{
  "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": 31.813573837280273,
      "learning_rate": 5e-06,
      "loss": 9.1179,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 27.920026779174805,
      "learning_rate": 1e-05,
      "loss": 8.6958,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 153.7402801513672,
      "learning_rate": 1.5e-05,
      "loss": 8.1658,
      "step": 15
    },
    {
      "epoch": 0.0,
      "grad_norm": 7.882206916809082,
      "learning_rate": 2e-05,
      "loss": 7.78,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.389701843261719,
      "learning_rate": 2.5e-05,
      "loss": 7.5562,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.589534759521484,
      "learning_rate": 3e-05,
      "loss": 7.4236,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.589334487915039,
      "learning_rate": 3.5e-05,
      "loss": 7.208,
      "step": 35
    },
    {
      "epoch": 0.01,
      "grad_norm": 53.45998001098633,
      "learning_rate": 4e-05,
      "loss": 6.9417,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 6.914029598236084,
      "learning_rate": 4.5e-05,
      "loss": 6.6974,
      "step": 45
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.271754264831543,
      "learning_rate": 5e-05,
      "loss": 6.4774,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 50.406898498535156,
      "learning_rate": 5.500000000000001e-05,
      "loss": 6.2127,
      "step": 55
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.283674240112305,
      "learning_rate": 6e-05,
      "loss": 6.1023,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.251434326171875,
      "learning_rate": 6.500000000000001e-05,
      "loss": 5.8973,
      "step": 65
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.821683406829834,
      "learning_rate": 7e-05,
      "loss": 5.6945,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.5881142616271973,
      "learning_rate": 7.500000000000001e-05,
      "loss": 5.4818,
      "step": 75
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.2939279079437256,
      "learning_rate": 8e-05,
      "loss": 5.3125,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.72703218460083,
      "learning_rate": 8.5e-05,
      "loss": 5.1715,
      "step": 85
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.7567694187164307,
      "learning_rate": 9e-05,
      "loss": 4.9456,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.2102720737457275,
      "learning_rate": 9.5e-05,
      "loss": 4.8999,
      "step": 95
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.9953739643096924,
      "learning_rate": 0.0001,
      "loss": 4.699,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.5861263275146484,
      "learning_rate": 9.999972205865686e-05,
      "loss": 4.58,
      "step": 105
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.5365688800811768,
      "learning_rate": 9.999888823771751e-05,
      "loss": 4.4493,
      "step": 110
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.0703377723693848,
      "learning_rate": 9.999749854645204e-05,
      "loss": 4.4265,
      "step": 115
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.441176414489746,
      "learning_rate": 9.99955530003106e-05,
      "loss": 4.3361,
      "step": 120
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.3856356143951416,
      "learning_rate": 9.99930516209231e-05,
      "loss": 4.3491,
      "step": 125
    },
    {
      "epoch": 0.03,
      "grad_norm": 55.84754943847656,
      "learning_rate": 9.998999443609897e-05,
      "loss": 4.1683,
      "step": 130
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.092285633087158,
      "learning_rate": 9.998638147982696e-05,
      "loss": 4.1626,
      "step": 135
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.619375228881836,
      "learning_rate": 9.998221279227467e-05,
      "loss": 4.0489,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.414034843444824,
      "learning_rate": 9.997748841978812e-05,
      "loss": 3.9887,
      "step": 145
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.347266435623169,
      "learning_rate": 9.997220841489122e-05,
      "loss": 3.8839,
      "step": 150
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.6054680347442627,
      "learning_rate": 9.996637283628528e-05,
      "loss": 3.8374,
      "step": 155
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.4477622509002686,
      "learning_rate": 9.995998174884821e-05,
      "loss": 3.7885,
      "step": 160
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.374764919281006,
      "learning_rate": 9.995303522363394e-05,
      "loss": 3.6678,
      "step": 165
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.28847599029541,
      "learning_rate": 9.99455333378715e-05,
      "loss": 3.7301,
      "step": 170
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.0600550174713135,
      "learning_rate": 9.993747617496428e-05,
      "loss": 3.6988,
      "step": 175
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.428739309310913,
      "learning_rate": 9.9928863824489e-05,
      "loss": 3.6342,
      "step": 180
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.138685941696167,
      "learning_rate": 9.99196963821948e-05,
      "loss": 3.507,
      "step": 185
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.2313568592071533,
      "learning_rate": 9.990997395000217e-05,
      "loss": 3.5209,
      "step": 190
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.0725700855255127,
      "learning_rate": 9.989969663600169e-05,
      "loss": 3.4709,
      "step": 195
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.8447208404541016,
      "learning_rate": 9.9888864554453e-05,
      "loss": 3.4876,
      "step": 200
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.315260410308838,
      "learning_rate": 9.987747782578342e-05,
      "loss": 3.5003,
      "step": 205
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.2463152408599854,
      "learning_rate": 9.986553657658668e-05,
      "loss": 3.4555,
      "step": 210
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.66461181640625,
      "learning_rate": 9.985304093962145e-05,
      "loss": 3.3826,
      "step": 215
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.163248062133789,
      "learning_rate": 9.983999105380988e-05,
      "loss": 3.329,
      "step": 220
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.7948611974716187,
      "learning_rate": 9.982638706423608e-05,
      "loss": 3.3666,
      "step": 225
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.7669121026992798,
      "learning_rate": 9.98122291221445e-05,
      "loss": 3.3379,
      "step": 230
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.7311700582504272,
      "learning_rate": 9.979751738493826e-05,
      "loss": 3.3163,
      "step": 235
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.7457441091537476,
      "learning_rate": 9.978225201617732e-05,
      "loss": 3.292,
      "step": 240
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.000087738037109,
      "learning_rate": 9.976643318557678e-05,
      "loss": 3.2588,
      "step": 245
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.4156339168548584,
      "learning_rate": 9.975006106900495e-05,
      "loss": 3.244,
      "step": 250
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.8009788990020752,
      "learning_rate": 9.973313584848132e-05,
      "loss": 3.2657,
      "step": 255
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.0066516399383545,
      "learning_rate": 9.971565771217464e-05,
      "loss": 3.1446,
      "step": 260
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.2172601222991943,
      "learning_rate": 9.969762685440076e-05,
      "loss": 3.3102,
      "step": 265
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.6497077941894531,
      "learning_rate": 9.967904347562054e-05,
      "loss": 3.1381,
      "step": 270
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.7990940809249878,
      "learning_rate": 9.965990778243755e-05,
      "loss": 3.2215,
      "step": 275
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.077507734298706,
      "learning_rate": 9.964021998759577e-05,
      "loss": 3.2076,
      "step": 280
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.731491208076477,
      "learning_rate": 9.961998030997733e-05,
      "loss": 3.273,
      "step": 285
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.7491103410720825,
      "learning_rate": 9.95991889745999e-05,
      "loss": 3.2319,
      "step": 290
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.697852373123169,
      "learning_rate": 9.957784621261441e-05,
      "loss": 3.1617,
      "step": 295
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.6772892475128174,
      "learning_rate": 9.955595226130226e-05,
      "loss": 3.0351,
      "step": 300
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.686403512954712,
      "learning_rate": 9.953350736407282e-05,
      "loss": 3.1364,
      "step": 305
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.605745553970337,
      "learning_rate": 9.951051177046069e-05,
      "loss": 3.1433,
      "step": 310
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.5012344121932983,
      "learning_rate": 9.948696573612292e-05,
      "loss": 3.1276,
      "step": 315
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.677473783493042,
      "learning_rate": 9.946286952283618e-05,
      "loss": 3.1291,
      "step": 320
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.7051786184310913,
      "learning_rate": 9.943822339849381e-05,
      "loss": 3.0843,
      "step": 325
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.4949860572814941,
      "learning_rate": 9.941302763710288e-05,
      "loss": 3.0474,
      "step": 330
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.5325379371643066,
      "learning_rate": 9.938728251878116e-05,
      "loss": 3.1196,
      "step": 335
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.5314396619796753,
      "learning_rate": 9.936098832975393e-05,
      "loss": 3.0491,
      "step": 340
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.7919921875,
      "learning_rate": 9.933414536235091e-05,
      "loss": 2.9929,
      "step": 345
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.5717966556549072,
      "learning_rate": 9.93067539150029e-05,
      "loss": 3.0304,
      "step": 350
    },
    {
      "epoch": 0.07,
      "grad_norm": 4.389132976531982,
      "learning_rate": 9.927881429223853e-05,
      "loss": 3.0005,
      "step": 355
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.6020395755767822,
      "learning_rate": 9.925032680468085e-05,
      "loss": 3.071,
      "step": 360
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.7732460498809814,
      "learning_rate": 9.922129176904388e-05,
      "loss": 3.0712,
      "step": 365
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.502739667892456,
      "learning_rate": 9.919170950812911e-05,
      "loss": 3.0481,
      "step": 370
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.4535770416259766,
      "learning_rate": 9.916158035082184e-05,
      "loss": 3.0377,
      "step": 375
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.457908272743225,
      "learning_rate": 9.913090463208763e-05,
      "loss": 3.0151,
      "step": 380
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.3857684135437012,
      "learning_rate": 9.90996826929685e-05,
      "loss": 3.0602,
      "step": 385
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.4621527194976807,
      "learning_rate": 9.906791488057916e-05,
      "loss": 2.9867,
      "step": 390
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.4058239459991455,
      "learning_rate": 9.903560154810313e-05,
      "loss": 2.9423,
      "step": 395
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.3318129777908325,
      "learning_rate": 9.900274305478887e-05,
      "loss": 2.9254,
      "step": 400
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.401201844215393,
      "learning_rate": 9.896933976594572e-05,
      "loss": 3.0216,
      "step": 405
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.4442485570907593,
      "learning_rate": 9.893539205293989e-05,
      "loss": 2.9923,
      "step": 410
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.4190860986709595,
      "learning_rate": 9.890090029319028e-05,
      "loss": 2.9611,
      "step": 415
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.4127558469772339,
      "learning_rate": 9.886586487016433e-05,
      "loss": 2.979,
      "step": 420
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.3861864805221558,
      "learning_rate": 9.883028617337378e-05,
      "loss": 2.9438,
      "step": 425
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.4450087547302246,
      "learning_rate": 9.879416459837022e-05,
      "loss": 2.9116,
      "step": 430
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.407228708267212,
      "learning_rate": 9.875750054674082e-05,
      "loss": 2.9427,
      "step": 435
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.5297348499298096,
      "learning_rate": 9.872029442610382e-05,
      "loss": 2.9826,
      "step": 440
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.481693148612976,
      "learning_rate": 9.8682546650104e-05,
      "loss": 2.9041,
      "step": 445
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.3595296144485474,
      "learning_rate": 9.864425763840802e-05,
      "loss": 2.9036,
      "step": 450
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.3935821056365967,
      "learning_rate": 9.860542781669988e-05,
      "loss": 2.9413,
      "step": 455
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.4735407829284668,
      "learning_rate": 9.85660576166761e-05,
      "loss": 2.973,
      "step": 460
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.3592067956924438,
      "learning_rate": 9.852614747604093e-05,
      "loss": 2.9061,
      "step": 465
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.4290657043457031,
      "learning_rate": 9.848569783850145e-05,
      "loss": 2.9136,
      "step": 470
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.2641561031341553,
      "learning_rate": 9.844470915376278e-05,
      "loss": 2.8821,
      "step": 475
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.3433492183685303,
      "learning_rate": 9.840318187752292e-05,
      "loss": 2.9061,
      "step": 480
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.3268264532089233,
      "learning_rate": 9.836111647146771e-05,
      "loss": 2.9067,
      "step": 485
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.4298189878463745,
      "learning_rate": 9.831851340326577e-05,
      "loss": 2.93,
      "step": 490
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.2761307954788208,
      "learning_rate": 9.82753731465633e-05,
      "loss": 2.8199,
      "step": 495
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.3382200002670288,
      "learning_rate": 9.823169618097871e-05,
      "loss": 2.854,
      "step": 500
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.5193979740142822,
      "learning_rate": 9.81874829920974e-05,
      "loss": 2.8094,
      "step": 505
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2841696739196777,
      "learning_rate": 9.814273407146623e-05,
      "loss": 2.8539,
      "step": 510
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.268577218055725,
      "learning_rate": 9.809744991658829e-05,
      "loss": 2.8192,
      "step": 515
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2177411317825317,
