{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 2950,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03389830508474576,
      "grad_norm": 7.922511100769043,
      "learning_rate": 6.081081081081082e-06,
      "loss": 1.3016,
      "step": 10
    },
    {
      "epoch": 0.06779661016949153,
      "grad_norm": 1.777317762374878,
      "learning_rate": 1.2837837837837838e-05,
      "loss": 0.6145,
      "step": 20
    },
    {
      "epoch": 0.1016949152542373,
      "grad_norm": 2.9913039207458496,
      "learning_rate": 1.9594594594594595e-05,
      "loss": 0.2696,
      "step": 30
    },
    {
      "epoch": 0.13559322033898305,
      "grad_norm": 1.0988717079162598,
      "learning_rate": 2.635135135135135e-05,
      "loss": 0.1885,
      "step": 40
    },
    {
      "epoch": 0.1694915254237288,
      "grad_norm": 1.2467349767684937,
      "learning_rate": 3.310810810810811e-05,
      "loss": 0.1516,
      "step": 50
    },
    {
      "epoch": 0.2033898305084746,
      "grad_norm": 1.4213334321975708,
      "learning_rate": 3.986486486486487e-05,
      "loss": 0.1406,
      "step": 60
    },
    {
      "epoch": 0.23728813559322035,
      "grad_norm": 1.0713083744049072,
      "learning_rate": 4.662162162162162e-05,
      "loss": 0.1316,
      "step": 70
    },
    {
      "epoch": 0.2711864406779661,
      "grad_norm": 1.506921410560608,
      "learning_rate": 5.337837837837838e-05,
      "loss": 0.1413,
      "step": 80
    },
    {
      "epoch": 0.3050847457627119,
      "grad_norm": 0.9623873233795166,
      "learning_rate": 6.013513513513514e-05,
      "loss": 0.1204,
      "step": 90
    },
    {
      "epoch": 0.3389830508474576,
      "grad_norm": 1.5513378381729126,
      "learning_rate": 6.68918918918919e-05,
      "loss": 0.1194,
      "step": 100
    },
    {
      "epoch": 0.3728813559322034,
      "grad_norm": 1.0387235879898071,
      "learning_rate": 7.364864864864865e-05,
      "loss": 0.1111,
      "step": 110
    },
    {
      "epoch": 0.4067796610169492,
      "grad_norm": 1.3981765508651733,
      "learning_rate": 8.040540540540541e-05,
      "loss": 0.119,
      "step": 120
    },
    {
      "epoch": 0.4406779661016949,
      "grad_norm": 1.1575907468795776,
      "learning_rate": 8.716216216216216e-05,
      "loss": 0.0932,
      "step": 130
    },
    {
      "epoch": 0.4745762711864407,
      "grad_norm": 0.8324375748634338,
      "learning_rate": 9.391891891891892e-05,
      "loss": 0.0997,
      "step": 140
    },
    {
      "epoch": 0.5084745762711864,
      "grad_norm": 1.0497596263885498,
      "learning_rate": 9.99999685729623e-05,
      "loss": 0.0795,
      "step": 150
    },
    {
      "epoch": 0.5423728813559322,
      "grad_norm": 0.9868571758270264,
      "learning_rate": 9.999619737623951e-05,
      "loss": 0.0759,
      "step": 160
    },
    {
      "epoch": 0.576271186440678,
      "grad_norm": 0.6950494050979614,
      "learning_rate": 9.99861413151757e-05,
      "loss": 0.0647,
      "step": 170
    },
    {
      "epoch": 0.6101694915254238,
      "grad_norm": 0.9549587965011597,
      "learning_rate": 9.99698016538866e-05,
      "loss": 0.0708,
      "step": 180
    },
    {
      "epoch": 0.6440677966101694,
      "grad_norm": 0.6263039112091064,
      "learning_rate": 9.994718044637956e-05,
      "loss": 0.0711,
      "step": 190
    },
    {
      "epoch": 0.6779661016949152,
      "grad_norm": 0.6637316346168518,
      "learning_rate": 9.991828053629519e-05,
      "loss": 0.0635,
      "step": 200
    },
    {
      "epoch": 0.711864406779661,
      "grad_norm": 0.8776123523712158,
      "learning_rate": 9.988310555655009e-05,
      "loss": 0.0609,
      "step": 210
    },
    {
      "epoch": 0.7457627118644068,
      "grad_norm": 0.9095442891120911,
      "learning_rate": 9.984165992888009e-05,
      "loss": 0.0532,
      "step": 220
    },
    {
      "epoch": 0.7796610169491526,
      "grad_norm": 0.8120779395103455,
      "learning_rate": 9.979394886328434e-05,
      "loss": 0.0522,
      "step": 230
    },
    {
      "epoch": 0.8135593220338984,
      "grad_norm": 0.7972075939178467,
      "learning_rate": 9.973997835737049e-05,
      "loss": 0.0564,
      "step": 240
    },
    {
      "epoch": 0.847457627118644,
      "grad_norm": 0.8452187776565552,
      "learning_rate": 9.967975519560067e-05,
      "loss": 0.0522,
      "step": 250
    },
    {
      "epoch": 0.8813559322033898,
      "grad_norm": 0.8456110954284668,
      "learning_rate": 9.961328694843865e-05,
      "loss": 0.049,
      "step": 260
    },
    {
      "epoch": 0.9152542372881356,
      "grad_norm": 0.47546839714050293,
      "learning_rate": 9.954058197139824e-05,
      "loss": 0.0503,
      "step": 270
    },
    {
      "epoch": 0.9491525423728814,
      "grad_norm": 0.7819595336914062,
      "learning_rate": 9.946164940399288e-05,
      "loss": 0.0515,
      "step": 280
    },
    {
      "epoch": 0.9830508474576272,
      "grad_norm": 0.7775952219963074,
      "learning_rate": 9.937649916858673e-05,
      "loss": 0.05,
      "step": 290
    },
    {
      "epoch": 1.0169491525423728,
      "grad_norm": 0.7178511619567871,
      "learning_rate": 9.928514196914744e-05,
      "loss": 0.0529,
      "step": 300
    },
    {
      "epoch": 1.0508474576271187,
      "grad_norm": 0.6638413667678833,
      "learning_rate": 9.91875892899005e-05,
      "loss": 0.053,
      "step": 310
    },
    {
      "epoch": 1.0847457627118644,
      "grad_norm": 0.6948956251144409,
      "learning_rate": 9.908385339388564e-05,
      "loss": 0.0484,
      "step": 320
    },
    {
      "epoch": 1.11864406779661,
      "grad_norm": 0.47976964712142944,
      "learning_rate": 9.897394732141531e-05,
      "loss": 0.0477,
      "step": 330
    },
    {
      "epoch": 1.152542372881356,
      "grad_norm": 0.786380410194397,
      "learning_rate": 9.885788488843535e-05,
      "loss": 0.0464,
      "step": 340
    },
    {
      "epoch": 1.1864406779661016,
      "grad_norm": 0.7638571858406067,
      "learning_rate": 9.873568068478826e-05,
      "loss": 0.0493,
      "step": 350
    },
    {
      "epoch": 1.2203389830508475,
      "grad_norm": 0.8229843974113464,
      "learning_rate": 9.860735007237922e-05,
      "loss": 0.0488,
      "step": 360
    },
    {
      "epoch": 1.2542372881355932,
      "grad_norm": 0.6507530808448792,
      "learning_rate": 9.847290918324487e-05,
      "loss": 0.0442,
