_input_tokens_seen": 1384807944,
      "step": 26915,
      "train_runtime": 165681.5825,
      "train_tokens_per_second": 8358.249
    },
    {
      "epoch": 1.812550498249394,
      "grad_norm": 0.6084602350509514,
      "learning_rate": 2.1533669983823835e-07,
      "loss": 0.29628739356994627,
      "num_input_tokens_seen": 1385084792,
      "step": 26920,
      "train_runtime": 165715.2179,
      "train_tokens_per_second": 8358.223
    },
    {
      "epoch": 1.812887153245354,
      "grad_norm": 0.5523003978749602,
      "learning_rate": 2.1456976544579732e-07,
      "loss": 0.2712899684906006,
      "num_input_tokens_seen": 1385379072,
      "step": 26925,
      "train_runtime": 165749.731,
      "train_tokens_per_second": 8358.258
    },
    {
      "epoch": 1.8132238082413143,
      "grad_norm": 0.6605273085910873,
      "learning_rate": 2.138041692836623e-07,
      "loss": 0.2913766384124756,
      "num_input_tokens_seen": 1385629184,
      "step": 26930,
      "train_runtime": 165779.3411,
      "train_tokens_per_second": 8358.274
    },
    {
      "epoch": 1.8135604632372746,
      "grad_norm": 0.6930566053609906,
      "learning_rate": 2.1303991156592818e-07,
      "loss": 0.3048421382904053,
      "num_input_tokens_seen": 1385866240,
      "step": 26935,
      "train_runtime": 165808.6707,
      "train_tokens_per_second": 8358.225
    },
    {
      "epoch": 1.8138971182332346,
      "grad_norm": 0.6490378632303658,
      "learning_rate": 2.122769925063195e-07,
      "loss": 0.28559825420379636,
      "num_input_tokens_seen": 1386123544,
      "step": 26940,
      "train_runtime": 165838.9092,
      "train_tokens_per_second": 8358.253
    },
    {
      "epoch": 1.8142337732291947,
      "grad_norm": 0.752443075467216,
      "learning_rate": 2.1151541231818174e-07,
      "loss": 0.29168314933776857,
      "num_input_tokens_seen": 1386362664,
      "step": 26945,
      "train_runtime": 165870.8739,
      "train_tokens_per_second": 8358.084
    },
    {
      "epoch": 1.814570428225155,
      "grad_norm": 0.7134220774370073,
      "learning_rate": 2.1075517121449063e-07,
      "loss": 0.31829442977905276,
      "num_input_tokens_seen": 1386619872,
      "step": 26950,
      "train_runtime": 165901.3615,
      "train_tokens_per_second": 8358.098
    },
    {
      "epoch": 1.8149070832211152,
      "grad_norm": 0.6706854821420166,
      "learning_rate": 2.0999626940784334e-07,
      "loss": 0.26757619380950926,
      "num_input_tokens_seen": 1386854760,
      "step": 26955,
      "train_runtime": 165930.5124,
      "train_tokens_per_second": 8358.045
    },
    {
      "epoch": 1.8152437382170752,
      "grad_norm": 0.7388333553272171,
      "learning_rate": 2.0923870711046568e-07,
      "loss": 0.28675568103790283,
      "num_input_tokens_seen": 1387118080,
      "step": 26960,
      "train_runtime": 165960.3518,
      "train_tokens_per_second": 8358.129
    },
    {
      "epoch": 1.8155803932130352,
      "grad_norm": 0.7130875110358111,
      "learning_rate": 2.0848248453420704e-07,
      "loss": 0.282489275932312,
      "num_input_tokens_seen": 1387396160,
      "step": 26965,
      "train_runtime": 165986.7196,
      "train_tokens_per_second": 8358.477
    },
    {
      "epoch": 1.8159170482089955,
      "grad_norm": 0.7715206728047204,
      "learning_rate": 2.077276018905422e-07,
      "loss": 0.29452803134918215,
      "num_input_tokens_seen": 1387652608,
      "step": 26970,
      "train_runtime": 166016.5811,
      "train_tokens_per_second": 8358.518
    },
    {
      "epoch": 1.8162537032049557,
      "grad_norm": 0.6833449679724676,
      "learning_rate": 2.0697405939057225e-07,
      "loss": 0.30560061931610105,
      "num_input_tokens_seen": 1387879768,
      "step": 26975,
      "train_runtime": 166046.1568,
      "train_tokens_per_second": 8358.397
    },
    {
      "epoch": 1.8165903582009157,
      "grad_norm": 0.5768563615217351,
      "learning_rate": 2.0622185724502253e-07,
      "loss": 0.28327476978302,
      "num_input_tokens_seen": 1388133648,
      "step": 26980,
      "train_runtime": 166080.6717,
      "train_tokens_per_second": 8358.189
    },
    {
      "epoch": 1.8169270131968758,
      "grad_norm": 0.6830234135402892,
      "learning_rate": 2.0547099566424367e-07,
      "loss": 0.28138298988342286,
      "num_input_tokens_seen": 1388385552,
      "step": 26985,
      "train_runtime": 166110.2063,
      "train_tokens_per_second": 8358.219
    },
    {
      "epoch": 1.817263668192836,
      "grad_norm": 0.691867403862184,
      "learning_rate": 2.0472147485821325e-07,
      "loss": 0.28491902351379395,
      "num_input_tokens_seen": 1388631352,
      "step": 26990,
      "train_runtime": 166137.9587,
      "train_tokens_per_second": 8358.303
    },
    {
      "epoch": 1.8176003231887963,
      "grad_norm": 0.5994111701077457,
      "learning_rate": 2.0397329503653084e-07,
      "loss": 0.2698211193084717,
      "num_input_tokens_seen": 1388884488,
      "step": 26995,
      "train_runtime": 166168.3652,
      "train_tokens_per_second": 8358.297
    },
    {
      "epoch": 1.8179369781847563,
      "grad_norm": 0.7296968268578183,
      "learning_rate": 2.03226456408423e-07,
      "loss": 0.3060991048812866,
      "num_input_tokens_seen": 1389108784,
      "step": 27000,
      "train_runtime": 166196.1992,
      "train_tokens_per_second": 8358.246
    },
    {
      "epoch": 1.8182736331807163,
      "grad_norm": 0.6387386114230853,
      "learning_rate": 2.024809591827409e-07,
      "loss": 0.30579261779785155,
      "num_input_tokens_seen": 1389373096,
      "step": 27005,
      "train_runtime": 166227.3918,
      "train_tokens_per_second": 8358.268
    },
    {
      "epoch": 1.8186102881766766,
      "grad_norm": 0.7096753048523624,
      "learning_rate": 2.0173680356796066e-07,
      "loss": 0.2677801847457886,
      "num_input_tokens_seen": 1389607144,
      "step": 27010,
      "train_runtime": 166256.4254,
      "train_tokens_per_second": 8358.216
    },
    {
      "epoch": 1.8189469431726368,
      "grad_norm": 0.731282229645248,
      "learning_rate": 2.00993989772183e-07,
      "loss": 0.29690213203430177,
      "num_input_tokens_seen": 1389863632,
      "step": 27015,
      "train_runtime": 166292.4801,
      "train_tokens_per_second": 8357.946
    },
    {
      "epoch": 1.8192835981685969,
      "grad_norm": 0.7077629321902659,
      "learning_rate": 2.0025251800313284e-07,
      "loss": 0.30224099159240725,
      "num_input_tokens_seen": 1390103608,
      "step": 27020,
      "train_runtime": 166321.6912,
      "train_tokens_per_second": 8357.921
    },
    {
      "epoch": 1.8196202531645569,
      "grad_norm": 0.6836798250157011,
      "learning_rate": 1.9951238846816045e-07,
      "loss": 0.3087285041809082,
      "num_input_tokens_seen": 1390394208,
      "step": 27025,
      "train_runtime": 166355.106,
      "train_tokens_per_second": 8357.989
    },
    {
      "epoch": 1.8199569081605171,
      "grad_norm": 0.6948724914858133,
      "learning_rate": 1.987736013742414e-07,
      "loss": 0.31884472370147704,
      "num_input_tokens_seen": 1390648264,
      "step": 27030,
      "train_runtime": 166384.8704,
      "train_tokens_per_second": 8358.021
    },
    {
      "epoch": 1.8202935631564774,
      "grad_norm": 0.6887173930041717,
      "learning_rate": 1.9803615692797375e-07,
      "loss": 0.27652673721313475,
      "num_input_tokens_seen": 1390909168,
      "step": 27035,
      "train_runtime": 166416.3379,
      "train_tokens_per_second": 8358.009
    },
    {
      "epoch": 1.8206302181524374,
      "grad_norm": 0.6000507738370874,
      "learning_rate": 1.9730005533558417e-07,
      "loss": 0.28652157783508303,
      "num_input_tokens_seen": 1391186688,
      "step": 27040,
      "train_runtime": 166449.0435,
      "train_tokens_per_second": 8358.034
    },
    {
      "epoch": 1.8209668731483974,
      "grad_norm": 0.6177067206356486,
      "learning_rate": 1.9656529680291802e-07,
      "loss": 0.280538272857666,
      "num_input_tokens_seen": 1391455168,
      "step": 27045,
      "train_runtime": 166479.084,
      "train_tokens_per_second": 8358.138
    },
    {
      "epoch": 1.8213035281443577,
      "grad_norm": 0.6204337720350389,
      "learning_rate": 1.9583188153544986e-07,
      "loss": 0.284575366973877,
      "num_input_tokens_seen": 1391692184,
      "step": 27050,
      "train_runtime": 166510.266,
      "train_tokens_per_second": 8357.996
    },
    {
      "epoch": 1.821640183140318,
      "grad_norm": 0.7084622944019282,
      "learning_rate": 1.9509980973827724e-07,
      "loss": 0.2941887617111206,
      "num_input_tokens_seen": 1391954296,
      "step": 27055,
      "train_runtime": 166541.7703,
      "train_tokens_per_second": 8357.989
    },
    {
      "epoch": 1.821976838136278,
      "grad_norm": 0.6395859234921267,
      "learning_rate": 1.943690816161209e-07,
      "loss": 0.300445032119751,
      "num_input_tokens_seen": 1392230448,
      "step": 27060,
      "train_runtime": 166572.4539,
      "train_tokens_per_second": 8358.107
    },
    {
      "epoch": 1.822313493132238,
      "grad_norm": 0.6285541322214141,
      "learning_rate": 1.9363969737332677e-07,
      "loss": 0.2658878803253174,
      "num_input_tokens_seen": 1392497248,
      "step": 27065,
      "train_runtime": 166603.9971,
      "train_tokens_per_second": 8358.126
    },
    {
      "epoch": 1.8226501481281983,
      "grad_norm": 0.5753714649455663,
      "learning_rate": 1.9291165721386507e-07,
      "loss": 0.27535560131073,
      "num_input_tokens_seen": 1392737160,
      "step": 27070,
      "train_runtime": 166633.0704,
      "train_tokens_per_second": 8358.108
    },
    {
      "epoch": 1.8229868031241585,
      "grad_norm": 0.6579280121773382,
      "learning_rate": 1.9218496134132957e-07,
      "loss": 0.3031576871871948,
      "num_input_tokens_seen": 1393012296,
      "step": 27075,
      "train_runtime": 166664.8111,
      "train_tokens_per_second": 8358.167
    },
    {
      "epoch": 1.8233234581201185,
      "grad_norm": 0.6236859931188852,
      "learning_rate": 1.9145960995893831e-07,
      "loss": 0.2870443105697632,
      "num_input_tokens_seen": 1393276592,
      "step": 27080,
      "train_runtime": 166692.2673,
      "train_tokens_per_second": 8358.376
    },
    {
      "epoch": 1.8236601131160786,
      "grad_norm": 0.6082355272243751,
      "learning_rate": 1.9073560326953344e-07,
      "loss": 0.29740471839904786,
      "num_input_tokens_seen": 1393548720,
      "step": 27085,
      "train_runtime": 166724.6998,
      "train_tokens_per_second": 8358.382
    },
    {
      "epoch": 1.8239967681120388,
      "grad_norm": 0.6612118392370018,
      "learning_rate": 1.9001294147558303e-07,
      "loss": 0.30573444366455077,
      "num_input_tokens_seen": 1393799960,
      "step": 27090,
      "train_runtime": 166754.2131,
      "train_tokens_per_second": 8358.409
    },
    {
      "epoch": 1.824333423107999,
      "grad_norm": 0.7411477730346385,
      "learning_rate": 1.892916247791743e-07,
      "loss": 0.3048961877822876,
      "num_input_tokens_seen": 1394045320,
      "step": 27095,
      "train_runtime": 166784.5264,
      "train_tokens_per_second": 8358.361
    },
    {
      "epoch": 1.824670078103959,
      "grad_norm": 0.6105790209696156,
      "learning_rate": 1.885716533820231e-07,
      "loss": 0.2926558256149292,
      "num_input_tokens_seen": 1394298736,
      "step": 27100,
      "train_runtime": 166817.1696,
      "train_tokens_per_second": 8358.245
    },
    {
      "epoch": 1.8250067330999191,
      "grad_norm": 0.6483474890975706,
      "learning_rate": 1.878530274854662e-07,
      "loss": 0.2878259181976318,
      "num_input_tokens_seen": 1394554912,
      "step": 27105,
      "train_runtime": 166848.2386,
      "train_tokens_per_second": 8358.224
    },
    {
      "epoch": 1.8253433880958794,
      "grad_norm": 0.5752131812958418,
      "learning_rate": 1.8713574729046557e-07,
      "loss": 0.3138455629348755,
      "num_input_tokens_seen": 1394841336,
      "step": 27110,
      "train_runtime": 166879.9971,
      "train_tokens_per_second": 8358.349
    },
    {
      "epoch": 1.8256800430918396,
      "grad_norm": 0.6928346091982974,
      "learning_rate": 1.8641981299760582e-07,
      "loss": 0.2970116138458252,
      "num_input_tokens_seen": 1395091680,
      "step": 27115,
      "train_runtime": 166910.8619,
      "train_tokens_per_second": 8358.304
    },
    {
      "epoch": 1.8260166980877997,
      "grad_norm": 0.7434494944184638,
      "learning_rate": 1.857052248070962e-07,
      "loss": 0.3048053741455078,
      "num_input_tokens_seen": 1395315856,
      "step": 27120,
      "train_runtime": 166939.9883,
      "train_tokens_per_second": 8358.188
    },
    {
      "epoch": 1.8263533530837597,
      "grad_norm": 0.6558597464279147,
      "learning_rate": 1.8499198291876907e-07,
      "loss": 0.289397668838501,
      "num_input_tokens_seen": 1395542352,
      "step": 27125,
      "train_runtime": 166967.871,
      "train_tokens_per_second": 8358.149
    },
    {
      "epoch": 1.82669000807972,
      "grad_norm": 0.5688663947531563,
      "learning_rate": 1.8428008753207937e-07,
      "loss": 0.2969072341918945,
      "num_input_tokens_seen": 1395806832,
      "step": 27130,
      "train_runtime": 166995.8159,
      "train_tokens_per_second": 8358.334
    },
    {
      "epoch": 1.8270266630756802,
      "grad_norm": 0.669109369785162,
      "learning_rate": 1.8356953884610673e-07,
      "loss": 0.30004134178161623,
      "num_input_tokens_seen": 1396060360,
      "step": 27135,
      "train_runtime": 167024.4626,
      "train_tokens_per_second": 8358.419
    },
    {
      "epoch": 1.8273633180716402,
      "grad_norm": 0.6489252067609378,
      "learning_rate": 1.8286033705955554e-07,
      "loss": 0.2909057378768921,
      "num_input_tokens_seen": 1396330616,
      "step": 27140,
      "train_runtime": 167056.7558,
      "train_tokens_per_second": 8358.421
    },
    {
      "epoch": 1.8276999730676002,
      "grad_norm": 0.714117394349501,
      "learning_rate": 1.8215248237074879e-07,
      "loss": 0.30047130584716797,
      "num_input_tokens_seen": 1396594168,
      "step": 27145,
      "train_runtime": 167083.6118,
      "train_tokens_per_second": 8358.654
    },
    {
      "epoch": 1.8280366280635605,
      "grad_norm": 0.5544295507624538,
      "learning_rate": 1.8144597497763816e-07,
      "loss": 0.2934086799621582,
      "num_input_tokens_seen": 1396855296,
      "step": 27150,
      "train_runtime": 167116.1503,
      "train_tokens_per_second": 8358.589
    },
    {
      "epoch": 1.8283732830595207,
      "grad_norm": 0.6271774417389512,
      "learning_rate": 1.8074081507779506e-07,
      "loss": 0.2876414775848389,
      "num_input_tokens_seen": 1397126952,
      "step": 27155,
      "train_runtime": 167145.8543,
      "train_tokens_per_second": 8358.729
    },
    {
      "epoch": 1.8287099380554808,
      "grad_norm": 0.662308957813309,
      "learning_rate": 1.8003700286841563e-07,
      "loss": 0.29808611869812013,
      "num_input_tokens_seen": 1397388688,
      "step": 27160,
      "train_runtime": 167176.5394,
      "train_tokens_per_second": 8358.761
    },
    {
      "epoch": 1.8290465930514408,
      "grad_norm": 0.7783122379188856,
      "learning_rate": 1.793345385463191e-07,
      "loss": 0.3017151355743408,
      "num_input_tokens_seen": 1397656920,
      "step": 27165,
      "train_runtime": 167204.7443,
      "train_tokens_per_second": 8358.955
    },
    {
      "epoch": 1.829383248047401,
      "grad_norm": 0.7658095539223514,
      "learning_rate": 1.7863342230794667e-07,
      "loss": 0.309536075592041,
      "num_input_tokens_seen": 1397922632,
      "step": 27170,
      "train_runtime": 167238.1247,
      "train_tokens_per_second": 8358.875
    },
    {
      "epoch": 1.8297199030433613,
      "grad_norm": 0.6270033726856385,
      "learning_rate": 1.7793365434936316e-07,
      "loss": 0.3031143665313721,
      "num_input_tokens_seen": 1398196888,
      "step": 27175,
      "train_runtime": 167271.5198,
      "train_tokens_per_second": 8358.846
    },
    {
      "epoch": 1.8300565580393213,
      "grad_norm": 0.664211829170988,
      "learning_rate": 1.7723523486625704e-07,
      "loss": 0.29147729873657224,
      "num_input_tokens_seen": 1398440712,
      "step": 27180,
      "train_runtime": 167300.8087,
      "train_tokens_per_second": 8358.84
    },
    {
      "epoch": 1.8303932130352814,
      "grad_norm": 0.6146336558744057,
      "learning_rate": 1.7653816405393766e-07,
      "loss": 0.3021785497665405,
      "num_input_tokens_seen": 1398694744,
      "step": 27185,
      "train_runtime": 167329.4527,
      "train_tokens_per_second": 8358.927
    },
    {
      "epoch": 1.8307298680312416,
      "grad_norm": 0.7131557199032027,
      "learning_rate": 1.758424421073418e-07,
      "loss": 0.3074542284011841,
      "num_input_tokens_seen": 1398971632,
      "step": 27190,
      "train_runtime": 167361.814,
      "train_tokens_per_second": 8358.966
    },
    {
      "epoch": 1.8310665230272019,
      "grad_norm": 0.6351209848950977,
      "learning_rate": 1.7514806922102222e-07,
      "loss": 0.28278307914733886,
      "num_input_tokens_seen": 1399188320,
      "step": 27195,
      "train_runtime": 167391.0506,
      "train_tokens_per_second": 8358.8
    },
    {
      "epoch": 1.8314031780231619,
      "grad_norm": 0.6224365674600529,
      "learning_rate": 1.7445504558916027e-07,
      "loss": 0.28804335594177244,
      "num_input_tokens_seen": 1399446896,
      "step": 27200,
      "train_runtime": 167416.6624,
      "train_tokens_per_second": 8359.066
    },
    {
      "epoch": 1.831739833019122,
      "grad_norm": 0.6373426586445394,
      "learning_rate": 1.7376337140555754e-07,
      "loss": 0.279966402053833,
      "num_input_tokens_seen": 1399687376,
      "step": 27205,
      "train_runtime": 167448.5243,
      "train_tokens_per_second": 8358.911
    },
    {
      "epoch": 1.8320764880150822,
      "grad_norm": 0.602124579855926,
      "learning_rate": 1.7307304686363825e-07,
      "loss": 0.31307082176208495,
      "num_input_tokens_seen": 1399947088,
      "step": 27210,
      "train_runtime": 167480.6715,
      "train_tokens_per_second": 8358.858
    },
    {
      "epoch": 1.8324131430110424,
      "grad_norm": 0.6138934039001713,
      "learning_rate": 1.723840721564496e-07,
      "loss": 0.2807272672653198,
      "num_input_tokens_seen": 1400215616,
      "step": 27215,
      "train_runtime": 167512.7099,
      "train_tokens_per_second": 8358.862
    },
    {
      "epoch": 1.8327497980070024,
      "grad_norm": 0.9072835960951174,
      "learning_rate": 1.7169644747666081e-07,
      "loss": 0.295215368270874,
      "num_input_tokens_seen": 1400492288,
      "step": 27220,
      "train_runtime": 167543.4337,
      "train_tokens_per_second": 8358.98
    },
    {
      "epoch": 1.8330864530029625,
      "grad_norm": 0.608343442482064,
      "learning_rate": 1.710101730165642e-07,
      "loss": 0.29801416397094727,
      "num_input_tokens_seen": 1400750624,
      "step": 27225,
      "train_runtime": 167577.457,
      "train_tokens_per_second": 8358.825
    },
    {
      "epoch": 1.8334231079989227,
      "grad_norm": 0.6726919206400542,
      "learning_rate": 1.7032524896807402e-07,
      "loss": 0.28414487838745117,
      "num_input_tokens_seen": 1400988224,
      "step": 27230,
      "train_runtime": 167608.2249,
      "train_tokens_per_second": 8358.708
    },
    {
      "epoch": 1.833759762994883,
      "grad_norm": 0.6626439388270664,
      "learning_rate": 1.6964167552272702e-07,
      "loss": 0.3009357213973999,
      "num_input_tokens_seen": 1401228400,
      "step": 27235,
      "train_runtime": 167635.0001,
      "train_tokens_per_second": 8358.806
    },
    {
      "epoch": 1.834096417990843,
      "grad_norm": 0.6437375856818925,
      "learning_rate": 1.6895945287168314e-07,
      "loss": 0.29804329872131347,
      "num_input_tokens_seen": 1401481816,
      "step": 27240,
      "train_runtime": 167663.8132,
      "train_tokens_per_second": 8358.881
    },
    {
      "epoch": 1.834433072986803,
      "grad_norm": 0.6796456920823095,
      "learning_rate": 1.682785812057225e-07,
      "loss": 0.3101332426071167,
      "num_input_tokens_seen": 1401725096,
      "step": 27245,
      "train_runtime": 167698.5374,
      "train_tokens_per_second": 8358.601
    },
    {
      "epoch": 1.8347697279827633,
      "grad_norm": 0.66711543735236,
      "learning_rate": 1.6759906071525e-07,
      "loss": 0.2989849090576172,
      "num_input_tokens_seen": 1401976600,
      "step": 27250,
      "train_runtime": 167727.6007,
      "train_tokens_per_second": 8358.652
    },
    {
      "epoch": 1.8351063829787235,
      "grad_norm": 0.6656729397141194,
      "learning_rate": 1.669208915902909e-07,
      "loss": 0.3268181324005127,
      "num_input_tokens_seen": 1402267344,
      "step": 27255,
      "train_runtime": 167759.1635,
      "train_tokens_per_second": 8358.812
    },
    {
      "epoch": 1.8354430379746836,
      "grad_norm": 0.7124408416615807,
      "learning_rate": 1.6624407402049237e-07,
      "loss": 0.28125953674316406,
      "num_input_tokens_seen": 1402517472,
      "step": 27260,
      "train_runtime": 167792.3969,
      "train_tokens_per_second": 8358.647
    },
    {
      "epoch": 1.8357796929706436,
      "grad_norm": 0.6521591645143701,
      "learning_rate": 1.655686081951252e-07,