      "learning_rate": 9.805163103091708e-05,
      "loss": 2.8578,
      "step": 520
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.3120864629745483,
      "learning_rate": 9.800527792385112e-05,
      "loss": 2.8267,
      "step": 525
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2718641757965088,
      "learning_rate": 9.79583911107282e-05,
      "loss": 2.9266,
      "step": 530
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.300713300704956,
      "learning_rate": 9.791097111281968e-05,
      "loss": 2.8243,
      "step": 535
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2401148080825806,
      "learning_rate": 9.786301845732467e-05,
      "loss": 2.9235,
      "step": 540
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2712082862854004,
      "learning_rate": 9.781453367736418e-05,
      "loss": 2.8,
      "step": 545
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2234678268432617,
      "learning_rate": 9.776551731197524e-05,
      "loss": 2.8874,
      "step": 550
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2925634384155273,
      "learning_rate": 9.771596990610478e-05,
      "loss": 2.8941,
      "step": 555
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.286523461341858,
      "learning_rate": 9.766589201060372e-05,
      "loss": 2.8159,
      "step": 560
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2230117321014404,
      "learning_rate": 9.761528418222077e-05,
      "loss": 2.8885,
      "step": 565
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2012773752212524,
      "learning_rate": 9.756414698359624e-05,
      "loss": 2.8415,
      "step": 570
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2559715509414673,
      "learning_rate": 9.75124809832558e-05,
      "loss": 2.8911,
      "step": 575
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.252892255783081,
      "learning_rate": 9.746028675560413e-05,
      "loss": 2.8023,
      "step": 580
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2187546491622925,
      "learning_rate": 9.740756488091861e-05,
      "loss": 2.8552,
      "step": 585
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2338809967041016,
      "learning_rate": 9.735431594534277e-05,
      "loss": 2.8293,
      "step": 590
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2111319303512573,
      "learning_rate": 9.730054054087983e-05,
      "loss": 2.7828,
      "step": 595
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.204712152481079,
      "learning_rate": 9.724623926538612e-05,
      "loss": 2.8412,
      "step": 600
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2021050453186035,
      "learning_rate": 9.719141272256443e-05,
      "loss": 2.7744,
      "step": 605
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.1791361570358276,
      "learning_rate": 9.713606152195726e-05,
      "loss": 2.8102,
      "step": 610
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2355018854141235,
      "learning_rate": 9.708018627894011e-05,
      "loss": 2.8044,
      "step": 615
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.9050341844558716,
      "learning_rate": 9.702378761471456e-05,
      "loss": 2.8365,
      "step": 620
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.1793168783187866,
      "learning_rate": 9.696686615630146e-05,
      "loss": 2.7307,
      "step": 625
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.618408441543579,
      "learning_rate": 9.690942253653385e-05,
      "loss": 2.7531,
      "step": 630
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.1792315244674683,
      "learning_rate": 9.685145739405002e-05,
      "loss": 2.8224,
      "step": 635
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2722834348678589,
      "learning_rate": 9.679297137328634e-05,
      "loss": 2.7523,
      "step": 640
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2244315147399902,
      "learning_rate": 9.673396512447013e-05,
      "loss": 2.7402,
      "step": 645
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.217962622642517,
      "learning_rate": 9.667443930361247e-05,
      "loss": 2.7718,
      "step": 650
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1462675333023071,
      "learning_rate": 9.661439457250076e-05,
      "loss": 2.7529,
      "step": 655
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1446068286895752,
      "learning_rate": 9.655383159869158e-05,
      "loss": 2.7839,
      "step": 660
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1892576217651367,
      "learning_rate": 9.649275105550309e-05,
      "loss": 2.7048,
      "step": 665
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1554418802261353,
      "learning_rate": 9.643115362200762e-05,
      "loss": 2.7552,
      "step": 670
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1442291736602783,
      "learning_rate": 9.636903998302409e-05,
      "loss": 2.8037,
      "step": 675
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.131379246711731,
      "learning_rate": 9.630641082911045e-05,
      "loss": 2.6973,
      "step": 680
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1673518419265747,
      "learning_rate": 9.624326685655593e-05,
      "loss": 2.7727,
      "step": 685
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1787452697753906,
      "learning_rate": 9.617960876737337e-05,
      "loss": 2.7412,
      "step": 690
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1861211061477661,
      "learning_rate": 9.611543726929134e-05,
      "loss": 2.7139,
      "step": 695
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2275561094284058,
      "learning_rate": 9.605075307574635e-05,
      "loss": 2.8026,
      "step": 700
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1469069719314575,
      "learning_rate": 9.598555690587487e-05,
      "loss": 2.7468,
      "step": 705
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1670516729354858,
      "learning_rate": 9.591984948450532e-05,
      "loss": 2.7824,
      "step": 710
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1783413887023926,
      "learning_rate": 9.585363154215008e-05,
      "loss": 2.8328,
      "step": 715
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.108250379562378,
      "learning_rate": 9.578690381499728e-05,
      "loss": 2.7159,
      "step": 720
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.202819585800171,
      "learning_rate": 9.571966704490271e-05,
      "loss": 2.7425,
      "step": 725
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1453441381454468,
      "learning_rate": 9.565192197938148e-05,
      "loss": 2.7279,
      "step": 730
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1307199001312256,
      "learning_rate": 9.558366937159977e-05,
      "loss": 2.7687,
      "step": 735
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1222865581512451,
      "learning_rate": 9.551490998036646e-05,
      "loss": 2.7295,
      "step": 740
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.175696849822998,
      "learning_rate": 9.544564457012463e-05,
      "loss": 2.7889,
      "step": 745
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.1392408609390259,
      "learning_rate": 9.537587391094314e-05,
      "loss": 2.7533,
      "step": 750
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.142823576927185,
      "learning_rate": 9.5305598778508e-05,
      "loss": 2.6836,
      "step": 755
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.1527529954910278,
      "learning_rate": 9.52348199541138e-05,
      "loss": 2.7062,
      "step": 760
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.2020232677459717,
      "learning_rate": 9.516353822465504e-05,
      "loss": 2.7576,
      "step": 765
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.1423379182815552,
      "learning_rate": 9.509175438261726e-05,
      "loss": 2.7109,
      "step": 770
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.208016037940979,
      "learning_rate": 9.501946922606838e-05,
      "loss": 2.774,
      "step": 775
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.1515254974365234,
      "learning_rate": 9.494668355864973e-05,
      "loss": 2.7202,
      "step": 780
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.1667124032974243,
      "learning_rate": 9.487339818956716e-05,
      "loss": 2.6929,
      "step": 785
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.1528996229171753,
      "learning_rate": 9.479961393358203e-05,
      "loss": 2.7247,
      "step": 790
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1446163654327393,
      "learning_rate": 9.472533161100215e-05,
      "loss": 2.7139,
      "step": 795
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1616708040237427,
      "learning_rate": 9.465055204767265e-05,
      "loss": 2.7139,
      "step": 800
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.0954192876815796,
      "learning_rate": 9.457527607496685e-05,
      "loss": 2.7367,
      "step": 805
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1313517093658447,
      "learning_rate": 9.44995045297769e-05,
      "loss": 2.7652,
      "step": 810
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.102845311164856,
      "learning_rate": 9.442323825450464e-05,
      "loss": 2.649,
      "step": 815
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1344150304794312,
      "learning_rate": 9.43464780970521e-05,
      "loss": 2.6686,
      "step": 820
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1737732887268066,
      "learning_rate": 9.426922491081212e-05,
      "loss": 2.7243,
      "step": 825
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.5500872135162354,
      "learning_rate": 9.419147955465888e-05,
      "loss": 2.682,
      "step": 830
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1714314222335815,
      "learning_rate": 9.411324289293832e-05,
      "loss": 2.6775,
      "step": 835
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1273174285888672,
      "learning_rate": 9.403451579545859e-05,
      "loss": 2.6827,
      "step": 840
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.0785410404205322,
      "learning_rate": 9.395529913748025e-05,
      "loss": 2.6474,
      "step": 845
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1376752853393555,
      "learning_rate": 9.387559379970672e-05,
      "loss": 2.7519,
      "step": 850
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.923419713973999,
      "learning_rate": 9.379540066827431e-05,
      "loss": 2.6756,
      "step": 855
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1519583463668823,
      "learning_rate": 9.371472063474248e-05,
      "loss": 2.7166,
      "step": 860
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.0922170877456665,
      "learning_rate": 9.363355459608394e-05,
      "loss": 2.6794,
      "step": 865
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2151093482971191,
      "learning_rate": 9.355190345467457e-05,
      "loss": 2.6482,
      "step": 870
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.0864477157592773,
      "learning_rate": 9.346976811828352e-05,
      "loss": 2.6631,
      "step": 875
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1231117248535156,
      "learning_rate": 9.338714950006297e-05,
      "loss": 2.6855,
      "step": 880
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.0890246629714966,
      "learning_rate": 9.330404851853817e-05,
      "loss": 2.7035,
      "step": 885
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1322407722473145,
      "learning_rate": 9.3220466097597e-05,
      "loss": 2.6365,
      "step": 890
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.0793037414550781,
      "learning_rate": 9.313640316647991e-05,
      "loss": 2.7133,
      "step": 895
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.0490481853485107,
      "learning_rate": 9.305186065976945e-05,
      "loss": 2.6949,
      "step": 900
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.1165353059768677,
      "learning_rate": 9.296683951737993e-05,
      "loss": 2.6312,
      "step": 905
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.0708872079849243,
      "learning_rate": 9.288134068454697e-05,
      "loss": 2.6534,
      "step": 910
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.0487841367721558,
      "learning_rate": 9.2795365111817e-05,
      "loss": 2.6838,
      "step": 915
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.0860440731048584,
      "learning_rate": 9.270891375503665e-05,
      "loss": 2.7283,
      "step": 920
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.076627254486084,
      "learning_rate": 9.262198757534218e-05,
      "loss": 2.6791,
      "step": 925
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.0581265687942505,
      "learning_rate": 9.253458753914874e-05,
      "loss": 2.6602,
      "step": 930
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.0864975452423096,
      "learning_rate": 9.244671461813969e-05,
      "loss": 2.6428,
      "step": 935
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.1001484394073486,
      "learning_rate": 9.235836978925572e-05,
      "loss": 2.7087,
      "step": 940
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.130135178565979,
      "learning_rate": 9.226955403468406e-05,
      "loss": 2.6837,
      "step": 945
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.0518337488174438,
      "learning_rate": 9.21802683418475e-05,
      "loss": 2.6547,
      "step": 950
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.0764212608337402,
      "learning_rate": 9.209051370339347e-05,
      "loss": 2.6823,
      "step": 955
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.0864237546920776,
      "learning_rate": 9.200029111718295e-05,
      "loss": 2.6558,
      "step": 960
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.0547573566436768,
      "learning_rate": 9.190960158627941e-05,
      "loss": 2.6308,
      "step": 965
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.0916203260421753,
      "learning_rate": 9.181844611893766e-05,
      "loss": 2.6403,
      "step": 970
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.1068625450134277,
      "learning_rate": 9.172682572859261e-05,
      "loss": 2.6449,
      "step": 975
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.0739907026290894,
      "learning_rate": 9.163474143384806e-05,
      "loss": 2.6834,
      "step": 980
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.0589396953582764,
      "learning_rate": 9.154219425846528e-05,
      "loss": 2.6742,
      "step": 985
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0412026643753052,
      "learning_rate": 9.144918523135175e-05,
      "loss": 2.6834,
      "step": 990
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.122963309288025,
      "learning_rate": 9.13557153865496e-05,
      "loss": 2.6321,
      "step": 995
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.151918888092041,
      "learning_rate": 9.12617857632242e-05,
      "loss": 2.6733,
      "step": 1000
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0950418710708618,
      "learning_rate": 9.116739740565259e-05,
      "loss": 2.6735,
      "step": 1005
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0299748182296753,
      "learning_rate": 9.107255136321184e-05,
      "loss": 2.639,
      "step": 1010
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0754122734069824,
      "learning_rate": 9.09772486903674e-05,
      "loss": 2.6601,
      "step": 1015
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0586258172988892,
