{
  "best_metric": 0.006617749575525522,
  "best_model_checkpoint": "vit_epochs5_batch64_lr5e-05_size224_tiles1_seed1_classic_image_classification/checkpoint-1410",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1410,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01773049645390071,
      "grad_norm": 1.1057369709014893,
      "learning_rate": 4.9822695035461e-05,
      "loss": 0.6106,
      "step": 5
    },
    {
      "epoch": 0.03546099290780142,
      "grad_norm": 1.056671142578125,
      "learning_rate": 4.964539007092199e-05,
      "loss": 0.462,
      "step": 10
    },
    {
      "epoch": 0.05319148936170213,
      "grad_norm": 0.9491887092590332,
      "learning_rate": 4.946808510638298e-05,
      "loss": 0.31,
      "step": 15
    },
    {
      "epoch": 0.07092198581560284,
      "grad_norm": 1.2302058935165405,
      "learning_rate": 4.929078014184397e-05,
      "loss": 0.2197,
      "step": 20
    },
    {
      "epoch": 0.08865248226950355,
      "grad_norm": 1.1420009136199951,
      "learning_rate": 4.911347517730497e-05,
      "loss": 0.1468,
      "step": 25
    },
    {
      "epoch": 0.10638297872340426,
      "grad_norm": 1.3249993324279785,
      "learning_rate": 4.893617021276596e-05,
      "loss": 0.1098,
      "step": 30
    },
    {
      "epoch": 0.12411347517730496,
      "grad_norm": 0.8271305561065674,
      "learning_rate": 4.875886524822695e-05,
      "loss": 0.0994,
      "step": 35
    },
    {
      "epoch": 0.14184397163120568,
      "grad_norm": 1.1268686056137085,
      "learning_rate": 4.858156028368794e-05,
      "loss": 0.0718,
      "step": 40
    },
    {
      "epoch": 0.1595744680851064,
      "grad_norm": 0.8098769187927246,
      "learning_rate": 4.840425531914894e-05,
      "loss": 0.0861,
      "step": 45
    },
    {
      "epoch": 0.1773049645390071,
      "grad_norm": 1.5605119466781616,
      "learning_rate": 4.822695035460993e-05,
      "loss": 0.0882,
      "step": 50
    },
    {
      "epoch": 0.1950354609929078,
      "grad_norm": 0.6594899892807007,
      "learning_rate": 4.804964539007092e-05,
      "loss": 0.0567,
      "step": 55
    },
    {
      "epoch": 0.2127659574468085,
      "grad_norm": 0.3054638206958771,
      "learning_rate": 4.787234042553192e-05,
      "loss": 0.043,
      "step": 60
    },
    {
      "epoch": 0.23049645390070922,
      "grad_norm": 1.3071125745773315,
      "learning_rate": 4.769503546099291e-05,
      "loss": 0.0675,
      "step": 65
    },
    {
      "epoch": 0.24822695035460993,
      "grad_norm": 1.4442698955535889,
      "learning_rate": 4.751773049645391e-05,
      "loss": 0.0818,
      "step": 70
    },
    {
      "epoch": 0.26595744680851063,
      "grad_norm": 1.443562388420105,
      "learning_rate": 4.734042553191489e-05,
      "loss": 0.09,
      "step": 75
    },
    {
      "epoch": 0.28368794326241137,
      "grad_norm": 1.0713045597076416,
      "learning_rate": 4.716312056737589e-05,
      "loss": 0.0766,
      "step": 80
    },
    {
      "epoch": 0.30141843971631205,
      "grad_norm": 0.24775153398513794,
      "learning_rate": 4.698581560283688e-05,
      "loss": 0.067,
      "step": 85
    },
    {
      "epoch": 0.3191489361702128,
      "grad_norm": 1.436522126197815,
      "learning_rate": 4.680851063829788e-05,
      "loss": 0.0792,
      "step": 90
    },
    {
      "epoch": 0.33687943262411346,
      "grad_norm": 0.49839749932289124,
      "learning_rate": 4.663120567375887e-05,
      "loss": 0.0595,
      "step": 95
    },
    {
      "epoch": 0.3546099290780142,
      "grad_norm": 0.6618617177009583,
      "learning_rate": 4.645390070921986e-05,
      "loss": 0.036,
      "step": 100
    },
    {
      "epoch": 0.3723404255319149,
      "grad_norm": 0.8074954748153687,
      "learning_rate": 4.627659574468085e-05,
      "loss": 0.0589,
      "step": 105
    },
    {
      "epoch": 0.3900709219858156,
      "grad_norm": 1.512774109840393,
      "learning_rate": 4.609929078014185e-05,
      "loss": 0.0477,
      "step": 110
    },
    {
      "epoch": 0.4078014184397163,
      "grad_norm": 1.3695027828216553,
      "learning_rate": 4.592198581560284e-05,
      "loss": 0.1018,
      "step": 115
    },
    {
      "epoch": 0.425531914893617,
      "grad_norm": 0.7358906865119934,
      "learning_rate": 4.574468085106383e-05,
      "loss": 0.0689,
      "step": 120
    },
    {
      "epoch": 0.4432624113475177,
      "grad_norm": 0.2363295555114746,
      "learning_rate": 4.556737588652483e-05,
      "loss": 0.0644,
      "step": 125
    },
    {
      "epoch": 0.46099290780141844,
      "grad_norm": 2.4520492553710938,
      "learning_rate": 4.539007092198582e-05,
      "loss": 0.0489,
      "step": 130
    },
    {
      "epoch": 0.4787234042553192,
      "grad_norm": 1.8659327030181885,
      "learning_rate": 4.5212765957446815e-05,
      "loss": 0.0544,
      "step": 135
    },
    {
      "epoch": 0.49645390070921985,
      "grad_norm": 0.5137922167778015,
      "learning_rate": 4.50354609929078e-05,
      "loss": 0.0733,
      "step": 140
    },
    {
      "epoch": 0.5141843971631206,
      "grad_norm": 0.2559826076030731,
      "learning_rate": 4.48581560283688e-05,
      "loss": 0.0349,
      "step": 145
    },
    {
      "epoch": 0.5319148936170213,
      "grad_norm": 1.2222139835357666,
      "learning_rate": 4.468085106382979e-05,
      "loss": 0.061,
      "step": 150
    },
    {
      "epoch": 0.549645390070922,
      "grad_norm": 0.19817987084388733,
      "learning_rate": 4.4503546099290785e-05,
      "loss": 0.0548,
      "step": 155
    },
    {
      "epoch": 0.5673758865248227,
      "grad_norm": 0.6621349453926086,
      "learning_rate": 4.432624113475177e-05,
      "loss": 0.0409,
      "step": 160
    },
    {
      "epoch": 0.5851063829787234,
      "grad_norm": 0.8852491974830627,
      "learning_rate": 4.414893617021277e-05,
      "loss": 0.027,
      "step": 165
    },
    {
      "epoch": 0.6028368794326241,
      "grad_norm": 0.1130673736333847,
      "learning_rate": 4.3971631205673764e-05,
      "loss": 0.0271,
      "step": 170
    },
    {
      "epoch": 0.6205673758865248,
      "grad_norm": 2.4764528274536133,
      "learning_rate": 4.3794326241134755e-05,
      "loss": 0.0578,
      "step": 175
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 1.4849659204483032,