      "step": 370
    },
    {
      "epoch": 1.288135593220339,
      "grad_norm": 0.6550291180610657,
      "learning_rate": 9.833237491752557e-05,
      "loss": 0.0422,
      "step": 380
    },
    {
      "epoch": 1.3220338983050848,
      "grad_norm": 0.4493582844734192,
      "learning_rate": 9.818576494134071e-05,
      "loss": 0.0436,
      "step": 390
    },
    {
      "epoch": 1.3559322033898304,
      "grad_norm": 0.4366756081581116,
      "learning_rate": 9.803309768456819e-05,
      "loss": 0.0483,
      "step": 400
    },
    {
      "epoch": 1.3898305084745763,
      "grad_norm": 0.518522322177887,
      "learning_rate": 9.787439233852752e-05,
      "loss": 0.04,
      "step": 410
    },
    {
      "epoch": 1.423728813559322,
      "grad_norm": 0.5709708333015442,
      "learning_rate": 9.770966885356736e-05,
      "loss": 0.0389,
      "step": 420
    },
    {
      "epoch": 1.457627118644068,
      "grad_norm": 0.44815006852149963,
      "learning_rate": 9.753894793655767e-05,
      "loss": 0.0414,
      "step": 430
    },
    {
      "epoch": 1.4915254237288136,
      "grad_norm": 0.6205833554267883,
      "learning_rate": 9.736225104828666e-05,
      "loss": 0.0488,
      "step": 440
    },
    {
      "epoch": 1.5254237288135593,
      "grad_norm": 0.38849759101867676,
      "learning_rate": 9.717960040076305e-05,
      "loss": 0.0419,
      "step": 450
    },
    {
      "epoch": 1.559322033898305,
      "grad_norm": 0.3880470395088196,
      "learning_rate": 9.699101895442388e-05,
      "loss": 0.0438,
      "step": 460
    },
    {
      "epoch": 1.5932203389830508,
      "grad_norm": 0.7112048268318176,
      "learning_rate": 9.679653041524818e-05,
      "loss": 0.0424,
      "step": 470
    },
    {
      "epoch": 1.6271186440677967,
      "grad_norm": 0.8996948003768921,
      "learning_rate": 9.659615923177702e-05,
      "loss": 0.0441,
      "step": 480
    },
    {
      "epoch": 1.6610169491525424,
      "grad_norm": 0.6175360083580017,
      "learning_rate": 9.638993059204016e-05,
      "loss": 0.0429,
      "step": 490
    },
    {
      "epoch": 1.694915254237288,
      "grad_norm": 0.39414605498313904,
      "learning_rate": 9.617787042038964e-05,
      "loss": 0.0377,
      "step": 500
    },
    {
      "epoch": 1.7288135593220337,
      "grad_norm": 0.5390213131904602,
      "learning_rate": 9.596000537424107e-05,
      "loss": 0.0407,
      "step": 510
    },
    {
      "epoch": 1.7627118644067796,
      "grad_norm": 0.4397003948688507,
      "learning_rate": 9.573636284072252e-05,
      "loss": 0.0386,
      "step": 520
    },
    {
      "epoch": 1.7966101694915255,
      "grad_norm": 0.6200686097145081,
      "learning_rate": 9.550697093323173e-05,
      "loss": 0.0392,
      "step": 530
    },
    {
      "epoch": 1.8305084745762712,
      "grad_norm": 0.5365989804267883,
      "learning_rate": 9.527185848790216e-05,
      "loss": 0.0355,
      "step": 540
    },
    {
      "epoch": 1.8644067796610169,
      "grad_norm": 0.4744288921356201,
      "learning_rate": 9.503105505997805e-05,
      "loss": 0.0387,
      "step": 550
    },
    {
      "epoch": 1.8983050847457628,
      "grad_norm": 0.5821052193641663,
      "learning_rate": 9.478459092009909e-05,
      "loss": 0.0392,
      "step": 560
    },
    {
      "epoch": 1.9322033898305084,
      "grad_norm": 0.5957028269767761,
      "learning_rate": 9.453249705049523e-05,
      "loss": 0.0392,
      "step": 570
    },
    {
      "epoch": 1.9661016949152543,
      "grad_norm": 0.54462730884552,
      "learning_rate": 9.427480514109202e-05,
      "loss": 0.0355,
      "step": 580
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2646660804748535,
      "learning_rate": 9.401154758552691e-05,
      "loss": 0.045,
      "step": 590
    },
    {
      "epoch": 2.0338983050847457,
      "grad_norm": 0.4895438551902771,
      "learning_rate": 9.37427574770772e-05,
      "loss": 0.0394,
      "step": 600
    },
    {
      "epoch": 2.0677966101694913,
      "grad_norm": 0.7774738669395447,
      "learning_rate": 9.346846860449996e-05,
      "loss": 0.0412,
      "step": 610
    },
    {
      "epoch": 2.1016949152542375,
      "grad_norm": 0.7907915115356445,
      "learning_rate": 9.318871544778457e-05,
      "loss": 0.0434,
      "step": 620
    },
    {
      "epoch": 2.135593220338983,
      "grad_norm": 0.5634192228317261,
      "learning_rate": 9.290353317381834e-05,
      "loss": 0.0445,
      "step": 630
    },
    {
      "epoch": 2.169491525423729,
      "grad_norm": 0.524152398109436,
      "learning_rate": 9.261295763196579e-05,
      "loss": 0.0347,
      "step": 640
    },
    {
      "epoch": 2.2033898305084745,
      "grad_norm": 0.48940563201904297,
      "learning_rate": 9.231702534956213e-05,
      "loss": 0.0387,
      "step": 650
    },
    {
      "epoch": 2.23728813559322,
      "grad_norm": 0.4866064488887787,
      "learning_rate": 9.201577352732157e-05,
      "loss": 0.0386,
      "step": 660
    },
    {
      "epoch": 2.2711864406779663,
      "grad_norm": 0.379842609167099,
      "learning_rate": 9.170924003466086e-05,
      "loss": 0.0341,
      "step": 670
    },
    {
      "epoch": 2.305084745762712,
      "grad_norm": 0.41267237067222595,
      "learning_rate": 9.13974634049389e-05,
      "loss": 0.0379,
      "step": 680
    },
    {
      "epoch": 2.3389830508474576,
      "grad_norm": 0.6153244376182556,
      "learning_rate": 9.108048283061283e-05,
      "loss": 0.0331,
      "step": 690
    },
    {
      "epoch": 2.3728813559322033,
      "grad_norm": 0.48824194073677063,
      "learning_rate": 9.075833815831128e-05,
      "loss": 0.0448,
      "step": 700
    },
    {
      "epoch": 2.406779661016949,
      "grad_norm": 0.47475725412368774,
      "learning_rate": 9.043106988382526e-05,
      "loss": 0.0373,
      "step": 710
    },
    {
      "epoch": 2.440677966101695,
      "grad_norm": 0.44530612230300903,
      "learning_rate": 9.009871914701779e-05,
      "loss": 0.036,
      "step": 720
    },
    {
      "epoch": 2.4745762711864407,
      "grad_norm": 0.704857349395752,
      "learning_rate": 8.976132772665211e-05,
      "loss": 0.0405,
      "step": 730
    },
    {
      "epoch": 2.5084745762711864,
      "grad_norm": 0.5021880269050598,
      "learning_rate": 8.941893803513994e-05,
      "loss": 0.039,
      "step": 740
    },
    {
      "epoch": 2.542372881355932,
