{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.998592210229939,
  "eval_steps": 100,
  "global_step": 3195,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009385265133740028,
      "grad_norm": 0.4665657877922058,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 2.9241,
      "step": 10
    },
    {
      "epoch": 0.018770530267480056,
      "grad_norm": 0.36958208680152893,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 2.9288,
      "step": 20
    },
    {
      "epoch": 0.028155795401220086,
      "grad_norm": 0.3911522626876831,
      "learning_rate": 5.2e-06,
      "loss": 2.8784,
      "step": 30
    },
    {
      "epoch": 0.03754106053496011,
      "grad_norm": 0.6396021842956543,
      "learning_rate": 7.2e-06,
      "loss": 2.9314,
      "step": 40
    },
    {
      "epoch": 0.04692632566870014,
      "grad_norm": 0.5952326059341431,
      "learning_rate": 9.2e-06,
      "loss": 2.8987,
      "step": 50
    },
    {
      "epoch": 0.05631159080244017,
      "grad_norm": 0.570318341255188,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 2.7701,
      "step": 60
    },
    {
      "epoch": 0.0656968559361802,
      "grad_norm": 0.5945647358894348,
      "learning_rate": 1.32e-05,
      "loss": 2.7628,
      "step": 70
    },
    {
      "epoch": 0.07508212106992022,
      "grad_norm": 0.5424015522003174,
      "learning_rate": 1.52e-05,
      "loss": 2.8025,
      "step": 80
    },
    {
      "epoch": 0.08446738620366025,
      "grad_norm": 0.5131893157958984,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 2.6489,
      "step": 90
    },
    {
      "epoch": 0.09385265133740028,
      "grad_norm": 0.550221860408783,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 2.6995,
      "step": 100
    },
    {
      "epoch": 0.09385265133740028,
      "eval_loss": 2.63047194480896,
      "eval_runtime": 937.5041,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.506,
      "step": 100
    },
    {
      "epoch": 0.1032379164711403,
      "grad_norm": 0.42704904079437256,
      "learning_rate": 2.12e-05,
      "loss": 2.6395,
      "step": 110
    },
    {
      "epoch": 0.11262318160488034,
      "grad_norm": 0.44245535135269165,
      "learning_rate": 2.32e-05,
      "loss": 2.6048,
      "step": 120
    },
    {
      "epoch": 0.12200844673862037,
      "grad_norm": 0.3578210473060608,
      "learning_rate": 2.5200000000000003e-05,
      "loss": 2.5695,
      "step": 130
    },
    {
      "epoch": 0.1313937118723604,
      "grad_norm": 0.5043906569480896,
      "learning_rate": 2.7200000000000004e-05,
      "loss": 2.5239,
      "step": 140
    },
    {
      "epoch": 0.14077897700610043,
      "grad_norm": 0.4712482988834381,
      "learning_rate": 2.9199999999999998e-05,
      "loss": 2.4199,
      "step": 150
    },
    {
      "epoch": 0.15016424213984045,
      "grad_norm": 0.5192570090293884,
      "learning_rate": 3.12e-05,
      "loss": 2.5526,
      "step": 160
    },
    {
      "epoch": 0.15954950727358047,
      "grad_norm": 0.48030856251716614,
      "learning_rate": 3.32e-05,
      "loss": 2.4876,
      "step": 170
    },
    {
      "epoch": 0.1689347724073205,
      "grad_norm": 0.9590848088264465,
      "learning_rate": 3.52e-05,
      "loss": 2.439,
      "step": 180
    },
    {
      "epoch": 0.17832003754106054,
      "grad_norm": 0.5423749685287476,
      "learning_rate": 3.72e-05,
      "loss": 2.3547,
      "step": 190
    },
    {
      "epoch": 0.18770530267480057,
      "grad_norm": 0.4835156798362732,
      "learning_rate": 3.9200000000000004e-05,
      "loss": 2.4199,
      "step": 200
    },
    {
      "epoch": 0.18770530267480057,
      "eval_loss": 2.3978819847106934,
      "eval_runtime": 937.0546,
      "eval_samples_per_second": 1.011,
      "eval_steps_per_second": 0.506,
      "step": 200
    },
    {
      "epoch": 0.1970905678085406,
      "grad_norm": 0.7263341546058655,
      "learning_rate": 4.12e-05,
      "loss": 2.4937,
      "step": 210
    },
    {
      "epoch": 0.2064758329422806,
      "grad_norm": 0.6794207096099854,
      "learning_rate": 4.32e-05,
      "loss": 2.3535,
      "step": 220
    },
    {
      "epoch": 0.21586109807602064,
      "grad_norm": 0.6615617275238037,
      "learning_rate": 4.52e-05,
      "loss": 2.3753,
      "step": 230
    },
    {
      "epoch": 0.22524636320976069,
      "grad_norm": 0.7630943059921265,
      "learning_rate": 4.72e-05,
      "loss": 2.3971,
      "step": 240
    },
    {
      "epoch": 0.2346316283435007,
      "grad_norm": 0.7536003589630127,
      "learning_rate": 4.92e-05,
      "loss": 2.3282,
      "step": 250
    },
    {
      "epoch": 0.24401689347724073,
      "grad_norm": 1.0338555574417114,
      "learning_rate": 5.1200000000000004e-05,
      "loss": 2.333,
      "step": 260
    },
    {
      "epoch": 0.25340215861098075,
      "grad_norm": 0.7328284382820129,
      "learning_rate": 5.3200000000000006e-05,
      "loss": 2.3966,
      "step": 270
    },
    {
      "epoch": 0.2627874237447208,
      "grad_norm": 0.8608214855194092,
      "learning_rate": 5.520000000000001e-05,
      "loss": 2.2628,
      "step": 280
    },
    {
      "epoch": 0.2721726888784608,
      "grad_norm": 0.8262203931808472,
      "learning_rate": 5.72e-05,
      "loss": 2.3092,
      "step": 290
    },
    {
      "epoch": 0.28155795401220085,
      "grad_norm": 0.8401615619659424,
      "learning_rate": 5.92e-05,
      "loss": 2.2722,
      "step": 300
    },
    {
      "epoch": 0.28155795401220085,
      "eval_loss": 2.217956066131592,
      "eval_runtime": 936.7717,
      "eval_samples_per_second": 1.011,
      "eval_steps_per_second": 0.506,
      "step": 300
    },
    {
      "epoch": 0.29094321914594085,
      "grad_norm": 0.7364845275878906,
      "learning_rate": 6.12e-05,
      "loss": 2.2179,
      "step": 310
    },
    {
      "epoch": 0.3003284842796809,
      "grad_norm": 0.8354003429412842,
      "learning_rate": 6.3e-05,
      "loss": 2.3654,
      "step": 320
    },
    {
      "epoch": 0.30971374941342095,
      "grad_norm": 1.0776605606079102,
      "learning_rate": 6.500000000000001e-05,
      "loss": 2.2143,
      "step": 330
    },
    {
      "epoch": 0.31909901454716094,
      "grad_norm": 0.8842081427574158,
      "learning_rate": 6.7e-05,
      "loss": 2.2705,
      "step": 340
    },
    {
      "epoch": 0.328484279680901,
      "grad_norm": 1.1752427816390991,
      "learning_rate": 6.9e-05,
      "loss": 2.2334,
      "step": 350
    },
    {
      "epoch": 0.337869544814641,
      "grad_norm": 0.7227160334587097,
      "learning_rate": 7.1e-05,
      "loss": 2.228,
      "step": 360
    },
    {
      "epoch": 0.34725480994838104,
      "grad_norm": 1.1053409576416016,
      "learning_rate": 7.3e-05,
      "loss": 2.2071,
      "step": 370
    },
    {
      "epoch": 0.3566400750821211,
      "grad_norm": 0.9307537078857422,
      "learning_rate": 7.500000000000001e-05,
      "loss": 2.2334,
      "step": 380
    },
    {
      "epoch": 0.3660253402158611,
      "grad_norm": 0.9264342188835144,
      "learning_rate": 7.7e-05,
      "loss": 2.2815,
      "step": 390
    },
    {
      "epoch": 0.37541060534960113,
      "grad_norm": 1.0281509160995483,
      "learning_rate": 7.900000000000001e-05,
      "loss": 2.0762,
      "step": 400
    },
    {
      "epoch": 0.37541060534960113,
      "eval_loss": 2.12506103515625,
      "eval_runtime": 936.7185,
      "eval_samples_per_second": 1.011,
      "eval_steps_per_second": 0.506,
      "step": 400
    },
    {
      "epoch": 0.38479587048334113,
      "grad_norm": 0.979566216468811,
      "learning_rate": 8.1e-05,
      "loss": 2.1801,
      "step": 410
    },
    {
