{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.16324626865671643,
  "eval_steps": 500,
  "global_step": 700,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0002332089552238806,
      "grad_norm": 423.7213439941406,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 5.5445,
      "step": 1
    },
    {
      "epoch": 0.0004664179104477612,
      "grad_norm": 401.9790954589844,
      "learning_rate": 4.000000000000001e-06,
      "loss": 5.1829,
      "step": 2
    },
    {
      "epoch": 0.0006996268656716418,
      "grad_norm": 266.18048095703125,
      "learning_rate": 6e-06,
      "loss": 4.2153,
      "step": 3
    },
    {
      "epoch": 0.0009328358208955224,
      "grad_norm": 226.71279907226562,
      "learning_rate": 8.000000000000001e-06,
      "loss": 3.8925,
      "step": 4
    },
    {
      "epoch": 0.0011660447761194029,
      "grad_norm": 138.92372131347656,
      "learning_rate": 1e-05,
      "loss": 3.5884,
      "step": 5
    },
    {
      "epoch": 0.0013992537313432835,
      "grad_norm": 104.29601287841797,
      "learning_rate": 1.2e-05,
      "loss": 3.1472,
      "step": 6
    },
    {
      "epoch": 0.0016324626865671641,
      "grad_norm": 91.28285217285156,
      "learning_rate": 1.4e-05,
      "loss": 2.9841,
      "step": 7
    },
    {
      "epoch": 0.0018656716417910447,
      "grad_norm": 38.12488555908203,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.6387,
      "step": 8
    },
    {
      "epoch": 0.0020988805970149254,
      "grad_norm": 33.630401611328125,
      "learning_rate": 1.8e-05,
      "loss": 2.5758,
      "step": 9
    },
    {
      "epoch": 0.0023320895522388058,
      "grad_norm": 22.647951126098633,
      "learning_rate": 2e-05,
      "loss": 2.3714,
      "step": 10
    },
    {
      "epoch": 0.0025652985074626866,
      "grad_norm": 21.309112548828125,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 2.648,
      "step": 11
    },
    {
      "epoch": 0.002798507462686567,
      "grad_norm": 9.581689834594727,
      "learning_rate": 2.4e-05,
      "loss": 2.3512,
      "step": 12
    },
    {
      "epoch": 0.003031716417910448,
      "grad_norm": 17.20270538330078,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 2.5495,
      "step": 13
    },
    {
      "epoch": 0.0032649253731343282,
      "grad_norm": 17.363622665405273,
      "learning_rate": 2.8e-05,
      "loss": 2.48,
      "step": 14
    },
    {
      "epoch": 0.003498134328358209,
      "grad_norm": 7.987298488616943,
      "learning_rate": 3.0000000000000004e-05,
      "loss": 2.4462,
      "step": 15
    },
    {
      "epoch": 0.0037313432835820895,
      "grad_norm": 7.776955604553223,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 2.3571,
      "step": 16
    },
    {
      "epoch": 0.00396455223880597,
      "grad_norm": 9.465243339538574,
      "learning_rate": 3.4e-05,
      "loss": 2.4514,
      "step": 17
    },
    {
      "epoch": 0.004197761194029851,
      "grad_norm": 8.567023277282715,
      "learning_rate": 3.6e-05,
      "loss": 2.1793,
      "step": 18
    },
    {
      "epoch": 0.004430970149253731,
      "grad_norm": 8.37545394897461,
      "learning_rate": 3.8e-05,
      "loss": 2.4488,
      "step": 19
    },
    {
      "epoch": 0.0046641791044776115,
      "grad_norm": 6.3132219314575195,
      "learning_rate": 4e-05,
      "loss": 2.2925,
      "step": 20
    },
    {
      "epoch": 0.004897388059701493,
      "grad_norm": 6.038559436798096,
      "learning_rate": 3.99999945818515e-05,
      "loss": 2.2954,
      "step": 21
    },
    {
      "epoch": 0.005130597014925373,
      "grad_norm": 5.058603763580322,
      "learning_rate": 3.999997832740892e-05,
      "loss": 2.2947,
      "step": 22
    },
    {
      "epoch": 0.005363805970149254,
      "grad_norm": 5.408638954162598,
      "learning_rate": 3.9999951236681086e-05,
      "loss": 2.3726,
      "step": 23
    },
    {
      "epoch": 0.005597014925373134,
      "grad_norm": 5.363165855407715,
      "learning_rate": 3.999991330968265e-05,
      "loss": 2.2435,
      "step": 24
    },
    {
      "epoch": 0.005830223880597015,
      "grad_norm": 4.440342426300049,
      "learning_rate": 3.999986454643418e-05,
      "loss": 2.2964,
      "step": 25
    },
    {
      "epoch": 0.006063432835820896,
      "grad_norm": 4.5738725662231445,
      "learning_rate": 3.99998049469621e-05,
      "loss": 2.2094,
      "step": 26
    },
    {
      "epoch": 0.006296641791044776,
      "grad_norm": 9.806268692016602,
      "learning_rate": 3.999973451129869e-05,
      "loss": 2.271,
      "step": 27
    },
    {
      "epoch": 0.0065298507462686565,
      "grad_norm": 112.9478530883789,
      "learning_rate": 3.999965323948212e-05,
      "loss": 2.3381,
      "step": 28
    },
    {
      "epoch": 0.006763059701492537,
      "grad_norm": 4.859926223754883,
      "learning_rate": 3.9999561131556416e-05,
      "loss": 2.3136,
      "step": 29
    },
    {
      "epoch": 0.006996268656716418,
      "grad_norm": 4.374765872955322,
      "learning_rate": 3.9999458187571496e-05,
      "loss": 2.2018,
      "step": 30
    },
    {
      "epoch": 0.0072294776119402986,
      "grad_norm": 4.2191057205200195,
      "learning_rate": 3.999934440758313e-05,
      "loss": 2.1621,
      "step": 31
    },
    {
      "epoch": 0.007462686567164179,
      "grad_norm": 4.751776695251465,
      "learning_rate": 3.9999219791652964e-05,
      "loss": 2.3259,
      "step": 32
    },
    {
      "epoch": 0.007695895522388059,
      "grad_norm": 4.612663745880127,
      "learning_rate": 3.9999084339848515e-05,
      "loss": 2.2397,
      "step": 33
    },
    {
      "epoch": 0.00792910447761194,
      "grad_norm": 15.328047752380371,
      "learning_rate": 3.999893805224318e-05,
      "loss": 2.263,
      "step": 34
    },
    {
      "epoch": 0.00816231343283582,
      "grad_norm": 5.5305705070495605,
      "learning_rate": 3.999878092891622e-05,
      "loss": 2.2587,
      "step": 35
    },
    {
      "epoch": 0.008395522388059701,
      "grad_norm": 8.774163246154785,
      "learning_rate": 3.999861296995275e-05,
      "loss": 2.2607,
      "step": 36
    },
    {
      "epoch": 0.008628731343283583,
      "grad_norm": 5.19516134262085,
      "learning_rate": 3.99984341754438e-05,
      "loss": 2.2421,
      "step": 37
    },
    {
      "epoch": 0.008861940298507462,
      "grad_norm": 3.969237804412842,
      "learning_rate": 3.999824454548622e-05,
      "loss": 2.233,
      "step": 38
    },
    {
      "epoch": 0.009095149253731344,
      "grad_norm": 3.5150845050811768,
      "learning_rate": 3.999804408018277e-05,
      "loss": 2.2537,
      "step": 39
    },
    {
      "epoch": 0.009328358208955223,
      "grad_norm": 3.506343364715576,
      "learning_rate": 3.999783277964205e-05,
      "loss": 2.2368,
      "step": 40
    },
    {
      "epoch": 0.009561567164179104,
      "grad_norm": 3.9643306732177734,
      "learning_rate": 3.9997610643978555e-05,
      "loss": 2.2454,
      "step": 41
    },
    {
      "epoch": 0.009794776119402986,
      "grad_norm": 4.097018718719482,
      "learning_rate": 3.9997377673312644e-05,
      "loss": 2.2105,
      "step": 42
    },
    {
      "epoch": 0.010027985074626865,
      "grad_norm": 3.586857318878174,
      "learning_rate": 3.9997133867770544e-05,
      "loss": 2.2916,
      "step": 43
    },
    {
      "epoch": 0.010261194029850746,
      "grad_norm": 5.354259967803955,
      "learning_rate": 3.9996879227484344e-05,
      "loss": 2.3228,
      "step": 44
    },
    {
      "epoch": 0.010494402985074628,
      "grad_norm": 3.6894421577453613,
      "learning_rate": 3.999661375259202e-05,
      "loss": 2.2154,
      "step": 45
    },
    {
      "epoch": 0.010727611940298507,
      "grad_norm": 3.81148624420166,
      "learning_rate": 3.99963374432374e-05,
      "loss": 2.2654,
      "step": 46
    },
    {
      "epoch": 0.010960820895522388,
      "grad_norm": 3.892672061920166,
      "learning_rate": 3.999605029957021e-05,
      "loss": 2.0993,
      "step": 47
    },
    {
      "epoch": 0.011194029850746268,
      "grad_norm": 3.519918441772461,
      "learning_rate": 3.999575232174601e-05,
      "loss": 2.1821,
      "step": 48
    },
    {
      "epoch": 0.01142723880597015,
      "grad_norm": 16.402650833129883,
      "learning_rate": 3.9995443509926266e-05,
      "loss": 2.1084,
      "step": 49
    },
    {
      "epoch": 0.01166044776119403,
      "grad_norm": 4.631513595581055,
      "learning_rate": 3.999512386427828e-05,
      "loss": 2.2714,
      "step": 50
    },
    {
      "epoch": 0.01189365671641791,
      "grad_norm": 4.3803534507751465,
      "learning_rate": 3.9994793384975254e-05,
      "loss": 2.263,
      "step": 51
    },
    {
      "epoch": 0.012126865671641791,
      "grad_norm": 5.122961521148682,
      "learning_rate": 3.999445207219624e-05,
      "loss": 1.9787,
      "step": 52
    },
    {
      "epoch": 0.012360074626865671,
      "grad_norm": 48.300960540771484,
      "learning_rate": 3.999409992612616e-05,
      "loss": 2.281,
      "step": 53
    },
    {
      "epoch": 0.012593283582089552,
      "grad_norm": 6.367133617401123,
      "learning_rate": 3.999373694695583e-05,
      "loss": 2.2563,
      "step": 54
    },
    {
      "epoch": 0.012826492537313433,
      "grad_norm": 5.271441459655762,
      "learning_rate": 3.999336313488191e-05,
      "loss": 2.2302,
      "step": 55
    },
    {
      "epoch": 0.013059701492537313,
      "grad_norm": 3.770447254180908,
      "learning_rate": 3.999297849010692e-05,
      "loss": 2.2824,
      "step": 56
    },
    {
      "epoch": 0.013292910447761194,
      "grad_norm": 20.51272201538086,
      "learning_rate": 3.999258301283929e-05,
      "loss": 2.2094,
      "step": 57
    },
    {
      "epoch": 0.013526119402985074,
      "grad_norm": 33.304561614990234,
      "learning_rate": 3.999217670329328e-05,
      "loss": 2.3444,
      "step": 58
    },
    {
      "epoch": 0.013759328358208955,
      "grad_norm": 7.211132049560547,
      "learning_rate": 3.999175956168905e-05,
      "loss": 2.0804,
      "step": 59
    },
    {
      "epoch": 0.013992537313432836,
      "grad_norm": 4.490359306335449,
      "learning_rate": 3.9991331588252594e-05,
      "loss": 2.2384,
      "step": 60
    },
    {
      "epoch": 0.014225746268656716,
      "grad_norm": 5.349526882171631,
      "learning_rate": 3.9990892783215806e-05,
      "loss": 2.2158,
      "step": 61
    },
    {
      "epoch": 0.014458955223880597,
      "grad_norm": 6.723339557647705,
      "learning_rate": 3.9990443146816445e-05,
      "loss": 2.2596,
      "step": 62
    },
    {
      "epoch": 0.014692164179104478,
      "grad_norm": 4.265007019042969,
      "learning_rate": 3.99899826792981e-05,
      "loss": 2.2496,
      "step": 63
    },
    {
      "epoch": 0.014925373134328358,
      "grad_norm": 3.5939512252807617,
      "learning_rate": 3.998951138091031e-05,
      "loss": 2.1458,
      "step": 64
    },
    {
      "epoch": 0.01515858208955224,
      "grad_norm": 8.321859359741211,
      "learning_rate": 3.9989029251908377e-05,
      "loss": 2.1587,
      "step": 65
    },
    {
      "epoch": 0.015391791044776119,
      "grad_norm": 10.590601921081543,
      "learning_rate": 3.9988536292553556e-05,
      "loss": 2.2836,
      "step": 66
    },
    {
      "epoch": 0.015625,
      "grad_norm": 3.6230194568634033,
      "learning_rate": 3.998803250311294e-05,
      "loss": 2.1383,
      "step": 67
    },
    {
      "epoch": 0.01585820895522388,
      "grad_norm": 6.873433589935303,
      "learning_rate": 3.9987517883859476e-05,
      "loss": 2.1058,
      "step": 68
    },
    {
      "epoch": 0.016091417910447763,
      "grad_norm": 4.433073997497559,
      "learning_rate": 3.998699243507201e-05,
      "loss": 2.2397,
      "step": 69
    },
    {
      "epoch": 0.01632462686567164,
      "grad_norm": 5.387369632720947,
      "learning_rate": 3.998645615703521e-05,
      "loss": 2.2282,
      "step": 70
    },
    {
      "epoch": 0.01655783582089552,
      "grad_norm": 4.005219459533691,
      "learning_rate": 3.9985909050039675e-05,
      "loss": 2.2086,
      "step": 71
    },
    {
      "epoch": 0.016791044776119403,
      "grad_norm": 3.682222843170166,
      "learning_rate": 3.998535111438181e-05,
      "loss": 2.0818,
      "step": 72
    },
    {
      "epoch": 0.017024253731343284,
      "grad_norm": 3.637570858001709,
      "learning_rate": 3.9984782350363924e-05,
      "loss": 2.1803,
      "step": 73
    },
    {
      "epoch": 0.017257462686567165,
      "grad_norm": 3.78757381439209,
      "learning_rate": 3.998420275829418e-05,
      "loss": 2.1749,
      "step": 74
    },
    {
      "epoch": 0.017490671641791043,
      "grad_norm": 3.294189453125,
      "learning_rate": 3.9983612338486604e-05,
      "loss": 2.0996,
      "step": 75
    },
    {
      "epoch": 0.017723880597014924,
      "grad_norm": 3.8346738815307617,
      "learning_rate": 3.99830110912611e-05,
      "loss": 2.12,
      "step": 76
    },
    {
      "epoch": 0.017957089552238806,
      "grad_norm": 5.455776691436768,
      "learning_rate": 3.9982399016943426e-05,
      "loss": 2.2213,
      "step": 77
    },
    {
      "epoch": 0.018190298507462687,
      "grad_norm": 4.453072547912598,
      "learning_rate": 3.998177611586523e-05,
      "loss": 2.1554,
      "step": 78
    },
    {
      "epoch": 0.01842350746268657,
      "grad_norm": 5.480541706085205,
      "learning_rate": 3.998114238836399e-05,
      "loss": 2.0593,
      "step": 79
    },
    {
      "epoch": 0.018656716417910446,
      "grad_norm": 4.714183330535889,
      "learning_rate": 3.9980497834783075e-05,
      "loss": 2.2054,
      "step": 80
    },
    {
      "epoch": 0.018889925373134327,
      "grad_norm": 7.966398239135742,
      "learning_rate": 3.9979842455471714e-05,
      "loss": 2.1411,
      "step": 81
    },
    {
      "epoch": 0.01912313432835821,
      "grad_norm": 4.027999401092529,
      "learning_rate": 3.9979176250785e-05,
      "loss": 2.0637,
      "step": 82
    },
    {
      "epoch": 0.01935634328358209,
      "grad_norm": 7.136476516723633,
      "learning_rate": 3.99784992210839e-05,
      "loss": 2.1932,
      "step": 83
    },
    {
      "epoch": 0.01958955223880597,
      "grad_norm": 6.693317890167236,
      "learning_rate": 3.997781136673523e-05,
      "loss": 2.3287,
      "step": 84
    },
    {
      "epoch": 0.019822761194029852,
      "grad_norm": 3.998831033706665,
      "learning_rate": 3.9977112688111684e-05,
      "loss": 2.1086,
      "step": 85
    },
    {
      "epoch": 0.02005597014925373,
      "grad_norm": 5.019552230834961,
      "learning_rate": 3.997640318559182e-05,
      "loss": 2.245,
      "step": 86
    },
    {
      "epoch": 0.02028917910447761,
      "grad_norm": 3.880352735519409,
      "learning_rate": 3.9975682859560044e-05,
      "loss": 2.0994,
      "step": 87
    },
    {
      "epoch": 0.020522388059701493,
      "grad_norm": 4.001659393310547,
      "learning_rate": 3.997495171040665e-05,
      "loss": 2.1649,
      "step": 88
    },
    {
      "epoch": 0.020755597014925374,
