{
  "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": 0.9370151162147522,
      "learning_rate": 8.333333333333333e-07,
      "loss": 4.115976333618164,
      "step": 2
    },
    {
      "epoch": 0.01702127659574468,
      "grad_norm": 0.41166678071022034,
      "learning_rate": 2.4999999999999998e-06,
      "loss": 1.878818154335022,
      "step": 4
    },
    {
      "epoch": 0.02553191489361702,
      "grad_norm": 0.4669341444969177,
      "learning_rate": 4.166666666666667e-06,
      "loss": 2.0509164333343506,
      "step": 6
    },
    {
      "epoch": 0.03404255319148936,
      "grad_norm": 0.6062728762626648,
      "learning_rate": 5.833333333333334e-06,
      "loss": 1.7545194625854492,
      "step": 8
    },
    {
      "epoch": 0.0425531914893617,
      "grad_norm": 0.37637069821357727,
      "learning_rate": 7.5e-06,
      "loss": 1.7690430879592896,
      "step": 10
    },
    {
      "epoch": 0.05106382978723404,
      "grad_norm": 2.065159797668457,
      "learning_rate": 9.166666666666668e-06,
      "loss": 2.355275869369507,
      "step": 12
    },
    {
      "epoch": 0.059574468085106386,
      "grad_norm": 1.091937780380249,
      "learning_rate": 1.0833333333333334e-05,
      "loss": 1.876306176185608,
      "step": 14
    },
    {
      "epoch": 0.06808510638297872,
      "grad_norm": 0.49564921855926514,
      "learning_rate": 1.25e-05,
      "loss": 1.785620927810669,
      "step": 16
    },
    {
      "epoch": 0.07659574468085106,
      "grad_norm": 1.3353960514068604,
      "learning_rate": 1.4166666666666666e-05,
      "loss": 1.523728370666504,
      "step": 18
    },
    {
      "epoch": 0.0851063829787234,
      "grad_norm": 0.30626556277275085,
      "learning_rate": 1.5833333333333333e-05,
      "loss": 1.510640025138855,
      "step": 20
    },
    {
      "epoch": 0.09361702127659574,
      "grad_norm": 0.1502075493335724,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 1.4279940128326416,
      "step": 22
    },
    {
      "epoch": 0.10212765957446808,
      "grad_norm": 0.8587890863418579,
      "learning_rate": 1.9166666666666667e-05,
      "loss": 1.1794553995132446,
      "step": 24
    },
    {
      "epoch": 0.11063829787234042,
      "grad_norm": 1.0492326021194458,
      "learning_rate": 2.0833333333333333e-05,
      "loss": 1.2579714059829712,
      "step": 26
    },
    {
      "epoch": 0.11914893617021277,
      "grad_norm": 0.264726847410202,
      "learning_rate": 2.25e-05,
      "loss": 0.9510668516159058,
      "step": 28
    },
    {
      "epoch": 0.1276595744680851,
      "grad_norm": 0.2859492301940918,
      "learning_rate": 2.4166666666666667e-05,
      "loss": 1.656806230545044,
      "step": 30
    },
    {
      "epoch": 0.13617021276595745,
      "grad_norm": 1.1864591836929321,
      "learning_rate": 2.5833333333333336e-05,
      "loss": 0.683425784111023,
      "step": 32
    },
    {
      "epoch": 0.14468085106382977,
      "grad_norm": 0.22737205028533936,
      "learning_rate": 2.75e-05,
      "loss": 1.2381809949874878,
      "step": 34
    },
    {
      "epoch": 0.15319148936170213,
      "grad_norm": 0.32182246446609497,
      "learning_rate": 2.9166666666666666e-05,
      "loss": 1.422176480293274,
      "step": 36
    },
    {
      "epoch": 0.16170212765957448,
      "grad_norm": 0.36561885476112366,
      "learning_rate": 2.9998734788806287e-05,
      "loss": 1.2621259689331055,
      "step": 38
    },
    {
      "epoch": 0.1702127659574468,
      "grad_norm": 0.7920733690261841,
      "learning_rate": 2.9988614605803806e-05,
      "loss": 1.2012425661087036,
      "step": 40
    },
    {
      "epoch": 0.17872340425531916,
      "grad_norm": 0.19826872646808624,
      "learning_rate": 2.99683822733885e-05,
      "loss": 1.2804166078567505,
      "step": 42
    },
    {
      "epoch": 0.18723404255319148,
      "grad_norm": 0.5331127643585205,
      "learning_rate": 2.9938053852362484e-05,
      "loss": 1.1188151836395264,
      "step": 44
    },
    {
      "epoch": 0.19574468085106383,
      "grad_norm": 0.22702525556087494,
      "learning_rate": 2.989765341799095e-05,
      "loss": 1.338016152381897,
      "step": 46
    },
    {
      "epoch": 0.20425531914893616,
      "grad_norm": 1.6932013034820557,
      "learning_rate": 2.9847213040890793e-05,
      "loss": 0.8224107027053833,
      "step": 48
    },
    {
      "epoch": 0.2127659574468085,
      "grad_norm": 0.49766504764556885,
      "learning_rate": 2.9786772761572335e-05,
      "loss": 0.9487060308456421,
      "step": 50
    },
    {
      "epoch": 0.22127659574468084,
      "grad_norm": 0.20401130616664886,
      "learning_rate": 2.9716380558654445e-05,
      "loss": 1.3093581199645996,
      "step": 52
    },
    {
      "epoch": 0.2297872340425532,
      "grad_norm": 0.4012207090854645,
      "learning_rate": 2.9636092310778195e-05,
      "loss": 1.2255353927612305,
      "step": 54
    },
    {
      "epoch": 0.23829787234042554,
      "grad_norm": 0.8393444418907166,
      "learning_rate": 2.954597175224938e-05,
      "loss": 0.8253700137138367,
      "step": 56
    },
    {
      "epoch": 0.24680851063829787,
      "grad_norm": 0.27147796750068665,
      "learning_rate": 2.9446090422445016e-05,
      "loss": 1.0494465827941895,
      "step": 58
    },
    {
      "epoch": 0.2553191489361702,
      "grad_norm": 0.42597004771232605,
      "learning_rate": 2.9336527609024072e-05,
      "loss": 1.358299970626831,
      "step": 60
    },
    {
      "epoch": 0.26382978723404255,
      "grad_norm": 0.20699705183506012,
      "learning_rate": 2.9217370284987434e-05,
      "loss": 1.335315465927124,
      "step": 62
    },
    {
      "epoch": 0.2723404255319149,
      "grad_norm": 0.16389895975589752,
      "learning_rate": 2.9088713039637117e-05,
      "loss": 1.2662416696548462,
      "step": 64
    },
    {
      "epoch": 0.28085106382978725,
      "grad_norm": 0.25043565034866333,
      "learning_rate": 2.8950658003489534e-05,
      "loss": 1.2614431381225586,
      "step": 66
    },
    {
      "epoch": 0.28936170212765955,
      "grad_norm": 0.19853799045085907,
      "learning_rate": 2.880331476720238e-05,
      "loss": 1.2736550569534302,
      "step": 68
    },
    {
      "epoch": 0.2978723404255319,
      "grad_norm": 2.3273322582244873,
      "learning_rate": 2.8646800294579517e-05,
      "loss": 1.2623916864395142,
      "step": 70
    },
    {
      "epoch": 0.30638297872340425,
      "grad_norm": 0.25601401925086975,
      "learning_rate": 2.848123882972295e-05,
      "loss": 1.2526882886886597,
      "step": 72
    },
    {
      "epoch": 0.3148936170212766,
      "grad_norm": 0.342012494802475,
      "learning_rate": 2.8306761798405526e-05,
      "loss": 0.9114877581596375,
      "step": 74
    },
    {
      "epoch": 0.32340425531914896,
      "grad_norm": 0.5187206864356995,
      "learning_rate": 2.812350770374273e-05,
      "loss": 1.2244638204574585,
      "step": 76
    },
    {
      "epoch": 0.33191489361702126,
      "grad_norm": 0.2703041434288025,
      "learning_rate": 2.793162201624631e-05,
      "loss": 1.2725831270217896,
      "step": 78
    },
    {
      "epoch": 0.3404255319148936,
      "grad_norm": 0.1924225091934204,
      "learning_rate": 2.77312570583471e-05,
      "loss": 0.7258127927780151,
      "step": 80
    },
    {
      "epoch": 0.34893617021276596,
      "grad_norm": 0.25936806201934814,
      "learning_rate": 2.752257188347862e-05,
      "loss": 1.2518136501312256,
      "step": 82
    },
    {
      "epoch": 0.3574468085106383,
      "grad_norm": 0.1409609615802765,
      "learning_rate": 2.730573214981751e-05,
      "loss": 0.9943649172782898,
      "step": 84
    },
    {
      "epoch": 0.3659574468085106,
      "grad_norm": 0.18737079203128815,
      "learning_rate": 2.7080909988780982e-05,
      "loss": 1.001371145248413,
      "step": 86
    },
    {
      "epoch": 0.37446808510638296,
      "grad_norm": 0.512141227722168,
      "learning_rate": 2.684828386838569e-05,
      "loss": 1.2457740306854248,
      "step": 88
    },
    {
      "epoch": 0.3829787234042553,
      "grad_norm": 0.7956476807594299,
