{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.999819969034674,
  "eval_steps": 500,
  "global_step": 3471,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "eval_loss": 3.0020999908447266,
      "eval_runtime": 146.9072,
      "eval_samples_per_second": 21.0,
      "eval_steps_per_second": 21.0,
      "step": 0
    },
    {
      "epoch": 0.0002880495445216577,
      "grad_norm": 2.435948610305786,
      "learning_rate": 0.0,
      "loss": 2.9427003860473633,
      "lr/ungrouped": 2.8735632183908047e-07,
      "step": 1
    },
    {
      "epoch": 0.0028804954452165774,
      "grad_norm": 1.9128899574279785,
      "learning_rate": 2.586206896551724e-06,
      "loss": 3.016257815890842,
      "lr/ungrouped": 2.8735632183908046e-06,
      "step": 10
    },
    {
      "epoch": 0.005760990890433155,
      "grad_norm": 1.8352476358413696,
      "learning_rate": 5.45977011494253e-06,
      "loss": 3.0308181762695314,
      "lr/ungrouped": 5.747126436781609e-06,
      "step": 20
    },
    {
      "epoch": 0.008641486335649732,
      "grad_norm": 1.6436556577682495,
      "learning_rate": 8.333333333333334e-06,
      "loss": 3.029275894165039,
      "lr/ungrouped": 8.620689655172414e-06,
      "step": 30
    },
    {
      "epoch": 0.01152198178086631,
      "grad_norm": 1.4728389978408813,
      "learning_rate": 1.1206896551724138e-05,
      "loss": 2.81824951171875,
      "lr/ungrouped": 1.1494252873563218e-05,
      "step": 40
    },
    {
      "epoch": 0.014402477226082887,
      "grad_norm": 1.3213839530944824,
      "learning_rate": 1.4080459770114942e-05,
      "loss": 2.7225183486938476,
      "lr/ungrouped": 1.4367816091954022e-05,
      "step": 50
    },
    {
      "epoch": 0.017282972671299465,
      "grad_norm": 1.4416816234588623,
      "learning_rate": 1.6954022988505746e-05,
      "loss": 2.6364492416381835,
      "lr/ungrouped": 1.7241379310344828e-05,
      "step": 60
    },
    {
      "epoch": 0.02016346811651604,
      "grad_norm": 1.2685229778289795,
      "learning_rate": 1.9827586206896554e-05,
      "loss": 2.52068977355957,
      "lr/ungrouped": 2.0114942528735632e-05,
      "step": 70
    },
    {
      "epoch": 0.02304396356173262,
      "grad_norm": 1.4016832113265991,
      "learning_rate": 2.2701149425287358e-05,
      "loss": 2.3681804656982424,
      "lr/ungrouped": 2.2988505747126437e-05,
      "step": 80
    },
    {
      "epoch": 0.025924459006949195,
      "grad_norm": 1.3321669101715088,
      "learning_rate": 2.5574712643678162e-05,
      "loss": 2.2856060028076173,
      "lr/ungrouped": 2.5862068965517244e-05,
      "step": 90
    },
    {
      "epoch": 0.028804954452165774,
      "grad_norm": 1.393796682357788,
      "learning_rate": 2.844827586206897e-05,
      "loss": 2.054812431335449,
      "lr/ungrouped": 2.8735632183908045e-05,
      "step": 100
    },
    {
      "epoch": 0.031685449897382346,
      "grad_norm": 1.3842580318450928,
      "learning_rate": 3.132183908045977e-05,
      "loss": 1.7950876235961915,
      "lr/ungrouped": 3.160919540229885e-05,
      "step": 110
    },
    {
      "epoch": 0.03456594534259893,
      "grad_norm": 1.629372239112854,
      "learning_rate": 3.419540229885058e-05,
      "loss": 1.5896265029907226,
      "lr/ungrouped": 3.4482758620689657e-05,
      "step": 120
    },
    {
      "epoch": 0.037446440787815505,
      "grad_norm": 1.680420160293579,
      "learning_rate": 3.7068965517241385e-05,
      "loss": 1.40800142288208,
      "lr/ungrouped": 3.735632183908046e-05,
      "step": 130
    },
    {
      "epoch": 0.04032693623303208,
      "grad_norm": 1.6829431056976318,
      "learning_rate": 3.9942528735632186e-05,
      "loss": 1.231640338897705,
      "lr/ungrouped": 4.0229885057471265e-05,
      "step": 140
    },
    {
      "epoch": 0.043207431678248656,
      "grad_norm": 1.4036335945129395,
      "learning_rate": 4.2816091954022994e-05,
      "loss": 0.8847078323364258,
      "lr/ungrouped": 4.3103448275862066e-05,
      "step": 150
    },
    {
      "epoch": 0.04608792712346524,
      "grad_norm": 1.434554100036621,
      "learning_rate": 4.5689655172413794e-05,
      "loss": 0.8526058197021484,
      "lr/ungrouped": 4.597701149425287e-05,
      "step": 160
    },
    {
      "epoch": 0.048968422568681814,
      "grad_norm": 1.7466213703155518,
      "learning_rate": 4.85632183908046e-05,
      "loss": 0.94129638671875,
      "lr/ungrouped": 4.885057471264368e-05,
      "step": 170
    },
    {
      "epoch": 0.05184891801389839,
      "grad_norm": 1.315833330154419,
      "learning_rate": 4.999971626621519e-05,
      "loss": 0.764661169052124,
      "lr/ungrouped": 4.999959142368992e-05,
      "step": 180
    },
    {
      "epoch": 0.054729413459114966,
      "grad_norm": 1.0953325033187866,
      "learning_rate": 4.999744643457886e-05,
      "loss": 0.7581181049346923,
      "lr/ungrouped": 4.999709461682445e-05,
      "step": 190
    },
    {
      "epoch": 0.05760990890433155,
      "grad_norm": 1.5300097465515137,
      "learning_rate": 4.999290697739397e-05,
      "loss": 0.7554854869842529,
      "lr/ungrouped": 4.999232821635345e-05,
      "step": 200
    },
    {
      "epoch": 0.060490404349548124,
      "grad_norm": 1.1001440286636353,
      "learning_rate": 4.998609830681734e-05,
      "loss": 0.7553323745727539,
      "lr/ungrouped": 4.9985292655038905e-05,
      "step": 210
    },
    {
      "epoch": 0.06337089979476469,
      "grad_norm": 1.6377240419387817,
      "learning_rate": 4.997702104103742e-05,
      "loss": 0.6583240985870361,
      "lr/ungrouped": 4.997598857166966e-05,
      "step": 220
    },
    {
      "epoch": 0.06625139523998128,
      "grad_norm": 1.3325226306915283,
      "learning_rate": 4.996567600421816e-05,
      "loss": 0.6379648685455322,
      "lr/ungrouped": 4.9964416811003414e-05,
      "step": 230
    },
    {
      "epoch": 0.06913189068519786,
      "grad_norm": 0.752729594707489,
      "learning_rate": 4.995206422642419e-05,
      "loss": 0.7104136943817139,
      "lr/ungrouped": 4.995057842369002e-05,
      "step": 240
    },
    {
      "epoch": 0.07201238613041443,
      "grad_norm": 0.933250367641449,
      "learning_rate": 4.99361869435273e-05,
      "loss": 0.6339001178741455,
      "lr/ungrouped": 4.993447466617611e-05,
      "step": 250
    },
    {
      "epoch": 0.07489288157563101,
      "grad_norm": 0.9432796239852905,
      "learning_rate": 4.991804559709421e-05,
      "loss": 0.6346065044403076,
      "lr/ungrouped": 4.991610700059102e-05,
      "step": 260
    },
    {
      "epoch": 0.07777337702084759,
      "grad_norm": 0.9591599106788635,
      "learning_rate": 4.989764183425568e-05,
      "loss": 0.527804183959961,
      "lr/ungrouped": 4.9895477094613994e-05,
      "step": 270
    },
    {
      "epoch": 0.08065387246606416,
      "grad_norm": 1.5225812196731567,
      "learning_rate": 4.987497750755702e-05,
      "loss": 0.7571865081787109,
      "lr/ungrouped": 4.9872586821322844e-05,
      "step": 280
    },
    {
      "epoch": 0.08353436791128074,
      "grad_norm": 1.242102026939392,
      "learning_rate": 4.985005467478978e-05,
      "loss": 0.620611572265625,
      "lr/ungrouped": 4.98474382590238e-05,
      "step": 290
    },
    {
      "epoch": 0.08641486335649731,
      "grad_norm": 0.9705703854560852,
      "learning_rate": 4.982287559880504e-05,
      "loss": 0.6448197841644288,
      "lr/ungrouped": 4.982003369106287e-05,
      "step": 300
    },
    {
      "epoch": 0.0892953588017139,
      "grad_norm": 0.9389514327049255,
      "learning_rate": 4.979344274730786e-05,
      "loss": 0.6706052303314209,
      "lr/ungrouped": 4.9790375605618523e-05,
      "step": 310
    },
    {
      "epoch": 0.09217585424693048,
      "grad_norm": 0.9404653906822205,
      "learning_rate": 4.9761758792633266e-05,
      "loss": 0.5455225944519043,
      "lr/ungrouped": 4.975846669547572e-05,
      "step": 320
    },
    {
      "epoch": 0.09505634969214705,
      "grad_norm": 0.8755877017974854,
      "learning_rate": 4.972782661150361e-05,
      "loss": 0.6003330230712891,
      "lr/ungrouped": 4.972430985778152e-05,
      "step": 330
    },
    {
      "epoch": 0.09793684513736363,
      "grad_norm": 0.8913218379020691,
      "learning_rate": 4.9691649284767406e-05,
      "loss": 0.6236235618591308,
      "lr/ungrouped": 4.968790819378193e-05,
      "step": 340
    },
    {
      "epoch": 0.1008173405825802,
      "grad_norm": 0.9349474310874939,
      "learning_rate": 4.965323009711954e-05,
      "loss": 0.5250090599060059,
      "lr/ungrouped": 4.96492650085404e-05,
      "step": 350
    },
    {
      "epoch": 0.10369783602779678,
      "grad_norm": 0.8367001414299011,
      "learning_rate": 4.961257253680311e-05,
      "loss": 0.6016794681549072,
      "lr/ungrouped": 4.960838381063774e-05,
      "step": 360
    },
    {
      "epoch": 0.10657833147301336,
      "grad_norm": 0.9123567938804626,
      "learning_rate": 4.956968029529268e-05,
      "loss": 0.6599809169769287,
      "lr/ungrouped": 4.956526831185353e-05,
      "step": 370
    },
    {
      "epoch": 0.10945882691822993,
      "grad_norm": 1.1716485023498535,
      "learning_rate": 4.952455726695913e-05,
      "loss": 0.6317523002624512,
      "lr/ungrouped": 4.9519922426829104e-05,
      "step": 380
    },
    {
      "epoch": 0.11233932236344651,
      "grad_norm": 1.0180355310440063,
      "learning_rate": 4.947720754871602e-05,
      "loss": 0.6037989139556885,
      "lr/ungrouped": 4.947235027271219e-05,
      "step": 390
    },
    {
      "epoch": 0.1152198178086631,
      "grad_norm": 1.019775629043579,
      "learning_rate": 4.9427635439647704e-05,
      "loss": 0.5504507064819336,
      "lr/ungrouped": 4.942255616878302e-05,
      "step": 400
    },
    {
      "epoch": 0.11810031325387967,
      "grad_norm": 1.1898387670516968,
      "learning_rate": 4.937584544061891e-05,
      "loss": 0.5491292953491211,
      "lr/ungrouped": 4.9370544636062185e-05,
      "step": 410
    },
    {
      "epoch": 0.12098080869909625,
      "grad_norm": 0.8084814548492432,
      "learning_rate": 4.9321842253866136e-05,
