{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8313847752663029,
  "eval_steps": 500,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 33.651363372802734,
      "learning_rate": 5e-06,
      "loss": 11.4242,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 30.245094299316406,
      "learning_rate": 1e-05,
      "loss": 10.7984,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 20.595664978027344,
      "learning_rate": 1.5e-05,
      "loss": 9.6784,
      "step": 15
    },
    {
      "epoch": 0.0,
      "grad_norm": 18.10934829711914,
      "learning_rate": 2e-05,
      "loss": 8.6549,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.764612197875977,
      "learning_rate": 2.5e-05,
      "loss": 8.0527,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.847819805145264,
      "learning_rate": 3e-05,
      "loss": 7.8584,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.7772746086120605,
      "learning_rate": 3.5e-05,
      "loss": 7.6997,
      "step": 35
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.401936054229736,
      "learning_rate": 4e-05,
      "loss": 7.5018,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.469693660736084,
      "learning_rate": 4.5e-05,
      "loss": 7.2811,
      "step": 45
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.165266036987305,
      "learning_rate": 5e-05,
      "loss": 7.047,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.453402519226074,
      "learning_rate": 5.500000000000001e-05,
      "loss": 6.8175,
      "step": 55
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.7616753578186035,
      "learning_rate": 6e-05,
      "loss": 6.6509,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.575658798217773,
      "learning_rate": 6.500000000000001e-05,
      "loss": 6.4582,
      "step": 65
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.6263885498046875,
      "learning_rate": 7e-05,
      "loss": 6.2432,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 7.119388580322266,
      "learning_rate": 7.500000000000001e-05,
      "loss": 6.0652,
      "step": 75
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.671682834625244,
      "learning_rate": 8e-05,
      "loss": 5.8913,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.266168117523193,
      "learning_rate": 8.5e-05,
      "loss": 5.7924,
      "step": 85
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.612922668457031,
      "learning_rate": 9e-05,
      "loss": 5.6806,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.099239349365234,
      "learning_rate": 9.5e-05,
      "loss": 5.5746,
      "step": 95
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.313576698303223,
      "learning_rate": 0.0001,
      "loss": 5.4404,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.374126434326172,
      "learning_rate": 9.999972205865686e-05,
      "loss": 5.3661,
      "step": 105
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.098925590515137,
      "learning_rate": 9.999888823771751e-05,
      "loss": 5.2502,
      "step": 110
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.676019668579102,
      "learning_rate": 9.999749854645204e-05,
      "loss": 5.2054,
      "step": 115
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.790042400360107,
      "learning_rate": 9.99955530003106e-05,
      "loss": 5.1922,
      "step": 120
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.1792755126953125,
      "learning_rate": 9.99930516209231e-05,
      "loss": 5.0362,
      "step": 125
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.925230026245117,
      "learning_rate": 9.998999443609897e-05,
      "loss": 4.8993,
      "step": 130
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.8192741870880127,
      "learning_rate": 9.998638147982696e-05,
      "loss": 4.8546,
      "step": 135
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.455966472625732,
      "learning_rate": 9.998221279227467e-05,
      "loss": 4.9245,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.5653443336486816,
      "learning_rate": 9.997748841978812e-05,
      "loss": 4.794,
      "step": 145
    },
    {
      "epoch": 0.03,
      "grad_norm": 7.422428131103516,
      "learning_rate": 9.997220841489122e-05,
      "loss": 4.6852,
      "step": 150
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.473473072052002,
      "learning_rate": 9.996637283628528e-05,
      "loss": 4.7194,
      "step": 155
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.07218074798584,
      "learning_rate": 9.995998174884821e-05,
      "loss": 4.5755,
      "step": 160
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.4391303062438965,
      "learning_rate": 9.995303522363394e-05,
      "loss": 4.6125,
      "step": 165
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.6765010356903076,
      "learning_rate": 9.99455333378715e-05,
      "loss": 4.5488,
      "step": 170
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.9314346313476562,
      "learning_rate": 9.993747617496428e-05,
      "loss": 4.5681,
      "step": 175
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.8918819427490234,
      "learning_rate": 9.9928863824489e-05,
      "loss": 4.5552,
      "step": 180
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.8918302059173584,
      "learning_rate": 9.99196963821948e-05,
      "loss": 4.5485,
      "step": 185
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.52258563041687,
      "learning_rate": 9.990997395000217e-05,
      "loss": 4.3666,
      "step": 190
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.107878684997559,
      "learning_rate": 9.989969663600169e-05,
      "loss": 4.4963,
      "step": 195
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.0174808502197266,
      "learning_rate": 9.9888864554453e-05,
      "loss": 4.4471,
      "step": 200
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.8166937828063965,
      "learning_rate": 9.987747782578342e-05,
      "loss": 4.3326,
      "step": 205
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.0829238891601562,
      "learning_rate": 9.986553657658668e-05,
      "loss": 4.3306,
      "step": 210
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.6593828201293945,
      "learning_rate": 9.985304093962145e-05,
      "loss": 4.2864,
      "step": 215
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.727086067199707,
      "learning_rate": 9.983999105380988e-05,
      "loss": 4.2608,
      "step": 220
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.664433240890503,
      "learning_rate": 9.982638706423608e-05,
      "loss": 4.3274,
      "step": 225
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.2005085945129395,
      "learning_rate": 9.98122291221445e-05,
      "loss": 4.29,
      "step": 230
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.2342798709869385,
      "learning_rate": 9.979751738493826e-05,
      "loss": 4.1954,
      "step": 235
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.773253917694092,
      "learning_rate": 9.978225201617732e-05,
      "loss": 4.2651,
      "step": 240
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.033334255218506,
      "learning_rate": 9.976643318557678e-05,
      "loss": 4.1532,
      "step": 245
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.3765692710876465,
      "learning_rate": 9.975006106900495e-05,
      "loss": 4.1574,
      "step": 250
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.4881410598754883,
      "learning_rate": 9.973313584848132e-05,
      "loss": 4.151,
      "step": 255
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.6022536754608154,
      "learning_rate": 9.971565771217464e-05,
      "loss": 4.2039,
      "step": 260
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.6714437007904053,
      "learning_rate": 9.969762685440076e-05,
      "loss": 4.1488,
      "step": 265
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.723792552947998,
      "learning_rate": 9.967904347562054e-05,
      "loss": 4.0716,
      "step": 270
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.407099723815918,
      "learning_rate": 9.965990778243755e-05,
      "loss": 4.0903,
      "step": 275
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.4060311317443848,
      "learning_rate": 9.964021998759577e-05,
      "loss": 4.0282,
      "step": 280
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.4491236209869385,
      "learning_rate": 9.961998030997733e-05,
      "loss": 4.0502,
      "step": 285
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.302164316177368,
      "learning_rate": 9.95991889745999e-05,
      "loss": 4.0588,
      "step": 290
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.6794662475585938,
      "learning_rate": 9.957784621261441e-05,
      "loss": 4.033,
      "step": 295
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.262129306793213,
      "learning_rate": 9.955595226130226e-05,
      "loss": 3.9834,
      "step": 300
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.1817383766174316,
      "learning_rate": 9.953350736407282e-05,
      "loss": 3.9371,
      "step": 305
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.4953038692474365,
      "learning_rate": 9.951051177046069e-05,
      "loss": 4.0102,
      "step": 310
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.575176477432251,
      "learning_rate": 9.948696573612292e-05,
      "loss": 4.0396,
      "step": 315
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.594761610031128,
      "learning_rate": 9.946286952283618e-05,
      "loss": 3.9192,
      "step": 320
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.27431058883667,
      "learning_rate": 9.943822339849381e-05,
      "loss": 4.0117,
      "step": 325
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.3896944522857666,
      "learning_rate": 9.941302763710288e-05,
      "loss": 3.9799,
      "step": 330
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.9870325326919556,
      "learning_rate": 9.938728251878116e-05,
      "loss": 3.8637,
      "step": 335
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.1930291652679443,
      "learning_rate": 9.936098832975393e-05,
      "loss": 3.7962,
      "step": 340
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.4338083267211914,
      "learning_rate": 9.933414536235091e-05,
      "loss": 3.9224,
      "step": 345
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.0370845794677734,
      "learning_rate": 9.93067539150029e-05,
      "loss": 3.8586,
      "step": 350
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.9598731994628906,
      "learning_rate": 9.927881429223853e-05,
      "loss": 3.8631,
      "step": 355
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.95335853099823,
      "learning_rate": 9.925032680468085e-05,
      "loss": 3.7463,
      "step": 360
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.5965161323547363,
      "learning_rate": 9.922129176904388e-05,
      "loss": 3.8511,
      "step": 365
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.8875285387039185,
      "learning_rate": 9.919170950812911e-05,
      "loss": 3.7626,
      "step": 370
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.4620866775512695,
      "learning_rate": 9.916158035082184e-05,
      "loss": 3.7785,
      "step": 375
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.9820605516433716,
      "learning_rate": 9.913090463208763e-05,
      "loss": 3.8284,
      "step": 380
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.7589184045791626,
      "learning_rate": 9.90996826929685e-05,
      "loss": 3.7592,
      "step": 385
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.8934298753738403,
      "learning_rate": 9.906791488057916e-05,
      "loss": 3.7244,
      "step": 390
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.0055394172668457,
      "learning_rate": 9.903560154810313e-05,
      "loss": 3.7211,
      "step": 395
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.9094047546386719,
      "learning_rate": 9.900274305478887e-05,
      "loss": 3.7496,
      "step": 400
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.7309576272964478,
      "learning_rate": 9.896933976594572e-05,
      "loss": 3.7648,
      "step": 405
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.0965123176574707,
      "learning_rate": 9.893539205293989e-05,
      "loss": 3.6743,
      "step": 410
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.9884498119354248,
      "learning_rate": 9.890090029319028e-05,
      "loss": 3.6636,
      "step": 415
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.9006081819534302,
      "learning_rate": 9.886586487016433e-05,
      "loss": 3.6416,
      "step": 420
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.8273619413375854,
      "learning_rate": 9.883028617337378e-05,
      "loss": 3.6083,
      "step": 425
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.13328218460083,
      "learning_rate": 9.879416459837022e-05,
      "loss": 3.6626,
      "step": 430
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.8524399995803833,
      "learning_rate": 9.875750054674082e-05,
      "loss": 3.6508,
      "step": 435
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.8650786876678467,
      "learning_rate": 9.872029442610382e-05,
      "loss": 3.6299,
      "step": 440
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.6650742292404175,
      "learning_rate": 9.8682546650104e-05,
      "loss": 3.5859,
      "step": 445
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.5603471994400024,
      "learning_rate": 9.864425763840802e-05,
      "loss": 3.6087,
      "step": 450
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.5650691986083984,
      "learning_rate": 9.860542781669988e-05,
      "loss": 3.6397,
      "step": 455
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.7188957929611206,
      "learning_rate": 9.85660576166761e-05,
      "loss": 3.6476,
      "step": 460
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.6306991577148438,
      "learning_rate": 9.852614747604093e-05,
      "loss": 3.6055,
      "step": 465
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.1953701972961426,
      "learning_rate": 9.848569783850145e-05,
      "loss": 3.5826,
      "step": 470
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.1616299152374268,
      "learning_rate": 9.844470915376278e-05,
      "loss": 3.6174,
      "step": 475
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.7074434757232666,
      "learning_rate": 9.840318187752292e-05,
      "loss": 3.4243,
      "step": 480
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.6088547706604004,
      "learning_rate": 9.836111647146771e-05,
      "loss": 3.4957,
      "step": 485
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.8437894582748413,
      "learning_rate": 9.831851340326577e-05,
      "loss": 3.5407,
      "step": 490
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.6059656143188477,
      "learning_rate": 9.82753731465633e-05,
      "loss": 3.4262,
      "step": 495
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.987560510635376,
      "learning_rate": 9.823169618097871e-05,
      "loss": 3.5446,
      "step": 500
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.5124210119247437,
      "learning_rate": 9.81874829920974e-05,
      "loss": 3.4832,
      "step": 505
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5314714908599854,
      "learning_rate": 9.814273407146623e-05,
      "loss": 3.4532,
      "step": 510
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.6071770191192627,
      "learning_rate": 9.809744991658829e-05,
      "loss": 3.4115,
