{
  "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": 35.52880096435547,
      "learning_rate": 5e-06,
      "loss": 13.913,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 27.649765014648438,
      "learning_rate": 1e-05,
      "loss": 13.4311,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 16.23821258544922,
      "learning_rate": 1.5e-05,
      "loss": 12.2243,
      "step": 15
    },
    {
      "epoch": 0.0,
      "grad_norm": 11.717832565307617,
      "learning_rate": 2e-05,
      "loss": 10.8725,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.961292266845703,
      "learning_rate": 2.5e-05,
      "loss": 9.8498,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.860273838043213,
      "learning_rate": 3e-05,
      "loss": 9.0782,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.4377548694610596,
      "learning_rate": 3.5e-05,
      "loss": 8.516,
      "step": 35
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.771414041519165,
      "learning_rate": 4e-05,
      "loss": 8.1584,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5369564294815063,
      "learning_rate": 4.5e-05,
      "loss": 7.9712,
      "step": 45
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.1230754852294922,
      "learning_rate": 5e-05,
      "loss": 7.9062,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8305569291114807,
      "learning_rate": 5.500000000000001e-05,
      "loss": 7.8964,
      "step": 55
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7593777775764465,
      "learning_rate": 6e-05,
      "loss": 7.8687,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.6929013133049011,
      "learning_rate": 6.500000000000001e-05,
      "loss": 7.8158,
      "step": 65
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3948832750320435,
      "learning_rate": 7e-05,
      "loss": 7.722,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.8355083465576172,
      "learning_rate": 7.500000000000001e-05,
      "loss": 7.5869,
      "step": 75
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.5804309844970703,
      "learning_rate": 8e-05,
      "loss": 7.4761,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4202959537506104,
      "learning_rate": 8.5e-05,
      "loss": 7.3409,
      "step": 85
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6333388090133667,
      "learning_rate": 9e-05,
      "loss": 7.148,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2520595788955688,
      "learning_rate": 9.5e-05,
      "loss": 7.0504,
      "step": 95
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.1779870986938477,
      "learning_rate": 0.0001,
      "loss": 6.8871,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.1255111694335938,
      "learning_rate": 9.999972205865686e-05,
      "loss": 6.7967,
      "step": 105
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3718830347061157,
      "learning_rate": 9.999888823771751e-05,
      "loss": 6.6346,
      "step": 110
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3273447751998901,
      "learning_rate": 9.999749854645204e-05,
      "loss": 6.485,
      "step": 115
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.462202310562134,
      "learning_rate": 9.99955530003106e-05,
      "loss": 6.4167,
      "step": 120
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.3976832628250122,
      "learning_rate": 9.99930516209231e-05,
      "loss": 6.2702,
      "step": 125
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.9608620405197144,
      "learning_rate": 9.998999443609897e-05,
      "loss": 6.2682,
      "step": 130
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.4369174242019653,
      "learning_rate": 9.998638147982696e-05,
      "loss": 6.099,
      "step": 135
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.3051707744598389,
      "learning_rate": 9.998221279227467e-05,
      "loss": 6.0768,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.865341305732727,
      "learning_rate": 9.997748841978812e-05,
      "loss": 5.9998,
      "step": 145
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.810935616493225,
      "learning_rate": 9.997220841489122e-05,
      "loss": 5.8743,
      "step": 150
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.2615277767181396,
      "learning_rate": 9.996637283628528e-05,
      "loss": 5.8808,
      "step": 155
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.114527702331543,
      "learning_rate": 9.995998174884821e-05,
      "loss": 5.7969,
      "step": 160
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.029925584793091,
      "learning_rate": 9.995303522363394e-05,
      "loss": 5.7036,
      "step": 165
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.2772178649902344,
      "learning_rate": 9.99455333378715e-05,
      "loss": 5.6773,
      "step": 170
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.9576845169067383,
      "learning_rate": 9.993747617496428e-05,
      "loss": 5.6935,
      "step": 175
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.008150577545166,
      "learning_rate": 9.9928863824489e-05,
      "loss": 5.6154,
      "step": 180
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.7725847959518433,
      "learning_rate": 9.99196963821948e-05,
      "loss": 5.6392,
      "step": 185
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.334122896194458,
      "learning_rate": 9.990997395000217e-05,
      "loss": 5.548,
      "step": 190
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.318779468536377,
      "learning_rate": 9.989969663600169e-05,
      "loss": 5.4917,
      "step": 195
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.765774130821228,
      "learning_rate": 9.9888864554453e-05,
      "loss": 5.4576,
      "step": 200
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.9322354793548584,
      "learning_rate": 9.987747782578342e-05,
      "loss": 5.4416,
      "step": 205
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.2729501724243164,
      "learning_rate": 9.986553657658668e-05,
      "loss": 5.4067,
      "step": 210
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.7171967029571533,
      "learning_rate": 9.985304093962145e-05,
      "loss": 5.2989,
      "step": 215
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.9652690887451172,
      "learning_rate": 9.983999105380988e-05,
      "loss": 5.2895,
      "step": 220
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.945024847984314,
      "learning_rate": 9.982638706423608e-05,
      "loss": 5.2129,
      "step": 225
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.4888384342193604,
      "learning_rate": 9.98122291221445e-05,
      "loss": 5.2575,
      "step": 230
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.6458323001861572,
      "learning_rate": 9.979751738493826e-05,
      "loss": 5.2492,
      "step": 235
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.904470920562744,
      "learning_rate": 9.978225201617732e-05,
      "loss": 5.2304,
      "step": 240
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.577502489089966,
      "learning_rate": 9.976643318557678e-05,
      "loss": 5.1574,
      "step": 245
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.2518582344055176,
      "learning_rate": 9.975006106900495e-05,
      "loss": 5.1468,
      "step": 250
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.210632562637329,
      "learning_rate": 9.973313584848132e-05,
      "loss": 5.1903,
      "step": 255
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.289008855819702,
      "learning_rate": 9.971565771217464e-05,
      "loss": 5.0538,
      "step": 260
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.9316537380218506,
      "learning_rate": 9.969762685440076e-05,
      "loss": 5.0137,
      "step": 265
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.8232439756393433,
      "learning_rate": 9.967904347562054e-05,
      "loss": 5.0465,
      "step": 270
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.632053852081299,
      "learning_rate": 9.965990778243755e-05,
      "loss": 4.9713,
      "step": 275
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.431483745574951,
      "learning_rate": 9.964021998759577e-05,
      "loss": 4.9716,
      "step": 280
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.3636586666107178,
      "learning_rate": 9.961998030997733e-05,
      "loss": 5.0058,
      "step": 285
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.994718551635742,
      "learning_rate": 9.95991889745999e-05,
      "loss": 4.9891,
      "step": 290
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.4368574619293213,
      "learning_rate": 9.957784621261441e-05,
      "loss": 4.9757,
      "step": 295
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.8038134574890137,
      "learning_rate": 9.955595226130226e-05,
      "loss": 4.9921,
      "step": 300
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.0562264919281006,
      "learning_rate": 9.953350736407282e-05,
      "loss": 4.9411,
      "step": 305
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.29128360748291,
      "learning_rate": 9.951051177046069e-05,
      "loss": 4.9772,
      "step": 310
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.8482352495193481,
      "learning_rate": 9.948696573612292e-05,
      "loss": 4.9491,
      "step": 315
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.426596164703369,
      "learning_rate": 9.946286952283618e-05,
      "loss": 4.9207,
      "step": 320
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.412191152572632,
      "learning_rate": 9.943822339849381e-05,
      "loss": 4.8813,
      "step": 325
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.520648956298828,
      "learning_rate": 9.941302763710288e-05,
      "loss": 4.8326,
      "step": 330
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.1452250480651855,
      "learning_rate": 9.938728251878116e-05,
      "loss": 4.8624,
      "step": 335
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.7072198390960693,
      "learning_rate": 9.936098832975393e-05,
      "loss": 4.7228,
      "step": 340
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.610759973526001,
      "learning_rate": 9.933414536235091e-05,
      "loss": 4.7795,
      "step": 345
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.72112238407135,
      "learning_rate": 9.93067539150029e-05,
      "loss": 4.8786,
      "step": 350
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.8034067153930664,
      "learning_rate": 9.927881429223853e-05,
      "loss": 4.7931,
      "step": 355
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.0501575469970703,
      "learning_rate": 9.925032680468085e-05,
      "loss": 4.7893,
      "step": 360
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.5287954807281494,
      "learning_rate": 9.922129176904388e-05,
      "loss": 4.7427,
      "step": 365
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.001993417739868,
      "learning_rate": 9.919170950812911e-05,