      "grad_norm": 0.43946319818496704,
      "learning_rate": 8.907159311320985e-05,
      "loss": 0.0377,
      "step": 750
    },
    {
      "epoch": 2.576271186440678,
      "grad_norm": 0.4081028997898102,
      "learning_rate": 8.87193366244969e-05,
      "loss": 0.0337,
      "step": 760
    },
    {
      "epoch": 2.610169491525424,
      "grad_norm": 0.42831340432167053,
      "learning_rate": 8.836221285005364e-05,
      "loss": 0.0331,
      "step": 770
    },
    {
      "epoch": 2.6440677966101696,
      "grad_norm": 0.5115790963172913,
      "learning_rate": 8.800026668278378e-05,
      "loss": 0.0383,
      "step": 780
    },
    {
      "epoch": 2.6779661016949152,
      "grad_norm": 0.7489039897918701,
      "learning_rate": 8.763354362179887e-05,
      "loss": 0.037,
      "step": 790
    },
    {
      "epoch": 2.711864406779661,
      "grad_norm": 0.47204285860061646,
      "learning_rate": 8.726208976669868e-05,
      "loss": 0.0366,
      "step": 800
    },
    {
      "epoch": 2.7457627118644066,
      "grad_norm": 0.3575058579444885,
      "learning_rate": 8.688595181177617e-05,
      "loss": 0.03,
      "step": 810
    },
    {
      "epoch": 2.7796610169491527,
      "grad_norm": 0.3760571777820587,
      "learning_rate": 8.650517704014773e-05,
      "loss": 0.0376,
      "step": 820
    },
    {
      "epoch": 2.8135593220338984,
      "grad_norm": 0.3222888708114624,
      "learning_rate": 8.611981331780942e-05,
      "loss": 0.0373,
      "step": 830
    },
    {
      "epoch": 2.847457627118644,
      "grad_norm": 0.3633357286453247,
      "learning_rate": 8.572990908761974e-05,
      "loss": 0.0358,
      "step": 840
    },
    {
      "epoch": 2.8813559322033897,
      "grad_norm": 0.3399844765663147,
      "learning_rate": 8.533551336321015e-05,
      "loss": 0.0359,
      "step": 850
    },
    {
      "epoch": 2.915254237288136,
      "grad_norm": 0.7424901127815247,
      "learning_rate": 8.493667572282371e-05,
      "loss": 0.0338,
      "step": 860
    },
    {
      "epoch": 2.9491525423728815,
      "grad_norm": 0.4179316759109497,
      "learning_rate": 8.453344630308269e-05,
      "loss": 0.0358,
      "step": 870
    },
    {
      "epoch": 2.983050847457627,
      "grad_norm": 0.4843442738056183,
      "learning_rate": 8.412587579268613e-05,
      "loss": 0.0353,
      "step": 880
    },
    {
      "epoch": 3.016949152542373,
      "grad_norm": 0.5782058835029602,
      "learning_rate": 8.371401542603791e-05,
      "loss": 0.0373,
      "step": 890
    },
    {
      "epoch": 3.0508474576271185,
      "grad_norm": 0.5099455714225769,
      "learning_rate": 8.329791697680612e-05,
      "loss": 0.0336,
      "step": 900
    },
    {
      "epoch": 3.084745762711864,
      "grad_norm": 0.31090185046195984,
      "learning_rate": 8.287763275141496e-05,
      "loss": 0.0326,
      "step": 910
    },
    {
      "epoch": 3.1186440677966103,
      "grad_norm": 0.22995054721832275,
      "learning_rate": 8.245321558246928e-05,
      "loss": 0.0342,
      "step": 920
    },
    {
      "epoch": 3.152542372881356,
      "grad_norm": 0.46085304021835327,
      "learning_rate": 8.202471882211327e-05,
      "loss": 0.0314,
      "step": 930
    },
    {
      "epoch": 3.1864406779661016,
      "grad_norm": 0.47579386830329895,
      "learning_rate": 8.159219633532368e-05,
      "loss": 0.0335,
      "step": 940
    },
    {
      "epoch": 3.2203389830508473,
      "grad_norm": 0.3809256851673126,
      "learning_rate": 8.115570249313864e-05,
      "loss": 0.0322,
      "step": 950
    },
    {
      "epoch": 3.2542372881355934,
      "grad_norm": 0.38978174328804016,
      "learning_rate": 8.071529216582284e-05,
      "loss": 0.0313,
      "step": 960
    },
    {
      "epoch": 3.288135593220339,
      "grad_norm": 0.3661576211452484,
      "learning_rate": 8.027102071596999e-05,
      "loss": 0.0378,
      "step": 970
    },
    {
      "epoch": 3.3220338983050848,
      "grad_norm": 0.28337371349334717,
      "learning_rate": 7.982294399154331e-05,
      "loss": 0.0338,
      "step": 980
    },
    {
      "epoch": 3.3559322033898304,
      "grad_norm": 0.48508763313293457,
      "learning_rate": 7.93711183188551e-05,
      "loss": 0.0374,
      "step": 990
    },
    {
      "epoch": 3.389830508474576,
      "grad_norm": 0.4197925925254822,
      "learning_rate": 7.891560049548614e-05,
      "loss": 0.0336,
      "step": 1000
    },
    {
      "epoch": 3.423728813559322,
      "grad_norm": 0.4868261516094208,
      "learning_rate": 7.845644778314583e-05,
      "loss": 0.0396,
      "step": 1010
    },
    {
      "epoch": 3.457627118644068,
      "grad_norm": 0.264763206243515,
      "learning_rate": 7.799371790047395e-05,
      "loss": 0.0317,
      "step": 1020
    },
    {
      "epoch": 3.4915254237288136,
      "grad_norm": 0.2507479190826416,
      "learning_rate": 7.752746901578517e-05,
      "loss": 0.0308,
      "step": 1030
    },
    {
      "epoch": 3.5254237288135593,
      "grad_norm": 0.40973591804504395,
      "learning_rate": 7.705775973975675e-05,
      "loss": 0.035,
      "step": 1040
    },
    {
      "epoch": 3.559322033898305,
      "grad_norm": 0.34160512685775757,
      "learning_rate": 7.658464911806087e-05,
      "loss": 0.0287,
      "step": 1050
    },
    {
      "epoch": 3.593220338983051,
      "grad_norm": 0.3527885973453522,
      "learning_rate": 7.61081966239421e-05,
      "loss": 0.026,
      "step": 1060
    },
    {
      "epoch": 3.6271186440677967,
      "grad_norm": 0.4634409248828888,
      "learning_rate": 7.562846215074129e-05,
      "loss": 0.03,
      "step": 1070
    },
    {
      "epoch": 3.6610169491525424,
      "grad_norm": 0.3833574652671814,
      "learning_rate": 7.514550600436657e-05,
      "loss": 0.032,
      "step": 1080
    },
    {
      "epoch": 3.694915254237288,
      "grad_norm": 0.4073950946331024,
      "learning_rate": 7.465938889571232e-05,
      "loss": 0.0268,
      "step": 1090
    },
    {
      "epoch": 3.7288135593220337,
      "grad_norm": 0.590704083442688,
      "learning_rate": 7.417017193302769e-05,
      "loss": 0.0298,
      "step": 1100
    },
    {
      "epoch": 3.7627118644067794,
      "grad_norm": 0.3835419714450836,
      "learning_rate": 7.367791661423461e-05,
      "loss": 0.0305,
      "step": 1110
    },
    {
      "epoch": 3.7966101694915255,
      "grad_norm": 0.4307284951210022,