      "learning_rate": 9.08814904466614e-05,
      "loss": 2.6266,
      "step": 1020
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0555306673049927,
      "learning_rate": 9.078527769670085e-05,
      "loss": 2.6555,
      "step": 1025
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0900399684906006,
      "learning_rate": 9.068861151014575e-05,
      "loss": 2.6175,
      "step": 1030
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0734572410583496,
      "learning_rate": 9.05914929616973e-05,
      "loss": 2.6134,
      "step": 1035
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.1126459836959839,
      "learning_rate": 9.04939231310859e-05,
      "loss": 2.6445,
      "step": 1040
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0274248123168945,
      "learning_rate": 9.039590310305914e-05,
      "loss": 2.7062,
      "step": 1045
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0258575677871704,
      "learning_rate": 9.029743396736974e-05,
      "loss": 2.5811,
      "step": 1050
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0055421590805054,
      "learning_rate": 9.019851681876348e-05,
      "loss": 2.6982,
      "step": 1055
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0245190858840942,
      "learning_rate": 9.009915275696693e-05,
      "loss": 2.6102,
      "step": 1060
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.5464237928390503,
      "learning_rate": 8.999934288667534e-05,
      "loss": 2.6405,
      "step": 1065
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0202620029449463,
      "learning_rate": 8.989908831754028e-05,
      "loss": 2.5989,
      "step": 1070
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.009238362312317,
      "learning_rate": 8.979839016415735e-05,
      "loss": 2.6014,
      "step": 1075
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.04855215549469,
      "learning_rate": 8.969724954605373e-05,
      "loss": 2.6536,
      "step": 1080
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0804136991500854,
      "learning_rate": 8.959566758767581e-05,
      "loss": 2.6686,
      "step": 1085
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0398406982421875,
      "learning_rate": 8.949364541837661e-05,
      "loss": 2.616,
      "step": 1090
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0304359197616577,
      "learning_rate": 8.939118417240329e-05,
      "loss": 2.6174,
      "step": 1095
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.080388069152832,
      "learning_rate": 8.92882849888845e-05,
      "loss": 2.627,
      "step": 1100
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0364700555801392,
      "learning_rate": 8.918494901181773e-05,
      "loss": 2.6303,
      "step": 1105
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0459529161453247,
      "learning_rate": 8.908117739005659e-05,
      "loss": 2.5842,
      "step": 1110
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0185799598693848,
      "learning_rate": 8.897697127729805e-05,
      "loss": 2.6425,
      "step": 1115
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.9739039540290833,
      "learning_rate": 8.887233183206957e-05,
      "loss": 2.5828,
      "step": 1120
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.059308409690857,
      "learning_rate": 8.876726021771627e-05,
      "loss": 2.5769,
      "step": 1125
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0338994264602661,
      "learning_rate": 8.866175760238798e-05,
      "loss": 2.6329,
      "step": 1130
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0227488279342651,
      "learning_rate": 8.855582515902625e-05,
      "loss": 2.5905,
      "step": 1135
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.9925109148025513,
      "learning_rate": 8.844946406535131e-05,
      "loss": 2.5524,
      "step": 1140
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0282485485076904,
      "learning_rate": 8.834267550384893e-05,
      "loss": 2.581,
      "step": 1145
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0401309728622437,
      "learning_rate": 8.823546066175741e-05,
      "loss": 2.6801,
      "step": 1150
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0400043725967407,
      "learning_rate": 8.81278207310542e-05,
      "loss": 2.6248,
      "step": 1155
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.9869978427886963,
      "learning_rate": 8.801975690844278e-05,
      "loss": 2.682,
      "step": 1160
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0918792486190796,
      "learning_rate": 8.791127039533934e-05,
      "loss": 2.5757,
      "step": 1165
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0104856491088867,
      "learning_rate": 8.780236239785935e-05,
      "loss": 2.58,
      "step": 1170
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.059522271156311,
      "learning_rate": 8.76930341268042e-05,
      "loss": 2.6208,
      "step": 1175
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0344668626785278,
      "learning_rate": 8.758328679764776e-05,
      "loss": 2.6312,
      "step": 1180
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0061168670654297,
      "learning_rate": 8.747312163052284e-05,
      "loss": 2.5952,
      "step": 1185
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0210862159729004,
      "learning_rate": 8.736253985020761e-05,
      "loss": 2.6657,
      "step": 1190
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0098751783370972,
      "learning_rate": 8.725154268611203e-05,
      "loss": 2.6222,
      "step": 1195
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.041806936264038,
      "learning_rate": 8.714013137226411e-05,
      "loss": 2.5781,
      "step": 1200
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0064157247543335,
      "learning_rate": 8.702830714729628e-05,
      "loss": 2.5136,
      "step": 1205
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.9981704354286194,
      "learning_rate": 8.691607125443153e-05,
      "loss": 2.611,
      "step": 1210
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0080243349075317,
      "learning_rate": 8.680342494146967e-05,
      "loss": 2.5375,
      "step": 1215
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.0076395273208618,
      "learning_rate": 8.66903694607734e-05,
      "loss": 2.6269,
      "step": 1220
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.9885242581367493,
      "learning_rate": 8.65769060692544e-05,
      "loss": 2.5712,
      "step": 1225
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.0156185626983643,
      "learning_rate": 8.646303602835936e-05,
      "loss": 2.6381,
      "step": 1230
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.99817955493927,
      "learning_rate": 8.634876060405597e-05,
      "loss": 2.6589,
      "step": 1235
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.9817842841148376,
      "learning_rate": 8.623408106681884e-05,
      "loss": 2.5775,
      "step": 1240
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.9927291870117188,
      "learning_rate": 8.611899869161535e-05,
      "loss": 2.5987,
      "step": 1245
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.006893515586853,
      "learning_rate": 8.600351475789147e-05,
      "loss": 2.565,
      "step": 1250
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1008330583572388,
      "learning_rate": 8.588763054955764e-05,
      "loss": 2.6247,
      "step": 1255
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.0435577630996704,
      "learning_rate": 8.57713473549743e-05,
      "loss": 2.6331,
      "step": 1260
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.0769011974334717,
      "learning_rate": 8.565466646693778e-05,
      "loss": 2.6576,
      "step": 1265
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.0972942113876343,
      "learning_rate": 8.553758918266578e-05,
      "loss": 2.6234,
      "step": 1270
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.964900016784668,
      "learning_rate": 8.5420116803783e-05,
      "loss": 2.5506,
      "step": 1275
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.9919426441192627,
      "learning_rate": 8.530225063630668e-05,
      "loss": 2.5138,
      "step": 1280
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.0192852020263672,
      "learning_rate": 8.518399199063205e-05,
      "loss": 2.6286,
      "step": 1285
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.030393362045288,
      "learning_rate": 8.50653421815178e-05,
      "loss": 2.5331,
      "step": 1290
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.0030856132507324,
      "learning_rate": 8.494630252807138e-05,
      "loss": 2.6121,
      "step": 1295
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.058526635169983,
      "learning_rate": 8.482687435373449e-05,
      "loss": 2.6006,
      "step": 1300
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.9560089111328125,
      "learning_rate": 8.470705898626817e-05,
      "loss": 2.6244,
      "step": 1305
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.9747763872146606,
      "learning_rate": 8.458685775773822e-05,
      "loss": 2.5592,
      "step": 1310
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.0110520124435425,
      "learning_rate": 8.446627200450025e-05,
      "loss": 2.6079,
      "step": 1315
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.9854313731193542,
      "learning_rate": 8.434530306718493e-05,
      "loss": 2.5953,
      "step": 1320
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.98088538646698,
      "learning_rate": 8.4223952290683e-05,
      "loss": 2.598,
      "step": 1325
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.9649909734725952,
      "learning_rate": 8.41022210241304e-05,
      "loss": 2.5269,
      "step": 1330
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0427321195602417,
      "learning_rate": 8.398011062089316e-05,
      "loss": 2.5821,
      "step": 1335
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.9849219918251038,
      "learning_rate": 8.385762243855249e-05,
      "loss": 2.5982,
      "step": 1340
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.015129804611206,
      "learning_rate": 8.373475783888958e-05,
      "loss": 2.5903,
      "step": 1345
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.9560769200325012,
      "learning_rate": 8.36115181878705e-05,
      "loss": 2.5121,
      "step": 1350
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.9451593160629272,
      "learning_rate": 8.348790485563101e-05,
      "loss": 2.605,
      "step": 1355
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.0395113229751587,
      "learning_rate": 8.336391921646134e-05,
      "loss": 2.6172,
      "step": 1360
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.9910452961921692,
      "learning_rate": 8.323956264879089e-05,
      "loss": 2.5662,
      "step": 1365
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.9585237503051758,
      "learning_rate": 8.311483653517294e-05,
      "loss": 2.5895,
      "step": 1370
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.9870867133140564,
      "learning_rate": 8.298974226226919e-05,
      "loss": 2.6034,
      "step": 1375
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.9943922758102417,
      "learning_rate": 8.28642812208345e-05,
      "loss": 2.5244,
      "step": 1380
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.010125756263733,
      "learning_rate": 8.273845480570123e-05,
      "loss": 2.6087,
      "step": 1385
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.9950689077377319,
      "learning_rate": 8.26122644157639e-05,
      "loss": 2.5111,
      "step": 1390
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.0169402360916138,
      "learning_rate": 8.248571145396362e-05,
      "loss": 2.5522,
      "step": 1395
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.9725683927536011,
      "learning_rate": 8.235879732727236e-05,
      "loss": 2.5374,
      "step": 1400
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.9679161310195923,
      "learning_rate": 8.223152344667745e-05,
      "loss": 2.5739,
      "step": 1405
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.9871531128883362,
      "learning_rate": 8.21038912271658e-05,
      "loss": 2.5947,
      "step": 1410
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.9824729561805725,
      "learning_rate": 8.197590208770824e-05,
      "loss": 2.5648,
      "step": 1415
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.9750092625617981,
      "learning_rate": 8.184755745124371e-05,
      "loss": 2.5038,
      "step": 1420
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.9619457721710205,
      "learning_rate": 8.171885874466342e-05,
      "loss": 2.5755,
      "step": 1425
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.993933916091919,
      "learning_rate": 8.158980739879507e-05,
      "loss": 2.5726,
      "step": 1430
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.9767422080039978,
      "learning_rate": 8.146040484838677e-05,
      "loss": 2.6105,
      "step": 1435
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.0110574960708618,
      "learning_rate": 8.133065253209132e-05,
      "loss": 2.5235,
      "step": 1440
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.9937053918838501,
      "learning_rate": 8.120055189245e-05,
      "loss": 2.538,
      "step": 1445
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.9573049545288086,
      "learning_rate": 8.10701043758767e-05,
      "loss": 2.524,
      "step": 1450
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.9540808796882629,
      "learning_rate": 8.093931143264174e-05,
      "loss": 2.5675,
      "step": 1455
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.008131504058838,
      "learning_rate": 8.080817451685576e-05,
      "loss": 2.5015,
      "step": 1460
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.9936932921409607,
      "learning_rate": 8.067669508645356e-05,
      "loss": 2.5818,
      "step": 1465
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9977306723594666,
      "learning_rate": 8.054487460317797e-05,
      "loss": 2.5082,
      "step": 1470
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9737963080406189,
      "learning_rate": 8.041271453256345e-05,
      "loss": 2.5276,
      "step": 1475
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9630906581878662,
      "learning_rate": 8.02802163439199e-05,
      "loss": 2.44,
      "step": 1480
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9698747396469116,
      "learning_rate": 8.01473815103163e-05,
      "loss": 2.5131,
      "step": 1485
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9699703454971313,
      "learning_rate": 8.001421150856434e-05,
      "loss": 2.4736,
      "step": 1490
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9490015506744385,
      "learning_rate": 7.988070781920197e-05,
      "loss": 2.4978,
      "step": 1495
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9748291969299316,
      "learning_rate": 7.9746871926477e-05,
      "loss": 2.5235,
      "step": 1500
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9626737236976624,
      "learning_rate": 7.961270531833052e-05,
      "loss": 2.4812,
      "step": 1505
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9995949864387512,
      "learning_rate": 7.947820948638045e-05,
      "loss": 2.558,
      "step": 1510
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.9817675352096558,
      "learning_rate": 7.934338592590486e-05,
      "loss": 2.5591,