      "epoch": 0.3941811356170812,
      "grad_norm": 1.4208861589431763,
      "learning_rate": 8.3e-05,
      "loss": 2.263,
      "step": 420
    },
    {
      "epoch": 0.40356640075082123,
      "grad_norm": 1.0267932415008545,
      "learning_rate": 8.5e-05,
      "loss": 2.1086,
      "step": 430
    },
    {
      "epoch": 0.4129516658845612,
      "grad_norm": 1.010489583015442,
      "learning_rate": 8.7e-05,
      "loss": 2.1647,
      "step": 440
    },
    {
      "epoch": 0.4223369310183013,
      "grad_norm": 0.784968912601471,
      "learning_rate": 8.900000000000001e-05,
      "loss": 2.2024,
      "step": 450
    },
    {
      "epoch": 0.43172219615204127,
      "grad_norm": 1.1498106718063354,
      "learning_rate": 9.1e-05,
      "loss": 2.0905,
      "step": 460
    },
    {
      "epoch": 0.4411074612857813,
      "grad_norm": 0.7570444345474243,
      "learning_rate": 9.300000000000001e-05,
      "loss": 2.0965,
      "step": 470
    },
    {
      "epoch": 0.45049272641952137,
      "grad_norm": 1.162133812904358,
      "learning_rate": 9.5e-05,
      "loss": 2.1354,
      "step": 480
    },
    {
      "epoch": 0.45987799155326137,
      "grad_norm": 0.963750422000885,
      "learning_rate": 9.7e-05,
      "loss": 2.1286,
      "step": 490
    },
    {
      "epoch": 0.4692632566870014,
      "grad_norm": 0.9272373914718628,
      "learning_rate": 9.900000000000001e-05,
      "loss": 1.9652,
      "step": 500
    },
    {
      "epoch": 0.4692632566870014,
      "eval_loss": 2.085766077041626,
      "eval_runtime": 937.6717,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.506,
      "step": 500
    },
    {
      "epoch": 0.4786485218207414,
      "grad_norm": 0.7409124374389648,
      "learning_rate": 9.999915070025401e-05,
      "loss": 2.099,
      "step": 510
    },
    {
      "epoch": 0.48803378695448146,
      "grad_norm": 0.8032079339027405,
      "learning_rate": 9.999235647539953e-05,
      "loss": 2.1874,
      "step": 520
    },
    {
      "epoch": 0.4974190520882215,
      "grad_norm": 1.00968599319458,
      "learning_rate": 9.997876894893606e-05,
      "loss": 2.2589,
      "step": 530
    },
    {
      "epoch": 0.5068043172219615,
      "grad_norm": 0.7752834558486938,
      "learning_rate": 9.995838996722914e-05,
      "loss": 2.1808,
      "step": 540
    },
    {
      "epoch": 0.5161895823557016,
      "grad_norm": 0.8999218940734863,
      "learning_rate": 9.993122229951354e-05,
      "loss": 2.2034,
      "step": 550
    },
    {
      "epoch": 0.5255748474894416,
      "grad_norm": 0.8991349339485168,
      "learning_rate": 9.989726963751682e-05,
      "loss": 2.1284,
      "step": 560
    },
    {
      "epoch": 0.5349601126231815,
      "grad_norm": 0.9258519411087036,
      "learning_rate": 9.985653659495773e-05,
      "loss": 2.0642,
      "step": 570
    },
    {
      "epoch": 0.5443453777569216,
      "grad_norm": 0.7985184788703918,
      "learning_rate": 9.980902870691931e-05,
      "loss": 1.9404,
      "step": 580
    },
    {
      "epoch": 0.5537306428906617,
      "grad_norm": 0.9277619123458862,
      "learning_rate": 9.975475242909667e-05,
      "loss": 1.9928,
      "step": 590
    },
    {
      "epoch": 0.5631159080244017,
      "grad_norm": 0.8406923413276672,
      "learning_rate": 9.969371513691982e-05,
      "loss": 2.1893,
      "step": 600
    },
    {
      "epoch": 0.5631159080244017,
      "eval_loss": 2.0628976821899414,
      "eval_runtime": 937.447,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.506,
      "step": 600
    },
    {
      "epoch": 0.5725011731581418,
      "grad_norm": 0.678254246711731,
      "learning_rate": 9.962592512455138e-05,
      "loss": 2.1181,
      "step": 610
    },
    {
      "epoch": 0.5818864382918817,
      "grad_norm": 0.7161825895309448,
      "learning_rate": 9.955139160375959e-05,
      "loss": 2.1328,
      "step": 620
    },
    {
      "epoch": 0.5912717034256217,
      "grad_norm": 0.7554607391357422,
      "learning_rate": 9.947012470266645e-05,
      "loss": 2.0865,
      "step": 630
    },
    {
      "epoch": 0.6006569685593618,
      "grad_norm": 0.9590722322463989,
      "learning_rate": 9.938213546437154e-05,
      "loss": 2.1012,
      "step": 640
    },
    {
      "epoch": 0.6100422336931018,
      "grad_norm": 0.8377043008804321,
      "learning_rate": 9.928743584545132e-05,
      "loss": 2.1155,
      "step": 650
    },
    {
      "epoch": 0.6194274988268419,
      "grad_norm": 0.7528682947158813,
      "learning_rate": 9.91860387143345e-05,
      "loss": 2.2455,
      "step": 660
    },
    {
      "epoch": 0.6288127639605818,
      "grad_norm": 0.7164850234985352,
      "learning_rate": 9.907795784955327e-05,
      "loss": 2.155,
      "step": 670
    },
    {
      "epoch": 0.6381980290943219,
      "grad_norm": 0.7665808200836182,
      "learning_rate": 9.896320793787106e-05,
      "loss": 2.0722,
      "step": 680
    },
    {
      "epoch": 0.6475832942280619,
      "grad_norm": 0.8012193441390991,
      "learning_rate": 9.884180457228678e-05,
      "loss": 2.1045,
      "step": 690
    },
    {
      "epoch": 0.656968559361802,
      "grad_norm": 0.8847366571426392,
      "learning_rate": 9.871376424991589e-05,
      "loss": 2.0153,
      "step": 700
    },
    {
      "epoch": 0.656968559361802,
      "eval_loss": 2.047290802001953,
      "eval_runtime": 937.2712,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.506,
      "step": 700
    },
    {
      "epoch": 0.666353824495542,
      "grad_norm": 0.7482985854148865,
      "learning_rate": 9.85791043697488e-05,
      "loss": 1.981,
      "step": 710
    },
    {
      "epoch": 0.675739089629282,
      "grad_norm": 0.776764988899231,
      "learning_rate": 9.843784323028638e-05,
      "loss": 2.062,
      "step": 720
    },
    {
      "epoch": 0.685124354763022,
      "grad_norm": 0.8061379790306091,
      "learning_rate": 9.82900000270536e-05,
      "loss": 1.9919,
      "step": 730
    },
    {
      "epoch": 0.6945096198967621,
      "grad_norm": 0.7650445699691772,
      "learning_rate": 9.813559484999102e-05,
      "loss": 2.0104,
      "step": 740
    },
    {
      "epoch": 0.7038948850305021,
      "grad_norm": 1.2102171182632446,
      "learning_rate": 9.797464868072488e-05,
      "loss": 2.0407,
      "step": 750
    },
    {
      "epoch": 0.7132801501642422,
      "grad_norm": 0.7476488351821899,
      "learning_rate": 9.780718338971591e-05,
      "loss": 1.8859,
      "step": 760
    },
    {
      "epoch": 0.7226654152979821,
      "grad_norm": 0.7931784391403198,
      "learning_rate": 9.763322173328753e-05,
      "loss": 2.1804,
      "step": 770
    },
    {
      "epoch": 0.7320506804317222,
      "grad_norm": 1.1580551862716675,
      "learning_rate": 9.745278735053343e-05,
      "loss": 2.1195,
      "step": 780
    },
    {
      "epoch": 0.7414359455654622,
      "grad_norm": 0.7719026803970337,
      "learning_rate": 9.726590476010548e-05,
      "loss": 1.97,
      "step": 790
    },
    {
      "epoch": 0.7508212106992023,
      "grad_norm": 0.9482008218765259,
      "learning_rate": 9.707259935688187e-05,
      "loss": 1.9911,
      "step": 800
    },
    {
      "epoch": 0.7508212106992023,
      "eval_loss": 2.0318124294281006,
      "eval_runtime": 937.1438,
      "eval_samples_per_second": 1.011,
      "eval_steps_per_second": 0.506,
      "step": 800
    },
    {
      "epoch": 0.7602064758329423,
      "grad_norm": 0.7328064441680908,
      "learning_rate": 9.687289740851622e-05,
      "loss": 2.1643,
      "step": 810
    },
    {
      "epoch": 0.7695917409666823,
      "grad_norm": 0.7916940450668335,
      "learning_rate": 9.666682605186835e-05,