      "learning_rate": 7.318268481919716e-05,
      "loss": 0.0312,
      "step": 1120
    },
    {
      "epoch": 3.830508474576271,
      "grad_norm": 0.5295498371124268,
      "learning_rate": 7.268453880194292e-05,
      "loss": 0.0329,
      "step": 1130
    },
    {
      "epoch": 3.864406779661017,
      "grad_norm": 0.5060334801673889,
      "learning_rate": 7.218354118283715e-05,
      "loss": 0.0346,
      "step": 1140
    },
    {
      "epoch": 3.898305084745763,
      "grad_norm": 0.3661726415157318,
      "learning_rate": 7.167975494071093e-05,
      "loss": 0.034,
      "step": 1150
    },
    {
      "epoch": 3.9322033898305087,
      "grad_norm": 0.382707804441452,
      "learning_rate": 7.11732434049444e-05,
      "loss": 0.0314,
      "step": 1160
    },
    {
      "epoch": 3.9661016949152543,
      "grad_norm": 0.44602781534194946,
      "learning_rate": 7.066407024750578e-05,
      "loss": 0.0346,
      "step": 1170
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.8751994371414185,
      "learning_rate": 7.015229947494734e-05,
      "loss": 0.0321,
      "step": 1180
    },
    {
      "epoch": 4.033898305084746,
      "grad_norm": 0.31756699085235596,
      "learning_rate": 6.963799542035932e-05,
      "loss": 0.0319,
      "step": 1190
    },
    {
      "epoch": 4.067796610169491,
      "grad_norm": 0.336243599653244,
      "learning_rate": 6.912122273528291e-05,
      "loss": 0.0296,
      "step": 1200
    },
    {
      "epoch": 4.101694915254237,
      "grad_norm": 0.3985385298728943,
      "learning_rate": 6.860204638158305e-05,
      "loss": 0.028,
      "step": 1210
    },
    {
      "epoch": 4.135593220338983,
      "grad_norm": 0.5406593084335327,
      "learning_rate": 6.808053162328227e-05,
      "loss": 0.034,
      "step": 1220
    },
    {
      "epoch": 4.169491525423728,
      "grad_norm": 0.3683408200740814,
      "learning_rate": 6.755674401835657e-05,
      "loss": 0.0312,
      "step": 1230
    },
    {
      "epoch": 4.203389830508475,
      "grad_norm": 0.5468569993972778,
      "learning_rate": 6.703074941049431e-05,
      "loss": 0.0286,
      "step": 1240
    },
    {
      "epoch": 4.237288135593221,
      "grad_norm": 0.5490782260894775,
      "learning_rate": 6.650261392081929e-05,
      "loss": 0.0322,
      "step": 1250
    },
    {
      "epoch": 4.271186440677966,
      "grad_norm": 0.2997434437274933,
      "learning_rate": 6.597240393957875e-05,
      "loss": 0.0299,
      "step": 1260
    },
    {
      "epoch": 4.305084745762712,
      "grad_norm": 0.43356242775917053,
      "learning_rate": 6.544018611779778e-05,
      "loss": 0.0331,
      "step": 1270
    },
    {
      "epoch": 4.338983050847458,
      "grad_norm": 0.33595308661460876,
      "learning_rate": 6.490602735890071e-05,
      "loss": 0.0275,
      "step": 1280
    },
    {
      "epoch": 4.372881355932203,
      "grad_norm": 0.4883277714252472,
      "learning_rate": 6.436999481030104e-05,
      "loss": 0.0321,
      "step": 1290
    },
    {
      "epoch": 4.406779661016949,
      "grad_norm": 0.25386956334114075,
      "learning_rate": 6.383215585496037e-05,
      "loss": 0.0322,
      "step": 1300
    },
    {
      "epoch": 4.440677966101695,
      "grad_norm": 0.30296579003334045,
      "learning_rate": 6.329257810291806e-05,
      "loss": 0.023,
      "step": 1310
    },
    {
      "epoch": 4.47457627118644,
      "grad_norm": 0.47214221954345703,
      "learning_rate": 6.275132938279213e-05,
      "loss": 0.0248,
      "step": 1320
    },
    {
      "epoch": 4.508474576271187,
      "grad_norm": 0.40692755579948425,
      "learning_rate": 6.220847773325268e-05,
      "loss": 0.0325,
      "step": 1330
    },
    {
      "epoch": 4.5423728813559325,
      "grad_norm": 0.26799410581588745,
      "learning_rate": 6.166409139446903e-05,
      "loss": 0.0305,
      "step": 1340
    },
    {
      "epoch": 4.576271186440678,
      "grad_norm": 0.2558389902114868,
      "learning_rate": 6.111823879953145e-05,
      "loss": 0.0265,
      "step": 1350
    },
    {
      "epoch": 4.610169491525424,
      "grad_norm": 0.28272953629493713,
      "learning_rate": 6.0570988565848685e-05,
      "loss": 0.025,
      "step": 1360
    },
    {
      "epoch": 4.6440677966101696,
      "grad_norm": 0.4986218214035034,
      "learning_rate": 6.002240948652218e-05,
      "loss": 0.0251,
      "step": 1370
    },
    {
      "epoch": 4.677966101694915,
      "grad_norm": 0.5608692169189453,
      "learning_rate": 5.94725705216984e-05,
      "loss": 0.0271,
      "step": 1380
    },
    {
      "epoch": 4.711864406779661,
      "grad_norm": 0.5304986834526062,
      "learning_rate": 5.8921540789900064e-05,
      "loss": 0.0268,
      "step": 1390
    },
    {
      "epoch": 4.745762711864407,
      "grad_norm": 0.45950591564178467,
      "learning_rate": 5.836938955933743e-05,
      "loss": 0.0272,
      "step": 1400
    },
    {
      "epoch": 4.779661016949152,
      "grad_norm": 0.31691136956214905,
      "learning_rate": 5.781618623920081e-05,
      "loss": 0.0312,
      "step": 1410
    },
    {
      "epoch": 4.813559322033898,
      "grad_norm": 0.33825352787971497,
      "learning_rate": 5.726200037093542e-05,
      "loss": 0.0289,
      "step": 1420
    },
    {
      "epoch": 4.847457627118644,
      "grad_norm": 0.31809505820274353,
      "learning_rate": 5.670690161949952e-05,
      "loss": 0.0269,
      "step": 1430
    },
    {
      "epoch": 4.88135593220339,
      "grad_norm": 0.27987194061279297,
      "learning_rate": 5.6150959764606944e-05,
      "loss": 0.026,
      "step": 1440
    },
    {
      "epoch": 4.915254237288136,
      "grad_norm": 0.33458834886550903,
      "learning_rate": 5.559424469195551e-05,
      "loss": 0.0256,
      "step": 1450
    },
    {
      "epoch": 4.9491525423728815,
      "grad_norm": 0.5636442303657532,
      "learning_rate": 5.503682638444173e-05,
      "loss": 0.0279,
      "step": 1460
    },
    {
      "epoch": 4.983050847457627,
      "grad_norm": 0.400234192609787,
      "learning_rate": 5.4478774913363596e-05,
      "loss": 0.0273,
      "step": 1470
    },
    {
      "epoch": 5.016949152542373,
      "grad_norm": 0.5593984127044678,
      "learning_rate": 5.392016042961205e-05,
      "loss": 0.0299,
      "step": 1480
    },
    {
      "epoch": 5.0508474576271185,
      "grad_norm": 0.42886337637901306,
      "learning_rate": 5.336105315485264e-05,