      "learning_rate": 4.3617021276595746e-05,
      "loss": 0.0425,
      "step": 180
    },
    {
      "epoch": 0.6560283687943262,
      "grad_norm": 0.5148060321807861,
      "learning_rate": 4.343971631205674e-05,
      "loss": 0.036,
      "step": 185
    },
    {
      "epoch": 0.6737588652482269,
      "grad_norm": 1.4764983654022217,
      "learning_rate": 4.3262411347517734e-05,
      "loss": 0.0815,
      "step": 190
    },
    {
      "epoch": 0.6914893617021277,
      "grad_norm": 0.46336182951927185,
      "learning_rate": 4.3085106382978725e-05,
      "loss": 0.0225,
      "step": 195
    },
    {
      "epoch": 0.7092198581560284,
      "grad_norm": 0.6195293664932251,
      "learning_rate": 4.2907801418439716e-05,
      "loss": 0.0228,
      "step": 200
    },
    {
      "epoch": 0.7269503546099291,
      "grad_norm": 0.27533644437789917,
      "learning_rate": 4.273049645390071e-05,
      "loss": 0.0709,
      "step": 205
    },
    {
      "epoch": 0.7446808510638298,
      "grad_norm": 0.6618344187736511,
      "learning_rate": 4.2553191489361704e-05,
      "loss": 0.0708,
      "step": 210
    },
    {
      "epoch": 0.7624113475177305,
      "grad_norm": 0.9032989740371704,
      "learning_rate": 4.2375886524822695e-05,
      "loss": 0.0396,
      "step": 215
    },
    {
      "epoch": 0.7801418439716312,
      "grad_norm": 2.292132616043091,
      "learning_rate": 4.219858156028369e-05,
      "loss": 0.0927,
      "step": 220
    },
    {
      "epoch": 0.7978723404255319,
      "grad_norm": 2.1669278144836426,
      "learning_rate": 4.2021276595744684e-05,
      "loss": 0.1068,
      "step": 225
    },
    {
      "epoch": 0.8156028368794326,
      "grad_norm": 1.218011498451233,
      "learning_rate": 4.1843971631205674e-05,
      "loss": 0.0444,
      "step": 230
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 1.6065305471420288,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.0443,
      "step": 235
    },
    {
      "epoch": 0.851063829787234,
      "grad_norm": 1.1235682964324951,
      "learning_rate": 4.148936170212766e-05,
      "loss": 0.0683,
      "step": 240
    },
    {
      "epoch": 0.8687943262411347,
      "grad_norm": 1.7434436082839966,
      "learning_rate": 4.1312056737588654e-05,
      "loss": 0.0857,
      "step": 245
    },
    {
      "epoch": 0.8865248226950354,
      "grad_norm": 0.5483965277671814,
      "learning_rate": 4.1134751773049644e-05,
      "loss": 0.0437,
      "step": 250
    },
    {
      "epoch": 0.9042553191489362,
      "grad_norm": 1.176393985748291,
      "learning_rate": 4.095744680851064e-05,
      "loss": 0.0771,
      "step": 255
    },
    {
      "epoch": 0.9219858156028369,
      "grad_norm": 1.270770788192749,
      "learning_rate": 4.078014184397163e-05,
      "loss": 0.0371,
      "step": 260
    },
    {
      "epoch": 0.9397163120567376,
      "grad_norm": 1.3304009437561035,
      "learning_rate": 4.0602836879432624e-05,
      "loss": 0.0815,
      "step": 265
    },
    {
      "epoch": 0.9574468085106383,
      "grad_norm": 0.5169529914855957,
      "learning_rate": 4.0425531914893614e-05,
      "loss": 0.0511,
      "step": 270
    },
    {
      "epoch": 0.975177304964539,
      "grad_norm": 0.4900430142879486,
      "learning_rate": 4.024822695035461e-05,
      "loss": 0.0304,
      "step": 275
    },
    {
      "epoch": 0.9929078014184397,
      "grad_norm": 1.8383599519729614,
      "learning_rate": 4.007092198581561e-05,
      "loss": 0.0449,
      "step": 280
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9957144968228167,
      "eval_loss": 0.018303971737623215,
      "eval_runtime": 99.5406,
      "eval_samples_per_second": 67.982,
      "eval_steps_per_second": 1.065,
      "step": 282
    },
    {
      "epoch": 1.0106382978723405,
      "grad_norm": 0.5246375203132629,
      "learning_rate": 3.9893617021276594e-05,
      "loss": 0.0246,
      "step": 285
    },
    {
      "epoch": 1.0283687943262412,
      "grad_norm": 1.0449841022491455,
      "learning_rate": 3.971631205673759e-05,
      "loss": 0.0212,
      "step": 290
    },
    {
      "epoch": 1.0460992907801419,
      "grad_norm": 0.26453474164009094,
      "learning_rate": 3.953900709219858e-05,
      "loss": 0.0358,
      "step": 295
    },
    {
      "epoch": 1.0638297872340425,
      "grad_norm": 0.6182503700256348,
      "learning_rate": 3.936170212765958e-05,
      "loss": 0.0387,
      "step": 300
    },
    {
      "epoch": 1.0815602836879432,
      "grad_norm": 0.46771693229675293,
      "learning_rate": 3.918439716312057e-05,
      "loss": 0.0285,
      "step": 305
    },
    {
      "epoch": 1.099290780141844,
      "grad_norm": 1.8450125455856323,
      "learning_rate": 3.900709219858156e-05,
      "loss": 0.0475,
      "step": 310
    },
    {
      "epoch": 1.1170212765957448,
      "grad_norm": 0.7594661712646484,
      "learning_rate": 3.882978723404255e-05,
      "loss": 0.0156,
      "step": 315
    },
    {
      "epoch": 1.1347517730496455,
      "grad_norm": 0.29846104979515076,
      "learning_rate": 3.865248226950355e-05,
      "loss": 0.0375,
      "step": 320
    },
    {
      "epoch": 1.1524822695035462,
      "grad_norm": 1.215070128440857,
      "learning_rate": 3.847517730496454e-05,
      "loss": 0.0626,
      "step": 325
    },
    {
      "epoch": 1.1702127659574468,
      "grad_norm": 0.27688533067703247,
      "learning_rate": 3.829787234042553e-05,
      "loss": 0.0327,
      "step": 330
    },
    {
      "epoch": 1.1879432624113475,
      "grad_norm": 1.9927847385406494,
      "learning_rate": 3.812056737588653e-05,
      "loss": 0.0557,
      "step": 335
    },
    {
      "epoch": 1.2056737588652482,
      "grad_norm": 4.12082576751709,
      "learning_rate": 3.794326241134752e-05,
      "loss": 0.0669,
      "step": 340
    },
    {
      "epoch": 1.2234042553191489,
      "grad_norm": 1.2266093492507935,
      "learning_rate": 3.776595744680852e-05,
      "loss": 0.0575,
      "step": 345
    },
    {
      "epoch": 1.2411347517730495,
      "grad_norm": 0.7239733934402466,
      "learning_rate": 3.75886524822695e-05,
      "loss": 0.0316,
      "step": 350
    },
    {
      "epoch": 1.2588652482269502,
      "grad_norm": 1.3613947629928589,