      "learning_rate": 2.6608038451576528e-05,
      "loss": 0.9049152135848999,
      "step": 90
    },
    {
      "epoch": 0.39148936170212767,
      "grad_norm": 0.17900533974170685,
      "learning_rate": 2.636036444963769e-05,
      "loss": 1.0495070219039917,
      "step": 92
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.28203219175338745,
      "learning_rate": 2.6105458470802563e-05,
      "loss": 1.163374900817871,
      "step": 94
    },
    {
      "epoch": 0.4085106382978723,
      "grad_norm": 0.2998636066913605,
      "learning_rate": 2.5843522864182394e-05,
      "loss": 1.064250111579895,
      "step": 96
    },
    {
      "epoch": 0.41702127659574467,
      "grad_norm": 0.34955132007598877,
      "learning_rate": 2.557476555913785e-05,
      "loss": 1.0862053632736206,
      "step": 98
    },
    {
      "epoch": 0.425531914893617,
      "grad_norm": 0.5751879215240479,
      "learning_rate": 2.5299399900220803e-05,
      "loss": 0.9672101736068726,
      "step": 100
    },
    {
      "epoch": 0.4340425531914894,
      "grad_norm": 0.556722104549408,
      "learning_rate": 2.5017644477817424e-05,
      "loss": 1.38407564163208,
      "step": 102
    },
    {
      "epoch": 0.4425531914893617,
      "grad_norm": 0.24831920862197876,
      "learning_rate": 2.47297229546271e-05,
      "loss": 1.0887106657028198,
      "step": 104
    },
    {
      "epoch": 0.451063829787234,
      "grad_norm": 0.39581331610679626,
      "learning_rate": 2.4435863888114814e-05,
      "loss": 0.8722383975982666,
      "step": 106
    },
    {
      "epoch": 0.4595744680851064,
      "grad_norm": 0.16548699140548706,
      "learning_rate": 2.4136300549077976e-05,
      "loss": 1.2673208713531494,
      "step": 108
    },
    {
      "epoch": 0.46808510638297873,
      "grad_norm": 0.20999404788017273,
      "learning_rate": 2.3831270736471703e-05,
      "loss": 1.2440522909164429,
      "step": 110
    },
    {
      "epoch": 0.4765957446808511,
      "grad_norm": 0.18370361626148224,
      "learning_rate": 2.352101658863959e-05,
      "loss": 1.1419363021850586,
      "step": 112
    },
    {
      "epoch": 0.4851063829787234,
      "grad_norm": 0.24067164957523346,
      "learning_rate": 2.32057843910998e-05,
      "loss": 1.0633783340454102,
      "step": 114
    },
    {
      "epoch": 0.49361702127659574,
      "grad_norm": 0.47072628140449524,
      "learning_rate": 2.288582438103903e-05,
      "loss": 1.179775595664978,
      "step": 116
    },
    {
      "epoch": 0.502127659574468,
      "grad_norm": 0.2161823809146881,
      "learning_rate": 2.256139054866955e-05,
      "loss": 1.2002534866333008,
      "step": 118
    },
    {
      "epoch": 0.5106382978723404,
      "grad_norm": 1.2126752138137817,
      "learning_rate": 2.2232740435607067e-05,
      "loss": 0.9621443748474121,
      "step": 120
    },
    {
      "epoch": 0.5191489361702127,
      "grad_norm": 0.1435491442680359,
      "learning_rate": 2.19001349304294e-05,
      "loss": 1.253274917602539,
      "step": 122
    },
    {
      "epoch": 0.5276595744680851,
      "grad_norm": 0.19416141510009766,
      "learning_rate": 2.156383806157826e-05,
      "loss": 1.3183720111846924,
      "step": 124
    },
    {
      "epoch": 0.5361702127659574,
      "grad_norm": 0.14461560547351837,
      "learning_rate": 2.1224116787768552e-05,
      "loss": 1.238662838935852,
      "step": 126
    },
    {
      "epoch": 0.5446808510638298,
      "grad_norm": 0.155701145529747,
      "learning_rate": 2.0881240786071588e-05,
      "loss": 1.1723605394363403,
      "step": 128
    },
    {
      "epoch": 0.5531914893617021,
      "grad_norm": 0.17010867595672607,
      "learning_rate": 2.05354822378404e-05,
      "loss": 1.021695852279663,
      "step": 130
    },
    {
      "epoch": 0.5617021276595745,
      "grad_norm": 0.17941494286060333,
      "learning_rate": 2.018711561264714e-05,
      "loss": 1.2407881021499634,
      "step": 132
    },
    {
      "epoch": 0.5702127659574469,
      "grad_norm": 0.3246072828769684,
      "learning_rate": 1.9836417450403978e-05,
      "loss": 1.186881422996521,
      "step": 134
    },
    {
      "epoch": 0.5787234042553191,
      "grad_norm": 0.32305997610092163,
      "learning_rate": 1.9483666141840615e-05,
      "loss": 0.965923547744751,
      "step": 136
    },
    {
      "epoch": 0.5872340425531914,
      "grad_norm": 0.15171490609645844,
      "learning_rate": 1.9129141707512508e-05,
      "loss": 0.5913432836532593,
      "step": 138
    },
    {
      "epoch": 0.5957446808510638,
      "grad_norm": 0.37349367141723633,
      "learning_rate": 1.8773125575515364e-05,
      "loss": 1.1055043935775757,
      "step": 140
    },
    {
      "epoch": 0.6042553191489362,
      "grad_norm": 0.15140922367572784,
      "learning_rate": 1.8415900358082268e-05,
      "loss": 1.3598577976226807,
      "step": 142
    },
    {
      "epoch": 0.6127659574468085,
      "grad_norm": 0.5196365118026733,
      "learning_rate": 1.805774962724083e-05,
      "loss": 0.9218302369117737,
      "step": 144
    },
    {
      "epoch": 0.6212765957446809,
      "grad_norm": 0.15164780616760254,
      "learning_rate": 1.7698957689708426e-05,
      "loss": 1.1118155717849731,
      "step": 146
    },
    {
      "epoch": 0.6297872340425532,
      "grad_norm": 0.2984839677810669,
      "learning_rate": 1.7339809361204252e-05,
      "loss": 1.2171759605407715,
      "step": 148
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 0.28176164627075195,
      "learning_rate": 1.6980589740357294e-05,
      "loss": 1.1775646209716797,
      "step": 150
    },
    {
      "epoch": 0.6468085106382979,
      "grad_norm": 0.19051885604858398,
      "learning_rate": 1.6621583982389707e-05,
      "loss": 1.1114767789840698,
      "step": 152
    },
    {
      "epoch": 0.6553191489361702,
      "grad_norm": 0.28726738691329956,
      "learning_rate": 1.6263077072755326e-05,
      "loss": 0.7064566612243652,
      "step": 154
    },
    {
      "epoch": 0.6638297872340425,
      "grad_norm": 0.3889693021774292,
      "learning_rate": 1.5905353600912898e-05,
      "loss": 0.8802153468132019,
      "step": 156
    },
    {
      "epoch": 0.6723404255319149,
      "grad_norm": 0.28420913219451904,
      "learning_rate": 1.5548697534413646e-05,
      "loss": 0.9915045499801636,
      "step": 158
    },
    {
      "epoch": 0.6808510638297872,
      "grad_norm": 0.2218930721282959,
      "learning_rate": 1.5193391993482582e-05,
      "loss": 1.2188233137130737,
      "step": 160
    },
    {
      "epoch": 0.6893617021276596,
      "grad_norm": 0.1267762929201126,
      "learning_rate": 1.4839719026272377e-05,
      "loss": 1.0931644439697266,
      "step": 162
    },
    {
      "epoch": 0.6978723404255319,
      "grad_norm": 0.20958174765110016,
      "learning_rate": 1.4487959384968272e-05,
      "loss": 0.875884473323822,
      "step": 164
    },
    {
      "epoch": 0.7063829787234043,
      "grad_norm": 0.2075873166322708,
      "learning_rate": 1.4138392302921813e-05,
      "loss": 0.9931239485740662,
      "step": 166
    },
    {
      "epoch": 0.7148936170212766,
      "grad_norm": 0.45486778020858765,
      "learning_rate": 1.3791295272990175e-05,
      "loss": 1.1112878322601318,
      "step": 168
    },
    {
      "epoch": 0.723404255319149,
      "grad_norm": 0.12606321275234222,
      "learning_rate": 1.344694382725718e-05,
      "loss": 0.8394916653633118,
      "step": 170
    },
    {
      "epoch": 0.7319148936170212,
      "grad_norm": 1.405964732170105,
      "learning_rate": 1.3105611318310818e-05,
      "loss": 0.8859961032867432,
      "step": 172
    },
    {
      "epoch": 0.7404255319148936,
      "grad_norm": 0.24160844087600708,
      "learning_rate": 1.2767568702250844e-05,
      "loss": 1.1839947700500488,
      "step": 174
    },
    {
      "epoch": 0.7489361702127659,
      "grad_norm": 0.28286510705947876,
      "learning_rate": 1.2433084323598791e-05,
      "loss": 0.6160851120948792,
      "step": 176
    },
    {
      "epoch": 0.7574468085106383,
      "grad_norm": 0.36609017848968506,
      "learning_rate": 1.2102423702281116e-05,
