{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.998630870885172,
  "eval_steps": 500,
  "global_step": 11196,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0035716411691172094,
      "grad_norm": 47.84096509103244,
      "learning_rate": 6.3e-07,
      "loss": 2.0895,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.22,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 10
    },
    {
      "epoch": 0.007143282338234419,
      "grad_norm": 9.599267571025768,
      "learning_rate": 1.33e-06,
      "loss": 1.8176,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.22,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 20
    },
    {
      "epoch": 0.010714923507351629,
      "grad_norm": 5.205599049164542,
      "learning_rate": 2.0299999999999996e-06,
      "loss": 1.4788,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.22,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 30
    },
    {
      "epoch": 0.014286564676468837,
      "grad_norm": 2.205719923387632,
      "learning_rate": 2.73e-06,
      "loss": 1.2848,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 40
    },
    {
      "epoch": 0.017858205845586048,
      "grad_norm": 1.8447838629310436,
      "learning_rate": 3.4299999999999998e-06,
      "loss": 1.169,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 50
    },
    {
      "epoch": 0.021429847014703258,
      "grad_norm": 2.1579365048105044,
      "learning_rate": 4.129999999999999e-06,
      "loss": 1.0719,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 60
    },
    {
      "epoch": 0.025001488183820465,
      "grad_norm": 2.7295524544918397,
      "learning_rate": 4.8299999999999995e-06,
      "loss": 1.0057,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 70
    },
    {
      "epoch": 0.028573129352937675,
      "grad_norm": 2.7450280014903705,
      "learning_rate": 5.53e-06,
      "loss": 0.9462,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 80
    },
    {
      "epoch": 0.03214477052205488,
      "grad_norm": 3.5632078056569534,
      "learning_rate": 6.23e-06,
      "loss": 0.8855,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 90
    },
    {
      "epoch": 0.035716411691172095,
      "grad_norm": 3.4382364016435805,
      "learning_rate": 6.93e-06,
      "loss": 0.8407,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 100
    },
    {
      "epoch": 0.0392880528602893,
      "grad_norm": 2.9311709402300705,
      "learning_rate": 6.999988637086991e-06,
      "loss": 0.7935,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 110
    },
    {
      "epoch": 0.042859694029406516,
      "grad_norm": 2.796049740599755,
      "learning_rate": 6.9999493579762545e-06,
      "loss": 0.7523,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 120
    },
    {
      "epoch": 0.04643133519852372,
      "grad_norm": 3.1001912029157466,
      "learning_rate": 6.999882022699704e-06,
      "loss": 0.7096,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 130
    },
    {
      "epoch": 0.05000297636764093,
      "grad_norm": 2.720368644013231,
      "learning_rate": 6.999786631797113e-06,
      "loss": 0.6689,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 140
    },
    {
      "epoch": 0.05357461753675814,
      "grad_norm": 2.570815950070848,
      "learning_rate": 6.999663186033147e-06,
      "loss": 0.6353,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 150
    },
    {
      "epoch": 0.05714625870587535,
      "grad_norm": 2.4115369873146384,
      "learning_rate": 6.999511686397369e-06,
      "loss": 0.6019,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 160
    },
    {
      "epoch": 0.060717899874992556,
      "grad_norm": 2.3377949765181993,
      "learning_rate": 6.999332134104226e-06,
      "loss": 0.5704,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 170
    },
    {
      "epoch": 0.06428954104410976,
      "grad_norm": 1.5878939024656435,
      "learning_rate": 6.999124530593037e-06,
      "loss": 0.5523,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 180
    },
    {
      "epoch": 0.06786118221322698,
      "grad_norm": 1.3264919157704183,
      "learning_rate": 6.998888877527986e-06,
      "loss": 0.5435,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 190
    },
    {
      "epoch": 0.07143282338234419,
      "grad_norm": 2.038010532729759,
      "learning_rate": 6.9986251767981075e-06,
      "loss": 0.522,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 200
    },
    {
      "epoch": 0.0750044645514614,
      "grad_norm": 1.21415374871142,
      "learning_rate": 6.9983334305172685e-06,
      "loss": 0.51,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 210
    },
    {
      "epoch": 0.0785761057205786,
      "grad_norm": 1.614164511359785,
      "learning_rate": 6.998013641024154e-06,
      "loss": 0.4999,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 220
    },
    {
      "epoch": 0.08214774688969581,
      "grad_norm": 1.8682436211450861,
      "learning_rate": 6.997665810882251e-06,
      "loss": 0.495,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 230
    },
    {
      "epoch": 0.08571938805881303,
      "grad_norm": 1.2962880024167205,
      "learning_rate": 6.997289942879822e-06,
      "loss": 0.4771,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 240
    },
    {
      "epoch": 0.08929102922793024,
      "grad_norm": 1.087472607402264,
      "learning_rate": 6.996886040029885e-06,
      "loss": 0.4806,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 250
    },
    {
      "epoch": 0.09286267039704744,
      "grad_norm": 1.0059439297326096,
      "learning_rate": 6.996454105570193e-06,
      "loss": 0.4723,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 260
    },
    {
      "epoch": 0.09643431156616465,
      "grad_norm": 0.8919607073801068,
      "learning_rate": 6.9959941429632016e-06,
      "loss": 0.4619,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 270
    },
    {
      "epoch": 0.10000595273528186,
      "grad_norm": 0.9964374765321111,
      "learning_rate": 6.995506155896048e-06,
      "loss": 0.4596,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 280
    },
    {
      "epoch": 0.10357759390439907,
      "grad_norm": 1.6592621687619997,
      "learning_rate": 6.994990148280515e-06,
      "loss": 0.4583,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 290
    },
    {
      "epoch": 0.10714923507351629,
      "grad_norm": 0.9205505376855222,
      "learning_rate": 6.994446124253003e-06,
      "loss": 0.4494,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 300
    },
    {
      "epoch": 0.11072087624263349,
      "grad_norm": 0.8244439874224572,
      "learning_rate": 6.993874088174499e-06,
      "loss": 0.4397,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 310
    },
    {
      "epoch": 0.1142925174117507,
      "grad_norm": 0.8112163055151138,
      "learning_rate": 6.993274044630538e-06,
      "loss": 0.4474,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 320
    },
    {
      "epoch": 0.1178641585808679,
      "grad_norm": 0.9568110158706775,
      "learning_rate": 6.992645998431166e-06,
      "loss": 0.4381,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 330
    },
    {
      "epoch": 0.12143579974998511,
      "grad_norm": 0.894726666953806,
      "learning_rate": 6.991989954610905e-06,
      "loss": 0.4297,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 340
    },
    {
      "epoch": 0.12500744091910232,
      "grad_norm": 0.8386265568617023,
      "learning_rate": 6.991305918428709e-06,
      "loss": 0.4323,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 350
    },
    {
      "epoch": 0.12857908208821953,
      "grad_norm": 0.7834041126531694,
      "learning_rate": 6.990593895367922e-06,
      "loss": 0.43,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 360
    },
    {
      "epoch": 0.13215072325733673,
      "grad_norm": 0.6738833957249635,
      "learning_rate": 6.98985389113624e-06,
      "loss": 0.4248,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 370
    },
    {
      "epoch": 0.13572236442645397,
      "grad_norm": 0.6793575164378994,
      "learning_rate": 6.989085911665656e-06,
      "loss": 0.425,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 380
    },
    {
      "epoch": 0.13929400559557117,
      "grad_norm": 0.7442286270617003,
      "learning_rate": 6.988289963112419e-06,
      "loss": 0.4176,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 390
    },
    {
      "epoch": 0.14286564676468838,
      "grad_norm": 0.6670476894638441,
      "learning_rate": 6.987466051856985e-06,
      "loss": 0.4187,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 400
    },
    {
      "epoch": 0.1464372879338056,
      "grad_norm": 0.6158756183627782,
      "learning_rate": 6.9866141845039574e-06,
      "loss": 0.4083,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 410
    },
    {
      "epoch": 0.1500089291029228,
      "grad_norm": 0.8089591515808892,
      "learning_rate": 6.9857343678820465e-06,
      "loss": 0.4113,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 420
    },
    {
      "epoch": 0.15358057027204,
      "grad_norm": 0.6783726635892394,
      "learning_rate": 6.984826609044006e-06,
      "loss": 0.41,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 430
    },
    {
      "epoch": 0.1571522114411572,
      "grad_norm": 0.7629974327324691,
      "learning_rate": 6.983890915266577e-06,
      "loss": 0.4134,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 440
    },
    {
      "epoch": 0.16072385261027441,
      "grad_norm": 0.7450263962040445,
      "learning_rate": 6.982927294050436e-06,
      "loss": 0.4152,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 450
    },
    {
      "epoch": 0.16429549377939162,
      "grad_norm": 0.6928100693829363,
      "learning_rate": 6.981935753120128e-06,
      "loss": 0.4113,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 460
    },
    {
      "epoch": 0.16786713494850883,
      "grad_norm": 0.7438608324903054,
      "learning_rate": 6.980916300424006e-06,
      "loss": 0.4076,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 470
    },
    {
      "epoch": 0.17143877611762606,
      "grad_norm": 0.7341733676360269,
      "learning_rate": 6.979868944134169e-06,
      "loss": 0.4047,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 480
    },
    {
      "epoch": 0.17501041728674327,
      "grad_norm": 0.6106239719972382,
      "learning_rate": 6.9787936926463975e-06,
      "loss": 0.4093,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 490
    },
    {
      "epoch": 0.17858205845586048,
      "grad_norm": 0.5925020448722477,
      "learning_rate": 6.977690554580082e-06,
      "loss": 0.4011,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 500
    },
    {
      "epoch": 0.18215369962497768,
      "grad_norm": 0.5871269087366053,
      "learning_rate": 6.976559538778159e-06,
      "loss": 0.3971,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 510
    },
    {
      "epoch": 0.1857253407940949,
      "grad_norm": 0.8321256719515702,
      "learning_rate": 6.975400654307034e-06,
      "loss": 0.3997,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 520
    },
    {
      "epoch": 0.1892969819632121,
      "grad_norm": 0.5233429979550782,
      "learning_rate": 6.974213910456515e-06,
      "loss": 0.4004,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 530
    },
    {
      "epoch": 0.1928686231323293,
      "grad_norm": 0.6812462304184482,
      "learning_rate": 6.9729993167397355e-06,
      "loss": 0.3937,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 540
    },
    {
      "epoch": 0.1964402643014465,
      "grad_norm": 0.6409024032491494,
      "learning_rate": 6.971756882893076e-06,
      "loss": 0.3889,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 550
    },
    {
      "epoch": 0.20001190547056372,
      "grad_norm": 0.7193181432834939,
      "learning_rate": 6.9704866188760895e-06,
      "loss": 0.3992,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 560
    },
    {
      "epoch": 0.20358354663968092,
      "grad_norm": 0.7830346863841446,
      "learning_rate": 6.9691885348714184e-06,
      "loss": 0.3937,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 570
    },
    {
      "epoch": 0.20715518780879813,
      "grad_norm": 0.5652448193001737,
      "learning_rate": 6.967862641284718e-06,
      "loss": 0.3886,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 580
    },
    {
      "epoch": 0.21072682897791536,
      "grad_norm": 0.5748926945027741,
      "learning_rate": 6.966508948744569e-06,
      "loss": 0.3848,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 590
    },
    {
      "epoch": 0.21429847014703257,
      "grad_norm": 0.6665986237846571,
      "learning_rate": 6.96512746810239e-06,
      "loss": 0.3924,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 600
    },
    {
      "epoch": 0.21787011131614978,
      "grad_norm": 0.6823372205985191,
      "learning_rate": 6.963718210432355e-06,
      "loss": 0.3826,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 610
    },
    {
      "epoch": 0.22144175248526698,
      "grad_norm": 0.58994938065327,
      "learning_rate": 6.962281187031305e-06,
      "loss": 0.389,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 620
    },
    {
      "epoch": 0.2250133936543842,
      "grad_norm": 0.6052840463581863,
      "learning_rate": 6.9608164094186545e-06,
      "loss": 0.3835,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 630
    },
    {
      "epoch": 0.2285850348235014,
      "grad_norm": 0.7055587121927881,
      "learning_rate": 6.959323889336299e-06,
      "loss": 0.3793,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 640
    },
    {
      "epoch": 0.2321566759926186,
      "grad_norm": 0.6473817353018219,
      "learning_rate": 6.957803638748524e-06,
      "loss": 0.3768,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 650
    },
    {
      "epoch": 0.2357283171617358,
      "grad_norm": 0.5453880438030917,
      "learning_rate": 6.9562556698419065e-06,
      "loss": 0.3858,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 660
    },
    {
      "epoch": 0.23929995833085302,
      "grad_norm": 0.539099324403936,
      "learning_rate": 6.954679995025219e-06,
      "loss": 0.3882,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 670
    },
    {
      "epoch": 0.24287159949997023,
      "grad_norm": 0.5764837722625463,
      "learning_rate": 6.953076626929326e-06,
      "loss": 0.3808,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 680
    },
    {
      "epoch": 0.24644324066908743,
      "grad_norm": 0.5646676416233323,
      "learning_rate": 6.95144557840709e-06,
      "loss": 0.3801,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 690
    },
    {
      "epoch": 0.25001488183820464,
      "grad_norm": 0.6097644788446263,
      "learning_rate": 6.949786862533261e-06,
      "loss": 0.3716,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 700
    },
    {
      "epoch": 0.2535865230073219,
      "grad_norm": 0.526307204275848,
      "learning_rate": 6.9481004926043784e-06,
      "loss": 0.3778,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 710
    },
    {
      "epoch": 0.25715816417643905,
      "grad_norm": 0.578906592299905,
      "learning_rate": 6.946386482138658e-06,
      "loss": 0.3787,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 720
    },
    {
      "epoch": 0.2607298053455563,
      "grad_norm": 0.549626668018743,
      "learning_rate": 6.944644844875887e-06,
      "loss": 0.3702,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 730
    },
    {
      "epoch": 0.26430144651467347,
      "grad_norm": 0.6165546258876643,
      "learning_rate": 6.942875594777315e-06,
      "loss": 0.3721,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 740
    },
    {
      "epoch": 0.2678730876837907,
      "grad_norm": 0.6930131839825238,
      "learning_rate": 6.94107874602554e-06,
      "loss": 0.3703,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 750
    },
    {
      "epoch": 0.27144472885290793,
      "grad_norm": 0.60045008248484,
      "learning_rate": 6.939254313024393e-06,
      "loss": 0.3704,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 760
    },
    {
      "epoch": 0.2750163700220251,
      "grad_norm": 0.5827829638301874,
      "learning_rate": 6.9374023103988305e-06,
      "loss": 0.3708,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 770
    },
    {
      "epoch": 0.27858801119114235,
      "grad_norm": 0.5545642688802787,
      "learning_rate": 6.935522752994806e-06,
      "loss": 0.3749,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 780
    },
    {
      "epoch": 0.2821596523602595,
      "grad_norm": 0.5429199324119214,
      "learning_rate": 6.933615655879159e-06,
      "loss": 0.3748,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 790
    },
    {
      "epoch": 0.28573129352937676,
      "grad_norm": 0.5151541889971357,
      "learning_rate": 6.931681034339491e-06,
      "loss": 0.3692,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 800
    },
    {
      "epoch": 0.28930293469849394,
      "grad_norm": 0.5071957665722244,
      "learning_rate": 6.929718903884045e-06,
      "loss": 0.372,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 810
    },
    {
      "epoch": 0.2928745758676112,
      "grad_norm": 0.6233887048007566,
      "learning_rate": 6.92772928024158e-06,
      "loss": 0.3689,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 820
    },
    {
      "epoch": 0.29644621703672835,
      "grad_norm": 0.522671393367273,
      "learning_rate": 6.9257121793612425e-06,
      "loss": 0.3648,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 830
    },
    {
      "epoch": 0.3000178582058456,
      "grad_norm": 0.5410930274218302,
      "learning_rate": 6.923667617412443e-06,
      "loss": 0.3684,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 840
    },
    {
      "epoch": 0.30358949937496277,
      "grad_norm": 0.5662127188946626,
      "learning_rate": 6.921595610784726e-06,
      "loss": 0.367,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 850
    },
    {
      "epoch": 0.30716114054408,
      "grad_norm": 0.6258807561777707,
      "learning_rate": 6.9194961760876324e-06,
      "loss": 0.3655,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 860
    },
    {
      "epoch": 0.31073278171319724,
      "grad_norm": 0.5662982884434634,
      "learning_rate": 6.917369330150575e-06,
      "loss": 0.3668,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 870
    },
    {
      "epoch": 0.3143044228823144,
      "grad_norm": 0.5937482674450755,
      "learning_rate": 6.915215090022699e-06,
      "loss": 0.3648,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 880
    },
    {
      "epoch": 0.31787606405143165,
      "grad_norm": 0.5546463119504276,
      "learning_rate": 6.913033472972744e-06,
      "loss": 0.3642,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 890
    },
    {
      "epoch": 0.32144770522054883,
      "grad_norm": 0.5674302479041585,
      "learning_rate": 6.910824496488908e-06,
      "loss": 0.3765,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 900
    },
    {
      "epoch": 0.32501934638966606,
      "grad_norm": 0.5205528325154691,
      "learning_rate": 6.90858817827871e-06,
      "loss": 0.36,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 910
    },
    {
      "epoch": 0.32859098755878324,
      "grad_norm": 0.529377222438568,
      "learning_rate": 6.906324536268841e-06,
      "loss": 0.3677,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 920
    },
    {
      "epoch": 0.3321626287279005,
      "grad_norm": 0.6122529498214767,
      "learning_rate": 6.9040335886050235e-06,
      "loss": 0.3665,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 930
    },
    {
      "epoch": 0.33573426989701766,
      "grad_norm": 0.6135486838026446,
      "learning_rate": 6.901715353651869e-06,
      "loss": 0.3601,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 940
    },
    {
      "epoch": 0.3393059110661349,
      "grad_norm": 0.4990872913204794,
      "learning_rate": 6.899369849992728e-06,
      "loss": 0.3642,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 950
    },
    {
      "epoch": 0.3428775522352521,
      "grad_norm": 0.5348432895185071,
      "learning_rate": 6.896997096429542e-06,
      "loss": 0.3559,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 960
    },
    {
      "epoch": 0.3464491934043693,
      "grad_norm": 0.5440123718353397,
      "learning_rate": 6.89459711198269e-06,
      "loss": 0.3644,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 970
    },
    {
      "epoch": 0.35002083457348654,
      "grad_norm": 0.5253113764115817,
      "learning_rate": 6.89216991589084e-06,
      "loss": 0.359,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 980
    },
    {
      "epoch": 0.3535924757426037,
      "grad_norm": 0.6010219056393492,
      "learning_rate": 6.889715527610791e-06,
      "loss": 0.3586,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 990
    },
    {
      "epoch": 0.35716411691172095,
      "grad_norm": 0.49806273998200895,
      "learning_rate": 6.887233966817321e-06,
      "loss": 0.3588,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1000
    },
    {
      "epoch": 0.36073575808083813,
      "grad_norm": 0.5185542033740208,
      "learning_rate": 6.884725253403029e-06,
      "loss": 0.3559,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1010
    },
    {
      "epoch": 0.36430739924995537,
      "grad_norm": 0.6091051912479017,
      "learning_rate": 6.882189407478168e-06,
      "loss": 0.3593,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1020
    },
    {
      "epoch": 0.36787904041907254,
      "grad_norm": 0.490572831948872,
      "learning_rate": 6.8796264493704955e-06,
      "loss": 0.3533,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1030
    },
    {
      "epoch": 0.3714506815881898,
      "grad_norm": 0.4991028642294145,
      "learning_rate": 6.877036399625101e-06,
      "loss": 0.3586,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1040
    },
    {
      "epoch": 0.37502232275730696,
      "grad_norm": 0.47906335050403054,
      "learning_rate": 6.874419279004248e-06,
      "loss": 0.3581,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1050
    },
    {
      "epoch": 0.3785939639264242,
      "grad_norm": 0.49469149452681016,
      "learning_rate": 6.871775108487201e-06,
      "loss": 0.3596,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1060
    },
    {
      "epoch": 0.3821656050955414,
      "grad_norm": 0.5292248202385722,
      "learning_rate": 6.869103909270065e-06,
      "loss": 0.3583,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1070
    },
    {
      "epoch": 0.3857372462646586,
      "grad_norm": 0.5117377218602205,
      "learning_rate": 6.866405702765607e-06,
      "loss": 0.365,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1080
    },
    {
      "epoch": 0.38930888743377584,
      "grad_norm": 0.5004559418367507,
      "learning_rate": 6.863680510603094e-06,
      "loss": 0.3455,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1090
    },
    {
      "epoch": 0.392880528602893,
      "grad_norm": 0.57089126215835,
      "learning_rate": 6.860928354628108e-06,
      "loss": 0.349,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1100
    },
    {
      "epoch": 0.39645216977201025,
      "grad_norm": 0.5820503861826878,
      "learning_rate": 6.8581492569023836e-06,
      "loss": 0.3551,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1110
    },
    {
      "epoch": 0.40002381094112743,
      "grad_norm": 0.5570132816377671,
      "learning_rate": 6.85534323970362e-06,
      "loss": 0.3513,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1120
    },
    {
      "epoch": 0.40359545211024467,
      "grad_norm": 0.5464001911986117,
      "learning_rate": 6.852510325525313e-06,
      "loss": 0.3613,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1130
    },
    {
      "epoch": 0.40716709327936185,
      "grad_norm": 0.4827396326230109,
      "learning_rate": 6.849650537076562e-06,
      "loss": 0.3504,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1140
    },
    {
      "epoch": 0.4107387344484791,
      "grad_norm": 0.4999437570126179,
      "learning_rate": 6.8467638972819e-06,
      "loss": 0.3526,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1150
    },
    {
      "epoch": 0.41431037561759626,
      "grad_norm": 0.48618334288528897,
      "learning_rate": 6.843850429281102e-06,
      "loss": 0.3541,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1160
    },
    {
      "epoch": 0.4178820167867135,
      "grad_norm": 0.48710415139914987,
      "learning_rate": 6.840910156429005e-06,
      "loss": 0.3502,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1170
    },
    {
      "epoch": 0.42145365795583073,
      "grad_norm": 0.5301441193838501,
      "learning_rate": 6.8379431022953156e-06,
      "loss": 0.3451,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1180
    },
    {
      "epoch": 0.4250252991249479,
      "grad_norm": 0.4847505852864843,
      "learning_rate": 6.834949290664424e-06,
      "loss": 0.3549,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1190
    },
    {
      "epoch": 0.42859694029406514,
      "grad_norm": 0.49082072305818153,
      "learning_rate": 6.831928745535214e-06,
      "loss": 0.3539,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1200
    },
    {
      "epoch": 0.4321685814631823,
      "grad_norm": 0.46047367228279323,
      "learning_rate": 6.828881491120869e-06,
      "loss": 0.3545,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1210
    },
    {
      "epoch": 0.43574022263229956,
      "grad_norm": 0.48672379907591806,
      "learning_rate": 6.825807551848678e-06,
      "loss": 0.347,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1220
    },
    {
      "epoch": 0.43931186380141674,
      "grad_norm": 0.4448810458268107,
      "learning_rate": 6.8227069523598405e-06,
      "loss": 0.3559,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1230
    },
    {
      "epoch": 0.44288350497053397,
      "grad_norm": 0.49497043238088634,
      "learning_rate": 6.8195797175092675e-06,
      "loss": 0.3546,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1240
    },
    {
      "epoch": 0.44645514613965115,
      "grad_norm": 0.4732540784083669,
      "learning_rate": 6.816425872365385e-06,
      "loss": 0.349,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1250
    },
    {
      "epoch": 0.4500267873087684,
      "grad_norm": 0.5192314130895199,
      "learning_rate": 6.813245442209932e-06,
      "loss": 0.3515,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1260
    },
    {
      "epoch": 0.45359842847788556,
      "grad_norm": 0.4792403319160851,
      "learning_rate": 6.810038452537754e-06,
      "loss": 0.3382,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1270
    },
    {
      "epoch": 0.4571700696470028,
      "grad_norm": 0.5060077004358183,
      "learning_rate": 6.806804929056606e-06,
      "loss": 0.3554,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1280
    },
    {
      "epoch": 0.46074171081612003,
      "grad_norm": 0.520248105755976,
      "learning_rate": 6.80354489768694e-06,
      "loss": 0.3464,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1290
    },
    {
      "epoch": 0.4643133519852372,
      "grad_norm": 0.4598019361746213,
      "learning_rate": 6.8002583845617e-06,
      "loss": 0.3526,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1300
    },
    {
      "epoch": 0.46788499315435444,
      "grad_norm": 0.4724933886093667,
      "learning_rate": 6.796945416026112e-06,
      "loss": 0.3512,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1310
    },
    {
      "epoch": 0.4714566343234716,
      "grad_norm": 0.4660281073410152,
      "learning_rate": 6.793606018637473e-06,
      "loss": 0.3476,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1320
    },
    {
      "epoch": 0.47502827549258886,
      "grad_norm": 0.5045848706762358,
      "learning_rate": 6.790240219164937e-06,
      "loss": 0.3453,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1330
    },
    {
      "epoch": 0.47859991666170604,
      "grad_norm": 0.5194987277056717,
      "learning_rate": 6.786848044589305e-06,
      "loss": 0.3474,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1340
    },
    {
      "epoch": 0.48217155783082327,
      "grad_norm": 0.5656264203505013,
      "learning_rate": 6.7834295221028e-06,
      "loss": 0.3424,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1350
    },
    {
      "epoch": 0.48574319899994045,
      "grad_norm": 0.4894023655199842,
      "learning_rate": 6.779984679108858e-06,
      "loss": 0.3515,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1360
    },
    {
      "epoch": 0.4893148401690577,
      "grad_norm": 0.5037132785096825,
      "learning_rate": 6.776513543221901e-06,
      "loss": 0.3346,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1370
    },
    {
      "epoch": 0.49288648133817486,
      "grad_norm": 0.5190300836703553,
      "learning_rate": 6.773016142267126e-06,
      "loss": 0.3476,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1380
    },
    {
      "epoch": 0.4964581225072921,
      "grad_norm": 0.4435043093847383,
      "learning_rate": 6.7694925042802655e-06,
      "loss": 0.3366,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1390
    },
    {