      "learning_rate": 3.74113475177305e-05,
      "loss": 0.0512,
      "step": 355
    },
    {
      "epoch": 1.2765957446808511,
      "grad_norm": 2.1938693523406982,
      "learning_rate": 3.723404255319149e-05,
      "loss": 0.0511,
      "step": 360
    },
    {
      "epoch": 1.2943262411347518,
      "grad_norm": 0.5384643077850342,
      "learning_rate": 3.705673758865249e-05,
      "loss": 0.0479,
      "step": 365
    },
    {
      "epoch": 1.3120567375886525,
      "grad_norm": 0.4367295801639557,
      "learning_rate": 3.687943262411347e-05,
      "loss": 0.0595,
      "step": 370
    },
    {
      "epoch": 1.3297872340425532,
      "grad_norm": 1.2368284463882446,
      "learning_rate": 3.670212765957447e-05,
      "loss": 0.0332,
      "step": 375
    },
    {
      "epoch": 1.3475177304964538,
      "grad_norm": 1.422599196434021,
      "learning_rate": 3.6524822695035466e-05,
      "loss": 0.0433,
      "step": 380
    },
    {
      "epoch": 1.3652482269503547,
      "grad_norm": 1.9355744123458862,
      "learning_rate": 3.634751773049646e-05,
      "loss": 0.0401,
      "step": 385
    },
    {
      "epoch": 1.3829787234042552,
      "grad_norm": 0.11844053119421005,
      "learning_rate": 3.617021276595745e-05,
      "loss": 0.0308,
      "step": 390
    },
    {
      "epoch": 1.400709219858156,
      "grad_norm": 1.5698013305664062,
      "learning_rate": 3.599290780141844e-05,
      "loss": 0.024,
      "step": 395
    },
    {
      "epoch": 1.4184397163120568,
      "grad_norm": 0.24424462020397186,
      "learning_rate": 3.5815602836879437e-05,
      "loss": 0.0284,
      "step": 400
    },
    {
      "epoch": 1.4361702127659575,
      "grad_norm": 0.8079735040664673,
      "learning_rate": 3.563829787234043e-05,
      "loss": 0.0289,
      "step": 405
    },
    {
      "epoch": 1.4539007092198581,
      "grad_norm": 0.5713645815849304,
      "learning_rate": 3.546099290780142e-05,
      "loss": 0.0489,
      "step": 410
    },
    {
      "epoch": 1.4716312056737588,
      "grad_norm": 1.418959140777588,
      "learning_rate": 3.528368794326241e-05,
      "loss": 0.0494,
      "step": 415
    },
    {
      "epoch": 1.4893617021276595,
      "grad_norm": 1.3852524757385254,
      "learning_rate": 3.5106382978723407e-05,
      "loss": 0.0416,
      "step": 420
    },
    {
      "epoch": 1.5070921985815602,
      "grad_norm": 1.9135431051254272,
      "learning_rate": 3.49290780141844e-05,
      "loss": 0.0244,
      "step": 425
    },
    {
      "epoch": 1.524822695035461,
      "grad_norm": 0.7146717309951782,
      "learning_rate": 3.4751773049645395e-05,
      "loss": 0.0512,
      "step": 430
    },
    {
      "epoch": 1.5425531914893615,
      "grad_norm": 0.1944599747657776,
      "learning_rate": 3.4574468085106386e-05,
      "loss": 0.0191,
      "step": 435
    },
    {
      "epoch": 1.5602836879432624,
      "grad_norm": 0.32298266887664795,
      "learning_rate": 3.4397163120567377e-05,
      "loss": 0.0221,
      "step": 440
    },
    {
      "epoch": 1.5780141843971631,
      "grad_norm": 0.7677968740463257,
      "learning_rate": 3.4219858156028374e-05,
      "loss": 0.023,
      "step": 445
    },
    {
      "epoch": 1.5957446808510638,
      "grad_norm": 0.19664119184017181,
      "learning_rate": 3.4042553191489365e-05,
      "loss": 0.0402,
      "step": 450
    },
    {
      "epoch": 1.6134751773049647,
      "grad_norm": 0.8602023124694824,
      "learning_rate": 3.3865248226950356e-05,
      "loss": 0.0343,
      "step": 455
    },
    {
      "epoch": 1.6312056737588652,
      "grad_norm": 2.0019989013671875,
      "learning_rate": 3.3687943262411347e-05,
      "loss": 0.0521,
      "step": 460
    },
    {
      "epoch": 1.648936170212766,
      "grad_norm": 0.42928043007850647,
      "learning_rate": 3.3510638297872344e-05,
      "loss": 0.0434,
      "step": 465
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.31418949365615845,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0125,
      "step": 470
    },
    {
      "epoch": 1.6843971631205674,
      "grad_norm": 0.6069945096969604,
      "learning_rate": 3.3156028368794326e-05,
      "loss": 0.055,
      "step": 475
    },
    {
      "epoch": 1.702127659574468,
      "grad_norm": 2.3644161224365234,
      "learning_rate": 3.2978723404255317e-05,
      "loss": 0.0652,
      "step": 480
    },
    {
      "epoch": 1.7198581560283688,
      "grad_norm": 0.054500628262758255,
      "learning_rate": 3.2801418439716314e-05,
      "loss": 0.0398,
      "step": 485
    },
    {
      "epoch": 1.7375886524822695,
      "grad_norm": 2.016355037689209,
      "learning_rate": 3.262411347517731e-05,
      "loss": 0.0788,
      "step": 490
    },
    {
      "epoch": 1.7553191489361701,
      "grad_norm": 0.6187741160392761,
      "learning_rate": 3.2446808510638296e-05,
      "loss": 0.0579,
      "step": 495
    },
    {
      "epoch": 1.773049645390071,
      "grad_norm": 1.820343255996704,
      "learning_rate": 3.226950354609929e-05,
      "loss": 0.0632,
      "step": 500
    },
    {
      "epoch": 1.7907801418439715,
      "grad_norm": 0.4283825755119324,
      "learning_rate": 3.2092198581560284e-05,
      "loss": 0.0449,
      "step": 505
    },
    {
      "epoch": 1.8085106382978724,
      "grad_norm": 1.2107046842575073,
      "learning_rate": 3.191489361702128e-05,
      "loss": 0.0293,
      "step": 510
    },
    {
      "epoch": 1.826241134751773,
      "grad_norm": 0.751215398311615,
      "learning_rate": 3.173758865248227e-05,
      "loss": 0.0436,
      "step": 515
    },
    {
      "epoch": 1.8439716312056738,
      "grad_norm": 0.8796170353889465,
      "learning_rate": 3.156028368794326e-05,
      "loss": 0.052,
      "step": 520
    },
    {
      "epoch": 1.8617021276595744,
      "grad_norm": 0.14203386008739471,
      "learning_rate": 3.1382978723404254e-05,
      "loss": 0.0103,
      "step": 525
    },
    {
      "epoch": 1.8794326241134751,
      "grad_norm": 1.044580340385437,
      "learning_rate": 3.120567375886525e-05,
      "loss": 0.0269,
      "step": 530
    },
    {
      "epoch": 1.897163120567376,
      "grad_norm": 2.8254494667053223,
      "learning_rate": 3.102836879432624e-05,
      "loss": 0.0662,
      "step": 535
    },
    {
