{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1206,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008291873963515755,
      "grad_norm": 24.363085939314125,
      "learning_rate": 8.264462809917357e-08,
      "loss": 1.4091,
      "step": 1
    },
    {
      "epoch": 0.0041459369817578775,
      "grad_norm": 23.543346063299293,
      "learning_rate": 4.132231404958678e-07,
      "loss": 1.4056,
      "step": 5
    },
    {
      "epoch": 0.008291873963515755,
      "grad_norm": 8.313966376086048,
      "learning_rate": 8.264462809917356e-07,
      "loss": 1.3082,
      "step": 10
    },
    {
      "epoch": 0.012437810945273632,
      "grad_norm": 10.10778485014052,
      "learning_rate": 1.2396694214876035e-06,
      "loss": 1.1527,
      "step": 15
    },
    {
      "epoch": 0.01658374792703151,
      "grad_norm": 3.1507455192980616,
      "learning_rate": 1.6528925619834712e-06,
      "loss": 1.0115,
      "step": 20
    },
    {
      "epoch": 0.020729684908789386,
      "grad_norm": 2.949214823531174,
      "learning_rate": 2.066115702479339e-06,
      "loss": 0.9662,
      "step": 25
    },
    {
      "epoch": 0.024875621890547265,
      "grad_norm": 2.3865704411346944,
      "learning_rate": 2.479338842975207e-06,
      "loss": 0.9301,
      "step": 30
    },
    {
      "epoch": 0.02902155887230514,
      "grad_norm": 2.367353362561768,
      "learning_rate": 2.8925619834710743e-06,
      "loss": 0.9066,
      "step": 35
    },
    {
      "epoch": 0.03316749585406302,
      "grad_norm": 2.168359757453863,
      "learning_rate": 3.3057851239669424e-06,
      "loss": 0.8864,
      "step": 40
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 2.352027725705367,
      "learning_rate": 3.71900826446281e-06,
      "loss": 0.8776,
      "step": 45
    },
    {
      "epoch": 0.04145936981757877,
      "grad_norm": 2.221157238277382,
      "learning_rate": 4.132231404958678e-06,
      "loss": 0.8738,
      "step": 50
    },
    {
      "epoch": 0.04560530679933665,
      "grad_norm": 2.254296003816909,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.8654,
      "step": 55
    },
    {
      "epoch": 0.04975124378109453,
      "grad_norm": 2.2549051162675258,
      "learning_rate": 4.958677685950414e-06,
      "loss": 0.8603,
      "step": 60
    },
    {
      "epoch": 0.0538971807628524,
      "grad_norm": 2.421942208846336,
      "learning_rate": 5.371900826446281e-06,
      "loss": 0.8514,
      "step": 65
    },
    {
      "epoch": 0.05804311774461028,
      "grad_norm": 2.2852906820767833,
      "learning_rate": 5.785123966942149e-06,
      "loss": 0.8448,
      "step": 70
    },
    {
      "epoch": 0.06218905472636816,
      "grad_norm": 2.5429700413304293,
      "learning_rate": 6.198347107438017e-06,
      "loss": 0.8342,
      "step": 75
    },
    {
      "epoch": 0.06633499170812604,
      "grad_norm": 2.41053818480909,
      "learning_rate": 6.611570247933885e-06,
      "loss": 0.8233,
      "step": 80
    },
    {
      "epoch": 0.07048092868988391,
      "grad_norm": 2.467306342713205,
      "learning_rate": 7.0247933884297525e-06,
      "loss": 0.8191,
      "step": 85
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 2.4370364722694924,
      "learning_rate": 7.43801652892562e-06,
      "loss": 0.8283,
      "step": 90
    },
    {
      "epoch": 0.07877280265339967,
      "grad_norm": 2.3448779174579175,
      "learning_rate": 7.851239669421489e-06,
      "loss": 0.8109,
      "step": 95
    },
    {
      "epoch": 0.08291873963515754,
      "grad_norm": 2.366708369360123,
      "learning_rate": 8.264462809917356e-06,
      "loss": 0.8175,
      "step": 100
    },
    {
      "epoch": 0.08706467661691543,
      "grad_norm": 2.7803642859629445,
      "learning_rate": 8.677685950413224e-06,
      "loss": 0.8078,
      "step": 105
    },
    {
      "epoch": 0.0912106135986733,
      "grad_norm": 2.402987189273672,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.8036,
      "step": 110
    },
    {
      "epoch": 0.09535655058043117,
      "grad_norm": 2.485685260898161,
      "learning_rate": 9.50413223140496e-06,
      "loss": 0.801,
      "step": 115
    },
    {
      "epoch": 0.09950248756218906,
      "grad_norm": 3.0171204848637294,
      "learning_rate": 9.917355371900828e-06,
      "loss": 0.7926,
      "step": 120
    },
    {
      "epoch": 0.10364842454394693,
      "grad_norm": 2.3850376443520602,
      "learning_rate": 9.999664652243188e-06,
      "loss": 0.7988,
      "step": 125
    },
    {
      "epoch": 0.1077943615257048,
      "grad_norm": 2.489350579580748,
      "learning_rate": 9.99830237907608e-06,
      "loss": 0.7869,
      "step": 130
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 2.4001253138099843,
      "learning_rate": 9.995892506564461e-06,
      "loss": 0.7817,
      "step": 135
    },
    {
      "epoch": 0.11608623548922056,
      "grad_norm": 2.4718753541913037,
      "learning_rate": 9.992435539796e-06,
      "loss": 0.7853,
      "step": 140
    },
    {
      "epoch": 0.12023217247097844,
      "grad_norm": 2.4377853310772966,
      "learning_rate": 9.987932203319917e-06,
      "loss": 0.7634,
      "step": 145
    },
    {
      "epoch": 0.12437810945273632,
      "grad_norm": 2.8232961214963024,