      "epoch": 0.5000297636764093,
      "grad_norm": 0.5324417067919307,
      "learning_rate": 6.7659426575073794e-06,
      "loss": 0.3456,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1400
    },
    {
      "epoch": 0.5036014048455265,
      "grad_norm": 0.494609813535511,
      "learning_rate": 6.76236663040462e-06,
      "loss": 0.3389,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1410
    },
    {
      "epoch": 0.5071730460146437,
      "grad_norm": 0.4723453866802145,
      "learning_rate": 6.758764451638004e-06,
      "loss": 0.3419,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1420
    },
    {
      "epoch": 0.510744687183761,
      "grad_norm": 0.4986566417970195,
      "learning_rate": 6.7551361500831845e-06,
      "loss": 0.3359,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1430
    },
    {
      "epoch": 0.5143163283528781,
      "grad_norm": 0.4900882038508289,
      "learning_rate": 6.75148175482522e-06,
      "loss": 0.3453,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1440
    },
    {
      "epoch": 0.5178879695219953,
      "grad_norm": 0.48044517729584096,
      "learning_rate": 6.747801295158339e-06,
      "loss": 0.3442,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1450
    },
    {
      "epoch": 0.5214596106911126,
      "grad_norm": 0.5288761906532677,
      "learning_rate": 6.744094800585709e-06,
      "loss": 0.3375,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1460
    },
    {
      "epoch": 0.5250312518602298,
      "grad_norm": 0.530447136190816,
      "learning_rate": 6.740362300819195e-06,
      "loss": 0.3382,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1470
    },
    {
      "epoch": 0.5286028930293469,
      "grad_norm": 0.45351586743181693,
      "learning_rate": 6.736603825779125e-06,
      "loss": 0.3444,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1480
    },
    {
      "epoch": 0.5321745341984642,
      "grad_norm": 0.5208363356822504,
      "learning_rate": 6.732819405594048e-06,
      "loss": 0.3347,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1490
    },
    {
      "epoch": 0.5357461753675814,
      "grad_norm": 0.45820411460318944,
      "learning_rate": 6.729009070600497e-06,
      "loss": 0.3389,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1500
    },
    {
      "epoch": 0.5393178165366986,
      "grad_norm": 0.4831990656531665,
      "learning_rate": 6.725172851342737e-06,
      "loss": 0.3355,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1510
    },
    {
      "epoch": 0.5428894577058159,
      "grad_norm": 0.4699654688524428,
      "learning_rate": 6.721310778572529e-06,
      "loss": 0.3318,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1520
    },
    {
      "epoch": 0.546461098874933,
      "grad_norm": 0.47152717294043456,
      "learning_rate": 6.71742288324888e-06,
      "loss": 0.3412,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1530
    },
    {
      "epoch": 0.5500327400440502,
      "grad_norm": 0.4994080875975566,
      "learning_rate": 6.713509196537794e-06,
      "loss": 0.3376,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1540
    },
    {
      "epoch": 0.5536043812131675,
      "grad_norm": 0.48606980053716015,
      "learning_rate": 6.709569749812023e-06,
      "loss": 0.3412,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1550
    },
    {
      "epoch": 0.5571760223822847,
      "grad_norm": 0.47736374810187937,
      "learning_rate": 6.705604574650815e-06,
      "loss": 0.3359,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1560
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 0.46009726498403986,
      "learning_rate": 6.701613702839661e-06,
      "loss": 0.3372,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1570
    },
    {
      "epoch": 0.564319304720519,
      "grad_norm": 0.49517486088050466,
      "learning_rate": 6.6975971663700434e-06,
      "loss": 0.328,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1580
    },
    {
      "epoch": 0.5678909458896363,
      "grad_norm": 0.4746416834193424,
      "learning_rate": 6.693554997439175e-06,
      "loss": 0.3366,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1590
    },
    {
      "epoch": 0.5714625870587535,
      "grad_norm": 0.4611224689559148,
      "learning_rate": 6.689487228449739e-06,
      "loss": 0.3409,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1600
    },
    {
      "epoch": 0.5750342282278708,
      "grad_norm": 0.4809305461747918,
      "learning_rate": 6.685393892009639e-06,
      "loss": 0.3398,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1610
    },
    {
      "epoch": 0.5786058693969879,
      "grad_norm": 0.44334262590360224,
      "learning_rate": 6.681275020931727e-06,
      "loss": 0.3382,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1620
    },
    {
      "epoch": 0.5821775105661051,
      "grad_norm": 0.5398313870854491,
      "learning_rate": 6.677130648233548e-06,
      "loss": 0.3464,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1630
    },
    {
      "epoch": 0.5857491517352224,
      "grad_norm": 0.490905419595684,
      "learning_rate": 6.6729608071370695e-06,
      "loss": 0.3377,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1640
    },
    {
      "epoch": 0.5893207929043396,
      "grad_norm": 0.5468845395496524,
      "learning_rate": 6.668765531068419e-06,
      "loss": 0.3363,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1650
    },
    {
      "epoch": 0.5928924340734567,
      "grad_norm": 0.48192206630092693,
      "learning_rate": 6.664544853657616e-06,
      "loss": 0.3298,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1660
    },
    {
      "epoch": 0.5964640752425739,
      "grad_norm": 0.45744945668619447,
      "learning_rate": 6.660298808738299e-06,
      "loss": 0.334,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1670
    },
    {
      "epoch": 0.6000357164116912,
      "grad_norm": 0.44653921337558566,
      "learning_rate": 6.656027430347457e-06,
      "loss": 0.3355,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1680
    },
    {
      "epoch": 0.6036073575808084,
      "grad_norm": 0.5388041006651919,
      "learning_rate": 6.6517307527251576e-06,
      "loss": 0.3372,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1690
    },
    {
      "epoch": 0.6071789987499255,
      "grad_norm": 0.4494250389449916,
      "learning_rate": 6.647408810314272e-06,
      "loss": 0.3289,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1700
    },
    {
      "epoch": 0.6107506399190428,
      "grad_norm": 0.5041032492465202,
      "learning_rate": 6.643061637760195e-06,
      "loss": 0.329,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1710
    },
    {
      "epoch": 0.61432228108816,
      "grad_norm": 0.4542350053071805,
      "learning_rate": 6.638689269910572e-06,
      "loss": 0.3269,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1720
    },
    {
      "epoch": 0.6178939222572772,
      "grad_norm": 0.4717924107240841,
      "learning_rate": 6.634291741815018e-06,
      "loss": 0.336,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1730
    },
    {
      "epoch": 0.6214655634263945,
      "grad_norm": 0.49412268954326644,
      "learning_rate": 6.629869088724838e-06,
      "loss": 0.3356,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1740
    },
    {
      "epoch": 0.6250372045955116,
      "grad_norm": 0.4881525370343868,
      "learning_rate": 6.6254213460927385e-06,
      "loss": 0.3294,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1750
    },
    {
      "epoch": 0.6286088457646288,
      "grad_norm": 0.5027799485224533,
      "learning_rate": 6.620948549572555e-06,
      "loss": 0.3344,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1760
    },
    {
      "epoch": 0.6321804869337461,
      "grad_norm": 0.48657084247101007,
      "learning_rate": 6.616450735018952e-06,
      "loss": 0.3346,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1770
    },
    {
      "epoch": 0.6357521281028633,
      "grad_norm": 0.4974458887948958,
      "learning_rate": 6.611927938487148e-06,
      "loss": 0.3372,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1780
    },
    {
      "epoch": 0.6393237692719804,
      "grad_norm": 0.48881133320324655,
      "learning_rate": 6.607380196232617e-06,
      "loss": 0.332,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1790
    },
    {
      "epoch": 0.6428954104410977,
      "grad_norm": 0.436305609913343,
      "learning_rate": 6.6028075447108054e-06,
      "loss": 0.3285,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1800
    },
    {
      "epoch": 0.6464670516102149,
      "grad_norm": 0.48349336881385824,
      "learning_rate": 6.598210020576833e-06,
      "loss": 0.329,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1810
    },
    {
      "epoch": 0.6500386927793321,
      "grad_norm": 0.45307654128523317,
      "learning_rate": 6.593587660685206e-06,
      "loss": 0.323,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1820
    },
    {
      "epoch": 0.6536103339484494,
      "grad_norm": 0.4663165053604926,
      "learning_rate": 6.588940502089516e-06,
      "loss": 0.3354,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1830
    },
    {
      "epoch": 0.6571819751175665,
      "grad_norm": 0.49842281430038393,
      "learning_rate": 6.584268582042145e-06,
      "loss": 0.3329,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1840
    },
    {
      "epoch": 0.6607536162866837,
      "grad_norm": 0.4758384481874692,
      "learning_rate": 6.579571937993967e-06,
      "loss": 0.3332,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1850
    },
    {
      "epoch": 0.664325257455801,
      "grad_norm": 0.4548392499179828,
      "learning_rate": 6.574850607594048e-06,
      "loss": 0.3356,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1860
    },
    {
      "epoch": 0.6678968986249182,
      "grad_norm": 0.4723502380387424,
      "learning_rate": 6.570104628689345e-06,
      "loss": 0.3228,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1870
    },
    {
      "epoch": 0.6714685397940353,
      "grad_norm": 0.4827017868525288,
      "learning_rate": 6.5653340393244e-06,
      "loss": 0.3299,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1880
    },
    {
      "epoch": 0.6750401809631525,
      "grad_norm": 0.4853339734974653,
      "learning_rate": 6.560538877741036e-06,
      "loss": 0.3354,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1890
    },
    {
      "epoch": 0.6786118221322698,
      "grad_norm": 0.490622912910108,
      "learning_rate": 6.55571918237805e-06,
      "loss": 0.3289,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1900
    },
    {
      "epoch": 0.682183463301387,
      "grad_norm": 0.4360684992956346,
      "learning_rate": 6.5508749918709115e-06,
      "loss": 0.3304,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1910
    },
    {
      "epoch": 0.6857551044705043,
      "grad_norm": 0.4706600602999743,
      "learning_rate": 6.54600634505144e-06,
      "loss": 0.3268,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1920
    },
    {
      "epoch": 0.6893267456396214,
      "grad_norm": 0.4674898732610253,
      "learning_rate": 6.541113280947503e-06,
      "loss": 0.328,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1930
    },
    {
      "epoch": 0.6928983868087386,
      "grad_norm": 0.48582636452480815,
      "learning_rate": 6.536195838782704e-06,
      "loss": 0.3317,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1940
    },
    {
      "epoch": 0.6964700279778558,
      "grad_norm": 0.46098715412427904,
      "learning_rate": 6.5312540579760625e-06,
      "loss": 0.3335,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1950
    },
    {
      "epoch": 0.7000416691469731,
      "grad_norm": 0.4468284228904629,
      "learning_rate": 6.526287978141699e-06,
      "loss": 0.3269,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1960
    },
    {
      "epoch": 0.7036133103160902,
      "grad_norm": 0.4754658451034331,
      "learning_rate": 6.521297639088522e-06,
      "loss": 0.3271,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1970
    },
    {
      "epoch": 0.7071849514852074,
      "grad_norm": 0.45435138107973977,
      "learning_rate": 6.516283080819904e-06,
      "loss": 0.3299,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1980
    },
    {
      "epoch": 0.7107565926543247,
      "grad_norm": 0.449811476238864,
      "learning_rate": 6.511244343533364e-06,
      "loss": 0.3239,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 1990
    },
    {
      "epoch": 0.7143282338234419,
      "grad_norm": 0.43297219624166366,
      "learning_rate": 6.506181467620243e-06,
      "loss": 0.3289,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2000
    },
    {
      "epoch": 0.717899874992559,
      "grad_norm": 0.4396876046033842,
      "learning_rate": 6.5010944936653805e-06,
      "loss": 0.3247,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2010
    },
    {
      "epoch": 0.7214715161616763,
      "grad_norm": 0.4557973023145909,
      "learning_rate": 6.495983462446792e-06,
      "loss": 0.3256,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2020
    },
    {
      "epoch": 0.7250431573307935,
      "grad_norm": 0.5048088833153063,
      "learning_rate": 6.490848414935338e-06,
      "loss": 0.3242,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2030
    },
    {
      "epoch": 0.7286147984999107,
      "grad_norm": 0.46964680827760763,
      "learning_rate": 6.485689392294398e-06,
      "loss": 0.3337,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2040
    },
    {
      "epoch": 0.732186439669028,
      "grad_norm": 0.4182978089039866,
      "learning_rate": 6.480506435879541e-06,
      "loss": 0.3254,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2050
    },
    {
      "epoch": 0.7357580808381451,
      "grad_norm": 0.5087956625646197,
      "learning_rate": 6.475299587238192e-06,
      "loss": 0.3283,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2060
    },
    {
      "epoch": 0.7393297220072623,
      "grad_norm": 0.4433568329887951,
      "learning_rate": 6.470068888109301e-06,
      "loss": 0.328,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2070
    },
    {
      "epoch": 0.7429013631763796,
      "grad_norm": 0.40440774246534084,
      "learning_rate": 6.4648143804230044e-06,
      "loss": 0.3323,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2080
    },
    {
      "epoch": 0.7464730043454968,
      "grad_norm": 0.4776400773257993,
      "learning_rate": 6.459536106300296e-06,
      "loss": 0.33,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2090
    },
    {
      "epoch": 0.7500446455146139,
      "grad_norm": 0.48086135057212126,
      "learning_rate": 6.454234108052685e-06,
      "loss": 0.3284,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2100
    },
    {
      "epoch": 0.7536162866837312,
      "grad_norm": 0.45259232721638404,
      "learning_rate": 6.448908428181852e-06,
      "loss": 0.326,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2110
    },
    {
      "epoch": 0.7571879278528484,
      "grad_norm": 0.46227294968779126,
      "learning_rate": 6.443559109379319e-06,
      "loss": 0.3326,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2120
    },
    {
      "epoch": 0.7607595690219656,
      "grad_norm": 0.4815320401008922,
      "learning_rate": 6.4381861945261e-06,
      "loss": 0.327,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2130
    },
    {
      "epoch": 0.7643312101910829,
      "grad_norm": 0.46393575252024677,
      "learning_rate": 6.432789726692356e-06,
      "loss": 0.3302,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2140
    },
    {
      "epoch": 0.7679028513602,
      "grad_norm": 0.4195186834772109,
      "learning_rate": 6.427369749137057e-06,
      "loss": 0.3162,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2150
    },
    {
      "epoch": 0.7714744925293172,
      "grad_norm": 0.4537338691534887,
      "learning_rate": 6.421926305307627e-06,
      "loss": 0.3234,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2160
    },
    {
      "epoch": 0.7750461336984344,
      "grad_norm": 0.48382227197517164,
      "learning_rate": 6.416459438839601e-06,
      "loss": 0.3256,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2170
    },
    {
      "epoch": 0.7786177748675517,
      "grad_norm": 0.46259622419983354,
      "learning_rate": 6.410969193556274e-06,
      "loss": 0.3241,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2180
    },
    {
      "epoch": 0.7821894160366688,
      "grad_norm": 0.4300803430769861,
      "learning_rate": 6.40545561346835e-06,
      "loss": 0.3242,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2190
    },
    {
      "epoch": 0.785761057205786,
      "grad_norm": 0.42886334543184423,
      "learning_rate": 6.399918742773586e-06,
      "loss": 0.3203,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2200
    },
    {
      "epoch": 0.7893326983749033,
      "grad_norm": 0.436336380040389,
      "learning_rate": 6.394358625856444e-06,
      "loss": 0.3258,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2210
    },
    {
      "epoch": 0.7929043395440205,
      "grad_norm": 0.5165620808176424,
      "learning_rate": 6.388775307287727e-06,
      "loss": 0.3099,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2220
    },
    {
      "epoch": 0.7964759807131376,
      "grad_norm": 0.4104870809975225,
      "learning_rate": 6.383168831824231e-06,
      "loss": 0.3279,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2230
    },
    {
      "epoch": 0.8000476218822549,
      "grad_norm": 0.4312051340095396,
      "learning_rate": 6.377539244408378e-06,
      "loss": 0.3297,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2240
    },
    {
      "epoch": 0.8036192630513721,
      "grad_norm": 0.42973894851523453,
      "learning_rate": 6.371886590167861e-06,
      "loss": 0.3259,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2250
    },
    {
      "epoch": 0.8071909042204893,
      "grad_norm": 0.450430593545997,
      "learning_rate": 6.36621091441528e-06,
      "loss": 0.3287,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2260
    },
    {
      "epoch": 0.8107625453896066,
      "grad_norm": 0.4843313728163245,
      "learning_rate": 6.360512262647777e-06,
      "loss": 0.3195,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2270
    },
    {
      "epoch": 0.8143341865587237,
      "grad_norm": 0.4484783923353552,
      "learning_rate": 6.354790680546679e-06,
      "loss": 0.3171,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2280
    },
    {
      "epoch": 0.8179058277278409,
      "grad_norm": 0.4215719667731261,
      "learning_rate": 6.349046213977121e-06,
      "loss": 0.3169,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2290
    },
    {
      "epoch": 0.8214774688969582,
      "grad_norm": 0.4508726622466319,
      "learning_rate": 6.343278908987684e-06,
      "loss": 0.319,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2300
    },
    {
      "epoch": 0.8250491100660754,
      "grad_norm": 0.4361713908568727,
      "learning_rate": 6.337488811810029e-06,
      "loss": 0.3211,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2310
    },
    {
      "epoch": 0.8286207512351925,
      "grad_norm": 0.4816418184058236,
      "learning_rate": 6.33167596885852e-06,
      "loss": 0.3235,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2320
    },
    {
      "epoch": 0.8321923924043098,
      "grad_norm": 0.4589272175981858,
      "learning_rate": 6.325840426729856e-06,
      "loss": 0.3197,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2330
    },
    {
      "epoch": 0.835764033573427,
      "grad_norm": 0.43673536428623677,
      "learning_rate": 6.319982232202696e-06,
      "loss": 0.3205,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2340
    },
    {
      "epoch": 0.8393356747425442,
      "grad_norm": 0.4214827560121218,
      "learning_rate": 6.314101432237283e-06,
      "loss": 0.3198,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2350
    },
    {
      "epoch": 0.8429073159116615,
      "grad_norm": 0.43810471273328333,
      "learning_rate": 6.308198073975073e-06,
      "loss": 0.3146,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2360
    },
    {
      "epoch": 0.8464789570807786,
      "grad_norm": 0.45322674327393886,
      "learning_rate": 6.302272204738348e-06,
      "loss": 0.3146,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2370
    },
    {
      "epoch": 0.8500505982498958,
      "grad_norm": 0.44384180344721963,
      "learning_rate": 6.296323872029845e-06,
      "loss": 0.3113,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2380
    },
    {
      "epoch": 0.853622239419013,
      "grad_norm": 0.4318032922963888,
      "learning_rate": 6.29035312353237e-06,
      "loss": 0.3216,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2390
    },
    {
      "epoch": 0.8571938805881303,
      "grad_norm": 0.48276893773226826,
      "learning_rate": 6.284360007108418e-06,
      "loss": 0.3218,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2400
    },
    {
      "epoch": 0.8607655217572474,
      "grad_norm": 0.4432412247091493,
      "learning_rate": 6.278344570799791e-06,
      "loss": 0.323,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2410
    },
    {
      "epoch": 0.8643371629263646,
      "grad_norm": 0.4357439381549616,
      "learning_rate": 6.272306862827208e-06,
      "loss": 0.3127,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2420
    },
    {
      "epoch": 0.8679088040954819,
      "grad_norm": 0.4287466917293321,
      "learning_rate": 6.2662469315899215e-06,
      "loss": 0.3201,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2430
    },
    {
      "epoch": 0.8714804452645991,
      "grad_norm": 0.46939300578506366,
      "learning_rate": 6.260164825665332e-06,
      "loss": 0.3123,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2440
    },
    {
      "epoch": 0.8750520864337163,
      "grad_norm": 0.4358908165077067,
      "learning_rate": 6.254060593808594e-06,
      "loss": 0.3154,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2450
    },
    {
      "epoch": 0.8786237276028335,
      "grad_norm": 0.4691550559602913,
      "learning_rate": 6.247934284952225e-06,
      "loss": 0.3154,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2460
    },
    {
      "epoch": 0.8821953687719507,
      "grad_norm": 0.4614380220064387,
      "learning_rate": 6.241785948205721e-06,
      "loss": 0.3218,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2470
    },
    {
      "epoch": 0.8857670099410679,
      "grad_norm": 0.4490215516280512,
      "learning_rate": 6.235615632855151e-06,
      "loss": 0.313,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2480
    },
    {
      "epoch": 0.8893386511101852,
      "grad_norm": 0.45310821557712655,
      "learning_rate": 6.229423388362772e-06,
      "loss": 0.3239,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2490
    },
    {
      "epoch": 0.8929102922793023,
      "grad_norm": 0.4246377711390903,
      "learning_rate": 6.223209264366629e-06,
      "loss": 0.3178,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2500
    },
    {
      "epoch": 0.8964819334484195,
      "grad_norm": 0.437170171634512,
      "learning_rate": 6.216973310680151e-06,
      "loss": 0.3161,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2510
    },
    {
      "epoch": 0.9000535746175368,
      "grad_norm": 0.4100710907843204,
      "learning_rate": 6.2107155772917645e-06,
      "loss": 0.322,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2520
    },
    {
      "epoch": 0.903625215786654,
      "grad_norm": 0.4297080236276428,
      "learning_rate": 6.204436114364483e-06,
      "loss": 0.3088,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2530
    },
    {
      "epoch": 0.9071968569557711,
      "grad_norm": 0.4578128481232775,
      "learning_rate": 6.198134972235506e-06,
      "loss": 0.3116,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2540
    },
    {
      "epoch": 0.9107684981248884,
      "grad_norm": 0.4619368696236996,
      "learning_rate": 6.191812201415817e-06,
      "loss": 0.3108,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2550
    },
    {
      "epoch": 0.9143401392940056,
      "grad_norm": 0.4316502369623513,
      "learning_rate": 6.185467852589783e-06,
      "loss": 0.3076,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2560
    },
    {
      "epoch": 0.9179117804631228,
      "grad_norm": 0.46879215227469134,
      "learning_rate": 6.179101976614738e-06,
      "loss": 0.3079,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2570
    },
    {
      "epoch": 0.9214834216322401,
      "grad_norm": 0.45391009388979736,
      "learning_rate": 6.172714624520584e-06,
      "loss": 0.3117,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2580
    },
    {
      "epoch": 0.9250550628013572,
      "grad_norm": 0.41117818979235166,
      "learning_rate": 6.16630584750938e-06,
      "loss": 0.3085,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2590
    },
    {
      "epoch": 0.9286267039704744,
      "grad_norm": 0.41393645126366063,
      "learning_rate": 6.159875696954928e-06,
      "loss": 0.3158,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2600
    },
    {
      "epoch": 0.9321983451395917,
      "grad_norm": 0.43546721450403336,
      "learning_rate": 6.153424224402368e-06,
      "loss": 0.3114,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2610
    },
    {
      "epoch": 0.9357699863087089,
      "grad_norm": 0.47547951834025626,
      "learning_rate": 6.146951481567755e-06,
      "loss": 0.3159,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2620
    },
    {
      "epoch": 0.939341627477826,
      "grad_norm": 0.41613918798314703,
      "learning_rate": 6.1404575203376536e-06,
      "loss": 0.318,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2630
    },
    {
      "epoch": 0.9429132686469432,
      "grad_norm": 0.43579838968445417,
      "learning_rate": 6.133942392768719e-06,
      "loss": 0.3159,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2640
    },
    {
      "epoch": 0.9464849098160605,
      "grad_norm": 0.42779957417782555,
      "learning_rate": 6.1274061510872765e-06,
      "loss": 0.3171,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2650
    },
    {
      "epoch": 0.9500565509851777,
      "grad_norm": 0.4571516272464143,
      "learning_rate": 6.120848847688909e-06,
      "loss": 0.308,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2660
    },
    {
      "epoch": 0.953628192154295,
      "grad_norm": 0.4259836126699497,
      "learning_rate": 6.11427053513803e-06,
      "loss": 0.3081,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2670
    },
    {
      "epoch": 0.9571998333234121,
      "grad_norm": 0.45984729995716905,
      "learning_rate": 6.107671266167468e-06,
      "loss": 0.3163,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2680
    },
    {
      "epoch": 0.9607714744925293,
      "grad_norm": 0.4786910501462735,
      "learning_rate": 6.1010510936780425e-06,
      "loss": 0.3122,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2690
    },
    {
      "epoch": 0.9643431156616465,
      "grad_norm": 0.45855998268064024,
      "learning_rate": 6.094410070738137e-06,
      "loss": 0.3111,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2700
    },
    {
      "epoch": 0.9679147568307638,
      "grad_norm": 0.40749766688739447,
      "learning_rate": 6.087748250583276e-06,
      "loss": 0.3098,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2710
    },
    {
      "epoch": 0.9714863979998809,
      "grad_norm": 0.3971023012451066,
      "learning_rate": 6.081065686615698e-06,
      "loss": 0.3121,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2720
    },
    {
      "epoch": 0.9750580391689981,
      "grad_norm": 0.43110701814916075,
      "learning_rate": 6.074362432403927e-06,
      "loss": 0.3165,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2730
    },
    {
      "epoch": 0.9786296803381154,
      "grad_norm": 0.44353114381202113,
      "learning_rate": 6.067638541682344e-06,
      "loss": 0.3085,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2740
    },
    {
      "epoch": 0.9822013215072326,
      "grad_norm": 0.3889830628797175,
      "learning_rate": 6.060894068350758e-06,
      "loss": 0.3111,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2750
    },
    {
      "epoch": 0.9857729626763497,
      "grad_norm": 0.45027635846439507,
      "learning_rate": 6.0541290664739644e-06,
      "loss": 0.3149,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2760
    },
    {
      "epoch": 0.989344603845467,
      "grad_norm": 0.4203375605922012,
      "learning_rate": 6.047343590281326e-06,
      "loss": 0.3122,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2770
    },
    {
      "epoch": 0.9929162450145842,
      "grad_norm": 0.42453145771429407,
      "learning_rate": 6.040537694166331e-06,
      "loss": 0.3122,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2780
    },
    {
      "epoch": 0.9964878861837014,
      "grad_norm": 0.48274276529679444,
      "learning_rate": 6.0337114326861495e-06,
      "loss": 0.3125,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2790
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.4713684607192772,
      "learning_rate": 6.026864860561211e-06,
      "loss": 0.3211,