      "loss": 4.6804,
      "step": 370
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.9089677333831787,
      "learning_rate": 9.916158035082184e-05,
      "loss": 4.7201,
      "step": 375
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.031919479370117,
      "learning_rate": 9.913090463208763e-05,
      "loss": 4.5473,
      "step": 380
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.068758964538574,
      "learning_rate": 9.90996826929685e-05,
      "loss": 4.637,
      "step": 385
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.9847966432571411,
      "learning_rate": 9.906791488057916e-05,
      "loss": 4.6055,
      "step": 390
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.768234372138977,
      "learning_rate": 9.903560154810313e-05,
      "loss": 4.6769,
      "step": 395
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.6738271713256836,
      "learning_rate": 9.900274305478887e-05,
      "loss": 4.6804,
      "step": 400
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.7261288166046143,
      "learning_rate": 9.896933976594572e-05,
      "loss": 4.7266,
      "step": 405
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.4627445936203003,
      "learning_rate": 9.893539205293989e-05,
      "loss": 4.6272,
      "step": 410
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.5622859001159668,
      "learning_rate": 9.890090029319028e-05,
      "loss": 4.6135,
      "step": 415
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.9910693168640137,
      "learning_rate": 9.886586487016433e-05,
      "loss": 4.5982,
      "step": 420
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.1940295696258545,
      "learning_rate": 9.883028617337378e-05,
      "loss": 4.6473,
      "step": 425
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.3309640884399414,
      "learning_rate": 9.879416459837022e-05,
      "loss": 4.5478,
      "step": 430
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.0572140216827393,
      "learning_rate": 9.875750054674082e-05,
      "loss": 4.5874,
      "step": 435
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.139042854309082,
      "learning_rate": 9.872029442610382e-05,
      "loss": 4.5338,
      "step": 440
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.662381887435913,
      "learning_rate": 9.8682546650104e-05,
      "loss": 4.5524,
      "step": 445
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.8986903429031372,
      "learning_rate": 9.864425763840802e-05,
      "loss": 4.6181,
      "step": 450
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.9163240194320679,
      "learning_rate": 9.860542781669988e-05,
      "loss": 4.5864,
      "step": 455
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.5074962377548218,
      "learning_rate": 9.85660576166761e-05,
      "loss": 4.456,
      "step": 460
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.8273178339004517,
      "learning_rate": 9.852614747604093e-05,
      "loss": 4.4551,
      "step": 465
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.927778720855713,
      "learning_rate": 9.848569783850145e-05,
      "loss": 4.423,
      "step": 470
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.81380295753479,
      "learning_rate": 9.844470915376278e-05,
      "loss": 4.4809,
      "step": 475
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.4844441413879395,
      "learning_rate": 9.840318187752292e-05,
      "loss": 4.4751,
      "step": 480
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.5609166622161865,
      "learning_rate": 9.836111647146771e-05,
      "loss": 4.4466,
      "step": 485
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.7383570671081543,
      "learning_rate": 9.831851340326577e-05,
      "loss": 4.4678,
      "step": 490
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.1461851596832275,
      "learning_rate": 9.82753731465633e-05,
      "loss": 4.423,
      "step": 495
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.8402971029281616,
      "learning_rate": 9.823169618097871e-05,
      "loss": 4.4111,
      "step": 500
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.8172292709350586,
      "learning_rate": 9.81874829920974e-05,
      "loss": 4.4032,
      "step": 505
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.862669825553894,
      "learning_rate": 9.814273407146623e-05,
      "loss": 4.4133,
      "step": 510
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.58668851852417,
      "learning_rate": 9.809744991658829e-05,
      "loss": 4.4304,
      "step": 515
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.267015218734741,
      "learning_rate": 9.805163103091708e-05,
      "loss": 4.4452,
      "step": 520
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.892336130142212,
      "learning_rate": 9.800527792385112e-05,
      "loss": 4.3714,
      "step": 525
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.2870523929595947,
      "learning_rate": 9.79583911107282e-05,
      "loss": 4.3809,
      "step": 530
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.7279003858566284,
      "learning_rate": 9.791097111281968e-05,
      "loss": 4.4289,
      "step": 535
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.5451871156692505,
      "learning_rate": 9.786301845732467e-05,
      "loss": 4.3959,
      "step": 540
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.040153741836548,
      "learning_rate": 9.781453367736418e-05,
      "loss": 4.3958,
      "step": 545
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.7074291706085205,
      "learning_rate": 9.776551731197524e-05,
      "loss": 4.445,
      "step": 550
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.0845582485198975,
      "learning_rate": 9.771596990610478e-05,
      "loss": 4.321,
      "step": 555
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.7364312410354614,
      "learning_rate": 9.766589201060372e-05,
      "loss": 4.3121,
      "step": 560
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.838957667350769,
      "learning_rate": 9.761528418222077e-05,
      "loss": 4.3749,
      "step": 565
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.750223994255066,
      "learning_rate": 9.756414698359624e-05,
      "loss": 4.3366,
      "step": 570
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.4889832735061646,
      "learning_rate": 9.75124809832558e-05,
      "loss": 4.3415,
      "step": 575
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.8685040473937988,
      "learning_rate": 9.746028675560413e-05,
      "loss": 4.3689,
      "step": 580
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.534946084022522,
      "learning_rate": 9.740756488091861e-05,
      "loss": 4.3685,
      "step": 585
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5127125978469849,
      "learning_rate": 9.735431594534277e-05,
      "loss": 4.3401,
      "step": 590
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5884028673171997,
      "learning_rate": 9.730054054087983e-05,
      "loss": 4.3424,
      "step": 595
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.3165289163589478,
      "learning_rate": 9.724623926538612e-05,
      "loss": 4.2607,
      "step": 600
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.4285833835601807,
      "learning_rate": 9.719141272256443e-05,
      "loss": 4.3209,
      "step": 605
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.1819188594818115,
      "learning_rate": 9.713606152195726e-05,
      "loss": 4.3364,
      "step": 610
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.726144552230835,
      "learning_rate": 9.708018627894011e-05,
      "loss": 4.3645,
      "step": 615
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.9843640327453613,
      "learning_rate": 9.702378761471456e-05,
      "loss": 4.238,
      "step": 620
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.620071291923523,
      "learning_rate": 9.696686615630146e-05,
      "loss": 4.2553,
      "step": 625
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.4403271675109863,
      "learning_rate": 9.690942253653385e-05,
      "loss": 4.3202,
      "step": 630
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.4145867824554443,
      "learning_rate": 9.685145739405002e-05,
      "loss": 4.2826,
      "step": 635
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.439998745918274,
      "learning_rate": 9.679297137328634e-05,
      "loss": 4.2418,
      "step": 640
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.131375789642334,
      "learning_rate": 9.673396512447013e-05,
      "loss": 4.2036,
      "step": 645
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.8813244104385376,
      "learning_rate": 9.667443930361247e-05,
      "loss": 4.1535,
      "step": 650
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4201704263687134,
      "learning_rate": 9.661439457250076e-05,
      "loss": 4.2536,
      "step": 655
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3782165050506592,
      "learning_rate": 9.655383159869158e-05,
      "loss": 4.1932,
      "step": 660
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.094642162322998,
      "learning_rate": 9.649275105550309e-05,
      "loss": 4.1604,
      "step": 665
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.7626591920852661,
      "learning_rate": 9.643115362200762e-05,
      "loss": 4.0819,
      "step": 670
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.609782099723816,
      "learning_rate": 9.636903998302409e-05,
      "loss": 4.1571,
      "step": 675
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.287858486175537,
      "learning_rate": 9.630641082911045e-05,
      "loss": 4.1096,
      "step": 680
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.6848094463348389,
      "learning_rate": 9.624326685655593e-05,
      "loss": 4.2902,
      "step": 685
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.5279600620269775,
      "learning_rate": 9.617960876737337e-05,
      "loss": 4.1994,
      "step": 690
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3572136163711548,
      "learning_rate": 9.611543726929134e-05,
      "loss": 4.1961,
      "step": 695
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6097009181976318,
      "learning_rate": 9.605075307574635e-05,
      "loss": 4.1777,
      "step": 700
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.7889363765716553,
      "learning_rate": 9.598555690587487e-05,
      "loss": 4.1743,
      "step": 705
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.6500908136367798,
      "learning_rate": 9.591984948450532e-05,
      "loss": 4.1654,
      "step": 710
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.0012271404266357,
      "learning_rate": 9.585363154215008e-05,
      "loss": 4.1007,
      "step": 715
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.140477418899536,
      "learning_rate": 9.578690381499728e-05,
      "loss": 4.2186,
      "step": 720
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.5439093112945557,