      "loss": 0.2871826410293579,
      "num_input_tokens_seen": 1402784976,
      "step": 27265,
      "train_runtime": 167823.3694,
      "train_tokens_per_second": 8358.699
    },
    {
      "epoch": 1.8361163479666038,
      "grad_norm": 0.6771500113503154,
      "learning_rate": 1.6489449430308058e-07,
      "loss": 0.3271247148513794,
      "num_input_tokens_seen": 1403058336,
      "step": 27270,
      "train_runtime": 167853.4959,
      "train_tokens_per_second": 8358.827
    },
    {
      "epoch": 1.836453002962564,
      "grad_norm": 0.6181372314589809,
      "learning_rate": 1.6422173253287323e-07,
      "loss": 0.31166090965271,
      "num_input_tokens_seen": 1403324664,
      "step": 27275,
      "train_runtime": 167885.4725,
      "train_tokens_per_second": 8358.821
    },
    {
      "epoch": 1.8367896579585241,
      "grad_norm": 0.7111830243589163,
      "learning_rate": 1.6355032307263764e-07,
      "loss": 0.2877988815307617,
      "num_input_tokens_seen": 1403577688,
      "step": 27280,
      "train_runtime": 167918.0215,
      "train_tokens_per_second": 8358.708
    },
    {
      "epoch": 1.8371263129544841,
      "grad_norm": 0.6346462829143176,
      "learning_rate": 1.628802661101314e-07,
      "loss": 0.274484658241272,
      "num_input_tokens_seen": 1403826456,
      "step": 27285,
      "train_runtime": 167946.8311,
      "train_tokens_per_second": 8358.755
    },
    {
      "epoch": 1.8374629679504444,
      "grad_norm": 0.6123578575646134,
      "learning_rate": 1.6221156183273524e-07,
      "loss": 0.29307241439819337,
      "num_input_tokens_seen": 1404080280,
      "step": 27290,
      "train_runtime": 167976.4211,
      "train_tokens_per_second": 8358.794
    },
    {
      "epoch": 1.8377996229464046,
      "grad_norm": 0.7187925007189334,
      "learning_rate": 1.615442104274484e-07,
      "loss": 0.2953474521636963,
      "num_input_tokens_seen": 1404362872,
      "step": 27295,
      "train_runtime": 168007.0158,
      "train_tokens_per_second": 8358.954
    },
    {
      "epoch": 1.8381362779423647,
      "grad_norm": 0.6562192460938727,
      "learning_rate": 1.6087821208089492e-07,
      "loss": 0.29661567211151124,
      "num_input_tokens_seen": 1404626928,
      "step": 27300,
      "train_runtime": 168036.3859,
      "train_tokens_per_second": 8359.064
    },
    {
      "epoch": 1.8384729329383247,
      "grad_norm": 0.5845289835162022,
      "learning_rate": 1.6021356697931866e-07,
      "loss": 0.2748464584350586,
      "num_input_tokens_seen": 1404902656,
      "step": 27305,
      "train_runtime": 168066.9002,
      "train_tokens_per_second": 8359.187
    },
    {
      "epoch": 1.838809587934285,
      "grad_norm": 0.5946993546114681,
      "learning_rate": 1.5955027530858537e-07,
      "loss": 0.27166571617126467,
      "num_input_tokens_seen": 1405167456,
      "step": 27310,
      "train_runtime": 168097.4438,
      "train_tokens_per_second": 8359.243
    },
    {
      "epoch": 1.8391462429302452,
      "grad_norm": 0.5775718061572994,
      "learning_rate": 1.5888833725418275e-07,
      "loss": 0.269016695022583,
      "num_input_tokens_seen": 1405419776,
      "step": 27315,
      "train_runtime": 168124.4515,
      "train_tokens_per_second": 8359.401
    },
    {
      "epoch": 1.8394828979262052,
      "grad_norm": 0.6341184280842755,
      "learning_rate": 1.582277530012194e-07,
      "loss": 0.25913591384887696,
      "num_input_tokens_seen": 1405718416,
      "step": 27320,
      "train_runtime": 168155.2146,
      "train_tokens_per_second": 8359.648
    },
    {
      "epoch": 1.8398195529221653,
      "grad_norm": 0.7935496511553164,
      "learning_rate": 1.5756852273442536e-07,
      "loss": 0.29579505920410154,
      "num_input_tokens_seen": 1405956560,
      "step": 27325,
      "train_runtime": 168182.2304,
      "train_tokens_per_second": 8359.721
    },
    {
      "epoch": 1.8401562079181255,
      "grad_norm": 0.6069987895490597,
      "learning_rate": 1.569106466381526e-07,
      "loss": 0.31722095012664797,
      "num_input_tokens_seen": 1406216792,
      "step": 27330,
      "train_runtime": 168211.9463,
      "train_tokens_per_second": 8359.791
    },
    {
      "epoch": 1.8404928629140858,
      "grad_norm": 0.6879213634840055,
      "learning_rate": 1.5625412489637337e-07,
      "loss": 0.27366271018981936,
      "num_input_tokens_seen": 1406480048,
      "step": 27335,
      "train_runtime": 168242.1292,
      "train_tokens_per_second": 8359.856
    },
    {
      "epoch": 1.8408295179100458,
      "grad_norm": 0.6486497026161603,
      "learning_rate": 1.555989576926842e-07,
      "loss": 0.28552708625793455,
      "num_input_tokens_seen": 1406740784,
      "step": 27340,
      "train_runtime": 168279.2329,
      "train_tokens_per_second": 8359.563
    },
    {
      "epoch": 1.8411661729060058,
      "grad_norm": 0.6645531873628024,
      "learning_rate": 1.5494514521029736e-07,
      "loss": 0.29195704460144045,
      "num_input_tokens_seen": 1407002264,
      "step": 27345,
      "train_runtime": 168310.5687,
      "train_tokens_per_second": 8359.56
    },
    {
      "epoch": 1.841502827901966,
      "grad_norm": 0.6455881785747429,
      "learning_rate": 1.5429268763205108e-07,
      "loss": 0.301835823059082,
      "num_input_tokens_seen": 1407262408,
      "step": 27350,
      "train_runtime": 168340.9206,
      "train_tokens_per_second": 8359.598
    },
    {
      "epoch": 1.8418394828979263,
      "grad_norm": 0.6921140262258009,
      "learning_rate": 1.5364158514040328e-07,
      "loss": 0.3242556810379028,
      "num_input_tokens_seen": 1407515024,
      "step": 27355,
      "train_runtime": 168372.9108,
      "train_tokens_per_second": 8359.51
    },
    {
      "epoch": 1.8421761378938863,
      "grad_norm": 0.7926638034372919,
      "learning_rate": 1.5299183791743223e-07,
      "loss": 0.2945744276046753,
      "num_input_tokens_seen": 1407762816,
      "step": 27360,
      "train_runtime": 168405.1321,
      "train_tokens_per_second": 8359.382
    },
    {
      "epoch": 1.8425127928898464,
      "grad_norm": 0.7906450727227767,
      "learning_rate": 1.523434461448381e-07,
      "loss": 0.32512893676757815,
      "num_input_tokens_seen": 1408039184,
      "step": 27365,
      "train_runtime": 168435.0112,
      "train_tokens_per_second": 8359.54
    },
    {
      "epoch": 1.8428494478858066,
      "grad_norm": 0.6315089228481756,
      "learning_rate": 1.5169641000394086e-07,
      "loss": 0.29710228443145753,
      "num_input_tokens_seen": 1408313416,
      "step": 27370,
      "train_runtime": 168468.4729,
      "train_tokens_per_second": 8359.507
    },
    {
      "epoch": 1.8431861028817669,
      "grad_norm": 0.6312206434884472,
      "learning_rate": 1.5105072967568246e-07,
      "loss": 0.2969668388366699,
      "num_input_tokens_seen": 1408556016,
      "step": 27375,
      "train_runtime": 168495.2871,
      "train_tokens_per_second": 8359.617
    },
    {
      "epoch": 1.843522757877727,
      "grad_norm": 0.6086820643023647,
      "learning_rate": 1.504064053406251e-07,
      "loss": 0.296018385887146,
      "num_input_tokens_seen": 1408804192,
      "step": 27380,
      "train_runtime": 168521.1509,
      "train_tokens_per_second": 8359.806
    },
    {
      "epoch": 1.843859412873687,
      "grad_norm": 0.6830826390943389,
      "learning_rate": 1.4976343717895192e-07,
      "loss": 0.298221492767334,
      "num_input_tokens_seen": 1409087152,
      "step": 27385,
      "train_runtime": 168555.4997,
      "train_tokens_per_second": 8359.782
    },
    {
      "epoch": 1.8441960678696472,
      "grad_norm": 0.686576396948177,
      "learning_rate": 1.491218253704685e-07,
      "loss": 0.30627970695495604,
      "num_input_tokens_seen": 1409343176,
      "step": 27390,
      "train_runtime": 168587.5465,
      "train_tokens_per_second": 8359.711
    },
    {
      "epoch": 1.8445327228656074,
      "grad_norm": 0.6036825428865756,
      "learning_rate": 1.484815700945974e-07,
      "loss": 0.31020827293395997,
      "num_input_tokens_seen": 1409601280,
      "step": 27395,
      "train_runtime": 168615.1306,
      "train_tokens_per_second": 8359.874
    },
    {
      "epoch": 1.8448693778615675,
      "grad_norm": 0.8536675994905177,
      "learning_rate": 1.478426715303849e-07,
      "loss": 0.29754033088684084,
      "num_input_tokens_seen": 1409836296,
      "step": 27400,
      "train_runtime": 168644.1666,
      "train_tokens_per_second": 8359.828
    },
    {
      "epoch": 1.8452060328575275,
      "grad_norm": 0.6414574357696662,
      "learning_rate": 1.4720512985649694e-07,
      "loss": 0.2844900131225586,
      "num_input_tokens_seen": 1410077936,
      "step": 27405,
      "train_runtime": 168673.2506,
      "train_tokens_per_second": 8359.82
    },
    {
      "epoch": 1.8455426878534877,
      "grad_norm": 0.9044133972096227,
      "learning_rate": 1.4656894525122034e-07,
      "loss": 0.30521335601806643,
      "num_input_tokens_seen": 1410296520,
      "step": 27410,
      "train_runtime": 168704.1518,
      "train_tokens_per_second": 8359.584
    },
    {
      "epoch": 1.845879342849448,
      "grad_norm": 0.6910285531728951,
      "learning_rate": 1.459341178924617e-07,
      "loss": 0.3017052888870239,
      "num_input_tokens_seen": 1410559104,
      "step": 27415,
      "train_runtime": 168735.7148,
      "train_tokens_per_second": 8359.576
    },
    {
      "epoch": 1.846215997845408,
      "grad_norm": 0.8534954957219579,
      "learning_rate": 1.453006479577479e-07,
      "loss": 0.28848450183868407,
      "num_input_tokens_seen": 1410811856,
      "step": 27420,
      "train_runtime": 168766.8776,
      "train_tokens_per_second": 8359.53
    },
    {
      "epoch": 1.846552652841368,
      "grad_norm": 0.5743725805523999,
      "learning_rate": 1.4466853562422778e-07,
      "loss": 0.3029740810394287,
      "num_input_tokens_seen": 1411043464,
      "step": 27425,
      "train_runtime": 168796.194,
      "train_tokens_per_second": 8359.451
    },
    {
      "epoch": 1.8468893078373283,
      "grad_norm": 0.6771213685199623,
      "learning_rate": 1.440377810686694e-07,
      "loss": 0.2874499797821045,
      "num_input_tokens_seen": 1411320832,
      "step": 27430,
      "train_runtime": 168826.6871,
      "train_tokens_per_second": 8359.584
    },
    {
      "epoch": 1.8472259628332885,
      "grad_norm": 0.713441908277151,
      "learning_rate": 1.4340838446746052e-07,
      "loss": 0.29938862323760984,
      "num_input_tokens_seen": 1411594184,
      "step": 27435,
      "train_runtime": 168857.0302,
      "train_tokens_per_second": 8359.7
    },
    {
      "epoch": 1.8475626178292486,
      "grad_norm": 0.709424279955867,
      "learning_rate": 1.4278034599661094e-07,
      "loss": 0.2813431739807129,
      "num_input_tokens_seen": 1411845664,
      "step": 27440,
      "train_runtime": 168888.457,
      "train_tokens_per_second": 8359.634
    },
    {
      "epoch": 1.8478992728252086,
      "grad_norm": 0.713577780392702,
      "learning_rate": 1.421536658317485e-07,
      "loss": 0.2671042442321777,
      "num_input_tokens_seen": 1412075432,
      "step": 27445,
      "train_runtime": 168914.9125,
      "train_tokens_per_second": 8359.685
    },
    {
      "epoch": 1.8482359278211689,
      "grad_norm": 0.6521315998891405,
      "learning_rate": 1.41528344148123e-07,
      "loss": 0.29294869899749754,
      "num_input_tokens_seen": 1412367576,
      "step": 27450,
      "train_runtime": 168949.4842,
      "train_tokens_per_second": 8359.703
    },
    {
      "epoch": 1.848572582817129,
      "grad_norm": 0.8595621225858232,
      "learning_rate": 1.40904381120604e-07,
      "loss": 0.2685142278671265,
      "num_input_tokens_seen": 1412629104,
      "step": 27455,
      "train_runtime": 168981.2491,
      "train_tokens_per_second": 8359.68
    },
    {
      "epoch": 1.8489092378130891,
      "grad_norm": 0.6631559670266449,
      "learning_rate": 1.4028177692367973e-07,
      "loss": 0.2970660448074341,
      "num_input_tokens_seen": 1412875168,
      "step": 27460,
      "train_runtime": 169010.3327,
      "train_tokens_per_second": 8359.697
    },
    {
      "epoch": 1.8492458928090492,
      "grad_norm": 0.7161151240736936,
      "learning_rate": 1.3966053173146032e-07,
      "loss": 0.29160304069519044,
      "num_input_tokens_seen": 1413150208,
      "step": 27465,
      "train_runtime": 169044.2885,
      "train_tokens_per_second": 8359.645
    },
    {
      "epoch": 1.8495825478050094,
      "grad_norm": 0.6177342121481179,
      "learning_rate": 1.3904064571767462e-07,
      "loss": 0.29656195640563965,
      "num_input_tokens_seen": 1413418488,
      "step": 27470,
      "train_runtime": 169074.6285,
      "train_tokens_per_second": 8359.731
    },
    {
      "epoch": 1.8499192028009697,
      "grad_norm": 0.5952438528755347,
      "learning_rate": 1.3842211905567227e-07,
      "loss": 0.2738644123077393,
      "num_input_tokens_seen": 1413684352,
      "step": 27475,
      "train_runtime": 169109.5901,
      "train_tokens_per_second": 8359.575
    },
    {
      "epoch": 1.8502558577969297,
      "grad_norm": 0.6994585377432212,
      "learning_rate": 1.378049519184216e-07,
      "loss": 0.2825650215148926,
      "num_input_tokens_seen": 1413933328,
      "step": 27480,
      "train_runtime": 169138.6889,
      "train_tokens_per_second": 8359.609
    },
    {
      "epoch": 1.8505925127928897,
      "grad_norm": 0.6217462961460664,
      "learning_rate": 1.3718914447851116e-07,
      "loss": 0.2660493850708008,
      "num_input_tokens_seen": 1414205352,
      "step": 27485,
      "train_runtime": 169171.6324,
      "train_tokens_per_second": 8359.589
    },
    {
      "epoch": 1.85092916778885,
      "grad_norm": 0.6375960399534866,
      "learning_rate": 1.3657469690815106e-07,
      "loss": 0.2900256872177124,
      "num_input_tokens_seen": 1414493472,
      "step": 27490,
      "train_runtime": 169205.223,
      "train_tokens_per_second": 8359.632
    },
    {
      "epoch": 1.8512658227848102,
      "grad_norm": 0.5970428755654892,
      "learning_rate": 1.359616093791677e-07,
      "loss": 0.3065678119659424,
      "num_input_tokens_seen": 1414768408,
      "step": 27495,
      "train_runtime": 169233.7341,
      "train_tokens_per_second": 8359.849
    },
    {
      "epoch": 1.8516024777807703,
      "grad_norm": 0.6631510961062705,
      "learning_rate": 1.353498820630106e-07,
      "loss": 0.2946380615234375,
      "num_input_tokens_seen": 1415021672,
      "step": 27500,
      "train_runtime": 169267.2081,
      "train_tokens_per_second": 8359.692
    },
    {
      "epoch": 1.8519391327767303,
      "grad_norm": 0.6528403078921385,
      "learning_rate": 1.3473951513074625e-07,
      "loss": 0.31060593128204345,
      "num_input_tokens_seen": 1415270272,
      "step": 27505,
      "train_runtime": 169298.782,
      "train_tokens_per_second": 8359.601
    },
    {
      "epoch": 1.8522757877726905,
      "grad_norm": 0.625132873120097,
      "learning_rate": 1.341305087530631e-07,
      "loss": 0.2826535701751709,
      "num_input_tokens_seen": 1415530752,
      "step": 27510,
      "train_runtime": 169329.2273,
      "train_tokens_per_second": 8359.636
    },
    {
      "epoch": 1.8526124427686508,
      "grad_norm": 0.7971832393972776,
      "learning_rate": 1.3352286310026663e-07,
      "loss": 0.3159646034240723,
      "num_input_tokens_seen": 1415773400,
      "step": 27515,
      "train_runtime": 169359.2017,
      "train_tokens_per_second": 8359.589
    },
    {
      "epoch": 1.8529490977646108,
      "grad_norm": 0.6704572480180694,
      "learning_rate": 1.329165783422831e-07,
      "loss": 0.3092250108718872,
      "num_input_tokens_seen": 1416023936,
      "step": 27520,
      "train_runtime": 169388.0082,
      "train_tokens_per_second": 8359.647
    },
    {
      "epoch": 1.8532857527605708,
      "grad_norm": 0.6555681851350031,
      "learning_rate": 1.3231165464865914e-07,
      "loss": 0.29388065338134767,
      "num_input_tokens_seen": 1416288176,
      "step": 27525,
      "train_runtime": 169421.6459,
      "train_tokens_per_second": 8359.547
    },
    {
      "epoch": 1.853622407756531,
      "grad_norm": 0.6061566524514764,
      "learning_rate": 1.3170809218855884e-07,
      "loss": 0.28916311264038086,
      "num_input_tokens_seen": 1416565928,
      "step": 27530,
      "train_runtime": 169451.994,
      "train_tokens_per_second": 8359.689
    },
    {
      "epoch": 1.8539590627524913,
      "grad_norm": 0.6640187380470108,
      "learning_rate": 1.3110589113076554e-07,
      "loss": 0.29465248584747317,
      "num_input_tokens_seen": 1416815680,
      "step": 27535,
      "train_runtime": 169482.5325,
      "train_tokens_per_second": 8359.656
    },
    {
      "epoch": 1.8542957177484514,
      "grad_norm": 0.6366128642691923,
      "learning_rate": 1.3050505164368566e-07,
      "loss": 0.28653922080993655,
      "num_input_tokens_seen": 1417087968,
      "step": 27540,
      "train_runtime": 169513.9371,
      "train_tokens_per_second": 8359.714
    },
    {
      "epoch": 1.8546323727444114,
      "grad_norm": 0.6617786211616568,
      "learning_rate": 1.299055738953392e-07,
      "loss": 0.2848658084869385,
      "num_input_tokens_seen": 1417310824,
      "step": 27545,
      "train_runtime": 169541.1305,
      "train_tokens_per_second": 8359.687
    },
    {
      "epoch": 1.8549690277403716,
      "grad_norm": 0.6778198148636001,
      "learning_rate": 1.293074580533704e-07,
      "loss": 0.2821719884872437,
      "num_input_tokens_seen": 1417548184,
      "step": 27550,
      "train_runtime": 169570.21,
      "train_tokens_per_second": 8359.653
    },
    {
      "epoch": 1.855305682736332,
      "grad_norm": 0.6836192742473766,
      "learning_rate": 1.2871070428503885e-07,
      "loss": 0.31565265655517577,
      "num_input_tokens_seen": 1417791584,
      "step": 27555,
      "train_runtime": 169603.103,
      "train_tokens_per_second": 8359.467
    },
    {
      "epoch": 1.855642337732292,
      "grad_norm": 0.5660042291143903,
      "learning_rate": 1.2811531275722543e-07,
      "loss": 0.3169874906539917,
      "num_input_tokens_seen": 1418071984,
      "step": 27560,
      "train_runtime": 169634.8171,
      "train_tokens_per_second": 8359.557
    },
    {
      "epoch": 1.855978992728252,
      "grad_norm": 0.7480710286090106,
      "learning_rate": 1.275212836364298e-07,
      "loss": 0.3053633213043213,
      "num_input_tokens_seen": 1418323752,
      "step": 27565,
      "train_runtime": 169667.6111,
      "train_tokens_per_second": 8359.425
    },
    {
      "epoch": 1.8563156477242122,
      "grad_norm": 0.6981017657511316,
      "learning_rate": 1.2692861708877014e-07,
      "loss": 0.2882718563079834,
      "num_input_tokens_seen": 1418551672,
      "step": 27570,
      "train_runtime": 169699.2019,
      "train_tokens_per_second": 8359.212
    },
    {
      "epoch": 1.8566523027201725,
      "grad_norm": 0.7111456883968743,
      "learning_rate": 1.2633731327998332e-07,
      "loss": 0.2841503143310547,
      "num_input_tokens_seen": 1418816168,
      "step": 27575,
      "train_runtime": 169727.1356,
      "train_tokens_per_second": 8359.395
    },
    {
      "epoch": 1.8569889577161325,
      "grad_norm": 0.8301723328675424,
      "learning_rate": 1.25747372375426e-07,
      "loss": 0.28955535888671874,
      "num_input_tokens_seen": 1419056968,
      "step": 27580,
      "train_runtime": 169757.6004,
      "train_tokens_per_second": 8359.313
    },
    {
      "epoch": 1.8573256127120925,
      "grad_norm": 0.7209966789409151,
      "learning_rate": 1.251587945400723e-07,
      "loss": 0.2950953722000122,
      "num_input_tokens_seen": 1419310528,
      "step": 27585,
      "train_runtime": 169788.9064,
      "train_tokens_per_second": 8359.265
    },
    {
      "epoch": 1.8576622677080528,
      "grad_norm": 0.6595023297015018,
      "learning_rate": 1.2457157993851832e-07,
      "loss": 0.2897037506103516,
      "num_input_tokens_seen": 1419575672,
      "step": 27590,
      "train_runtime": 169822.7447,
      "train_tokens_per_second": 8359.161
    },
    {
      "epoch": 1.857998922704013,
      "grad_norm": 0.6960625628492969,
      "learning_rate": 1.2398572873497438e-07,
      "loss": 0.29905047416687014,
      "num_input_tokens_seen": 1419825024,
      "step": 27595,
      "train_runtime": 169850.9198,
      "train_tokens_per_second": 8359.242
    },
    {
      "epoch": 1.858335577699973,
      "grad_norm": 0.7186942192910873,
      "learning_rate": 1.2340124109327334e-07,
      "loss": 0.302016019821167,
      "num_input_tokens_seen": 1420069760,
      "step": 27600,
      "train_runtime": 169881.5531,
      "train_tokens_per_second": 8359.176
    },
    {
      "epoch": 1.858672232695933,
      "grad_norm": 0.724108708038881,
      "learning_rate": 1.22818117176865e-07,
      "loss": 0.31241559982299805,
      "num_input_tokens_seen": 1420323344,
      "step": 27605,
      "train_runtime": 169913.7137,
      "train_tokens_per_second": 8359.086
    },
    {
      "epoch": 1.8590088876918933,
      "grad_norm": 0.615898829092621,
      "learning_rate": 1.2223635714881777e-07,
      "loss": 0.3039550304412842,
      "num_input_tokens_seen": 1420564072,
      "step": 27610,
      "train_runtime": 169942.5293,
      "train_tokens_per_second": 8359.085
    },
    {
      "epoch": 1.8593455426878536,
      "grad_norm": 0.5854338027549183,
