{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 39.42652329749104,
  "eval_steps": 500,
  "global_step": 5500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.14336917562724014,
      "grad_norm": 0.7797384262084961,
      "learning_rate": 4e-05,
      "loss": 2.2646,
      "step": 20
    },
    {
      "epoch": 0.2867383512544803,
      "grad_norm": 0.8766279220581055,
      "learning_rate": 8e-05,
      "loss": 2.2106,
      "step": 40
    },
    {
      "epoch": 0.43010752688172044,
      "grad_norm": 0.9374061822891235,
      "learning_rate": 0.00012,
      "loss": 1.9441,
      "step": 60
    },
    {
      "epoch": 0.5734767025089605,
      "grad_norm": 0.9437981247901917,
      "learning_rate": 0.00016,
      "loss": 1.8573,
      "step": 80
    },
    {
      "epoch": 0.7168458781362007,
      "grad_norm": 1.16326904296875,
      "learning_rate": 0.0002,
      "loss": 1.7739,
      "step": 100
    },
    {
      "epoch": 0.8602150537634409,
      "grad_norm": 1.1639043092727661,
      "learning_rate": 0.00019988455988455988,
      "loss": 1.6996,
      "step": 120
    },
    {
      "epoch": 1.003584229390681,
      "grad_norm": 1.3743420839309692,
      "learning_rate": 0.00019976911976911978,
      "loss": 1.6048,
      "step": 140
    },
    {
      "epoch": 1.146953405017921,
      "grad_norm": 1.3700010776519775,
      "learning_rate": 0.00019965367965367966,
      "loss": 1.5738,
      "step": 160
    },
    {
      "epoch": 1.2903225806451613,
      "grad_norm": 1.8980423212051392,
      "learning_rate": 0.00019953823953823956,
      "loss": 1.4982,
      "step": 180
    },
    {
      "epoch": 1.4336917562724014,
      "grad_norm": 1.8157719373703003,
      "learning_rate": 0.00019942279942279943,
      "loss": 1.5173,
      "step": 200
    },
    {
      "epoch": 1.5770609318996416,
      "grad_norm": 2.1088435649871826,
      "learning_rate": 0.0001993073593073593,
      "loss": 1.4963,
      "step": 220
    },
    {
      "epoch": 1.7204301075268817,
      "grad_norm": 2.1570374965667725,
      "learning_rate": 0.0001991919191919192,
      "loss": 1.4934,
      "step": 240
    },
    {
      "epoch": 1.863799283154122,
      "grad_norm": 2.2034499645233154,
      "learning_rate": 0.00019907647907647908,
      "loss": 1.429,
      "step": 260
    },
    {
      "epoch": 2.007168458781362,
      "grad_norm": 1.7870151996612549,
      "learning_rate": 0.00019896103896103898,
      "loss": 1.4812,
      "step": 280
    },
    {
      "epoch": 2.150537634408602,
      "grad_norm": 2.705448627471924,
      "learning_rate": 0.00019884559884559885,
      "loss": 1.3495,
      "step": 300
    },
    {
      "epoch": 2.293906810035842,
      "grad_norm": 2.151088237762451,
      "learning_rate": 0.00019873015873015875,
      "loss": 1.3401,
      "step": 320
    },
    {
      "epoch": 2.4372759856630823,
      "grad_norm": 2.273388385772705,
      "learning_rate": 0.00019861471861471863,
      "loss": 1.261,
      "step": 340
    },
    {
      "epoch": 2.5806451612903225,
      "grad_norm": 3.103841543197632,
      "learning_rate": 0.00019849927849927853,
      "loss": 1.2768,
      "step": 360
    },
    {
      "epoch": 2.7240143369175627,
      "grad_norm": 2.259409189224243,
      "learning_rate": 0.00019838383838383837,
      "loss": 1.2932,
      "step": 380
    },
    {
      "epoch": 2.867383512544803,
      "grad_norm": 3.061185359954834,
      "learning_rate": 0.00019826839826839827,
      "loss": 1.2528,
      "step": 400
    },
    {
      "epoch": 3.010752688172043,
      "grad_norm": 2.311734199523926,
      "learning_rate": 0.00019815295815295815,
      "loss": 1.2575,
      "step": 420
    },
    {
      "epoch": 3.154121863799283,
      "grad_norm": 3.4611690044403076,
      "learning_rate": 0.00019803751803751805,
      "loss": 1.1404,
      "step": 440
    },
    {
      "epoch": 3.2974910394265233,
      "grad_norm": 3.1576485633850098,
      "learning_rate": 0.00019792207792207792,
      "loss": 1.1195,
      "step": 460
    },
    {
      "epoch": 3.4408602150537635,
      "grad_norm": 2.9822025299072266,
      "learning_rate": 0.00019780663780663782,
      "loss": 1.1651,
      "step": 480
    },
    {
      "epoch": 3.5842293906810037,
      "grad_norm": 2.700014591217041,
      "learning_rate": 0.0001976911976911977,
      "loss": 1.1024,
      "step": 500
    },
    {
      "epoch": 3.727598566308244,
      "grad_norm": 3.5420925617218018,
      "learning_rate": 0.0001975757575757576,
      "loss": 1.1593,
      "step": 520
    },
    {
      "epoch": 3.870967741935484,
      "grad_norm": 3.261657238006592,
      "learning_rate": 0.00019746031746031747,
      "loss": 1.1687,
      "step": 540
    },
    {
      "epoch": 4.014336917562724,
      "grad_norm": 2.6333651542663574,
      "learning_rate": 0.00019734487734487734,
      "loss": 1.0741,
      "step": 560
    },
    {
      "epoch": 4.157706093189964,
      "grad_norm": 3.958665609359741,
      "learning_rate": 0.00019722943722943722,
      "loss": 0.9966,
      "step": 580
    },
    {
      "epoch": 4.301075268817204,
      "grad_norm": 3.4680724143981934,
      "learning_rate": 0.00019711399711399712,
      "loss": 0.9532,
      "step": 600
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 3.7936246395111084,
      "learning_rate": 0.000196998556998557,
      "loss": 1.0428,
      "step": 620
    },
    {
      "epoch": 4.587813620071684,
      "grad_norm": 3.4765727519989014,
      "learning_rate": 0.0001968831168831169,
      "loss": 1.0067,
      "step": 640
    },
    {
      "epoch": 4.731182795698925,
      "grad_norm": 4.042802810668945,
      "learning_rate": 0.00019676767676767677,
      "loss": 1.0318,
      "step": 660
    },
    {
      "epoch": 4.874551971326165,
      "grad_norm": 3.597938060760498,
      "learning_rate": 0.00019665223665223667,
      "loss": 0.9431,
      "step": 680
    },
    {
      "epoch": 5.017921146953405,