      "loss": 0.0282,
      "step": 1490
    },
    {
      "epoch": 5.084745762711864,
      "grad_norm": 0.4065008759498596,
      "learning_rate": 5.280152337269807e-05,
      "loss": 0.0282,
      "step": 1500
    },
    {
      "epoch": 5.11864406779661,
      "grad_norm": 0.46473270654678345,
      "learning_rate": 5.224164141987313e-05,
      "loss": 0.027,
      "step": 1510
    },
    {
      "epoch": 5.1525423728813555,
      "grad_norm": 0.35750046372413635,
      "learning_rate": 5.168147767737289e-05,
      "loss": 0.0288,
      "step": 1520
    },
    {
      "epoch": 5.186440677966102,
      "grad_norm": 0.3990558385848999,
      "learning_rate": 5.1121102561615295e-05,
      "loss": 0.0294,
      "step": 1530
    },
    {
      "epoch": 5.220338983050848,
      "grad_norm": 0.32793498039245605,
      "learning_rate": 5.056058651558936e-05,
      "loss": 0.028,
      "step": 1540
    },
    {
      "epoch": 5.254237288135593,
      "grad_norm": 0.6837785243988037,
      "learning_rate": 5e-05,
      "loss": 0.0256,
      "step": 1550
    },
    {
      "epoch": 5.288135593220339,
      "grad_norm": 0.49780961871147156,
      "learning_rate": 4.943941348441065e-05,
      "loss": 0.0258,
      "step": 1560
    },
    {
      "epoch": 5.322033898305085,
      "grad_norm": 0.529035210609436,
      "learning_rate": 4.8878897438384716e-05,
      "loss": 0.0281,
      "step": 1570
    },
    {
      "epoch": 5.3559322033898304,
      "grad_norm": 0.27274975180625916,
      "learning_rate": 4.831852232262713e-05,
      "loss": 0.0233,
      "step": 1580
    },
    {
      "epoch": 5.389830508474576,
      "grad_norm": 0.5764862895011902,
      "learning_rate": 4.775835858012689e-05,
      "loss": 0.0235,
      "step": 1590
    },
    {
      "epoch": 5.423728813559322,
      "grad_norm": 0.3733164370059967,
      "learning_rate": 4.7198476627301955e-05,
      "loss": 0.0269,
      "step": 1600
    },
    {
      "epoch": 5.4576271186440675,
      "grad_norm": 0.4069574177265167,
      "learning_rate": 4.663894684514737e-05,
      "loss": 0.0267,
      "step": 1610
    },
    {
      "epoch": 5.491525423728813,
      "grad_norm": 0.37277278304100037,
      "learning_rate": 4.6079839570387954e-05,
      "loss": 0.0239,
      "step": 1620
    },
    {
      "epoch": 5.52542372881356,
      "grad_norm": 0.43740156292915344,
      "learning_rate": 4.552122508663641e-05,
      "loss": 0.0218,
      "step": 1630
    },
    {
      "epoch": 5.559322033898305,
      "grad_norm": 0.4995250105857849,
      "learning_rate": 4.496317361555828e-05,
      "loss": 0.0275,
      "step": 1640
    },
    {
      "epoch": 5.593220338983051,
      "grad_norm": 0.3874273896217346,
      "learning_rate": 4.44057553080445e-05,
      "loss": 0.0218,
      "step": 1650
    },
    {
      "epoch": 5.627118644067797,
      "grad_norm": 0.3819812536239624,
      "learning_rate": 4.384904023539306e-05,
      "loss": 0.0217,
      "step": 1660
    },
    {
      "epoch": 5.661016949152542,
      "grad_norm": 0.44215914607048035,
      "learning_rate": 4.32930983805005e-05,
      "loss": 0.028,
      "step": 1670
    },
    {
      "epoch": 5.694915254237288,
      "grad_norm": 0.25850358605384827,
      "learning_rate": 4.273799962906459e-05,
      "loss": 0.0255,
      "step": 1680
    },
    {
      "epoch": 5.728813559322034,
      "grad_norm": 0.3987925052642822,
      "learning_rate": 4.218381376079921e-05,
      "loss": 0.0267,
      "step": 1690
    },
    {
      "epoch": 5.762711864406779,
      "grad_norm": 0.25257599353790283,
      "learning_rate": 4.16306104406626e-05,
      "loss": 0.0244,
      "step": 1700
    },
    {
      "epoch": 5.796610169491525,
      "grad_norm": 0.33424732089042664,
      "learning_rate": 4.1078459210099954e-05,
      "loss": 0.0254,
      "step": 1710
    },
    {
      "epoch": 5.830508474576272,
      "grad_norm": 0.22258403897285461,
      "learning_rate": 4.0527429478301596e-05,
      "loss": 0.0236,
      "step": 1720
    },
    {
      "epoch": 5.864406779661017,
      "grad_norm": 0.34541794657707214,
      "learning_rate": 3.997759051347782e-05,
      "loss": 0.0206,
      "step": 1730
    },
    {
      "epoch": 5.898305084745763,
      "grad_norm": 0.38338717818260193,
      "learning_rate": 3.9429011434151306e-05,
      "loss": 0.031,
      "step": 1740
    },
    {
      "epoch": 5.932203389830509,
      "grad_norm": 0.5410785675048828,
      "learning_rate": 3.8881761200468546e-05,
      "loss": 0.0269,
      "step": 1750
    },
    {
      "epoch": 5.966101694915254,
      "grad_norm": 0.3703366219997406,
      "learning_rate": 3.833590860553098e-05,
      "loss": 0.0265,
      "step": 1760
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.3290477991104126,
      "learning_rate": 3.779152226674733e-05,
      "loss": 0.0231,
      "step": 1770
    },
    {
      "epoch": 6.033898305084746,
      "grad_norm": 0.3305765986442566,
      "learning_rate": 3.724867061720787e-05,
      "loss": 0.0237,
      "step": 1780
    },
    {
      "epoch": 6.067796610169491,
      "grad_norm": 0.2830105125904083,
      "learning_rate": 3.6707421897081947e-05,
      "loss": 0.0253,
      "step": 1790
    },
    {
      "epoch": 6.101694915254237,
      "grad_norm": 0.4226081073284149,
      "learning_rate": 3.6167844145039644e-05,
      "loss": 0.021,
      "step": 1800
    },
    {
      "epoch": 6.135593220338983,
      "grad_norm": 0.3365839123725891,
      "learning_rate": 3.563000518969898e-05,
      "loss": 0.0236,
      "step": 1810
    },
    {
      "epoch": 6.169491525423728,
      "grad_norm": 0.3108862042427063,
      "learning_rate": 3.509397264109929e-05,
      "loss": 0.0214,
      "step": 1820
    },
    {
      "epoch": 6.203389830508475,
      "grad_norm": 0.22683779895305634,
      "learning_rate": 3.455981388220223e-05,
      "loss": 0.0194,
      "step": 1830
    },
    {
      "epoch": 6.237288135593221,
      "grad_norm": 0.38497164845466614,
      "learning_rate": 3.4027596060421254e-05,
      "loss": 0.0241,
      "step": 1840
    },
    {
      "epoch": 6.271186440677966,
      "grad_norm": 0.22206397354602814,
      "learning_rate": 3.3497386079180725e-05,
      "loss": 0.019,
      "step": 1850
    },
    {
      "epoch": 6.305084745762712,
      "grad_norm": 0.3224073648452759,
      "learning_rate": 3.29692505895057e-05,
      "loss": 0.0231,
