{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 705,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00851063829787234,
      "grad_norm": 1.2625373601913452,
      "learning_rate": 8.333333333333333e-07,
      "loss": 4.0039777755737305,
      "step": 2
    },
    {
      "epoch": 0.01702127659574468,
      "grad_norm": 0.8850640654563904,
      "learning_rate": 2.4999999999999998e-06,
      "loss": 2.092726945877075,
      "step": 4
    },
    {
      "epoch": 0.02553191489361702,
      "grad_norm": 0.42879173159599304,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.9236807823181152,
      "step": 6
    },
    {
      "epoch": 0.03404255319148936,
      "grad_norm": 0.8568687438964844,
      "learning_rate": 5.833333333333334e-06,
      "loss": 1.7322841882705688,
      "step": 8
    },
    {
      "epoch": 0.0425531914893617,
      "grad_norm": 0.3813435435295105,
      "learning_rate": 7.5e-06,
      "loss": 1.7557320594787598,
      "step": 10
    },
    {
      "epoch": 0.05106382978723404,
      "grad_norm": 3.289834499359131,
      "learning_rate": 9.166666666666668e-06,
      "loss": 2.0681214332580566,
      "step": 12
    },
    {
      "epoch": 0.059574468085106386,
      "grad_norm": 0.8551011085510254,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 1.6496400833129883,
      "step": 14
    },
    {
      "epoch": 0.06808510638297872,
      "grad_norm": 0.6510457396507263,
      "learning_rate": 1.25e-05,
      "loss": 1.78928804397583,
      "step": 16
    },
    {
      "epoch": 0.07659574468085106,
      "grad_norm": 0.9583740830421448,
      "learning_rate": 1.4166666666666666e-05,
      "loss": 1.4100672006607056,
      "step": 18
    },
    {
      "epoch": 0.0851063829787234,
      "grad_norm": 0.25010502338409424,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 1.4046533107757568,
      "step": 20
    },
    {
      "epoch": 0.09361702127659574,
      "grad_norm": 0.2605890929698944,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 1.4094287157058716,
      "step": 22
    },
    {
      "epoch": 0.10212765957446808,
      "grad_norm": 1.3786605596542358,
      "learning_rate": 1.9166666666666667e-05,
      "loss": 1.0651377439498901,
      "step": 24
    },
    {
      "epoch": 0.11063829787234042,
      "grad_norm": 0.5269911289215088,
      "learning_rate": 2.0833333333333333e-05,
      "loss": 1.187571406364441,
      "step": 26
    },
    {
      "epoch": 0.11914893617021277,
      "grad_norm": 0.24545100331306458,
      "learning_rate": 2.25e-05,
      "loss": 0.8422402739524841,
      "step": 28
    },
    {
      "epoch": 0.1276595744680851,
      "grad_norm": 0.3340352475643158,
      "learning_rate": 2.4166666666666667e-05,
      "loss": 1.3952103853225708,
      "step": 30
    },
    {
      "epoch": 0.13617021276595745,
      "grad_norm": 1.0275627374649048,
      "learning_rate": 2.5833333333333336e-05,
      "loss": 0.6625706553459167,
      "step": 32
    },
    {
      "epoch": 0.14468085106382977,
      "grad_norm": 0.27028778195381165,
      "learning_rate": 2.75e-05,
      "loss": 1.233126163482666,
      "step": 34
    },
    {
      "epoch": 0.15319148936170213,
      "grad_norm": 0.3007853627204895,
      "learning_rate": 2.9166666666666666e-05,
      "loss": 1.522036075592041,
      "step": 36
    },
    {
      "epoch": 0.16170212765957448,
      "grad_norm": 0.25563573837280273,
      "learning_rate": 2.9998734788806287e-05,
      "loss": 1.4075392484664917,
      "step": 38
    },
    {
      "epoch": 0.1702127659574468,
      "grad_norm": 0.3948342502117157,
      "learning_rate": 2.9988614605803806e-05,
      "loss": 1.0875834226608276,
      "step": 40
    },
    {
      "epoch": 0.17872340425531916,
      "grad_norm": 0.189285010099411,
      "learning_rate": 2.99683822733885e-05,
      "loss": 1.2082504034042358,
      "step": 42
    },
    {
      "epoch": 0.18723404255319148,
      "grad_norm": 0.6968986988067627,
      "learning_rate": 2.9938053852362484e-05,
      "loss": 1.2259453535079956,
      "step": 44
    },
    {
      "epoch": 0.19574468085106383,
      "grad_norm": 0.2155912071466446,
      "learning_rate": 2.989765341799095e-05,
      "loss": 1.3428634405136108,
      "step": 46
    },
    {
      "epoch": 0.20425531914893616,
      "grad_norm": 1.266939401626587,
      "learning_rate": 2.9847213040890793e-05,
      "loss": 0.7827966213226318,
      "step": 48
    },
    {
      "epoch": 0.2127659574468085,
      "grad_norm": 0.738777756690979,
      "learning_rate": 2.9786772761572335e-05,
      "loss": 0.9074859023094177,
      "step": 50
    },
    {
      "epoch": 0.22127659574468084,
      "grad_norm": 0.2421567142009735,
      "learning_rate": 2.9716380558654445e-05,
      "loss": 1.321467399597168,
      "step": 52
    },
    {
      "epoch": 0.2297872340425532,
      "grad_norm": 0.16079238057136536,
      "learning_rate": 2.9636092310778195e-05,
      "loss": 1.1002169847488403,
      "step": 54
    },
    {
      "epoch": 0.23829787234042554,
      "grad_norm": 1.2075369358062744,
      "learning_rate": 2.954597175224938e-05,
      "loss": 0.8815129399299622,
      "step": 56
    },
    {
      "epoch": 0.24680851063829787,
      "grad_norm": 0.17142559587955475,
      "learning_rate": 2.9446090422445016e-05,
      "loss": 1.0324469804763794,
      "step": 58
    },
    {
      "epoch": 0.2553191489361702,
      "grad_norm": 0.1944112330675125,
      "learning_rate": 2.9336527609024072e-05,
      "loss": 1.3160098791122437,
      "step": 60
    },
    {
      "epoch": 0.26382978723404255,
      "grad_norm": 0.3625730574131012,
      "learning_rate": 2.9217370284987434e-05,
      "loss": 1.2711020708084106,
      "step": 62
    },
    {
      "epoch": 0.2723404255319149,
      "grad_norm": 0.1674957275390625,
      "learning_rate": 2.9088713039637117e-05,
      "loss": 1.2588579654693604,
      "step": 64
    },
    {
      "epoch": 0.28085106382978725,
      "grad_norm": 0.2627532184123993,
      "learning_rate": 2.8950658003489534e-05,
      "loss": 1.371834635734558,
      "step": 66
    },
    {
      "epoch": 0.28936170212765955,
      "grad_norm": 0.1840338557958603,
      "learning_rate": 2.880331476720238e-05,
      "loss": 1.227813720703125,
      "step": 68
    },
    {
      "epoch": 0.2978723404255319,
      "grad_norm": 0.5450706481933594,
      "learning_rate": 2.8646800294579517e-05,
      "loss": 1.4304167032241821,
      "step": 70
    },
    {
      "epoch": 0.30638297872340425,
      "grad_norm": 0.197848379611969,
      "learning_rate": 2.848123882972295e-05,
      "loss": 1.328366756439209,
      "step": 72
    },
    {
      "epoch": 0.3148936170212766,
      "grad_norm": 0.2317323237657547,
      "learning_rate": 2.8306761798405526e-05,
      "loss": 0.8880937695503235,
      "step": 74
    },
    {
      "epoch": 0.32340425531914896,
      "grad_norm": 2.356330156326294,
      "learning_rate": 2.812350770374273e-05,
      "loss": 1.3181480169296265,
      "step": 76
    },
    {
      "epoch": 0.33191489361702126,
      "grad_norm": 0.24451768398284912,
      "learning_rate": 2.793162201624631e-05,
      "loss": 1.2563281059265137,
      "step": 78
    },
    {
      "epoch": 0.3404255319148936,
      "grad_norm": 0.38318371772766113,
      "learning_rate": 2.77312570583471e-05,
      "loss": 0.903084933757782,
      "step": 80
    },
    {
      "epoch": 0.34893617021276596,
      "grad_norm": 0.3278236389160156,
      "learning_rate": 2.752257188347862e-05,
      "loss": 1.2202683687210083,
      "step": 82
    },
    {
      "epoch": 0.3574468085106383,
      "grad_norm": 0.12423935532569885,
      "learning_rate": 2.730573214981751e-05,
      "loss": 1.021187424659729,
      "step": 84
    },
    {
      "epoch": 0.3659574468085106,
      "grad_norm": 0.15058887004852295,
      "learning_rate": 2.7080909988780982e-05,
      "loss": 0.955544650554657,
      "step": 86
    },
    {
      "epoch": 0.37446808510638296,
      "grad_norm": 0.2629765272140503,
      "learning_rate": 2.684828386838569e-05,
      "loss": 1.1776865720748901,
      "step": 88
    },
    {
      "epoch": 0.3829787234042553,
      "grad_norm": 0.5266962647438049,