      "grad_norm": 3.7627108097076416,
      "learning_rate": 0.00019653679653679654,
      "loss": 1.0065,
      "step": 700
    },
    {
      "epoch": 5.161290322580645,
      "grad_norm": 3.363626003265381,
      "learning_rate": 0.00019642135642135644,
      "loss": 0.8925,
      "step": 720
    },
    {
      "epoch": 5.304659498207886,
      "grad_norm": 3.777928352355957,
      "learning_rate": 0.00019630591630591631,
      "loss": 0.8798,
      "step": 740
    },
    {
      "epoch": 5.448028673835125,
      "grad_norm": 3.517575740814209,
      "learning_rate": 0.0001961904761904762,
      "loss": 0.8555,
      "step": 760
    },
    {
      "epoch": 5.591397849462366,
      "grad_norm": 3.2290761470794678,
      "learning_rate": 0.0001960750360750361,
      "loss": 0.8826,
      "step": 780
    },
    {
      "epoch": 5.734767025089606,
      "grad_norm": 3.73039174079895,
      "learning_rate": 0.00019595959595959596,
      "loss": 0.9197,
      "step": 800
    },
    {
      "epoch": 5.878136200716845,
      "grad_norm": 3.3132483959198,
      "learning_rate": 0.00019584415584415586,
      "loss": 0.9388,
      "step": 820
    },
    {
      "epoch": 6.021505376344086,
      "grad_norm": 3.7642362117767334,
      "learning_rate": 0.00019572871572871574,
      "loss": 0.8208,
      "step": 840
    },
    {
      "epoch": 6.164874551971327,
      "grad_norm": 4.448329448699951,
      "learning_rate": 0.00019561327561327564,
      "loss": 0.7378,
      "step": 860
    },
    {
      "epoch": 6.308243727598566,
      "grad_norm": 3.517230987548828,
      "learning_rate": 0.0001954978354978355,
      "loss": 0.7511,
      "step": 880
    },
    {
      "epoch": 6.451612903225806,
      "grad_norm": 4.085216999053955,
      "learning_rate": 0.0001953823953823954,
      "loss": 0.7659,
      "step": 900
    },
    {
      "epoch": 6.594982078853047,
      "grad_norm": 4.3206634521484375,
      "learning_rate": 0.00019526695526695528,
      "loss": 0.8262,
      "step": 920
    },
    {
      "epoch": 6.738351254480286,
      "grad_norm": 3.6010711193084717,
      "learning_rate": 0.00019515151515151516,
      "loss": 0.8121,
      "step": 940
    },
    {
      "epoch": 6.881720430107527,
      "grad_norm": 4.308924198150635,
      "learning_rate": 0.00019503607503607503,
      "loss": 0.83,
      "step": 960
    },
    {
      "epoch": 7.025089605734767,
      "grad_norm": 3.5463600158691406,
      "learning_rate": 0.00019492063492063493,
      "loss": 0.7746,
      "step": 980
    },
    {
      "epoch": 7.168458781362007,
      "grad_norm": 3.1309635639190674,
      "learning_rate": 0.0001948051948051948,
      "loss": 0.6994,
      "step": 1000
    },
    {
      "epoch": 7.311827956989247,
      "grad_norm": 4.161702632904053,
      "learning_rate": 0.0001946897546897547,
      "loss": 0.6635,
      "step": 1020
    },
    {
      "epoch": 7.455197132616488,
      "grad_norm": 3.9364986419677734,
      "learning_rate": 0.00019457431457431458,
      "loss": 0.7223,
      "step": 1040
    },
    {
      "epoch": 7.598566308243727,
      "grad_norm": 3.674201250076294,
      "learning_rate": 0.00019445887445887448,
      "loss": 0.737,
      "step": 1060
    },
    {
      "epoch": 7.741935483870968,
      "grad_norm": 5.211554050445557,
      "learning_rate": 0.00019434343434343435,
      "loss": 0.6624,
      "step": 1080
    },
    {
      "epoch": 7.885304659498208,
      "grad_norm": 3.496736764907837,
      "learning_rate": 0.00019422799422799425,
      "loss": 0.6788,
      "step": 1100
    },
    {
      "epoch": 8.028673835125447,
      "grad_norm": 4.182429790496826,
      "learning_rate": 0.0001941125541125541,
      "loss": 0.6958,
      "step": 1120
    },
    {
      "epoch": 8.172043010752688,
      "grad_norm": 3.506474018096924,
      "learning_rate": 0.000193997113997114,
      "loss": 0.5827,
      "step": 1140
    },
    {
      "epoch": 8.315412186379929,
      "grad_norm": 4.012503623962402,
      "learning_rate": 0.00019388167388167387,
      "loss": 0.5956,
      "step": 1160
    },
    {
      "epoch": 8.45878136200717,
      "grad_norm": 4.3530497550964355,
      "learning_rate": 0.00019376623376623377,
      "loss": 0.6149,
      "step": 1180
    },
    {
      "epoch": 8.602150537634408,
      "grad_norm": 3.8034679889678955,
      "learning_rate": 0.00019365079365079365,
      "loss": 0.6482,
      "step": 1200
    },
    {
      "epoch": 8.745519713261649,
      "grad_norm": 4.107527256011963,
      "learning_rate": 0.00019353535353535355,
      "loss": 0.601,
      "step": 1220
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 3.9704201221466064,
      "learning_rate": 0.00019341991341991342,
      "loss": 0.6424,
      "step": 1240
    },
    {
      "epoch": 9.03225806451613,
      "grad_norm": 5.5366315841674805,
      "learning_rate": 0.00019330447330447332,
      "loss": 0.6291,
      "step": 1260
    },
    {
      "epoch": 9.175627240143369,
      "grad_norm": 4.817638397216797,
      "learning_rate": 0.0001931890331890332,
      "loss": 0.5392,
      "step": 1280
    },
    {
      "epoch": 9.31899641577061,
      "grad_norm": 4.172895431518555,
      "learning_rate": 0.00019307359307359307,
      "loss": 0.5134,
      "step": 1300
    },
    {
      "epoch": 9.46236559139785,
      "grad_norm": 4.357290267944336,
      "learning_rate": 0.00019295815295815297,
      "loss": 0.5292,
      "step": 1320
    },
    {
      "epoch": 9.60573476702509,
      "grad_norm": 4.0718512535095215,
      "learning_rate": 0.00019284271284271284,
      "loss": 0.5353,
      "step": 1340
    },
    {
      "epoch": 9.74910394265233,
      "grad_norm": 4.535477638244629,
      "learning_rate": 0.00019272727272727274,
      "loss": 0.5357,
      "step": 1360
    },
    {
      "epoch": 9.89247311827957,
      "grad_norm": 5.310904026031494,
      "learning_rate": 0.00019261183261183262,
      "loss": 0.5955,
      "step": 1380
    },
    {
      "epoch": 10.03584229390681,
