{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.998505231689088,
  "eval_steps": 500,
  "global_step": 2090,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01195814648729447,
      "grad_norm": 8.200420379638672,
      "learning_rate": 4.999929391798332e-05,
      "loss": 1.6433,
      "step": 5
    },
    {
      "epoch": 0.02391629297458894,
      "grad_norm": 1.7042728662490845,
      "learning_rate": 4.999717571181742e-05,
      "loss": 0.3662,
      "step": 10
    },
    {
      "epoch": 0.03587443946188341,
      "grad_norm": 1.3026138544082642,
      "learning_rate": 4.999364550115248e-05,
      "loss": 0.2349,
      "step": 15
    },
    {
      "epoch": 0.04783258594917788,
      "grad_norm": 0.6563483476638794,
      "learning_rate": 4.998870348539797e-05,
      "loss": 0.1672,
      "step": 20
    },
    {
      "epoch": 0.059790732436472344,
      "grad_norm": 0.7150379419326782,
      "learning_rate": 4.998234994371135e-05,
      "loss": 0.147,
      "step": 25
    },
    {
      "epoch": 0.07174887892376682,
      "grad_norm": 0.6974969506263733,
      "learning_rate": 4.997458523498236e-05,
      "loss": 0.1399,
      "step": 30
    },
    {
      "epoch": 0.08370702541106129,
      "grad_norm": 0.7378151416778564,
      "learning_rate": 4.996540979781269e-05,
      "loss": 0.1002,
      "step": 35
    },
    {
      "epoch": 0.09566517189835576,
      "grad_norm": 0.6553875803947449,
      "learning_rate": 4.995482415049123e-05,
      "loss": 0.1146,
      "step": 40
    },
    {
      "epoch": 0.10762331838565023,
      "grad_norm": 0.4195781350135803,
      "learning_rate": 4.99428288909648e-05,
      "loss": 0.0978,
      "step": 45
    },
    {
      "epoch": 0.11958146487294469,
      "grad_norm": 0.886047899723053,
      "learning_rate": 4.992942469680436e-05,
      "loss": 0.0925,
      "step": 50
    },
    {
      "epoch": 0.13153961136023917,
      "grad_norm": 0.4746655821800232,
      "learning_rate": 4.991461232516675e-05,
      "loss": 0.0904,
      "step": 55
    },
    {
      "epoch": 0.14349775784753363,
      "grad_norm": 0.8347558379173279,
      "learning_rate": 4.989839261275191e-05,
      "loss": 0.1096,
      "step": 60
    },
    {
      "epoch": 0.1554559043348281,
      "grad_norm": 0.4383985698223114,
      "learning_rate": 4.988076647575562e-05,
      "loss": 0.0872,
      "step": 65
    },
    {
      "epoch": 0.16741405082212257,
      "grad_norm": 0.5418385863304138,
      "learning_rate": 4.986173490981773e-05,
      "loss": 0.092,
      "step": 70
    },
    {
      "epoch": 0.17937219730941703,
      "grad_norm": 0.679784893989563,
      "learning_rate": 4.9841298989965984e-05,
      "loss": 0.1006,
      "step": 75
    },
    {
      "epoch": 0.19133034379671152,
      "grad_norm": 0.6306747794151306,
      "learning_rate": 4.981945987055521e-05,
      "loss": 0.0918,
      "step": 80
    },
    {
      "epoch": 0.20328849028400597,
      "grad_norm": 0.8185198307037354,
      "learning_rate": 4.979621878520216e-05,
      "loss": 0.0767,
      "step": 85
    },
    {
      "epoch": 0.21524663677130046,
      "grad_norm": 1.059411644935608,
      "learning_rate": 4.9771577046715846e-05,
      "loss": 0.0953,
      "step": 90
    },
    {
      "epoch": 0.22720478325859492,
      "grad_norm": 0.6877408027648926,
      "learning_rate": 4.9745536047023324e-05,
      "loss": 0.0775,
      "step": 95
    },
    {
      "epoch": 0.23916292974588937,
      "grad_norm": 0.5024028420448303,
      "learning_rate": 4.971809725709112e-05,
      "loss": 0.0703,
      "step": 100
    },
    {
      "epoch": 0.25112107623318386,
      "grad_norm": 0.5661141872406006,
      "learning_rate": 4.968926222684213e-05,
      "loss": 0.0774,
      "step": 105
    },
    {
      "epoch": 0.26307922272047835,
      "grad_norm": 0.6358644366264343,
      "learning_rate": 4.965903258506806e-05,
      "loss": 0.0851,
      "step": 110
    },
    {
      "epoch": 0.2750373692077728,
      "grad_norm": 1.1677368879318237,
      "learning_rate": 4.962741003933742e-05,
      "loss": 0.0833,
      "step": 115
    },
    {
      "epoch": 0.28699551569506726,
      "grad_norm": 0.6264662146568298,
      "learning_rate": 4.959439637589909e-05,
      "loss": 0.0793,
      "step": 120
    },
    {
      "epoch": 0.29895366218236175,
      "grad_norm": 0.5868094563484192,
      "learning_rate": 4.9559993459581375e-05,
      "loss": 0.0658,
      "step": 125
    },
    {
      "epoch": 0.3109118086696562,
      "grad_norm": 0.5872901678085327,
      "learning_rate": 4.952420323368673e-05,
      "loss": 0.0737,
      "step": 130
    },
    {
      "epoch": 0.32286995515695066,
      "grad_norm": 0.6091463565826416,
      "learning_rate": 4.948702771988195e-05,
      "loss": 0.0644,
      "step": 135
    },
    {
      "epoch": 0.33482810164424515,
      "grad_norm": 0.6292822360992432,
      "learning_rate": 4.9448469018083965e-05,
      "loss": 0.0554,
      "step": 140
    },
    {
      "epoch": 0.34678624813153963,
      "grad_norm": 0.9520431756973267,
      "learning_rate": 4.9408529306341255e-05,
      "loss": 0.0888,
      "step": 145
    },
    {
      "epoch": 0.35874439461883406,
      "grad_norm": 0.673231840133667,
      "learning_rate": 4.936721084071079e-05,
      "loss": 0.0657,
      "step": 150
    },
    {
      "epoch": 0.37070254110612855,
      "grad_norm": 0.6284344792366028,
      "learning_rate": 4.932451595513062e-05,
      "loss": 0.0626,
      "step": 155
    },
    {
      "epoch": 0.38266068759342303,
      "grad_norm": 1.076338291168213,
      "learning_rate": 4.928044706128803e-05,
      "loss": 0.0721,
      "step": 160
    },
    {
      "epoch": 0.39461883408071746,
      "grad_norm": 0.4863973557949066,
      "learning_rate": 4.923500664848326e-05,
      "loss": 0.0568,
      "step": 165
    },
    {
      "epoch": 0.40657698056801195,
      "grad_norm": 0.8602327704429626,
      "learning_rate": 4.9188197283489015e-05,
      "loss": 0.0659,
      "step": 170
    },
    {
      "epoch": 0.41853512705530643,
      "grad_norm": 0.4428125321865082,
      "learning_rate": 4.9140021610405326e-05,
      "loss": 0.0587,
      "step": 175
    },
    {
      "epoch": 0.4304932735426009,
      "grad_norm": 0.6152353286743164,
      "learning_rate": 4.9090482350510336e-05,
      "loss": 0.079,
      "step": 180
    },
    {
      "epoch": 0.44245142002989535,
      "grad_norm": 0.6276563405990601,
      "learning_rate": 4.9039582302106465e-05,
      "loss": 0.0588,
      "step": 185
    },
    {
      "epoch": 0.45440956651718983,
      "grad_norm": 0.705228328704834,
      "learning_rate": 4.898732434036244e-05,
      "loss": 0.0563,
      "step": 190
    },
    {
      "epoch": 0.4663677130044843,
      "grad_norm": 1.0965617895126343,
      "learning_rate": 4.89337114171508e-05,
      "loss": 0.08,
      "step": 195
    },
    {
      "epoch": 0.47832585949177875,
      "grad_norm": 0.7665426731109619,
      "learning_rate": 4.887874656088124e-05,
      "loss": 0.0667,
      "step": 200
    },
    {
      "epoch": 0.49028400597907323,
      "grad_norm": 0.6774755120277405,
      "learning_rate": 4.882243287632947e-05,
      "loss": 0.0671,
      "step": 205
    },
    {
      "epoch": 0.5022421524663677,
      "grad_norm": 0.5077477693557739,
      "learning_rate": 4.8764773544461886e-05,
      "loss": 0.0698,
      "step": 210
    },
    {
      "epoch": 0.5142002989536621,
      "grad_norm": 0.4905177354812622,
      "learning_rate": 4.870577182225589e-05,
      "loss": 0.0623,
      "step": 215
    },
    {
      "epoch": 0.5261584454409567,
      "grad_norm": 0.6313666701316833,
      "learning_rate": 4.864543104251587e-05,
      "loss": 0.0783,
      "step": 220
    },
    {
      "epoch": 0.5381165919282511,
      "grad_norm": 0.4347245991230011,
      "learning_rate": 4.858375461368499e-05,
      "loss": 0.0601,
      "step": 225
    },
    {
      "epoch": 0.5500747384155455,
      "grad_norm": 0.4309779405593872,
      "learning_rate": 4.8520746019652605e-05,
      "loss": 0.0621,
      "step": 230
    },
    {
      "epoch": 0.5620328849028401,
      "grad_norm": 0.7200354337692261,
      "learning_rate": 4.8456408819557564e-05,
      "loss": 0.0633,
      "step": 235
    },
    {
      "epoch": 0.5739910313901345,
      "grad_norm": 0.6128847599029541,
      "learning_rate": 4.839074664758704e-05,
      "loss": 0.053,
      "step": 240
    },
    {
      "epoch": 0.585949177877429,
      "grad_norm": 0.32561948895454407,
      "learning_rate": 4.8323763212771354e-05,
      "loss": 0.0523,
      "step": 245
    },
    {
      "epoch": 0.5979073243647235,
      "grad_norm": 0.5633959770202637,
      "learning_rate": 4.825546229877439e-05,
      "loss": 0.0575,
      "step": 250
    },
    {
      "epoch": 0.6098654708520179,
      "grad_norm": 0.6730295419692993,
      "learning_rate": 4.818584776367993e-05,
      "loss": 0.0616,
      "step": 255
    },
    {
      "epoch": 0.6218236173393124,
      "grad_norm": 0.7559422254562378,
      "learning_rate": 4.811492353977366e-05,