      "step": 1515
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9922047257423401,
      "learning_rate": 7.92082361358254e-05,
      "loss": 2.4919,
      "step": 1520
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9521270394325256,
      "learning_rate": 7.907276161869065e-05,
      "loss": 2.4929,
      "step": 1525
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9504519104957581,
      "learning_rate": 7.893696388065936e-05,
      "loss": 2.4731,
      "step": 1530
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0164660215377808,
      "learning_rate": 7.88008444314838e-05,
      "loss": 2.5618,
      "step": 1535
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9919765591621399,
      "learning_rate": 7.866440478449283e-05,
      "loss": 2.5814,
      "step": 1540
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9793297648429871,
      "learning_rate": 7.852764645657522e-05,
      "loss": 2.5629,
      "step": 1545
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9306464791297913,
      "learning_rate": 7.839057096816271e-05,
      "loss": 2.5159,
      "step": 1550
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9663394689559937,
      "learning_rate": 7.82531798432131e-05,
      "loss": 2.5165,
      "step": 1555
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9504035115242004,
      "learning_rate": 7.811547460919333e-05,
      "loss": 2.5199,
      "step": 1560
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9541032314300537,
      "learning_rate": 7.797745679706254e-05,
      "loss": 2.533,
      "step": 1565
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9573409557342529,
      "learning_rate": 7.783912794125496e-05,
      "loss": 2.5546,
      "step": 1570
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9424445033073425,
      "learning_rate": 7.770048957966291e-05,
      "loss": 2.5258,
      "step": 1575
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9768951535224915,
      "learning_rate": 7.756154325361967e-05,
      "loss": 2.4991,
      "step": 1580
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9881746172904968,
      "learning_rate": 7.74222905078824e-05,
      "loss": 2.5044,
      "step": 1585
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9817689657211304,
      "learning_rate": 7.728273289061489e-05,
      "loss": 2.5459,
      "step": 1590
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9682257771492004,
      "learning_rate": 7.714287195337044e-05,
      "loss": 2.5957,
      "step": 1595
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9901173710823059,
      "learning_rate": 7.700270925107448e-05,
      "loss": 2.4403,
      "step": 1600
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.948213517665863,
      "learning_rate": 7.686224634200742e-05,
      "loss": 2.4842,
      "step": 1605
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9362388849258423,
      "learning_rate": 7.672148478778722e-05,
      "loss": 2.4878,
      "step": 1610
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9562108516693115,
      "learning_rate": 7.658042615335212e-05,
      "loss": 2.4728,
      "step": 1615
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9186726212501526,
      "learning_rate": 7.643907200694318e-05,
      "loss": 2.4542,
      "step": 1620
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9547454714775085,
      "learning_rate": 7.629742392008684e-05,
      "loss": 2.5156,
      "step": 1625
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9697511196136475,
      "learning_rate": 7.615548346757749e-05,
      "loss": 2.5298,
      "step": 1630
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.95559161901474,
      "learning_rate": 7.60132522274599e-05,
      "loss": 2.5401,
      "step": 1635
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9713429808616638,
      "learning_rate": 7.587073178101178e-05,
      "loss": 2.5195,
      "step": 1640
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.9706266522407532,
      "learning_rate": 7.572792371272609e-05,
      "loss": 2.5245,
      "step": 1645
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.978418231010437,
      "learning_rate": 7.55848296102935e-05,
      "loss": 2.4615,
      "step": 1650
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.956576406955719,
      "learning_rate": 7.544145106458465e-05,
      "loss": 2.524,
      "step": 1655
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9353023171424866,
      "learning_rate": 7.529778966963259e-05,
      "loss": 2.3879,
      "step": 1660
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.0076223611831665,
      "learning_rate": 7.515384702261496e-05,
      "loss": 2.491,
      "step": 1665
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9393713474273682,
      "learning_rate": 7.500962472383627e-05,
      "loss": 2.4818,
      "step": 1670
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.957146942615509,
      "learning_rate": 7.486512437671011e-05,
      "loss": 2.5437,
      "step": 1675
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9632449746131897,
      "learning_rate": 7.472034758774128e-05,
      "loss": 2.5207,
      "step": 1680
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9379749894142151,
      "learning_rate": 7.457529596650797e-05,
      "loss": 2.5303,
      "step": 1685
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9668945074081421,
      "learning_rate": 7.442997112564392e-05,
      "loss": 2.5195,
      "step": 1690
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9343632459640503,
      "learning_rate": 7.428437468082037e-05,
      "loss": 2.4692,
      "step": 1695
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9466119408607483,
      "learning_rate": 7.413850825072817e-05,
      "loss": 2.5595,
      "step": 1700
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9706186056137085,
      "learning_rate": 7.39923734570598e-05,
      "loss": 2.5421,
      "step": 1705
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9418413639068604,
      "learning_rate": 7.384597192449126e-05,
      "loss": 2.5056,
      "step": 1710
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9419413805007935,
      "learning_rate": 7.369930528066412e-05,
      "loss": 2.5314,
      "step": 1715
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9597350358963013,
      "learning_rate": 7.355237515616732e-05,
      "loss": 2.4869,
      "step": 1720
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9583315849304199,
      "learning_rate": 7.340518318451914e-05,
      "loss": 2.5011,
      "step": 1725
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9204713702201843,
      "learning_rate": 7.325773100214893e-05,
      "loss": 2.5222,
      "step": 1730
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9271583557128906,
      "learning_rate": 7.311002024837899e-05,
      "loss": 2.5486,
      "step": 1735
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9411045908927917,
      "learning_rate": 7.296205256540633e-05,
      "loss": 2.4855,
      "step": 1740
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9310178756713867,
      "learning_rate": 7.281382959828443e-05,
      "loss": 2.594,
      "step": 1745
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9783688187599182,
      "learning_rate": 7.26653529949049e-05,
      "loss": 2.5684,
      "step": 1750
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9621866345405579,
      "learning_rate": 7.25166244059792e-05,
      "loss": 2.441,
      "step": 1755
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9207419753074646,
      "learning_rate": 7.236764548502029e-05,
      "loss": 2.4502,
      "step": 1760
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9349749684333801,
      "learning_rate": 7.221841788832421e-05,
      "loss": 2.5613,
      "step": 1765
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9797950387001038,
      "learning_rate": 7.206894327495173e-05,
      "loss": 2.4534,
      "step": 1770
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9534410834312439,
      "learning_rate": 7.191922330670982e-05,
      "loss": 2.4731,
      "step": 1775
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.939933717250824,
      "learning_rate": 7.176925964813326e-05,
      "loss": 2.4995,
      "step": 1780
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9610041379928589,
      "learning_rate": 7.161905396646607e-05,
      "loss": 2.4542,
      "step": 1785
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9338921904563904,
      "learning_rate": 7.146860793164299e-05,
      "loss": 2.504,
      "step": 1790
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.918339729309082,
      "learning_rate": 7.131792321627098e-05,
      "loss": 2.4966,
      "step": 1795
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9426106214523315,
      "learning_rate": 7.116700149561048e-05,
      "loss": 2.5555,
      "step": 1800
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9430469870567322,
      "learning_rate": 7.101584444755696e-05,
      "loss": 2.5065,
      "step": 1805
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9407751560211182,
      "learning_rate": 7.086445375262212e-05,
      "loss": 2.4893,
      "step": 1810
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9642773866653442,
      "learning_rate": 7.071283109391528e-05,
      "loss": 2.467,
      "step": 1815
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9692126512527466,
      "learning_rate": 7.056097815712466e-05,
      "loss": 2.5445,
      "step": 1820
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9674531817436218,
      "learning_rate": 7.040889663049862e-05,
      "loss": 2.4867,
      "step": 1825
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9319456219673157,
      "learning_rate": 7.025658820482693e-05,
      "loss": 2.4393,
      "step": 1830
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9372115135192871,
      "learning_rate": 7.010405457342192e-05,
      "loss": 2.5165,
      "step": 1835
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9726076126098633,
      "learning_rate": 6.995129743209967e-05,
      "loss": 2.4505,
      "step": 1840
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9362107515335083,
      "learning_rate": 6.97983184791612e-05,
      "loss": 2.4647,
      "step": 1845
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9178482890129089,
      "learning_rate": 6.964511941537355e-05,
      "loss": 2.4714,
      "step": 1850
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9568119645118713,
      "learning_rate": 6.949170194395083e-05,
      "loss": 2.5276,
      "step": 1855
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9537916779518127,
      "learning_rate": 6.933806777053536e-05,
      "loss": 2.4758,
      "step": 1860
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9236088991165161,
      "learning_rate": 6.918421860317872e-05,
      "loss": 2.4279,
      "step": 1865
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9248093962669373,
      "learning_rate": 6.903015615232263e-05,
      "loss": 2.4253,
      "step": 1870
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9502087831497192,
      "learning_rate": 6.887588213078012e-05,
      "loss": 2.4653,
      "step": 1875
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9479555487632751,
      "learning_rate": 6.87213982537163e-05,
      "loss": 2.4596,
      "step": 1880
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9418471455574036,
      "learning_rate": 6.856670623862943e-05,
      "loss": 2.5136,
      "step": 1885
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9373717308044434,
      "learning_rate": 6.841180780533179e-05,
      "loss": 2.4837,
      "step": 1890
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9609478116035461,
      "learning_rate": 6.82567046759305e-05,
      "loss": 2.422,
      "step": 1895
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9436773061752319,
      "learning_rate": 6.810139857480844e-05,
      "loss": 2.4756,
      "step": 1900
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.947167694568634,
      "learning_rate": 6.794589122860509e-05,
      "loss": 2.405,
      "step": 1905
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9732680916786194,
      "learning_rate": 6.779018436619725e-05,
      "loss": 2.4835,
      "step": 1910
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.950404703617096,
      "learning_rate": 6.763427971867992e-05,
      "loss": 2.5281,
      "step": 1915
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9311380982398987,
      "learning_rate": 6.747817901934699e-05,
      "loss": 2.4992,
      "step": 1920
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9213589429855347,
      "learning_rate": 6.732188400367197e-05,
      "loss": 2.4271,
      "step": 1925
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9262340664863586,
      "learning_rate": 6.716539640928871e-05,
      "loss": 2.4826,
      "step": 1930
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9462832808494568,
      "learning_rate": 6.70087179759721e-05,
      "loss": 2.4917,
      "step": 1935
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9576349854469299,
      "learning_rate": 6.685185044561874e-05,
      "loss": 2.4312,
      "step": 1940
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9459937810897827,
      "learning_rate": 6.669479556222747e-05,
      "loss": 2.5343,
      "step": 1945
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9218082427978516,
      "learning_rate": 6.653755507188013e-05,
      "loss": 2.5057,
      "step": 1950
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.950690507888794,
      "learning_rate": 6.638013072272205e-05,
      "loss": 2.4339,
      "step": 1955
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9415127635002136,
      "learning_rate": 6.622252426494259e-05,
      "loss": 2.4953,
      "step": 1960
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.898391604423523,
      "learning_rate": 6.606473745075581e-05,
      "loss": 2.4518,
      "step": 1965
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9288300275802612,
      "learning_rate": 6.590677203438084e-05,
      "loss": 2.4957,
      "step": 1970
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.92172771692276,
      "learning_rate": 6.574862977202252e-05,
      "loss": 2.4942,
      "step": 1975
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9391177892684937,
      "learning_rate": 6.559031242185174e-05,
      "loss": 2.5103,
      "step": 1980
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9494156837463379,
      "learning_rate": 6.543182174398597e-05,
      "loss": 2.4935,
      "step": 1985
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9366905689239502,
      "learning_rate": 6.52731595004697e-05,
      "loss": 2.5083,
      "step": 1990
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9476346373558044,
      "learning_rate": 6.51143274552548e-05,
      "loss": 2.4404,
      "step": 1995
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9403136968612671,
      "learning_rate": 6.495532737418098e-05,
      "loss": 2.4399,
      "step": 2000
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9438544511795044,
      "learning_rate": 6.479616102495605e-05,
      "loss": 2.4434,
      "step": 2005
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9154673218727112,
      "learning_rate": 6.463683017713638e-05,
      "loss": 2.4772,
      "step": 2010
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.925992488861084,
      "learning_rate": 6.447733660210715e-05,