      "grad_norm": 4.394368648529053,
      "learning_rate": 0.00019249639249639252,
      "loss": 0.5394,
      "step": 1400
    },
    {
      "epoch": 10.17921146953405,
      "grad_norm": 4.731152057647705,
      "learning_rate": 0.0001923809523809524,
      "loss": 0.4465,
      "step": 1420
    },
    {
      "epoch": 10.32258064516129,
      "grad_norm": 4.1888041496276855,
      "learning_rate": 0.0001922655122655123,
      "loss": 0.4734,
      "step": 1440
    },
    {
      "epoch": 10.46594982078853,
      "grad_norm": 4.412322044372559,
      "learning_rate": 0.00019215007215007217,
      "loss": 0.4749,
      "step": 1460
    },
    {
      "epoch": 10.609318996415771,
      "grad_norm": 4.3348612785339355,
      "learning_rate": 0.00019203463203463204,
      "loss": 0.4751,
      "step": 1480
    },
    {
      "epoch": 10.75268817204301,
      "grad_norm": 4.405585765838623,
      "learning_rate": 0.00019191919191919191,
      "loss": 0.5176,
      "step": 1500
    },
    {
      "epoch": 10.89605734767025,
      "grad_norm": 3.9980640411376953,
      "learning_rate": 0.00019180375180375181,
      "loss": 0.5078,
      "step": 1520
    },
    {
      "epoch": 11.039426523297491,
      "grad_norm": 4.7061028480529785,
      "learning_rate": 0.0001916883116883117,
      "loss": 0.4775,
      "step": 1540
    },
    {
      "epoch": 11.182795698924732,
      "grad_norm": 5.084146022796631,
      "learning_rate": 0.0001915728715728716,
      "loss": 0.3826,
      "step": 1560
    },
    {
      "epoch": 11.32616487455197,
      "grad_norm": 3.631321907043457,
      "learning_rate": 0.00019145743145743146,
      "loss": 0.4259,
      "step": 1580
    },
    {
      "epoch": 11.469534050179211,
      "grad_norm": 5.307252407073975,
      "learning_rate": 0.00019134199134199136,
      "loss": 0.409,
      "step": 1600
    },
    {
      "epoch": 11.612903225806452,
      "grad_norm": 4.273385047912598,
      "learning_rate": 0.00019122655122655124,
      "loss": 0.4716,
      "step": 1620
    },
    {
      "epoch": 11.756272401433693,
      "grad_norm": 3.9993953704833984,
      "learning_rate": 0.00019111111111111114,
      "loss": 0.4482,
      "step": 1640
    },
    {
      "epoch": 11.899641577060931,
      "grad_norm": 4.456310749053955,
      "learning_rate": 0.00019099567099567098,
      "loss": 0.4285,
      "step": 1660
    },
    {
      "epoch": 12.043010752688172,
      "grad_norm": 3.430715322494507,
      "learning_rate": 0.00019088023088023088,
      "loss": 0.4146,
      "step": 1680
    },
    {
      "epoch": 12.186379928315413,
      "grad_norm": 4.638120174407959,
      "learning_rate": 0.00019076479076479076,
      "loss": 0.3632,
      "step": 1700
    },
    {
      "epoch": 12.329749103942653,
      "grad_norm": 4.701437950134277,
      "learning_rate": 0.00019064935064935066,
      "loss": 0.3433,
      "step": 1720
    },
    {
      "epoch": 12.473118279569892,
      "grad_norm": 4.129784107208252,
      "learning_rate": 0.00019053391053391053,
      "loss": 0.3695,
      "step": 1740
    },
    {
      "epoch": 12.616487455197133,
      "grad_norm": 4.153416633605957,
      "learning_rate": 0.00019041847041847043,
      "loss": 0.3807,
      "step": 1760
    },
    {
      "epoch": 12.759856630824373,
      "grad_norm": 4.782583713531494,
      "learning_rate": 0.0001903030303030303,
      "loss": 0.4431,
      "step": 1780
    },
    {
      "epoch": 12.903225806451612,
      "grad_norm": 3.8527839183807373,
      "learning_rate": 0.0001901875901875902,
      "loss": 0.3609,
      "step": 1800
    },
    {
      "epoch": 13.046594982078853,
      "grad_norm": 3.863771438598633,
      "learning_rate": 0.00019007215007215008,
      "loss": 0.3711,
      "step": 1820
    },
    {
      "epoch": 13.189964157706093,
      "grad_norm": 4.283368110656738,
      "learning_rate": 0.00018995670995670995,
      "loss": 0.3066,
      "step": 1840
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 4.724956512451172,
      "learning_rate": 0.00018984126984126985,
      "loss": 0.3394,
      "step": 1860
    },
    {
      "epoch": 13.476702508960573,
      "grad_norm": 4.810844421386719,
      "learning_rate": 0.00018972582972582973,
      "loss": 0.31,
      "step": 1880
    },
    {
      "epoch": 13.620071684587813,
      "grad_norm": 4.2984466552734375,
      "learning_rate": 0.00018961038961038963,
      "loss": 0.3296,
      "step": 1900
    },
    {
      "epoch": 13.763440860215054,
      "grad_norm": 4.326981067657471,
      "learning_rate": 0.0001894949494949495,
      "loss": 0.3428,
      "step": 1920
    },
    {
      "epoch": 13.906810035842295,
      "grad_norm": 3.757113218307495,
      "learning_rate": 0.0001893795093795094,
      "loss": 0.3637,
      "step": 1940
    },
    {
      "epoch": 14.050179211469533,
      "grad_norm": 4.3763861656188965,
      "learning_rate": 0.00018926406926406928,
      "loss": 0.3518,
      "step": 1960
    },
    {
      "epoch": 14.193548387096774,
      "grad_norm": 4.564425468444824,
      "learning_rate": 0.00018914862914862918,
      "loss": 0.2795,
      "step": 1980
    },
    {
      "epoch": 14.336917562724015,
      "grad_norm": 4.2235283851623535,
      "learning_rate": 0.00018903318903318905,
      "loss": 0.2836,
      "step": 2000
    },
    {
      "epoch": 14.480286738351255,
      "grad_norm": 5.298072338104248,
      "learning_rate": 0.00018891774891774892,
      "loss": 0.2968,
      "step": 2020
    },
    {
      "epoch": 14.623655913978494,
      "grad_norm": 4.486312389373779,
      "learning_rate": 0.0001888023088023088,
      "loss": 0.2891,
      "step": 2040
    },
    {
      "epoch": 14.767025089605735,
      "grad_norm": 9.727706909179688,
      "learning_rate": 0.0001886868686868687,
      "loss": 0.3302,
      "step": 2060
    },
    {
      "epoch": 14.910394265232975,
      "grad_norm": 4.170314788818359,
      "learning_rate": 0.00018857142857142857,