      "learning_rate": 1.2165596117181931e-07,
      "loss": 0.28476815223693847,
      "num_input_tokens_seen": 1420843528,
      "step": 27615,
      "train_runtime": 169973.07,
      "train_tokens_per_second": 8359.227
    },
    {
      "epoch": 1.8596821976838136,
      "grad_norm": 0.652826919183899,
      "learning_rate": 1.2107692940817594e-07,
      "loss": 0.3045393466949463,
      "num_input_tokens_seen": 1421097616,
      "step": 27620,
      "train_runtime": 169998.6158,
      "train_tokens_per_second": 8359.466
    },
    {
      "epoch": 1.8600188526797736,
      "grad_norm": 0.7330234719786909,
      "learning_rate": 1.2049926201981088e-07,
      "loss": 0.28773958683013917,
      "num_input_tokens_seen": 1421352808,
      "step": 27625,
      "train_runtime": 170029.3361,
      "train_tokens_per_second": 8359.456
    },
    {
      "epoch": 1.8603555076757339,
      "grad_norm": 0.6134251284457598,
      "learning_rate": 1.199229591682677e-07,
      "loss": 0.3080180883407593,
      "num_input_tokens_seen": 1421630760,
      "step": 27630,
      "train_runtime": 170061.7637,
      "train_tokens_per_second": 8359.497
    },
    {
      "epoch": 1.8606921626716941,
      "grad_norm": 0.6871566614619191,
      "learning_rate": 1.193480210147069e-07,
      "loss": 0.3114048480987549,
      "num_input_tokens_seen": 1421899552,
      "step": 27635,
      "train_runtime": 170091.3421,
      "train_tokens_per_second": 8359.623
    },
    {
      "epoch": 1.8610288176676542,
      "grad_norm": 0.6935358181452087,
      "learning_rate": 1.1877444771991042e-07,
      "loss": 0.28173537254333497,
      "num_input_tokens_seen": 1422171648,
      "step": 27640,
      "train_runtime": 170122.1086,
      "train_tokens_per_second": 8359.711
    },
    {
      "epoch": 1.8613654726636142,
      "grad_norm": 0.6902386980676435,
      "learning_rate": 1.1820223944427279e-07,
      "loss": 0.27820525169372556,
      "num_input_tokens_seen": 1422411672,
      "step": 27645,
      "train_runtime": 170154.6027,
      "train_tokens_per_second": 8359.525
    },
    {
      "epoch": 1.8617021276595744,
      "grad_norm": 0.62711399561758,
      "learning_rate": 1.1763139634781262e-07,
      "loss": 0.2948620796203613,
      "num_input_tokens_seen": 1422687840,
      "step": 27650,
      "train_runtime": 170186.9139,
      "train_tokens_per_second": 8359.561
    },
    {
      "epoch": 1.8620387826555347,
      "grad_norm": 0.6323444192773547,
      "learning_rate": 1.1706191859016392e-07,
      "loss": 0.3267491340637207,
      "num_input_tokens_seen": 1422942576,
      "step": 27655,
      "train_runtime": 170218.6744,
      "train_tokens_per_second": 8359.497
    },
    {
      "epoch": 1.8623754376514947,
      "grad_norm": 0.6753353075325804,
      "learning_rate": 1.1649380633057927e-07,
      "loss": 0.2987608194351196,
      "num_input_tokens_seen": 1423192544,
      "step": 27660,
      "train_runtime": 170248.4114,
      "train_tokens_per_second": 8359.506
    },
    {
      "epoch": 1.8627120926474547,
      "grad_norm": 0.6885897862367918,
      "learning_rate": 1.1592705972792939e-07,
      "loss": 0.27299914360046384,
      "num_input_tokens_seen": 1423445512,
      "step": 27665,
      "train_runtime": 170277.1691,
      "train_tokens_per_second": 8359.579
    },
    {
      "epoch": 1.863048747643415,
      "grad_norm": 0.6471103463326967,
      "learning_rate": 1.1536167894070249e-07,
      "loss": 0.3060754299163818,
      "num_input_tokens_seen": 1423712960,
      "step": 27670,
      "train_runtime": 170306.5856,
      "train_tokens_per_second": 8359.706
    },
    {
      "epoch": 1.8633854026393752,
      "grad_norm": 0.6317101052333661,
      "learning_rate": 1.1479766412700655e-07,
      "loss": 0.31825709342956543,
      "num_input_tokens_seen": 1423979800,
      "step": 27675,
      "train_runtime": 170339.1344,
      "train_tokens_per_second": 8359.675
    },
    {
      "epoch": 1.8637220576353353,
      "grad_norm": 0.7168097661870647,
      "learning_rate": 1.1423501544456595e-07,
      "loss": 0.31889991760253905,
      "num_input_tokens_seen": 1424231240,
      "step": 27680,
      "train_runtime": 170371.1201,
      "train_tokens_per_second": 8359.581
    },
    {
      "epoch": 1.8640587126312953,
      "grad_norm": 0.6502339743988153,
      "learning_rate": 1.1367373305072315e-07,
      "loss": 0.28447527885437013,
      "num_input_tokens_seen": 1424488920,
      "step": 27685,
      "train_runtime": 170399.2507,
      "train_tokens_per_second": 8359.714
    },
    {
      "epoch": 1.8643953676272556,
      "grad_norm": 0.6256570262463526,
      "learning_rate": 1.1311381710244096e-07,
      "loss": 0.26739528179168703,
      "num_input_tokens_seen": 1424746832,
      "step": 27690,
      "train_runtime": 170432.9123,
      "train_tokens_per_second": 8359.576
    },
    {
      "epoch": 1.8647320226232158,
      "grad_norm": 0.7520783022609009,
      "learning_rate": 1.1255526775629521e-07,
      "loss": 0.2719376802444458,
      "num_input_tokens_seen": 1424987216,
      "step": 27695,
      "train_runtime": 170458.4008,
      "train_tokens_per_second": 8359.736
    },
    {
      "epoch": 1.8650686776191758,
      "grad_norm": 0.6651916457233429,
      "learning_rate": 1.1199808516848487e-07,
      "loss": 0.30454182624816895,
      "num_input_tokens_seen": 1425233840,
      "step": 27700,
      "train_runtime": 170490.5222,
      "train_tokens_per_second": 8359.607
    },
    {
      "epoch": 1.8654053326151359,
      "grad_norm": 0.6737702076141017,
      "learning_rate": 1.1144226949482306e-07,
      "loss": 0.2796602487564087,
      "num_input_tokens_seen": 1425479240,
      "step": 27705,
      "train_runtime": 170518.9871,
      "train_tokens_per_second": 8359.651
    },
    {
      "epoch": 1.865741987611096,
      "grad_norm": 0.6421541971760195,
      "learning_rate": 1.1088782089074157e-07,
      "loss": 0.2778012275695801,
      "num_input_tokens_seen": 1425737160,
      "step": 27710,
      "train_runtime": 170552.2963,
      "train_tokens_per_second": 8359.531
    },
    {
      "epoch": 1.8660786426070564,
      "grad_norm": 0.7064251508918062,
      "learning_rate": 1.1033473951129137e-07,
      "loss": 0.3050196170806885,
      "num_input_tokens_seen": 1425995088,
      "step": 27715,
      "train_runtime": 170585.7643,
      "train_tokens_per_second": 8359.403
    },
    {
      "epoch": 1.8664152976030164,
      "grad_norm": 0.6673001390753529,
      "learning_rate": 1.0978302551113929e-07,
      "loss": 0.31004350185394286,
      "num_input_tokens_seen": 1426265080,
      "step": 27720,
      "train_runtime": 170616.1357,
      "train_tokens_per_second": 8359.497
    },
    {
      "epoch": 1.8667519525989764,
      "grad_norm": 0.6487286273717094,
      "learning_rate": 1.0923267904457025e-07,
      "loss": 0.29609601497650145,
      "num_input_tokens_seen": 1426536208,
      "step": 27725,
      "train_runtime": 170647.0826,
      "train_tokens_per_second": 8359.57
    },
    {
      "epoch": 1.8670886075949367,
      "grad_norm": 0.7694798950617916,
      "learning_rate": 1.086837002654867e-07,
      "loss": 0.3116004467010498,
      "num_input_tokens_seen": 1426778168,
      "step": 27730,
      "train_runtime": 170674.2042,
      "train_tokens_per_second": 8359.659
    },
    {
      "epoch": 1.867425262590897,
      "grad_norm": 0.6769414819378207,
      "learning_rate": 1.0813608932740915e-07,
      "loss": 0.30857923030853274,
      "num_input_tokens_seen": 1427057848,
      "step": 27735,
      "train_runtime": 170707.2946,
      "train_tokens_per_second": 8359.677
    },
    {
      "epoch": 1.867761917586857,
      "grad_norm": 0.6048038576131638,
      "learning_rate": 1.0758984638347569e-07,
      "loss": 0.2895212173461914,
      "num_input_tokens_seen": 1427343376,
      "step": 27740,
      "train_runtime": 170738.6537,
      "train_tokens_per_second": 8359.814
    },
    {
      "epoch": 1.868098572582817,
      "grad_norm": 0.842875606628196,
      "learning_rate": 1.0704497158644022e-07,
      "loss": 0.32908222675323484,
      "num_input_tokens_seen": 1427595384,
      "step": 27745,
      "train_runtime": 170766.8228,
      "train_tokens_per_second": 8359.911
    },
    {
      "epoch": 1.8684352275787772,
      "grad_norm": 0.5996434204345845,
      "learning_rate": 1.0650146508867642e-07,
      "loss": 0.2838188648223877,
      "num_input_tokens_seen": 1427855248,
      "step": 27750,
      "train_runtime": 170796.8255,
      "train_tokens_per_second": 8359.964
    },
    {
      "epoch": 1.8687718825747375,
      "grad_norm": 0.6572371796535794,
      "learning_rate": 1.0595932704217327e-07,
      "loss": 0.29073357582092285,
      "num_input_tokens_seen": 1428115888,
      "step": 27755,
      "train_runtime": 170826.4886,
      "train_tokens_per_second": 8360.038
    },
    {
      "epoch": 1.8691085375706975,
      "grad_norm": 0.5933540796516603,
      "learning_rate": 1.0541855759853837e-07,
      "loss": 0.2728950023651123,
      "num_input_tokens_seen": 1428395728,
      "step": 27760,
      "train_runtime": 170857.1598,
      "train_tokens_per_second": 8360.175
    },
    {
      "epoch": 1.8694451925666575,
      "grad_norm": 0.5902125002398455,
      "learning_rate": 1.0487915690899685e-07,
      "loss": 0.2974870681762695,
      "num_input_tokens_seen": 1428654248,
      "step": 27765,
      "train_runtime": 170891.5763,
      "train_tokens_per_second": 8360.004
    },
    {
      "epoch": 1.8697818475626178,
      "grad_norm": 0.6444277459107227,
      "learning_rate": 1.0434112512438921e-07,
      "loss": 0.291212797164917,
      "num_input_tokens_seen": 1428900352,
      "step": 27770,
      "train_runtime": 170921.1057,
      "train_tokens_per_second": 8359.999
    },
    {
      "epoch": 1.870118502558578,
      "grad_norm": 0.713356292477479,
      "learning_rate": 1.0380446239517506e-07,
      "loss": 0.27637970447540283,
      "num_input_tokens_seen": 1429124192,
      "step": 27775,
      "train_runtime": 170950.153,
      "train_tokens_per_second": 8359.888
    },
    {
      "epoch": 1.870455157554538,
      "grad_norm": 0.7896566432327069,
      "learning_rate": 1.032691688714299e-07,
      "loss": 0.29166626930236816,
      "num_input_tokens_seen": 1429328080,
      "step": 27780,
      "train_runtime": 170979.5214,
      "train_tokens_per_second": 8359.645
    },
    {
      "epoch": 1.870791812550498,
      "grad_norm": 0.7019892607682099,
      "learning_rate": 1.0273524470284734e-07,
      "loss": 0.31655895709991455,
      "num_input_tokens_seen": 1429602056,
      "step": 27785,
      "train_runtime": 171009.1885,
      "train_tokens_per_second": 8359.797
    },
    {
      "epoch": 1.8711284675464583,
      "grad_norm": 0.6656024089826869,
      "learning_rate": 1.0220269003873739e-07,
      "loss": 0.2848045825958252,
      "num_input_tokens_seen": 1429868584,
      "step": 27790,
      "train_runtime": 171035.9965,
      "train_tokens_per_second": 8360.045
    },
    {
      "epoch": 1.8714651225424186,
      "grad_norm": 1.1322323734605937,
      "learning_rate": 1.0167150502802703e-07,
      "loss": 0.2794020175933838,
      "num_input_tokens_seen": 1430141200,
      "step": 27795,
      "train_runtime": 171064.3112,
      "train_tokens_per_second": 8360.255
    },
    {
      "epoch": 1.8718017775383786,
      "grad_norm": 0.6664880918497217,
      "learning_rate": 1.0114168981926075e-07,
      "loss": 0.31267361640930175,
      "num_input_tokens_seen": 1430404024,
      "step": 27800,
      "train_runtime": 171097.5254,
      "train_tokens_per_second": 8360.168
    },
    {
      "epoch": 1.8721384325343386,
      "grad_norm": 0.6335536709657157,
      "learning_rate": 1.006132445605995e-07,
      "loss": 0.2943727493286133,
      "num_input_tokens_seen": 1430649456,
      "step": 27805,
      "train_runtime": 171128.0099,
      "train_tokens_per_second": 8360.113
    },
    {
      "epoch": 1.872475087530299,
      "grad_norm": 0.6853183988025989,
      "learning_rate": 1.0008616939982118e-07,
      "loss": 0.28060824871063234,
      "num_input_tokens_seen": 1430913056,
      "step": 27810,
      "train_runtime": 171157.2242,
      "train_tokens_per_second": 8360.226
    },
    {
      "epoch": 1.8728117425262591,
      "grad_norm": 0.6911874347028841,
      "learning_rate": 9.956046448432177e-08,
      "loss": 0.2828995227813721,
      "num_input_tokens_seen": 1431170576,
      "step": 27815,
      "train_runtime": 171189.3387,
      "train_tokens_per_second": 8360.162
    },
    {
      "epoch": 1.8731483975222192,
      "grad_norm": 0.6722504586316683,
      "learning_rate": 9.90361299611109e-08,
      "loss": 0.2942378044128418,
      "num_input_tokens_seen": 1431447832,
      "step": 27820,
      "train_runtime": 171223.4201,
      "train_tokens_per_second": 8360.117
    },
    {
      "epoch": 1.8734850525181792,
      "grad_norm": 0.7158791364922268,
      "learning_rate": 9.851316597681959e-08,
      "loss": 0.31331501007080076,
      "num_input_tokens_seen": 1431695832,
      "step": 27825,
      "train_runtime": 171249.5387,
      "train_tokens_per_second": 8360.29
    },
    {
      "epoch": 1.8738217075141395,
      "grad_norm": 0.6415459401583329,
      "learning_rate": 9.799157267769088e-08,
      "loss": 0.2883334159851074,
      "num_input_tokens_seen": 1431968112,
      "step": 27830,
      "train_runtime": 171276.2362,
      "train_tokens_per_second": 8360.577
    },
    {
      "epoch": 1.8741583625100997,
      "grad_norm": 0.583518866328104,
      "learning_rate": 9.747135020958753e-08,
      "loss": 0.2998178005218506,
      "num_input_tokens_seen": 1432232064,
      "step": 27835,
      "train_runtime": 171306.4825,
      "train_tokens_per_second": 8360.641
    },
    {
      "epoch": 1.8744950175060597,
      "grad_norm": 0.6530259589317329,
      "learning_rate": 9.695249871798818e-08,
      "loss": 0.27299132347106936,
      "num_input_tokens_seen": 1432473696,
      "step": 27840,
      "train_runtime": 171336.0778,
      "train_tokens_per_second": 8360.607
    },
    {
      "epoch": 1.8748316725020198,
      "grad_norm": 0.628568425668609,
      "learning_rate": 9.643501834798786e-08,
      "loss": 0.2936988353729248,
      "num_input_tokens_seen": 1432715320,
      "step": 27845,
      "train_runtime": 171365.6453,
      "train_tokens_per_second": 8360.575
    },
    {
      "epoch": 1.8751683274979802,
      "grad_norm": 0.6493738907144728,
      "learning_rate": 9.591890924429858e-08,
      "loss": 0.29607720375061036,
      "num_input_tokens_seen": 1433001632,
      "step": 27850,
      "train_runtime": 171397.2974,
      "train_tokens_per_second": 8360.701
    },
    {
      "epoch": 1.8755049824939403,
      "grad_norm": 0.6966150606857303,
      "learning_rate": 9.540417155124826e-08,
      "loss": 0.2797668933868408,
      "num_input_tokens_seen": 1433230032,
      "step": 27855,
      "train_runtime": 171427.0139,
      "train_tokens_per_second": 8360.584
    },
    {
      "epoch": 1.8758416374899003,
      "grad_norm": 0.6056753319348104,
      "learning_rate": 9.489080541278117e-08,
      "loss": 0.2972454786300659,
      "num_input_tokens_seen": 1433453800,
      "step": 27860,
      "train_runtime": 171454.9985,
      "train_tokens_per_second": 8360.525
    },
    {
      "epoch": 1.8761782924858605,
      "grad_norm": 0.7330618051770845,
      "learning_rate": 9.437881097246026e-08,
      "loss": 0.30346534252166746,
      "num_input_tokens_seen": 1433714672,
      "step": 27865,
      "train_runtime": 171485.6484,
      "train_tokens_per_second": 8360.552
    },
    {
      "epoch": 1.8765149474818208,
      "grad_norm": 0.6525489375354859,
      "learning_rate": 9.3868188373461e-08,
      "loss": 0.2962563753128052,
      "num_input_tokens_seen": 1433967032,
      "step": 27870,
      "train_runtime": 171515.29,
      "train_tokens_per_second": 8360.578
    },
    {
      "epoch": 1.8768516024777808,
      "grad_norm": 0.8000297024015047,
      "learning_rate": 9.335893775857918e-08,
      "loss": 0.3038081169128418,
      "num_input_tokens_seen": 1434210952,
      "step": 27875,
      "train_runtime": 171543.2297,
      "train_tokens_per_second": 8360.639
    },
    {
      "epoch": 1.8771882574737409,
      "grad_norm": 0.6522608352201992,
      "learning_rate": 9.285105927022309e-08,
      "loss": 0.2902660608291626,
      "num_input_tokens_seen": 1434467544,
      "step": 27880,
      "train_runtime": 171575.0568,
      "train_tokens_per_second": 8360.583
    },
    {
      "epoch": 1.877524912469701,
      "grad_norm": 0.7006782465454449,
      "learning_rate": 9.234455305042133e-08,
      "loss": 0.28328585624694824,
      "num_input_tokens_seen": 1434694072,
      "step": 27885,
      "train_runtime": 171602.6172,
      "train_tokens_per_second": 8360.561
    },
    {
      "epoch": 1.8778615674656614,
      "grad_norm": 0.7508504965604864,
      "learning_rate": 9.183941924081563e-08,
      "loss": 0.2881436347961426,
      "num_input_tokens_seen": 1434953080,
      "step": 27890,
      "train_runtime": 171632.9064,
      "train_tokens_per_second": 8360.594
    },
    {
      "epoch": 1.8781982224616214,
      "grad_norm": 0.6625638543565165,
      "learning_rate": 9.133565798266574e-08,
      "loss": 0.28098812103271487,
      "num_input_tokens_seen": 1435222416,
      "step": 27895,
      "train_runtime": 171662.5444,
      "train_tokens_per_second": 8360.72
    },
    {
      "epoch": 1.8785348774575814,
      "grad_norm": 0.6138041190145097,
      "learning_rate": 9.083326941684622e-08,
      "loss": 0.31454222202301024,
      "num_input_tokens_seen": 1435503648,
      "step": 27900,
      "train_runtime": 171689.4753,
      "train_tokens_per_second": 8361.046
    },
    {
      "epoch": 1.8788715324535417,
      "grad_norm": 0.7020660148701215,
      "learning_rate": 9.033225368384912e-08,
      "loss": 0.3046130895614624,
      "num_input_tokens_seen": 1435737248,
      "step": 27905,
      "train_runtime": 171718.5238,
      "train_tokens_per_second": 8360.992
    },
    {
      "epoch": 1.879208187449502,
      "grad_norm": 0.6371900629940745,
      "learning_rate": 8.98326109237807e-08,
      "loss": 0.2739833116531372,
      "num_input_tokens_seen": 1435983072,
      "step": 27910,
      "train_runtime": 171748.9217,
      "train_tokens_per_second": 8360.944
    },
    {
      "epoch": 1.879544842445462,
      "grad_norm": 0.6069671180898712,
      "learning_rate": 8.933434127636586e-08,
      "loss": 0.3169389724731445,
      "num_input_tokens_seen": 1436249368,
      "step": 27915,
      "train_runtime": 171780.5142,
      "train_tokens_per_second": 8360.956
    },
    {
      "epoch": 1.879881497441422,
      "grad_norm": 0.6319584994083526,
      "learning_rate": 8.883744488094314e-08,
      "loss": 0.3080581188201904,
      "num_input_tokens_seen": 1436499456,
      "step": 27920,
      "train_runtime": 171809.4554,
      "train_tokens_per_second": 8361.003
    },
    {
      "epoch": 1.8802181524373822,
      "grad_norm": 0.6275066493638901,
      "learning_rate": 8.834192187646917e-08,
      "loss": 0.31001996994018555,
      "num_input_tokens_seen": 1436779856,
      "step": 27925,
      "train_runtime": 171839.5408,
      "train_tokens_per_second": 8361.171
    },
    {
      "epoch": 1.8805548074333425,
      "grad_norm": 0.7413279231173322,
      "learning_rate": 8.784777240151365e-08,
      "loss": 0.3019866466522217,
      "num_input_tokens_seen": 1437049424,
      "step": 27930,
      "train_runtime": 171873.9378,
      "train_tokens_per_second": 8361.067
    },
    {
      "epoch": 1.8808914624293025,
      "grad_norm": 0.6715834840845668,
      "learning_rate": 8.73549965942655e-08,
      "loss": 0.3016371250152588,
      "num_input_tokens_seen": 1437312064,
      "step": 27935,
      "train_runtime": 171903.3209,
      "train_tokens_per_second": 8361.165
    },
    {
      "epoch": 1.8812281174252625,
      "grad_norm": 1.0464276702959467,
      "learning_rate": 8.686359459252725e-08,
      "loss": 0.2919389963150024,
      "num_input_tokens_seen": 1437562368,
      "step": 27940,
      "train_runtime": 171936.6819,
      "train_tokens_per_second": 8360.999
    },
    {
      "epoch": 1.8815647724212228,
      "grad_norm": 0.545470540026027,
      "learning_rate": 8.637356653371787e-08,
      "loss": 0.28544371128082274,
      "num_input_tokens_seen": 1437836464,
      "step": 27945,
      "train_runtime": 171971.6389,
      "train_tokens_per_second": 8360.893
    },
    {
      "epoch": 1.881901427417183,
      "grad_norm": 0.7366164873093676,
      "learning_rate": 8.588491255487163e-08,
      "loss": 0.2741818428039551,
      "num_input_tokens_seen": 1438078008,
      "step": 27950,
      "train_runtime": 171999.6068,
      "train_tokens_per_second": 8360.938
    },
    {
      "epoch": 1.882238082413143,
      "grad_norm": 0.7084434505715027,
      "learning_rate": 8.539763279264035e-08,
      "loss": 0.28293395042419434,
      "num_input_tokens_seen": 1438322952,
      "step": 27955,
      "train_runtime": 172026.6746,
      "train_tokens_per_second": 8361.046
    },
    {
      "epoch": 1.882574737409103,
      "grad_norm": 0.642108535228344,
      "learning_rate": 8.49117273832889e-08,
      "loss": 0.3127452373504639,
      "num_input_tokens_seen": 1438603888,
      "step": 27960,
      "train_runtime": 172060.0409,
      "train_tokens_per_second": 8361.057
    },
    {
      "epoch": 1.8829113924050633,