      "grad_norm": 5.98531436920166,
      "learning_rate": 3.997420973852779e-05,
      "loss": 2.2158,
      "step": 89
    },
    {
      "epoch": 0.020988805970149255,
      "grad_norm": 3.6509780883789062,
      "learning_rate": 3.997345694432547e-05,
      "loss": 2.1527,
      "step": 90
    },
    {
      "epoch": 0.021222014925373133,
      "grad_norm": 41.5565299987793,
      "learning_rate": 3.997269332820755e-05,
      "loss": 2.1809,
      "step": 91
    },
    {
      "epoch": 0.021455223880597014,
      "grad_norm": 3.682171583175659,
      "learning_rate": 3.997191889058779e-05,
      "loss": 2.1234,
      "step": 92
    },
    {
      "epoch": 0.021688432835820896,
      "grad_norm": 5.654005527496338,
      "learning_rate": 3.997113363188579e-05,
      "loss": 2.2371,
      "step": 93
    },
    {
      "epoch": 0.021921641791044777,
      "grad_norm": 3.710350275039673,
      "learning_rate": 3.9970337552527e-05,
      "loss": 2.2371,
      "step": 94
    },
    {
      "epoch": 0.022154850746268658,
      "grad_norm": 3.153291940689087,
      "learning_rate": 3.996953065294276e-05,
      "loss": 1.9611,
      "step": 95
    },
    {
      "epoch": 0.022388059701492536,
      "grad_norm": 3.825709581375122,
      "learning_rate": 3.996871293357026e-05,
      "loss": 2.0728,
      "step": 96
    },
    {
      "epoch": 0.022621268656716417,
      "grad_norm": 4.941300868988037,
      "learning_rate": 3.9967884394852543e-05,
      "loss": 2.1803,
      "step": 97
    },
    {
      "epoch": 0.0228544776119403,
      "grad_norm": 3.1444172859191895,
      "learning_rate": 3.996704503723853e-05,
      "loss": 2.0425,
      "step": 98
    },
    {
      "epoch": 0.02308768656716418,
      "grad_norm": 4.754464626312256,
      "learning_rate": 3.9966194861183e-05,
      "loss": 2.1679,
      "step": 99
    },
    {
      "epoch": 0.02332089552238806,
      "grad_norm": 17.525373458862305,
      "learning_rate": 3.996533386714659e-05,
      "loss": 2.1369,
      "step": 100
    },
    {
      "epoch": 0.02355410447761194,
      "grad_norm": 3.344606876373291,
      "learning_rate": 3.996446205559579e-05,
      "loss": 2.0255,
      "step": 101
    },
    {
      "epoch": 0.02378731343283582,
      "grad_norm": 3.5711669921875,
      "learning_rate": 3.996357942700298e-05,
      "loss": 2.2018,
      "step": 102
    },
    {
      "epoch": 0.0240205223880597,
      "grad_norm": 3.2885384559631348,
      "learning_rate": 3.996268598184636e-05,
      "loss": 2.0955,
      "step": 103
    },
    {
      "epoch": 0.024253731343283583,
      "grad_norm": 3.517833709716797,
      "learning_rate": 3.996178172061003e-05,
      "loss": 2.1727,
      "step": 104
    },
    {
      "epoch": 0.024486940298507464,
      "grad_norm": 3.3114850521087646,
      "learning_rate": 3.9960866643783915e-05,
      "loss": 2.0719,
      "step": 105
    },
    {
      "epoch": 0.024720149253731342,
      "grad_norm": 3.343580961227417,
      "learning_rate": 3.995994075186382e-05,
      "loss": 2.0242,
      "step": 106
    },
    {
      "epoch": 0.024953358208955223,
      "grad_norm": 3.4693753719329834,
      "learning_rate": 3.9959004045351424e-05,
      "loss": 2.0062,
      "step": 107
    },
    {
      "epoch": 0.025186567164179104,
      "grad_norm": 3.1031532287597656,
      "learning_rate": 3.995805652475424e-05,
      "loss": 2.1026,
      "step": 108
    },
    {
      "epoch": 0.025419776119402986,
      "grad_norm": 5.138773441314697,
      "learning_rate": 3.995709819058564e-05,
      "loss": 2.2566,
      "step": 109
    },
    {
      "epoch": 0.025652985074626867,
      "grad_norm": 8.629913330078125,
      "learning_rate": 3.995612904336486e-05,
      "loss": 2.2888,
      "step": 110
    },
    {
      "epoch": 0.025886194029850745,
      "grad_norm": 3.2712810039520264,
      "learning_rate": 3.995514908361702e-05,
      "loss": 2.1317,
      "step": 111
    },
    {
      "epoch": 0.026119402985074626,
      "grad_norm": 4.156867027282715,
      "learning_rate": 3.995415831187306e-05,
      "loss": 2.03,
      "step": 112
    },
    {
      "epoch": 0.026352611940298507,
      "grad_norm": 3.8207807540893555,
      "learning_rate": 3.9953156728669805e-05,
      "loss": 2.0482,
      "step": 113
    },
    {
      "epoch": 0.02658582089552239,
      "grad_norm": 3.4664762020111084,
      "learning_rate": 3.995214433454991e-05,
      "loss": 1.9948,
      "step": 114
    },
    {
      "epoch": 0.02681902985074627,
      "grad_norm": 11.297024726867676,
      "learning_rate": 3.995112113006193e-05,
      "loss": 2.1873,
      "step": 115
    },
    {
      "epoch": 0.027052238805970148,
      "grad_norm": 3.348475456237793,
      "learning_rate": 3.995008711576024e-05,
      "loss": 2.0035,
      "step": 116
    },
    {
      "epoch": 0.02728544776119403,
      "grad_norm": 4.034727573394775,
      "learning_rate": 3.994904229220507e-05,
      "loss": 2.0467,
      "step": 117
    },
    {
      "epoch": 0.02751865671641791,
      "grad_norm": 3.6657028198242188,
      "learning_rate": 3.994798665996255e-05,
      "loss": 1.9762,
      "step": 118
    },
    {
      "epoch": 0.02775186567164179,
      "grad_norm": 3.8150887489318848,
      "learning_rate": 3.994692021960461e-05,
      "loss": 1.9928,
      "step": 119
    },
    {
      "epoch": 0.027985074626865673,
      "grad_norm": 3.2384026050567627,
      "learning_rate": 3.994584297170908e-05,
      "loss": 1.9952,
      "step": 120
    },
    {
      "epoch": 0.028218283582089554,
      "grad_norm": 3.292480230331421,
      "learning_rate": 3.994475491685963e-05,
      "loss": 2.1267,
      "step": 121
    },
    {
      "epoch": 0.02845149253731343,
      "grad_norm": 4.7834391593933105,
      "learning_rate": 3.994365605564577e-05,
      "loss": 2.0106,
      "step": 122
    },
    {
      "epoch": 0.028684701492537313,
      "grad_norm": 3.275834321975708,
      "learning_rate": 3.9942546388662896e-05,
      "loss": 2.0101,
      "step": 123
    },
    {
      "epoch": 0.028917910447761194,
      "grad_norm": 3.0013022422790527,
      "learning_rate": 3.9941425916512226e-05,
      "loss": 2.1472,
      "step": 124
    },
    {
      "epoch": 0.029151119402985076,
      "grad_norm": 3.1672825813293457,
      "learning_rate": 3.994029463980086e-05,
      "loss": 2.0675,
      "step": 125
    },
    {
      "epoch": 0.029384328358208957,
      "grad_norm": 3.963453769683838,
      "learning_rate": 3.993915255914173e-05,
      "loss": 2.169,
      "step": 126
    },
    {
      "epoch": 0.029617537313432835,
      "grad_norm": 6.512315273284912,
      "learning_rate": 3.993799967515365e-05,
      "loss": 2.019,
      "step": 127
    },
    {
      "epoch": 0.029850746268656716,
      "grad_norm": 3.4920895099639893,
      "learning_rate": 3.9936835988461255e-05,
      "loss": 2.0329,
      "step": 128
    },
    {
      "epoch": 0.030083955223880597,
      "grad_norm": 3.3934459686279297,
      "learning_rate": 3.993566149969505e-05,
      "loss": 1.9587,
      "step": 129
    },
    {
      "epoch": 0.03031716417910448,
      "grad_norm": 3.363553762435913,
      "learning_rate": 3.993447620949139e-05,
      "loss": 1.9899,
      "step": 130
    },
    {
      "epoch": 0.03055037313432836,
      "grad_norm": 3.218477487564087,
      "learning_rate": 3.993328011849249e-05,
      "loss": 2.0305,
      "step": 131
    },
    {
      "epoch": 0.030783582089552237,
      "grad_norm": 3.379997968673706,
      "learning_rate": 3.9932073227346404e-05,
      "loss": 2.1174,
      "step": 132
    },
    {
      "epoch": 0.03101679104477612,
      "grad_norm": 3.660712242126465,
      "learning_rate": 3.993085553670704e-05,
      "loss": 2.0363,
      "step": 133
    },
    {
      "epoch": 0.03125,
      "grad_norm": 3.401226282119751,
      "learning_rate": 3.992962704723416e-05,
      "loss": 1.9897,
      "step": 134
    },
    {
      "epoch": 0.03148320895522388,
      "grad_norm": 3.0702459812164307,
      "learning_rate": 3.99283877595934e-05,
      "loss": 2.0938,
      "step": 135
    },
    {
      "epoch": 0.03171641791044776,
      "grad_norm": 2.8834478855133057,
      "learning_rate": 3.992713767445619e-05,
      "loss": 2.0078,
      "step": 136
    },
    {
      "epoch": 0.03194962686567164,
      "grad_norm": 3.4496426582336426,
      "learning_rate": 3.992587679249987e-05,
      "loss": 2.1045,
      "step": 137
    },
    {
      "epoch": 0.032182835820895525,
      "grad_norm": 3.3012664318084717,
      "learning_rate": 3.9924605114407596e-05,
      "loss": 2.0047,
      "step": 138
    },
    {
      "epoch": 0.0324160447761194,
      "grad_norm": 5.3053669929504395,
      "learning_rate": 3.9923322640868385e-05,
      "loss": 1.9839,
      "step": 139
    },
    {
      "epoch": 0.03264925373134328,
      "grad_norm": 3.9441440105438232,
      "learning_rate": 3.992202937257709e-05,
      "loss": 2.1236,
      "step": 140
    },
    {
      "epoch": 0.032882462686567165,
      "grad_norm": 3.4359235763549805,
      "learning_rate": 3.992072531023444e-05,
      "loss": 2.0132,
      "step": 141
    },
    {
      "epoch": 0.03311567164179104,
      "grad_norm": 3.8967628479003906,
      "learning_rate": 3.991941045454698e-05,
      "loss": 1.9305,
      "step": 142
    },
    {
      "epoch": 0.03334888059701493,
      "grad_norm": 3.1435556411743164,
      "learning_rate": 3.991808480622713e-05,
      "loss": 2.0303,
      "step": 143
    },
    {
      "epoch": 0.033582089552238806,
      "grad_norm": 4.330624103546143,
      "learning_rate": 3.991674836599313e-05,
      "loss": 1.9745,
      "step": 144
    },
    {
      "epoch": 0.033815298507462684,
      "grad_norm": 5.981459140777588,
      "learning_rate": 3.9915401134569104e-05,
      "loss": 2.1993,
      "step": 145
    },
    {
      "epoch": 0.03404850746268657,
      "grad_norm": 3.1248579025268555,
      "learning_rate": 3.991404311268499e-05,
      "loss": 1.9074,
      "step": 146
    },
    {
      "epoch": 0.034281716417910446,
      "grad_norm": 16.910768508911133,
      "learning_rate": 3.991267430107659e-05,
      "loss": 1.9766,
      "step": 147
    },
    {
      "epoch": 0.03451492537313433,
      "grad_norm": 3.268249988555908,
      "learning_rate": 3.991129470048553e-05,
      "loss": 1.9185,
      "step": 148
    },
    {
      "epoch": 0.03474813432835821,
      "grad_norm": 3.2575478553771973,
      "learning_rate": 3.9909904311659315e-05,
      "loss": 1.9919,
      "step": 149
    },
    {
      "epoch": 0.034981343283582086,
      "grad_norm": 3.324004650115967,
      "learning_rate": 3.990850313535128e-05,
      "loss": 2.0377,
      "step": 150
    },
    {
      "epoch": 0.03521455223880597,
      "grad_norm": 3.281524658203125,
      "learning_rate": 3.99070911723206e-05,
      "loss": 2.0168,
      "step": 151
    },
    {
      "epoch": 0.03544776119402985,
      "grad_norm": 3.652379274368286,
      "learning_rate": 3.990566842333228e-05,
      "loss": 1.9907,
      "step": 152
    },
    {
      "epoch": 0.035680970149253734,
      "grad_norm": 3.1975131034851074,
      "learning_rate": 3.990423488915721e-05,
      "loss": 2.1955,
      "step": 153
    },
    {
      "epoch": 0.03591417910447761,
      "grad_norm": 3.227027177810669,
      "learning_rate": 3.990279057057209e-05,
      "loss": 1.9631,
      "step": 154
    },
    {
      "epoch": 0.03614738805970149,
      "grad_norm": 3.062188148498535,
      "learning_rate": 3.9901335468359475e-05,
      "loss": 1.8978,
      "step": 155
    },
    {
      "epoch": 0.036380597014925374,
      "grad_norm": 3.0640323162078857,
      "learning_rate": 3.9899869583307764e-05,
      "loss": 1.9497,
      "step": 156
    },
    {
      "epoch": 0.03661380597014925,
      "grad_norm": 3.414837121963501,
      "learning_rate": 3.989839291621119e-05,
      "loss": 1.9659,
      "step": 157
    },
    {
      "epoch": 0.03684701492537314,
      "grad_norm": 4.126245498657227,
      "learning_rate": 3.989690546786983e-05,
      "loss": 2.0608,
      "step": 158
    },
    {
      "epoch": 0.037080223880597014,
      "grad_norm": 3.0987002849578857,
      "learning_rate": 3.9895407239089615e-05,
      "loss": 1.9818,
      "step": 159
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 12.345720291137695,
      "learning_rate": 3.98938982306823e-05,
      "loss": 1.9607,
      "step": 160
    },
    {
      "epoch": 0.03754664179104478,
      "grad_norm": 3.713212251663208,
      "learning_rate": 3.989237844346549e-05,
      "loss": 2.0619,
      "step": 161
    },
    {
      "epoch": 0.037779850746268655,
      "grad_norm": 3.197726249694824,
      "learning_rate": 3.9890847878262635e-05,
      "loss": 2.0587,
      "step": 162
    },
    {
      "epoch": 0.03801305970149254,
      "grad_norm": 3.1207189559936523,
      "learning_rate": 3.9889306535903015e-05,
      "loss": 2.0071,
      "step": 163
    },
    {
      "epoch": 0.03824626865671642,
      "grad_norm": 3.6011950969696045,
      "learning_rate": 3.988775441722174e-05,
      "loss": 1.9336,
      "step": 164
    },
    {
      "epoch": 0.038479477611940295,
      "grad_norm": 3.658418893814087,
      "learning_rate": 3.9886191523059785e-05,
      "loss": 1.9423,
      "step": 165
    },
    {
      "epoch": 0.03871268656716418,
      "grad_norm": 3.060042381286621,
      "learning_rate": 3.9884617854263946e-05,
      "loss": 1.7841,
      "step": 166
    },
    {
      "epoch": 0.03894589552238806,
      "grad_norm": 3.0380563735961914,
      "learning_rate": 3.988303341168686e-05,
      "loss": 2.0729,
      "step": 167
    },
    {
      "epoch": 0.03917910447761194,
      "grad_norm": 2.719310998916626,
      "learning_rate": 3.9881438196187e-05,
      "loss": 1.8692,
      "step": 168
    },
    {
      "epoch": 0.03941231343283582,
      "grad_norm": 4.415019989013672,
      "learning_rate": 3.987983220862868e-05,
      "loss": 2.0537,
      "step": 169
    },
    {
      "epoch": 0.039645522388059705,
      "grad_norm": 2.7036170959472656,
      "learning_rate": 3.9878215449882036e-05,
      "loss": 1.9642,
      "step": 170
    },
    {
      "epoch": 0.03987873134328358,
      "grad_norm": 3.0154969692230225,
      "learning_rate": 3.987658792082307e-05,
      "loss": 1.9907,
      "step": 171
    },
    {
      "epoch": 0.04011194029850746,
      "grad_norm": 3.5329384803771973,
      "learning_rate": 3.987494962233359e-05,
      "loss": 1.838,
      "step": 172
    },
    {
      "epoch": 0.040345149253731345,
      "grad_norm": 3.2209582328796387,
      "learning_rate": 3.987330055530126e-05,
      "loss": 2.0288,
      "step": 173
    },
    {
      "epoch": 0.04057835820895522,
      "grad_norm": 3.63930344581604,
      "learning_rate": 3.987164072061956e-05,
      "loss": 1.9419,
      "step": 174
    },
    {
      "epoch": 0.04081156716417911,
      "grad_norm": 3.329287528991699,
      "learning_rate": 3.986997011918781e-05,
      "loss": 1.9797,
      "step": 175
    },
    {
      "epoch": 0.041044776119402986,
      "grad_norm": 2.8629071712493896,
      "learning_rate": 3.9868288751911174e-05,