      "loss": 1.0064830780029297,
      "step": 178
    },
    {
      "epoch": 0.7659574468085106,
      "grad_norm": 0.25650474429130554,
      "learning_rate": 1.1775849322854516e-05,
      "loss": 1.0609067678451538,
      "step": 180
    },
    {
      "epoch": 0.774468085106383,
      "grad_norm": 0.3248478174209595,
      "learning_rate": 1.1453620426140795e-05,
      "loss": 0.7533571720123291,
      "step": 182
    },
    {
      "epoch": 0.7829787234042553,
      "grad_norm": 0.22153323888778687,
      "learning_rate": 1.1135992803436695e-05,
      "loss": 0.9635902643203735,
      "step": 184
    },
    {
      "epoch": 0.7914893617021277,
      "grad_norm": 0.11977202445268631,
      "learning_rate": 1.0823218593461992e-05,
      "loss": 0.9765074849128723,
      "step": 186
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2166285663843155,
      "learning_rate": 1.0515546082207097e-05,
      "loss": 0.730889081954956,
      "step": 188
    },
    {
      "epoch": 0.8085106382978723,
      "grad_norm": 0.6851885914802551,
      "learning_rate": 1.0213219505838983e-05,
      "loss": 0.966530442237854,
      "step": 190
    },
    {
      "epoch": 0.8170212765957446,
      "grad_norm": 0.33713918924331665,
      "learning_rate": 9.91647885682201e-06,
      "loss": 1.3054033517837524,
      "step": 192
    },
    {
      "epoch": 0.825531914893617,
      "grad_norm": 0.196068674325943,
      "learning_rate": 9.625559693407413e-06,
      "loss": 1.2358617782592773,
      "step": 194
    },
    {
      "epoch": 0.8340425531914893,
      "grad_norm": 0.2631889283657074,
      "learning_rate": 9.340692952642789e-06,
      "loss": 0.789035975933075,
      "step": 196
    },
    {
      "epoch": 0.8425531914893617,
      "grad_norm": 0.165288507938385,
      "learning_rate": 9.062104767049956e-06,
      "loss": 1.2202699184417725,
      "step": 198
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 0.4081096649169922,
      "learning_rate": 8.790016285116763e-06,
      "loss": 1.005751132965088,
      "step": 200
    },
    {
      "epoch": 0.8595744680851064,
      "grad_norm": 0.16539311408996582,
      "learning_rate": 8.524643495745306e-06,
      "loss": 0.9970273971557617,
      "step": 202
    },
    {
      "epoch": 0.8680851063829788,
      "grad_norm": 0.1629868745803833,
      "learning_rate": 8.26619705679589e-06,
      "loss": 1.1567330360412598,
      "step": 204
    },
    {
      "epoch": 0.8765957446808511,
      "grad_norm": 0.13489580154418945,
      "learning_rate": 8.014882127862923e-06,
      "loss": 1.1192888021469116,
      "step": 206
    },
    {
      "epoch": 0.8851063829787233,
      "grad_norm": 0.1189538910984993,
      "learning_rate": 7.770898207415416e-06,
      "loss": 0.9340039491653442,
      "step": 208
    },
    {
      "epoch": 0.8936170212765957,
      "grad_norm": 0.12542982399463654,
      "learning_rate": 7.534438974431351e-06,
      "loss": 1.2615220546722412,
      "step": 210
    },
    {
      "epoch": 0.902127659574468,
      "grad_norm": 0.2067514955997467,
      "learning_rate": 7.305692134651742e-06,
      "loss": 0.7793570160865784,
      "step": 212
    },
    {
      "epoch": 0.9106382978723404,
      "grad_norm": 0.12745940685272217,
      "learning_rate": 7.084839271576291e-06,
      "loss": 0.9480925798416138,
      "step": 214
    },
    {
      "epoch": 0.9191489361702128,
      "grad_norm": 0.14118105173110962,
      "learning_rate": 6.872055702319054e-06,
      "loss": 0.7321507334709167,
      "step": 216
    },
    {
      "epoch": 0.9276595744680851,
      "grad_norm": 0.19066962599754333,
      "learning_rate": 6.667510338438419e-06,
      "loss": 1.2155194282531738,
      "step": 218
    },
    {
      "epoch": 0.9361702127659575,
      "grad_norm": 0.2428826540708542,
      "learning_rate": 6.471365551852012e-06,
      "loss": 0.39672887325286865,
      "step": 220
    },
    {
      "epoch": 0.9446808510638298,
      "grad_norm": 0.16802483797073364,
      "learning_rate": 6.2837770459428e-06,
      "loss": 1.254830241203308,
      "step": 222
    },
    {
      "epoch": 0.9531914893617022,
      "grad_norm": 0.1394779235124588,
      "learning_rate": 6.1048937319588676e-06,
      "loss": 1.138830304145813,
      "step": 224
    },
    {
      "epoch": 0.9617021276595744,
      "grad_norm": 0.2871001362800598,
      "learning_rate": 5.9348576108049065e-06,
      "loss": 0.8572854399681091,
      "step": 226
    },
    {
      "epoch": 0.9702127659574468,
      "grad_norm": 0.1366124153137207,
      "learning_rate": 5.773803660319234e-06,
      "loss": 1.0662239789962769,
      "step": 228
    },
    {
      "epoch": 0.9787234042553191,
      "grad_norm": 0.12555794417858124,
      "learning_rate": 5.621859728125884e-06,
      "loss": 1.222013235092163,
      "step": 230
    },
    {
      "epoch": 0.9872340425531915,
      "grad_norm": 0.29969581961631775,
      "learning_rate": 5.479146430146781e-06,
      "loss": 1.0417675971984863,
      "step": 232
    },
    {
      "epoch": 0.9957446808510638,
      "grad_norm": 0.4836138188838959,
      "learning_rate": 5.345777054854579e-06,
      "loss": 0.9495846629142761,
      "step": 234
    },
    {
      "epoch": 1.004255319148936,
      "grad_norm": 0.28257113695144653,
      "learning_rate": 5.221857473342149e-06,
      "loss": 0.843216061592102,
      "step": 236
    },
    {
      "epoch": 1.0127659574468084,
      "grad_norm": 0.1423388570547104,
      "learning_rate": 5.10748605528015e-06,
      "loss": 0.9633752107620239,
      "step": 238
    },
    {
      "epoch": 1.0212765957446808,
      "grad_norm": 0.12806400656700134,
      "learning_rate": 5.002753590829349e-06,
      "loss": 0.6067866683006287,
      "step": 240
    },
    {
      "epoch": 1.0297872340425531,
      "grad_norm": 0.243302121758461,
      "learning_rate": 4.9077432185697e-06,
      "loss": 0.9341376423835754,
      "step": 242
    },
    {
      "epoch": 1.0382978723404255,
      "grad_norm": 0.22620677947998047,
      "learning_rate": 4.822530359503393e-06,
      "loss": 1.046706199645996,
      "step": 244
    },
    {
      "epoch": 1.0468085106382978,
      "grad_norm": 0.1606007069349289,
      "learning_rate": 4.747182657184251e-06,
      "loss": 0.6613105535507202,
      "step": 246
    },
    {
      "epoch": 1.0553191489361702,
      "grad_norm": 0.3230328857898712,
      "learning_rate": 4.681759924021033e-06,
      "loss": 0.7598414421081543,
      "step": 248
    },
    {
      "epoch": 1.0638297872340425,
      "grad_norm": 0.13726186752319336,
      "learning_rate": 4.626314093797213e-06,
      "loss": 0.7334257364273071,
      "step": 250
    },
    {
      "epoch": 1.0723404255319149,
      "grad_norm": 0.3747357726097107,
      "learning_rate": 4.580889180444988e-06,
      "loss": 0.5062970519065857,
      "step": 252
    },
    {
      "epoch": 1.0808510638297872,
      "grad_norm": 0.15663602948188782,
      "learning_rate": 4.545521243106197e-06,
      "loss": 1.0599923133850098,
      "step": 254
    },
    {
      "epoch": 1.0893617021276596,
      "grad_norm": 0.1843012273311615,
      "learning_rate": 4.520238357507899e-06,
      "loss": 0.940389096736908,
      "step": 256
    },
    {
      "epoch": 1.097872340425532,
      "grad_norm": 0.22409912943840027,
      "learning_rate": 4.505060593675342e-06,
      "loss": 1.0018244981765747,
      "step": 258
    },
    {
      "epoch": 1.1063829787234043,
      "grad_norm": 0.45610666275024414,
      "learning_rate": 4.5e-06,
      "loss": 0.8150784969329834,
      "step": 260
    },
    {
      "epoch": 1.1148936170212767,
      "grad_norm": 0.19215510785579681,
      "learning_rate": 4.505060593675342e-06,
      "loss": 0.7950176000595093,
      "step": 262
    },
    {
      "epoch": 1.123404255319149,
      "grad_norm": 0.23423391580581665,
      "learning_rate": 4.520238357507899e-06,
      "loss": 0.7621920704841614,
      "step": 264
    },
    {
      "epoch": 1.1319148936170214,
      "grad_norm": 0.15685321390628815,
      "learning_rate": 4.545521243106197e-06,
      "loss": 0.8206462264060974,
      "step": 266
    },
    {
      "epoch": 1.1404255319148937,