      "loss": 0.6774951457977295,
      "lr/ungrouped": 4.9316320396900193e-05,
      "step": 420
    },
    {
      "epoch": 0.12386130414431282,
      "grad_norm": 0.8823770880699158,
      "learning_rate": 4.9265630782570704e-05,
      "loss": 0.5574594020843506,
      "lr/ungrouped": 4.925988837454866e-05,
      "step": 430
    },
    {
      "epoch": 0.12674179958952939,
      "grad_norm": 1.1832399368286133,
      "learning_rate": 4.920721613041357e-05,
      "loss": 0.5654309749603271,
      "lr/ungrouped": 4.920125369271332e-05,
      "step": 440
    },
    {
      "epoch": 0.12962229503474598,
      "grad_norm": 1.1205772161483765,
      "learning_rate": 4.9146603601111943e-05,
      "loss": 0.6521910667419434,
      "lr/ungrouped": 4.9140421675088815e-05,
      "step": 450
    },
    {
      "epoch": 0.13250279047996255,
      "grad_norm": 1.040226697921753,
      "learning_rate": 4.9083798697937754e-05,
      "loss": 0.6186479091644287,
      "lr/ungrouped": 4.9077397844875375e-05,
      "step": 460
    },
    {
      "epoch": 0.13538328592517912,
      "grad_norm": 1.0655864477157593,
      "learning_rate": 4.9018807123217955e-05,
      "loss": 0.7008211612701416,
      "lr/ungrouped": 4.9012187924277266e-05,
      "step": 470
    },
    {
      "epoch": 0.13826378137039572,
      "grad_norm": 0.8568135499954224,
      "learning_rate": 4.895163477781683e-05,
      "loss": 0.5307230949401855,
      "lr/ungrouped": 4.894479783398334e-05,
      "step": 480
    },
    {
      "epoch": 0.14114427681561229,
      "grad_norm": 0.98490309715271,
      "learning_rate": 4.888228776060016e-05,
      "loss": 0.5620987415313721,
      "lr/ungrouped": 4.887523369262937e-05,
      "step": 490
    },
    {
      "epoch": 0.14402477226082885,
      "grad_norm": 0.9300552010536194,
      "learning_rate": 4.8810772367881575e-05,
      "loss": 0.5468950748443604,
      "lr/ungrouped": 4.8803501816242606e-05,
      "step": 500
    },
    {
      "epoch": 0.14690526770604545,
      "grad_norm": 1.1576457023620605,
      "learning_rate": 4.8737095092850773e-05,
      "loss": 0.6259415149688721,
      "lr/ungrouped": 4.8729608717668265e-05,
      "step": 510
    },
    {
      "epoch": 0.14978576315126202,
      "grad_norm": 1.122401475906372,
      "learning_rate": 4.866126262498409e-05,
      "loss": 0.5051443099975585,
      "lr/ungrouped": 4.865356110597819e-05,
      "step": 520
    },
    {
      "epoch": 0.1526662585964786,
      "grad_norm": 1.1125292778015137,
      "learning_rate": 4.858328184943704e-05,
      "loss": 0.5950119972229004,
      "lr/ungrouped": 4.8575365885861744e-05,
      "step": 530
    },
    {
      "epoch": 0.15554675404169518,
      "grad_norm": 1.1249250173568726,
      "learning_rate": 4.8503159846419236e-05,
      "loss": 0.5435703754425049,
      "lr/ungrouped": 4.8495030156998896e-05,
      "step": 540
    },
    {
      "epoch": 0.15842724948691175,
      "grad_norm": 0.9522327780723572,
      "learning_rate": 4.8420903890551526e-05,
      "loss": 0.5966388702392578,
      "lr/ungrouped": 4.8412561213415566e-05,
      "step": 550
    },
    {
      "epoch": 0.16130774493212832,
      "grad_norm": 1.2842515707015991,
      "learning_rate": 4.833652145020551e-05,
      "loss": 0.5642343521118164,
      "lr/ungrouped": 4.8327966542821426e-05,
      "step": 560
    },
    {
      "epoch": 0.1641882403773449,
      "grad_norm": 0.9321410655975342,
      "learning_rate": 4.825002018682545e-05,
      "loss": 0.5413059234619141,
      "lr/ungrouped": 4.824125382593003e-05,
      "step": 570
    },
    {
      "epoch": 0.1670687358225615,
      "grad_norm": 1.3501596450805664,
      "learning_rate": 4.816140795423265e-05,
      "loss": 0.5266181945800781,
      "lr/ungrouped": 4.8152430935761456e-05,
      "step": 580
    },
    {
      "epoch": 0.16994923126777806,
      "grad_norm": 0.9427384734153748,
      "learning_rate": 4.807069279791237e-05,
      "loss": 0.5448482990264892,
      "lr/ungrouped": 4.806150593692748e-05,
      "step": 590
    },
    {
      "epoch": 0.17282972671299462,
      "grad_norm": 1.1746304035186768,
      "learning_rate": 4.797788295428336e-05,
      "loss": 0.5196105003356933,
      "lr/ungrouped": 4.796848708489935e-05,
      "step": 600
    },
    {
      "epoch": 0.17571022215821122,
      "grad_norm": 0.15966133773326874,
      "learning_rate": 4.788298684995003e-05,
      "loss": 0.4818408489227295,
      "lr/ungrouped": 4.7873382825258236e-05,
      "step": 610
    },
    {
      "epoch": 0.1785907176034278,
      "grad_norm": 0.8140650987625122,
      "learning_rate": 4.778601310093735e-05,
      "loss": 0.6027480125427246,
      "lr/ungrouped": 4.777620179292844e-05,
      "step": 620
    },
    {
      "epoch": 0.18147121304864436,
      "grad_norm": 0.9272246956825256,
      "learning_rate": 4.7686970511908594e-05,
      "loss": 0.42555975914001465,
      "lr/ungrouped": 4.767695281139336e-05,
      "step": 630
    },
    {
      "epoch": 0.18435170849386096,
      "grad_norm": 0.8695138096809387,
      "learning_rate": 4.758586807536588e-05,
      "loss": 0.5902394771575927,
      "lr/ungrouped": 4.75756448918944e-05,
      "step": 640
    },
    {
      "epoch": 0.18723220393907752,
      "grad_norm": 0.8543628454208374,
      "learning_rate": 4.748271497083376e-05,
      "loss": 0.5532349109649658,
      "lr/ungrouped": 4.747228723261278e-05,
      "step": 650
    },
    {
      "epoch": 0.1901126993842941,
      "grad_norm": 0.8089218139648438,
      "learning_rate": 4.7377520564025734e-05,
      "loss": 0.5382043361663819,
      "lr/ungrouped": 4.7366889217834396e-05,
      "step": 660
    },
    {
      "epoch": 0.1929931948295107,
      "grad_norm": 1.8423728942871094,
      "learning_rate": 4.72702944059939e-05,
      "loss": 0.6316296100616455,
      "lr/ungrouped": 4.72594604170978e-05,
      "step": 670
    },
    {
      "epoch": 0.19587369027472726,
      "grad_norm": 0.8306758999824524,
      "learning_rate": 4.716104623226178e-05,
      "loss": 0.5056098461151123,
      "lr/ungrouped": 4.715001058432531e-05,
      "step": 680
    },
    {
      "epoch": 0.19875418571994383,
      "grad_norm": 1.0015183687210083,
      "learning_rate": 4.7049785961940384e-05,
      "loss": 0.600749921798706,
      "lr/ungrouped": 4.703854965693742e-05,
      "step": 690
    },
    {
      "epoch": 0.2016346811651604,
      "grad_norm": 0.9881550669670105,
      "learning_rate": 4.6936523696827615e-05,
      "loss": 0.622192096710205,
      "lr/ungrouped": 4.692508775495057e-05,
      "step": 700
    },
    {
      "epoch": 0.204515176610377,
      "grad_norm": 0.9981948137283325,
      "learning_rate": 4.68212697204911e-05,
      "loss": 0.520939826965332,
      "lr/ungrouped": 4.680963518005826e-05,
      "step": 710
    },
    {
      "epoch": 0.20739567205559356,
      "grad_norm": 0.9479494690895081,
      "learning_rate": 4.6704034497334445e-05,
      "loss": 0.5735066890716553,
      "lr/ungrouped": 4.669220241469573e-05,
      "step": 720
    },
    {
      "epoch": 0.21027616750081013,
      "grad_norm": 0.9721952676773071,
      "learning_rate": 4.658482867164723e-05,
      "loss": 0.6063258171081543,
      "lr/ungrouped": 4.6572800121088234e-05,
      "step": 730
    },
    {
      "epoch": 0.21315666294602673,
      "grad_norm": 0.9246804118156433,
      "learning_rate": 4.646366306663844e-05,
      "loss": 0.53493332862854,
      "lr/ungrouped": 4.645143914028295e-05,
      "step": 740
    },
    {
      "epoch": 0.2160371583912433,
      "grad_norm": 0.8607037663459778,
      "learning_rate": 4.6340548683453876e-05,
      "loss": 0.5157440185546875,
      "lr/ungrouped": 4.6328130491164675e-05,
      "step": 750
    },
    {
      "epoch": 0.21891765383645986,
      "grad_norm": 0.9400234222412109,
      "learning_rate": 4.6215496700177276e-05,
      "loss": 0.5758402824401856,
      "lr/ungrouped": 4.620288536945539e-05,
      "step": 760
    },
    {
      "epoch": 0.22179814928167646,
      "grad_norm": 1.1869113445281982,
      "learning_rate": 4.608851847081542e-05,
      "loss": 0.5269617080688477,
      "lr/ungrouped": 4.607571514669774e-05,
      "step": 770
    },
    {
      "epoch": 0.22467864472689303,
      "grad_norm": 1.053968906402588,
      "learning_rate": 4.595962552426722e-05,
      "loss": 0.5329038619995117,
      "lr/ungrouped": 4.594663136922256e-05,
      "step": 780
    },
    {
      "epoch": 0.2275591401721096,
      "grad_norm": 1.169480323791504,
      "learning_rate": 4.582882956327703e-05,
      "loss": 0.5033031463623047,
      "lr/ungrouped": 4.581564575710053e-05,
      "step": 790
    },
    {
      "epoch": 0.2304396356173262,
      "grad_norm": 1.0367141962051392,
      "learning_rate": 4.569614246337202e-05,
      "loss": 0.5166784286499023,
      "lr/ungrouped": 4.568277020307809e-05,
      "step": 800
    },
    {
      "epoch": 0.23332013106254276,
      "grad_norm": 1.02375328540802,
      "learning_rate": 4.5561576271784015e-05,
      "loss": 0.564197301864624,
      "lr/ungrouped": 4.55480167714976e-05,
      "step": 810
    },
    {
      "epoch": 0.23620062650775933,
      "grad_norm": 1.7949998378753662,
      "learning_rate": 4.5425143206355614e-05,
      "loss": 0.5752384185791015,
      "lr/ungrouped": 4.541139769720201e-05,
      "step": 820
    },
    {
      "epoch": 0.2390811219529759,
      "grad_norm": 1.053128957748413,
      "learning_rate": 4.528685565443094e-05,
      "loss": 0.5962469100952148,
      "lr/ungrouped": 4.527292538442397e-05,
      "step": 830
    },
    {
      "epoch": 0.2419616173981925,
      "grad_norm": 0.9810224175453186,
      "learning_rate": 4.514672617173091e-05,
      "loss": 0.5856562614440918,
      "lr/ungrouped": 4.5132612405659625e-05,
      "step": 840
    },
    {
      "epoch": 0.24484211284340907,
      "grad_norm": 0.8318721652030945,
      "learning_rate": 4.500476748121324e-05,
      "loss": 0.47009835243225095,
      "lr/ungrouped": 4.49904715005271e-05,
      "step": 850
    },
    {
      "epoch": 0.24772260828862563,
      "grad_norm": 1.1943013668060303,
      "learning_rate": 4.4860992471917286e-05,