      "loss": 2.166,
      "step": 820
    },
    {
      "epoch": 0.7789770061004223,
      "grad_norm": 0.9480199217796326,
      "learning_rate": 9.645441328931654e-05,
      "loss": 2.108,
      "step": 830
    },
    {
      "epoch": 0.7883622712341624,
      "grad_norm": 0.8467483520507812,
      "learning_rate": 9.62356879849525e-05,
      "loss": 2.0794,
      "step": 840
    },
    {
      "epoch": 0.7977475363679024,
      "grad_norm": 0.6742168664932251,
      "learning_rate": 9.601067986065909e-05,
      "loss": 2.1227,
      "step": 850
    },
    {
      "epoch": 0.8071328015016425,
      "grad_norm": 0.7015961408615112,
      "learning_rate": 9.577941949207146e-05,
      "loss": 2.0288,
      "step": 860
    },
    {
      "epoch": 0.8165180666353824,
      "grad_norm": 0.8289031386375427,
      "learning_rate": 9.556596544693951e-05,
      "loss": 2.037,
      "step": 870
    },
    {
      "epoch": 0.8259033317691225,
      "grad_norm": 0.8550999760627747,
      "learning_rate": 9.53229130894619e-05,
      "loss": 2.1908,
      "step": 880
    },
    {
      "epoch": 0.8352885969028625,
      "grad_norm": 0.7619972229003906,
      "learning_rate": 9.50737019461194e-05,
      "loss": 2.1759,
      "step": 890
    },
    {
      "epoch": 0.8446738620366026,
      "grad_norm": 0.7303412556648254,
      "learning_rate": 9.481836588141808e-05,
      "loss": 2.1041,
      "step": 900
    },
    {
      "epoch": 0.8446738620366026,
      "eval_loss": 2.019794225692749,
      "eval_runtime": 938.1446,
      "eval_samples_per_second": 1.009,
      "eval_steps_per_second": 0.505,
      "step": 900
    },
    {
      "epoch": 0.8540591271703426,
      "grad_norm": 0.6646521687507629,
      "learning_rate": 9.455693959216005e-05,
      "loss": 2.0648,
      "step": 910
    },
    {
      "epoch": 0.8634443923040825,
      "grad_norm": 0.910926878452301,
      "learning_rate": 9.428945860272858e-05,
      "loss": 2.0945,
      "step": 920
    },
    {
      "epoch": 0.8728296574378226,
      "grad_norm": 0.7175999283790588,
      "learning_rate": 9.401595926026077e-05,
      "loss": 2.0488,
      "step": 930
    },
    {
      "epoch": 0.8822149225715626,
      "grad_norm": 1.090782642364502,
      "learning_rate": 9.373647872970852e-05,
      "loss": 1.9902,
      "step": 940
    },
    {
      "epoch": 0.8916001877053027,
      "grad_norm": 1.0074050426483154,
      "learning_rate": 9.345105498878826e-05,
      "loss": 1.9974,
      "step": 950
    },
    {
      "epoch": 0.9009854528390427,
      "grad_norm": 0.6742076277732849,
      "learning_rate": 9.315972682282031e-05,
      "loss": 2.0359,
      "step": 960
    },
    {
      "epoch": 0.9103707179727827,
      "grad_norm": 0.9036829471588135,
      "learning_rate": 9.286253381945837e-05,
      "loss": 2.1047,
      "step": 970
    },
    {
      "epoch": 0.9197559831065227,
      "grad_norm": 1.02723228931427,
      "learning_rate": 9.255951636331028e-05,
      "loss": 1.9049,
      "step": 980
    },
    {
      "epoch": 0.9291412482402628,
      "grad_norm": 0.9634993076324463,
      "learning_rate": 9.225071563045007e-05,
      "loss": 2.045,
      "step": 990
    },
    {
      "epoch": 0.9385265133740028,
      "grad_norm": 0.8380990624427795,
      "learning_rate": 9.193617358282277e-05,
      "loss": 2.0488,
      "step": 1000
    },
    {
      "epoch": 0.9385265133740028,
      "eval_loss": 2.0117270946502686,
      "eval_runtime": 939.385,
      "eval_samples_per_second": 1.008,
      "eval_steps_per_second": 0.505,
      "step": 1000
    },
    {
      "epoch": 0.9479117785077429,
      "grad_norm": 0.8317145705223083,
      "learning_rate": 9.161593296254235e-05,
      "loss": 2.0196,
      "step": 1010
    },
    {
      "epoch": 0.9572970436414828,
      "grad_norm": 0.9297407865524292,
      "learning_rate": 9.129003728608367e-05,
      "loss": 2.0798,
      "step": 1020
    },
    {
      "epoch": 0.9666823087752229,
      "grad_norm": 0.6938600540161133,
      "learning_rate": 9.095853083836902e-05,
      "loss": 2.1225,
      "step": 1030
    },
    {
      "epoch": 0.9760675739089629,
      "grad_norm": 0.7786221504211426,
      "learning_rate": 9.062145866675048e-05,
      "loss": 2.0098,
      "step": 1040
    },
    {
      "epoch": 0.985452839042703,
      "grad_norm": 0.7615213394165039,
      "learning_rate": 9.027886657488862e-05,
      "loss": 2.1385,
      "step": 1050
    },
    {
      "epoch": 0.994838104176443,
      "grad_norm": 0.9867497086524963,
      "learning_rate": 8.993080111652831e-05,
      "loss": 2.1039,
      "step": 1060
    },
    {
      "epoch": 1.004223369310183,
      "grad_norm": 0.747900664806366,
      "learning_rate": 8.95773095891727e-05,
      "loss": 2.104,
      "step": 1070
    },
    {
      "epoch": 1.013608634443923,
      "grad_norm": 0.6495920419692993,
      "learning_rate": 8.921844002765613e-05,
      "loss": 2.0333,
      "step": 1080
    },
    {
      "epoch": 1.022993899577663,
      "grad_norm": 0.8173234462738037,
      "learning_rate": 8.885424119761684e-05,
      "loss": 2.0524,
      "step": 1090
    },
    {
      "epoch": 1.0323791647114031,
      "grad_norm": 0.8507694602012634,
      "learning_rate": 8.848476258887031e-05,
      "loss": 1.897,
      "step": 1100
    },
    {
      "epoch": 1.0323791647114031,
      "eval_loss": 2.001809597015381,
      "eval_runtime": 940.385,
      "eval_samples_per_second": 1.007,
      "eval_steps_per_second": 0.504,
      "step": 1100
    },
    {
      "epoch": 1.0417644298451432,
      "grad_norm": 1.4220669269561768,
      "learning_rate": 8.814775911179585e-05,
      "loss": 1.9925,
      "step": 1110
    },
    {
      "epoch": 1.0511496949788832,
      "grad_norm": 0.7997964024543762,
      "learning_rate": 8.776838783200623e-05,
      "loss": 1.9378,
      "step": 1120
    },
    {
      "epoch": 1.0605349601126233,
      "grad_norm": 2.4897735118865967,
      "learning_rate": 8.738388432665424e-05,
      "loss": 1.855,
      "step": 1130
    },
    {
      "epoch": 1.069920225246363,
      "grad_norm": 0.8852785229682922,
      "learning_rate": 8.699430084469276e-05,
      "loss": 2.0958,
      "step": 1140
    },
    {
      "epoch": 1.0793054903801031,
      "grad_norm": 0.698052167892456,
      "learning_rate": 8.65996903253766e-05,
      "loss": 2.1623,
      "step": 1150
    },
    {
      "epoch": 1.0886907555138432,
      "grad_norm": 0.8183425068855286,
      "learning_rate": 8.620010639106853e-05,
      "loss": 2.0938,
      "step": 1160
    },
    {
      "epoch": 1.0980760206475833,
      "grad_norm": 0.8702631592750549,
      "learning_rate": 8.57956033399528e-05,
      "loss": 1.842,
      "step": 1170
    },
    {
      "epoch": 1.1074612857813233,
      "grad_norm": 0.7535381317138672,
      "learning_rate": 8.538623613865678e-05,
      "loss": 2.069,
      "step": 1180
    },
    {
      "epoch": 1.1168465509150634,
      "grad_norm": 0.7525476813316345,
      "learning_rate": 8.497206041478162e-05,
      "loss": 2.0564,
      "step": 1190
    },
    {
      "epoch": 1.1262318160488034,
      "grad_norm": 0.80061274766922,
      "learning_rate": 8.455313244934324e-05,
      "loss": 2.0298,
      "step": 1200
    },
    {
      "epoch": 1.1262318160488034,
      "eval_loss": 1.9951938390731812,
      "eval_runtime": 939.3487,
      "eval_samples_per_second": 1.008,
      "eval_steps_per_second": 0.505,
      "step": 1200
    },
    {
      "epoch": 1.1356170811825435,
      "grad_norm": 0.7788612246513367,
      "learning_rate": 8.412950916912451e-05,
      "loss": 2.1235,
      "step": 1210
    },
    {
      "epoch": 1.1450023463162835,
      "grad_norm": 0.7296183705329895,