      "loss": 0.3139,
      "step": 2080
    },
    {
      "epoch": 15.053763440860216,
      "grad_norm": 4.840059757232666,
      "learning_rate": 0.00018845598845598847,
      "loss": 0.2774,
      "step": 2100
    },
    {
      "epoch": 15.197132616487455,
      "grad_norm": 4.770188808441162,
      "learning_rate": 0.00018834054834054834,
      "loss": 0.2681,
      "step": 2120
    },
    {
      "epoch": 15.340501792114695,
      "grad_norm": 4.672384262084961,
      "learning_rate": 0.00018822510822510825,
      "loss": 0.2499,
      "step": 2140
    },
    {
      "epoch": 15.483870967741936,
      "grad_norm": 4.933983325958252,
      "learning_rate": 0.00018810966810966812,
      "loss": 0.2398,
      "step": 2160
    },
    {
      "epoch": 15.627240143369175,
      "grad_norm": 3.96673583984375,
      "learning_rate": 0.00018799422799422802,
      "loss": 0.2917,
      "step": 2180
    },
    {
      "epoch": 15.770609318996415,
      "grad_norm": 4.215606212615967,
      "learning_rate": 0.0001878787878787879,
      "loss": 0.272,
      "step": 2200
    },
    {
      "epoch": 15.913978494623656,
      "grad_norm": 4.834962844848633,
      "learning_rate": 0.00018776334776334777,
      "loss": 0.2717,
      "step": 2220
    },
    {
      "epoch": 16.057347670250895,
      "grad_norm": 3.783409595489502,
      "learning_rate": 0.00018764790764790764,
      "loss": 0.2594,
      "step": 2240
    },
    {
      "epoch": 16.200716845878137,
      "grad_norm": 3.0773093700408936,
      "learning_rate": 0.00018753246753246754,
      "loss": 0.2007,
      "step": 2260
    },
    {
      "epoch": 16.344086021505376,
      "grad_norm": 3.63726806640625,
      "learning_rate": 0.00018741702741702741,
      "loss": 0.2213,
      "step": 2280
    },
    {
      "epoch": 16.487455197132615,
      "grad_norm": 2.6716272830963135,
      "learning_rate": 0.00018730158730158731,
      "loss": 0.2239,
      "step": 2300
    },
    {
      "epoch": 16.630824372759857,
      "grad_norm": 3.748936414718628,
      "learning_rate": 0.0001871861471861472,
      "loss": 0.2455,
      "step": 2320
    },
    {
      "epoch": 16.774193548387096,
      "grad_norm": 4.2871785163879395,
      "learning_rate": 0.0001870707070707071,
      "loss": 0.2833,
      "step": 2340
    },
    {
      "epoch": 16.91756272401434,
      "grad_norm": 4.30325984954834,
      "learning_rate": 0.00018695526695526696,
      "loss": 0.2538,
      "step": 2360
    },
    {
      "epoch": 17.060931899641577,
      "grad_norm": 3.659284830093384,
      "learning_rate": 0.00018683982683982686,
      "loss": 0.2033,
      "step": 2380
    },
    {
      "epoch": 17.204301075268816,
      "grad_norm": 3.3626983165740967,
      "learning_rate": 0.00018672438672438674,
      "loss": 0.1804,
      "step": 2400
    },
    {
      "epoch": 17.34767025089606,
      "grad_norm": 4.500549793243408,
      "learning_rate": 0.0001866089466089466,
      "loss": 0.2,
      "step": 2420
    },
    {
      "epoch": 17.491039426523297,
      "grad_norm": 3.9919979572296143,
      "learning_rate": 0.0001864935064935065,
      "loss": 0.2469,
      "step": 2440
    },
    {
      "epoch": 17.634408602150536,
      "grad_norm": 3.7051095962524414,
      "learning_rate": 0.00018637806637806638,
      "loss": 0.1967,
      "step": 2460
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 5.194777011871338,
      "learning_rate": 0.00018626262626262628,
      "loss": 0.2353,
      "step": 2480
    },
    {
      "epoch": 17.921146953405017,
      "grad_norm": 4.292696952819824,
      "learning_rate": 0.00018614718614718616,
      "loss": 0.2291,
      "step": 2500
    },
    {
      "epoch": 18.06451612903226,
      "grad_norm": 5.281682968139648,
      "learning_rate": 0.00018603174603174606,
      "loss": 0.2102,
      "step": 2520
    },
    {
      "epoch": 18.2078853046595,
      "grad_norm": 4.829092502593994,
      "learning_rate": 0.00018591630591630593,
      "loss": 0.1822,
      "step": 2540
    },
    {
      "epoch": 18.351254480286737,
      "grad_norm": 4.637266635894775,
      "learning_rate": 0.00018580086580086583,
      "loss": 0.1774,
      "step": 2560
    },
    {
      "epoch": 18.49462365591398,
      "grad_norm": 4.328069686889648,
      "learning_rate": 0.00018568542568542568,
      "loss": 0.1867,
      "step": 2580
    },
    {
      "epoch": 18.63799283154122,
      "grad_norm": 3.5773091316223145,
      "learning_rate": 0.00018556998556998558,
      "loss": 0.1957,
      "step": 2600
    },
    {
      "epoch": 18.781362007168457,
      "grad_norm": 3.868898391723633,
      "learning_rate": 0.00018545454545454545,
      "loss": 0.2005,
      "step": 2620
    },
    {
      "epoch": 18.9247311827957,
      "grad_norm": 3.5077199935913086,
      "learning_rate": 0.00018533910533910535,
      "loss": 0.2185,
      "step": 2640
    },
    {
      "epoch": 19.06810035842294,
      "grad_norm": 4.56404972076416,
      "learning_rate": 0.00018522366522366523,
      "loss": 0.1939,
      "step": 2660
    },
    {
      "epoch": 19.211469534050178,
      "grad_norm": 3.897749662399292,
      "learning_rate": 0.00018510822510822513,
      "loss": 0.1669,
      "step": 2680
    },
    {
      "epoch": 19.35483870967742,
      "grad_norm": 5.491626262664795,
      "learning_rate": 0.000184992784992785,
      "loss": 0.1685,
      "step": 2700
    },
    {
      "epoch": 19.49820788530466,
      "grad_norm": 4.0989813804626465,
      "learning_rate": 0.0001848773448773449,
      "loss": 0.1582,
      "step": 2720
    },
    {
      "epoch": 19.6415770609319,
      "grad_norm": 3.87003755569458,
      "learning_rate": 0.00018476190476190478,
      "loss": 0.1644,
      "step": 2740
    },
    {
      "epoch": 19.78494623655914,
      "grad_norm": 4.180914402008057,
      "learning_rate": 0.00018464646464646465,
      "loss": 0.1731,
      "step": 2760
    },
    {
      "epoch": 19.92831541218638,