      "loss": 2.3922,
      "step": 2015
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.926508903503418,
      "learning_rate": 6.431768207306272e-05,
      "loss": 2.4002,
      "step": 2020
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9747362732887268,
      "learning_rate": 6.415786836498684e-05,
      "loss": 2.4693,
      "step": 2025
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.918900728225708,
      "learning_rate": 6.399789725463298e-05,
      "loss": 2.4109,
      "step": 2030
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9118623733520508,
      "learning_rate": 6.383777052050458e-05,
      "loss": 2.4303,
      "step": 2035
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9325087666511536,
      "learning_rate": 6.367748994283518e-05,
      "loss": 2.4623,
      "step": 2040
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9554570913314819,
      "learning_rate": 6.351705730356877e-05,
      "loss": 2.4295,
      "step": 2045
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9043391942977905,
      "learning_rate": 6.335647438633987e-05,
      "loss": 2.3834,
      "step": 2050
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9207665920257568,
      "learning_rate": 6.319574297645374e-05,
      "loss": 2.4103,
      "step": 2055
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9363177418708801,
      "learning_rate": 6.303486486086654e-05,
      "loss": 2.3831,
      "step": 2060
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9334292411804199,
      "learning_rate": 6.287384182816546e-05,
      "loss": 2.5353,
      "step": 2065
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9631732106208801,
      "learning_rate": 6.271267566854883e-05,
      "loss": 2.4478,
      "step": 2070
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9069787859916687,
      "learning_rate": 6.255136817380618e-05,
      "loss": 2.4026,
      "step": 2075
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.915317952632904,
      "learning_rate": 6.23899211372984e-05,
      "loss": 2.4569,
      "step": 2080
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9367572665214539,
      "learning_rate": 6.222833635393772e-05,
      "loss": 2.4625,
      "step": 2085
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9076709151268005,
      "learning_rate": 6.206661562016782e-05,
      "loss": 2.3889,
      "step": 2090
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.909893810749054,
      "learning_rate": 6.190476073394382e-05,
      "loss": 2.4695,
      "step": 2095
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9313353896141052,
      "learning_rate": 6.17427734947123e-05,
      "loss": 2.5082,
      "step": 2100
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9223794937133789,
      "learning_rate": 6.158065570339127e-05,
      "loss": 2.4231,
      "step": 2105
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9507770538330078,
      "learning_rate": 6.141840916235021e-05,
      "loss": 2.4544,
      "step": 2110
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9403271675109863,
      "learning_rate": 6.125603567539001e-05,
      "loss": 2.4226,
      "step": 2115
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9203287363052368,
      "learning_rate": 6.109353704772284e-05,
      "loss": 2.419,
      "step": 2120
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9149782657623291,
      "learning_rate": 6.0930915085952164e-05,
      "loss": 2.4555,
      "step": 2125
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.910736620426178,
      "learning_rate": 6.076817159805267e-05,
      "loss": 2.4473,
      "step": 2130
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9063466191291809,
      "learning_rate": 6.06053083933501e-05,
      "loss": 2.4813,
      "step": 2135
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9377574324607849,
      "learning_rate": 6.044232728250116e-05,
      "loss": 2.4192,
      "step": 2140
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9170573353767395,
      "learning_rate": 6.027923007747339e-05,
      "loss": 2.391,
      "step": 2145
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9315197467803955,
      "learning_rate": 6.011601859152506e-05,
      "loss": 2.3765,
      "step": 2150
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9108867645263672,
      "learning_rate": 5.995269463918495e-05,
      "loss": 2.4168,
      "step": 2155
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9182575941085815,
      "learning_rate": 5.97892600362322e-05,
      "loss": 2.441,
      "step": 2160
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9704192876815796,
      "learning_rate": 5.962571659967614e-05,
      "loss": 2.4847,
      "step": 2165
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9109625220298767,
      "learning_rate": 5.946206614773606e-05,
      "loss": 2.4149,
      "step": 2170
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.949019730091095,
      "learning_rate": 5.929831049982103e-05,
      "loss": 2.4266,
      "step": 2175
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9033789038658142,
      "learning_rate": 5.9134451476509633e-05,
      "loss": 2.4469,
      "step": 2180
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9018952250480652,
      "learning_rate": 5.897049089952974e-05,
      "loss": 2.4051,
      "step": 2185
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.941893458366394,
      "learning_rate": 5.880643059173826e-05,
      "loss": 2.438,
      "step": 2190
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9332031011581421,
      "learning_rate": 5.864227237710093e-05,
      "loss": 2.393,
      "step": 2195
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9140224456787109,
      "learning_rate": 5.847801808067189e-05,
      "loss": 2.4287,
      "step": 2200
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9171401858329773,
      "learning_rate": 5.831366952857357e-05,
      "loss": 2.4062,
      "step": 2205
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9194997549057007,
      "learning_rate": 5.814922854797622e-05,
      "loss": 2.4013,
      "step": 2210
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9855097532272339,
      "learning_rate": 5.798469696707775e-05,
      "loss": 2.4905,
      "step": 2215
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.946471095085144,
      "learning_rate": 5.782007661508331e-05,
      "loss": 2.4176,
      "step": 2220
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9259464740753174,
      "learning_rate": 5.765536932218495e-05,
      "loss": 2.4195,
      "step": 2225
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9670231342315674,
      "learning_rate": 5.7490576919541315e-05,
      "loss": 2.433,
      "step": 2230
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.911496639251709,
      "learning_rate": 5.732570123925729e-05,
      "loss": 2.4162,
      "step": 2235
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.925327718257904,
      "learning_rate": 5.7160744114363593e-05,
      "loss": 2.4379,
      "step": 2240
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9388567805290222,
      "learning_rate": 5.699570737879641e-05,
      "loss": 2.4124,
      "step": 2245
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9244475960731506,
      "learning_rate": 5.683059286737702e-05,
      "loss": 2.4292,
      "step": 2250
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9055370092391968,
      "learning_rate": 5.666540241579139e-05,
      "loss": 2.3783,
      "step": 2255
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9099582433700562,
      "learning_rate": 5.6500137860569766e-05,
      "loss": 2.4469,
      "step": 2260
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.8946197628974915,
      "learning_rate": 5.633480103906624e-05,
      "loss": 2.4016,
      "step": 2265
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9014321565628052,
      "learning_rate": 5.616939378943834e-05,
      "loss": 2.4056,
      "step": 2270
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9044644832611084,
      "learning_rate": 5.6003917950626595e-05,
      "loss": 2.3981,
      "step": 2275
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.94088214635849,
      "learning_rate": 5.583837536233407e-05,
      "loss": 2.4835,
      "step": 2280
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.8926177620887756,
      "learning_rate": 5.567276786500596e-05,
      "loss": 2.4517,
      "step": 2285
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8916089534759521,
      "learning_rate": 5.5507097299809054e-05,
      "loss": 2.3458,
      "step": 2290
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.937061607837677,
      "learning_rate": 5.534136550861133e-05,
      "loss": 2.4264,
      "step": 2295
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9375845193862915,
      "learning_rate": 5.5175574333961465e-05,
      "loss": 2.4403,
      "step": 2300
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9123733639717102,
      "learning_rate": 5.500972561906832e-05,
      "loss": 2.4226,
      "step": 2305
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9240120053291321,
      "learning_rate": 5.484382120778048e-05,
      "loss": 2.436,
      "step": 2310
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9178423881530762,
      "learning_rate": 5.467786294456575e-05,
      "loss": 2.4285,
      "step": 2315
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9010103940963745,
      "learning_rate": 5.451185267449061e-05,
      "loss": 2.3652,
      "step": 2320
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9071227312088013,
      "learning_rate": 5.43457922431998e-05,
      "loss": 2.4407,
      "step": 2325
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9325879216194153,
      "learning_rate": 5.417968349689566e-05,
      "loss": 2.4164,
      "step": 2330
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9494773149490356,
      "learning_rate": 5.401352828231772e-05,
      "loss": 2.441,
      "step": 2335
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9466855525970459,
      "learning_rate": 5.384732844672211e-05,
      "loss": 2.4787,
      "step": 2340
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.907010555267334,
      "learning_rate": 5.368108583786107e-05,
      "loss": 2.3742,
      "step": 2345
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.93699711561203,
      "learning_rate": 5.3514802303962344e-05,
      "loss": 2.3629,
      "step": 2350
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.030379295349121,
      "learning_rate": 5.334847969370868e-05,
      "loss": 2.4036,
      "step": 2355
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.907166063785553,
      "learning_rate": 5.3182119856217284e-05,
      "loss": 2.4725,
      "step": 2360
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9387956261634827,
      "learning_rate": 5.3015724641019214e-05,
      "loss": 2.3503,
      "step": 2365
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9360509514808655,
      "learning_rate": 5.284929589803884e-05,
      "loss": 2.4759,
      "step": 2370
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9375712871551514,
      "learning_rate": 5.2682835477573336e-05,
      "loss": 2.4025,
      "step": 2375
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.916762113571167,
      "learning_rate": 5.2516345230271965e-05,
      "loss": 2.3794,
      "step": 2380
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9070376753807068,
      "learning_rate": 5.234982700711569e-05,
      "loss": 2.4983,
      "step": 2385
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9349552392959595,
      "learning_rate": 5.218328265939643e-05,
      "loss": 2.4623,
      "step": 2390
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9127172827720642,
      "learning_rate": 5.201671403869657e-05,
      "loss": 2.4297,
      "step": 2395
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9283477067947388,
      "learning_rate": 5.1850122996868366e-05,
      "loss": 2.4022,
      "step": 2400
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9371105432510376,
      "learning_rate": 5.168351138601334e-05,
      "loss": 2.4772,
      "step": 2405
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9107518196105957,
      "learning_rate": 5.1516881058461675e-05,
      "loss": 2.3453,
      "step": 2410
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9191110730171204,
      "learning_rate": 5.135023386675166e-05,
      "loss": 2.4052,
      "step": 2415
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.904039740562439,
      "learning_rate": 5.118357166360906e-05,
      "loss": 2.3583,
      "step": 2420
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9445516467094421,
      "learning_rate": 5.101689630192655e-05,
      "loss": 2.4346,
      "step": 2425
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9454602003097534,
      "learning_rate": 5.085020963474307e-05,
      "loss": 2.3718,
      "step": 2430
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9205687642097473,
      "learning_rate": 5.068351351522329e-05,
      "loss": 2.4452,
      "step": 2435
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9437626004219055,
      "learning_rate": 5.0516809796636935e-05,
      "loss": 2.4275,
      "step": 2440
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9340523481369019,
      "learning_rate": 5.035010033233821e-05,
      "loss": 2.4543,
      "step": 2445
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9576351046562195,
      "learning_rate": 5.018338697574523e-05,
      "loss": 2.3797,
      "step": 2450
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.888482391834259,
      "learning_rate": 5.0016671580319354e-05,
      "loss": 2.4352,
      "step": 2455
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9680615663528442,
      "learning_rate": 4.984995599954461e-05,
      "loss": 2.4499,
      "step": 2460
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9344144463539124,
      "learning_rate": 4.968324208690712e-05,
      "loss": 2.4594,
      "step": 2465
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9282997846603394,
      "learning_rate": 4.951653169587441e-05,
      "loss": 2.3162,
      "step": 2470
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8972924947738647,
      "learning_rate": 4.93498266798749e-05,
      "loss": 2.428,
      "step": 2475
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9192982912063599,
      "learning_rate": 4.918312889227722e-05,
      "loss": 2.4126,
      "step": 2480
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9322929382324219,
      "learning_rate": 4.901644018636966e-05,
      "loss": 2.3799,
      "step": 2485
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9455317258834839,
      "learning_rate": 4.8849762415339526e-05,
      "loss": 2.4297,
      "step": 2490
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9278294444084167,
      "learning_rate": 4.868309743225256e-05,
      "loss": 2.3809,
      "step": 2495
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.921036958694458,
      "learning_rate": 4.851644709003233e-05,
      "loss": 2.4302,
      "step": 2500
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.933717668056488,
      "learning_rate": 4.834981324143964e-05,
      "loss": 2.4012,
      "step": 2505
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9190297722816467,
      "learning_rate": 4.818319773905191e-05,
      "loss": 2.3523,
      "step": 2510
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9000710844993591,