      "grad_norm": 0.4606432020664215,
      "learning_rate": 4.580889180444988e-06,
      "loss": 0.9801955819129944,
      "step": 268
    },
    {
      "epoch": 1.148936170212766,
      "grad_norm": 0.21424169838428497,
      "learning_rate": 4.6263140937972124e-06,
      "loss": 0.8219439387321472,
      "step": 270
    },
    {
      "epoch": 1.1574468085106382,
      "grad_norm": 0.2082989364862442,
      "learning_rate": 4.6817599240210315e-06,
      "loss": 0.6543676257133484,
      "step": 272
    },
    {
      "epoch": 1.1659574468085105,
      "grad_norm": 0.1622384637594223,
      "learning_rate": 4.747182657184251e-06,
      "loss": 0.9012349247932434,
      "step": 274
    },
    {
      "epoch": 1.174468085106383,
      "grad_norm": 0.13174618780612946,
      "learning_rate": 4.822530359503391e-06,
      "loss": 0.7129980325698853,
      "step": 276
    },
    {
      "epoch": 1.1829787234042553,
      "grad_norm": 0.2709389925003052,
      "learning_rate": 4.9077432185697e-06,
      "loss": 0.5482591986656189,
      "step": 278
    },
    {
      "epoch": 1.1914893617021276,
      "grad_norm": 0.2576906979084015,
      "learning_rate": 5.002753590829348e-06,
      "loss": 0.5497639775276184,
      "step": 280
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.1526031494140625,
      "learning_rate": 5.1074860552801466e-06,
      "loss": 0.6242255568504333,
      "step": 282
    },
    {
      "epoch": 1.2085106382978723,
      "grad_norm": 0.22073198854923248,
      "learning_rate": 5.2218574733421455e-06,
      "loss": 0.693662703037262,
      "step": 284
    },
    {
      "epoch": 1.2170212765957447,
      "grad_norm": 0.22191239893436432,
      "learning_rate": 5.345777054854578e-06,
      "loss": 0.9797627329826355,
      "step": 286
    },
    {
      "epoch": 1.225531914893617,
      "grad_norm": 0.17303584516048431,
      "learning_rate": 5.479146430146783e-06,
      "loss": 0.9435557126998901,
      "step": 288
    },
    {
      "epoch": 1.2340425531914894,
      "grad_norm": 0.2218431681394577,
      "learning_rate": 5.6218597281258834e-06,
      "loss": 0.6460038423538208,
      "step": 290
    },
    {
      "epoch": 1.2425531914893617,
      "grad_norm": 0.22332793474197388,
      "learning_rate": 5.773803660319232e-06,
      "loss": 0.8468654155731201,
      "step": 292
    },
    {
      "epoch": 1.251063829787234,
      "grad_norm": 0.13609401881694794,
      "learning_rate": 5.934857610804904e-06,
      "loss": 0.8879528641700745,
      "step": 294
    },
    {
      "epoch": 1.2595744680851064,
      "grad_norm": 0.18097440898418427,
      "learning_rate": 6.1048937319588676e-06,
      "loss": 0.7949476838111877,
      "step": 296
    },
    {
      "epoch": 1.2680851063829788,
      "grad_norm": 0.3170534074306488,
      "learning_rate": 6.283777045942798e-06,
      "loss": 0.8318912386894226,
      "step": 298
    },
    {
      "epoch": 1.2765957446808511,
      "grad_norm": 0.19770559668540955,
      "learning_rate": 6.471365551852011e-06,
      "loss": 1.0059137344360352,
      "step": 300
    },
    {
      "epoch": 1.2851063829787235,
      "grad_norm": 0.14630961418151855,
      "learning_rate": 6.66751033843842e-06,
      "loss": 0.39432206749916077,
      "step": 302
    },
    {
      "epoch": 1.2936170212765958,
      "grad_norm": 0.3450065553188324,
      "learning_rate": 6.872055702319048e-06,
      "loss": 0.9369863271713257,
      "step": 304
    },
    {
      "epoch": 1.302127659574468,
      "grad_norm": 0.18490372598171234,
      "learning_rate": 7.084839271576289e-06,
      "loss": 0.9826689958572388,
      "step": 306
    },
    {
      "epoch": 1.3106382978723405,
      "grad_norm": 0.1866351068019867,
      "learning_rate": 7.30569213465174e-06,
      "loss": 0.5573344826698303,
      "step": 308
    },
    {
      "epoch": 1.3191489361702127,
      "grad_norm": 0.1976342350244522,
      "learning_rate": 7.534438974431351e-06,
      "loss": 0.5633378624916077,
      "step": 310
    },
    {
      "epoch": 1.327659574468085,
      "grad_norm": 0.3162063956260681,
      "learning_rate": 7.770898207415414e-06,
      "loss": 0.7375507354736328,
      "step": 312
    },
    {
      "epoch": 1.3361702127659574,
      "grad_norm": 0.2016829252243042,
      "learning_rate": 8.014882127862923e-06,
      "loss": 0.4668198823928833,
      "step": 314
    },
    {
      "epoch": 1.3446808510638297,
      "grad_norm": 1.459976315498352,
      "learning_rate": 8.266197056795886e-06,
      "loss": 0.7280799150466919,
      "step": 316
    },
    {
      "epoch": 1.353191489361702,
      "grad_norm": 0.3398215174674988,
      "learning_rate": 8.5246434957453e-06,
      "loss": 0.5291551351547241,
      "step": 318
    },
    {
      "epoch": 1.3617021276595744,
      "grad_norm": 0.19973713159561157,
      "learning_rate": 8.790016285116766e-06,
      "loss": 0.5350017547607422,
      "step": 320
    },
    {
      "epoch": 1.3702127659574468,
      "grad_norm": 0.14308671653270721,
      "learning_rate": 9.062104767049957e-06,
      "loss": 0.9615387916564941,
      "step": 322
    },
    {
      "epoch": 1.3787234042553191,
      "grad_norm": 0.1939127892255783,
      "learning_rate": 9.340692952642787e-06,
      "loss": 0.6444461345672607,
      "step": 324
    },
    {
      "epoch": 1.3872340425531915,
      "grad_norm": 0.18625043332576752,
      "learning_rate": 9.625559693407413e-06,
      "loss": 0.9194571375846863,
      "step": 326
    },
    {
      "epoch": 1.3957446808510638,
      "grad_norm": 0.2951974868774414,
      "learning_rate": 9.916478856822006e-06,
      "loss": 0.9841423034667969,
      "step": 328
    },
    {
      "epoch": 1.4042553191489362,
      "grad_norm": 0.16953378915786743,
      "learning_rate": 1.0213219505838985e-05,
      "loss": 0.8577695488929749,
      "step": 330
    },
    {
      "epoch": 1.4127659574468086,
      "grad_norm": 0.2879262864589691,
      "learning_rate": 1.0515546082207094e-05,
      "loss": 0.8521637916564941,
      "step": 332
    },
    {
      "epoch": 1.421276595744681,
      "grad_norm": 0.22187970578670502,
      "learning_rate": 1.082321859346199e-05,
      "loss": 0.6609964966773987,
      "step": 334
    },
    {
      "epoch": 1.4297872340425533,
      "grad_norm": 0.19994378089904785,
      "learning_rate": 1.1135992803436696e-05,
      "loss": 0.5523483753204346,
      "step": 336
    },
    {
      "epoch": 1.4382978723404256,
      "grad_norm": 0.19332869350910187,
      "learning_rate": 1.1453620426140791e-05,
      "loss": 0.7419446110725403,
      "step": 338
    },
    {
      "epoch": 1.4468085106382977,
      "grad_norm": 0.17565295100212097,
      "learning_rate": 1.1775849322854508e-05,
      "loss": 0.7033047676086426,
      "step": 340
    },
    {
      "epoch": 1.4553191489361703,
      "grad_norm": 0.3124133050441742,
      "learning_rate": 1.210242370228112e-05,
      "loss": 0.5836398005485535,
      "step": 342
    },
    {
      "epoch": 1.4638297872340424,
      "grad_norm": 0.21754996478557587,
      "learning_rate": 1.2433084323598791e-05,
      "loss": 0.9612702131271362,
      "step": 344
    },
    {
      "epoch": 1.472340425531915,
      "grad_norm": 0.21039743721485138,
      "learning_rate": 1.2767568702250838e-05,
      "loss": 0.9444953799247742,
      "step": 346
    },
    {
      "epoch": 1.4808510638297872,
      "grad_norm": 0.42962566018104553,
      "learning_rate": 1.3105611318310811e-05,
      "loss": 0.8060452938079834,
      "step": 348
    },
    {
      "epoch": 1.4893617021276595,
      "grad_norm": 0.19513703882694244,
      "learning_rate": 1.3446943827257184e-05,
      "loss": 0.8150795698165894,
      "step": 350
    },
    {
      "epoch": 1.4978723404255319,
      "grad_norm": 0.2640067934989929,
      "learning_rate": 1.3791295272990172e-05,
      "loss": 0.3241785168647766,
      "step": 352
    },
    {
      "epoch": 1.5063829787234042,
      "grad_norm": 0.25144851207733154,
      "learning_rate": 1.413839230292182e-05,
      "loss": 0.7416596412658691,
      "step": 354
    },
    {
      "epoch": 1.5148936170212766,
      "grad_norm": 0.16774724423885345,