      "learning_rate": 2.6608038451576528e-05,
      "loss": 0.8726862668991089,
      "step": 90
    },
    {
      "epoch": 0.39148936170212767,
      "grad_norm": 0.20310066640377045,
      "learning_rate": 2.636036444963769e-05,
      "loss": 1.2601391077041626,
      "step": 92
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.21515996754169464,
      "learning_rate": 2.6105458470802563e-05,
      "loss": 1.246147871017456,
      "step": 94
    },
    {
      "epoch": 0.4085106382978723,
      "grad_norm": 0.28013795614242554,
      "learning_rate": 2.5843522864182394e-05,
      "loss": 1.1205939054489136,
      "step": 96
    },
    {
      "epoch": 0.41702127659574467,
      "grad_norm": 0.27731189131736755,
      "learning_rate": 2.557476555913785e-05,
      "loss": 1.029671549797058,
      "step": 98
    },
    {
      "epoch": 0.425531914893617,
      "grad_norm": 0.5619083046913147,
      "learning_rate": 2.5299399900220803e-05,
      "loss": 1.0204272270202637,
      "step": 100
    },
    {
      "epoch": 0.4340425531914894,
      "grad_norm": 0.18958072364330292,
      "learning_rate": 2.5017644477817424e-05,
      "loss": 1.2801953554153442,
      "step": 102
    },
    {
      "epoch": 0.4425531914893617,
      "grad_norm": 0.14192377030849457,
      "learning_rate": 2.47297229546271e-05,
      "loss": 0.9845929741859436,
      "step": 104
    },
    {
      "epoch": 0.451063829787234,
      "grad_norm": 1.0689587593078613,
      "learning_rate": 2.4435863888114814e-05,
      "loss": 0.8531012535095215,
      "step": 106
    },
    {
      "epoch": 0.4595744680851064,
      "grad_norm": 0.22902530431747437,
      "learning_rate": 2.4136300549077976e-05,
      "loss": 1.256201148033142,
      "step": 108
    },
    {
      "epoch": 0.46808510638297873,
      "grad_norm": 0.15747293829917908,
      "learning_rate": 2.3831270736471703e-05,
      "loss": 1.2275390625,
      "step": 110
    },
    {
      "epoch": 0.4765957446808511,
      "grad_norm": 0.2964925467967987,
      "learning_rate": 2.352101658863959e-05,
      "loss": 1.5093128681182861,
      "step": 112
    },
    {
      "epoch": 0.4851063829787234,
      "grad_norm": 0.23254844546318054,
      "learning_rate": 2.32057843910998e-05,
      "loss": 1.0485801696777344,
      "step": 114
    },
    {
      "epoch": 0.49361702127659574,
      "grad_norm": 0.2164650410413742,
      "learning_rate": 2.288582438103903e-05,
      "loss": 1.0241464376449585,
      "step": 116
    },
    {
      "epoch": 0.502127659574468,
      "grad_norm": 0.2963908612728119,
      "learning_rate": 2.256139054866955e-05,
      "loss": 1.259883999824524,
      "step": 118
    },
    {
      "epoch": 0.5106382978723404,
      "grad_norm": 0.3361697494983673,
      "learning_rate": 2.2232740435607067e-05,
      "loss": 0.8736612796783447,
      "step": 120
    },
    {
      "epoch": 0.5191489361702127,
      "grad_norm": 0.2727033495903015,
      "learning_rate": 2.19001349304294e-05,
      "loss": 1.2946780920028687,
      "step": 122
    },
    {
      "epoch": 0.5276595744680851,
      "grad_norm": 0.2193780094385147,
      "learning_rate": 2.156383806157826e-05,
      "loss": 1.2013747692108154,
      "step": 124
    },
    {
      "epoch": 0.5361702127659574,
      "grad_norm": 0.18866108357906342,
      "learning_rate": 2.1224116787768552e-05,
      "loss": 1.2230784893035889,
      "step": 126
    },
    {
      "epoch": 0.5446808510638298,
      "grad_norm": 0.4698742926120758,
      "learning_rate": 2.0881240786071588e-05,
      "loss": 1.0054576396942139,
      "step": 128
    },
    {
      "epoch": 0.5531914893617021,
      "grad_norm": 0.2406734824180603,
      "learning_rate": 2.05354822378404e-05,
      "loss": 1.0494941473007202,
      "step": 130
    },
    {
      "epoch": 0.5617021276595745,
      "grad_norm": 0.20684203505516052,
      "learning_rate": 2.018711561264714e-05,
      "loss": 1.2933894395828247,
      "step": 132
    },
    {
      "epoch": 0.5702127659574469,
      "grad_norm": 0.18240053951740265,
      "learning_rate": 1.9836417450403978e-05,
      "loss": 1.172606110572815,
      "step": 134
    },
    {
      "epoch": 0.5787234042553191,
      "grad_norm": 0.27221980690956116,
      "learning_rate": 1.9483666141840615e-05,
      "loss": 0.9525983929634094,
      "step": 136
    },
    {
      "epoch": 0.5872340425531914,
      "grad_norm": 0.45265427231788635,
      "learning_rate": 1.9129141707512508e-05,
      "loss": 0.6214677095413208,
      "step": 138
    },
    {
      "epoch": 0.5957446808510638,
      "grad_norm": 0.27331340312957764,
      "learning_rate": 1.8773125575515364e-05,
      "loss": 1.0034734010696411,
      "step": 140
    },
    {
      "epoch": 0.6042553191489362,
      "grad_norm": 0.5351057648658752,
      "learning_rate": 1.8415900358082268e-05,
      "loss": 1.2574905157089233,
      "step": 142
    },
    {
      "epoch": 0.6127659574468085,
      "grad_norm": 2.15321946144104,
      "learning_rate": 1.805774962724083e-05,
      "loss": 1.039708137512207,
      "step": 144
    },
    {
      "epoch": 0.6212765957446809,
      "grad_norm": 0.19406694173812866,
      "learning_rate": 1.7698957689708426e-05,
      "loss": 1.1023411750793457,
      "step": 146
    },
    {
      "epoch": 0.6297872340425532,
      "grad_norm": 0.28949815034866333,
      "learning_rate": 1.7339809361204252e-05,
      "loss": 1.2206919193267822,
      "step": 148
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 0.17495916783809662,
      "learning_rate": 1.6980589740357294e-05,
      "loss": 1.327172875404358,
      "step": 150
    },
    {
      "epoch": 0.6468085106382979,
      "grad_norm": 0.3632895052433014,
      "learning_rate": 1.6621583982389707e-05,
      "loss": 1.3268358707427979,
      "step": 152
    },
    {
      "epoch": 0.6553191489361702,
      "grad_norm": 0.3797944486141205,
      "learning_rate": 1.6263077072755326e-05,
      "loss": 0.7203034162521362,
      "step": 154
    },
    {
      "epoch": 0.6638297872340425,
      "grad_norm": 0.3124215602874756,
      "learning_rate": 1.5905353600912898e-05,
      "loss": 0.9463112354278564,
      "step": 156
    },
    {
      "epoch": 0.6723404255319149,
      "grad_norm": 0.20629045367240906,
      "learning_rate": 1.5548697534413646e-05,
      "loss": 0.9798778891563416,
      "step": 158
    },
    {
      "epoch": 0.6808510638297872,
      "grad_norm": 0.17236708104610443,
      "learning_rate": 1.5193391993482582e-05,
      "loss": 1.1495308876037598,
      "step": 160
    },
    {
      "epoch": 0.6893617021276596,
      "grad_norm": 0.17647439241409302,
      "learning_rate": 1.4839719026272377e-05,
      "loss": 1.0284472703933716,
      "step": 162
    },
    {
      "epoch": 0.6978723404255319,
      "grad_norm": 0.21520790457725525,
      "learning_rate": 1.4487959384968272e-05,
      "loss": 0.8173097968101501,
      "step": 164
    },
    {
      "epoch": 0.7063829787234043,
      "grad_norm": 0.22225292026996613,
      "learning_rate": 1.4138392302921813e-05,
      "loss": 0.994484007358551,
      "step": 166
    },
    {
      "epoch": 0.7148936170212766,
      "grad_norm": 0.8280470967292786,
      "learning_rate": 1.3791295272990175e-05,
      "loss": 1.1443301439285278,
      "step": 168
    },
    {
      "epoch": 0.723404255319149,
      "grad_norm": 0.3155645430088043,
      "learning_rate": 1.344694382725718e-05,
      "loss": 1.0067280530929565,
      "step": 170
    },
    {
      "epoch": 0.7319148936170212,
      "grad_norm": 0.269070029258728,
      "learning_rate": 1.3105611318310818e-05,
      "loss": 0.9691381454467773,
      "step": 172
    },
    {
      "epoch": 0.7404255319148936,
      "grad_norm": 0.14885057508945465,
      "learning_rate": 1.2767568702250844e-05,
      "loss": 1.0925947427749634,
      "step": 174
    },
    {
      "epoch": 0.7489361702127659,
      "grad_norm": 1.3091343641281128,
      "learning_rate": 1.2433084323598791e-05,
      "loss": 0.6399118304252625,
      "step": 176
    },
    {
      "epoch": 0.7574468085106383,
      "grad_norm": 0.5745394825935364,
      "learning_rate": 1.2102423702281116e-05,