      "step": 515
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.7526746988296509,
      "learning_rate": 9.805163103091708e-05,
      "loss": 3.5478,
      "step": 520
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.570514440536499,
      "learning_rate": 9.800527792385112e-05,
      "loss": 3.4841,
      "step": 525
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.8370064496994019,
      "learning_rate": 9.79583911107282e-05,
      "loss": 3.4776,
      "step": 530
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.590645432472229,
      "learning_rate": 9.791097111281968e-05,
      "loss": 3.399,
      "step": 535
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5733847618103027,
      "learning_rate": 9.786301845732467e-05,
      "loss": 3.3824,
      "step": 540
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.6162751913070679,
      "learning_rate": 9.781453367736418e-05,
      "loss": 3.3776,
      "step": 545
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5934056043624878,
      "learning_rate": 9.776551731197524e-05,
      "loss": 3.342,
      "step": 550
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.523496150970459,
      "learning_rate": 9.771596990610478e-05,
      "loss": 3.3919,
      "step": 555
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.703430414199829,
      "learning_rate": 9.766589201060372e-05,
      "loss": 3.3063,
      "step": 560
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5598702430725098,
      "learning_rate": 9.761528418222077e-05,
      "loss": 3.3171,
      "step": 565
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.7453784942626953,
      "learning_rate": 9.756414698359624e-05,
      "loss": 3.379,
      "step": 570
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.4527804851531982,
      "learning_rate": 9.75124809832558e-05,
      "loss": 3.397,
      "step": 575
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5606558322906494,
      "learning_rate": 9.746028675560413e-05,
      "loss": 3.3644,
      "step": 580
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.6888071298599243,
      "learning_rate": 9.740756488091861e-05,
      "loss": 3.2505,
      "step": 585
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5528062582015991,
      "learning_rate": 9.735431594534277e-05,
      "loss": 3.3397,
      "step": 590
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.6364246606826782,
      "learning_rate": 9.730054054087983e-05,
      "loss": 3.3461,
      "step": 595
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.4486875534057617,
      "learning_rate": 9.724623926538612e-05,
      "loss": 3.2633,
      "step": 600
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.5004665851593018,
      "learning_rate": 9.719141272256443e-05,
      "loss": 3.2701,
      "step": 605
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.489261269569397,
      "learning_rate": 9.713606152195726e-05,
      "loss": 3.267,
      "step": 610
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.4302242994308472,
      "learning_rate": 9.708018627894011e-05,
      "loss": 3.3144,
      "step": 615
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.3415112495422363,
      "learning_rate": 9.702378761471456e-05,
      "loss": 3.2477,
      "step": 620
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.5592623949050903,
      "learning_rate": 9.696686615630146e-05,
      "loss": 3.2573,
      "step": 625
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.563595175743103,
      "learning_rate": 9.690942253653385e-05,
      "loss": 3.2603,
      "step": 630
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.4028782844543457,
      "learning_rate": 9.685145739405002e-05,
      "loss": 3.3102,
      "step": 635
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.8662109375,
      "learning_rate": 9.679297137328634e-05,
      "loss": 3.2572,
      "step": 640
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.7463732957839966,
      "learning_rate": 9.673396512447013e-05,
      "loss": 3.2952,
      "step": 645
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4323670864105225,
      "learning_rate": 9.667443930361247e-05,
      "loss": 3.2638,
      "step": 650
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3809984922409058,
      "learning_rate": 9.661439457250076e-05,
      "loss": 3.2625,
      "step": 655
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.5374635457992554,
      "learning_rate": 9.655383159869158e-05,
      "loss": 3.29,
      "step": 660
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4163179397583008,
      "learning_rate": 9.649275105550309e-05,
      "loss": 3.1998,
      "step": 665
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4354526996612549,
      "learning_rate": 9.643115362200762e-05,
      "loss": 3.2943,
      "step": 670
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4390583038330078,
      "learning_rate": 9.636903998302409e-05,
      "loss": 3.2955,
      "step": 675
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4252122640609741,
      "learning_rate": 9.630641082911045e-05,
      "loss": 3.2063,
      "step": 680
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3900178670883179,
      "learning_rate": 9.624326685655593e-05,
      "loss": 3.2456,
      "step": 685
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.5450825691223145,
      "learning_rate": 9.617960876737337e-05,
      "loss": 3.2088,
      "step": 690
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4761152267456055,
      "learning_rate": 9.611543726929134e-05,
      "loss": 3.1838,
      "step": 695
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.3841191530227661,
      "learning_rate": 9.605075307574635e-05,
      "loss": 3.147,
      "step": 700
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.4704729318618774,
      "learning_rate": 9.598555690587487e-05,
      "loss": 3.1682,
      "step": 705
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2939262390136719,
      "learning_rate": 9.591984948450532e-05,
      "loss": 3.1114,
      "step": 710
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6519654989242554,
      "learning_rate": 9.585363154215008e-05,
      "loss": 3.2082,
      "step": 715
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2991009950637817,
      "learning_rate": 9.578690381499728e-05,
      "loss": 3.1081,
      "step": 720
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2765209674835205,
      "learning_rate": 9.571966704490271e-05,
      "loss": 3.1507,
      "step": 725
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2271565198898315,
      "learning_rate": 9.565192197938148e-05,
      "loss": 3.202,
      "step": 730
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.3048691749572754,
      "learning_rate": 9.558366937159977e-05,
      "loss": 3.1312,
      "step": 735
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.8889031410217285,
      "learning_rate": 9.551490998036646e-05,
      "loss": 3.1106,
      "step": 740
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.3917661905288696,
      "learning_rate": 9.544564457012463e-05,
      "loss": 3.1339,
      "step": 745
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.3188519477844238,
      "learning_rate": 9.537587391094314e-05,
      "loss": 3.1738,
      "step": 750
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.403839111328125,
      "learning_rate": 9.5305598778508e-05,
      "loss": 3.1675,
      "step": 755
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.42656409740448,
      "learning_rate": 9.52348199541138e-05,
      "loss": 3.1208,
      "step": 760
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.6658726930618286,
      "learning_rate": 9.516353822465504e-05,
      "loss": 3.0526,
      "step": 765
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.266996145248413,
      "learning_rate": 9.509175438261726e-05,
      "loss": 3.0473,
      "step": 770
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.3163378238677979,
      "learning_rate": 9.501946922606838e-05,
      "loss": 3.0853,
      "step": 775
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.3675296306610107,
      "learning_rate": 9.494668355864973e-05,
      "loss": 3.0912,
      "step": 780
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.2658545970916748,
      "learning_rate": 9.487339818956716e-05,
      "loss": 3.1622,
      "step": 785
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.3492169380187988,
      "learning_rate": 9.479961393358203e-05,
      "loss": 3.1041,
      "step": 790
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3128236532211304,
      "learning_rate": 9.472533161100215e-05,
      "loss": 3.0637,
      "step": 795
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3049265146255493,
      "learning_rate": 9.465055204767265e-05,
      "loss": 3.0836,
      "step": 800
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2383729219436646,
      "learning_rate": 9.457527607496685e-05,
      "loss": 3.0421,
      "step": 805
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2715247869491577,
      "learning_rate": 9.44995045297769e-05,
      "loss": 3.0051,
      "step": 810
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.276710867881775,
      "learning_rate": 9.442323825450464e-05,
      "loss": 3.0368,
      "step": 815
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3114001750946045,
      "learning_rate": 9.43464780970521e-05,
      "loss": 3.06,
      "step": 820
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2944623231887817,
      "learning_rate": 9.426922491081212e-05,
      "loss": 3.084,
      "step": 825
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2431820631027222,
      "learning_rate": 9.419147955465888e-05,
      "loss": 3.0199,
      "step": 830
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3521535396575928,
      "learning_rate": 9.411324289293832e-05,
      "loss": 3.0438,
      "step": 835
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3010791540145874,
      "learning_rate": 9.403451579545859e-05,
      "loss": 3.0329,
      "step": 840
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.430463194847107,
      "learning_rate": 9.395529913748025e-05,
      "loss": 3.003,
      "step": 845
    },
    {
      "epoch": 0.18,
      "grad_norm": 19.969221115112305,
      "learning_rate": 9.387559379970672e-05,
      "loss": 3.0045,
      "step": 850
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2505444288253784,
      "learning_rate": 9.379540066827431e-05,
      "loss": 3.023,
      "step": 855
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.25497305393219,
      "learning_rate": 9.371472063474248e-05,
      "loss": 3.0441,
      "step": 860
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1969456672668457,
      "learning_rate": 9.363355459608394e-05,
      "loss": 2.9503,
      "step": 865
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1699817180633545,
      "learning_rate": 9.355190345467457e-05,
      "loss": 2.9636,
      "step": 870
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.318147897720337,
      "learning_rate": 9.346976811828352e-05,
      "loss": 2.9839,
      "step": 875
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2377957105636597,
      "learning_rate": 9.338714950006297e-05,
      "loss": 3.065,
      "step": 880
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.4408152103424072,
      "learning_rate": 9.330404851853817e-05,
      "loss": 2.9526,
      "step": 885
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2080788612365723,
      "learning_rate": 9.3220466097597e-05,
      "loss": 3.0015,
      "step": 890
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2468167543411255,
      "learning_rate": 9.313640316647991e-05,
      "loss": 3.0341,
      "step": 895
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2537471055984497,
      "learning_rate": 9.305186065976945e-05,
      "loss": 3.0205,
      "step": 900
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.183099389076233,
      "learning_rate": 9.296683951737993e-05,
      "loss": 3.0408,
      "step": 905
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.1875871419906616,
      "learning_rate": 9.288134068454697e-05,
      "loss": 2.9291,
      "step": 910
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2846369743347168,
      "learning_rate": 9.2795365111817e-05,
      "loss": 2.9141,
      "step": 915
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.204579472541809,
      "learning_rate": 9.270891375503665e-05,
      "loss": 2.9797,
      "step": 920
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.171580195426941,
      "learning_rate": 9.262198757534218e-05,
      "loss": 2.9523,
      "step": 925
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.176835298538208,
      "learning_rate": 9.253458753914874e-05,
      "loss": 2.9909,
      "step": 930
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2722320556640625,
      "learning_rate": 9.244671461813969e-05,
      "loss": 2.9957,
      "step": 935
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.291190505027771,
      "learning_rate": 9.235836978925572e-05,
      "loss": 3.0211,
      "step": 940
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.1778265237808228,
      "learning_rate": 9.226955403468406e-05,
      "loss": 2.9882,
      "step": 945
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.176194667816162,
      "learning_rate": 9.21802683418475e-05,
      "loss": 2.938,
      "step": 950
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.1951394081115723,
      "learning_rate": 9.209051370339347e-05,
      "loss": 2.961,
      "step": 955
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.1604056358337402,
      "learning_rate": 9.200029111718295e-05,
      "loss": 2.8819,
      "step": 960
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2189394235610962,
      "learning_rate": 9.190960158627941e-05,
      "loss": 2.9453,
      "step": 965
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.1493133306503296,
      "learning_rate": 9.181844611893766e-05,
      "loss": 2.9281,
      "step": 970
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.230989933013916,
      "learning_rate": 9.172682572859261e-05,
      "loss": 2.9844,
      "step": 975
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.1456234455108643,
      "learning_rate": 9.163474143384806e-05,
      "loss": 3.028,
      "step": 980
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.1728698015213013,
      "learning_rate": 9.154219425846528e-05,
      "loss": 2.9739,
      "step": 985
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.1567491292953491,
      "learning_rate": 9.144918523135175e-05,
      "loss": 2.857,
      "step": 990
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.1249184608459473,
      "learning_rate": 9.13557153865496e-05,
      "loss": 2.9362,
      "step": 995
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.1881022453308105,
      "learning_rate": 9.12617857632242e-05,
      "loss": 2.9454,
      "step": 1000
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.1303149461746216,
      "learning_rate": 9.116739740565259e-05,
      "loss": 2.8987,
      "step": 1005
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.099196434020996,
      "learning_rate": 9.107255136321184e-05,
      "loss": 2.8858,
      "step": 1010
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.1457173824310303,
      "learning_rate": 9.09772486903674e-05,
      "loss": 2.871,
      "step": 1015