      "epoch": 1.9148936170212765,
      "grad_norm": 0.6130526065826416,
      "learning_rate": 3.085106382978723e-05,
      "loss": 0.0277,
      "step": 540
    },
    {
      "epoch": 1.9326241134751774,
      "grad_norm": 0.12123987823724747,
      "learning_rate": 3.067375886524823e-05,
      "loss": 0.035,
      "step": 545
    },
    {
      "epoch": 1.950354609929078,
      "grad_norm": 0.5822201371192932,
      "learning_rate": 3.0496453900709222e-05,
      "loss": 0.0294,
      "step": 550
    },
    {
      "epoch": 1.9680851063829787,
      "grad_norm": 0.28792017698287964,
      "learning_rate": 3.0319148936170216e-05,
      "loss": 0.0543,
      "step": 555
    },
    {
      "epoch": 1.9858156028368794,
      "grad_norm": 1.7587871551513672,
      "learning_rate": 3.0141843971631207e-05,
      "loss": 0.04,
      "step": 560
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9980789123688488,
      "eval_loss": 0.010071107186377048,
      "eval_runtime": 99.7696,
      "eval_samples_per_second": 67.826,
      "eval_steps_per_second": 1.062,
      "step": 564
    },
    {
      "epoch": 2.00354609929078,
      "grad_norm": 0.3479292690753937,
      "learning_rate": 2.99645390070922e-05,
      "loss": 0.0726,
      "step": 565
    },
    {
      "epoch": 2.021276595744681,
      "grad_norm": 0.10121797770261765,
      "learning_rate": 2.9787234042553192e-05,
      "loss": 0.0185,
      "step": 570
    },
    {
      "epoch": 2.0390070921985815,
      "grad_norm": 0.9174487590789795,
      "learning_rate": 2.9609929078014186e-05,
      "loss": 0.0637,
      "step": 575
    },
    {
      "epoch": 2.0567375886524824,
      "grad_norm": 0.4720062017440796,
      "learning_rate": 2.9432624113475177e-05,
      "loss": 0.0246,
      "step": 580
    },
    {
      "epoch": 2.074468085106383,
      "grad_norm": 0.3442896008491516,
      "learning_rate": 2.925531914893617e-05,
      "loss": 0.0159,
      "step": 585
    },
    {
      "epoch": 2.0921985815602837,
      "grad_norm": 1.1636303663253784,
      "learning_rate": 2.9078014184397162e-05,
      "loss": 0.018,
      "step": 590
    },
    {
      "epoch": 2.1099290780141846,
      "grad_norm": 0.7616550326347351,
      "learning_rate": 2.8900709219858156e-05,
      "loss": 0.0392,
      "step": 595
    },
    {
      "epoch": 2.127659574468085,
      "grad_norm": 0.17084839940071106,
      "learning_rate": 2.8723404255319154e-05,
      "loss": 0.0172,
      "step": 600
    },
    {
      "epoch": 2.145390070921986,
      "grad_norm": 0.4418025612831116,
      "learning_rate": 2.8546099290780144e-05,
      "loss": 0.0287,
      "step": 605
    },
    {
      "epoch": 2.1631205673758864,
      "grad_norm": 0.05007610097527504,
      "learning_rate": 2.836879432624114e-05,
      "loss": 0.0311,
      "step": 610
    },
    {
      "epoch": 2.1808510638297873,
      "grad_norm": 2.390089273452759,
      "learning_rate": 2.819148936170213e-05,
      "loss": 0.0393,
      "step": 615
    },
    {
      "epoch": 2.198581560283688,
      "grad_norm": 1.8016951084136963,
      "learning_rate": 2.8014184397163124e-05,
      "loss": 0.0518,
      "step": 620
    },
    {
      "epoch": 2.2163120567375887,
      "grad_norm": 0.39676979184150696,
      "learning_rate": 2.7836879432624114e-05,
      "loss": 0.0319,
      "step": 625
    },
    {
      "epoch": 2.2340425531914896,
      "grad_norm": 0.2608576714992523,
      "learning_rate": 2.765957446808511e-05,
      "loss": 0.0148,
      "step": 630
    },
    {
      "epoch": 2.25177304964539,
      "grad_norm": 1.5164140462875366,
      "learning_rate": 2.74822695035461e-05,
      "loss": 0.04,
      "step": 635
    },
    {
      "epoch": 2.269503546099291,
      "grad_norm": 0.3103680908679962,
      "learning_rate": 2.7304964539007094e-05,
      "loss": 0.0234,
      "step": 640
    },
    {
      "epoch": 2.2872340425531914,
      "grad_norm": 0.4016372561454773,
      "learning_rate": 2.7127659574468084e-05,
      "loss": 0.0171,
      "step": 645
    },
    {
      "epoch": 2.3049645390070923,
      "grad_norm": 0.5997406840324402,
      "learning_rate": 2.695035460992908e-05,
      "loss": 0.0387,
      "step": 650
    },
    {
      "epoch": 2.3226950354609928,
      "grad_norm": 0.8444051742553711,
      "learning_rate": 2.6773049645390076e-05,
      "loss": 0.0539,
      "step": 655
    },
    {
      "epoch": 2.3404255319148937,
      "grad_norm": 0.108007051050663,
      "learning_rate": 2.6595744680851064e-05,
      "loss": 0.0315,
      "step": 660
    },
    {
      "epoch": 2.3581560283687946,
      "grad_norm": 0.21093517541885376,
      "learning_rate": 2.641843971631206e-05,
      "loss": 0.0175,
      "step": 665
    },
    {
      "epoch": 2.375886524822695,
      "grad_norm": 0.4845324754714966,
      "learning_rate": 2.624113475177305e-05,
      "loss": 0.0185,
      "step": 670
    },
    {
      "epoch": 2.393617021276596,
      "grad_norm": 2.6781203746795654,
      "learning_rate": 2.6063829787234046e-05,
      "loss": 0.0562,
      "step": 675
    },
    {
      "epoch": 2.4113475177304964,
      "grad_norm": 0.301878958940506,
      "learning_rate": 2.5886524822695034e-05,
      "loss": 0.0298,
      "step": 680
    },
    {
      "epoch": 2.4290780141843973,
      "grad_norm": 0.9099262952804565,
      "learning_rate": 2.570921985815603e-05,
      "loss": 0.0421,
      "step": 685
    },
    {
      "epoch": 2.4468085106382977,
      "grad_norm": 0.11847200244665146,
      "learning_rate": 2.5531914893617022e-05,
      "loss": 0.0174,
      "step": 690
    },
    {
      "epoch": 2.4645390070921986,
      "grad_norm": 0.12315840274095535,
      "learning_rate": 2.5354609929078016e-05,
      "loss": 0.0129,
      "step": 695
    },
    {
      "epoch": 2.482269503546099,
      "grad_norm": 0.955663800239563,
      "learning_rate": 2.5177304964539007e-05,
      "loss": 0.0239,
      "step": 700
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.13527634739875793,
      "learning_rate": 2.5e-05,
      "loss": 0.0217,
      "step": 705
    },
    {
      "epoch": 2.5177304964539005,
      "grad_norm": 0.24646982550621033,
      "learning_rate": 2.4822695035460995e-05,
      "loss": 0.0388,
      "step": 710
    },
    {
      "epoch": 2.5354609929078014,
      "grad_norm": 0.09928718209266663,