      "loss": 0.583559513092041,
      "lr/ungrouped": 4.484651557460978e-05,
      "step": 860
    },
    {
      "epoch": 0.2506031037338422,
      "grad_norm": 1.0103408098220825,
      "learning_rate": 4.471541419779377e-05,
      "loss": 0.565358018875122,
      "lr/ungrouped": 4.470075769828461e-05,
      "step": 870
    },
    {
      "epoch": 0.25348359917905877,
      "grad_norm": 1.0658060312271118,
      "learning_rate": 4.456804587651961e-05,
      "loss": 0.4926689624786377,
      "lr/ungrouped": 4.455321110553534e-05,
      "step": 880
    },
    {
      "epoch": 0.2563640946242754,
      "grad_norm": 1.0738447904586792,
      "learning_rate": 4.4418900888297766e-05,
      "loss": 0.5356546401977539,
      "lr/ungrouped": 4.4403889192750973e-05,
      "step": 890
    },
    {
      "epoch": 0.25924459006949196,
      "grad_norm": 1.0322351455688477,
      "learning_rate": 4.426799277464243e-05,
      "loss": 0.46140599250793457,
      "lr/ungrouped": 4.425280551750945e-05,
      "step": 900
    },
    {
      "epoch": 0.26212508551470853,
      "grad_norm": 1.064253807067871,
      "learning_rate": 4.411533523714954e-05,
      "loss": 0.5007781505584716,
      "lr/ungrouped": 4.409997379734668e-05,
      "step": 910
    },
    {
      "epoch": 0.2650055809599251,
      "grad_norm": 0.9836868047714233,
      "learning_rate": 4.3960942136252706e-05,
      "loss": 0.5272809028625488,
      "lr/ungrouped": 4.3945407908511093e-05,
      "step": 920
    },
    {
      "epoch": 0.26788607640514167,
      "grad_norm": 0.9986889958381653,
      "learning_rate": 4.3804827489964845e-05,
      "loss": 0.530044412612915,
      "lr/ungrouped": 4.378912188470373e-05,
      "step": 930
    },
    {
      "epoch": 0.27076657185035824,
      "grad_norm": 1.0644898414611816,
      "learning_rate": 4.364700547260533e-05,
      "loss": 0.5202998161315918,
      "lr/ungrouped": 4.363112991580409e-05,
      "step": 940
    },
    {
      "epoch": 0.27364706729557486,
      "grad_norm": 1.0076950788497925,
      "learning_rate": 4.348749041351311e-05,
      "loss": 0.559341287612915,
      "lr/ungrouped": 4.3471446346581756e-05,
      "step": 950
    },
    {
      "epoch": 0.27652756274079143,
      "grad_norm": 0.8769247531890869,
      "learning_rate": 4.332629679574566e-05,
      "loss": 0.5032573223114014,
      "lr/ungrouped": 4.331008567539395e-05,
      "step": 960
    },
    {
      "epoch": 0.279408058186008,
      "grad_norm": 1.532835841178894,
      "learning_rate": 4.316343925476402e-05,
      "loss": 0.5314249038696289,
      "lr/ungrouped": 4.31470625528692e-05,
      "step": 970
    },
    {
      "epoch": 0.28228855363122457,
      "grad_norm": 0.8993820548057556,
      "learning_rate": 4.299893257710394e-05,
      "loss": 0.5449575424194336,
      "lr/ungrouped": 4.298239178057713e-05,
      "step": 980
    },
    {
      "epoch": 0.28516904907644114,
      "grad_norm": 0.8594847321510315,
      "learning_rate": 4.2832791699033404e-05,
      "loss": 0.40355558395385743,
      "lr/ungrouped": 4.2816088309684595e-05,
      "step": 990
    },
    {
      "epoch": 0.2880495445216577,
      "grad_norm": 1.1475166082382202,
      "learning_rate": 4.266503170519646e-05,
      "loss": 0.49099025726318357,
      "lr/ungrouped": 4.2648167239598115e-05,
      "step": 1000
    },
    {
      "epoch": 0.2909300399668743,
      "grad_norm": 1.0785480737686157,
      "learning_rate": 4.249566782724361e-05,
      "loss": 0.5238031387329102,
      "lr/ungrouped": 4.2478643816593024e-05,
      "step": 1010
    },
    {
      "epoch": 0.2938105354120909,
      "grad_norm": 0.9834842681884766,
      "learning_rate": 4.232471544244894e-05,
      "loss": 0.4776494026184082,
      "lr/ungrouped": 4.230753343242915e-05,
      "step": 1020
    },
    {
      "epoch": 0.29669103085730747,
      "grad_norm": 0.6954681277275085,
      "learning_rate": 4.2152190072313824e-05,
      "loss": 0.5106285572052002,
      "lr/ungrouped": 4.213485162295335e-05,
      "step": 1030
    },
    {
      "epoch": 0.29957152630252404,
      "grad_norm": 1.213712453842163,
      "learning_rate": 4.1978107381157796e-05,
      "loss": 0.5357083320617676,
      "lr/ungrouped": 4.196061406668893e-05,
      "step": 1040
    },
    {
      "epoch": 0.3024520217477406,
      "grad_norm": 1.0161744356155396,
      "learning_rate": 4.1802483174696214e-05,
      "loss": 0.5518830776214599,
      "lr/ungrouped": 4.1784836583412135e-05,
      "step": 1050
    },
    {
      "epoch": 0.3053325171929572,
      "grad_norm": 0.9562483429908752,
      "learning_rate": 4.162533339860527e-05,
      "loss": 0.5182131767272949,
      "lr/ungrouped": 4.160753513271579e-05,
      "step": 1060
    },
    {
      "epoch": 0.30821301263817374,
      "grad_norm": 1.014853596687317,
      "learning_rate": 4.144667413707413e-05,
      "loss": 0.4750497817993164,
      "lr/ungrouped": 4.142872581256028e-05,
      "step": 1070
    },
    {
      "epoch": 0.31109350808339037,
      "grad_norm": 1.1462771892547607,
      "learning_rate": 4.1266521611344654e-05,
      "loss": 0.5645370483398438,
      "lr/ungrouped": 4.124842485781193e-05,
      "step": 1080
    },
    {
      "epoch": 0.31397400352860694,
      "grad_norm": 1.0921515226364136,
      "learning_rate": 4.108489217823859e-05,
      "loss": 0.44844784736633303,
      "lr/ungrouped": 4.106664863876898e-05,
      "step": 1090
    },
    {
      "epoch": 0.3168544989738235,
      "grad_norm": 0.9067047834396362,
      "learning_rate": 4.090180232867242e-05,
      "loss": 0.510321855545044,
      "lr/ungrouped": 4.0883413659675254e-05,
      "step": 1100
    },
    {
      "epoch": 0.3197349944190401,
      "grad_norm": 1.0348539352416992,
      "learning_rate": 4.0717268686160137e-05,
      "loss": 0.5465463161468506,
      "lr/ungrouped": 4.069873655722165e-05,
      "step": 1110
    },
    {
      "epoch": 0.32261548986425664,
      "grad_norm": 1.16339111328125,
      "learning_rate": 4.053130800530386e-05,
      "loss": 0.49945316314697263,
      "lr/ungrouped": 4.051263409903567e-05,
      "step": 1120
    },
    {
      "epoch": 0.3254959853094732,
      "grad_norm": 0.9570099115371704,
      "learning_rate": 4.03439371702727e-05,
      "loss": 0.5009670257568359,
      "lr/ungrouped": 4.032512318215896e-05,
      "step": 1130
    },
    {
      "epoch": 0.3283764807546898,
      "grad_norm": 0.9747746586799622,
      "learning_rate": 4.015517319326968e-05,
      "loss": 0.4737854480743408,
      "lr/ungrouped": 4.013622083151321e-05,
      "step": 1140
    },
    {
      "epoch": 0.3312569761999064,
      "grad_norm": 1.0171892642974854,
      "learning_rate": 3.996503321298719e-05,
      "loss": 0.4660749912261963,
      "lr/ungrouped": 3.994594419835433e-05,
      "step": 1150
    },
    {
      "epoch": 0.334137471645123,
      "grad_norm": 1.356640338897705,
      "learning_rate": 3.9773534493050866e-05,
      "loss": 0.5911624908447266,
      "lr/ungrouped": 3.975431055871525e-05,
      "step": 1160
    },
    {
      "epoch": 0.33701796709033954,
      "grad_norm": 0.9333807826042175,
      "learning_rate": 3.9580694420452155e-05,
      "loss": 0.5451563835144043,
      "lr/ungrouped": 3.9561337311837366e-05,
      "step": 1170
    },
    {
      "epoch": 0.3398984625355561,
      "grad_norm": 0.945462703704834,
      "learning_rate": 3.938653050396967e-05,
      "loss": 0.5237153530120849,
      "lr/ungrouped": 3.936704197859076e-05,
      "step": 1180
    },
    {
      "epoch": 0.3427789579807727,
      "grad_norm": 0.9627684950828552,
      "learning_rate": 3.91910603725795e-05,
      "loss": 0.5417716026306152,
      "lr/ungrouped": 3.91714421998834e-05,
      "step": 1190
    },
    {
      "epoch": 0.34565945342598925,
      "grad_norm": 0.9419501423835754,
      "learning_rate": 3.899430177385458e-05,
      "loss": 0.4466389179229736,
      "lr/ungrouped": 3.897455573505949e-05,
      "step": 1200
    },
    {
      "epoch": 0.3485399488712059,
      "grad_norm": 0.9696369171142578,
      "learning_rate": 3.879627257235337e-05,
      "loss": 0.460161828994751,
      "lr/ungrouped": 3.877640046028696e-05,
      "step": 1210
    },
    {
      "epoch": 0.35142044431642244,
      "grad_norm": 0.8624915480613708,
      "learning_rate": 3.8596990747997776e-05,
      "loss": 0.5080221176147461,
      "lr/ungrouped": 3.857699436693447e-05,
      "step": 1220
    },
    {
      "epoch": 0.354300939761639,
      "grad_norm": 0.8985052108764648,
      "learning_rate": 3.839647439444073e-05,
      "loss": 0.46982293128967284,
      "lr/ungrouped": 3.837635555993787e-05,
      "step": 1230
    },
    {
      "epoch": 0.3571814352068556,
      "grad_norm": 0.7898502945899963,
      "learning_rate": 3.819474171742337e-05,
      "loss": 0.547551965713501,
      "lr/ungrouped": 3.8174502256156374e-05,
      "step": 1240
    },
    {
      "epoch": 0.36006193065207215,
      "grad_norm": 0.8851045370101929,
      "learning_rate": 3.799181103312207e-05,
      "loss": 0.5325418949127197,
      "lr/ungrouped": 3.797145278271858e-05,
      "step": 1250
    },
    {
      "epoch": 0.3629424260972887,
      "grad_norm": 0.9912987351417542,
      "learning_rate": 3.778770076648543e-05,
      "loss": 0.5237712860107422,
      "lr/ungrouped": 3.7767225575358434e-05,
      "step": 1260
    },
    {
      "epoch": 0.3658229215425053,
      "grad_norm": 1.044459581375122,
      "learning_rate": 3.7582429449561396e-05,
      "loss": 0.4483062744140625,
      "lr/ungrouped": 3.756183917674148e-05,
      "step": 1270
    },
    {
      "epoch": 0.3687034169877219,
      "grad_norm": 0.9795230031013489,
      "learning_rate": 3.737601571981466e-05,
      "loss": 0.5435775756835938,
      "lr/ungrouped": 3.735531223478113e-05,
      "step": 1280
    },
    {
      "epoch": 0.3715839124329385,
      "grad_norm": 0.886932909488678,
      "learning_rate": 3.7168478318434476e-05,
      "loss": 0.5326513290405274,
      "lr/ungrouped": 3.714766350094569e-05,
      "step": 1290
    },
    {
      "epoch": 0.37446440787815505,
      "grad_norm": 1.123531460762024,
      "learning_rate": 3.695983608863308e-05,