      "learning_rate": 8.370124813893962e-05,
      "loss": 2.1001,
      "step": 1220
    },
    {
      "epoch": 1.1543876114500236,
      "grad_norm": 0.9630563855171204,
      "learning_rate": 8.326840755381176e-05,
      "loss": 2.1847,
      "step": 1230
    },
    {
      "epoch": 1.1637728765837636,
      "grad_norm": 0.8061946630477905,
      "learning_rate": 8.283104623106525e-05,
      "loss": 2.0888,
      "step": 1240
    },
    {
      "epoch": 1.1731581417175034,
      "grad_norm": 0.8088260889053345,
      "learning_rate": 8.238922360233297e-05,
      "loss": 1.9784,
      "step": 1250
    },
    {
      "epoch": 1.1825434068512435,
      "grad_norm": 0.8710635900497437,
      "learning_rate": 8.194299970548045e-05,
      "loss": 1.915,
      "step": 1260
    },
    {
      "epoch": 1.1919286719849835,
      "grad_norm": 0.6845631003379822,
      "learning_rate": 8.149243517644745e-05,
      "loss": 2.2073,
      "step": 1270
    },
    {
      "epoch": 1.2013139371187236,
      "grad_norm": 0.697823166847229,
      "learning_rate": 8.103759124100839e-05,
      "loss": 2.0622,
      "step": 1280
    },
    {
      "epoch": 1.2106992022524636,
      "grad_norm": 0.7662308216094971,
      "learning_rate": 8.057852970645254e-05,
      "loss": 2.0764,
      "step": 1290
    },
    {
      "epoch": 1.2200844673862037,
      "grad_norm": 0.7569838166236877,
      "learning_rate": 8.011531295318526e-05,
      "loss": 2.0989,
      "step": 1300
    },
    {
      "epoch": 1.2200844673862037,
      "eval_loss": 1.9889544248580933,
      "eval_runtime": 940.2035,
      "eval_samples_per_second": 1.007,
      "eval_steps_per_second": 0.504,
      "step": 1300
    },
    {
      "epoch": 1.2294697325199437,
      "grad_norm": 0.7303356528282166,
      "learning_rate": 7.964800392625129e-05,
      "loss": 1.9281,
      "step": 1310
    },
    {
      "epoch": 1.2388549976536838,
      "grad_norm": 0.8145589232444763,
      "learning_rate": 7.917666612678138e-05,
      "loss": 2.0838,
      "step": 1320
    },
    {
      "epoch": 1.2482402627874238,
      "grad_norm": 0.9209080934524536,
      "learning_rate": 7.870136360336328e-05,
      "loss": 2.0761,
      "step": 1330
    },
    {
      "epoch": 1.2576255279211637,
      "grad_norm": 0.8099146485328674,
      "learning_rate": 7.822216094333847e-05,
      "loss": 2.098,
      "step": 1340
    },
    {
      "epoch": 1.267010793054904,
      "grad_norm": 0.7984501719474792,
      "learning_rate": 7.773912326402543e-05,
      "loss": 1.8043,
      "step": 1350
    },
    {
      "epoch": 1.2763960581886438,
      "grad_norm": 0.713470995426178,
      "learning_rate": 7.72523162038713e-05,
      "loss": 1.9197,
      "step": 1360
    },
    {
      "epoch": 1.2857813233223838,
      "grad_norm": 0.8801192045211792,
      "learning_rate": 7.676180591353219e-05,
      "loss": 2.1053,
      "step": 1370
    },
    {
      "epoch": 1.2951665884561239,
      "grad_norm": 0.7982367873191833,
      "learning_rate": 7.626765904688447e-05,
      "loss": 2.2708,
      "step": 1380
    },
    {
      "epoch": 1.304551853589864,
      "grad_norm": 0.840467095375061,
      "learning_rate": 7.576994275196712e-05,
      "loss": 2.0068,
      "step": 1390
    },
    {
      "epoch": 1.313937118723604,
      "grad_norm": 0.8295313715934753,
      "learning_rate": 7.526872466185742e-05,
      "loss": 1.8695,
      "step": 1400
    },
    {
      "epoch": 1.313937118723604,
      "eval_loss": 1.9837820529937744,
      "eval_runtime": 938.2236,
      "eval_samples_per_second": 1.009,
      "eval_steps_per_second": 0.505,
      "step": 1400
    },
    {
      "epoch": 1.323322383857344,
      "grad_norm": 0.7913984060287476,
      "learning_rate": 7.476407288548036e-05,
      "loss": 1.9639,
      "step": 1410
    },
    {
      "epoch": 1.332707648991084,
      "grad_norm": 0.819983720779419,
      "learning_rate": 7.425605599835361e-05,
      "loss": 2.0229,
      "step": 1420
    },
    {
      "epoch": 1.342092914124824,
      "grad_norm": 0.8383524417877197,
      "learning_rate": 7.374474303326896e-05,
      "loss": 1.9001,
      "step": 1430
    },
    {
      "epoch": 1.3514781792585642,
      "grad_norm": 1.1012581586837769,
      "learning_rate": 7.323020347091177e-05,
      "loss": 1.9938,
      "step": 1440
    },
    {
      "epoch": 1.360863444392304,
      "grad_norm": 0.8445969223976135,
      "learning_rate": 7.271250723041932e-05,
      "loss": 2.0726,
      "step": 1450
    },
    {
      "epoch": 1.370248709526044,
      "grad_norm": 0.8882044553756714,
      "learning_rate": 7.21917246598798e-05,
      "loss": 2.0695,
      "step": 1460
    },
    {
      "epoch": 1.379633974659784,
      "grad_norm": 0.7796063423156738,
      "learning_rate": 7.1667926526773e-05,
      "loss": 2.0607,
      "step": 1470
    },
    {
      "epoch": 1.3890192397935242,
      "grad_norm": 0.7294184565544128,
      "learning_rate": 7.114118400835382e-05,
      "loss": 2.1345,
      "step": 1480
    },
    {
      "epoch": 1.3984045049272642,
      "grad_norm": 0.7649526000022888,
      "learning_rate": 7.061156868198028e-05,
      "loss": 1.999,
      "step": 1490
    },
    {
      "epoch": 1.4077897700610043,
      "grad_norm": 0.8375232219696045,
      "learning_rate": 7.007915251538708e-05,
      "loss": 2.1573,
      "step": 1500
    },
    {
      "epoch": 1.4077897700610043,
      "eval_loss": 1.976365566253662,
      "eval_runtime": 937.4612,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.506,
      "step": 1500
    },
    {
      "epoch": 1.4171750351947443,
      "grad_norm": 0.7321649193763733,
      "learning_rate": 6.954400785690622e-05,
      "loss": 2.0845,
      "step": 1510
    },
    {
      "epoch": 1.4265603003284844,
      "grad_norm": 0.778896152973175,
      "learning_rate": 6.900620742563562e-05,
      "loss": 1.9401,
      "step": 1520
    },
    {
      "epoch": 1.4359455654622244,
      "grad_norm": 0.7842182517051697,
      "learning_rate": 6.846582430155783e-05,
      "loss": 1.8992,
      "step": 1530
    },
    {
      "epoch": 1.4453308305959642,
      "grad_norm": 0.6991093754768372,
      "learning_rate": 6.792293191560914e-05,
      "loss": 2.0625,
      "step": 1540
    },
    {
      "epoch": 1.4547160957297043,
      "grad_norm": 0.9950138330459595,
      "learning_rate": 6.737760403970152e-05,
      "loss": 2.0905,
      "step": 1550
    },
    {
      "epoch": 1.4641013608634443,
      "grad_norm": 0.6939354538917542,
      "learning_rate": 6.682991477669781e-05,
      "loss": 2.2633,
      "step": 1560
    },
    {
      "epoch": 1.4734866259971844,
      "grad_norm": 0.842707633972168,
      "learning_rate": 6.627993855034228e-05,
      "loss": 1.8811,
      "step": 1570
    },
    {
      "epoch": 1.4828718911309244,
      "grad_norm": 0.8008860945701599,
      "learning_rate": 6.572775009514725e-05,
      "loss": 1.8528,
      "step": 1580
    },
    {
      "epoch": 1.4922571562646645,
      "grad_norm": 0.7409046292304993,
      "learning_rate": 6.517342444623784e-05,
      "loss": 1.9773,
      "step": 1590
    },
    {
      "epoch": 1.5016424213984045,
      "grad_norm": 0.7854930758476257,
      "learning_rate": 6.461703692915553e-05,
      "loss": 2.0183,
      "step": 1600
    },
    {
      "epoch": 1.5016424213984045,
      "eval_loss": 1.9713027477264404,
      "eval_runtime": 938.8142,
      "eval_samples_per_second": 1.009,
      "eval_steps_per_second": 0.505,
      "step": 1600
    },
    {
      "epoch": 1.5110276865321446,
      "grad_norm": 0.8054217100143433,
      "learning_rate": 6.405866314962252e-05,
      "loss": 2.1303,
      "step": 1610
    },
    {
      "epoch": 1.5204129516658846,