      "loss": 1.8674,
      "step": 176
    },
    {
      "epoch": 0.04127798507462686,
      "grad_norm": 2.942261219024658,
      "learning_rate": 3.9866596619700645e-05,
      "loss": 1.9551,
      "step": 177
    },
    {
      "epoch": 0.04151119402985075,
      "grad_norm": 2.76924991607666,
      "learning_rate": 3.986489372347304e-05,
      "loss": 1.9898,
      "step": 178
    },
    {
      "epoch": 0.041744402985074626,
      "grad_norm": 3.235534191131592,
      "learning_rate": 3.986318006415101e-05,
      "loss": 2.075,
      "step": 179
    },
    {
      "epoch": 0.04197761194029851,
      "grad_norm": 2.618192434310913,
      "learning_rate": 3.986145564266305e-05,
      "loss": 1.8968,
      "step": 180
    },
    {
      "epoch": 0.04221082089552239,
      "grad_norm": 2.873217821121216,
      "learning_rate": 3.9859720459943466e-05,
      "loss": 1.8983,
      "step": 181
    },
    {
      "epoch": 0.042444029850746266,
      "grad_norm": 16.802047729492188,
      "learning_rate": 3.985797451693242e-05,
      "loss": 2.148,
      "step": 182
    },
    {
      "epoch": 0.04267723880597015,
      "grad_norm": 4.490235805511475,
      "learning_rate": 3.985621781457588e-05,
      "loss": 1.9332,
      "step": 183
    },
    {
      "epoch": 0.04291044776119403,
      "grad_norm": 3.088561773300171,
      "learning_rate": 3.985445035382565e-05,
      "loss": 1.9186,
      "step": 184
    },
    {
      "epoch": 0.043143656716417914,
      "grad_norm": 2.596893310546875,
      "learning_rate": 3.985267213563937e-05,
      "loss": 1.8679,
      "step": 185
    },
    {
      "epoch": 0.04337686567164179,
      "grad_norm": 4.679966449737549,
      "learning_rate": 3.985088316098051e-05,
      "loss": 2.1873,
      "step": 186
    },
    {
      "epoch": 0.04361007462686567,
      "grad_norm": 2.7401936054229736,
      "learning_rate": 3.984908343081836e-05,
      "loss": 2.0631,
      "step": 187
    },
    {
      "epoch": 0.043843283582089554,
      "grad_norm": 3.5176198482513428,
      "learning_rate": 3.984727294612805e-05,
      "loss": 2.1258,
      "step": 188
    },
    {
      "epoch": 0.04407649253731343,
      "grad_norm": 3.1320960521698,
      "learning_rate": 3.984545170789051e-05,
      "loss": 1.9133,
      "step": 189
    },
    {
      "epoch": 0.044309701492537316,
      "grad_norm": 2.9197463989257812,
      "learning_rate": 3.984361971709252e-05,
      "loss": 1.9789,
      "step": 190
    },
    {
      "epoch": 0.044542910447761194,
      "grad_norm": 2.8054819107055664,
      "learning_rate": 3.9841776974726684e-05,
      "loss": 1.9821,
      "step": 191
    },
    {
      "epoch": 0.04477611940298507,
      "grad_norm": 5.703899383544922,
      "learning_rate": 3.983992348179141e-05,
      "loss": 2.0093,
      "step": 192
    },
    {
      "epoch": 0.04500932835820896,
      "grad_norm": 2.9593870639801025,
      "learning_rate": 3.9838059239290984e-05,
      "loss": 1.8733,
      "step": 193
    },
    {
      "epoch": 0.045242537313432835,
      "grad_norm": 3.220139980316162,
      "learning_rate": 3.983618424823545e-05,
      "loss": 1.9619,
      "step": 194
    },
    {
      "epoch": 0.04547574626865672,
      "grad_norm": 2.6724166870117188,
      "learning_rate": 3.9834298509640724e-05,
      "loss": 1.9359,
      "step": 195
    },
    {
      "epoch": 0.0457089552238806,
      "grad_norm": 3.129688262939453,
      "learning_rate": 3.983240202452851e-05,
      "loss": 1.9413,
      "step": 196
    },
    {
      "epoch": 0.045942164179104475,
      "grad_norm": 2.4414985179901123,
      "learning_rate": 3.9830494793926354e-05,
      "loss": 1.9618,
      "step": 197
    },
    {
      "epoch": 0.04617537313432836,
      "grad_norm": 3.0320217609405518,
      "learning_rate": 3.982857681886763e-05,
      "loss": 2.0387,
      "step": 198
    },
    {
      "epoch": 0.04640858208955224,
      "grad_norm": 2.540480852127075,
      "learning_rate": 3.982664810039153e-05,
      "loss": 1.9275,
      "step": 199
    },
    {
      "epoch": 0.04664179104477612,
      "grad_norm": 2.731313943862915,
      "learning_rate": 3.982470863954306e-05,
      "loss": 1.984,
      "step": 200
    },
    {
      "epoch": 0.046875,
      "grad_norm": 2.7708122730255127,
      "learning_rate": 3.982275843737304e-05,
      "loss": 1.9202,
      "step": 201
    },
    {
      "epoch": 0.04710820895522388,
      "grad_norm": 5.1076250076293945,
      "learning_rate": 3.982079749493812e-05,
      "loss": 2.103,
      "step": 202
    },
    {
      "epoch": 0.04734141791044776,
      "grad_norm": 2.4482932090759277,
      "learning_rate": 3.981882581330076e-05,
      "loss": 1.9025,
      "step": 203
    },
    {
      "epoch": 0.04757462686567164,
      "grad_norm": 2.5416674613952637,
      "learning_rate": 3.981684339352928e-05,
      "loss": 1.865,
      "step": 204
    },
    {
      "epoch": 0.047807835820895525,
      "grad_norm": 2.681642770767212,
      "learning_rate": 3.981485023669774e-05,
      "loss": 1.9344,
      "step": 205
    },
    {
      "epoch": 0.0480410447761194,
      "grad_norm": 2.6333141326904297,
      "learning_rate": 3.98128463438861e-05,
      "loss": 1.9102,
      "step": 206
    },
    {
      "epoch": 0.04827425373134328,
      "grad_norm": 4.268520355224609,
      "learning_rate": 3.981083171618007e-05,
      "loss": 1.8022,
      "step": 207
    },
    {
      "epoch": 0.048507462686567165,
      "grad_norm": 2.3018503189086914,
      "learning_rate": 3.980880635467122e-05,
      "loss": 1.8682,
      "step": 208
    },
    {
      "epoch": 0.04874067164179104,
      "grad_norm": 2.244959831237793,
      "learning_rate": 3.980677026045692e-05,
      "loss": 1.8426,
      "step": 209
    },
    {
      "epoch": 0.04897388059701493,
      "grad_norm": 2.2810251712799072,
      "learning_rate": 3.980472343464036e-05,
      "loss": 1.9129,
      "step": 210
    },
    {
      "epoch": 0.049207089552238806,
      "grad_norm": 2.5715503692626953,
      "learning_rate": 3.9802665878330525e-05,
      "loss": 1.8085,
      "step": 211
    },
    {
      "epoch": 0.049440298507462684,
      "grad_norm": 2.3552956581115723,
      "learning_rate": 3.9800597592642246e-05,
      "loss": 1.8618,
      "step": 212
    },
    {
      "epoch": 0.04967350746268657,
      "grad_norm": 2.410694122314453,
      "learning_rate": 3.9798518578696146e-05,
      "loss": 1.9434,
      "step": 213
    },
    {
      "epoch": 0.049906716417910446,
      "grad_norm": 2.56752347946167,
      "learning_rate": 3.979642883761866e-05,
      "loss": 1.9068,
      "step": 214
    },
    {
      "epoch": 0.05013992537313433,
      "grad_norm": 2.2785165309906006,
      "learning_rate": 3.979432837054204e-05,
      "loss": 2.1213,
      "step": 215
    },
    {
      "epoch": 0.05037313432835821,
      "grad_norm": 3.460172653198242,
      "learning_rate": 3.979221717860436e-05,
      "loss": 2.0173,
      "step": 216
    },
    {
      "epoch": 0.050606343283582086,
      "grad_norm": 2.3723466396331787,
      "learning_rate": 3.9790095262949496e-05,
      "loss": 1.9182,
      "step": 217
    },
    {
      "epoch": 0.05083955223880597,
      "grad_norm": 2.329787015914917,
      "learning_rate": 3.9787962624727126e-05,
      "loss": 1.817,
      "step": 218
    },
    {
      "epoch": 0.05107276119402985,
      "grad_norm": 2.4414517879486084,
      "learning_rate": 3.9785819265092744e-05,
      "loss": 1.9061,
      "step": 219
    },
    {
      "epoch": 0.051305970149253734,
      "grad_norm": 2.1740472316741943,
      "learning_rate": 3.978366518520766e-05,
      "loss": 1.9097,
      "step": 220
    },
    {
      "epoch": 0.05153917910447761,
      "grad_norm": 2.4254770278930664,
      "learning_rate": 3.978150038623898e-05,
      "loss": 1.8126,
      "step": 221
    },
    {
      "epoch": 0.05177238805970149,
      "grad_norm": 4.020961284637451,
      "learning_rate": 3.9779324869359624e-05,
      "loss": 2.0289,
      "step": 222
    },
    {
      "epoch": 0.052005597014925374,
      "grad_norm": 3.6008048057556152,
      "learning_rate": 3.9777138635748323e-05,
      "loss": 1.928,
      "step": 223
    },
    {
      "epoch": 0.05223880597014925,
      "grad_norm": 2.236170768737793,
      "learning_rate": 3.9774941686589616e-05,
      "loss": 1.8885,
      "step": 224
    },
    {
      "epoch": 0.05247201492537314,
      "grad_norm": 13.37221908569336,
      "learning_rate": 3.9772734023073834e-05,
      "loss": 2.04,
      "step": 225
    },
    {
      "epoch": 0.052705223880597014,
      "grad_norm": 2.5587236881256104,
      "learning_rate": 3.977051564639712e-05,
      "loss": 1.9722,
      "step": 226
    },
    {
      "epoch": 0.05293843283582089,
      "grad_norm": 2.701770782470703,
      "learning_rate": 3.976828655776144e-05,
      "loss": 1.8886,
      "step": 227
    },
    {
      "epoch": 0.05317164179104478,
      "grad_norm": 4.411864280700684,
      "learning_rate": 3.9766046758374535e-05,
      "loss": 1.9268,
      "step": 228
    },
    {
      "epoch": 0.053404850746268655,
      "grad_norm": 2.2130813598632812,
      "learning_rate": 3.976379624944996e-05,
      "loss": 1.9045,
      "step": 229
    },
    {
      "epoch": 0.05363805970149254,
      "grad_norm": 2.2269232273101807,
      "learning_rate": 3.976153503220708e-05,
      "loss": 1.9103,
      "step": 230
    },
    {
      "epoch": 0.05387126865671642,
      "grad_norm": 2.3306562900543213,
      "learning_rate": 3.9759263107871045e-05,
      "loss": 1.7871,
      "step": 231
    },
    {
      "epoch": 0.054104477611940295,
      "grad_norm": 2.066474676132202,
      "learning_rate": 3.975698047767283e-05,
      "loss": 1.8891,
      "step": 232
    },
    {
      "epoch": 0.05433768656716418,
      "grad_norm": 2.4459681510925293,
      "learning_rate": 3.975468714284919e-05,
      "loss": 1.8943,
      "step": 233
    },
    {
      "epoch": 0.05457089552238806,
      "grad_norm": 2.5987977981567383,
      "learning_rate": 3.97523831046427e-05,
      "loss": 1.8035,
      "step": 234
    },
    {
      "epoch": 0.05480410447761194,
      "grad_norm": 2.882322072982788,
      "learning_rate": 3.9750068364301705e-05,
      "loss": 2.0535,
      "step": 235
    },
    {
      "epoch": 0.05503731343283582,
      "grad_norm": 1.9460241794586182,
      "learning_rate": 3.974774292308037e-05,
      "loss": 1.7767,
      "step": 236
    },
    {
      "epoch": 0.055270522388059705,
      "grad_norm": 2.2095730304718018,
      "learning_rate": 3.9745406782238664e-05,
      "loss": 1.9401,
      "step": 237
    },
    {
      "epoch": 0.05550373134328358,
      "grad_norm": 2.2746217250823975,
      "learning_rate": 3.974305994304233e-05,
      "loss": 1.8318,
      "step": 238
    },
    {
      "epoch": 0.05573694029850746,
      "grad_norm": 2.2539775371551514,
      "learning_rate": 3.974070240676294e-05,
      "loss": 1.8619,
      "step": 239
    },
    {
      "epoch": 0.055970149253731345,
      "grad_norm": 2.116060733795166,
      "learning_rate": 3.9738334174677816e-05,
      "loss": 1.7938,
      "step": 240
    },
    {
      "epoch": 0.05620335820895522,
      "grad_norm": 1.9295035600662231,
      "learning_rate": 3.9735955248070116e-05,
      "loss": 1.8597,
      "step": 241
    },
    {
      "epoch": 0.05643656716417911,
      "grad_norm": 2.2538259029388428,
      "learning_rate": 3.9733565628228785e-05,
      "loss": 2.0752,
      "step": 242
    },
    {
      "epoch": 0.056669776119402986,
      "grad_norm": 1.9746248722076416,
      "learning_rate": 3.9731165316448535e-05,
      "loss": 1.8209,
      "step": 243
    },
    {
      "epoch": 0.05690298507462686,
      "grad_norm": 1.9035216569900513,
      "learning_rate": 3.97287543140299e-05,
      "loss": 1.7395,
      "step": 244
    },
    {
      "epoch": 0.05713619402985075,
      "grad_norm": 1.9367563724517822,
      "learning_rate": 3.9726332622279206e-05,
      "loss": 1.8738,
      "step": 245
    },
    {
      "epoch": 0.057369402985074626,
      "grad_norm": 2.2633564472198486,
      "learning_rate": 3.972390024250855e-05,
      "loss": 1.7889,
      "step": 246
    },
    {
      "epoch": 0.05760261194029851,
      "grad_norm": 2.0598018169403076,
      "learning_rate": 3.972145717603583e-05,
      "loss": 1.8099,
      "step": 247
    },
    {
      "epoch": 0.05783582089552239,
      "grad_norm": 2.243903398513794,
      "learning_rate": 3.971900342418475e-05,
      "loss": 1.9669,
      "step": 248
    },
    {
      "epoch": 0.058069029850746266,
      "grad_norm": 2.4976234436035156,
      "learning_rate": 3.9716538988284774e-05,
      "loss": 1.8317,
      "step": 249
    },
    {
      "epoch": 0.05830223880597015,
      "grad_norm": 2.58784818649292,
      "learning_rate": 3.9714063869671175e-05,
      "loss": 1.9311,
      "step": 250
    },
    {
      "epoch": 0.05853544776119403,
      "grad_norm": 1.965406060218811,
      "learning_rate": 3.971157806968501e-05,
      "loss": 1.9202,
      "step": 251
    },
    {
      "epoch": 0.058768656716417914,
      "grad_norm": 2.051205635070801,
      "learning_rate": 3.9709081589673123e-05,
      "loss": 1.8593,
      "step": 252
    },
    {
      "epoch": 0.05900186567164179,
      "grad_norm": 2.1563780307769775,
      "learning_rate": 3.9706574430988144e-05,
      "loss": 1.8526,
      "step": 253
    },
    {
      "epoch": 0.05923507462686567,
      "grad_norm": 2.34407114982605,
      "learning_rate": 3.970405659498848e-05,
      "loss": 2.0071,
      "step": 254
    },
    {
      "epoch": 0.059468283582089554,
      "grad_norm": 19.88642692565918,
      "learning_rate": 3.970152808303835e-05,
      "loss": 1.986,
      "step": 255
    },
    {
      "epoch": 0.05970149253731343,
      "grad_norm": 1.892634391784668,
      "learning_rate": 3.969898889650773e-05,
      "loss": 1.8909,
      "step": 256
    },
    {
      "epoch": 0.059934701492537316,
      "grad_norm": 2.4715380668640137,
      "learning_rate": 3.969643903677238e-05,
      "loss": 1.8654,
      "step": 257
    },
    {
      "epoch": 0.060167910447761194,
      "grad_norm": 2.0501859188079834,
      "learning_rate": 3.969387850521386e-05,
      "loss": 1.9859,
      "step": 258
    },
    {
      "epoch": 0.06040111940298507,
      "grad_norm": 2.007378101348877,
      "learning_rate": 3.969130730321951e-05,
      "loss": 1.9306,
      "step": 259
    },
    {
      "epoch": 0.06063432835820896,
      "grad_norm": 1.8277643918991089,
      "learning_rate": 3.9688725432182434e-05,
      "loss": 1.8319,
      "step": 260
    },
    {
      "epoch": 0.060867537313432835,
      "grad_norm": 1.9453307390213013,
      "learning_rate": 3.9686132893501534e-05,
      "loss": 1.8207,
      "step": 261
    },
    {
      "epoch": 0.06110074626865672,
      "grad_norm": 1.9022923707962036,
      "learning_rate": 3.968352968858149e-05,
      "loss": 1.7967,
      "step": 262
    },
    {
      "epoch": 0.0613339552238806,
      "grad_norm": 2.0632996559143066,
      "learning_rate": 3.968091581883274e-05,
      "loss": 1.9198,
      "step": 263
    },
    {
      "epoch": 0.061567164179104475,