      "learning_rate": 9.982383440995146e-06,
      "loss": 0.7688,
      "step": 150
    },
    {
      "epoch": 0.1285240464344942,
      "grad_norm": 2.25234786421426,
      "learning_rate": 9.975790415792497e-06,
      "loss": 0.7707,
      "step": 155
    },
    {
      "epoch": 0.13266998341625208,
      "grad_norm": 2.653089890533132,
      "learning_rate": 9.968154509550914e-06,
      "loss": 0.759,
      "step": 160
    },
    {
      "epoch": 0.13681592039800994,
      "grad_norm": 2.4692695376219946,
      "learning_rate": 9.959477322687852e-06,
      "loss": 0.7678,
      "step": 165
    },
    {
      "epoch": 0.14096185737976782,
      "grad_norm": 2.465268528338133,
      "learning_rate": 9.949760673863846e-06,
      "loss": 0.7512,
      "step": 170
    },
    {
      "epoch": 0.1451077943615257,
      "grad_norm": 2.487803804934397,
      "learning_rate": 9.93900659960133e-06,
      "loss": 0.7366,
      "step": 175
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 2.55421871649192,
      "learning_rate": 9.927217353857809e-06,
      "loss": 0.7463,
      "step": 180
    },
    {
      "epoch": 0.15339966832504145,
      "grad_norm": 2.639039784633591,
      "learning_rate": 9.914395407553444e-06,
      "loss": 0.7443,
      "step": 185
    },
    {
      "epoch": 0.15754560530679934,
      "grad_norm": 2.501978124475521,
      "learning_rate": 9.900543448053164e-06,
      "loss": 0.7197,
      "step": 190
    },
    {
      "epoch": 0.16169154228855723,
      "grad_norm": 2.3337568323918214,
      "learning_rate": 9.885664378603432e-06,
      "loss": 0.7332,
      "step": 195
    },
    {
      "epoch": 0.16583747927031509,
      "grad_norm": 2.378783071529505,
      "learning_rate": 9.869761317723744e-06,
      "loss": 0.7136,
      "step": 200
    },
    {
      "epoch": 0.16998341625207297,
      "grad_norm": 2.277527337184093,
      "learning_rate": 9.85283759855301e-06,
      "loss": 0.7201,
      "step": 205
    },
    {
      "epoch": 0.17412935323383086,
      "grad_norm": 2.2079190034247365,
      "learning_rate": 9.834896768150963e-06,
      "loss": 0.7174,
      "step": 210
    },
    {
      "epoch": 0.17827529021558872,
      "grad_norm": 2.4371753499407434,
      "learning_rate": 9.81594258675473e-06,
      "loss": 0.7341,
      "step": 215
    },
    {
      "epoch": 0.1824212271973466,
      "grad_norm": 2.2055041342497255,
      "learning_rate": 9.795979026990717e-06,
      "loss": 0.7024,
      "step": 220
    },
    {
      "epoch": 0.1865671641791045,
      "grad_norm": 2.406773347160676,
      "learning_rate": 9.775010273041975e-06,
      "loss": 0.7173,
      "step": 225
    },
    {
      "epoch": 0.19071310116086235,
      "grad_norm": 2.2838878169624173,
      "learning_rate": 9.753040719771249e-06,
      "loss": 0.6933,
      "step": 230
    },
    {
      "epoch": 0.19485903814262023,
      "grad_norm": 2.3948306779315756,
      "learning_rate": 9.730074971799837e-06,
      "loss": 0.7093,
      "step": 235
    },
    {
      "epoch": 0.19900497512437812,
      "grad_norm": 2.3419375521446595,
      "learning_rate": 9.706117842542517e-06,
      "loss": 0.7026,
      "step": 240
    },
    {
      "epoch": 0.20315091210613598,
      "grad_norm": 2.386424037340903,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.6928,
      "step": 245
    },
    {
      "epoch": 0.20729684908789386,
      "grad_norm": 2.352824469415509,
      "learning_rate": 9.655249731699973e-06,
      "loss": 0.6758,
      "step": 250
    },
    {
      "epoch": 0.21144278606965175,
      "grad_norm": 2.448238996649246,
      "learning_rate": 9.628349411614503e-06,
      "loss": 0.6846,
      "step": 255
    },
    {
      "epoch": 0.2155887230514096,
      "grad_norm": 2.18778612059073,
      "learning_rate": 9.600479031008072e-06,
      "loss": 0.6783,
      "step": 260
    },
    {
      "epoch": 0.2197346600331675,
      "grad_norm": 2.4786933731659384,
      "learning_rate": 9.571644431262463e-06,
      "loss": 0.6862,
      "step": 265
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 2.1664093593847165,
      "learning_rate": 9.54185165585114e-06,
      "loss": 0.6598,
      "step": 270
    },
    {
      "epoch": 0.22802653399668324,
      "grad_norm": 2.3439695927983024,
      "learning_rate": 9.511106949072588e-06,
      "loss": 0.6538,
      "step": 275
    },
    {
      "epoch": 0.23217247097844113,
      "grad_norm": 2.573211098478694,
      "learning_rate": 9.479416754741577e-06,
      "loss": 0.6619,
      "step": 280
    },
    {
      "epoch": 0.236318407960199,
      "grad_norm": 2.7234227642496616,
      "learning_rate": 9.446787714838579e-06,
      "loss": 0.6562,
      "step": 285
    },
    {
      "epoch": 0.24046434494195687,
      "grad_norm": 2.5205068601141525,
      "learning_rate": 9.413226668117679e-06,
      "loss": 0.6461,
      "step": 290
    },
    {
      "epoch": 0.24461028192371476,
      "grad_norm": 2.229496276338444,
      "learning_rate": 9.37874064867323e-06,
      "loss": 0.6469,
      "step": 295
    },
    {
      "epoch": 0.24875621890547264,
      "grad_norm": 2.3470411543105265,
      "learning_rate": 9.343336884465577e-06,
      "loss": 0.6426,
      "step": 300
    },
    {