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.095299482345581,
      "learning_rate": 9.08814904466614e-05,
      "loss": 2.8584,
      "step": 1020
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.2056212425231934,
      "learning_rate": 9.078527769670085e-05,
      "loss": 2.9021,
      "step": 1025
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.2899084091186523,
      "learning_rate": 9.068861151014575e-05,
      "loss": 2.9313,
      "step": 1030
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.143584132194519,
      "learning_rate": 9.05914929616973e-05,
      "loss": 2.8577,
      "step": 1035
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.1004834175109863,
      "learning_rate": 9.04939231310859e-05,
      "loss": 2.9285,
      "step": 1040
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.25986909866333,
      "learning_rate": 9.039590310305914e-05,
      "loss": 2.8859,
      "step": 1045
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3628671169281006,
      "learning_rate": 9.029743396736974e-05,
      "loss": 2.8753,
      "step": 1050
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0711766481399536,
      "learning_rate": 9.019851681876348e-05,
      "loss": 2.9071,
      "step": 1055
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0614064931869507,
      "learning_rate": 9.009915275696693e-05,
      "loss": 2.8429,
      "step": 1060
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.2664223909378052,
      "learning_rate": 8.999934288667534e-05,
      "loss": 2.8943,
      "step": 1065
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.1698609590530396,
      "learning_rate": 8.989908831754028e-05,
      "loss": 2.8691,
      "step": 1070
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.1064506769180298,
      "learning_rate": 8.979839016415735e-05,
      "loss": 2.8696,
      "step": 1075
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.049668788909912,
      "learning_rate": 8.969724954605373e-05,
      "loss": 2.8282,
      "step": 1080
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.2249704599380493,
      "learning_rate": 8.959566758767581e-05,
      "loss": 2.7954,
      "step": 1085
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0685900449752808,
      "learning_rate": 8.949364541837661e-05,
      "loss": 2.8621,
      "step": 1090
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.3482073545455933,
      "learning_rate": 8.939118417240329e-05,
      "loss": 2.8579,
      "step": 1095
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.239876389503479,
      "learning_rate": 8.92882849888845e-05,
      "loss": 2.812,
      "step": 1100
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.229067087173462,
      "learning_rate": 8.918494901181773e-05,
      "loss": 2.8575,
      "step": 1105
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.3079419136047363,
      "learning_rate": 8.908117739005659e-05,
      "loss": 2.848,
      "step": 1110
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0963923931121826,
      "learning_rate": 8.897697127729805e-05,
      "loss": 2.8002,
      "step": 1115
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.166383147239685,
      "learning_rate": 8.887233183206957e-05,
      "loss": 2.8857,
      "step": 1120
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.1164528131484985,
      "learning_rate": 8.876726021771627e-05,
      "loss": 2.8646,
      "step": 1125
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.123231291770935,
      "learning_rate": 8.866175760238798e-05,
      "loss": 2.8596,
      "step": 1130
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.088853359222412,
      "learning_rate": 8.855582515902625e-05,
      "loss": 2.8162,
      "step": 1135
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.1311272382736206,
      "learning_rate": 8.844946406535131e-05,
      "loss": 2.8188,
      "step": 1140
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.2540462017059326,
      "learning_rate": 8.834267550384893e-05,
      "loss": 2.836,
      "step": 1145
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.156281590461731,
      "learning_rate": 8.823546066175741e-05,
      "loss": 2.8382,
      "step": 1150
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.1692214012145996,
      "learning_rate": 8.81278207310542e-05,
      "loss": 2.8032,
      "step": 1155
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.1173639297485352,
      "learning_rate": 8.801975690844278e-05,
      "loss": 2.7956,
      "step": 1160
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0640108585357666,
      "learning_rate": 8.791127039533934e-05,
      "loss": 2.8234,
      "step": 1165
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.5127233266830444,
      "learning_rate": 8.780236239785935e-05,
      "loss": 2.8631,
      "step": 1170
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0999369621276855,
      "learning_rate": 8.76930341268042e-05,
      "loss": 2.8476,
      "step": 1175
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0751299858093262,
      "learning_rate": 8.758328679764776e-05,
      "loss": 2.7966,
      "step": 1180
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.1676278114318848,
      "learning_rate": 8.747312163052284e-05,
      "loss": 2.8598,
      "step": 1185
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0812705755233765,
      "learning_rate": 8.736253985020761e-05,
      "loss": 2.8554,
      "step": 1190
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0541287660598755,
      "learning_rate": 8.725154268611203e-05,
      "loss": 2.8371,
      "step": 1195
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0639691352844238,
      "learning_rate": 8.714013137226411e-05,
      "loss": 2.7449,
      "step": 1200
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0529950857162476,
      "learning_rate": 8.702830714729628e-05,
      "loss": 2.8088,
      "step": 1205
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.1315858364105225,
      "learning_rate": 8.691607125443153e-05,
      "loss": 2.8446,
      "step": 1210
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0378223657608032,
      "learning_rate": 8.680342494146967e-05,
      "loss": 2.8384,
      "step": 1215
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0306965112686157,
      "learning_rate": 8.66903694607734e-05,
      "loss": 2.847,
      "step": 1220
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0947153568267822,
      "learning_rate": 8.65769060692544e-05,
      "loss": 2.8501,
      "step": 1225
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1506305932998657,
      "learning_rate": 8.646303602835936e-05,
      "loss": 2.8494,
      "step": 1230
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.056742548942566,
      "learning_rate": 8.634876060405597e-05,
      "loss": 2.8271,
      "step": 1235
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2626043558120728,
      "learning_rate": 8.623408106681884e-05,
      "loss": 2.7732,
      "step": 1240
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2176762819290161,
      "learning_rate": 8.611899869161535e-05,
      "loss": 2.8197,
      "step": 1245
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.0657603740692139,
      "learning_rate": 8.600351475789147e-05,
      "loss": 2.7657,
      "step": 1250
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.0771147012710571,
      "learning_rate": 8.588763054955764e-05,
      "loss": 2.8071,
      "step": 1255
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.0495407581329346,
      "learning_rate": 8.57713473549743e-05,
      "loss": 2.7559,
      "step": 1260
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.0748529434204102,
      "learning_rate": 8.565466646693778e-05,
      "loss": 2.7885,
      "step": 1265
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1235424280166626,
      "learning_rate": 8.553758918266578e-05,
      "loss": 2.8258,
      "step": 1270
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.035912036895752,
      "learning_rate": 8.5420116803783e-05,
      "loss": 2.7846,
      "step": 1275
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.0870811939239502,
      "learning_rate": 8.530225063630668e-05,
      "loss": 2.8605,
      "step": 1280
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.0321590900421143,
      "learning_rate": 8.518399199063205e-05,
      "loss": 2.7339,
      "step": 1285
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.1152360439300537,
      "learning_rate": 8.50653421815178e-05,
      "loss": 2.8021,
      "step": 1290
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.0527422428131104,
      "learning_rate": 8.494630252807138e-05,
      "loss": 2.8167,
      "step": 1295
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.065908432006836,
      "learning_rate": 8.482687435373449e-05,
      "loss": 2.7533,
      "step": 1300
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.055224895477295,
      "learning_rate": 8.470705898626817e-05,
      "loss": 2.746,
      "step": 1305
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.0720412731170654,
      "learning_rate": 8.458685775773822e-05,
      "loss": 2.7908,
      "step": 1310
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.1304291486740112,
      "learning_rate": 8.446627200450025e-05,
      "loss": 2.7551,
      "step": 1315
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.0969698429107666,
      "learning_rate": 8.434530306718493e-05,
      "loss": 2.755,
      "step": 1320
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0863304138183594,
      "learning_rate": 8.4223952290683e-05,
      "loss": 2.8039,
      "step": 1325
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.048060655593872,
      "learning_rate": 8.41022210241304e-05,
      "loss": 2.7631,
      "step": 1330
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.062782645225525,
      "learning_rate": 8.398011062089316e-05,
      "loss": 2.8196,
      "step": 1335
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.044792890548706,
      "learning_rate": 8.385762243855249e-05,
      "loss": 2.7331,
      "step": 1340
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0595911741256714,
      "learning_rate": 8.373475783888958e-05,
      "loss": 2.7563,
      "step": 1345
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.037611961364746,
      "learning_rate": 8.36115181878705e-05,
      "loss": 2.7296,
      "step": 1350
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0513596534729004,
      "learning_rate": 8.348790485563101e-05,
      "loss": 2.7668,
      "step": 1355
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0393058061599731,
      "learning_rate": 8.336391921646134e-05,
      "loss": 2.7692,
      "step": 1360
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0555886030197144,
      "learning_rate": 8.323956264879089e-05,
      "loss": 2.7927,
      "step": 1365
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.018600583076477,
      "learning_rate": 8.311483653517294e-05,
      "loss": 2.7877,
      "step": 1370
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.0081210136413574,
      "learning_rate": 8.298974226226919e-05,
      "loss": 2.7696,
      "step": 1375
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.0363234281539917,
      "learning_rate": 8.28642812208345e-05,
      "loss": 2.7195,
      "step": 1380
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.0745787620544434,
      "learning_rate": 8.273845480570123e-05,
      "loss": 2.6861,
      "step": 1385
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.073278784751892,
      "learning_rate": 8.26122644157639e-05,
      "loss": 2.719,
      "step": 1390
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.0157283544540405,
      "learning_rate": 8.248571145396362e-05,
      "loss": 2.7545,
      "step": 1395
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.1474837064743042,
      "learning_rate": 8.235879732727236e-05,
      "loss": 2.8279,
      "step": 1400
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.0333293676376343,
      "learning_rate": 8.223152344667745e-05,
      "loss": 2.7702,
      "step": 1405
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.987920880317688,
      "learning_rate": 8.21038912271658e-05,
      "loss": 2.7351,
      "step": 1410
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.0156697034835815,
      "learning_rate": 8.197590208770824e-05,
      "loss": 2.7674,
      "step": 1415
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.0461132526397705,
      "learning_rate": 8.184755745124371e-05,
      "loss": 2.745,
      "step": 1420
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.0527997016906738,
      "learning_rate": 8.171885874466342e-05,
      "loss": 2.799,
      "step": 1425
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.162582278251648,
      "learning_rate": 8.158980739879507e-05,
      "loss": 2.7565,
      "step": 1430
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.0373525619506836,
      "learning_rate": 8.146040484838677e-05,
      "loss": 2.7538,
      "step": 1435
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.0482909679412842,
      "learning_rate": 8.133065253209132e-05,
      "loss": 2.8521,
      "step": 1440
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.0852783918380737,
      "learning_rate": 8.120055189245e-05,
      "loss": 2.7346,
      "step": 1445
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.078517198562622,
      "learning_rate": 8.10701043758767e-05,
      "loss": 2.6777,
      "step": 1450
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.0317991971969604,
      "learning_rate": 8.093931143264174e-05,
      "loss": 2.7016,
      "step": 1455
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.9878939986228943,
      "learning_rate": 8.080817451685576e-05,
      "loss": 2.8025,
      "step": 1460
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.9930393695831299,
      "learning_rate": 8.067669508645356e-05,
      "loss": 2.735,
      "step": 1465
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0263429880142212,
      "learning_rate": 8.054487460317797e-05,
      "loss": 2.7556,
      "step": 1470
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0251388549804688,
      "learning_rate": 8.041271453256345e-05,
      "loss": 2.7951,
      "step": 1475
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0688138008117676,
      "learning_rate": 8.02802163439199e-05,
      "loss": 2.6784,
      "step": 1480
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0590801239013672,
      "learning_rate": 8.01473815103163e-05,
      "loss": 2.7043,
      "step": 1485
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0488477945327759,
      "learning_rate": 8.001421150856434e-05,
      "loss": 2.7299,
      "step": 1490
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0498743057250977,
      "learning_rate": 7.988070781920197e-05,
      "loss": 2.7261,
      "step": 1495
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0502474308013916,
      "learning_rate": 7.9746871926477e-05,
      "loss": 2.756,
      "step": 1500
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0085452795028687,
      "learning_rate": 7.961270531833052e-05,
      "loss": 2.7687,
      "step": 1505
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0181148052215576,
      "learning_rate": 7.947820948638045e-05,
      "loss": 2.7631,
      "step": 1510
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9904955625534058,
      "learning_rate": 7.934338592590486e-05,
      "loss": 2.7268,
      "step": 1515