      "grad_norm": 4.430409908294678,
      "learning_rate": 0.00018453102453102452,
      "loss": 0.191,
      "step": 2780
    },
    {
      "epoch": 20.07168458781362,
      "grad_norm": 4.21690559387207,
      "learning_rate": 0.00018441558441558442,
      "loss": 0.1707,
      "step": 2800
    },
    {
      "epoch": 20.21505376344086,
      "grad_norm": 3.604231595993042,
      "learning_rate": 0.0001843001443001443,
      "loss": 0.1358,
      "step": 2820
    },
    {
      "epoch": 20.3584229390681,
      "grad_norm": 3.138463020324707,
      "learning_rate": 0.0001841847041847042,
      "loss": 0.1588,
      "step": 2840
    },
    {
      "epoch": 20.50179211469534,
      "grad_norm": 4.0394978523254395,
      "learning_rate": 0.00018406926406926407,
      "loss": 0.1727,
      "step": 2860
    },
    {
      "epoch": 20.64516129032258,
      "grad_norm": 3.6658740043640137,
      "learning_rate": 0.00018395382395382397,
      "loss": 0.1759,
      "step": 2880
    },
    {
      "epoch": 20.788530465949822,
      "grad_norm": 3.8693015575408936,
      "learning_rate": 0.00018383838383838384,
      "loss": 0.1486,
      "step": 2900
    },
    {
      "epoch": 20.93189964157706,
      "grad_norm": 4.257044315338135,
      "learning_rate": 0.00018372294372294375,
      "loss": 0.1338,
      "step": 2920
    },
    {
      "epoch": 21.0752688172043,
      "grad_norm": 2.618070125579834,
      "learning_rate": 0.00018360750360750362,
      "loss": 0.1527,
      "step": 2940
    },
    {
      "epoch": 21.218637992831543,
      "grad_norm": 4.323324203491211,
      "learning_rate": 0.0001834920634920635,
      "loss": 0.1519,
      "step": 2960
    },
    {
      "epoch": 21.36200716845878,
      "grad_norm": 3.53774356842041,
      "learning_rate": 0.0001833766233766234,
      "loss": 0.1175,
      "step": 2980
    },
    {
      "epoch": 21.50537634408602,
      "grad_norm": 5.719931125640869,
      "learning_rate": 0.00018326118326118327,
      "loss": 0.1323,
      "step": 3000
    },
    {
      "epoch": 21.648745519713263,
      "grad_norm": 4.327314376831055,
      "learning_rate": 0.00018314574314574317,
      "loss": 0.1344,
      "step": 3020
    },
    {
      "epoch": 21.7921146953405,
      "grad_norm": 4.0956501960754395,
      "learning_rate": 0.00018303030303030304,
      "loss": 0.1339,
      "step": 3040
    },
    {
      "epoch": 21.93548387096774,
      "grad_norm": 4.5580830574035645,
      "learning_rate": 0.00018291486291486294,
      "loss": 0.1707,
      "step": 3060
    },
    {
      "epoch": 22.078853046594983,
      "grad_norm": 2.9596970081329346,
      "learning_rate": 0.00018279942279942281,
      "loss": 0.137,
      "step": 3080
    },
    {
      "epoch": 22.22222222222222,
      "grad_norm": 3.2811834812164307,
      "learning_rate": 0.00018268398268398272,
      "loss": 0.1299,
      "step": 3100
    },
    {
      "epoch": 22.365591397849464,
      "grad_norm": 2.986440420150757,
      "learning_rate": 0.00018256854256854256,
      "loss": 0.1318,
      "step": 3120
    },
    {
      "epoch": 22.508960573476703,
      "grad_norm": 3.6425325870513916,
      "learning_rate": 0.00018245310245310246,
      "loss": 0.1234,
      "step": 3140
    },
    {
      "epoch": 22.65232974910394,
      "grad_norm": 4.4187846183776855,
      "learning_rate": 0.00018233766233766234,
      "loss": 0.1333,
      "step": 3160
    },
    {
      "epoch": 22.795698924731184,
      "grad_norm": 4.019857406616211,
      "learning_rate": 0.00018222222222222224,
      "loss": 0.126,
      "step": 3180
    },
    {
      "epoch": 22.939068100358423,
      "grad_norm": 3.8513457775115967,
      "learning_rate": 0.0001821067821067821,
      "loss": 0.1346,
      "step": 3200
    },
    {
      "epoch": 23.08243727598566,
      "grad_norm": 2.8631300926208496,
      "learning_rate": 0.000181991341991342,
      "loss": 0.1151,
      "step": 3220
    },
    {
      "epoch": 23.225806451612904,
      "grad_norm": 2.056180477142334,
      "learning_rate": 0.00018187590187590188,
      "loss": 0.093,
      "step": 3240
    },
    {
      "epoch": 23.369175627240143,
      "grad_norm": 4.408998966217041,
      "learning_rate": 0.00018176046176046178,
      "loss": 0.1404,
      "step": 3260
    },
    {
      "epoch": 23.512544802867385,
      "grad_norm": 4.436832904815674,
      "learning_rate": 0.00018164502164502166,
      "loss": 0.1199,
      "step": 3280
    },
    {
      "epoch": 23.655913978494624,
      "grad_norm": 3.343065023422241,
      "learning_rate": 0.00018152958152958153,
      "loss": 0.1341,
      "step": 3300
    },
    {
      "epoch": 23.799283154121863,
      "grad_norm": 3.611084461212158,
      "learning_rate": 0.0001814141414141414,
      "loss": 0.1149,
      "step": 3320
    },
    {
      "epoch": 23.942652329749105,
      "grad_norm": 5.188615322113037,
      "learning_rate": 0.0001812987012987013,
      "loss": 0.1342,
      "step": 3340
    },
    {
      "epoch": 24.086021505376344,
      "grad_norm": 3.88657546043396,
      "learning_rate": 0.00018118326118326118,
      "loss": 0.1084,
      "step": 3360
    },
    {
      "epoch": 24.229390681003583,
      "grad_norm": 2.9031708240509033,
      "learning_rate": 0.00018106782106782108,
      "loss": 0.1054,
      "step": 3380
    },
    {
      "epoch": 24.372759856630825,
      "grad_norm": 3.283027172088623,
      "learning_rate": 0.00018095238095238095,
      "loss": 0.1049,
      "step": 3400
    },
    {
      "epoch": 24.516129032258064,
      "grad_norm": 3.191681385040283,
      "learning_rate": 0.00018083694083694085,
      "loss": 0.114,
      "step": 3420
    },
    {
      "epoch": 24.659498207885306,
      "grad_norm": 2.726076602935791,
      "learning_rate": 0.00018072150072150073,
      "loss": 0.1119,
      "step": 3440
    },
    {
      "epoch": 24.802867383512545,
      "grad_norm": 3.3401198387145996,
      "learning_rate": 0.00018060606060606063,
      "loss": 0.1098,