      "loss": 0.0642,
      "step": 260
    },
    {
      "epoch": 0.6337817638266069,
      "grad_norm": 0.7299085855484009,
      "learning_rate": 4.804269363332112e-05,
      "loss": 0.062,
      "step": 265
    },
    {
      "epoch": 0.6457399103139013,
      "grad_norm": 0.5151236653327942,
      "learning_rate": 4.796916212434135e-05,
      "loss": 0.0635,
      "step": 270
    },
    {
      "epoch": 0.6576980568011959,
      "grad_norm": 0.7765339612960815,
      "learning_rate": 4.789433316637644e-05,
      "loss": 0.0727,
      "step": 275
    },
    {
      "epoch": 0.6696562032884903,
      "grad_norm": 0.4541449546813965,
      "learning_rate": 4.78182109862569e-05,
      "loss": 0.0539,
      "step": 280
    },
    {
      "epoch": 0.6816143497757847,
      "grad_norm": 0.4902482330799103,
      "learning_rate": 4.774079988386296e-05,
      "loss": 0.0635,
      "step": 285
    },
    {
      "epoch": 0.6935724962630793,
      "grad_norm": 0.563753068447113,
      "learning_rate": 4.7662104231881574e-05,
      "loss": 0.0504,
      "step": 290
    },
    {
      "epoch": 0.7055306427503737,
      "grad_norm": 0.6252387762069702,
      "learning_rate": 4.758212847555953e-05,
      "loss": 0.0499,
      "step": 295
    },
    {
      "epoch": 0.7174887892376681,
      "grad_norm": 0.3774847984313965,
      "learning_rate": 4.750087713245227e-05,
      "loss": 0.0527,
      "step": 300
    },
    {
      "epoch": 0.7294469357249627,
      "grad_norm": 1.1581354141235352,
      "learning_rate": 4.7418354792168794e-05,
      "loss": 0.0654,
      "step": 305
    },
    {
      "epoch": 0.7414050822122571,
      "grad_norm": 0.7672606706619263,
      "learning_rate": 4.733456611611233e-05,
      "loss": 0.0697,
      "step": 310
    },
    {
      "epoch": 0.7533632286995515,
      "grad_norm": 0.5647212266921997,
      "learning_rate": 4.724951583721707e-05,
      "loss": 0.0493,
      "step": 315
    },
    {
      "epoch": 0.7653213751868461,
      "grad_norm": 0.49768275022506714,
      "learning_rate": 4.716320875968081e-05,
      "loss": 0.0479,
      "step": 320
    },
    {
      "epoch": 0.7772795216741405,
      "grad_norm": 0.882534921169281,
      "learning_rate": 4.7075649758693565e-05,
      "loss": 0.0592,
      "step": 325
    },
    {
      "epoch": 0.7892376681614349,
      "grad_norm": 0.6543036103248596,
      "learning_rate": 4.698684378016222e-05,
      "loss": 0.0563,
      "step": 330
    },
    {
      "epoch": 0.8011958146487295,
      "grad_norm": 0.49713218212127686,
      "learning_rate": 4.689679584043115e-05,
      "loss": 0.0522,
      "step": 335
    },
    {
      "epoch": 0.8131539611360239,
      "grad_norm": 0.8511777520179749,
      "learning_rate": 4.68055110259988e-05,
      "loss": 0.083,
      "step": 340
    },
    {
      "epoch": 0.8251121076233184,
      "grad_norm": 0.4030969440937042,
      "learning_rate": 4.671299449323045e-05,
      "loss": 0.053,
      "step": 345
    },
    {
      "epoch": 0.8370702541106129,
      "grad_norm": 0.5079150795936584,
      "learning_rate": 4.66192514680669e-05,
      "loss": 0.0621,
      "step": 350
    },
    {
      "epoch": 0.8490284005979073,
      "grad_norm": 0.9145223498344421,
      "learning_rate": 4.6524287245729295e-05,
      "loss": 0.0514,
      "step": 355
    },
    {
      "epoch": 0.8609865470852018,
      "grad_norm": 0.852425217628479,
      "learning_rate": 4.642810719041999e-05,
      "loss": 0.0712,
      "step": 360
    },
    {
      "epoch": 0.8729446935724963,
      "grad_norm": 0.6567260026931763,
      "learning_rate": 4.63307167350196e-05,
      "loss": 0.0671,
      "step": 365
    },
    {
      "epoch": 0.8849028400597907,
      "grad_norm": 0.4592411518096924,
      "learning_rate": 4.6232121380780034e-05,
      "loss": 0.0616,
      "step": 370
    },
    {
      "epoch": 0.8968609865470852,
      "grad_norm": 0.7522163987159729,
      "learning_rate": 4.613232669701384e-05,
      "loss": 0.0613,
      "step": 375
    },
    {
      "epoch": 0.9088191330343797,
      "grad_norm": 0.6464064121246338,
      "learning_rate": 4.6031338320779534e-05,
      "loss": 0.0596,
      "step": 380
    },
    {
      "epoch": 0.9207772795216741,
      "grad_norm": 0.4108300507068634,
      "learning_rate": 4.592916195656322e-05,
      "loss": 0.0365,
      "step": 385
    },
    {
      "epoch": 0.9327354260089686,
      "grad_norm": 0.6339195966720581,
      "learning_rate": 4.582580337595636e-05,
      "loss": 0.0664,
      "step": 390
    },
    {
      "epoch": 0.9446935724962631,
      "grad_norm": 0.8676311373710632,
      "learning_rate": 4.572126841732976e-05,
      "loss": 0.0623,
      "step": 395
    },
    {
      "epoch": 0.9566517189835575,
      "grad_norm": 0.5879262685775757,
      "learning_rate": 4.561556298550379e-05,
      "loss": 0.0642,
      "step": 400
    },
    {
      "epoch": 0.968609865470852,
      "grad_norm": 0.6106139421463013,
      "learning_rate": 4.550869305141478e-05,
      "loss": 0.054,
      "step": 405
    },
    {
      "epoch": 0.9805680119581465,
      "grad_norm": 0.8377204537391663,
      "learning_rate": 4.5400664651777835e-05,
      "loss": 0.0641,
      "step": 410
    },
    {
      "epoch": 0.992526158445441,
      "grad_norm": 0.4505299925804138,
      "learning_rate": 4.529148388874577e-05,
      "loss": 0.0749,
      "step": 415
    },
    {
      "epoch": 1.0044843049327354,
      "grad_norm": 0.2847709059715271,
      "learning_rate": 4.518115692956445e-05,
      "loss": 0.0545,
      "step": 420
    },
    {
      "epoch": 1.0164424514200299,
      "grad_norm": 0.6187481880187988,
      "learning_rate": 4.5069690006224424e-05,
      "loss": 0.0531,
      "step": 425
    },
    {
      "epoch": 1.0284005979073243,
      "grad_norm": 0.49963054060935974,
      "learning_rate": 4.49570894151089e-05,
      "loss": 0.0453,
      "step": 430
    },
    {
      "epoch": 1.0403587443946187,
      "grad_norm": 0.48253875970840454,
      "learning_rate": 4.4843361516638075e-05,
      "loss": 0.0596,
      "step": 435
    },
    {
      "epoch": 1.0523168908819134,
      "grad_norm": 0.584597647190094,
      "learning_rate": 4.4728512734909844e-05,
      "loss": 0.038,
      "step": 440
    },
    {
      "epoch": 1.0642750373692078,
      "grad_norm": 0.8537632822990417,
      "learning_rate": 4.4612549557336974e-05,
      "loss": 0.051,
      "step": 445
    },
    {
      "epoch": 1.0762331838565022,
      "grad_norm": 0.46478036046028137,
      "learning_rate": 4.449547853428061e-05,
      "loss": 0.0495,
      "step": 450
    },
    {
      "epoch": 1.0881913303437967,
      "grad_norm": 0.558437705039978,
      "learning_rate": 4.437730627868027e-05,
      "loss": 0.0473,
      "step": 455
    },
    {
      "epoch": 1.100149476831091,
      "grad_norm": 0.6500675082206726,
      "learning_rate": 4.4258039465680326e-05,
      "loss": 0.0627,
      "step": 460
    },
    {
      "epoch": 1.1121076233183858,
      "grad_norm": 0.5839610695838928,
      "learning_rate": 4.413768483225292e-05,
      "loss": 0.0407,
      "step": 465
    },
    {
      "epoch": 1.1240657698056802,
      "grad_norm": 0.46521812677383423,
      "learning_rate": 4.4016249176817424e-05,
      "loss": 0.0515,
      "step": 470
    },
    {
      "epoch": 1.1360239162929746,
      "grad_norm": 0.7125497460365295,
      "learning_rate": 4.389373935885646e-05,
      "loss": 0.0537,
      "step": 475
    },
    {
      "epoch": 1.147982062780269,
      "grad_norm": 0.569307267665863,
      "learning_rate": 4.377016229852836e-05,
      "loss": 0.0366,
      "step": 480
    },
    {
      "epoch": 1.1599402092675635,
      "grad_norm": 0.476139098405838,
      "learning_rate": 4.364552497627632e-05,
      "loss": 0.0558,
      "step": 485
    },
    {
      "epoch": 1.171898355754858,
      "grad_norm": 0.6692591309547424,
      "learning_rate": 4.351983443243409e-05,
      "loss": 0.0427,
      "step": 490
    },
    {
      "epoch": 1.1838565022421526,
      "grad_norm": 0.6070113182067871,
      "learning_rate": 4.3393097766828293e-05,
      "loss": 0.0366,
      "step": 495
    },
    {
      "epoch": 1.195814648729447,
      "grad_norm": 0.6990951299667358,
      "learning_rate": 4.3265322138377354e-05,
      "loss": 0.0407,
      "step": 500
    },
    {
      "epoch": 1.2077727952167414,
      "grad_norm": 0.8403803706169128,
      "learning_rate": 4.313651476468715e-05,
      "loss": 0.0414,
      "step": 505
    },
    {
      "epoch": 1.2197309417040358,
      "grad_norm": 0.43457669019699097,
      "learning_rate": 4.3006682921643296e-05,
      "loss": 0.0443,
      "step": 510
    },
    {
      "epoch": 1.2316890881913303,
      "grad_norm": 0.8741814494132996,
      "learning_rate": 4.287583394300015e-05,
      "loss": 0.0479,
      "step": 515
    },
    {
      "epoch": 1.2436472346786247,
      "grad_norm": 0.8148631453514099,
      "learning_rate": 4.274397521996658e-05,
      "loss": 0.0645,
      "step": 520
    },
    {
      "epoch": 1.2556053811659194,