      "learning_rate": 9.571966704490271e-05,
      "loss": 4.1307,
      "step": 725
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.7154203653335571,
      "learning_rate": 9.565192197938148e-05,
      "loss": 4.1912,
      "step": 730
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.53816556930542,
      "learning_rate": 9.558366937159977e-05,
      "loss": 4.2425,
      "step": 735
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.3727734088897705,
      "learning_rate": 9.551490998036646e-05,
      "loss": 4.2211,
      "step": 740
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.5313302278518677,
      "learning_rate": 9.544564457012463e-05,
      "loss": 4.1654,
      "step": 745
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.9445953369140625,
      "learning_rate": 9.537587391094314e-05,
      "loss": 4.2171,
      "step": 750
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.5916863679885864,
      "learning_rate": 9.5305598778508e-05,
      "loss": 4.1061,
      "step": 755
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.3092683553695679,
      "learning_rate": 9.52348199541138e-05,
      "loss": 4.2096,
      "step": 760
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.3142445087432861,
      "learning_rate": 9.516353822465504e-05,
      "loss": 4.1241,
      "step": 765
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.357508897781372,
      "learning_rate": 9.509175438261726e-05,
      "loss": 4.1138,
      "step": 770
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.302988052368164,
      "learning_rate": 9.501946922606838e-05,
      "loss": 4.1249,
      "step": 775
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.1296546459198,
      "learning_rate": 9.494668355864973e-05,
      "loss": 4.1248,
      "step": 780
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.0370583534240723,
      "learning_rate": 9.487339818956716e-05,
      "loss": 4.1247,
      "step": 785
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.6055999994277954,
      "learning_rate": 9.479961393358203e-05,
      "loss": 3.9874,
      "step": 790
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.8076801300048828,
      "learning_rate": 9.472533161100215e-05,
      "loss": 4.1358,
      "step": 795
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.474900245666504,
      "learning_rate": 9.465055204767265e-05,
      "loss": 4.0485,
      "step": 800
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3105722665786743,
      "learning_rate": 9.457527607496685e-05,
      "loss": 4.1364,
      "step": 805
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3706189393997192,
      "learning_rate": 9.44995045297769e-05,
      "loss": 4.0926,
      "step": 810
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.493538737297058,
      "learning_rate": 9.442323825450464e-05,
      "loss": 4.0733,
      "step": 815
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.29449462890625,
      "learning_rate": 9.43464780970521e-05,
      "loss": 4.0523,
      "step": 820
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.291658878326416,
      "learning_rate": 9.426922491081212e-05,
      "loss": 4.1155,
      "step": 825
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2312451601028442,
      "learning_rate": 9.419147955465888e-05,
      "loss": 4.006,
      "step": 830
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.4466781616210938,
      "learning_rate": 9.411324289293832e-05,
      "loss": 4.1482,
      "step": 835
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2307556867599487,
      "learning_rate": 9.403451579545859e-05,
      "loss": 4.086,
      "step": 840
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3094658851623535,
      "learning_rate": 9.395529913748025e-05,
      "loss": 4.0652,
      "step": 845
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.4318170547485352,
      "learning_rate": 9.387559379970672e-05,
      "loss": 4.1354,
      "step": 850
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2737038135528564,
      "learning_rate": 9.379540066827431e-05,
      "loss": 4.0278,
      "step": 855
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1153333187103271,
      "learning_rate": 9.371472063474248e-05,
      "loss": 4.0898,
      "step": 860
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2244912385940552,
      "learning_rate": 9.363355459608394e-05,
      "loss": 3.9959,
      "step": 865
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1208878755569458,
      "learning_rate": 9.355190345467457e-05,
      "loss": 4.0623,
      "step": 870
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1840647459030151,
      "learning_rate": 9.346976811828352e-05,
      "loss": 4.045,
      "step": 875
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.4008674621582031,
      "learning_rate": 9.338714950006297e-05,
      "loss": 4.0959,
      "step": 880
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3445881605148315,
      "learning_rate": 9.330404851853817e-05,
      "loss": 4.02,
      "step": 885
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1713682413101196,
      "learning_rate": 9.3220466097597e-05,
      "loss": 4.0312,
      "step": 890
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2331210374832153,
      "learning_rate": 9.313640316647991e-05,
      "loss": 4.0289,
      "step": 895
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.279057502746582,
      "learning_rate": 9.305186065976945e-05,
      "loss": 4.0998,
      "step": 900
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3163552284240723,
      "learning_rate": 9.296683951737993e-05,
      "loss": 4.0448,
      "step": 905
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3067126274108887,
      "learning_rate": 9.288134068454697e-05,
      "loss": 4.029,
      "step": 910
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.145782470703125,
      "learning_rate": 9.2795365111817e-05,
      "loss": 4.0305,
      "step": 915
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.6717885732650757,
      "learning_rate": 9.270891375503665e-05,
      "loss": 3.9886,
      "step": 920
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3663681745529175,
      "learning_rate": 9.262198757534218e-05,
      "loss": 3.9891,
      "step": 925
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.4202651977539062,
      "learning_rate": 9.253458753914874e-05,
      "loss": 3.9358,
      "step": 930
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2852869033813477,
      "learning_rate": 9.244671461813969e-05,
      "loss": 4.0289,
      "step": 935
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.138818383216858,
      "learning_rate": 9.235836978925572e-05,
      "loss": 4.0046,
      "step": 940
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.1434671878814697,
      "learning_rate": 9.226955403468406e-05,
      "loss": 4.0873,
      "step": 945
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3317791223526,
      "learning_rate": 9.21802683418475e-05,
      "loss": 3.963,
      "step": 950
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2954212427139282,
      "learning_rate": 9.209051370339347e-05,
      "loss": 3.9772,
      "step": 955
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.265075922012329,
      "learning_rate": 9.200029111718295e-05,
      "loss": 3.9925,
      "step": 960
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2149906158447266,
      "learning_rate": 9.190960158627941e-05,
      "loss": 4.0012,
      "step": 965
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2477954626083374,
      "learning_rate": 9.181844611893766e-05,
      "loss": 3.9611,
      "step": 970
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.324196696281433,
      "learning_rate": 9.172682572859261e-05,
      "loss": 3.9677,
      "step": 975
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3045045137405396,
      "learning_rate": 9.163474143384806e-05,
      "loss": 3.918,
      "step": 980
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.8825814723968506,
      "learning_rate": 9.154219425846528e-05,
      "loss": 4.0018,
      "step": 985
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4024105072021484,
      "learning_rate": 9.144918523135175e-05,
      "loss": 3.8843,
      "step": 990
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.1396621465682983,
      "learning_rate": 9.13557153865496e-05,
      "loss": 3.9899,
      "step": 995
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4695543050765991,
      "learning_rate": 9.12617857632242e-05,
      "loss": 3.9376,
      "step": 1000
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0880409479141235,
      "learning_rate": 9.116739740565259e-05,
      "loss": 3.9491,
      "step": 1005
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.1712977886199951,
      "learning_rate": 9.107255136321184e-05,
      "loss": 3.8714,
      "step": 1010
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.2302933931350708,
      "learning_rate": 9.09772486903674e-05,
      "loss": 3.8528,
      "step": 1015
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.258522629737854,
      "learning_rate": 9.08814904466614e-05,
      "loss": 3.9281,
      "step": 1020
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.213845133781433,
      "learning_rate": 9.078527769670085e-05,
      "loss": 3.9133,
      "step": 1025
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.2096773386001587,
      "learning_rate": 9.068861151014575e-05,
      "loss": 3.9051,
      "step": 1030
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3456385135650635,
      "learning_rate": 9.05914929616973e-05,
      "loss": 3.8499,
      "step": 1035
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3120944499969482,
      "learning_rate": 9.04939231310859e-05,
      "loss": 3.932,
      "step": 1040
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.2625679969787598,
      "learning_rate": 9.039590310305914e-05,
      "loss": 3.851,
      "step": 1045
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.229634165763855,
      "learning_rate": 9.029743396736974e-05,
      "loss": 3.8906,
      "step": 1050
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3108216524124146,
      "learning_rate": 9.019851681876348e-05,
      "loss": 3.8976,
      "step": 1055
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.382756233215332,
      "learning_rate": 9.009915275696693e-05,
      "loss": 3.9716,
      "step": 1060
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.498990774154663,
      "learning_rate": 8.999934288667534e-05,
      "loss": 3.931,
      "step": 1065
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.2947123050689697,
      "learning_rate": 8.989908831754028e-05,
      "loss": 3.8841,
      "step": 1070
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.247287392616272,
      "learning_rate": 8.979839016415735e-05,
      "loss": 3.7915,