      "learning_rate": 4.801660243524261e-05,
      "loss": 2.4201,
      "step": 2515
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9176121950149536,
      "learning_rate": 4.7850029182160626e-05,
      "loss": 2.3806,
      "step": 2520
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9202491044998169,
      "learning_rate": 4.768347983170973e-05,
      "loss": 2.4076,
      "step": 2525
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9020574688911438,
      "learning_rate": 4.7516956235527884e-05,
      "loss": 2.4504,
      "step": 2530
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.894968569278717,
      "learning_rate": 4.735046024496682e-05,
      "loss": 2.404,
      "step": 2535
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9300612807273865,
      "learning_rate": 4.7183993711071286e-05,
      "loss": 2.3997,
      "step": 2540
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.929145097732544,
      "learning_rate": 4.7017558484558554e-05,
      "loss": 2.4052,
      "step": 2545
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9086617231369019,
      "learning_rate": 4.6851156415797844e-05,
      "loss": 2.402,
      "step": 2550
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.894952118396759,
      "learning_rate": 4.6684789354789746e-05,
      "loss": 2.4397,
      "step": 2555
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9135778546333313,
      "learning_rate": 4.651845915114563e-05,
      "loss": 2.4014,
      "step": 2560
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9323270916938782,
      "learning_rate": 4.6352167654067095e-05,
      "loss": 2.4016,
      "step": 2565
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.911676824092865,
      "learning_rate": 4.618591671232544e-05,
      "loss": 2.3717,
      "step": 2570
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.881301999092102,
      "learning_rate": 4.601970817424106e-05,
      "loss": 2.3498,
      "step": 2575
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9046157002449036,
      "learning_rate": 4.585354388766292e-05,
      "loss": 2.3722,
      "step": 2580
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9167553186416626,
      "learning_rate": 4.568742569994802e-05,
      "loss": 2.3989,
      "step": 2585
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9558010697364807,
      "learning_rate": 4.552135545794086e-05,
      "loss": 2.4688,
      "step": 2590
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.930412232875824,
      "learning_rate": 4.535533500795288e-05,
      "loss": 2.4169,
      "step": 2595
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.95481938123703,
      "learning_rate": 4.5189366195741953e-05,
      "loss": 2.4134,
      "step": 2600
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8904323577880859,
      "learning_rate": 4.502345086649186e-05,
      "loss": 2.3209,
      "step": 2605
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9117818474769592,
      "learning_rate": 4.485759086479179e-05,
      "loss": 2.3939,
      "step": 2610
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9135521650314331,
      "learning_rate": 4.469178803461579e-05,
      "loss": 2.3403,
      "step": 2615
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8931097984313965,
      "learning_rate": 4.4526044219302326e-05,
      "loss": 2.3794,
      "step": 2620
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9080528020858765,
      "learning_rate": 4.4360361261533745e-05,
      "loss": 2.3682,
      "step": 2625
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8922243118286133,
      "learning_rate": 4.419474100331579e-05,
      "loss": 2.3913,
      "step": 2630
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9073590636253357,
      "learning_rate": 4.402918528595715e-05,
      "loss": 2.332,
      "step": 2635
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8992304801940918,
      "learning_rate": 4.386369595004896e-05,
      "loss": 2.4228,
      "step": 2640
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9212525486946106,
      "learning_rate": 4.3698274835444354e-05,
      "loss": 2.3724,
      "step": 2645
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9063347578048706,
      "learning_rate": 4.3532923781238e-05,
      "loss": 2.3823,
      "step": 2650
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9009177684783936,
      "learning_rate": 4.336764462574566e-05,
      "loss": 2.3531,
      "step": 2655
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9231202602386475,
      "learning_rate": 4.320243920648376e-05,
      "loss": 2.4197,
      "step": 2660
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9398454427719116,
      "learning_rate": 4.303730936014894e-05,
      "loss": 2.392,
      "step": 2665
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8874997496604919,
      "learning_rate": 4.287225692259765e-05,
      "loss": 2.3254,
      "step": 2670
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.937724232673645,
      "learning_rate": 4.270728372882575e-05,
      "loss": 2.416,
      "step": 2675
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9113511443138123,
      "learning_rate": 4.254239161294804e-05,
      "loss": 2.4305,
      "step": 2680
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9297624230384827,
      "learning_rate": 4.237758240817802e-05,
      "loss": 2.3571,
      "step": 2685
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.8927726745605469,
      "learning_rate": 4.2212857946807336e-05,
      "loss": 2.3647,
      "step": 2690
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9231744408607483,
      "learning_rate": 4.2048220060185516e-05,
      "loss": 2.4092,
      "step": 2695
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9141883850097656,
      "learning_rate": 4.188367057869957e-05,
      "loss": 2.3413,
      "step": 2700
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9077816605567932,
      "learning_rate": 4.171921133175365e-05,
      "loss": 2.4024,
      "step": 2705
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9256351590156555,
      "learning_rate": 4.155484414774872e-05,
      "loss": 2.4133,
      "step": 2710
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9212350249290466,
      "learning_rate": 4.139057085406221e-05,
      "loss": 2.3808,
      "step": 2715
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9098197221755981,
      "learning_rate": 4.1226393277027726e-05,
      "loss": 2.4289,
      "step": 2720
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9494619369506836,
      "learning_rate": 4.106231324191471e-05,
      "loss": 2.4195,
      "step": 2725
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8974795937538147,
      "learning_rate": 4.089833257290817e-05,
      "loss": 2.4012,
      "step": 2730
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8910524249076843,
      "learning_rate": 4.073445309308842e-05,
      "loss": 2.3182,
      "step": 2735
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8898556232452393,
      "learning_rate": 4.0570676624410756e-05,
      "loss": 2.4341,
      "step": 2740
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9277542233467102,
      "learning_rate": 4.040700498768525e-05,
      "loss": 2.387,
      "step": 2745
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9058116674423218,
      "learning_rate": 4.024344000255648e-05,
      "loss": 2.3898,
      "step": 2750
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9045486450195312,
      "learning_rate": 4.0079983487483313e-05,
      "loss": 2.3977,
      "step": 2755
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9825146794319153,
      "learning_rate": 3.9916637259718683e-05,
      "loss": 2.3811,
      "step": 2760
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.879527747631073,
      "learning_rate": 3.9753403135289396e-05,
      "loss": 2.3712,
      "step": 2765
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9239062666893005,
      "learning_rate": 3.9590282928975914e-05,
      "loss": 2.3904,
      "step": 2770
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9276576638221741,
      "learning_rate": 3.942727845429221e-05,
      "loss": 2.3147,
      "step": 2775
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9070415496826172,
      "learning_rate": 3.926439152346558e-05,
      "loss": 2.3752,
      "step": 2780
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9174474477767944,
      "learning_rate": 3.910162394741653e-05,
      "loss": 2.3734,
      "step": 2785
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9264554977416992,
      "learning_rate": 3.893897753573861e-05,
      "loss": 2.3948,
      "step": 2790
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9267555475234985,
      "learning_rate": 3.877645409667829e-05,
      "loss": 2.394,
      "step": 2795
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9059020280838013,
      "learning_rate": 3.861405543711491e-05,
      "loss": 2.3697,
      "step": 2800
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9132973551750183,
      "learning_rate": 3.8451783362540507e-05,
      "loss": 2.409,
      "step": 2805
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.920639157295227,
      "learning_rate": 3.828963967703983e-05,
      "loss": 2.3474,
      "step": 2810
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9337471723556519,
      "learning_rate": 3.8127626183270223e-05,
      "loss": 2.3821,
      "step": 2815
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9225865006446838,
      "learning_rate": 3.796574468244161e-05,
      "loss": 2.4106,
      "step": 2820
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9328856468200684,
      "learning_rate": 3.7803996974296444e-05,
      "loss": 2.4665,
      "step": 2825
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9074559807777405,
      "learning_rate": 3.7642384857089776e-05,
      "loss": 2.3724,
      "step": 2830
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9395449161529541,
      "learning_rate": 3.748091012756915e-05,
      "loss": 2.3418,
      "step": 2835
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9058288931846619,
      "learning_rate": 3.731957458095467e-05,
      "loss": 2.3908,
      "step": 2840
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.0165904760360718,
      "learning_rate": 3.71583800109191e-05,
      "loss": 2.3796,
      "step": 2845
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9278620481491089,
      "learning_rate": 3.699732820956784e-05,
      "loss": 2.329,
      "step": 2850
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9070698618888855,
      "learning_rate": 3.6836420967419057e-05,
      "loss": 2.4322,
      "step": 2855
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9063517451286316,
      "learning_rate": 3.6675660073383745e-05,
      "loss": 2.3743,
      "step": 2860
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8901442885398865,
      "learning_rate": 3.6515047314745856e-05,
      "loss": 2.317,
      "step": 2865
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9144355058670044,
      "learning_rate": 3.6354584477142437e-05,
      "loss": 2.4009,
      "step": 2870
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9156922101974487,
      "learning_rate": 3.6194273344543736e-05,
      "loss": 2.3737,
      "step": 2875
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9051995277404785,
      "learning_rate": 3.6034115699233425e-05,
      "loss": 2.4593,
      "step": 2880
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9019202589988708,
      "learning_rate": 3.5874113321788736e-05,
      "loss": 2.3608,
      "step": 2885
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9227021336555481,
      "learning_rate": 3.571426799106071e-05,
      "loss": 2.3829,
      "step": 2890
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9123120903968811,
      "learning_rate": 3.555458148415437e-05,
      "loss": 2.3874,
      "step": 2895
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9297784566879272,
      "learning_rate": 3.539505557640901e-05,
      "loss": 2.349,
      "step": 2900
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9303633570671082,
      "learning_rate": 3.523569204137843e-05,
      "loss": 2.4062,
      "step": 2905
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9044543504714966,
      "learning_rate": 3.5076492650811246e-05,
      "loss": 2.4351,
      "step": 2910
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9213395714759827,
      "learning_rate": 3.491745917463113e-05,
      "loss": 2.316,
      "step": 2915
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9098991751670837,
      "learning_rate": 3.475859338091721e-05,
      "loss": 2.3523,
      "step": 2920
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9128162264823914,
      "learning_rate": 3.4599897035884374e-05,
      "loss": 2.3663,
      "step": 2925
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9543825387954712,
      "learning_rate": 3.444137190386363e-05,
      "loss": 2.3587,
      "step": 2930
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9522660970687866,
      "learning_rate": 3.4283019747282514e-05,
      "loss": 2.3508,
      "step": 2935
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8969595432281494,
      "learning_rate": 3.412484232664545e-05,
      "loss": 2.3126,
      "step": 2940
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9274249076843262,
      "learning_rate": 3.396684140051424e-05,
      "loss": 2.3236,
      "step": 2945
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8981707692146301,
      "learning_rate": 3.3809018725488466e-05,
      "loss": 2.3745,
      "step": 2950
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.947674036026001,
      "learning_rate": 3.365137605618598e-05,
      "loss": 2.3715,
      "step": 2955
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.891778290271759,
      "learning_rate": 3.3493915145223395e-05,
      "loss": 2.3635,
      "step": 2960
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9314977526664734,
      "learning_rate": 3.3336637743196584e-05,
      "loss": 2.2918,
      "step": 2965
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9266039729118347,
      "learning_rate": 3.317954559866126e-05,
      "loss": 2.4019,
      "step": 2970
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8894385099411011,
      "learning_rate": 3.302264045811344e-05,
      "loss": 2.3935,
      "step": 2975
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9218305945396423,
      "learning_rate": 3.286592406597021e-05,
      "loss": 2.3447,
      "step": 2980
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9280163645744324,
      "learning_rate": 3.270939816455012e-05,
      "loss": 2.3861,
      "step": 2985
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9048585891723633,
      "learning_rate": 3.255306449405395e-05,
      "loss": 2.4027,
      "step": 2990
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8996835947036743,
      "learning_rate": 3.2396924792545304e-05,
      "loss": 2.3135,
      "step": 2995
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9637964367866516,
      "learning_rate": 3.224098079593132e-05,
      "loss": 2.4027,
      "step": 3000
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8991714715957642,
      "learning_rate": 3.2085234237943354e-05,
      "loss": 2.3699,
      "step": 3005
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9499838948249817,
      "learning_rate": 3.19296868501177e-05,