      "epoch": 0.25290215588723053,
      "grad_norm": 2.3905607702569704,
      "learning_rate": 9.307022795806125e-06,
      "loss": 0.6278,
      "step": 305
    },
    {
      "epoch": 0.2570480928689884,
      "grad_norm": 2.159892061781749,
      "learning_rate": 9.26980599380213e-06,
      "loss": 0.6404,
      "step": 310
    },
    {
      "epoch": 0.26119402985074625,
      "grad_norm": 2.573673941571869,
      "learning_rate": 9.231694278761455e-06,
      "loss": 0.629,
      "step": 315
    },
    {
      "epoch": 0.26533996683250416,
      "grad_norm": 2.1819450147556494,
      "learning_rate": 9.192695638557723e-06,
      "loss": 0.6297,
      "step": 320
    },
    {
      "epoch": 0.269485903814262,
      "grad_norm": 2.307987631760188,
      "learning_rate": 9.1528182469561e-06,
      "loss": 0.6254,
      "step": 325
    },
    {
      "epoch": 0.2736318407960199,
      "grad_norm": 2.4579207582542906,
      "learning_rate": 9.112070461900178e-06,
      "loss": 0.6322,
      "step": 330
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 2.3690336709142064,
      "learning_rate": 9.070460823760197e-06,
      "loss": 0.6164,
      "step": 335
    },
    {
      "epoch": 0.28192371475953565,
      "grad_norm": 2.7180849820218067,
      "learning_rate": 9.027998053543079e-06,
      "loss": 0.6298,
      "step": 340
    },
    {
      "epoch": 0.2860696517412935,
      "grad_norm": 2.4829451807056118,
      "learning_rate": 8.984691051064576e-06,
      "loss": 0.6162,
      "step": 345
    },
    {
      "epoch": 0.2902155887230514,
      "grad_norm": 2.459002185250392,
      "learning_rate": 8.94054889308395e-06,
      "loss": 0.6045,
      "step": 350
    },
    {
      "epoch": 0.2943615257048093,
      "grad_norm": 2.2572208925258597,
      "learning_rate": 8.895580831401563e-06,
      "loss": 0.6112,
      "step": 355
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 2.276783038981904,
      "learning_rate": 8.849796290919787e-06,
      "loss": 0.6115,
      "step": 360
    },
    {
      "epoch": 0.30265339966832505,
      "grad_norm": 2.415622228387365,
      "learning_rate": 8.803204867667624e-06,
      "loss": 0.5991,
      "step": 365
    },
    {
      "epoch": 0.3067993366500829,
      "grad_norm": 2.045217334174227,
      "learning_rate": 8.755816326789469e-06,
      "loss": 0.5909,
      "step": 370
    },
    {
      "epoch": 0.31094527363184077,
      "grad_norm": 2.3646147099166526,
      "learning_rate": 8.70764060049842e-06,
      "loss": 0.5929,
      "step": 375
    },
    {
      "epoch": 0.3150912106135987,
      "grad_norm": 2.179575894376864,
      "learning_rate": 8.658687785994579e-06,
      "loss": 0.602,
      "step": 380
    },
    {
      "epoch": 0.31923714759535654,
      "grad_norm": 2.303527412372746,
      "learning_rate": 8.608968143348765e-06,
      "loss": 0.5911,
      "step": 385
    },
    {
      "epoch": 0.32338308457711445,
      "grad_norm": 2.184412340951817,
      "learning_rate": 8.558492093352098e-06,
      "loss": 0.5834,
      "step": 390
    },
    {
      "epoch": 0.3275290215588723,
      "grad_norm": 2.2982310068640963,
      "learning_rate": 8.50727021533189e-06,
      "loss": 0.573,
      "step": 395
    },
    {
      "epoch": 0.33167495854063017,
      "grad_norm": 2.3823718371379288,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.5704,
      "step": 400
    },
    {
      "epoch": 0.3358208955223881,
      "grad_norm": 2.308301112102417,
      "learning_rate": 8.402632071874348e-06,
      "loss": 0.5676,
      "step": 405
    },
    {
      "epoch": 0.33996683250414594,
      "grad_norm": 2.1617090999565782,
      "learning_rate": 8.349237737653288e-06,
      "loss": 0.5709,
      "step": 410
    },
    {
      "epoch": 0.3441127694859038,
      "grad_norm": 2.280161952890472,
      "learning_rate": 8.29514143324466e-06,
      "loss": 0.5779,
      "step": 415
    },
    {
      "epoch": 0.3482587064676617,
      "grad_norm": 2.2490286708106577,
      "learning_rate": 8.24035449674863e-06,
      "loss": 0.5564,
      "step": 420
    },
    {
      "epoch": 0.3524046434494196,
      "grad_norm": 2.2807466234509777,
      "learning_rate": 8.184888411015655e-06,
      "loss": 0.5554,
      "step": 425
    },
    {
      "epoch": 0.35655058043117743,
      "grad_norm": 2.180528318028485,
      "learning_rate": 8.128754801239781e-06,
      "loss": 0.5606,
      "step": 430
    },
    {
      "epoch": 0.36069651741293535,
      "grad_norm": 2.3314447428823,
      "learning_rate": 8.071965432522107e-06,
      "loss": 0.5593,
      "step": 435
    },
    {
      "epoch": 0.3648424543946932,
      "grad_norm": 2.2300833962237974,
      "learning_rate": 8.01453220740492e-06,
      "loss": 0.5513,
      "step": 440
    },
    {
      "epoch": 0.36898839137645106,
      "grad_norm": 2.4375150414786044,
      "learning_rate": 7.956467163377037e-06,
      "loss": 0.5451,
      "step": 445
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 2.345552450223764,
      "learning_rate": 7.89778247035085e-06,
      "loss": 0.5513,
      "step": 450
    },
    {
      "epoch": 0.37728026533996684,
      "grad_norm": 2.302225232727345,
      "learning_rate": 7.838490428111625e-06,
      "loss": 0.5454,
      "step": 455
    },