      "learning_rate": 2.4645390070921986e-05,
      "loss": 0.0117,
      "step": 715
    },
    {
      "epoch": 2.5531914893617023,
      "grad_norm": 2.1757736206054688,
      "learning_rate": 2.446808510638298e-05,
      "loss": 0.0205,
      "step": 720
    },
    {
      "epoch": 2.5709219858156027,
      "grad_norm": 0.4307613968849182,
      "learning_rate": 2.429078014184397e-05,
      "loss": 0.0495,
      "step": 725
    },
    {
      "epoch": 2.5886524822695036,
      "grad_norm": 0.07029764354228973,
      "learning_rate": 2.4113475177304965e-05,
      "loss": 0.0568,
      "step": 730
    },
    {
      "epoch": 2.6063829787234045,
      "grad_norm": 1.3018437623977661,
      "learning_rate": 2.393617021276596e-05,
      "loss": 0.0389,
      "step": 735
    },
    {
      "epoch": 2.624113475177305,
      "grad_norm": 0.2572399973869324,
      "learning_rate": 2.3758865248226954e-05,
      "loss": 0.0392,
      "step": 740
    },
    {
      "epoch": 2.6418439716312054,
      "grad_norm": 1.5049391984939575,
      "learning_rate": 2.3581560283687945e-05,
      "loss": 0.0273,
      "step": 745
    },
    {
      "epoch": 2.6595744680851063,
      "grad_norm": 1.8658090829849243,
      "learning_rate": 2.340425531914894e-05,
      "loss": 0.0503,
      "step": 750
    },
    {
      "epoch": 2.6773049645390072,
      "grad_norm": 1.4984087944030762,
      "learning_rate": 2.322695035460993e-05,
      "loss": 0.0256,
      "step": 755
    },
    {
      "epoch": 2.6950354609929077,
      "grad_norm": 1.7155365943908691,
      "learning_rate": 2.3049645390070924e-05,
      "loss": 0.0371,
      "step": 760
    },
    {
      "epoch": 2.7127659574468086,
      "grad_norm": 2.706315517425537,
      "learning_rate": 2.2872340425531915e-05,
      "loss": 0.0289,
      "step": 765
    },
    {
      "epoch": 2.7304964539007095,
      "grad_norm": 0.5071002244949341,
      "learning_rate": 2.269503546099291e-05,
      "loss": 0.0144,
      "step": 770
    },
    {
      "epoch": 2.74822695035461,
      "grad_norm": 2.5877788066864014,
      "learning_rate": 2.25177304964539e-05,
      "loss": 0.0462,
      "step": 775
    },
    {
      "epoch": 2.7659574468085104,
      "grad_norm": 0.13342903554439545,
      "learning_rate": 2.2340425531914894e-05,
      "loss": 0.0311,
      "step": 780
    },
    {
      "epoch": 2.7836879432624113,
      "grad_norm": 0.5681129693984985,
      "learning_rate": 2.2163120567375885e-05,
      "loss": 0.0496,
      "step": 785
    },
    {
      "epoch": 2.801418439716312,
      "grad_norm": 1.3924553394317627,
      "learning_rate": 2.1985815602836882e-05,
      "loss": 0.0302,
      "step": 790
    },
    {
      "epoch": 2.8191489361702127,
      "grad_norm": 0.9623014330863953,
      "learning_rate": 2.1808510638297873e-05,
      "loss": 0.0166,
      "step": 795
    },
    {
      "epoch": 2.8368794326241136,
      "grad_norm": 1.2445993423461914,
      "learning_rate": 2.1631205673758867e-05,
      "loss": 0.0237,
      "step": 800
    },
    {
      "epoch": 2.854609929078014,
      "grad_norm": 0.8575385212898254,
      "learning_rate": 2.1453900709219858e-05,
      "loss": 0.0753,
      "step": 805
    },
    {
      "epoch": 2.872340425531915,
      "grad_norm": 1.026553988456726,
      "learning_rate": 2.1276595744680852e-05,
      "loss": 0.0266,
      "step": 810
    },
    {
      "epoch": 2.8900709219858154,
      "grad_norm": 1.974367380142212,
      "learning_rate": 2.1099290780141846e-05,
      "loss": 0.0302,
      "step": 815
    },
    {
      "epoch": 2.9078014184397163,
      "grad_norm": 1.4900615215301514,
      "learning_rate": 2.0921985815602837e-05,
      "loss": 0.0499,
      "step": 820
    },
    {
      "epoch": 2.925531914893617,
      "grad_norm": 2.0212316513061523,
      "learning_rate": 2.074468085106383e-05,
      "loss": 0.0411,
      "step": 825
    },
    {
      "epoch": 2.9432624113475176,
      "grad_norm": 1.2716954946517944,
      "learning_rate": 2.0567375886524822e-05,
      "loss": 0.0538,
      "step": 830
    },
    {
      "epoch": 2.9609929078014185,
      "grad_norm": 1.8260325193405151,
      "learning_rate": 2.0390070921985816e-05,
      "loss": 0.0405,
      "step": 835
    },
    {
      "epoch": 2.978723404255319,
      "grad_norm": 0.23081009089946747,
      "learning_rate": 2.0212765957446807e-05,
      "loss": 0.0255,
      "step": 840
    },
    {
      "epoch": 2.99645390070922,
      "grad_norm": 1.7825428247451782,
      "learning_rate": 2.0035460992907805e-05,
      "loss": 0.0303,
      "step": 845
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9985222402837298,
      "eval_loss": 0.008113781921565533,
      "eval_runtime": 100.8534,
      "eval_samples_per_second": 67.097,
      "eval_steps_per_second": 1.051,
      "step": 846
    },
    {
      "epoch": 3.0141843971631204,
      "grad_norm": 1.1453741788864136,
      "learning_rate": 1.9858156028368796e-05,
      "loss": 0.0207,
      "step": 850
    },
    {
      "epoch": 3.0319148936170213,
      "grad_norm": 0.37254324555397034,
      "learning_rate": 1.968085106382979e-05,
      "loss": 0.0222,
      "step": 855
    },
    {
      "epoch": 3.049645390070922,
      "grad_norm": 2.6893270015716553,
      "learning_rate": 1.950354609929078e-05,
      "loss": 0.0606,
      "step": 860
    },
    {
      "epoch": 3.0673758865248226,
      "grad_norm": 2.4312515258789062,
      "learning_rate": 1.9326241134751775e-05,
      "loss": 0.0251,
      "step": 865
    },
    {
      "epoch": 3.0851063829787235,
      "grad_norm": 1.2454279661178589,
      "learning_rate": 1.9148936170212766e-05,
      "loss": 0.0173,
      "step": 870
    },
    {
      "epoch": 3.102836879432624,
      "grad_norm": 0.6530167460441589,
      "learning_rate": 1.897163120567376e-05,
      "loss": 0.0156,
      "step": 875
    },
    {
      "epoch": 3.120567375886525,
      "grad_norm": 0.12946569919586182,
      "learning_rate": 1.879432624113475e-05,
      "loss": 0.0225,
      "step": 880
    },
    {
      "epoch": 3.1382978723404253,
      "grad_norm": 0.357861191034317,
      "learning_rate": 1.8617021276595745e-05,
      "loss": 0.0121,
      "step": 885
    },
    {
      "epoch": 3.1560283687943262,
      "grad_norm": 0.1260479986667633,