      "step": 1860
    },
    {
      "epoch": 6.338983050847458,
      "grad_norm": 0.28747010231018066,
      "learning_rate": 3.244325598164344e-05,
      "loss": 0.0206,
      "step": 1870
    },
    {
      "epoch": 6.372881355932203,
      "grad_norm": 0.3315770924091339,
      "learning_rate": 3.191946837671773e-05,
      "loss": 0.0192,
      "step": 1880
    },
    {
      "epoch": 6.406779661016949,
      "grad_norm": 0.38061845302581787,
      "learning_rate": 3.139795361841696e-05,
      "loss": 0.0179,
      "step": 1890
    },
    {
      "epoch": 6.440677966101695,
      "grad_norm": 0.2752900719642639,
      "learning_rate": 3.08787772647171e-05,
      "loss": 0.0217,
      "step": 1900
    },
    {
      "epoch": 6.47457627118644,
      "grad_norm": 0.270625501871109,
      "learning_rate": 3.03620045796407e-05,
      "loss": 0.0246,
      "step": 1910
    },
    {
      "epoch": 6.508474576271187,
      "grad_norm": 0.2529767155647278,
      "learning_rate": 2.9847700525052687e-05,
      "loss": 0.0192,
      "step": 1920
    },
    {
      "epoch": 6.5423728813559325,
      "grad_norm": 0.25423118472099304,
      "learning_rate": 2.933592975249423e-05,
      "loss": 0.0209,
      "step": 1930
    },
    {
      "epoch": 6.576271186440678,
      "grad_norm": 0.20597638189792633,
      "learning_rate": 2.8826756595055593e-05,
      "loss": 0.024,
      "step": 1940
    },
    {
      "epoch": 6.610169491525424,
      "grad_norm": 0.2644325792789459,
      "learning_rate": 2.8320245059289084e-05,
      "loss": 0.0249,
      "step": 1950
    },
    {
      "epoch": 6.6440677966101696,
      "grad_norm": 0.3659607768058777,
      "learning_rate": 2.781645881716286e-05,
      "loss": 0.0189,
      "step": 1960
    },
    {
      "epoch": 6.677966101694915,
      "grad_norm": 0.2306346893310547,
      "learning_rate": 2.731546119805708e-05,
      "loss": 0.0224,
      "step": 1970
    },
    {
      "epoch": 6.711864406779661,
      "grad_norm": 0.30413874983787537,
      "learning_rate": 2.6817315180802848e-05,
      "loss": 0.023,
      "step": 1980
    },
    {
      "epoch": 6.745762711864407,
      "grad_norm": 0.2451293170452118,
      "learning_rate": 2.63220833857654e-05,
      "loss": 0.0189,
      "step": 1990
    },
    {
      "epoch": 6.779661016949152,
      "grad_norm": 0.3022925555706024,
      "learning_rate": 2.5829828066972318e-05,
      "loss": 0.0217,
      "step": 2000
    },
    {
      "epoch": 6.813559322033898,
      "grad_norm": 0.2744346261024475,
      "learning_rate": 2.5340611104287682e-05,
      "loss": 0.0215,
      "step": 2010
    },
    {
      "epoch": 6.847457627118644,
      "grad_norm": 0.38032010197639465,
      "learning_rate": 2.4854493995633465e-05,
      "loss": 0.0194,
      "step": 2020
    },
    {
      "epoch": 6.88135593220339,
      "grad_norm": 0.3329201340675354,
      "learning_rate": 2.437153784925873e-05,
      "loss": 0.0217,
      "step": 2030
    },
    {
      "epoch": 6.915254237288136,
      "grad_norm": 0.3519807457923889,
      "learning_rate": 2.3891803376057914e-05,
      "loss": 0.0229,
      "step": 2040
    },
    {
      "epoch": 6.9491525423728815,
      "grad_norm": 0.3398759961128235,
      "learning_rate": 2.3415350881939134e-05,
      "loss": 0.0201,
      "step": 2050
    },
    {
      "epoch": 6.983050847457627,
      "grad_norm": 0.5052980780601501,
      "learning_rate": 2.2942240260243248e-05,
      "loss": 0.0212,
      "step": 2060
    },
    {
      "epoch": 7.016949152542373,
      "grad_norm": 0.26014694571495056,
      "learning_rate": 2.2472530984214823e-05,
      "loss": 0.02,
      "step": 2070
    },
    {
      "epoch": 7.0508474576271185,
      "grad_norm": 0.20524539053440094,
      "learning_rate": 2.2006282099526053e-05,
      "loss": 0.0203,
      "step": 2080
    },
    {
      "epoch": 7.084745762711864,
      "grad_norm": 0.26466068625450134,
      "learning_rate": 2.15435522168542e-05,
      "loss": 0.0194,
      "step": 2090
    },
    {
      "epoch": 7.11864406779661,
      "grad_norm": 0.5399367809295654,
      "learning_rate": 2.108439950451387e-05,
      "loss": 0.0216,
      "step": 2100
    },
    {
      "epoch": 7.1525423728813555,
      "grad_norm": 0.39390575885772705,
      "learning_rate": 2.0628881681144912e-05,
      "loss": 0.0199,
      "step": 2110
    },
    {
      "epoch": 7.186440677966102,
      "grad_norm": 0.17156942188739777,
      "learning_rate": 2.0177056008456702e-05,
      "loss": 0.0177,
      "step": 2120
    },
    {
      "epoch": 7.220338983050848,
      "grad_norm": 0.337688684463501,
      "learning_rate": 1.9728979284030025e-05,
      "loss": 0.0204,
      "step": 2130
    },
    {
      "epoch": 7.254237288135593,
      "grad_norm": 0.4261310398578644,
      "learning_rate": 1.928470783417718e-05,
      "loss": 0.0221,
      "step": 2140
    },
    {
      "epoch": 7.288135593220339,
      "grad_norm": 0.3683595359325409,
      "learning_rate": 1.8844297506861375e-05,
      "loss": 0.018,
      "step": 2150
    },
    {
      "epoch": 7.322033898305085,
      "grad_norm": 0.31089678406715393,
      "learning_rate": 1.840780366467632e-05,
      "loss": 0.0182,
      "step": 2160
    },
    {
      "epoch": 7.3559322033898304,
      "grad_norm": 0.3290519118309021,
      "learning_rate": 1.797528117788674e-05,
      "loss": 0.0173,
      "step": 2170
    },
    {
      "epoch": 7.389830508474576,
      "grad_norm": 0.21714557707309723,
      "learning_rate": 1.754678441753072e-05,
      "loss": 0.0181,
      "step": 2180
    },
    {
      "epoch": 7.423728813559322,
      "grad_norm": 0.2562347948551178,
      "learning_rate": 1.7122367248585048e-05,
      "loss": 0.0192,
      "step": 2190
    },
    {
      "epoch": 7.4576271186440675,
      "grad_norm": 0.22622661292552948,
      "learning_rate": 1.6702083023193887e-05,
      "loss": 0.018,
      "step": 2200
    },
    {
      "epoch": 7.491525423728813,
      "grad_norm": 0.2597718834877014,
      "learning_rate": 1.6285984573962098e-05,
      "loss": 0.0238,
      "step": 2210
    },
    {
      "epoch": 7.52542372881356,
      "grad_norm": 0.32258540391921997,
      "learning_rate": 1.587412420731387e-05,
      "loss": 0.0207,
      "step": 2220
    },
    {
      "epoch": 7.559322033898305,
      "grad_norm": 0.27688607573509216,
      "learning_rate": 1.5466553696917323e-05,
      "loss": 0.0206,
      "step": 2230