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.027920126914978,
      "learning_rate": 7.92082361358254e-05,
      "loss": 2.7101,
      "step": 1520
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.058929443359375,
      "learning_rate": 7.907276161869065e-05,
      "loss": 2.7147,
      "step": 1525
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.049456238746643,
      "learning_rate": 7.893696388065936e-05,
      "loss": 2.7167,
      "step": 1530
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0867774486541748,
      "learning_rate": 7.88008444314838e-05,
      "loss": 2.6422,
      "step": 1535
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9677130579948425,
      "learning_rate": 7.866440478449283e-05,
      "loss": 2.6352,
      "step": 1540
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0104931592941284,
      "learning_rate": 7.852764645657522e-05,
      "loss": 2.8134,
      "step": 1545
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9881729483604431,
      "learning_rate": 7.839057096816271e-05,
      "loss": 2.7757,
      "step": 1550
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0977373123168945,
      "learning_rate": 7.82531798432131e-05,
      "loss": 2.7072,
      "step": 1555
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0084450244903564,
      "learning_rate": 7.811547460919333e-05,
      "loss": 2.6915,
      "step": 1560
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9954885840415955,
      "learning_rate": 7.797745679706254e-05,
      "loss": 2.7586,
      "step": 1565
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.036213755607605,
      "learning_rate": 7.783912794125496e-05,
      "loss": 2.6901,
      "step": 1570
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.0960286855697632,
      "learning_rate": 7.770048957966291e-05,
      "loss": 2.7356,
      "step": 1575
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.081369161605835,
      "learning_rate": 7.756154325361967e-05,
      "loss": 2.6979,
      "step": 1580
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.055607795715332,
      "learning_rate": 7.74222905078824e-05,
      "loss": 2.7447,
      "step": 1585
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.969778299331665,
      "learning_rate": 7.728273289061489e-05,
      "loss": 2.6753,
      "step": 1590
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9912227392196655,
      "learning_rate": 7.714287195337044e-05,
      "loss": 2.7152,
      "step": 1595
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.969498872756958,
      "learning_rate": 7.700270925107448e-05,
      "loss": 2.7074,
      "step": 1600
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9954387545585632,
      "learning_rate": 7.686224634200742e-05,
      "loss": 2.7113,
      "step": 1605
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.036186933517456,
      "learning_rate": 7.672148478778722e-05,
      "loss": 2.6632,
      "step": 1610
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9863171577453613,
      "learning_rate": 7.658042615335212e-05,
      "loss": 2.7085,
      "step": 1615
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9429605603218079,
      "learning_rate": 7.643907200694318e-05,
      "loss": 2.6517,
      "step": 1620
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.014510154724121,
      "learning_rate": 7.629742392008684e-05,
      "loss": 2.6171,
      "step": 1625
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9857035279273987,
      "learning_rate": 7.615548346757749e-05,
      "loss": 2.6967,
      "step": 1630
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.0688436031341553,
      "learning_rate": 7.60132522274599e-05,
      "loss": 2.7761,
      "step": 1635
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9699895977973938,
      "learning_rate": 7.587073178101178e-05,
      "loss": 2.7205,
      "step": 1640
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.1228586435317993,
      "learning_rate": 7.572792371272609e-05,
      "loss": 2.6913,
      "step": 1645
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.030258297920227,
      "learning_rate": 7.55848296102935e-05,
      "loss": 2.6817,
      "step": 1650
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.0188491344451904,
      "learning_rate": 7.544145106458465e-05,
      "loss": 2.7072,
      "step": 1655
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9991633892059326,
      "learning_rate": 7.529778966963259e-05,
      "loss": 2.6532,
      "step": 1660
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.0799708366394043,
      "learning_rate": 7.515384702261496e-05,
      "loss": 2.6856,
      "step": 1665
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9435780644416809,
      "learning_rate": 7.500962472383627e-05,
      "loss": 2.6041,
      "step": 1670
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.0160300731658936,
      "learning_rate": 7.486512437671011e-05,
      "loss": 2.7019,
      "step": 1675
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.114219307899475,
      "learning_rate": 7.472034758774128e-05,
      "loss": 2.7862,
      "step": 1680
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9939204454421997,
      "learning_rate": 7.457529596650797e-05,
      "loss": 2.6545,
      "step": 1685
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.975845456123352,
      "learning_rate": 7.442997112564392e-05,
      "loss": 2.7047,
      "step": 1690
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9775164723396301,
      "learning_rate": 7.428437468082037e-05,
      "loss": 2.7103,
      "step": 1695
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9839438796043396,
      "learning_rate": 7.413850825072817e-05,
      "loss": 2.7212,
      "step": 1700
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9818946123123169,
      "learning_rate": 7.39923734570598e-05,
      "loss": 2.71,
      "step": 1705
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9547445178031921,
      "learning_rate": 7.384597192449126e-05,
      "loss": 2.7184,
      "step": 1710
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.0153919458389282,
      "learning_rate": 7.369930528066412e-05,
      "loss": 2.6053,
      "step": 1715
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9527406096458435,
      "learning_rate": 7.355237515616732e-05,
      "loss": 2.6794,
      "step": 1720
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9509900808334351,
      "learning_rate": 7.340518318451914e-05,
      "loss": 2.6821,
      "step": 1725
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.0353055000305176,
      "learning_rate": 7.325773100214893e-05,
      "loss": 2.6917,
      "step": 1730
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9707476496696472,
      "learning_rate": 7.311002024837899e-05,
      "loss": 2.6346,
      "step": 1735
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.999159574508667,
      "learning_rate": 7.296205256540633e-05,
      "loss": 2.6769,
      "step": 1740
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9343658089637756,
      "learning_rate": 7.281382959828443e-05,
      "loss": 2.6112,
      "step": 1745
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.0174192190170288,
      "learning_rate": 7.26653529949049e-05,
      "loss": 2.7012,
      "step": 1750
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.0045185089111328,
      "learning_rate": 7.25166244059792e-05,
      "loss": 2.6414,
      "step": 1755
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9818803668022156,
      "learning_rate": 7.236764548502029e-05,
      "loss": 2.6397,
      "step": 1760
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.0778367519378662,
      "learning_rate": 7.221841788832421e-05,
      "loss": 2.635,
      "step": 1765
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.0316153764724731,
      "learning_rate": 7.206894327495173e-05,
      "loss": 2.6314,
      "step": 1770
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9475739002227783,
      "learning_rate": 7.191922330670982e-05,
      "loss": 2.595,
      "step": 1775
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9840036034584045,
      "learning_rate": 7.176925964813326e-05,
      "loss": 2.6984,
      "step": 1780
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9781378507614136,
      "learning_rate": 7.161905396646607e-05,
      "loss": 2.6671,
      "step": 1785
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9534721970558167,
      "learning_rate": 7.146860793164299e-05,
      "loss": 2.6378,
      "step": 1790
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9708443880081177,
      "learning_rate": 7.131792321627098e-05,
      "loss": 2.6401,
      "step": 1795
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9724230766296387,
      "learning_rate": 7.116700149561048e-05,
      "loss": 2.6484,
      "step": 1800
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.0079472064971924,
      "learning_rate": 7.101584444755696e-05,
      "loss": 2.6751,
      "step": 1805
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9262291789054871,
      "learning_rate": 7.086445375262212e-05,
      "loss": 2.5996,
      "step": 1810
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.0142055749893188,
      "learning_rate": 7.071283109391528e-05,
      "loss": 2.7081,
      "step": 1815
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.0082738399505615,
      "learning_rate": 7.056097815712466e-05,
      "loss": 2.6156,
      "step": 1820
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9950462579727173,
      "learning_rate": 7.040889663049862e-05,
      "loss": 2.6308,
      "step": 1825
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.0031845569610596,
      "learning_rate": 7.025658820482693e-05,
      "loss": 2.6845,
      "step": 1830
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.0086175203323364,
      "learning_rate": 7.010405457342192e-05,
      "loss": 2.6379,
      "step": 1835
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.97476726770401,
      "learning_rate": 6.995129743209967e-05,
      "loss": 2.6356,
      "step": 1840
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9677056074142456,
      "learning_rate": 6.97983184791612e-05,
      "loss": 2.6214,
      "step": 1845
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.002587914466858,
      "learning_rate": 6.964511941537355e-05,
      "loss": 2.6255,
      "step": 1850
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9613837003707886,
      "learning_rate": 6.949170194395083e-05,
      "loss": 2.6151,
      "step": 1855
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.0320087671279907,
      "learning_rate": 6.933806777053536e-05,
      "loss": 2.6616,
      "step": 1860
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.0032234191894531,
      "learning_rate": 6.918421860317872e-05,
      "loss": 2.6229,
      "step": 1865
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9621163606643677,
      "learning_rate": 6.903015615232263e-05,
      "loss": 2.6176,
      "step": 1870
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9764354825019836,
      "learning_rate": 6.887588213078012e-05,
      "loss": 2.6557,
      "step": 1875
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.0155274868011475,
      "learning_rate": 6.87213982537163e-05,
      "loss": 2.6642,
      "step": 1880
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9893097281455994,
      "learning_rate": 6.856670623862943e-05,
      "loss": 2.6603,
      "step": 1885
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9779693484306335,
      "learning_rate": 6.841180780533179e-05,
      "loss": 2.6607,
      "step": 1890
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9833638072013855,
      "learning_rate": 6.82567046759305e-05,
      "loss": 2.6829,
      "step": 1895
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.0706186294555664,
      "learning_rate": 6.810139857480844e-05,
      "loss": 2.6703,
      "step": 1900
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.0078755617141724,
      "learning_rate": 6.794589122860509e-05,
      "loss": 2.6086,
      "step": 1905
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9727465510368347,
      "learning_rate": 6.779018436619725e-05,
      "loss": 2.6492,
      "step": 1910
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.985817015171051,
      "learning_rate": 6.763427971867992e-05,
      "loss": 2.5795,
      "step": 1915
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.0108627080917358,
      "learning_rate": 6.747817901934699e-05,
      "loss": 2.6732,
      "step": 1920
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9770353436470032,
      "learning_rate": 6.732188400367197e-05,
      "loss": 2.616,
      "step": 1925
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9645941853523254,
      "learning_rate": 6.716539640928871e-05,
      "loss": 2.6159,
      "step": 1930
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9376698136329651,
      "learning_rate": 6.70087179759721e-05,
      "loss": 2.6428,
      "step": 1935
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9465940594673157,
      "learning_rate": 6.685185044561874e-05,
      "loss": 2.6459,
      "step": 1940
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9306380748748779,
      "learning_rate": 6.669479556222747e-05,
      "loss": 2.633,
      "step": 1945
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9376333355903625,
      "learning_rate": 6.653755507188013e-05,
      "loss": 2.5994,
      "step": 1950
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9646631479263306,
      "learning_rate": 6.638013072272205e-05,
      "loss": 2.6609,
      "step": 1955
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9547147154808044,
      "learning_rate": 6.622252426494259e-05,
      "loss": 2.638,
      "step": 1960
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9426686763763428,
      "learning_rate": 6.606473745075581e-05,
      "loss": 2.5625,
      "step": 1965
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9773340821266174,
      "learning_rate": 6.590677203438084e-05,
      "loss": 2.6222,
      "step": 1970
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9521613121032715,
      "learning_rate": 6.574862977202252e-05,
      "loss": 2.6725,
      "step": 1975
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9960840940475464,
      "learning_rate": 6.559031242185174e-05,
      "loss": 2.629,
      "step": 1980
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9541398286819458,
      "learning_rate": 6.543182174398597e-05,
      "loss": 2.604,
      "step": 1985
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9481289386749268,
      "learning_rate": 6.52731595004697e-05,
      "loss": 2.5867,
      "step": 1990
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9873711466789246,
      "learning_rate": 6.51143274552548e-05,
      "loss": 2.6228,
      "step": 1995
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.999569296836853,
      "learning_rate": 6.495532737418098e-05,
      "loss": 2.6684,
      "step": 2000
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.019819736480713,
      "learning_rate": 6.479616102495605e-05,
      "loss": 2.571,
      "step": 2005
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9891926050186157,
      "learning_rate": 6.463683017713638e-05,
      "loss": 2.7003,
      "step": 2010
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9509793519973755,
      "learning_rate": 6.447733660210715e-05,