      "grad_norm": 0.7766031622886658,
      "learning_rate": 4.261111420078843e-05,
      "loss": 0.0423,
      "step": 525
    },
    {
      "epoch": 1.2675635276532138,
      "grad_norm": 0.7706691026687622,
      "learning_rate": 4.247725839032781e-05,
      "loss": 0.0517,
      "step": 530
    },
    {
      "epoch": 1.2795216741405082,
      "grad_norm": 0.438267320394516,
      "learning_rate": 4.234241534963916e-05,
      "loss": 0.0483,
      "step": 535
    },
    {
      "epoch": 1.2914798206278026,
      "grad_norm": 0.6633365154266357,
      "learning_rate": 4.220659269554217e-05,
      "loss": 0.0438,
      "step": 540
    },
    {
      "epoch": 1.303437967115097,
      "grad_norm": 0.42005932331085205,
      "learning_rate": 4.2069798100191525e-05,
      "loss": 0.0522,
      "step": 545
    },
    {
      "epoch": 1.3153961136023917,
      "grad_norm": 0.4571685194969177,
      "learning_rate": 4.193203929064353e-05,
      "loss": 0.0457,
      "step": 550
    },
    {
      "epoch": 1.3273542600896862,
      "grad_norm": 0.5415948629379272,
      "learning_rate": 4.1793324048419626e-05,
      "loss": 0.0475,
      "step": 555
    },
    {
      "epoch": 1.3393124065769806,
      "grad_norm": 0.4834306836128235,
      "learning_rate": 4.1653660209066835e-05,
      "loss": 0.0298,
      "step": 560
    },
    {
      "epoch": 1.351270553064275,
      "grad_norm": 0.7590181827545166,
      "learning_rate": 4.1513055661715214e-05,
      "loss": 0.043,
      "step": 565
    },
    {
      "epoch": 1.3632286995515694,
      "grad_norm": 0.8852930068969727,
      "learning_rate": 4.137151834863213e-05,
      "loss": 0.0503,
      "step": 570
    },
    {
      "epoch": 1.375186846038864,
      "grad_norm": 0.6908103227615356,
      "learning_rate": 4.1229056264773705e-05,
      "loss": 0.0494,
      "step": 575
    },
    {
      "epoch": 1.3871449925261585,
      "grad_norm": 0.45172494649887085,
      "learning_rate": 4.108567745733318e-05,
      "loss": 0.0559,
      "step": 580
    },
    {
      "epoch": 1.399103139013453,
      "grad_norm": 0.5985448956489563,
      "learning_rate": 4.094139002528635e-05,
      "loss": 0.0471,
      "step": 585
    },
    {
      "epoch": 1.4110612855007474,
      "grad_norm": 0.6178925037384033,
      "learning_rate": 4.0796202118934105e-05,
      "loss": 0.0622,
      "step": 590
    },
    {
      "epoch": 1.4230194319880418,
      "grad_norm": 0.5416215062141418,
      "learning_rate": 4.065012193944201e-05,
      "loss": 0.0485,
      "step": 595
    },
    {
      "epoch": 1.4349775784753362,
      "grad_norm": 0.5193672776222229,
      "learning_rate": 4.050315773837708e-05,
      "loss": 0.0484,
      "step": 600
    },
    {
      "epoch": 1.4469357249626307,
      "grad_norm": 0.576651394367218,
      "learning_rate": 4.03553178172417e-05,
      "loss": 0.0432,
      "step": 605
    },
    {
      "epoch": 1.4588938714499253,
      "grad_norm": 0.6012284755706787,
      "learning_rate": 4.020661052700461e-05,
      "loss": 0.0384,
      "step": 610
    },
    {
      "epoch": 1.4708520179372198,
      "grad_norm": 0.42638862133026123,
      "learning_rate": 4.005704426762931e-05,
      "loss": 0.0488,
      "step": 615
    },
    {
      "epoch": 1.4828101644245142,
      "grad_norm": 0.7791224122047424,
      "learning_rate": 3.990662748759946e-05,
      "loss": 0.05,
      "step": 620
    },
    {
      "epoch": 1.4947683109118086,
      "grad_norm": 0.4531303942203522,
      "learning_rate": 3.9755368683441735e-05,
      "loss": 0.047,
      "step": 625
    },
    {
      "epoch": 1.506726457399103,
      "grad_norm": 0.5733367800712585,
      "learning_rate": 3.960327639924586e-05,
      "loss": 0.0509,
      "step": 630
    },
    {
      "epoch": 1.5186846038863977,
      "grad_norm": 0.6013754606246948,
      "learning_rate": 3.945035922618197e-05,
      "loss": 0.0477,
      "step": 635
    },
    {
      "epoch": 1.5306427503736921,
      "grad_norm": 0.8139174580574036,
      "learning_rate": 3.9296625802015356e-05,
      "loss": 0.0578,
      "step": 640
    },
    {
      "epoch": 1.5426008968609866,
      "grad_norm": 0.730939507484436,
      "learning_rate": 3.9142084810618495e-05,
      "loss": 0.0547,
      "step": 645
    },
    {
      "epoch": 1.554559043348281,
      "grad_norm": 0.6104177832603455,
      "learning_rate": 3.898674498148058e-05,
      "loss": 0.0455,
      "step": 650
    },
    {
      "epoch": 1.5665171898355754,
      "grad_norm": 0.7068261504173279,
      "learning_rate": 3.883061508921439e-05,
      "loss": 0.038,
      "step": 655
    },
    {
      "epoch": 1.57847533632287,
      "grad_norm": 0.8497127890586853,
      "learning_rate": 3.867370395306068e-05,
      "loss": 0.0538,
      "step": 660
    },
    {
      "epoch": 1.5904334828101643,
      "grad_norm": 0.5805686116218567,
      "learning_rate": 3.851602043638994e-05,
      "loss": 0.0537,
      "step": 665
    },
    {
      "epoch": 1.602391629297459,
      "grad_norm": 0.48102009296417236,
      "learning_rate": 3.8357573446201825e-05,
      "loss": 0.0355,
      "step": 670
    },
    {
      "epoch": 1.6143497757847534,
      "grad_norm": 0.5563987493515015,
      "learning_rate": 3.8198371932621965e-05,
      "loss": 0.0467,
      "step": 675
    },
    {
      "epoch": 1.6263079222720478,
      "grad_norm": 0.3860163986682892,
      "learning_rate": 3.8038424888396416e-05,
      "loss": 0.0443,
      "step": 680
    },
    {
      "epoch": 1.6382660687593424,
      "grad_norm": 0.20400579273700714,
      "learning_rate": 3.78777413483837e-05,
      "loss": 0.0375,
      "step": 685
    },
    {
      "epoch": 1.6502242152466366,
      "grad_norm": 0.43344926834106445,
      "learning_rate": 3.771633038904446e-05,
      "loss": 0.0526,
      "step": 690
    },
    {
      "epoch": 1.6621823617339313,
      "grad_norm": 0.5572287440299988,
      "learning_rate": 3.7554201127928744e-05,
      "loss": 0.0423,
      "step": 695
    },
    {
      "epoch": 1.6741405082212257,
      "grad_norm": 0.799628734588623,
      "learning_rate": 3.739136272316102e-05,
      "loss": 0.0552,
      "step": 700
    },
    {
      "epoch": 1.6860986547085202,
      "grad_norm": 0.6148493885993958,
      "learning_rate": 3.72278243729228e-05,
      "loss": 0.0369,
      "step": 705
    },
    {
      "epoch": 1.6980568011958148,
      "grad_norm": 0.6144611239433289,
      "learning_rate": 3.706359531493316e-05,
      "loss": 0.0468,
      "step": 710
    },
    {
      "epoch": 1.710014947683109,
      "grad_norm": 0.5393456220626831,
      "learning_rate": 3.689868482592684e-05,
      "loss": 0.0476,
      "step": 715
    },
    {
      "epoch": 1.7219730941704037,
      "grad_norm": 1.1401357650756836,
      "learning_rate": 3.67331022211303e-05,
      "loss": 0.0412,
      "step": 720
    },
    {
      "epoch": 1.733931240657698,
      "grad_norm": 0.4808436930179596,
      "learning_rate": 3.6566856853735516e-05,
      "loss": 0.05,
      "step": 725
    },
    {
      "epoch": 1.7458893871449925,
      "grad_norm": 0.37553855776786804,
      "learning_rate": 3.6399958114371595e-05,
      "loss": 0.0351,
      "step": 730
    },
    {
      "epoch": 1.757847533632287,
      "grad_norm": 0.41987690329551697,
      "learning_rate": 3.623241543057444e-05,
      "loss": 0.0468,
      "step": 735
    },
    {
      "epoch": 1.7698056801195814,
      "grad_norm": 0.6836597323417664,
      "learning_rate": 3.6064238266254145e-05,
      "loss": 0.0505,
      "step": 740
    },
    {
      "epoch": 1.781763826606876,
      "grad_norm": 1.0485986471176147,
      "learning_rate": 3.5895436121160386e-05,
      "loss": 0.0555,
      "step": 745
    },
    {
      "epoch": 1.7937219730941703,
      "grad_norm": 0.6411017775535583,
      "learning_rate": 3.5726018530345915e-05,
      "loss": 0.0477,
      "step": 750
    },
    {
      "epoch": 1.805680119581465,
      "grad_norm": 0.8012114763259888,
      "learning_rate": 3.555599506362784e-05,
      "loss": 0.0652,
      "step": 755
    },
    {
      "epoch": 1.8176382660687593,
      "grad_norm": 0.6524213552474976,
      "learning_rate": 3.5385375325047166e-05,
      "loss": 0.0346,
      "step": 760
    },
    {
      "epoch": 1.8295964125560538,
      "grad_norm": 0.5920986533164978,
      "learning_rate": 3.52141689523262e-05,
      "loss": 0.0545,
      "step": 765
    },
    {
      "epoch": 1.8415545590433484,
      "grad_norm": 0.38302087783813477,
      "learning_rate": 3.504238561632424e-05,
      "loss": 0.0274,
      "step": 770
    },
    {
      "epoch": 1.8535127055306426,
      "grad_norm": 0.5103805661201477,
      "learning_rate": 3.487003502049122e-05,
      "loss": 0.0466,
      "step": 775
    },
    {
      "epoch": 1.8654708520179373,
      "grad_norm": 0.2805655002593994,
      "learning_rate": 3.469712690031962e-05,
      "loss": 0.0428,
      "step": 780
    },
    {
      "epoch": 1.8774289985052317,
      "grad_norm": 0.6101044416427612,
      "learning_rate": 3.4523671022794616e-05,