      "learning_rate": 1.4487959384968276e-05,
      "loss": 0.518079400062561,
      "step": 356
    },
    {
      "epoch": 1.523404255319149,
      "grad_norm": 0.1518772393465042,
      "learning_rate": 1.4839719026272373e-05,
      "loss": 0.5670950412750244,
      "step": 358
    },
    {
      "epoch": 1.5319148936170213,
      "grad_norm": 0.23323562741279602,
      "learning_rate": 1.5193391993482579e-05,
      "loss": 1.1480307579040527,
      "step": 360
    },
    {
      "epoch": 1.5404255319148936,
      "grad_norm": 0.1399996131658554,
      "learning_rate": 1.5548697534413636e-05,
      "loss": 0.5208263397216797,
      "step": 362
    },
    {
      "epoch": 1.548936170212766,
      "grad_norm": 0.19347982108592987,
      "learning_rate": 1.5905353600912894e-05,
      "loss": 0.9330644607543945,
      "step": 364
    },
    {
      "epoch": 1.5574468085106383,
      "grad_norm": 0.15133146941661835,
      "learning_rate": 1.6263077072755323e-05,
      "loss": 0.8360081911087036,
      "step": 366
    },
    {
      "epoch": 1.5659574468085107,
      "grad_norm": 0.18284079432487488,
      "learning_rate": 1.6621583982389707e-05,
      "loss": 0.7362462878227234,
      "step": 368
    },
    {
      "epoch": 1.574468085106383,
      "grad_norm": 0.19372136890888214,
      "learning_rate": 1.6980589740357294e-05,
      "loss": 0.6744348406791687,
      "step": 370
    },
    {
      "epoch": 1.5829787234042554,
      "grad_norm": 0.12256433069705963,
      "learning_rate": 1.733980936120425e-05,
      "loss": 0.9976662397384644,
      "step": 372
    },
    {
      "epoch": 1.5914893617021275,
      "grad_norm": 0.20529648661613464,
      "learning_rate": 1.7698957689708416e-05,
      "loss": 1.0515512228012085,
      "step": 374
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.16672298312187195,
      "learning_rate": 1.805774962724083e-05,
      "loss": 0.7891988754272461,
      "step": 376
    },
    {
      "epoch": 1.6085106382978722,
      "grad_norm": 0.1434396356344223,
      "learning_rate": 1.8415900358082265e-05,
      "loss": 0.6114462018013,
      "step": 378
    },
    {
      "epoch": 1.6170212765957448,
      "grad_norm": 0.16634748876094818,
      "learning_rate": 1.8773125575515357e-05,
      "loss": 0.5973749756813049,
      "step": 380
    },
    {
      "epoch": 1.625531914893617,
      "grad_norm": 0.13240687549114227,
      "learning_rate": 1.912914170751251e-05,
      "loss": 0.7943100929260254,
      "step": 382
    },
    {
      "epoch": 1.6340425531914895,
      "grad_norm": 0.12580837309360504,
      "learning_rate": 1.9483666141840612e-05,
      "loss": 0.9919679760932922,
      "step": 384
    },
    {
      "epoch": 1.6425531914893616,
      "grad_norm": 0.35519301891326904,
      "learning_rate": 1.9836417450403974e-05,
      "loss": 0.8425494432449341,
      "step": 386
    },
    {
      "epoch": 1.6510638297872342,
      "grad_norm": 0.2488897144794464,
      "learning_rate": 2.0187115612647133e-05,
      "loss": 0.7412829399108887,
      "step": 388
    },
    {
      "epoch": 1.6595744680851063,
      "grad_norm": 0.5074354410171509,
      "learning_rate": 2.05354822378404e-05,
      "loss": 0.7992886900901794,
      "step": 390
    },
    {
      "epoch": 1.6680851063829787,
      "grad_norm": 0.12573714554309845,
      "learning_rate": 2.0881240786071595e-05,
      "loss": 0.909046471118927,
      "step": 392
    },
    {
      "epoch": 1.676595744680851,
      "grad_norm": 0.33457598090171814,
      "learning_rate": 2.122411678776854e-05,
      "loss": 0.6243407726287842,
      "step": 394
    },
    {
      "epoch": 1.6851063829787234,
      "grad_norm": 0.15411563217639923,
      "learning_rate": 2.1563838061578258e-05,
      "loss": 0.8719924688339233,
      "step": 396
    },
    {
      "epoch": 1.6936170212765957,
      "grad_norm": 0.13289184868335724,
      "learning_rate": 2.1900134930429396e-05,
      "loss": 1.0623129606246948,
      "step": 398
    },
    {
      "epoch": 1.702127659574468,
      "grad_norm": 0.3402920663356781,
      "learning_rate": 2.223274043560706e-05,
      "loss": 0.9607300162315369,
      "step": 400
    },
    {
      "epoch": 1.7106382978723405,
      "grad_norm": 0.20696532726287842,
      "learning_rate": 2.256139054866954e-05,
      "loss": 0.4819542169570923,
      "step": 402
    },
    {
      "epoch": 1.7191489361702128,
      "grad_norm": 0.1925085484981537,
      "learning_rate": 2.2885824381039028e-05,
      "loss": 0.6816413402557373,
      "step": 404
    },
    {
      "epoch": 1.7276595744680852,
      "grad_norm": 0.5462839603424072,
      "learning_rate": 2.3205784391099808e-05,
      "loss": 0.5322574377059937,
      "step": 406
    },
    {
      "epoch": 1.7361702127659573,
      "grad_norm": 0.12952132523059845,
      "learning_rate": 2.352101658863958e-05,
      "loss": 0.8366644382476807,
      "step": 408
    },
    {
      "epoch": 1.7446808510638299,
      "grad_norm": 0.15898458659648895,
      "learning_rate": 2.3831270736471703e-05,
      "loss": 0.8219179511070251,
      "step": 410
    },
    {
      "epoch": 1.753191489361702,
      "grad_norm": 0.10134760290384293,
      "learning_rate": 2.4136300549077973e-05,
      "loss": 0.8943849802017212,
      "step": 412
    },
    {
      "epoch": 1.7617021276595746,
      "grad_norm": 0.21209417283535004,
      "learning_rate": 2.443586388811481e-05,
      "loss": 0.7499610781669617,
      "step": 414
    },
    {
      "epoch": 1.7702127659574467,
      "grad_norm": 0.47646665573120117,
      "learning_rate": 2.47297229546271e-05,
      "loss": 0.6121576428413391,
      "step": 416
    },
    {
      "epoch": 1.7787234042553193,
      "grad_norm": 0.26675277948379517,
      "learning_rate": 2.5017644477817418e-05,
      "loss": 0.7867974638938904,
      "step": 418
    },
    {
      "epoch": 1.7872340425531914,
      "grad_norm": 0.840756356716156,
      "learning_rate": 2.5299399900220807e-05,
      "loss": 0.6045423150062561,
      "step": 420
    },
    {
      "epoch": 1.795744680851064,
      "grad_norm": 0.400618314743042,
      "learning_rate": 2.5574765559137848e-05,
      "loss": 0.7998336553573608,
      "step": 422
    },
    {
      "epoch": 1.804255319148936,
      "grad_norm": 0.15446501970291138,
      "learning_rate": 2.584352286418239e-05,
      "loss": 0.7926132678985596,
      "step": 424
    },
    {
      "epoch": 1.8127659574468085,
      "grad_norm": 0.15063929557800293,
      "learning_rate": 2.6105458470802563e-05,
      "loss": 0.8235700130462646,
      "step": 426
    },
    {
      "epoch": 1.8212765957446808,
      "grad_norm": 0.11289381980895996,
      "learning_rate": 2.6360364449637683e-05,
      "loss": 0.8320169448852539,
      "step": 428
    },
    {
      "epoch": 1.8297872340425532,
      "grad_norm": 0.14190740883350372,
      "learning_rate": 2.6608038451576528e-05,
      "loss": 0.8250552415847778,
      "step": 430
    },
    {
      "epoch": 1.8382978723404255,
      "grad_norm": 0.3685653507709503,
      "learning_rate": 2.684828386838569e-05,
      "loss": 0.8021817207336426,
      "step": 432
    },
    {
      "epoch": 1.8468085106382979,
      "grad_norm": 0.12853652238845825,
      "learning_rate": 2.7080909988780982e-05,
      "loss": 0.7879712581634521,
      "step": 434
    },
    {
      "epoch": 1.8553191489361702,
      "grad_norm": 0.3018054962158203,
      "learning_rate": 2.7305732149817502e-05,
      "loss": 0.7762024998664856,
      "step": 436
    },
    {
      "epoch": 1.8638297872340426,
      "grad_norm": 0.14753182232379913,
      "learning_rate": 2.7522571883478617e-05,
      "loss": 1.2395200729370117,
      "step": 438
    },
    {
      "epoch": 1.872340425531915,
      "grad_norm": 0.1134040355682373,
      "learning_rate": 2.77312570583471e-05,
      "loss": 0.8664835691452026,
      "step": 440
    },
    {
      "epoch": 1.8808510638297873,
      "grad_norm": 0.19325752556324005,
      "learning_rate": 2.7931622016246304e-05,
      "loss": 1.0493484735488892,
      "step": 442
    },
    {
      "epoch": 1.8893617021276596,
      "grad_norm": 0.15366853773593903,