      "step": 1075
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.2066890001296997,
      "learning_rate": 8.969724954605373e-05,
      "loss": 3.922,
      "step": 1080
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.2116167545318604,
      "learning_rate": 8.959566758767581e-05,
      "loss": 3.8854,
      "step": 1085
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.1037461757659912,
      "learning_rate": 8.949364541837661e-05,
      "loss": 3.8996,
      "step": 1090
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.324324369430542,
      "learning_rate": 8.939118417240329e-05,
      "loss": 3.9283,
      "step": 1095
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.3123801946640015,
      "learning_rate": 8.92882849888845e-05,
      "loss": 3.8173,
      "step": 1100
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.2384531497955322,
      "learning_rate": 8.918494901181773e-05,
      "loss": 3.9028,
      "step": 1105
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.1092414855957031,
      "learning_rate": 8.908117739005659e-05,
      "loss": 3.783,
      "step": 1110
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.040840983390808,
      "learning_rate": 8.897697127729805e-05,
      "loss": 3.867,
      "step": 1115
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.2289977073669434,
      "learning_rate": 8.887233183206957e-05,
      "loss": 3.8533,
      "step": 1120
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.9840658903121948,
      "learning_rate": 8.876726021771627e-05,
      "loss": 3.851,
      "step": 1125
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.1351091861724854,
      "learning_rate": 8.866175760238798e-05,
      "loss": 3.792,
      "step": 1130
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4369754791259766,
      "learning_rate": 8.855582515902625e-05,
      "loss": 3.8564,
      "step": 1135
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.2083522081375122,
      "learning_rate": 8.844946406535131e-05,
      "loss": 3.8876,
      "step": 1140
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.2528421878814697,
      "learning_rate": 8.834267550384893e-05,
      "loss": 3.7933,
      "step": 1145
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.289838433265686,
      "learning_rate": 8.823546066175741e-05,
      "loss": 3.8005,
      "step": 1150
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.1734122037887573,
      "learning_rate": 8.81278207310542e-05,
      "loss": 3.7707,
      "step": 1155
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.1725057363510132,
      "learning_rate": 8.801975690844278e-05,
      "loss": 3.8424,
      "step": 1160
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.9867896437644958,
      "learning_rate": 8.791127039533934e-05,
      "loss": 3.7997,
      "step": 1165
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.1472907066345215,
      "learning_rate": 8.780236239785935e-05,
      "loss": 3.8205,
      "step": 1170
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.1044718027114868,
      "learning_rate": 8.76930341268042e-05,
      "loss": 3.789,
      "step": 1175
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.2645314931869507,
      "learning_rate": 8.758328679764776e-05,
      "loss": 3.9305,
      "step": 1180
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.1153810024261475,
      "learning_rate": 8.747312163052284e-05,
      "loss": 3.7777,
      "step": 1185
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.1703031063079834,
      "learning_rate": 8.736253985020761e-05,
      "loss": 3.819,
      "step": 1190
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.2039633989334106,
      "learning_rate": 8.725154268611203e-05,
      "loss": 3.742,
      "step": 1195
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.057329773902893,
      "learning_rate": 8.714013137226411e-05,
      "loss": 3.7937,
      "step": 1200
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.1288293600082397,
      "learning_rate": 8.702830714729628e-05,
      "loss": 3.8379,
      "step": 1205
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.081825613975525,
      "learning_rate": 8.691607125443153e-05,
      "loss": 3.8629,
      "step": 1210
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.1750764846801758,
      "learning_rate": 8.680342494146967e-05,
      "loss": 3.83,
      "step": 1215
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.1391602754592896,
      "learning_rate": 8.66903694607734e-05,
      "loss": 3.8315,
      "step": 1220
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.4016056060791016,
      "learning_rate": 8.65769060692544e-05,
      "loss": 3.7994,
      "step": 1225
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.0848188400268555,
      "learning_rate": 8.646303602835936e-05,
      "loss": 3.7555,
      "step": 1230
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.025410532951355,
      "learning_rate": 8.634876060405597e-05,
      "loss": 3.8574,
      "step": 1235
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1180078983306885,
      "learning_rate": 8.623408106681884e-05,
      "loss": 3.8189,
      "step": 1240
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.166062593460083,
      "learning_rate": 8.611899869161535e-05,
      "loss": 3.6929,
      "step": 1245
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2254574298858643,
      "learning_rate": 8.600351475789147e-05,
      "loss": 3.8875,
      "step": 1250
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.343238353729248,
      "learning_rate": 8.588763054955764e-05,
      "loss": 3.7674,
      "step": 1255
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1699622869491577,
      "learning_rate": 8.57713473549743e-05,
      "loss": 3.7034,
      "step": 1260
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.0328434705734253,
      "learning_rate": 8.565466646693778e-05,
      "loss": 3.8051,
      "step": 1265
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2840907573699951,
      "learning_rate": 8.553758918266578e-05,
      "loss": 3.756,
      "step": 1270
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.2463798522949219,
      "learning_rate": 8.5420116803783e-05,
      "loss": 3.8213,
      "step": 1275
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.457021951675415,
      "learning_rate": 8.530225063630668e-05,
      "loss": 3.8307,
      "step": 1280
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.196196436882019,
      "learning_rate": 8.518399199063205e-05,
      "loss": 3.7173,
      "step": 1285
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.2081055641174316,
      "learning_rate": 8.50653421815178e-05,
      "loss": 3.76,
      "step": 1290
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.0329689979553223,
      "learning_rate": 8.494630252807138e-05,
      "loss": 3.7521,
      "step": 1295
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.0779180526733398,
      "learning_rate": 8.482687435373449e-05,
      "loss": 3.694,
      "step": 1300
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.2917273044586182,
      "learning_rate": 8.470705898626817e-05,
      "loss": 3.7758,
      "step": 1305
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.575312852859497,
      "learning_rate": 8.458685775773822e-05,
      "loss": 3.7461,
      "step": 1310
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.2756739854812622,
      "learning_rate": 8.446627200450025e-05,
      "loss": 3.764,
      "step": 1315
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.1574435234069824,
      "learning_rate": 8.434530306718493e-05,
      "loss": 3.776,
      "step": 1320
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.1152746677398682,
      "learning_rate": 8.4223952290683e-05,
      "loss": 3.8227,
      "step": 1325
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.215476393699646,
      "learning_rate": 8.41022210241304e-05,
      "loss": 3.684,
      "step": 1330
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.2368085384368896,
      "learning_rate": 8.398011062089316e-05,
      "loss": 3.7279,
      "step": 1335
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.1022911071777344,
      "learning_rate": 8.385762243855249e-05,
      "loss": 3.7327,
      "step": 1340
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.9960315227508545,
      "learning_rate": 8.373475783888958e-05,
      "loss": 3.6113,
      "step": 1345
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.3093613386154175,
      "learning_rate": 8.36115181878705e-05,
      "loss": 3.7385,
      "step": 1350
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.1868112087249756,
      "learning_rate": 8.348790485563101e-05,
      "loss": 3.7108,
      "step": 1355
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.20425546169281,
      "learning_rate": 8.336391921646134e-05,
      "loss": 3.6743,
      "step": 1360
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.1599810123443604,
      "learning_rate": 8.323956264879089e-05,
      "loss": 3.7433,
      "step": 1365
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.2243181467056274,
      "learning_rate": 8.311483653517294e-05,
      "loss": 3.6934,
      "step": 1370
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.073176383972168,
      "learning_rate": 8.298974226226919e-05,
      "loss": 3.697,
      "step": 1375
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.175176739692688,
      "learning_rate": 8.28642812208345e-05,
      "loss": 3.7262,
      "step": 1380
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.068116307258606,
      "learning_rate": 8.273845480570123e-05,
      "loss": 3.6871,
      "step": 1385
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.1474370956420898,
      "learning_rate": 8.26122644157639e-05,
      "loss": 3.7306,
      "step": 1390
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.140490174293518,
      "learning_rate": 8.248571145396362e-05,
      "loss": 3.6987,
      "step": 1395
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.1814143657684326,
      "learning_rate": 8.235879732727236e-05,
      "loss": 3.7757,
      "step": 1400
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.2589894533157349,
      "learning_rate": 8.223152344667745e-05,
      "loss": 3.7821,
      "step": 1405
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.0849429368972778,
      "learning_rate": 8.21038912271658e-05,
      "loss": 3.6301,
      "step": 1410
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.1057863235473633,
      "learning_rate": 8.197590208770824e-05,
      "loss": 3.7313,
      "step": 1415
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.1446341276168823,
      "learning_rate": 8.184755745124371e-05,
      "loss": 3.7866,
      "step": 1420
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.1112569570541382,
      "learning_rate": 8.171885874466342e-05,
      "loss": 3.6763,
      "step": 1425