      "loss": 0.0355,
      "step": 785
    },
    {
      "epoch": 1.8893871449925261,
      "grad_norm": 0.442029744386673,
      "learning_rate": 3.4349677185842245e-05,
      "loss": 0.0531,
      "step": 790
    },
    {
      "epoch": 1.9013452914798208,
      "grad_norm": 0.8262781500816345,
      "learning_rate": 3.4175155217776055e-05,
      "loss": 0.0515,
      "step": 795
    },
    {
      "epoch": 1.913303437967115,
      "grad_norm": 0.4850008487701416,
      "learning_rate": 3.4000114976741906e-05,
      "loss": 0.0406,
      "step": 800
    },
    {
      "epoch": 1.9252615844544096,
      "grad_norm": 0.7991802096366882,
      "learning_rate": 3.38245663501611e-05,
      "loss": 0.0388,
      "step": 805
    },
    {
      "epoch": 1.937219730941704,
      "grad_norm": 0.7038184404373169,
      "learning_rate": 3.364851925417191e-05,
      "loss": 0.0401,
      "step": 810
    },
    {
      "epoch": 1.9491778774289985,
      "grad_norm": 0.5613796710968018,
      "learning_rate": 3.347198363306942e-05,
      "loss": 0.048,
      "step": 815
    },
    {
      "epoch": 1.961136023916293,
      "grad_norm": 0.48176127672195435,
      "learning_rate": 3.32949694587438e-05,
      "loss": 0.0357,
      "step": 820
    },
    {
      "epoch": 1.9730941704035874,
      "grad_norm": 0.4175252616405487,
      "learning_rate": 3.311748673011709e-05,
      "loss": 0.0406,
      "step": 825
    },
    {
      "epoch": 1.985052316890882,
      "grad_norm": 0.6962341070175171,
      "learning_rate": 3.293954547257832e-05,
      "loss": 0.0482,
      "step": 830
    },
    {
      "epoch": 1.9970104633781762,
      "grad_norm": 0.4897420406341553,
      "learning_rate": 3.276115573741724e-05,
      "loss": 0.0457,
      "step": 835
    },
    {
      "epoch": 2.008968609865471,
      "grad_norm": 0.5920233726501465,
      "learning_rate": 3.258232760125657e-05,
      "loss": 0.0378,
      "step": 840
    },
    {
      "epoch": 2.0209267563527655,
      "grad_norm": 0.6143045425415039,
      "learning_rate": 3.240307116548279e-05,
      "loss": 0.0432,
      "step": 845
    },
    {
      "epoch": 2.0328849028400597,
      "grad_norm": 0.580260694026947,
      "learning_rate": 3.222339655567556e-05,
      "loss": 0.0459,
      "step": 850
    },
    {
      "epoch": 2.0448430493273544,
      "grad_norm": 0.5552830100059509,
      "learning_rate": 3.2043313921035743e-05,
      "loss": 0.0292,
      "step": 855
    },
    {
      "epoch": 2.0568011958146486,
      "grad_norm": 0.5925277471542358,
      "learning_rate": 3.186283343381213e-05,
      "loss": 0.0291,
      "step": 860
    },
    {
      "epoch": 2.0687593423019432,
      "grad_norm": 0.6500665545463562,
      "learning_rate": 3.168196528872682e-05,
      "loss": 0.0376,
      "step": 865
    },
    {
      "epoch": 2.0807174887892375,
      "grad_norm": 0.328319787979126,
      "learning_rate": 3.150071970239941e-05,
      "loss": 0.0272,
      "step": 870
    },
    {
      "epoch": 2.092675635276532,
      "grad_norm": 0.49454575777053833,
      "learning_rate": 3.13191069127698e-05,
      "loss": 0.0337,
      "step": 875
    },
    {
      "epoch": 2.1046337817638268,
      "grad_norm": 0.466003954410553,
      "learning_rate": 3.1137137178519985e-05,
      "loss": 0.03,
      "step": 880
    },
    {
      "epoch": 2.116591928251121,
      "grad_norm": 0.5952199101448059,
      "learning_rate": 3.0954820778494516e-05,
      "loss": 0.034,
      "step": 885
    },
    {
      "epoch": 2.1285500747384156,
      "grad_norm": 0.5215238928794861,
      "learning_rate": 3.07721680111199e-05,
      "loss": 0.0298,
      "step": 890
    },
    {
      "epoch": 2.14050822122571,
      "grad_norm": 0.6787840723991394,
      "learning_rate": 3.0589189193822895e-05,
      "loss": 0.0255,
      "step": 895
    },
    {
      "epoch": 2.1524663677130045,
      "grad_norm": 0.5151398777961731,
      "learning_rate": 3.040589466244768e-05,
      "loss": 0.0291,
      "step": 900
    },
    {
      "epoch": 2.164424514200299,
      "grad_norm": 0.6676251888275146,
      "learning_rate": 3.0222294770672053e-05,
      "loss": 0.0281,
      "step": 905
    },
    {
      "epoch": 2.1763826606875933,
      "grad_norm": 0.46516361832618713,
      "learning_rate": 3.0038399889422553e-05,
      "loss": 0.0209,
      "step": 910
    },
    {
      "epoch": 2.188340807174888,
      "grad_norm": 0.9628980755805969,
      "learning_rate": 2.9854220406288668e-05,
      "loss": 0.0374,
      "step": 915
    },
    {
      "epoch": 2.200298953662182,
      "grad_norm": 0.4208248257637024,
      "learning_rate": 2.966976672493607e-05,
      "loss": 0.0329,
      "step": 920
    },
    {
      "epoch": 2.212257100149477,
      "grad_norm": 0.960185706615448,
      "learning_rate": 2.948504926451896e-05,
      "loss": 0.0413,
      "step": 925
    },
    {
      "epoch": 2.2242152466367715,
      "grad_norm": 0.6481686234474182,
      "learning_rate": 2.9300078459091462e-05,
      "loss": 0.0348,
      "step": 930
    },
    {
      "epoch": 2.2361733931240657,
      "grad_norm": 0.688122570514679,
      "learning_rate": 2.9114864757018352e-05,
      "loss": 0.0294,
      "step": 935
    },
    {
      "epoch": 2.2481315396113604,
      "grad_norm": 0.4938519299030304,
      "learning_rate": 2.8929418620384753e-05,
      "loss": 0.0359,
      "step": 940
    },
    {
      "epoch": 2.2600896860986546,
      "grad_norm": 0.37619251012802124,
      "learning_rate": 2.8743750524405254e-05,
      "loss": 0.0251,
      "step": 945
    },
    {
      "epoch": 2.2720478325859492,
      "grad_norm": 0.6733412146568298,
      "learning_rate": 2.8557870956832132e-05,
      "loss": 0.0368,
      "step": 950
    },
    {
      "epoch": 2.2840059790732434,
      "grad_norm": 0.36500585079193115,
      "learning_rate": 2.8371790417362987e-05,
      "loss": 0.0256,
      "step": 955
    },
    {
      "epoch": 2.295964125560538,
      "grad_norm": 0.36346250772476196,
      "learning_rate": 2.8185519417047624e-05,
      "loss": 0.0226,
      "step": 960
    },
    {
      "epoch": 2.3079222720478327,
      "grad_norm": 0.8299899101257324,
      "learning_rate": 2.799906847769433e-05,
      "loss": 0.029,
      "step": 965
    },
    {
      "epoch": 2.319880418535127,
      "grad_norm": 0.9291535019874573,
      "learning_rate": 2.781244813127552e-05,
      "loss": 0.0279,
      "step": 970
    },
    {
      "epoch": 2.3318385650224216,
      "grad_norm": 0.42336174845695496,
      "learning_rate": 2.762566891933285e-05,
      "loss": 0.0371,
      "step": 975
    },
    {
      "epoch": 2.343796711509716,
      "grad_norm": 0.7763833999633789,
      "learning_rate": 2.7438741392381705e-05,
      "loss": 0.0359,
      "step": 980
    },
    {
      "epoch": 2.3557548579970105,
      "grad_norm": 0.9157342910766602,
      "learning_rate": 2.7251676109315338e-05,
      "loss": 0.0328,
      "step": 985
    },
    {
      "epoch": 2.367713004484305,
      "grad_norm": 1.0131932497024536,
      "learning_rate": 2.7064483636808313e-05,
      "loss": 0.0265,
      "step": 990
    },
    {
      "epoch": 2.3796711509715993,
      "grad_norm": 0.4146224856376648,
      "learning_rate": 2.6877174548719706e-05,
      "loss": 0.0361,
      "step": 995
    },
    {
      "epoch": 2.391629297458894,
      "grad_norm": 0.9083970189094543,
      "learning_rate": 2.668975942549583e-05,
      "loss": 0.0279,
      "step": 1000
    },
    {
      "epoch": 2.403587443946188,
      "grad_norm": 0.8512977957725525,
      "learning_rate": 2.650224885357251e-05,
      "loss": 0.0227,
      "step": 1005
    },
    {
      "epoch": 2.415545590433483,
      "grad_norm": 0.7963445782661438,
      "learning_rate": 2.6314653424777193e-05,
      "loss": 0.0327,
      "step": 1010
    },
    {
      "epoch": 2.4275037369207775,
      "grad_norm": 1.0128358602523804,
      "learning_rate": 2.612698373573056e-05,
      "loss": 0.0357,
      "step": 1015
    },
    {
      "epoch": 2.4394618834080717,
      "grad_norm": 0.5992169380187988,
      "learning_rate": 2.593925038724802e-05,
      "loss": 0.0341,
      "step": 1020
    },
    {
      "epoch": 2.4514200298953663,
      "grad_norm": 0.5387618541717529,
      "learning_rate": 2.575146398374087e-05,
      "loss": 0.0247,
      "step": 1025
    },
    {
      "epoch": 2.4633781763826605,
      "grad_norm": 0.5542154908180237,
      "learning_rate": 2.5563635132617302e-05,
      "loss": 0.0229,
      "step": 1030
    },
    {
      "epoch": 2.475336322869955,
      "grad_norm": 0.4752875566482544,
      "learning_rate": 2.5375774443683265e-05,
      "loss": 0.0432,
      "step": 1035
    },
    {
      "epoch": 2.4872944693572494,
      "grad_norm": 0.4711446762084961,
      "learning_rate": 2.518789252854305e-05,
      "loss": 0.0198,
      "step": 1040
    },
    {
      "epoch": 2.499252615844544,
      "grad_norm": 0.7046725153923035,
      "learning_rate": 2.5e-05,
      "loss": 0.0322,
      "step": 1045
    },
    {
      "epoch": 2.5112107623318387,
      "grad_norm": 1.0238018035888672,