      "grad_norm": 0.5764868188304534,
      "learning_rate": 8.442719646270082e-08,
      "loss": 0.28748598098754885,
      "num_input_tokens_seen": 1438858976,
      "step": 27965,
      "train_runtime": 172087.5873,
      "train_tokens_per_second": 8361.201
    },
    {
      "epoch": 1.8832480474010236,
      "grad_norm": 0.640610272018838,
      "learning_rate": 8.394404016637159e-08,
      "loss": 0.28852548599243166,
      "num_input_tokens_seen": 1439112424,
      "step": 27970,
      "train_runtime": 172115.0986,
      "train_tokens_per_second": 8361.337
    },
    {
      "epoch": 1.8835847023969836,
      "grad_norm": 0.6729805763134907,
      "learning_rate": 8.346225862941592e-08,
      "loss": 0.2824716567993164,
      "num_input_tokens_seen": 1439351152,
      "step": 27975,
      "train_runtime": 172142.0883,
      "train_tokens_per_second": 8361.413
    },
    {
      "epoch": 1.8839213573929436,
      "grad_norm": 0.6239557324777648,
      "learning_rate": 8.298185198656161e-08,
      "loss": 0.3113942384719849,
      "num_input_tokens_seen": 1439626664,
      "step": 27980,
      "train_runtime": 172172.3439,
      "train_tokens_per_second": 8361.544
    },
    {
      "epoch": 1.884258012388904,
      "grad_norm": 0.6312813526979082,
      "learning_rate": 8.25028203721534e-08,
      "loss": 0.27808122634887694,
      "num_input_tokens_seen": 1439861568,
      "step": 27985,
      "train_runtime": 172201.5909,
      "train_tokens_per_second": 8361.488
    },
    {
      "epoch": 1.8845946673848641,
      "grad_norm": 0.6585657705453929,
      "learning_rate": 8.202516392015081e-08,
      "loss": 0.2969624996185303,
      "num_input_tokens_seen": 1440132408,
      "step": 27990,
      "train_runtime": 172232.928,
      "train_tokens_per_second": 8361.539
    },
    {
      "epoch": 1.8849313223808242,
      "grad_norm": 1.3734288521267668,
      "learning_rate": 8.154888276412865e-08,
      "loss": 0.3078592777252197,
      "num_input_tokens_seen": 1440375624,
      "step": 27995,
      "train_runtime": 172261.4356,
      "train_tokens_per_second": 8361.568
    },
    {
      "epoch": 1.8852679773767842,
      "grad_norm": 0.6272189547251327,
      "learning_rate": 8.10739770372776e-08,
      "loss": 0.2678866386413574,
      "num_input_tokens_seen": 1440652320,
      "step": 28000,
      "train_runtime": 172286.8661,
      "train_tokens_per_second": 8361.939
    },
    {
      "epoch": 1.8856046323727444,
      "grad_norm": 0.549250362085971,
      "learning_rate": 8.060044687240365e-08,
      "loss": 0.2751821756362915,
      "num_input_tokens_seen": 1440944280,
      "step": 28005,
      "train_runtime": 172320.6513,
      "train_tokens_per_second": 8361.994
    },
    {
      "epoch": 1.8859412873687047,
      "grad_norm": 0.5880050401758855,
      "learning_rate": 8.012829240192754e-08,
      "loss": 0.3051964998245239,
      "num_input_tokens_seen": 1441194184,
      "step": 28010,
      "train_runtime": 172351.0567,
      "train_tokens_per_second": 8361.969
    },
    {
      "epoch": 1.8862779423646647,
      "grad_norm": 0.5989400517804865,
      "learning_rate": 7.9657513757887e-08,
      "loss": 0.26883363723754883,
      "num_input_tokens_seen": 1441457792,
      "step": 28015,
      "train_runtime": 172378.0145,
      "train_tokens_per_second": 8362.19
    },
    {
      "epoch": 1.8866145973606248,
      "grad_norm": 0.5808122795929139,
      "learning_rate": 7.918811107193225e-08,
      "loss": 0.27985458374023436,
      "num_input_tokens_seen": 1441745680,
      "step": 28020,
      "train_runtime": 172410.1431,
      "train_tokens_per_second": 8362.302
    },
    {
      "epoch": 1.886951252356585,
      "grad_norm": 0.74116355614521,
      "learning_rate": 7.872008447533109e-08,
      "loss": 0.30963473320007323,
      "num_input_tokens_seen": 1442007880,
      "step": 28025,
      "train_runtime": 172440.6935,
      "train_tokens_per_second": 8362.341
    },
    {
      "epoch": 1.8872879073525453,
      "grad_norm": 0.7115990328455432,
      "learning_rate": 7.825343409896547e-08,
      "loss": 0.3052254915237427,
      "num_input_tokens_seen": 1442264008,
      "step": 28030,
      "train_runtime": 172474.2274,
      "train_tokens_per_second": 8362.2
    },
    {
      "epoch": 1.8876245623485053,
      "grad_norm": 0.5921860098481945,
      "learning_rate": 7.778816007333268e-08,
      "loss": 0.30379338264465333,
      "num_input_tokens_seen": 1442525904,
      "step": 28035,
      "train_runtime": 172503.1968,
      "train_tokens_per_second": 8362.314
    },
    {
      "epoch": 1.8879612173444653,
      "grad_norm": 0.7036202819409253,
      "learning_rate": 7.73242625285453e-08,
      "loss": 0.27389404773712156,
      "num_input_tokens_seen": 1442786376,
      "step": 28040,
      "train_runtime": 172531.5184,
      "train_tokens_per_second": 8362.451
    },
    {
      "epoch": 1.8882978723404256,
      "grad_norm": 0.6341700625869245,
      "learning_rate": 7.686174159433069e-08,
      "loss": 0.30092084407806396,
      "num_input_tokens_seen": 1443034448,
      "step": 28045,
      "train_runtime": 172563.0257,
      "train_tokens_per_second": 8362.362
    },
    {
      "epoch": 1.8886345273363858,
      "grad_norm": 0.5891787410854975,
      "learning_rate": 7.640059740003091e-08,
      "loss": 0.2716386318206787,
      "num_input_tokens_seen": 1443297264,
      "step": 28050,
      "train_runtime": 172593.4413,
      "train_tokens_per_second": 8362.411
    },
    {
      "epoch": 1.8889711823323458,
      "grad_norm": 0.7884019818538921,
      "learning_rate": 7.594083007460396e-08,
      "loss": 0.3018721342086792,
      "num_input_tokens_seen": 1443531280,
      "step": 28055,
      "train_runtime": 172623.7844,
      "train_tokens_per_second": 8362.297
    },
    {
      "epoch": 1.8893078373283059,
      "grad_norm": 0.6288179383622766,
      "learning_rate": 7.54824397466225e-08,
      "loss": 0.28793673515319823,
      "num_input_tokens_seen": 1443806464,
      "step": 28060,
      "train_runtime": 172654.1223,
      "train_tokens_per_second": 8362.421
    },
    {
      "epoch": 1.8896444923242661,
      "grad_norm": 0.6813386907275393,
      "learning_rate": 7.502542654427347e-08,
      "loss": 0.2864276647567749,
      "num_input_tokens_seen": 1444051192,
      "step": 28065,
      "train_runtime": 172681.4945,
      "train_tokens_per_second": 8362.513
    },
    {
      "epoch": 1.8899811473202264,
      "grad_norm": 0.6412087690977896,
      "learning_rate": 7.456979059535907e-08,
      "loss": 0.2956748723983765,
      "num_input_tokens_seen": 1444307384,
      "step": 28070,
      "train_runtime": 172709.8888,
      "train_tokens_per_second": 8362.621
    },
    {
      "epoch": 1.8903178023161864,
      "grad_norm": 0.7716978314825256,
      "learning_rate": 7.411553202729682e-08,
      "loss": 0.2955905914306641,
      "num_input_tokens_seen": 1444523336,
      "step": 28075,
      "train_runtime": 172738.4828,
      "train_tokens_per_second": 8362.487
    },
    {
      "epoch": 1.8906544573121464,
      "grad_norm": 0.6496381092811266,
      "learning_rate": 7.3662650967119e-08,
      "loss": 0.28202505111694337,
      "num_input_tokens_seen": 1444762512,
      "step": 28080,
      "train_runtime": 172769.7575,
      "train_tokens_per_second": 8362.358
    },
    {
      "epoch": 1.8909911123081067,
      "grad_norm": 0.6208795860790407,
      "learning_rate": 7.321114754147151e-08,
      "loss": 0.27320065498352053,
      "num_input_tokens_seen": 1445032432,
      "step": 28085,
      "train_runtime": 172799.2456,
      "train_tokens_per_second": 8362.493
    },
    {
      "epoch": 1.891327767304067,
      "grad_norm": 0.7380939508263371,
      "learning_rate": 7.276102187661615e-08,
      "loss": 0.3193091869354248,
      "num_input_tokens_seen": 1445288856,
      "step": 28090,
      "train_runtime": 172828.0552,
      "train_tokens_per_second": 8362.582
    },
    {
      "epoch": 1.891664422300027,
      "grad_norm": 0.6214240676447897,
      "learning_rate": 7.231227409842944e-08,
      "loss": 0.29962007999420165,
      "num_input_tokens_seen": 1445561832,
      "step": 28095,
      "train_runtime": 172858.0289,
      "train_tokens_per_second": 8362.712
    },
    {
      "epoch": 1.892001077295987,
      "grad_norm": 0.6139294507435995,
      "learning_rate": 7.186490433240156e-08,
      "loss": 0.31186728477478026,
      "num_input_tokens_seen": 1445792400,
      "step": 28100,
      "train_runtime": 172887.8279,
      "train_tokens_per_second": 8362.604
    },
    {
      "epoch": 1.8923377322919472,
      "grad_norm": 0.7373100880581402,
      "learning_rate": 7.14189127036391e-08,
      "loss": 0.2869421482086182,
      "num_input_tokens_seen": 1446043864,
      "step": 28105,
      "train_runtime": 172917.6046,
      "train_tokens_per_second": 8362.618
    },
    {
      "epoch": 1.8926743872879075,
      "grad_norm": 0.7399910566824656,
      "learning_rate": 7.097429933686118e-08,
      "loss": 0.31017193794250486,
      "num_input_tokens_seen": 1446283080,
      "step": 28110,
      "train_runtime": 172948.9682,
      "train_tokens_per_second": 8362.485
    },
    {
      "epoch": 1.8930110422838675,
      "grad_norm": 0.8000770375931352,
      "learning_rate": 7.053106435640333e-08,
      "loss": 0.30927815437316897,
      "num_input_tokens_seen": 1446558816,
      "step": 28115,
      "train_runtime": 172978.9877,
      "train_tokens_per_second": 8362.627
    },
    {
      "epoch": 1.8933476972798275,
      "grad_norm": 0.6100792107252664,
      "learning_rate": 7.008920788621366e-08,
      "loss": 0.29720795154571533,
      "num_input_tokens_seen": 1446817496,
      "step": 28120,
      "train_runtime": 173005.8506,
      "train_tokens_per_second": 8362.824
    },
    {
      "epoch": 1.8936843522757878,
      "grad_norm": 0.6874319986157098,
      "learning_rate": 6.964873004985717e-08,
      "loss": 0.29988856315612794,
      "num_input_tokens_seen": 1447075896,
      "step": 28125,
      "train_runtime": 173036.3231,
      "train_tokens_per_second": 8362.845
    },
    {
      "epoch": 1.894021007271748,
      "grad_norm": 0.6676469702324634,
      "learning_rate": 6.920963097051148e-08,
      "loss": 0.3000833749771118,
      "num_input_tokens_seen": 1447330296,
      "step": 28130,
      "train_runtime": 173064.5119,
      "train_tokens_per_second": 8362.953
    },
    {
      "epoch": 1.894357662267708,
      "grad_norm": 0.7328502805578202,
      "learning_rate": 6.877191077096945e-08,
      "loss": 0.28713183403015136,
      "num_input_tokens_seen": 1447578128,
      "step": 28135,
      "train_runtime": 173091.6256,
      "train_tokens_per_second": 8363.074
    },
    {
      "epoch": 1.894694317263668,
      "grad_norm": 0.7209965911108205,
      "learning_rate": 6.833556957363818e-08,
      "loss": 0.3032160997390747,
      "num_input_tokens_seen": 1447831488,
      "step": 28140,
      "train_runtime": 173120.1182,
      "train_tokens_per_second": 8363.161
    },
    {
      "epoch": 1.8950309722596284,
      "grad_norm": 0.6487910766072128,
      "learning_rate": 6.790060750053895e-08,
      "loss": 0.2913081407546997,
      "num_input_tokens_seen": 1448086024,
      "step": 28145,
      "train_runtime": 173151.7315,
      "train_tokens_per_second": 8363.104
    },
    {
      "epoch": 1.8953676272555886,
      "grad_norm": 0.5746335024861681,
      "learning_rate": 6.746702467330723e-08,
      "loss": 0.2694845199584961,
      "num_input_tokens_seen": 1448358376,
      "step": 28150,
      "train_runtime": 173183.05,
      "train_tokens_per_second": 8363.165
    },
    {
      "epoch": 1.8957042822515486,
      "grad_norm": 0.6044873948114485,
      "learning_rate": 6.703482121319383e-08,
      "loss": 0.29488723278045653,
      "num_input_tokens_seen": 1448618416,
      "step": 28155,
      "train_runtime": 173207.2617,
      "train_tokens_per_second": 8363.497
    },
    {
      "epoch": 1.8960409372475087,
      "grad_norm": 0.6628795990703278,
      "learning_rate": 6.660399724106204e-08,
      "loss": 0.30855042934417726,
      "num_input_tokens_seen": 1448899128,
      "step": 28160,
      "train_runtime": 173236.6657,
      "train_tokens_per_second": 8363.698
    },
    {
      "epoch": 1.896377592243469,
      "grad_norm": 0.6251799058934738,
      "learning_rate": 6.617455287739161e-08,
      "loss": 0.29090232849121095,
      "num_input_tokens_seen": 1449174440,
      "step": 28165,
      "train_runtime": 173273.6608,
      "train_tokens_per_second": 8363.501
    },
    {
      "epoch": 1.8967142472394292,
      "grad_norm": 0.6062828212002128,
      "learning_rate": 6.574648824227369e-08,
      "loss": 0.3005927324295044,
      "num_input_tokens_seen": 1449439920,
      "step": 28170,
      "train_runtime": 173305.0668,
      "train_tokens_per_second": 8363.517
    },
    {
      "epoch": 1.8970509022353892,
      "grad_norm": 0.6619354348653433,
      "learning_rate": 6.5319803455417e-08,
      "loss": 0.3071671724319458,
      "num_input_tokens_seen": 1449700232,
      "step": 28175,
      "train_runtime": 173338.1296,
      "train_tokens_per_second": 8363.424
    },
    {
      "epoch": 1.8973875572313492,
      "grad_norm": 0.7291761942642166,
      "learning_rate": 6.489449863614106e-08,
      "loss": 0.32452781200408937,
      "num_input_tokens_seen": 1449955424,
      "step": 28180,
      "train_runtime": 173368.4513,
      "train_tokens_per_second": 8363.433
    },
    {
      "epoch": 1.8977242122273095,
      "grad_norm": 0.601464000092221,
      "learning_rate": 6.44705739033824e-08,
      "loss": 0.295467734336853,
      "num_input_tokens_seen": 1450213576,
      "step": 28185,
      "train_runtime": 173401.4199,
      "train_tokens_per_second": 8363.332
    },
    {
      "epoch": 1.8980608672232697,
      "grad_norm": 0.7166403572754645,
      "learning_rate": 6.404802937568899e-08,
      "loss": 0.2911070346832275,
      "num_input_tokens_seen": 1450452016,
      "step": 28190,
      "train_runtime": 173435.3719,
      "train_tokens_per_second": 8363.069
    },
    {
      "epoch": 1.8983975222192297,
      "grad_norm": 0.6003405288033064,
      "learning_rate": 6.362686517122463e-08,
      "loss": 0.2703869342803955,
      "num_input_tokens_seen": 1450703768,
      "step": 28195,
      "train_runtime": 173463.3682,
      "train_tokens_per_second": 8363.171
    },
    {
      "epoch": 1.8987341772151898,
      "grad_norm": 0.6484101391084709,
      "learning_rate": 6.320708140776676e-08,
      "loss": 0.2854024410247803,
      "num_input_tokens_seen": 1450956328,
      "step": 28200,
      "train_runtime": 173497.2109,
      "train_tokens_per_second": 8362.995
    },
    {
      "epoch": 1.89907083221115,
      "grad_norm": 0.6256511326716879,
      "learning_rate": 6.278867820270596e-08,
      "loss": 0.3061659574508667,
      "num_input_tokens_seen": 1451233224,
      "step": 28205,
      "train_runtime": 173523.9897,
      "train_tokens_per_second": 8363.3
    },
    {
      "epoch": 1.8994074872071103,
      "grad_norm": 0.7361092483119689,
      "learning_rate": 6.237165567304749e-08,
      "loss": 0.29479451179504396,
      "num_input_tokens_seen": 1451478768,
      "step": 28210,
      "train_runtime": 173556.387,
      "train_tokens_per_second": 8363.154
    },
    {
      "epoch": 1.8997441422030703,
      "grad_norm": 0.5768041348922974,
      "learning_rate": 6.19560139354114e-08,
      "loss": 0.3085838556289673,
      "num_input_tokens_seen": 1451775976,
      "step": 28215,
      "train_runtime": 173590.6848,
      "train_tokens_per_second": 8363.214
    },
    {
      "epoch": 1.9000807971990303,
      "grad_norm": 0.6133101566274559,
      "learning_rate": 6.154175310602916e-08,
      "loss": 0.3101807117462158,
      "num_input_tokens_seen": 1452042136,
      "step": 28220,
      "train_runtime": 173619.807,
      "train_tokens_per_second": 8363.344
    },
    {
      "epoch": 1.9004174521949906,
      "grad_norm": 0.6789391032674356,
      "learning_rate": 6.112887330074812e-08,
      "loss": 0.2886934757232666,
      "num_input_tokens_seen": 1452280408,
      "step": 28225,
      "train_runtime": 173650.859,
      "train_tokens_per_second": 8363.22
    },
    {
      "epoch": 1.9007541071909508,
      "grad_norm": 0.7482517064791319,
      "learning_rate": 6.07173746350287e-08,
      "loss": 0.2905031681060791,
      "num_input_tokens_seen": 1452548216,
      "step": 28230,
      "train_runtime": 173684.1794,
      "train_tokens_per_second": 8363.158
    },
    {
      "epoch": 1.9010907621869109,
      "grad_norm": 0.6305788946379265,
      "learning_rate": 6.030725722394548e-08,
      "loss": 0.3051170349121094,
      "num_input_tokens_seen": 1452820656,
      "step": 28235,
      "train_runtime": 173714.3716,
      "train_tokens_per_second": 8363.273
    },
    {
      "epoch": 1.901427417182871,
      "grad_norm": 0.7817188353852436,
      "learning_rate": 5.98985211821862e-08,
      "loss": 0.31268725395202634,
      "num_input_tokens_seen": 1453068200,
      "step": 28240,
      "train_runtime": 173748.6852,
      "train_tokens_per_second": 8363.046
    },
    {
      "epoch": 1.9017640721788311,
      "grad_norm": 0.6607385171559538,
      "learning_rate": 5.9491166624052185e-08,
      "loss": 0.28387644290924074,
      "num_input_tokens_seen": 1453349240,
      "step": 28245,
      "train_runtime": 173778.367,
      "train_tokens_per_second": 8363.235
    },
    {
      "epoch": 1.9021007271747914,
      "grad_norm": 0.6814552299070368,
      "learning_rate": 5.908519366345955e-08,
      "loss": 0.2934247970581055,
      "num_input_tokens_seen": 1453595832,
      "step": 28250,
      "train_runtime": 173814.8541,
      "train_tokens_per_second": 8362.898
    },
    {
      "epoch": 1.9024373821707514,
      "grad_norm": 0.6313679996327783,
      "learning_rate": 5.868060241393747e-08,
      "loss": 0.3159658670425415,
      "num_input_tokens_seen": 1453881456,
      "step": 28255,
      "train_runtime": 173844.7177,
      "train_tokens_per_second": 8363.104
    },
    {
      "epoch": 1.9027740371667115,
      "grad_norm": 0.7054072806490601,
      "learning_rate": 5.8277392988627665e-08,
      "loss": 0.30750389099121095,
      "num_input_tokens_seen": 1454131648,
      "step": 28260,
      "train_runtime": 173876.3091,
      "train_tokens_per_second": 8363.023
    },
    {
      "epoch": 1.9031106921626717,
      "grad_norm": 0.8052618487208849,
      "learning_rate": 5.787556550028772e-08,
      "loss": 0.27507839202880857,
      "num_input_tokens_seen": 1454370296,
      "step": 28265,
      "train_runtime": 173904.318,
      "train_tokens_per_second": 8363.049
    },
    {
      "epoch": 1.903447347158632,
      "grad_norm": 0.6051956223827512,
      "learning_rate": 5.747512006128553e-08,
      "loss": 0.2998154878616333,
      "num_input_tokens_seen": 1454655792,
      "step": 28270,
      "train_runtime": 173937.6668,
      "train_tokens_per_second": 8363.087
    },
    {
      "epoch": 1.903784002154592,
      "grad_norm": 0.7214280069571142,
      "learning_rate": 5.707605678360595e-08,
      "loss": 0.2961782693862915,
      "num_input_tokens_seen": 1454923472,
      "step": 28275,
      "train_runtime": 173968.0263,
      "train_tokens_per_second": 8363.166
    },
    {
      "epoch": 1.904120657150552,
      "grad_norm": 0.5891759604969541,
      "learning_rate": 5.6678375778845276e-08,
      "loss": 0.2948455333709717,
      "num_input_tokens_seen": 1455202584,
      "step": 28280,
      "train_runtime": 173999.8269,
      "train_tokens_per_second": 8363.242
    },
    {
      "epoch": 1.9044573121465123,
      "grad_norm": 0.578324714855599,
      "learning_rate": 5.6282077158213435e-08,
      "loss": 0.2771736145019531,
      "num_input_tokens_seen": 1455472800,
      "step": 28285,
      "train_runtime": 174026.6018,
      "train_tokens_per_second": 8363.508
    },
    {
      "epoch": 1.9047939671424725,
      "grad_norm": 0.6792249274784136,
      "learning_rate": 5.588716103253511e-08,
      "loss": 0.3008836507797241,
      "num_input_tokens_seen": 1455741984,
      "step": 28290,
      "train_runtime": 174060.5216,
      "train_tokens_per_second": 8363.424
    },
    {
      "epoch": 1.9051306221384325,
      "grad_norm": 0.6787192511979049,
      "learning_rate": 5.549362751224585e-08,
      "loss": 0.3226099967956543,
      "num_input_tokens_seen": 1455981752,
      "step": 28295,
      "train_runtime": 174095.0545,
      "train_tokens_per_second": 8363.143
    },
    {
      "epoch": 1.9054672771343926,
      "grad_norm": 0.7725707231018043,
      "learning_rate": 5.5101476707397075e-08,
      "loss": 0.29603376388549807,
      "num_input_tokens_seen": 1456246336,
      "step": 28300,
      "train_runtime": 174126.7403,
      "train_tokens_per_second": 8363.14
    },
    {
      "epoch": 1.9058039321303528,
      "grad_norm": 0.584446634807504,
      "learning_rate": 5.471070872765216e-08,
      "loss": 0.3060909271240234,
      "num_input_tokens_seen": 1456504136,
      "step": 28305,
      "train_runtime": 174155.6331,
      "train_tokens_per_second": 8363.233
    },
    {
      "epoch": 1.906140587126313,
      "grad_norm": 0.7829902655656001,
      "learning_rate": 5.432132368228815e-08,
      "loss": 0.3088386535644531,
      "num_input_tokens_seen": 1456782176,
      "step": 28310,
      "train_runtime": 174184.7382,
      "train_tokens_per_second": 8363.432
    },
    {
      "epoch": 1.906477242122273,