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.0507431030273438,
      "learning_rate": 8.158980739879507e-05,
      "loss": 3.7954,
      "step": 1430
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.2522321939468384,
      "learning_rate": 8.146040484838677e-05,
      "loss": 3.6357,
      "step": 1435
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.060377836227417,
      "learning_rate": 8.133065253209132e-05,
      "loss": 3.6315,
      "step": 1440
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.035387396812439,
      "learning_rate": 8.120055189245e-05,
      "loss": 3.6919,
      "step": 1445
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.0047308206558228,
      "learning_rate": 8.10701043758767e-05,
      "loss": 3.6445,
      "step": 1450
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.10782790184021,
      "learning_rate": 8.093931143264174e-05,
      "loss": 3.6457,
      "step": 1455
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.0221713781356812,
      "learning_rate": 8.080817451685576e-05,
      "loss": 3.743,
      "step": 1460
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.0856046676635742,
      "learning_rate": 8.067669508645356e-05,
      "loss": 3.6569,
      "step": 1465
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.1059452295303345,
      "learning_rate": 8.054487460317797e-05,
      "loss": 3.7369,
      "step": 1470
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.1251044273376465,
      "learning_rate": 8.041271453256345e-05,
      "loss": 3.6478,
      "step": 1475
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0519945621490479,
      "learning_rate": 8.02802163439199e-05,
      "loss": 3.6413,
      "step": 1480
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.137978196144104,
      "learning_rate": 8.01473815103163e-05,
      "loss": 3.6362,
      "step": 1485
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.045811653137207,
      "learning_rate": 8.001421150856434e-05,
      "loss": 3.6628,
      "step": 1490
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.188126802444458,
      "learning_rate": 7.988070781920197e-05,
      "loss": 3.7502,
      "step": 1495
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.1069587469100952,
      "learning_rate": 7.9746871926477e-05,
      "loss": 3.6779,
      "step": 1500
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0773390531539917,
      "learning_rate": 7.961270531833052e-05,
      "loss": 3.6953,
      "step": 1505
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0676312446594238,
      "learning_rate": 7.947820948638045e-05,
      "loss": 3.7452,
      "step": 1510
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0171785354614258,
      "learning_rate": 7.934338592590486e-05,
      "loss": 3.6311,
      "step": 1515
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0171067714691162,
      "learning_rate": 7.92082361358254e-05,
      "loss": 3.6453,
      "step": 1520
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.1444722414016724,
      "learning_rate": 7.907276161869065e-05,
      "loss": 3.6623,
      "step": 1525
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.1217622756958008,
      "learning_rate": 7.893696388065936e-05,
      "loss": 3.6992,
      "step": 1530
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0387611389160156,
      "learning_rate": 7.88008444314838e-05,
      "loss": 3.6233,
      "step": 1535
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.9862802028656006,
      "learning_rate": 7.866440478449283e-05,
      "loss": 3.5765,
      "step": 1540
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.1087441444396973,
      "learning_rate": 7.852764645657522e-05,
      "loss": 3.6472,
      "step": 1545
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0384817123413086,
      "learning_rate": 7.839057096816271e-05,
      "loss": 3.7186,
      "step": 1550
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0545164346694946,
      "learning_rate": 7.82531798432131e-05,
      "loss": 3.6654,
      "step": 1555
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.1012707948684692,
      "learning_rate": 7.811547460919333e-05,
      "loss": 3.6155,
      "step": 1560
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9383161067962646,
      "learning_rate": 7.797745679706254e-05,
      "loss": 3.6554,
      "step": 1565
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9553996920585632,
      "learning_rate": 7.783912794125496e-05,
      "loss": 3.6888,
      "step": 1570
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.117811679840088,
      "learning_rate": 7.770048957966291e-05,
      "loss": 3.6646,
      "step": 1575
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.0329620838165283,
      "learning_rate": 7.756154325361967e-05,
      "loss": 3.5898,
      "step": 1580
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.8986888527870178,
      "learning_rate": 7.74222905078824e-05,
      "loss": 3.585,
      "step": 1585
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9581660628318787,
      "learning_rate": 7.728273289061489e-05,
      "loss": 3.6762,
      "step": 1590
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9644125699996948,
      "learning_rate": 7.714287195337044e-05,
      "loss": 3.5898,
      "step": 1595
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.0127379894256592,
      "learning_rate": 7.700270925107448e-05,
      "loss": 3.7051,
      "step": 1600
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.0924148559570312,
      "learning_rate": 7.686224634200742e-05,
      "loss": 3.7056,
      "step": 1605
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.0332456827163696,
      "learning_rate": 7.672148478778722e-05,
      "loss": 3.6393,
      "step": 1610
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.129856824874878,
      "learning_rate": 7.658042615335212e-05,
      "loss": 3.6422,
      "step": 1615
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.075238585472107,
      "learning_rate": 7.643907200694318e-05,
      "loss": 3.5333,
      "step": 1620
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.1756746768951416,
      "learning_rate": 7.629742392008684e-05,
      "loss": 3.5834,
      "step": 1625
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.1802172660827637,
      "learning_rate": 7.615548346757749e-05,
      "loss": 3.5762,
      "step": 1630
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.1827501058578491,
      "learning_rate": 7.60132522274599e-05,
      "loss": 3.6299,
      "step": 1635
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.2258820533752441,
      "learning_rate": 7.587073178101178e-05,
      "loss": 3.7182,
      "step": 1640
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.035818338394165,
      "learning_rate": 7.572792371272609e-05,
      "loss": 3.5232,
      "step": 1645
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.0153076648712158,
      "learning_rate": 7.55848296102935e-05,
      "loss": 3.6142,
      "step": 1650
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.1396682262420654,
      "learning_rate": 7.544145106458465e-05,
      "loss": 3.6179,
      "step": 1655
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9990943074226379,
      "learning_rate": 7.529778966963259e-05,
      "loss": 3.6462,
      "step": 1660
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.1192935705184937,
      "learning_rate": 7.515384702261496e-05,
      "loss": 3.6308,
      "step": 1665
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.3080693483352661,
      "learning_rate": 7.500962472383627e-05,
      "loss": 3.6369,
      "step": 1670
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.019212245941162,
      "learning_rate": 7.486512437671011e-05,
      "loss": 3.6282,
      "step": 1675
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.9433236122131348,
      "learning_rate": 7.472034758774128e-05,
      "loss": 3.5127,
      "step": 1680
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.0534073114395142,
      "learning_rate": 7.457529596650797e-05,
      "loss": 3.6323,
      "step": 1685
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.1378173828125,
      "learning_rate": 7.442997112564392e-05,
      "loss": 3.5825,
      "step": 1690
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.0004600286483765,
      "learning_rate": 7.428437468082037e-05,
      "loss": 3.524,
      "step": 1695
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.1865243911743164,
      "learning_rate": 7.413850825072817e-05,
      "loss": 3.6502,
      "step": 1700
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.0776444673538208,
      "learning_rate": 7.39923734570598e-05,
      "loss": 3.5899,
      "step": 1705
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.0138128995895386,
      "learning_rate": 7.384597192449126e-05,
      "loss": 3.6673,
      "step": 1710
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.1027905941009521,
      "learning_rate": 7.369930528066412e-05,
      "loss": 3.6207,
      "step": 1715
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.036793828010559,
      "learning_rate": 7.355237515616732e-05,
      "loss": 3.5804,
      "step": 1720
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.0655146837234497,
      "learning_rate": 7.340518318451914e-05,
      "loss": 3.6235,
      "step": 1725
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9288371205329895,
      "learning_rate": 7.325773100214893e-05,
      "loss": 3.5938,
      "step": 1730
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9640668630599976,
      "learning_rate": 7.311002024837899e-05,
      "loss": 3.5945,
      "step": 1735
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9216758012771606,
      "learning_rate": 7.296205256540633e-05,
      "loss": 3.6156,
      "step": 1740
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.1537082195281982,
      "learning_rate": 7.281382959828443e-05,
      "loss": 3.5763,
      "step": 1745
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9384089112281799,
      "learning_rate": 7.26653529949049e-05,
      "loss": 3.6599,
      "step": 1750
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.0455132722854614,
      "learning_rate": 7.25166244059792e-05,
      "loss": 3.592,
      "step": 1755
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.1599528789520264,
      "learning_rate": 7.236764548502029e-05,
      "loss": 3.6206,
      "step": 1760
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.2807819843292236,
      "learning_rate": 7.221841788832421e-05,
      "loss": 3.573,
      "step": 1765
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.999480128288269,
      "learning_rate": 7.206894327495173e-05,
      "loss": 3.6414,
      "step": 1770
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.0415242910385132,
      "learning_rate": 7.191922330670982e-05,
      "loss": 3.5397,
      "step": 1775
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9353726506233215,