    },
    {
      "epoch": 7.593220338983051,
      "grad_norm": 0.326158344745636,
      "learning_rate": 1.5063324277176305e-05,
      "loss": 0.0187,
      "step": 2240
    },
    {
      "epoch": 7.627118644067797,
      "grad_norm": 0.22293926775455475,
      "learning_rate": 1.4664486636789871e-05,
      "loss": 0.0181,
      "step": 2250
    },
    {
      "epoch": 7.661016949152542,
      "grad_norm": 0.5270984172821045,
      "learning_rate": 1.4270090912380274e-05,
      "loss": 0.0206,
      "step": 2260
    },
    {
      "epoch": 7.694915254237288,
      "grad_norm": 0.2740112245082855,
      "learning_rate": 1.3880186682190583e-05,
      "loss": 0.0188,
      "step": 2270
    },
    {
      "epoch": 7.728813559322034,
      "grad_norm": 0.22564932703971863,
      "learning_rate": 1.3494822959852271e-05,
      "loss": 0.0165,
      "step": 2280
    },
    {
      "epoch": 7.762711864406779,
      "grad_norm": 0.22905248403549194,
      "learning_rate": 1.3114048188223838e-05,
      "loss": 0.0205,
      "step": 2290
    },
    {
      "epoch": 7.796610169491525,
      "grad_norm": 0.3510766923427582,
      "learning_rate": 1.2737910233301332e-05,
      "loss": 0.0167,
      "step": 2300
    },
    {
      "epoch": 7.830508474576272,
      "grad_norm": 0.25208619236946106,
      "learning_rate": 1.2366456378201135e-05,
      "loss": 0.0178,
      "step": 2310
    },
    {
      "epoch": 7.864406779661017,
      "grad_norm": 0.27848222851753235,
      "learning_rate": 1.1999733317216222e-05,
      "loss": 0.0208,
      "step": 2320
    },
    {
      "epoch": 7.898305084745763,
      "grad_norm": 0.2933652997016907,
      "learning_rate": 1.1637787149946377e-05,
      "loss": 0.0171,
      "step": 2330
    },
    {
      "epoch": 7.932203389830509,
      "grad_norm": 0.28821244835853577,
      "learning_rate": 1.128066337550312e-05,
      "loss": 0.0204,
      "step": 2340
    },
    {
      "epoch": 7.966101694915254,
      "grad_norm": 0.3546352684497833,
      "learning_rate": 1.0928406886790155e-05,
      "loss": 0.0192,
      "step": 2350
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.7775828242301941,
      "learning_rate": 1.0581061964860085e-05,
      "loss": 0.018,
      "step": 2360
    },
    {
      "epoch": 8.033898305084746,
      "grad_norm": 0.3042224943637848,
      "learning_rate": 1.0238672273347894e-05,
      "loss": 0.0163,
      "step": 2370
    },
    {
      "epoch": 8.067796610169491,
      "grad_norm": 0.24432986974716187,
      "learning_rate": 9.901280852982203e-06,
      "loss": 0.0215,
      "step": 2380
    },
    {
      "epoch": 8.101694915254237,
      "grad_norm": 0.25356370210647583,
      "learning_rate": 9.568930116174734e-06,
      "loss": 0.0198,
      "step": 2390
    },
    {
      "epoch": 8.135593220338983,
      "grad_norm": 0.2504992187023163,
      "learning_rate": 9.241661841688732e-06,
      "loss": 0.0141,
      "step": 2400
    },
    {
      "epoch": 8.169491525423728,
      "grad_norm": 0.33074647188186646,
      "learning_rate": 8.919517169387165e-06,
      "loss": 0.0205,
      "step": 2410
    },
    {
      "epoch": 8.203389830508474,
      "grad_norm": 0.336225301027298,
      "learning_rate": 8.602536595061111e-06,
      "loss": 0.0203,
      "step": 2420
    },
    {
      "epoch": 8.23728813559322,
      "grad_norm": 0.17460206151008606,
      "learning_rate": 8.290759965339145e-06,
      "loss": 0.0199,
      "step": 2430
    },
    {
      "epoch": 8.271186440677965,
      "grad_norm": 0.26632779836654663,
      "learning_rate": 7.98422647267844e-06,
      "loss": 0.0159,
      "step": 2440
    },
    {
      "epoch": 8.305084745762711,
      "grad_norm": 0.22440572082996368,
      "learning_rate": 7.682974650437879e-06,
      "loss": 0.0193,
      "step": 2450
    },
    {
      "epoch": 8.338983050847457,
      "grad_norm": 0.30192360281944275,
      "learning_rate": 7.387042368034219e-06,
      "loss": 0.0183,
      "step": 2460
    },
    {
      "epoch": 8.372881355932204,
      "grad_norm": 0.2981124818325043,
      "learning_rate": 7.096466826181669e-06,
      "loss": 0.0203,
      "step": 2470
    },
    {
      "epoch": 8.40677966101695,
      "grad_norm": 0.1849370151758194,
      "learning_rate": 6.811284552215435e-06,
      "loss": 0.0155,
      "step": 2480
    },
    {
      "epoch": 8.440677966101696,
      "grad_norm": 0.2095867246389389,
      "learning_rate": 6.531531395500046e-06,
      "loss": 0.0181,
      "step": 2490
    },
    {
      "epoch": 8.474576271186441,
      "grad_norm": 0.3721981644630432,
      "learning_rate": 6.257242522922813e-06,
      "loss": 0.0186,
      "step": 2500
    },
    {
      "epoch": 8.508474576271187,
      "grad_norm": 0.19706839323043823,
      "learning_rate": 5.988452414473106e-06,
      "loss": 0.0184,
      "step": 2510
    },
    {
      "epoch": 8.542372881355933,
      "grad_norm": 0.2606714367866516,
      "learning_rate": 5.725194858907989e-06,
      "loss": 0.0172,
      "step": 2520
    },
    {
      "epoch": 8.576271186440678,
      "grad_norm": 0.1873464286327362,
      "learning_rate": 5.467502949504783e-06,
      "loss": 0.0181,
      "step": 2530
    },
    {
      "epoch": 8.610169491525424,
      "grad_norm": 0.2513989806175232,
      "learning_rate": 5.215409079900923e-06,
      "loss": 0.0171,
      "step": 2540
    },
    {
      "epoch": 8.64406779661017,
      "grad_norm": 0.3158385157585144,
      "learning_rate": 4.9689449400219635e-06,
      "loss": 0.0173,
      "step": 2550
    },
    {
      "epoch": 8.677966101694915,
      "grad_norm": 0.22146488726139069,
      "learning_rate": 4.7281415120978504e-06,
      "loss": 0.0162,
      "step": 2560
    },
    {
      "epoch": 8.711864406779661,
      "grad_norm": 0.23577848076820374,
      "learning_rate": 4.4930290667682786e-06,
      "loss": 0.0172,
      "step": 2570
    },
    {
      "epoch": 8.745762711864407,
      "grad_norm": 0.41719919443130493,
      "learning_rate": 4.263637159277495e-06,
      "loss": 0.0173,
      "step": 2580
    },
    {
      "epoch": 8.779661016949152,
      "grad_norm": 0.3312488794326782,
      "learning_rate": 4.039994625758925e-06,
      "loss": 0.0182,
      "step": 2590
    },
    {
      "epoch": 8.813559322033898,
      "grad_norm": 0.2829158306121826,
      "learning_rate": 3.822129579610356e-06,
      "loss": 0.0181,
      "step": 2600
    },
    {