      "learning_rate": 2.8123507703742727e-05,
      "loss": 0.8971328139305115,
      "step": 444
    },
    {
      "epoch": 1.8978723404255318,
      "grad_norm": 0.1713661104440689,
      "learning_rate": 2.8306761798405522e-05,
      "loss": 0.9821380376815796,
      "step": 446
    },
    {
      "epoch": 1.9063829787234043,
      "grad_norm": 0.4014803469181061,
      "learning_rate": 2.848123882972295e-05,
      "loss": 1.0046873092651367,
      "step": 448
    },
    {
      "epoch": 1.9148936170212765,
      "grad_norm": 0.1893383413553238,
      "learning_rate": 2.8646800294579514e-05,
      "loss": 0.647858738899231,
      "step": 450
    },
    {
      "epoch": 1.923404255319149,
      "grad_norm": 0.5622705221176147,
      "learning_rate": 2.8803314767202376e-05,
      "loss": 0.5715224742889404,
      "step": 452
    },
    {
      "epoch": 1.9319148936170212,
      "grad_norm": 0.14517684280872345,
      "learning_rate": 2.8950658003489534e-05,
      "loss": 0.8555135726928711,
      "step": 454
    },
    {
      "epoch": 1.9404255319148938,
      "grad_norm": 0.16155880689620972,
      "learning_rate": 2.908871303963711e-05,
      "loss": 0.9127392768859863,
      "step": 456
    },
    {
      "epoch": 1.9489361702127659,
      "grad_norm": 0.12455891817808151,
      "learning_rate": 2.9217370284987434e-05,
      "loss": 0.9905153512954712,
      "step": 458
    },
    {
      "epoch": 1.9574468085106385,
      "grad_norm": 0.1255619376897812,
      "learning_rate": 2.9336527609024075e-05,
      "loss": 1.022163987159729,
      "step": 460
    },
    {
      "epoch": 1.9659574468085106,
      "grad_norm": 0.38456299901008606,
      "learning_rate": 2.9446090422445016e-05,
      "loss": 0.8251804709434509,
      "step": 462
    },
    {
      "epoch": 1.974468085106383,
      "grad_norm": 0.12075914442539215,
      "learning_rate": 2.9545971752249376e-05,
      "loss": 1.0136280059814453,
      "step": 464
    },
    {
      "epoch": 1.9829787234042553,
      "grad_norm": 0.20235367119312286,
      "learning_rate": 2.9636092310778195e-05,
      "loss": 0.9567205905914307,
      "step": 466
    },
    {
      "epoch": 1.9914893617021276,
      "grad_norm": 0.542694091796875,
      "learning_rate": 2.9716380558654445e-05,
      "loss": 0.6730895042419434,
      "step": 468
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.11073193699121475,
      "learning_rate": 2.9786772761572338e-05,
      "loss": 0.6498197317123413,
      "step": 470
    },
    {
      "epoch": 2.008510638297872,
      "grad_norm": 0.11644601076841354,
      "learning_rate": 2.9847213040890793e-05,
      "loss": 0.4679168164730072,
      "step": 472
    },
    {
      "epoch": 2.0170212765957447,
      "grad_norm": 0.14129982888698578,
      "learning_rate": 2.989765341799095e-05,
      "loss": 0.7698261737823486,
      "step": 474
    },
    {
      "epoch": 2.025531914893617,
      "grad_norm": 0.13531845808029175,
      "learning_rate": 2.9938053852362484e-05,
      "loss": 0.47750866413116455,
      "step": 476
    },
    {
      "epoch": 2.0340425531914894,
      "grad_norm": 0.37192943692207336,
      "learning_rate": 2.99683822733885e-05,
      "loss": 0.6370930075645447,
      "step": 478
    },
    {
      "epoch": 2.0425531914893615,
      "grad_norm": 0.13162164390087128,
      "learning_rate": 2.9988614605803806e-05,
      "loss": 0.44695600867271423,
      "step": 480
    },
    {
      "epoch": 2.051063829787234,
      "grad_norm": 0.20202530920505524,
      "learning_rate": 2.9998734788806287e-05,
      "loss": 0.4687855839729309,
      "step": 482
    },
    {
      "epoch": 2.0595744680851062,
      "grad_norm": 0.09361886233091354,
      "learning_rate": 2.9998734788806287e-05,
      "loss": 0.28051936626434326,
      "step": 484
    },
    {
      "epoch": 2.068085106382979,
      "grad_norm": 0.1879085749387741,
      "learning_rate": 2.9988614605803806e-05,
      "loss": 0.6579060554504395,
      "step": 486
    },
    {
      "epoch": 2.076595744680851,
      "grad_norm": 0.24100597202777863,
      "learning_rate": 2.99683822733885e-05,
      "loss": 0.5791319608688354,
      "step": 488
    },
    {
      "epoch": 2.0851063829787235,
      "grad_norm": 0.15436111390590668,
      "learning_rate": 2.9938053852362484e-05,
      "loss": 0.5440673828125,
      "step": 490
    },
    {
      "epoch": 2.0936170212765957,
      "grad_norm": 0.1665167361497879,
      "learning_rate": 2.989765341799095e-05,
      "loss": 0.7786204218864441,
      "step": 492
    },
    {
      "epoch": 2.1021276595744682,
      "grad_norm": 0.12963709235191345,
      "learning_rate": 2.9847213040890793e-05,
      "loss": 0.6289861798286438,
      "step": 494
    },
    {
      "epoch": 2.1106382978723404,
      "grad_norm": 0.23455171287059784,
      "learning_rate": 2.9786772761572338e-05,
      "loss": 0.7049335241317749,
      "step": 496
    },
    {
      "epoch": 2.119148936170213,
      "grad_norm": 0.10971709340810776,
      "learning_rate": 2.9716380558654445e-05,
      "loss": 0.9018527269363403,
      "step": 498
    },
    {
      "epoch": 2.127659574468085,
      "grad_norm": 0.3122045695781708,
      "learning_rate": 2.9636092310778195e-05,
      "loss": 0.6065503358840942,
      "step": 500
    },
    {
      "epoch": 2.1361702127659576,
      "grad_norm": 0.15959368646144867,
      "learning_rate": 2.9545971752249383e-05,
      "loss": 0.5845508575439453,
      "step": 502
    },
    {
      "epoch": 2.1446808510638298,
      "grad_norm": 0.12507954239845276,
      "learning_rate": 2.9446090422445016e-05,
      "loss": 0.6952933669090271,
      "step": 504
    },
    {
      "epoch": 2.153191489361702,
      "grad_norm": 0.12080808728933334,
      "learning_rate": 2.933652760902408e-05,
      "loss": 0.7144219279289246,
      "step": 506
    },
    {
      "epoch": 2.1617021276595745,
      "grad_norm": 0.3896176815032959,
      "learning_rate": 2.9217370284987434e-05,
      "loss": 0.5674622654914856,
      "step": 508
    },
    {
      "epoch": 2.1702127659574466,
      "grad_norm": 0.14054062962532043,
      "learning_rate": 2.908871303963712e-05,
      "loss": 0.28315988183021545,
      "step": 510
    },
    {
      "epoch": 2.178723404255319,
      "grad_norm": 0.1403239667415619,
      "learning_rate": 2.8950658003489534e-05,
      "loss": 0.430463582277298,
      "step": 512
    },
    {
      "epoch": 2.1872340425531913,
      "grad_norm": 0.14137110114097595,
      "learning_rate": 2.8803314767202387e-05,
      "loss": 0.4114861488342285,
      "step": 514
    },
    {
      "epoch": 2.195744680851064,
      "grad_norm": 0.1280188113451004,
      "learning_rate": 2.864680029457952e-05,
      "loss": 0.5308094620704651,
      "step": 516
    },
    {
      "epoch": 2.204255319148936,
      "grad_norm": 0.25649434328079224,
      "learning_rate": 2.8481238829722946e-05,
      "loss": 0.6915931701660156,
      "step": 518
    },
    {
      "epoch": 2.2127659574468086,
      "grad_norm": 0.10003609210252762,
      "learning_rate": 2.8306761798405533e-05,
      "loss": 0.4150570333003998,
      "step": 520
    },
    {
      "epoch": 2.2212765957446807,
      "grad_norm": 0.15974430739879608,
      "learning_rate": 2.812350770374273e-05,
      "loss": 0.36256328225135803,
      "step": 522
    },
    {
      "epoch": 2.2297872340425533,
      "grad_norm": 0.38845136761665344,
      "learning_rate": 2.793162201624631e-05,
      "loss": 0.37502366304397583,
      "step": 524
    },
    {
      "epoch": 2.2382978723404254,
      "grad_norm": 0.1723078191280365,
      "learning_rate": 2.7731257058347098e-05,
      "loss": 0.7607256770133972,
      "step": 526
    },
    {
      "epoch": 2.246808510638298,
      "grad_norm": 0.255744993686676,
      "learning_rate": 2.752257188347862e-05,
      "loss": 0.6797502040863037,
      "step": 528
    },
    {
      "epoch": 2.25531914893617,
      "grad_norm": 0.3702438771724701,
      "learning_rate": 2.7305732149817512e-05,
      "loss": 0.7978914380073547,
      "step": 530
    },
    {
      "epoch": 2.2638297872340427,
      "grad_norm": 0.26327913999557495,
      "learning_rate": 2.708090998878098e-05,