      "loss": 0.5240880966186523,
      "lr/ungrouped": 3.693891182855573e-05,
      "step": 1300
    },
    {
      "epoch": 0.3773449033233716,
      "grad_norm": 1.0670071840286255,
      "learning_rate": 3.6750107973934825e-05,
      "loss": 0.48720593452453614,
      "lr/ungrouped": 3.672907617107237e-05,
      "step": 1310
    },
    {
      "epoch": 0.3802253987685882,
      "grad_norm": 0.9110984206199646,
      "learning_rate": 3.653931301645622e-05,
      "loss": 0.4474952220916748,
      "lr/ungrouped": 3.6518175580376395e-05,
      "step": 1320
    },
    {
      "epoch": 0.38310589421380475,
      "grad_norm": 1.0211994647979736,
      "learning_rate": 3.632747035517701e-05,
      "loss": 0.445147180557251,
      "lr/ungrouped": 3.630622920503844e-05,
      "step": 1330
    },
    {
      "epoch": 0.3859863896590214,
      "grad_norm": 1.063029170036316,
      "learning_rate": 3.6114599224202464e-05,
      "loss": 0.5867525100708008,
      "lr/ungrouped": 3.609325628858042e-05,
      "step": 1340
    },
    {
      "epoch": 0.38886688510423795,
      "grad_norm": 1.0554449558258057,
      "learning_rate": 3.590071895101703e-05,
      "loss": 0.4353907585144043,
      "lr/ungrouped": 3.587927616772834e-05,
      "step": 1350
    },
    {
      "epoch": 0.3917473805494545,
      "grad_norm": 1.1792598962783813,
      "learning_rate": 3.568584895472954e-05,
      "loss": 0.5592319011688233,
      "lr/ungrouped": 3.5664308270656604e-05,
      "step": 1360
    },
    {
      "epoch": 0.3946278759946711,
      "grad_norm": 1.0061612129211426,
      "learning_rate": 3.5470008744310015e-05,
      "loss": 0.4510073661804199,
      "lr/ungrouped": 3.544837211522407e-05,
      "step": 1370
    },
    {
      "epoch": 0.39750837143988765,
      "grad_norm": 0.8598170280456543,
      "learning_rate": 3.5253217916818375e-05,
      "loss": 0.5544373512268066,
      "lr/ungrouped": 3.523148730720193e-05,
      "step": 1380
    },
    {
      "epoch": 0.4003888668851042,
      "grad_norm": 1.0894895792007446,
      "learning_rate": 3.503549615562518e-05,
      "loss": 0.5787700176239013,
      "lr/ungrouped": 3.501367353849363e-05,
      "step": 1390
    },
    {
      "epoch": 0.4032693623303208,
      "grad_norm": 0.9866119027137756,
      "learning_rate": 3.481686322862443e-05,
      "loss": 0.5516094207763672,
      "lr/ungrouped": 3.479495058534694e-05,
      "step": 1400
    },
    {
      "epoch": 0.4061498577755374,
      "grad_norm": 1.0283448696136475,
      "learning_rate": 3.45973389864388e-05,
      "loss": 0.49516887664794923,
      "lr/ungrouped": 3.457533830655838e-05,
      "step": 1410
    },
    {
      "epoch": 0.409030353220754,
      "grad_norm": 0.8972716331481934,
      "learning_rate": 3.4376943360617295e-05,
      "loss": 0.4186561584472656,
      "lr/ungrouped": 3.435485664167019e-05,
      "step": 1420
    },
    {
      "epoch": 0.41191084866597055,
      "grad_norm": 0.9961366653442383,
      "learning_rate": 3.4155696361825565e-05,
      "loss": 0.4232534408569336,
      "lr/ungrouped": 3.413352560915988e-05,
      "step": 1430
    },
    {
      "epoch": 0.4147913441111871,
      "grad_norm": 1.2983286380767822,
      "learning_rate": 3.3933618078029115e-05,
      "loss": 0.4506828308105469,
      "lr/ungrouped": 3.3911365304622714e-05,
      "step": 1440
    },
    {
      "epoch": 0.4176718395564037,
      "grad_norm": 1.0538884401321411,
      "learning_rate": 3.371072867266934e-05,
      "loss": 0.49685029983520507,
      "lr/ungrouped": 3.368839589894712e-05,
      "step": 1450
    },
    {
      "epoch": 0.42055233500162026,
      "grad_norm": 1.069190502166748,
      "learning_rate": 3.3487048382832876e-05,
      "loss": 0.5658205032348633,
      "lr/ungrouped": 3.3464637636483304e-05,
      "step": 1460
    },
    {
      "epoch": 0.4234328304468369,
      "grad_norm": 0.9561610817909241,
      "learning_rate": 3.326259751741414e-05,
      "loss": 0.5299627304077148,
      "lr/ungrouped": 3.324011083320515e-05,
      "step": 1470
    },
    {
      "epoch": 0.42631332589205345,
      "grad_norm": 1.0791432857513428,
      "learning_rate": 3.303739645527144e-05,
      "loss": 0.43665251731872556,
      "lr/ungrouped": 3.301483587486569e-05,
      "step": 1480
    },
    {
      "epoch": 0.42919382133727,
      "grad_norm": 1.0314401388168335,
      "learning_rate": 3.281146564337662e-05,
      "loss": 0.45025482177734377,
      "lr/ungrouped": 3.278883321514613e-05,
      "step": 1490
    },
    {
      "epoch": 0.4320743167824866,
      "grad_norm": 1.140481948852539,
      "learning_rate": 3.25848255949587e-05,
      "loss": 0.4550967216491699,
      "lr/ungrouped": 3.256212337379886e-05,
      "step": 1500
    },
    {
      "epoch": 0.43495481222770316,
      "grad_norm": 1.1562436819076538,
      "learning_rate": 3.2357496887641284e-05,
      "loss": 0.5989727973937988,
      "lr/ungrouped": 3.2334726934784265e-05,
      "step": 1510
    },
    {
      "epoch": 0.4378353076729197,
      "grad_norm": 1.045846939086914,
      "learning_rate": 3.2129500161574304e-05,
      "loss": 0.5038487911224365,
      "lr/ungrouped": 3.210666454440195e-05,
      "step": 1520
    },
    {
      "epoch": 0.4407158031181363,
      "grad_norm": 0.9375977516174316,
      "learning_rate": 3.1900856117559965e-05,
      "loss": 0.47698178291320803,
      "lr/ungrouped": 3.1877956909416065e-05,
      "step": 1530
    },
    {
      "epoch": 0.4435962985633529,
      "grad_norm": 1.0946990251541138,
      "learning_rate": 3.167158551517326e-05,
      "loss": 0.4762615203857422,
      "lr/ungrouped": 3.1648624795175286e-05,
      "step": 1540
    },
    {
      "epoch": 0.4464767940085695,
      "grad_norm": 1.1144299507141113,
      "learning_rate": 3.1441709170877105e-05,
      "loss": 0.4851114749908447,
      "lr/ungrouped": 3.141868902372747e-05,
      "step": 1550
    },
    {
      "epoch": 0.44935728945378606,
      "grad_norm": 1.282989263534546,
      "learning_rate": 3.121124795613233e-05,
      "loss": 0.5381083011627197,
      "lr/ungrouped": 3.118817047192907e-05,
      "step": 1560
    },
    {
      "epoch": 0.4522377848990026,
      "grad_norm": 0.9744628667831421,
      "learning_rate": 3.098022279550265e-05,
      "loss": 0.5239127159118653,
      "lr/ungrouped": 3.0957090069549714e-05,
      "step": 1570
    },
    {
      "epoch": 0.4551182803442192,
      "grad_norm": 0.8819881677627563,
      "learning_rate": 3.0748654664754886e-05,
      "loss": 0.4305234432220459,
      "lr/ungrouped": 3.072546879737182e-05,
      "step": 1580
    },
    {
      "epoch": 0.45799877578943576,
      "grad_norm": 0.8542435169219971,
      "learning_rate": 3.0516564588954395e-05,
      "loss": 0.5580653667449951,
      "lr/ungrouped": 3.0493327685285723e-05,
      "step": 1590
    },
    {
      "epoch": 0.4608792712346524,
      "grad_norm": 0.7599719762802124,
      "learning_rate": 3.0283973640556218e-05,
      "loss": 0.5187072277069091,
      "lr/ungrouped": 3.0260687810380233e-05,
      "step": 1600
    },
    {
      "epoch": 0.46375976667986896,
      "grad_norm": 0.9886980056762695,
      "learning_rate": 3.005090293749174e-05,
      "loss": 0.6065029621124267,
      "lr/ungrouped": 3.0027570295029007e-05,
      "step": 1610
    },
    {
      "epoch": 0.4666402621250855,
      "grad_norm": 0.9663744568824768,
      "learning_rate": 2.9817373641251344e-05,
      "loss": 0.4928086280822754,
      "lr/ungrouped": 2.9793996304972704e-05,
      "step": 1620
    },
    {
      "epoch": 0.4695207575703021,
      "grad_norm": 0.931624174118042,
      "learning_rate": 2.958340695496306e-05,
      "loss": 0.4987185955047607,
      "lr/ungrouped": 2.9559987047397303e-05,
      "step": 1630
    },
    {
      "epoch": 0.47240125301551866,
      "grad_norm": 0.8112949728965759,
      "learning_rate": 2.9349024121467434e-05,
      "loss": 0.38908233642578127,
      "lr/ungrouped": 2.9325563769008573e-05,
      "step": 1640
    },
    {
      "epoch": 0.47528174846073523,
      "grad_norm": 0.9888959527015686,
      "learning_rate": 2.9114246421388804e-05,
      "loss": 0.46108264923095704,
      "lr/ungrouped": 2.909074775410302e-05,
      "step": 1650
    },
    {
      "epoch": 0.4781622439059518,
      "grad_norm": 1.1524813175201416,
      "learning_rate": 2.8879095171203145e-05,
      "loss": 0.45529804229736326,
      "lr/ungrouped": 2.88555603226354e-05,
      "step": 1660
    },
    {
      "epoch": 0.4810427393511684,
      "grad_norm": 0.9758327007293701,
      "learning_rate": 2.864359172130266e-05,
      "loss": 0.4319746017456055,
      "lr/ungrouped": 2.862002282828295e-05,
      "step": 1670
    },
    {
      "epoch": 0.483923234796385,
      "grad_norm": 0.9985355138778687,
      "learning_rate": 2.8407757454057248e-05,
      "loss": 0.5377474308013916,
      "lr/ungrouped": 2.8384156656506626e-05,
      "step": 1680
    },
    {
      "epoch": 0.48680373024160156,
      "grad_norm": 1.3406964540481567,
      "learning_rate": 2.8171613781873168e-05,
      "loss": 0.5202048301696778,
      "lr/ungrouped": 2.8147983222609426e-05,
      "step": 1690
    },
    {
      "epoch": 0.48968422568681813,
      "grad_norm": 1.2826424837112427,
      "learning_rate": 2.793518214524886e-05,
      "loss": 0.5007876396179199,
      "lr/ungrouped": 2.7911523969791997e-05,
      "step": 1700
    },
    {
      "epoch": 0.4925647211320347,
      "grad_norm": 2.1532509326934814,
      "learning_rate": 2.7698484010828322e-05,
      "loss": 0.4914586067199707,
      "lr/ungrouped": 2.7674800367205704e-05,
      "step": 1710
    },
    {
      "epoch": 0.49544521657725127,
      "grad_norm": 1.2262667417526245,
      "learning_rate": 2.746154086945199e-05,
      "loss": 0.5049422740936279,
      "lr/ungrouped": 2.743783390800337e-05,
      "step": 1720
    },
    {
      "epoch": 0.49832571202246784,
      "grad_norm": 1.0331627130508423,
      "learning_rate": 2.7224374234205564e-05,
      "loss": 0.49055981636047363,
      "lr/ungrouped": 2.7200646107387824e-05,
      "step": 1730
    },
    {
      "epoch": 0.5012062074676844,