    {
      "epoch": 0.3814262023217247,
      "grad_norm": 2.3901207498612274,
      "learning_rate": 7.77860346373957e-06,
      "loss": 0.5354,
      "step": 460
    },
    {
      "epoch": 0.3855721393034826,
      "grad_norm": 2.1013639490757527,
      "learning_rate": 7.718134129005238e-06,
      "loss": 0.5338,
      "step": 465
    },
    {
      "epoch": 0.38971807628524047,
      "grad_norm": 2.1095521859807604,
      "learning_rate": 7.657095097738793e-06,
      "loss": 0.5281,
      "step": 470
    },
    {
      "epoch": 0.3938640132669983,
      "grad_norm": 2.1090684046109778,
      "learning_rate": 7.59549916317367e-06,
      "loss": 0.5269,
      "step": 475
    },
    {
      "epoch": 0.39800995024875624,
      "grad_norm": 2.1756383908491466,
      "learning_rate": 7.533359235265248e-06,
      "loss": 0.5404,
      "step": 480
    },
    {
      "epoch": 0.4021558872305141,
      "grad_norm": 2.067392945377543,
      "learning_rate": 7.470688337985029e-06,
      "loss": 0.5257,
      "step": 485
    },
    {
      "epoch": 0.40630182421227196,
      "grad_norm": 2.178779204316918,
      "learning_rate": 7.407499606590934e-06,
      "loss": 0.5108,
      "step": 490
    },
    {
      "epoch": 0.41044776119402987,
      "grad_norm": 2.1482896888352303,
      "learning_rate": 7.343806284874268e-06,
      "loss": 0.5195,
      "step": 495
    },
    {
      "epoch": 0.41459369817578773,
      "grad_norm": 2.122877082306712,
      "learning_rate": 7.279621722383939e-06,
      "loss": 0.5238,
      "step": 500
    },
    {
      "epoch": 0.4187396351575456,
      "grad_norm": 2.513475103077436,
      "learning_rate": 7.214959371628522e-06,
      "loss": 0.5239,
      "step": 505
    },
    {
      "epoch": 0.4228855721393035,
      "grad_norm": 2.256666838754752,
      "learning_rate": 7.149832785256718e-06,
      "loss": 0.5111,
      "step": 510
    },
    {
      "epoch": 0.42703150912106136,
      "grad_norm": 2.126071871777049,
      "learning_rate": 7.084255613216855e-06,
      "loss": 0.5109,
      "step": 515
    },
    {
      "epoch": 0.4311774461028192,
      "grad_norm": 2.2958892839783003,
      "learning_rate": 7.018241599895974e-06,
      "loss": 0.5075,
      "step": 520
    },
    {
      "epoch": 0.43532338308457713,
      "grad_norm": 2.1249877107238166,
      "learning_rate": 6.95180458123913e-06,
      "loss": 0.5021,
      "step": 525
    },
    {
      "epoch": 0.439469320066335,
      "grad_norm": 2.268170350249536,
      "learning_rate": 6.8849584818494984e-06,
      "loss": 0.5075,
      "step": 530
    },
    {
      "epoch": 0.44361525704809285,
      "grad_norm": 2.1889766909888304,
      "learning_rate": 6.817717312069913e-06,
      "loss": 0.4951,
      "step": 535
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 2.116309228498843,
      "learning_rate": 6.750095165046415e-06,
      "loss": 0.4993,
      "step": 540
    },
    {
      "epoch": 0.4519071310116086,
      "grad_norm": 2.2014038727267735,
      "learning_rate": 6.682106213774459e-06,
      "loss": 0.5014,
      "step": 545
    },
    {
      "epoch": 0.4560530679933665,
      "grad_norm": 2.1791180363289833,
      "learning_rate": 6.6137647081283776e-06,
      "loss": 0.5004,
      "step": 550
    },
    {
      "epoch": 0.4601990049751244,
      "grad_norm": 2.181037493892715,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.4869,
      "step": 555
    },
    {
      "epoch": 0.46434494195688225,
      "grad_norm": 2.1234780810442317,
      "learning_rate": 6.476081399670212e-06,
      "loss": 0.487,
      "step": 560
    },
    {
      "epoch": 0.4684908789386401,
      "grad_norm": 2.1577583891835137,
      "learning_rate": 6.406768454044581e-06,
      "loss": 0.4925,
      "step": 565
    },
    {
      "epoch": 0.472636815920398,
      "grad_norm": 2.1435505057127053,
      "learning_rate": 6.337160662369519e-06,
      "loss": 0.4852,
      "step": 570
    },
    {
      "epoch": 0.4767827529021559,
      "grad_norm": 2.1004015267155056,
      "learning_rate": 6.267272613813789e-06,
      "loss": 0.4894,
      "step": 575
    },
    {
      "epoch": 0.48092868988391374,
      "grad_norm": 2.275508822904309,
      "learning_rate": 6.19711895628548e-06,
      "loss": 0.4783,
      "step": 580
    },
    {
      "epoch": 0.48507462686567165,
      "grad_norm": 2.0957567591112456,
      "learning_rate": 6.126714393361939e-06,
      "loss": 0.4811,
      "step": 585
    },
    {
      "epoch": 0.4892205638474295,
      "grad_norm": 2.2453086852427138,
      "learning_rate": 6.056073681208038e-06,
      "loss": 0.4778,
      "step": 590
    },
    {
      "epoch": 0.49336650082918737,
      "grad_norm": 2.086146491253835,
      "learning_rate": 5.985211625483403e-06,
      "loss": 0.468,
      "step": 595
    },
    {
      "epoch": 0.4975124378109453,
      "grad_norm": 2.08746099925638,
      "learning_rate": 5.914143078239293e-06,
      "loss": 0.4702,
      "step": 600
    },
    {
      "epoch": 0.5016583747927031,
      "grad_norm": 2.183555664254409,
      "learning_rate": 5.842882934805731e-06,
      "loss": 0.4683,
      "step": 605
    },
    {
      "epoch": 0.5058043117744611,
      "grad_norm": 2.141477284770924,
      "learning_rate": 5.771446130669589e-06,