      "memory/device_mem_reserved(gib)": 48.61,
      "memory/max_mem_active(gib)": 43.25,
      "memory/max_mem_allocated(gib)": 42.69,
      "step": 2800
    },
    {
      "epoch": 1.0035716411691171,
      "grad_norm": 0.42558419555930044,
      "learning_rate": 6.019998032674756e-06,
      "loss": 0.3189,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2810
    },
    {
      "epoch": 1.0071432823382345,
      "grad_norm": 0.43216906153711016,
      "learning_rate": 6.0131110040724e-06,
      "loss": 0.3112,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2820
    },
    {
      "epoch": 1.0107149235073516,
      "grad_norm": 0.45448985837389827,
      "learning_rate": 6.006203829961688e-06,
      "loss": 0.3075,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2830
    },
    {
      "epoch": 1.014286564676469,
      "grad_norm": 0.41525982725225924,
      "learning_rate": 5.999276565711656e-06,
      "loss": 0.2989,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2840
    },
    {
      "epoch": 1.017858205845586,
      "grad_norm": 0.4198260335440655,
      "learning_rate": 5.992329266852385e-06,
      "loss": 0.3071,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2850
    },
    {
      "epoch": 1.0214298470147032,
      "grad_norm": 0.42331773555378355,
      "learning_rate": 5.98536198907456e-06,
      "loss": 0.3025,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2860
    },
    {
      "epoch": 1.0250014881838205,
      "grad_norm": 0.42277960958705246,
      "learning_rate": 5.978374788229015e-06,
      "loss": 0.3058,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2870
    },
    {
      "epoch": 1.0285731293529377,
      "grad_norm": 0.41102681519195966,
      "learning_rate": 5.971367720326295e-06,
      "loss": 0.3077,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2880
    },
    {
      "epoch": 1.0321447705220548,
      "grad_norm": 0.4156887229302285,
      "learning_rate": 5.9643408415362015e-06,
      "loss": 0.2976,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2890
    },
    {
      "epoch": 1.0357164116911721,
      "grad_norm": 0.40926887059288125,
      "learning_rate": 5.9572942081873415e-06,
      "loss": 0.3073,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2900
    },
    {
      "epoch": 1.0392880528602892,
      "grad_norm": 0.4181073533586528,
      "learning_rate": 5.950227876766679e-06,
      "loss": 0.3048,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2910
    },
    {
      "epoch": 1.0428596940294066,
      "grad_norm": 0.43301206694152716,
      "learning_rate": 5.943141903919084e-06,
      "loss": 0.3102,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2920
    },
    {
      "epoch": 1.0464313351985237,
      "grad_norm": 0.428644958607888,
      "learning_rate": 5.93603634644687e-06,
      "loss": 0.3069,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2930
    },
    {
      "epoch": 1.0500029763676408,
      "grad_norm": 0.4289954783985394,
      "learning_rate": 5.928911261309345e-06,
      "loss": 0.3062,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2940
    },
    {
      "epoch": 1.0535746175367582,
      "grad_norm": 0.40640231743067806,
      "learning_rate": 5.921766705622358e-06,
      "loss": 0.3057,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2950
    },
    {
      "epoch": 1.0571462587058753,
      "grad_norm": 0.39786241335597483,
      "learning_rate": 5.914602736657834e-06,
      "loss": 0.3014,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2960
    },
    {
      "epoch": 1.0607178998749927,
      "grad_norm": 0.42829381973477715,
      "learning_rate": 5.9074194118433185e-06,
      "loss": 0.2951,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2970
    },
    {
      "epoch": 1.0642895410441098,
      "grad_norm": 0.40677472977856,
      "learning_rate": 5.900216788761514e-06,
      "loss": 0.299,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2980
    },
    {
      "epoch": 1.067861182213227,
      "grad_norm": 0.40052402118235286,
      "learning_rate": 5.892994925149829e-06,
      "loss": 0.3081,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.4,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 2990
    },
    {
      "epoch": 1.0714328233823442,
      "grad_norm": 0.41164871053260205,
      "learning_rate": 5.885753878899898e-06,
      "loss": 0.2983,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3000
    },
    {
      "epoch": 1.0750044645514614,
      "grad_norm": 0.4398438793790258,
      "learning_rate": 5.878493708057132e-06,
      "loss": 0.296,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3010
    },
    {
      "epoch": 1.0785761057205785,
      "grad_norm": 0.4145451367489992,
      "learning_rate": 5.871214470820245e-06,
      "loss": 0.2953,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3020
    },
    {
      "epoch": 1.0821477468896958,
      "grad_norm": 0.3902504116999639,
      "learning_rate": 5.863916225540795e-06,
      "loss": 0.302,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3030
    },
    {
      "epoch": 1.085719388058813,
      "grad_norm": 0.4038545931689053,
      "learning_rate": 5.8565990307227055e-06,
      "loss": 0.2892,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3040
    },
    {
      "epoch": 1.0892910292279303,
      "grad_norm": 0.4002310932995622,
      "learning_rate": 5.849262945021807e-06,
      "loss": 0.2993,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3050
    },
    {
      "epoch": 1.0928626703970474,
      "grad_norm": 0.4200267477903967,
      "learning_rate": 5.84190802724536e-06,
      "loss": 0.2959,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3060
    },
    {
      "epoch": 1.0964343115661646,
      "grad_norm": 0.4118492608691234,
      "learning_rate": 5.834534336351588e-06,
      "loss": 0.2904,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3070
    },
    {
      "epoch": 1.100005952735282,
      "grad_norm": 0.37731728704380824,
      "learning_rate": 5.827141931449201e-06,
      "loss": 0.2931,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3080
    },
    {
      "epoch": 1.103577593904399,
      "grad_norm": 0.4257361126495842,
      "learning_rate": 5.819730871796925e-06,
      "loss": 0.3007,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3090
    },
    {
      "epoch": 1.1071492350735164,
      "grad_norm": 0.4627051529190304,
      "learning_rate": 5.812301216803024e-06,
      "loss": 0.2945,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3100
    },
    {
      "epoch": 1.1107208762426335,
      "grad_norm": 0.39052506920739677,
      "learning_rate": 5.804853026024826e-06,
      "loss": 0.2844,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3110
    },
    {
      "epoch": 1.1142925174117506,
      "grad_norm": 0.4183858283818629,
      "learning_rate": 5.797386359168247e-06,
      "loss": 0.2969,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3120
    },
    {
      "epoch": 1.117864158580868,
      "grad_norm": 0.38122190274713647,
      "learning_rate": 5.789901276087306e-06,
      "loss": 0.2903,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3130
    },
    {
      "epoch": 1.121435799749985,
      "grad_norm": 0.41997541586634773,
      "learning_rate": 5.782397836783651e-06,
      "loss": 0.2859,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3140
    },
    {
      "epoch": 1.1250074409191022,
      "grad_norm": 0.49411147636632047,
      "learning_rate": 5.774876101406081e-06,
      "loss": 0.2915,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3150
    },
    {
      "epoch": 1.1285790820882196,
      "grad_norm": 0.38342852409381073,
      "learning_rate": 5.767336130250054e-06,
      "loss": 0.2922,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3160
    },
    {
      "epoch": 1.1321507232573367,
      "grad_norm": 0.4209796316280021,
      "learning_rate": 5.759777983757208e-06,
      "loss": 0.2888,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3170
    },
    {
      "epoch": 1.135722364426454,
      "grad_norm": 0.4048539945241082,
      "learning_rate": 5.752201722514882e-06,
      "loss": 0.292,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3180
    },
    {
      "epoch": 1.1392940055955711,
      "grad_norm": 0.41621200257599267,
      "learning_rate": 5.7446074072556255e-06,
      "loss": 0.2877,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3190
    },
    {
      "epoch": 1.1428656467646885,
      "grad_norm": 0.45972371472591095,
      "learning_rate": 5.736995098856708e-06,
      "loss": 0.2907,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3200
    },
    {
      "epoch": 1.1464372879338056,
      "grad_norm": 0.4150892225825858,
      "learning_rate": 5.7293648583396364e-06,
      "loss": 0.2819,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3210
    },
    {
      "epoch": 1.1500089291029227,
      "grad_norm": 0.3953394886577025,
      "learning_rate": 5.7217167468696675e-06,
      "loss": 0.2869,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3220
    },
    {
      "epoch": 1.15358057027204,
      "grad_norm": 0.4043756771843379,
      "learning_rate": 5.714050825755311e-06,
      "loss": 0.2883,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3230
    },
    {
      "epoch": 1.1571522114411572,
      "grad_norm": 0.42513927984802846,
      "learning_rate": 5.706367156447842e-06,
      "loss": 0.2936,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3240
    },
    {
      "epoch": 1.1607238526102743,
      "grad_norm": 0.37727727646775006,
      "learning_rate": 5.698665800540812e-06,
      "loss": 0.2972,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3250
    },
    {
      "epoch": 1.1642954937793917,
      "grad_norm": 0.38826025836911765,
      "learning_rate": 5.690946819769544e-06,
      "loss": 0.2954,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3260
    },
    {
      "epoch": 1.1678671349485088,
      "grad_norm": 0.405329832879052,
      "learning_rate": 5.683210276010654e-06,
      "loss": 0.2915,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3270
    },
    {
      "epoch": 1.1714387761176261,
      "grad_norm": 0.4363084476787587,
      "learning_rate": 5.6754562312815345e-06,
      "loss": 0.291,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3280
    },
    {
      "epoch": 1.1750104172867433,
      "grad_norm": 0.4133239785031364,
      "learning_rate": 5.667684747739879e-06,
      "loss": 0.2974,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3290
    },
    {
      "epoch": 1.1785820584558604,
      "grad_norm": 0.38573210471666897,
      "learning_rate": 5.6598958876831675e-06,
      "loss": 0.2907,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3300
    },
    {
      "epoch": 1.1821536996249777,
      "grad_norm": 0.39168356395863324,
      "learning_rate": 5.652089713548175e-06,
      "loss": 0.2875,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3310
    },
    {
      "epoch": 1.1857253407940949,
      "grad_norm": 0.43833999524703754,
      "learning_rate": 5.644266287910469e-06,
      "loss": 0.2918,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3320
    },
    {
      "epoch": 1.1892969819632122,
      "grad_norm": 0.3739394107569196,
      "learning_rate": 5.636425673483907e-06,
      "loss": 0.2938,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3330
    },
    {
      "epoch": 1.1928686231323293,
      "grad_norm": 0.39393210278890395,
      "learning_rate": 5.628567933120135e-06,
      "loss": 0.287,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3340
    },
    {
      "epoch": 1.1964402643014465,
      "grad_norm": 0.42110108390228834,
      "learning_rate": 5.620693129808084e-06,
      "loss": 0.2839,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3350
    },
    {
      "epoch": 1.2000119054705638,
      "grad_norm": 0.45198733665344365,
      "learning_rate": 5.612801326673463e-06,
      "loss": 0.2956,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3360
    },
    {
      "epoch": 1.203583546639681,
      "grad_norm": 0.40498561338650063,
      "learning_rate": 5.604892586978256e-06,
      "loss": 0.2904,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3370
    },
    {
      "epoch": 1.207155187808798,
      "grad_norm": 0.44100061374306926,
      "learning_rate": 5.5969669741202115e-06,
      "loss": 0.2863,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3380
    },
    {
      "epoch": 1.2107268289779154,
      "grad_norm": 0.4159877214300284,
      "learning_rate": 5.589024551632338e-06,
      "loss": 0.2835,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3390
    },
    {
      "epoch": 1.2142984701470325,
      "grad_norm": 0.4114576141382402,
      "learning_rate": 5.58106538318239e-06,
      "loss": 0.2913,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3400
    },
    {
      "epoch": 1.2178701113161499,
      "grad_norm": 0.38152511059940936,
      "learning_rate": 5.5730895325723645e-06,
      "loss": 0.2839,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3410
    },
    {
      "epoch": 1.221441752485267,
      "grad_norm": 0.3988441810677571,
      "learning_rate": 5.5650970637379795e-06,
      "loss": 0.291,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3420
    },
    {
      "epoch": 1.225013393654384,
      "grad_norm": 0.40024145076228973,
      "learning_rate": 5.557088040748174e-06,
      "loss": 0.2849,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3430
    },
    {
      "epoch": 1.2285850348235015,
      "grad_norm": 18.581620195482078,
      "learning_rate": 5.549062527804584e-06,
      "loss": 0.2826,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3440
    },
    {
      "epoch": 1.2321566759926186,
      "grad_norm": 0.4030023053420153,
      "learning_rate": 5.5410205892410326e-06,
      "loss": 0.2806,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3450
    },
    {
      "epoch": 1.235728317161736,
      "grad_norm": 0.4515723074035561,
      "learning_rate": 5.53296228952301e-06,
      "loss": 0.2917,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3460
    },
    {
      "epoch": 1.239299958330853,
      "grad_norm": 0.39728025388029203,
      "learning_rate": 5.524887693247166e-06,
      "loss": 0.2947,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3470
    },
    {
      "epoch": 1.2428715994999702,
      "grad_norm": 0.3871968955057113,
      "learning_rate": 5.516796865140783e-06,
      "loss": 0.2879,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3480
    },
    {
      "epoch": 1.2464432406690875,
      "grad_norm": 0.4087098376285781,
      "learning_rate": 5.508689870061262e-06,
      "loss": 0.2877,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3490
    },
    {
      "epoch": 1.2500148818382046,
      "grad_norm": 0.3882691465242462,
      "learning_rate": 5.500566772995597e-06,
      "loss": 0.2797,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3500
    },
    {
      "epoch": 1.2535865230073218,
      "grad_norm": 0.4185654530957896,
      "learning_rate": 5.4924276390598625e-06,
      "loss": 0.2868,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3510
    },
    {
      "epoch": 1.257158164176439,
      "grad_norm": 0.4035064939077187,
      "learning_rate": 5.484272533498688e-06,
      "loss": 0.2881,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3520
    },
    {
      "epoch": 1.2607298053455562,
      "grad_norm": 0.3866791184579453,
      "learning_rate": 5.476101521684729e-06,
      "loss": 0.2796,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3530
    },
    {
      "epoch": 1.2643014465146734,
      "grad_norm": 0.4247476728802867,
      "learning_rate": 5.467914669118151e-06,
      "loss": 0.2818,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3540
    },
    {
      "epoch": 1.2678730876837907,
      "grad_norm": 0.4457076719062088,
      "learning_rate": 5.459712041426103e-06,
      "loss": 0.2821,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3550
    },
    {
      "epoch": 1.271444728852908,
      "grad_norm": 0.39931687643920494,
      "learning_rate": 5.451493704362188e-06,
      "loss": 0.2822,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3560
    },
    {
      "epoch": 1.2750163700220252,
      "grad_norm": 0.40299377985016827,
      "learning_rate": 5.443259723805939e-06,
      "loss": 0.2825,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3570
    },
    {
      "epoch": 1.2785880111911423,
      "grad_norm": 0.39622897594218864,
      "learning_rate": 5.435010165762289e-06,
      "loss": 0.2885,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3580
    },
    {
      "epoch": 1.2821596523602596,
      "grad_norm": 0.43394101353228975,
      "learning_rate": 5.426745096361043e-06,
      "loss": 0.2886,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3590
    },
    {
      "epoch": 1.2857312935293768,
      "grad_norm": 0.3824521403588983,
      "learning_rate": 5.418464581856346e-06,
      "loss": 0.2822,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3600
    },
    {
      "epoch": 1.2893029346984939,
      "grad_norm": 0.41532444786322925,
      "learning_rate": 5.4101686886261575e-06,
      "loss": 0.286,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3610
    },
    {
      "epoch": 1.2928745758676112,
      "grad_norm": 0.45648561913824676,
      "learning_rate": 5.4018574831717105e-06,
      "loss": 0.2843,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3620
    },
    {
      "epoch": 1.2964462170367284,
      "grad_norm": 0.40279789970652746,
      "learning_rate": 5.393531032116988e-06,
      "loss": 0.2803,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3630
    },
    {
      "epoch": 1.3000178582058455,
      "grad_norm": 0.41292629848412293,
      "learning_rate": 5.38518940220818e-06,
      "loss": 0.2837,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3640
    },
    {
      "epoch": 1.3035894993749628,
      "grad_norm": 0.3842527104062671,
      "learning_rate": 5.376832660313156e-06,
      "loss": 0.2837,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3650
    },
    {
      "epoch": 1.30716114054408,
      "grad_norm": 0.4265057399549405,
      "learning_rate": 5.368460873420923e-06,
      "loss": 0.2823,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3660
    },
    {
      "epoch": 1.3107327817131973,
      "grad_norm": 0.3861802078618478,
      "learning_rate": 5.360074108641092e-06,
      "loss": 0.2846,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3670
    },
    {
      "epoch": 1.3143044228823144,
      "grad_norm": 0.4162455814696529,
      "learning_rate": 5.3516724332033415e-06,
      "loss": 0.2817,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3680
    },
    {
      "epoch": 1.3178760640514318,
      "grad_norm": 0.41457828657902673,
      "learning_rate": 5.343255914456873e-06,
      "loss": 0.2824,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3690
    },
    {
      "epoch": 1.3214477052205489,
      "grad_norm": 0.3826207041057534,
      "learning_rate": 5.334824619869876e-06,
      "loss": 0.2961,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3700
    },
    {
      "epoch": 1.325019346389666,
      "grad_norm": 0.40031209082106967,
      "learning_rate": 5.326378617028985e-06,
      "loss": 0.2791,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3710
    },
    {
      "epoch": 1.3285909875587834,
      "grad_norm": 0.3708661212761246,
      "learning_rate": 5.3179179736387385e-06,
      "loss": 0.2874,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3720
    },
    {
      "epoch": 1.3321626287279005,
      "grad_norm": 0.38510593028444307,
      "learning_rate": 5.3094427575210366e-06,
      "loss": 0.2867,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3730
    },
    {
      "epoch": 1.3357342698970176,
      "grad_norm": 0.3819530538365875,
      "learning_rate": 5.300953036614595e-06,
      "loss": 0.2804,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3740
    },
    {
      "epoch": 1.339305911066135,
      "grad_norm": 0.3794797389478174,
      "learning_rate": 5.2924488789744055e-06,
      "loss": 0.2851,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3750
    },
    {
      "epoch": 1.342877552235252,
      "grad_norm": 0.41346602960853757,
      "learning_rate": 5.283930352771183e-06,
      "loss": 0.2762,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3760
    },
    {
      "epoch": 1.3464491934043692,
      "grad_norm": 0.4037130991475417,
      "learning_rate": 5.275397526290827e-06,
      "loss": 0.2865,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3770
    },
    {
      "epoch": 1.3500208345734865,
      "grad_norm": 0.42726973214522196,
      "learning_rate": 5.266850467933866e-06,
      "loss": 0.281,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3780
    },
    {
      "epoch": 1.3535924757426037,
      "grad_norm": 0.43061681622670994,
      "learning_rate": 5.258289246214919e-06,
      "loss": 0.2811,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3790
    },
    {
      "epoch": 1.357164116911721,
      "grad_norm": 0.4380949060565935,
      "learning_rate": 5.249713929762136e-06,
      "loss": 0.2828,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3800
    },
    {
      "epoch": 1.3607357580808381,
      "grad_norm": 0.47620516316901934,
      "learning_rate": 5.241124587316655e-06,
      "loss": 0.2793,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3810
    },
    {
      "epoch": 1.3643073992499555,
      "grad_norm": 0.3966834578414604,
      "learning_rate": 5.2325212877320475e-06,
      "loss": 0.2832,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3820
    },
    {
      "epoch": 1.3678790404190726,
      "grad_norm": 0.41841887208116907,
      "learning_rate": 5.223904099973769e-06,
      "loss": 0.2776,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3830
    },
    {
      "epoch": 1.3714506815881897,
      "grad_norm": 0.3887158923993312,
      "learning_rate": 5.215273093118605e-06,
      "loss": 0.2834,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3840
    },
    {
      "epoch": 1.375022322757307,
      "grad_norm": 0.38605915567228644,
      "learning_rate": 5.2066283363541145e-06,
      "loss": 0.2824,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3850
    },
    {
      "epoch": 1.3785939639264242,
      "grad_norm": 0.3735164396659493,
      "learning_rate": 5.197969898978083e-06,
      "loss": 0.2846,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3860
    },
    {
      "epoch": 1.3821656050955413,
      "grad_norm": 0.38639038722023733,
      "learning_rate": 5.189297850397956e-06,
      "loss": 0.2833,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3870
    },
    {
      "epoch": 1.3857372462646587,
      "grad_norm": 0.38176503042886195,
      "learning_rate": 5.180612260130293e-06,
      "loss": 0.291,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3880
    },
    {
      "epoch": 1.3893088874337758,
      "grad_norm": 0.4012566858870177,
      "learning_rate": 5.171913197800206e-06,
      "loss": 0.2709,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3890
    },
    {
      "epoch": 1.392880528602893,
      "grad_norm": 0.4285411453903363,
      "learning_rate": 5.1632007331407955e-06,
      "loss": 0.275,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3900
    },
    {
      "epoch": 1.3964521697720103,
      "grad_norm": 0.43737881415448876,
      "learning_rate": 5.1544749359926055e-06,
      "loss": 0.2811,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3910
    },
    {
      "epoch": 1.4000238109411274,
      "grad_norm": 0.4077413602492253,
      "learning_rate": 5.1457358763030504e-06,
      "loss": 0.2783,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3920
    },
    {
      "epoch": 1.4035954521102447,
      "grad_norm": 0.3921345244240013,
      "learning_rate": 5.13698362412586e-06,
      "loss": 0.2887,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3930
    },
    {
      "epoch": 1.4071670932793618,
      "grad_norm": 0.41078975111561067,
      "learning_rate": 5.128218249620516e-06,
      "loss": 0.2779,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3940
    },
    {
      "epoch": 1.4107387344484792,
      "grad_norm": 0.4733932988022064,
      "learning_rate": 5.119439823051695e-06,
      "loss": 0.2808,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3950
    },
    {
      "epoch": 1.4143103756175963,
      "grad_norm": 0.42977751530797087,
      "learning_rate": 5.110648414788696e-06,
      "loss": 0.2824,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3960
    },
    {
      "epoch": 1.4178820167867134,
      "grad_norm": 0.4072718628555276,
      "learning_rate": 5.101844095304886e-06,
      "loss": 0.2783,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3970
    },
    {
      "epoch": 1.4214536579558308,
      "grad_norm": 0.3867375629277297,
      "learning_rate": 5.093026935177127e-06,
      "loss": 0.2739,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3980
    },
    {
      "epoch": 1.425025299124948,
      "grad_norm": 0.4270140563186022,
      "learning_rate": 5.084197005085215e-06,
      "loss": 0.2839,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 3990
    },
    {
      "epoch": 1.428596940294065,
      "grad_norm": 0.3952561361354359,
      "learning_rate": 5.075354375811313e-06,
      "loss": 0.2824,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4000
    },
    {
      "epoch": 1.4321685814631824,
      "grad_norm": 0.42309739207423486,
      "learning_rate": 5.066499118239381e-06,
      "loss": 0.2841,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4010
    },
    {
      "epoch": 1.4357402226322995,
      "grad_norm": 0.4393277747654901,
      "learning_rate": 5.05763130335461e-06,
      "loss": 0.2768,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4020
    },
    {
      "epoch": 1.4393118638014166,
      "grad_norm": 0.42584412565699653,
      "learning_rate": 5.048751002242853e-06,
      "loss": 0.2863,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4030
    },
    {
      "epoch": 1.442883504970534,
      "grad_norm": 0.4411596536003725,
      "learning_rate": 5.039858286090056e-06,
      "loss": 0.2849,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4040
    },
    {
      "epoch": 1.446455146139651,
      "grad_norm": 0.41130998429424936,
      "learning_rate": 5.030953226181682e-06,
      "loss": 0.2799,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4050
    },
    {
      "epoch": 1.4500267873087684,
      "grad_norm": 0.4154103752019673,
      "learning_rate": 5.022035893902147e-06,
      "loss": 0.2823,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4060
    },
    {
      "epoch": 1.4535984284778856,
      "grad_norm": 0.4208114402245905,
      "learning_rate": 5.013106360734244e-06,
      "loss": 0.2691,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4070
    },
    {
      "epoch": 1.457170069647003,
      "grad_norm": 0.43416877401764065,
      "learning_rate": 5.0041646982585694e-06,
      "loss": 0.2874,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4080
    },
    {
      "epoch": 1.46074171081612,
      "grad_norm": 0.4893174023257388,
      "learning_rate": 4.9952109781529494e-06,
      "loss": 0.2787,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4090
    },
    {
      "epoch": 1.4643133519852372,
      "grad_norm": 0.4375592068849344,
      "learning_rate": 4.986245272191866e-06,
      "loss": 0.2846,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4100
    },
    {
      "epoch": 1.4678849931543545,
      "grad_norm": 0.41720630958175864,
      "learning_rate": 4.977267652245887e-06,
      "loss": 0.2831,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4110
    },
    {
      "epoch": 1.4714566343234716,
      "grad_norm": 0.38893766128889445,
      "learning_rate": 4.968278190281076e-06,
      "loss": 0.2801,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4120
    },
    {
      "epoch": 1.4750282754925887,
      "grad_norm": 0.40586605542464455,
      "learning_rate": 4.959276958358434e-06,
      "loss": 0.2782,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4130
    },
    {
      "epoch": 1.478599916661706,
      "grad_norm": 0.39331036478589604,
      "learning_rate": 4.9502640286333e-06,
      "loss": 0.2805,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4140
    },
    {
      "epoch": 1.4821715578308232,
      "grad_norm": 0.42690856798454524,
      "learning_rate": 4.941239473354793e-06,
      "loss": 0.2757,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4150
    },
    {
      "epoch": 1.4857431989999403,
      "grad_norm": 0.42353429308127194,
      "learning_rate": 4.932203364865227e-06,
      "loss": 0.2853,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4160
    },
    {
      "epoch": 1.4893148401690577,
      "grad_norm": 0.46112729709385236,
      "learning_rate": 4.923155775599519e-06,
      "loss": 0.2679,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4170
    },
    {
      "epoch": 1.4928864813381748,
      "grad_norm": 0.4161270630543364,
      "learning_rate": 4.914096778084623e-06,
      "loss": 0.2813,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4180
    },
    {
      "epoch": 1.4964581225072922,
      "grad_norm": 0.4339245152353466,
      "learning_rate": 4.905026444938945e-06,
      "loss": 0.2706,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4190
    },
    {
      "epoch": 1.5000297636764093,
      "grad_norm": 0.40007840075097373,