      "grad_norm": 0.9930821657180786,
      "learning_rate": 2.6987005638466705e-05,
      "loss": 0.4328910827636719,
      "lr/ungrouped": 2.6963258500658405e-05,
      "step": 1740
    },
    {
      "epoch": 0.504086702912901,
      "grad_norm": 1.0581098794937134,
      "learning_rate": 2.6749456633949932e-05,
      "loss": 0.5578174114227294,
      "lr/ungrouped": 2.6725692641255734e-05,
      "step": 1750
    },
    {
      "epoch": 0.5069671983581175,
      "grad_norm": 1.2751959562301636,
      "learning_rate": 2.651174878874985e-05,
      "loss": 0.5080609321594238,
      "lr/ungrouped": 2.6487970098804742e-05,
      "step": 1760
    },
    {
      "epoch": 0.5098476938033342,
      "grad_norm": 0.8918941020965576,
      "learning_rate": 2.6273903685382907e-05,
      "loss": 0.5544655799865723,
      "lr/ungrouped": 2.6250112457156296e-05,
      "step": 1770
    },
    {
      "epoch": 0.5127281892485508,
      "grad_norm": 1.1249994039535522,
      "learning_rate": 2.6035942918827793e-05,
      "loss": 0.5208686351776123,
      "lr/ungrouped": 2.601214131242749e-05,
      "step": 1780
    },
    {
      "epoch": 0.5156086846937673,
      "grad_norm": 1.0296849012374878,
      "learning_rate": 2.579788809456477e-05,
      "loss": 0.4910425186157227,
      "lr/ungrouped": 2.5774078271040864e-05,
      "step": 1790
    },
    {
      "epoch": 0.5184891801389839,
      "grad_norm": 1.0689972639083862,
      "learning_rate": 2.5559760826613977e-05,
      "loss": 0.4991764068603516,
      "lr/ungrouped": 2.553594494776264e-05,
      "step": 1800
    },
    {
      "epoch": 0.5213696755842004,
      "grad_norm": 1.2130671739578247,
      "learning_rate": 2.5321582735573056e-05,
      "loss": 0.5711346626281738,
      "lr/ungrouped": 2.5297762963740247e-05,
      "step": 1810
    },
    {
      "epoch": 0.5242501710294171,
      "grad_norm": 0.897699773311615,
      "learning_rate": 2.5083375446654083e-05,
      "loss": 0.508837080001831,
      "lr/ungrouped": 2.5059553944539204e-05,
      "step": 1820
    },
    {
      "epoch": 0.5271306664746337,
      "grad_norm": 0.878955066204071,
      "learning_rate": 2.4845160587720134e-05,
      "loss": 0.5688987731933594,
      "lr/ungrouped": 2.4821339518179692e-05,
      "step": 1830
    },
    {
      "epoch": 0.5300111619198502,
      "grad_norm": 0.8825750946998596,
      "learning_rate": 2.4606959787321595e-05,
      "loss": 0.6016063213348388,
      "lr/ungrouped": 2.4583141313172842e-05,
      "step": 1840
    },
    {
      "epoch": 0.5328916573650668,
      "grad_norm": 1.1181200742721558,
      "learning_rate": 2.4368794672732442e-05,
      "loss": 0.3941169023513794,
      "lr/ungrouped": 2.434498095655696e-05,
      "step": 1850
    },
    {
      "epoch": 0.5357721528102833,
      "grad_norm": 1.0510873794555664,
      "learning_rate": 2.4130686867986544e-05,
      "loss": 0.5355625152587891,
      "lr/ungrouped": 2.4106880071933924e-05,
      "step": 1860
    },
    {
      "epoch": 0.5386526482555,
      "grad_norm": 1.037467122077942,
      "learning_rate": 2.38926579919144e-05,
      "loss": 0.5746402740478516,
      "lr/ungrouped": 2.3868860277505932e-05,
      "step": 1870
    },
    {
      "epoch": 0.5415331437007165,
      "grad_norm": 1.060508131980896,
      "learning_rate": 2.3654729656180204e-05,
      "loss": 0.5412226676940918,
      "lr/ungrouped": 2.3630943184112605e-05,
      "step": 1880
    },
    {
      "epoch": 0.5444136391459331,
      "grad_norm": 1.1407394409179688,
      "learning_rate": 2.3416923463319686e-05,
      "loss": 0.4937282085418701,
      "lr/ungrouped": 2.339315039326895e-05,
      "step": 1890
    },
    {
      "epoch": 0.5472941345911497,
      "grad_norm": 0.9881560802459717,
      "learning_rate": 2.317926100477869e-05,
      "loss": 0.5151865959167481,
      "lr/ungrouped": 2.315550349520396e-05,
      "step": 1900
    },
    {
      "epoch": 0.5501746300363662,
      "grad_norm": 0.9714236259460449,
      "learning_rate": 2.2941763858952788e-05,
      "loss": 0.5674307346343994,
      "lr/ungrouped": 2.2918024066900433e-05,
      "step": 1910
    },
    {
      "epoch": 0.5530551254815829,
      "grad_norm": 1.0508372783660889,
      "learning_rate": 2.2704453589228143e-05,
      "loss": 0.5401134490966797,
      "lr/ungrouped": 2.268073367013586e-05,
      "step": 1920
    },
    {
      "epoch": 0.5559356209267994,
      "grad_norm": 1.160943627357483,
      "learning_rate": 2.246735174202358e-05,
      "loss": 0.513639497756958,
      "lr/ungrouped": 2.2443653849524723e-05,
      "step": 1930
    },
    {
      "epoch": 0.558816116372016,
      "grad_norm": 1.0371874570846558,
      "learning_rate": 2.2230479844834395e-05,
      "loss": 0.5705047607421875,
      "lr/ungrouped": 2.220680613056243e-05,
      "step": 1940
    },
    {
      "epoch": 0.5616966118172325,
      "grad_norm": 0.9895194172859192,
      "learning_rate": 2.1993859404277693e-05,
      "loss": 0.558998966217041,
      "lr/ungrouped": 2.1970212017670837e-05,
      "step": 1950
    },
    {
      "epoch": 0.5645771072624491,
      "grad_norm": 1.0365338325500488,
      "learning_rate": 2.1757511904139793e-05,
      "loss": 0.4781347274780273,
      "lr/ungrouped": 2.173389299224587e-05,
      "step": 1960
    },
    {
      "epoch": 0.5674576027076658,
      "grad_norm": 1.6648669242858887,
      "learning_rate": 2.152145880342555e-05,
      "loss": 0.5348897933959961,
      "lr/ungrouped": 2.149787051070702e-05,
      "step": 1970
    },
    {
      "epoch": 0.5703380981528823,
      "grad_norm": 1.0151875019073486,
      "learning_rate": 2.1285721534410047e-05,
      "loss": 0.5148487091064453,
      "lr/ungrouped": 2.1262166002549344e-05,
      "step": 1980
    },
    {
      "epoch": 0.5732185935980989,
      "grad_norm": 1.114115834236145,
      "learning_rate": 2.105032150069264e-05,
      "loss": 0.5177653312683106,
      "lr/ungrouped": 2.102680086839769e-05,
      "step": 1990
    },
    {
      "epoch": 0.5760990890433154,
      "grad_norm": 1.0472909212112427,
      "learning_rate": 2.0815280075253635e-05,
      "loss": 0.4812153339385986,
      "lr/ungrouped": 2.0791796478063704e-05,
      "step": 2000
    },
    {
      "epoch": 0.578979584488532,
      "grad_norm": 0.8699169158935547,
      "learning_rate": 2.0580618598513785e-05,
      "loss": 0.5446981430053711,
      "lr/ungrouped": 2.0557174168605536e-05,
      "step": 2010
    },
    {
      "epoch": 0.5818600799337486,
      "grad_norm": 1.2035599946975708,
      "learning_rate": 2.0346358376396628e-05,
      "loss": 0.5508872509002686,
      "lr/ungrouped": 2.032295524239057e-05,
      "step": 2020
    },
    {
      "epoch": 0.5847405753789652,
      "grad_norm": 1.1340709924697876,
      "learning_rate": 2.0112520678394107e-05,
      "loss": 0.5180424690246582,
      "lr/ungrouped": 2.0089160965161312e-05,
      "step": 2030
    },
    {
      "epoch": 0.5876210708241818,
      "grad_norm": 1.0939905643463135,
      "learning_rate": 1.987912673563536e-05,
      "loss": 0.4519155979156494,
      "lr/ungrouped": 1.9855812564104546e-05,
      "step": 2040
    },
    {
      "epoch": 0.5905015662693983,
      "grad_norm": 1.2461870908737183,
      "learning_rate": 1.9646197738959102e-05,
      "loss": 0.5145881652832032,
      "lr/ungrouped": 1.962293122592405e-05,
      "step": 2050
    },
    {
      "epoch": 0.5933820617146149,
      "grad_norm": 0.9741010069847107,
      "learning_rate": 1.9413754836989574e-05,
      "loss": 0.5592999458312988,
      "lr/ungrouped": 1.9390538094916944e-05,
      "step": 2060
    },
    {
      "epoch": 0.5962625571598315,
      "grad_norm": 1.1677906513214111,
      "learning_rate": 1.9181819134216387e-05,
      "loss": 0.5196237087249755,
      "lr/ungrouped": 1.9158654271053956e-05,
      "step": 2070
    },
    {
      "epoch": 0.5991430526050481,
      "grad_norm": 1.2987737655639648,
      "learning_rate": 1.895041168907839e-05,
      "loss": 0.5596961498260498,
      "lr/ungrouped": 1.8927300808063588e-05,
      "step": 2080
    },
    {
      "epoch": 0.6020235480502647,
      "grad_norm": 1.0141394138336182,
      "learning_rate": 1.8719553512051624e-05,
      "loss": 0.5235199451446533,
      "lr/ungrouped": 1.869649871152061e-05,
      "step": 2090
    },
    {
      "epoch": 0.6049040434954812,
      "grad_norm": 1.2024528980255127,
      "learning_rate": 1.8489265563741725e-05,
      "loss": 0.5100419521331787,
      "lr/ungrouped": 1.8466268936938893e-05,
      "step": 2100
    },
    {
      "epoch": 0.6077845389406978,
      "grad_norm": 1.0439091920852661,
      "learning_rate": 1.8259568752980817e-05,
      "loss": 0.44300432205200196,
      "lr/ungrouped": 1.82366323878687e-05,
      "step": 2110
    },
    {
      "epoch": 0.6106650343859144,
      "grad_norm": 1.0874866247177124,
      "learning_rate": 1.8030483934929122e-05,
      "loss": 0.48569788932800295,
      "lr/ungrouped": 1.800760991399884e-05,
      "step": 2120
    },
    {
      "epoch": 0.613545529831131,
      "grad_norm": 1.2774713039398193,
      "learning_rate": 1.780203190918138e-05,
      "loss": 0.48184938430786134,
      "lr/ungrouped": 1.7779222309263555e-05,
      "step": 2130
    },
    {
      "epoch": 0.6164260252763475,
      "grad_norm": 1.2759746313095093,
      "learning_rate": 1.7574233417878414e-05,
      "loss": 0.4664162635803223,
      "lr/ungrouped": 1.7551490309954623e-05,
      "step": 2140
    },
    {
      "epoch": 0.6193065207215641,
      "grad_norm": 0.8512833714485168,
      "learning_rate": 1.7347109143823863e-05,
      "loss": 0.4700309276580811,
      "lr/ungrouped": 1.732443459283858e-05,
      "step": 2150
    },
    {
      "epoch": 0.6221870161667807,
      "grad_norm": 1.0298893451690674,
      "learning_rate": 1.712067970860627e-05,
      "loss": 0.462721586227417,
      "lr/ungrouped": 1.709807577327937e-05,
      "step": 2160
    },
    {
      "epoch": 0.6250675116119973,
      "grad_norm": 1.1021806001663208,
      "learning_rate": 1.689496567072678e-05,
      "loss": 0.5450207710266113,
      "lr/ungrouped": 1.6872434403366662e-05,
      "step": 2170
    },