      "learning_rate": 2.4812107471456954e-05,
      "loss": 0.0431,
      "step": 1050
    },
    {
      "epoch": 2.523168908819133,
      "grad_norm": 0.3083067834377289,
      "learning_rate": 2.4624225556316744e-05,
      "loss": 0.025,
      "step": 1055
    },
    {
      "epoch": 2.5351270553064276,
      "grad_norm": 0.36663854122161865,
      "learning_rate": 2.44363648673827e-05,
      "loss": 0.0245,
      "step": 1060
    },
    {
      "epoch": 2.547085201793722,
      "grad_norm": 0.6427327394485474,
      "learning_rate": 2.4248536016259135e-05,
      "loss": 0.0284,
      "step": 1065
    },
    {
      "epoch": 2.5590433482810164,
      "grad_norm": 0.35747870802879333,
      "learning_rate": 2.4060749612751988e-05,
      "loss": 0.0186,
      "step": 1070
    },
    {
      "epoch": 2.5710014947683106,
      "grad_norm": 0.4468722939491272,
      "learning_rate": 2.387301626426944e-05,
      "loss": 0.0378,
      "step": 1075
    },
    {
      "epoch": 2.5829596412556053,
      "grad_norm": 0.5151249170303345,
      "learning_rate": 2.368534657522281e-05,
      "loss": 0.0318,
      "step": 1080
    },
    {
      "epoch": 2.5949177877429,
      "grad_norm": 0.7595095038414001,
      "learning_rate": 2.3497751146427493e-05,
      "loss": 0.0433,
      "step": 1085
    },
    {
      "epoch": 2.606875934230194,
      "grad_norm": 0.6947106719017029,
      "learning_rate": 2.3310240574504185e-05,
      "loss": 0.0361,
      "step": 1090
    },
    {
      "epoch": 2.618834080717489,
      "grad_norm": 0.5724536776542664,
      "learning_rate": 2.3122825451280296e-05,
      "loss": 0.0365,
      "step": 1095
    },
    {
      "epoch": 2.6307922272047835,
      "grad_norm": 0.9276584386825562,
      "learning_rate": 2.2935516363191693e-05,
      "loss": 0.0295,
      "step": 1100
    },
    {
      "epoch": 2.6427503736920777,
      "grad_norm": 0.7914692759513855,
      "learning_rate": 2.2748323890684665e-05,
      "loss": 0.0314,
      "step": 1105
    },
    {
      "epoch": 2.6547085201793723,
      "grad_norm": 0.7331878542900085,
      "learning_rate": 2.2561258607618297e-05,
      "loss": 0.0331,
      "step": 1110
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.383178174495697,
      "learning_rate": 2.2374331080667164e-05,
      "loss": 0.0439,
      "step": 1115
    },
    {
      "epoch": 2.678624813153961,
      "grad_norm": 0.3744180202484131,
      "learning_rate": 2.2187551868724485e-05,
      "loss": 0.0208,
      "step": 1120
    },
    {
      "epoch": 2.6905829596412554,
      "grad_norm": 0.6185144186019897,
      "learning_rate": 2.200093152230568e-05,
      "loss": 0.0299,
      "step": 1125
    },
    {
      "epoch": 2.70254110612855,
      "grad_norm": 0.559657633304596,
      "learning_rate": 2.1814480582952375e-05,
      "loss": 0.0285,
      "step": 1130
    },
    {
      "epoch": 2.7144992526158447,
      "grad_norm": 0.5390679240226746,
      "learning_rate": 2.1628209582637022e-05,
      "loss": 0.0286,
      "step": 1135
    },
    {
      "epoch": 2.726457399103139,
      "grad_norm": 0.8902041912078857,
      "learning_rate": 2.1442129043167874e-05,
      "loss": 0.0354,
      "step": 1140
    },
    {
      "epoch": 2.7384155455904335,
      "grad_norm": 0.37249886989593506,
      "learning_rate": 2.125624947559475e-05,
      "loss": 0.0281,
      "step": 1145
    },
    {
      "epoch": 2.750373692077728,
      "grad_norm": 0.49380356073379517,
      "learning_rate": 2.1070581379615253e-05,
      "loss": 0.0244,
      "step": 1150
    },
    {
      "epoch": 2.7623318385650224,
      "grad_norm": 0.8165044188499451,
      "learning_rate": 2.088513524298165e-05,
      "loss": 0.0372,
      "step": 1155
    },
    {
      "epoch": 2.774289985052317,
      "grad_norm": 0.5371778011322021,
      "learning_rate": 2.0699921540908544e-05,
      "loss": 0.0234,
      "step": 1160
    },
    {
      "epoch": 2.7862481315396113,
      "grad_norm": 0.4927629828453064,
      "learning_rate": 2.0514950735481052e-05,
      "loss": 0.0316,
      "step": 1165
    },
    {
      "epoch": 2.798206278026906,
      "grad_norm": 0.8102060556411743,
      "learning_rate": 2.033023327506393e-05,
      "loss": 0.0447,
      "step": 1170
    },
    {
      "epoch": 2.8101644245142,
      "grad_norm": 0.5554202198982239,
      "learning_rate": 2.0145779593711338e-05,
      "loss": 0.0316,
      "step": 1175
    },
    {
      "epoch": 2.8221225710014948,
      "grad_norm": 0.4643726050853729,
      "learning_rate": 1.9961600110577456e-05,
      "loss": 0.0252,
      "step": 1180
    },
    {
      "epoch": 2.8340807174887894,
      "grad_norm": 0.8105862140655518,
      "learning_rate": 1.9777705229327952e-05,
      "loss": 0.0315,
      "step": 1185
    },
    {
      "epoch": 2.8460388639760836,
      "grad_norm": 0.34342893958091736,
      "learning_rate": 1.9594105337552323e-05,
      "loss": 0.0254,
      "step": 1190
    },
    {
      "epoch": 2.8579970104633783,
      "grad_norm": 0.6640718579292297,
      "learning_rate": 1.9410810806177104e-05,
      "loss": 0.0228,
      "step": 1195
    },
    {
      "epoch": 2.8699551569506725,
      "grad_norm": 0.6938194036483765,
      "learning_rate": 1.9227831988880107e-05,
      "loss": 0.0264,
      "step": 1200
    },
    {
      "epoch": 2.881913303437967,
      "grad_norm": 0.19737358391284943,
      "learning_rate": 1.9045179221505497e-05,
      "loss": 0.0157,
      "step": 1205
    },
    {
      "epoch": 2.8938714499252614,
      "grad_norm": 0.7193530201911926,
      "learning_rate": 1.8862862821480025e-05,
      "loss": 0.0334,
      "step": 1210
    },
    {
      "epoch": 2.905829596412556,
      "grad_norm": 0.461428701877594,
      "learning_rate": 1.8680893087230204e-05,
      "loss": 0.029,
      "step": 1215
    },
    {
      "epoch": 2.9177877428998507,
      "grad_norm": 0.5498121976852417,
      "learning_rate": 1.8499280297600594e-05,
      "loss": 0.029,
      "step": 1220
    },
    {
      "epoch": 2.929745889387145,
      "grad_norm": 0.6723641753196716,
      "learning_rate": 1.831803471127318e-05,
      "loss": 0.0263,
      "step": 1225
    },
    {
      "epoch": 2.9417040358744395,
      "grad_norm": 0.5063521862030029,
      "learning_rate": 1.813716656618788e-05,
      "loss": 0.0297,
      "step": 1230
    },
    {
      "epoch": 2.953662182361734,
      "grad_norm": 0.5885164141654968,
      "learning_rate": 1.795668607896426e-05,
      "loss": 0.036,
      "step": 1235
    },
    {
      "epoch": 2.9656203288490284,
      "grad_norm": 0.5046067833900452,
      "learning_rate": 1.7776603444324445e-05,
      "loss": 0.0361,
      "step": 1240
    },
    {
      "epoch": 2.977578475336323,
      "grad_norm": 0.5021582841873169,
      "learning_rate": 1.759692883451721e-05,
      "loss": 0.0252,
      "step": 1245
    },
    {
      "epoch": 2.9895366218236172,
      "grad_norm": 0.7973111271858215,
      "learning_rate": 1.741767239874344e-05,
      "loss": 0.0271,
      "step": 1250
    },
    {
      "epoch": 3.001494768310912,
      "grad_norm": 1.7499645948410034,
      "learning_rate": 1.723884426258277e-05,
      "loss": 0.0318,
      "step": 1255
    },
    {
      "epoch": 3.013452914798206,
      "grad_norm": 0.6428520679473877,
      "learning_rate": 1.7060454527421688e-05,
      "loss": 0.0193,
      "step": 1260
    },
    {
      "epoch": 3.0254110612855007,
      "grad_norm": 0.697303295135498,
      "learning_rate": 1.6882513269882917e-05,
      "loss": 0.0235,
      "step": 1265
    },
    {
      "epoch": 3.0373692077727954,
      "grad_norm": 0.5823438763618469,
      "learning_rate": 1.670503054125621e-05,
      "loss": 0.0183,
      "step": 1270
    },
    {
      "epoch": 3.0493273542600896,
      "grad_norm": 0.6484075784683228,
      "learning_rate": 1.6528016366930592e-05,
      "loss": 0.0138,
      "step": 1275
    },
    {
      "epoch": 3.0612855007473843,
      "grad_norm": 0.8780815601348877,
      "learning_rate": 1.63514807458281e-05,
      "loss": 0.0244,
      "step": 1280
    },
    {
      "epoch": 3.0732436472346785,
      "grad_norm": 0.4560433626174927,
      "learning_rate": 1.61754336498389e-05,
      "loss": 0.0117,
      "step": 1285
    },
    {
      "epoch": 3.085201793721973,
      "grad_norm": 0.2612263858318329,
      "learning_rate": 1.59998850232581e-05,
      "loss": 0.0142,
      "step": 1290
    },
    {
      "epoch": 3.0971599402092678,
      "grad_norm": 0.6966968178749084,
      "learning_rate": 1.5824844782223954e-05,
      "loss": 0.0141,
      "step": 1295
    },
    {
      "epoch": 3.109118086696562,
      "grad_norm": 0.4598274827003479,
      "learning_rate": 1.5650322814157764e-05,
      "loss": 0.0181,
      "step": 1300
    },
    {
      "epoch": 3.1210762331838566,
      "grad_norm": 0.3780767321586609,
      "learning_rate": 1.5476328977205397e-05,
      "loss": 0.0098,
      "step": 1305
    },
    {
      "epoch": 3.133034379671151,
      "grad_norm": 0.41301998496055603,
      "learning_rate": 1.5302873099680377e-05,