      "loss": 0.5080232620239258,
      "step": 532
    },
    {
      "epoch": 2.272340425531915,
      "grad_norm": 0.14510440826416016,
      "learning_rate": 2.68482838683857e-05,
      "loss": 0.2904958724975586,
      "step": 534
    },
    {
      "epoch": 2.2808510638297874,
      "grad_norm": 0.1330595761537552,
      "learning_rate": 2.6608038451576528e-05,
      "loss": 0.5988547205924988,
      "step": 536
    },
    {
      "epoch": 2.2893617021276595,
      "grad_norm": 0.15609091520309448,
      "learning_rate": 2.6360364449637686e-05,
      "loss": 0.8217217922210693,
      "step": 538
    },
    {
      "epoch": 2.297872340425532,
      "grad_norm": 0.27816882729530334,
      "learning_rate": 2.6105458470802563e-05,
      "loss": 0.29454290866851807,
      "step": 540
    },
    {
      "epoch": 2.3063829787234043,
      "grad_norm": 0.16807164251804352,
      "learning_rate": 2.5843522864182397e-05,
      "loss": 0.2661294937133789,
      "step": 542
    },
    {
      "epoch": 2.3148936170212764,
      "grad_norm": 0.2340618073940277,
      "learning_rate": 2.557476555913786e-05,
      "loss": 0.4138932228088379,
      "step": 544
    },
    {
      "epoch": 2.323404255319149,
      "grad_norm": 0.12209529429674149,
      "learning_rate": 2.52993999002208e-05,
      "loss": 0.47122570872306824,
      "step": 546
    },
    {
      "epoch": 2.331914893617021,
      "grad_norm": 0.22724129259586334,
      "learning_rate": 2.501764447781743e-05,
      "loss": 0.5940241813659668,
      "step": 548
    },
    {
      "epoch": 2.3404255319148937,
      "grad_norm": 0.10988523811101913,
      "learning_rate": 2.4729722954627106e-05,
      "loss": 0.443272203207016,
      "step": 550
    },
    {
      "epoch": 2.348936170212766,
      "grad_norm": 0.3907153904438019,
      "learning_rate": 2.4435863888114814e-05,
      "loss": 0.612157940864563,
      "step": 552
    },
    {
      "epoch": 2.3574468085106384,
      "grad_norm": 0.15928088128566742,
      "learning_rate": 2.413630054907798e-05,
      "loss": 0.6884116530418396,
      "step": 554
    },
    {
      "epoch": 2.3659574468085105,
      "grad_norm": 0.12084610015153885,
      "learning_rate": 2.383127073647171e-05,
      "loss": 0.6459460854530334,
      "step": 556
    },
    {
      "epoch": 2.374468085106383,
      "grad_norm": 0.14340610802173615,
      "learning_rate": 2.3521016588639598e-05,
      "loss": 0.5485546588897705,
      "step": 558
    },
    {
      "epoch": 2.382978723404255,
      "grad_norm": 0.1330912858247757,
      "learning_rate": 2.32057843910998e-05,
      "loss": 0.6871886253356934,
      "step": 560
    },
    {
      "epoch": 2.391489361702128,
      "grad_norm": 0.1461075246334076,
      "learning_rate": 2.2885824381039024e-05,
      "loss": 0.5943357944488525,
      "step": 562
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.4632057547569275,
      "learning_rate": 2.2561390548669552e-05,
      "loss": 0.44520270824432373,
      "step": 564
    },
    {
      "epoch": 2.4085106382978725,
      "grad_norm": 0.16760534048080444,
      "learning_rate": 2.2232740435607067e-05,
      "loss": 0.4427604377269745,
      "step": 566
    },
    {
      "epoch": 2.4170212765957446,
      "grad_norm": 0.17341409623622894,
      "learning_rate": 2.1900134930429403e-05,
      "loss": 0.7208576798439026,
      "step": 568
    },
    {
      "epoch": 2.425531914893617,
      "grad_norm": 0.12514737248420715,
      "learning_rate": 2.1563838061578264e-05,
      "loss": 0.6061448454856873,
      "step": 570
    },
    {
      "epoch": 2.4340425531914893,
      "grad_norm": 0.275291383266449,
      "learning_rate": 2.122411678776855e-05,
      "loss": 0.48493388295173645,
      "step": 572
    },
    {
      "epoch": 2.4425531914893615,
      "grad_norm": 0.35213977098464966,
      "learning_rate": 2.0881240786071588e-05,
      "loss": 0.6101267337799072,
      "step": 574
    },
    {
      "epoch": 2.451063829787234,
      "grad_norm": 0.3173239827156067,
      "learning_rate": 2.0535482237840398e-05,
      "loss": 0.6204816102981567,
      "step": 576
    },
    {
      "epoch": 2.4595744680851066,
      "grad_norm": 0.12788061797618866,
      "learning_rate": 2.018711561264715e-05,
      "loss": 0.5858104825019836,
      "step": 578
    },
    {
      "epoch": 2.4680851063829787,
      "grad_norm": 0.21373364329338074,
      "learning_rate": 1.9836417450403978e-05,
      "loss": 0.5342880487442017,
      "step": 580
    },
    {
      "epoch": 2.476595744680851,
      "grad_norm": 0.1782151162624359,
      "learning_rate": 1.948366614184062e-05,
      "loss": 0.6852924823760986,
      "step": 582
    },
    {
      "epoch": 2.4851063829787234,
      "grad_norm": 0.12118978798389435,
      "learning_rate": 1.9129141707512514e-05,
      "loss": 0.5905125737190247,
      "step": 584
    },
    {
      "epoch": 2.4936170212765956,
      "grad_norm": 0.6266365647315979,
      "learning_rate": 1.877312557551536e-05,
      "loss": 0.6729872822761536,
      "step": 586
    },
    {
      "epoch": 2.502127659574468,
      "grad_norm": 0.13392919301986694,
      "learning_rate": 1.841590035808227e-05,
      "loss": 0.7322168350219727,
      "step": 588
    },
    {
      "epoch": 2.5106382978723403,
      "grad_norm": 0.27286240458488464,
      "learning_rate": 1.8057749627240822e-05,
      "loss": 0.6285769939422607,
      "step": 590
    },
    {
      "epoch": 2.519148936170213,
      "grad_norm": 0.10404688119888306,
      "learning_rate": 1.7698957689708436e-05,
      "loss": 0.766042172908783,
      "step": 592
    },
    {
      "epoch": 2.527659574468085,
      "grad_norm": 0.11238724738359451,
      "learning_rate": 1.7339809361204255e-05,
      "loss": 0.43586933612823486,
      "step": 594
    },
    {
      "epoch": 2.5361702127659576,
      "grad_norm": 0.12042172253131866,
      "learning_rate": 1.6980589740357287e-05,
      "loss": 0.6923111081123352,
      "step": 596
    },
    {
      "epoch": 2.5446808510638297,
      "grad_norm": 0.2841528058052063,
      "learning_rate": 1.6621583982389714e-05,
      "loss": 0.47686851024627686,
      "step": 598
    },
    {
      "epoch": 2.5531914893617023,
      "grad_norm": 0.11969015002250671,
      "learning_rate": 1.626307707275533e-05,
      "loss": 0.5013985633850098,
      "step": 600
    },
    {
      "epoch": 2.5617021276595744,
      "grad_norm": 0.09876365959644318,
      "learning_rate": 1.59053536009129e-05,
      "loss": 0.6609531044960022,
      "step": 602
    },
    {
      "epoch": 2.570212765957447,
      "grad_norm": 0.4649478495121002,
      "learning_rate": 1.5548697534413653e-05,
      "loss": 0.6927770972251892,
      "step": 604
    },
    {
      "epoch": 2.578723404255319,
      "grad_norm": 0.1714295893907547,
      "learning_rate": 1.5193391993482573e-05,
      "loss": 0.8361954689025879,
      "step": 606
    },
    {
      "epoch": 2.5872340425531917,
      "grad_norm": 0.15755438804626465,
      "learning_rate": 1.483971902627239e-05,
      "loss": 0.5158759951591492,
      "step": 608
    },
    {
      "epoch": 2.595744680851064,
      "grad_norm": 0.11576741188764572,
      "learning_rate": 1.4487959384968272e-05,
      "loss": 0.6141366362571716,
      "step": 610
    },
    {
      "epoch": 2.604255319148936,
      "grad_norm": 0.2397107034921646,
      "learning_rate": 1.4138392302921813e-05,
      "loss": 0.669286847114563,
      "step": 612
    },
    {
      "epoch": 2.6127659574468085,
      "grad_norm": 0.17677916586399078,
      "learning_rate": 1.3791295272990178e-05,
      "loss": 0.8155961036682129,
      "step": 614
    },
    {
      "epoch": 2.621276595744681,
      "grad_norm": 0.10534735023975372,
      "learning_rate": 1.344694382725719e-05,
      "loss": 0.6486954689025879,
      "step": 616
    },
    {
      "epoch": 2.629787234042553,
      "grad_norm": 0.17969125509262085,
      "learning_rate": 1.3105611318310828e-05,
      "loss": 0.7624613046646118,
      "step": 618
    },
    {
      "epoch": 2.6382978723404253,
      "grad_norm": 0.10477960854768753,
      "learning_rate": 1.2767568702250834e-05,