      "grad_norm": 0.6168081956101524,
      "learning_rate": 5.393332168019572e-08,
      "loss": 0.2845174789428711,
      "num_input_tokens_seen": 1457055448,
      "step": 28315,
      "train_runtime": 174215.3052,
      "train_tokens_per_second": 8363.533
    },
    {
      "epoch": 1.9068138971182331,
      "grad_norm": 0.6506271383876459,
      "learning_rate": 5.354670282987695e-08,
      "loss": 0.323134183883667,
      "num_input_tokens_seen": 1457346416,
      "step": 28320,
      "train_runtime": 174247.5859,
      "train_tokens_per_second": 8363.653
    },
    {
      "epoch": 1.9071505521141934,
      "grad_norm": 0.6387969334887728,
      "learning_rate": 5.316146723945037e-08,
      "loss": 0.31044270992279055,
      "num_input_tokens_seen": 1457608992,
      "step": 28325,
      "train_runtime": 174276.8766,
      "train_tokens_per_second": 8363.754
    },
    {
      "epoch": 1.9074872071101536,
      "grad_norm": 0.7040201375155568,
      "learning_rate": 5.277761501664535e-08,
      "loss": 0.2872446537017822,
      "num_input_tokens_seen": 1457856888,
      "step": 28330,
      "train_runtime": 174307.6913,
      "train_tokens_per_second": 8363.698
    },
    {
      "epoch": 1.9078238621061137,
      "grad_norm": 0.8013818407966221,
      "learning_rate": 5.2395146268804934e-08,
      "loss": 0.31221890449523926,
      "num_input_tokens_seen": 1458137800,
      "step": 28335,
      "train_runtime": 174335.5297,
      "train_tokens_per_second": 8363.974
    },
    {
      "epoch": 1.9081605171020737,
      "grad_norm": 0.5702481371070891,
      "learning_rate": 5.201406110288465e-08,
      "loss": 0.29315435886383057,
      "num_input_tokens_seen": 1458394816,
      "step": 28340,
      "train_runtime": 174369.3543,
      "train_tokens_per_second": 8363.825
    },
    {
      "epoch": 1.908497172098034,
      "grad_norm": 0.630143604554746,
      "learning_rate": 5.1634359625454266e-08,
      "loss": 0.26510460376739503,
      "num_input_tokens_seen": 1458642088,
      "step": 28345,
      "train_runtime": 174402.5299,
      "train_tokens_per_second": 8363.652
    },
    {
      "epoch": 1.9088338270939942,
      "grad_norm": 0.7557887116111003,
      "learning_rate": 5.125604194269606e-08,
      "loss": 0.31952414512634275,
      "num_input_tokens_seen": 1458893440,
      "step": 28350,
      "train_runtime": 174428.3668,
      "train_tokens_per_second": 8363.854
    },
    {
      "epoch": 1.9091704820899542,
      "grad_norm": 0.731121902531662,
      "learning_rate": 5.087910816040542e-08,
      "loss": 0.31671741008758547,
      "num_input_tokens_seen": 1459142080,
      "step": 28355,
      "train_runtime": 174454.6724,
      "train_tokens_per_second": 8364.018
    },
    {
      "epoch": 1.9095071370859142,
      "grad_norm": 0.5706520147065023,
      "learning_rate": 5.0503558383990235e-08,
      "loss": 0.28488037586212156,
      "num_input_tokens_seen": 1459421088,
      "step": 28360,
      "train_runtime": 174483.3831,
      "train_tokens_per_second": 8364.241
    },
    {
      "epoch": 1.9098437920818745,
      "grad_norm": 0.7767258337738469,
      "learning_rate": 5.0129392718472614e-08,
      "loss": 0.29333393573760985,
      "num_input_tokens_seen": 1459681944,
      "step": 28365,
      "train_runtime": 174512.8812,
      "train_tokens_per_second": 8364.322
    },
    {
      "epoch": 1.9101804470778347,
      "grad_norm": 0.7047570944130624,
      "learning_rate": 4.9756611268486077e-08,
      "loss": 0.3190404653549194,
      "num_input_tokens_seen": 1459956656,
      "step": 28370,
      "train_runtime": 174544.9874,
      "train_tokens_per_second": 8364.357
    },
    {
      "epoch": 1.9105171020737948,
      "grad_norm": 0.7647517629006011,
      "learning_rate": 4.9385214138277795e-08,
      "loss": 0.30591492652893065,
      "num_input_tokens_seen": 1460206040,
      "step": 28375,
      "train_runtime": 174571.8566,
      "train_tokens_per_second": 8364.499
    },
    {
      "epoch": 1.9108537570697548,
      "grad_norm": 0.6450248860947574,
      "learning_rate": 4.901520143170857e-08,
      "loss": 0.310720157623291,
      "num_input_tokens_seen": 1460476760,
      "step": 28380,
      "train_runtime": 174605.1737,
      "train_tokens_per_second": 8364.453
    },
    {
      "epoch": 1.911190412065715,
      "grad_norm": 0.7496129083421398,
      "learning_rate": 4.864657325225064e-08,
      "loss": 0.3003356456756592,
      "num_input_tokens_seen": 1460728680,
      "step": 28385,
      "train_runtime": 174636.4215,
      "train_tokens_per_second": 8364.399
    },
    {
      "epoch": 1.9115270670616753,
      "grad_norm": 0.6762826151639594,
      "learning_rate": 4.827932970298932e-08,
      "loss": 0.2791886806488037,
      "num_input_tokens_seen": 1460963432,
      "step": 28390,
      "train_runtime": 174668.746,
      "train_tokens_per_second": 8364.195
    },
    {
      "epoch": 1.9118637220576353,
      "grad_norm": 0.7405588914114161,
      "learning_rate": 4.7913470886624125e-08,
      "loss": 0.2993204355239868,
      "num_input_tokens_seen": 1461215192,
      "step": 28395,
      "train_runtime": 174701.4259,
      "train_tokens_per_second": 8364.071
    },
    {
      "epoch": 1.9122003770535954,
      "grad_norm": 0.6814770686108627,
      "learning_rate": 4.754899690546544e-08,
      "loss": 0.27663254737854004,
      "num_input_tokens_seen": 1461474472,
      "step": 28400,
      "train_runtime": 174734.9633,
      "train_tokens_per_second": 8363.95
    },
    {
      "epoch": 1.9125370320495556,
      "grad_norm": 0.6025396904369373,
      "learning_rate": 4.718590786143729e-08,
      "loss": 0.2974004507064819,
      "num_input_tokens_seen": 1461740520,
      "step": 28405,
      "train_runtime": 174766.7352,
      "train_tokens_per_second": 8363.952
    },
    {
      "epoch": 1.9128736870455159,
      "grad_norm": 0.6505520527247218,
      "learning_rate": 4.6824203856076776e-08,
      "loss": 0.31267476081848145,
      "num_input_tokens_seen": 1462025392,
      "step": 28410,
      "train_runtime": 174794.6184,
      "train_tokens_per_second": 8364.247
    },
    {
      "epoch": 1.9132103420414759,
      "grad_norm": 0.6226357723924448,
      "learning_rate": 4.646388499053356e-08,
      "loss": 0.2701855182647705,
      "num_input_tokens_seen": 1462276224,
      "step": 28415,
      "train_runtime": 174821.905,
      "train_tokens_per_second": 8364.376
    },
    {
      "epoch": 1.913546997037436,
      "grad_norm": 0.7062648122639825,
      "learning_rate": 4.6104951365568696e-08,
      "loss": 0.2738612174987793,
      "num_input_tokens_seen": 1462530560,
      "step": 28420,
      "train_runtime": 174853.3909,
      "train_tokens_per_second": 8364.325
    },
    {
      "epoch": 1.9138836520333962,
      "grad_norm": 0.7238301964873903,
      "learning_rate": 4.574740308155801e-08,
      "loss": 0.30540382862091064,
      "num_input_tokens_seen": 1462770888,
      "step": 28425,
      "train_runtime": 174883.0515,
      "train_tokens_per_second": 8364.28
    },
    {
      "epoch": 1.9142203070293564,
      "grad_norm": 0.6468459777347114,
      "learning_rate": 4.539124023848762e-08,
      "loss": 0.29903900623321533,
      "num_input_tokens_seen": 1463050760,
      "step": 28430,
      "train_runtime": 174914.4862,
      "train_tokens_per_second": 8364.377
    },
    {
      "epoch": 1.9145569620253164,
      "grad_norm": 0.6371649430293056,
      "learning_rate": 4.503646293595787e-08,
      "loss": 0.323260760307312,
      "num_input_tokens_seen": 1463310816,
      "step": 28435,
      "train_runtime": 174946.1029,
      "train_tokens_per_second": 8364.352
    },
    {
      "epoch": 1.9148936170212765,
      "grad_norm": 0.7015844556416977,
      "learning_rate": 4.468307127318105e-08,
      "loss": 0.2891608715057373,
      "num_input_tokens_seen": 1463580296,
      "step": 28440,
      "train_runtime": 174977.181,
      "train_tokens_per_second": 8364.407
    },
    {
      "epoch": 1.9152302720172367,
      "grad_norm": 0.7211529139079911,
      "learning_rate": 4.4331065348982014e-08,
      "loss": 0.28575563430786133,
      "num_input_tokens_seen": 1463855264,
      "step": 28445,
      "train_runtime": 175002.1257,
      "train_tokens_per_second": 8364.786
    },
    {
      "epoch": 1.915566927013197,
      "grad_norm": 0.6749547219934094,
      "learning_rate": 4.3980445261798125e-08,
      "loss": 0.3017213582992554,
      "num_input_tokens_seen": 1464108416,
      "step": 28450,
      "train_runtime": 175034.7959,
      "train_tokens_per_second": 8364.671
    },
    {
      "epoch": 1.915903582009157,
      "grad_norm": 0.7419150062854378,
      "learning_rate": 4.363121110967927e-08,
      "loss": 0.2856736660003662,
      "num_input_tokens_seen": 1464343320,
      "step": 28455,
      "train_runtime": 175063.8881,
      "train_tokens_per_second": 8364.622
    },
    {
      "epoch": 1.916240237005117,
      "grad_norm": 0.6729599530519617,
      "learning_rate": 4.328336299028735e-08,
      "loss": 0.276474928855896,
      "num_input_tokens_seen": 1464582320,
      "step": 28460,
      "train_runtime": 175096.8953,
      "train_tokens_per_second": 8364.411
    },
    {
      "epoch": 1.9165768920010773,
      "grad_norm": 0.796158855438293,
      "learning_rate": 4.293690100089731e-08,
      "loss": 0.29607014656066893,
      "num_input_tokens_seen": 1464827360,
      "step": 28465,
      "train_runtime": 175126.345,
      "train_tokens_per_second": 8364.403
    },
    {
      "epoch": 1.9169135469970375,
      "grad_norm": 0.6559950123704406,
      "learning_rate": 4.259182523839556e-08,
      "loss": 0.2929547786712646,
      "num_input_tokens_seen": 1465096056,
      "step": 28470,
      "train_runtime": 175156.2272,
      "train_tokens_per_second": 8364.51
    },
    {
      "epoch": 1.9172502019929976,
      "grad_norm": 0.7259846875277434,
      "learning_rate": 4.224813579928211e-08,
      "loss": 0.31351165771484374,
      "num_input_tokens_seen": 1465347448,
      "step": 28475,
      "train_runtime": 175188.6544,
      "train_tokens_per_second": 8364.397
    },
    {
      "epoch": 1.9175868569889576,
      "grad_norm": 0.6323427491033254,
      "learning_rate": 4.190583277966842e-08,
      "loss": 0.2921279430389404,
      "num_input_tokens_seen": 1465602424,
      "step": 28480,
      "train_runtime": 175221.6146,
      "train_tokens_per_second": 8364.279
    },
    {
      "epoch": 1.9179235119849178,
      "grad_norm": 0.6202881413746567,
      "learning_rate": 4.1564916275278477e-08,
      "loss": 0.27725677490234374,
      "num_input_tokens_seen": 1465845552,
      "step": 28485,
      "train_runtime": 175251.8941,
      "train_tokens_per_second": 8364.221
    },
    {
      "epoch": 1.918260166980878,
      "grad_norm": 0.6569337712568537,
      "learning_rate": 4.122538638144824e-08,
      "loss": 0.28941941261291504,
      "num_input_tokens_seen": 1466105584,
      "step": 28490,
      "train_runtime": 175282.9162,
      "train_tokens_per_second": 8364.224
    },
    {
      "epoch": 1.9185968219768381,
      "grad_norm": 0.6647791296995852,
      "learning_rate": 4.08872431931262e-08,
      "loss": 0.27710888385772703,
      "num_input_tokens_seen": 1466343904,
      "step": 28495,
      "train_runtime": 175309.4764,
      "train_tokens_per_second": 8364.316
    },
    {
      "epoch": 1.9189334769727981,
      "grad_norm": 0.6874243141044926,
      "learning_rate": 4.055048680487339e-08,
      "loss": 0.29281494617462156,
      "num_input_tokens_seen": 1466611472,
      "step": 28500,
      "train_runtime": 175340.0199,
      "train_tokens_per_second": 8364.385
    },
    {
      "epoch": 1.9192701319687584,
      "grad_norm": 0.661444803831243,
      "learning_rate": 4.0215117310862806e-08,
      "loss": 0.316835880279541,
      "num_input_tokens_seen": 1466832024,
      "step": 28505,
      "train_runtime": 175371.4456,
      "train_tokens_per_second": 8364.144
    },
    {
      "epoch": 1.9196067869647186,
      "grad_norm": 0.7752674807720837,
      "learning_rate": 3.9881134804878316e-08,
      "loss": 0.3136550188064575,
      "num_input_tokens_seen": 1467060776,
      "step": 28510,
      "train_runtime": 175400.3258,
      "train_tokens_per_second": 8364.071
    },
    {
      "epoch": 1.9199434419606787,
      "grad_norm": 0.6480524689804303,
      "learning_rate": 3.9548539380318e-08,
      "loss": 0.31084351539611815,
      "num_input_tokens_seen": 1467344024,
      "step": 28515,
      "train_runtime": 175430.4503,
      "train_tokens_per_second": 8364.249
    },
    {
      "epoch": 1.9202800969566387,
      "grad_norm": 0.7273406300813666,
      "learning_rate": 3.921733113019077e-08,
      "loss": 0.3014693737030029,
      "num_input_tokens_seen": 1467572672,
      "step": 28520,
      "train_runtime": 175458.5819,
      "train_tokens_per_second": 8364.211
    },
    {
      "epoch": 1.920616751952599,
      "grad_norm": 0.6349316435673338,
      "learning_rate": 3.888751014711867e-08,
      "loss": 0.3122251987457275,
      "num_input_tokens_seen": 1467810096,
      "step": 28525,
      "train_runtime": 175488.2407,
      "train_tokens_per_second": 8364.151
    },
    {
      "epoch": 1.9209534069485592,
      "grad_norm": 0.6279370540054833,
      "learning_rate": 3.855907652333402e-08,
      "loss": 0.28517923355102537,
      "num_input_tokens_seen": 1468062872,
      "step": 28530,
      "train_runtime": 175517.6031,
      "train_tokens_per_second": 8364.192
    },
    {
      "epoch": 1.9212900619445192,
      "grad_norm": 0.6238055618076058,
      "learning_rate": 3.8232030350683344e-08,
      "loss": 0.29549684524536135,
      "num_input_tokens_seen": 1468332504,
      "step": 28535,
      "train_runtime": 175549.2758,
      "train_tokens_per_second": 8364.219
    },
    {
      "epoch": 1.9216267169404793,
      "grad_norm": 0.5945717314853067,
      "learning_rate": 3.7906371720622946e-08,
      "loss": 0.3014315843582153,
      "num_input_tokens_seen": 1468598360,
      "step": 28540,
      "train_runtime": 175580.2351,
      "train_tokens_per_second": 8364.258
    },
    {
      "epoch": 1.9219633719364395,
      "grad_norm": 0.6405582833993597,
      "learning_rate": 3.758210072422275e-08,
      "loss": 0.2730912685394287,
      "num_input_tokens_seen": 1468849160,
      "step": 28545,
      "train_runtime": 175613.4681,
      "train_tokens_per_second": 8364.103
    },
    {
      "epoch": 1.9223000269323998,
      "grad_norm": 0.6713205865140677,
      "learning_rate": 3.72592174521641e-08,
      "loss": 0.3019833564758301,
      "num_input_tokens_seen": 1469108712,
      "step": 28550,
      "train_runtime": 175639.3916,
      "train_tokens_per_second": 8364.346
    },
    {
      "epoch": 1.9226366819283598,
      "grad_norm": 0.6482299918492885,
      "learning_rate": 3.693772199474033e-08,
      "loss": 0.3081575870513916,
      "num_input_tokens_seen": 1469372048,
      "step": 28555,
      "train_runtime": 175668.271,
      "train_tokens_per_second": 8364.47
    },
    {
      "epoch": 1.9229733369243198,
      "grad_norm": 0.7819079748862763,
      "learning_rate": 3.661761444185674e-08,
      "loss": 0.2989888906478882,
      "num_input_tokens_seen": 1469614656,
      "step": 28560,
      "train_runtime": 175701.0975,
      "train_tokens_per_second": 8364.288
    },
    {
      "epoch": 1.92330999192028,
      "grad_norm": 0.6260975987251955,
      "learning_rate": 3.6298894883030066e-08,
      "loss": 0.2980659484863281,
      "num_input_tokens_seen": 1469879784,
      "step": 28565,
      "train_runtime": 175734.0959,
      "train_tokens_per_second": 8364.227
    },
    {
      "epoch": 1.9236466469162403,
      "grad_norm": 0.7130489078926029,
      "learning_rate": 3.5981563407389007e-08,
      "loss": 0.3248549699783325,
      "num_input_tokens_seen": 1470136800,
      "step": 28570,
      "train_runtime": 175771.1722,
      "train_tokens_per_second": 8363.924
    },
    {
      "epoch": 1.9239833019122003,
      "grad_norm": 0.6812550881956241,
      "learning_rate": 3.566562010367536e-08,
      "loss": 0.27317888736724855,
      "num_input_tokens_seen": 1470380400,
      "step": 28575,
      "train_runtime": 175800.1359,
      "train_tokens_per_second": 8363.932
    },
    {
      "epoch": 1.9243199569081604,
      "grad_norm": 0.6017860829770323,
      "learning_rate": 3.535106506024011e-08,
      "loss": 0.2937533140182495,
      "num_input_tokens_seen": 1470640528,
      "step": 28580,
      "train_runtime": 175827.064,
      "train_tokens_per_second": 8364.131
    },
    {
      "epoch": 1.9246566119041206,
      "grad_norm": 0.6240215368231828,
      "learning_rate": 3.5037898365049005e-08,
      "loss": 0.26291799545288086,
      "num_input_tokens_seen": 1470897528,
      "step": 28585,
      "train_runtime": 175855.1201,
      "train_tokens_per_second": 8364.258
    },
    {
      "epoch": 1.9249932669000809,
      "grad_norm": 0.7491648035697425,
      "learning_rate": 3.472612010567755e-08,
      "loss": 0.2939670324325562,
      "num_input_tokens_seen": 1471142936,
      "step": 28590,
      "train_runtime": 175890.5141,
      "train_tokens_per_second": 8363.97
    },
    {
      "epoch": 1.925329921896041,
      "grad_norm": 0.5731153413298199,
      "learning_rate": 3.441573036931323e-08,
      "loss": 0.29923434257507325,
      "num_input_tokens_seen": 1471413320,
      "step": 28595,
      "train_runtime": 175923.563,
      "train_tokens_per_second": 8363.935
    },
    {
      "epoch": 1.925666576892001,
      "grad_norm": 0.6923071648631414,
      "learning_rate": 3.410672924275604e-08,
      "loss": 0.27001328468322755,
      "num_input_tokens_seen": 1471685976,
      "step": 28600,
      "train_runtime": 175954.4133,
      "train_tokens_per_second": 8364.019
    },
    {
      "epoch": 1.9260032318879612,
      "grad_norm": 0.7379190159835388,
      "learning_rate": 3.3799116812416876e-08,
      "loss": 0.3102369546890259,
      "num_input_tokens_seen": 1471937328,
      "step": 28605,
      "train_runtime": 175984.6182,
      "train_tokens_per_second": 8364.011
    },
    {
      "epoch": 1.9263398868839214,
      "grad_norm": 0.7562694868465994,
      "learning_rate": 3.349289316431803e-08,
      "loss": 0.31123788356781007,
      "num_input_tokens_seen": 1472206072,
      "step": 28610,
      "train_runtime": 176016.905,
      "train_tokens_per_second": 8364.004
    },
    {
      "epoch": 1.9266765418798815,
      "grad_norm": 0.7144686492841298,
      "learning_rate": 3.3188058384095446e-08,
      "loss": 0.2918006420135498,
      "num_input_tokens_seen": 1472465192,
      "step": 28615,
      "train_runtime": 176049.0682,
      "train_tokens_per_second": 8363.948
    },
    {
      "epoch": 1.9270131968758415,
      "grad_norm": 0.7141502949995416,
      "learning_rate": 3.2884612556993714e-08,
      "loss": 0.3175919532775879,
      "num_input_tokens_seen": 1472722112,
      "step": 28620,
      "train_runtime": 176075.9881,
      "train_tokens_per_second": 8364.128
    },
    {
      "epoch": 1.9273498518718017,
      "grad_norm": 0.7087495748108501,
      "learning_rate": 3.258255576787161e-08,
      "loss": 0.30709455013275144,
      "num_input_tokens_seen": 1472974712,
      "step": 28625,
      "train_runtime": 176106.8521,
      "train_tokens_per_second": 8364.097
    },
    {
      "epoch": 1.927686506867762,
      "grad_norm": 0.7940745409664285,
      "learning_rate": 3.22818881011977e-08,
      "loss": 0.3141803741455078,
      "num_input_tokens_seen": 1473262784,
      "step": 28630,
      "train_runtime": 176139.4887,
      "train_tokens_per_second": 8364.182
    },
    {
      "epoch": 1.928023161863722,
      "grad_norm": 1.2169072229649933,
      "learning_rate": 3.198260964105249e-08,
      "loss": 0.28010993003845214,
      "num_input_tokens_seen": 1473518312,
      "step": 28635,
      "train_runtime": 176170.7038,
      "train_tokens_per_second": 8364.151
    },
    {
      "epoch": 1.928359816859682,
      "grad_norm": 0.7342850031686914,
      "learning_rate": 3.168472047112903e-08,
      "loss": 0.3078049898147583,
      "num_input_tokens_seen": 1473758752,
      "step": 28640,
      "train_runtime": 176199.8269,
      "train_tokens_per_second": 8364.133
    },
    {
      "epoch": 1.9286964718556423,
      "grad_norm": 0.8693611587513813,
      "learning_rate": 3.138822067473068e-08,
      "loss": 0.2899463415145874,
      "num_input_tokens_seen": 1474007424,
      "step": 28645,
      "train_runtime": 176232.0601,
      "train_tokens_per_second": 8364.014
    },
    {
      "epoch": 1.9290331268516026,
      "grad_norm": 0.8212834898442424,
      "learning_rate": 3.10931103347728e-08,
      "loss": 0.2897956848144531,
      "num_input_tokens_seen": 1474261680,
      "step": 28650,
      "train_runtime": 176262.3098,
      "train_tokens_per_second": 8364.021
    },
    {
      "epoch": 1.9293697818475626,
      "grad_norm": 0.6423199372845801,
      "learning_rate": 3.0799389533781587e-08,
      "loss": 0.3189815044403076,
      "num_input_tokens_seen": 1474519136,
      "step": 28655,
      "train_runtime": 176291.6812,
      "train_tokens_per_second": 8364.088
    },
    {
      "epoch": 1.9297064368435226,
      "grad_norm": 0.5779845512396371,
      "learning_rate": 3.050705835389578e-08,
      "loss": 0.28115997314453123,
      "num_input_tokens_seen": 1474766560,
      "step": 28660,
      "train_runtime": 176321.6562,
      "train_tokens_per_second": 8364.069
    },