      "loss": 1.0284110307693481,
      "step": 178
    },
    {
      "epoch": 0.7659574468085106,
      "grad_norm": 0.4978311061859131,
      "learning_rate": 1.1775849322854516e-05,
      "loss": 1.0611108541488647,
      "step": 180
    },
    {
      "epoch": 0.774468085106383,
      "grad_norm": 0.3034535050392151,
      "learning_rate": 1.1453620426140795e-05,
      "loss": 0.7678611278533936,
      "step": 182
    },
    {
      "epoch": 0.7829787234042553,
      "grad_norm": 0.1695900857448578,
      "learning_rate": 1.1135992803436695e-05,
      "loss": 0.954198956489563,
      "step": 184
    },
    {
      "epoch": 0.7914893617021277,
      "grad_norm": 0.1768021285533905,
      "learning_rate": 1.0823218593461992e-05,
      "loss": 1.0221017599105835,
      "step": 186
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.22744205594062805,
      "learning_rate": 1.0515546082207097e-05,
      "loss": 0.7189929485321045,
      "step": 188
    },
    {
      "epoch": 0.8085106382978723,
      "grad_norm": 0.3399372100830078,
      "learning_rate": 1.0213219505838983e-05,
      "loss": 1.0886459350585938,
      "step": 190
    },
    {
      "epoch": 0.8170212765957446,
      "grad_norm": 0.17291869223117828,
      "learning_rate": 9.91647885682201e-06,
      "loss": 1.2072498798370361,
      "step": 192
    },
    {
      "epoch": 0.825531914893617,
      "grad_norm": 0.19541242718696594,
      "learning_rate": 9.625559693407413e-06,
      "loss": 1.1986221075057983,
      "step": 194
    },
    {
      "epoch": 0.8340425531914893,
      "grad_norm": 0.14983496069908142,
      "learning_rate": 9.340692952642789e-06,
      "loss": 0.720963180065155,
      "step": 196
    },
    {
      "epoch": 0.8425531914893617,
      "grad_norm": 0.22478389739990234,
      "learning_rate": 9.062104767049956e-06,
      "loss": 1.1722772121429443,
      "step": 198
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 0.22518520057201385,
      "learning_rate": 8.790016285116763e-06,
      "loss": 1.0002360343933105,
      "step": 200
    },
    {
      "epoch": 0.8595744680851064,
      "grad_norm": 0.21948722004890442,
      "learning_rate": 8.524643495745306e-06,
      "loss": 0.9713231325149536,
      "step": 202
    },
    {
      "epoch": 0.8680851063829788,
      "grad_norm": 0.15189974009990692,
      "learning_rate": 8.26619705679589e-06,
      "loss": 1.3201754093170166,
      "step": 204
    },
    {
      "epoch": 0.8765957446808511,
      "grad_norm": 0.24632404744625092,
      "learning_rate": 8.014882127862923e-06,
      "loss": 1.282568335533142,
      "step": 206
    },
    {
      "epoch": 0.8851063829787233,
      "grad_norm": 0.13049790263175964,
      "learning_rate": 7.770898207415416e-06,
      "loss": 0.9211198091506958,
      "step": 208
    },
    {
      "epoch": 0.8936170212765957,
      "grad_norm": 0.19982703030109406,
      "learning_rate": 7.534438974431351e-06,
      "loss": 1.167230248451233,
      "step": 210
    },
    {
      "epoch": 0.902127659574468,
      "grad_norm": 0.18156394362449646,
      "learning_rate": 7.305692134651742e-06,
      "loss": 0.7129737138748169,
      "step": 212
    },
    {
      "epoch": 0.9106382978723404,
      "grad_norm": 0.26700806617736816,
      "learning_rate": 7.084839271576291e-06,
      "loss": 0.8392055034637451,
      "step": 214
    },
    {
      "epoch": 0.9191489361702128,
      "grad_norm": 0.1967747062444687,
      "learning_rate": 6.872055702319054e-06,
      "loss": 0.5692006349563599,
      "step": 216
    },
    {
      "epoch": 0.9276595744680851,
      "grad_norm": 0.1744726002216339,
      "learning_rate": 6.667510338438419e-06,
      "loss": 1.1953270435333252,
      "step": 218
    },
    {
      "epoch": 0.9361702127659575,
      "grad_norm": 0.38672754168510437,
      "learning_rate": 6.471365551852012e-06,
      "loss": 0.3942694067955017,
      "step": 220
    },
    {
      "epoch": 0.9446808510638298,
      "grad_norm": 0.19522197544574738,
      "learning_rate": 6.2837770459428e-06,
      "loss": 1.3081778287887573,
      "step": 222
    },
    {
      "epoch": 0.9531914893617022,
      "grad_norm": 0.22507460415363312,
      "learning_rate": 6.1048937319588676e-06,
      "loss": 1.3682390451431274,
      "step": 224
    },
    {
      "epoch": 0.9617021276595744,
      "grad_norm": 0.303733766078949,
      "learning_rate": 5.9348576108049065e-06,
      "loss": 0.894014835357666,
      "step": 226
    },
    {
      "epoch": 0.9702127659574468,
      "grad_norm": 0.205647811293602,
      "learning_rate": 5.773803660319234e-06,
      "loss": 1.0522630214691162,
      "step": 228
    },
    {
      "epoch": 0.9787234042553191,
      "grad_norm": 0.21294599771499634,
      "learning_rate": 5.621859728125884e-06,
      "loss": 1.2654610872268677,
      "step": 230
    },
    {
      "epoch": 0.9872340425531915,
      "grad_norm": 0.3174782395362854,
      "learning_rate": 5.479146430146781e-06,
      "loss": 1.0047165155410767,
      "step": 232
    },
    {
      "epoch": 0.9957446808510638,
      "grad_norm": 1.1156642436981201,
      "learning_rate": 5.345777054854579e-06,
      "loss": 0.9647661447525024,
      "step": 234
    },
    {
      "epoch": 1.004255319148936,
      "grad_norm": 0.3331040143966675,
      "learning_rate": 5.221857473342149e-06,
      "loss": 0.9735086560249329,
      "step": 236
    },
    {
      "epoch": 1.0127659574468084,
      "grad_norm": 0.18369407951831818,
      "learning_rate": 5.10748605528015e-06,
      "loss": 0.8922760486602783,
      "step": 238
    },
    {
      "epoch": 1.0212765957446808,
      "grad_norm": 0.08912570774555206,
      "learning_rate": 5.002753590829349e-06,
      "loss": 0.6222575306892395,
      "step": 240
    },
    {
      "epoch": 1.0297872340425531,
      "grad_norm": 0.15733036398887634,
      "learning_rate": 4.9077432185697e-06,
      "loss": 0.8671458959579468,
      "step": 242
    },
    {
      "epoch": 1.0382978723404255,
      "grad_norm": 0.22573904693126678,
      "learning_rate": 4.822530359503393e-06,
      "loss": 0.9555956721305847,
      "step": 244
    },
    {
      "epoch": 1.0468085106382978,
      "grad_norm": 0.13159972429275513,
      "learning_rate": 4.747182657184251e-06,
      "loss": 0.6556552052497864,
      "step": 246
    },
    {
      "epoch": 1.0553191489361702,
      "grad_norm": 0.19305525720119476,
      "learning_rate": 4.681759924021033e-06,
      "loss": 0.7056201100349426,
      "step": 248
    },
    {
      "epoch": 1.0638297872340425,
      "grad_norm": 0.15075285732746124,
      "learning_rate": 4.626314093797213e-06,
      "loss": 0.6993778347969055,
      "step": 250
    },
    {
      "epoch": 1.0723404255319149,
      "grad_norm": 0.2154897153377533,
      "learning_rate": 4.580889180444988e-06,
      "loss": 0.5365909337997437,
      "step": 252
    },
    {
      "epoch": 1.0808510638297872,
      "grad_norm": 0.17837639153003693,
      "learning_rate": 4.545521243106197e-06,
      "loss": 0.8809426426887512,
      "step": 254
    },
    {
      "epoch": 1.0893617021276596,
      "grad_norm": 0.13702534139156342,
      "learning_rate": 4.520238357507899e-06,
      "loss": 0.8317233324050903,
      "step": 256
    },
    {
      "epoch": 1.097872340425532,
      "grad_norm": 0.23964199423789978,
      "learning_rate": 4.505060593675342e-06,
      "loss": 0.9430326819419861,
      "step": 258
    },
    {
      "epoch": 1.1063829787234043,
      "grad_norm": 0.2120589166879654,
      "learning_rate": 4.5e-06,
      "loss": 0.7370553612709045,
      "step": 260
    },
    {
      "epoch": 1.1148936170212767,
      "grad_norm": 0.21252022683620453,
      "learning_rate": 4.505060593675342e-06,
      "loss": 0.7691786289215088,
      "step": 262
    },
    {
      "epoch": 1.123404255319149,
      "grad_norm": 0.17547725141048431,
      "learning_rate": 4.520238357507899e-06,
      "loss": 0.7104212045669556,
      "step": 264
    },
    {
      "epoch": 1.1319148936170214,
      "grad_norm": 0.1989092081785202,
      "learning_rate": 4.545521243106197e-06,
      "loss": 0.8642681837081909,
      "step": 266
    },
    {
      "epoch": 1.1404255319148937,