      "loss": 2.5304,
      "step": 2015
    },
    {
      "epoch": 0.42,
      "grad_norm": 12.407959938049316,
      "learning_rate": 6.431768207306272e-05,
      "loss": 2.6177,
      "step": 2020
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.0190647840499878,
      "learning_rate": 6.415786836498684e-05,
      "loss": 2.5533,
      "step": 2025
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9944144487380981,
      "learning_rate": 6.399789725463298e-05,
      "loss": 2.5744,
      "step": 2030
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9825426340103149,
      "learning_rate": 6.383777052050458e-05,
      "loss": 2.6318,
      "step": 2035
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9628515839576721,
      "learning_rate": 6.367748994283518e-05,
      "loss": 2.6367,
      "step": 2040
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9694194197654724,
      "learning_rate": 6.351705730356877e-05,
      "loss": 2.5802,
      "step": 2045
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.967638373374939,
      "learning_rate": 6.335647438633987e-05,
      "loss": 2.5357,
      "step": 2050
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9158982038497925,
      "learning_rate": 6.319574297645374e-05,
      "loss": 2.5934,
      "step": 2055
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9469335675239563,
      "learning_rate": 6.303486486086654e-05,
      "loss": 2.5442,
      "step": 2060
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9547492861747742,
      "learning_rate": 6.287384182816546e-05,
      "loss": 2.639,
      "step": 2065
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9499625563621521,
      "learning_rate": 6.271267566854883e-05,
      "loss": 2.5463,
      "step": 2070
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9230909943580627,
      "learning_rate": 6.255136817380618e-05,
      "loss": 2.5663,
      "step": 2075
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9761818647384644,
      "learning_rate": 6.23899211372984e-05,
      "loss": 2.5928,
      "step": 2080
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9458242654800415,
      "learning_rate": 6.222833635393772e-05,
      "loss": 2.5068,
      "step": 2085
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.0186399221420288,
      "learning_rate": 6.206661562016782e-05,
      "loss": 2.5338,
      "step": 2090
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9645049571990967,
      "learning_rate": 6.190476073394382e-05,
      "loss": 2.5828,
      "step": 2095
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9071171283721924,
      "learning_rate": 6.17427734947123e-05,
      "loss": 2.5709,
      "step": 2100
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9457060694694519,
      "learning_rate": 6.158065570339127e-05,
      "loss": 2.569,
      "step": 2105
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.0017690658569336,
      "learning_rate": 6.141840916235021e-05,
      "loss": 2.6378,
      "step": 2110
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9879325032234192,
      "learning_rate": 6.125603567539001e-05,
      "loss": 2.5915,
      "step": 2115
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9789354205131531,
      "learning_rate": 6.109353704772284e-05,
      "loss": 2.6959,
      "step": 2120
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9717521071434021,
      "learning_rate": 6.0930915085952164e-05,
      "loss": 2.6171,
      "step": 2125
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.026318907737732,
      "learning_rate": 6.076817159805267e-05,
      "loss": 2.59,
      "step": 2130
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9704822301864624,
      "learning_rate": 6.06053083933501e-05,
      "loss": 2.6399,
      "step": 2135
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.938707709312439,
      "learning_rate": 6.044232728250116e-05,
      "loss": 2.5744,
      "step": 2140
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9552039504051208,
      "learning_rate": 6.027923007747339e-05,
      "loss": 2.603,
      "step": 2145
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9553301334381104,
      "learning_rate": 6.011601859152506e-05,
      "loss": 2.6054,
      "step": 2150
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9410426616668701,
      "learning_rate": 5.995269463918495e-05,
      "loss": 2.6622,
      "step": 2155
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.929993212223053,
      "learning_rate": 5.97892600362322e-05,
      "loss": 2.6446,
      "step": 2160
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9774343967437744,
      "learning_rate": 5.962571659967614e-05,
      "loss": 2.6214,
      "step": 2165
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9398266673088074,
      "learning_rate": 5.946206614773606e-05,
      "loss": 2.5305,
      "step": 2170
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.005797266960144,
      "learning_rate": 5.929831049982103e-05,
      "loss": 2.5693,
      "step": 2175
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9824793338775635,
      "learning_rate": 5.9134451476509633e-05,
      "loss": 2.6274,
      "step": 2180
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9403955936431885,
      "learning_rate": 5.897049089952974e-05,
      "loss": 2.5396,
      "step": 2185
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.966221034526825,
      "learning_rate": 5.880643059173826e-05,
      "loss": 2.555,
      "step": 2190
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9708110690116882,
      "learning_rate": 5.864227237710093e-05,
      "loss": 2.5647,
      "step": 2195
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9578514695167542,
      "learning_rate": 5.847801808067189e-05,
      "loss": 2.6166,
      "step": 2200
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9706119298934937,
      "learning_rate": 5.831366952857357e-05,
      "loss": 2.6013,
      "step": 2205
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.961341917514801,
      "learning_rate": 5.814922854797622e-05,
      "loss": 2.6273,
      "step": 2210
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9282676577568054,
      "learning_rate": 5.798469696707775e-05,
      "loss": 2.5886,
      "step": 2215
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.1473058462142944,
      "learning_rate": 5.782007661508331e-05,
      "loss": 2.5553,
      "step": 2220
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9114869832992554,
      "learning_rate": 5.765536932218495e-05,
      "loss": 2.53,
      "step": 2225
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9566113352775574,
      "learning_rate": 5.7490576919541315e-05,
      "loss": 2.5858,
      "step": 2230
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9558666944503784,
      "learning_rate": 5.732570123925729e-05,
      "loss": 2.5127,
      "step": 2235
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9059987664222717,
      "learning_rate": 5.7160744114363593e-05,
      "loss": 2.5475,
      "step": 2240
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9068989157676697,
      "learning_rate": 5.699570737879641e-05,
      "loss": 2.5572,
      "step": 2245
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9678176045417786,
      "learning_rate": 5.683059286737702e-05,
      "loss": 2.5725,
      "step": 2250
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9395660161972046,
      "learning_rate": 5.666540241579139e-05,
      "loss": 2.5859,
      "step": 2255
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9277108311653137,
      "learning_rate": 5.6500137860569766e-05,
      "loss": 2.541,
      "step": 2260
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9472150206565857,
      "learning_rate": 5.633480103906624e-05,
      "loss": 2.6044,
      "step": 2265
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9881107807159424,
      "learning_rate": 5.616939378943834e-05,
      "loss": 2.5783,
      "step": 2270
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9077834486961365,
      "learning_rate": 5.6003917950626595e-05,
      "loss": 2.5131,
      "step": 2275
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9303057193756104,
      "learning_rate": 5.583837536233407e-05,
      "loss": 2.5234,
      "step": 2280
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9464006423950195,
      "learning_rate": 5.567276786500596e-05,
      "loss": 2.6048,
      "step": 2285
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9303953051567078,
      "learning_rate": 5.5507097299809054e-05,
      "loss": 2.6002,
      "step": 2290
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9646936655044556,
      "learning_rate": 5.534136550861133e-05,
      "loss": 2.5923,
      "step": 2295
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9685410261154175,
      "learning_rate": 5.5175574333961465e-05,
      "loss": 2.5573,
      "step": 2300
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9414950609207153,
      "learning_rate": 5.500972561906832e-05,
      "loss": 2.6158,
      "step": 2305
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9068962931632996,
      "learning_rate": 5.484382120778048e-05,
      "loss": 2.6208,
      "step": 2310
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9170411229133606,
      "learning_rate": 5.467786294456575e-05,
      "loss": 2.523,
      "step": 2315
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9522758722305298,
      "learning_rate": 5.451185267449061e-05,
      "loss": 2.5167,
      "step": 2320
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9827879071235657,
      "learning_rate": 5.43457922431998e-05,
      "loss": 2.6251,
      "step": 2325
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9485974907875061,
      "learning_rate": 5.417968349689566e-05,
      "loss": 2.5533,
      "step": 2330
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9093174338340759,
      "learning_rate": 5.401352828231772e-05,
      "loss": 2.5801,
      "step": 2335
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9550360441207886,
      "learning_rate": 5.384732844672211e-05,
      "loss": 2.5791,
      "step": 2340
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.01358962059021,
      "learning_rate": 5.368108583786107e-05,
      "loss": 2.5222,
      "step": 2345
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9926130175590515,
      "learning_rate": 5.3514802303962344e-05,
      "loss": 2.5862,
      "step": 2350
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.913351833820343,
      "learning_rate": 5.334847969370868e-05,
      "loss": 2.4312,
      "step": 2355
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9434849619865417,
      "learning_rate": 5.3182119856217284e-05,
      "loss": 2.6087,
      "step": 2360
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9408784508705139,
      "learning_rate": 5.3015724641019214e-05,
      "loss": 2.5432,
      "step": 2365
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9512544870376587,
      "learning_rate": 5.284929589803884e-05,
      "loss": 2.5393,
      "step": 2370
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9654476642608643,
      "learning_rate": 5.2682835477573336e-05,
      "loss": 2.5588,
      "step": 2375
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9294528961181641,
      "learning_rate": 5.2516345230271965e-05,
      "loss": 2.5511,
      "step": 2380
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9224638342857361,
      "learning_rate": 5.234982700711569e-05,
      "loss": 2.5216,
      "step": 2385
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9053465127944946,
      "learning_rate": 5.218328265939643e-05,
      "loss": 2.5536,
      "step": 2390
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9090244174003601,
      "learning_rate": 5.201671403869657e-05,
      "loss": 2.5365,
      "step": 2395
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9613234400749207,
      "learning_rate": 5.1850122996868366e-05,
      "loss": 2.572,
      "step": 2400
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9468548893928528,
      "learning_rate": 5.168351138601334e-05,
      "loss": 2.509,
      "step": 2405
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.098836898803711,
      "learning_rate": 5.1516881058461675e-05,
      "loss": 2.5212,
      "step": 2410
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9981148838996887,
      "learning_rate": 5.135023386675166e-05,
      "loss": 2.5296,
      "step": 2415
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9370298385620117,
      "learning_rate": 5.118357166360906e-05,
      "loss": 2.565,
      "step": 2420
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9485419988632202,
      "learning_rate": 5.101689630192655e-05,
      "loss": 2.4705,
      "step": 2425
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9375891089439392,
      "learning_rate": 5.085020963474307e-05,
      "loss": 2.4834,
      "step": 2430
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9355273246765137,
      "learning_rate": 5.068351351522329e-05,
      "loss": 2.6003,
      "step": 2435
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9307159781455994,
      "learning_rate": 5.0516809796636935e-05,
      "loss": 2.4923,
      "step": 2440
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.951213002204895,
      "learning_rate": 5.035010033233821e-05,
      "loss": 2.5471,
      "step": 2445
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8841201663017273,
      "learning_rate": 5.018338697574523e-05,
      "loss": 2.4939,
      "step": 2450
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9634934067726135,
      "learning_rate": 5.0016671580319354e-05,
      "loss": 2.5294,
      "step": 2455
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9498656988143921,
      "learning_rate": 4.984995599954461e-05,
      "loss": 2.5306,
      "step": 2460
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9849341511726379,
      "learning_rate": 4.968324208690712e-05,
      "loss": 2.5498,
      "step": 2465
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9472889304161072,
      "learning_rate": 4.951653169587441e-05,
      "loss": 2.5471,
      "step": 2470
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.919108510017395,
      "learning_rate": 4.93498266798749e-05,
      "loss": 2.5413,
      "step": 2475
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9519367814064026,
      "learning_rate": 4.918312889227722e-05,
      "loss": 2.5299,
      "step": 2480
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9095039367675781,
      "learning_rate": 4.901644018636966e-05,
      "loss": 2.5499,
      "step": 2485
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9608018398284912,
      "learning_rate": 4.8849762415339526e-05,
      "loss": 2.4624,
      "step": 2490
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9327159523963928,
      "learning_rate": 4.868309743225256e-05,
      "loss": 2.5808,
      "step": 2495
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9636819362640381,
      "learning_rate": 4.851644709003233e-05,
      "loss": 2.5551,
      "step": 2500
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9298324584960938,
      "learning_rate": 4.834981324143964e-05,
      "loss": 2.4952,
      "step": 2505
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9329404830932617,
      "learning_rate": 4.818319773905191e-05,
      "loss": 2.5115,
      "step": 2510
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9319917559623718,