      "loss": 0.4692,
      "step": 610
    },
    {
      "epoch": 0.5099502487562189,
      "grad_norm": 1.999530939511801,
      "learning_rate": 5.6998476383442345e-06,
      "loss": 0.4672,
      "step": 615
    },
    {
      "epoch": 0.5140961857379768,
      "grad_norm": 2.3861168673535476,
      "learning_rate": 5.628102464231429e-06,
      "loss": 0.4647,
      "step": 620
    },
    {
      "epoch": 0.5182421227197347,
      "grad_norm": 2.112621163685018,
      "learning_rate": 5.556225645476119e-06,
      "loss": 0.4569,
      "step": 625
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 2.2162792041456023,
      "learning_rate": 5.4842322468147926e-06,
      "loss": 0.4636,
      "step": 630
    },
    {
      "epoch": 0.5265339966832504,
      "grad_norm": 2.1411819631797444,
      "learning_rate": 5.412137357418037e-06,
      "loss": 0.4609,
      "step": 635
    },
    {
      "epoch": 0.5306799336650083,
      "grad_norm": 2.0506863143987277,
      "learning_rate": 5.339956087727985e-06,
      "loss": 0.4501,
      "step": 640
    },
    {
      "epoch": 0.5348258706467661,
      "grad_norm": 2.0133156015781193,
      "learning_rate": 5.2677035662913116e-06,
      "loss": 0.4507,
      "step": 645
    },
    {
      "epoch": 0.538971807628524,
      "grad_norm": 2.0603049501006336,
      "learning_rate": 5.195394936588409e-06,
      "loss": 0.441,
      "step": 650
    },
    {
      "epoch": 0.543117744610282,
      "grad_norm": 2.012154333900926,
      "learning_rate": 5.123045353859465e-06,
      "loss": 0.4577,
      "step": 655
    },
    {
      "epoch": 0.5472636815920398,
      "grad_norm": 2.0263592161269184,
      "learning_rate": 5.050669981928056e-06,
      "loss": 0.4568,
      "step": 660
    },
    {
      "epoch": 0.5514096185737977,
      "grad_norm": 2.0917399842428157,
      "learning_rate": 4.978283990022936e-06,
      "loss": 0.4532,
      "step": 665
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 2.188577844097009,
      "learning_rate": 4.905902549598719e-06,
      "loss": 0.4382,
      "step": 670
    },
    {
      "epoch": 0.5597014925373134,
      "grad_norm": 2.199659099615686,
      "learning_rate": 4.833540831156062e-06,
      "loss": 0.4437,
      "step": 675
    },
    {
      "epoch": 0.5638474295190713,
      "grad_norm": 2.0640167310940027,
      "learning_rate": 4.761214001062079e-06,
      "loss": 0.4379,
      "step": 680
    },
    {
      "epoch": 0.5679933665008292,
      "grad_norm": 2.179786737229522,
      "learning_rate": 4.688937218371592e-06,
      "loss": 0.4409,
      "step": 685
    },
    {
      "epoch": 0.572139303482587,
      "grad_norm": 2.074519322035315,
      "learning_rate": 4.616725631649938e-06,
      "loss": 0.4383,
      "step": 690
    },
    {
      "epoch": 0.5762852404643449,
      "grad_norm": 2.1399873216394036,
      "learning_rate": 4.544594375797969e-06,
      "loss": 0.4288,
      "step": 695
    },
    {
      "epoch": 0.5804311774461028,
      "grad_norm": 2.1123595173850935,
      "learning_rate": 4.472558568879901e-06,
      "loss": 0.4351,
      "step": 700
    },
    {
      "epoch": 0.5845771144278606,
      "grad_norm": 2.031607056740364,
      "learning_rate": 4.400633308954713e-06,
      "loss": 0.4244,
      "step": 705
    },
    {
      "epoch": 0.5887230514096186,
      "grad_norm": 1.9626506911231658,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.4192,
      "step": 710
    },
    {
      "epoch": 0.5928689883913765,
      "grad_norm": 2.10360277006623,
      "learning_rate": 4.257174703311032e-06,
      "loss": 0.4191,
      "step": 715
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 2.008208706865011,
      "learning_rate": 4.185671425229477e-06,
      "loss": 0.4191,
      "step": 720
    },
    {
      "epoch": 0.6011608623548922,
      "grad_norm": 2.0304909176887977,
      "learning_rate": 4.11433882311277e-06,
      "loss": 0.4193,
      "step": 725
    },
    {
      "epoch": 0.6053067993366501,
      "grad_norm": 2.0261313089901303,
      "learning_rate": 4.043191847634469e-06,
      "loss": 0.4164,
      "step": 730
    },
    {
      "epoch": 0.6094527363184079,
      "grad_norm": 2.0660700135752945,
      "learning_rate": 3.9722454105624545e-06,
      "loss": 0.4121,
      "step": 735
    },
    {
      "epoch": 0.6135986733001658,
      "grad_norm": 1.9536775589165198,
      "learning_rate": 3.901514381633555e-06,
      "loss": 0.4184,
      "step": 740
    },
    {
      "epoch": 0.6177446102819237,
      "grad_norm": 2.096984004473796,
      "learning_rate": 3.831013585436985e-06,
      "loss": 0.4209,
      "step": 745
    },
    {
      "epoch": 0.6218905472636815,
      "grad_norm": 1.9585159850477776,
      "learning_rate": 3.7607577983072486e-06,
      "loss": 0.4076,
      "step": 750
    },
    {
      "epoch": 0.6260364842454395,
      "grad_norm": 2.021551204900084,
      "learning_rate": 3.6907617452271394e-06,
      "loss": 0.4054,
      "step": 755
    },
    {
      "epoch": 0.6301824212271974,
      "grad_norm": 2.2587727736565095,
      "learning_rate": 3.621040096741526e-06,
      "loss": 0.4024,
      "step": 760
    },
    {
      "epoch": 0.6343283582089553,
      "grad_norm": 2.1288856262867517,