      "grad_norm": 1.8707062005996704,
      "learning_rate": 3.967829128567155e-05,
      "loss": 1.9269,
      "step": 264
    },
    {
      "epoch": 0.06180037313432836,
      "grad_norm": 2.633587121963501,
      "learning_rate": 3.9675656090519894e-05,
      "loss": 1.9065,
      "step": 265
    },
    {
      "epoch": 0.06203358208955224,
      "grad_norm": 3.348646402359009,
      "learning_rate": 3.967301023480559e-05,
      "loss": 2.0197,
      "step": 266
    },
    {
      "epoch": 0.06226679104477612,
      "grad_norm": 6.585485935211182,
      "learning_rate": 3.967035371996218e-05,
      "loss": 2.0851,
      "step": 267
    },
    {
      "epoch": 0.0625,
      "grad_norm": 2.538081169128418,
      "learning_rate": 3.966768654742901e-05,
      "loss": 2.0437,
      "step": 268
    },
    {
      "epoch": 0.06273320895522388,
      "grad_norm": 7.334850311279297,
      "learning_rate": 3.96650087186512e-05,
      "loss": 1.9436,
      "step": 269
    },
    {
      "epoch": 0.06296641791044776,
      "grad_norm": 2.735170602798462,
      "learning_rate": 3.966232023507964e-05,
      "loss": 1.9254,
      "step": 270
    },
    {
      "epoch": 0.06319962686567164,
      "grad_norm": 2.1415090560913086,
      "learning_rate": 3.965962109817097e-05,
      "loss": 1.8887,
      "step": 271
    },
    {
      "epoch": 0.06343283582089553,
      "grad_norm": 2.2606241703033447,
      "learning_rate": 3.965691130938764e-05,
      "loss": 1.667,
      "step": 272
    },
    {
      "epoch": 0.06366604477611941,
      "grad_norm": 2.0732829570770264,
      "learning_rate": 3.965419087019785e-05,
      "loss": 1.9087,
      "step": 273
    },
    {
      "epoch": 0.06389925373134328,
      "grad_norm": 1.9969362020492554,
      "learning_rate": 3.9651459782075575e-05,
      "loss": 1.8597,
      "step": 274
    },
    {
      "epoch": 0.06413246268656717,
      "grad_norm": 2.24550199508667,
      "learning_rate": 3.964871804650056e-05,
      "loss": 1.9181,
      "step": 275
    },
    {
      "epoch": 0.06436567164179105,
      "grad_norm": 5.610985279083252,
      "learning_rate": 3.964596566495831e-05,
      "loss": 1.9888,
      "step": 276
    },
    {
      "epoch": 0.06459888059701492,
      "grad_norm": 2.051719903945923,
      "learning_rate": 3.964320263894012e-05,
      "loss": 1.8188,
      "step": 277
    },
    {
      "epoch": 0.0648320895522388,
      "grad_norm": 2.001046657562256,
      "learning_rate": 3.964042896994303e-05,
      "loss": 1.9464,
      "step": 278
    },
    {
      "epoch": 0.06506529850746269,
      "grad_norm": 2.2559609413146973,
      "learning_rate": 3.963764465946985e-05,
      "loss": 1.8127,
      "step": 279
    },
    {
      "epoch": 0.06529850746268656,
      "grad_norm": 1.9646382331848145,
      "learning_rate": 3.9634849709029165e-05,
      "loss": 1.9007,
      "step": 280
    },
    {
      "epoch": 0.06553171641791045,
      "grad_norm": 2.2789816856384277,
      "learning_rate": 3.963204412013533e-05,
      "loss": 1.9662,
      "step": 281
    },
    {
      "epoch": 0.06576492537313433,
      "grad_norm": 2.1956863403320312,
      "learning_rate": 3.962922789430844e-05,
      "loss": 1.9932,
      "step": 282
    },
    {
      "epoch": 0.06599813432835822,
      "grad_norm": 2.1003928184509277,
      "learning_rate": 3.962640103307439e-05,
      "loss": 1.7965,
      "step": 283
    },
    {
      "epoch": 0.06623134328358209,
      "grad_norm": 1.9648340940475464,
      "learning_rate": 3.9623563537964784e-05,
      "loss": 1.8397,
      "step": 284
    },
    {
      "epoch": 0.06646455223880597,
      "grad_norm": 1.8762441873550415,
      "learning_rate": 3.962071541051704e-05,
      "loss": 1.9145,
      "step": 285
    },
    {
      "epoch": 0.06669776119402986,
      "grad_norm": 1.9080212116241455,
      "learning_rate": 3.9617856652274315e-05,
      "loss": 1.9347,
      "step": 286
    },
    {
      "epoch": 0.06693097014925373,
      "grad_norm": 2.0662126541137695,
      "learning_rate": 3.961498726478552e-05,
      "loss": 1.9161,
      "step": 287
    },
    {
      "epoch": 0.06716417910447761,
      "grad_norm": 1.8274857997894287,
      "learning_rate": 3.9612107249605335e-05,
      "loss": 1.8803,
      "step": 288
    },
    {
      "epoch": 0.0673973880597015,
      "grad_norm": 2.2975552082061768,
      "learning_rate": 3.960921660829419e-05,
      "loss": 1.8408,
      "step": 289
    },
    {
      "epoch": 0.06763059701492537,
      "grad_norm": 23.154972076416016,
      "learning_rate": 3.960631534241829e-05,
      "loss": 2.3806,
      "step": 290
    },
    {
      "epoch": 0.06786380597014925,
      "grad_norm": 1.8874152898788452,
      "learning_rate": 3.960340345354957e-05,
      "loss": 1.799,
      "step": 291
    },
    {
      "epoch": 0.06809701492537314,
      "grad_norm": 2.8265576362609863,
      "learning_rate": 3.960048094326574e-05,
      "loss": 2.0515,
      "step": 292
    },
    {
      "epoch": 0.06833022388059702,
      "grad_norm": 2.8480818271636963,
      "learning_rate": 3.9597547813150264e-05,
      "loss": 2.0439,
      "step": 293
    },
    {
      "epoch": 0.06856343283582089,
      "grad_norm": 1.9650859832763672,
      "learning_rate": 3.959460406479235e-05,
      "loss": 1.8352,
      "step": 294
    },
    {
      "epoch": 0.06879664179104478,
      "grad_norm": 1.9850744009017944,
      "learning_rate": 3.9591649699786965e-05,
      "loss": 1.8353,
      "step": 295
    },
    {
      "epoch": 0.06902985074626866,
      "grad_norm": 2.029048204421997,
      "learning_rate": 3.9588684719734824e-05,
      "loss": 1.9234,
      "step": 296
    },
    {
      "epoch": 0.06926305970149253,
      "grad_norm": 2.312969207763672,
      "learning_rate": 3.958570912624241e-05,
      "loss": 1.7853,
      "step": 297
    },
    {
      "epoch": 0.06949626865671642,
      "grad_norm": 2.661008834838867,
      "learning_rate": 3.958272292092193e-05,
      "loss": 1.9652,
      "step": 298
    },
    {
      "epoch": 0.0697294776119403,
      "grad_norm": 2.3096468448638916,
      "learning_rate": 3.9579726105391364e-05,
      "loss": 2.0013,
      "step": 299
    },
    {
      "epoch": 0.06996268656716417,
      "grad_norm": 1.9524552822113037,
      "learning_rate": 3.9576718681274426e-05,
      "loss": 1.834,
      "step": 300
    },
    {
      "epoch": 0.07019589552238806,
      "grad_norm": 2.630711793899536,
      "learning_rate": 3.957370065020059e-05,
      "loss": 1.8423,
      "step": 301
    },
    {
      "epoch": 0.07042910447761194,
      "grad_norm": 1.9605114459991455,
      "learning_rate": 3.957067201380506e-05,
      "loss": 1.8462,
      "step": 302
    },
    {
      "epoch": 0.07066231343283583,
      "grad_norm": 2.745546817779541,
      "learning_rate": 3.9567632773728796e-05,
      "loss": 1.6943,
      "step": 303
    },
    {
      "epoch": 0.0708955223880597,
      "grad_norm": 4.500966548919678,
      "learning_rate": 3.956458293161851e-05,
      "loss": 2.0578,
      "step": 304
    },
    {
      "epoch": 0.07112873134328358,
      "grad_norm": 3.483400583267212,
      "learning_rate": 3.9561522489126654e-05,
      "loss": 1.9156,
      "step": 305
    },
    {
      "epoch": 0.07136194029850747,
      "grad_norm": 1.9230903387069702,
      "learning_rate": 3.955845144791142e-05,
      "loss": 1.7679,
      "step": 306
    },
    {
      "epoch": 0.07159514925373134,
      "grad_norm": 2.5835723876953125,
      "learning_rate": 3.955536980963673e-05,
      "loss": 1.8439,
      "step": 307
    },
    {
      "epoch": 0.07182835820895522,
      "grad_norm": 1.9240423440933228,
      "learning_rate": 3.9552277575972275e-05,
      "loss": 1.7954,
      "step": 308
    },
    {
      "epoch": 0.07206156716417911,
      "grad_norm": 1.8923357725143433,
      "learning_rate": 3.9549174748593475e-05,
      "loss": 1.789,
      "step": 309
    },
    {
      "epoch": 0.07229477611940298,
      "grad_norm": 6.578820705413818,
      "learning_rate": 3.954606132918148e-05,
      "loss": 1.974,
      "step": 310
    },
    {
      "epoch": 0.07252798507462686,
      "grad_norm": 1.958921194076538,
      "learning_rate": 3.954293731942319e-05,
      "loss": 1.8797,
      "step": 311
    },
    {
      "epoch": 0.07276119402985075,
      "grad_norm": 2.3170156478881836,
      "learning_rate": 3.9539802721011246e-05,
      "loss": 1.945,
      "step": 312
    },
    {
      "epoch": 0.07299440298507463,
      "grad_norm": 3.415703535079956,
      "learning_rate": 3.9536657535644004e-05,
      "loss": 2.0047,
      "step": 313
    },
    {
      "epoch": 0.0732276119402985,
      "grad_norm": 1.8777929544448853,
      "learning_rate": 3.953350176502559e-05,
      "loss": 1.8979,
      "step": 314
    },
    {
      "epoch": 0.07346082089552239,
      "grad_norm": 1.8773884773254395,
      "learning_rate": 3.953033541086583e-05,
      "loss": 2.0219,
      "step": 315
    },
    {
      "epoch": 0.07369402985074627,
      "grad_norm": 1.8622405529022217,
      "learning_rate": 3.952715847488032e-05,
      "loss": 1.933,
      "step": 316
    },
    {
      "epoch": 0.07392723880597014,
      "grad_norm": 1.7881577014923096,
      "learning_rate": 3.952397095879036e-05,
      "loss": 1.8632,
      "step": 317
    },
    {
      "epoch": 0.07416044776119403,
      "grad_norm": 1.768904209136963,
      "learning_rate": 3.9520772864323e-05,
      "loss": 1.8257,
      "step": 318
    },
    {
      "epoch": 0.07439365671641791,
      "grad_norm": 1.7308861017227173,
      "learning_rate": 3.9517564193211004e-05,
      "loss": 1.8176,
      "step": 319
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 1.817916989326477,
      "learning_rate": 3.951434494719289e-05,
      "loss": 1.8935,
      "step": 320
    },
    {
      "epoch": 0.07486007462686567,
      "grad_norm": 1.8161803483963013,
      "learning_rate": 3.951111512801288e-05,
      "loss": 1.8158,
      "step": 321
    },
    {
      "epoch": 0.07509328358208955,
      "grad_norm": 1.7080978155136108,
      "learning_rate": 3.950787473742095e-05,
      "loss": 1.7454,
      "step": 322
    },
    {
      "epoch": 0.07532649253731344,
      "grad_norm": 2.499462842941284,
      "learning_rate": 3.950462377717279e-05,
      "loss": 2.0163,
      "step": 323
    },
    {
      "epoch": 0.07555970149253731,
      "grad_norm": 1.9035282135009766,
      "learning_rate": 3.950136224902982e-05,
      "loss": 1.7977,
      "step": 324
    },
    {
      "epoch": 0.0757929104477612,
      "grad_norm": 1.6562100648880005,
      "learning_rate": 3.949809015475917e-05,
      "loss": 1.7032,
      "step": 325
    },
    {
      "epoch": 0.07602611940298508,
      "grad_norm": 1.8899322748184204,
      "learning_rate": 3.949480749613372e-05,
      "loss": 1.7767,
      "step": 326
    },
    {
      "epoch": 0.07625932835820895,
      "grad_norm": 2.7696213722229004,
      "learning_rate": 3.949151427493207e-05,
      "loss": 1.9997,
      "step": 327
    },
    {
      "epoch": 0.07649253731343283,
      "grad_norm": 1.9022033214569092,
      "learning_rate": 3.948821049293853e-05,
      "loss": 1.8976,
      "step": 328
    },
    {
      "epoch": 0.07672574626865672,
      "grad_norm": 2.007911443710327,
      "learning_rate": 3.948489615194313e-05,
      "loss": 1.8529,
      "step": 329
    },
    {
      "epoch": 0.07695895522388059,
      "grad_norm": 1.833503007888794,
      "learning_rate": 3.948157125374164e-05,
      "loss": 1.9123,
      "step": 330
    },
    {
      "epoch": 0.07719216417910447,
      "grad_norm": 2.152531623840332,
      "learning_rate": 3.947823580013553e-05,
      "loss": 1.9037,
      "step": 331
    },
    {
      "epoch": 0.07742537313432836,
      "grad_norm": 1.8107256889343262,
      "learning_rate": 3.947488979293201e-05,
      "loss": 1.92,
      "step": 332
    },
    {
      "epoch": 0.07765858208955224,
      "grad_norm": 1.9272944927215576,
      "learning_rate": 3.947153323394399e-05,
      "loss": 1.9718,
      "step": 333
    },
    {
      "epoch": 0.07789179104477612,
      "grad_norm": 2.0863277912139893,
      "learning_rate": 3.94681661249901e-05,
      "loss": 1.8168,
      "step": 334
    },
    {
      "epoch": 0.078125,
      "grad_norm": 1.8380833864212036,
      "learning_rate": 3.94647884678947e-05,
      "loss": 1.8322,
      "step": 335
    },
    {
      "epoch": 0.07835820895522388,
      "grad_norm": 3.2511706352233887,
      "learning_rate": 3.946140026448785e-05,
      "loss": 1.9787,
      "step": 336
    },
    {
      "epoch": 0.07859141791044776,
      "grad_norm": 1.7913483381271362,
      "learning_rate": 3.945800151660532e-05,
      "loss": 1.8353,
      "step": 337
    },
    {
      "epoch": 0.07882462686567164,
      "grad_norm": 1.7866766452789307,
      "learning_rate": 3.945459222608861e-05,
      "loss": 1.815,
      "step": 338
    },
    {
      "epoch": 0.07905783582089553,
      "grad_norm": 5.889623641967773,
      "learning_rate": 3.945117239478492e-05,
      "loss": 2.0594,
      "step": 339
    },
    {
      "epoch": 0.07929104477611941,
      "grad_norm": 2.011286735534668,
      "learning_rate": 3.944774202454718e-05,
      "loss": 1.9879,
      "step": 340
    },
    {
      "epoch": 0.07952425373134328,
      "grad_norm": 2.1573173999786377,
      "learning_rate": 3.9444301117233995e-05,
      "loss": 2.0311,
      "step": 341
    },
    {
      "epoch": 0.07975746268656717,
      "grad_norm": 1.7714496850967407,
      "learning_rate": 3.944084967470971e-05,
      "loss": 1.877,
      "step": 342
    },
    {
      "epoch": 0.07999067164179105,
      "grad_norm": 1.8621400594711304,
      "learning_rate": 3.943738769884437e-05,
      "loss": 1.9594,
      "step": 343
    },
    {
      "epoch": 0.08022388059701492,
      "grad_norm": 1.8687459230422974,
      "learning_rate": 3.943391519151372e-05,
      "loss": 1.8247,
      "step": 344
    },
    {
      "epoch": 0.0804570895522388,
      "grad_norm": 1.7938812971115112,
      "learning_rate": 3.9430432154599215e-05,
      "loss": 1.7671,
      "step": 345
    },
    {
      "epoch": 0.08069029850746269,
      "grad_norm": 1.7832001447677612,
      "learning_rate": 3.9426938589988024e-05,
      "loss": 1.8371,
      "step": 346
    },
    {
      "epoch": 0.08092350746268656,
      "grad_norm": 1.8771873712539673,
      "learning_rate": 3.9423434499573004e-05,
      "loss": 1.8227,
      "step": 347
    },
    {
      "epoch": 0.08115671641791045,
      "grad_norm": 1.7765568494796753,
      "learning_rate": 3.941991988525273e-05,
      "loss": 1.7482,
      "step": 348
    },
    {
      "epoch": 0.08138992537313433,
      "grad_norm": 1.8241585493087769,
      "learning_rate": 3.941639474893147e-05,
      "loss": 1.8088,
      "step": 349
    },
    {
      "epoch": 0.08162313432835822,
      "grad_norm": 1.8858530521392822,
      "learning_rate": 3.9412859092519184e-05,
      "loss": 1.9021,
      "step": 350
    },
    {
      "epoch": 0.08185634328358209,
      "grad_norm": 2.71389102935791,
      "learning_rate": 3.940931291793156e-05,
      "loss": 1.9762,