    {
      "epoch": 0.6279480070572139,
      "grad_norm": 1.132588267326355,
      "learning_rate": 1.666998752373254e-05,
      "loss": 0.5511212348937988,
      "lr/ungrouped": 1.6647530970049756e-05,
      "step": 2180
    },
    {
      "epoch": 0.6308285025024304,
      "grad_norm": 0.9780980348587036,
      "learning_rate": 1.6445765694356002e-05,
      "loss": 0.5909686088562012,
      "lr/ungrouped": 1.6423385893277536e-05,
      "step": 2190
    },
    {
      "epoch": 0.633708997947647,
      "grad_norm": 1.138410210609436,
      "learning_rate": 1.622232054066028e-05,
      "loss": 0.44877986907958983,
      "lr/ungrouped": 1.6200019524144417e-05,
      "step": 2200
    },
    {
      "epoch": 0.6365894933928635,
      "grad_norm": 1.1982225179672241,
      "learning_rate": 1.59996723501908e-05,
      "loss": 0.446318244934082,
      "lr/ungrouped": 1.5977452143042627e-05,
      "step": 2210
    },
    {
      "epoch": 0.6394699888380802,
      "grad_norm": 1.1241157054901123,
      "learning_rate": 1.577784133813323e-05,
      "loss": 0.5420880794525147,
      "lr/ungrouped": 1.5755703957820812e-05,
      "step": 2220
    },
    {
      "epoch": 0.6423504842832968,
      "grad_norm": 1.0700193643569946,
      "learning_rate": 1.555684764547813e-05,
      "loss": 0.4542251110076904,
      "lr/ungrouped": 1.5534795101949328e-05,
      "step": 2230
    },
    {
      "epoch": 0.6452309797285133,
      "grad_norm": 1.065657615661621,
      "learning_rate": 1.5336711337192227e-05,
      "loss": 0.536868667602539,
      "lr/ungrouped": 1.5314745632692203e-05,
      "step": 2240
    },
    {
      "epoch": 0.6481114751737299,
      "grad_norm": 1.1254609823226929,
      "learning_rate": 1.5117452400396629e-05,
      "loss": 0.5238078594207763,
      "lr/ungrouped": 1.5095575529286055e-05,
      "step": 2250
    },
    {
      "epoch": 0.6509919706189464,
      "grad_norm": 0.8902557492256165,
      "learning_rate": 1.4899090742552135e-05,
      "loss": 0.4986436367034912,
      "lr/ungrouped": 1.4877304691126123e-05,
      "step": 2260
    },
    {
      "epoch": 0.653872466064163,
      "grad_norm": 1.2387348413467407,
      "learning_rate": 1.468164618965171e-05,
      "loss": 0.5431238651275635,
      "lr/ungrouped": 1.4659952935959456e-05,
      "step": 2270
    },
    {
      "epoch": 0.6567529615093796,
      "grad_norm": 1.0034769773483276,
      "learning_rate": 1.4465138484420443e-05,
      "loss": 0.4437143802642822,
      "lr/ungrouped": 1.4443539998085647e-05,
      "step": 2280
    },
    {
      "epoch": 0.6596334569545962,
      "grad_norm": 0.8418393731117249,
      "learning_rate": 1.4249587284522998e-05,
      "loss": 0.5147094249725341,
      "lr/ungrouped": 1.4228085526564999e-05,
      "step": 2290
    },
    {
      "epoch": 0.6625139523998128,
      "grad_norm": 1.0145301818847656,
      "learning_rate": 1.4035012160778768e-05,
      "loss": 0.42089481353759767,
      "lr/ungrouped": 1.4013609083434579e-05,
      "step": 2300
    },
    {
      "epoch": 0.6653944478450293,
      "grad_norm": 1.0261805057525635,
      "learning_rate": 1.382143259538507e-05,
      "loss": 0.5229764938354492,
      "lr/ungrouped": 1.3800130141932006e-05,
      "step": 2310
    },
    {
      "epoch": 0.668274943290246,
      "grad_norm": 1.0287816524505615,
      "learning_rate": 1.3608867980148147e-05,
      "loss": 0.4696493148803711,
      "lr/ungrouped": 1.3587668084727474e-05,
      "step": 2320
    },
    {
      "epoch": 0.6711554387354625,
      "grad_norm": 1.1746306419372559,
      "learning_rate": 1.3397337614722577e-05,
      "loss": 0.505794620513916,
      "lr/ungrouped": 1.3376242202163868e-05,
      "step": 2330
    },
    {
      "epoch": 0.6740359341806791,
      "grad_norm": 1.1598416566848755,
      "learning_rate": 1.3186860704858955e-05,
      "loss": 0.4999171257019043,
      "lr/ungrouped": 1.3165871690505337e-05,
      "step": 2340
    },
    {
      "epoch": 0.6769164296258956,
      "grad_norm": 1.0474591255187988,
      "learning_rate": 1.2977456360660117e-05,
      "loss": 0.542423677444458,
      "lr/ungrouped": 1.2956575650194387e-05,
      "step": 2350
    },
    {
      "epoch": 0.6797969250711122,
      "grad_norm": 0.9387258291244507,
      "learning_rate": 1.2769143594846062e-05,
      "loss": 0.42920703887939454,
      "lr/ungrouped": 1.2748373084117593e-05,
      "step": 2360
    },
    {
      "epoch": 0.6826774205163288,
      "grad_norm": 1.1878985166549683,
      "learning_rate": 1.256194132102766e-05,
      "loss": 0.42487411499023436,
      "lr/ungrouped": 1.254128289588039e-05,
      "step": 2370
    },
    {
      "epoch": 0.6855579159615454,
      "grad_norm": 1.5014756917953491,
      "learning_rate": 1.2355868351989509e-05,
      "loss": 0.49875149726867674,
      "lr/ungrouped": 1.2335323888090586e-05,
      "step": 2380
    },
    {
      "epoch": 0.688438411406762,
      "grad_norm": 1.1662882566452026,
      "learning_rate": 1.2150943397981724e-05,
      "loss": 0.49876132011413576,
      "lr/ungrouped": 1.2130514760651282e-05,
      "step": 2390
    },
    {
      "epoch": 0.6913189068519785,
      "grad_norm": 0.9223743677139282,
      "learning_rate": 1.194718506502122e-05,
      "loss": 0.506182861328125,
      "lr/ungrouped": 1.1926874109063e-05,
      "step": 2400
    },
    {
      "epoch": 0.6941994022971951,
      "grad_norm": 1.344974160194397,
      "learning_rate": 1.1744611853202376e-05,
      "loss": 0.5172177791595459,
      "lr/ungrouped": 1.1724420422735313e-05,
      "step": 2410
    },
    {
      "epoch": 0.6970798977424117,
      "grad_norm": 1.1166934967041016,
      "learning_rate": 1.1543242155017322e-05,
      "loss": 0.4599472522735596,
      "lr/ungrouped": 1.1523172083308154e-05,
      "step": 2420
    },
    {
      "epoch": 0.6999603931876283,
      "grad_norm": 0.9860295057296753,
      "learning_rate": 1.1343094253686032e-05,
      "loss": 0.5041580677032471,
      "lr/ungrouped": 1.1323147362982761e-05,
      "step": 2430
    },
    {
      "epoch": 0.7028408886328449,
      "grad_norm": 0.9514420032501221,
      "learning_rate": 1.1144186321496277e-05,
      "loss": 0.5381842613220215,
      "lr/ungrouped": 1.1124364422862826e-05,
      "step": 2440
    },
    {
      "epoch": 0.7057213840780614,
      "grad_norm": 0.9792171120643616,
      "learning_rate": 1.0946536418153716e-05,
      "loss": 0.47804694175720214,
      "lr/ungrouped": 1.0926841311305392e-05,
      "step": 2450
    },
    {
      "epoch": 0.708601879523278,
      "grad_norm": 1.1488412618637085,
      "learning_rate": 1.0750162489142157e-05,
      "loss": 0.5548057556152344,
      "lr/ungrouped": 1.0730595962282339e-05,
      "step": 2460
    },
    {
      "epoch": 0.7114823749684945,
      "grad_norm": 1.1042265892028809,
      "learning_rate": 1.0555082364094221e-05,
      "loss": 0.4182869911193848,
      "lr/ungrouped": 1.053564619375193e-05,
      "step": 2470
    },
    {
      "epoch": 0.7143628704137112,
      "grad_norm": 1.070912480354309,
      "learning_rate": 1.0361313755172506e-05,
      "loss": 0.46544060707092283,
      "lr/ungrouped": 1.0342009706041133e-05,
      "step": 2480
    },
    {
      "epoch": 0.7172433658589278,
      "grad_norm": 1.0181386470794678,
      "learning_rate": 1.0168874255461431e-05,
      "loss": 0.540457820892334,
      "lr/ungrouped": 1.0149704080238543e-05,
      "step": 2490
    },
    {
      "epoch": 0.7201238613041443,
      "grad_norm": 1.11381196975708,
      "learning_rate": 9.977781337369873e-06,
      "loss": 0.516001844406128,
      "lr/ungrouped": 9.958746776598018e-06,
      "step": 2500
    },
    {
      "epoch": 0.7230043567493609,
      "grad_norm": 1.2050833702087402,
      "learning_rate": 9.788052351044788e-06,
      "loss": 0.5464937210083007,
      "lr/ungrouped": 9.769155132953522e-06,
      "step": 2510
    },
    {
      "epoch": 0.7258848521945774,
      "grad_norm": 1.1345531940460205,
      "learning_rate": 9.599704522795899e-06,
      "loss": 0.5882165431976318,
      "lr/ungrouped": 9.580946363144821e-06,
      "step": 2520
    },
    {
      "epoch": 0.7287653476397941,
      "grad_norm": 1.0768378973007202,
      "learning_rate": 9.412754953531663e-06,
      "loss": 0.5241845130920411,
      "lr/ungrouped": 9.394137555454677e-06,
      "step": 2530
    },
    {
      "epoch": 0.7316458430850106,
      "grad_norm": 1.1802098751068115,
      "learning_rate": 9.22722061720659e-06,
      "loss": 0.5538426876068115,
      "lr/ungrouped": 9.20874567105725e-06,
      "step": 2540
    },
    {
      "epoch": 0.7345263385302272,
      "grad_norm": 1.0653661489486694,
      "learning_rate": 9.043118359280111e-06,
      "loss": 0.5484429836273194,
      "lr/ungrouped": 9.024787542478133e-06,
      "step": 2550
    },
    {
      "epoch": 0.7374068339754438,
      "grad_norm": 1.093384027481079,
      "learning_rate": 8.860464895187112e-06,
      "loss": 0.4375253200531006,
      "lr/ungrouped": 8.842279872066118e-06,
      "step": 2560
    },
    {
      "epoch": 0.7402873294206603,
      "grad_norm": 1.1435723304748535,
      "learning_rate": 8.679276808820258e-06,
      "loss": 0.6201212406158447,
      "lr/ungrouped": 8.6612392304766e-06,
      "step": 2570
    },
    {
      "epoch": 0.743167824865877,
      "grad_norm": 1.0707294940948486,
      "learning_rate": 8.499570551024282e-06,
      "loss": 0.5173251628875732,
      "lr/ungrouped": 8.481682055167203e-06,
      "step": 2580
    },
    {
      "epoch": 0.7460483203110935,
      "grad_norm": 1.0999010801315308,
      "learning_rate": 8.32136243810233e-06,
      "loss": 0.5098042488098145,
      "lr/ungrouped": 8.303624648905197e-06,
      "step": 2590
    },
    {
      "epoch": 0.7489288157563101,
      "grad_norm": 0.9520456194877625,
      "learning_rate": 8.144668650334539e-06,
      "loss": 0.42621121406555174,
      "lr/ungrouped": 8.127083178287443e-06,
      "step": 2600
    },
    {
      "epoch": 0.7518093112015266,
      "grad_norm": 1.2338683605194092,
      "learning_rate": 7.969505230508962e-06,
      "loss": 0.4682492733001709,
      "lr/ungrouped": 7.952073672272465e-06,