      "epoch": 8.847457627118644,
      "grad_norm": 0.24710486829280853,
      "learning_rate": 3.6100694079598485e-06,
      "loss": 0.0166,
      "step": 2610
    },
    {
      "epoch": 8.88135593220339,
      "grad_norm": 0.1250915676355362,
      "learning_rate": 3.4038407682229768e-06,
      "loss": 0.0153,
      "step": 2620
    },
    {
      "epoch": 8.915254237288135,
      "grad_norm": 0.1759193390607834,
      "learning_rate": 3.2034695847518227e-06,
      "loss": 0.0161,
      "step": 2630
    },
    {
      "epoch": 8.94915254237288,
      "grad_norm": 0.23805756866931915,
      "learning_rate": 3.0089810455761316e-06,
      "loss": 0.0165,
      "step": 2640
    },
    {
      "epoch": 8.983050847457626,
      "grad_norm": 0.19876998662948608,
      "learning_rate": 2.8203995992369503e-06,
      "loss": 0.0156,
      "step": 2650
    },
    {
      "epoch": 9.016949152542374,
      "grad_norm": 0.21514378488063812,
      "learning_rate": 2.637748951713348e-06,
      "loss": 0.0199,
      "step": 2660
    },
    {
      "epoch": 9.05084745762712,
      "grad_norm": 0.29321616888046265,
      "learning_rate": 2.4610520634423374e-06,
      "loss": 0.0158,
      "step": 2670
    },
    {
      "epoch": 9.084745762711865,
      "grad_norm": 0.23713384568691254,
      "learning_rate": 2.290331146432645e-06,
      "loss": 0.018,
      "step": 2680
    },
    {
      "epoch": 9.11864406779661,
      "grad_norm": 0.40180814266204834,
      "learning_rate": 2.125607661472495e-06,
      "loss": 0.017,
      "step": 2690
    },
    {
      "epoch": 9.152542372881356,
      "grad_norm": 0.2338988482952118,
      "learning_rate": 1.9669023154318235e-06,
      "loss": 0.0172,
      "step": 2700
    },
    {
      "epoch": 9.186440677966102,
      "grad_norm": 0.3559688329696655,
      "learning_rate": 1.814235058659297e-06,
      "loss": 0.016,
      "step": 2710
    },
    {
      "epoch": 9.220338983050848,
      "grad_norm": 0.191647008061409,
      "learning_rate": 1.6676250824744444e-06,
      "loss": 0.0161,
      "step": 2720
    },
    {
      "epoch": 9.254237288135593,
      "grad_norm": 0.20268139243125916,
      "learning_rate": 1.5270908167551223e-06,
      "loss": 0.0165,
      "step": 2730
    },
    {
      "epoch": 9.288135593220339,
      "grad_norm": 0.1706833839416504,
      "learning_rate": 1.3926499276207872e-06,
      "loss": 0.0191,
      "step": 2740
    },
    {
      "epoch": 9.322033898305085,
      "grad_norm": 0.1586904525756836,
      "learning_rate": 1.2643193152117428e-06,
      "loss": 0.0148,
      "step": 2750
    },
    {
      "epoch": 9.35593220338983,
      "grad_norm": 0.18658769130706787,
      "learning_rate": 1.1421151115646621e-06,
      "loss": 0.0142,
      "step": 2760
    },
    {
      "epoch": 9.389830508474576,
      "grad_norm": 0.3142736852169037,
      "learning_rate": 1.0260526785846936e-06,
      "loss": 0.0146,
      "step": 2770
    },
    {
      "epoch": 9.423728813559322,
      "grad_norm": 0.35053491592407227,
      "learning_rate": 9.161466061143653e-07,
      "loss": 0.0167,
      "step": 2780
    },
    {
      "epoch": 9.457627118644067,
      "grad_norm": 0.28281068801879883,
      "learning_rate": 8.124107100995093e-07,
      "loss": 0.0185,
      "step": 2790
    },
    {
      "epoch": 9.491525423728813,
      "grad_norm": 0.23286263644695282,
      "learning_rate": 7.14858030852572e-07,
      "loss": 0.0142,
      "step": 2800
    },
    {
      "epoch": 9.525423728813559,
      "grad_norm": 0.2240029126405716,
      "learning_rate": 6.235008314132762e-07,
      "loss": 0.0162,
      "step": 2810
    },
    {
      "epoch": 9.559322033898304,
      "grad_norm": 0.17313669621944427,
      "learning_rate": 5.383505960071256e-07,
      "loss": 0.0131,
      "step": 2820
    },
    {
      "epoch": 9.59322033898305,
      "grad_norm": 0.2640604078769684,
      "learning_rate": 4.5941802860176017e-07,
      "loss": 0.0159,
      "step": 2830
    },
    {
      "epoch": 9.627118644067796,
      "grad_norm": 0.26407647132873535,
      "learning_rate": 3.8671305156135484e-07,
      "loss": 0.0128,
      "step": 2840
    },
    {
      "epoch": 9.661016949152543,
      "grad_norm": 0.19562020897865295,
      "learning_rate": 3.2024480439934445e-07,
      "loss": 0.0145,
      "step": 2850
    },
    {
      "epoch": 9.694915254237289,
      "grad_norm": 0.1296786069869995,
      "learning_rate": 2.6002164262951546e-07,
      "loss": 0.0153,
      "step": 2860
    },
    {
      "epoch": 9.728813559322035,
      "grad_norm": 0.14787475764751434,
      "learning_rate": 2.06051136715657e-07,
      "loss": 0.0141,
      "step": 2870
    },
    {
      "epoch": 9.76271186440678,
      "grad_norm": 0.20423457026481628,
      "learning_rate": 1.5834007111991122e-07,
      "loss": 0.0188,
      "step": 2880
    },
    {
      "epoch": 9.796610169491526,
      "grad_norm": 0.2300773411989212,
      "learning_rate": 1.168944434499053e-07,
      "loss": 0.0158,
      "step": 2890
    },
    {
      "epoch": 9.830508474576272,
      "grad_norm": 0.39848440885543823,
      "learning_rate": 8.171946370481576e-08,
      "loss": 0.0168,
      "step": 2900
    },
    {
      "epoch": 9.864406779661017,
      "grad_norm": 0.1742592751979828,
      "learning_rate": 5.281955362045343e-08,
      "loss": 0.0191,
      "step": 2910
    },
    {
      "epoch": 9.898305084745763,
      "grad_norm": 0.21877504885196686,
      "learning_rate": 3.019834611339145e-08,
      "loss": 0.0146,
      "step": 2920
    },
    {
      "epoch": 9.932203389830509,
      "grad_norm": 0.19842731952667236,
      "learning_rate": 1.38586848243083e-08,
      "loss": 0.0144,
      "step": 2930
    },
    {
      "epoch": 9.966101694915254,
      "grad_norm": 0.1726018637418747,
      "learning_rate": 3.802623760501556e-09,
      "loss": 0.017,
      "step": 2940
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.1511099338531494,
      "learning_rate": 3.1427037711084796e-11,
      "loss": 0.0179,
      "step": 2950
    },
    {
      "epoch": 10.0,
      "step": 2950,
      "total_flos": 0.0,
      "train_loss": 0.03957842677326526,
      "train_runtime": 3172.5088,
      "train_samples_per_second": 45.431,
      "train_steps_per_second": 0.93
    }
  ],
  "logging_steps": 10,
  "max_steps": 2950,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 49,
  "trial_name": null,
  "trial_params": null
}