      "learning_rate": 4.895944848871756e-06,
      "loss": 0.2807,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4200
    },
    {
      "epoch": 1.5036014048455266,
      "grad_norm": 0.4009324464084967,
      "learning_rate": 4.886852062682613e-06,
      "loss": 0.2728,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4210
    },
    {
      "epoch": 1.5071730460146437,
      "grad_norm": 0.39202018659117227,
      "learning_rate": 4.8777481592607775e-06,
      "loss": 0.2765,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4220
    },
    {
      "epoch": 1.5107446871837609,
      "grad_norm": 0.4112480071981047,
      "learning_rate": 4.868633211584625e-06,
      "loss": 0.2701,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4230
    },
    {
      "epoch": 1.5143163283528782,
      "grad_norm": 0.41390732975997946,
      "learning_rate": 4.859507292721066e-06,
      "loss": 0.2801,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4240
    },
    {
      "epoch": 1.5178879695219953,
      "grad_norm": 0.4401971189377111,
      "learning_rate": 4.850370475824957e-06,
      "loss": 0.2791,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4250
    },
    {
      "epoch": 1.5214596106911125,
      "grad_norm": 0.40944175778293085,
      "learning_rate": 4.841222834138513e-06,
      "loss": 0.2725,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4260
    },
    {
      "epoch": 1.5250312518602298,
      "grad_norm": 0.4006029556225281,
      "learning_rate": 4.832064440990725e-06,
      "loss": 0.2731,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4270
    },
    {
      "epoch": 1.528602893029347,
      "grad_norm": 0.3895187645181517,
      "learning_rate": 4.8228953697967705e-06,
      "loss": 0.2799,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4280
    },
    {
      "epoch": 1.532174534198464,
      "grad_norm": 0.43446583393388194,
      "learning_rate": 4.813715694057419e-06,
      "loss": 0.2706,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4290
    },
    {
      "epoch": 1.5357461753675814,
      "grad_norm": 0.4365210838133416,
      "learning_rate": 4.804525487358454e-06,
      "loss": 0.2748,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4300
    },
    {
      "epoch": 1.5393178165366987,
      "grad_norm": 0.46735370374372087,
      "learning_rate": 4.7953248233700725e-06,
      "loss": 0.272,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4310
    },
    {
      "epoch": 1.5428894577058159,
      "grad_norm": 0.42778351347035765,
      "learning_rate": 4.7861137758463e-06,
      "loss": 0.268,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4320
    },
    {
      "epoch": 1.546461098874933,
      "grad_norm": 0.40402901670951247,
      "learning_rate": 4.776892418624396e-06,
      "loss": 0.2784,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4330
    },
    {
      "epoch": 1.5500327400440503,
      "grad_norm": 0.42578918369148505,
      "learning_rate": 4.767660825624269e-06,
      "loss": 0.2739,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4340
    },
    {
      "epoch": 1.5536043812131675,
      "grad_norm": 0.41895323402321494,
      "learning_rate": 4.758419070847877e-06,
      "loss": 0.2783,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4350
    },
    {
      "epoch": 1.5571760223822846,
      "grad_norm": 0.4321907650265535,
      "learning_rate": 4.749167228378634e-06,
      "loss": 0.2727,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4360
    },
    {
      "epoch": 1.560747663551402,
      "grad_norm": 0.449385172204246,
      "learning_rate": 4.739905372380822e-06,
      "loss": 0.2754,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4370
    },
    {
      "epoch": 1.564319304720519,
      "grad_norm": 0.40983330954131525,
      "learning_rate": 4.730633577098992e-06,
      "loss": 0.2653,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4380
    },
    {
      "epoch": 1.5678909458896362,
      "grad_norm": 0.4099672865787947,
      "learning_rate": 4.721351916857368e-06,
      "loss": 0.2743,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4390
    },
    {
      "epoch": 1.5714625870587535,
      "grad_norm": 0.449638292162348,
      "learning_rate": 4.7120604660592564e-06,
      "loss": 0.2798,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4400
    },
    {
      "epoch": 1.5750342282278709,
      "grad_norm": 0.4888696074836512,
      "learning_rate": 4.702759299186444e-06,
      "loss": 0.2785,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4410
    },
    {
      "epoch": 1.5786058693969878,
      "grad_norm": 0.41140229166260467,
      "learning_rate": 4.693448490798603e-06,
      "loss": 0.2768,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4420
    },
    {
      "epoch": 1.5821775105661051,
      "grad_norm": 0.3965457289700049,
      "learning_rate": 4.684128115532697e-06,
      "loss": 0.2852,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4430
    },
    {
      "epoch": 1.5857491517352225,
      "grad_norm": 0.40893457249676096,
      "learning_rate": 4.674798248102374e-06,
      "loss": 0.2767,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4440
    },
    {
      "epoch": 1.5893207929043396,
      "grad_norm": 0.40947270681652376,
      "learning_rate": 4.665458963297376e-06,
      "loss": 0.2747,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4450
    },
    {
      "epoch": 1.5928924340734567,
      "grad_norm": 0.43217919095328655,
      "learning_rate": 4.656110335982939e-06,
      "loss": 0.2688,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4460
    },
    {
      "epoch": 1.596464075242574,
      "grad_norm": 0.4048417118383221,
      "learning_rate": 4.6467524410991815e-06,
      "loss": 0.2737,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4470
    },
    {
      "epoch": 1.6000357164116912,
      "grad_norm": 0.40615664984204264,
      "learning_rate": 4.637385353660521e-06,
      "loss": 0.2747,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4480
    },
    {
      "epoch": 1.6036073575808083,
      "grad_norm": 0.4556450981794944,
      "learning_rate": 4.62800914875506e-06,
      "loss": 0.2771,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4490
    },
    {
      "epoch": 1.6071789987499256,
      "grad_norm": 0.4456452650392903,
      "learning_rate": 4.618623901543989e-06,
      "loss": 0.2688,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4500
    },
    {
      "epoch": 1.6107506399190428,
      "grad_norm": 0.4072795804451177,
      "learning_rate": 4.609229687260981e-06,
      "loss": 0.2683,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4510
    },
    {
      "epoch": 1.61432228108816,
      "grad_norm": 0.3987863022162168,
      "learning_rate": 4.599826581211594e-06,
      "loss": 0.2668,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4520
    },
    {
      "epoch": 1.6178939222572772,
      "grad_norm": 0.42547111660934306,
      "learning_rate": 4.590414658772663e-06,
      "loss": 0.2763,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4530
    },
    {
      "epoch": 1.6214655634263946,
      "grad_norm": 0.41478885940190585,
      "learning_rate": 4.580993995391694e-06,
      "loss": 0.2765,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4540
    },
    {
      "epoch": 1.6250372045955115,
      "grad_norm": 0.4198144043488199,
      "learning_rate": 4.5715646665862645e-06,
      "loss": 0.2698,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4550
    },
    {
      "epoch": 1.6286088457646288,
      "grad_norm": 0.4179570184140089,
      "learning_rate": 4.562126747943415e-06,
      "loss": 0.2753,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4560
    },
    {
      "epoch": 1.6321804869337462,
      "grad_norm": 0.4407922084504109,
      "learning_rate": 4.5526803151190415e-06,
      "loss": 0.2758,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4570
    },
    {
      "epoch": 1.6357521281028633,
      "grad_norm": 0.5289684644845591,
      "learning_rate": 4.543225443837295e-06,
      "loss": 0.2788,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4580
    },
    {
      "epoch": 1.6393237692719804,
      "grad_norm": 0.4428322641436808,
      "learning_rate": 4.5337622098899675e-06,
      "loss": 0.2732,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4590
    },
    {
      "epoch": 1.6428954104410978,
      "grad_norm": 0.4552104082227244,
      "learning_rate": 4.5242906891358866e-06,
      "loss": 0.2698,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4600
    },
    {
      "epoch": 1.646467051610215,
      "grad_norm": 0.4677528492713541,
      "learning_rate": 4.514810957500311e-06,
      "loss": 0.2703,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4610
    },
    {
      "epoch": 1.650038692779332,
      "grad_norm": 0.45043763627660394,
      "learning_rate": 4.5053230909743195e-06,
      "loss": 0.265,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4620
    },
    {
      "epoch": 1.6536103339484494,
      "grad_norm": 0.451616136811452,
      "learning_rate": 4.495827165614198e-06,
      "loss": 0.2777,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4630
    },
    {
      "epoch": 1.6571819751175665,
      "grad_norm": 0.4281517840545686,
      "learning_rate": 4.486323257540838e-06,
      "loss": 0.2754,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4640
    },
    {
      "epoch": 1.6607536162866836,
      "grad_norm": 0.420818894121793,
      "learning_rate": 4.476811442939116e-06,
      "loss": 0.276,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4650
    },
    {
      "epoch": 1.664325257455801,
      "grad_norm": 0.4149963360632429,
      "learning_rate": 4.467291798057295e-06,
      "loss": 0.2781,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4660
    },
    {
      "epoch": 1.6678968986249183,
      "grad_norm": 0.4336018015531387,
      "learning_rate": 4.4577643992064e-06,
      "loss": 0.2654,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4670
    },
    {
      "epoch": 1.6714685397940352,
      "grad_norm": 0.41195956330157535,
      "learning_rate": 4.44822932275962e-06,
      "loss": 0.2725,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4680
    },
    {
      "epoch": 1.6750401809631525,
      "grad_norm": 0.4234398479845422,
      "learning_rate": 4.438686645151683e-06,
      "loss": 0.2787,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4690
    },
    {
      "epoch": 1.67861182213227,
      "grad_norm": 0.42731758155771205,
      "learning_rate": 4.429136442878253e-06,
      "loss": 0.2715,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4700
    },
    {
      "epoch": 1.682183463301387,
      "grad_norm": 0.42562808585130063,
      "learning_rate": 4.419578792495311e-06,
      "loss": 0.2737,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4710
    },
    {
      "epoch": 1.6857551044705041,
      "grad_norm": 0.4421430973069684,
      "learning_rate": 4.410013770618545e-06,
      "loss": 0.27,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4720
    },
    {
      "epoch": 1.6893267456396215,
      "grad_norm": 0.43191492739674137,
      "learning_rate": 4.4004414539227295e-06,
      "loss": 0.2712,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4730
    },
    {
      "epoch": 1.6928983868087386,
      "grad_norm": 0.4334260693657613,
      "learning_rate": 4.3908619191411235e-06,
      "loss": 0.2748,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4740
    },
    {
      "epoch": 1.6964700279778557,
      "grad_norm": 0.41696073544153367,
      "learning_rate": 4.3812752430648395e-06,
      "loss": 0.2779,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4750
    },
    {
      "epoch": 1.700041669146973,
      "grad_norm": 0.4348794170591351,
      "learning_rate": 4.371681502542241e-06,
      "loss": 0.2711,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4760
    },
    {
      "epoch": 1.7036133103160902,
      "grad_norm": 0.492407893353495,
      "learning_rate": 4.362080774478316e-06,
      "loss": 0.2707,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4770
    },
    {
      "epoch": 1.7071849514852073,
      "grad_norm": 0.45024503624278867,
      "learning_rate": 4.352473135834074e-06,
      "loss": 0.2742,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4780
    },
    {
      "epoch": 1.7107565926543247,
      "grad_norm": 0.45700989872478603,
      "learning_rate": 4.342858663625912e-06,
      "loss": 0.2682,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4790
    },
    {
      "epoch": 1.714328233823442,
      "grad_norm": 0.41115833866014856,
      "learning_rate": 4.333237434925012e-06,
      "loss": 0.2731,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4800
    },
    {
      "epoch": 1.717899874992559,
      "grad_norm": 0.4229273901374492,
      "learning_rate": 4.323609526856713e-06,
      "loss": 0.2696,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4810
    },
    {
      "epoch": 1.7214715161616763,
      "grad_norm": 0.42165606422875274,
      "learning_rate": 4.3139750165999e-06,
      "loss": 0.2702,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4820
    },
    {
      "epoch": 1.7250431573307936,
      "grad_norm": 0.42215858922988486,
      "learning_rate": 4.3043339813863825e-06,
      "loss": 0.269,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4830
    },
    {
      "epoch": 1.7286147984999107,
      "grad_norm": 0.4083178726418061,
      "learning_rate": 4.294686498500272e-06,
      "loss": 0.2793,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4840
    },
    {
      "epoch": 1.7321864396690279,
      "grad_norm": 0.46153757788947947,
      "learning_rate": 4.285032645277368e-06,
      "loss": 0.2706,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4850
    },
    {
      "epoch": 1.7357580808381452,
      "grad_norm": 0.47338863262420977,
      "learning_rate": 4.275372499104534e-06,
      "loss": 0.2734,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4860
    },
    {
      "epoch": 1.7393297220072623,
      "grad_norm": 0.4257009544961603,
      "learning_rate": 4.265706137419082e-06,
      "loss": 0.2732,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4870
    },
    {
      "epoch": 1.7429013631763794,
      "grad_norm": 0.432685871073769,
      "learning_rate": 4.256033637708144e-06,
      "loss": 0.2778,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4880
    },
    {
      "epoch": 1.7464730043454968,
      "grad_norm": 0.46972048535868977,
      "learning_rate": 4.246355077508058e-06,
      "loss": 0.2756,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4890
    },
    {
      "epoch": 1.750044645514614,
      "grad_norm": 0.43325958278072135,
      "learning_rate": 4.2366705344037455e-06,
      "loss": 0.2741,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4900
    },
    {
      "epoch": 1.753616286683731,
      "grad_norm": 0.44379386526215314,
      "learning_rate": 4.226980086028085e-06,
      "loss": 0.2721,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4910
    },
    {
      "epoch": 1.7571879278528484,
      "grad_norm": 0.5008628676566734,
      "learning_rate": 4.217283810061295e-06,
      "loss": 0.2786,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4920
    },
    {
      "epoch": 1.7607595690219657,
      "grad_norm": 0.4265989351603266,
      "learning_rate": 4.207581784230304e-06,
      "loss": 0.2733,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4930
    },
    {
      "epoch": 1.7643312101910829,
      "grad_norm": 0.4410937655399,
      "learning_rate": 4.197874086308141e-06,
      "loss": 0.2765,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4940
    },
    {
      "epoch": 1.7679028513602,
      "grad_norm": 0.43772683050066985,
      "learning_rate": 4.188160794113295e-06,
      "loss": 0.2621,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4950
    },
    {
      "epoch": 1.7714744925293173,
      "grad_norm": 0.44596352642090287,
      "learning_rate": 4.178441985509104e-06,
      "loss": 0.2699,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4960
    },
    {
      "epoch": 1.7750461336984344,
      "grad_norm": 0.43919374270562106,
      "learning_rate": 4.168717738403124e-06,
      "loss": 0.2724,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4970
    },
    {
      "epoch": 1.7786177748675516,
      "grad_norm": 0.409663218208272,
      "learning_rate": 4.158988130746512e-06,
      "loss": 0.2709,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4980
    },
    {
      "epoch": 1.782189416036669,
      "grad_norm": 0.40778352844560883,
      "learning_rate": 4.149253240533389e-06,
      "loss": 0.2708,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 4990
    },
    {
      "epoch": 1.785761057205786,
      "grad_norm": 0.3988447330223531,
      "learning_rate": 4.139513145800226e-06,
      "loss": 0.2674,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5000
    },
    {
      "epoch": 1.7893326983749032,
      "grad_norm": 0.4197454963211274,
      "learning_rate": 4.129767924625215e-06,
      "loss": 0.2727,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5010
    },
    {
      "epoch": 1.7929043395440205,
      "grad_norm": 0.45641210622438827,
      "learning_rate": 4.12001765512764e-06,
      "loss": 0.2567,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5020
    },
    {
      "epoch": 1.7964759807131376,
      "grad_norm": 0.4218050555962391,
      "learning_rate": 4.110262415467254e-06,
      "loss": 0.2755,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5030
    },
    {
      "epoch": 1.8000476218822548,
      "grad_norm": 0.41944541731559504,
      "learning_rate": 4.100502283843651e-06,
      "loss": 0.278,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5040
    },
    {
      "epoch": 1.803619263051372,
      "grad_norm": 0.4238198211524272,
      "learning_rate": 4.0907373384956415e-06,
      "loss": 0.2731,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5050
    },
    {
      "epoch": 1.8071909042204894,
      "grad_norm": 0.42139886595109105,
      "learning_rate": 4.080967657700623e-06,
      "loss": 0.277,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5060
    },
    {
      "epoch": 1.8107625453896066,
      "grad_norm": 0.4294097732318922,
      "learning_rate": 4.071193319773952e-06,
      "loss": 0.2666,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5070
    },
    {
      "epoch": 1.8143341865587237,
      "grad_norm": 0.4088369335354677,
      "learning_rate": 4.06141440306832e-06,
      "loss": 0.265,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5080
    },
    {
      "epoch": 1.817905827727841,
      "grad_norm": 0.4325090741382579,
      "learning_rate": 4.051630985973116e-06,
      "loss": 0.2649,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5090
    },
    {
      "epoch": 1.8214774688969582,
      "grad_norm": 0.4782429273760244,
      "learning_rate": 4.041843146913814e-06,
      "loss": 0.2666,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5100
    },
    {
      "epoch": 1.8250491100660753,
      "grad_norm": 0.44116784442799345,
      "learning_rate": 4.03205096435133e-06,
      "loss": 0.2696,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5110
    },
    {
      "epoch": 1.8286207512351926,
      "grad_norm": 0.41397328741000383,
      "learning_rate": 4.022254516781399e-06,
      "loss": 0.2717,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5120
    },
    {
      "epoch": 1.8321923924043098,
      "grad_norm": 0.45756084112038137,
      "learning_rate": 4.012453882733945e-06,
      "loss": 0.2678,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5130
    },
    {
      "epoch": 1.8357640335734269,
      "grad_norm": 0.4307519036701089,
      "learning_rate": 4.00264914077245e-06,
      "loss": 0.2685,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5140
    },
    {
      "epoch": 1.8393356747425442,
      "grad_norm": 0.46282872085965715,
      "learning_rate": 3.99284036949333e-06,
      "loss": 0.2684,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5150
    },
    {
      "epoch": 1.8429073159116616,
      "grad_norm": 0.44521556978523347,
      "learning_rate": 3.983027647525297e-06,
      "loss": 0.2629,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5160
    },
    {
      "epoch": 1.8464789570807785,
      "grad_norm": 0.4335775967545023,
      "learning_rate": 3.973211053528732e-06,
      "loss": 0.2638,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5170
    },
    {
      "epoch": 1.8500505982498958,
      "grad_norm": 0.4484527647303034,
      "learning_rate": 3.963390666195058e-06,
      "loss": 0.2601,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5180
    },
    {
      "epoch": 1.8536222394190132,
      "grad_norm": 0.4147772650920792,
      "learning_rate": 3.953566564246102e-06,
      "loss": 0.2708,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5190
    },
    {
      "epoch": 1.8571938805881303,
      "grad_norm": 0.42238734977530235,
      "learning_rate": 3.9437388264334726e-06,
      "loss": 0.2711,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5200
    },
    {
      "epoch": 1.8607655217572474,
      "grad_norm": 0.4676602900276513,
      "learning_rate": 3.933907531537919e-06,
      "loss": 0.2724,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5210
    },
    {
      "epoch": 1.8643371629263648,
      "grad_norm": 0.4550684970676724,
      "learning_rate": 3.92407275836871e-06,
      "loss": 0.2625,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5220
    },
    {
      "epoch": 1.8679088040954819,
      "grad_norm": 0.39876614644325137,
      "learning_rate": 3.914234585762992e-06,
      "loss": 0.2699,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5230
    },
    {
      "epoch": 1.871480445264599,
      "grad_norm": 0.45192453682747524,
      "learning_rate": 3.9043930925851644e-06,
      "loss": 0.2616,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5240
    },
    {
      "epoch": 1.8750520864337163,
      "grad_norm": 0.4260929665151792,
      "learning_rate": 3.894548357726243e-06,
      "loss": 0.2645,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5250
    },
    {
      "epoch": 1.8786237276028335,
      "grad_norm": 0.43944022835650526,
      "learning_rate": 3.884700460103234e-06,
      "loss": 0.2652,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5260
    },
    {
      "epoch": 1.8821953687719506,
      "grad_norm": 0.45508403385278373,
      "learning_rate": 3.874849478658489e-06,
      "loss": 0.272,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5270
    },
    {
      "epoch": 1.885767009941068,
      "grad_norm": 0.41091398036572885,
      "learning_rate": 3.864995492359088e-06,
      "loss": 0.2628,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5280
    },
    {
      "epoch": 1.8893386511101853,
      "grad_norm": 0.4605821484406852,
      "learning_rate": 3.8551385801961935e-06,
      "loss": 0.2737,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5290
    },
    {
      "epoch": 1.8929102922793022,
      "grad_norm": 0.43068610457940865,
      "learning_rate": 3.845278821184424e-06,
      "loss": 0.2681,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5300
    },
    {
      "epoch": 1.8964819334484195,
      "grad_norm": 0.43164071390264,
      "learning_rate": 3.835416294361219e-06,
      "loss": 0.2662,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5310
    },
    {
      "epoch": 1.9000535746175369,
      "grad_norm": 0.3967008822555253,
      "learning_rate": 3.825551078786203e-06,
      "loss": 0.2726,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5320
    },
    {
      "epoch": 1.903625215786654,
      "grad_norm": 0.44872440715524753,
      "learning_rate": 3.815683253540557e-06,
      "loss": 0.2595,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5330
    },
    {
      "epoch": 1.9071968569557711,
      "grad_norm": 0.47048031848498323,
      "learning_rate": 3.8058128977263796e-06,
      "loss": 0.2625,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5340
    },
    {
      "epoch": 1.9107684981248885,
      "grad_norm": 0.43319523625383444,
      "learning_rate": 3.7959400904660573e-06,
      "loss": 0.2614,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5350
    },
    {
      "epoch": 1.9143401392940056,
      "grad_norm": 0.4520347805433191,
      "learning_rate": 3.786064910901623e-06,
      "loss": 0.2581,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5360
    },
    {
      "epoch": 1.9179117804631227,
      "grad_norm": 0.44455950032323127,
      "learning_rate": 3.7761874381941305e-06,
      "loss": 0.2587,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5370
    },
    {
      "epoch": 1.92148342163224,
      "grad_norm": 0.4238561873196431,
      "learning_rate": 3.7663077515230163e-06,
      "loss": 0.2629,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5380
    },
    {
      "epoch": 1.9250550628013572,
      "grad_norm": 0.46891711405964687,
      "learning_rate": 3.75642593008546e-06,
      "loss": 0.2594,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5390
    },
    {
      "epoch": 1.9286267039704743,
      "grad_norm": 0.40249725381883916,
      "learning_rate": 3.7465420530957588e-06,
      "loss": 0.2673,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5400
    },
    {
      "epoch": 1.9321983451395917,
      "grad_norm": 0.4369159681597768,
      "learning_rate": 3.736656199784684e-06,
      "loss": 0.2626,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5410
    },
    {
      "epoch": 1.935769986308709,
      "grad_norm": 0.4618371306142006,
      "learning_rate": 3.72676844939885e-06,
      "loss": 0.2674,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5420
    },
    {
      "epoch": 1.939341627477826,
      "grad_norm": 0.44720354687845026,
      "learning_rate": 3.7168788812000813e-06,
      "loss": 0.2699,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5430
    },
    {
      "epoch": 1.9429132686469432,
      "grad_norm": 0.4779469192627861,
      "learning_rate": 3.706987574464772e-06,
      "loss": 0.2673,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5440
    },
    {
      "epoch": 1.9464849098160606,
      "grad_norm": 0.4639961764942175,
      "learning_rate": 3.6970946084832493e-06,
      "loss": 0.269,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5450
    },
    {
      "epoch": 1.9500565509851777,
      "grad_norm": 0.4658367692686749,
      "learning_rate": 3.687200062559149e-06,
      "loss": 0.2601,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5460
    },
    {
      "epoch": 1.9536281921542948,
      "grad_norm": 0.45772338543604596,
      "learning_rate": 3.677304016008766e-06,
      "loss": 0.2596,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5470
    },
    {
      "epoch": 1.9571998333234122,
      "grad_norm": 0.4632287878695277,
      "learning_rate": 3.6674065481604254e-06,
      "loss": 0.2681,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5480
    },
    {
      "epoch": 1.9607714744925293,
      "grad_norm": 0.47016891116157244,
      "learning_rate": 3.657507738353846e-06,
      "loss": 0.2644,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5490
    },
    {
      "epoch": 1.9643431156616464,
      "grad_norm": 0.439306529444033,
      "learning_rate": 3.6476076659395063e-06,
      "loss": 0.263,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5500
    },
    {
      "epoch": 1.9679147568307638,
      "grad_norm": 0.4217601827104545,
      "learning_rate": 3.6377064102780027e-06,
      "loss": 0.2623,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5510
    },
    {
      "epoch": 1.971486397999881,
      "grad_norm": 0.41695850691495356,
      "learning_rate": 3.627804050739418e-06,
      "loss": 0.2646,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5520
    },
    {
      "epoch": 1.975058039168998,
      "grad_norm": 0.4599681552034561,
      "learning_rate": 3.6179006667026843e-06,
      "loss": 0.2692,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5530
    },
    {
      "epoch": 1.9786296803381154,
      "grad_norm": 0.4485808768326826,
      "learning_rate": 3.607996337554946e-06,
      "loss": 0.2615,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5540
    },
    {
      "epoch": 1.9822013215072327,
      "grad_norm": 0.40761040542444305,
      "learning_rate": 3.5980911426909237e-06,
      "loss": 0.2638,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5550
    },
    {
      "epoch": 1.9857729626763496,
      "grad_norm": 0.4222409436766008,
      "learning_rate": 3.588185161512277e-06,
      "loss": 0.2682,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5560
    },
    {
      "epoch": 1.989344603845467,
      "grad_norm": 0.4469348281535693,
      "learning_rate": 3.5782784734269684e-06,
      "loss": 0.2652,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5570
    },
    {
      "epoch": 1.9929162450145843,
      "grad_norm": 0.4436372969867896,
      "learning_rate": 3.56837115784863e-06,
      "loss": 0.2654,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5580
    },
    {
      "epoch": 1.9964878861837014,
      "grad_norm": 0.4450124638999612,
      "learning_rate": 3.558463294195921e-06,
      "loss": 0.2654,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5590
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.4587295790990368,
      "learning_rate": 3.548554961891895e-06,
      "loss": 0.2733,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5600