      "learning_rate": 4.801660243524261e-05,
      "loss": 2.5493,
      "step": 2515
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9572451114654541,
      "learning_rate": 4.7850029182160626e-05,
      "loss": 2.5974,
      "step": 2520
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.938522219657898,
      "learning_rate": 4.768347983170973e-05,
      "loss": 2.5079,
      "step": 2525
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9182351231575012,
      "learning_rate": 4.7516956235527884e-05,
      "loss": 2.4939,
      "step": 2530
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9615866541862488,
      "learning_rate": 4.735046024496682e-05,
      "loss": 2.5146,
      "step": 2535
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9523360729217529,
      "learning_rate": 4.7183993711071286e-05,
      "loss": 2.4768,
      "step": 2540
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9364380240440369,
      "learning_rate": 4.7017558484558554e-05,
      "loss": 2.55,
      "step": 2545
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9449701905250549,
      "learning_rate": 4.6851156415797844e-05,
      "loss": 2.4937,
      "step": 2550
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9214823246002197,
      "learning_rate": 4.6684789354789746e-05,
      "loss": 2.5316,
      "step": 2555
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8947305679321289,
      "learning_rate": 4.651845915114563e-05,
      "loss": 2.5681,
      "step": 2560
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9033908843994141,
      "learning_rate": 4.6352167654067095e-05,
      "loss": 2.5183,
      "step": 2565
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.0538268089294434,
      "learning_rate": 4.618591671232544e-05,
      "loss": 2.5556,
      "step": 2570
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9056328535079956,
      "learning_rate": 4.601970817424106e-05,
      "loss": 2.5046,
      "step": 2575
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.1014292240142822,
      "learning_rate": 4.585354388766292e-05,
      "loss": 2.5109,
      "step": 2580
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8942855596542358,
      "learning_rate": 4.568742569994802e-05,
      "loss": 2.4971,
      "step": 2585
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8742932677268982,
      "learning_rate": 4.552135545794086e-05,
      "loss": 2.5116,
      "step": 2590
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9332443475723267,
      "learning_rate": 4.535533500795288e-05,
      "loss": 2.511,
      "step": 2595
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9206485152244568,
      "learning_rate": 4.5189366195741953e-05,
      "loss": 2.5107,
      "step": 2600
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9191185832023621,
      "learning_rate": 4.502345086649186e-05,
      "loss": 2.4824,
      "step": 2605
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9247689247131348,
      "learning_rate": 4.485759086479179e-05,
      "loss": 2.4865,
      "step": 2610
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9270159006118774,
      "learning_rate": 4.469178803461579e-05,
      "loss": 2.5851,
      "step": 2615
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9007619023323059,
      "learning_rate": 4.4526044219302326e-05,
      "loss": 2.4677,
      "step": 2620
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9385355114936829,
      "learning_rate": 4.4360361261533745e-05,
      "loss": 2.4913,
      "step": 2625
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.936024010181427,
      "learning_rate": 4.419474100331579e-05,
      "loss": 2.5147,
      "step": 2630
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9851381778717041,
      "learning_rate": 4.402918528595715e-05,
      "loss": 2.5625,
      "step": 2635
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9795789122581482,
      "learning_rate": 4.386369595004896e-05,
      "loss": 2.5276,
      "step": 2640
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9688675403594971,
      "learning_rate": 4.3698274835444354e-05,
      "loss": 2.5187,
      "step": 2645
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9564111828804016,
      "learning_rate": 4.3532923781238e-05,
      "loss": 2.5103,
      "step": 2650
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9104586243629456,
      "learning_rate": 4.336764462574566e-05,
      "loss": 2.5078,
      "step": 2655
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9217368364334106,
      "learning_rate": 4.320243920648376e-05,
      "loss": 2.5819,
      "step": 2660
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8980958461761475,
      "learning_rate": 4.303730936014894e-05,
      "loss": 2.5107,
      "step": 2665
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9054094552993774,
      "learning_rate": 4.287225692259765e-05,
      "loss": 2.4817,
      "step": 2670
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9648101925849915,
      "learning_rate": 4.270728372882575e-05,
      "loss": 2.4727,
      "step": 2675
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.0163148641586304,
      "learning_rate": 4.254239161294804e-05,
      "loss": 2.5593,
      "step": 2680
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9777525067329407,
      "learning_rate": 4.237758240817802e-05,
      "loss": 2.5291,
      "step": 2685
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9313981533050537,
      "learning_rate": 4.2212857946807336e-05,
      "loss": 2.5546,
      "step": 2690
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9364113807678223,
      "learning_rate": 4.2048220060185516e-05,
      "loss": 2.4344,
      "step": 2695
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.912443995475769,
      "learning_rate": 4.188367057869957e-05,
      "loss": 2.5345,
      "step": 2700
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9088121652603149,
      "learning_rate": 4.171921133175365e-05,
      "loss": 2.5508,
      "step": 2705
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9182174801826477,
      "learning_rate": 4.155484414774872e-05,
      "loss": 2.4666,
      "step": 2710
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9237582087516785,
      "learning_rate": 4.139057085406221e-05,
      "loss": 2.52,
      "step": 2715
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9367585778236389,
      "learning_rate": 4.1226393277027726e-05,
      "loss": 2.5637,
      "step": 2720
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9298025965690613,
      "learning_rate": 4.106231324191471e-05,
      "loss": 2.5733,
      "step": 2725
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.921440064907074,
      "learning_rate": 4.089833257290817e-05,
      "loss": 2.509,
      "step": 2730
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9271067380905151,
      "learning_rate": 4.073445309308842e-05,
      "loss": 2.5282,
      "step": 2735
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9048083424568176,
      "learning_rate": 4.0570676624410756e-05,
      "loss": 2.5263,
      "step": 2740
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9049689173698425,
      "learning_rate": 4.040700498768525e-05,
      "loss": 2.5656,
      "step": 2745
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9169425368309021,
      "learning_rate": 4.024344000255648e-05,
      "loss": 2.5417,
      "step": 2750
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9237762093544006,
      "learning_rate": 4.0079983487483313e-05,
      "loss": 2.4893,
      "step": 2755
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9239086508750916,
      "learning_rate": 3.9916637259718683e-05,
      "loss": 2.5609,
      "step": 2760
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9628281593322754,
      "learning_rate": 3.9753403135289396e-05,
      "loss": 2.5858,
      "step": 2765
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9475969672203064,
      "learning_rate": 3.9590282928975914e-05,
      "loss": 2.5439,
      "step": 2770
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9971244931221008,
      "learning_rate": 3.942727845429221e-05,
      "loss": 2.5218,
      "step": 2775
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9252620935440063,
      "learning_rate": 3.926439152346558e-05,
      "loss": 2.5145,
      "step": 2780
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9607737064361572,
      "learning_rate": 3.910162394741653e-05,
      "loss": 2.4541,
      "step": 2785
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9223071932792664,
      "learning_rate": 3.893897753573861e-05,
      "loss": 2.4738,
      "step": 2790
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9555774927139282,
      "learning_rate": 3.877645409667829e-05,
      "loss": 2.431,
      "step": 2795
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9491199254989624,
      "learning_rate": 3.861405543711491e-05,
      "loss": 2.4294,
      "step": 2800
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9491716623306274,
      "learning_rate": 3.8451783362540507e-05,
      "loss": 2.4642,
      "step": 2805
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9213913679122925,
      "learning_rate": 3.828963967703983e-05,
      "loss": 2.4871,
      "step": 2810
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9365127682685852,
      "learning_rate": 3.8127626183270223e-05,
      "loss": 2.4805,
      "step": 2815
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.908562958240509,
      "learning_rate": 3.796574468244161e-05,
      "loss": 2.4572,
      "step": 2820
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9270069599151611,
      "learning_rate": 3.7803996974296444e-05,
      "loss": 2.5138,
      "step": 2825
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9667962193489075,
      "learning_rate": 3.7642384857089776e-05,
      "loss": 2.5384,
      "step": 2830
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9173279404640198,
      "learning_rate": 3.748091012756915e-05,
      "loss": 2.5528,
      "step": 2835
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9413408041000366,
      "learning_rate": 3.731957458095467e-05,
      "loss": 2.4776,
      "step": 2840
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9140501618385315,
      "learning_rate": 3.71583800109191e-05,
      "loss": 2.5965,
      "step": 2845
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9697785377502441,
      "learning_rate": 3.699732820956784e-05,
      "loss": 2.4439,
      "step": 2850
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9393496513366699,
      "learning_rate": 3.6836420967419057e-05,
      "loss": 2.4883,
      "step": 2855
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9118742942810059,
      "learning_rate": 3.6675660073383745e-05,
      "loss": 2.5305,
      "step": 2860
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8981103301048279,
      "learning_rate": 3.6515047314745856e-05,
      "loss": 2.4762,
      "step": 2865
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9413728713989258,
      "learning_rate": 3.6354584477142437e-05,
      "loss": 2.5435,
      "step": 2870
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9207038879394531,
      "learning_rate": 3.6194273344543736e-05,
      "loss": 2.5114,
      "step": 2875
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9038093090057373,
      "learning_rate": 3.6034115699233425e-05,
      "loss": 2.4749,
      "step": 2880
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9292318224906921,
      "learning_rate": 3.5874113321788736e-05,
      "loss": 2.4526,
      "step": 2885
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8991361856460571,
      "learning_rate": 3.571426799106071e-05,
      "loss": 2.4961,
      "step": 2890
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9043452739715576,
      "learning_rate": 3.555458148415437e-05,
      "loss": 2.468,
      "step": 2895
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8990312218666077,
      "learning_rate": 3.539505557640901e-05,
      "loss": 2.4527,
      "step": 2900
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9166343808174133,
      "learning_rate": 3.523569204137843e-05,
      "loss": 2.5101,
      "step": 2905
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9148818850517273,
      "learning_rate": 3.5076492650811246e-05,
      "loss": 2.4654,
      "step": 2910
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8907680511474609,
      "learning_rate": 3.491745917463113e-05,
      "loss": 2.513,
      "step": 2915
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8982478380203247,
      "learning_rate": 3.475859338091721e-05,
      "loss": 2.4826,
      "step": 2920
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9068254828453064,
      "learning_rate": 3.4599897035884374e-05,
      "loss": 2.4805,
      "step": 2925
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9060400128364563,
      "learning_rate": 3.444137190386363e-05,
      "loss": 2.4454,
      "step": 2930
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9024885296821594,
      "learning_rate": 3.4283019747282514e-05,
      "loss": 2.4804,
      "step": 2935
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8852449655532837,
      "learning_rate": 3.412484232664545e-05,
      "loss": 2.4749,
      "step": 2940
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9173283576965332,
      "learning_rate": 3.396684140051424e-05,
      "loss": 2.4272,
      "step": 2945
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8929095268249512,
      "learning_rate": 3.3809018725488466e-05,
      "loss": 2.4746,
      "step": 2950
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9127289652824402,
      "learning_rate": 3.365137605618598e-05,
      "loss": 2.4833,
      "step": 2955
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9289007782936096,
      "learning_rate": 3.3493915145223395e-05,
      "loss": 2.4848,
      "step": 2960
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.0071487426757812,
      "learning_rate": 3.3336637743196584e-05,
      "loss": 2.4652,
      "step": 2965
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.938108503818512,
      "learning_rate": 3.317954559866126e-05,
      "loss": 2.4782,
      "step": 2970
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9593190550804138,
      "learning_rate": 3.302264045811344e-05,
      "loss": 2.4936,
      "step": 2975
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8893548846244812,
      "learning_rate": 3.286592406597021e-05,
      "loss": 2.4851,
      "step": 2980
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8929252624511719,
      "learning_rate": 3.270939816455012e-05,
      "loss": 2.4784,
      "step": 2985
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9195278286933899,
      "learning_rate": 3.255306449405395e-05,
      "loss": 2.5345,
      "step": 2990
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9475163221359253,
      "learning_rate": 3.2396924792545304e-05,
      "loss": 2.4952,
      "step": 2995
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9016882181167603,
      "learning_rate": 3.224098079593132e-05,
      "loss": 2.4655,
      "step": 3000
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9258350729942322,
      "learning_rate": 3.2085234237943354e-05,
      "loss": 2.4314,
      "step": 3005
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9156048893928528,
      "learning_rate": 3.19296868501177e-05,
      "loss": 2.5102,
      "step": 3010