      "learning_rate": 3.55160746588254e-06,
      "loss": 0.4071,
      "step": 765
    },
    {
      "epoch": 0.6384742951907131,
      "grad_norm": 2.0991322193698485,
      "learning_rate": 3.482478405106803e-06,
      "loss": 0.4119,
      "step": 770
    },
    {
      "epoch": 0.642620232172471,
      "grad_norm": 2.1933844049539153,
      "learning_rate": 3.4136674032453787e-06,
      "loss": 0.4057,
      "step": 775
    },
    {
      "epoch": 0.6467661691542289,
      "grad_norm": 1.9800816023792092,
      "learning_rate": 3.34518888246703e-06,
      "loss": 0.4081,
      "step": 780
    },
    {
      "epoch": 0.6509121061359867,
      "grad_norm": 2.134679203262353,
      "learning_rate": 3.2770571952554674e-06,
      "loss": 0.4055,
      "step": 785
    },
    {
      "epoch": 0.6550580431177446,
      "grad_norm": 2.0810403529065824,
      "learning_rate": 3.2092866214011984e-06,
      "loss": 0.3953,
      "step": 790
    },
    {
      "epoch": 0.6592039800995025,
      "grad_norm": 1.9681607355510164,
      "learning_rate": 3.141891365008609e-06,
      "loss": 0.4043,
      "step": 795
    },
    {
      "epoch": 0.6633499170812603,
      "grad_norm": 2.0452177198825683,
      "learning_rate": 3.0748855515189104e-06,
      "loss": 0.4,
      "step": 800
    },
    {
      "epoch": 0.6674958540630183,
      "grad_norm": 2.0817931631950914,
      "learning_rate": 3.00828322474958e-06,
      "loss": 0.3938,
      "step": 805
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 2.0579281745324876,
      "learning_rate": 2.942098343950891e-06,
      "loss": 0.3848,
      "step": 810
    },
    {
      "epoch": 0.675787728026534,
      "grad_norm": 2.0378046645796446,
      "learning_rate": 2.8763447808801914e-06,
      "loss": 0.3895,
      "step": 815
    },
    {
      "epoch": 0.6799336650082919,
      "grad_norm": 2.089276342993108,
      "learning_rate": 2.8110363168944976e-06,
      "loss": 0.3947,
      "step": 820
    },
    {
      "epoch": 0.6840796019900498,
      "grad_norm": 1.9512325130647332,
      "learning_rate": 2.7461866400620506e-06,
      "loss": 0.3981,
      "step": 825
    },
    {
      "epoch": 0.6882255389718076,
      "grad_norm": 2.006701131264969,
      "learning_rate": 2.6818093422934254e-06,
      "loss": 0.3908,
      "step": 830
    },
    {
      "epoch": 0.6923714759535655,
      "grad_norm": 2.134162418048178,
      "learning_rate": 2.617917916492776e-06,
      "loss": 0.3922,
      "step": 835
    },
    {
      "epoch": 0.6965174129353234,
      "grad_norm": 2.1121757305097226,
      "learning_rate": 2.5545257537298497e-06,
      "loss": 0.3903,
      "step": 840
    },
    {
      "epoch": 0.7006633499170812,
      "grad_norm": 2.06027998760199,
      "learning_rate": 2.491646140433346e-06,
      "loss": 0.3795,
      "step": 845
    },
    {
      "epoch": 0.7048092868988391,
      "grad_norm": 2.1221824714111035,
      "learning_rate": 2.4292922556061877e-06,
      "loss": 0.3891,
      "step": 850
    },
    {
      "epoch": 0.7089552238805971,
      "grad_norm": 2.0687201549733905,
      "learning_rate": 2.367477168063326e-06,
      "loss": 0.3804,
      "step": 855
    },
    {
      "epoch": 0.7131011608623549,
      "grad_norm": 2.1096088311491323,
      "learning_rate": 2.3062138336926406e-06,
      "loss": 0.3861,
      "step": 860
    },
    {
      "epoch": 0.7172470978441128,
      "grad_norm": 1.9369468070487719,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.38,
      "step": 865
    },
    {
      "epoch": 0.7213930348258707,
      "grad_norm": 2.0347831655952198,
      "learning_rate": 2.185393667115513e-06,
      "loss": 0.3697,
      "step": 870
    },
    {
      "epoch": 0.7255389718076285,
      "grad_norm": 2.0345030208735735,
      "learning_rate": 2.125862157732245e-06,
      "loss": 0.3686,
      "step": 875
    },
    {
      "epoch": 0.7296849087893864,
      "grad_norm": 1.9521605057602363,
      "learning_rate": 2.066933041860059e-06,
      "loss": 0.3804,
      "step": 880
    },
    {
      "epoch": 0.7338308457711443,
      "grad_norm": 2.0778756164223364,
      "learning_rate": 2.0086186705130545e-06,
      "loss": 0.3804,
      "step": 885
    },
    {
      "epoch": 0.7379767827529021,
      "grad_norm": 2.1591475619282847,
      "learning_rate": 1.9509312658603954e-06,
      "loss": 0.3752,
      "step": 890
    },
    {
      "epoch": 0.74212271973466,
      "grad_norm": 2.0870382174991993,
      "learning_rate": 1.8938829186646484e-06,
      "loss": 0.3705,
      "step": 895
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 1.9907262732293516,
      "learning_rate": 1.8374855857476687e-06,
      "loss": 0.3646,
      "step": 900
    },
    {
      "epoch": 0.7504145936981758,
      "grad_norm": 2.112014948477697,
      "learning_rate": 1.7817510874845585e-06,
      "loss": 0.378,
      "step": 905
    },
    {
      "epoch": 0.7545605306799337,
      "grad_norm": 2.00612301158268,
      "learning_rate": 1.7266911053262196e-06,
      "loss": 0.3722,
      "step": 910
    },
    {
      "epoch": 0.7587064676616916,
      "grad_norm": 1.9805581686834959,
      "learning_rate": 1.6723171793510363e-06,
      "loss": 0.3571,
      "step": 915
    },
    {
      "epoch": 0.7628524046434494,