    },
    {
      "epoch": 2.0035716411691173,
      "grad_norm": 0.4581938127499719,
      "learning_rate": 3.5386462403633622e-06,
      "loss": 0.2723,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5610
    },
    {
      "epoch": 2.0071432823382342,
      "grad_norm": 0.4332836163918156,
      "learning_rate": 3.5287372090402544e-06,
      "loss": 0.2648,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5620
    },
    {
      "epoch": 2.0107149235073516,
      "grad_norm": 0.46166257438776565,
      "learning_rate": 3.518827947354983e-06,
      "loss": 0.2609,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5630
    },
    {
      "epoch": 2.014286564676469,
      "grad_norm": 0.4258612846530011,
      "learning_rate": 3.5089185347418104e-06,
      "loss": 0.2524,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5640
    },
    {
      "epoch": 2.017858205845586,
      "grad_norm": 0.451199865672739,
      "learning_rate": 3.499009050636206e-06,
      "loss": 0.2615,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5650
    },
    {
      "epoch": 2.021429847014703,
      "grad_norm": 0.4539388452090233,
      "learning_rate": 3.4890995744742136e-06,
      "loss": 0.257,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5660
    },
    {
      "epoch": 2.0250014881838205,
      "grad_norm": 0.46272019900125183,
      "learning_rate": 3.4791901856918133e-06,
      "loss": 0.26,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5670
    },
    {
      "epoch": 2.028573129352938,
      "grad_norm": 0.4462580527431944,
      "learning_rate": 3.469280963724284e-06,
      "loss": 0.2624,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5680
    },
    {
      "epoch": 2.0321447705220548,
      "grad_norm": 0.4313640422424715,
      "learning_rate": 3.4593719880055683e-06,
      "loss": 0.252,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5690
    },
    {
      "epoch": 2.035716411691172,
      "grad_norm": 0.4320905225982968,
      "learning_rate": 3.449463337967633e-06,
      "loss": 0.2624,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5700
    },
    {
      "epoch": 2.0392880528602895,
      "grad_norm": 0.45081082335764194,
      "learning_rate": 3.439555093039836e-06,
      "loss": 0.2596,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5710
    },
    {
      "epoch": 2.0428596940294064,
      "grad_norm": 0.44641947710146607,
      "learning_rate": 3.4296473326482884e-06,
      "loss": 0.2654,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5720
    },
    {
      "epoch": 2.0464313351985237,
      "grad_norm": 0.4744778002332598,
      "learning_rate": 3.4197401362152145e-06,
      "loss": 0.2623,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5730
    },
    {
      "epoch": 2.050002976367641,
      "grad_norm": 0.40222803242906663,
      "learning_rate": 3.4098335831583214e-06,
      "loss": 0.2615,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5740
    },
    {
      "epoch": 2.053574617536758,
      "grad_norm": 0.4291805662714492,
      "learning_rate": 3.3999277528901547e-06,
      "loss": 0.2612,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5750
    },
    {
      "epoch": 2.0571462587058753,
      "grad_norm": 0.4198663575448227,
      "learning_rate": 3.390022724817469e-06,
      "loss": 0.257,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5760
    },
    {
      "epoch": 2.0607178998749927,
      "grad_norm": 0.43469615242755727,
      "learning_rate": 3.380118578340587e-06,
      "loss": 0.2506,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5770
    },
    {
      "epoch": 2.0642895410441096,
      "grad_norm": 0.4257751104947171,
      "learning_rate": 3.370215392852766e-06,
      "loss": 0.2551,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5780
    },
    {
      "epoch": 2.067861182213227,
      "grad_norm": 0.44934709851583765,
      "learning_rate": 3.3603132477395556e-06,
      "loss": 0.2647,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5790
    },
    {
      "epoch": 2.0714328233823442,
      "grad_norm": 0.4151266377740739,
      "learning_rate": 3.350412222378173e-06,
      "loss": 0.2553,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5800
    },
    {
      "epoch": 2.0750044645514616,
      "grad_norm": 0.4131070130410974,
      "learning_rate": 3.3405123961368523e-06,
      "loss": 0.2531,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5810
    },
    {
      "epoch": 2.0785761057205785,
      "grad_norm": 0.4469057886718626,
      "learning_rate": 3.330613848374218e-06,
      "loss": 0.2529,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5820
    },
    {
      "epoch": 2.082147746889696,
      "grad_norm": 0.4470454898990664,
      "learning_rate": 3.3207166584386465e-06,
      "loss": 0.2596,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5830
    },
    {
      "epoch": 2.085719388058813,
      "grad_norm": 0.43300846252821945,
      "learning_rate": 3.31082090566763e-06,
      "loss": 0.2465,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5840
    },
    {
      "epoch": 2.08929102922793,
      "grad_norm": 0.4158002323170265,
      "learning_rate": 3.3009266693871378e-06,
      "loss": 0.2571,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5850
    },
    {
      "epoch": 2.0928626703970474,
      "grad_norm": 0.4410294971525737,
      "learning_rate": 3.2910340289109845e-06,
      "loss": 0.2537,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5860
    },
    {
      "epoch": 2.0964343115661648,
      "grad_norm": 0.425419914687579,
      "learning_rate": 3.281143063540192e-06,
      "loss": 0.2485,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5870
    },
    {
      "epoch": 2.1000059527352817,
      "grad_norm": 0.3930814964109948,
      "learning_rate": 3.2712538525623556e-06,
      "loss": 0.2509,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5880
    },
    {
      "epoch": 2.103577593904399,
      "grad_norm": 0.39312197575007407,
      "learning_rate": 3.2613664752510058e-06,
      "loss": 0.2579,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5890
    },
    {
      "epoch": 2.1071492350735164,
      "grad_norm": 0.39166055787352044,
      "learning_rate": 3.251481010864975e-06,
      "loss": 0.253,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5900
    },
    {
      "epoch": 2.1107208762426337,
      "grad_norm": 0.4168399489824956,
      "learning_rate": 3.2415975386477585e-06,
      "loss": 0.2426,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5910
    },
    {
      "epoch": 2.1142925174117506,
      "grad_norm": 0.4003519724030952,
      "learning_rate": 3.2317161378268857e-06,
      "loss": 0.2555,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5920
    },
    {
      "epoch": 2.117864158580868,
      "grad_norm": 0.4031809919204729,
      "learning_rate": 3.22183688761328e-06,
      "loss": 0.2488,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5930
    },
    {
      "epoch": 2.1214357997499853,
      "grad_norm": 0.4234261622652059,
      "learning_rate": 3.211959867200624e-06,
      "loss": 0.2445,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5940
    },
    {
      "epoch": 2.125007440919102,
      "grad_norm": 0.4612705252248991,
      "learning_rate": 3.2020851557647266e-06,
      "loss": 0.2503,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5950
    },
    {
      "epoch": 2.1285790820882196,
      "grad_norm": 0.42876076853178796,
      "learning_rate": 3.192212832462889e-06,
      "loss": 0.2507,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5960
    },
    {
      "epoch": 2.132150723257337,
      "grad_norm": 0.45700635843544063,
      "learning_rate": 3.1823429764332676e-06,
      "loss": 0.2476,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5970
    },
    {
      "epoch": 2.135722364426454,
      "grad_norm": 0.43185044728972843,
      "learning_rate": 3.1724756667942405e-06,
      "loss": 0.2504,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5980
    },
    {
      "epoch": 2.139294005595571,
      "grad_norm": 0.38819831543066835,
      "learning_rate": 3.1626109826437737e-06,
      "loss": 0.2464,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 5990
    },
    {
      "epoch": 2.1428656467646885,
      "grad_norm": 0.4246096563981283,
      "learning_rate": 3.1527490030587865e-06,
      "loss": 0.2498,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6000
    },
    {
      "epoch": 2.1464372879338054,
      "grad_norm": 0.4478475877616365,
      "learning_rate": 3.142889807094518e-06,
      "loss": 0.241,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6010
    },
    {
      "epoch": 2.1500089291029227,
      "grad_norm": 0.42555214861086776,
      "learning_rate": 3.133033473783894e-06,
      "loss": 0.2457,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6020
    },
    {
      "epoch": 2.15358057027204,
      "grad_norm": 0.41306054422231764,
      "learning_rate": 3.123180082136892e-06,
      "loss": 0.2477,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6030
    },
    {
      "epoch": 2.157152211441157,
      "grad_norm": 0.42793129046487083,
      "learning_rate": 3.11332971113991e-06,
      "loss": 0.2531,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6040
    },
    {
      "epoch": 2.1607238526102743,
      "grad_norm": 0.4284276032645389,
      "learning_rate": 3.1034824397551324e-06,
      "loss": 0.2564,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6050
    },
    {
      "epoch": 2.1642954937793917,
      "grad_norm": 0.4546595301716484,
      "learning_rate": 3.0936383469198925e-06,
      "loss": 0.2551,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6060
    },
    {
      "epoch": 2.167867134948509,
      "grad_norm": 0.3960043742406359,
      "learning_rate": 3.083797511546048e-06,
      "loss": 0.2508,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6070
    },
    {
      "epoch": 2.171438776117626,
      "grad_norm": 0.44211103699658133,
      "learning_rate": 3.073960012519343e-06,
      "loss": 0.2505,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6080
    },
    {
      "epoch": 2.1750104172867433,
      "grad_norm": 0.4384592528383511,
      "learning_rate": 3.0641259286987764e-06,
      "loss": 0.2571,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6090
    },
    {
      "epoch": 2.1785820584558606,
      "grad_norm": 0.4153547793008788,
      "learning_rate": 3.05429533891597e-06,
      "loss": 0.2502,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6100
    },
    {
      "epoch": 2.1821536996249775,
      "grad_norm": 0.43122353801146596,
      "learning_rate": 3.0444683219745377e-06,
      "loss": 0.2472,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6110
    },
    {
      "epoch": 2.185725340794095,
      "grad_norm": 0.46468804115251483,
      "learning_rate": 3.0346449566494538e-06,
      "loss": 0.2518,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6120
    },
    {
      "epoch": 2.189296981963212,
      "grad_norm": 0.426327090760231,
      "learning_rate": 3.024825321686419e-06,
      "loss": 0.2537,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6130
    },
    {
      "epoch": 2.192868623132329,
      "grad_norm": 0.43925587873676897,
      "learning_rate": 3.0150094958012315e-06,
      "loss": 0.2466,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6140
    },
    {
      "epoch": 2.1964402643014465,
      "grad_norm": 0.44205546017901753,
      "learning_rate": 3.0051975576791544e-06,
      "loss": 0.2439,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6150
    },
    {
      "epoch": 2.200011905470564,
      "grad_norm": 0.4229463218493798,
      "learning_rate": 2.9953895859742865e-06,
      "loss": 0.2555,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6160
    },
    {
      "epoch": 2.203583546639681,
      "grad_norm": 0.43407788987427626,
      "learning_rate": 2.985585659308932e-06,
      "loss": 0.2504,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6170
    },
    {
      "epoch": 2.207155187808798,
      "grad_norm": 0.4354528910386347,
      "learning_rate": 2.975785856272969e-06,
      "loss": 0.2467,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6180
    },
    {
      "epoch": 2.2107268289779154,
      "grad_norm": 0.43331128792320145,
      "learning_rate": 2.965990255423217e-06,
      "loss": 0.2439,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6190
    },
    {
      "epoch": 2.2142984701470327,
      "grad_norm": 0.42629909041017044,
      "learning_rate": 2.956198935282816e-06,
      "loss": 0.2512,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6200
    },
    {
      "epoch": 2.2178701113161496,
      "grad_norm": 0.406376228889923,
      "learning_rate": 2.9464119743405863e-06,
      "loss": 0.2445,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6210
    },
    {
      "epoch": 2.221441752485267,
      "grad_norm": 0.45906603584954947,
      "learning_rate": 2.936629451050405e-06,
      "loss": 0.2515,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6220
    },
    {
      "epoch": 2.2250133936543843,
      "grad_norm": 0.44547898712147144,
      "learning_rate": 2.926851443830579e-06,
      "loss": 0.245,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6230
    },
    {
      "epoch": 2.2285850348235012,
      "grad_norm": 0.5384322504281486,
      "learning_rate": 2.9170780310632086e-06,
      "loss": 0.2454,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6240
    },
    {
      "epoch": 2.2321566759926186,
      "grad_norm": 0.42570030625635924,
      "learning_rate": 2.907309291093569e-06,
      "loss": 0.2405,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6250
    },
    {
      "epoch": 2.235728317161736,
      "grad_norm": 0.436114018629106,
      "learning_rate": 2.8975453022294748e-06,
      "loss": 0.252,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6260
    },
    {
      "epoch": 2.239299958330853,
      "grad_norm": 0.41311020815509447,
      "learning_rate": 2.8877861427406562e-06,
      "loss": 0.2555,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6270
    },
    {
      "epoch": 2.24287159949997,
      "grad_norm": 0.40408747470000445,
      "learning_rate": 2.8780318908581304e-06,
      "loss": 0.2487,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6280
    },
    {
      "epoch": 2.2464432406690875,
      "grad_norm": 0.4185113340584532,
      "learning_rate": 2.8682826247735742e-06,
      "loss": 0.2486,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6290
    },
    {
      "epoch": 2.2500148818382044,
      "grad_norm": 0.4218486087745524,
      "learning_rate": 2.8585384226386955e-06,
      "loss": 0.2407,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6300
    },
    {
      "epoch": 2.2535865230073218,
      "grad_norm": 0.40512251740904315,
      "learning_rate": 2.8487993625646116e-06,
      "loss": 0.248,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6310
    },
    {
      "epoch": 2.257158164176439,
      "grad_norm": 0.4409482835365694,
      "learning_rate": 2.8390655226212196e-06,
      "loss": 0.2492,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6320
    },
    {
      "epoch": 2.2607298053455565,
      "grad_norm": 0.4073725088603519,
      "learning_rate": 2.82933698083657e-06,
      "loss": 0.2404,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6330
    },
    {
      "epoch": 2.2643014465146734,
      "grad_norm": 0.4369684228847736,
      "learning_rate": 2.8196138151962426e-06,
      "loss": 0.2427,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6340
    },
    {
      "epoch": 2.2678730876837907,
      "grad_norm": 0.44579174359074303,
      "learning_rate": 2.809896103642721e-06,
      "loss": 0.2426,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6350
    },
    {
      "epoch": 2.271444728852908,
      "grad_norm": 0.45771064161651553,
      "learning_rate": 2.80018392407477e-06,
      "loss": 0.2435,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6360
    },
    {
      "epoch": 2.275016370022025,
      "grad_norm": 0.4403774918989739,
      "learning_rate": 2.7904773543468075e-06,
      "loss": 0.2438,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6370
    },
    {
      "epoch": 2.2785880111911423,
      "grad_norm": 0.42598744927312987,
      "learning_rate": 2.780776472268282e-06,
      "loss": 0.2503,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6380
    },
    {
      "epoch": 2.2821596523602596,
      "grad_norm": 0.4276886507005743,
      "learning_rate": 2.771081355603049e-06,
      "loss": 0.2503,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6390
    },
    {
      "epoch": 2.285731293529377,
      "grad_norm": 0.4920717543136929,
      "learning_rate": 2.7613920820687467e-06,
      "loss": 0.2435,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6400
    },
    {
      "epoch": 2.289302934698494,
      "grad_norm": 0.43701925036033384,
      "learning_rate": 2.7517087293361747e-06,
      "loss": 0.2475,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6410
    },
    {
      "epoch": 2.2928745758676112,
      "grad_norm": 0.4346436136227055,
      "learning_rate": 2.7420313750286722e-06,
      "loss": 0.2461,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6420
    },
    {
      "epoch": 2.2964462170367286,
      "grad_norm": 0.4229939188749013,
      "learning_rate": 2.7323600967214882e-06,
      "loss": 0.2421,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6430
    },
    {
      "epoch": 2.3000178582058455,
      "grad_norm": 0.5912201414304545,
      "learning_rate": 2.7226949719411746e-06,
      "loss": 0.2453,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6440
    },
    {
      "epoch": 2.303589499374963,
      "grad_norm": 0.4253354692500375,
      "learning_rate": 2.7130360781649473e-06,
      "loss": 0.2455,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6450
    },
    {
      "epoch": 2.30716114054408,
      "grad_norm": 0.4342807176023442,
      "learning_rate": 2.703383492820078e-06,
      "loss": 0.2443,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6460
    },
    {
      "epoch": 2.310732781713197,
      "grad_norm": 0.44986044561831723,
      "learning_rate": 2.6937372932832662e-06,
      "loss": 0.2465,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6470
    },
    {
      "epoch": 2.3143044228823144,
      "grad_norm": 0.4577282072825657,
      "learning_rate": 2.684097556880026e-06,
      "loss": 0.2432,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6480
    },
    {
      "epoch": 2.3178760640514318,
      "grad_norm": 0.41883367404578103,
      "learning_rate": 2.674464360884056e-06,
      "loss": 0.2446,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6490
    },
    {
      "epoch": 2.3214477052205487,
      "grad_norm": 0.42225480419162303,
      "learning_rate": 2.664837782516631e-06,
      "loss": 0.2585,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6500
    },
    {
      "epoch": 2.325019346389666,
      "grad_norm": 0.43200294151141205,
      "learning_rate": 2.6552178989459744e-06,
      "loss": 0.241,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6510
    },
    {
      "epoch": 2.3285909875587834,
      "grad_norm": 0.4172939355364942,
      "learning_rate": 2.6456047872866464e-06,
      "loss": 0.2498,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6520
    },
    {
      "epoch": 2.3321626287279003,
      "grad_norm": 0.4336551198473009,
      "learning_rate": 2.6359985245989188e-06,
      "loss": 0.2492,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6530
    },
    {
      "epoch": 2.3357342698970176,
      "grad_norm": 0.45451056073122864,
      "learning_rate": 2.626399187888164e-06,
      "loss": 0.2427,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6540
    },
    {
      "epoch": 2.339305911066135,
      "grad_norm": 0.42567264012171563,
      "learning_rate": 2.616806854104232e-06,
      "loss": 0.2475,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6550
    },
    {
      "epoch": 2.3428775522352523,
      "grad_norm": 0.4748131514352899,
      "learning_rate": 2.6072216001408377e-06,
      "loss": 0.2383,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6560
    },
    {
      "epoch": 2.346449193404369,
      "grad_norm": 0.4477560986216873,
      "learning_rate": 2.5976435028349436e-06,
      "loss": 0.2492,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6570
    },
    {
      "epoch": 2.3500208345734865,
      "grad_norm": 0.4394971651973961,
      "learning_rate": 2.588072638966141e-06,
      "loss": 0.2437,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6580
    },
    {
      "epoch": 2.353592475742604,
      "grad_norm": 0.4470690760524359,
      "learning_rate": 2.578509085256038e-06,
      "loss": 0.2439,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6590
    },
    {
      "epoch": 2.357164116911721,
      "grad_norm": 0.43968237151683753,
      "learning_rate": 2.568952918367645e-06,
      "loss": 0.2457,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6600
    },
    {
      "epoch": 2.360735758080838,
      "grad_norm": 0.4391386784963587,
      "learning_rate": 2.5594042149047564e-06,
      "loss": 0.2423,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6610
    },
    {
      "epoch": 2.3643073992499555,
      "grad_norm": 0.42975324407653287,
      "learning_rate": 2.549863051411339e-06,
      "loss": 0.2462,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6620
    },
    {
      "epoch": 2.3678790404190724,
      "grad_norm": 0.45538626525983805,
      "learning_rate": 2.540329504370918e-06,
      "loss": 0.2406,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6630
    },
    {
      "epoch": 2.3714506815881897,
      "grad_norm": 0.442390876010859,
      "learning_rate": 2.530803650205964e-06,
      "loss": 0.2466,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6640
    },
    {
      "epoch": 2.375022322757307,
      "grad_norm": 0.44546208871045945,
      "learning_rate": 2.5212855652772796e-06,
      "loss": 0.2454,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6650
    },
    {
      "epoch": 2.3785939639264244,
      "grad_norm": 0.4205593813710197,
      "learning_rate": 2.5117753258833893e-06,
      "loss": 0.2477,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6660
    },
    {
      "epoch": 2.3821656050955413,
      "grad_norm": 0.4276254011439656,
      "learning_rate": 2.5022730082599236e-06,
      "loss": 0.2464,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6670
    },
    {
      "epoch": 2.3857372462646587,
      "grad_norm": 0.4278204935672946,
      "learning_rate": 2.492778688579016e-06,
      "loss": 0.2543,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6680
    },
    {
      "epoch": 2.389308887433776,
      "grad_norm": 0.4343823003511441,
      "learning_rate": 2.4832924429486827e-06,
      "loss": 0.2341,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6690
    },
    {
      "epoch": 2.392880528602893,
      "grad_norm": 0.4323058611549525,
      "learning_rate": 2.473814347412218e-06,
      "loss": 0.238,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6700
    },
    {
      "epoch": 2.3964521697720103,
      "grad_norm": 0.4565089465307684,
      "learning_rate": 2.464344477947585e-06,
      "loss": 0.2444,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6710
    },
    {
      "epoch": 2.4000238109411276,
      "grad_norm": 0.4734833348995525,
      "learning_rate": 2.4548829104668033e-06,
      "loss": 0.2418,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6720
    },
    {
      "epoch": 2.4035954521102445,
      "grad_norm": 0.46249619301612205,
      "learning_rate": 2.4454297208153435e-06,
      "loss": 0.2521,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6730
    },
    {
      "epoch": 2.407167093279362,
      "grad_norm": 0.43223468376106816,
      "learning_rate": 2.4359849847715185e-06,
      "loss": 0.2416,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6740
    },
    {
      "epoch": 2.410738734448479,
      "grad_norm": 0.4508896819243592,
      "learning_rate": 2.426548778045875e-06,
      "loss": 0.2445,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6750
    },
    {
      "epoch": 2.414310375617596,
      "grad_norm": 0.4495937778511734,
      "learning_rate": 2.4171211762805884e-06,
      "loss": 0.246,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6760
    },
    {
      "epoch": 2.4178820167867134,
      "grad_norm": 0.43179341290400897,
      "learning_rate": 2.4077022550488544e-06,
      "loss": 0.2417,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6770
    },
    {
      "epoch": 2.421453657955831,
      "grad_norm": 0.4705685298645784,
      "learning_rate": 2.398292089854285e-06,
      "loss": 0.2381,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6780
    },
    {
      "epoch": 2.4250252991249477,
      "grad_norm": 0.43061736432599096,
      "learning_rate": 2.3888907561302995e-06,
      "loss": 0.2477,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6790
    },
    {
      "epoch": 2.428596940294065,
      "grad_norm": 0.4934734522126448,
      "learning_rate": 2.379498329239526e-06,
      "loss": 0.2459,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6800
    },
    {
      "epoch": 2.4321685814631824,
      "grad_norm": 0.4765838098586213,
      "learning_rate": 2.370114884473194e-06,
      "loss": 0.2482,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6810
    },
    {
      "epoch": 2.4357402226322997,
      "grad_norm": 0.4687408325113669,
      "learning_rate": 2.3607404970505284e-06,
      "loss": 0.2408,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6820
    },
    {
      "epoch": 2.4393118638014166,
      "grad_norm": 0.4281004117147228,
      "learning_rate": 2.351375242118148e-06,
      "loss": 0.2505,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6830
    },
    {
      "epoch": 2.442883504970534,
      "grad_norm": 0.4370567386193378,
      "learning_rate": 2.342019194749471e-06,
      "loss": 0.249,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6840
    },
    {
      "epoch": 2.4464551461396513,
      "grad_norm": 0.4777666587818985,
      "learning_rate": 2.3326724299440986e-06,
      "loss": 0.2441,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6850
    },
    {
      "epoch": 2.450026787308768,
      "grad_norm": 0.450289306534585,
      "learning_rate": 2.323335022627225e-06,
      "loss": 0.2466,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6860
    },
    {
      "epoch": 2.4535984284778856,
      "grad_norm": 0.4472240813142968,
      "learning_rate": 2.314007047649033e-06,
      "loss": 0.2336,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6870
    },
    {
      "epoch": 2.457170069647003,
      "grad_norm": 0.4377670065814464,
      "learning_rate": 2.304688579784093e-06,
      "loss": 0.2519,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6880
    },
    {
      "epoch": 2.4607417108161203,
      "grad_norm": 0.4491227795973426,
      "learning_rate": 2.295379693730766e-06,
      "loss": 0.2436,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6890
    },
    {
      "epoch": 2.464313351985237,
      "grad_norm": 0.476378700259544,
      "learning_rate": 2.2860804641106033e-06,
      "loss": 0.249,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6900
    },
    {
      "epoch": 2.4678849931543545,
      "grad_norm": 0.4508043014114273,
      "learning_rate": 2.276790965467748e-06,
      "loss": 0.2473,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6910
    },
    {
      "epoch": 2.471456634323472,
      "grad_norm": 0.42106773935779124,
      "learning_rate": 2.2675112722683406e-06,
      "loss": 0.245,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6920
    },
    {
      "epoch": 2.4750282754925887,
      "grad_norm": 0.49568541004831845,
      "learning_rate": 2.2582414588999175e-06,
      "loss": 0.2432,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6930
    },
    {
      "epoch": 2.478599916661706,
      "grad_norm": 0.44834524118595986,
      "learning_rate": 2.248981599670815e-06,
      "loss": 0.2456,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6940
    },
    {
      "epoch": 2.4821715578308234,
      "grad_norm": 0.4382727602006904,
      "learning_rate": 2.2397317688095792e-06,
      "loss": 0.2404,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6950
    },
    {
      "epoch": 2.4857431989999403,
      "grad_norm": 0.4357541382580279,
      "learning_rate": 2.230492040464365e-06,
      "loss": 0.2506,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6960
    },
    {
      "epoch": 2.4893148401690577,
      "grad_norm": 0.4494934844393061,
      "learning_rate": 2.221262488702344e-06,
      "loss": 0.2328,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6970
    },
    {
      "epoch": 2.492886481338175,
      "grad_norm": 0.44511789317832856,
      "learning_rate": 2.2120431875091107e-06,
      "loss": 0.2462,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6980
    },
    {
      "epoch": 2.496458122507292,
      "grad_norm": 0.4637114798209286,
      "learning_rate": 2.202834210788089e-06,
      "loss": 0.2359,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 6990
    },
    {
      "epoch": 2.5000297636764093,
      "grad_norm": 0.454258990361472,
      "learning_rate": 2.1936356323599417e-06,
      "loss": 0.2464,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7000
    },
    {
      "epoch": 2.5036014048455266,