      "step": 3460
    },
    {
      "epoch": 24.946236559139784,
      "grad_norm": 3.8354945182800293,
      "learning_rate": 0.0001804906204906205,
      "loss": 0.1212,
      "step": 3480
    },
    {
      "epoch": 25.089605734767026,
      "grad_norm": 3.6096622943878174,
      "learning_rate": 0.00018037518037518038,
      "loss": 0.0942,
      "step": 3500
    },
    {
      "epoch": 25.232974910394265,
      "grad_norm": 2.9408082962036133,
      "learning_rate": 0.00018025974025974025,
      "loss": 0.0873,
      "step": 3520
    },
    {
      "epoch": 25.376344086021504,
      "grad_norm": 2.9690041542053223,
      "learning_rate": 0.00018014430014430015,
      "loss": 0.0999,
      "step": 3540
    },
    {
      "epoch": 25.519713261648747,
      "grad_norm": 2.772616147994995,
      "learning_rate": 0.00018002886002886002,
      "loss": 0.1108,
      "step": 3560
    },
    {
      "epoch": 25.663082437275985,
      "grad_norm": 2.993304491043091,
      "learning_rate": 0.00017991341991341992,
      "loss": 0.1065,
      "step": 3580
    },
    {
      "epoch": 25.806451612903224,
      "grad_norm": 3.176945209503174,
      "learning_rate": 0.0001797979797979798,
      "loss": 0.1009,
      "step": 3600
    },
    {
      "epoch": 25.949820788530467,
      "grad_norm": 3.697052001953125,
      "learning_rate": 0.0001796825396825397,
      "loss": 0.135,
      "step": 3620
    },
    {
      "epoch": 26.093189964157705,
      "grad_norm": 2.9682657718658447,
      "learning_rate": 0.00017956709956709957,
      "loss": 0.0872,
      "step": 3640
    },
    {
      "epoch": 26.236559139784948,
      "grad_norm": 2.0252315998077393,
      "learning_rate": 0.00017945165945165947,
      "loss": 0.0871,
      "step": 3660
    },
    {
      "epoch": 26.379928315412187,
      "grad_norm": 2.12572979927063,
      "learning_rate": 0.00017933621933621934,
      "loss": 0.0881,
      "step": 3680
    },
    {
      "epoch": 26.523297491039425,
      "grad_norm": 3.1188809871673584,
      "learning_rate": 0.00017922077922077922,
      "loss": 0.0943,
      "step": 3700
    },
    {
      "epoch": 26.666666666666668,
      "grad_norm": 3.05329966545105,
      "learning_rate": 0.00017910533910533912,
      "loss": 0.1057,
      "step": 3720
    },
    {
      "epoch": 26.810035842293907,
      "grad_norm": 3.4610469341278076,
      "learning_rate": 0.000178989898989899,
      "loss": 0.0933,
      "step": 3740
    },
    {
      "epoch": 26.953405017921146,
      "grad_norm": 4.512834072113037,
      "learning_rate": 0.0001788744588744589,
      "loss": 0.1219,
      "step": 3760
    },
    {
      "epoch": 27.096774193548388,
      "grad_norm": 6.541393756866455,
      "learning_rate": 0.00017875901875901877,
      "loss": 0.0882,
      "step": 3780
    },
    {
      "epoch": 27.240143369175627,
      "grad_norm": 3.331212043762207,
      "learning_rate": 0.00017864357864357867,
      "loss": 0.081,
      "step": 3800
    },
    {
      "epoch": 27.38351254480287,
      "grad_norm": 2.9885828495025635,
      "learning_rate": 0.00017852813852813854,
      "loss": 0.0919,
      "step": 3820
    },
    {
      "epoch": 27.526881720430108,
      "grad_norm": 3.2344679832458496,
      "learning_rate": 0.00017841269841269844,
      "loss": 0.0774,
      "step": 3840
    },
    {
      "epoch": 27.670250896057347,
      "grad_norm": 4.0804290771484375,
      "learning_rate": 0.0001782972582972583,
      "loss": 0.1038,
      "step": 3860
    },
    {
      "epoch": 27.81362007168459,
      "grad_norm": 4.064199924468994,
      "learning_rate": 0.0001781818181818182,
      "loss": 0.0962,
      "step": 3880
    },
    {
      "epoch": 27.956989247311828,
      "grad_norm": 3.9135751724243164,
      "learning_rate": 0.00017806637806637806,
      "loss": 0.0937,
      "step": 3900
    },
    {
      "epoch": 28.100358422939067,
      "grad_norm": 2.8442928791046143,
      "learning_rate": 0.00017795093795093796,
      "loss": 0.0873,
      "step": 3920
    },
    {
      "epoch": 28.24372759856631,
      "grad_norm": 2.8450095653533936,
      "learning_rate": 0.00017783549783549784,
      "loss": 0.0865,
      "step": 3940
    },
    {
      "epoch": 28.387096774193548,
      "grad_norm": 5.168785572052002,
      "learning_rate": 0.00017772005772005774,
      "loss": 0.0819,
      "step": 3960
    },
    {
      "epoch": 28.530465949820787,
      "grad_norm": 2.829857587814331,
      "learning_rate": 0.0001776046176046176,
      "loss": 0.0777,
      "step": 3980
    },
    {
      "epoch": 28.67383512544803,
      "grad_norm": 3.645796060562134,
      "learning_rate": 0.0001774891774891775,
      "loss": 0.0784,
      "step": 4000
    },
    {
      "epoch": 28.817204301075268,
      "grad_norm": 2.782454252243042,
      "learning_rate": 0.00017737373737373738,
      "loss": 0.0907,
      "step": 4020
    },
    {
      "epoch": 28.96057347670251,
      "grad_norm": 3.160844564437866,
      "learning_rate": 0.00017725829725829726,
      "loss": 0.0893,
      "step": 4040
    },
    {
      "epoch": 29.10394265232975,
      "grad_norm": 2.9098074436187744,
      "learning_rate": 0.00017714285714285713,
      "loss": 0.0696,
      "step": 4060
    },
    {
      "epoch": 29.247311827956988,
      "grad_norm": 2.2926838397979736,
      "learning_rate": 0.00017702741702741703,
      "loss": 0.0762,
      "step": 4080
    },
    {
      "epoch": 29.39068100358423,
      "grad_norm": 3.952270030975342,
      "learning_rate": 0.0001769119769119769,
      "loss": 0.0754,
      "step": 4100
    },
    {
      "epoch": 29.53405017921147,
      "grad_norm": 2.679051160812378,
      "learning_rate": 0.0001767965367965368,
      "loss": 0.0753,
      "step": 4120
    },
    {
      "epoch": 29.677419354838708,
      "grad_norm": 3.0351693630218506,
      "learning_rate": 0.00017668109668109668,
      "loss": 0.0889,
      "step": 4140
    },
    {
      "epoch": 29.82078853046595,