    {
      "epoch": 1.9300430918394829,
      "grad_norm": 0.7432928491536824,
      "learning_rate": 3.021611687686443e-08,
      "loss": 0.29572300910949706,
      "num_input_tokens_seen": 1475015480,
      "step": 28665,
      "train_runtime": 176353.5109,
      "train_tokens_per_second": 8363.97
    },
    {
      "epoch": 1.9303797468354431,
      "grad_norm": 0.7483505342482756,
      "learning_rate": 2.992656518404857e-08,
      "loss": 0.31209144592285154,
      "num_input_tokens_seen": 1475255080,
      "step": 28670,
      "train_runtime": 176381.992,
      "train_tokens_per_second": 8363.978
    },
    {
      "epoch": 1.9307164018314031,
      "grad_norm": 0.5865581337986642,
      "learning_rate": 2.96384033564201e-08,
      "loss": 0.28619658946990967,
      "num_input_tokens_seen": 1475500856,
      "step": 28675,
      "train_runtime": 176412.341,
      "train_tokens_per_second": 8363.932
    },
    {
      "epoch": 1.9310530568273632,
      "grad_norm": 0.595994916347019,
      "learning_rate": 2.935163147456288e-08,
      "loss": 0.2630106687545776,
      "num_input_tokens_seen": 1475767336,
      "step": 28680,
      "train_runtime": 176440.8912,
      "train_tokens_per_second": 8364.089
    },
    {
      "epoch": 1.9313897118233234,
      "grad_norm": 0.7926672683723244,
      "learning_rate": 2.906624961867166e-08,
      "loss": 0.30887184143066404,
      "num_input_tokens_seen": 1476023576,
      "step": 28685,
      "train_runtime": 176471.4633,
      "train_tokens_per_second": 8364.092
    },
    {
      "epoch": 1.9317263668192837,
      "grad_norm": 0.7398141507670686,
      "learning_rate": 2.8782257868553154e-08,
      "loss": 0.29244370460510255,
      "num_input_tokens_seen": 1476277848,
      "step": 28690,
      "train_runtime": 176503.6034,
      "train_tokens_per_second": 8364.01
    },
    {
      "epoch": 1.9320630218152437,
      "grad_norm": 0.705346121881099,
      "learning_rate": 2.849965630362328e-08,
      "loss": 0.2926915407180786,
      "num_input_tokens_seen": 1476522648,
      "step": 28695,
      "train_runtime": 176532.3509,
      "train_tokens_per_second": 8364.034
    },
    {
      "epoch": 1.9323996768112037,
      "grad_norm": 0.7302196855256424,
      "learning_rate": 2.8218445002912153e-08,
      "loss": 0.29307539463043214,
      "num_input_tokens_seen": 1476785936,
      "step": 28700,
      "train_runtime": 176561.4714,
      "train_tokens_per_second": 8364.146
    },
    {
      "epoch": 1.932736331807164,
      "grad_norm": 0.7317278879844824,
      "learning_rate": 2.7938624045059094e-08,
      "loss": 0.3010687351226807,
      "num_input_tokens_seen": 1477059640,
      "step": 28705,
      "train_runtime": 176597.7117,
      "train_tokens_per_second": 8363.979
    },
    {
      "epoch": 1.9330729868031242,
      "grad_norm": 0.7191750431710353,
      "learning_rate": 2.766019350831428e-08,
      "loss": 0.30344333648681643,
      "num_input_tokens_seen": 1477353992,
      "step": 28710,
      "train_runtime": 176628.6099,
      "train_tokens_per_second": 8364.183
    },
    {
      "epoch": 1.9334096417990843,
      "grad_norm": 0.6692609673232055,
      "learning_rate": 2.7383153470541546e-08,
      "loss": 0.315699028968811,
      "num_input_tokens_seen": 1477636000,
      "step": 28715,
      "train_runtime": 176663.3693,
      "train_tokens_per_second": 8364.133
    },
    {
      "epoch": 1.9337462967950443,
      "grad_norm": 0.7639793845204258,
      "learning_rate": 2.710750400921336e-08,
      "loss": 0.2854224681854248,
      "num_input_tokens_seen": 1477867192,
      "step": 28720,
      "train_runtime": 176693.9845,
      "train_tokens_per_second": 8363.993
    },
    {
      "epoch": 1.9340829517910045,
      "grad_norm": 0.6505924752163026,
      "learning_rate": 2.683324520141417e-08,
      "loss": 0.298730206489563,
      "num_input_tokens_seen": 1478131096,
      "step": 28725,
      "train_runtime": 176727.4131,
      "train_tokens_per_second": 8363.904
    },
    {
      "epoch": 1.9344196067869648,
      "grad_norm": 0.7019914275721029,
      "learning_rate": 2.65603771238393e-08,
      "loss": 0.30492196083068845,
      "num_input_tokens_seen": 1478363992,
      "step": 28730,
      "train_runtime": 176758.8706,
      "train_tokens_per_second": 8363.733
    },
    {
      "epoch": 1.9347562617829248,
      "grad_norm": 0.5926718192259186,
      "learning_rate": 2.6288899852796036e-08,
      "loss": 0.29820098876953127,
      "num_input_tokens_seen": 1478648336,
      "step": 28735,
      "train_runtime": 176789.813,
      "train_tokens_per_second": 8363.877
    },
    {
      "epoch": 1.9350929167788848,
      "grad_norm": 0.6585839731470671,
      "learning_rate": 2.6018813464202543e-08,
      "loss": 0.29645524024963377,
      "num_input_tokens_seen": 1478902392,
      "step": 28740,
      "train_runtime": 176816.4128,
      "train_tokens_per_second": 8364.056
    },
    {
      "epoch": 1.935429571774845,
      "grad_norm": 0.6352093755750375,
      "learning_rate": 2.575011803358618e-08,
      "loss": 0.2993162631988525,
      "num_input_tokens_seen": 1479176664,
      "step": 28745,
      "train_runtime": 176849.3486,
      "train_tokens_per_second": 8364.049
    },
    {
      "epoch": 1.9357662267708053,
      "grad_norm": 0.5302682365430811,
      "learning_rate": 2.5482813636087955e-08,
      "loss": 0.27205429077148435,
      "num_input_tokens_seen": 1479411112,
      "step": 28750,
      "train_runtime": 176877.7056,
      "train_tokens_per_second": 8364.034
    },
    {
      "epoch": 1.9361028817667654,
      "grad_norm": 0.7088347956938078,
      "learning_rate": 2.5216900346458073e-08,
      "loss": 0.29427251815795896,
      "num_input_tokens_seen": 1479667016,
      "step": 28755,
      "train_runtime": 176910.0539,
      "train_tokens_per_second": 8363.951
    },
    {
      "epoch": 1.9364395367627254,
      "grad_norm": 0.6559964059995803,
      "learning_rate": 2.4952378239058716e-08,
      "loss": 0.29413092136383057,
      "num_input_tokens_seen": 1479951208,
      "step": 28760,
      "train_runtime": 176945.5835,
      "train_tokens_per_second": 8363.878
    },
    {
      "epoch": 1.9367761917586857,
      "grad_norm": 0.7361793421030431,
      "learning_rate": 2.4689247387862934e-08,
      "loss": 0.2923002243041992,
      "num_input_tokens_seen": 1480212480,
      "step": 28765,
      "train_runtime": 176973.2495,
      "train_tokens_per_second": 8364.046
    },
    {
      "epoch": 1.937112846754646,
      "grad_norm": 0.6645253573654699,
      "learning_rate": 2.4427507866453537e-08,
      "loss": 0.2789115905761719,
      "num_input_tokens_seen": 1480417280,
      "step": 28770,
      "train_runtime": 177002.5393,
      "train_tokens_per_second": 8363.819
    },
    {
      "epoch": 1.937449501750606,
      "grad_norm": 0.7652371924275769,
      "learning_rate": 2.416715974802586e-08,
      "loss": 0.2915948867797852,
      "num_input_tokens_seen": 1480667320,
      "step": 28775,
      "train_runtime": 177034.1316,
      "train_tokens_per_second": 8363.739
    },
    {
      "epoch": 1.937786156746566,
      "grad_norm": 0.56988049145146,
      "learning_rate": 2.3908203105384997e-08,
      "loss": 0.30509414672851565,
      "num_input_tokens_seen": 1480938272,
      "step": 28780,
      "train_runtime": 177064.663,
      "train_tokens_per_second": 8363.827
    },
    {
      "epoch": 1.9381228117425262,
      "grad_norm": 0.7194144636803746,
      "learning_rate": 2.365063801094747e-08,
      "loss": 0.2874755859375,
      "num_input_tokens_seen": 1481190224,
      "step": 28785,
      "train_runtime": 177093.844,
      "train_tokens_per_second": 8363.872
    },
    {
      "epoch": 1.9384594667384865,
      "grad_norm": 0.6039107947561102,
      "learning_rate": 2.3394464536740658e-08,
      "loss": 0.3072074890136719,
      "num_input_tokens_seen": 1481464448,
      "step": 28790,
      "train_runtime": 177127.6755,
      "train_tokens_per_second": 8363.823
    },
    {
      "epoch": 1.9387961217344465,
      "grad_norm": 0.7364705292494335,
      "learning_rate": 2.3139682754402816e-08,
      "loss": 0.32218711376190184,
      "num_input_tokens_seen": 1481743856,
      "step": 28795,
      "train_runtime": 177160.4428,
      "train_tokens_per_second": 8363.853
    },
    {
      "epoch": 1.9391327767304065,
      "grad_norm": 0.7100424303411628,
      "learning_rate": 2.2886292735182502e-08,
      "loss": 0.29333910942077634,
      "num_input_tokens_seen": 1482015808,
      "step": 28800,
      "train_runtime": 177191.1055,
      "train_tokens_per_second": 8363.94
    },
    {
      "epoch": 1.9394694317263668,
      "grad_norm": 0.6880414817462062,
      "learning_rate": 2.2634294549939705e-08,
      "loss": 0.28635692596435547,
      "num_input_tokens_seen": 1482291024,
      "step": 28805,
      "train_runtime": 177221.8055,
      "train_tokens_per_second": 8364.044
    },
    {
      "epoch": 1.939806086722327,
      "grad_norm": 0.7092942350850139,
      "learning_rate": 2.2383688269144723e-08,
      "loss": 0.35138978958129885,
      "num_input_tokens_seen": 1482554648,
      "step": 28810,
      "train_runtime": 177252.1344,
      "train_tokens_per_second": 8364.1
    },
    {
      "epoch": 1.940142741718287,
      "grad_norm": 0.6900041347541296,
      "learning_rate": 2.2134473962878712e-08,
      "loss": 0.28131470680236814,
      "num_input_tokens_seen": 1482801440,
      "step": 28815,
      "train_runtime": 177286.546,
      "train_tokens_per_second": 8363.869
    },
    {
      "epoch": 1.940479396714247,
      "grad_norm": 0.8222980810497796,
      "learning_rate": 2.1886651700833705e-08,
      "loss": 0.2803335189819336,
      "num_input_tokens_seen": 1483015120,
      "step": 28820,
      "train_runtime": 177313.0594,
      "train_tokens_per_second": 8363.823
    },
    {
      "epoch": 1.9408160517102075,
      "grad_norm": 0.6331058093093895,
      "learning_rate": 2.1640221552312603e-08,
      "loss": 0.298292350769043,
      "num_input_tokens_seen": 1483263960,
      "step": 28825,
      "train_runtime": 177340.1125,
      "train_tokens_per_second": 8363.951
    },
    {
      "epoch": 1.9411527067061676,
      "grad_norm": 0.6835693554209017,
      "learning_rate": 2.1395183586229163e-08,
      "loss": 0.3051080465316772,
      "num_input_tokens_seen": 1483541208,
      "step": 28830,
      "train_runtime": 177376.5468,
      "train_tokens_per_second": 8363.796
    },
    {
      "epoch": 1.9414893617021276,
      "grad_norm": 0.6328869194846295,
      "learning_rate": 2.1151537871106353e-08,
      "loss": 0.2829404354095459,
      "num_input_tokens_seen": 1483791128,
      "step": 28835,
      "train_runtime": 177408.6086,
      "train_tokens_per_second": 8363.693
    },
    {
      "epoch": 1.9418260166980879,
      "grad_norm": 0.6710599673014853,
      "learning_rate": 2.0909284475080226e-08,
      "loss": 0.28106050491333007,
      "num_input_tokens_seen": 1484043696,
      "step": 28840,
      "train_runtime": 177439.2656,
      "train_tokens_per_second": 8363.671
    },
    {
      "epoch": 1.942162671694048,
      "grad_norm": 0.6973936845164039,
      "learning_rate": 2.0668423465894928e-08,
      "loss": 0.283825945854187,
      "num_input_tokens_seen": 1484294416,
      "step": 28845,
      "train_runtime": 177468.912,
      "train_tokens_per_second": 8363.687
    },
    {
      "epoch": 1.9424993266900081,
      "grad_norm": 0.666945603427092,
      "learning_rate": 2.0428954910907684e-08,
      "loss": 0.27664687633514407,
      "num_input_tokens_seen": 1484522304,
      "step": 28850,
      "train_runtime": 177498.9907,
      "train_tokens_per_second": 8363.553
    },
    {
      "epoch": 1.9428359816859682,
      "grad_norm": 0.738055769155565,
      "learning_rate": 2.019087887708382e-08,
      "loss": 0.30380749702453613,
      "num_input_tokens_seen": 1484777584,
      "step": 28855,
      "train_runtime": 177530.5645,
      "train_tokens_per_second": 8363.504
    },
    {
      "epoch": 1.9431726366819284,
      "grad_norm": 0.6361020528370082,
      "learning_rate": 1.9954195431001744e-08,
      "loss": 0.29081239700317385,
      "num_input_tokens_seen": 1485050992,
      "step": 28860,
      "train_runtime": 177563.9868,
      "train_tokens_per_second": 8363.47
    },
    {
      "epoch": 1.9435092916778887,
      "grad_norm": 0.6738247619894314,
      "learning_rate": 1.9718904638848513e-08,
      "loss": 0.2886855125427246,
      "num_input_tokens_seen": 1485287488,
      "step": 28865,
      "train_runtime": 177592.9626,
      "train_tokens_per_second": 8363.437
    },
    {
      "epoch": 1.9438459466738487,
      "grad_norm": 0.5885807554962978,
      "learning_rate": 1.9485006566422602e-08,
      "loss": 0.29042792320251465,
      "num_input_tokens_seen": 1485571632,
      "step": 28870,
      "train_runtime": 177626.7562,
      "train_tokens_per_second": 8363.445
    },
    {
      "epoch": 1.9441826016698087,
      "grad_norm": 0.6553128295855747,
      "learning_rate": 1.9252501279132806e-08,
      "loss": 0.30260353088378905,
      "num_input_tokens_seen": 1485816552,
      "step": 28875,
      "train_runtime": 177659.2719,
      "train_tokens_per_second": 8363.293
    },
    {
      "epoch": 1.944519256665769,
      "grad_norm": 0.6126180830424264,
      "learning_rate": 1.902138884199878e-08,
      "loss": 0.30224220752716063,
      "num_input_tokens_seen": 1486088232,
      "step": 28880,
      "train_runtime": 177690.682,
      "train_tokens_per_second": 8363.344
    },
    {
      "epoch": 1.9448559116617292,
      "grad_norm": 0.6650825142536669,
      "learning_rate": 1.879166931964993e-08,
      "loss": 0.30625553131103517,
      "num_input_tokens_seen": 1486342128,
      "step": 28885,
      "train_runtime": 177725.7188,
      "train_tokens_per_second": 8363.123
    },
    {
      "epoch": 1.9451925666576892,
      "grad_norm": 0.6917697309698548,
      "learning_rate": 1.856334277632765e-08,
      "loss": 0.29618086814880373,
      "num_input_tokens_seen": 1486617744,
      "step": 28890,
      "train_runtime": 177759.2416,
      "train_tokens_per_second": 8363.097
    },
    {
      "epoch": 1.9455292216536493,
      "grad_norm": 0.6177096100535745,
      "learning_rate": 1.8336409275880872e-08,
      "loss": 0.3007068634033203,
      "num_input_tokens_seen": 1486890760,
      "step": 28895,
      "train_runtime": 177790.1311,
      "train_tokens_per_second": 8363.179
    },
    {
      "epoch": 1.9458658766496095,
      "grad_norm": 0.6651778323009966,
      "learning_rate": 1.8110868881772713e-08,
      "loss": 0.2878331184387207,
      "num_input_tokens_seen": 1487125880,
      "step": 28900,
      "train_runtime": 177820.1725,
      "train_tokens_per_second": 8363.089
    },
    {
      "epoch": 1.9462025316455698,
      "grad_norm": 0.6187672503067327,
      "learning_rate": 1.788672165707328e-08,
      "loss": 0.2772597551345825,
      "num_input_tokens_seen": 1487385896,
      "step": 28905,
      "train_runtime": 177853.782,
      "train_tokens_per_second": 8362.97
    },
    {
      "epoch": 1.9465391866415298,
      "grad_norm": 0.6445517794258447,
      "learning_rate": 1.7663967664465763e-08,
      "loss": 0.271146559715271,
      "num_input_tokens_seen": 1487632792,
      "step": 28910,
      "train_runtime": 177882.3527,
      "train_tokens_per_second": 8363.015
    },
    {
      "epoch": 1.9468758416374898,
      "grad_norm": 0.6774015492895618,
      "learning_rate": 1.7442606966242005e-08,
      "loss": 0.3164908170700073,
      "num_input_tokens_seen": 1487915080,
      "step": 28915,
      "train_runtime": 177913.1738,
      "train_tokens_per_second": 8363.153
    },
    {
      "epoch": 1.94721249663345,
      "grad_norm": 0.6235589975928585,
      "learning_rate": 1.722263962430526e-08,
      "loss": 0.30228657722473146,
      "num_input_tokens_seen": 1488184632,
      "step": 28920,
      "train_runtime": 177943.7302,
      "train_tokens_per_second": 8363.232
    },
    {
      "epoch": 1.9475491516294103,
      "grad_norm": 0.715216782098826,
      "learning_rate": 1.7004065700168547e-08,
      "loss": 0.28380913734436036,
      "num_input_tokens_seen": 1488452856,
      "step": 28925,
      "train_runtime": 177972.7242,
      "train_tokens_per_second": 8363.376
    },
    {
      "epoch": 1.9478858066253704,
      "grad_norm": 0.6400428230808546,
      "learning_rate": 1.6786885254954644e-08,
      "loss": 0.30549752712249756,
      "num_input_tokens_seen": 1488738336,
      "step": 28930,
      "train_runtime": 178003.3244,
      "train_tokens_per_second": 8363.542
    },
    {
      "epoch": 1.9482224616213304,
      "grad_norm": 0.7530329271746429,
      "learning_rate": 1.6571098349398296e-08,
      "loss": 0.29172630310058595,
      "num_input_tokens_seen": 1488974024,
      "step": 28935,
      "train_runtime": 178036.465,
      "train_tokens_per_second": 8363.309
    },
    {
      "epoch": 1.9485591166172906,
      "grad_norm": 0.8077143971263997,
      "learning_rate": 1.635670504384346e-08,
      "loss": 0.314191198348999,
      "num_input_tokens_seen": 1489184288,
      "step": 28940,
      "train_runtime": 178066.5525,
      "train_tokens_per_second": 8363.077
    },
    {
      "epoch": 1.948895771613251,
      "grad_norm": 0.8612993181094634,
      "learning_rate": 1.6143705398243837e-08,
      "loss": 0.286503267288208,
      "num_input_tokens_seen": 1489397184,
      "step": 28945,
      "train_runtime": 178094.7187,
      "train_tokens_per_second": 8362.95
    },
    {
      "epoch": 1.949232426609211,
      "grad_norm": 0.7542978573022324,
      "learning_rate": 1.5932099472165674e-08,
      "loss": 0.2979788303375244,
      "num_input_tokens_seen": 1489644776,
      "step": 28950,
      "train_runtime": 178124.9964,
      "train_tokens_per_second": 8362.918
    },
    {
      "epoch": 1.949569081605171,
      "grad_norm": 0.6349699530278812,
      "learning_rate": 1.572188732478164e-08,
      "loss": 0.2871803045272827,
      "num_input_tokens_seen": 1489918168,
      "step": 28955,
      "train_runtime": 178156.4345,
      "train_tokens_per_second": 8362.977
    },
    {
      "epoch": 1.9499057366011312,
      "grad_norm": 0.6671664663405085,
      "learning_rate": 1.551306901487859e-08,
      "loss": 0.2986912250518799,
      "num_input_tokens_seen": 1490170056,
      "step": 28960,
      "train_runtime": 178189.1612,
      "train_tokens_per_second": 8362.855
    },
    {
      "epoch": 1.9502423915970915,
      "grad_norm": 0.630981550004515,
      "learning_rate": 1.5305644600852043e-08,
      "loss": 0.2847066640853882,
      "num_input_tokens_seen": 1490455360,
      "step": 28965,
      "train_runtime": 178219.6756,
      "train_tokens_per_second": 8363.024
    },
    {
      "epoch": 1.9505790465930515,
      "grad_norm": 0.6527474889922451,
      "learning_rate": 1.5099614140706154e-08,
      "loss": 0.2909574508666992,
      "num_input_tokens_seen": 1490702696,
      "step": 28970,
      "train_runtime": 178248.8994,
      "train_tokens_per_second": 8363.04
    },
    {
      "epoch": 1.9509157015890115,
      "grad_norm": 0.7401277467041066,
      "learning_rate": 1.489497769205761e-08,
      "loss": 0.31166274547576905,
      "num_input_tokens_seen": 1490933800,
      "step": 28975,
      "train_runtime": 178277.5823,
      "train_tokens_per_second": 8362.991
    },
    {
      "epoch": 1.9512523565849718,
      "grad_norm": 0.5810144969500624,
      "learning_rate": 1.4691735312132304e-08,
      "loss": 0.289442253112793,
      "num_input_tokens_seen": 1491204160,
      "step": 28980,
      "train_runtime": 178309.8384,
      "train_tokens_per_second": 8362.994
    },
    {
      "epoch": 1.951589011580932,
      "grad_norm": 0.7374730925051637,
      "learning_rate": 1.4489887057766439e-08,
      "loss": 0.300986123085022,
      "num_input_tokens_seen": 1491455432,
      "step": 28985,
      "train_runtime": 178340.9188,
      "train_tokens_per_second": 8362.946
    },
    {
      "epoch": 1.951925666576892,
      "grad_norm": 0.6237332168804023,
      "learning_rate": 1.428943298540597e-08,
      "loss": 0.3017923831939697,
      "num_input_tokens_seen": 1491723872,
      "step": 28990,
      "train_runtime": 178372.4383,
      "train_tokens_per_second": 8362.973
    },
    {
      "epoch": 1.952262321572852,
      "grad_norm": 0.5608689213015369,
      "learning_rate": 1.4090373151107173e-08,
      "loss": 0.30710830688476565,
      "num_input_tokens_seen": 1491999128,
      "step": 28995,
      "train_runtime": 178400.2731,
      "train_tokens_per_second": 8363.211
    },
    {
      "epoch": 1.9525989765688123,
      "grad_norm": 0.6530083114831734,
      "learning_rate": 1.3892707610536627e-08,
      "loss": 0.24876198768615723,
      "num_input_tokens_seen": 1492235960,
      "step": 29000,
      "train_runtime": 178428.7979,
      "train_tokens_per_second": 8363.201
    },
    {
      "epoch": 1.9529356315647726,
      "grad_norm": 0.7572943612636176,
      "learning_rate": 1.3696436418970672e-08,
      "loss": 0.29105844497680666,
      "num_input_tokens_seen": 1492493240,
      "step": 29005,
      "train_runtime": 178459.6522,
      "train_tokens_per_second": 8363.197
    },
    {
      "epoch": 1.9532722865607326,
      "grad_norm": 0.7869166125382482,
      "learning_rate": 1.3501559631296512e-08,
      "loss": 0.3104537487030029,
      "num_input_tokens_seen": 1492721968,
      "step": 29010,
      "train_runtime": 178486.3811,