      "learning_rate": 7.176925964813326e-05,
      "loss": 3.6543,
      "step": 1780
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9449454545974731,
      "learning_rate": 7.161905396646607e-05,
      "loss": 3.5735,
      "step": 1785
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9990008473396301,
      "learning_rate": 7.146860793164299e-05,
      "loss": 3.5328,
      "step": 1790
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.05113685131073,
      "learning_rate": 7.131792321627098e-05,
      "loss": 3.5401,
      "step": 1795
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9677740335464478,
      "learning_rate": 7.116700149561048e-05,
      "loss": 3.577,
      "step": 1800
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9737529754638672,
      "learning_rate": 7.101584444755696e-05,
      "loss": 3.5982,
      "step": 1805
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9464470148086548,
      "learning_rate": 7.086445375262212e-05,
      "loss": 3.4938,
      "step": 1810
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9980365633964539,
      "learning_rate": 7.071283109391528e-05,
      "loss": 3.554,
      "step": 1815
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9686562418937683,
      "learning_rate": 7.056097815712466e-05,
      "loss": 3.494,
      "step": 1820
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9243375658988953,
      "learning_rate": 7.040889663049862e-05,
      "loss": 3.5385,
      "step": 1825
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.0450866222381592,
      "learning_rate": 7.025658820482693e-05,
      "loss": 3.604,
      "step": 1830
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.4128719568252563,
      "learning_rate": 7.010405457342192e-05,
      "loss": 3.6343,
      "step": 1835
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9513347148895264,
      "learning_rate": 6.995129743209967e-05,
      "loss": 3.4879,
      "step": 1840
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.8341937065124512,
      "learning_rate": 6.97983184791612e-05,
      "loss": 3.5948,
      "step": 1845
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9494589567184448,
      "learning_rate": 6.964511941537355e-05,
      "loss": 3.4882,
      "step": 1850
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9091277718544006,
      "learning_rate": 6.949170194395083e-05,
      "loss": 3.6224,
      "step": 1855
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9695673584938049,
      "learning_rate": 6.933806777053536e-05,
      "loss": 3.5524,
      "step": 1860
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.0510302782058716,
      "learning_rate": 6.918421860317872e-05,
      "loss": 3.5681,
      "step": 1865
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9663901329040527,
      "learning_rate": 6.903015615232263e-05,
      "loss": 3.5877,
      "step": 1870
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.8571495413780212,
      "learning_rate": 6.887588213078012e-05,
      "loss": 3.4845,
      "step": 1875
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9498350620269775,
      "learning_rate": 6.87213982537163e-05,
      "loss": 3.6167,
      "step": 1880
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9438138604164124,
      "learning_rate": 6.856670623862943e-05,
      "loss": 3.6043,
      "step": 1885
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.8803019523620605,
      "learning_rate": 6.841180780533179e-05,
      "loss": 3.5352,
      "step": 1890
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9197990894317627,
      "learning_rate": 6.82567046759305e-05,
      "loss": 3.5607,
      "step": 1895
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9768213033676147,
      "learning_rate": 6.810139857480844e-05,
      "loss": 3.5614,
      "step": 1900
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.995598316192627,
      "learning_rate": 6.794589122860509e-05,
      "loss": 3.4622,
      "step": 1905
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.8904504776000977,
      "learning_rate": 6.779018436619725e-05,
      "loss": 3.4692,
      "step": 1910
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9814966320991516,
      "learning_rate": 6.763427971867992e-05,
      "loss": 3.5498,
      "step": 1915
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9564259052276611,
      "learning_rate": 6.747817901934699e-05,
      "loss": 3.5178,
      "step": 1920
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.0320782661437988,
      "learning_rate": 6.732188400367197e-05,
      "loss": 3.585,
      "step": 1925
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.04094660282135,
      "learning_rate": 6.716539640928871e-05,
      "loss": 3.6019,
      "step": 1930
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9763439893722534,
      "learning_rate": 6.70087179759721e-05,
      "loss": 3.5412,
      "step": 1935
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9772986173629761,
      "learning_rate": 6.685185044561874e-05,
      "loss": 3.4934,
      "step": 1940
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9123693108558655,
      "learning_rate": 6.669479556222747e-05,
      "loss": 3.5201,
      "step": 1945
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.0170071125030518,
      "learning_rate": 6.653755507188013e-05,
      "loss": 3.4898,
      "step": 1950
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9429283142089844,
      "learning_rate": 6.638013072272205e-05,
      "loss": 3.5073,
      "step": 1955
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9589207768440247,
      "learning_rate": 6.622252426494259e-05,
      "loss": 3.5705,
      "step": 1960
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9630715847015381,
      "learning_rate": 6.606473745075581e-05,
      "loss": 3.5051,
      "step": 1965
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9157508611679077,
      "learning_rate": 6.590677203438084e-05,
      "loss": 3.4814,
      "step": 1970
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9560885429382324,
      "learning_rate": 6.574862977202252e-05,
      "loss": 3.4907,
      "step": 1975
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.8377314209938049,
      "learning_rate": 6.559031242185174e-05,
      "loss": 3.4926,
      "step": 1980
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.048364281654358,
      "learning_rate": 6.543182174398597e-05,
      "loss": 3.4779,
      "step": 1985
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9476514458656311,
      "learning_rate": 6.52731595004697e-05,
      "loss": 3.5114,
      "step": 1990
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.9079380631446838,
      "learning_rate": 6.51143274552548e-05,
      "loss": 3.4975,
      "step": 1995
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9813776612281799,
      "learning_rate": 6.495532737418098e-05,
      "loss": 3.529,
      "step": 2000
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.8631868362426758,
      "learning_rate": 6.479616102495605e-05,
      "loss": 3.5393,
      "step": 2005
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9297569394111633,
      "learning_rate": 6.463683017713638e-05,
      "loss": 3.5401,
      "step": 2010
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9774935245513916,
      "learning_rate": 6.447733660210715e-05,
      "loss": 3.4963,
      "step": 2015
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.0512644052505493,
      "learning_rate": 6.431768207306272e-05,
      "loss": 3.506,
      "step": 2020
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9815961122512817,
      "learning_rate": 6.415786836498684e-05,
      "loss": 3.5325,
      "step": 2025
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.0608659982681274,
      "learning_rate": 6.399789725463298e-05,
      "loss": 3.4853,
      "step": 2030
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9495364427566528,
      "learning_rate": 6.383777052050458e-05,
      "loss": 3.5141,
      "step": 2035
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.97361820936203,
      "learning_rate": 6.367748994283518e-05,
      "loss": 3.5453,
      "step": 2040
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9082745909690857,
      "learning_rate": 6.351705730356877e-05,
      "loss": 3.4761,
      "step": 2045
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.8285298943519592,
      "learning_rate": 6.335647438633987e-05,
      "loss": 3.5024,
      "step": 2050
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.8225826025009155,
      "learning_rate": 6.319574297645374e-05,
      "loss": 3.5347,
      "step": 2055
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9011406898498535,
      "learning_rate": 6.303486486086654e-05,
      "loss": 3.5243,
      "step": 2060
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9474372267723083,
      "learning_rate": 6.287384182816546e-05,
      "loss": 3.4822,
      "step": 2065
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.0432955026626587,
      "learning_rate": 6.271267566854883e-05,
      "loss": 3.5558,
      "step": 2070
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.8962119817733765,
      "learning_rate": 6.255136817380618e-05,
      "loss": 3.506,
      "step": 2075
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.8903326392173767,
      "learning_rate": 6.23899211372984e-05,
      "loss": 3.5284,
      "step": 2080
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9178863167762756,
      "learning_rate": 6.222833635393772e-05,
      "loss": 3.5501,
      "step": 2085
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.8843467235565186,
      "learning_rate": 6.206661562016782e-05,
      "loss": 3.47,
      "step": 2090
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.8805639147758484,
      "learning_rate": 6.190476073394382e-05,
      "loss": 3.5827,
      "step": 2095
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.8859938383102417,
      "learning_rate": 6.17427734947123e-05,
      "loss": 3.4916,
      "step": 2100
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9646418690681458,
      "learning_rate": 6.158065570339127e-05,
      "loss": 3.5029,
      "step": 2105
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.8537634015083313,
      "learning_rate": 6.141840916235021e-05,
      "loss": 3.5093,
      "step": 2110
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.956503689289093,
      "learning_rate": 6.125603567539001e-05,
      "loss": 3.5195,
      "step": 2115
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.0732929706573486,
      "learning_rate": 6.109353704772284e-05,
      "loss": 3.5231,
      "step": 2120
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9116024374961853,
      "learning_rate": 6.0930915085952164e-05,
      "loss": 3.4417,
      "step": 2125
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.8402517437934875,