      "loss": 2.3901,
      "step": 3010
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9127416014671326,
      "learning_rate": 3.177434036177636e-05,
      "loss": 2.4039,
      "step": 3015
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9323961734771729,
      "learning_rate": 3.1619196500007804e-05,
      "loss": 2.3456,
      "step": 3020
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9023792743682861,
      "learning_rate": 3.146425698964776e-05,
      "loss": 2.3986,
      "step": 3025
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9215497374534607,
      "learning_rate": 3.1309523553260046e-05,
      "loss": 2.3066,
      "step": 3030
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9214668869972229,
      "learning_rate": 3.115499791111743e-05,
      "loss": 2.3897,
      "step": 3035
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8906429409980774,
      "learning_rate": 3.10006817811825e-05,
      "loss": 2.2407,
      "step": 3040
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9099827408790588,
      "learning_rate": 3.084657687908855e-05,
      "loss": 2.3565,
      "step": 3045
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8930659890174866,
      "learning_rate": 3.069268491812052e-05,
      "loss": 2.3377,
      "step": 3050
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.904613733291626,
      "learning_rate": 3.0539007609195934e-05,
      "loss": 2.3979,
      "step": 3055
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9138222932815552,
      "learning_rate": 3.0385546660845908e-05,
      "loss": 2.341,
      "step": 3060
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.928210437297821,
      "learning_rate": 3.0232303779196132e-05,
      "loss": 2.3547,
      "step": 3065
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9129458665847778,
      "learning_rate": 3.0079280667947885e-05,
      "loss": 2.3852,
      "step": 3070
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9085707664489746,
      "learning_rate": 2.9926479028359132e-05,
      "loss": 2.4033,
      "step": 3075
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.919783890247345,
      "learning_rate": 2.97739005592256e-05,
      "loss": 2.3779,
      "step": 3080
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9020473957061768,
      "learning_rate": 2.962154695686187e-05,
      "loss": 2.311,
      "step": 3085
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9025293588638306,
      "learning_rate": 2.9469419915082536e-05,
      "loss": 2.3531,
      "step": 3090
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9105976223945618,
      "learning_rate": 2.9317521125183368e-05,
      "loss": 2.379,
      "step": 3095
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9467954039573669,
      "learning_rate": 2.9165852275922524e-05,
      "loss": 2.2659,
      "step": 3100
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9154301881790161,
      "learning_rate": 2.901441505350174e-05,
      "loss": 2.4311,
      "step": 3105
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9146058559417725,
      "learning_rate": 2.886321114154762e-05,
      "loss": 2.3807,
      "step": 3110
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9149680137634277,
      "learning_rate": 2.87122422210929e-05,
      "loss": 2.3241,
      "step": 3115
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9131379127502441,
      "learning_rate": 2.8561509970557736e-05,
      "loss": 2.254,
      "step": 3120
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9278967380523682,
      "learning_rate": 2.8411016065731146e-05,
      "loss": 2.3627,
      "step": 3125
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9049065113067627,
      "learning_rate": 2.826076217975222e-05,
      "loss": 2.3892,
      "step": 3130
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9476072788238525,
      "learning_rate": 2.8110749983091632e-05,
      "loss": 2.3093,
      "step": 3135
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9162594079971313,
      "learning_rate": 2.7960981143533053e-05,
      "loss": 2.3172,
      "step": 3140
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9001609086990356,
      "learning_rate": 2.781145732615457e-05,
      "loss": 2.3749,
      "step": 3145
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8743308186531067,
      "learning_rate": 2.7662180193310218e-05,
      "loss": 2.3167,
      "step": 3150
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9399544596672058,
      "learning_rate": 2.751315140461145e-05,
      "loss": 2.2892,
      "step": 3155
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8993040919303894,
      "learning_rate": 2.7364372616908744e-05,
      "loss": 2.4035,
      "step": 3160
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8989415168762207,
      "learning_rate": 2.7215845484273152e-05,
      "loss": 2.3562,
      "step": 3165
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9252132773399353,
      "learning_rate": 2.7067571657977893e-05,
      "loss": 2.3413,
      "step": 3170
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9166437387466431,
      "learning_rate": 2.691955278648003e-05,
      "loss": 2.3385,
      "step": 3175
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9225659370422363,
      "learning_rate": 2.6771790515402112e-05,
      "loss": 2.369,
      "step": 3180
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9323848485946655,
      "learning_rate": 2.6624286487513916e-05,
      "loss": 2.3854,
      "step": 3185
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9322205185890198,
      "learning_rate": 2.6477042342714137e-05,
      "loss": 2.3311,
      "step": 3190
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9424782991409302,
      "learning_rate": 2.633005971801219e-05,
      "loss": 2.3474,
      "step": 3195
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9437265396118164,
      "learning_rate": 2.6183340247510013e-05,
      "loss": 2.3294,
      "step": 3200
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9069112539291382,
      "learning_rate": 2.6036885562383856e-05,
      "loss": 2.3599,
      "step": 3205
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9054490327835083,
      "learning_rate": 2.5890697290866206e-05,
      "loss": 2.3053,
      "step": 3210
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8967999815940857,
      "learning_rate": 2.5744777058227642e-05,
      "loss": 2.3403,
      "step": 3215
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9110304713249207,
      "learning_rate": 2.5599126486758777e-05,
      "loss": 2.2913,
      "step": 3220
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9111791849136353,
      "learning_rate": 2.5453747195752243e-05,
      "loss": 2.2663,
      "step": 3225
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.90242999792099,
      "learning_rate": 2.530864080148464e-05,
      "loss": 2.3761,
      "step": 3230
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9307655692100525,
      "learning_rate": 2.5163808917198615e-05,
      "loss": 2.2882,
      "step": 3235
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9393335580825806,
      "learning_rate": 2.501925315308492e-05,
      "loss": 2.3714,
      "step": 3240
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9044473171234131,
      "learning_rate": 2.4874975116264477e-05,
      "loss": 2.3814,
      "step": 3245
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.928254246711731,
      "learning_rate": 2.4730976410770534e-05,
      "loss": 2.3812,
      "step": 3250
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9389528036117554,
      "learning_rate": 2.458725863753084e-05,
      "loss": 2.3669,
      "step": 3255
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9334967732429504,
      "learning_rate": 2.4443823394349834e-05,
      "loss": 2.3799,
      "step": 3260
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9133898019790649,
      "learning_rate": 2.430067227589088e-05,
      "loss": 2.4019,
      "step": 3265
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8969634771347046,
      "learning_rate": 2.4157806873658517e-05,
      "loss": 2.3894,
      "step": 3270
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8983045220375061,
      "learning_rate": 2.401522877598087e-05,
      "loss": 2.3523,
      "step": 3275
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9043636322021484,
      "learning_rate": 2.3872939567991827e-05,
      "loss": 2.3632,
      "step": 3280
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8893569111824036,
      "learning_rate": 2.373094083161353e-05,
      "loss": 2.3254,
      "step": 3285
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9358165264129639,
      "learning_rate": 2.358923414553877e-05,
      "loss": 2.3605,
      "step": 3290
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8925164937973022,
      "learning_rate": 2.3447821085213405e-05,
      "loss": 2.3003,
      "step": 3295
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9067592024803162,
      "learning_rate": 2.3306703222818878e-05,
      "loss": 2.3638,
      "step": 3300
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9226784110069275,
      "learning_rate": 2.3165882127254705e-05,
      "loss": 2.2942,
      "step": 3305
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9251599907875061,
      "learning_rate": 2.302535936412108e-05,
      "loss": 2.381,
      "step": 3310
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9351403713226318,
      "learning_rate": 2.2885136495701415e-05,
      "loss": 2.3022,
      "step": 3315
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9234192967414856,
      "learning_rate": 2.274521508094501e-05,
      "loss": 2.3531,
      "step": 3320
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8894725441932678,
      "learning_rate": 2.2605596675449698e-05,
      "loss": 2.3283,
      "step": 3325
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8899527192115784,
      "learning_rate": 2.246628283144457e-05,
      "loss": 2.3818,
      "step": 3330
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9192507863044739,
      "learning_rate": 2.232727509777269e-05,
      "loss": 2.341,
      "step": 3335
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9314063191413879,
      "learning_rate": 2.2188575019873932e-05,
      "loss": 2.3511,
      "step": 3340
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9228346347808838,
      "learning_rate": 2.2050184139767704e-05,
      "loss": 2.3609,
      "step": 3345
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9301695823669434,
      "learning_rate": 2.191210399603591e-05,
      "loss": 2.3318,
      "step": 3350
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9185614585876465,
      "learning_rate": 2.1774336123805772e-05,
      "loss": 2.4049,
      "step": 3355
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9196087121963501,
      "learning_rate": 2.1636882054732776e-05,
      "loss": 2.3333,
      "step": 3360
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.921023964881897,
      "learning_rate": 2.1499743316983684e-05,
      "loss": 2.3288,
      "step": 3365
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9260315299034119,
      "learning_rate": 2.1362921435219473e-05,
      "loss": 2.3624,
      "step": 3370
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9450960159301758,
      "learning_rate": 2.1226417930578464e-05,
      "loss": 2.4221,
      "step": 3375
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9169435501098633,
      "learning_rate": 2.109023432065935e-05,
      "loss": 2.3724,
      "step": 3380
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9358202219009399,
      "learning_rate": 2.095437211950434e-05,
      "loss": 2.3792,
      "step": 3385
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9272687435150146,
      "learning_rate": 2.0818832837582352e-05,
      "loss": 2.3538,
      "step": 3390
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9306479096412659,
      "learning_rate": 2.068361798177218e-05,
      "loss": 2.3273,
      "step": 3395
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9368727207183838,
      "learning_rate": 2.0548729055345778e-05,
      "loss": 2.324,
      "step": 3400
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9212680459022522,
      "learning_rate": 2.0414167557951514e-05,
      "loss": 2.3944,
      "step": 3405
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.917975902557373,
      "learning_rate": 2.0279934985597527e-05,
      "loss": 2.3215,
      "step": 3410
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9334181547164917,
      "learning_rate": 2.0146032830635054e-05,
      "loss": 2.3373,
      "step": 3415
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9223392009735107,
      "learning_rate": 2.001246258174192e-05,
      "loss": 2.4093,
      "step": 3420
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9029899835586548,
      "learning_rate": 1.9879225723905886e-05,
      "loss": 2.3304,
      "step": 3425
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9155861735343933,
      "learning_rate": 1.9746323738408203e-05,
      "loss": 2.3056,
      "step": 3430
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9587869048118591,
      "learning_rate": 1.9613758102807117e-05,
      "loss": 2.3085,
      "step": 3435
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9450895190238953,
      "learning_rate": 1.9481530290921474e-05,
      "loss": 2.3667,
      "step": 3440
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9262810945510864,
      "learning_rate": 1.934964177281428e-05,
      "loss": 2.2548,
      "step": 3445
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9066665768623352,
      "learning_rate": 1.9218094014776434e-05,
      "loss": 2.3256,
      "step": 3450
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.925977349281311,
      "learning_rate": 1.9086888479310333e-05,
      "loss": 2.3263,
      "step": 3455
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9015002846717834,
      "learning_rate": 1.895602662511371e-05,
      "loss": 2.193,
      "step": 3460
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.900550365447998,
      "learning_rate": 1.8825509907063327e-05,
      "loss": 2.3363,
      "step": 3465
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.932928204536438,
      "learning_rate": 1.8695339776198872e-05,
      "loss": 2.3329,
      "step": 3470
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9403447508811951,
      "learning_rate": 1.8565517679706783e-05,
      "loss": 2.3214,
      "step": 3475
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8891700506210327,
      "learning_rate": 1.8436045060904174e-05,
      "loss": 2.2526,
      "step": 3480
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9113773703575134,
      "learning_rate": 1.830692335922279e-05,
      "loss": 2.344,
      "step": 3485
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8912283182144165,
      "learning_rate": 1.8178154010192994e-05,
      "loss": 2.3739,
      "step": 3490
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.925250232219696,
      "learning_rate": 1.8049738445427822e-05,
      "loss": 2.3043,
      "step": 3495
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.925449013710022,
      "learning_rate": 1.7921678092607052e-05,
      "loss": 2.3278,
      "step": 3500
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9105491638183594,
      "learning_rate": 1.7793974375461352e-05,
      "loss": 2.2991,
      "step": 3505