      "step": 2610
    },
    {
      "epoch": 0.7546898066467432,
      "grad_norm": 1.1173771619796753,
      "learning_rate": 7.795888082464966e-06,
      "loss": 0.4147955417633057,
      "lr/ungrouped": 7.778612020725109e-06,
      "step": 2620
    },
    {
      "epoch": 0.7575703020919599,
      "grad_norm": 1.0187040567398071,
      "learning_rate": 7.623832969649275e-06,
      "loss": 0.5439486980438233,
      "lr/ungrouped": 7.6067139729739335e-06,
      "step": 2630
    },
    {
      "epoch": 0.7604507975371764,
      "grad_norm": 1.1298575401306152,
      "learning_rate": 7.453355513684706e-06,
      "loss": 0.5126908779144287,
      "lr/ungrouped": 7.436395136381116e-06,
      "step": 2640
    },
    {
      "epoch": 0.763331292982393,
      "grad_norm": 1.1339991092681885,
      "learning_rate": 7.284471192951847e-06,
      "loss": 0.49756360054016113,
      "lr/ungrouped": 7.267670974925539e-06,
      "step": 2650
    },
    {
      "epoch": 0.7662117884276095,
      "grad_norm": 1.2817519903182983,
      "learning_rate": 7.11719534118368e-06,
      "loss": 0.4942440986633301,
      "lr/ungrouped": 7.100556807798614e-06,
      "step": 2660
    },
    {
      "epoch": 0.7690922838728261,
      "grad_norm": 1.183738112449646,
      "learning_rate": 6.951543146073414e-06,
      "loss": 0.3188255548477173,
      "lr/ungrouped": 6.935067808013501e-06,
      "step": 2670
    },
    {
      "epoch": 0.7719727793180428,
      "grad_norm": 1.0017569065093994,
      "learning_rate": 6.787529647895441e-06,
      "loss": 0.44585723876953126,
      "lr/ungrouped": 6.771219001027415e-06,
      "step": 2680
    },
    {
      "epoch": 0.7748532747632593,
      "grad_norm": 0.9212375283241272,
      "learning_rate": 6.625169738139861e-06,
      "loss": 0.4645206451416016,
      "lr/ungrouped": 6.609025263377414e-06,
      "step": 2690
    },
    {
      "epoch": 0.7777337702084759,
      "grad_norm": 0.8730063438415527,
      "learning_rate": 6.464478158160361e-06,
      "loss": 0.4745302200317383,
      "lr/ungrouped": 6.448501321329722e-06,
      "step": 2700
    },
    {
      "epoch": 0.7806142656536924,
      "grad_norm": 1.132372260093689,
      "learning_rate": 6.305469497835803e-06,
      "loss": 0.46212496757507326,
      "lr/ungrouped": 6.289661749542611e-06,
      "step": 2710
    },
    {
      "epoch": 0.783494761098909,
      "grad_norm": 1.0981274843215942,
      "learning_rate": 6.148158194245551e-06,
      "loss": 0.48363475799560546,
      "lr/ungrouped": 6.132520969743194e-06,
      "step": 2720
    },
    {
      "epoch": 0.7863752565441255,
      "grad_norm": 0.9928323030471802,
      "learning_rate": 5.992558530358638e-06,
      "loss": 0.5386440753936768,
      "lr/ungrouped": 5.9770932494178976e-06,
      "step": 2730
    },
    {
      "epoch": 0.7892557519893422,
      "grad_norm": 1.0599957704544067,
      "learning_rate": 5.838684633737018e-06,
      "loss": 0.4258011817932129,
      "lr/ungrouped": 5.82339270051718e-06,
      "step": 2740
    },
    {
      "epoch": 0.7921362474345588,
      "grad_norm": 1.2112878561019897,
      "learning_rate": 5.686550475252766e-06,
      "loss": 0.49411768913269044,
      "lr/ungrouped": 5.671433278174151e-06,
      "step": 2750
    },
    {
      "epoch": 0.7950167428797753,
      "grad_norm": 1.191348910331726,
      "learning_rate": 5.536169867819701e-06,
      "loss": 0.48946733474731446,
      "lr/ungrouped": 5.521228779437568e-06,
      "step": 2760
    },
    {
      "epoch": 0.7978972383249919,
      "grad_norm": 1.105289101600647,
      "learning_rate": 5.387556465139184e-06,
      "loss": 0.4951319694519043,
      "lr/ungrouped": 5.372792842019145e-06,
      "step": 2770
    },
    {
      "epoch": 0.8007777337702084,
      "grad_norm": 0.7638000845909119,
      "learning_rate": 5.240723760460481e-06,
      "loss": 0.40555453300476074,
      "lr/ungrouped": 5.226138943055287e-06,
      "step": 2780
    },
    {
      "epoch": 0.8036582292154251,
      "grad_norm": 1.061234474182129,
      "learning_rate": 5.095685085355625e-06,
      "loss": 0.5597369194030761,
      "lr/ungrouped": 5.081280397883509e-06,
      "step": 2790
    },
    {
      "epoch": 0.8065387246606416,
      "grad_norm": 1.0556143522262573,
      "learning_rate": 4.952453608509e-06,
      "loss": 0.5111080646514893,
      "lr/ungrouped": 4.938230358833426e-06,
      "step": 2800
    },
    {
      "epoch": 0.8094192201058582,
      "grad_norm": 1.1975961923599243,
      "learning_rate": 4.8110423345217145e-06,
      "loss": 0.4770818710327148,
      "lr/ungrouped": 4.797001814032609e-06,
      "step": 2810
    },
    {
      "epoch": 0.8122997155510748,
      "grad_norm": 1.054638385772705,
      "learning_rate": 4.6714641027308044e-06,
      "loss": 0.5621273994445801,
      "lr/ungrouped": 4.657607586227345e-06,
      "step": 2820
    },
    {
      "epoch": 0.8151802109962913,
      "grad_norm": 0.919845461845398,
      "learning_rate": 4.533731586043557e-06,
      "loss": 0.5498823165893555,
      "lr/ungrouped": 4.5200603316184055e-06,
      "step": 2830
    },
    {
      "epoch": 0.818060706441508,
      "grad_norm": 1.1536132097244263,
      "learning_rate": 4.39785728978685e-06,
      "loss": 0.5180442333221436,
      "lr/ungrouped": 4.384372538711931e-06,
      "step": 2840
    },
    {
      "epoch": 0.8209412018867245,
      "grad_norm": 1.3467572927474976,
      "learning_rate": 4.263853550571734e-06,
      "loss": 0.5023393630981445,
      "lr/ungrouped": 4.2505565271855205e-06,
      "step": 2850
    },
    {
      "epoch": 0.8238216973319411,
      "grad_norm": 1.1213948726654053,
      "learning_rate": 4.131732535173382e-06,
      "loss": 0.4931967258453369,
      "lr/ungrouped": 4.118624446769745e-06,
      "step": 2860
    },
    {
      "epoch": 0.8267021927771576,
      "grad_norm": 1.1812928915023804,
      "learning_rate": 4.001506239426339e-06,
      "loss": 0.4181187629699707,
      "lr/ungrouped": 3.988588276144947e-06,
      "step": 2870
    },
    {
      "epoch": 0.8295826882223742,
      "grad_norm": 1.0258709192276,
      "learning_rate": 3.873186487135455e-06,
      "loss": 0.4609867572784424,
      "lr/ungrouped": 3.860459821853679e-06,
      "step": 2880
    },
    {
      "epoch": 0.8324631836675909,
      "grad_norm": 0.9248396158218384,
      "learning_rate": 3.7467849290022727e-06,
      "loss": 0.36834762096405027,
      "lr/ungrouped": 3.734250717228735e-06,
      "step": 2890
    },
    {
      "epoch": 0.8353436791128074,
      "grad_norm": 1.080792784690857,
      "learning_rate": 3.6223130415672853e-06,
      "loss": 0.47798714637756345,
      "lr/ungrouped": 3.6099724213369048e-06,
      "step": 2900
    },
    {
      "epoch": 0.838224174558024,
      "grad_norm": 1.1009042263031006,
      "learning_rate": 3.499782126167853e-06,
      "loss": 0.4232470512390137,
      "lr/ungrouped": 3.487636217938567e-06,
      "step": 2910
    },
    {
      "epoch": 0.8411046700032405,
      "grad_norm": 1.2341742515563965,
      "learning_rate": 3.379203307912171e-06,
      "loss": 0.5825160026550293,
      "lr/ungrouped": 3.3672532144631676e-06,
      "step": 2920
    },
    {
      "epoch": 0.8439851654484571,
      "grad_norm": 1.2406553030014038,
      "learning_rate": 3.26058753466916e-06,
      "loss": 0.4204261779785156,
      "lr/ungrouped": 3.2488343410007476e-06,
      "step": 2930
    },
    {
      "epoch": 0.8468656608936738,
      "grad_norm": 1.0230709314346313,
      "learning_rate": 3.1439455760744112e-06,
      "loss": 0.5102825164794922,
      "lr/ungrouped": 3.132390349309547e-06,
      "step": 2940
    },
    {
      "epoch": 0.8497461563388903,
      "grad_norm": 1.3116899728775024,
      "learning_rate": 3.0292880225524513e-06,
      "loss": 0.5189552307128906,
      "lr/ungrouped": 3.0179318118398025e-06,
      "step": 2950
    },
    {
      "epoch": 0.8526266517841069,
      "grad_norm": 0.9516841173171997,
      "learning_rate": 2.916625284355104e-06,
      "loss": 0.4937589168548584,
      "lr/ungrouped": 2.905469120773835e-06,
      "step": 2960
    },
    {
      "epoch": 0.8555071472293234,
      "grad_norm": 1.1842330694198608,
      "learning_rate": 2.805967590616379e-06,
      "loss": 0.46395196914672854,
      "lr/ungrouped": 2.7950124870824964e-06,
      "step": 2970
    },
    {
      "epoch": 0.85838764267454,
      "grad_norm": 1.4125162363052368,
      "learning_rate": 2.6973249884236495e-06,
      "loss": 0.5278787612915039,
      "lr/ungrouped": 2.6865719395980732e-06,
      "step": 2980
    },
    {
      "epoch": 0.8612681381197566,
      "grad_norm": 1.0791908502578735,
      "learning_rate": 2.590707341905488e-06,
      "loss": 0.531835412979126,
      "lr/ungrouped": 2.5801573241037184e-06,
      "step": 2990
    },
    {
      "epoch": 0.8641486335649732,
      "grad_norm": 0.9116361141204834,
      "learning_rate": 2.4861243313360566e-06,
      "loss": 0.49050092697143555,
      "lr/ungrouped": 2.475778302439524e-06,
      "step": 3000
    },
    {
      "epoch": 0.8670291290101898,
      "grad_norm": 1.0779783725738525,
      "learning_rate": 2.383585452256146e-06,
      "loss": 0.49513607025146483,
      "lr/ungrouped": 2.3734443516252645e-06,
      "step": 3010
    },
    {
      "epoch": 0.8699096244554063,
      "grad_norm": 1.047575831413269,
      "learning_rate": 2.283100014611103e-06,
      "loss": 0.48196020126342776,
      "lr/ungrouped": 2.2731647629999575e-06,
      "step": 3020
    },
    {
      "epoch": 0.8727901199006229,
      "grad_norm": 1.3714956045150757,
      "learning_rate": 2.184677141905467e-06,
      "loss": 0.5195942401885987,
      "lr/ungrouped": 2.1749486413782437e-06,
      "step": 3030
    },
    {
      "epoch": 0.8756706153458395,
      "grad_norm": 1.0408002138137817,
      "learning_rate": 2.0883257703746855e-06,
      "loss": 0.5287550926208496,
      "lr/ungrouped": 2.078804904223733e-06,
      "step": 3040
    },
    {
      "epoch": 0.8785511107910561,
      "grad_norm": 0.9607561230659485,
      "learning_rate": 1.9940546481736933e-06,
      "loss": 0.5245569705963135,