      "loss": 0.0095,
      "step": 1310
    },
    {
      "epoch": 3.1449925261584455,
      "grad_norm": 0.4434594213962555,
      "learning_rate": 1.5129964979508792e-05,
      "loss": 0.0151,
      "step": 1315
    },
    {
      "epoch": 3.15695067264574,
      "grad_norm": 0.7352045774459839,
      "learning_rate": 1.495761438367577e-05,
      "loss": 0.0292,
      "step": 1320
    },
    {
      "epoch": 3.1689088191330343,
      "grad_norm": 0.6801838874816895,
      "learning_rate": 1.4785831047673799e-05,
      "loss": 0.0162,
      "step": 1325
    },
    {
      "epoch": 3.180866965620329,
      "grad_norm": 0.4488966464996338,
      "learning_rate": 1.4614624674952842e-05,
      "loss": 0.0103,
      "step": 1330
    },
    {
      "epoch": 3.192825112107623,
      "grad_norm": 0.644514799118042,
      "learning_rate": 1.4444004936372165e-05,
      "loss": 0.0225,
      "step": 1335
    },
    {
      "epoch": 3.204783258594918,
      "grad_norm": 0.5824326276779175,
      "learning_rate": 1.4273981469654093e-05,
      "loss": 0.0174,
      "step": 1340
    },
    {
      "epoch": 3.216741405082212,
      "grad_norm": 0.4944539964199066,
      "learning_rate": 1.4104563878839621e-05,
      "loss": 0.0202,
      "step": 1345
    },
    {
      "epoch": 3.2286995515695067,
      "grad_norm": 0.6498035788536072,
      "learning_rate": 1.3935761733745865e-05,
      "loss": 0.0119,
      "step": 1350
    },
    {
      "epoch": 3.2406576980568014,
      "grad_norm": 0.38544780015945435,
      "learning_rate": 1.3767584569425562e-05,
      "loss": 0.0127,
      "step": 1355
    },
    {
      "epoch": 3.2526158445440956,
      "grad_norm": 1.0216518640518188,
      "learning_rate": 1.3600041885628409e-05,
      "loss": 0.0145,
      "step": 1360
    },
    {
      "epoch": 3.2645739910313902,
      "grad_norm": 0.5669509768486023,
      "learning_rate": 1.3433143146264493e-05,
      "loss": 0.0116,
      "step": 1365
    },
    {
      "epoch": 3.2765321375186844,
      "grad_norm": 0.37746644020080566,
      "learning_rate": 1.3266897778869702e-05,
      "loss": 0.0103,
      "step": 1370
    },
    {
      "epoch": 3.288490284005979,
      "grad_norm": 1.114930272102356,
      "learning_rate": 1.3101315174073162e-05,
      "loss": 0.0152,
      "step": 1375
    },
    {
      "epoch": 3.3004484304932733,
      "grad_norm": 1.2389074563980103,
      "learning_rate": 1.2936404685066852e-05,
      "loss": 0.0143,
      "step": 1380
    },
    {
      "epoch": 3.312406576980568,
      "grad_norm": 0.8430703282356262,
      "learning_rate": 1.2772175627077205e-05,
      "loss": 0.0166,
      "step": 1385
    },
    {
      "epoch": 3.3243647234678626,
      "grad_norm": 1.1820229291915894,
      "learning_rate": 1.2608637276838986e-05,
      "loss": 0.0182,
      "step": 1390
    },
    {
      "epoch": 3.336322869955157,
      "grad_norm": 0.7732896208763123,
      "learning_rate": 1.244579887207126e-05,
      "loss": 0.0181,
      "step": 1395
    },
    {
      "epoch": 3.3482810164424515,
      "grad_norm": 1.0598599910736084,
      "learning_rate": 1.2283669610955542e-05,
      "loss": 0.0175,
      "step": 1400
    },
    {
      "epoch": 3.360239162929746,
      "grad_norm": 0.669745683670044,
      "learning_rate": 1.2122258651616306e-05,
      "loss": 0.0248,
      "step": 1405
    },
    {
      "epoch": 3.3721973094170403,
      "grad_norm": 0.5088245868682861,
      "learning_rate": 1.1961575111603588e-05,
      "loss": 0.0175,
      "step": 1410
    },
    {
      "epoch": 3.384155455904335,
      "grad_norm": 0.8206422328948975,
      "learning_rate": 1.1801628067378031e-05,
      "loss": 0.0177,
      "step": 1415
    },
    {
      "epoch": 3.396113602391629,
      "grad_norm": 0.19338686764240265,
      "learning_rate": 1.1642426553798174e-05,
      "loss": 0.0192,
      "step": 1420
    },
    {
      "epoch": 3.408071748878924,
      "grad_norm": 0.602216899394989,
      "learning_rate": 1.148397956361007e-05,
      "loss": 0.0127,
      "step": 1425
    },
    {
      "epoch": 3.420029895366218,
      "grad_norm": 1.2466075420379639,
      "learning_rate": 1.1326296046939333e-05,
      "loss": 0.0192,
      "step": 1430
    },
    {
      "epoch": 3.4319880418535127,
      "grad_norm": 0.3795795440673828,
      "learning_rate": 1.1169384910785614e-05,
      "loss": 0.0095,
      "step": 1435
    },
    {
      "epoch": 3.4439461883408073,
      "grad_norm": 0.8901958465576172,
      "learning_rate": 1.1013255018519425e-05,
      "loss": 0.015,
      "step": 1440
    },
    {
      "epoch": 3.4559043348281016,
      "grad_norm": 0.8561021685600281,
      "learning_rate": 1.0857915189381513e-05,
      "loss": 0.0178,
      "step": 1445
    },
    {
      "epoch": 3.467862481315396,
      "grad_norm": 0.7785664200782776,
      "learning_rate": 1.0703374197984653e-05,
      "loss": 0.022,
      "step": 1450
    },
    {
      "epoch": 3.4798206278026904,
      "grad_norm": 0.9649757742881775,
      "learning_rate": 1.0549640773818029e-05,
      "loss": 0.0177,
      "step": 1455
    },
    {
      "epoch": 3.491778774289985,
      "grad_norm": 0.9792755842208862,
      "learning_rate": 1.0396723600754143e-05,
      "loss": 0.0241,
      "step": 1460
    },
    {
      "epoch": 3.5037369207772793,
      "grad_norm": 0.9604386687278748,
      "learning_rate": 1.0244631316558267e-05,
      "loss": 0.0327,
      "step": 1465
    },
    {
      "epoch": 3.515695067264574,
      "grad_norm": 0.82301265001297,
      "learning_rate": 1.0093372512400551e-05,
      "loss": 0.0264,
      "step": 1470
    },
    {
      "epoch": 3.5276532137518686,
      "grad_norm": 0.7299047112464905,
      "learning_rate": 9.942955732370707e-06,
      "loss": 0.0191,
      "step": 1475
    },
    {
      "epoch": 3.539611360239163,
      "grad_norm": 0.37303781509399414,
      "learning_rate": 9.793389472995393e-06,
      "loss": 0.0191,
      "step": 1480
    },
    {
      "epoch": 3.5515695067264574,
      "grad_norm": 0.5559309720993042,
      "learning_rate": 9.644682182758306e-06,
      "loss": 0.0211,
      "step": 1485
    },
    {
      "epoch": 3.563527653213752,
      "grad_norm": 0.4034077823162079,
      "learning_rate": 9.496842261622921e-06,
      "loss": 0.0138,
      "step": 1490
    },
    {
      "epoch": 3.5754857997010463,
      "grad_norm": 0.8548296093940735,
      "learning_rate": 9.349878060557999e-06,
      "loss": 0.0137,
      "step": 1495
    },
    {
      "epoch": 3.587443946188341,
      "grad_norm": 0.4712274670600891,
      "learning_rate": 9.203797881065906e-06,
      "loss": 0.0195,
      "step": 1500
    },
    {
      "epoch": 3.599402092675635,
      "grad_norm": 0.6298921704292297,
      "learning_rate": 9.058609974713655e-06,
      "loss": 0.0141,
      "step": 1505
    },
    {
      "epoch": 3.61136023916293,
      "grad_norm": 0.6767133474349976,
      "learning_rate": 8.914322542666822e-06,
      "loss": 0.0213,
      "step": 1510
    },
    {
      "epoch": 3.623318385650224,
      "grad_norm": 0.9857379794120789,
      "learning_rate": 8.770943735226303e-06,
      "loss": 0.0232,
      "step": 1515
    },
    {
      "epoch": 3.6352765321375187,
      "grad_norm": 0.6573039889335632,
      "learning_rate": 8.628481651367876e-06,
      "loss": 0.0189,
      "step": 1520
    },
    {
      "epoch": 3.6472346786248133,
      "grad_norm": 0.6067800521850586,
      "learning_rate": 8.486944338284797e-06,
      "loss": 0.0151,
      "step": 1525
    },
    {
      "epoch": 3.6591928251121075,
      "grad_norm": 0.41472485661506653,
      "learning_rate": 8.346339790933166e-06,
      "loss": 0.011,
      "step": 1530
    },
    {
      "epoch": 3.671150971599402,
      "grad_norm": 0.3811599314212799,
      "learning_rate": 8.206675951580381e-06,
      "loss": 0.0127,
      "step": 1535
    },
    {
      "epoch": 3.683109118086697,
      "grad_norm": 0.3842083811759949,
      "learning_rate": 8.067960709356478e-06,
      "loss": 0.0222,
      "step": 1540
    },
    {
      "epoch": 3.695067264573991,
      "grad_norm": 0.8651444911956787,
      "learning_rate": 7.930201899808475e-06,
      "loss": 0.0246,
      "step": 1545
    },
    {
      "epoch": 3.7070254110612852,
      "grad_norm": 0.5229407548904419,
      "learning_rate": 7.793407304457836e-06,
      "loss": 0.0115,
      "step": 1550
    },
    {
      "epoch": 3.71898355754858,
      "grad_norm": 0.9863315224647522,
      "learning_rate": 7.657584650360847e-06,
      "loss": 0.0189,
      "step": 1555
    },
    {
      "epoch": 3.7309417040358746,
      "grad_norm": 0.7004222273826599,
      "learning_rate": 7.522741609672193e-06,
      "loss": 0.0116,
      "step": 1560
    },
    {
      "epoch": 3.7428998505231688,
      "grad_norm": 0.6254629492759705,
      "learning_rate": 7.388885799211573e-06,
      "loss": 0.0184,
      "step": 1565
    },
    {
      "epoch": 3.7548579970104634,
      "grad_norm": 0.6357623338699341,
      "learning_rate": 7.256024780033418e-06,
      "loss": 0.0119,
      "step": 1570
    },