      "train_tokens_per_second": 8363.226
    },
    {
      "epoch": 1.9536089415566926,
      "grad_norm": 0.6372715512055034,
      "learning_rate": 1.3308077302010557e-08,
      "loss": 0.31415011882781985,
      "num_input_tokens_seen": 1493005440,
      "step": 29015,
      "train_runtime": 178516.9735,
      "train_tokens_per_second": 8363.381
    },
    {
      "epoch": 1.9539455965526529,
      "grad_norm": 0.7250990895544565,
      "learning_rate": 1.3115989485218972e-08,
      "loss": 0.29164671897888184,
      "num_input_tokens_seen": 1493259512,
      "step": 29020,
      "train_runtime": 178543.8434,
      "train_tokens_per_second": 8363.545
    },
    {
      "epoch": 1.9542822515486131,
      "grad_norm": 0.6999705254417848,
      "learning_rate": 1.2925296234638784e-08,
      "loss": 0.28792333602905273,
      "num_input_tokens_seen": 1493519728,
      "step": 29025,
      "train_runtime": 178571.5135,
      "train_tokens_per_second": 8363.707
    },
    {
      "epoch": 1.9546189065445732,
      "grad_norm": 0.6086394052818093,
      "learning_rate": 1.2735997603597339e-08,
      "loss": 0.28763833045959475,
      "num_input_tokens_seen": 1493807856,
      "step": 29030,
      "train_runtime": 178605.349,
      "train_tokens_per_second": 8363.735
    },
    {
      "epoch": 1.9549555615405332,
      "grad_norm": 0.6729546121827489,
      "learning_rate": 1.2548093645030623e-08,
      "loss": 0.291399621963501,
      "num_input_tokens_seen": 1494076216,
      "step": 29035,
      "train_runtime": 178635.1614,
      "train_tokens_per_second": 8363.842
    },
    {
      "epoch": 1.9552922165364934,
      "grad_norm": 0.6236600595542107,
      "learning_rate": 1.236158441148605e-08,
      "loss": 0.2833369255065918,
      "num_input_tokens_seen": 1494360896,
      "step": 29040,
      "train_runtime": 178660.8247,
      "train_tokens_per_second": 8364.234
    },
    {
      "epoch": 1.9556288715324537,
      "grad_norm": 0.6901504378135351,
      "learning_rate": 1.2176469955119119e-08,
      "loss": 0.32641329765319826,
      "num_input_tokens_seen": 1494632568,
      "step": 29045,
      "train_runtime": 178692.3145,
      "train_tokens_per_second": 8364.28
    },
    {
      "epoch": 1.9559655265284137,
      "grad_norm": 0.6755471006546183,
      "learning_rate": 1.199275032769842e-08,
      "loss": 0.31005043983459474,
      "num_input_tokens_seen": 1494907840,
      "step": 29050,
      "train_runtime": 178722.8584,
      "train_tokens_per_second": 8364.391
    },
    {
      "epoch": 1.9563021815243737,
      "grad_norm": 0.6554362547355489,
      "learning_rate": 1.1810425580598971e-08,
      "loss": 0.304477334022522,
      "num_input_tokens_seen": 1495190096,
      "step": 29055,
      "train_runtime": 178755.1996,
      "train_tokens_per_second": 8364.457
    },
    {
      "epoch": 1.956638836520334,
      "grad_norm": 0.7156923718500092,
      "learning_rate": 1.1629495764807763e-08,
      "loss": 0.2974085330963135,
      "num_input_tokens_seen": 1495476304,
      "step": 29060,
      "train_runtime": 178783.2803,
      "train_tokens_per_second": 8364.744
    },
    {
      "epoch": 1.9569754915162942,
      "grad_norm": 0.632330982233734,
      "learning_rate": 1.1449960930921544e-08,
      "loss": 0.2895530700683594,
      "num_input_tokens_seen": 1495726272,
      "step": 29065,
      "train_runtime": 178816.0586,
      "train_tokens_per_second": 8364.608
    },
    {
      "epoch": 1.9573121465122543,
      "grad_norm": 0.6581024270428759,
      "learning_rate": 1.1271821129146266e-08,
      "loss": 0.28275628089904786,
      "num_input_tokens_seen": 1495989968,
      "step": 29070,
      "train_runtime": 178850.9356,
      "train_tokens_per_second": 8364.451
    },
    {
      "epoch": 1.9576488015082143,
      "grad_norm": 0.674533308041232,
      "learning_rate": 1.1095076409298189e-08,
      "loss": 0.28294646739959717,
      "num_input_tokens_seen": 1496243024,
      "step": 29075,
      "train_runtime": 178886.4563,
      "train_tokens_per_second": 8364.205
    },
    {
      "epoch": 1.9579854565041745,
      "grad_norm": 0.7364653914898187,
      "learning_rate": 1.0919726820803889e-08,
      "loss": 0.28847699165344237,
      "num_input_tokens_seen": 1496487320,
      "step": 29080,
      "train_runtime": 178917.1569,
      "train_tokens_per_second": 8364.135
    },
    {
      "epoch": 1.9583221115001348,
      "grad_norm": 0.6505416554572111,
      "learning_rate": 1.0745772412698585e-08,
      "loss": 0.29737024307250975,
      "num_input_tokens_seen": 1496734944,
      "step": 29085,
      "train_runtime": 178950.8946,
      "train_tokens_per_second": 8363.942
    },
    {
      "epoch": 1.9586587664960948,
      "grad_norm": 0.6452942871741933,
      "learning_rate": 1.057321323362892e-08,
      "loss": 0.28976590633392335,
      "num_input_tokens_seen": 1497019064,
      "step": 29090,
      "train_runtime": 178982.3482,
      "train_tokens_per_second": 8364.06
    },
    {
      "epoch": 1.9589954214920549,
      "grad_norm": 0.6032729174043552,
      "learning_rate": 1.0402049331849629e-08,
      "loss": 0.2917973279953003,
      "num_input_tokens_seen": 1497294912,
      "step": 29095,
      "train_runtime": 179012.4686,
      "train_tokens_per_second": 8364.193
    },
    {
      "epoch": 1.959332076488015,
      "grad_norm": 0.68156765450217,
      "learning_rate": 1.0232280755226865e-08,
      "loss": 0.3063060998916626,
      "num_input_tokens_seen": 1497543416,
      "step": 29100,
      "train_runtime": 179045.0852,
      "train_tokens_per_second": 8364.058
    },
    {
      "epoch": 1.9596687314839754,
      "grad_norm": 0.5799748023958279,
      "learning_rate": 1.006390755123543e-08,
      "loss": 0.2897289752960205,
      "num_input_tokens_seen": 1497785920,
      "step": 29105,
      "train_runtime": 179075.8399,
      "train_tokens_per_second": 8363.975
    },
    {
      "epoch": 1.9600053864799354,
      "grad_norm": 0.8199008766367752,
      "learning_rate": 9.89692976696044e-09,
      "loss": 0.31242401599884034,
      "num_input_tokens_seen": 1498035128,
      "step": 29110,
      "train_runtime": 179106.6683,
      "train_tokens_per_second": 8363.927
    },
    {
      "epoch": 1.9603420414758954,
      "grad_norm": 0.5992485666906531,
      "learning_rate": 9.731347449097316e-09,
      "loss": 0.3134958744049072,
      "num_input_tokens_seen": 1498287616,
      "step": 29115,
      "train_runtime": 179138.6687,
      "train_tokens_per_second": 8363.843
    },
    {
      "epoch": 1.9606786964718557,
      "grad_norm": 0.6673546418349985,
      "learning_rate": 9.56716064394958e-09,
      "loss": 0.30624971389770506,
      "num_input_tokens_seen": 1498569120,
      "step": 29120,
      "train_runtime": 179168.5334,
      "train_tokens_per_second": 8364.02
    },
    {
      "epoch": 1.961015351467816,
      "grad_norm": 0.7557045621262098,
      "learning_rate": 9.404369397432166e-09,
      "loss": 0.27965855598449707,
      "num_input_tokens_seen": 1498825464,
      "step": 29125,
      "train_runtime": 179201.7562,
      "train_tokens_per_second": 8363.899
    },
    {
      "epoch": 1.961352006463776,
      "grad_norm": 0.6572126546465014,
      "learning_rate": 9.242973755068663e-09,
      "loss": 0.29594902992248534,
      "num_input_tokens_seen": 1499095464,
      "step": 29130,
      "train_runtime": 179234.2321,
      "train_tokens_per_second": 8363.89
    },
    {
      "epoch": 1.961688661459736,
      "grad_norm": 0.7849441216872376,
      "learning_rate": 9.082973761993518e-09,
      "loss": 0.29764485359191895,
      "num_input_tokens_seen": 1499345152,
      "step": 29135,
      "train_runtime": 179263.5608,
      "train_tokens_per_second": 8363.915
    },
    {
      "epoch": 1.9620253164556962,
      "grad_norm": 0.7792089819231262,
      "learning_rate": 8.924369462949834e-09,
      "loss": 0.32599496841430664,
      "num_input_tokens_seen": 1499574080,
      "step": 29140,
      "train_runtime": 179291.0874,
      "train_tokens_per_second": 8363.908
    },
    {
      "epoch": 1.9623619714516565,
      "grad_norm": 0.6585619411045989,
      "learning_rate": 8.767160902291016e-09,
      "loss": 0.2845829963684082,
      "num_input_tokens_seen": 1499830432,
      "step": 29145,
      "train_runtime": 179323.9807,
      "train_tokens_per_second": 8363.803
    },
    {
      "epoch": 1.9626986264476165,
      "grad_norm": 0.5913634968736563,
      "learning_rate": 8.61134812397968e-09,
      "loss": 0.29850144386291505,
      "num_input_tokens_seen": 1500093688,
      "step": 29150,
      "train_runtime": 179355.8483,
      "train_tokens_per_second": 8363.785
    },
    {
      "epoch": 1.9630352814435765,
      "grad_norm": 0.6808016573758802,
      "learning_rate": 8.45693117158819e-09,
      "loss": 0.2752715587615967,
      "num_input_tokens_seen": 1500340056,
      "step": 29155,
      "train_runtime": 179385.4069,
      "train_tokens_per_second": 8363.78
    },
    {
      "epoch": 1.9633719364395368,
      "grad_norm": 0.6821953002896898,
      "learning_rate": 8.303910088299228e-09,
      "loss": 0.2894306182861328,
      "num_input_tokens_seen": 1500589848,
      "step": 29160,
      "train_runtime": 179417.045,
      "train_tokens_per_second": 8363.697
    },
    {
      "epoch": 1.963708591435497,
      "grad_norm": 0.6736327069623957,
      "learning_rate": 8.152284916904674e-09,
      "loss": 0.29739718437194823,
      "num_input_tokens_seen": 1500859216,
      "step": 29165,
      "train_runtime": 179446.5067,
      "train_tokens_per_second": 8363.825
    },
    {
      "epoch": 1.964045246431457,
      "grad_norm": 0.6333812883762733,
      "learning_rate": 8.002055699805612e-09,
      "loss": 0.2785855770111084,
      "num_input_tokens_seen": 1501095592,
      "step": 29170,
      "train_runtime": 179472.6061,
      "train_tokens_per_second": 8363.926
    },
    {
      "epoch": 1.964381901427417,
      "grad_norm": 0.6746285691671621,
      "learning_rate": 7.853222479013434e-09,
      "loss": 0.2850207328796387,
      "num_input_tokens_seen": 1501383672,
      "step": 29175,
      "train_runtime": 179503.871,
      "train_tokens_per_second": 8364.074
    },
    {
      "epoch": 1.9647185564233773,
      "grad_norm": 0.6351087161482677,
      "learning_rate": 7.705785296148737e-09,
      "loss": 0.29580020904541016,
      "num_input_tokens_seen": 1501641608,
      "step": 29180,
      "train_runtime": 179533.3731,
      "train_tokens_per_second": 8364.136
    },
    {
      "epoch": 1.9650552114193376,
      "grad_norm": 0.6769307324988518,
      "learning_rate": 7.55974419244132e-09,
      "loss": 0.3071397066116333,
      "num_input_tokens_seen": 1501914392,
      "step": 29185,
      "train_runtime": 179560.9241,
      "train_tokens_per_second": 8364.372
    },
    {
      "epoch": 1.9653918664152976,
      "grad_norm": 0.6591447220285305,
      "learning_rate": 7.415099208732402e-09,
      "loss": 0.27669696807861327,
      "num_input_tokens_seen": 1502179080,
      "step": 29190,
      "train_runtime": 179592.5072,
      "train_tokens_per_second": 8364.375
    },
    {
      "epoch": 1.9657285214112576,
      "grad_norm": 0.5815216097046829,
      "learning_rate": 7.2718503854707355e-09,
      "loss": 0.3035935401916504,
      "num_input_tokens_seen": 1502451160,
      "step": 29195,
      "train_runtime": 179620.6376,
      "train_tokens_per_second": 8364.58
    },
    {
      "epoch": 1.966065176407218,
      "grad_norm": 0.6025751930703044,
      "learning_rate": 7.129997762715391e-09,
      "loss": 0.263226842880249,
      "num_input_tokens_seen": 1502721400,
      "step": 29200,
      "train_runtime": 179654.6477,
      "train_tokens_per_second": 8364.501
    },
    {
      "epoch": 1.9664018314031781,
      "grad_norm": 0.7015180011541339,
      "learning_rate": 6.98954138013519e-09,
      "loss": 0.2902121067047119,
      "num_input_tokens_seen": 1502959568,
      "step": 29205,
      "train_runtime": 179683.4672,
      "train_tokens_per_second": 8364.484
    },
    {
      "epoch": 1.9667384863991382,
      "grad_norm": 0.6432801976148382,
      "learning_rate": 6.850481277008159e-09,
      "loss": 0.30002832412719727,
      "num_input_tokens_seen": 1503222208,
      "step": 29210,
      "train_runtime": 179718.4268,
      "train_tokens_per_second": 8364.319
    },
    {
      "epoch": 1.9670751413950982,
      "grad_norm": 0.6760490112077181,
      "learning_rate": 6.712817492222079e-09,
      "loss": 0.31739010810852053,
      "num_input_tokens_seen": 1503484272,
      "step": 29215,
      "train_runtime": 179750.6897,
      "train_tokens_per_second": 8364.275
    },
    {
      "epoch": 1.9674117963910585,
      "grad_norm": 0.5977054435026813,
      "learning_rate": 6.576550064273934e-09,
      "loss": 0.30227208137512207,
      "num_input_tokens_seen": 1503746624,
      "step": 29220,
      "train_runtime": 179781.2299,
      "train_tokens_per_second": 8364.314
    },
    {
      "epoch": 1.9677484513870187,
      "grad_norm": 0.5740523337910926,
      "learning_rate": 6.44167903127102e-09,
      "loss": 0.30749878883361814,
      "num_input_tokens_seen": 1504014408,
      "step": 29225,
      "train_runtime": 179811.5012,
      "train_tokens_per_second": 8364.395
    },
    {
      "epoch": 1.9680851063829787,
      "grad_norm": 0.6583571277393592,
      "learning_rate": 6.3082044309287215e-09,
      "loss": 0.280363655090332,
      "num_input_tokens_seen": 1504281944,
      "step": 29230,
      "train_runtime": 179840.2206,
      "train_tokens_per_second": 8364.547
    },
    {
      "epoch": 1.9684217613789388,
      "grad_norm": 0.6889982532341788,
      "learning_rate": 6.176126300573848e-09,
      "loss": 0.2949554920196533,
      "num_input_tokens_seen": 1504529664,
      "step": 29235,
      "train_runtime": 179870.1265,
      "train_tokens_per_second": 8364.533
    },
    {
      "epoch": 1.968758416374899,
      "grad_norm": 0.6872578378249476,
      "learning_rate": 6.045444677140744e-09,
      "loss": 0.2868474245071411,
      "num_input_tokens_seen": 1504793704,
      "step": 29240,
      "train_runtime": 179902.2318,
      "train_tokens_per_second": 8364.508
    },
    {
      "epoch": 1.9690950713708593,
      "grad_norm": 0.7384819481325343,
      "learning_rate": 5.91615959717462e-09,
      "loss": 0.2898238658905029,
      "num_input_tokens_seen": 1505050312,
      "step": 29245,
      "train_runtime": 179936.0755,
      "train_tokens_per_second": 8364.361
    },
    {
      "epoch": 1.9694317263668193,
      "grad_norm": 0.6303992451000372,
      "learning_rate": 5.78827109682989e-09,
      "loss": 0.2685997486114502,
      "num_input_tokens_seen": 1505318560,
      "step": 29250,
      "train_runtime": 179965.7698,
      "train_tokens_per_second": 8364.472
    },
    {
      "epoch": 1.9697683813627793,
      "grad_norm": 0.6565767214270553,
      "learning_rate": 5.661779211869056e-09,
      "loss": 0.2876578807830811,
      "num_input_tokens_seen": 1505568160,
      "step": 29255,
      "train_runtime": 180000.446,
      "train_tokens_per_second": 8364.247
    },
    {
      "epoch": 1.9701050363587396,
      "grad_norm": 0.7098402369573917,
      "learning_rate": 5.536683977666601e-09,
      "loss": 0.3161795139312744,
      "num_input_tokens_seen": 1505848520,
      "step": 29260,
      "train_runtime": 180030.934,
      "train_tokens_per_second": 8364.388
    },
    {
      "epoch": 1.9704416913546998,
      "grad_norm": 0.6138555620857502,
      "learning_rate": 5.412985429203988e-09,
      "loss": 0.2911818981170654,
      "num_input_tokens_seen": 1506094288,
      "step": 29265,
      "train_runtime": 180059.5421,
      "train_tokens_per_second": 8364.424
    },
    {
      "epoch": 1.9707783463506598,
      "grad_norm": 0.6440375128288922,
      "learning_rate": 5.290683601073543e-09,
      "loss": 0.2755396842956543,
      "num_input_tokens_seen": 1506371736,
      "step": 29270,
      "train_runtime": 180089.6213,
      "train_tokens_per_second": 8364.567
    },
    {
      "epoch": 1.9711150013466199,
      "grad_norm": 0.7107370143607851,
      "learning_rate": 5.169778527476243e-09,
      "loss": 0.3043062210083008,
      "num_input_tokens_seen": 1506649872,
      "step": 29275,
      "train_runtime": 180120.6571,
      "train_tokens_per_second": 8364.67
    },
    {
      "epoch": 1.9714516563425801,
      "grad_norm": 0.797644435053485,
      "learning_rate": 5.050270242222821e-09,
      "loss": 0.3170661926269531,
      "num_input_tokens_seen": 1506880704,
      "step": 29280,
      "train_runtime": 180148.6301,
      "train_tokens_per_second": 8364.653
    },
    {
      "epoch": 1.9717883113385404,
      "grad_norm": 0.6846605074606379,
      "learning_rate": 4.932158778733765e-09,
      "loss": 0.2894470453262329,
      "num_input_tokens_seen": 1507139304,
      "step": 29285,
      "train_runtime": 180181.0002,
      "train_tokens_per_second": 8364.585
    },
    {
      "epoch": 1.9721249663345004,
      "grad_norm": 0.7645302943582444,
      "learning_rate": 4.8154441700387636e-09,
      "loss": 0.3222166061401367,
      "num_input_tokens_seen": 1507378752,
      "step": 29290,
      "train_runtime": 180211.9724,
      "train_tokens_per_second": 8364.476
    },
    {
      "epoch": 1.9724616213304604,
      "grad_norm": 0.6653676095510115,
      "learning_rate": 4.7001264487761545e-09,
      "loss": 0.30617542266845704,
      "num_input_tokens_seen": 1507648368,
      "step": 29295,
      "train_runtime": 180244.5317,
      "train_tokens_per_second": 8364.461
    },
    {
      "epoch": 1.9727982763264207,
      "grad_norm": 0.6771516079983738,
      "learning_rate": 4.586205647194031e-09,
      "loss": 0.3050475358963013,
      "num_input_tokens_seen": 1507908488,
      "step": 29300,
      "train_runtime": 180275.6214,
      "train_tokens_per_second": 8364.461
    },
    {
      "epoch": 1.973134931322381,
      "grad_norm": 0.7850328737099909,
      "learning_rate": 4.4736817971507974e-09,
      "loss": 0.30063343048095703,
      "num_input_tokens_seen": 1508155864,
      "step": 29305,
      "train_runtime": 180306.7505,
      "train_tokens_per_second": 8364.389
    },
    {
      "epoch": 1.973471586318341,
      "grad_norm": 0.6567769495784996,
      "learning_rate": 4.362554930112395e-09,
      "loss": 0.2905814409255981,
      "num_input_tokens_seen": 1508415944,
      "step": 29310,
      "train_runtime": 180339.3424,
      "train_tokens_per_second": 8364.32
    },
    {
      "epoch": 1.973808241314301,
      "grad_norm": 0.601809244875547,
      "learning_rate": 4.252825077155631e-09,
      "loss": 0.2922177314758301,
      "num_input_tokens_seen": 1508681056,
      "step": 29315,
      "train_runtime": 180369.5805,
      "train_tokens_per_second": 8364.387
    },
    {
      "epoch": 1.9741448963102612,
      "grad_norm": 0.666718009690019,
      "learning_rate": 4.1444922689665156e-09,
      "loss": 0.29201231002807615,
      "num_input_tokens_seen": 1508926992,
      "step": 29320,
      "train_runtime": 180400.5841,
      "train_tokens_per_second": 8364.313
    },
    {
      "epoch": 1.9744815513062215,
      "grad_norm": 0.7887076177073186,
      "learning_rate": 4.037556535839149e-09,
      "loss": 0.26700987815856936,
      "num_input_tokens_seen": 1509200552,
      "step": 29325,
      "train_runtime": 180432.0843,
      "train_tokens_per_second": 8364.369
    },
    {
      "epoch": 1.9748182063021815,
      "grad_norm": 0.6279954406950227,
      "learning_rate": 3.932017907677943e-09,
      "loss": 0.2826359272003174,
      "num_input_tokens_seen": 1509463568,
      "step": 29330,
      "train_runtime": 180463.7421,
      "train_tokens_per_second": 8364.359
    },
    {
      "epoch": 1.9751548612981416,
      "grad_norm": 0.5954039329199299,
      "learning_rate": 3.827876413997067e-09,
      "loss": 0.31002748012542725,
      "num_input_tokens_seen": 1509698440,
      "step": 29335,
      "train_runtime": 180492.7305,
      "train_tokens_per_second": 8364.317
    },
    {
      "epoch": 1.9754915162941018,
      "grad_norm": 0.7854335650009864,
      "learning_rate": 3.725132083918781e-09,
      "loss": 0.29423749446868896,
      "num_input_tokens_seen": 1509938208,
      "step": 29340,
      "train_runtime": 180522.174,
      "train_tokens_per_second": 8364.281
    },
    {
      "epoch": 1.975828171290062,
      "grad_norm": 0.6546906216173584,
      "learning_rate": 3.623784946175102e-09,
      "loss": 0.31098506450653074,
      "num_input_tokens_seen": 1510160128,
      "step": 29345,
      "train_runtime": 180551.6374,
      "train_tokens_per_second": 8364.145
    },
    {
      "epoch": 1.976164826286022,
      "grad_norm": 0.6889761751757516,
      "learning_rate": 3.5238350291083577e-09,
      "loss": 0.292008638381958,
      "num_input_tokens_seen": 1510400904,
      "step": 29350,
      "train_runtime": 180579.1426,
      "train_tokens_per_second": 8364.205
    },
    {
      "epoch": 1.976501481281982,
      "grad_norm": 0.6561070936924952,
      "learning_rate": 3.4252823606678588e-09,
      "loss": 0.3112313985824585,
      "num_input_tokens_seen": 1510676064,
      "step": 29355,
      "train_runtime": 180609.453,
      "train_tokens_per_second": 8364.324
    },
    {
      "epoch": 1.9768381362779424,
      "grad_norm": 0.7036187745894608,
      "learning_rate": 3.328126968414336e-09,
      "loss": 0.3141892671585083,
      "num_input_tokens_seen": 1510894352,
      "step": 29360,