      "step": 351
    },
    {
      "epoch": 0.08208955223880597,
      "grad_norm": 1.7988463640213013,
      "learning_rate": 3.940575622708996e-05,
      "loss": 1.8389,
      "step": 352
    },
    {
      "epoch": 0.08232276119402986,
      "grad_norm": 6.051729202270508,
      "learning_rate": 3.9402189021921453e-05,
      "loss": 1.9242,
      "step": 353
    },
    {
      "epoch": 0.08255597014925373,
      "grad_norm": 2.4854869842529297,
      "learning_rate": 3.939861130435881e-05,
      "loss": 1.9259,
      "step": 354
    },
    {
      "epoch": 0.08278917910447761,
      "grad_norm": 3.8117482662200928,
      "learning_rate": 3.939502307634047e-05,
      "loss": 1.9278,
      "step": 355
    },
    {
      "epoch": 0.0830223880597015,
      "grad_norm": 2.6823151111602783,
      "learning_rate": 3.9391424339810614e-05,
      "loss": 1.8894,
      "step": 356
    },
    {
      "epoch": 0.08325559701492537,
      "grad_norm": 1.8398593664169312,
      "learning_rate": 3.9387815096719084e-05,
      "loss": 1.8438,
      "step": 357
    },
    {
      "epoch": 0.08348880597014925,
      "grad_norm": 1.9887045621871948,
      "learning_rate": 3.938419534902141e-05,
      "loss": 1.8411,
      "step": 358
    },
    {
      "epoch": 0.08372201492537314,
      "grad_norm": 1.9320147037506104,
      "learning_rate": 3.9380565098678835e-05,
      "loss": 1.8433,
      "step": 359
    },
    {
      "epoch": 0.08395522388059702,
      "grad_norm": 2.1372487545013428,
      "learning_rate": 3.937692434765828e-05,
      "loss": 1.8415,
      "step": 360
    },
    {
      "epoch": 0.08418843283582089,
      "grad_norm": 3.710360527038574,
      "learning_rate": 3.9373273097932354e-05,
      "loss": 1.7969,
      "step": 361
    },
    {
      "epoch": 0.08442164179104478,
      "grad_norm": 3.8473570346832275,
      "learning_rate": 3.936961135147937e-05,
      "loss": 1.9017,
      "step": 362
    },
    {
      "epoch": 0.08465485074626866,
      "grad_norm": 2.6215438842773438,
      "learning_rate": 3.9365939110283306e-05,
      "loss": 1.8847,
      "step": 363
    },
    {
      "epoch": 0.08488805970149253,
      "grad_norm": 2.3148863315582275,
      "learning_rate": 3.9362256376333846e-05,
      "loss": 1.8978,
      "step": 364
    },
    {
      "epoch": 0.08512126865671642,
      "grad_norm": 3.2686333656311035,
      "learning_rate": 3.9358563151626336e-05,
      "loss": 1.9277,
      "step": 365
    },
    {
      "epoch": 0.0853544776119403,
      "grad_norm": 1.895060420036316,
      "learning_rate": 3.9354859438161835e-05,
      "loss": 1.8471,
      "step": 366
    },
    {
      "epoch": 0.08558768656716417,
      "grad_norm": 2.4638521671295166,
      "learning_rate": 3.935114523794706e-05,
      "loss": 1.7614,
      "step": 367
    },
    {
      "epoch": 0.08582089552238806,
      "grad_norm": 1.8638423681259155,
      "learning_rate": 3.9347420552994415e-05,
      "loss": 1.9506,
      "step": 368
    },
    {
      "epoch": 0.08605410447761194,
      "grad_norm": 2.3636820316314697,
      "learning_rate": 3.934368538532201e-05,
      "loss": 1.8791,
      "step": 369
    },
    {
      "epoch": 0.08628731343283583,
      "grad_norm": 6.352964878082275,
      "learning_rate": 3.9339939736953595e-05,
      "loss": 1.9904,
      "step": 370
    },
    {
      "epoch": 0.0865205223880597,
      "grad_norm": 2.037658214569092,
      "learning_rate": 3.933618360991863e-05,
      "loss": 1.8709,
      "step": 371
    },
    {
      "epoch": 0.08675373134328358,
      "grad_norm": 1.7510515451431274,
      "learning_rate": 3.933241700625223e-05,
      "loss": 1.7455,
      "step": 372
    },
    {
      "epoch": 0.08698694029850747,
      "grad_norm": 1.7622164487838745,
      "learning_rate": 3.93286399279952e-05,
      "loss": 1.8339,
      "step": 373
    },
    {
      "epoch": 0.08722014925373134,
      "grad_norm": 3.107625722885132,
      "learning_rate": 3.9324852377194026e-05,
      "loss": 1.7994,
      "step": 374
    },
    {
      "epoch": 0.08745335820895522,
      "grad_norm": 2.1422181129455566,
      "learning_rate": 3.932105435590084e-05,
      "loss": 1.9355,
      "step": 375
    },
    {
      "epoch": 0.08768656716417911,
      "grad_norm": 2.186555862426758,
      "learning_rate": 3.931724586617349e-05,
      "loss": 1.9485,
      "step": 376
    },
    {
      "epoch": 0.08791977611940298,
      "grad_norm": 16.650178909301758,
      "learning_rate": 3.9313426910075455e-05,
      "loss": 1.9388,
      "step": 377
    },
    {
      "epoch": 0.08815298507462686,
      "grad_norm": 2.001224994659424,
      "learning_rate": 3.930959748967591e-05,
      "loss": 1.958,
      "step": 378
    },
    {
      "epoch": 0.08838619402985075,
      "grad_norm": 1.8503843545913696,
      "learning_rate": 3.93057576070497e-05,
      "loss": 1.9768,
      "step": 379
    },
    {
      "epoch": 0.08861940298507463,
      "grad_norm": 1.828757405281067,
      "learning_rate": 3.9301907264277305e-05,
      "loss": 1.7807,
      "step": 380
    },
    {
      "epoch": 0.0888526119402985,
      "grad_norm": 4.222869396209717,
      "learning_rate": 3.929804646344492e-05,
      "loss": 1.8592,
      "step": 381
    },
    {
      "epoch": 0.08908582089552239,
      "grad_norm": 4.99127721786499,
      "learning_rate": 3.929417520664438e-05,
      "loss": 1.865,
      "step": 382
    },
    {
      "epoch": 0.08931902985074627,
      "grad_norm": 2.110808849334717,
      "learning_rate": 3.929029349597318e-05,
      "loss": 1.9565,
      "step": 383
    },
    {
      "epoch": 0.08955223880597014,
      "grad_norm": 2.6213815212249756,
      "learning_rate": 3.92864013335345e-05,
      "loss": 1.8469,
      "step": 384
    },
    {
      "epoch": 0.08978544776119403,
      "grad_norm": 2.161807060241699,
      "learning_rate": 3.928249872143716e-05,
      "loss": 1.8154,
      "step": 385
    },
    {
      "epoch": 0.09001865671641791,
      "grad_norm": 1.9784412384033203,
      "learning_rate": 3.927858566179567e-05,
      "loss": 1.9335,
      "step": 386
    },
    {
      "epoch": 0.09025186567164178,
      "grad_norm": 3.250617742538452,
      "learning_rate": 3.927466215673017e-05,
      "loss": 1.99,
      "step": 387
    },
    {
      "epoch": 0.09048507462686567,
      "grad_norm": 3.5758609771728516,
      "learning_rate": 3.927072820836647e-05,
      "loss": 2.0221,
      "step": 388
    },
    {
      "epoch": 0.09071828358208955,
      "grad_norm": 1.728752851486206,
      "learning_rate": 3.9266783818836056e-05,
      "loss": 1.836,
      "step": 389
    },
    {
      "epoch": 0.09095149253731344,
      "grad_norm": 1.9743226766586304,
      "learning_rate": 3.9262828990276054e-05,
      "loss": 1.7504,
      "step": 390
    },
    {
      "epoch": 0.09118470149253731,
      "grad_norm": 1.8036150932312012,
      "learning_rate": 3.925886372482923e-05,
      "loss": 1.9098,
      "step": 391
    },
    {
      "epoch": 0.0914179104477612,
      "grad_norm": 1.6875373125076294,
      "learning_rate": 3.9254888024644054e-05,
      "loss": 1.8294,
      "step": 392
    },
    {
      "epoch": 0.09165111940298508,
      "grad_norm": 1.8700023889541626,
      "learning_rate": 3.92509018918746e-05,
      "loss": 1.8578,
      "step": 393
    },
    {
      "epoch": 0.09188432835820895,
      "grad_norm": 1.7668216228485107,
      "learning_rate": 3.924690532868061e-05,
      "loss": 1.9039,
      "step": 394
    },
    {
      "epoch": 0.09211753731343283,
      "grad_norm": 1.9266901016235352,
      "learning_rate": 3.924289833722749e-05,
      "loss": 1.8882,
      "step": 395
    },
    {
      "epoch": 0.09235074626865672,
      "grad_norm": 1.8543881177902222,
      "learning_rate": 3.923888091968629e-05,
      "loss": 1.9744,
      "step": 396
    },
    {
      "epoch": 0.09258395522388059,
      "grad_norm": 1.7179049253463745,
      "learning_rate": 3.92348530782337e-05,
      "loss": 1.8302,
      "step": 397
    },
    {
      "epoch": 0.09281716417910447,
      "grad_norm": 3.277365207672119,
      "learning_rate": 3.923081481505206e-05,
      "loss": 1.8717,
      "step": 398
    },
    {
      "epoch": 0.09305037313432836,
      "grad_norm": 1.743199348449707,
      "learning_rate": 3.9226766132329385e-05,
      "loss": 1.9405,
      "step": 399
    },
    {
      "epoch": 0.09328358208955224,
      "grad_norm": 1.809787631034851,
      "learning_rate": 3.922270703225929e-05,
      "loss": 1.9448,
      "step": 400
    },
    {
      "epoch": 0.09351679104477612,
      "grad_norm": 1.7982717752456665,
      "learning_rate": 3.9218637517041046e-05,
      "loss": 1.9609,
      "step": 401
    },
    {
      "epoch": 0.09375,
      "grad_norm": 1.6124346256256104,
      "learning_rate": 3.9214557588879604e-05,
      "loss": 1.8055,
      "step": 402
    },
    {
      "epoch": 0.09398320895522388,
      "grad_norm": 1.7059659957885742,
      "learning_rate": 3.921046724998551e-05,
      "loss": 1.8394,
      "step": 403
    },
    {
      "epoch": 0.09421641791044776,
      "grad_norm": 1.623435616493225,
      "learning_rate": 3.9206366502574976e-05,
      "loss": 1.8205,
      "step": 404
    },
    {
      "epoch": 0.09444962686567164,
      "grad_norm": 1.7674553394317627,
      "learning_rate": 3.920225534886986e-05,
      "loss": 1.8586,
      "step": 405
    },
    {
      "epoch": 0.09468283582089553,
      "grad_norm": 1.7021158933639526,
      "learning_rate": 3.919813379109762e-05,
      "loss": 1.8585,
      "step": 406
    },
    {
      "epoch": 0.09491604477611941,
      "grad_norm": 1.6854712963104248,
      "learning_rate": 3.9194001831491404e-05,
      "loss": 1.7138,
      "step": 407
    },
    {
      "epoch": 0.09514925373134328,
      "grad_norm": 1.8121768236160278,
      "learning_rate": 3.9189859472289956e-05,
      "loss": 1.8628,
      "step": 408
    },
    {
      "epoch": 0.09538246268656717,
      "grad_norm": 1.8775556087493896,
      "learning_rate": 3.918570671573766e-05,
      "loss": 1.8956,
      "step": 409
    },
    {
      "epoch": 0.09561567164179105,
      "grad_norm": 1.7699267864227295,
      "learning_rate": 3.9181543564084554e-05,
      "loss": 1.8609,
      "step": 410
    },
    {
      "epoch": 0.09584888059701492,
      "grad_norm": 1.7189266681671143,
      "learning_rate": 3.9177370019586295e-05,
      "loss": 1.9685,
      "step": 411
    },
    {
      "epoch": 0.0960820895522388,
      "grad_norm": 1.7291022539138794,
      "learning_rate": 3.917318608450416e-05,
      "loss": 1.9328,
      "step": 412
    },
    {
      "epoch": 0.09631529850746269,
      "grad_norm": 2.582749366760254,
      "learning_rate": 3.9168991761105086e-05,
      "loss": 1.8823,
      "step": 413
    },
    {
      "epoch": 0.09654850746268656,
      "grad_norm": 5.4985833168029785,
      "learning_rate": 3.91647870516616e-05,
      "loss": 1.8894,
      "step": 414
    },
    {
      "epoch": 0.09678171641791045,
      "grad_norm": 1.7926266193389893,
      "learning_rate": 3.9160571958451885e-05,
      "loss": 1.955,
      "step": 415
    },
    {
      "epoch": 0.09701492537313433,
      "grad_norm": 1.7279105186462402,
      "learning_rate": 3.915634648375974e-05,
      "loss": 1.9612,
      "step": 416
    },
    {
      "epoch": 0.09724813432835822,
      "grad_norm": 1.769933819770813,
      "learning_rate": 3.91521106298746e-05,
      "loss": 1.9393,
      "step": 417
    },
    {
      "epoch": 0.09748134328358209,
      "grad_norm": 1.9194711446762085,
      "learning_rate": 3.91478643990915e-05,
      "loss": 1.941,
      "step": 418
    },
    {
      "epoch": 0.09771455223880597,
      "grad_norm": 1.7428562641143799,
      "learning_rate": 3.9143607793711114e-05,
      "loss": 1.7563,
      "step": 419
    },
    {
      "epoch": 0.09794776119402986,
      "grad_norm": 1.826330304145813,
      "learning_rate": 3.913934081603973e-05,
      "loss": 1.7407,
      "step": 420
    },
    {
      "epoch": 0.09818097014925373,
      "grad_norm": 1.7170411348342896,
      "learning_rate": 3.913506346838927e-05,
      "loss": 1.8246,
      "step": 421
    },
    {
      "epoch": 0.09841417910447761,
      "grad_norm": 1.664249300956726,
      "learning_rate": 3.9130775753077254e-05,
      "loss": 1.8438,
      "step": 422
    },
    {
      "epoch": 0.0986473880597015,
      "grad_norm": 1.5838645696640015,
      "learning_rate": 3.9126477672426845e-05,
      "loss": 1.8022,
      "step": 423
    },
    {
      "epoch": 0.09888059701492537,
      "grad_norm": 1.9592511653900146,
      "learning_rate": 3.912216922876679e-05,
      "loss": 1.867,
      "step": 424
    },
    {
      "epoch": 0.09911380597014925,
      "grad_norm": 1.7943328619003296,
      "learning_rate": 3.911785042443148e-05,
      "loss": 1.8517,
      "step": 425
    },
    {
      "epoch": 0.09934701492537314,
      "grad_norm": 1.706250548362732,
      "learning_rate": 3.91135212617609e-05,
      "loss": 1.8178,
      "step": 426
    },
    {
      "epoch": 0.09958022388059702,
      "grad_norm": 1.7921274900436401,
      "learning_rate": 3.910918174310066e-05,
      "loss": 1.7517,
      "step": 427
    },
    {
      "epoch": 0.09981343283582089,
      "grad_norm": 1.630629062652588,
      "learning_rate": 3.910483187080196e-05,
      "loss": 1.7661,
      "step": 428
    },
    {
      "epoch": 0.10004664179104478,
      "grad_norm": 1.6972260475158691,
      "learning_rate": 3.910047164722165e-05,
      "loss": 1.9308,
      "step": 429
    },
    {
      "epoch": 0.10027985074626866,
      "grad_norm": 2.070817708969116,
      "learning_rate": 3.909610107472215e-05,
      "loss": 1.9332,
      "step": 430
    },
    {
      "epoch": 0.10051305970149253,
      "grad_norm": 1.921702265739441,
      "learning_rate": 3.90917201556715e-05,
      "loss": 2.0213,
      "step": 431
    },
    {
      "epoch": 0.10074626865671642,
      "grad_norm": 2.562431812286377,
      "learning_rate": 3.908732889244336e-05,
      "loss": 1.8974,
      "step": 432
    },
    {
      "epoch": 0.1009794776119403,
      "grad_norm": 1.7326966524124146,
      "learning_rate": 3.908292728741696e-05,
      "loss": 1.8752,
      "step": 433
    },
    {
      "epoch": 0.10121268656716417,
      "grad_norm": 1.9353653192520142,
      "learning_rate": 3.907851534297717e-05,
      "loss": 1.716,
      "step": 434
    },
    {
      "epoch": 0.10144589552238806,
      "grad_norm": 1.7853633165359497,
      "learning_rate": 3.9074093061514446e-05,
      "loss": 1.7676,
      "step": 435
    },
    {
      "epoch": 0.10167910447761194,
      "grad_norm": 2.518691062927246,
      "learning_rate": 3.906966044542484e-05,
      "loss": 1.7454,
      "step": 436
    },
    {
      "epoch": 0.10191231343283583,
      "grad_norm": 10.521029472351074,
      "learning_rate": 3.9065217497110016e-05,
      "loss": 2.1409,
      "step": 437
    },
    {
      "epoch": 0.1021455223880597,
      "grad_norm": 1.7788546085357666,
      "learning_rate": 3.906076421897722e-05,
      "loss": 1.8026,
      "step": 438
    },
    {
      "epoch": 0.10237873134328358,
      "grad_norm": 1.800396203994751,