    {
      "epoch": 3.766816143497758,
      "grad_norm": 0.6104183793067932,
      "learning_rate": 7.124166056999854e-06,
      "loss": 0.0089,
      "step": 1575
    },
    {
      "epoch": 3.7787742899850523,
      "grad_norm": 1.3286741971969604,
      "learning_rate": 6.993317078356709e-06,
      "loss": 0.0221,
      "step": 1580
    },
    {
      "epoch": 3.790732436472347,
      "grad_norm": 1.1910456418991089,
      "learning_rate": 6.863485235312853e-06,
      "loss": 0.0182,
      "step": 1585
    },
    {
      "epoch": 3.802690582959641,
      "grad_norm": 0.7423701286315918,
      "learning_rate": 6.7346778616226515e-06,
      "loss": 0.0268,
      "step": 1590
    },
    {
      "epoch": 3.814648729446936,
      "grad_norm": 0.7771933078765869,
      "learning_rate": 6.606902233171711e-06,
      "loss": 0.0217,
      "step": 1595
    },
    {
      "epoch": 3.82660687593423,
      "grad_norm": 0.7063082456588745,
      "learning_rate": 6.480165567565913e-06,
      "loss": 0.0171,
      "step": 1600
    },
    {
      "epoch": 3.8385650224215246,
      "grad_norm": 0.5673719644546509,
      "learning_rate": 6.354475023723686e-06,
      "loss": 0.0142,
      "step": 1605
    },
    {
      "epoch": 3.8505231689088193,
      "grad_norm": 0.941227376461029,
      "learning_rate": 6.229837701471644e-06,
      "loss": 0.0177,
      "step": 1610
    },
    {
      "epoch": 3.8624813153961135,
      "grad_norm": 0.24737170338630676,
      "learning_rate": 6.106260641143546e-06,
      "loss": 0.0212,
      "step": 1615
    },
    {
      "epoch": 3.874439461883408,
      "grad_norm": 0.5687448382377625,
      "learning_rate": 5.983750823182574e-06,
      "loss": 0.0144,
      "step": 1620
    },
    {
      "epoch": 3.886397608370703,
      "grad_norm": 0.9398664236068726,
      "learning_rate": 5.86231516774709e-06,
      "loss": 0.0149,
      "step": 1625
    },
    {
      "epoch": 3.898355754857997,
      "grad_norm": 0.33037155866622925,
      "learning_rate": 5.741960534319677e-06,
      "loss": 0.0123,
      "step": 1630
    },
    {
      "epoch": 3.910313901345291,
      "grad_norm": 0.6622065305709839,
      "learning_rate": 5.622693721319727e-06,
      "loss": 0.0124,
      "step": 1635
    },
    {
      "epoch": 3.922272047832586,
      "grad_norm": 0.5076555013656616,
      "learning_rate": 5.5045214657193925e-06,
      "loss": 0.0141,
      "step": 1640
    },
    {
      "epoch": 3.9342301943198805,
      "grad_norm": 0.14695705473423004,
      "learning_rate": 5.387450442663025e-06,
      "loss": 0.0142,
      "step": 1645
    },
    {
      "epoch": 3.9461883408071747,
      "grad_norm": 0.37793174386024475,
      "learning_rate": 5.271487265090163e-06,
      "loss": 0.0113,
      "step": 1650
    },
    {
      "epoch": 3.9581464872944694,
      "grad_norm": 0.6981557607650757,
      "learning_rate": 5.156638483361934e-06,
      "loss": 0.0179,
      "step": 1655
    },
    {
      "epoch": 3.970104633781764,
      "grad_norm": 0.8722692728042603,
      "learning_rate": 5.0429105848911e-06,
      "loss": 0.0167,
      "step": 1660
    },
    {
      "epoch": 3.9820627802690582,
      "grad_norm": 0.971043586730957,
      "learning_rate": 4.930309993775578e-06,
      "loss": 0.0115,
      "step": 1665
    },
    {
      "epoch": 3.994020926756353,
      "grad_norm": 0.6254692673683167,
      "learning_rate": 4.8188430704355605e-06,
      "loss": 0.0115,
      "step": 1670
    },
    {
      "epoch": 4.0059790732436475,
      "grad_norm": 0.2083253413438797,
      "learning_rate": 4.708516111254238e-06,
      "loss": 0.0149,
      "step": 1675
    },
    {
      "epoch": 4.017937219730942,
      "grad_norm": 0.7459127306938171,
      "learning_rate": 4.59933534822217e-06,
      "loss": 0.0107,
      "step": 1680
    },
    {
      "epoch": 4.029895366218236,
      "grad_norm": 0.6895865797996521,
      "learning_rate": 4.491306948585219e-06,
      "loss": 0.0147,
      "step": 1685
    },
    {
      "epoch": 4.041853512705531,
      "grad_norm": 0.45793023705482483,
      "learning_rate": 4.384437014496215e-06,
      "loss": 0.009,
      "step": 1690
    },
    {
      "epoch": 4.053811659192825,
      "grad_norm": 0.30743464827537537,
      "learning_rate": 4.278731582670239e-06,
      "loss": 0.0073,
      "step": 1695
    },
    {
      "epoch": 4.0657698056801195,
      "grad_norm": 1.2472518682479858,
      "learning_rate": 4.1741966240436445e-06,
      "loss": 0.0094,
      "step": 1700
    },
    {
      "epoch": 4.077727952167414,
      "grad_norm": 0.48823294043540955,
      "learning_rate": 4.070838043436786e-06,
      "loss": 0.008,
      "step": 1705
    },
    {
      "epoch": 4.089686098654709,
      "grad_norm": 0.6218563914299011,
      "learning_rate": 3.968661679220468e-06,
      "loss": 0.0101,
      "step": 1710
    },
    {
      "epoch": 4.101644245142003,
      "grad_norm": 0.15734516084194183,
      "learning_rate": 3.867673302986161e-06,
      "loss": 0.0081,
      "step": 1715
    },
    {
      "epoch": 4.113602391629297,
      "grad_norm": 0.6841528415679932,
      "learning_rate": 3.7678786192199694e-06,
      "loss": 0.0113,
      "step": 1720
    },
    {
      "epoch": 4.125560538116592,
      "grad_norm": 0.21844132244586945,
      "learning_rate": 3.669283264980408e-06,
      "loss": 0.0104,
      "step": 1725
    },
    {
      "epoch": 4.1375186846038865,
      "grad_norm": 1.151573896408081,
      "learning_rate": 3.571892809580013e-06,
      "loss": 0.0154,
      "step": 1730
    },
    {
      "epoch": 4.149476831091181,
      "grad_norm": 0.45680588483810425,
      "learning_rate": 3.475712754270716e-06,
      "loss": 0.0107,
      "step": 1735
    },
    {
      "epoch": 4.161434977578475,
      "grad_norm": 0.23463399708271027,
      "learning_rate": 3.3807485319331034e-06,
      "loss": 0.0111,
      "step": 1740
    },
    {
      "epoch": 4.17339312406577,
      "grad_norm": 0.34237194061279297,
      "learning_rate": 3.2870055067695556e-06,
      "loss": 0.0087,
      "step": 1745
    },
    {
      "epoch": 4.185351270553064,
      "grad_norm": 0.26900872588157654,
      "learning_rate": 3.194488974001203e-06,
      "loss": 0.0086,
      "step": 1750
    },
    {
      "epoch": 4.197309417040358,
      "grad_norm": 0.1915324628353119,
      "learning_rate": 3.103204159568851e-06,
      "loss": 0.004,
      "step": 1755
    },
    {
      "epoch": 4.2092675635276535,
      "grad_norm": 0.6967126131057739,
      "learning_rate": 3.013156219837776e-06,
      "loss": 0.0137,
      "step": 1760
    },
    {
      "epoch": 4.221225710014948,
      "grad_norm": 1.3966867923736572,
      "learning_rate": 2.9243502413064368e-06,
      "loss": 0.0099,
      "step": 1765
    },
    {
      "epoch": 4.233183856502242,
      "grad_norm": 0.257445365190506,
      "learning_rate": 2.8367912403191977e-06,
      "loss": 0.0087,
      "step": 1770
    },
    {
      "epoch": 4.245142002989537,
      "grad_norm": 0.27273523807525635,
      "learning_rate": 2.750484162782929e-06,
      "loss": 0.0087,
      "step": 1775
    },
    {
      "epoch": 4.257100149476831,
      "grad_norm": 0.5466504096984863,
      "learning_rate": 2.6654338838876665e-06,
      "loss": 0.0128,
      "step": 1780
    },
    {
      "epoch": 4.2690582959641254,
      "grad_norm": 0.27948179841041565,
      "learning_rate": 2.581645207831204e-06,
      "loss": 0.0081,
      "step": 1785
    },
    {
      "epoch": 4.28101644245142,
      "grad_norm": 0.5633261203765869,
      "learning_rate": 2.499122867547729e-06,
      "loss": 0.0122,
      "step": 1790
    },
    {
      "epoch": 4.292974588938715,
      "grad_norm": 0.5617706775665283,
      "learning_rate": 2.4178715244404794e-06,
      "loss": 0.008,
      "step": 1795
    },
    {
      "epoch": 4.304932735426009,
      "grad_norm": 0.23696863651275635,
      "learning_rate": 2.3378957681184283e-06,
      "loss": 0.0061,
      "step": 1800
    },
    {
      "epoch": 4.316890881913303,
      "grad_norm": 0.37975013256073,
      "learning_rate": 2.2592001161370392e-06,
      "loss": 0.0044,
      "step": 1805
    },
    {
      "epoch": 4.328849028400598,
      "grad_norm": 0.8079380393028259,
      "learning_rate": 2.1817890137430934e-06,
      "loss": 0.0128,
      "step": 1810
    },
    {
      "epoch": 4.3408071748878925,
      "grad_norm": 0.4358161687850952,
      "learning_rate": 2.1056668336235622e-06,
      "loss": 0.0145,
      "step": 1815
    },
    {
      "epoch": 4.352765321375187,
      "grad_norm": 0.13853619992733002,
      "learning_rate": 2.030837875658656e-06,
      "loss": 0.01,
      "step": 1820
    },
    {
      "epoch": 4.364723467862481,
      "grad_norm": 0.1564410775899887,
      "learning_rate": 1.9573063666788875e-06,
      "loss": 0.0075,
      "step": 1825
    },
    {
      "epoch": 4.376681614349776,
      "grad_norm": 0.8492725491523743,
      "learning_rate": 1.8850764602263426e-06,
      "loss": 0.012,
      "step": 1830
    },
    {
      "epoch": 4.38863976083707,
      "grad_norm": 0.4738127589225769,