      "grad_norm": 0.42621406491701536,
      "learning_rate": 2.1844475259619763e-06,
      "loss": 0.238,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7010
    },
    {
      "epoch": 2.5071730460146435,
      "grad_norm": 0.4742468073415194,
      "learning_rate": 2.1752699652475546e-06,
      "loss": 0.2417,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7020
    },
    {
      "epoch": 2.510744687183761,
      "grad_norm": 0.47342536962727527,
      "learning_rate": 2.1661030237855023e-06,
      "loss": 0.2351,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7030
    },
    {
      "epoch": 2.514316328352878,
      "grad_norm": 0.47877566175565467,
      "learning_rate": 2.1569467750595193e-06,
      "loss": 0.2452,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7040
    },
    {
      "epoch": 2.517887969521995,
      "grad_norm": 0.46978072487842876,
      "learning_rate": 2.147801292467592e-06,
      "loss": 0.2443,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7050
    },
    {
      "epoch": 2.5214596106911125,
      "grad_norm": 0.4735918423996795,
      "learning_rate": 2.1386666493214007e-06,
      "loss": 0.2378,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7060
    },
    {
      "epoch": 2.52503125186023,
      "grad_norm": 0.4567127424051271,
      "learning_rate": 2.12954291884574e-06,
      "loss": 0.2384,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7070
    },
    {
      "epoch": 2.5286028930293467,
      "grad_norm": 0.4147541911140241,
      "learning_rate": 2.1204301741779235e-06,
      "loss": 0.2453,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7080
    },
    {
      "epoch": 2.532174534198464,
      "grad_norm": 0.43912608187760216,
      "learning_rate": 2.1113284883671994e-06,
      "loss": 0.2361,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7090
    },
    {
      "epoch": 2.5357461753675814,
      "grad_norm": 0.44929631139927123,
      "learning_rate": 2.102237934374169e-06,
      "loss": 0.2401,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7100
    },
    {
      "epoch": 2.5393178165366987,
      "grad_norm": 0.4882310550461796,
      "learning_rate": 2.0931585850701984e-06,
      "loss": 0.2377,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7110
    },
    {
      "epoch": 2.542889457705816,
      "grad_norm": 0.4581959273553773,
      "learning_rate": 2.084090513236835e-06,
      "loss": 0.2337,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7120
    },
    {
      "epoch": 2.546461098874933,
      "grad_norm": 0.45309648854952933,
      "learning_rate": 2.075033791565223e-06,
      "loss": 0.2445,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7130
    },
    {
      "epoch": 2.5500327400440503,
      "grad_norm": 0.44638496641251035,
      "learning_rate": 2.0659884926555226e-06,
      "loss": 0.2393,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7140
    },
    {
      "epoch": 2.5536043812131677,
      "grad_norm": 0.48769533117288066,
      "learning_rate": 2.0569546890163285e-06,
      "loss": 0.2442,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7150
    },
    {
      "epoch": 2.5571760223822846,
      "grad_norm": 0.43511656727356074,
      "learning_rate": 2.0479324530640867e-06,
      "loss": 0.2384,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7160
    },
    {
      "epoch": 2.560747663551402,
      "grad_norm": 0.4820233106672117,
      "learning_rate": 2.038921857122515e-06,
      "loss": 0.2419,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7170
    },
    {
      "epoch": 2.5643193047205193,
      "grad_norm": 0.48561125119478454,
      "learning_rate": 2.02992297342202e-06,
      "loss": 0.2312,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7180
    },
    {
      "epoch": 2.567890945889636,
      "grad_norm": 0.4432776127145546,
      "learning_rate": 2.020935874099126e-06,
      "loss": 0.2402,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7190
    },
    {
      "epoch": 2.5714625870587535,
      "grad_norm": 0.454164264161987,
      "learning_rate": 2.011960631195889e-06,
      "loss": 0.2466,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7200
    },
    {
      "epoch": 2.575034228227871,
      "grad_norm": 0.4854707049194837,
      "learning_rate": 2.0029973166593196e-06,
      "loss": 0.245,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7210
    },
    {
      "epoch": 2.5786058693969878,
      "grad_norm": 0.46000696889932535,
      "learning_rate": 1.994046002340814e-06,
      "loss": 0.243,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7220
    },
    {
      "epoch": 2.582177510566105,
      "grad_norm": 0.4357766462500482,
      "learning_rate": 1.9851067599955707e-06,
      "loss": 0.2516,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7230
    },
    {
      "epoch": 2.5857491517352225,
      "grad_norm": 0.44408294376392227,
      "learning_rate": 1.9761796612820146e-06,
      "loss": 0.2432,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7240
    },
    {
      "epoch": 2.5893207929043394,
      "grad_norm": 0.43382954307970867,
      "learning_rate": 1.967264777761229e-06,
      "loss": 0.241,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7250
    },
    {
      "epoch": 2.5928924340734567,
      "grad_norm": 0.46704708085378005,
      "learning_rate": 1.958362180896376e-06,
      "loss": 0.2352,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7260
    },
    {
      "epoch": 2.596464075242574,
      "grad_norm": 0.4388547369132919,
      "learning_rate": 1.9494719420521254e-06,
      "loss": 0.2406,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7270
    },
    {
      "epoch": 2.600035716411691,
      "grad_norm": 0.4280307854060593,
      "learning_rate": 1.9405941324940857e-06,
      "loss": 0.2414,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7280
    },
    {
      "epoch": 2.6036073575808083,
      "grad_norm": 0.45466648840686463,
      "learning_rate": 1.931728823388228e-06,
      "loss": 0.244,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7290
    },
    {
      "epoch": 2.6071789987499256,
      "grad_norm": 0.44941873865770565,
      "learning_rate": 1.9228760858003155e-06,
      "loss": 0.2356,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7300
    },
    {
      "epoch": 2.6107506399190425,
      "grad_norm": 0.4776784235080879,
      "learning_rate": 1.9140359906953443e-06,
      "loss": 0.2348,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7310
    },
    {
      "epoch": 2.61432228108816,
      "grad_norm": 0.4859122793447613,
      "learning_rate": 1.9052086089369585e-06,
      "loss": 0.2335,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7320
    },
    {
      "epoch": 2.6178939222572772,
      "grad_norm": 0.4645265007751813,
      "learning_rate": 1.8963940112868904e-06,
      "loss": 0.2431,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7330
    },
    {
      "epoch": 2.6214655634263946,
      "grad_norm": 0.47714483697616294,
      "learning_rate": 1.8875922684043977e-06,
      "loss": 0.2438,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7340
    },
    {
      "epoch": 2.6250372045955115,
      "grad_norm": 0.4921155235969745,
      "learning_rate": 1.878803450845688e-06,
      "loss": 0.2367,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7350
    },
    {
      "epoch": 2.628608845764629,
      "grad_norm": 0.4582649026095651,
      "learning_rate": 1.870027629063357e-06,
      "loss": 0.2428,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7360
    },
    {
      "epoch": 2.632180486933746,
      "grad_norm": 0.44087571867035075,
      "learning_rate": 1.8612648734058292e-06,
      "loss": 0.2433,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7370
    },
    {
      "epoch": 2.6357521281028635,
      "grad_norm": 0.5367973728899978,
      "learning_rate": 1.852515254116782e-06,
      "loss": 0.2462,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7380
    },
    {
      "epoch": 2.6393237692719804,
      "grad_norm": 0.47523519421713606,
      "learning_rate": 1.843778841334594e-06,
      "loss": 0.2404,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7390
    },
    {
      "epoch": 2.6428954104410978,
      "grad_norm": 0.45351487278591784,
      "learning_rate": 1.8350557050917794e-06,
      "loss": 0.237,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7400
    },
    {
      "epoch": 2.646467051610215,
      "grad_norm": 0.45566439391509844,
      "learning_rate": 1.8263459153144216e-06,
      "loss": 0.2378,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7410
    },
    {
      "epoch": 2.650038692779332,
      "grad_norm": 0.5427961932037867,
      "learning_rate": 1.8176495418216173e-06,
      "loss": 0.2327,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7420
    },
    {
      "epoch": 2.6536103339484494,
      "grad_norm": 0.48467724978186316,
      "learning_rate": 1.808966654324919e-06,
      "loss": 0.2453,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7430
    },
    {
      "epoch": 2.6571819751175667,
      "grad_norm": 0.4689333195988378,
      "learning_rate": 1.800297322427771e-06,
      "loss": 0.2433,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7440
    },
    {
      "epoch": 2.6607536162866836,
      "grad_norm": 0.4870900784001144,
      "learning_rate": 1.791641615624952e-06,
      "loss": 0.2441,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7450
    },
    {
      "epoch": 2.664325257455801,
      "grad_norm": 0.45290609880585164,
      "learning_rate": 1.7829996033020236e-06,
      "loss": 0.2461,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7460
    },
    {
      "epoch": 2.6678968986249183,
      "grad_norm": 0.4809762662064227,
      "learning_rate": 1.7743713547347693e-06,
      "loss": 0.2332,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7470
    },
    {
      "epoch": 2.671468539794035,
      "grad_norm": 0.4803894889697133,
      "learning_rate": 1.7657569390886364e-06,
      "loss": 0.2403,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7480
    },
    {
      "epoch": 2.6750401809631525,
      "grad_norm": 0.43418058879650556,
      "learning_rate": 1.757156425418191e-06,
      "loss": 0.2471,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7490
    },
    {
      "epoch": 2.67861182213227,
      "grad_norm": 0.4545794361949984,
      "learning_rate": 1.7485698826665536e-06,
      "loss": 0.2395,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7500
    },
    {
      "epoch": 2.682183463301387,
      "grad_norm": 0.44328983850304654,
      "learning_rate": 1.7399973796648535e-06,
      "loss": 0.2419,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7510
    },
    {
      "epoch": 2.685755104470504,
      "grad_norm": 0.48089511751970304,
      "learning_rate": 1.731438985131677e-06,
      "loss": 0.2381,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7520
    },
    {
      "epoch": 2.6893267456396215,
      "grad_norm": 0.5182314821606652,
      "learning_rate": 1.7228947676725114e-06,
      "loss": 0.2393,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7530
    },
    {
      "epoch": 2.6928983868087384,
      "grad_norm": 0.4896768129268529,
      "learning_rate": 1.7143647957791976e-06,
      "loss": 0.2428,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7540
    },
    {
      "epoch": 2.6964700279778557,
      "grad_norm": 0.43406888099995405,
      "learning_rate": 1.7058491378293887e-06,
      "loss": 0.2464,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7550
    },
    {
      "epoch": 2.700041669146973,
      "grad_norm": 0.45987710855301045,
      "learning_rate": 1.6973478620859872e-06,
      "loss": 0.2396,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7560
    },
    {
      "epoch": 2.70361331031609,
      "grad_norm": 0.5081238736314797,
      "learning_rate": 1.688861036696607e-06,
      "loss": 0.239,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7570
    },
    {
      "epoch": 2.7071849514852073,
      "grad_norm": 0.47124511315850837,
      "learning_rate": 1.6803887296930296e-06,
      "loss": 0.2427,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7580
    },
    {
      "epoch": 2.7107565926543247,
      "grad_norm": 0.4766186555462526,
      "learning_rate": 1.6719310089906508e-06,
      "loss": 0.2367,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7590
    },
    {
      "epoch": 2.714328233823442,
      "grad_norm": 0.5154643063486422,
      "learning_rate": 1.6634879423879397e-06,
      "loss": 0.2418,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7600
    },
    {
      "epoch": 2.717899874992559,
      "grad_norm": 0.47619250886490805,
      "learning_rate": 1.655059597565901e-06,
      "loss": 0.2388,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7610
    },
    {
      "epoch": 2.7214715161616763,
      "grad_norm": 0.4727081854520604,
      "learning_rate": 1.6466460420875199e-06,
      "loss": 0.2389,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7620
    },
    {
      "epoch": 2.7250431573307936,
      "grad_norm": 0.4739130851871452,
      "learning_rate": 1.638247343397233e-06,
      "loss": 0.2379,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7630
    },
    {
      "epoch": 2.728614798499911,
      "grad_norm": 0.4921562232207917,
      "learning_rate": 1.6298635688203832e-06,
      "loss": 0.2486,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7640
    },
    {
      "epoch": 2.732186439669028,
      "grad_norm": 0.4757143273606086,
      "learning_rate": 1.6214947855626752e-06,
      "loss": 0.2397,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7650
    },
    {
      "epoch": 2.735758080838145,
      "grad_norm": 0.47786652916433087,
      "learning_rate": 1.6131410607096414e-06,
      "loss": 0.2424,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7660
    },
    {
      "epoch": 2.7393297220072625,
      "grad_norm": 0.4357313446728837,
      "learning_rate": 1.6048024612261076e-06,
      "loss": 0.2422,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7670
    },
    {
      "epoch": 2.7429013631763794,
      "grad_norm": 0.43258578669352915,
      "learning_rate": 1.5964790539556484e-06,
      "loss": 0.2467,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7680
    },
    {
      "epoch": 2.746473004345497,
      "grad_norm": 0.45798160471764066,
      "learning_rate": 1.5881709056200545e-06,
      "loss": 0.2451,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7690
    },
    {
      "epoch": 2.750044645514614,
      "grad_norm": 0.5046607545679609,
      "learning_rate": 1.579878082818802e-06,
      "loss": 0.2431,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7700
    },
    {
      "epoch": 2.753616286683731,
      "grad_norm": 0.4561696903127055,
      "learning_rate": 1.5716006520285138e-06,
      "loss": 0.2416,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7710
    },
    {
      "epoch": 2.7571879278528484,
      "grad_norm": 0.4582998703554197,
      "learning_rate": 1.5633386796024254e-06,
      "loss": 0.2481,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7720
    },
    {
      "epoch": 2.7607595690219657,
      "grad_norm": 0.4856769517926892,
      "learning_rate": 1.55509223176986e-06,
      "loss": 0.2428,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7730
    },
    {
      "epoch": 2.7643312101910826,
      "grad_norm": 0.4665717722179424,
      "learning_rate": 1.5468613746356885e-06,
      "loss": 0.246,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7740
    },
    {
      "epoch": 2.7679028513602,
      "grad_norm": 0.46949182547775437,
      "learning_rate": 1.538646174179806e-06,
      "loss": 0.2317,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7750
    },
    {
      "epoch": 2.7714744925293173,
      "grad_norm": 0.48292030377173556,
      "learning_rate": 1.5304466962566037e-06,
      "loss": 0.2396,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7760
    },
    {
      "epoch": 2.7750461336984342,
      "grad_norm": 0.4431681729827931,
      "learning_rate": 1.5222630065944332e-06,
      "loss": 0.2422,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7770
    },
    {
      "epoch": 2.7786177748675516,
      "grad_norm": 0.4660840922860918,
      "learning_rate": 1.5140951707950896e-06,
      "loss": 0.2407,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7780
    },
    {
      "epoch": 2.782189416036669,
      "grad_norm": 0.4322294580280508,
      "learning_rate": 1.5059432543332798e-06,
      "loss": 0.2406,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7790
    },
    {
      "epoch": 2.785761057205786,
      "grad_norm": 0.4534373647030015,
      "learning_rate": 1.4978073225560966e-06,
      "loss": 0.2375,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7800
    },
    {
      "epoch": 2.789332698374903,
      "grad_norm": 0.4509567338324847,
      "learning_rate": 1.489687440682497e-06,
      "loss": 0.2426,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7810
    },
    {
      "epoch": 2.7929043395440205,
      "grad_norm": 0.5264855086764537,
      "learning_rate": 1.4815836738027832e-06,
      "loss": 0.2265,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7820
    },
    {
      "epoch": 2.7964759807131374,
      "grad_norm": 0.46188466281106827,
      "learning_rate": 1.473496086878072e-06,
      "loss": 0.2455,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7830
    },
    {
      "epoch": 2.8000476218822548,
      "grad_norm": 0.4550270822364228,
      "learning_rate": 1.4654247447397833e-06,
      "loss": 0.2487,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7840
    },
    {
      "epoch": 2.803619263051372,
      "grad_norm": 0.4812751842028551,
      "learning_rate": 1.457369712089114e-06,
      "loss": 0.2432,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7850
    },
    {
      "epoch": 2.8071909042204894,
      "grad_norm": 0.4300511348882846,
      "learning_rate": 1.4493310534965194e-06,
      "loss": 0.2477,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7860
    },
    {
      "epoch": 2.810762545389607,
      "grad_norm": 0.44046616488515516,
      "learning_rate": 1.4413088334012023e-06,
      "loss": 0.2366,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7870
    },
    {
      "epoch": 2.8143341865587237,
      "grad_norm": 0.46437456660895376,
      "learning_rate": 1.4333031161105897e-06,
      "loss": 0.2356,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7880
    },
    {
      "epoch": 2.817905827727841,
      "grad_norm": 0.4753228834616267,
      "learning_rate": 1.4253139657998184e-06,
      "loss": 0.2353,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7890
    },
    {
      "epoch": 2.8214774688969584,
      "grad_norm": 0.4851133893845524,
      "learning_rate": 1.4173414465112214e-06,
      "loss": 0.2369,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7900
    },
    {
      "epoch": 2.8250491100660753,
      "grad_norm": 0.47493639702473806,
      "learning_rate": 1.4093856221538177e-06,
      "loss": 0.2401,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7910
    },
    {
      "epoch": 2.8286207512351926,
      "grad_norm": 0.4618709950690347,
      "learning_rate": 1.4014465565027934e-06,
      "loss": 0.2423,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7920
    },
    {
      "epoch": 2.83219239240431,
      "grad_norm": 0.43927959015901663,
      "learning_rate": 1.3935243131989971e-06,
      "loss": 0.2383,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7930
    },
    {
      "epoch": 2.835764033573427,
      "grad_norm": 0.4484129148789068,
      "learning_rate": 1.385618955748424e-06,
      "loss": 0.239,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7940
    },
    {
      "epoch": 2.8393356747425442,
      "grad_norm": 0.5001001943658824,
      "learning_rate": 1.3777305475217123e-06,
      "loss": 0.2391,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7950
    },
    {
      "epoch": 2.8429073159116616,
      "grad_norm": 0.4653892594419888,
      "learning_rate": 1.3698591517536295e-06,
      "loss": 0.2334,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7960
    },
    {
      "epoch": 2.8464789570807785,
      "grad_norm": 0.45427703489652743,
      "learning_rate": 1.3620048315425715e-06,
      "loss": 0.2347,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7970
    },
    {
      "epoch": 2.850050598249896,
      "grad_norm": 0.46031626412232396,
      "learning_rate": 1.3541676498500502e-06,
      "loss": 0.2309,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7980
    },
    {
      "epoch": 2.853622239419013,
      "grad_norm": 0.4781032434861627,
      "learning_rate": 1.3463476695001968e-06,
      "loss": 0.242,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 7990
    },
    {
      "epoch": 2.85719388058813,
      "grad_norm": 0.449958891165902,
      "learning_rate": 1.3385449531792503e-06,
      "loss": 0.242,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8000
    },
    {
      "epoch": 2.8607655217572474,
      "grad_norm": 0.4739933313720244,
      "learning_rate": 1.3307595634350577e-06,
      "loss": 0.2438,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8010
    },
    {
      "epoch": 2.8643371629263648,
      "grad_norm": 0.4777799159455414,
      "learning_rate": 1.3229915626765764e-06,
      "loss": 0.2338,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8020
    },
    {
      "epoch": 2.8679088040954817,
      "grad_norm": 0.4870133963570842,
      "learning_rate": 1.3152410131733706e-06,
      "loss": 0.2411,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8030
    },
    {
      "epoch": 2.871480445264599,
      "grad_norm": 0.46036598057169154,
      "learning_rate": 1.3075079770551103e-06,
      "loss": 0.2328,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8040
    },
    {
      "epoch": 2.8750520864337163,
      "grad_norm": 0.45474146002760457,
      "learning_rate": 1.299792516311075e-06,
      "loss": 0.2356,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8050
    },
    {
      "epoch": 2.8786237276028332,
      "grad_norm": 0.4650118946273859,
      "learning_rate": 1.2920946927896608e-06,
      "loss": 0.2367,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8060
    },
    {
      "epoch": 2.8821953687719506,
      "grad_norm": 0.4918407064463534,
      "learning_rate": 1.2844145681978763e-06,
      "loss": 0.2435,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8070
    },
    {
      "epoch": 2.885767009941068,
      "grad_norm": 0.45031935108227794,
      "learning_rate": 1.276752204100857e-06,
      "loss": 0.2342,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8080
    },
    {
      "epoch": 2.8893386511101853,
      "grad_norm": 0.4511846978714396,
      "learning_rate": 1.2691076619213641e-06,
      "loss": 0.2451,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8090
    },
    {
      "epoch": 2.892910292279302,
      "grad_norm": 0.49634809036650396,
      "learning_rate": 1.261481002939296e-06,
      "loss": 0.2397,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8100
    },
    {
      "epoch": 2.8964819334484195,
      "grad_norm": 0.4735229254072526,
      "learning_rate": 1.2538722882911975e-06,
      "loss": 0.2376,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8110
    },
    {
      "epoch": 2.900053574617537,
      "grad_norm": 0.4586289523437465,
      "learning_rate": 1.2462815789697694e-06,
      "loss": 0.2444,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8120
    },
    {
      "epoch": 2.9036252157866542,
      "grad_norm": 0.4689993067162779,
      "learning_rate": 1.2387089358233753e-06,
      "loss": 0.2312,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8130
    },
    {
      "epoch": 2.907196856955771,
      "grad_norm": 0.45127761849831977,
      "learning_rate": 1.2311544195555625e-06,
      "loss": 0.2345,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8140
    },
    {
      "epoch": 2.9107684981248885,
      "grad_norm": 0.4551924773089047,
      "learning_rate": 1.2236180907245657e-06,
      "loss": 0.2333,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8150
    },
    {
      "epoch": 2.914340139294006,
      "grad_norm": 0.4747350168287828,
      "learning_rate": 1.2161000097428276e-06,
      "loss": 0.2299,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8160
    },
    {
      "epoch": 2.9179117804631227,
      "grad_norm": 0.4789946852521992,
      "learning_rate": 1.2086002368765154e-06,
      "loss": 0.2309,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8170
    },
    {
      "epoch": 2.92148342163224,
      "grad_norm": 0.44972106394869,
      "learning_rate": 1.2011188322450313e-06,
      "loss": 0.2351,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8180
    },
    {
      "epoch": 2.9250550628013574,
      "grad_norm": 0.4752319084408135,
      "learning_rate": 1.1936558558205381e-06,
      "loss": 0.2316,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8190
    },
    {
      "epoch": 2.9286267039704743,
      "grad_norm": 0.4538625576215163,
      "learning_rate": 1.186211367427475e-06,
      "loss": 0.2395,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8200
    },
    {
      "epoch": 2.9321983451395917,
      "grad_norm": 0.43101707019643254,
      "learning_rate": 1.1787854267420764e-06,
      "loss": 0.2347,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8210
    },
    {
      "epoch": 2.935769986308709,
      "grad_norm": 0.4565611978219319,
      "learning_rate": 1.1713780932918943e-06,
      "loss": 0.2397,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8220
    },
    {
      "epoch": 2.939341627477826,
      "grad_norm": 0.4675562969199988,
      "learning_rate": 1.1639894264553263e-06,
      "loss": 0.2427,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8230
    },
    {
      "epoch": 2.9429132686469432,
      "grad_norm": 0.47564898045704657,
      "learning_rate": 1.1566194854611319e-06,
      "loss": 0.2395,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8240
    },
    {
      "epoch": 2.9464849098160606,
      "grad_norm": 0.4539496439942422,
      "learning_rate": 1.1492683293879607e-06,
      "loss": 0.2415,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8250
    },
    {
      "epoch": 2.9500565509851775,
      "grad_norm": 0.4425681909675462,
      "learning_rate": 1.1419360171638813e-06,
      "loss": 0.2328,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8260
    },
    {
      "epoch": 2.953628192154295,
      "grad_norm": 0.5071600153926035,
      "learning_rate": 1.1346226075659083e-06,
      "loss": 0.2321,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8270
    },
    {
      "epoch": 2.957199833323412,
      "grad_norm": 0.4678630904942017,
      "learning_rate": 1.1273281592195246e-06,
      "loss": 0.2407,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8280
    },
    {
      "epoch": 2.960771474492529,
      "grad_norm": 0.4622258694530203,
      "learning_rate": 1.1200527305982229e-06,
      "loss": 0.2372,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8290
    },
    {
      "epoch": 2.9643431156616464,
      "grad_norm": 0.4561377388624881,
      "learning_rate": 1.1127963800230261e-06,
      "loss": 0.2358,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8300
    },
    {
      "epoch": 2.967914756830764,
      "grad_norm": 0.45829660432747915,
      "learning_rate": 1.1055591656620247e-06,
      "loss": 0.2352,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8310
    },
    {
      "epoch": 2.9714863979998807,
      "grad_norm": 0.4557869621482418,
      "learning_rate": 1.0983411455299138e-06,
      "loss": 0.2376,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8320
    },
    {
      "epoch": 2.975058039168998,
      "grad_norm": 0.4681011674673779,
      "learning_rate": 1.09114237748752e-06,
      "loss": 0.2425,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8330
    },
    {
      "epoch": 2.9786296803381154,
      "grad_norm": 0.4732661149841976,
      "learning_rate": 1.0839629192413425e-06,
      "loss": 0.2349,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8340
    },
    {
      "epoch": 2.9822013215072327,
      "grad_norm": 0.4476543017032875,
      "learning_rate": 1.0768028283430959e-06,
      "loss": 0.237,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8350
    },
    {
      "epoch": 2.9857729626763496,
      "grad_norm": 0.4680863834511507,
      "learning_rate": 1.0696621621892357e-06,
      "loss": 0.2417,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8360
    },
    {
      "epoch": 2.989344603845467,
      "grad_norm": 0.4578896945136068,
      "learning_rate": 1.062540978020508e-06,
      "loss": 0.2385,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8370
    },
    {
      "epoch": 2.9929162450145843,
      "grad_norm": 0.4549383047870384,
      "learning_rate": 1.0554393329214897e-06,
      "loss": 0.2388,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8380
    },
    {
      "epoch": 2.9964878861837017,
      "grad_norm": 0.4693977536348272,
      "learning_rate": 1.0483572838201275e-06,
      "loss": 0.2387,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8390
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4767978710071707,
      "learning_rate": 1.0412948874872829e-06,
      "loss": 0.2462,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.43,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8400
    },
    {
      "epoch": 3.0035716411691173,