      "grad_norm": 0.7017131447792053,
      "learning_rate": 6.349837898326784e-05,
      "loss": 2.0846,
      "step": 1620
    },
    {
      "epoch": 1.5297982167996245,
      "grad_norm": 0.8393527865409851,
      "learning_rate": 6.293626056531693e-05,
      "loss": 1.8327,
      "step": 1630
    },
    {
      "epoch": 1.5391834819333647,
      "grad_norm": 0.8798466920852661,
      "learning_rate": 6.237238428024572e-05,
      "loss": 1.8657,
      "step": 1640
    },
    {
      "epoch": 1.5485687470671046,
      "grad_norm": 0.7530277371406555,
      "learning_rate": 6.180682675140121e-05,
      "loss": 2.245,
      "step": 1650
    },
    {
      "epoch": 1.5579540122008448,
      "grad_norm": 0.7642443776130676,
      "learning_rate": 6.123966483058916e-05,
      "loss": 1.9058,
      "step": 1660
    },
    {
      "epoch": 1.5673392773345847,
      "grad_norm": 0.7459161281585693,
      "learning_rate": 6.067097558763106e-05,
      "loss": 1.9482,
      "step": 1670
    },
    {
      "epoch": 1.5767245424683247,
      "grad_norm": 0.7460825443267822,
      "learning_rate": 6.0100836299891314e-05,
      "loss": 2.127,
      "step": 1680
    },
    {
      "epoch": 1.5861098076020648,
      "grad_norm": 0.710259735584259,
      "learning_rate": 5.9529324441776314e-05,
      "loss": 2.1407,
      "step": 1690
    },
    {
      "epoch": 1.5954950727358048,
      "grad_norm": 0.7227075695991516,
      "learning_rate": 5.8956517674206605e-05,
      "loss": 1.9229,
      "step": 1700
    },
    {
      "epoch": 1.5954950727358048,
      "eval_loss": 1.967227816581726,
      "eval_runtime": 937.7597,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.505,
      "step": 1700
    },
    {
      "epoch": 1.6048803378695449,
      "grad_norm": 1.0246224403381348,
      "learning_rate": 5.838249383406387e-05,
      "loss": 2.0563,
      "step": 1710
    },
    {
      "epoch": 1.6142656030032847,
      "grad_norm": 0.8386335968971252,
      "learning_rate": 5.780733092361388e-05,
      "loss": 1.8553,
      "step": 1720
    },
    {
      "epoch": 1.623650868137025,
      "grad_norm": 0.7936443090438843,
      "learning_rate": 5.723110709990707e-05,
      "loss": 2.1631,
      "step": 1730
    },
    {
      "epoch": 1.6330361332707648,
      "grad_norm": 0.7047923803329468,
      "learning_rate": 5.6653900664157934e-05,
      "loss": 1.9989,
      "step": 1740
    },
    {
      "epoch": 1.642421398404505,
      "grad_norm": 0.8624520897865295,
      "learning_rate": 5.6075790051105023e-05,
      "loss": 2.0198,
      "step": 1750
    },
    {
      "epoch": 1.651806663538245,
      "grad_norm": 0.8698344826698303,
      "learning_rate": 5.5496853818352614e-05,
      "loss": 2.1045,
      "step": 1760
    },
    {
      "epoch": 1.661191928671985,
      "grad_norm": 0.78273606300354,
      "learning_rate": 5.491717063569582e-05,
      "loss": 1.9399,
      "step": 1770
    },
    {
      "epoch": 1.670577193805725,
      "grad_norm": 0.7217704057693481,
      "learning_rate": 5.433681927443043e-05,
      "loss": 2.0161,
      "step": 1780
    },
    {
      "epoch": 1.679962458939465,
      "grad_norm": 0.7136641144752502,
      "learning_rate": 5.375587859664885e-05,
      "loss": 2.1437,
      "step": 1790
    },
    {
      "epoch": 1.689347724073205,
      "grad_norm": 0.7752694487571716,
      "learning_rate": 5.317442754452379e-05,
      "loss": 1.9732,
      "step": 1800
    },
    {
      "epoch": 1.689347724073205,
      "eval_loss": 1.9616819620132446,
      "eval_runtime": 938.0777,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.505,
      "step": 1800
    },
    {
      "epoch": 1.698732989206945,
      "grad_norm": 0.7387381196022034,
      "learning_rate": 5.2592545129581185e-05,
      "loss": 1.8547,
      "step": 1810
    },
    {
      "epoch": 1.7081182543406852,
      "grad_norm": 0.9277381300926208,
      "learning_rate": 5.2010310421963415e-05,
      "loss": 1.8679,
      "step": 1820
    },
    {
      "epoch": 1.717503519474425,
      "grad_norm": 0.7474396824836731,
      "learning_rate": 5.142780253968481e-05,
      "loss": 2.1186,
      "step": 1830
    },
    {
      "epoch": 1.7268887846081653,
      "grad_norm": 0.8091953992843628,
      "learning_rate": 5.084510063788056e-05,
      "loss": 2.0762,
      "step": 1840
    },
    {
      "epoch": 1.7362740497419051,
      "grad_norm": 0.7326928973197937,
      "learning_rate": 5.02622838980505e-05,
      "loss": 1.9626,
      "step": 1850
    },
    {
      "epoch": 1.7456593148756452,
      "grad_norm": 0.6803217530250549,
      "learning_rate": 4.967943151729945e-05,
      "loss": 1.8606,
      "step": 1860
    },
    {
      "epoch": 1.7550445800093852,
      "grad_norm": 0.7370252013206482,
      "learning_rate": 4.9096622697575394e-05,
      "loss": 1.9649,
      "step": 1870
    },
    {
      "epoch": 1.7644298451431253,
      "grad_norm": 0.7405309677124023,
      "learning_rate": 4.851393663490689e-05,
      "loss": 2.0119,
      "step": 1880
    },
    {
      "epoch": 1.7738151102768653,
      "grad_norm": 0.8256239891052246,
      "learning_rate": 4.793145250864151e-05,
      "loss": 1.9313,
      "step": 1890
    },
    {
      "epoch": 1.7832003754106054,
      "grad_norm": 0.8547274470329285,
      "learning_rate": 4.7349249470686266e-05,
      "loss": 1.6835,
      "step": 1900
    },
    {
      "epoch": 1.7832003754106054,
      "eval_loss": 1.9573733806610107,
      "eval_runtime": 937.9887,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.505,
      "step": 1900
    },
    {
      "epoch": 1.7925856405443454,
      "grad_norm": 0.9643909335136414,
      "learning_rate": 4.676740663475198e-05,
      "loss": 1.8968,
      "step": 1910
    },
    {
      "epoch": 1.8019709056780853,
      "grad_norm": 0.8356963992118835,
      "learning_rate": 4.6186003065602827e-05,
      "loss": 1.9651,
      "step": 1920
    },
    {
      "epoch": 1.8113561708118255,
      "grad_norm": 0.7793695330619812,
      "learning_rate": 4.560511776831235e-05,
      "loss": 2.0038,
      "step": 1930
    },
    {
      "epoch": 1.8207414359455654,
      "grad_norm": 0.7343499660491943,
      "learning_rate": 4.502482967752786e-05,
      "loss": 1.7593,
      "step": 1940
    },
    {
      "epoch": 1.8301267010793056,
      "grad_norm": 0.7215515971183777,
      "learning_rate": 4.444521764674411e-05,
      "loss": 2.0668,
      "step": 1950
    },
    {
      "epoch": 1.8395119662130455,
      "grad_norm": 0.8646712303161621,
      "learning_rate": 4.3866360437588294e-05,
      "loss": 1.9422,
      "step": 1960
    },
    {
      "epoch": 1.8488972313467855,
      "grad_norm": 0.8201174736022949,
      "learning_rate": 4.328833670911724e-05,
      "loss": 2.2159,
      "step": 1970
    },
    {
      "epoch": 1.8582824964805256,
      "grad_norm": 0.8474441766738892,
      "learning_rate": 4.2711225007128765e-05,
      "loss": 2.0485,
      "step": 1980
    },
    {
      "epoch": 1.8676677616142656,
      "grad_norm": 0.679102897644043,
      "learning_rate": 4.213510375348837e-05,
      "loss": 1.9853,
      "step": 1990
    },
    {
      "epoch": 1.8770530267480057,
      "grad_norm": 0.7910708785057068,
      "learning_rate": 4.15600512354726e-05,
      "loss": 1.9874,
      "step": 2000
    },
    {
      "epoch": 1.8770530267480057,
      "eval_loss": 1.9538992643356323,
      "eval_runtime": 937.8684,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.505,
      "step": 2000
    },
    {
      "epoch": 1.8864382918817455,
      "grad_norm": 0.6856018900871277,
      "learning_rate": 4.0986145595131055e-05,
      "loss": 1.9927,
      "step": 2010
    },
    {
      "epoch": 1.8958235570154858,