      "learning_rate": 1.8141522363200797e-06,
      "loss": 0.0053,
      "step": 1835
    },
    {
      "epoch": 4.400597907324364,
      "grad_norm": 0.7158594131469727,
      "learning_rate": 1.7445377012256126e-06,
      "loss": 0.008,
      "step": 1840
    },
    {
      "epoch": 4.4125560538116595,
      "grad_norm": 0.2946886718273163,
      "learning_rate": 1.6762367872286522e-06,
      "loss": 0.0047,
      "step": 1845
    },
    {
      "epoch": 4.424514200298954,
      "grad_norm": 0.4199522137641907,
      "learning_rate": 1.6092533524129622e-06,
      "loss": 0.0105,
      "step": 1850
    },
    {
      "epoch": 4.436472346786248,
      "grad_norm": 0.6322415471076965,
      "learning_rate": 1.5435911804424357e-06,
      "loss": 0.006,
      "step": 1855
    },
    {
      "epoch": 4.448430493273543,
      "grad_norm": 1.0315839052200317,
      "learning_rate": 1.479253980347392e-06,
      "loss": 0.0095,
      "step": 1860
    },
    {
      "epoch": 4.460388639760837,
      "grad_norm": 1.0432811975479126,
      "learning_rate": 1.4162453863150183e-06,
      "loss": 0.0074,
      "step": 1865
    },
    {
      "epoch": 4.472346786248131,
      "grad_norm": 0.3731776773929596,
      "learning_rate": 1.3545689574841342e-06,
      "loss": 0.0061,
      "step": 1870
    },
    {
      "epoch": 4.484304932735426,
      "grad_norm": 0.2885247468948364,
      "learning_rate": 1.2942281777441168e-06,
      "loss": 0.015,
      "step": 1875
    },
    {
      "epoch": 4.496263079222721,
      "grad_norm": 0.5823212265968323,
      "learning_rate": 1.2352264555381132e-06,
      "loss": 0.009,
      "step": 1880
    },
    {
      "epoch": 4.508221225710015,
      "grad_norm": 0.9644975066184998,
      "learning_rate": 1.1775671236705365e-06,
      "loss": 0.0123,
      "step": 1885
    },
    {
      "epoch": 4.520179372197309,
      "grad_norm": 0.7990277409553528,
      "learning_rate": 1.121253439118769e-06,
      "loss": 0.0079,
      "step": 1890
    },
    {
      "epoch": 4.532137518684604,
      "grad_norm": 0.6153960227966309,
      "learning_rate": 1.0662885828492036e-06,
      "loss": 0.0084,
      "step": 1895
    },
    {
      "epoch": 4.5440956651718984,
      "grad_norm": 0.8191128373146057,
      "learning_rate": 1.0126756596375686e-06,
      "loss": 0.0144,
      "step": 1900
    },
    {
      "epoch": 4.556053811659193,
      "grad_norm": 0.6022669672966003,
      "learning_rate": 9.604176978935343e-07,
      "loss": 0.0067,
      "step": 1905
    },
    {
      "epoch": 4.568011958146487,
      "grad_norm": 0.26910826563835144,
      "learning_rate": 9.095176494896663e-07,
      "loss": 0.0052,
      "step": 1910
    },
    {
      "epoch": 4.579970104633782,
      "grad_norm": 1.1590691804885864,
      "learning_rate": 8.599783895946761e-07,
      "loss": 0.0127,
      "step": 1915
    },
    {
      "epoch": 4.591928251121076,
      "grad_norm": 0.6298155784606934,
      "learning_rate": 8.118027165109926e-07,
      "loss": 0.0085,
      "step": 1920
    },
    {
      "epoch": 4.60388639760837,
      "grad_norm": 0.516044557094574,
      "learning_rate": 7.649933515167407e-07,
      "loss": 0.0084,
      "step": 1925
    },
    {
      "epoch": 4.6158445440956655,
      "grad_norm": 1.4137001037597656,
      "learning_rate": 7.195529387119815e-07,
      "loss": 0.01,
      "step": 1930
    },
    {
      "epoch": 4.62780269058296,
      "grad_norm": 0.350608229637146,
      "learning_rate": 6.754840448693789e-07,
      "loss": 0.0048,
      "step": 1935
    },
    {
      "epoch": 4.639760837070254,
      "grad_norm": 0.13270160555839539,
      "learning_rate": 6.327891592892127e-07,
      "loss": 0.0056,
      "step": 1940
    },
    {
      "epoch": 4.651718983557549,
      "grad_norm": 0.6810560822486877,
      "learning_rate": 5.914706936587494e-07,
      "loss": 0.0094,
      "step": 1945
    },
    {
      "epoch": 4.663677130044843,
      "grad_norm": 0.6775500178337097,
      "learning_rate": 5.515309819160403e-07,
      "loss": 0.0088,
      "step": 1950
    },
    {
      "epoch": 4.675635276532137,
      "grad_norm": 0.29312002658843994,
      "learning_rate": 5.129722801180542e-07,
      "loss": 0.0112,
      "step": 1955
    },
    {
      "epoch": 4.687593423019432,
      "grad_norm": 0.6175358891487122,
      "learning_rate": 4.75796766313269e-07,
      "loss": 0.0077,
      "step": 1960
    },
    {
      "epoch": 4.699551569506727,
      "grad_norm": 0.7972726821899414,
      "learning_rate": 4.4000654041862764e-07,
      "loss": 0.0079,
      "step": 1965
    },
    {
      "epoch": 4.711509715994021,
      "grad_norm": 0.38660648465156555,
      "learning_rate": 4.0560362410091704e-07,
      "loss": 0.0053,
      "step": 1970
    },
    {
      "epoch": 4.723467862481315,
      "grad_norm": 0.236992746591568,
      "learning_rate": 3.7258996066258103e-07,
      "loss": 0.0091,
      "step": 1975
    },
    {
      "epoch": 4.73542600896861,
      "grad_norm": 0.015732845291495323,
      "learning_rate": 3.4096741493194197e-07,
      "loss": 0.006,
      "step": 1980
    },
    {
      "epoch": 4.747384155455904,
      "grad_norm": 0.10387540608644485,
      "learning_rate": 3.107377731578709e-07,
      "loss": 0.0064,
      "step": 1985
    },
    {
      "epoch": 4.759342301943199,
      "grad_norm": 0.5767621397972107,
      "learning_rate": 2.819027429088822e-07,
      "loss": 0.007,
      "step": 1990
    },
    {
      "epoch": 4.771300448430493,
      "grad_norm": 0.5412309765815735,
      "learning_rate": 2.544639529766829e-07,
      "loss": 0.0071,
      "step": 1995
    },
    {
      "epoch": 4.783258594917788,
      "grad_norm": 0.6424384117126465,
      "learning_rate": 2.284229532841603e-07,
      "loss": 0.0076,
      "step": 2000
    },
    {
      "epoch": 4.795216741405082,
      "grad_norm": 1.127122163772583,
      "learning_rate": 2.0378121479783796e-07,
      "loss": 0.0173,
      "step": 2005
    },
    {
      "epoch": 4.807174887892376,
      "grad_norm": 0.31246551871299744,
      "learning_rate": 1.8054012944479227e-07,
      "loss": 0.01,
      "step": 2010
    },
    {
      "epoch": 4.819133034379671,
      "grad_norm": 0.3126978874206543,
      "learning_rate": 1.5870101003402084e-07,
      "loss": 0.0119,
      "step": 2015
    },
    {
      "epoch": 4.831091180866966,
      "grad_norm": 0.7156423330307007,
      "learning_rate": 1.3826509018227128e-07,
      "loss": 0.0129,
      "step": 2020
    },
    {
      "epoch": 4.84304932735426,
      "grad_norm": 0.2048042267560959,
      "learning_rate": 1.192335242443915e-07,
      "loss": 0.0048,
      "step": 2025
    },
    {
      "epoch": 4.855007473841555,
      "grad_norm": 0.8227670788764954,
      "learning_rate": 1.0160738724809549e-07,
      "loss": 0.0087,
      "step": 2030
    },
    {
      "epoch": 4.866965620328849,
      "grad_norm": 0.5995686054229736,
      "learning_rate": 8.538767483325383e-08,
      "loss": 0.0079,
      "step": 2035
    },
    {
      "epoch": 4.878923766816143,
      "grad_norm": 0.4633730947971344,
      "learning_rate": 7.057530319564409e-08,
      "loss": 0.0129,
      "step": 2040
    },
    {
      "epoch": 4.8908819133034385,
      "grad_norm": 0.495402067899704,
      "learning_rate": 5.717110903520617e-08,
      "loss": 0.0097,
      "step": 2045
    },
    {
      "epoch": 4.902840059790733,
      "grad_norm": 0.3420732021331787,
      "learning_rate": 4.517584950877452e-08,
      "loss": 0.0057,
      "step": 2050
    },
    {
      "epoch": 4.914798206278027,
      "grad_norm": 0.24426616728305817,
      "learning_rate": 3.459020218731512e-08,
      "loss": 0.0098,
      "step": 2055
    },
    {
      "epoch": 4.926756352765321,
      "grad_norm": 0.8651923537254333,
      "learning_rate": 2.5414765017642284e-08,
      "loss": 0.0204,
      "step": 2060
    },
    {
      "epoch": 4.938714499252616,
      "grad_norm": 0.4833781123161316,
      "learning_rate": 1.7650056288651127e-08,
      "loss": 0.0067,
      "step": 2065
    },
    {
      "epoch": 4.95067264573991,
      "grad_norm": 0.12067227810621262,
      "learning_rate": 1.1296514602038288e-08,
      "loss": 0.0067,
      "step": 2070
    },
    {
      "epoch": 4.962630792227205,
      "grad_norm": 0.16125398874282837,
      "learning_rate": 6.354498847521706e-09,
      "loss": 0.0145,
      "step": 2075
    },
    {
      "epoch": 4.974588938714499,
      "grad_norm": 0.9371150732040405,
      "learning_rate": 2.8242881825846223e-09,
      "loss": 0.009,
      "step": 2080
    },
    {
      "epoch": 4.986547085201794,
      "grad_norm": 0.573594868183136,
      "learning_rate": 7.06082016682652e-10,
      "loss": 0.0118,
      "step": 2085
    },
    {
      "epoch": 4.998505231689088,
      "grad_norm": 0.4264791011810303,
      "learning_rate": 0.0,
      "loss": 0.0051,
      "step": 2090
    }
  ],
  "logging_steps": 5,
  "max_steps": 2090,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.993704594412995e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}