      "lr/ungrouped": 1.984742280839347e-06,
      "step": 3050
    },
    {
      "epoch": 0.8814316062362726,
      "grad_norm": 0.8116582632064819,
      "learning_rate": 1.9018723345826411e-06,
      "loss": 0.4748085975646973,
      "lr/ungrouped": 1.8927693115747464e-06,
      "step": 3060
    },
    {
      "epoch": 0.8843121016814892,
      "grad_norm": 1.112561583518982,
      "learning_rate": 1.8117871992298085e-06,
      "loss": 0.5299274921417236,
      "lr/ungrouped": 1.802894347050918e-06,
      "step": 3070
    },
    {
      "epoch": 0.8871925971267058,
      "grad_norm": 1.0939245223999023,
      "learning_rate": 1.7238074213316107e-06,
      "loss": 0.4996906280517578,
      "lr/ungrouped": 1.7151255474019857e-06,
      "step": 3080
    },
    {
      "epoch": 0.8900730925719224,
      "grad_norm": 1.239974021911621,
      "learning_rate": 1.6379409889500513e-06,
      "loss": 0.49614405632019043,
      "lr/ungrouped": 1.6294708815343174e-06,
      "step": 3090
    },
    {
      "epoch": 0.892953588017139,
      "grad_norm": 1.1982766389846802,
      "learning_rate": 1.554195698267391e-06,
      "loss": 0.5115400314331054,
      "lr/ungrouped": 1.5459381264029904e-06,
      "step": 3100
    },
    {
      "epoch": 0.8958340834623555,
      "grad_norm": 1.0616167783737183,
      "learning_rate": 1.4725791528783396e-06,
      "loss": 0.4672398090362549,
      "lr/ungrouped": 1.4645348663056918e-06,
      "step": 3110
    },
    {
      "epoch": 0.8987145789075721,
      "grad_norm": 1.0786339044570923,
      "learning_rate": 1.3930987630996677e-06,
      "loss": 0.5120093822479248,
      "lr/ungrouped": 1.385268492194111e-06,
      "step": 3120
    },
    {
      "epoch": 0.9015950743527886,
      "grad_norm": 1.0152426958084106,
      "learning_rate": 1.315761745297403e-06,
      "loss": 0.5579086780548096,
      "lr/ungrouped": 1.3081462010028717e-06,
      "step": 3130
    },
    {
      "epoch": 0.9044755697980053,
      "grad_norm": 1.384515404701233,
      "learning_rate": 1.2405751212316297e-06,
      "loss": 0.4666877269744873,
      "lr/ungrouped": 1.2331749949960974e-06,
      "step": 3140
    },
    {
      "epoch": 0.9073560652432219,
      "grad_norm": 1.0240039825439453,
      "learning_rate": 1.1675457174189302e-06,
      "loss": 0.5212295055389404,
      "lr/ungrouped": 1.1603616811316636e-06,
      "step": 3150
    },
    {
      "epoch": 0.9102365606884384,
      "grad_norm": 1.2126448154449463,
      "learning_rate": 1.0966801645126202e-06,
      "loss": 0.5673628807067871,
      "lr/ungrouped": 1.0897128704431242e-06,
      "step": 3160
    },
    {
      "epoch": 0.913117056133655,
      "grad_norm": 1.1971455812454224,
      "learning_rate": 1.0279848967006623e-06,
      "loss": 0.6138933658599853,
      "lr/ungrouped": 1.021234977439503e-06,
      "step": 3170
    },
    {
      "epoch": 0.9159975515788715,
      "grad_norm": 1.2284088134765625,
      "learning_rate": 9.614661511215456e-07,
      "loss": 0.48476076126098633,
      "lr/ungrouped": 9.54934219522885e-07,
      "step": 3180
    },
    {
      "epoch": 0.9188780470240882,
      "grad_norm": 0.91191565990448,
      "learning_rate": 8.9712996729793e-07,
      "loss": 0.4882956027984619,
      "lr/ungrouped": 8.908166164239035e-07,
      "step": 3190
    },
    {
      "epoch": 0.9217585424693048,
      "grad_norm": 0.966507613658905,
      "learning_rate": 8.349821865883317e-07,
      "loss": 0.5204824447631836,
      "lr/ungrouped": 8.288879896551832e-07,
      "step": 3200
    },
    {
      "epoch": 0.9246390379145213,
      "grad_norm": 1.2123650312423706,
      "learning_rate": 7.750284516567452e-07,
      "loss": 0.4546231269836426,
      "lr/ungrouped": 7.691539619827881e-07,
      "step": 3210
    },
    {
      "epoch": 0.9275195333597379,
      "grad_norm": 1.1479333639144897,
      "learning_rate": 7.17274205960311e-07,
      "loss": 0.4170845031738281,
      "lr/ungrouped": 7.1161995691571e-07,
      "step": 3220
    },
    {
      "epoch": 0.9304000288049544,
      "grad_norm": 1.034443974494934,
      "learning_rate": 6.617246932551113e-07,
      "loss": 0.5262176513671875,
      "lr/ungrouped": 6.562911982134212e-07,
      "step": 3230
    },
    {
      "epoch": 0.933280524250171,
      "grad_norm": 1.176424264907837,
      "learning_rate": 6.083849571200368e-07,
      "loss": 0.5241230487823486,
      "lr/ungrouped": 6.031727094116175e-07,
      "step": 3240
    },
    {
      "epoch": 0.9361610196953876,
      "grad_norm": 0.979678213596344,
      "learning_rate": 5.572598404988749e-07,
      "loss": 0.4910717010498047,
      "lr/ungrouped": 5.522693133660856e-07,
      "step": 3250
    },
    {
      "epoch": 0.9390415151406042,
      "grad_norm": 1.5239194631576538,
      "learning_rate": 5.083539852605929e-07,
      "loss": 0.5448841094970703,
      "lr/ungrouped": 5.035856318148291e-07,
      "step": 3260
    },
    {
      "epoch": 0.9419220105858208,
      "grad_norm": 1.359035849571228,
      "learning_rate": 4.616718317778823e-07,
      "loss": 0.4855171203613281,
      "lr/ungrouped": 4.571260849584397e-07,
      "step": 3270
    },
    {
      "epoch": 0.9448025060310373,
      "grad_norm": 1.365057110786438,
      "learning_rate": 4.1721761852400175e-07,
      "loss": 0.42261877059936526,
      "lr/ungrouped": 4.128948910587599e-07,
      "step": 3280
    },
    {
      "epoch": 0.947683001476254,
      "grad_norm": 1.3075063228607178,
      "learning_rate": 3.749953816879398e-07,
      "loss": 0.4199540138244629,
      "lr/ungrouped": 3.708960660559008e-07,
      "step": 3290
    },
    {
      "epoch": 0.9505634969214705,
      "grad_norm": 1.2111417055130005,
      "learning_rate": 3.3500895480796657e-07,
      "loss": 0.5039636611938476,
      "lr/ungrouped": 3.3113342320360285e-07,
      "step": 3300
    },
    {
      "epoch": 0.9534439923666871,
      "grad_norm": 1.030943751335144,
      "learning_rate": 2.972619684235539e-07,
      "loss": 0.5595455646514893,
      "lr/ungrouped": 2.936105727230298e-07,
      "step": 3310
    },
    {
      "epoch": 0.9563244878119036,
      "grad_norm": 1.2582727670669556,
      "learning_rate": 2.617578497457529e-07,
      "loss": 0.5251143932342529,
      "lr/ungrouped": 2.583309214749696e-07,
      "step": 3320
    },
    {
      "epoch": 0.9592049832571202,
      "grad_norm": 1.4242674112319946,
      "learning_rate": 2.284998223460205e-07,
      "loss": 0.4252504348754883,
      "lr/ungrouped": 2.2529767265051793e-07,
      "step": 3330
    },
    {
      "epoch": 0.9620854787023368,
      "grad_norm": 0.9866174459457397,
      "learning_rate": 1.974909058635399e-07,
      "loss": 0.4454331398010254,
      "lr/ungrouped": 1.94513825480247e-07,
      "step": 3340
    },
    {
      "epoch": 0.9649659741475534,
      "grad_norm": 1.1806648969650269,
      "learning_rate": 1.687339157310508e-07,
      "loss": 0.4342671871185303,
      "lr/ungrouped": 1.6598217496187895e-07,
      "step": 3350
    },
    {
      "epoch": 0.96784646959277,
      "grad_norm": 1.094575047492981,
      "learning_rate": 1.4223146291922062e-07,
      "loss": 0.4138617515563965,
      "lr/ungrouped": 1.3970531160654167e-07,
      "step": 3360
    },
    {
      "epoch": 0.9707269650379865,
      "grad_norm": 1.0167882442474365,
      "learning_rate": 1.179859536995953e-07,
      "loss": 0.5147321224212646,
      "lr/ungrouped": 1.1568562120353455e-07,
      "step": 3370
    },
    {
      "epoch": 0.9736074604832031,
      "grad_norm": 1.0953649282455444,
      "learning_rate": 9.599958942611287e-08,
      "loss": 0.4610922813415527,
      "lr/ungrouped": 9.392528460374362e-08,
      "step": 3380
    },
    {
      "epoch": 0.9764879559284196,
      "grad_norm": 1.111059308052063,
      "learning_rate": 7.627436633523832e-08,
      "loss": 0.4920982360839844,
      "lr/ungrouped": 7.442627752160258e-08,
      "step": 3390
    },
    {
      "epoch": 0.9793684513736363,
      "grad_norm": 0.9982632994651794,
      "learning_rate": 5.881207536471145e-08,
      "loss": 0.5243843555450439,
      "lr/ungrouped": 5.7190370355733605e-08,
      "step": 3400
    },
    {
      "epoch": 0.9822489468188529,
      "grad_norm": 1.0663832426071167,
      "learning_rate": 4.3614301990946296e-08,
      "loss": 0.47223596572875975,
      "lr/ungrouped": 4.22191280281925e-08,
      "step": 3410
    },
    {
      "epoch": 0.9851294422640694,
      "grad_norm": 1.0918127298355103,
      "learning_rate": 3.0682426085079655e-08,
      "loss": 0.5449328422546387,
      "lr/ungrouped": 2.9513909842382448e-08,
      "step": 3420
    },
    {
      "epoch": 0.988009937709286,
      "grad_norm": 1.0994895696640015,
      "learning_rate": 2.0017621787671303e-08,
      "loss": 0.5656056880950928,
      "lr/ungrouped": 1.9075869359647115e-08,
      "step": 3430
    },
    {
      "epoch": 0.9908904331545025,
      "grad_norm": 0.9849516749382019,
      "learning_rate": 1.1620857402125374e-08,
      "loss": 0.4636025905609131,
      "lr/ungrouped": 1.090595429451835e-08,
      "step": 3440
    },
    {
      "epoch": 0.9937709285997192,
      "grad_norm": 1.1691488027572632,
      "learning_rate": 5.4928953067495685e-09,
      "loss": 0.4662496566772461,
      "lr/ungrouped": 5.004906428685008e-09,
      "step": 3450
    },
    {
      "epoch": 0.9966514240449357,
      "grad_norm": 1.2294414043426514,
      "learning_rate": 1.6342918855494216e-09,
      "loss": 0.5150485038757324,
      "lr/ungrouped": 1.373261543632931e-09,
      "step": 3460
    },
    {
      "epoch": 0.9995319194901523,
      "grad_norm": 0.9555907249450684,
      "learning_rate": 4.539747770204716e-11,
      "loss": 0.46833095550537107,
      "lr/ungrouped": 1.134937200053532e-11,
      "step": 3470
    },
    {
      "epoch": 0.999819969034674,
      "eval_loss": 0.32731592655181885,
      "eval_runtime": 148.1143,
      "eval_samples_per_second": 20.829,
      "eval_steps_per_second": 20.829,
      "step": 3471
    }
  ],
  "logging_steps": 10,
  "max_steps": 3471,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.967823583130419e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}