      "loss": 0.44974949955940247,
      "step": 620
    },
    {
      "epoch": 2.646808510638298,
      "grad_norm": 0.41400146484375,
      "learning_rate": 1.2433084323598806e-05,
      "loss": 0.5666506886482239,
      "step": 622
    },
    {
      "epoch": 2.65531914893617,
      "grad_norm": 0.12478289753198624,
      "learning_rate": 1.2102423702281114e-05,
      "loss": 0.7800987362861633,
      "step": 624
    },
    {
      "epoch": 2.6638297872340426,
      "grad_norm": 0.16413263976573944,
      "learning_rate": 1.1775849322854515e-05,
      "loss": 0.5514877438545227,
      "step": 626
    },
    {
      "epoch": 2.6723404255319148,
      "grad_norm": 0.09944586455821991,
      "learning_rate": 1.1453620426140796e-05,
      "loss": 0.554164469242096,
      "step": 628
    },
    {
      "epoch": 2.6808510638297873,
      "grad_norm": 0.2592844069004059,
      "learning_rate": 1.1135992803436701e-05,
      "loss": 0.5563373565673828,
      "step": 630
    },
    {
      "epoch": 2.6893617021276595,
      "grad_norm": 0.2102140486240387,
      "learning_rate": 1.0823218593462006e-05,
      "loss": 0.5627699494361877,
      "step": 632
    },
    {
      "epoch": 2.697872340425532,
      "grad_norm": 0.1483510136604309,
      "learning_rate": 1.0515546082207089e-05,
      "loss": 0.698486328125,
      "step": 634
    },
    {
      "epoch": 2.706382978723404,
      "grad_norm": 0.11229369789361954,
      "learning_rate": 1.0213219505839e-05,
      "loss": 0.7639659643173218,
      "step": 636
    },
    {
      "epoch": 2.7148936170212767,
      "grad_norm": 0.10941766947507858,
      "learning_rate": 9.916478856822011e-06,
      "loss": 0.6158884763717651,
      "step": 638
    },
    {
      "epoch": 2.723404255319149,
      "grad_norm": 0.1750131994485855,
      "learning_rate": 9.625559693407416e-06,
      "loss": 0.9152920842170715,
      "step": 640
    },
    {
      "epoch": 2.731914893617021,
      "grad_norm": 0.18939398229122162,
      "learning_rate": 9.34069295264279e-06,
      "loss": 0.8749684691429138,
      "step": 642
    },
    {
      "epoch": 2.7404255319148936,
      "grad_norm": 0.0939684808254242,
      "learning_rate": 9.062104767049962e-06,
      "loss": 0.8344287872314453,
      "step": 644
    },
    {
      "epoch": 2.748936170212766,
      "grad_norm": 0.10787810385227203,
      "learning_rate": 8.790016285116773e-06,
      "loss": 0.45629745721817017,
      "step": 646
    },
    {
      "epoch": 2.7574468085106383,
      "grad_norm": 0.10236497223377228,
      "learning_rate": 8.5246434957453e-06,
      "loss": 0.3841492533683777,
      "step": 648
    },
    {
      "epoch": 2.7659574468085104,
      "grad_norm": 0.31643790006637573,
      "learning_rate": 8.266197056795886e-06,
      "loss": 0.4197019338607788,
      "step": 650
    },
    {
      "epoch": 2.774468085106383,
      "grad_norm": 0.15869994461536407,
      "learning_rate": 8.014882127862926e-06,
      "loss": 0.6872882843017578,
      "step": 652
    },
    {
      "epoch": 2.7829787234042556,
      "grad_norm": 0.15020446479320526,
      "learning_rate": 7.77089820741542e-06,
      "loss": 0.6434124112129211,
      "step": 654
    },
    {
      "epoch": 2.7914893617021277,
      "grad_norm": 0.13434921205043793,
      "learning_rate": 7.534438974431356e-06,
      "loss": 0.5578271150588989,
      "step": 656
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.1459910273551941,
      "learning_rate": 7.305692134651748e-06,
      "loss": 0.6800917983055115,
      "step": 658
    },
    {
      "epoch": 2.8085106382978724,
      "grad_norm": 0.13331423699855804,
      "learning_rate": 7.0848392715763e-06,
      "loss": 0.4372290074825287,
      "step": 660
    },
    {
      "epoch": 2.8170212765957445,
      "grad_norm": 0.11432936787605286,
      "learning_rate": 6.872055702319048e-06,
      "loss": 0.3426898121833801,
      "step": 662
    },
    {
      "epoch": 2.825531914893617,
      "grad_norm": 0.47585824131965637,
      "learning_rate": 6.66751033843842e-06,
      "loss": 0.5213098526000977,
      "step": 664
    },
    {
      "epoch": 2.8340425531914892,
      "grad_norm": 0.23138658702373505,
      "learning_rate": 6.471365551852014e-06,
      "loss": 0.40890708565711975,
      "step": 666
    },
    {
      "epoch": 2.842553191489362,
      "grad_norm": 0.1170782670378685,
      "learning_rate": 6.283777045942801e-06,
      "loss": 0.6171141266822815,
      "step": 668
    },
    {
      "epoch": 2.851063829787234,
      "grad_norm": 0.2987634837627411,
      "learning_rate": 6.104893731958872e-06,
      "loss": 0.79531329870224,
      "step": 670
    },
    {
      "epoch": 2.8595744680851065,
      "grad_norm": 0.16172188520431519,
      "learning_rate": 5.934857610804912e-06,
      "loss": 0.6109257936477661,
      "step": 672
    },
    {
      "epoch": 2.8680851063829786,
      "grad_norm": 0.19967441260814667,
      "learning_rate": 5.773803660319231e-06,
      "loss": 0.7840122580528259,
      "step": 674
    },
    {
      "epoch": 2.876595744680851,
      "grad_norm": 0.09976833313703537,
      "learning_rate": 5.6218597281258834e-06,
      "loss": 0.4490019679069519,
      "step": 676
    },
    {
      "epoch": 2.8851063829787233,
      "grad_norm": 0.1243966668844223,
      "learning_rate": 5.479146430146783e-06,
      "loss": 0.5424228310585022,
      "step": 678
    },
    {
      "epoch": 2.8936170212765955,
      "grad_norm": 0.17045721411705017,
      "learning_rate": 5.3457770548545805e-06,
      "loss": 0.5151770114898682,
      "step": 680
    },
    {
      "epoch": 2.902127659574468,
      "grad_norm": 0.1732950657606125,
      "learning_rate": 5.22185747334215e-06,
      "loss": 0.7383885979652405,
      "step": 682
    },
    {
      "epoch": 2.9106382978723406,
      "grad_norm": 0.21424034237861633,
      "learning_rate": 5.107486055280145e-06,
      "loss": 0.4885507822036743,
      "step": 684
    },
    {
      "epoch": 2.9191489361702128,
      "grad_norm": 1.2460668087005615,
      "learning_rate": 5.002753590829352e-06,
      "loss": 0.29081299901008606,
      "step": 686
    },
    {
      "epoch": 2.927659574468085,
      "grad_norm": 0.11700358986854553,
      "learning_rate": 4.9077432185697e-06,
      "loss": 0.436556339263916,
      "step": 688
    },
    {
      "epoch": 2.9361702127659575,
      "grad_norm": 0.2099040299654007,
      "learning_rate": 4.822530359503391e-06,
      "loss": 0.759993314743042,
      "step": 690
    },
    {
      "epoch": 2.94468085106383,
      "grad_norm": 0.12797491252422333,
      "learning_rate": 4.747182657184251e-06,
      "loss": 0.6529436111450195,
      "step": 692
    },
    {
      "epoch": 2.953191489361702,
      "grad_norm": 0.122190460562706,
      "learning_rate": 4.681759924021033e-06,
      "loss": 0.6622499227523804,
      "step": 694
    },
    {
      "epoch": 2.9617021276595743,
      "grad_norm": 0.19574476778507233,
      "learning_rate": 4.626314093797213e-06,
      "loss": 0.5456458926200867,
      "step": 696
    },
    {
      "epoch": 2.970212765957447,
      "grad_norm": 0.15566527843475342,
      "learning_rate": 4.580889180444988e-06,
      "loss": 0.738301157951355,
      "step": 698
    },
    {
      "epoch": 2.978723404255319,
      "grad_norm": 0.13139484822750092,
      "learning_rate": 4.5455212431061985e-06,
      "loss": 0.4914497435092926,
      "step": 700
    },
    {
      "epoch": 2.9872340425531916,
      "grad_norm": 0.24261365830898285,
      "learning_rate": 4.520238357507899e-06,
      "loss": 0.5938633680343628,
      "step": 702
    },
    {
      "epoch": 2.9957446808510637,
      "grad_norm": 0.09112333506345749,
      "learning_rate": 4.505060593675342e-06,
      "loss": 0.4166713058948517,
      "step": 704
    },
    {
      "epoch": 3.0,
      "step": 705,
      "total_flos": 3.0241141054567875e+18,
      "train_loss": 0.8476084547685393,
      "train_runtime": 9225.3576,
      "train_samples_per_second": 4.891,
      "train_steps_per_second": 0.076
    }
  ],
  "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": 3.0241141054567875e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}