      "grad_norm": 2.2097012996673584,
      "learning_rate": 0.00017656565656565658,
      "loss": 0.0882,
      "step": 4160
    },
    {
      "epoch": 29.96415770609319,
      "grad_norm": 3.2388463020324707,
      "learning_rate": 0.00017645021645021645,
      "loss": 0.0828,
      "step": 4180
    },
    {
      "epoch": 30.107526881720432,
      "grad_norm": 2.688284158706665,
      "learning_rate": 0.00017633477633477635,
      "loss": 0.0639,
      "step": 4200
    },
    {
      "epoch": 30.25089605734767,
      "grad_norm": 3.341397523880005,
      "learning_rate": 0.00017621933621933623,
      "loss": 0.0651,
      "step": 4220
    },
    {
      "epoch": 30.39426523297491,
      "grad_norm": 3.0759265422821045,
      "learning_rate": 0.0001761038961038961,
      "loss": 0.0728,
      "step": 4240
    },
    {
      "epoch": 30.537634408602152,
      "grad_norm": 2.7288103103637695,
      "learning_rate": 0.000175988455988456,
      "loss": 0.0719,
      "step": 4260
    },
    {
      "epoch": 30.68100358422939,
      "grad_norm": 3.3915891647338867,
      "learning_rate": 0.00017587301587301588,
      "loss": 0.0758,
      "step": 4280
    },
    {
      "epoch": 30.82437275985663,
      "grad_norm": 3.2848050594329834,
      "learning_rate": 0.00017575757575757578,
      "loss": 0.0842,
      "step": 4300
    },
    {
      "epoch": 30.967741935483872,
      "grad_norm": 3.1872403621673584,
      "learning_rate": 0.00017564213564213565,
      "loss": 0.0895,
      "step": 4320
    },
    {
      "epoch": 31.11111111111111,
      "grad_norm": 2.675920248031616,
      "learning_rate": 0.00017552669552669555,
      "loss": 0.067,
      "step": 4340
    },
    {
      "epoch": 31.25448028673835,
      "grad_norm": 3.840548276901245,
      "learning_rate": 0.00017541125541125542,
      "loss": 0.0614,
      "step": 4360
    },
    {
      "epoch": 31.397849462365592,
      "grad_norm": 3.0870628356933594,
      "learning_rate": 0.00017529581529581532,
      "loss": 0.0726,
      "step": 4380
    },
    {
      "epoch": 31.54121863799283,
      "grad_norm": 3.956441640853882,
      "learning_rate": 0.00017518037518037517,
      "loss": 0.0673,
      "step": 4400
    },
    {
      "epoch": 31.684587813620073,
      "grad_norm": 2.798611640930176,
      "learning_rate": 0.00017506493506493507,
      "loss": 0.0755,
      "step": 4420
    },
    {
      "epoch": 31.827956989247312,
      "grad_norm": 3.071885585784912,
      "learning_rate": 0.00017494949494949494,
      "loss": 0.0697,
      "step": 4440
    },
    {
      "epoch": 31.97132616487455,
      "grad_norm": 4.090572834014893,
      "learning_rate": 0.00017483405483405485,
      "loss": 0.0833,
      "step": 4460
    },
    {
      "epoch": 32.11469534050179,
      "grad_norm": 3.011218309402466,
      "learning_rate": 0.00017471861471861472,
      "loss": 0.0625,
      "step": 4480
    },
    {
      "epoch": 32.25806451612903,
      "grad_norm": 2.3517768383026123,
      "learning_rate": 0.00017460317460317462,
      "loss": 0.0658,
      "step": 4500
    },
    {
      "epoch": 32.401433691756274,
      "grad_norm": 2.255727767944336,
      "learning_rate": 0.0001744877344877345,
      "loss": 0.0595,
      "step": 4520
    },
    {
      "epoch": 32.54480286738351,
      "grad_norm": 3.6209774017333984,
      "learning_rate": 0.0001743722943722944,
      "loss": 0.0696,
      "step": 4540
    },
    {
      "epoch": 32.68817204301075,
      "grad_norm": 2.5406808853149414,
      "learning_rate": 0.00017425685425685427,
      "loss": 0.0689,
      "step": 4560
    },
    {
      "epoch": 32.831541218637994,
      "grad_norm": 3.2799172401428223,
      "learning_rate": 0.00017414141414141414,
      "loss": 0.0737,
      "step": 4580
    },
    {
      "epoch": 32.97491039426523,
      "grad_norm": 2.7221455574035645,
      "learning_rate": 0.00017402597402597401,
      "loss": 0.0786,
      "step": 4600
    },
    {
      "epoch": 33.11827956989247,
      "grad_norm": 1.5500837564468384,
      "learning_rate": 0.00017391053391053391,
      "loss": 0.0558,
      "step": 4620
    },
    {
      "epoch": 33.261648745519715,
      "grad_norm": 2.5242912769317627,
      "learning_rate": 0.0001737950937950938,
      "loss": 0.0578,
      "step": 4640
    },
    {
      "epoch": 33.40501792114696,
      "grad_norm": 3.8142666816711426,
      "learning_rate": 0.0001736796536796537,
      "loss": 0.062,
      "step": 4660
    },
    {
      "epoch": 33.54838709677419,
      "grad_norm": 3.5291833877563477,
      "learning_rate": 0.00017356421356421356,
      "loss": 0.0669,
      "step": 4680
    },
    {
      "epoch": 33.691756272401435,
      "grad_norm": 2.1203930377960205,
      "learning_rate": 0.00017344877344877346,
      "loss": 0.0658,
      "step": 4700
    },
    {
      "epoch": 33.83512544802868,
      "grad_norm": 3.0345406532287598,
      "learning_rate": 0.00017333333333333334,
      "loss": 0.0652,
      "step": 4720
    },
    {
      "epoch": 33.97849462365591,
      "grad_norm": 3.3054096698760986,
      "learning_rate": 0.00017321789321789324,
      "loss": 0.0662,
      "step": 4740
    },
    {
      "epoch": 34.121863799283155,
      "grad_norm": 2.796600580215454,
      "learning_rate": 0.0001731024531024531,
      "loss": 0.053,
      "step": 4760
    },
    {
      "epoch": 34.2652329749104,
      "grad_norm": 2.0653796195983887,
      "learning_rate": 0.00017298701298701298,
      "loss": 0.0626,
      "step": 4780
    },
    {
      "epoch": 34.40860215053763,
      "grad_norm": 1.7784148454666138,
      "learning_rate": 0.00017287157287157288,
      "loss": 0.0527,
      "step": 4800
    },
    {
      "epoch": 34.551971326164875,
      "grad_norm": 1.5484881401062012,
      "learning_rate": 0.00017275613275613276,
      "loss": 0.0597,
      "step": 4820
    },
    {
      "epoch": 34.69534050179212,
      "grad_norm": 2.585890769958496,
      "learning_rate": 0.00017264069264069266,