      "grad_norm": 0.17653805017471313,
      "learning_rate": 4.580889180444988e-06,
      "loss": 0.770044207572937,
      "step": 268
    },
    {
      "epoch": 1.148936170212766,
      "grad_norm": 0.18561692535877228,
      "learning_rate": 4.6263140937972124e-06,
      "loss": 0.9661275148391724,
      "step": 270
    },
    {
      "epoch": 1.1574468085106382,
      "grad_norm": 0.2783913314342499,
      "learning_rate": 4.6817599240210315e-06,
      "loss": 0.6838027238845825,
      "step": 272
    },
    {
      "epoch": 1.1659574468085105,
      "grad_norm": 0.17354853451251984,
      "learning_rate": 4.747182657184251e-06,
      "loss": 0.8435700535774231,
      "step": 274
    },
    {
      "epoch": 1.174468085106383,
      "grad_norm": 0.19762302935123444,
      "learning_rate": 4.822530359503391e-06,
      "loss": 0.6485944986343384,
      "step": 276
    },
    {
      "epoch": 1.1829787234042553,
      "grad_norm": 0.3056742250919342,
      "learning_rate": 4.9077432185697e-06,
      "loss": 0.5239860415458679,
      "step": 278
    },
    {
      "epoch": 1.1914893617021276,
      "grad_norm": 0.14598001539707184,
      "learning_rate": 5.002753590829348e-06,
      "loss": 0.524198055267334,
      "step": 280
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.16965726017951965,
      "learning_rate": 5.1074860552801466e-06,
      "loss": 0.8808120489120483,
      "step": 282
    },
    {
      "epoch": 1.2085106382978723,
      "grad_norm": 0.24560458958148956,
      "learning_rate": 5.2218574733421455e-06,
      "loss": 0.6656435132026672,
      "step": 284
    },
    {
      "epoch": 1.2170212765957447,
      "grad_norm": 0.23104996979236603,
      "learning_rate": 5.345777054854578e-06,
      "loss": 0.9856959581375122,
      "step": 286
    },
    {
      "epoch": 1.225531914893617,
      "grad_norm": 0.16903841495513916,
      "learning_rate": 5.479146430146783e-06,
      "loss": 0.8170779943466187,
      "step": 288
    },
    {
      "epoch": 1.2340425531914894,
      "grad_norm": 0.17953285574913025,
      "learning_rate": 5.6218597281258834e-06,
      "loss": 0.44949835538864136,
      "step": 290
    },
    {
      "epoch": 1.2425531914893617,
      "grad_norm": 0.4803825914859772,
      "learning_rate": 5.773803660319232e-06,
      "loss": 0.9233723282814026,
      "step": 292
    },
    {
      "epoch": 1.251063829787234,
      "grad_norm": 0.15944814682006836,
      "learning_rate": 5.934857610804904e-06,
      "loss": 0.846575140953064,
      "step": 294
    },
    {
      "epoch": 1.2595744680851064,
      "grad_norm": 0.17765574157238007,
      "learning_rate": 6.1048937319588676e-06,
      "loss": 0.6937251687049866,
      "step": 296
    },
    {
      "epoch": 1.2680851063829788,
      "grad_norm": 0.19115029275417328,
      "learning_rate": 6.283777045942798e-06,
      "loss": 0.8138055205345154,
      "step": 298
    },
    {
      "epoch": 1.2765957446808511,
      "grad_norm": 0.2789772152900696,
      "learning_rate": 6.471365551852011e-06,
      "loss": 0.9583751559257507,
      "step": 300
    },
    {
      "epoch": 1.2851063829787235,
      "grad_norm": 0.15571321547031403,
      "learning_rate": 6.66751033843842e-06,
      "loss": 0.38521289825439453,
      "step": 302
    },
    {
      "epoch": 1.2936170212765958,
      "grad_norm": 0.17477093636989594,
      "learning_rate": 6.872055702319048e-06,
      "loss": 0.9158145189285278,
      "step": 304
    },
    {
      "epoch": 1.302127659574468,
      "grad_norm": 0.18948765099048615,
      "learning_rate": 7.084839271576289e-06,
      "loss": 0.9535613059997559,
      "step": 306
    },
    {
      "epoch": 1.3106382978723405,
      "grad_norm": 0.4815928637981415,
      "learning_rate": 7.30569213465174e-06,
      "loss": 0.5246633887290955,
      "step": 308
    },
    {
      "epoch": 1.3191489361702127,
      "grad_norm": 0.4748033881187439,
      "learning_rate": 7.534438974431351e-06,
      "loss": 0.5024292469024658,
      "step": 310
    },
    {
      "epoch": 1.327659574468085,
      "grad_norm": 0.2643539607524872,
      "learning_rate": 7.770898207415414e-06,
      "loss": 0.5716067552566528,
      "step": 312
    },
    {
      "epoch": 1.3361702127659574,
      "grad_norm": 0.2274530827999115,
      "learning_rate": 8.014882127862923e-06,
      "loss": 0.5070632100105286,
      "step": 314
    },
    {
      "epoch": 1.3446808510638297,
      "grad_norm": 0.2692515254020691,
      "learning_rate": 8.266197056795886e-06,
      "loss": 0.6399731040000916,
      "step": 316
    },
    {
      "epoch": 1.353191489361702,
      "grad_norm": 0.17816181480884552,
      "learning_rate": 8.5246434957453e-06,
      "loss": 0.5720309019088745,
      "step": 318
    },
    {
      "epoch": 1.3617021276595744,
      "grad_norm": 0.427095890045166,
      "learning_rate": 8.790016285116766e-06,
      "loss": 0.49866631627082825,
      "step": 320
    },
    {
      "epoch": 1.3702127659574468,
      "grad_norm": 0.2689511179924011,
      "learning_rate": 9.062104767049957e-06,
      "loss": 0.8831391334533691,
      "step": 322
    },
    {
      "epoch": 1.3787234042553191,
      "grad_norm": 0.2776535451412201,
      "learning_rate": 9.340692952642787e-06,
      "loss": 0.6558242440223694,
      "step": 324
    },
    {
      "epoch": 1.3872340425531915,
      "grad_norm": 0.18160879611968994,
      "learning_rate": 9.625559693407413e-06,
      "loss": 0.8320285677909851,
      "step": 326
    },
    {
      "epoch": 1.3957446808510638,
      "grad_norm": 0.2418782263994217,
      "learning_rate": 9.916478856822006e-06,
      "loss": 0.6791567802429199,
      "step": 328
    },
    {
      "epoch": 1.4042553191489362,
      "grad_norm": 0.30612000823020935,
      "learning_rate": 1.0213219505838985e-05,
      "loss": 0.7776432633399963,
      "step": 330
    },
    {
      "epoch": 1.4127659574468086,
      "grad_norm": 0.2482786476612091,
      "learning_rate": 1.0515546082207094e-05,
      "loss": 0.8018679022789001,
      "step": 332
    },
    {
      "epoch": 1.421276595744681,
      "grad_norm": 0.2703081965446472,
      "learning_rate": 1.082321859346199e-05,
      "loss": 0.6729298830032349,
      "step": 334
    },
    {
      "epoch": 1.4297872340425533,
      "grad_norm": 0.21901671588420868,
      "learning_rate": 1.1135992803436696e-05,
      "loss": 0.6063901782035828,
      "step": 336
    },
    {
      "epoch": 1.4382978723404256,
      "grad_norm": 0.20921635627746582,
      "learning_rate": 1.1453620426140791e-05,
      "loss": 0.6617997288703918,
      "step": 338
    },
    {
      "epoch": 1.4468085106382977,
      "grad_norm": 0.19303618371486664,
      "learning_rate": 1.1775849322854508e-05,
      "loss": 0.5969347953796387,
      "step": 340
    },
    {
      "epoch": 1.4553191489361703,
      "grad_norm": 0.3104172646999359,
      "learning_rate": 1.210242370228112e-05,
      "loss": 0.6071085333824158,
      "step": 342
    },
    {
      "epoch": 1.4638297872340424,
      "grad_norm": 0.22875197231769562,
      "learning_rate": 1.2433084323598791e-05,
      "loss": 0.9178237915039062,
      "step": 344
    },
    {
      "epoch": 1.472340425531915,
      "grad_norm": 0.19736064970493317,
      "learning_rate": 1.2767568702250838e-05,
      "loss": 0.8496959209442139,
      "step": 346
    },
    {
      "epoch": 1.4808510638297872,
      "grad_norm": 0.24981622397899628,
      "learning_rate": 1.3105611318310811e-05,
      "loss": 0.7014995813369751,
      "step": 348
    },
    {
      "epoch": 1.4893617021276595,
      "grad_norm": 0.2884751558303833,
      "learning_rate": 1.3446943827257184e-05,
      "loss": 0.7825354337692261,
      "step": 350
    },
    {
      "epoch": 1.4978723404255319,
      "grad_norm": 0.42386531829833984,
      "learning_rate": 1.3791295272990172e-05,
      "loss": 0.33443060517311096,
      "step": 352
    },
    {
      "epoch": 1.5063829787234042,
      "grad_norm": 0.16379332542419434,
      "learning_rate": 1.413839230292182e-05,
      "loss": 0.6189748048782349,
      "step": 354
    },
    {
      "epoch": 1.5148936170212766,
      "grad_norm": 0.2462461143732071,