      "learning_rate": 6.076817159805267e-05,
      "loss": 3.4525,
      "step": 2130
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.928032398223877,
      "learning_rate": 6.06053083933501e-05,
      "loss": 3.4627,
      "step": 2135
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.9409497976303101,
      "learning_rate": 6.044232728250116e-05,
      "loss": 3.5356,
      "step": 2140
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.1118775606155396,
      "learning_rate": 6.027923007747339e-05,
      "loss": 3.4684,
      "step": 2145
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9488819241523743,
      "learning_rate": 6.011601859152506e-05,
      "loss": 3.495,
      "step": 2150
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.8869526982307434,
      "learning_rate": 5.995269463918495e-05,
      "loss": 3.4509,
      "step": 2155
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.8808426856994629,
      "learning_rate": 5.97892600362322e-05,
      "loss": 3.4256,
      "step": 2160
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9103177785873413,
      "learning_rate": 5.962571659967614e-05,
      "loss": 3.4464,
      "step": 2165
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9502519369125366,
      "learning_rate": 5.946206614773606e-05,
      "loss": 3.4684,
      "step": 2170
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.091119408607483,
      "learning_rate": 5.929831049982103e-05,
      "loss": 3.5087,
      "step": 2175
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9751874804496765,
      "learning_rate": 5.9134451476509633e-05,
      "loss": 3.3998,
      "step": 2180
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9701889753341675,
      "learning_rate": 5.897049089952974e-05,
      "loss": 3.5021,
      "step": 2185
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.1176971197128296,
      "learning_rate": 5.880643059173826e-05,
      "loss": 3.459,
      "step": 2190
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.982542097568512,
      "learning_rate": 5.864227237710093e-05,
      "loss": 3.3349,
      "step": 2195
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9292042255401611,
      "learning_rate": 5.847801808067189e-05,
      "loss": 3.422,
      "step": 2200
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.939666748046875,
      "learning_rate": 5.831366952857357e-05,
      "loss": 3.4601,
      "step": 2205
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9310621619224548,
      "learning_rate": 5.814922854797622e-05,
      "loss": 3.4736,
      "step": 2210
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9938127994537354,
      "learning_rate": 5.798469696707775e-05,
      "loss": 3.5266,
      "step": 2215
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.0024408102035522,
      "learning_rate": 5.782007661508331e-05,
      "loss": 3.5674,
      "step": 2220
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9908074140548706,
      "learning_rate": 5.765536932218495e-05,
      "loss": 3.4635,
      "step": 2225
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9737017750740051,
      "learning_rate": 5.7490576919541315e-05,
      "loss": 3.4767,
      "step": 2230
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.8988921046257019,
      "learning_rate": 5.732570123925729e-05,
      "loss": 3.4504,
      "step": 2235
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.8467938899993896,
      "learning_rate": 5.7160744114363593e-05,
      "loss": 3.4944,
      "step": 2240
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.0359325408935547,
      "learning_rate": 5.699570737879641e-05,
      "loss": 3.4789,
      "step": 2245
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.0919708013534546,
      "learning_rate": 5.683059286737702e-05,
      "loss": 3.4604,
      "step": 2250
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.081618309020996,
      "learning_rate": 5.666540241579139e-05,
      "loss": 3.4671,
      "step": 2255
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9050670266151428,
      "learning_rate": 5.6500137860569766e-05,
      "loss": 3.5274,
      "step": 2260
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9571909308433533,
      "learning_rate": 5.633480103906624e-05,
      "loss": 3.5328,
      "step": 2265
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.8787181973457336,
      "learning_rate": 5.616939378943834e-05,
      "loss": 3.4774,
      "step": 2270
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.8168122172355652,
      "learning_rate": 5.6003917950626595e-05,
      "loss": 3.4833,
      "step": 2275
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9859710931777954,
      "learning_rate": 5.583837536233407e-05,
      "loss": 3.4339,
      "step": 2280
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9478222131729126,
      "learning_rate": 5.567276786500596e-05,
      "loss": 3.4099,
      "step": 2285
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.020331621170044,
      "learning_rate": 5.5507097299809054e-05,
      "loss": 3.4251,
      "step": 2290
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9651831388473511,
      "learning_rate": 5.534136550861133e-05,
      "loss": 3.5135,
      "step": 2295
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.0853240489959717,
      "learning_rate": 5.5175574333961465e-05,
      "loss": 3.4398,
      "step": 2300
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.000195026397705,
      "learning_rate": 5.500972561906832e-05,
      "loss": 3.4936,
      "step": 2305
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8865319490432739,
      "learning_rate": 5.484382120778048e-05,
      "loss": 3.5513,
      "step": 2310
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8763468861579895,
      "learning_rate": 5.467786294456575e-05,
      "loss": 3.398,
      "step": 2315
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.861831545829773,
      "learning_rate": 5.451185267449061e-05,
      "loss": 3.4215,
      "step": 2320
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9110276699066162,
      "learning_rate": 5.43457922431998e-05,
      "loss": 3.4848,
      "step": 2325
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.135488748550415,
      "learning_rate": 5.417968349689566e-05,
      "loss": 3.4418,
      "step": 2330
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9430781602859497,
      "learning_rate": 5.401352828231772e-05,
      "loss": 3.4883,
      "step": 2335
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.8660580515861511,
      "learning_rate": 5.384732844672211e-05,
      "loss": 3.4674,
      "step": 2340
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.8980595469474792,
      "learning_rate": 5.368108583786107e-05,
      "loss": 3.5509,
      "step": 2345
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.8920612931251526,
      "learning_rate": 5.3514802303962344e-05,
      "loss": 3.4593,
      "step": 2350
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.3392480611801147,
      "learning_rate": 5.334847969370868e-05,
      "loss": 3.5069,
      "step": 2355
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.1390236616134644,
      "learning_rate": 5.3182119856217284e-05,
      "loss": 3.4267,
      "step": 2360
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.8977495431900024,
      "learning_rate": 5.3015724641019214e-05,
      "loss": 3.4173,
      "step": 2365
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.8903648853302002,
      "learning_rate": 5.284929589803884e-05,
      "loss": 3.4099,
      "step": 2370
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.926061749458313,
      "learning_rate": 5.2682835477573336e-05,
      "loss": 3.3682,
      "step": 2375
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.061814546585083,
      "learning_rate": 5.2516345230271965e-05,
      "loss": 3.4584,
      "step": 2380
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.6410764455795288,
      "learning_rate": 5.234982700711569e-05,
      "loss": 3.4347,
      "step": 2385
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8893712759017944,
      "learning_rate": 5.218328265939643e-05,
      "loss": 3.453,
      "step": 2390
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.878666341304779,
      "learning_rate": 5.201671403869657e-05,
      "loss": 3.4353,
      "step": 2395
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8680508732795715,
      "learning_rate": 5.1850122996868366e-05,
      "loss": 3.355,
      "step": 2400
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.049208402633667,
      "learning_rate": 5.168351138601334e-05,
      "loss": 3.4118,
      "step": 2405
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9009265899658203,
      "learning_rate": 5.1516881058461675e-05,
      "loss": 3.4656,
      "step": 2410
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8777670860290527,
      "learning_rate": 5.135023386675166e-05,
      "loss": 3.3768,
      "step": 2415
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8929083943367004,
      "learning_rate": 5.118357166360906e-05,
      "loss": 3.4334,
      "step": 2420
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8720307946205139,
      "learning_rate": 5.101689630192655e-05,
      "loss": 3.414,
      "step": 2425
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8738992214202881,
      "learning_rate": 5.085020963474307e-05,
      "loss": 3.4644,
      "step": 2430
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6024993658065796,
      "learning_rate": 5.068351351522329e-05,
      "loss": 3.3989,
      "step": 2435
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9233328700065613,
      "learning_rate": 5.0516809796636935e-05,
      "loss": 3.493,
      "step": 2440
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9974755644798279,
      "learning_rate": 5.035010033233821e-05,
      "loss": 3.4787,
      "step": 2445
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9158952832221985,
      "learning_rate": 5.018338697574523e-05,
      "loss": 3.4315,
      "step": 2450
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8736887574195862,
      "learning_rate": 5.0016671580319354e-05,
      "loss": 3.5099,
      "step": 2455
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8213635087013245,
      "learning_rate": 4.984995599954461e-05,
      "loss": 3.3743,
      "step": 2460
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9550583362579346,
      "learning_rate": 4.968324208690712e-05,
      "loss": 3.3991,
      "step": 2465
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.8275715708732605,
      "learning_rate": 4.951653169587441e-05,
      "loss": 3.4345,
      "step": 2470
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9144462943077087,
      "learning_rate": 4.93498266798749e-05,
      "loss": 3.4527,
      "step": 2475
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9548483490943909,