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.949779212474823,
      "learning_rate": 1.7666628713756417e-05,
      "loss": 2.3522,
      "step": 3510
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.944808304309845,
      "learning_rate": 1.7539642523277228e-05,
      "loss": 2.3239,
      "step": 3515
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9059569239616394,
      "learning_rate": 1.7413017215812273e-05,
      "loss": 2.2738,
      "step": 3520
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9568001627922058,
      "learning_rate": 1.728675419913788e-05,
      "loss": 2.2816,
      "step": 3525
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9053666591644287,
      "learning_rate": 1.716085487700253e-05,
      "loss": 2.2909,
      "step": 3530
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9108089804649353,
      "learning_rate": 1.703532064911131e-05,
      "loss": 2.3613,
      "step": 3535
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9420007467269897,
      "learning_rate": 1.6910152911110283e-05,
      "loss": 2.3443,
      "step": 3540
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9437248706817627,
      "learning_rate": 1.6785353054571024e-05,
      "loss": 2.3616,
      "step": 3545
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9269158840179443,
      "learning_rate": 1.666092246697512e-05,
      "loss": 2.3475,
      "step": 3550
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9224128723144531,
      "learning_rate": 1.6536862531698766e-05,
      "loss": 2.3165,
      "step": 3555
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9019190669059753,
      "learning_rate": 1.6413174627997328e-05,
      "loss": 2.3543,
      "step": 3560
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.174714207649231,
      "learning_rate": 1.6289860130990147e-05,
      "loss": 2.3201,
      "step": 3565
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9111523628234863,
      "learning_rate": 1.6166920411645064e-05,
      "loss": 2.3747,
      "step": 3570
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.913772463798523,
      "learning_rate": 1.6044356836763315e-05,
      "loss": 2.3417,
      "step": 3575
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9173497557640076,
      "learning_rate": 1.5922170768964285e-05,
      "loss": 2.3649,
      "step": 3580
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9145777821540833,
      "learning_rate": 1.5800363566670362e-05,
      "loss": 2.3156,
      "step": 3585
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9274765849113464,
      "learning_rate": 1.5678936584091852e-05,
      "loss": 2.2597,
      "step": 3590
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9273658394813538,
      "learning_rate": 1.5557891171211892e-05,
      "loss": 2.3189,
      "step": 3595
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9246422052383423,
      "learning_rate": 1.5437228673771465e-05,
      "loss": 2.2589,
      "step": 3600
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.946138322353363,
      "learning_rate": 1.5316950433254445e-05,
      "loss": 2.2714,
      "step": 3605
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9329668283462524,
      "learning_rate": 1.5197057786872649e-05,
      "loss": 2.2799,
      "step": 3610
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9475228786468506,
      "learning_rate": 1.5077552067551015e-05,
      "loss": 2.3693,
      "step": 3615
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9241766929626465,
      "learning_rate": 1.4958434603912747e-05,
      "loss": 2.3682,
      "step": 3620
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9258906841278076,
      "learning_rate": 1.4839706720264546e-05,
      "loss": 2.2912,
      "step": 3625
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9170982837677002,
      "learning_rate": 1.4721369736581924e-05,
      "loss": 2.2662,
      "step": 3630
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9363173246383667,
      "learning_rate": 1.4603424968494484e-05,
      "loss": 2.3692,
      "step": 3635
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9334496259689331,
      "learning_rate": 1.448587372727132e-05,
      "loss": 2.286,
      "step": 3640
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9272920489311218,
      "learning_rate": 1.4368717319806419e-05,
      "loss": 2.3078,
      "step": 3645
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9488955140113831,
      "learning_rate": 1.4251957048604152e-05,
      "loss": 2.3554,
      "step": 3650
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.918768584728241,
      "learning_rate": 1.413559421176479e-05,
      "loss": 2.3845,
      "step": 3655
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9137639999389648,
      "learning_rate": 1.4019630102970056e-05,
      "loss": 2.3445,
      "step": 3660
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9241656064987183,
      "learning_rate": 1.3904066011468753e-05,
      "loss": 2.3188,
      "step": 3665
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9146557450294495,
      "learning_rate": 1.3788903222062433e-05,
      "loss": 2.3191,
      "step": 3670
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.947729229927063,
      "learning_rate": 1.3674143015091118e-05,
      "loss": 2.353,
      "step": 3675
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9175273180007935,
      "learning_rate": 1.355978666641905e-05,
      "loss": 2.3044,
      "step": 3680
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9403489828109741,
      "learning_rate": 1.3445835447420507e-05,
      "loss": 2.3539,
      "step": 3685
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9238479137420654,
      "learning_rate": 1.3332290624965688e-05,
      "loss": 2.2995,
      "step": 3690
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9429262280464172,
      "learning_rate": 1.3219153461406609e-05,
      "loss": 2.3249,
      "step": 3695
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8907161355018616,
      "learning_rate": 1.3106425214563078e-05,
      "loss": 2.3333,
      "step": 3700
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9309660792350769,
      "learning_rate": 1.2994107137708716e-05,
      "loss": 2.3351,
      "step": 3705
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9421401619911194,
      "learning_rate": 1.2882200479556988e-05,
      "loss": 2.3823,
      "step": 3710
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9495819807052612,
      "learning_rate": 1.2770706484247397e-05,
      "loss": 2.3302,
      "step": 3715
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9074979424476624,
      "learning_rate": 1.2659626391331564e-05,
      "loss": 2.2637,
      "step": 3720
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9183170199394226,
      "learning_rate": 1.2548961435759493e-05,
      "loss": 2.3008,
      "step": 3725
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9259270429611206,
      "learning_rate": 1.2438712847865846e-05,
      "loss": 2.3748,
      "step": 3730
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9168004989624023,
      "learning_rate": 1.2328881853356244e-05,
      "loss": 2.2474,
      "step": 3735
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8991266489028931,
      "learning_rate": 1.221946967329365e-05,
      "loss": 2.346,
      "step": 3740
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9321264028549194,
      "learning_rate": 1.2110477524084796e-05,
      "loss": 2.3265,
      "step": 3745
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9339208602905273,
      "learning_rate": 1.2001906617466657e-05,
      "loss": 2.3527,
      "step": 3750
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8979222178459167,
      "learning_rate": 1.1893758160492978e-05,
      "loss": 2.2731,
      "step": 3755
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9186643362045288,
      "learning_rate": 1.1786033355520859e-05,
      "loss": 2.368,
      "step": 3760
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9406339526176453,
      "learning_rate": 1.1678733400197373e-05,
      "loss": 2.3329,
      "step": 3765
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9028125405311584,
      "learning_rate": 1.1571859487446263e-05,
      "loss": 2.3374,
      "step": 3770
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9233968257904053,
      "learning_rate": 1.1465412805454695e-05,
      "loss": 2.3105,
      "step": 3775
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9305930733680725,
      "learning_rate": 1.1359394537660011e-05,
      "loss": 2.3692,
      "step": 3780
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9098057746887207,
      "learning_rate": 1.125380586273661e-05,
      "loss": 2.3132,
      "step": 3785
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9284851551055908,
      "learning_rate": 1.1148647954582808e-05,
      "loss": 2.2891,
      "step": 3790
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.911727249622345,
      "learning_rate": 1.1043921982307819e-05,
      "loss": 2.3675,
      "step": 3795
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9202545881271362,
      "learning_rate": 1.0939629110218735e-05,
      "loss": 2.3074,
      "step": 3800
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9392959475517273,
      "learning_rate": 1.0835770497807596e-05,
      "loss": 2.3653,
      "step": 3805
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9099482297897339,
      "learning_rate": 1.0732347299738493e-05,
      "loss": 2.3247,
      "step": 3810
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9176100492477417,
      "learning_rate": 1.0629360665834732e-05,
      "loss": 2.3071,
      "step": 3815
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.945218563079834,
      "learning_rate": 1.052681174106604e-05,
      "loss": 2.3201,
      "step": 3820
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9112182259559631,
      "learning_rate": 1.0424701665535852e-05,
      "loss": 2.3579,
      "step": 3825
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9128363728523254,
      "learning_rate": 1.0323031574468638e-05,
      "loss": 2.2977,
      "step": 3830
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9202331304550171,
      "learning_rate": 1.0221802598197261e-05,
      "loss": 2.339,
      "step": 3835
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9266894459724426,
      "learning_rate": 1.0121015862150423e-05,
      "loss": 2.2969,
      "step": 3840
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9341091513633728,
      "learning_rate": 1.0020672486840154e-05,
      "loss": 2.3475,
      "step": 3845
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9220916628837585,
      "learning_rate": 9.920773587849364e-06,
      "loss": 2.3231,
      "step": 3850
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9296770095825195,
      "learning_rate": 9.821320275819401e-06,
      "loss": 2.2958,
      "step": 3855
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9489056468009949,
      "learning_rate": 9.72231365643777e-06,
      "loss": 2.2493,
      "step": 3860
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9161087274551392,
      "learning_rate": 9.623754830425779e-06,
      "loss": 2.3089,
      "step": 3865
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9204024076461792,
      "learning_rate": 9.52564489352632e-06,
      "loss": 2.2413,
      "step": 3870
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9348191022872925,
      "learning_rate": 9.427984936491702e-06,
      "loss": 2.29,
      "step": 3875
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9372574090957642,
      "learning_rate": 9.330776045071509e-06,
      "loss": 2.2761,
      "step": 3880
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9387868642807007,
      "learning_rate": 9.23401930000054e-06,
      "loss": 2.2701,
      "step": 3885
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9021495580673218,
      "learning_rate": 9.137715776986772e-06,
      "loss": 2.2859,
      "step": 3890
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9116939306259155,
      "learning_rate": 9.041866546699434e-06,
      "loss": 2.2792,
      "step": 3895
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9261936545372009,
      "learning_rate": 8.946472674757078e-06,
      "loss": 2.2656,
      "step": 3900
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9336845278739929,
      "learning_rate": 8.851535221715735e-06,
      "loss": 2.3391,
      "step": 3905
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9195213913917542,
      "learning_rate": 8.757055243057132e-06,
      "loss": 2.2762,
      "step": 3910
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9221490621566772,
      "learning_rate": 8.663033789176967e-06,
      "loss": 2.3548,
      "step": 3915
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9198946356773376,
      "learning_rate": 8.5694719053732e-06,
      "loss": 2.3239,
      "step": 3920
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9193469285964966,
      "learning_rate": 8.476370631834458e-06,
      "loss": 2.2747,
      "step": 3925
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9079307317733765,
      "learning_rate": 8.383731003628452e-06,
      "loss": 2.347,
      "step": 3930
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9317283630371094,
      "learning_rate": 8.291554050690508e-06,
      "loss": 2.2833,
      "step": 3935
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9155171513557434,
      "learning_rate": 8.199840797812058e-06,
      "loss": 2.3311,
      "step": 3940
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9332835674285889,
      "learning_rate": 8.108592264629295e-06,
      "loss": 2.3594,
      "step": 3945
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9451523423194885,
      "learning_rate": 8.017809465611803e-06,
      "loss": 2.2932,
      "step": 3950
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9449190497398376,
      "learning_rate": 7.927493410051324e-06,
      "loss": 2.3888,
      "step": 3955
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9128648042678833,
      "learning_rate": 7.837645102050473e-06,
      "loss": 2.2962,
      "step": 3960
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9178701639175415,
      "learning_rate": 7.748265540511635e-06,
      "loss": 2.3119,
      "step": 3965
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9139108061790466,
      "learning_rate": 7.65935571912582e-06,
      "loss": 2.259,
      "step": 3970
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9180004596710205,
      "learning_rate": 7.5709166263616405e-06,
      "loss": 2.2977,
      "step": 3975
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9333677887916565,
      "learning_rate": 7.482949245454302e-06,
      "loss": 2.2495,
      "step": 3980
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9381062984466553,
      "learning_rate": 7.3954545543946876e-06,
      "loss": 2.302,
      "step": 3985
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9359762072563171,
      "learning_rate": 7.308433525918468e-06,
      "loss": 2.2843,
      "step": 3990
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9028638005256653,
      "learning_rate": 7.221887127495313e-06,
      "loss": 2.23,
      "step": 3995
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9264029264450073,
      "learning_rate": 7.1358163213181114e-06,
      "loss": 2.3494,
      "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
}