      "grad_norm": 0.7743241786956787,
      "learning_rate": 4.041346481866768e-05,
      "loss": 1.9437,
      "step": 2020
    },
    {
      "epoch": 1.9052088221492256,
      "grad_norm": 0.7838016748428345,
      "learning_rate": 3.9842086725843625e-05,
      "loss": 1.953,
      "step": 2030
    },
    {
      "epoch": 1.9145940872829659,
      "grad_norm": 0.7300184965133667,
      "learning_rate": 3.9272088959402534e-05,
      "loss": 2.0461,
      "step": 2040
    },
    {
      "epoch": 1.9239793524167057,
      "grad_norm": 0.8169785141944885,
      "learning_rate": 3.8703548974519874e-05,
      "loss": 2.1075,
      "step": 2050
    },
    {
      "epoch": 1.9333646175504458,
      "grad_norm": 0.724233090877533,
      "learning_rate": 3.8136544028277894e-05,
      "loss": 1.897,
      "step": 2060
    },
    {
      "epoch": 1.9427498826841858,
      "grad_norm": 0.7821764945983887,
      "learning_rate": 3.757115116916727e-05,
      "loss": 2.1728,
      "step": 2070
    },
    {
      "epoch": 1.9521351478179259,
      "grad_norm": 0.7325600981712341,
      "learning_rate": 3.7007447226617366e-05,
      "loss": 1.9865,
      "step": 2080
    },
    {
      "epoch": 1.961520412951666,
      "grad_norm": 0.7352250814437866,
      "learning_rate": 3.6445508800556036e-05,
      "loss": 2.0352,
      "step": 2090
    },
    {
      "epoch": 1.970905678085406,
      "grad_norm": 0.6944911479949951,
      "learning_rate": 3.5885412251000745e-05,
      "loss": 1.7607,
      "step": 2100
    },
    {
      "epoch": 1.970905678085406,
      "eval_loss": 1.9512391090393066,
      "eval_runtime": 935.8673,
      "eval_samples_per_second": 1.012,
      "eval_steps_per_second": 0.506,
      "step": 2100
    },
    {
      "epoch": 1.980290943219146,
      "grad_norm": 1.345037579536438,
      "learning_rate": 3.532723368768228e-05,
      "loss": 1.8189,
      "step": 2110
    },
    {
      "epoch": 1.9896762083528858,
      "grad_norm": 0.8876499533653259,
      "learning_rate": 3.477104895970234e-05,
      "loss": 2.0414,
      "step": 2120
    },
    {
      "epoch": 1.999061473486626,
      "grad_norm": 0.832295835018158,
      "learning_rate": 3.4216933645226776e-05,
      "loss": 1.9307,
      "step": 2130
    },
    {
      "epoch": 2.008446738620366,
      "grad_norm": 0.7709248661994934,
      "learning_rate": 3.3664963041215406e-05,
      "loss": 2.0266,
      "step": 2140
    },
    {
      "epoch": 2.017832003754106,
      "grad_norm": 0.8151468634605408,
      "learning_rate": 3.311521215319021e-05,
      "loss": 1.9638,
      "step": 2150
    },
    {
      "epoch": 2.027217268887846,
      "grad_norm": 0.731529951095581,
      "learning_rate": 3.256775568504305e-05,
      "loss": 1.9209,
      "step": 2160
    },
    {
      "epoch": 2.0366025340215863,
      "grad_norm": 0.7643239498138428,
      "learning_rate": 3.202266802888439e-05,
      "loss": 2.0578,
      "step": 2170
    },
    {
      "epoch": 2.045987799155326,
      "grad_norm": 0.7743598222732544,
      "learning_rate": 3.148002325493445e-05,
      "loss": 1.9511,
      "step": 2180
    },
    {
      "epoch": 2.055373064289066,
      "grad_norm": 0.7745042443275452,
      "learning_rate": 3.0939895101457916e-05,
      "loss": 1.9773,
      "step": 2190
    },
    {
      "epoch": 2.0647583294228062,
      "grad_norm": 0.8307595252990723,
      "learning_rate": 3.0402356964744027e-05,
      "loss": 1.9459,
      "step": 2200
    },
    {
      "epoch": 2.0647583294228062,
      "eval_loss": 1.947997808456421,
      "eval_runtime": 937.6638,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.506,
      "step": 2200
    },
    {
      "epoch": 2.074143594556546,
      "grad_norm": 0.7969174981117249,
      "learning_rate": 2.986748188913287e-05,
      "loss": 1.8188,
      "step": 2210
    },
    {
      "epoch": 2.0835288596902863,
      "grad_norm": 2.2084882259368896,
      "learning_rate": 2.9335342557089668e-05,
      "loss": 1.8969,
      "step": 2220
    },
    {
      "epoch": 2.092914124824026,
      "grad_norm": 0.7004146575927734,
      "learning_rate": 2.8806011279328256e-05,
      "loss": 1.7638,
      "step": 2230
    },
    {
      "epoch": 2.1022993899577664,
      "grad_norm": 0.6949226260185242,
      "learning_rate": 2.827955998498482e-05,
      "loss": 2.1413,
      "step": 2240
    },
    {
      "epoch": 2.1116846550915063,
      "grad_norm": 0.733250081539154,
      "learning_rate": 2.775606021184396e-05,
      "loss": 2.0401,
      "step": 2250
    },
    {
      "epoch": 2.1210699202252465,
      "grad_norm": 0.845836341381073,
      "learning_rate": 2.7235583096617346e-05,
      "loss": 1.9184,
      "step": 2260
    },
    {
      "epoch": 2.1304551853589864,
      "grad_norm": 0.7574362754821777,
      "learning_rate": 2.6718199365277397e-05,
      "loss": 2.0152,
      "step": 2270
    },
    {
      "epoch": 2.139840450492726,
      "grad_norm": 0.9965910911560059,
      "learning_rate": 2.6203979323446454e-05,
      "loss": 1.8746,
      "step": 2280
    },
    {
      "epoch": 2.1492257156264665,
      "grad_norm": 0.8181989192962646,
      "learning_rate": 2.5692992846843206e-05,
      "loss": 1.8114,
      "step": 2290
    },
    {
      "epoch": 2.1586109807602063,
      "grad_norm": 0.9191744327545166,
      "learning_rate": 2.5185309371787513e-05,
      "loss": 1.7611,
      "step": 2300
    },
    {
      "epoch": 2.1586109807602063,
      "eval_loss": 1.9463104009628296,
      "eval_runtime": 936.8197,
      "eval_samples_per_second": 1.011,
      "eval_steps_per_second": 0.506,
      "step": 2300
    },
    {
      "epoch": 2.1679962458939466,
      "grad_norm": 0.7393763065338135,
      "learning_rate": 2.468099788576482e-05,
      "loss": 1.9138,
      "step": 2310
    },
    {
      "epoch": 2.1773815110276864,
      "grad_norm": 0.8604278564453125,
      "learning_rate": 2.418012691805191e-05,
      "loss": 1.9187,
      "step": 2320
    },
    {
      "epoch": 2.1867667761614267,
      "grad_norm": 0.8256701827049255,
      "learning_rate": 2.3682764530404365e-05,
      "loss": 1.9313,
      "step": 2330
    },
    {
      "epoch": 2.1961520412951665,
      "grad_norm": 0.8674435019493103,
      "learning_rate": 2.3188978307808125e-05,
      "loss": 2.1127,
      "step": 2340
    },
    {
      "epoch": 2.2055373064289068,
      "grad_norm": 0.8510717153549194,
      "learning_rate": 2.2698835349295472e-05,
      "loss": 1.9931,
      "step": 2350
    },
    {
      "epoch": 2.2149225715626466,
      "grad_norm": 0.8951923847198486,
      "learning_rate": 2.2212402258827115e-05,
      "loss": 1.8811,
      "step": 2360
    },
    {
      "epoch": 2.224307836696387,
      "grad_norm": 0.7499418258666992,
      "learning_rate": 2.172974513624176e-05,
      "loss": 1.9194,
      "step": 2370
    },
    {
      "epoch": 2.2336931018301267,
      "grad_norm": 0.7999364137649536,
      "learning_rate": 2.1250929568273774e-05,
      "loss": 1.9925,
      "step": 2380
    },
    {
      "epoch": 2.2430783669638665,
      "grad_norm": 0.8435518145561218,
      "learning_rate": 2.0776020619641024e-05,
      "loss": 1.9746,
      "step": 2390
    },
    {
      "epoch": 2.252463632097607,
      "grad_norm": 0.8621445298194885,
      "learning_rate": 2.0305082824203343e-05,
      "loss": 1.8491,
      "step": 2400
    },
    {
      "epoch": 2.252463632097607,
      "eval_loss": 1.9441428184509277,
      "eval_runtime": 935.9838,
      "eval_samples_per_second": 1.012,
      "eval_steps_per_second": 0.506,
      "step": 2400
    },
    {
      "epoch": 2.2618488972313466,
      "grad_norm": 0.8402264714241028,
      "learning_rate": 1.9838180176193178e-05,
      "loss": 1.9554,