      "grad_norm": 0.5082675672275083,
      "learning_rate": 1.0342522005362807e-06,
      "loss": 0.2459,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8410
    },
    {
      "epoch": 3.0071432823382342,
      "grad_norm": 0.4434911719606514,
      "learning_rate": 1.0272292794224478e-06,
      "loss": 0.2385,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8420
    },
    {
      "epoch": 3.0107149235073516,
      "grad_norm": 0.47493763800133504,
      "learning_rate": 1.0202261804426686e-06,
      "loss": 0.2345,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8430
    },
    {
      "epoch": 3.014286564676469,
      "grad_norm": 0.45270030835499836,
      "learning_rate": 1.013242959734929e-06,
      "loss": 0.2258,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8440
    },
    {
      "epoch": 3.017858205845586,
      "grad_norm": 0.4945126721474151,
      "learning_rate": 1.0062796732778665e-06,
      "loss": 0.2357,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8450
    },
    {
      "epoch": 3.021429847014703,
      "grad_norm": 0.4584599172625007,
      "learning_rate": 9.99336376890321e-07,
      "loss": 0.231,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8460
    },
    {
      "epoch": 3.0250014881838205,
      "grad_norm": 0.4741731531048592,
      "learning_rate": 9.924131262308933e-07,
      "loss": 0.234,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8470
    },
    {
      "epoch": 3.028573129352938,
      "grad_norm": 0.42977088432539023,
      "learning_rate": 9.85509976797491e-07,
      "loss": 0.2367,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8480
    },
    {
      "epoch": 3.0321447705220548,
      "grad_norm": 0.4549490544845592,
      "learning_rate": 9.78626983926886e-07,
      "loss": 0.2261,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8490
    },
    {
      "epoch": 3.035716411691172,
      "grad_norm": 0.43517186698776544,
      "learning_rate": 9.717642027942746e-07,
      "loss": 0.2368,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8500
    },
    {
      "epoch": 3.0392880528602895,
      "grad_norm": 0.48096149960619133,
      "learning_rate": 9.64921688412833e-07,
      "loss": 0.2339,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8510
    },
    {
      "epoch": 3.0428596940294064,
      "grad_norm": 0.4695639710779253,
      "learning_rate": 9.58099495633272e-07,
      "loss": 0.2397,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8520
    },
    {
      "epoch": 3.0464313351985237,
      "grad_norm": 0.46406480819585466,
      "learning_rate": 9.512976791434055e-07,
      "loss": 0.2367,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8530
    },
    {
      "epoch": 3.050002976367641,
      "grad_norm": 0.43549316441900676,
      "learning_rate": 9.445162934677041e-07,
      "loss": 0.2362,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8540
    },
    {
      "epoch": 3.053574617536758,
      "grad_norm": 0.4776504686587063,
      "learning_rate": 9.377553929668628e-07,
      "loss": 0.2357,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8550
    },
    {
      "epoch": 3.0571462587058753,
      "grad_norm": 0.4957631348783547,
      "learning_rate": 9.310150318373668e-07,
      "loss": 0.2317,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8560
    },
    {
      "epoch": 3.0607178998749927,
      "grad_norm": 0.5121306681504587,
      "learning_rate": 9.242952641110502e-07,
      "loss": 0.225,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8570
    },
    {
      "epoch": 3.0642895410441096,
      "grad_norm": 0.4355193727970268,
      "learning_rate": 9.17596143654667e-07,
      "loss": 0.2301,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8580
    },
    {
      "epoch": 3.067861182213227,
      "grad_norm": 0.4570296964534929,
      "learning_rate": 9.109177241694655e-07,
      "loss": 0.2399,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8590
    },
    {
      "epoch": 3.0714328233823442,
      "grad_norm": 0.4830423290838575,
      "learning_rate": 9.042600591907455e-07,
      "loss": 0.2304,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8600
    },
    {
      "epoch": 3.0750044645514616,
      "grad_norm": 0.4429461104459678,
      "learning_rate": 8.976232020874352e-07,
      "loss": 0.2284,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8610
    },
    {
      "epoch": 3.0785761057205785,
      "grad_norm": 0.4705885398084829,
      "learning_rate": 8.910072060616686e-07,
      "loss": 0.2282,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8620
    },
    {
      "epoch": 3.082147746889696,
      "grad_norm": 0.48482873146757927,
      "learning_rate": 8.844121241483485e-07,
      "loss": 0.2351,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8630
    },
    {
      "epoch": 3.085719388058813,
      "grad_norm": 0.45871074100809267,
      "learning_rate": 8.778380092147283e-07,
      "loss": 0.222,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8640
    },
    {
      "epoch": 3.08929102922793,
      "grad_norm": 0.42580067592131254,
      "learning_rate": 8.712849139599894e-07,
      "loss": 0.2328,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8650
    },
    {
      "epoch": 3.0928626703970474,
      "grad_norm": 0.4683497978752016,
      "learning_rate": 8.647528909148116e-07,
      "loss": 0.2295,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8660
    },
    {
      "epoch": 3.0964343115661648,
      "grad_norm": 0.46173709254753664,
      "learning_rate": 8.582419924409602e-07,
      "loss": 0.2244,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8670
    },
    {
      "epoch": 3.1000059527352817,
      "grad_norm": 0.41918963310164564,
      "learning_rate": 8.51752270730862e-07,
      "loss": 0.2267,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8680
    },
    {
      "epoch": 3.103577593904399,
      "grad_norm": 0.4216737296789738,
      "learning_rate": 8.452837778071847e-07,
      "loss": 0.2336,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8690
    },
    {
      "epoch": 3.1071492350735164,
      "grad_norm": 0.4641138956232256,
      "learning_rate": 8.388365655224239e-07,
      "loss": 0.2293,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8700
    },
    {
      "epoch": 3.1107208762426337,
      "grad_norm": 0.466554529547195,
      "learning_rate": 8.324106855584883e-07,
      "loss": 0.2186,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8710
    },
    {
      "epoch": 3.1142925174117506,
      "grad_norm": 0.45252478767112264,
      "learning_rate": 8.260061894262803e-07,
      "loss": 0.232,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8720
    },
    {
      "epoch": 3.117864158580868,
      "grad_norm": 0.453030371983167,
      "learning_rate": 8.196231284652864e-07,
      "loss": 0.2249,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8730
    },
    {
      "epoch": 3.1214357997499853,
      "grad_norm": 0.434980320791229,
      "learning_rate": 8.132615538431668e-07,
      "loss": 0.2208,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8740
    },
    {
      "epoch": 3.125007440919102,
      "grad_norm": 0.4504272817775109,
      "learning_rate": 8.069215165553445e-07,
      "loss": 0.2269,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8750
    },
    {
      "epoch": 3.1285790820882196,
      "grad_norm": 0.4169338153102719,
      "learning_rate": 8.006030674245912e-07,
      "loss": 0.2272,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8760
    },
    {
      "epoch": 3.132150723257337,
      "grad_norm": 0.4381106314478239,
      "learning_rate": 7.9430625710063e-07,
      "loss": 0.2239,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8770
    },
    {
      "epoch": 3.135722364426454,
      "grad_norm": 0.46269561831814054,
      "learning_rate": 7.880311360597195e-07,
      "loss": 0.2267,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8780
    },
    {
      "epoch": 3.139294005595571,
      "grad_norm": 0.43404406950330415,
      "learning_rate": 7.817777546042548e-07,
      "loss": 0.223,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8790
    },
    {
      "epoch": 3.1428656467646885,
      "grad_norm": 0.45629754277622947,
      "learning_rate": 7.755461628623651e-07,
      "loss": 0.2266,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8800
    },
    {
      "epoch": 3.1464372879338054,
      "grad_norm": 0.48355699051102885,
      "learning_rate": 7.693364107875062e-07,
      "loss": 0.2175,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8810
    },
    {
      "epoch": 3.1500089291029227,
      "grad_norm": 0.4413863956425271,
      "learning_rate": 7.631485481580636e-07,
      "loss": 0.2221,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8820
    },
    {
      "epoch": 3.15358057027204,
      "grad_norm": 0.43695796833933653,
      "learning_rate": 7.569826245769594e-07,
      "loss": 0.2246,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8830
    },
    {
      "epoch": 3.157152211441157,
      "grad_norm": 0.4668782524426645,
      "learning_rate": 7.508386894712413e-07,
      "loss": 0.2301,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8840
    },
    {
      "epoch": 3.1607238526102743,
      "grad_norm": 0.45885921126660134,
      "learning_rate": 7.447167920916969e-07,
      "loss": 0.2332,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8850
    },
    {
      "epoch": 3.1642954937793917,
      "grad_norm": 0.42552829032477146,
      "learning_rate": 7.386169815124566e-07,
      "loss": 0.2322,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8860
    },
    {
      "epoch": 3.167867134948509,
      "grad_norm": 0.4633441040739579,
      "learning_rate": 7.325393066305984e-07,
      "loss": 0.2276,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8870
    },
    {
      "epoch": 3.171438776117626,
      "grad_norm": 0.48042671789935903,
      "learning_rate": 7.264838161657556e-07,
      "loss": 0.2276,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8880
    },
    {
      "epoch": 3.1750104172867433,
      "grad_norm": 0.4584169432718334,
      "learning_rate": 7.204505586597312e-07,
      "loss": 0.2344,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8890
    },
    {
      "epoch": 3.1785820584558606,
      "grad_norm": 0.4213870889501487,
      "learning_rate": 7.144395824761007e-07,
      "loss": 0.2274,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8900
    },
    {
      "epoch": 3.1821536996249775,
      "grad_norm": 0.43652677559804104,
      "learning_rate": 7.084509357998322e-07,
      "loss": 0.2244,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8910
    },
    {
      "epoch": 3.185725340794095,
      "grad_norm": 0.44176850021681147,
      "learning_rate": 7.024846666368969e-07,
      "loss": 0.2293,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8920
    },
    {
      "epoch": 3.189296981963212,
      "grad_norm": 0.43039405198823494,
      "learning_rate": 6.965408228138819e-07,
      "loss": 0.231,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8930
    },
    {
      "epoch": 3.192868623132329,
      "grad_norm": 0.4680348567715748,
      "learning_rate": 6.906194519776098e-07,
      "loss": 0.224,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8940
    },
    {
      "epoch": 3.1964402643014465,
      "grad_norm": 0.43030595828491525,
      "learning_rate": 6.847206015947576e-07,
      "loss": 0.2213,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8950
    },
    {
      "epoch": 3.200011905470564,
      "grad_norm": 0.44816634746331696,
      "learning_rate": 6.788443189514727e-07,
      "loss": 0.2329,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8960
    },
    {
      "epoch": 3.203583546639681,
      "grad_norm": 0.438369297856931,
      "learning_rate": 6.72990651152998e-07,
      "loss": 0.2279,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8970
    },
    {
      "epoch": 3.207155187808798,
      "grad_norm": 0.4379336123678947,
      "learning_rate": 6.67159645123289e-07,
      "loss": 0.2244,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8980
    },
    {
      "epoch": 3.2107268289779154,
      "grad_norm": 0.45049953635574025,
      "learning_rate": 6.613513476046447e-07,
      "loss": 0.2218,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 8990
    },
    {
      "epoch": 3.2142984701470327,
      "grad_norm": 0.4568418405772703,
      "learning_rate": 6.555658051573249e-07,
      "loss": 0.2288,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9000
    },
    {
      "epoch": 3.2178701113161496,
      "grad_norm": 0.4129714709714023,
      "learning_rate": 6.498030641591848e-07,
      "loss": 0.2224,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9010
    },
    {
      "epoch": 3.221441752485267,
      "grad_norm": 0.4770576540272976,
      "learning_rate": 6.440631708052969e-07,
      "loss": 0.2294,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9020
    },
    {
      "epoch": 3.2250133936543843,
      "grad_norm": 0.4508289702367081,
      "learning_rate": 6.383461711075827e-07,
      "loss": 0.2228,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9030
    },
    {
      "epoch": 3.2285850348235012,
      "grad_norm": 0.5081602228563172,
      "learning_rate": 6.32652110894448e-07,
      "loss": 0.2229,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9040
    },
    {
      "epoch": 3.2321566759926186,
      "grad_norm": 0.42729925975899286,
      "learning_rate": 6.269810358104077e-07,
      "loss": 0.2187,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9050
    },
    {
      "epoch": 3.235728317161736,
      "grad_norm": 0.455081900607026,
      "learning_rate": 6.213329913157267e-07,
      "loss": 0.2302,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9060
    },
    {
      "epoch": 3.239299958330853,
      "grad_norm": 0.4613482712305252,
      "learning_rate": 6.157080226860536e-07,
      "loss": 0.2338,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9070
    },
    {
      "epoch": 3.24287159949997,
      "grad_norm": 0.413583821978395,
      "learning_rate": 6.10106175012055e-07,
      "loss": 0.2269,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9080
    },
    {
      "epoch": 3.2464432406690875,
      "grad_norm": 0.4256358641740858,
      "learning_rate": 6.045274931990558e-07,
      "loss": 0.227,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9090
    },
    {
      "epoch": 3.2500148818382044,
      "grad_norm": 0.4410727602613116,
      "learning_rate": 5.98972021966683e-07,
      "loss": 0.2192,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9100
    },
    {
      "epoch": 3.2535865230073218,
      "grad_norm": 0.42510105742412124,
      "learning_rate": 5.934398058484991e-07,
      "loss": 0.2262,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9110
    },
    {
      "epoch": 3.257158164176439,
      "grad_norm": 0.4421749589580055,
      "learning_rate": 5.879308891916546e-07,
      "loss": 0.2279,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9120
    },
    {
      "epoch": 3.2607298053455565,
      "grad_norm": 0.4465359432618891,
      "learning_rate": 5.824453161565242e-07,
      "loss": 0.2189,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9130
    },
    {
      "epoch": 3.2643014465146734,
      "grad_norm": 0.4386778468888303,
      "learning_rate": 5.769831307163567e-07,
      "loss": 0.2212,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9140
    },
    {
      "epoch": 3.2678730876837907,
      "grad_norm": 0.43969024247923927,
      "learning_rate": 5.715443766569241e-07,
      "loss": 0.2211,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9150
    },
    {
      "epoch": 3.271444728852908,
      "grad_norm": 0.4841280042950919,
      "learning_rate": 5.661290975761674e-07,
      "loss": 0.2222,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9160
    },
    {
      "epoch": 3.275016370022025,
      "grad_norm": 0.43028249447306965,
      "learning_rate": 5.607373368838486e-07,
      "loss": 0.2225,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9170
    },
    {
      "epoch": 3.2785880111911423,
      "grad_norm": 0.45903939584627335,
      "learning_rate": 5.553691378012004e-07,
      "loss": 0.2295,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9180
    },
    {
      "epoch": 3.2821596523602596,
      "grad_norm": 0.4387835891178062,
      "learning_rate": 5.500245433605858e-07,
      "loss": 0.2292,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9190
    },
    {
      "epoch": 3.285731293529377,
      "grad_norm": 0.4390323394974518,
      "learning_rate": 5.447035964051456e-07,
      "loss": 0.2225,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9200
    },
    {
      "epoch": 3.289302934698494,
      "grad_norm": 0.45468304875658144,
      "learning_rate": 5.394063395884604e-07,
      "loss": 0.2263,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9210
    },
    {
      "epoch": 3.2928745758676112,
      "grad_norm": 0.49857421082659625,
      "learning_rate": 5.34132815374206e-07,
      "loss": 0.225,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9220
    },
    {
      "epoch": 3.2964462170367286,
      "grad_norm": 0.4995652805597166,
      "learning_rate": 5.288830660358142e-07,
      "loss": 0.2214,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9230
    },
    {
      "epoch": 3.3000178582058455,
      "grad_norm": 0.47464766483941245,
      "learning_rate": 5.23657133656133e-07,
      "loss": 0.2244,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9240
    },
    {
      "epoch": 3.303589499374963,
      "grad_norm": 0.42707525290171827,
      "learning_rate": 5.184550601270905e-07,
      "loss": 0.2245,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9250
    },
    {
      "epoch": 3.30716114054408,
      "grad_norm": 0.45332045123115117,
      "learning_rate": 5.132768871493571e-07,
      "loss": 0.2235,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9260
    },
    {
      "epoch": 3.310732781713197,
      "grad_norm": 0.4455639041063865,
      "learning_rate": 5.081226562320148e-07,
      "loss": 0.2258,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9270
    },
    {
      "epoch": 3.3143044228823144,
      "grad_norm": 0.4324319313455736,
      "learning_rate": 5.0299240869222e-07,
      "loss": 0.2224,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9280
    },
    {
      "epoch": 3.3178760640514318,
      "grad_norm": 0.42414698673635143,
      "learning_rate": 4.978861856548734e-07,
      "loss": 0.2243,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9290
    },
    {
      "epoch": 3.3214477052205487,
      "grad_norm": 0.42707846365134206,
      "learning_rate": 4.928040280522948e-07,
      "loss": 0.2382,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9300
    },
    {
      "epoch": 3.325019346389666,
      "grad_norm": 0.4246313252911472,
      "learning_rate": 4.877459766238902e-07,
      "loss": 0.2205,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9310
    },
    {
      "epoch": 3.3285909875587834,
      "grad_norm": 0.43253448134906214,
      "learning_rate": 4.827120719158243e-07,
      "loss": 0.2296,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9320
    },
    {
      "epoch": 3.3321626287279003,
      "grad_norm": 0.43450078224608335,
      "learning_rate": 4.777023542807018e-07,
      "loss": 0.229,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9330
    },
    {
      "epoch": 3.3357342698970176,
      "grad_norm": 0.4363961978427439,
      "learning_rate": 4.7271686387723625e-07,
      "loss": 0.2225,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9340
    },
    {
      "epoch": 3.339305911066135,
      "grad_norm": 0.45658061349483414,
      "learning_rate": 4.6775564066993293e-07,
      "loss": 0.2274,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9350
    },
    {
      "epoch": 3.3428775522352523,
      "grad_norm": 0.4509642094193982,
      "learning_rate": 4.6281872442876833e-07,
      "loss": 0.218,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9360
    },
    {
      "epoch": 3.346449193404369,
      "grad_norm": 0.4636424632943862,
      "learning_rate": 4.5790615472886965e-07,
      "loss": 0.2292,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9370
    },
    {
      "epoch": 3.3500208345734865,
      "grad_norm": 0.44021097993476904,
      "learning_rate": 4.5301797095019686e-07,
      "loss": 0.2238,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9380
    },
    {
      "epoch": 3.353592475742604,
      "grad_norm": 0.437952090120706,
      "learning_rate": 4.4815421227723026e-07,
      "loss": 0.2241,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9390
    },
    {
      "epoch": 3.357164116911721,
      "grad_norm": 0.4422016996505597,
      "learning_rate": 4.433149176986555e-07,
      "loss": 0.2262,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9400
    },
    {
      "epoch": 3.360735758080838,
      "grad_norm": 0.4416073497219435,
      "learning_rate": 4.385001260070462e-07,
      "loss": 0.2226,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9410
    },
    {
      "epoch": 3.3643073992499555,
      "grad_norm": 0.4566130926885388,
      "learning_rate": 4.3370987579856075e-07,
      "loss": 0.2265,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9420
    },
    {
      "epoch": 3.3678790404190724,
      "grad_norm": 0.43716008842107645,
      "learning_rate": 4.289442054726267e-07,
      "loss": 0.221,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9430
    },
    {
      "epoch": 3.3714506815881897,
      "grad_norm": 0.4295747135270123,
      "learning_rate": 4.2420315323163524e-07,
      "loss": 0.227,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9440
    },
    {
      "epoch": 3.375022322757307,
      "grad_norm": 0.4344921163984427,
      "learning_rate": 4.194867570806372e-07,
      "loss": 0.2257,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9450
    },
    {
      "epoch": 3.3785939639264244,
      "grad_norm": 0.41936065855618876,
      "learning_rate": 4.14795054827033e-07,
      "loss": 0.2281,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9460
    },
    {
      "epoch": 3.3821656050955413,
      "grad_norm": 0.42885109463625787,
      "learning_rate": 4.1012808408027487e-07,
      "loss": 0.2269,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9470
    },
    {
      "epoch": 3.3857372462646587,
      "grad_norm": 0.42821304096655227,
      "learning_rate": 4.054858822515636e-07,
      "loss": 0.235,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9480
    },
    {
      "epoch": 3.389308887433776,
      "grad_norm": 0.46047010134247357,
      "learning_rate": 4.008684865535473e-07,
      "loss": 0.2148,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9490
    },
    {
      "epoch": 3.392880528602893,
      "grad_norm": 0.450023215014394,
      "learning_rate": 3.9627593400002247e-07,
      "loss": 0.2186,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9500
    },
    {
      "epoch": 3.3964521697720103,
      "grad_norm": 0.4628252521903469,
      "learning_rate": 3.917082614056427e-07,
      "loss": 0.225,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9510
    },
    {
      "epoch": 3.4000238109411276,
      "grad_norm": 0.4377616291629553,
      "learning_rate": 3.871655053856167e-07,
      "loss": 0.2228,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9520
    },
    {
      "epoch": 3.4035954521102445,
      "grad_norm": 0.4204372477172826,
      "learning_rate": 3.8264770235541844e-07,
      "loss": 0.233,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9530
    },
    {
      "epoch": 3.407167093279362,
      "grad_norm": 0.45383296284555796,
      "learning_rate": 3.7815488853049534e-07,
      "loss": 0.2227,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9540
    },
    {
      "epoch": 3.410738734448479,
      "grad_norm": 0.4256157176449563,
      "learning_rate": 3.736870999259782e-07,
      "loss": 0.2255,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9550
    },
    {
      "epoch": 3.414310375617596,
      "grad_norm": 0.42932735183625526,
      "learning_rate": 3.6924437235638823e-07,
      "loss": 0.2272,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9560
    },
    {
      "epoch": 3.4178820167867134,
      "grad_norm": 0.4395411809664829,
      "learning_rate": 3.648267414353572e-07,
      "loss": 0.2226,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9570
    },
    {
      "epoch": 3.421453657955831,
      "grad_norm": 0.4789395353295623,
      "learning_rate": 3.60434242575336e-07,
      "loss": 0.2194,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9580
    },
    {
      "epoch": 3.4250252991249477,
      "grad_norm": 0.46026844417244916,
      "learning_rate": 3.5606691098731145e-07,
      "loss": 0.229,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9590
    },
    {
      "epoch": 3.428596940294065,
      "grad_norm": 0.4548661544045963,
      "learning_rate": 3.5172478168052915e-07,
      "loss": 0.2269,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9600
    },
    {
      "epoch": 3.4321685814631824,
      "grad_norm": 0.43693949940193977,
      "learning_rate": 3.47407889462206e-07,
      "loss": 0.2295,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9610
    },
    {
      "epoch": 3.4357402226322997,
      "grad_norm": 0.4358578582107585,
      "learning_rate": 3.431162689372535e-07,
      "loss": 0.2221,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9620
    },
    {
      "epoch": 3.4393118638014166,
      "grad_norm": 0.4277147932086469,
      "learning_rate": 3.3884995450800646e-07,
      "loss": 0.2322,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9630
    },
    {
      "epoch": 3.442883504970534,
      "grad_norm": 0.457090524162981,
      "learning_rate": 3.3460898037393604e-07,
      "loss": 0.2306,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9640
    },
    {
      "epoch": 3.4464551461396513,
      "grad_norm": 0.48622748466946314,
      "learning_rate": 3.3039338053138333e-07,
      "loss": 0.2257,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9650
    },
    {
      "epoch": 3.450026787308768,
      "grad_norm": 0.447418853195856,
      "learning_rate": 3.2620318877328716e-07,
      "loss": 0.2285,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9660
    },
    {
      "epoch": 3.4535984284778856,
      "grad_norm": 0.4494933647456064,
      "learning_rate": 3.22038438688908e-07,
      "loss": 0.2154,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9670
    },
    {
      "epoch": 3.457170069647003,
      "grad_norm": 0.4386760172527179,
      "learning_rate": 3.1789916366356237e-07,
      "loss": 0.2337,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9680
    },
    {
      "epoch": 3.4607417108161203,
      "grad_norm": 0.4472559976730384,
      "learning_rate": 3.137853968783562e-07,
      "loss": 0.2257,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9690
    },
    {
      "epoch": 3.464313351985237,
      "grad_norm": 0.4130599398053046,
      "learning_rate": 3.096971713099152e-07,
      "loss": 0.2307,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9700
    },
    {
      "epoch": 3.4678849931543545,
      "grad_norm": 0.44974189566383543,
      "learning_rate": 3.056345197301239e-07,
      "loss": 0.2293,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9710
    },
    {
      "epoch": 3.471456634323472,
      "grad_norm": 0.433647374405478,
      "learning_rate": 3.0159747470586114e-07,
      "loss": 0.227,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9720
    },
    {
      "epoch": 3.4750282754925887,
      "grad_norm": 0.4573097157873816,
      "learning_rate": 2.9758606859873936e-07,
      "loss": 0.2255,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9730
    },
    {
      "epoch": 3.478599916661706,
      "grad_norm": 0.4429742982193052,
      "learning_rate": 2.9360033356484463e-07,
      "loss": 0.2278,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9740
    },
    {
      "epoch": 3.4821715578308234,
      "grad_norm": 0.45668296151320453,
      "learning_rate": 2.8964030155448055e-07,
      "loss": 0.2225,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9750
    },
    {
      "epoch": 3.4857431989999403,
      "grad_norm": 0.44853320183811274,
      "learning_rate": 2.85706004311911e-07,
      "loss": 0.2332,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9760
    },
    {
      "epoch": 3.4893148401690577,
      "grad_norm": 0.4896496538364987,
      "learning_rate": 2.8179747337510424e-07,
      "loss": 0.2151,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9770
    },
    {
      "epoch": 3.492886481338175,
      "grad_norm": 0.45216838981680796,
      "learning_rate": 2.779147400754833e-07,
      "loss": 0.2284,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9780
    },
    {
      "epoch": 3.496458122507292,
      "grad_norm": 0.4446772622560835,
      "learning_rate": 2.740578355376736e-07,
      "loss": 0.2184,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9790
    },
    {
      "epoch": 3.5000297636764093,
      "grad_norm": 0.4468530070722961,
      "learning_rate": 2.7022679067925045e-07,
      "loss": 0.2292,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9800
    },
    {
      "epoch": 3.5036014048455266,
      "grad_norm": 0.434552824796451,
      "learning_rate": 2.664216362104964e-07,
      "loss": 0.2206,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9810