      "grad_norm": 2.1705390220024876,
      "learning_rate": 1.6186407058461622e-06,
      "loss": 0.3717,
      "step": 920
    },
    {
      "epoch": 0.7669983416252073,
      "grad_norm": 1.997191295027296,
      "learning_rate": 1.5656729349189742e-06,
      "loss": 0.3702,
      "step": 925
    },
    {
      "epoch": 0.7711442786069652,
      "grad_norm": 1.9480055411783619,
      "learning_rate": 1.5134249681391416e-06,
      "loss": 0.3662,
      "step": 930
    },
    {
      "epoch": 0.775290215588723,
      "grad_norm": 2.1202087648855708,
      "learning_rate": 1.4619077562118477e-06,
      "loss": 0.3624,
      "step": 935
    },
    {
      "epoch": 0.7794361525704809,
      "grad_norm": 2.0287381557784436,
      "learning_rate": 1.411132096682606e-06,
      "loss": 0.3647,
      "step": 940
    },
    {
      "epoch": 0.7835820895522388,
      "grad_norm": 2.1095441746045074,
      "learning_rate": 1.3611086316742057e-06,
      "loss": 0.3666,
      "step": 945
    },
    {
      "epoch": 0.7877280265339967,
      "grad_norm": 2.0271810521548077,
      "learning_rate": 1.3118478456562073e-06,
      "loss": 0.3644,
      "step": 950
    },
    {
      "epoch": 0.7918739635157546,
      "grad_norm": 2.017879350372662,
      "learning_rate": 1.2633600632474962e-06,
      "loss": 0.3555,
      "step": 955
    },
    {
      "epoch": 0.7960199004975125,
      "grad_norm": 2.0158894267700416,
      "learning_rate": 1.2156554470523364e-06,
      "loss": 0.3579,
      "step": 960
    },
    {
      "epoch": 0.8001658374792703,
      "grad_norm": 2.135171063836108,
      "learning_rate": 1.1687439955303764e-06,
      "loss": 0.3549,
      "step": 965
    },
    {
      "epoch": 0.8043117744610282,
      "grad_norm": 2.051748632880903,
      "learning_rate": 1.1226355409010686e-06,
      "loss": 0.3567,
      "step": 970
    },
    {
      "epoch": 0.8084577114427861,
      "grad_norm": 1.9895182337910944,
      "learning_rate": 1.0773397470829145e-06,
      "loss": 0.3506,
      "step": 975
    },
    {
      "epoch": 0.8126036484245439,
      "grad_norm": 2.0734006194975025,
      "learning_rate": 1.032866107667999e-06,
      "loss": 0.3527,
      "step": 980
    },
    {
      "epoch": 0.8167495854063018,
      "grad_norm": 1.9987662217078168,
      "learning_rate": 9.892239439322243e-07,
      "loss": 0.3581,
      "step": 985
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 2.0823516989471083,
      "learning_rate": 9.464224028816427e-07,
      "loss": 0.3665,
      "step": 990
    },
    {
      "epoch": 0.8250414593698175,
      "grad_norm": 2.051484969572584,
      "learning_rate": 9.044704553353323e-07,
      "loss": 0.3537,
      "step": 995
    },
    {
      "epoch": 0.8291873963515755,
      "grad_norm": 2.019033384913334,
      "learning_rate": 8.633768940451981e-07,
      "loss": 0.3484,
      "step": 1000
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 2.071090455759551,
      "learning_rate": 8.231503318530814e-07,
      "loss": 0.3515,
      "step": 1005
    },
    {
      "epoch": 0.8374792703150912,
      "grad_norm": 1.987573001153748,
      "learning_rate": 7.837991998855899e-07,
      "loss": 0.3458,
      "step": 1010
    },
    {
      "epoch": 0.8416252072968491,
      "grad_norm": 2.1342753070628544,
      "learning_rate": 7.453317457870096e-07,
      "loss": 0.3481,
      "step": 1015
    },
    {
      "epoch": 0.845771144278607,
      "grad_norm": 1.9443796196582224,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.3468,
      "step": 1020
    },
    {
      "epoch": 0.8499170812603648,
      "grad_norm": 2.1533371687865737,
      "learning_rate": 6.710799340291341e-07,
      "loss": 0.3516,
      "step": 1025
    },
    {
      "epoch": 0.8540630182421227,
      "grad_norm": 2.046122945823328,
      "learning_rate": 6.353111388835564e-07,
      "loss": 0.348,
      "step": 1030
    },
    {
      "epoch": 0.8582089552238806,
      "grad_norm": 2.0087071630631192,
      "learning_rate": 6.00457143372557e-07,
      "loss": 0.3464,
      "step": 1035
    },
    {
      "epoch": 0.8623548922056384,
      "grad_norm": 2.031472375135797,
      "learning_rate": 5.665252525809583e-07,
      "loss": 0.351,
      "step": 1040
    },
    {
      "epoch": 0.8665008291873963,
      "grad_norm": 2.085600740397554,
      "learning_rate": 5.335225783287051e-07,
      "loss": 0.3522,
      "step": 1045
    },
    {
      "epoch": 0.8706467661691543,
      "grad_norm": 2.0629885384115543,
      "learning_rate": 5.014560376802913e-07,
      "loss": 0.3507,
      "step": 1050
    },
    {
      "epoch": 0.8747927031509121,
      "grad_norm": 2.0514710921617265,
      "learning_rate": 4.703323514950042e-07,
      "loss": 0.3414,
      "step": 1055
    },
    {
      "epoch": 0.87893864013267,
      "grad_norm": 2.0910097285997415,
      "learning_rate": 4.401580430182928e-07,
      "loss": 0.3432,
      "step": 1060
    },
    {
      "epoch": 0.8830845771144279,
      "grad_norm": 2.007785736762621,
      "learning_rate": 4.1093943651455305e-07,
      "loss": 0.3479,
      "step": 1065
    },
    {
      "epoch": 0.8872305140961857,
      "grad_norm": 2.0849923696855326,
      "learning_rate": 3.826826559416219e-07,
      "loss": 0.3472,
      "step": 1070