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9461023807525635,
      "learning_rate": 3.177434036177636e-05,
      "loss": 2.4557,
      "step": 3015
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9364247918128967,
      "learning_rate": 3.1619196500007804e-05,
      "loss": 2.5106,
      "step": 3020
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9256249070167542,
      "learning_rate": 3.146425698964776e-05,
      "loss": 2.4673,
      "step": 3025
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9104235768318176,
      "learning_rate": 3.1309523553260046e-05,
      "loss": 2.5164,
      "step": 3030
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.947726309299469,
      "learning_rate": 3.115499791111743e-05,
      "loss": 2.5692,
      "step": 3035
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9046145677566528,
      "learning_rate": 3.10006817811825e-05,
      "loss": 2.4769,
      "step": 3040
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9253242015838623,
      "learning_rate": 3.084657687908855e-05,
      "loss": 2.4542,
      "step": 3045
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9397364854812622,
      "learning_rate": 3.069268491812052e-05,
      "loss": 2.4806,
      "step": 3050
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9234438538551331,
      "learning_rate": 3.0539007609195934e-05,
      "loss": 2.5438,
      "step": 3055
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9175778031349182,
      "learning_rate": 3.0385546660845908e-05,
      "loss": 2.467,
      "step": 3060
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9659436941146851,
      "learning_rate": 3.0232303779196132e-05,
      "loss": 2.5125,
      "step": 3065
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9433670043945312,
      "learning_rate": 3.0079280667947885e-05,
      "loss": 2.4722,
      "step": 3070
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9202114343643188,
      "learning_rate": 2.9926479028359132e-05,
      "loss": 2.489,
      "step": 3075
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9274362325668335,
      "learning_rate": 2.97739005592256e-05,
      "loss": 2.4452,
      "step": 3080
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9218001961708069,
      "learning_rate": 2.962154695686187e-05,
      "loss": 2.4078,
      "step": 3085
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9582597017288208,
      "learning_rate": 2.9469419915082536e-05,
      "loss": 2.4814,
      "step": 3090
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9216330051422119,
      "learning_rate": 2.9317521125183368e-05,
      "loss": 2.4246,
      "step": 3095
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.92444908618927,
      "learning_rate": 2.9165852275922524e-05,
      "loss": 2.513,
      "step": 3100
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9081312417984009,
      "learning_rate": 2.901441505350174e-05,
      "loss": 2.4436,
      "step": 3105
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9419301748275757,
      "learning_rate": 2.886321114154762e-05,
      "loss": 2.529,
      "step": 3110
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.920991837978363,
      "learning_rate": 2.87122422210929e-05,
      "loss": 2.4549,
      "step": 3115
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9346429109573364,
      "learning_rate": 2.8561509970557736e-05,
      "loss": 2.471,
      "step": 3120
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9041022658348083,
      "learning_rate": 2.8411016065731146e-05,
      "loss": 2.4544,
      "step": 3125
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8967710137367249,
      "learning_rate": 2.826076217975222e-05,
      "loss": 2.4794,
      "step": 3130
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9126041531562805,
      "learning_rate": 2.8110749983091632e-05,
      "loss": 2.4372,
      "step": 3135
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9003017544746399,
      "learning_rate": 2.7960981143533053e-05,
      "loss": 2.4927,
      "step": 3140
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8914475440979004,
      "learning_rate": 2.781145732615457e-05,
      "loss": 2.4697,
      "step": 3145
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.936227023601532,
      "learning_rate": 2.7662180193310218e-05,
      "loss": 2.461,
      "step": 3150
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8699093461036682,
      "learning_rate": 2.751315140461145e-05,
      "loss": 2.465,
      "step": 3155
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9162988662719727,
      "learning_rate": 2.7364372616908744e-05,
      "loss": 2.5081,
      "step": 3160
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9147549271583557,
      "learning_rate": 2.7215845484273152e-05,
      "loss": 2.4671,
      "step": 3165
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9275458455085754,
      "learning_rate": 2.7067571657977893e-05,
      "loss": 2.5228,
      "step": 3170
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9181550741195679,
      "learning_rate": 2.691955278648003e-05,
      "loss": 2.438,
      "step": 3175
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9136456847190857,
      "learning_rate": 2.6771790515402112e-05,
      "loss": 2.4914,
      "step": 3180
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9101700186729431,
      "learning_rate": 2.6624286487513916e-05,
      "loss": 2.4316,
      "step": 3185
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9077585339546204,
      "learning_rate": 2.6477042342714137e-05,
      "loss": 2.4393,
      "step": 3190
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.911003053188324,
      "learning_rate": 2.633005971801219e-05,
      "loss": 2.5048,
      "step": 3195
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9087389707565308,
      "learning_rate": 2.6183340247510013e-05,
      "loss": 2.4693,
      "step": 3200
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9216850399971008,
      "learning_rate": 2.6036885562383856e-05,
      "loss": 2.5016,
      "step": 3205
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9231410026550293,
      "learning_rate": 2.5890697290866206e-05,
      "loss": 2.445,
      "step": 3210
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9174930453300476,
      "learning_rate": 2.5744777058227642e-05,
      "loss": 2.4903,
      "step": 3215
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9033377766609192,
      "learning_rate": 2.5599126486758777e-05,
      "loss": 2.4601,
      "step": 3220
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9198935031890869,
      "learning_rate": 2.5453747195752243e-05,
      "loss": 2.459,
      "step": 3225
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8865978717803955,
      "learning_rate": 2.530864080148464e-05,
      "loss": 2.4891,
      "step": 3230
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8864363431930542,
      "learning_rate": 2.5163808917198615e-05,
      "loss": 2.4203,
      "step": 3235
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9178940653800964,
      "learning_rate": 2.501925315308492e-05,
      "loss": 2.4455,
      "step": 3240
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9083361625671387,
      "learning_rate": 2.4874975116264477e-05,
      "loss": 2.372,
      "step": 3245
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8891680836677551,
      "learning_rate": 2.4730976410770534e-05,
      "loss": 2.4615,
      "step": 3250
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9084404706954956,
      "learning_rate": 2.458725863753084e-05,
      "loss": 2.5105,
      "step": 3255
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.911205530166626,
      "learning_rate": 2.4443823394349834e-05,
      "loss": 2.4586,
      "step": 3260
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9256761074066162,
      "learning_rate": 2.430067227589088e-05,
      "loss": 2.4955,
      "step": 3265
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9415754675865173,
      "learning_rate": 2.4157806873658517e-05,
      "loss": 2.4188,
      "step": 3270
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9153307676315308,
      "learning_rate": 2.401522877598087e-05,
      "loss": 2.4196,
      "step": 3275
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8939723372459412,
      "learning_rate": 2.3872939567991827e-05,
      "loss": 2.421,
      "step": 3280
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9284604787826538,
      "learning_rate": 2.373094083161353e-05,
      "loss": 2.4798,
      "step": 3285
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8872914910316467,
      "learning_rate": 2.358923414553877e-05,
      "loss": 2.474,
      "step": 3290
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8954537510871887,
      "learning_rate": 2.3447821085213405e-05,
      "loss": 2.5011,
      "step": 3295
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8992528319358826,
      "learning_rate": 2.3306703222818878e-05,
      "loss": 2.4536,
      "step": 3300
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.922997236251831,
      "learning_rate": 2.3165882127254705e-05,
      "loss": 2.4254,
      "step": 3305
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9199411869049072,
      "learning_rate": 2.302535936412108e-05,
      "loss": 2.4285,
      "step": 3310
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9020093679428101,
      "learning_rate": 2.2885136495701415e-05,
      "loss": 2.4685,
      "step": 3315
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.894803524017334,
      "learning_rate": 2.274521508094501e-05,
      "loss": 2.3846,
      "step": 3320
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9010105133056641,
      "learning_rate": 2.2605596675449698e-05,
      "loss": 2.5001,
      "step": 3325
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8836818933486938,
      "learning_rate": 2.246628283144457e-05,
      "loss": 2.4844,
      "step": 3330
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.957069456577301,
      "learning_rate": 2.232727509777269e-05,
      "loss": 2.4861,
      "step": 3335
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9847378730773926,
      "learning_rate": 2.2188575019873932e-05,
      "loss": 2.4528,
      "step": 3340
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9106289744377136,
      "learning_rate": 2.2050184139767704e-05,
      "loss": 2.4389,
      "step": 3345
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9224086403846741,
      "learning_rate": 2.191210399603591e-05,
      "loss": 2.4644,
      "step": 3350
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9473267197608948,
      "learning_rate": 2.1774336123805772e-05,
      "loss": 2.3514,
      "step": 3355
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9258838295936584,
      "learning_rate": 2.1636882054732776e-05,
      "loss": 2.461,
      "step": 3360
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9247808456420898,
      "learning_rate": 2.1499743316983684e-05,
      "loss": 2.3768,
      "step": 3365
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9250257015228271,
      "learning_rate": 2.1362921435219473e-05,
      "loss": 2.4103,
      "step": 3370
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.917296290397644,
      "learning_rate": 2.1226417930578464e-05,
      "loss": 2.4581,
      "step": 3375
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9212003350257874,
      "learning_rate": 2.109023432065935e-05,
      "loss": 2.4713,
      "step": 3380
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9421318173408508,
      "learning_rate": 2.095437211950434e-05,
      "loss": 2.3969,
      "step": 3385
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.927157461643219,
      "learning_rate": 2.0818832837582352e-05,
      "loss": 2.4147,
      "step": 3390
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9017496705055237,
      "learning_rate": 2.068361798177218e-05,
      "loss": 2.4576,
      "step": 3395
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9253119826316833,
      "learning_rate": 2.0548729055345778e-05,
      "loss": 2.4253,
      "step": 3400
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9048140645027161,
      "learning_rate": 2.0414167557951514e-05,
      "loss": 2.4676,
      "step": 3405
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9291346073150635,
      "learning_rate": 2.0279934985597527e-05,
      "loss": 2.4913,
      "step": 3410
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9075619578361511,
      "learning_rate": 2.0146032830635054e-05,
      "loss": 2.4127,
      "step": 3415
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9002858400344849,
      "learning_rate": 2.001246258174192e-05,
      "loss": 2.4856,
      "step": 3420
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8779779672622681,
      "learning_rate": 1.9879225723905886e-05,
      "loss": 2.4432,
      "step": 3425
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9209660291671753,
      "learning_rate": 1.9746323738408203e-05,
      "loss": 2.4792,
      "step": 3430
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9068074226379395,
      "learning_rate": 1.9613758102807117e-05,
      "loss": 2.467,
      "step": 3435
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8899487853050232,
      "learning_rate": 1.9481530290921474e-05,
      "loss": 2.3785,
      "step": 3440
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9144604802131653,
      "learning_rate": 1.934964177281428e-05,
      "loss": 2.434,
      "step": 3445
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9262046217918396,
      "learning_rate": 1.9218094014776434e-05,
      "loss": 2.4116,
      "step": 3450
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9320298433303833,
      "learning_rate": 1.9086888479310333e-05,
      "loss": 2.4139,
      "step": 3455
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9386606812477112,
      "learning_rate": 1.895602662511371e-05,
      "loss": 2.4334,
      "step": 3460
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9101786613464355,
      "learning_rate": 1.8825509907063327e-05,
      "loss": 2.4237,
      "step": 3465
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9514603614807129,
      "learning_rate": 1.8695339776198872e-05,
      "loss": 2.4838,
      "step": 3470
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9128882884979248,
      "learning_rate": 1.8565517679706783e-05,
      "loss": 2.4205,
      "step": 3475
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9061757922172546,
      "learning_rate": 1.8436045060904174e-05,
      "loss": 2.4656,
      "step": 3480
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9159146547317505,
      "learning_rate": 1.830692335922279e-05,
      "loss": 2.4476,
      "step": 3485
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9187010526657104,
      "learning_rate": 1.8178154010192994e-05,
      "loss": 2.4313,
      "step": 3490
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9203282594680786,
      "learning_rate": 1.8049738445427822e-05,
      "loss": 2.4357,
      "step": 3495
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8951599597930908,
      "learning_rate": 1.7921678092607052e-05,
      "loss": 2.493,
      "step": 3500
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9371747374534607,
      "learning_rate": 1.7793974375461352e-05,
      "loss": 2.4493,
      "step": 3505
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8991261124610901,
      "learning_rate": 1.7666628713756417e-05,