    },
    {
      "epoch": 3.5071730460146435,
      "grad_norm": 0.4328854611895299,
      "learning_rate": 2.6264240263415113e-07,
      "loss": 0.2244,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9820
    },
    {
      "epoch": 3.510744687183761,
      "grad_norm": 0.4782923737873149,
      "learning_rate": 2.5888912024516743e-07,
      "loss": 0.2177,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9830
    },
    {
      "epoch": 3.514316328352878,
      "grad_norm": 0.4451929903668464,
      "learning_rate": 2.5516181913047134e-07,
      "loss": 0.2278,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9840
    },
    {
      "epoch": 3.517887969521995,
      "grad_norm": 0.461154244396955,
      "learning_rate": 2.5146052916871664e-07,
      "loss": 0.2269,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9850
    },
    {
      "epoch": 3.5214596106911125,
      "grad_norm": 0.44185695213818643,
      "learning_rate": 2.4778528003004823e-07,
      "loss": 0.2206,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9860
    },
    {
      "epoch": 3.52503125186023,
      "grad_norm": 0.44663018732053594,
      "learning_rate": 2.441361011758647e-07,
      "loss": 0.2211,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9870
    },
    {
      "epoch": 3.5286028930293467,
      "grad_norm": 0.4279233693907992,
      "learning_rate": 2.4051302185857915e-07,
      "loss": 0.2282,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9880
    },
    {
      "epoch": 3.532174534198464,
      "grad_norm": 0.5027365223037497,
      "learning_rate": 2.3691607112138646e-07,
      "loss": 0.2191,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9890
    },
    {
      "epoch": 3.5357461753675814,
      "grad_norm": 0.4283371243464985,
      "learning_rate": 2.3334527779803294e-07,
      "loss": 0.2231,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9900
    },
    {
      "epoch": 3.5393178165366987,
      "grad_norm": 0.451203596680216,
      "learning_rate": 2.2980067051258035e-07,
      "loss": 0.2208,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9910
    },
    {
      "epoch": 3.542889457705816,
      "grad_norm": 0.44280537011849175,
      "learning_rate": 2.2628227767917946e-07,
      "loss": 0.2168,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9920
    },
    {
      "epoch": 3.546461098874933,
      "grad_norm": 0.43243809254606863,
      "learning_rate": 2.2279012750184228e-07,
      "loss": 0.2279,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9930
    },
    {
      "epoch": 3.5500327400440503,
      "grad_norm": 0.4587688082684246,
      "learning_rate": 2.1932424797421556e-07,
      "loss": 0.2223,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9940
    },
    {
      "epoch": 3.5536043812131677,
      "grad_norm": 0.4638395466260308,
      "learning_rate": 2.1588466687935497e-07,
      "loss": 0.2275,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9950
    },
    {
      "epoch": 3.5571760223822846,
      "grad_norm": 0.4545490936757308,
      "learning_rate": 2.1247141178950523e-07,
      "loss": 0.2217,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9960
    },
    {
      "epoch": 3.560747663551402,
      "grad_norm": 0.47690321382830164,
      "learning_rate": 2.0908451006587663e-07,
      "loss": 0.2254,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9970
    },
    {
      "epoch": 3.5643193047205193,
      "grad_norm": 0.4392958225160946,
      "learning_rate": 2.0572398885842592e-07,
      "loss": 0.2148,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9980
    },
    {
      "epoch": 3.567890945889636,
      "grad_norm": 0.4239943589734696,
      "learning_rate": 2.0238987510564094e-07,
      "loss": 0.2238,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 9990
    },
    {
      "epoch": 3.5714625870587535,
      "grad_norm": 0.44038559554116197,
      "learning_rate": 1.99082195534322e-07,
      "loss": 0.2304,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10000
    },
    {
      "epoch": 3.575034228227871,
      "grad_norm": 0.4345253477997,
      "learning_rate": 1.9580097665936762e-07,
      "loss": 0.2287,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10010
    },
    {
      "epoch": 3.5786058693969878,
      "grad_norm": 0.4403790298956109,
      "learning_rate": 1.925462447835668e-07,
      "loss": 0.2268,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10020
    },
    {
      "epoch": 3.582177510566105,
      "grad_norm": 0.448263330638005,
      "learning_rate": 1.8931802599738046e-07,
      "loss": 0.2354,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10030
    },
    {
      "epoch": 3.5857491517352225,
      "grad_norm": 0.43699511153186094,
      "learning_rate": 1.8611634617873757e-07,
      "loss": 0.2273,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10040
    },
    {
      "epoch": 3.5893207929043394,
      "grad_norm": 0.4402260519814635,
      "learning_rate": 1.829412309928272e-07,
      "loss": 0.225,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10050
    },
    {
      "epoch": 3.5928924340734567,
      "grad_norm": 0.44922009323567325,
      "learning_rate": 1.7979270589189111e-07,
      "loss": 0.2191,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10060
    },
    {
      "epoch": 3.596464075242574,
      "grad_norm": 0.4348279357761571,
      "learning_rate": 1.7667079611502073e-07,
      "loss": 0.2248,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10070
    },
    {
      "epoch": 3.600035716411691,
      "grad_norm": 0.4224705514098165,
      "learning_rate": 1.7357552668795528e-07,
      "loss": 0.2255,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10080
    },
    {
      "epoch": 3.6036073575808083,
      "grad_norm": 0.4516505048559037,
      "learning_rate": 1.7050692242287923e-07,
      "loss": 0.2283,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10090
    },
    {
      "epoch": 3.6071789987499256,
      "grad_norm": 0.43040590998867473,
      "learning_rate": 1.6746500791822649e-07,
      "loss": 0.2199,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10100
    },
    {
      "epoch": 3.6107506399190425,
      "grad_norm": 0.47700956208234957,
      "learning_rate": 1.6444980755848065e-07,
      "loss": 0.2191,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10110
    },
    {
      "epoch": 3.61432228108816,
      "grad_norm": 0.4546842042912744,
      "learning_rate": 1.6146134551398033e-07,
      "loss": 0.2178,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10120
    },
    {
      "epoch": 3.6178939222572772,
      "grad_norm": 0.44712219892480254,
      "learning_rate": 1.584996457407249e-07,
      "loss": 0.2275,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10130
    },
    {
      "epoch": 3.6214655634263946,
      "grad_norm": 0.42639223794691933,
      "learning_rate": 1.555647319801844e-07,
      "loss": 0.2284,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10140
    },
    {
      "epoch": 3.6250372045955115,
      "grad_norm": 0.445529806494021,
      "learning_rate": 1.526566277591065e-07,
      "loss": 0.2211,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10150
    },
    {
      "epoch": 3.628608845764629,
      "grad_norm": 0.43147908453278766,
      "learning_rate": 1.4977535638932914e-07,
      "loss": 0.2275,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10160
    },
    {
      "epoch": 3.632180486933746,
      "grad_norm": 0.4399996543390641,
      "learning_rate": 1.469209409675953e-07,
      "loss": 0.2281,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10170
    },
    {
      "epoch": 3.6357521281028635,
      "grad_norm": 0.4523298177857587,
      "learning_rate": 1.4409340437536543e-07,
      "loss": 0.231,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10180
    },
    {
      "epoch": 3.6393237692719804,
      "grad_norm": 0.4450819890687884,
      "learning_rate": 1.4129276927863366e-07,
      "loss": 0.2253,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10190
    },
    {
      "epoch": 3.6428954104410978,
      "grad_norm": 0.4445668634013846,
      "learning_rate": 1.3851905812774945e-07,
      "loss": 0.2219,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10200
    },
    {
      "epoch": 3.646467051610215,
      "grad_norm": 0.4393809602135747,
      "learning_rate": 1.3577229315723405e-07,
      "loss": 0.2227,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10210
    },
    {
      "epoch": 3.650038692779332,
      "grad_norm": 0.4392939224137732,
      "learning_rate": 1.3305249638560396e-07,
      "loss": 0.2176,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10220
    },
    {
      "epoch": 3.6536103339484494,
      "grad_norm": 0.4550312708413194,
      "learning_rate": 1.3035968961519534e-07,
      "loss": 0.2304,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10230
    },
    {
      "epoch": 3.6571819751175667,
      "grad_norm": 0.44151330373786013,
      "learning_rate": 1.2769389443198603e-07,
      "loss": 0.2284,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10240
    },
    {
      "epoch": 3.6607536162866836,
      "grad_norm": 0.4168807111228616,
      "learning_rate": 1.2505513220542614e-07,
      "loss": 0.2294,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10250
    },
    {
      "epoch": 3.664325257455801,
      "grad_norm": 0.4464258397287314,
      "learning_rate": 1.2244342408826508e-07,
      "loss": 0.2313,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10260
    },
    {
      "epoch": 3.6678968986249183,
      "grad_norm": 0.45152627548033747,
      "learning_rate": 1.1985879101638148e-07,
      "loss": 0.2184,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10270
    },
    {
      "epoch": 3.671468539794035,
      "grad_norm": 0.47117490838138765,
      "learning_rate": 1.1730125370861632e-07,
      "loss": 0.2257,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10280
    },
    {
      "epoch": 3.6750401809631525,
      "grad_norm": 0.439951465915027,
      "learning_rate": 1.1477083266660642e-07,
      "loss": 0.2328,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10290
    },
    {
      "epoch": 3.67861182213227,
      "grad_norm": 0.4489829456467373,
      "learning_rate": 1.122675481746207e-07,
      "loss": 0.225,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10300
    },
    {
      "epoch": 3.682183463301387,
      "grad_norm": 0.44134052549882896,
      "learning_rate": 1.0979142029939548e-07,
      "loss": 0.2275,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10310
    },
    {
      "epoch": 3.685755104470504,
      "grad_norm": 0.46831622118801114,
      "learning_rate": 1.073424688899775e-07,
      "loss": 0.2237,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10320
    },
    {
      "epoch": 3.6893267456396215,
      "grad_norm": 0.45036094056520976,
      "learning_rate": 1.0492071357756033e-07,
      "loss": 0.2249,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10330
    },
    {
      "epoch": 3.6928983868087384,
      "grad_norm": 0.4659459059342654,
      "learning_rate": 1.0252617377533046e-07,
      "loss": 0.2284,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10340
    },
    {
      "epoch": 3.6964700279778557,
      "grad_norm": 0.4225585197970624,
      "learning_rate": 1.0015886867831115e-07,
      "loss": 0.2321,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10350
    },
    {
      "epoch": 3.700041669146973,
      "grad_norm": 0.44704385962317833,
      "learning_rate": 9.781881726320573e-08,
      "loss": 0.2256,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10360
    },
    {
      "epoch": 3.70361331031609,
      "grad_norm": 0.4549601644611548,
      "learning_rate": 9.550603828824816e-08,
      "loss": 0.2248,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10370
    },
    {
      "epoch": 3.7071849514852073,
      "grad_norm": 0.447148866393789,
      "learning_rate": 9.322055029305403e-08,
      "loss": 0.2286,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10380
    },
    {
      "epoch": 3.7107565926543247,
      "grad_norm": 0.45463473593785586,
      "learning_rate": 9.096237159846721e-08,
      "loss": 0.2227,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10390
    },
    {
      "epoch": 3.714328233823442,
      "grad_norm": 0.4412149959468804,
      "learning_rate": 8.873152030641751e-08,
      "loss": 0.2279,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10400
    },
    {
      "epoch": 3.717899874992559,
      "grad_norm": 0.4495044254264032,
      "learning_rate": 8.652801429977353e-08,
      "loss": 0.2251,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10410
    },
    {
      "epoch": 3.7214715161616763,
      "grad_norm": 0.43845028047131007,
      "learning_rate": 8.435187124219878e-08,
      "loss": 0.225,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10420
    },
    {
      "epoch": 3.7250431573307936,
      "grad_norm": 0.4599594623404225,
      "learning_rate": 8.220310857801189e-08,
      "loss": 0.2242,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10430
    },
    {
      "epoch": 3.728614798499911,
      "grad_norm": 0.4584359385494083,
      "learning_rate": 8.008174353204584e-08,
      "loss": 0.2351,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10440
    },
    {
      "epoch": 3.732186439669028,
      "grad_norm": 0.43163046353373913,
      "learning_rate": 7.798779310950854e-08,
      "loss": 0.2261,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10450
    },
    {
      "epoch": 3.735758080838145,
      "grad_norm": 0.43734187492828513,
      "learning_rate": 7.592127409584915e-08,
      "loss": 0.2289,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10460
    },
    {
      "epoch": 3.7393297220072625,
      "grad_norm": 0.43309527476787557,
      "learning_rate": 7.388220305662163e-08,
      "loss": 0.2285,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10470
    },
    {
      "epoch": 3.7429013631763794,
      "grad_norm": 0.4344606568256094,
      "learning_rate": 7.187059633735193e-08,
      "loss": 0.2331,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10480
    },
    {
      "epoch": 3.746473004345497,
      "grad_norm": 0.43446478923577986,
      "learning_rate": 6.988647006340892e-08,
      "loss": 0.2317,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10490
    },
    {
      "epoch": 3.750044645514614,
      "grad_norm": 0.4356415389064807,
      "learning_rate": 6.792984013987308e-08,
      "loss": 0.2297,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10500
    },
    {
      "epoch": 3.753616286683731,
      "grad_norm": 0.44211853419334296,
      "learning_rate": 6.600072225140902e-08,
      "loss": 0.2283,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10510
    },
    {
      "epoch": 3.7571879278528484,
      "grad_norm": 0.436758452421375,
      "learning_rate": 6.409913186214078e-08,
      "loss": 0.2349,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10520
    },
    {
      "epoch": 3.7607595690219657,
      "grad_norm": 0.43053855087980014,
      "learning_rate": 6.222508421552785e-08,
      "loss": 0.2297,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10530
    },
    {
      "epoch": 3.7643312101910826,
      "grad_norm": 0.4558435875308646,
      "learning_rate": 6.037859433424203e-08,
      "loss": 0.233,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10540
    },
    {
      "epoch": 3.7679028513602,
      "grad_norm": 0.4389567806893566,
      "learning_rate": 5.855967702004844e-08,
      "loss": 0.2186,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10550
    },
    {
      "epoch": 3.7714744925293173,
      "grad_norm": 0.4360877408318116,
      "learning_rate": 5.676834685368515e-08,
      "loss": 0.2266,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10560
    },
    {
      "epoch": 3.7750461336984342,
      "grad_norm": 0.4356712867921081,
      "learning_rate": 5.500461819474694e-08,
      "loss": 0.2294,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10570
    },
    {
      "epoch": 3.7786177748675516,
      "grad_norm": 0.43833812335077477,
      "learning_rate": 5.326850518157111e-08,
      "loss": 0.2279,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10580
    },
    {
      "epoch": 3.782189416036669,
      "grad_norm": 0.44419001553486764,
      "learning_rate": 5.156002173112356e-08,
      "loss": 0.2278,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10590
    },
    {
      "epoch": 3.785761057205786,
      "grad_norm": 0.4533306175595154,
      "learning_rate": 4.9879181538886154e-08,
      "loss": 0.2249,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10600
    },
    {
      "epoch": 3.789332698374903,
      "grad_norm": 0.43562953329475035,
      "learning_rate": 4.8225998078749054e-08,
      "loss": 0.2299,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10610
    },
    {
      "epoch": 3.7929043395440205,
      "grad_norm": 0.4515448260405365,
      "learning_rate": 4.660048460290078e-08,
      "loss": 0.2139,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10620
    },
    {
      "epoch": 3.7964759807131374,
      "grad_norm": 0.445160563686723,
      "learning_rate": 4.50026541417225e-08,
      "loss": 0.2329,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10630
    },
    {
      "epoch": 3.8000476218822548,
      "grad_norm": 0.4327457321668427,
      "learning_rate": 4.343251950368465e-08,
      "loss": 0.2363,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10640
    },
    {
      "epoch": 3.803619263051372,
      "grad_norm": 0.445610865255777,
      "learning_rate": 4.1890093275242846e-08,
      "loss": 0.2306,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10650
    },
    {
      "epoch": 3.8071909042204894,
      "grad_norm": 0.4360425724072501,
      "learning_rate": 4.0375387820737574e-08,
      "loss": 0.2355,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10660
    },
    {
      "epoch": 3.810762545389607,
      "grad_norm": 0.45104819913425176,
      "learning_rate": 3.888841528229553e-08,
      "loss": 0.2242,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10670
    },
    {
      "epoch": 3.8143341865587237,
      "grad_norm": 0.43201411969010634,
      "learning_rate": 3.742918757973129e-08,
      "loss": 0.2233,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10680
    },
    {
      "epoch": 3.817905827727841,
      "grad_norm": 0.4330792005779704,
      "learning_rate": 3.5997716410452127e-08,
      "loss": 0.223,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10690
    },
    {
      "epoch": 3.8214774688969584,
      "grad_norm": 0.4642141392644615,
      "learning_rate": 3.4594013249365507e-08,
      "loss": 0.2247,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10700
    },
    {
      "epoch": 3.8250491100660753,
      "grad_norm": 0.46034792920824497,
      "learning_rate": 3.3218089348785065e-08,
      "loss": 0.2279,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10710
    },
    {
      "epoch": 3.8286207512351926,
      "grad_norm": 0.443689997182439,
      "learning_rate": 3.1869955738341225e-08,
      "loss": 0.2303,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10720
    },
    {
      "epoch": 3.83219239240431,
      "grad_norm": 0.42628378709969905,
      "learning_rate": 3.054962322489341e-08,
      "loss": 0.2262,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10730
    },
    {
      "epoch": 3.835764033573427,
      "grad_norm": 0.4395794254721828,
      "learning_rate": 2.9257102392442946e-08,
      "loss": 0.2271,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10740
    },
    {
      "epoch": 3.8393356747425442,
      "grad_norm": 0.44880114793329107,
      "learning_rate": 2.7992403602047632e-08,
      "loss": 0.227,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10750
    },
    {
      "epoch": 3.8429073159116616,
      "grad_norm": 0.44743156837785386,
      "learning_rate": 2.6755536991739714e-08,
      "loss": 0.2214,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10760
    },
    {
      "epoch": 3.8464789570807785,
      "grad_norm": 0.4468589013028941,
      "learning_rate": 2.5546512476444294e-08,
      "loss": 0.2228,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10770
    },
    {
      "epoch": 3.850050598249896,
      "grad_norm": 0.4567894351522827,
      "learning_rate": 2.4365339747898893e-08,
      "loss": 0.2192,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10780
    },
    {
      "epoch": 3.853622239419013,
      "grad_norm": 0.4532132712813433,
      "learning_rate": 2.321202827457769e-08,
      "loss": 0.2303,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10790
    },
    {
      "epoch": 3.85719388058813,
      "grad_norm": 0.4490186477309506,
      "learning_rate": 2.2086587301613393e-08,
      "loss": 0.2304,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10800
    },
    {
      "epoch": 3.8607655217572474,
      "grad_norm": 0.43279214372519503,
      "learning_rate": 2.0989025850724206e-08,
      "loss": 0.2321,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10810
    },
    {
      "epoch": 3.8643371629263648,
      "grad_norm": 0.43545063828161296,
      "learning_rate": 1.991935272014311e-08,
      "loss": 0.2223,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10820
    },
    {
      "epoch": 3.8679088040954817,
      "grad_norm": 0.416421287059464,
      "learning_rate": 1.8877576484544022e-08,
      "loss": 0.2295,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10830
    },
    {
      "epoch": 3.871480445264599,
      "grad_norm": 0.44598171266797837,
      "learning_rate": 1.7863705494975736e-08,
      "loss": 0.2214,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10840
    },
    {
      "epoch": 3.8750520864337163,
      "grad_norm": 0.4293547056423775,
      "learning_rate": 1.6877747878793536e-08,
      "loss": 0.2241,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10850
    },
    {
      "epoch": 3.8786237276028332,
      "grad_norm": 0.44628839462942493,
      "learning_rate": 1.5919711539594706e-08,
      "loss": 0.2254,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10860
    },
    {
      "epoch": 3.8821953687719506,
      "grad_norm": 0.44532618818831393,
      "learning_rate": 1.4989604157154777e-08,
      "loss": 0.2322,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10870
    },
    {
      "epoch": 3.885767009941068,
      "grad_norm": 0.4620906666796364,
      "learning_rate": 1.4087433187367326e-08,
      "loss": 0.223,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10880
    },
    {
      "epoch": 3.8893386511101853,
      "grad_norm": 0.4519981254148374,
      "learning_rate": 1.3213205862181387e-08,
      "loss": 0.2339,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10890
    },
    {
      "epoch": 3.892910292279302,
      "grad_norm": 0.46209671155053056,
      "learning_rate": 1.2366929189546295e-08,
      "loss": 0.2286,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10900
    },
    {
      "epoch": 3.8964819334484195,
      "grad_norm": 0.4477262080879772,
      "learning_rate": 1.1548609953354549e-08,
      "loss": 0.2266,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10910
    },
    {
      "epoch": 3.900053574617537,
      "grad_norm": 0.41528706320813497,
      "learning_rate": 1.0758254713386262e-08,
      "loss": 0.2334,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10920
    },
    {
      "epoch": 3.9036252157866542,
      "grad_norm": 0.4380520934393717,
      "learning_rate": 9.995869805257852e-09,
      "loss": 0.2203,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10930
    },
    {
      "epoch": 3.907196856955771,
      "grad_norm": 0.44084027878315196,
      "learning_rate": 9.261461340372313e-09,
      "loss": 0.2237,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10940
    },
    {
      "epoch": 3.9107684981248885,
      "grad_norm": 0.4557862047592466,
      "learning_rate": 8.555035205867145e-09,
      "loss": 0.2226,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10950
    },
    {
      "epoch": 3.914340139294006,
      "grad_norm": 0.46256250332145743,
      "learning_rate": 7.876597064569668e-09,
      "loss": 0.2191,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10960
    },
    {
      "epoch": 3.9179117804631227,
      "grad_norm": 0.45890967205133926,
      "learning_rate": 7.226152354951165e-09,
      "loss": 0.2202,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10970
    },
    {
      "epoch": 3.92148342163224,
      "grad_norm": 0.44125688697910415,
      "learning_rate": 6.603706291082201e-09,
      "loss": 0.2245,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10980
    },
    {
      "epoch": 3.9250550628013574,
      "grad_norm": 0.43913591864768914,
      "learning_rate": 6.009263862592596e-09,
      "loss": 0.2209,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 10990
    },
    {
      "epoch": 3.9286267039704743,
      "grad_norm": 0.42891856210394647,
      "learning_rate": 5.442829834630236e-09,
      "loss": 0.229,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11000
    },
    {
      "epoch": 3.9321983451395917,
      "grad_norm": 0.43191273643462674,
      "learning_rate": 4.904408747821831e-09,
      "loss": 0.2241,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11010
    },
    {
      "epoch": 3.935769986308709,
      "grad_norm": 0.4321814680588804,
      "learning_rate": 4.3940049182394916e-09,
      "loss": 0.2292,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11020
    },
    {
      "epoch": 3.939341627477826,
      "grad_norm": 0.4440803966779172,
      "learning_rate": 3.911622437363427e-09,
      "loss": 0.2324,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11030
    },
    {
      "epoch": 3.9429132686469432,
      "grad_norm": 0.44630919988961,
      "learning_rate": 3.4572651720508605e-09,
      "loss": 0.2291,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11040
    },
    {
      "epoch": 3.9464849098160606,
      "grad_norm": 0.4479542040467652,
      "learning_rate": 3.0309367645030004e-09,
      "loss": 0.2312,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11050
    },
    {
      "epoch": 3.9500565509851775,
      "grad_norm": 0.44081290560579367,
      "learning_rate": 2.632640632239003e-09,
      "loss": 0.2228,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11060
    },
    {
      "epoch": 3.953628192154295,
      "grad_norm": 0.47217543215125773,
      "learning_rate": 2.262379968064887e-09,
      "loss": 0.2219,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11070
    },
    {
      "epoch": 3.957199833323412,
      "grad_norm": 0.43867083497842563,
      "learning_rate": 1.920157740050998e-09,
      "loss": 0.2305,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11080
    },
    {
      "epoch": 3.960771474492529,
      "grad_norm": 0.4391386544004875,
      "learning_rate": 1.6059766915067496e-09,
      "loss": 0.2272,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11090
    },
    {
      "epoch": 3.9643431156616464,
      "grad_norm": 0.45329886069059977,
      "learning_rate": 1.319839340958473e-09,
      "loss": 0.2257,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11100
    },
    {
      "epoch": 3.967914756830764,
      "grad_norm": 0.4698347155108858,
      "learning_rate": 1.0617479821303787e-09,
      "loss": 0.2253,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11110
    },
    {
      "epoch": 3.9714863979998807,
      "grad_norm": 0.4503276770969569,
      "learning_rate": 8.317046839243503e-10,
      "loss": 0.2277,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11120
    },
    {
      "epoch": 3.975058039168998,
      "grad_norm": 0.44723000513320477,
      "learning_rate": 6.297112904051771e-10,
      "loss": 0.2327,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11130
    },
    {
      "epoch": 3.9786296803381154,
      "grad_norm": 0.4400304568539704,
      "learning_rate": 4.5576942078501267e-10,
      "loss": 0.2251,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11140
    },
    {
      "epoch": 3.9822013215072327,
      "grad_norm": 0.44837484863154975,
      "learning_rate": 3.0988046941055057e-10,
      "loss": 0.2273,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11150
    },
    {
      "epoch": 3.9857729626763496,
      "grad_norm": 0.4270143306404482,
      "learning_rate": 1.9204560575136753e-10,
      "loss": 0.232,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11160
    },
    {
      "epoch": 3.989344603845467,
      "grad_norm": 0.44880561675935815,
      "learning_rate": 1.0226577439059748e-10,
      "loss": 0.2289,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11170
    },
    {
      "epoch": 3.9929162450145843,
      "grad_norm": 0.43633117229278284,
      "learning_rate": 4.054169501832572e-11,
      "loss": 0.2292,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11180
    },
    {
      "epoch": 3.9964878861837017,
      "grad_norm": 0.4732141126000996,
      "learning_rate": 6.873862424983113e-12,
      "loss": 0.2291,
      "memory/device_mem_reserved(gib)": 49.07,
      "memory/max_mem_active(gib)": 43.76,
      "memory/max_mem_allocated(gib)": 42.86,
      "step": 11190
    }
  ],
  "logging_steps": 10,
  "max_steps": 11196,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 400,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.217947929182208e+16,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}