      "train_runtime": 180635.4293,
      "train_tokens_per_second": 8364.33
    },
    {
      "epoch": 1.9771747912739026,
      "grad_norm": 0.7296627769058565,
      "learning_rate": 3.232368879517167e-09,
      "loss": 0.3143573760986328,
      "num_input_tokens_seen": 1511139448,
      "step": 29365,
      "train_runtime": 180665.1667,
      "train_tokens_per_second": 8364.31
    },
    {
      "epoch": 1.9775114462698626,
      "grad_norm": 0.6641672994552997,
      "learning_rate": 3.1380081207543766e-09,
      "loss": 0.2803871393203735,
      "num_input_tokens_seen": 1511394072,
      "step": 29370,
      "train_runtime": 180695.7017,
      "train_tokens_per_second": 8364.306
    },
    {
      "epoch": 1.9778481012658227,
      "grad_norm": 0.7477202191382726,
      "learning_rate": 3.0450447185137454e-09,
      "loss": 0.29802446365356444,
      "num_input_tokens_seen": 1511629648,
      "step": 29375,
      "train_runtime": 180731.4856,
      "train_tokens_per_second": 8363.953
    },
    {
      "epoch": 1.978184756261783,
      "grad_norm": 0.7293190230022246,
      "learning_rate": 2.953478698792256e-09,
      "loss": 0.2911405324935913,
      "num_input_tokens_seen": 1511903096,
      "step": 29380,
      "train_runtime": 180768.2077,
      "train_tokens_per_second": 8363.767
    },
    {
      "epoch": 1.9785214112577432,
      "grad_norm": 0.7119736837813693,
      "learning_rate": 2.8633100871960917e-09,
      "loss": 0.28659512996673586,
      "num_input_tokens_seen": 1512142968,
      "step": 29385,
      "train_runtime": 180797.723,
      "train_tokens_per_second": 8363.728
    },
    {
      "epoch": 1.9788580662537032,
      "grad_norm": 0.7040845099487462,
      "learning_rate": 2.774538908940638e-09,
      "loss": 0.29539806842803956,
      "num_input_tokens_seen": 1512385872,
      "step": 29390,
      "train_runtime": 180832.2931,
      "train_tokens_per_second": 8363.472
    },
    {
      "epoch": 1.9791947212496632,
      "grad_norm": 0.6138465812405117,
      "learning_rate": 2.687165188849927e-09,
      "loss": 0.29603164196014403,
      "num_input_tokens_seen": 1512662328,
      "step": 29395,
      "train_runtime": 180861.545,
      "train_tokens_per_second": 8363.648
    },
    {
      "epoch": 1.9795313762456235,
      "grad_norm": 0.6477386253752728,
      "learning_rate": 2.6011889513588574e-09,
      "loss": 0.29075088500976565,
      "num_input_tokens_seen": 1512923176,
      "step": 29400,
      "train_runtime": 180891.5032,
      "train_tokens_per_second": 8363.705
    },
    {
      "epoch": 1.9798680312415837,
      "grad_norm": 0.7089221414552866,
      "learning_rate": 2.5166102205093077e-09,
      "loss": 0.3138841152191162,
      "num_input_tokens_seen": 1513192784,
      "step": 29405,
      "train_runtime": 180922.7841,
      "train_tokens_per_second": 8363.749
    },
    {
      "epoch": 1.9802046862375438,
      "grad_norm": 0.6264556558794565,
      "learning_rate": 2.4334290199540254e-09,
      "loss": 0.29514188766479493,
      "num_input_tokens_seen": 1513455872,
      "step": 29410,
      "train_runtime": 180953.4014,
      "train_tokens_per_second": 8363.788
    },
    {
      "epoch": 1.9805413412335038,
      "grad_norm": 0.6237057864706744,
      "learning_rate": 2.351645372953848e-09,
      "loss": 0.28133296966552734,
      "num_input_tokens_seen": 1513701440,
      "step": 29415,
      "train_runtime": 180981.1175,
      "train_tokens_per_second": 8363.864
    },
    {
      "epoch": 1.980877996229464,
      "grad_norm": 0.7428202061240761,
      "learning_rate": 2.2712593023804795e-09,
      "loss": 0.29050416946411134,
      "num_input_tokens_seen": 1513945352,
      "step": 29420,
      "train_runtime": 181010.8569,
      "train_tokens_per_second": 8363.837
    },
    {
      "epoch": 1.9812146512254243,
      "grad_norm": 0.6835759888496816,
      "learning_rate": 2.1922708307120512e-09,
      "loss": 0.30541458129882815,
      "num_input_tokens_seen": 1514198208,
      "step": 29425,
      "train_runtime": 181044.6345,
      "train_tokens_per_second": 8363.673
    },
    {
      "epoch": 1.9815513062213843,
      "grad_norm": 0.6607617431502405,
      "learning_rate": 2.1146799800386698e-09,
      "loss": 0.3338462352752686,
      "num_input_tokens_seen": 1514474504,
      "step": 29430,
      "train_runtime": 181071.7334,
      "train_tokens_per_second": 8363.948
    },
    {
      "epoch": 1.9818879612173443,
      "grad_norm": 0.6625642609593971,
      "learning_rate": 2.0384867720579794e-09,
      "loss": 0.2755138397216797,
      "num_input_tokens_seen": 1514708328,
      "step": 29435,
      "train_runtime": 181101.6678,
      "train_tokens_per_second": 8363.856
    },
    {
      "epoch": 1.9822246162133046,
      "grad_norm": 0.650499762848263,
      "learning_rate": 1.96369122807738e-09,
      "loss": 0.30652446746826173,
      "num_input_tokens_seen": 1514979544,
      "step": 29440,
      "train_runtime": 181134.0211,
      "train_tokens_per_second": 8363.86
    },
    {
      "epoch": 1.9825612712092648,
      "grad_norm": 0.7180181987404669,
      "learning_rate": 1.8902933690123637e-09,
      "loss": 0.2881889343261719,
      "num_input_tokens_seen": 1515244592,
      "step": 29445,
      "train_runtime": 181168.4743,
      "train_tokens_per_second": 8363.732
    },
    {
      "epoch": 1.9828979262052249,
      "grad_norm": 0.7532271030536157,
      "learning_rate": 1.8182932153892884e-09,
      "loss": 0.3009000062942505,
      "num_input_tokens_seen": 1515501088,
      "step": 29450,
      "train_runtime": 181198.6634,
      "train_tokens_per_second": 8363.754
    },
    {
      "epoch": 1.983234581201185,
      "grad_norm": 0.6536701979937497,
      "learning_rate": 1.7476907873426041e-09,
      "loss": 0.28643455505371096,
      "num_input_tokens_seen": 1515775048,
      "step": 29455,
      "train_runtime": 181229.5421,
      "train_tokens_per_second": 8363.841
    },
    {
      "epoch": 1.9835712361971451,
      "grad_norm": 0.6699992724512239,
      "learning_rate": 1.6784861046159617e-09,
      "loss": 0.29562768936157224,
      "num_input_tokens_seen": 1516027160,
      "step": 29460,
      "train_runtime": 181265.0353,
      "train_tokens_per_second": 8363.594
    },
    {
      "epoch": 1.9839078911931054,
      "grad_norm": 0.6804544956221887,
      "learning_rate": 1.610679186561659e-09,
      "loss": 0.30770924091339114,
      "num_input_tokens_seen": 1516285720,
      "step": 29465,
      "train_runtime": 181295.0024,
      "train_tokens_per_second": 8363.638
    },
    {
      "epoch": 1.9842445461890654,
      "grad_norm": 0.6132458130893378,
      "learning_rate": 1.5442700521428599e-09,
      "loss": 0.2755927562713623,
      "num_input_tokens_seen": 1516520072,
      "step": 29470,
      "train_runtime": 181326.8365,
      "train_tokens_per_second": 8363.462
    },
    {
      "epoch": 1.9845812011850255,
      "grad_norm": 0.6726189922531981,
      "learning_rate": 1.4792587199297103e-09,
      "loss": 0.2916428089141846,
      "num_input_tokens_seen": 1516787400,
      "step": 29475,
      "train_runtime": 181356.0723,
      "train_tokens_per_second": 8363.588
    },
    {
      "epoch": 1.9849178561809857,
      "grad_norm": 0.7257399352469489,
      "learning_rate": 1.4156452081026672e-09,
      "loss": 0.31113319396972655,
      "num_input_tokens_seen": 1517060384,
      "step": 29480,
      "train_runtime": 181386.6215,
      "train_tokens_per_second": 8363.684
    },
    {
      "epoch": 1.985254511176946,
      "grad_norm": 0.771726005826126,
      "learning_rate": 1.3534295344513893e-09,
      "loss": 0.30374166965484617,
      "num_input_tokens_seen": 1517328144,
      "step": 29485,
      "train_runtime": 181420.6004,
      "train_tokens_per_second": 8363.593
    },
    {
      "epoch": 1.985591166172906,
      "grad_norm": 0.7519502354051919,
      "learning_rate": 1.2926117163741813e-09,
      "loss": 0.28894968032836915,
      "num_input_tokens_seen": 1517581616,
      "step": 29490,
      "train_runtime": 181450.4469,
      "train_tokens_per_second": 8363.615
    },
    {
      "epoch": 1.985927821168866,
      "grad_norm": 0.6778309006779999,
      "learning_rate": 1.233191770877995e-09,
      "loss": 0.30051801204681394,
      "num_input_tokens_seen": 1517832480,
      "step": 29495,
      "train_runtime": 181481.5907,
      "train_tokens_per_second": 8363.562
    },
    {
      "epoch": 1.9862644761648263,
      "grad_norm": 0.6055615778854064,
      "learning_rate": 1.1751697145800934e-09,
      "loss": 0.2895739316940308,
      "num_input_tokens_seen": 1518108344,
      "step": 29500,
      "train_runtime": 181514.1607,
      "train_tokens_per_second": 8363.581
    },
    {
      "epoch": 1.9866011311607865,
      "grad_norm": 0.722955230586742,
      "learning_rate": 1.1185455637052756e-09,
      "loss": 0.33818070888519286,
      "num_input_tokens_seen": 1518384968,
      "step": 29505,
      "train_runtime": 181545.157,
      "train_tokens_per_second": 8363.677
    },
    {
      "epoch": 1.9869377861567465,
      "grad_norm": 0.7798373038736839,
      "learning_rate": 1.063319334089763e-09,
      "loss": 0.3068626880645752,
      "num_input_tokens_seen": 1518632376,
      "step": 29510,
      "train_runtime": 181577.995,
      "train_tokens_per_second": 8363.527
    },
    {
      "epoch": 1.9872744411527066,
      "grad_norm": 0.7079638346824373,
      "learning_rate": 1.0094910411762027e-09,
      "loss": 0.2818891525268555,
      "num_input_tokens_seen": 1518894032,
      "step": 29515,
      "train_runtime": 181611.745,
      "train_tokens_per_second": 8363.413
    },
    {
      "epoch": 1.9876110961486668,
      "grad_norm": 0.8983091280526774,
      "learning_rate": 9.570607000175536e-10,
      "loss": 0.31378424167633057,
      "num_input_tokens_seen": 1519143880,
      "step": 29520,
      "train_runtime": 181642.6037,
      "train_tokens_per_second": 8363.368
    },
    {
      "epoch": 1.987947751144627,
      "grad_norm": 0.7451000826043283,
      "learning_rate": 9.060283252765311e-10,
      "loss": 0.2947085380554199,
      "num_input_tokens_seen": 1519392512,
      "step": 29525,
      "train_runtime": 181672.0013,
      "train_tokens_per_second": 8363.383
    },
    {
      "epoch": 1.988284406140587,
      "grad_norm": 0.6110034302295655,
      "learning_rate": 8.563939312233871e-10,
      "loss": 0.2806139707565308,
      "num_input_tokens_seen": 1519644504,
      "step": 29530,
      "train_runtime": 181705.352,
      "train_tokens_per_second": 8363.235
    },
    {
      "epoch": 1.9886210611365471,
      "grad_norm": 0.7258883064507552,
      "learning_rate": 8.081575317386847e-10,
      "loss": 0.2978924036026001,
      "num_input_tokens_seen": 1519888264,
      "step": 29535,
      "train_runtime": 181737.0818,
      "train_tokens_per_second": 8363.116
    },
    {
      "epoch": 1.9889577161325074,
      "grad_norm": 0.6223232170708429,
      "learning_rate": 7.613191403116338e-10,
      "loss": 0.30416758060455323,
      "num_input_tokens_seen": 1520150952,
      "step": 29540,
      "train_runtime": 181766.9841,
      "train_tokens_per_second": 8363.185
    },
    {
      "epoch": 1.9892943711284676,
      "grad_norm": 0.6561400853992311,
      "learning_rate": 7.158787700400905e-10,
      "loss": 0.301988410949707,
      "num_input_tokens_seen": 1520416672,
      "step": 29545,
      "train_runtime": 181800.5641,
      "train_tokens_per_second": 8363.102
    },
    {
      "epoch": 1.9896310261244277,
      "grad_norm": 0.6119790317439577,
      "learning_rate": 6.718364336316674e-10,
      "loss": 0.2906239748001099,
      "num_input_tokens_seen": 1520675376,
      "step": 29550,
      "train_runtime": 181833.2927,
      "train_tokens_per_second": 8363.02
    },
    {
      "epoch": 1.9899676811203877,
      "grad_norm": 0.673930214722912,
      "learning_rate": 6.291921434026238e-10,
      "loss": 0.28863816261291503,
      "num_input_tokens_seen": 1520936248,
      "step": 29555,
      "train_runtime": 181861.3648,
      "train_tokens_per_second": 8363.163
    },
    {
      "epoch": 1.990304336116348,
      "grad_norm": 0.5550642870703532,
      "learning_rate": 5.879459112778651e-10,
      "loss": 0.26832118034362795,
      "num_input_tokens_seen": 1521157256,
      "step": 29560,
      "train_runtime": 181887.4441,
      "train_tokens_per_second": 8363.179
    },
    {
      "epoch": 1.9906409911123082,
      "grad_norm": 0.5493975921533074,
      "learning_rate": 5.480977487920536e-10,
      "loss": 0.3061259984970093,
      "num_input_tokens_seen": 1521455624,
      "step": 29565,
      "train_runtime": 181913.2173,
      "train_tokens_per_second": 8363.634
    },
    {
      "epoch": 1.9909776461082682,
      "grad_norm": 0.6352279029764717,
      "learning_rate": 5.096476670890526e-10,
      "loss": 0.29147799015045167,
      "num_input_tokens_seen": 1521707944,
      "step": 29570,
      "train_runtime": 181942.9593,
      "train_tokens_per_second": 8363.654
    },
    {
      "epoch": 1.9913143011042282,
      "grad_norm": 0.7376967693577244,
      "learning_rate": 4.725956769202622e-10,
      "loss": 0.30047430992126467,
      "num_input_tokens_seen": 1521948216,
      "step": 29575,
      "train_runtime": 181972.9937,
      "train_tokens_per_second": 8363.594
    },
    {
      "epoch": 1.9916509561001885,
      "grad_norm": 0.7020330502284878,
      "learning_rate": 4.369417886485039e-10,
      "loss": 0.29830398559570315,
      "num_input_tokens_seen": 1522194000,
      "step": 29580,
      "train_runtime": 181999.7845,
      "train_tokens_per_second": 8363.713
    },
    {
      "epoch": 1.9919876110961487,
      "grad_norm": 0.6980635168209074,
      "learning_rate": 4.026860122430254e-10,
      "loss": 0.2862309694290161,
      "num_input_tokens_seen": 1522463280,
      "step": 29585,
      "train_runtime": 182031.5079,
      "train_tokens_per_second": 8363.735
    },
    {
      "epoch": 1.9923242660921088,
      "grad_norm": 0.7223083179004464,
      "learning_rate": 3.698283572839412e-10,
      "loss": 0.2967146635055542,
      "num_input_tokens_seen": 1522731696,
      "step": 29590,
      "train_runtime": 182063.5082,
      "train_tokens_per_second": 8363.739
    },
    {
      "epoch": 1.9926609210880688,
      "grad_norm": 0.7202757837547031,
      "learning_rate": 3.38368832959457e-10,
      "loss": 0.29752864837646487,
      "num_input_tokens_seen": 1522977240,
      "step": 29595,
      "train_runtime": 182094.0037,
      "train_tokens_per_second": 8363.687
    },
    {
      "epoch": 1.992997576084029,
      "grad_norm": 0.7467444630519463,
      "learning_rate": 3.0830744806753523e-10,
      "loss": 0.303085470199585,
      "num_input_tokens_seen": 1523232384,
      "step": 29600,
      "train_runtime": 182131.2624,
      "train_tokens_per_second": 8363.377
    },
    {
      "epoch": 1.9933342310799893,
      "grad_norm": 0.715331328977553,
      "learning_rate": 2.7964421101478457e-10,
      "loss": 0.2894284248352051,
      "num_input_tokens_seen": 1523488184,
      "step": 29605,
      "train_runtime": 182162.3103,
      "train_tokens_per_second": 8363.356
    },
    {
      "epoch": 1.9936708860759493,
      "grad_norm": 0.6385842339264021,
      "learning_rate": 2.5237912981646016e-10,
      "loss": 0.30422286987304686,
      "num_input_tokens_seen": 1523734576,
      "step": 29610,
      "train_runtime": 182193.6662,
      "train_tokens_per_second": 8363.269
    },
    {
      "epoch": 1.9940075410719094,
      "grad_norm": 0.8111416331164555,
      "learning_rate": 2.265122120975738e-10,
      "loss": 0.2924832820892334,
      "num_input_tokens_seen": 1523987072,
      "step": 29615,
      "train_runtime": 182225.6026,
      "train_tokens_per_second": 8363.189
    },
    {
      "epoch": 1.9943441960678696,
      "grad_norm": 0.58878827756117,
      "learning_rate": 2.020434650906733e-10,
      "loss": 0.2824302911758423,
      "num_input_tokens_seen": 1524248320,
      "step": 29620,
      "train_runtime": 182257.2425,
      "train_tokens_per_second": 8363.17
    },
    {
      "epoch": 1.9946808510638299,
      "grad_norm": 0.642268935587058,
      "learning_rate": 1.7897289563972854e-10,
      "loss": 0.29498944282531736,
      "num_input_tokens_seen": 1524526952,
      "step": 29625,
      "train_runtime": 182292.0089,
      "train_tokens_per_second": 8363.104
    },
    {
      "epoch": 1.99501750605979,
      "grad_norm": 0.6783982342493127,
      "learning_rate": 1.5730051019569037e-10,
      "loss": 0.2845705270767212,
      "num_input_tokens_seen": 1524773008,
      "step": 29630,
      "train_runtime": 182319.9797,
      "train_tokens_per_second": 8363.17
    },
    {
      "epoch": 1.99535416105575,
      "grad_norm": 0.5820580134371864,
      "learning_rate": 1.3702631481926633e-10,
      "loss": 0.2841642379760742,
      "num_input_tokens_seen": 1525032056,
      "step": 29635,
      "train_runtime": 182350.2735,
      "train_tokens_per_second": 8363.201
    },
    {
      "epoch": 1.9956908160517102,
      "grad_norm": 0.6410229576947637,
      "learning_rate": 1.1815031517981023e-10,
      "loss": 0.28531475067138673,
      "num_input_tokens_seen": 1525270240,
      "step": 29640,
      "train_runtime": 182380.0039,
      "train_tokens_per_second": 8363.144
    },
    {
      "epoch": 1.9960274710476704,
      "grad_norm": 0.623617537594343,
      "learning_rate": 1.0067251655643262e-10,
      "loss": 0.28848297595977784,
      "num_input_tokens_seen": 1525541960,
      "step": 29645,
      "train_runtime": 182413.0688,
      "train_tokens_per_second": 8363.118
    },
    {
      "epoch": 1.9963641260436304,
      "grad_norm": 0.7564295061127462,
      "learning_rate": 8.459292383633522e-11,
      "loss": 0.2902336835861206,
      "num_input_tokens_seen": 1525762952,
      "step": 29650,
      "train_runtime": 182442.3876,
      "train_tokens_per_second": 8362.985
    },
    {
      "epoch": 1.9967007810395905,
      "grad_norm": 0.7450784724871532,
      "learning_rate": 6.991154151592128e-11,
      "loss": 0.28372344970703123,
      "num_input_tokens_seen": 1525993008,
      "step": 29655,
      "train_runtime": 182467.2453,
      "train_tokens_per_second": 8363.107
    },
    {
      "epoch": 1.9970374360355507,
      "grad_norm": 0.6165546698932151,
      "learning_rate": 5.662837370135066e-11,
      "loss": 0.3064128398895264,
      "num_input_tokens_seen": 1526254872,
      "step": 29660,
      "train_runtime": 182497.1959,
      "train_tokens_per_second": 8363.169
    },
    {
      "epoch": 1.997374091031511,
      "grad_norm": 0.6932738944362068,
      "learning_rate": 4.47434241074296e-11,
      "loss": 0.2980318307876587,
      "num_input_tokens_seen": 1526507176,
      "step": 29665,
      "train_runtime": 182529.129,
      "train_tokens_per_second": 8363.088
    },
    {
      "epoch": 1.997710746027471,
      "grad_norm": 0.7595877128508282,
      "learning_rate": 3.4256696057055615e-11,
      "loss": 0.2937700510025024,
      "num_input_tokens_seen": 1526758568,
      "step": 29670,
      "train_runtime": 182559.9773,
      "train_tokens_per_second": 8363.052
    },
    {
      "epoch": 1.998047401023431,
      "grad_norm": 0.6654521551794409,
      "learning_rate": 2.5168192482882824e-11,
      "loss": 0.3022348880767822,
      "num_input_tokens_seen": 1527025848,
      "step": 29675,
      "train_runtime": 182588.3801,
      "train_tokens_per_second": 8363.215
    },
    {
      "epoch": 1.9983840560193913,
      "grad_norm": 0.6261714823331505,
      "learning_rate": 1.747791592676684e-11,
      "loss": 0.2663146734237671,
      "num_input_tokens_seen": 1527337896,
      "step": 29680,
      "train_runtime": 182616.0422,
      "train_tokens_per_second": 8363.657
    },
    {
      "epoch": 1.9987207110153515,
      "grad_norm": 0.6482410091751494,
      "learning_rate": 1.1185868539209665e-11,
      "loss": 0.2985062122344971,
      "num_input_tokens_seen": 1527589904,
      "step": 29685,
      "train_runtime": 182643.4534,
      "train_tokens_per_second": 8363.781
    },
    {
      "epoch": 1.9990573660113116,
      "grad_norm": 0.6852254647655432,
      "learning_rate": 6.29205207991479e-12,
      "loss": 0.309328556060791,
      "num_input_tokens_seen": 1527830240,
      "step": 29690,
      "train_runtime": 182676.1144,
      "train_tokens_per_second": 8363.602
    },
    {
      "epoch": 1.9993940210072716,
      "grad_norm": 0.7095386172687059,
      "learning_rate": 2.7964679166769815e-12,
      "loss": 0.2910658359527588,
      "num_input_tokens_seen": 1528082768,
      "step": 29695,
      "train_runtime": 182707.6448,
      "train_tokens_per_second": 8363.54
    },
    {
      "epoch": 1.9997306760032318,
      "grad_norm": 0.7225955814165942,
      "learning_rate": 6.991170281578364e-13,
      "loss": 0.2572767734527588,
      "num_input_tokens_seen": 1528327400,
      "step": 29700,
      "train_runtime": 182739.3454,
      "train_tokens_per_second": 8363.428
    },
    {
      "epoch": 2.0,
      "num_input_tokens_seen": 1528525792,
      "step": 29704,
      "total_flos": 2826770001362944.0,
      "train_loss": 0.34210769783975364,
      "train_runtime": 182910.8182,
      "train_samples_per_second": 0.65,
      "train_steps_per_second": 0.162
    }
  ],
  "logging_steps": 5,
  "max_steps": 29704,
  "num_input_tokens_seen": 1528525792,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2826770001362944.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}