      "learning_rate": 1.4487959384968276e-05,
      "loss": 0.46220940351486206,
      "step": 356
    },
    {
      "epoch": 1.523404255319149,
      "grad_norm": 0.23771870136260986,
      "learning_rate": 1.4839719026272373e-05,
      "loss": 0.4545080363750458,
      "step": 358
    },
    {
      "epoch": 1.5319148936170213,
      "grad_norm": 0.18759848177433014,
      "learning_rate": 1.5193391993482579e-05,
      "loss": 0.9656761288642883,
      "step": 360
    },
    {
      "epoch": 1.5404255319148936,
      "grad_norm": 0.18730829656124115,
      "learning_rate": 1.5548697534413636e-05,
      "loss": 0.44390690326690674,
      "step": 362
    },
    {
      "epoch": 1.548936170212766,
      "grad_norm": 0.3202720582485199,
      "learning_rate": 1.5905353600912894e-05,
      "loss": 0.9684057235717773,
      "step": 364
    },
    {
      "epoch": 1.5574468085106383,
      "grad_norm": 0.13575418293476105,
      "learning_rate": 1.6263077072755323e-05,
      "loss": 0.8047466278076172,
      "step": 366
    },
    {
      "epoch": 1.5659574468085107,
      "grad_norm": 0.36194950342178345,
      "learning_rate": 1.6621583982389707e-05,
      "loss": 0.6687026619911194,
      "step": 368
    },
    {
      "epoch": 1.574468085106383,
      "grad_norm": 0.12254664301872253,
      "learning_rate": 1.6980589740357294e-05,
      "loss": 0.5913136601448059,
      "step": 370
    },
    {
      "epoch": 1.5829787234042554,
      "grad_norm": 0.13009262084960938,
      "learning_rate": 1.733980936120425e-05,
      "loss": 0.7752522230148315,
      "step": 372
    },
    {
      "epoch": 1.5914893617021275,
      "grad_norm": 0.5959559679031372,
      "learning_rate": 1.7698957689708416e-05,
      "loss": 0.9753284454345703,
      "step": 374
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.5740659832954407,
      "learning_rate": 1.805774962724083e-05,
      "loss": 0.7318065762519836,
      "step": 376
    },
    {
      "epoch": 1.6085106382978722,
      "grad_norm": 0.5994616746902466,
      "learning_rate": 1.8415900358082265e-05,
      "loss": 0.5247442126274109,
      "step": 378
    },
    {
      "epoch": 1.6170212765957448,
      "grad_norm": 0.27797627449035645,
      "learning_rate": 1.8773125575515357e-05,
      "loss": 0.5938118696212769,
      "step": 380
    },
    {
      "epoch": 1.625531914893617,
      "grad_norm": 0.16901274025440216,
      "learning_rate": 1.912914170751251e-05,
      "loss": 0.761189341545105,
      "step": 382
    },
    {
      "epoch": 1.6340425531914895,
      "grad_norm": 0.2777624726295471,
      "learning_rate": 1.9483666141840612e-05,
      "loss": 0.9914259910583496,
      "step": 384
    },
    {
      "epoch": 1.6425531914893616,
      "grad_norm": 0.3253246247768402,
      "learning_rate": 1.9836417450403974e-05,
      "loss": 0.8401809930801392,
      "step": 386
    },
    {
      "epoch": 1.6510638297872342,
      "grad_norm": 0.16334128379821777,
      "learning_rate": 2.0187115612647133e-05,
      "loss": 0.694261372089386,
      "step": 388
    },
    {
      "epoch": 1.6595744680851063,
      "grad_norm": 0.16701728105545044,
      "learning_rate": 2.05354822378404e-05,
      "loss": 0.8252889513969421,
      "step": 390
    },
    {
      "epoch": 1.6680851063829787,
      "grad_norm": 0.10444586724042892,
      "learning_rate": 2.0881240786071595e-05,
      "loss": 0.8890981078147888,
      "step": 392
    },
    {
      "epoch": 1.676595744680851,
      "grad_norm": 0.4691241681575775,
      "learning_rate": 2.122411678776854e-05,
      "loss": 0.6046314835548401,
      "step": 394
    },
    {
      "epoch": 1.6851063829787234,
      "grad_norm": 0.11681725829839706,
      "learning_rate": 2.1563838061578258e-05,
      "loss": 0.8832032084465027,
      "step": 396
    },
    {
      "epoch": 1.6936170212765957,
      "grad_norm": 0.14530058205127716,
      "learning_rate": 2.1900134930429396e-05,
      "loss": 1.0986517667770386,
      "step": 398
    },
    {
      "epoch": 1.702127659574468,
      "grad_norm": 0.37800365686416626,
      "learning_rate": 2.223274043560706e-05,
      "loss": 0.9784438610076904,
      "step": 400
    },
    {
      "epoch": 1.7106382978723405,
      "grad_norm": 0.37573498487472534,
      "learning_rate": 2.256139054866954e-05,
      "loss": 0.47277578711509705,
      "step": 402
    },
    {
      "epoch": 1.7191489361702128,
      "grad_norm": 0.5493515729904175,
      "learning_rate": 2.2885824381039028e-05,
      "loss": 0.6571931838989258,
      "step": 404
    },
    {
      "epoch": 1.7276595744680852,
      "grad_norm": 0.24701279401779175,
      "learning_rate": 2.3205784391099808e-05,
      "loss": 0.5199674367904663,
      "step": 406
    },
    {
      "epoch": 1.7361702127659573,
      "grad_norm": 0.2061431109905243,
      "learning_rate": 2.352101658863958e-05,
      "loss": 0.7460405230522156,
      "step": 408
    },
    {
      "epoch": 1.7446808510638299,
      "grad_norm": 0.16800709068775177,
      "learning_rate": 2.3831270736471703e-05,
      "loss": 0.9135586023330688,
      "step": 410
    },
    {
      "epoch": 1.753191489361702,
      "grad_norm": 0.5365346074104309,
      "learning_rate": 2.4136300549077973e-05,
      "loss": 0.906153678894043,
      "step": 412
    },
    {
      "epoch": 1.7617021276595746,
      "grad_norm": 0.14748896658420563,
      "learning_rate": 2.443586388811481e-05,
      "loss": 0.7806675434112549,
      "step": 414
    },
    {
      "epoch": 1.7702127659574467,
      "grad_norm": 1.3085377216339111,
      "learning_rate": 2.47297229546271e-05,
      "loss": 0.6440402865409851,
      "step": 416
    },
    {
      "epoch": 1.7787234042553193,
      "grad_norm": 0.2483910322189331,
      "learning_rate": 2.5017644477817418e-05,
      "loss": 0.7717803120613098,
      "step": 418
    },
    {
      "epoch": 1.7872340425531914,
      "grad_norm": 0.4905950427055359,
      "learning_rate": 2.5299399900220807e-05,
      "loss": 0.6290927529335022,
      "step": 420
    },
    {
      "epoch": 1.795744680851064,
      "grad_norm": 0.11053664982318878,
      "learning_rate": 2.5574765559137848e-05,
      "loss": 0.7530769109725952,
      "step": 422
    },
    {
      "epoch": 1.804255319148936,
      "grad_norm": 0.11755101382732391,
      "learning_rate": 2.584352286418239e-05,
      "loss": 0.8429850935935974,
      "step": 424
    },
    {
      "epoch": 1.8127659574468085,
      "grad_norm": 0.14117616415023804,
      "learning_rate": 2.6105458470802563e-05,
      "loss": 0.8456153869628906,
      "step": 426
    },
    {
      "epoch": 1.8212765957446808,
      "grad_norm": 0.22968512773513794,
      "learning_rate": 2.6360364449637683e-05,
      "loss": 0.8607503771781921,
      "step": 428
    },
    {
      "epoch": 1.8297872340425532,
      "grad_norm": 0.3352075219154358,
      "learning_rate": 2.6608038451576528e-05,
      "loss": 0.9603878259658813,
      "step": 430
    },
    {
      "epoch": 1.8382978723404255,
      "grad_norm": 0.22618667781352997,
      "learning_rate": 2.684828386838569e-05,
      "loss": 0.7366024255752563,
      "step": 432
    },
    {
      "epoch": 1.8468085106382979,
      "grad_norm": 0.3134029805660248,
      "learning_rate": 2.7080909988780982e-05,
      "loss": 0.9086182117462158,
      "step": 434
    },
    {
      "epoch": 1.8553191489361702,
      "grad_norm": 0.12940463423728943,
      "learning_rate": 2.7305732149817502e-05,
      "loss": 0.642800509929657,
      "step": 436
    },
    {
      "epoch": 1.8638297872340426,
      "grad_norm": 0.3360910415649414,
      "learning_rate": 2.7522571883478617e-05,
      "loss": 1.0462186336517334,
      "step": 438
    },
    {
      "epoch": 1.872340425531915,
      "grad_norm": 0.18832530081272125,
      "learning_rate": 2.77312570583471e-05,
      "loss": 0.9516323804855347,
      "step": 440
    },
    {
      "epoch": 1.8808510638297873,
      "grad_norm": 0.12196290493011475,
      "learning_rate": 2.7931622016246304e-05,
      "loss": 0.9153946042060852,
      "step": 442
    },
    {
      "epoch": 1.8893617021276596,
      "grad_norm": 0.1658358871936798,
      "learning_rate": 2.8123507703742727e-05,