      "learning_rate": 1.8439716312056736e-05,
      "loss": 0.0496,
      "step": 890
    },
    {
      "epoch": 3.173758865248227,
      "grad_norm": 0.7033432722091675,
      "learning_rate": 1.8262411347517733e-05,
      "loss": 0.0313,
      "step": 895
    },
    {
      "epoch": 3.1914893617021276,
      "grad_norm": 1.0441181659698486,
      "learning_rate": 1.8085106382978724e-05,
      "loss": 0.0143,
      "step": 900
    },
    {
      "epoch": 3.2092198581560285,
      "grad_norm": 0.032710377126932144,
      "learning_rate": 1.7907801418439718e-05,
      "loss": 0.0091,
      "step": 905
    },
    {
      "epoch": 3.226950354609929,
      "grad_norm": 1.0562541484832764,
      "learning_rate": 1.773049645390071e-05,
      "loss": 0.0642,
      "step": 910
    },
    {
      "epoch": 3.24468085106383,
      "grad_norm": 0.16099953651428223,
      "learning_rate": 1.7553191489361703e-05,
      "loss": 0.0195,
      "step": 915
    },
    {
      "epoch": 3.2624113475177303,
      "grad_norm": 0.868291437625885,
      "learning_rate": 1.7375886524822697e-05,
      "loss": 0.0272,
      "step": 920
    },
    {
      "epoch": 3.280141843971631,
      "grad_norm": 0.9113226532936096,
      "learning_rate": 1.7198581560283688e-05,
      "loss": 0.0317,
      "step": 925
    },
    {
      "epoch": 3.297872340425532,
      "grad_norm": 1.1665312051773071,
      "learning_rate": 1.7021276595744682e-05,
      "loss": 0.0301,
      "step": 930
    },
    {
      "epoch": 3.3156028368794326,
      "grad_norm": 1.0826573371887207,
      "learning_rate": 1.6843971631205673e-05,
      "loss": 0.0129,
      "step": 935
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 1.054902195930481,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.0346,
      "step": 940
    },
    {
      "epoch": 3.351063829787234,
      "grad_norm": 2.4785568714141846,
      "learning_rate": 1.6489361702127658e-05,
      "loss": 0.0419,
      "step": 945
    },
    {
      "epoch": 3.368794326241135,
      "grad_norm": 0.26840367913246155,
      "learning_rate": 1.6312056737588656e-05,
      "loss": 0.0184,
      "step": 950
    },
    {
      "epoch": 3.3865248226950353,
      "grad_norm": 1.8092659711837769,
      "learning_rate": 1.6134751773049647e-05,
      "loss": 0.0374,
      "step": 955
    },
    {
      "epoch": 3.404255319148936,
      "grad_norm": 0.6401882767677307,
      "learning_rate": 1.595744680851064e-05,
      "loss": 0.0521,
      "step": 960
    },
    {
      "epoch": 3.421985815602837,
      "grad_norm": 0.3769932985305786,
      "learning_rate": 1.578014184397163e-05,
      "loss": 0.0176,
      "step": 965
    },
    {
      "epoch": 3.4397163120567376,
      "grad_norm": 1.343386173248291,
      "learning_rate": 1.5602836879432626e-05,
      "loss": 0.0187,
      "step": 970
    },
    {
      "epoch": 3.4574468085106385,
      "grad_norm": 0.7425516247749329,
      "learning_rate": 1.5425531914893617e-05,
      "loss": 0.0334,
      "step": 975
    },
    {
      "epoch": 3.475177304964539,
      "grad_norm": 0.5523159503936768,
      "learning_rate": 1.5248226950354611e-05,
      "loss": 0.0233,
      "step": 980
    },
    {
      "epoch": 3.49290780141844,
      "grad_norm": 0.3636256754398346,
      "learning_rate": 1.5070921985815603e-05,
      "loss": 0.0269,
      "step": 985
    },
    {
      "epoch": 3.5106382978723403,
      "grad_norm": 0.7461684346199036,
      "learning_rate": 1.4893617021276596e-05,
      "loss": 0.0313,
      "step": 990
    },
    {
      "epoch": 3.528368794326241,
      "grad_norm": 0.6355846524238586,
      "learning_rate": 1.4716312056737588e-05,
      "loss": 0.0216,
      "step": 995
    },
    {
      "epoch": 3.546099290780142,
      "grad_norm": 0.046597789973020554,
      "learning_rate": 1.4539007092198581e-05,
      "loss": 0.0247,
      "step": 1000
    },
    {
      "epoch": 3.5638297872340425,
      "grad_norm": 0.7104584574699402,
      "learning_rate": 1.4361702127659577e-05,
      "loss": 0.0375,
      "step": 1005
    },
    {
      "epoch": 3.581560283687943,
      "grad_norm": 0.7681288719177246,
      "learning_rate": 1.418439716312057e-05,
      "loss": 0.013,
      "step": 1010
    },
    {
      "epoch": 3.599290780141844,
      "grad_norm": 2.825113534927368,
      "learning_rate": 1.4007092198581562e-05,
      "loss": 0.0138,
      "step": 1015
    },
    {
      "epoch": 3.617021276595745,
      "grad_norm": 0.15184448659420013,
      "learning_rate": 1.3829787234042554e-05,
      "loss": 0.0278,
      "step": 1020
    },
    {
      "epoch": 3.6347517730496453,
      "grad_norm": 1.885351300239563,
      "learning_rate": 1.3652482269503547e-05,
      "loss": 0.047,
      "step": 1025
    },
    {
      "epoch": 3.652482269503546,
      "grad_norm": 1.271122932434082,
      "learning_rate": 1.347517730496454e-05,
      "loss": 0.0309,
      "step": 1030
    },
    {
      "epoch": 3.670212765957447,
      "grad_norm": 0.02030951902270317,
      "learning_rate": 1.3297872340425532e-05,
      "loss": 0.0072,
      "step": 1035
    },
    {
      "epoch": 3.6879432624113475,
      "grad_norm": 0.046280715614557266,
      "learning_rate": 1.3120567375886524e-05,
      "loss": 0.0219,
      "step": 1040
    },
    {
      "epoch": 3.705673758865248,
      "grad_norm": 2.204589605331421,
      "learning_rate": 1.2943262411347517e-05,
      "loss": 0.0402,
      "step": 1045
    },
    {
      "epoch": 3.723404255319149,
      "grad_norm": 0.6903723478317261,
      "learning_rate": 1.2765957446808511e-05,
      "loss": 0.0319,
      "step": 1050
    },
    {
      "epoch": 3.7411347517730498,
      "grad_norm": 0.8682162165641785,
      "learning_rate": 1.2588652482269504e-05,
      "loss": 0.0296,
      "step": 1055
    },
    {
      "epoch": 3.7588652482269502,
      "grad_norm": 0.9223999381065369,
      "learning_rate": 1.2411347517730498e-05,
      "loss": 0.0365,
      "step": 1060
    },
    {
      "epoch": 3.776595744680851,
      "grad_norm": 0.5410795211791992,
      "learning_rate": 1.223404255319149e-05,
      "loss": 0.0121,
      "step": 1065
    },
    {
      "epoch": 3.794326241134752,
      "grad_norm": 0.17826128005981445,
      "learning_rate": 1.2056737588652483e-05,
      "loss": 0.0231,
      "step": 1070
    },