      "learning_rate": 3.905630061343931e-05,
      "loss": 1.9245,
      "step": 439
    },
    {
      "epoch": 0.10261194029850747,
      "grad_norm": 2.207120180130005,
      "learning_rate": 3.905182668291474e-05,
      "loss": 1.893,
      "step": 440
    },
    {
      "epoch": 0.10284514925373134,
      "grad_norm": 1.800674557685852,
      "learning_rate": 3.904734242982754e-05,
      "loss": 1.8254,
      "step": 441
    },
    {
      "epoch": 0.10307835820895522,
      "grad_norm": 1.8203047513961792,
      "learning_rate": 3.904284785660737e-05,
      "loss": 1.7687,
      "step": 442
    },
    {
      "epoch": 0.10331156716417911,
      "grad_norm": 1.6651867628097534,
      "learning_rate": 3.903834296568941e-05,
      "loss": 1.7917,
      "step": 443
    },
    {
      "epoch": 0.10354477611940298,
      "grad_norm": 7.804349422454834,
      "learning_rate": 3.903382775951452e-05,
      "loss": 1.8551,
      "step": 444
    },
    {
      "epoch": 0.10377798507462686,
      "grad_norm": 1.8691987991333008,
      "learning_rate": 3.902930224052909e-05,
      "loss": 1.8034,
      "step": 445
    },
    {
      "epoch": 0.10401119402985075,
      "grad_norm": 2.092705488204956,
      "learning_rate": 3.9024766411185104e-05,
      "loss": 1.8435,
      "step": 446
    },
    {
      "epoch": 0.10424440298507463,
      "grad_norm": 2.1264102458953857,
      "learning_rate": 3.9020220273940156e-05,
      "loss": 1.8983,
      "step": 447
    },
    {
      "epoch": 0.1044776119402985,
      "grad_norm": 1.9079781770706177,
      "learning_rate": 3.901566383125741e-05,
      "loss": 1.8382,
      "step": 448
    },
    {
      "epoch": 0.10471082089552239,
      "grad_norm": 1.670935869216919,
      "learning_rate": 3.901109708560561e-05,
      "loss": 1.9516,
      "step": 449
    },
    {
      "epoch": 0.10494402985074627,
      "grad_norm": 1.6233859062194824,
      "learning_rate": 3.900652003945908e-05,
      "loss": 1.8913,
      "step": 450
    },
    {
      "epoch": 0.10517723880597014,
      "grad_norm": 1.7143909931182861,
      "learning_rate": 3.900193269529774e-05,
      "loss": 1.8275,
      "step": 451
    },
    {
      "epoch": 0.10541044776119403,
      "grad_norm": 2.0415287017822266,
      "learning_rate": 3.899733505560708e-05,
      "loss": 2.0339,
      "step": 452
    },
    {
      "epoch": 0.10564365671641791,
      "grad_norm": 1.6881966590881348,
      "learning_rate": 3.899272712287817e-05,
      "loss": 1.79,
      "step": 453
    },
    {
      "epoch": 0.10587686567164178,
      "grad_norm": 1.946357011795044,
      "learning_rate": 3.8988108899607654e-05,
      "loss": 1.8272,
      "step": 454
    },
    {
      "epoch": 0.10611007462686567,
      "grad_norm": 1.7551015615463257,
      "learning_rate": 3.8983480388297754e-05,
      "loss": 1.8929,
      "step": 455
    },
    {
      "epoch": 0.10634328358208955,
      "grad_norm": 1.6512070894241333,
      "learning_rate": 3.897884159145627e-05,
      "loss": 1.8676,
      "step": 456
    },
    {
      "epoch": 0.10657649253731344,
      "grad_norm": 1.7040420770645142,
      "learning_rate": 3.897419251159657e-05,
      "loss": 1.8826,
      "step": 457
    },
    {
      "epoch": 0.10680970149253731,
      "grad_norm": 1.7353508472442627,
      "learning_rate": 3.8969533151237594e-05,
      "loss": 1.6552,
      "step": 458
    },
    {
      "epoch": 0.1070429104477612,
      "grad_norm": 1.8035770654678345,
      "learning_rate": 3.896486351290384e-05,
      "loss": 1.8754,
      "step": 459
    },
    {
      "epoch": 0.10727611940298508,
      "grad_norm": 2.5990219116210938,
      "learning_rate": 3.896018359912541e-05,
      "loss": 2.0031,
      "step": 460
    },
    {
      "epoch": 0.10750932835820895,
      "grad_norm": 1.9087543487548828,
      "learning_rate": 3.895549341243794e-05,
      "loss": 1.8966,
      "step": 461
    },
    {
      "epoch": 0.10774253731343283,
      "grad_norm": 1.8487074375152588,
      "learning_rate": 3.895079295538264e-05,
      "loss": 2.0128,
      "step": 462
    },
    {
      "epoch": 0.10797574626865672,
      "grad_norm": 2.5845704078674316,
      "learning_rate": 3.894608223050629e-05,
      "loss": 1.9876,
      "step": 463
    },
    {
      "epoch": 0.10820895522388059,
      "grad_norm": 1.8744717836380005,
      "learning_rate": 3.8941361240361224e-05,
      "loss": 1.6917,
      "step": 464
    },
    {
      "epoch": 0.10844216417910447,
      "grad_norm": 1.7186557054519653,
      "learning_rate": 3.893662998750536e-05,
      "loss": 1.8475,
      "step": 465
    },
    {
      "epoch": 0.10867537313432836,
      "grad_norm": 1.842673897743225,
      "learning_rate": 3.8931888474502144e-05,
      "loss": 1.7338,
      "step": 466
    },
    {
      "epoch": 0.10890858208955224,
      "grad_norm": 2.291309118270874,
      "learning_rate": 3.892713670392062e-05,
      "loss": 1.9961,
      "step": 467
    },
    {
      "epoch": 0.10914179104477612,
      "grad_norm": 1.7876993417739868,
      "learning_rate": 3.8922374678335346e-05,
      "loss": 1.8137,
      "step": 468
    },
    {
      "epoch": 0.109375,
      "grad_norm": 25.753190994262695,
      "learning_rate": 3.891760240032646e-05,
      "loss": 1.8457,
      "step": 469
    },
    {
      "epoch": 0.10960820895522388,
      "grad_norm": 1.859543800354004,
      "learning_rate": 3.891281987247966e-05,
      "loss": 1.9267,
      "step": 470
    },
    {
      "epoch": 0.10984141791044776,
      "grad_norm": 1.7328011989593506,
      "learning_rate": 3.8908027097386205e-05,
      "loss": 1.8454,
      "step": 471
    },
    {
      "epoch": 0.11007462686567164,
      "grad_norm": 1.6649116277694702,
      "learning_rate": 3.890322407764287e-05,
      "loss": 1.8941,
      "step": 472
    },
    {
      "epoch": 0.11030783582089553,
      "grad_norm": 1.7180585861206055,
      "learning_rate": 3.8898410815852005e-05,
      "loss": 1.8905,
      "step": 473
    },
    {
      "epoch": 0.11054104477611941,
      "grad_norm": 1.7984118461608887,
      "learning_rate": 3.889358731462151e-05,
      "loss": 1.9018,
      "step": 474
    },
    {
      "epoch": 0.11077425373134328,
      "grad_norm": 1.6825501918792725,
      "learning_rate": 3.888875357656484e-05,
      "loss": 1.8236,
      "step": 475
    },
    {
      "epoch": 0.11100746268656717,
      "grad_norm": 1.8572235107421875,
      "learning_rate": 3.888390960430097e-05,
      "loss": 1.9564,
      "step": 476
    },
    {
      "epoch": 0.11124067164179105,
      "grad_norm": 1.6606448888778687,
      "learning_rate": 3.8879055400454445e-05,
      "loss": 1.8322,
      "step": 477
    },
    {
      "epoch": 0.11147388059701492,
      "grad_norm": 1.5973105430603027,
      "learning_rate": 3.887419096765534e-05,
      "loss": 1.8205,
      "step": 478
    },
    {
      "epoch": 0.1117070895522388,
      "grad_norm": 1.7587112188339233,
      "learning_rate": 3.886931630853928e-05,
      "loss": 1.9436,
      "step": 479
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 1.7078516483306885,
      "learning_rate": 3.886443142574742e-05,
      "loss": 1.9061,
      "step": 480
    },
    {
      "epoch": 0.11217350746268656,
      "grad_norm": 1.7035492658615112,
      "learning_rate": 3.885953632192648e-05,
      "loss": 1.8901,
      "step": 481
    },
    {
      "epoch": 0.11240671641791045,
      "grad_norm": 1.8233654499053955,
      "learning_rate": 3.885463099972869e-05,
      "loss": 1.859,
      "step": 482
    },
    {
      "epoch": 0.11263992537313433,
      "grad_norm": 1.6868548393249512,
      "learning_rate": 3.884971546181182e-05,
      "loss": 1.9083,
      "step": 483
    },
    {
      "epoch": 0.11287313432835822,
      "grad_norm": 1.7664008140563965,
      "learning_rate": 3.8844789710839185e-05,
      "loss": 1.9686,
      "step": 484
    },
    {
      "epoch": 0.11310634328358209,
      "grad_norm": 1.6839029788970947,
      "learning_rate": 3.883985374947964e-05,
      "loss": 1.8822,
      "step": 485
    },
    {
      "epoch": 0.11333955223880597,
      "grad_norm": 1.808986783027649,
      "learning_rate": 3.8834907580407544e-05,
      "loss": 1.8563,
      "step": 486
    },
    {
      "epoch": 0.11357276119402986,
      "grad_norm": 1.7399083375930786,
      "learning_rate": 3.882995120630282e-05,
      "loss": 1.8384,
      "step": 487
    },
    {
      "epoch": 0.11380597014925373,
      "grad_norm": 1.6091078519821167,
      "learning_rate": 3.8824984629850906e-05,
      "loss": 1.7762,
      "step": 488
    },
    {
      "epoch": 0.11403917910447761,
      "grad_norm": 2.2056517601013184,
      "learning_rate": 3.882000785374275e-05,
      "loss": 1.8974,
      "step": 489
    },
    {
      "epoch": 0.1142723880597015,
      "grad_norm": 1.8060996532440186,
      "learning_rate": 3.881502088067486e-05,
      "loss": 1.9111,
      "step": 490
    },
    {
      "epoch": 0.11450559701492537,
      "grad_norm": 2.0402112007141113,
      "learning_rate": 3.8810023713349246e-05,
      "loss": 1.852,
      "step": 491
    },
    {
      "epoch": 0.11473880597014925,
      "grad_norm": 1.71719491481781,
      "learning_rate": 3.880501635447345e-05,
      "loss": 1.8103,
      "step": 492
    },
    {
      "epoch": 0.11497201492537314,
      "grad_norm": 1.9916797876358032,
      "learning_rate": 3.879999880676053e-05,
      "loss": 1.963,
      "step": 493
    },
    {
      "epoch": 0.11520522388059702,
      "grad_norm": 1.7101480960845947,
      "learning_rate": 3.8794971072929074e-05,
      "loss": 1.9473,
      "step": 494
    },
    {
      "epoch": 0.11543843283582089,
      "grad_norm": 1.8247230052947998,
      "learning_rate": 3.8789933155703165e-05,
      "loss": 1.9005,
      "step": 495
    },
    {
      "epoch": 0.11567164179104478,
      "grad_norm": 1.821762204170227,
      "learning_rate": 3.878488505781244e-05,
      "loss": 1.9223,
      "step": 496
    },
    {
      "epoch": 0.11590485074626866,
      "grad_norm": 5.68617582321167,
      "learning_rate": 3.8779826781992034e-05,
      "loss": 2.0276,
      "step": 497
    },
    {
      "epoch": 0.11613805970149253,
      "grad_norm": 10.62319278717041,
      "learning_rate": 3.877475833098258e-05,
      "loss": 1.79,
      "step": 498
    },
    {
      "epoch": 0.11637126865671642,
      "grad_norm": 2.071319341659546,
      "learning_rate": 3.876967970753025e-05,
      "loss": 1.8204,
      "step": 499
    },
    {
      "epoch": 0.1166044776119403,
      "grad_norm": 1.9426172971725464,
      "learning_rate": 3.876459091438673e-05,
      "loss": 1.9042,
      "step": 500
    },
    {
      "epoch": 0.11683768656716417,
      "grad_norm": 2.147343158721924,
      "learning_rate": 3.875949195430917e-05,
      "loss": 1.7772,
      "step": 501
    },
    {
      "epoch": 0.11707089552238806,
      "grad_norm": 1.7090091705322266,
      "learning_rate": 3.8754382830060305e-05,
      "loss": 1.9179,
      "step": 502
    },
    {
      "epoch": 0.11730410447761194,
      "grad_norm": 2.1276614665985107,
      "learning_rate": 3.87492635444083e-05,
      "loss": 1.8173,
      "step": 503
    },
    {
      "epoch": 0.11753731343283583,
      "grad_norm": 1.6695350408554077,
      "learning_rate": 3.874413410012688e-05,
      "loss": 1.8033,
      "step": 504
    },
    {
      "epoch": 0.1177705223880597,
      "grad_norm": 1.795274257659912,
      "learning_rate": 3.873899449999524e-05,
      "loss": 1.731,
      "step": 505
    },
    {
      "epoch": 0.11800373134328358,
      "grad_norm": 1.7370485067367554,
      "learning_rate": 3.87338447467981e-05,
      "loss": 1.9384,
      "step": 506
    },
    {
      "epoch": 0.11823694029850747,
      "grad_norm": 1.8037909269332886,
      "learning_rate": 3.872868484332567e-05,
      "loss": 1.7818,
      "step": 507
    },
    {
      "epoch": 0.11847014925373134,
      "grad_norm": 2.1849019527435303,
      "learning_rate": 3.872351479237367e-05,
      "loss": 1.7373,
      "step": 508
    },
    {
      "epoch": 0.11870335820895522,
      "grad_norm": 1.72689688205719,
      "learning_rate": 3.8718334596743295e-05,
      "loss": 1.8295,
      "step": 509
    },
    {
      "epoch": 0.11893656716417911,
      "grad_norm": 1.8341068029403687,
      "learning_rate": 3.8713144259241265e-05,
      "loss": 1.9827,
      "step": 510
    },
    {
      "epoch": 0.11916977611940298,
      "grad_norm": 1.7365875244140625,
      "learning_rate": 3.8707943782679775e-05,
      "loss": 1.876,
      "step": 511
    },
    {
      "epoch": 0.11940298507462686,
      "grad_norm": 1.7286404371261597,
      "learning_rate": 3.8702733169876525e-05,
      "loss": 1.7407,
      "step": 512
    },
    {
      "epoch": 0.11963619402985075,
      "grad_norm": 1.8940424919128418,
      "learning_rate": 3.86975124236547e-05,
      "loss": 1.902,
      "step": 513
    },
    {
      "epoch": 0.11986940298507463,
      "grad_norm": 1.857895016670227,
      "learning_rate": 3.8692281546842984e-05,
      "loss": 2.0225,
      "step": 514
    },
    {
      "epoch": 0.1201026119402985,
      "grad_norm": 1.661617398262024,
      "learning_rate": 3.868704054227553e-05,
      "loss": 1.8023,
      "step": 515
    },
    {
      "epoch": 0.12033582089552239,
      "grad_norm": 2.279128313064575,
      "learning_rate": 3.8681789412792e-05,
      "loss": 1.9149,
      "step": 516
    },
    {
      "epoch": 0.12056902985074627,
      "grad_norm": 1.5851643085479736,
      "learning_rate": 3.8676528161237544e-05,
      "loss": 1.8286,
      "step": 517
    },
    {
      "epoch": 0.12080223880597014,
      "grad_norm": 1.827744722366333,
      "learning_rate": 3.8671256790462764e-05,
      "loss": 1.7488,
      "step": 518
    },
    {
      "epoch": 0.12103544776119403,
      "grad_norm": 2.884068727493286,
      "learning_rate": 3.866597530332379e-05,
      "loss": 1.8208,
      "step": 519
    },
    {
      "epoch": 0.12126865671641791,
      "grad_norm": 1.9382590055465698,
      "learning_rate": 3.8660683702682196e-05,
      "loss": 1.9297,
      "step": 520
    },
    {
      "epoch": 0.12150186567164178,
      "grad_norm": 1.6706973314285278,
      "learning_rate": 3.865538199140505e-05,
      "loss": 1.8927,
      "step": 521
    },
    {
      "epoch": 0.12173507462686567,
      "grad_norm": 1.7841709852218628,
      "learning_rate": 3.86500701723649e-05,
      "loss": 1.8215,
      "step": 522
    },
    {
      "epoch": 0.12196828358208955,
      "grad_norm": 2.0414628982543945,
      "learning_rate": 3.8644748248439775e-05,
      "loss": 1.8401,
      "step": 523
    },
    {
      "epoch": 0.12220149253731344,
      "grad_norm": 1.631481409072876,
      "learning_rate": 3.863941622251316e-05,
      "loss": 1.7098,
      "step": 524
    },
    {
      "epoch": 0.12243470149253731,
      "grad_norm": 1.5692790746688843,
      "learning_rate": 3.8634074097474036e-05,
      "loss": 1.7983,
      "step": 525
    },
    {
      "epoch": 0.1226679104477612,
      "grad_norm": 1.7986751794815063,
      "learning_rate": 3.862872187621685e-05,
      "loss": 1.89,
      "step": 526
    },
    {