      "step": 2410
    },
    {
      "epoch": 2.271234162365087,
      "grad_norm": 0.8322455883026123,
      "learning_rate": 1.9375376121519807e-05,
      "loss": 1.9463,
      "step": 2420
    },
    {
      "epoch": 2.2806194274988267,
      "grad_norm": 0.8249323964118958,
      "learning_rate": 1.891673354914761e-05,
      "loss": 1.8215,
      "step": 2430
    },
    {
      "epoch": 2.290004692632567,
      "grad_norm": 0.772278904914856,
      "learning_rate": 1.8462314782550578e-05,
      "loss": 1.9064,
      "step": 2440
    },
    {
      "epoch": 2.299389957766307,
      "grad_norm": 0.7047111392021179,
      "learning_rate": 1.8012181571243097e-05,
      "loss": 2.0491,
      "step": 2450
    },
    {
      "epoch": 2.308775222900047,
      "grad_norm": 0.9021138548851013,
      "learning_rate": 1.756639508238922e-05,
      "loss": 1.9212,
      "step": 2460
    },
    {
      "epoch": 2.318160488033787,
      "grad_norm": 0.9528132677078247,
      "learning_rate": 1.7125015892490753e-05,
      "loss": 2.0436,
      "step": 2470
    },
    {
      "epoch": 2.327545753167527,
      "grad_norm": 0.826156735420227,
      "learning_rate": 1.668810397915568e-05,
      "loss": 1.9951,
      "step": 2480
    },
    {
      "epoch": 2.336931018301267,
      "grad_norm": 0.7792008519172668,
      "learning_rate": 1.6255718712948143e-05,
      "loss": 1.9846,
      "step": 2490
    },
    {
      "epoch": 2.346316283435007,
      "grad_norm": 0.8387865424156189,
      "learning_rate": 1.5827918849320567e-05,
      "loss": 1.9121,
      "step": 2500
    },
    {
      "epoch": 2.346316283435007,
      "eval_loss": 1.9427493810653687,
      "eval_runtime": 936.6706,
      "eval_samples_per_second": 1.011,
      "eval_steps_per_second": 0.506,
      "step": 2500
    },
    {
      "epoch": 2.355701548568747,
      "grad_norm": 0.7230735421180725,
      "learning_rate": 1.5404762520629724e-05,
      "loss": 1.8782,
      "step": 2510
    },
    {
      "epoch": 2.365086813702487,
      "grad_norm": 0.7646508812904358,
      "learning_rate": 1.4986307228237268e-05,
      "loss": 1.8842,
      "step": 2520
    },
    {
      "epoch": 2.3744720788362272,
      "grad_norm": 0.7921754121780396,
      "learning_rate": 1.4572609834695971e-05,
      "loss": 2.1265,
      "step": 2530
    },
    {
      "epoch": 2.383857343969967,
      "grad_norm": 0.8904575109481812,
      "learning_rate": 1.4163726556023054e-05,
      "loss": 1.8978,
      "step": 2540
    },
    {
      "epoch": 2.3932426091037073,
      "grad_norm": 0.9155052900314331,
      "learning_rate": 1.3759712954060921e-05,
      "loss": 1.8854,
      "step": 2550
    },
    {
      "epoch": 2.402627874237447,
      "grad_norm": 0.8808565139770508,
      "learning_rate": 1.3360623928927291e-05,
      "loss": 1.8698,
      "step": 2560
    },
    {
      "epoch": 2.4120131393711874,
      "grad_norm": 0.9437252879142761,
      "learning_rate": 1.2966513711554744e-05,
      "loss": 1.7782,
      "step": 2570
    },
    {
      "epoch": 2.4213984045049273,
      "grad_norm": 0.853032112121582,
      "learning_rate": 1.2577435856321668e-05,
      "loss": 1.953,
      "step": 2580
    },
    {
      "epoch": 2.430783669638667,
      "grad_norm": 0.8684459328651428,
      "learning_rate": 1.219344323377482e-05,
      "loss": 2.2737,
      "step": 2590
    },
    {
      "epoch": 2.4401689347724074,
      "grad_norm": 0.7905233502388,
      "learning_rate": 1.1814588023444878e-05,
      "loss": 1.8849,
      "step": 2600
    },
    {
      "epoch": 2.4401689347724074,
      "eval_loss": 1.941327452659607,
      "eval_runtime": 937.1681,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.506,
      "step": 2600
    },
    {
      "epoch": 2.449554199906147,
      "grad_norm": 0.769902765750885,
      "learning_rate": 1.1440921706756092e-05,
      "loss": 2.057,
      "step": 2610
    },
    {
      "epoch": 2.4589394650398875,
      "grad_norm": 0.8708339333534241,
      "learning_rate": 1.1072495060030418e-05,
      "loss": 1.8389,
      "step": 2620
    },
    {
      "epoch": 2.4683247301736273,
      "grad_norm": 0.7585152387619019,
      "learning_rate": 1.0709358147587884e-05,
      "loss": 1.9889,
      "step": 2630
    },
    {
      "epoch": 2.4777099953073676,
      "grad_norm": 0.8954083919525146,
      "learning_rate": 1.0351560314943392e-05,
      "loss": 2.0466,
      "step": 2640
    },
    {
      "epoch": 2.4870952604411074,
      "grad_norm": 0.8597133755683899,
      "learning_rate": 9.999150182101319e-06,
      "loss": 1.8554,
      "step": 2650
    },
    {
      "epoch": 2.4964805255748477,
      "grad_norm": 0.7284257411956787,
      "learning_rate": 9.652175636948807e-06,
      "loss": 1.9854,
      "step": 2660
    },
    {
      "epoch": 2.5058657907085875,
      "grad_norm": 0.8639576435089111,
      "learning_rate": 9.310683828748251e-06,
      "loss": 1.924,
      "step": 2670
    },
    {
      "epoch": 2.5152510558423273,
      "grad_norm": 0.8123131990432739,
      "learning_rate": 8.974721161730553e-06,
      "loss": 1.9737,
      "step": 2680
    },
    {
      "epoch": 2.5246363209760676,
      "grad_norm": 0.8097216486930847,
      "learning_rate": 8.64433328878917e-06,
      "loss": 2.0566,
      "step": 2690
    },
    {
      "epoch": 2.534021586109808,
      "grad_norm": 0.8345467448234558,
      "learning_rate": 8.319565105276678e-06,
      "loss": 2.0679,
      "step": 2700
    },
    {
      "epoch": 2.534021586109808,
      "eval_loss": 1.9400410652160645,
      "eval_runtime": 937.8634,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.505,
      "step": 2700
    },
    {
      "epoch": 2.5434068512435477,
      "grad_norm": 0.8246074318885803,
      "learning_rate": 8.000460742903987e-06,
      "loss": 2.0611,
      "step": 2710
    },
    {
      "epoch": 2.5527921163772875,
      "grad_norm": 0.8071450591087341,
      "learning_rate": 7.687063563743413e-06,
      "loss": 1.8266,
      "step": 2720
    },
    {
      "epoch": 2.562177381511028,
      "grad_norm": 0.8657225966453552,
      "learning_rate": 7.379416154336455e-06,
      "loss": 2.0888,
      "step": 2730
    },
    {
      "epoch": 2.5715626466447676,
      "grad_norm": 0.914188027381897,
      "learning_rate": 7.077560319906695e-06,
      "loss": 1.8913,
      "step": 2740
    },
    {
      "epoch": 2.580947911778508,
      "grad_norm": 1.0282979011535645,
      "learning_rate": 6.781537078679134e-06,
      "loss": 2.0157,
      "step": 2750
    },
    {
      "epoch": 2.5903331769122477,
      "grad_norm": 0.7493969202041626,
      "learning_rate": 6.491386656306319e-06,
      "loss": 2.0123,
      "step": 2760
    },
    {
      "epoch": 2.5997184420459876,
      "grad_norm": 0.8242475986480713,
      "learning_rate": 6.2071484804021475e-06,
      "loss": 1.8812,
      "step": 2770
    },
    {
      "epoch": 2.609103707179728,
      "grad_norm": 0.9322003722190857,
      "learning_rate": 5.928861175184336e-06,
      "loss": 1.9338,
      "step": 2780
    },
    {
      "epoch": 2.618488972313468,
      "grad_norm": 0.7862038016319275,
      "learning_rate": 5.656562556225692e-06,
      "loss": 2.0133,
      "step": 2790
    },
    {
      "epoch": 2.627874237447208,
      "grad_norm": 0.7466335892677307,
      "learning_rate": 5.3902896253156365e-06,
      "loss": 1.9908,
      "step": 2800
    },
    {
      "epoch": 2.627874237447208,
      "eval_loss": 1.939355492591858,
      "eval_runtime": 937.8925,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.505,
      "step": 2800
    },
    {
      "epoch": 2.6372595025809478,
      "grad_norm": 0.7764114737510681,
      "learning_rate": 5.13007856543209e-06,