    {
      "epoch": 3.8120567375886525,
      "grad_norm": 0.47573089599609375,
      "learning_rate": 1.1879432624113477e-05,
      "loss": 0.0266,
      "step": 1075
    },
    {
      "epoch": 3.829787234042553,
      "grad_norm": 0.8923744559288025,
      "learning_rate": 1.170212765957447e-05,
      "loss": 0.0345,
      "step": 1080
    },
    {
      "epoch": 3.847517730496454,
      "grad_norm": 1.696198582649231,
      "learning_rate": 1.1524822695035462e-05,
      "loss": 0.0451,
      "step": 1085
    },
    {
      "epoch": 3.8652482269503547,
      "grad_norm": 0.4135494828224182,
      "learning_rate": 1.1347517730496454e-05,
      "loss": 0.0248,
      "step": 1090
    },
    {
      "epoch": 3.882978723404255,
      "grad_norm": 0.4472277760505676,
      "learning_rate": 1.1170212765957447e-05,
      "loss": 0.0125,
      "step": 1095
    },
    {
      "epoch": 3.900709219858156,
      "grad_norm": 3.610241174697876,
      "learning_rate": 1.0992907801418441e-05,
      "loss": 0.0263,
      "step": 1100
    },
    {
      "epoch": 3.918439716312057,
      "grad_norm": 0.06732177734375,
      "learning_rate": 1.0815602836879434e-05,
      "loss": 0.0153,
      "step": 1105
    },
    {
      "epoch": 3.9361702127659575,
      "grad_norm": 0.8051948547363281,
      "learning_rate": 1.0638297872340426e-05,
      "loss": 0.0217,
      "step": 1110
    },
    {
      "epoch": 3.953900709219858,
      "grad_norm": 0.24708445370197296,
      "learning_rate": 1.0460992907801419e-05,
      "loss": 0.0432,
      "step": 1115
    },
    {
      "epoch": 3.971631205673759,
      "grad_norm": 0.0366179421544075,
      "learning_rate": 1.0283687943262411e-05,
      "loss": 0.0065,
      "step": 1120
    },
    {
      "epoch": 3.9893617021276597,
      "grad_norm": 1.3547054529190063,
      "learning_rate": 1.0106382978723404e-05,
      "loss": 0.0489,
      "step": 1125
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9988177922269839,
      "eval_loss": 0.0067949132062494755,
      "eval_runtime": 99.24,
      "eval_samples_per_second": 68.188,
      "eval_steps_per_second": 1.068,
      "step": 1128
    },
    {
      "epoch": 4.00709219858156,
      "grad_norm": 0.2567829191684723,
      "learning_rate": 9.929078014184398e-06,
      "loss": 0.0132,
      "step": 1130
    },
    {
      "epoch": 4.024822695035461,
      "grad_norm": 0.8916745781898499,
      "learning_rate": 9.75177304964539e-06,
      "loss": 0.0217,
      "step": 1135
    },
    {
      "epoch": 4.042553191489362,
      "grad_norm": 0.13239595293998718,
      "learning_rate": 9.574468085106383e-06,
      "loss": 0.0171,
      "step": 1140
    },
    {
      "epoch": 4.060283687943262,
      "grad_norm": 0.2502301335334778,
      "learning_rate": 9.397163120567375e-06,
      "loss": 0.006,
      "step": 1145
    },
    {
      "epoch": 4.078014184397163,
      "grad_norm": 0.3878186047077179,
      "learning_rate": 9.219858156028368e-06,
      "loss": 0.0184,
      "step": 1150
    },
    {
      "epoch": 4.095744680851064,
      "grad_norm": 1.005374789237976,
      "learning_rate": 9.042553191489362e-06,
      "loss": 0.0633,
      "step": 1155
    },
    {
      "epoch": 4.113475177304965,
      "grad_norm": 1.9449095726013184,
      "learning_rate": 8.865248226950355e-06,
      "loss": 0.027,
      "step": 1160
    },
    {
      "epoch": 4.131205673758865,
      "grad_norm": 0.04838603362441063,
      "learning_rate": 8.687943262411349e-06,
      "loss": 0.0077,
      "step": 1165
    },
    {
      "epoch": 4.148936170212766,
      "grad_norm": 0.07912664860486984,
      "learning_rate": 8.510638297872341e-06,
      "loss": 0.0075,
      "step": 1170
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 0.24843692779541016,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.0134,
      "step": 1175
    },
    {
      "epoch": 4.184397163120567,
      "grad_norm": 0.7264091968536377,
      "learning_rate": 8.156028368794328e-06,
      "loss": 0.0367,
      "step": 1180
    },
    {
      "epoch": 4.202127659574468,
      "grad_norm": 1.442780613899231,
      "learning_rate": 7.97872340425532e-06,
      "loss": 0.0143,
      "step": 1185
    },
    {
      "epoch": 4.219858156028369,
      "grad_norm": 0.4630628228187561,
      "learning_rate": 7.801418439716313e-06,
      "loss": 0.0506,
      "step": 1190
    },
    {
      "epoch": 4.23758865248227,
      "grad_norm": 1.1415939331054688,
      "learning_rate": 7.6241134751773054e-06,
      "loss": 0.0266,
      "step": 1195
    },
    {
      "epoch": 4.25531914893617,
      "grad_norm": 0.8882013559341431,
      "learning_rate": 7.446808510638298e-06,
      "loss": 0.0159,
      "step": 1200
    },
    {
      "epoch": 4.273049645390071,
      "grad_norm": 0.12462817132472992,
      "learning_rate": 7.2695035460992904e-06,
      "loss": 0.0209,
      "step": 1205
    },
    {
      "epoch": 4.290780141843972,
      "grad_norm": 1.5256775617599487,
      "learning_rate": 7.092198581560285e-06,
      "loss": 0.0241,
      "step": 1210
    },
    {
      "epoch": 4.308510638297872,
      "grad_norm": 3.1360552310943604,
      "learning_rate": 6.914893617021277e-06,
      "loss": 0.0298,
      "step": 1215
    },
    {
      "epoch": 4.326241134751773,
      "grad_norm": 0.7372686266899109,
      "learning_rate": 6.73758865248227e-06,
      "loss": 0.0193,
      "step": 1220
    },
    {
      "epoch": 4.343971631205674,
      "grad_norm": 1.058393955230713,
      "learning_rate": 6.560283687943262e-06,
      "loss": 0.0293,
      "step": 1225
    },
    {
      "epoch": 4.361702127659575,
      "grad_norm": 0.09787444770336151,
      "learning_rate": 6.3829787234042555e-06,
      "loss": 0.0098,
      "step": 1230
    },
    {
      "epoch": 4.379432624113475,
      "grad_norm": 0.4723568856716156,
      "learning_rate": 6.205673758865249e-06,
      "loss": 0.0248,
      "step": 1235
    },
    {
      "epoch": 4.397163120567376,
      "grad_norm": 0.3988282382488251,
      "learning_rate": 6.028368794326241e-06,
      "loss": 0.0123,
      "step": 1240
    },
    {
      "epoch": 4.414893617021277,
      "grad_norm": 0.080900177359581,
      "learning_rate": 5.851063829787235e-06,
      "loss": 0.0361,
      "step": 1245
    },
    {
      "epoch": 4.432624113475177,