      "loss": 2.4306,
      "step": 3510
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9095786809921265,
      "learning_rate": 1.7539642523277228e-05,
      "loss": 2.4822,
      "step": 3515
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9121299386024475,
      "learning_rate": 1.7413017215812273e-05,
      "loss": 2.413,
      "step": 3520
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9428772926330566,
      "learning_rate": 1.728675419913788e-05,
      "loss": 2.4214,
      "step": 3525
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9310411214828491,
      "learning_rate": 1.716085487700253e-05,
      "loss": 2.4111,
      "step": 3530
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8934375047683716,
      "learning_rate": 1.703532064911131e-05,
      "loss": 2.4061,
      "step": 3535
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9108890891075134,
      "learning_rate": 1.6910152911110283e-05,
      "loss": 2.4731,
      "step": 3540
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9230520725250244,
      "learning_rate": 1.6785353054571024e-05,
      "loss": 2.3733,
      "step": 3545
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9198956489562988,
      "learning_rate": 1.666092246697512e-05,
      "loss": 2.5013,
      "step": 3550
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9111266136169434,
      "learning_rate": 1.6536862531698766e-05,
      "loss": 2.4381,
      "step": 3555
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9198375940322876,
      "learning_rate": 1.6413174627997328e-05,
      "loss": 2.5059,
      "step": 3560
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9285071492195129,
      "learning_rate": 1.6289860130990147e-05,
      "loss": 2.4141,
      "step": 3565
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9231954216957092,
      "learning_rate": 1.6166920411645064e-05,
      "loss": 2.4381,
      "step": 3570
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8866026401519775,
      "learning_rate": 1.6044356836763315e-05,
      "loss": 2.3406,
      "step": 3575
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9078811407089233,
      "learning_rate": 1.5922170768964285e-05,
      "loss": 2.4162,
      "step": 3580
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9162793755531311,
      "learning_rate": 1.5800363566670362e-05,
      "loss": 2.4726,
      "step": 3585
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8958753347396851,
      "learning_rate": 1.5678936584091852e-05,
      "loss": 2.4159,
      "step": 3590
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9262574315071106,
      "learning_rate": 1.5557891171211892e-05,
      "loss": 2.4923,
      "step": 3595
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9506419897079468,
      "learning_rate": 1.5437228673771465e-05,
      "loss": 2.4013,
      "step": 3600
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9091145992279053,
      "learning_rate": 1.5316950433254445e-05,
      "loss": 2.3992,
      "step": 3605
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9276121854782104,
      "learning_rate": 1.5197057786872649e-05,
      "loss": 2.4998,
      "step": 3610
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9171314239501953,
      "learning_rate": 1.5077552067551015e-05,
      "loss": 2.4324,
      "step": 3615
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9585117697715759,
      "learning_rate": 1.4958434603912747e-05,
      "loss": 2.5077,
      "step": 3620
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9010992050170898,
      "learning_rate": 1.4839706720264546e-05,
      "loss": 2.5017,
      "step": 3625
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9106521606445312,
      "learning_rate": 1.4721369736581924e-05,
      "loss": 2.3687,
      "step": 3630
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9210944771766663,
      "learning_rate": 1.4603424968494484e-05,
      "loss": 2.4641,
      "step": 3635
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9032136797904968,
      "learning_rate": 1.448587372727132e-05,
      "loss": 2.4258,
      "step": 3640
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9011669158935547,
      "learning_rate": 1.4368717319806419e-05,
      "loss": 2.4067,
      "step": 3645
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9185218214988708,
      "learning_rate": 1.4251957048604152e-05,
      "loss": 2.4289,
      "step": 3650
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8981814980506897,
      "learning_rate": 1.413559421176479e-05,
      "loss": 2.4876,
      "step": 3655
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9443618655204773,
      "learning_rate": 1.4019630102970056e-05,
      "loss": 2.5391,
      "step": 3660
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9296389818191528,
      "learning_rate": 1.3904066011468753e-05,
      "loss": 2.4167,
      "step": 3665
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9274140000343323,
      "learning_rate": 1.3788903222062433e-05,
      "loss": 2.4408,
      "step": 3670
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9020739793777466,
      "learning_rate": 1.3674143015091118e-05,
      "loss": 2.475,
      "step": 3675
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9183312058448792,
      "learning_rate": 1.355978666641905e-05,
      "loss": 2.4201,
      "step": 3680
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9057744145393372,
      "learning_rate": 1.3445835447420507e-05,
      "loss": 2.4336,
      "step": 3685
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8912809491157532,
      "learning_rate": 1.3332290624965688e-05,
      "loss": 2.4268,
      "step": 3690
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9013752937316895,
      "learning_rate": 1.3219153461406609e-05,
      "loss": 2.4332,
      "step": 3695
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9307184815406799,
      "learning_rate": 1.3106425214563078e-05,
      "loss": 2.3963,
      "step": 3700
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9287288784980774,
      "learning_rate": 1.2994107137708716e-05,
      "loss": 2.5023,
      "step": 3705
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9417144656181335,
      "learning_rate": 1.2882200479556988e-05,
      "loss": 2.4823,
      "step": 3710
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9298997521400452,
      "learning_rate": 1.2770706484247397e-05,
      "loss": 2.456,
      "step": 3715
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9269828796386719,
      "learning_rate": 1.2659626391331564e-05,
      "loss": 2.4669,
      "step": 3720
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9162706136703491,
      "learning_rate": 1.2548961435759493e-05,
      "loss": 2.4185,
      "step": 3725
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9016087055206299,
      "learning_rate": 1.2438712847865846e-05,
      "loss": 2.4509,
      "step": 3730
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9222176671028137,
      "learning_rate": 1.2328881853356244e-05,
      "loss": 2.4106,
      "step": 3735
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8969464898109436,
      "learning_rate": 1.221946967329365e-05,
      "loss": 2.4994,
      "step": 3740
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9177476763725281,
      "learning_rate": 1.2110477524084796e-05,
      "loss": 2.3859,
      "step": 3745
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9398115277290344,
      "learning_rate": 1.2001906617466657e-05,
      "loss": 2.3963,
      "step": 3750
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9646539092063904,
      "learning_rate": 1.1893758160492978e-05,
      "loss": 2.3666,
      "step": 3755
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9097631573677063,
      "learning_rate": 1.1786033355520859e-05,
      "loss": 2.3711,
      "step": 3760
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8921908140182495,
      "learning_rate": 1.1678733400197373e-05,
      "loss": 2.4005,
      "step": 3765
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.913013756275177,
      "learning_rate": 1.1571859487446263e-05,
      "loss": 2.4578,
      "step": 3770
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9429560303688049,
      "learning_rate": 1.1465412805454695e-05,
      "loss": 2.413,
      "step": 3775
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9588865637779236,
      "learning_rate": 1.1359394537660011e-05,
      "loss": 2.4039,
      "step": 3780
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9314767122268677,
      "learning_rate": 1.125380586273661e-05,
      "loss": 2.4643,
      "step": 3785
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9301546216011047,
      "learning_rate": 1.1148647954582808e-05,
      "loss": 2.3927,
      "step": 3790
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9069781303405762,
      "learning_rate": 1.1043921982307819e-05,
      "loss": 2.4396,
      "step": 3795
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9067623615264893,
      "learning_rate": 1.0939629110218735e-05,
      "loss": 2.4054,
      "step": 3800
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9323053956031799,
      "learning_rate": 1.0835770497807596e-05,
      "loss": 2.4356,
      "step": 3805
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9316867589950562,
      "learning_rate": 1.0732347299738493e-05,
      "loss": 2.3777,
      "step": 3810
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9224940538406372,
      "learning_rate": 1.0629360665834732e-05,
      "loss": 2.3806,
      "step": 3815
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9173339605331421,
      "learning_rate": 1.052681174106604e-05,
      "loss": 2.405,
      "step": 3820
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8999264240264893,
      "learning_rate": 1.0424701665535852e-05,
      "loss": 2.4004,
      "step": 3825
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8964149355888367,
      "learning_rate": 1.0323031574468638e-05,
      "loss": 2.4523,
      "step": 3830
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9076893925666809,
      "learning_rate": 1.0221802598197261e-05,
      "loss": 2.3863,
      "step": 3835
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9315918684005737,
      "learning_rate": 1.0121015862150423e-05,
      "loss": 2.3959,
      "step": 3840
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.938804030418396,
      "learning_rate": 1.0020672486840154e-05,
      "loss": 2.3878,
      "step": 3845
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.908531665802002,
      "learning_rate": 9.920773587849364e-06,
      "loss": 2.4283,
      "step": 3850
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9469609260559082,
      "learning_rate": 9.821320275819401e-06,
      "loss": 2.4166,
      "step": 3855
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9202249646186829,
      "learning_rate": 9.72231365643777e-06,
      "loss": 2.4384,
      "step": 3860
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9061878323554993,
      "learning_rate": 9.623754830425779e-06,
      "loss": 2.4621,
      "step": 3865
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9435011744499207,
      "learning_rate": 9.52564489352632e-06,
      "loss": 2.4808,
      "step": 3870
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8984233736991882,
      "learning_rate": 9.427984936491702e-06,
      "loss": 2.4471,
      "step": 3875
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9315876364707947,
      "learning_rate": 9.330776045071509e-06,
      "loss": 2.3882,
      "step": 3880
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9606491327285767,
      "learning_rate": 9.23401930000054e-06,
      "loss": 2.4099,
      "step": 3885
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9043254852294922,
      "learning_rate": 9.137715776986772e-06,
      "loss": 2.3666,
      "step": 3890
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9284046292304993,
      "learning_rate": 9.041866546699434e-06,
      "loss": 2.4225,
      "step": 3895
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9200391173362732,
      "learning_rate": 8.946472674757078e-06,
      "loss": 2.384,
      "step": 3900
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9081897139549255,
      "learning_rate": 8.851535221715735e-06,
      "loss": 2.4135,
      "step": 3905
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8964848518371582,
      "learning_rate": 8.757055243057132e-06,
      "loss": 2.3975,
      "step": 3910
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9136962890625,
      "learning_rate": 8.663033789176967e-06,
      "loss": 2.4181,
      "step": 3915
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9067717790603638,
      "learning_rate": 8.5694719053732e-06,
      "loss": 2.4189,
      "step": 3920
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.91167813539505,
      "learning_rate": 8.476370631834458e-06,
      "loss": 2.421,
      "step": 3925
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8846667408943176,
      "learning_rate": 8.383731003628452e-06,
      "loss": 2.4894,
      "step": 3930
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9231117963790894,
      "learning_rate": 8.291554050690508e-06,
      "loss": 2.5637,
      "step": 3935
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8884219527244568,
      "learning_rate": 8.199840797812058e-06,
      "loss": 2.411,
      "step": 3940
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9300855994224548,
      "learning_rate": 8.108592264629295e-06,
      "loss": 2.3647,
      "step": 3945
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9004824161529541,
      "learning_rate": 8.017809465611803e-06,
      "loss": 2.4171,
      "step": 3950
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9084776043891907,
      "learning_rate": 7.927493410051324e-06,
      "loss": 2.3586,
      "step": 3955
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9107391834259033,
      "learning_rate": 7.837645102050473e-06,
      "loss": 2.3917,
      "step": 3960
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9120099544525146,
      "learning_rate": 7.748265540511635e-06,
      "loss": 2.3377,
      "step": 3965
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9412475228309631,
      "learning_rate": 7.65935571912582e-06,
      "loss": 2.4913,
      "step": 3970
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9313330054283142,
      "learning_rate": 7.5709166263616405e-06,
      "loss": 2.435,
      "step": 3975
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9151592254638672,
      "learning_rate": 7.482949245454302e-06,
      "loss": 2.419,
      "step": 3980
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9209585189819336,
      "learning_rate": 7.3954545543946876e-06,
      "loss": 2.4214,
      "step": 3985
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8722422122955322,
      "learning_rate": 7.308433525918468e-06,
      "loss": 2.3465,
      "step": 3990
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9042191505432129,
      "learning_rate": 7.221887127495313e-06,
      "loss": 2.4304,
      "step": 3995
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9069880843162537,
      "learning_rate": 7.1358163213181114e-06,
      "loss": 2.3537,
      "step": 4000
    }
  ],
  "logging_steps": 5,
  "max_steps": 4811,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "total_flos": 1.1658574908358656e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}