      "loss": 0.8443633913993835,
      "step": 444
    },
    {
      "epoch": 1.8978723404255318,
      "grad_norm": 0.19309784471988678,
      "learning_rate": 2.8306761798405522e-05,
      "loss": 1.0077768564224243,
      "step": 446
    },
    {
      "epoch": 1.9063829787234043,
      "grad_norm": 0.1583261489868164,
      "learning_rate": 2.848123882972295e-05,
      "loss": 0.9472059607505798,
      "step": 448
    },
    {
      "epoch": 1.9148936170212765,
      "grad_norm": 0.5723341107368469,
      "learning_rate": 2.8646800294579514e-05,
      "loss": 0.6875810623168945,
      "step": 450
    },
    {
      "epoch": 1.923404255319149,
      "grad_norm": 0.339374303817749,
      "learning_rate": 2.8803314767202376e-05,
      "loss": 0.5654004812240601,
      "step": 452
    },
    {
      "epoch": 1.9319148936170212,
      "grad_norm": 0.5859424471855164,
      "learning_rate": 2.8950658003489534e-05,
      "loss": 0.8492231965065002,
      "step": 454
    },
    {
      "epoch": 1.9404255319148938,
      "grad_norm": 0.422346830368042,
      "learning_rate": 2.908871303963711e-05,
      "loss": 0.9699692130088806,
      "step": 456
    },
    {
      "epoch": 1.9489361702127659,
      "grad_norm": 0.1916320025920868,
      "learning_rate": 2.9217370284987434e-05,
      "loss": 0.7598567605018616,
      "step": 458
    },
    {
      "epoch": 1.9574468085106385,
      "grad_norm": 0.2646377384662628,
      "learning_rate": 2.9336527609024075e-05,
      "loss": 1.0349591970443726,
      "step": 460
    },
    {
      "epoch": 1.9659574468085106,
      "grad_norm": 0.19794800877571106,
      "learning_rate": 2.9446090422445016e-05,
      "loss": 0.7465431094169617,
      "step": 462
    },
    {
      "epoch": 1.974468085106383,
      "grad_norm": 0.11537999659776688,
      "learning_rate": 2.9545971752249376e-05,
      "loss": 0.8450900316238403,
      "step": 464
    },
    {
      "epoch": 1.9829787234042553,
      "grad_norm": 0.16437967121601105,
      "learning_rate": 2.9636092310778195e-05,
      "loss": 0.8825825452804565,
      "step": 466
    },
    {
      "epoch": 1.9914893617021276,
      "grad_norm": 0.1353980153799057,
      "learning_rate": 2.9716380558654445e-05,
      "loss": 0.6045427918434143,
      "step": 468
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.12124238908290863,
      "learning_rate": 2.9786772761572338e-05,
      "loss": 0.6623026728630066,
      "step": 470
    },
    {
      "epoch": 2.008510638297872,
      "grad_norm": 0.24585317075252533,
      "learning_rate": 2.9847213040890793e-05,
      "loss": 0.36443179845809937,
      "step": 472
    },
    {
      "epoch": 2.0170212765957447,
      "grad_norm": 0.13654032349586487,
      "learning_rate": 2.989765341799095e-05,
      "loss": 0.75254225730896,
      "step": 474
    },
    {
      "epoch": 2.025531914893617,
      "grad_norm": 0.5691254138946533,
      "learning_rate": 2.9938053852362484e-05,
      "loss": 0.47092172503471375,
      "step": 476
    },
    {
      "epoch": 2.0340425531914894,
      "grad_norm": 0.1691976934671402,
      "learning_rate": 2.99683822733885e-05,
      "loss": 0.6583607196807861,
      "step": 478
    },
    {
      "epoch": 2.0425531914893615,
      "grad_norm": 0.16407643258571625,
      "learning_rate": 2.9988614605803806e-05,
      "loss": 0.5634033679962158,
      "step": 480
    },
    {
      "epoch": 2.051063829787234,
      "grad_norm": 0.22440628707408905,
      "learning_rate": 2.9998734788806287e-05,
      "loss": 0.4582347571849823,
      "step": 482
    },
    {
      "epoch": 2.0595744680851062,
      "grad_norm": 0.07732467353343964,
      "learning_rate": 2.9998734788806287e-05,
      "loss": 0.24273234605789185,
      "step": 484
    },
    {
      "epoch": 2.068085106382979,
      "grad_norm": 0.46668562293052673,
      "learning_rate": 2.9988614605803806e-05,
      "loss": 0.6905072331428528,
      "step": 486
    },
    {
      "epoch": 2.076595744680851,
      "grad_norm": 0.14304345846176147,
      "learning_rate": 2.99683822733885e-05,
      "loss": 0.4559966027736664,
      "step": 488
    },
    {
      "epoch": 2.0851063829787235,
      "grad_norm": 0.370118111371994,
      "learning_rate": 2.9938053852362484e-05,
      "loss": 0.5135046243667603,
      "step": 490
    },
    {
      "epoch": 2.0936170212765957,
      "grad_norm": 0.15926745533943176,
      "learning_rate": 2.989765341799095e-05,
      "loss": 0.7977150678634644,
      "step": 492
    },
    {
      "epoch": 2.1021276595744682,
      "grad_norm": 0.13238710165023804,
      "learning_rate": 2.9847213040890793e-05,
      "loss": 0.6057368516921997,
      "step": 494
    },
    {
      "epoch": 2.1106382978723404,
      "grad_norm": 0.18792684376239777,
      "learning_rate": 2.9786772761572338e-05,
      "loss": 0.7654038071632385,
      "step": 496
    },
    {
      "epoch": 2.119148936170213,
      "grad_norm": 0.12919172644615173,
      "learning_rate": 2.9716380558654445e-05,
      "loss": 0.8083907961845398,
      "step": 498
    },
    {
      "epoch": 2.127659574468085,
      "grad_norm": 0.21281176805496216,
      "learning_rate": 2.9636092310778195e-05,
      "loss": 0.5940195322036743,
      "step": 500
    },
    {
      "epoch": 2.1361702127659576,
      "grad_norm": 0.1665232926607132,
      "learning_rate": 2.9545971752249383e-05,
      "loss": 0.5825145244598389,
      "step": 502
    },
    {
      "epoch": 2.1446808510638298,
      "grad_norm": 0.11076656728982925,
      "learning_rate": 2.9446090422445016e-05,
      "loss": 0.5060111284255981,
      "step": 504
    },
    {
      "epoch": 2.153191489361702,
      "grad_norm": 0.123874731361866,
      "learning_rate": 2.933652760902408e-05,
      "loss": 0.5895789265632629,
      "step": 506
    },
    {
      "epoch": 2.1617021276595745,
      "grad_norm": 0.24894492328166962,
      "learning_rate": 2.9217370284987434e-05,
      "loss": 0.3788684010505676,
      "step": 508
    },
    {
      "epoch": 2.1702127659574466,
      "grad_norm": 0.1470797210931778,
      "learning_rate": 2.908871303963712e-05,
      "loss": 0.3065789043903351,
      "step": 510
    },
    {
      "epoch": 2.178723404255319,
      "grad_norm": 0.2271808236837387,
      "learning_rate": 2.8950658003489534e-05,
      "loss": 0.3498597741127014,
      "step": 512
    },
    {
      "epoch": 2.1872340425531913,
      "grad_norm": 0.2549133896827698,
      "learning_rate": 2.8803314767202387e-05,
      "loss": 0.3588785231113434,
      "step": 514
    },
    {
      "epoch": 2.195744680851064,
      "grad_norm": 0.27323317527770996,
      "learning_rate": 2.864680029457952e-05,
      "loss": 0.4839470088481903,
      "step": 516
    },
    {
      "epoch": 2.204255319148936,
      "grad_norm": 0.7197582721710205,
      "learning_rate": 2.8481238829722946e-05,
      "loss": 0.6455433368682861,
      "step": 518
    },
    {
      "epoch": 2.2127659574468086,
      "grad_norm": 0.16234466433525085,
      "learning_rate": 2.8306761798405533e-05,
      "loss": 0.5413359999656677,
      "step": 520
    },
    {
      "epoch": 2.2212765957446807,
      "grad_norm": 0.21601392328739166,
      "learning_rate": 2.812350770374273e-05,
      "loss": 0.33399829268455505,
      "step": 522
    },
    {
      "epoch": 2.2297872340425533,
      "grad_norm": 0.3286589980125427,
      "learning_rate": 2.793162201624631e-05,
      "loss": 0.4370644688606262,
      "step": 524
    },
    {
      "epoch": 2.2382978723404254,
      "grad_norm": 0.28274786472320557,
      "learning_rate": 2.7731257058347098e-05,
      "loss": 0.6098963618278503,
      "step": 526
    },
    {
      "epoch": 2.246808510638298,
      "grad_norm": 0.1321384608745575,
      "learning_rate": 2.752257188347862e-05,
      "loss": 0.6518822908401489,
      "step": 528
    },
    {
      "epoch": 2.25531914893617,
      "grad_norm": 0.4378323256969452,
      "learning_rate": 2.7305732149817512e-05,
      "loss": 0.6883822679519653,
      "step": 530
    },
    {
      "epoch": 2.2638297872340427,
      "grad_norm": 0.12180285900831223,
      "learning_rate": 2.708090998878098e-05,
      "loss": 0.5799202919006348,
      "step": 532