      "loss": 1.9348,
      "step": 2810
    },
    {
      "epoch": 2.646644767714688,
      "grad_norm": 0.8519911766052246,
      "learning_rate": 4.875964735824645e-06,
      "loss": 1.9125,
      "step": 2820
    },
    {
      "epoch": 2.656030032848428,
      "grad_norm": 0.8778414726257324,
      "learning_rate": 4.627982667209818e-06,
      "loss": 2.0829,
      "step": 2830
    },
    {
      "epoch": 2.665415297982168,
      "grad_norm": 0.7963272929191589,
      "learning_rate": 4.386166057078639e-06,
      "loss": 1.989,
      "step": 2840
    },
    {
      "epoch": 2.674800563115908,
      "grad_norm": 0.790169894695282,
      "learning_rate": 4.150547765117746e-06,
      "loss": 1.9568,
      "step": 2850
    },
    {
      "epoch": 2.684185828249648,
      "grad_norm": 0.8890179395675659,
      "learning_rate": 3.921159808744085e-06,
      "loss": 2.104,
      "step": 2860
    },
    {
      "epoch": 2.693571093383388,
      "grad_norm": 0.8634796738624573,
      "learning_rate": 3.698033358754205e-06,
      "loss": 2.0033,
      "step": 2870
    },
    {
      "epoch": 2.7029563585171283,
      "grad_norm": 0.7101777791976929,
      "learning_rate": 3.481198735088581e-06,
      "loss": 1.9882,
      "step": 2880
    },
    {
      "epoch": 2.712341623650868,
      "grad_norm": 0.8689360022544861,
      "learning_rate": 3.270685402711471e-06,
      "loss": 1.9517,
      "step": 2890
    },
    {
      "epoch": 2.721726888784608,
      "grad_norm": 0.7733302712440491,
      "learning_rate": 3.0665219676071057e-06,
      "loss": 1.9557,
      "step": 2900
    },
    {
      "epoch": 2.721726888784608,
      "eval_loss": 1.9387511014938354,
      "eval_runtime": 937.6722,
      "eval_samples_per_second": 1.01,
      "eval_steps_per_second": 0.506,
      "step": 2900
    },
    {
      "epoch": 2.7311121539183483,
      "grad_norm": 0.8051707148551941,
      "learning_rate": 2.8687361728924056e-06,
      "loss": 1.9718,
      "step": 2910
    },
    {
      "epoch": 2.740497419052088,
      "grad_norm": 0.8936708569526672,
      "learning_rate": 2.6773548950471572e-06,
      "loss": 2.0474,
      "step": 2920
    },
    {
      "epoch": 2.7498826841858284,
      "grad_norm": 0.7301138639450073,
      "learning_rate": 2.492404140261795e-06,
      "loss": 1.9602,
      "step": 2930
    },
    {
      "epoch": 2.759267949319568,
      "grad_norm": 0.7741363644599915,
      "learning_rate": 2.3139090409034946e-06,
      "loss": 2.0386,
      "step": 2940
    },
    {
      "epoch": 2.768653214453308,
      "grad_norm": 0.771488606929779,
      "learning_rate": 2.1418938521010954e-06,
      "loss": 1.9822,
      "step": 2950
    },
    {
      "epoch": 2.7780384795870483,
      "grad_norm": 0.7883902788162231,
      "learning_rate": 1.9763819484490355e-06,
      "loss": 2.0061,
      "step": 2960
    },
    {
      "epoch": 2.7874237447207886,
      "grad_norm": 6.553975582122803,
      "learning_rate": 1.8173958208311526e-06,
      "loss": 1.9747,
      "step": 2970
    },
    {
      "epoch": 2.7968090098545284,
      "grad_norm": 0.8313648104667664,
      "learning_rate": 1.6649570733643982e-06,
      "loss": 2.0416,
      "step": 2980
    },
    {
      "epoch": 2.8061942749882682,
      "grad_norm": 0.867647647857666,
      "learning_rate": 1.5190864204631672e-06,
      "loss": 1.8462,
      "step": 2990
    },
    {
      "epoch": 2.8155795401220085,
      "grad_norm": 0.740106999874115,
      "learning_rate": 1.3798036840244667e-06,
      "loss": 1.9627,
      "step": 3000
    },
    {
      "epoch": 2.8155795401220085,
      "eval_loss": 1.9384320974349976,
      "eval_runtime": 936.8374,
      "eval_samples_per_second": 1.011,
      "eval_steps_per_second": 0.506,
      "step": 3000
    },
    {
      "epoch": 2.8249648052557483,
      "grad_norm": 0.7124555706977844,
      "learning_rate": 1.2471277907343703e-06,
      "loss": 2.1284,
      "step": 3010
    },
    {
      "epoch": 2.8343500703894886,
      "grad_norm": 0.809935986995697,
      "learning_rate": 1.1210767694961655e-06,
      "loss": 2.0541,
      "step": 3020
    },
    {
      "epoch": 2.8437353355232284,
      "grad_norm": 0.801823079586029,
      "learning_rate": 1.0016677489804171e-06,
      "loss": 1.9914,
      "step": 3030
    },
    {
      "epoch": 2.8531206006569687,
      "grad_norm": 0.7602815628051758,
      "learning_rate": 8.88916955297453e-07,
      "loss": 1.9802,
      "step": 3040
    },
    {
      "epoch": 2.8625058657907085,
      "grad_norm": 0.7440112829208374,
      "learning_rate": 7.8283970979241e-07,
      "loss": 1.9312,
      "step": 3050
    },
    {
      "epoch": 2.871891130924449,
      "grad_norm": 2.742288112640381,
      "learning_rate": 6.834504269632835e-07,
      "loss": 2.03,
      "step": 3060
    },
    {
      "epoch": 2.8812763960581886,
      "grad_norm": 0.815190315246582,
      "learning_rate": 5.907626125022159e-07,
      "loss": 1.9215,
      "step": 3070
    },
    {
      "epoch": 2.8906616611919285,
      "grad_norm": 0.8376407027244568,
      "learning_rate": 5.04788861460187e-07,
      "loss": 1.8776,
      "step": 3080
    },
    {
      "epoch": 2.9000469263256687,
      "grad_norm": 0.9095739126205444,
      "learning_rate": 4.255408565355612e-07,
      "loss": 1.9567,
      "step": 3090
    },
    {
      "epoch": 2.9094321914594086,
      "grad_norm": 0.858132541179657,
      "learning_rate": 3.530293664865514e-07,
      "loss": 1.8339,
      "step": 3100
    },
    {
      "epoch": 2.9094321914594086,
      "eval_loss": 1.9382846355438232,
      "eval_runtime": 937.0394,
      "eval_samples_per_second": 1.011,
      "eval_steps_per_second": 0.506,
      "step": 3100
    },
    {
      "epoch": 2.918817456593149,
      "grad_norm": 0.851668119430542,
      "learning_rate": 2.872642446678897e-07,
      "loss": 2.1549,
      "step": 3110
    },
    {
      "epoch": 2.9282027217268887,
      "grad_norm": 1.059181571006775,
      "learning_rate": 2.2825442769188188e-07,
      "loss": 2.0087,
      "step": 3120
    },
    {
      "epoch": 2.937587986860629,
      "grad_norm": 1.1176784038543701,
      "learning_rate": 1.7600793421402307e-07,
      "loss": 1.9762,
      "step": 3130
    },
    {
      "epoch": 2.9469732519943688,
      "grad_norm": 0.7648433446884155,
      "learning_rate": 1.305318638434083e-07,
      "loss": 1.9598,
      "step": 3140
    },
    {
      "epoch": 2.956358517128109,
      "grad_norm": 0.8515041470527649,
      "learning_rate": 9.183239617795436e-08,
      "loss": 2.0792,
      "step": 3150
    },
    {
      "epoch": 2.965743782261849,
      "grad_norm": 3.4551055431365967,
      "learning_rate": 5.991478996468236e-08,
      "loss": 1.9946,
      "step": 3160
    },
    {
      "epoch": 2.9751290473955887,
      "grad_norm": 0.7828194499015808,
      "learning_rate": 3.4783382385139565e-08,
      "loss": 1.9363,
      "step": 3170
    },
    {
      "epoch": 2.984514312529329,
      "grad_norm": 0.9214401245117188,
      "learning_rate": 1.644158846600963e-08,
      "loss": 1.9451,
      "step": 3180
    },
    {
      "epoch": 2.9938995776630692,
      "grad_norm": 0.8548945188522339,
      "learning_rate": 4.8919006150727195e-09,
      "loss": 1.8411,
      "step": 3190
    },
    {
      "epoch": 2.998592210229939,
      "step": 3195,
      "total_flos": 1.0917373877893988e+19,
      "train_loss": 2.069366264641751,
      "train_runtime": 97122.8592,
      "train_samples_per_second": 0.263,
      "train_steps_per_second": 0.033
    }
  ],
  "logging_steps": 10,
  "max_steps": 3195,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 1.0917373877893988e+19,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}