      "epoch": 0.12290111940298508,
      "grad_norm": 1.4750959873199463,
      "learning_rate": 3.86233595616415e-05,
      "loss": 1.6758,
      "step": 527
    },
    {
      "epoch": 0.12313432835820895,
      "grad_norm": 1.68661630153656,
      "learning_rate": 3.8617987156653376e-05,
      "loss": 1.7711,
      "step": 528
    },
    {
      "epoch": 0.12336753731343283,
      "grad_norm": 1.7725452184677124,
      "learning_rate": 3.861260466416333e-05,
      "loss": 1.7903,
      "step": 529
    },
    {
      "epoch": 0.12360074626865672,
      "grad_norm": 1.705694556236267,
      "learning_rate": 3.860721208708767e-05,
      "loss": 1.904,
      "step": 530
    },
    {
      "epoch": 0.12383395522388059,
      "grad_norm": 1.6583205461502075,
      "learning_rate": 3.860180942834817e-05,
      "loss": 1.8497,
      "step": 531
    },
    {
      "epoch": 0.12406716417910447,
      "grad_norm": 1.6668353080749512,
      "learning_rate": 3.859639669087209e-05,
      "loss": 1.8432,
      "step": 532
    },
    {
      "epoch": 0.12430037313432836,
      "grad_norm": 1.669429898262024,
      "learning_rate": 3.859097387759211e-05,
      "loss": 1.8019,
      "step": 533
    },
    {
      "epoch": 0.12453358208955224,
      "grad_norm": 1.7493246793746948,
      "learning_rate": 3.85855409914464e-05,
      "loss": 1.9772,
      "step": 534
    },
    {
      "epoch": 0.12476679104477612,
      "grad_norm": 1.9798468351364136,
      "learning_rate": 3.858009803537858e-05,
      "loss": 1.8996,
      "step": 535
    },
    {
      "epoch": 0.125,
      "grad_norm": 1.5825300216674805,
      "learning_rate": 3.8574645012337724e-05,
      "loss": 1.8415,
      "step": 536
    },
    {
      "epoch": 0.12523320895522388,
      "grad_norm": 2.2003815174102783,
      "learning_rate": 3.856918192527836e-05,
      "loss": 1.7921,
      "step": 537
    },
    {
      "epoch": 0.12546641791044777,
      "grad_norm": 1.6657588481903076,
      "learning_rate": 3.856370877716047e-05,
      "loss": 1.9312,
      "step": 538
    },
    {
      "epoch": 0.12569962686567165,
      "grad_norm": 1.6855183839797974,
      "learning_rate": 3.855822557094948e-05,
      "loss": 1.7429,
      "step": 539
    },
    {
      "epoch": 0.1259328358208955,
      "grad_norm": 1.6758406162261963,
      "learning_rate": 3.855273230961628e-05,
      "loss": 1.8712,
      "step": 540
    },
    {
      "epoch": 0.1261660447761194,
      "grad_norm": 1.76326584815979,
      "learning_rate": 3.85472289961372e-05,
      "loss": 1.8322,
      "step": 541
    },
    {
      "epoch": 0.12639925373134328,
      "grad_norm": 1.7951874732971191,
      "learning_rate": 3.8541715633494015e-05,
      "loss": 1.9478,
      "step": 542
    },
    {
      "epoch": 0.12663246268656717,
      "grad_norm": 2.0711734294891357,
      "learning_rate": 3.8536192224673955e-05,
      "loss": 1.8973,
      "step": 543
    },
    {
      "epoch": 0.12686567164179105,
      "grad_norm": 1.614438533782959,
      "learning_rate": 3.853065877266967e-05,
      "loss": 1.8301,
      "step": 544
    },
    {
      "epoch": 0.12709888059701493,
      "grad_norm": 2.362276077270508,
      "learning_rate": 3.8525115280479276e-05,
      "loss": 1.8669,
      "step": 545
    },
    {
      "epoch": 0.12733208955223882,
      "grad_norm": 1.9532853364944458,
      "learning_rate": 3.851956175110632e-05,
      "loss": 1.9104,
      "step": 546
    },
    {
      "epoch": 0.12756529850746268,
      "grad_norm": 1.6703084707260132,
      "learning_rate": 3.8513998187559785e-05,
      "loss": 1.8902,
      "step": 547
    },
    {
      "epoch": 0.12779850746268656,
      "grad_norm": 1.591282844543457,
      "learning_rate": 3.8508424592854085e-05,
      "loss": 1.7758,
      "step": 548
    },
    {
      "epoch": 0.12803171641791045,
      "grad_norm": 1.632678747177124,
      "learning_rate": 3.850284097000909e-05,
      "loss": 1.9323,
      "step": 549
    },
    {
      "epoch": 0.12826492537313433,
      "grad_norm": 1.8614696264266968,
      "learning_rate": 3.849724732205008e-05,
      "loss": 2.0605,
      "step": 550
    },
    {
      "epoch": 0.12849813432835822,
      "grad_norm": 3.961153507232666,
      "learning_rate": 3.8491643652007786e-05,
      "loss": 1.9562,
      "step": 551
    },
    {
      "epoch": 0.1287313432835821,
      "grad_norm": 1.6628211736679077,
      "learning_rate": 3.848602996291834e-05,
      "loss": 1.8023,
      "step": 552
    },
    {
      "epoch": 0.12896455223880596,
      "grad_norm": 3.4980955123901367,
      "learning_rate": 3.8480406257823346e-05,
      "loss": 1.8998,
      "step": 553
    },
    {
      "epoch": 0.12919776119402984,
      "grad_norm": 8.508882522583008,
      "learning_rate": 3.847477253976979e-05,
      "loss": 1.9506,
      "step": 554
    },
    {
      "epoch": 0.12943097014925373,
      "grad_norm": 2.1802289485931396,
      "learning_rate": 3.846912881181012e-05,
      "loss": 1.8399,
      "step": 555
    },
    {
      "epoch": 0.1296641791044776,
      "grad_norm": 1.6738145351409912,
      "learning_rate": 3.8463475077002187e-05,
      "loss": 1.863,
      "step": 556
    },
    {
      "epoch": 0.1298973880597015,
      "grad_norm": 1.7273435592651367,
      "learning_rate": 3.845781133840927e-05,
      "loss": 1.8877,
      "step": 557
    },
    {
      "epoch": 0.13013059701492538,
      "grad_norm": 1.8935929536819458,
      "learning_rate": 3.845213759910006e-05,
      "loss": 1.9345,
      "step": 558
    },
    {
      "epoch": 0.13036380597014927,
      "grad_norm": 1.716622233390808,
      "learning_rate": 3.844645386214868e-05,
      "loss": 1.7583,
      "step": 559
    },
    {
      "epoch": 0.13059701492537312,
      "grad_norm": 1.6881874799728394,
      "learning_rate": 3.8440760130634656e-05,
      "loss": 1.9046,
      "step": 560
    },
    {
      "epoch": 0.130830223880597,
      "grad_norm": 1.7108261585235596,
      "learning_rate": 3.843505640764295e-05,
      "loss": 1.8291,
      "step": 561
    },
    {
      "epoch": 0.1310634328358209,
      "grad_norm": 1.6400988101959229,
      "learning_rate": 3.8429342696263905e-05,
      "loss": 1.933,
      "step": 562
    },
    {
      "epoch": 0.13129664179104478,
      "grad_norm": 1.6172696352005005,
      "learning_rate": 3.8423618999593303e-05,
      "loss": 1.8152,
      "step": 563
    },
    {
      "epoch": 0.13152985074626866,
      "grad_norm": 1.9020593166351318,
      "learning_rate": 3.841788532073233e-05,
      "loss": 1.9087,
      "step": 564
    },
    {
      "epoch": 0.13176305970149255,
      "grad_norm": 1.5586692094802856,
      "learning_rate": 3.8412141662787584e-05,
      "loss": 1.8213,
      "step": 565
    },
    {
      "epoch": 0.13199626865671643,
      "grad_norm": 1.6323524713516235,
      "learning_rate": 3.840638802887105e-05,
      "loss": 1.7564,
      "step": 566
    },
    {
      "epoch": 0.1322294776119403,
      "grad_norm": 1.8810830116271973,
      "learning_rate": 3.8400624422100145e-05,
      "loss": 1.9664,
      "step": 567
    },
    {
      "epoch": 0.13246268656716417,
      "grad_norm": 1.555600881576538,
      "learning_rate": 3.8394850845597664e-05,
      "loss": 1.7495,
      "step": 568
    },
    {
      "epoch": 0.13269589552238806,
      "grad_norm": 1.6337941884994507,
      "learning_rate": 3.838906730249183e-05,
      "loss": 1.8366,
      "step": 569
    },
    {
      "epoch": 0.13292910447761194,
      "grad_norm": 1.5642575025558472,
      "learning_rate": 3.8383273795916245e-05,
      "loss": 1.8538,
      "step": 570
    },
    {
      "epoch": 0.13316231343283583,
      "grad_norm": 1.6527751684188843,
      "learning_rate": 3.837747032900992e-05,
      "loss": 1.8918,
      "step": 571
    },
    {
      "epoch": 0.1333955223880597,
      "grad_norm": 1.7977973222732544,
      "learning_rate": 3.837165690491726e-05,
      "loss": 1.8895,
      "step": 572
    },
    {
      "epoch": 0.13362873134328357,
      "grad_norm": 1.6657418012619019,
      "learning_rate": 3.8365833526788053e-05,
      "loss": 1.7777,
      "step": 573
    },
    {
      "epoch": 0.13386194029850745,
      "grad_norm": 1.6989332437515259,
      "learning_rate": 3.836000019777751e-05,
      "loss": 1.8374,
      "step": 574
    },
    {
      "epoch": 0.13409514925373134,
      "grad_norm": 1.7553014755249023,
      "learning_rate": 3.8354156921046204e-05,
      "loss": 1.9428,
      "step": 575
    },
    {
      "epoch": 0.13432835820895522,
      "grad_norm": 1.6305978298187256,
      "learning_rate": 3.834830369976011e-05,
      "loss": 1.9457,
      "step": 576
    },
    {
      "epoch": 0.1345615671641791,
      "grad_norm": 1.7655138969421387,
      "learning_rate": 3.83424405370906e-05,
      "loss": 1.9382,
      "step": 577
    },
    {
      "epoch": 0.134794776119403,
      "grad_norm": 1.9220948219299316,
      "learning_rate": 3.833656743621441e-05,
      "loss": 1.7962,
      "step": 578
    },
    {
      "epoch": 0.13502798507462688,
      "grad_norm": 1.6531621217727661,
      "learning_rate": 3.833068440031367e-05,
      "loss": 1.8319,
      "step": 579
    },
    {
      "epoch": 0.13526119402985073,
      "grad_norm": 1.7501474618911743,
      "learning_rate": 3.832479143257591e-05,
      "loss": 1.8661,
      "step": 580
    },
    {
      "epoch": 0.13549440298507462,
      "grad_norm": 1.5817334651947021,
      "learning_rate": 3.8318888536194025e-05,
      "loss": 1.8296,
      "step": 581
    },
    {
      "epoch": 0.1357276119402985,
      "grad_norm": 1.913205623626709,
      "learning_rate": 3.8312975714366285e-05,
      "loss": 1.7819,
      "step": 582
    },
    {
      "epoch": 0.1359608208955224,
      "grad_norm": 1.7539390325546265,
      "learning_rate": 3.830705297029634e-05,
      "loss": 1.9505,
      "step": 583
    },
    {
      "epoch": 0.13619402985074627,
      "grad_norm": 1.8320680856704712,
      "learning_rate": 3.830112030719324e-05,
      "loss": 1.9792,
      "step": 584
    },
    {
      "epoch": 0.13642723880597016,
      "grad_norm": 1.7389909029006958,
      "learning_rate": 3.8295177728271375e-05,
      "loss": 1.8987,
      "step": 585
    },
    {
      "epoch": 0.13666044776119404,
      "grad_norm": 1.8924716711044312,
      "learning_rate": 3.828922523675053e-05,
      "loss": 1.8432,
      "step": 586
    },
    {
      "epoch": 0.1368936567164179,
      "grad_norm": 1.8004634380340576,
      "learning_rate": 3.828326283585584e-05,
      "loss": 1.855,
      "step": 587
    },
    {
      "epoch": 0.13712686567164178,
      "grad_norm": 1.7374848127365112,
      "learning_rate": 3.8277290528817834e-05,
      "loss": 1.8606,
      "step": 588
    },
    {
      "epoch": 0.13736007462686567,
      "grad_norm": 1.995637059211731,
      "learning_rate": 3.82713083188724e-05,
      "loss": 1.784,
      "step": 589
    },
    {
      "epoch": 0.13759328358208955,
      "grad_norm": 1.804302453994751,
      "learning_rate": 3.826531620926078e-05,
      "loss": 1.9823,
      "step": 590
    },
    {
      "epoch": 0.13782649253731344,
      "grad_norm": 1.7690069675445557,
      "learning_rate": 3.825931420322959e-05,
      "loss": 1.8781,
      "step": 591
    },
    {
      "epoch": 0.13805970149253732,
      "grad_norm": 1.5160150527954102,
      "learning_rate": 3.825330230403081e-05,
      "loss": 1.7908,
      "step": 592
    },
    {
      "epoch": 0.13829291044776118,
      "grad_norm": 1.736472725868225,
      "learning_rate": 3.8247280514921765e-05,
      "loss": 1.8125,
      "step": 593
    },
    {
      "epoch": 0.13852611940298507,
      "grad_norm": 1.6031230688095093,
      "learning_rate": 3.8241248839165164e-05,
      "loss": 1.7838,
      "step": 594
    },
    {
      "epoch": 0.13875932835820895,
      "grad_norm": 1.8648014068603516,
      "learning_rate": 3.823520728002905e-05,
      "loss": 1.7746,
      "step": 595
    },
    {
      "epoch": 0.13899253731343283,
      "grad_norm": 1.6084344387054443,
      "learning_rate": 3.822915584078683e-05,
      "loss": 1.873,
      "step": 596
    },
    {
      "epoch": 0.13922574626865672,
      "grad_norm": 1.8837107419967651,
      "learning_rate": 3.822309452471727e-05,
      "loss": 1.907,
      "step": 597
    },
    {
      "epoch": 0.1394589552238806,
      "grad_norm": 1.6757920980453491,
      "learning_rate": 3.8217023335104475e-05,
      "loss": 1.8326,
      "step": 598
    },
    {
      "epoch": 0.1396921641791045,
      "grad_norm": 1.7303266525268555,
      "learning_rate": 3.821094227523791e-05,
      "loss": 1.8696,
      "step": 599
    },
    {
      "epoch": 0.13992537313432835,
      "grad_norm": 1.7140110731124878,
      "learning_rate": 3.820485134841237e-05,
      "loss": 2.0003,
      "step": 600
    },
    {
      "epoch": 0.14015858208955223,
      "grad_norm": 1.7962599992752075,
      "learning_rate": 3.8198750557928026e-05,
      "loss": 1.9732,
      "step": 601
    },
    {
      "epoch": 0.14039179104477612,
      "grad_norm": 1.75541090965271,
      "learning_rate": 3.819263990709037e-05,
      "loss": 1.929,
      "step": 602
    },
    {
      "epoch": 0.140625,
      "grad_norm": 1.4548100233078003,
      "learning_rate": 3.818651939921025e-05,
      "loss": 1.6609,
      "step": 603
    },
    {
      "epoch": 0.14085820895522388,
      "grad_norm": 1.8790242671966553,
      "learning_rate": 3.818038903760383e-05,
      "loss": 1.7907,
      "step": 604
    },
    {
      "epoch": 0.14109141791044777,
      "grad_norm": 1.7562451362609863,
      "learning_rate": 3.8174248825592655e-05,
      "loss": 1.9309,
      "step": 605
    },
    {
      "epoch": 0.14132462686567165,
      "grad_norm": 1.7391304969787598,
      "learning_rate": 3.8168098766503566e-05,
      "loss": 1.8939,
      "step": 606
    },
    {
      "epoch": 0.1415578358208955,
      "grad_norm": 1.9754254817962646,
      "learning_rate": 3.8161938863668764e-05,
      "loss": 1.7364,
      "step": 607
    },
    {
      "epoch": 0.1417910447761194,
      "grad_norm": 1.5888538360595703,
      "learning_rate": 3.815576912042577e-05,
      "loss": 1.8123,
      "step": 608
    },
    {
      "epoch": 0.14202425373134328,
      "grad_norm": 1.828532338142395,
      "learning_rate": 3.814958954011745e-05,
      "loss": 1.9907,
      "step": 609
    },
    {
      "epoch": 0.14225746268656717,
      "grad_norm": 1.7533674240112305,
      "learning_rate": 3.814340012609199e-05,
      "loss": 1.9634,
      "step": 610
    },
    {
      "epoch": 0.14249067164179105,
      "grad_norm": 1.7876921892166138,
      "learning_rate": 3.8137200881702905e-05,
      "loss": 1.8962,
      "step": 611
    },
    {
      "epoch": 0.14272388059701493,
      "grad_norm": 2.1552579402923584,
      "learning_rate": 3.8130991810309034e-05,
      "loss": 1.8537,
      "step": 612
    },
    {
      "epoch": 0.14295708955223882,
      "grad_norm": 1.636115550994873,
      "learning_rate": 3.812477291527455e-05,
      "loss": 1.8532,
      "step": 613
    },
    {
      "epoch": 0.14319029850746268,
      "grad_norm": 1.649437665939331,
      "learning_rate": 3.811854419996894e-05,