    },
    {
      "epoch": 2.272340425531915,
      "grad_norm": 0.2532312572002411,
      "learning_rate": 2.68482838683857e-05,
      "loss": 0.30189117789268494,
      "step": 534
    },
    {
      "epoch": 2.2808510638297874,
      "grad_norm": 0.1414760947227478,
      "learning_rate": 2.6608038451576528e-05,
      "loss": 0.5066512823104858,
      "step": 536
    },
    {
      "epoch": 2.2893617021276595,
      "grad_norm": 0.15154339373111725,
      "learning_rate": 2.6360364449637686e-05,
      "loss": 0.7329890131950378,
      "step": 538
    },
    {
      "epoch": 2.297872340425532,
      "grad_norm": 0.438917338848114,
      "learning_rate": 2.6105458470802563e-05,
      "loss": 0.23619456589221954,
      "step": 540
    },
    {
      "epoch": 2.3063829787234043,
      "grad_norm": 0.10420854389667511,
      "learning_rate": 2.5843522864182397e-05,
      "loss": 0.2426261454820633,
      "step": 542
    },
    {
      "epoch": 2.3148936170212764,
      "grad_norm": 0.3014877736568451,
      "learning_rate": 2.557476555913786e-05,
      "loss": 0.4293208420276642,
      "step": 544
    },
    {
      "epoch": 2.323404255319149,
      "grad_norm": 0.16749997437000275,
      "learning_rate": 2.52993999002208e-05,
      "loss": 0.317562460899353,
      "step": 546
    },
    {
      "epoch": 2.331914893617021,
      "grad_norm": 0.5269374847412109,
      "learning_rate": 2.501764447781743e-05,
      "loss": 0.7331305742263794,
      "step": 548
    },
    {
      "epoch": 2.3404255319148937,
      "grad_norm": 0.5243010520935059,
      "learning_rate": 2.4729722954627106e-05,
      "loss": 0.4494149088859558,
      "step": 550
    },
    {
      "epoch": 2.348936170212766,
      "grad_norm": 0.16172371804714203,
      "learning_rate": 2.4435863888114814e-05,
      "loss": 0.5373607873916626,
      "step": 552
    },
    {
      "epoch": 2.3574468085106384,
      "grad_norm": 0.13132323324680328,
      "learning_rate": 2.413630054907798e-05,
      "loss": 0.7187270522117615,
      "step": 554
    },
    {
      "epoch": 2.3659574468085105,
      "grad_norm": 0.16762499511241913,
      "learning_rate": 2.383127073647171e-05,
      "loss": 0.5384994745254517,
      "step": 556
    },
    {
      "epoch": 2.374468085106383,
      "grad_norm": 0.19120921194553375,
      "learning_rate": 2.3521016588639598e-05,
      "loss": 0.5281667709350586,
      "step": 558
    },
    {
      "epoch": 2.382978723404255,
      "grad_norm": 0.12356223165988922,
      "learning_rate": 2.32057843910998e-05,
      "loss": 0.47801631689071655,
      "step": 560
    },
    {
      "epoch": 2.391489361702128,
      "grad_norm": 0.11622758209705353,
      "learning_rate": 2.2885824381039024e-05,
      "loss": 0.5657482147216797,
      "step": 562
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.18093258142471313,
      "learning_rate": 2.2561390548669552e-05,
      "loss": 0.43637847900390625,
      "step": 564
    },
    {
      "epoch": 2.4085106382978725,
      "grad_norm": 0.1480812132358551,
      "learning_rate": 2.2232740435607067e-05,
      "loss": 0.4490112364292145,
      "step": 566
    },
    {
      "epoch": 2.4170212765957446,
      "grad_norm": 0.3488391935825348,
      "learning_rate": 2.1900134930429403e-05,
      "loss": 0.5125418901443481,
      "step": 568
    },
    {
      "epoch": 2.425531914893617,
      "grad_norm": 0.20989084243774414,
      "learning_rate": 2.1563838061578264e-05,
      "loss": 0.4947648346424103,
      "step": 570
    },
    {
      "epoch": 2.4340425531914893,
      "grad_norm": 0.1344413161277771,
      "learning_rate": 2.122411678776855e-05,
      "loss": 0.40764039754867554,
      "step": 572
    },
    {
      "epoch": 2.4425531914893615,
      "grad_norm": 0.5959489345550537,
      "learning_rate": 2.0881240786071588e-05,
      "loss": 0.402820885181427,
      "step": 574
    },
    {
      "epoch": 2.451063829787234,
      "grad_norm": 0.17302733659744263,
      "learning_rate": 2.0535482237840398e-05,
      "loss": 0.5862278342247009,
      "step": 576
    },
    {
      "epoch": 2.4595744680851066,
      "grad_norm": 0.18545182049274445,
      "learning_rate": 2.018711561264715e-05,
      "loss": 0.5663528442382812,
      "step": 578
    },
    {
      "epoch": 2.4680851063829787,
      "grad_norm": 0.1751205325126648,
      "learning_rate": 1.9836417450403978e-05,
      "loss": 0.3827342689037323,
      "step": 580
    },
    {
      "epoch": 2.476595744680851,
      "grad_norm": 0.25646570324897766,
      "learning_rate": 1.948366614184062e-05,
      "loss": 0.5597113966941833,
      "step": 582
    },
    {
      "epoch": 2.4851063829787234,
      "grad_norm": 0.1912815123796463,
      "learning_rate": 1.9129141707512514e-05,
      "loss": 0.6160271763801575,
      "step": 584
    },
    {
      "epoch": 2.4936170212765956,
      "grad_norm": 0.3505466878414154,
      "learning_rate": 1.877312557551536e-05,
      "loss": 0.5502729415893555,
      "step": 586
    },
    {
      "epoch": 2.502127659574468,
      "grad_norm": 0.17846280336380005,
      "learning_rate": 1.841590035808227e-05,
      "loss": 0.7185637354850769,
      "step": 588
    },
    {
      "epoch": 2.5106382978723403,
      "grad_norm": 0.15607386827468872,
      "learning_rate": 1.8057749627240822e-05,
      "loss": 0.6139054894447327,
      "step": 590
    },
    {
      "epoch": 2.519148936170213,
      "grad_norm": 0.13966147601604462,
      "learning_rate": 1.7698957689708436e-05,
      "loss": 0.7389253973960876,
      "step": 592
    },
    {
      "epoch": 2.527659574468085,
      "grad_norm": 0.14507059752941132,
      "learning_rate": 1.7339809361204255e-05,
      "loss": 0.35956406593322754,
      "step": 594
    },
    {
      "epoch": 2.5361702127659576,
      "grad_norm": 0.15395186841487885,
      "learning_rate": 1.6980589740357287e-05,
      "loss": 0.5801402926445007,
      "step": 596
    },
    {
      "epoch": 2.5446808510638297,
      "grad_norm": 0.13444769382476807,
      "learning_rate": 1.6621583982389714e-05,
      "loss": 0.4474773108959198,
      "step": 598
    },
    {
      "epoch": 2.5531914893617023,
      "grad_norm": 0.21395254135131836,
      "learning_rate": 1.626307707275533e-05,
      "loss": 0.5679066777229309,
      "step": 600
    },
    {
      "epoch": 2.5617021276595744,
      "grad_norm": 0.1135779544711113,
      "learning_rate": 1.59053536009129e-05,
      "loss": 0.5999655723571777,
      "step": 602
    },
    {
      "epoch": 2.570212765957447,
      "grad_norm": 0.11653319001197815,
      "learning_rate": 1.5548697534413653e-05,
      "loss": 0.577793300151825,
      "step": 604
    },
    {
      "epoch": 2.578723404255319,
      "grad_norm": 0.1222231388092041,
      "learning_rate": 1.5193391993482573e-05,
      "loss": 0.6178088188171387,
      "step": 606
    },
    {
      "epoch": 2.5872340425531917,
      "grad_norm": 0.1294461488723755,
      "learning_rate": 1.483971902627239e-05,
      "loss": 0.2757226824760437,
      "step": 608
    },
    {
      "epoch": 2.595744680851064,
      "grad_norm": 0.12525388598442078,
      "learning_rate": 1.4487959384968272e-05,
      "loss": 0.5488746762275696,
      "step": 610
    },
    {
      "epoch": 2.604255319148936,
      "grad_norm": 0.3998476564884186,
      "learning_rate": 1.4138392302921813e-05,
      "loss": 0.6341690421104431,
      "step": 612
    },
    {
      "epoch": 2.6127659574468085,
      "grad_norm": 0.1306917667388916,
      "learning_rate": 1.3791295272990178e-05,
      "loss": 0.6712806224822998,
      "step": 614
    },
    {
      "epoch": 2.621276595744681,
      "grad_norm": 0.08778780698776245,
      "learning_rate": 1.344694382725719e-05,
      "loss": 0.45013314485549927,
      "step": 616
    },
    {
      "epoch": 2.629787234042553,
      "grad_norm": 0.2586289048194885,
      "learning_rate": 1.3105611318310828e-05,
      "loss": 0.6572415828704834,
      "step": 618
    },
    {
      "epoch": 2.6382978723404253,
      "grad_norm": 0.10759896785020828,
      "learning_rate": 1.2767568702250834e-05,
      "loss": 0.41325774788856506,
      "step": 620
    },
    {
      "epoch": 2.646808510638298,