    },
    {
      "epoch": 0.8913764510779436,
      "grad_norm": 2.0132819974862843,
      "learning_rate": 3.5539362366724784e-07,
      "loss": 0.3399,
      "step": 1075
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 2.074275526756306,
      "learning_rate": 3.290780592278148e-07,
      "loss": 0.3416,
      "step": 1080
    },
    {
      "epoch": 0.8996683250414593,
      "grad_norm": 2.05724879908193,
      "learning_rate": 3.0374147812958387e-07,
      "loss": 0.3505,
      "step": 1085
    },
    {
      "epoch": 0.9038142620232172,
      "grad_norm": 2.157618991705193,
      "learning_rate": 2.7938919069268654e-07,
      "loss": 0.3505,
      "step": 1090
    },
    {
      "epoch": 0.9079601990049752,
      "grad_norm": 2.1499572962789775,
      "learning_rate": 2.5602630093813253e-07,
      "loss": 0.3546,
      "step": 1095
    },
    {
      "epoch": 0.912106135986733,
      "grad_norm": 2.0332963522319485,
      "learning_rate": 2.3365770551805223e-07,
      "loss": 0.3411,
      "step": 1100
    },
    {
      "epoch": 0.9162520729684909,
      "grad_norm": 1.9439734508841187,
      "learning_rate": 2.1228809268940164e-07,
      "loss": 0.3467,
      "step": 1105
    },
    {
      "epoch": 0.9203980099502488,
      "grad_norm": 2.01922575745273,
      "learning_rate": 1.919219413313478e-07,
      "loss": 0.3456,
      "step": 1110
    },
    {
      "epoch": 0.9245439469320066,
      "grad_norm": 2.0723878543911143,
      "learning_rate": 1.725635200065323e-07,
      "loss": 0.3351,
      "step": 1115
    },
    {
      "epoch": 0.9286898839137645,
      "grad_norm": 2.02921530849707,
      "learning_rate": 1.5421688606642392e-07,
      "loss": 0.3399,
      "step": 1120
    },
    {
      "epoch": 0.9328358208955224,
      "grad_norm": 2.0191708205978904,
      "learning_rate": 1.3688588480092913e-07,
      "loss": 0.3401,
      "step": 1125
    },
    {
      "epoch": 0.9369817578772802,
      "grad_norm": 2.1261706152763393,
      "learning_rate": 1.205741486324552e-07,
      "loss": 0.3417,
      "step": 1130
    },
    {
      "epoch": 0.9411276948590381,
      "grad_norm": 2.1058790836653287,
      "learning_rate": 1.0528509635458873e-07,
      "loss": 0.3404,
      "step": 1135
    },
    {
      "epoch": 0.945273631840796,
      "grad_norm": 2.074684841336411,
      "learning_rate": 9.102193241554757e-08,
      "loss": 0.3375,
      "step": 1140
    },
    {
      "epoch": 0.9494195688225538,
      "grad_norm": 1.991625852605994,
      "learning_rate": 7.778764624655433e-08,
      "loss": 0.3322,
      "step": 1145
    },
    {
      "epoch": 0.9535655058043118,
      "grad_norm": 2.104924653809554,
      "learning_rate": 6.558501163527964e-08,
      "loss": 0.336,
      "step": 1150
    },
    {
      "epoch": 0.9577114427860697,
      "grad_norm": 2.0384736104779226,
      "learning_rate": 5.44165861444812e-08,
      "loss": 0.3448,
      "step": 1155
    },
    {
      "epoch": 0.9618573797678275,
      "grad_norm": 2.0936118845033405,
      "learning_rate": 4.428471057596362e-08,
      "loss": 0.3381,
      "step": 1160
    },
    {
      "epoch": 0.9660033167495854,
      "grad_norm": 2.0037182601593466,
      "learning_rate": 3.519150847996422e-08,
      "loss": 0.3457,
      "step": 1165
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 2.09482647445627,
      "learning_rate": 2.713888571007739e-08,
      "loss": 0.3398,
      "step": 1170
    },
    {
      "epoch": 0.9742951907131011,
      "grad_norm": 2.085192366585127,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.3442,
      "step": 1175
    },
    {
      "epoch": 0.978441127694859,
      "grad_norm": 1.9611348789077625,
      "learning_rate": 1.4161910728816009e-08,
      "loss": 0.3373,
      "step": 1180
    },
    {
      "epoch": 0.9825870646766169,
      "grad_norm": 2.109979592147509,
      "learning_rate": 9.240278374995637e-09,
      "loss": 0.3365,
      "step": 1185
    },
    {
      "epoch": 0.9867330016583747,
      "grad_norm": 2.112199186520632,
      "learning_rate": 5.364664492337746e-09,
      "loss": 0.3407,
      "step": 1190
    },
    {
      "epoch": 0.9908789386401327,
      "grad_norm": 2.1961533969052236,
      "learning_rate": 2.5358813747500266e-09,
      "loss": 0.3407,
      "step": 1195
    },
    {
      "epoch": 0.9950248756218906,
      "grad_norm": 2.050107312724138,
      "learning_rate": 7.545219097987444e-10,
      "loss": 0.3484,
      "step": 1200
    },
    {
      "epoch": 0.9991708126036484,
      "grad_norm": 1.9981992682594107,
      "learning_rate": 2.0959454449243076e-11,
      "loss": 0.345,
      "step": 1205
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.3594575822353363,
      "eval_runtime": 0.9533,
      "eval_samples_per_second": 3.147,
      "eval_steps_per_second": 1.049,
      "step": 1206
    },
    {
      "epoch": 1.0,
      "step": 1206,
      "total_flos": 252459519836160.0,
      "train_loss": 0.5262548677858033,
      "train_runtime": 29272.1392,
      "train_samples_per_second": 1.318,
      "train_steps_per_second": 0.041
    }
  ],
  "logging_steps": 5,
  "max_steps": 1206,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 252459519836160.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}