      "loss": 0.0585,
      "step": 4840
    },
    {
      "epoch": 34.83870967741935,
      "grad_norm": 2.978422164916992,
      "learning_rate": 0.00017252525252525253,
      "loss": 0.0629,
      "step": 4860
    },
    {
      "epoch": 34.982078853046595,
      "grad_norm": 2.5631103515625,
      "learning_rate": 0.00017240981240981243,
      "loss": 0.078,
      "step": 4880
    },
    {
      "epoch": 35.12544802867384,
      "grad_norm": 2.0318007469177246,
      "learning_rate": 0.0001722943722943723,
      "loss": 0.0588,
      "step": 4900
    },
    {
      "epoch": 35.26881720430107,
      "grad_norm": 2.345930576324463,
      "learning_rate": 0.0001721789321789322,
      "loss": 0.0569,
      "step": 4920
    },
    {
      "epoch": 35.412186379928315,
      "grad_norm": 3.440155506134033,
      "learning_rate": 0.00017206349206349208,
      "loss": 0.0528,
      "step": 4940
    },
    {
      "epoch": 35.55555555555556,
      "grad_norm": 2.1735332012176514,
      "learning_rate": 0.00017194805194805195,
      "loss": 0.0576,
      "step": 4960
    },
    {
      "epoch": 35.69892473118279,
      "grad_norm": 2.1158435344696045,
      "learning_rate": 0.00017183261183261183,
      "loss": 0.0593,
      "step": 4980
    },
    {
      "epoch": 35.842293906810035,
      "grad_norm": 2.6468334197998047,
      "learning_rate": 0.00017171717171717173,
      "loss": 0.065,
      "step": 5000
    },
    {
      "epoch": 35.98566308243728,
      "grad_norm": 2.7233901023864746,
      "learning_rate": 0.0001716017316017316,
      "loss": 0.0557,
      "step": 5020
    },
    {
      "epoch": 36.12903225806452,
      "grad_norm": 2.840747594833374,
      "learning_rate": 0.0001714862914862915,
      "loss": 0.0483,
      "step": 5040
    },
    {
      "epoch": 36.272401433691755,
      "grad_norm": 2.235607624053955,
      "learning_rate": 0.00017137085137085138,
      "loss": 0.0504,
      "step": 5060
    },
    {
      "epoch": 36.415770609319,
      "grad_norm": 3.1793267726898193,
      "learning_rate": 0.00017125541125541128,
      "loss": 0.055,
      "step": 5080
    },
    {
      "epoch": 36.55913978494624,
      "grad_norm": 2.583932638168335,
      "learning_rate": 0.00017113997113997115,
      "loss": 0.0553,
      "step": 5100
    },
    {
      "epoch": 36.702508960573475,
      "grad_norm": 3.4074461460113525,
      "learning_rate": 0.00017102453102453105,
      "loss": 0.0591,
      "step": 5120
    },
    {
      "epoch": 36.84587813620072,
      "grad_norm": 2.3088393211364746,
      "learning_rate": 0.0001709090909090909,
      "loss": 0.0577,
      "step": 5140
    },
    {
      "epoch": 36.98924731182796,
      "grad_norm": 2.301844358444214,
      "learning_rate": 0.0001707936507936508,
      "loss": 0.0609,
      "step": 5160
    },
    {
      "epoch": 37.132616487455195,
      "grad_norm": 1.6736223697662354,
      "learning_rate": 0.00017067821067821067,
      "loss": 0.0483,
      "step": 5180
    },
    {
      "epoch": 37.27598566308244,
      "grad_norm": 2.2391276359558105,
      "learning_rate": 0.00017056277056277057,
      "loss": 0.044,
      "step": 5200
    },
    {
      "epoch": 37.41935483870968,
      "grad_norm": 4.5583062171936035,
      "learning_rate": 0.00017044733044733044,
      "loss": 0.0529,
      "step": 5220
    },
    {
      "epoch": 37.562724014336915,
      "grad_norm": 2.835345983505249,
      "learning_rate": 0.00017033189033189035,
      "loss": 0.0547,
      "step": 5240
    },
    {
      "epoch": 37.70609318996416,
      "grad_norm": 2.906676769256592,
      "learning_rate": 0.00017021645021645022,
      "loss": 0.0592,
      "step": 5260
    },
    {
      "epoch": 37.8494623655914,
      "grad_norm": 2.9492480754852295,
      "learning_rate": 0.00017010101010101012,
      "loss": 0.0597,
      "step": 5280
    },
    {
      "epoch": 37.992831541218635,
      "grad_norm": 2.3341236114501953,
      "learning_rate": 0.00016998556998557,
      "loss": 0.0597,
      "step": 5300
    },
    {
      "epoch": 38.13620071684588,
      "grad_norm": 2.3535149097442627,
      "learning_rate": 0.00016987012987012987,
      "loss": 0.0455,
      "step": 5320
    },
    {
      "epoch": 38.27956989247312,
      "grad_norm": 2.001614809036255,
      "learning_rate": 0.00016975468975468977,
      "loss": 0.048,
      "step": 5340
    },
    {
      "epoch": 38.422939068100355,
      "grad_norm": 2.9416966438293457,
      "learning_rate": 0.00016963924963924964,
      "loss": 0.0458,
      "step": 5360
    },
    {
      "epoch": 38.5663082437276,
      "grad_norm": 2.1588425636291504,
      "learning_rate": 0.00016952380952380954,
      "loss": 0.0503,
      "step": 5380
    },
    {
      "epoch": 38.70967741935484,
      "grad_norm": 1.9148027896881104,
      "learning_rate": 0.00016940836940836941,
      "loss": 0.056,
      "step": 5400
    },
    {
      "epoch": 38.85304659498208,
      "grad_norm": 3.3233532905578613,
      "learning_rate": 0.00016929292929292932,
      "loss": 0.0541,
      "step": 5420
    },
    {
      "epoch": 38.99641577060932,
      "grad_norm": 2.9712846279144287,
      "learning_rate": 0.0001691774891774892,
      "loss": 0.0584,
      "step": 5440
    },
    {
      "epoch": 39.13978494623656,
      "grad_norm": 3.8493432998657227,
      "learning_rate": 0.0001690620490620491,
      "loss": 0.0408,
      "step": 5460
    },
    {
      "epoch": 39.2831541218638,
      "grad_norm": 2.1640853881835938,
      "learning_rate": 0.00016894660894660896,
      "loss": 0.0521,
      "step": 5480
    },
    {
      "epoch": 39.42652329749104,
      "grad_norm": 3.0871174335479736,
      "learning_rate": 0.00016883116883116884,
      "loss": 0.0434,
      "step": 5500
    }
  ],
  "logging_steps": 20,
  "max_steps": 34750,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 250,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.880725930838835e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}