      "grad_norm": 0.21708400547504425,
      "learning_rate": 1.2433084323598806e-05,
      "loss": 0.4483058750629425,
      "step": 622
    },
    {
      "epoch": 2.65531914893617,
      "grad_norm": 0.17001259326934814,
      "learning_rate": 1.2102423702281114e-05,
      "loss": 0.6922531127929688,
      "step": 624
    },
    {
      "epoch": 2.6638297872340426,
      "grad_norm": 0.21257923543453217,
      "learning_rate": 1.1775849322854515e-05,
      "loss": 0.5280181169509888,
      "step": 626
    },
    {
      "epoch": 2.6723404255319148,
      "grad_norm": 0.18666359782218933,
      "learning_rate": 1.1453620426140796e-05,
      "loss": 0.5663700699806213,
      "step": 628
    },
    {
      "epoch": 2.6808510638297873,
      "grad_norm": 1.1582063436508179,
      "learning_rate": 1.1135992803436701e-05,
      "loss": 0.5217815637588501,
      "step": 630
    },
    {
      "epoch": 2.6893617021276595,
      "grad_norm": 0.15095171332359314,
      "learning_rate": 1.0823218593462006e-05,
      "loss": 0.5002037882804871,
      "step": 632
    },
    {
      "epoch": 2.697872340425532,
      "grad_norm": 0.258025199174881,
      "learning_rate": 1.0515546082207089e-05,
      "loss": 0.7550774812698364,
      "step": 634
    },
    {
      "epoch": 2.706382978723404,
      "grad_norm": 0.2747920751571655,
      "learning_rate": 1.0213219505839e-05,
      "loss": 0.7289509773254395,
      "step": 636
    },
    {
      "epoch": 2.7148936170212767,
      "grad_norm": 0.10981704294681549,
      "learning_rate": 9.916478856822011e-06,
      "loss": 0.4343630075454712,
      "step": 638
    },
    {
      "epoch": 2.723404255319149,
      "grad_norm": 0.10845521837472916,
      "learning_rate": 9.625559693407416e-06,
      "loss": 0.8249025940895081,
      "step": 640
    },
    {
      "epoch": 2.731914893617021,
      "grad_norm": 0.15325450897216797,
      "learning_rate": 9.34069295264279e-06,
      "loss": 0.7887688875198364,
      "step": 642
    },
    {
      "epoch": 2.7404255319148936,
      "grad_norm": 0.12251801788806915,
      "learning_rate": 9.062104767049962e-06,
      "loss": 0.8136211037635803,
      "step": 644
    },
    {
      "epoch": 2.748936170212766,
      "grad_norm": 0.12305980175733566,
      "learning_rate": 8.790016285116773e-06,
      "loss": 0.3323878347873688,
      "step": 646
    },
    {
      "epoch": 2.7574468085106383,
      "grad_norm": 0.16256316006183624,
      "learning_rate": 8.5246434957453e-06,
      "loss": 0.31487664580345154,
      "step": 648
    },
    {
      "epoch": 2.7659574468085104,
      "grad_norm": 0.29724588990211487,
      "learning_rate": 8.266197056795886e-06,
      "loss": 0.43204954266548157,
      "step": 650
    },
    {
      "epoch": 2.774468085106383,
      "grad_norm": 0.21929951012134552,
      "learning_rate": 8.014882127862926e-06,
      "loss": 0.562320351600647,
      "step": 652
    },
    {
      "epoch": 2.7829787234042556,
      "grad_norm": 0.11705265194177628,
      "learning_rate": 7.77089820741542e-06,
      "loss": 0.5608207583427429,
      "step": 654
    },
    {
      "epoch": 2.7914893617021277,
      "grad_norm": 0.14389561116695404,
      "learning_rate": 7.534438974431356e-06,
      "loss": 0.44889208674430847,
      "step": 656
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.10500220209360123,
      "learning_rate": 7.305692134651748e-06,
      "loss": 0.5233709216117859,
      "step": 658
    },
    {
      "epoch": 2.8085106382978724,
      "grad_norm": 0.11123310774564743,
      "learning_rate": 7.0848392715763e-06,
      "loss": 0.41444164514541626,
      "step": 660
    },
    {
      "epoch": 2.8170212765957445,
      "grad_norm": 0.14375989139080048,
      "learning_rate": 6.872055702319048e-06,
      "loss": 0.2203376591205597,
      "step": 662
    },
    {
      "epoch": 2.825531914893617,
      "grad_norm": 2.4339959621429443,
      "learning_rate": 6.66751033843842e-06,
      "loss": 0.4634183943271637,
      "step": 664
    },
    {
      "epoch": 2.8340425531914892,
      "grad_norm": 0.16652415692806244,
      "learning_rate": 6.471365551852014e-06,
      "loss": 0.36938074231147766,
      "step": 666
    },
    {
      "epoch": 2.842553191489362,
      "grad_norm": 0.13066215813159943,
      "learning_rate": 6.283777045942801e-06,
      "loss": 0.5424007177352905,
      "step": 668
    },
    {
      "epoch": 2.851063829787234,
      "grad_norm": 0.1719757467508316,
      "learning_rate": 6.104893731958872e-06,
      "loss": 0.7605326175689697,
      "step": 670
    },
    {
      "epoch": 2.8595744680851065,
      "grad_norm": 0.15954963862895966,
      "learning_rate": 5.934857610804912e-06,
      "loss": 0.4137082099914551,
      "step": 672
    },
    {
      "epoch": 2.8680851063829786,
      "grad_norm": 0.182533860206604,
      "learning_rate": 5.773803660319231e-06,
      "loss": 0.5575495362281799,
      "step": 674
    },
    {
      "epoch": 2.876595744680851,
      "grad_norm": 0.13435368239879608,
      "learning_rate": 5.6218597281258834e-06,
      "loss": 0.3841201961040497,
      "step": 676
    },
    {
      "epoch": 2.8851063829787233,
      "grad_norm": 0.09731018543243408,
      "learning_rate": 5.479146430146783e-06,
      "loss": 0.3572618365287781,
      "step": 678
    },
    {
      "epoch": 2.8936170212765955,
      "grad_norm": 0.12130768597126007,
      "learning_rate": 5.3457770548545805e-06,
      "loss": 0.4105460047721863,
      "step": 680
    },
    {
      "epoch": 2.902127659574468,
      "grad_norm": 0.4408089518547058,
      "learning_rate": 5.22185747334215e-06,
      "loss": 0.5799260139465332,
      "step": 682
    },
    {
      "epoch": 2.9106382978723406,
      "grad_norm": 0.12350846827030182,
      "learning_rate": 5.107486055280145e-06,
      "loss": 0.4056027829647064,
      "step": 684
    },
    {
      "epoch": 2.9191489361702128,
      "grad_norm": 3.8421390056610107,
      "learning_rate": 5.002753590829352e-06,
      "loss": 0.23622764647006989,
      "step": 686
    },
    {
      "epoch": 2.927659574468085,
      "grad_norm": 0.163557767868042,
      "learning_rate": 4.9077432185697e-06,
      "loss": 0.33212676644325256,
      "step": 688
    },
    {
      "epoch": 2.9361702127659575,
      "grad_norm": 0.143082395195961,
      "learning_rate": 4.822530359503391e-06,
      "loss": 0.7329955697059631,
      "step": 690
    },
    {
      "epoch": 2.94468085106383,
      "grad_norm": 0.2611314058303833,
      "learning_rate": 4.747182657184251e-06,
      "loss": 0.43970412015914917,
      "step": 692
    },
    {
      "epoch": 2.953191489361702,
      "grad_norm": 0.12711837887763977,
      "learning_rate": 4.681759924021033e-06,
      "loss": 0.642443835735321,
      "step": 694
    },
    {
      "epoch": 2.9617021276595743,
      "grad_norm": 0.30099472403526306,
      "learning_rate": 4.626314093797213e-06,
      "loss": 0.35701537132263184,
      "step": 696
    },
    {
      "epoch": 2.970212765957447,
      "grad_norm": 0.13871702551841736,
      "learning_rate": 4.580889180444988e-06,
      "loss": 0.48086193203926086,
      "step": 698
    },
    {
      "epoch": 2.978723404255319,
      "grad_norm": 0.1269788146018982,
      "learning_rate": 4.5455212431061985e-06,
      "loss": 0.4837539792060852,
      "step": 700
    },
    {
      "epoch": 2.9872340425531916,
      "grad_norm": 0.31502285599708557,
      "learning_rate": 4.520238357507899e-06,
      "loss": 0.5840956568717957,
      "step": 702
    },
    {
      "epoch": 2.9957446808510637,
      "grad_norm": 0.1223224624991417,
      "learning_rate": 4.505060593675342e-06,
      "loss": 0.34119483828544617,
      "step": 704
    },
    {
      "epoch": 3.0,
      "step": 705,
      "total_flos": 2.502111532985352e+18,
      "train_loss": 0.8120331417161523,
      "train_runtime": 7663.1544,
      "train_samples_per_second": 5.888,
      "train_steps_per_second": 0.092
    }
  ],
  "logging_steps": 2,
  "max_steps": 705,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 99999,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.502111532985352e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}