      "grad_norm": 0.28941839933395386,
      "learning_rate": 5.673758865248227e-06,
      "loss": 0.0173,
      "step": 1250
    },
    {
      "epoch": 4.450354609929078,
      "grad_norm": 0.16654981672763824,
      "learning_rate": 5.4964539007092206e-06,
      "loss": 0.0374,
      "step": 1255
    },
    {
      "epoch": 4.468085106382979,
      "grad_norm": 0.5223755240440369,
      "learning_rate": 5.319148936170213e-06,
      "loss": 0.0077,
      "step": 1260
    },
    {
      "epoch": 4.48581560283688,
      "grad_norm": 0.7028219103813171,
      "learning_rate": 5.1418439716312056e-06,
      "loss": 0.0299,
      "step": 1265
    },
    {
      "epoch": 4.50354609929078,
      "grad_norm": 0.3213287889957428,
      "learning_rate": 4.964539007092199e-06,
      "loss": 0.0569,
      "step": 1270
    },
    {
      "epoch": 4.5212765957446805,
      "grad_norm": 0.26597318053245544,
      "learning_rate": 4.787234042553191e-06,
      "loss": 0.0122,
      "step": 1275
    },
    {
      "epoch": 4.539007092198582,
      "grad_norm": 1.5103939771652222,
      "learning_rate": 4.609929078014184e-06,
      "loss": 0.021,
      "step": 1280
    },
    {
      "epoch": 4.556737588652482,
      "grad_norm": 0.4471294581890106,
      "learning_rate": 4.432624113475177e-06,
      "loss": 0.0361,
      "step": 1285
    },
    {
      "epoch": 4.574468085106383,
      "grad_norm": 0.21643757820129395,
      "learning_rate": 4.255319148936171e-06,
      "loss": 0.0294,
      "step": 1290
    },
    {
      "epoch": 4.592198581560284,
      "grad_norm": 2.008852958679199,
      "learning_rate": 4.078014184397164e-06,
      "loss": 0.0301,
      "step": 1295
    },
    {
      "epoch": 4.609929078014185,
      "grad_norm": 1.9454808235168457,
      "learning_rate": 3.9007092198581565e-06,
      "loss": 0.0468,
      "step": 1300
    },
    {
      "epoch": 4.627659574468085,
      "grad_norm": 0.3638037443161011,
      "learning_rate": 3.723404255319149e-06,
      "loss": 0.009,
      "step": 1305
    },
    {
      "epoch": 4.6453900709219855,
      "grad_norm": 0.5267730951309204,
      "learning_rate": 3.5460992907801423e-06,
      "loss": 0.0106,
      "step": 1310
    },
    {
      "epoch": 4.663120567375887,
      "grad_norm": 0.6058443188667297,
      "learning_rate": 3.368794326241135e-06,
      "loss": 0.0408,
      "step": 1315
    },
    {
      "epoch": 4.680851063829787,
      "grad_norm": 0.7102274894714355,
      "learning_rate": 3.1914893617021277e-06,
      "loss": 0.0177,
      "step": 1320
    },
    {
      "epoch": 4.698581560283688,
      "grad_norm": 0.03302247077226639,
      "learning_rate": 3.0141843971631207e-06,
      "loss": 0.0263,
      "step": 1325
    },
    {
      "epoch": 4.716312056737589,
      "grad_norm": 0.29536840319633484,
      "learning_rate": 2.8368794326241136e-06,
      "loss": 0.0176,
      "step": 1330
    },
    {
      "epoch": 4.73404255319149,
      "grad_norm": 0.303792268037796,
      "learning_rate": 2.6595744680851065e-06,
      "loss": 0.0149,
      "step": 1335
    },
    {
      "epoch": 4.75177304964539,
      "grad_norm": 1.4943726062774658,
      "learning_rate": 2.4822695035460995e-06,
      "loss": 0.0324,
      "step": 1340
    },
    {
      "epoch": 4.7695035460992905,
      "grad_norm": 0.2641482949256897,
      "learning_rate": 2.304964539007092e-06,
      "loss": 0.0258,
      "step": 1345
    },
    {
      "epoch": 4.787234042553192,
      "grad_norm": 1.3771953582763672,
      "learning_rate": 2.1276595744680853e-06,
      "loss": 0.0325,
      "step": 1350
    },
    {
      "epoch": 4.804964539007092,
      "grad_norm": 2.062373638153076,
      "learning_rate": 1.9503546099290782e-06,
      "loss": 0.0152,
      "step": 1355
    },
    {
      "epoch": 4.822695035460993,
      "grad_norm": 0.5422120094299316,
      "learning_rate": 1.7730496453900712e-06,
      "loss": 0.0387,
      "step": 1360
    },
    {
      "epoch": 4.840425531914894,
      "grad_norm": 0.02120036445558071,
      "learning_rate": 1.5957446808510639e-06,
      "loss": 0.0122,
      "step": 1365
    },
    {
      "epoch": 4.858156028368795,
      "grad_norm": 0.2176242172718048,
      "learning_rate": 1.4184397163120568e-06,
      "loss": 0.0205,
      "step": 1370
    },
    {
      "epoch": 4.875886524822695,
      "grad_norm": 2.1586310863494873,
      "learning_rate": 1.2411347517730497e-06,
      "loss": 0.0245,
      "step": 1375
    },
    {
      "epoch": 4.8936170212765955,
      "grad_norm": 0.3032481074333191,
      "learning_rate": 1.0638297872340427e-06,
      "loss": 0.0317,
      "step": 1380
    },
    {
      "epoch": 4.911347517730497,
      "grad_norm": 0.1574692726135254,
      "learning_rate": 8.865248226950356e-07,
      "loss": 0.0126,
      "step": 1385
    },
    {
      "epoch": 4.929078014184397,
      "grad_norm": 1.4468116760253906,
      "learning_rate": 7.092198581560284e-07,
      "loss": 0.0166,
      "step": 1390
    },
    {
      "epoch": 4.946808510638298,
      "grad_norm": 0.5175489783287048,
      "learning_rate": 5.319148936170213e-07,
      "loss": 0.0101,
      "step": 1395
    },
    {
      "epoch": 4.964539007092198,
      "grad_norm": 0.04873587563633919,
      "learning_rate": 3.546099290780142e-07,
      "loss": 0.0167,
      "step": 1400
    },
    {
      "epoch": 4.9822695035460995,
      "grad_norm": 3.589585304260254,
      "learning_rate": 1.773049645390071e-07,
      "loss": 0.0284,
      "step": 1405
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.1343780905008316,
      "learning_rate": 0.0,
      "loss": 0.0284,
      "step": 1410
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9989655681986109,
      "eval_loss": 0.006617749575525522,
      "eval_runtime": 99.2975,
      "eval_samples_per_second": 68.149,
      "eval_steps_per_second": 1.067,
      "step": 1410
    },
    {
      "epoch": 5.0,
      "step": 1410,
      "total_flos": 6.98086588443052e+18,
      "train_loss": 0.042465827911299596,
      "train_runtime": 3748.8226,
      "train_samples_per_second": 24.03,
      "train_steps_per_second": 0.376
    }
  ],
  "logging_steps": 5,
  "max_steps": 1410,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.98086588443052e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}