      "loss": 1.7605,
      "step": 614
    },
    {
      "epoch": 0.14342350746268656,
      "grad_norm": 1.8090143203735352,
      "learning_rate": 3.8112305667767014e-05,
      "loss": 1.7562,
      "step": 615
    },
    {
      "epoch": 0.14365671641791045,
      "grad_norm": 2.222144842147827,
      "learning_rate": 3.81060573220489e-05,
      "loss": 1.9434,
      "step": 616
    },
    {
      "epoch": 0.14388992537313433,
      "grad_norm": 1.8328090906143188,
      "learning_rate": 3.8099799166200056e-05,
      "loss": 1.8459,
      "step": 617
    },
    {
      "epoch": 0.14412313432835822,
      "grad_norm": 1.7580220699310303,
      "learning_rate": 3.809353120361122e-05,
      "loss": 1.8189,
      "step": 618
    },
    {
      "epoch": 0.1443563432835821,
      "grad_norm": 1.889229655265808,
      "learning_rate": 3.8087253437678494e-05,
      "loss": 1.7933,
      "step": 619
    },
    {
      "epoch": 0.14458955223880596,
      "grad_norm": 1.5743868350982666,
      "learning_rate": 3.808096587180325e-05,
      "loss": 1.6724,
      "step": 620
    },
    {
      "epoch": 0.14482276119402984,
      "grad_norm": 1.8446691036224365,
      "learning_rate": 3.807466850939218e-05,
      "loss": 1.9132,
      "step": 621
    },
    {
      "epoch": 0.14505597014925373,
      "grad_norm": 1.7169883251190186,
      "learning_rate": 3.80683613538573e-05,
      "loss": 1.8185,
      "step": 622
    },
    {
      "epoch": 0.1452891791044776,
      "grad_norm": 1.6484624147415161,
      "learning_rate": 3.806204440861592e-05,
      "loss": 1.8646,
      "step": 623
    },
    {
      "epoch": 0.1455223880597015,
      "grad_norm": 1.8274257183074951,
      "learning_rate": 3.805571767709065e-05,
      "loss": 1.7771,
      "step": 624
    },
    {
      "epoch": 0.14575559701492538,
      "grad_norm": 1.621960997581482,
      "learning_rate": 3.80493811627094e-05,
      "loss": 1.8819,
      "step": 625
    },
    {
      "epoch": 0.14598880597014927,
      "grad_norm": 1.627021074295044,
      "learning_rate": 3.8043034868905406e-05,
      "loss": 1.8997,
      "step": 626
    },
    {
      "epoch": 0.14622201492537312,
      "grad_norm": 1.9370806217193604,
      "learning_rate": 3.803667879911717e-05,
      "loss": 1.8961,
      "step": 627
    },
    {
      "epoch": 0.146455223880597,
      "grad_norm": 2.0215771198272705,
      "learning_rate": 3.80303129567885e-05,
      "loss": 1.844,
      "step": 628
    },
    {
      "epoch": 0.1466884328358209,
      "grad_norm": 2.4156980514526367,
      "learning_rate": 3.8023937345368524e-05,
      "loss": 1.8805,
      "step": 629
    },
    {
      "epoch": 0.14692164179104478,
      "grad_norm": 1.5813698768615723,
      "learning_rate": 3.8017551968311625e-05,
      "loss": 1.8105,
      "step": 630
    },
    {
      "epoch": 0.14715485074626866,
      "grad_norm": 1.6295543909072876,
      "learning_rate": 3.80111568290775e-05,
      "loss": 1.798,
      "step": 631
    },
    {
      "epoch": 0.14738805970149255,
      "grad_norm": 1.878781795501709,
      "learning_rate": 3.800475193113114e-05,
      "loss": 1.9143,
      "step": 632
    },
    {
      "epoch": 0.14762126865671643,
      "grad_norm": 1.614585041999817,
      "learning_rate": 3.799833727794281e-05,
      "loss": 1.8528,
      "step": 633
    },
    {
      "epoch": 0.1478544776119403,
      "grad_norm": 1.7429324388504028,
      "learning_rate": 3.7991912872988053e-05,
      "loss": 1.8455,
      "step": 634
    },
    {
      "epoch": 0.14808768656716417,
      "grad_norm": 1.757062554359436,
      "learning_rate": 3.798547871974771e-05,
      "loss": 1.9272,
      "step": 635
    },
    {
      "epoch": 0.14832089552238806,
      "grad_norm": 1.5766081809997559,
      "learning_rate": 3.797903482170791e-05,
      "loss": 1.746,
      "step": 636
    },
    {
      "epoch": 0.14855410447761194,
      "grad_norm": 1.5479782819747925,
      "learning_rate": 3.7972581182360045e-05,
      "loss": 1.777,
      "step": 637
    },
    {
      "epoch": 0.14878731343283583,
      "grad_norm": 5.7531867027282715,
      "learning_rate": 3.79661178052008e-05,
      "loss": 1.9667,
      "step": 638
    },
    {
      "epoch": 0.1490205223880597,
      "grad_norm": 1.5791025161743164,
      "learning_rate": 3.795964469373212e-05,
      "loss": 1.8993,
      "step": 639
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 1.6289030313491821,
      "learning_rate": 3.795316185146124e-05,
      "loss": 1.8488,
      "step": 640
    },
    {
      "epoch": 0.14948694029850745,
      "grad_norm": 1.6307135820388794,
      "learning_rate": 3.794666928190065e-05,
      "loss": 1.8025,
      "step": 641
    },
    {
      "epoch": 0.14972014925373134,
      "grad_norm": 1.5682588815689087,
      "learning_rate": 3.7940166988568134e-05,
      "loss": 1.9042,
      "step": 642
    },
    {
      "epoch": 0.14995335820895522,
      "grad_norm": 1.5684003829956055,
      "learning_rate": 3.7933654974986733e-05,
      "loss": 1.7298,
      "step": 643
    },
    {
      "epoch": 0.1501865671641791,
      "grad_norm": 1.9034098386764526,
      "learning_rate": 3.7927133244684733e-05,
      "loss": 2.0009,
      "step": 644
    },
    {
      "epoch": 0.150419776119403,
      "grad_norm": 6.363749027252197,
      "learning_rate": 3.792060180119572e-05,
      "loss": 1.7218,
      "step": 645
    },
    {
      "epoch": 0.15065298507462688,
      "grad_norm": 1.7217854261398315,
      "learning_rate": 3.791406064805852e-05,
      "loss": 1.8411,
      "step": 646
    },
    {
      "epoch": 0.15088619402985073,
      "grad_norm": 1.691464900970459,
      "learning_rate": 3.7907509788817234e-05,
      "loss": 1.822,
      "step": 647
    },
    {
      "epoch": 0.15111940298507462,
      "grad_norm": 1.7491888999938965,
      "learning_rate": 3.790094922702121e-05,
      "loss": 1.9398,
      "step": 648
    },
    {
      "epoch": 0.1513526119402985,
      "grad_norm": 1.6085827350616455,
      "learning_rate": 3.789437896622506e-05,
      "loss": 1.8517,
      "step": 649
    },
    {
      "epoch": 0.1515858208955224,
      "grad_norm": 1.8373286724090576,
      "learning_rate": 3.7887799009988643e-05,
      "loss": 1.8896,
      "step": 650
    },
    {
      "epoch": 0.15181902985074627,
      "grad_norm": 1.5964611768722534,
      "learning_rate": 3.7881209361877083e-05,
      "loss": 1.6167,
      "step": 651
    },
    {
      "epoch": 0.15205223880597016,
      "grad_norm": 1.6221952438354492,
      "learning_rate": 3.7874610025460745e-05,
      "loss": 1.9413,
      "step": 652
    },
    {
      "epoch": 0.15228544776119404,
      "grad_norm": 1.548662781715393,
      "learning_rate": 3.7868001004315253e-05,
      "loss": 1.7834,
      "step": 653
    },
    {
      "epoch": 0.1525186567164179,
      "grad_norm": 1.5606175661087036,
      "learning_rate": 3.786138230202147e-05,
      "loss": 1.753,
      "step": 654
    },
    {
      "epoch": 0.15275186567164178,
      "grad_norm": 1.5557639598846436,
      "learning_rate": 3.785475392216551e-05,
      "loss": 1.8477,
      "step": 655
    },
    {
      "epoch": 0.15298507462686567,
      "grad_norm": 6.305323600769043,
      "learning_rate": 3.784811586833872e-05,
      "loss": 1.9158,
      "step": 656
    },
    {
      "epoch": 0.15321828358208955,
      "grad_norm": 4.2686333656311035,
      "learning_rate": 3.7841468144137705e-05,
      "loss": 1.7661,
      "step": 657
    },
    {
      "epoch": 0.15345149253731344,
      "grad_norm": 1.8356473445892334,
      "learning_rate": 3.783481075316429e-05,
      "loss": 1.9685,
      "step": 658
    },
    {
      "epoch": 0.15368470149253732,
      "grad_norm": 1.760042667388916,
      "learning_rate": 3.782814369902556e-05,
      "loss": 1.9927,
      "step": 659
    },
    {
      "epoch": 0.15391791044776118,
      "grad_norm": 1.5548053979873657,
      "learning_rate": 3.7821466985333814e-05,
      "loss": 1.7311,
      "step": 660
    },
    {
      "epoch": 0.15415111940298507,
      "grad_norm": 1.560048222541809,
      "learning_rate": 3.781478061570661e-05,
      "loss": 1.7454,
      "step": 661
    },
    {
      "epoch": 0.15438432835820895,
      "grad_norm": 1.5317190885543823,
      "learning_rate": 3.7808084593766706e-05,
      "loss": 1.7318,
      "step": 662
    },
    {
      "epoch": 0.15461753731343283,
      "grad_norm": 1.569642424583435,
      "learning_rate": 3.780137892314211e-05,
      "loss": 1.8455,
      "step": 663
    },
    {
      "epoch": 0.15485074626865672,
      "grad_norm": 1.703273057937622,
      "learning_rate": 3.7794663607466054e-05,
      "loss": 1.8331,
      "step": 664
    },
    {
      "epoch": 0.1550839552238806,
      "grad_norm": 1.7371338605880737,
      "learning_rate": 3.7787938650377e-05,
      "loss": 1.9047,
      "step": 665
    },
    {
      "epoch": 0.1553171641791045,
      "grad_norm": 1.636391520500183,
      "learning_rate": 3.778120405551863e-05,
      "loss": 1.8462,
      "step": 666
    },
    {
      "epoch": 0.15555037313432835,
      "grad_norm": 1.675790548324585,
      "learning_rate": 3.777445982653984e-05,
      "loss": 1.7998,
      "step": 667
    },
    {
      "epoch": 0.15578358208955223,
      "grad_norm": 1.5780531167984009,
      "learning_rate": 3.7767705967094766e-05,
      "loss": 1.7959,
      "step": 668
    },
    {
      "epoch": 0.15601679104477612,
      "grad_norm": 1.6445425748825073,
      "learning_rate": 3.776094248084273e-05,
      "loss": 1.8562,
      "step": 669
    },
    {
      "epoch": 0.15625,
      "grad_norm": 1.603098750114441,
      "learning_rate": 3.775416937144831e-05,
      "loss": 1.8545,
      "step": 670
    },
    {
      "epoch": 0.15648320895522388,
      "grad_norm": 1.5723624229431152,
      "learning_rate": 3.774738664258126e-05,
      "loss": 1.8703,
      "step": 671
    },
    {
      "epoch": 0.15671641791044777,
      "grad_norm": 1.8484141826629639,
      "learning_rate": 3.774059429791658e-05,
      "loss": 1.8082,
      "step": 672
    },
    {
      "epoch": 0.15694962686567165,
      "grad_norm": 1.681832194328308,
      "learning_rate": 3.7733792341134455e-05,
      "loss": 1.9219,
      "step": 673
    },
    {
      "epoch": 0.1571828358208955,
      "grad_norm": 1.5479450225830078,
      "learning_rate": 3.772698077592028e-05,
      "loss": 1.7902,
      "step": 674
    },
    {
      "epoch": 0.1574160447761194,
      "grad_norm": 1.5456233024597168,
      "learning_rate": 3.7720159605964666e-05,
      "loss": 1.8561,
      "step": 675
    },
    {
      "epoch": 0.15764925373134328,
      "grad_norm": 1.5914347171783447,
      "learning_rate": 3.771332883496343e-05,
      "loss": 1.7913,
      "step": 676
    },
    {
      "epoch": 0.15788246268656717,
      "grad_norm": 1.572235107421875,
      "learning_rate": 3.7706488466617585e-05,
      "loss": 1.836,
      "step": 677
    },
    {
      "epoch": 0.15811567164179105,
      "grad_norm": 1.751052737236023,
      "learning_rate": 3.7699638504633334e-05,
      "loss": 1.7424,
      "step": 678
    },
    {
      "epoch": 0.15834888059701493,
      "grad_norm": 1.5815194845199585,
      "learning_rate": 3.769277895272209e-05,
      "loss": 1.7752,
      "step": 679
    },
    {
      "epoch": 0.15858208955223882,
      "grad_norm": 1.7180979251861572,
      "learning_rate": 3.768590981460047e-05,
      "loss": 1.9536,
      "step": 680
    },
    {
      "epoch": 0.15881529850746268,
      "grad_norm": 1.702512264251709,
      "learning_rate": 3.7679031093990273e-05,
      "loss": 1.7554,
      "step": 681
    },
    {
      "epoch": 0.15904850746268656,
      "grad_norm": 1.5992580652236938,
      "learning_rate": 3.7672142794618484e-05,
      "loss": 1.8401,
      "step": 682
    },
    {
      "epoch": 0.15928171641791045,
      "grad_norm": 1.6171188354492188,
      "learning_rate": 3.76652449202173e-05,
      "loss": 1.8938,
      "step": 683
    },
    {
      "epoch": 0.15951492537313433,
      "grad_norm": 1.5195794105529785,
      "learning_rate": 3.765833747452408e-05,
      "loss": 1.8607,
      "step": 684
    },
    {
      "epoch": 0.15974813432835822,
      "grad_norm": 1.5531786680221558,
      "learning_rate": 3.7651420461281375e-05,
      "loss": 1.855,
      "step": 685
    },
    {
      "epoch": 0.1599813432835821,
      "grad_norm": 1.8633657693862915,
      "learning_rate": 3.7644493884236934e-05,
      "loss": 1.792,
      "step": 686
    },
    {
      "epoch": 0.16021455223880596,
      "grad_norm": 1.5886945724487305,
      "learning_rate": 3.7637557747143676e-05,
      "loss": 1.8087,
      "step": 687
    },
    {
      "epoch": 0.16044776119402984,
      "grad_norm": 24.580108642578125,
      "learning_rate": 3.763061205375971e-05,
      "loss": 1.9446,
      "step": 688
    },
    {
      "epoch": 0.16068097014925373,
      "grad_norm": 1.6200324296951294,
      "learning_rate": 3.762365680784831e-05,
      "loss": 1.8471,
      "step": 689
    },
    {
      "epoch": 0.1609141791044776,
      "grad_norm": 1.6811875104904175,
      "learning_rate": 3.761669201317794e-05,
      "loss": 1.8605,
      "step": 690
    },
    {
      "epoch": 0.1611473880597015,
      "grad_norm": 1.7662079334259033,
      "learning_rate": 3.760971767352222e-05,
      "loss": 1.9124,
      "step": 691
    },
    {
      "epoch": 0.16138059701492538,
      "grad_norm": 1.7510358095169067,
      "learning_rate": 3.760273379265995e-05,
      "loss": 1.8007,
      "step": 692
    },
    {
      "epoch": 0.16161380597014927,
      "grad_norm": 1.6192445755004883,
      "learning_rate": 3.7595740374375104e-05,
      "loss": 1.9518,
      "step": 693
    },
    {
      "epoch": 0.16184701492537312,
      "grad_norm": 1.6131467819213867,
      "learning_rate": 3.7588737422456825e-05,
      "loss": 1.8764,
      "step": 694
    },
    {
      "epoch": 0.162080223880597,
      "grad_norm": 1.6501579284667969,
      "learning_rate": 3.75817249406994e-05,
      "loss": 1.8942,
      "step": 695
    },
    {
      "epoch": 0.1623134328358209,
      "grad_norm": 1.7367316484451294,
      "learning_rate": 3.757470293290231e-05,
      "loss": 1.8554,
      "step": 696
    },
    {
      "epoch": 0.16254664179104478,
      "grad_norm": 1.7118103504180908,
      "learning_rate": 3.7567671402870185e-05,
      "loss": 1.8493,
      "step": 697
    },
    {
      "epoch": 0.16277985074626866,
      "grad_norm": 1.6270357370376587,
      "learning_rate": 3.7560630354412805e-05,
      "loss": 1.7185,
      "step": 698
    },
    {
      "epoch": 0.16301305970149255,
      "grad_norm": 2.089033842086792,
      "learning_rate": 3.755357979134511e-05,
      "loss": 1.7234,
      "step": 699
    },
    {
      "epoch": 0.16324626865671643,
      "grad_norm": 1.6533550024032593,
      "learning_rate": 3.7546519717487216e-05,
      "loss": 1.8629,
      "step": 700
    }
  ],
  "logging_steps": 1,
  "max_steps": 4288,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 700,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.725376342523904e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}