      "learning_rate": 4.918312889227722e-05,
      "loss": 3.3642,
      "step": 2480
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9061784148216248,
      "learning_rate": 4.901644018636966e-05,
      "loss": 3.4216,
      "step": 2485
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9008503556251526,
      "learning_rate": 4.8849762415339526e-05,
      "loss": 3.4249,
      "step": 2490
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.996560275554657,
      "learning_rate": 4.868309743225256e-05,
      "loss": 3.4145,
      "step": 2495
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9953149557113647,
      "learning_rate": 4.851644709003233e-05,
      "loss": 3.544,
      "step": 2500
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.8433133363723755,
      "learning_rate": 4.834981324143964e-05,
      "loss": 3.4518,
      "step": 2505
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9231517910957336,
      "learning_rate": 4.818319773905191e-05,
      "loss": 3.3918,
      "step": 2510
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9039626121520996,
      "learning_rate": 4.801660243524261e-05,
      "loss": 3.3498,
      "step": 2515
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.8915175795555115,
      "learning_rate": 4.7850029182160626e-05,
      "loss": 3.386,
      "step": 2520
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.0723402500152588,
      "learning_rate": 4.768347983170973e-05,
      "loss": 3.3958,
      "step": 2525
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9475720524787903,
      "learning_rate": 4.7516956235527884e-05,
      "loss": 3.3698,
      "step": 2530
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8335372805595398,
      "learning_rate": 4.735046024496682e-05,
      "loss": 3.3022,
      "step": 2535
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8657789826393127,
      "learning_rate": 4.7183993711071286e-05,
      "loss": 3.4883,
      "step": 2540
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.009423017501831,
      "learning_rate": 4.7017558484558554e-05,
      "loss": 3.3994,
      "step": 2545
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9131430983543396,
      "learning_rate": 4.6851156415797844e-05,
      "loss": 3.4176,
      "step": 2550
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8437486886978149,
      "learning_rate": 4.6684789354789746e-05,
      "loss": 3.37,
      "step": 2555
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8856005072593689,
      "learning_rate": 4.651845915114563e-05,
      "loss": 3.4163,
      "step": 2560
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.8783836960792542,
      "learning_rate": 4.6352167654067095e-05,
      "loss": 3.3511,
      "step": 2565
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9118834733963013,
      "learning_rate": 4.618591671232544e-05,
      "loss": 3.444,
      "step": 2570
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8706138730049133,
      "learning_rate": 4.601970817424106e-05,
      "loss": 3.3979,
      "step": 2575
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8939517140388489,
      "learning_rate": 4.585354388766292e-05,
      "loss": 3.3708,
      "step": 2580
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8935962915420532,
      "learning_rate": 4.568742569994802e-05,
      "loss": 3.3578,
      "step": 2585
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8912794589996338,
      "learning_rate": 4.552135545794086e-05,
      "loss": 3.5254,
      "step": 2590
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8823184967041016,
      "learning_rate": 4.535533500795288e-05,
      "loss": 3.4387,
      "step": 2595
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9021984338760376,
      "learning_rate": 4.5189366195741953e-05,
      "loss": 3.3914,
      "step": 2600
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8895813226699829,
      "learning_rate": 4.502345086649186e-05,
      "loss": 3.4546,
      "step": 2605
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8331109285354614,
      "learning_rate": 4.485759086479179e-05,
      "loss": 3.4014,
      "step": 2610
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8797712922096252,
      "learning_rate": 4.469178803461579e-05,
      "loss": 3.4068,
      "step": 2615
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.8429707884788513,
      "learning_rate": 4.4526044219302326e-05,
      "loss": 3.473,
      "step": 2620
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.94316565990448,
      "learning_rate": 4.4360361261533745e-05,
      "loss": 3.3132,
      "step": 2625
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8813465237617493,
      "learning_rate": 4.419474100331579e-05,
      "loss": 3.4597,
      "step": 2630
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8824898600578308,
      "learning_rate": 4.402918528595715e-05,
      "loss": 3.4065,
      "step": 2635
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8799294829368591,
      "learning_rate": 4.386369595004896e-05,
      "loss": 3.4918,
      "step": 2640
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9282616972923279,
      "learning_rate": 4.3698274835444354e-05,
      "loss": 3.339,
      "step": 2645
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8519152402877808,
      "learning_rate": 4.3532923781238e-05,
      "loss": 3.448,
      "step": 2650
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9007513523101807,
      "learning_rate": 4.336764462574566e-05,
      "loss": 3.3405,
      "step": 2655
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.895147442817688,
      "learning_rate": 4.320243920648376e-05,
      "loss": 3.3407,
      "step": 2660
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.8479480147361755,
      "learning_rate": 4.303730936014894e-05,
      "loss": 3.4386,
      "step": 2665
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9062541723251343,
      "learning_rate": 4.287225692259765e-05,
      "loss": 3.3353,
      "step": 2670
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9131118655204773,
      "learning_rate": 4.270728372882575e-05,
      "loss": 3.3456,
      "step": 2675
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9297282695770264,
      "learning_rate": 4.254239161294804e-05,
      "loss": 3.4446,
      "step": 2680
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9151642322540283,
      "learning_rate": 4.237758240817802e-05,
      "loss": 3.4011,
      "step": 2685
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9209034442901611,
      "learning_rate": 4.2212857946807336e-05,
      "loss": 3.3808,
      "step": 2690
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.8558739423751831,
      "learning_rate": 4.2048220060185516e-05,
      "loss": 3.475,
      "step": 2695
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9316113591194153,
      "learning_rate": 4.188367057869957e-05,
      "loss": 3.4478,
      "step": 2700
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9140479564666748,
      "learning_rate": 4.171921133175365e-05,
      "loss": 3.4139,
      "step": 2705
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.8449094295501709,
      "learning_rate": 4.155484414774872e-05,
      "loss": 3.4462,
      "step": 2710
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9819065928459167,
      "learning_rate": 4.139057085406221e-05,
      "loss": 3.4237,
      "step": 2715
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9675536751747131,
      "learning_rate": 4.1226393277027726e-05,
      "loss": 3.4678,
      "step": 2720
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.955564022064209,
      "learning_rate": 4.106231324191471e-05,
      "loss": 3.4339,
      "step": 2725
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8302602767944336,
      "learning_rate": 4.089833257290817e-05,
      "loss": 3.3893,
      "step": 2730
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.0030620098114014,
      "learning_rate": 4.073445309308842e-05,
      "loss": 3.403,
      "step": 2735
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8429064750671387,
      "learning_rate": 4.0570676624410756e-05,
      "loss": 3.3268,
      "step": 2740
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9952207803726196,
      "learning_rate": 4.040700498768525e-05,
      "loss": 3.419,
      "step": 2745
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8415958881378174,
      "learning_rate": 4.024344000255648e-05,
      "loss": 3.3845,
      "step": 2750
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8832328915596008,
      "learning_rate": 4.0079983487483313e-05,
      "loss": 3.4452,
      "step": 2755
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.8209018707275391,
      "learning_rate": 3.9916637259718683e-05,
      "loss": 3.3467,
      "step": 2760
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.7963742613792419,
      "learning_rate": 3.9753403135289396e-05,
      "loss": 3.4648,
      "step": 2765
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.8554068207740784,
      "learning_rate": 3.9590282928975914e-05,
      "loss": 3.402,
      "step": 2770
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.8575777411460876,
      "learning_rate": 3.942727845429221e-05,
      "loss": 3.4545,
      "step": 2775
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9378578066825867,
      "learning_rate": 3.926439152346558e-05,
      "loss": 3.4027,
      "step": 2780
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.8902561664581299,
      "learning_rate": 3.910162394741653e-05,
      "loss": 3.4738,
      "step": 2785
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.8349782228469849,
      "learning_rate": 3.893897753573861e-05,
      "loss": 3.2739,
      "step": 2790
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9909433722496033,
      "learning_rate": 3.877645409667829e-05,
      "loss": 3.3343,
      "step": 2795
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9836429953575134,
      "learning_rate": 3.861405543711491e-05,
      "loss": 3.3772,
      "step": 2800
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9544408917427063,
      "learning_rate": 3.8451783362540507e-05,
      "loss": 3.4217,
      "step": 2805
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9274853467941284,
      "learning_rate": 3.828963967703983e-05,
      "loss": 3.3485,
      "step": 2810
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8512804508209229,
      "learning_rate": 3.8127626183270223e-05,
      "loss": 3.3691,
      "step": 2815
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8450992107391357,
      "learning_rate": 3.796574468244161e-05,
      "loss": 3.3275,
      "step": 2820
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8768813610076904,
      "learning_rate": 3.7803996974296444e-05,
      "loss": 3.3906,
      "step": 2825
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8551543354988098,
      "learning_rate": 3.7642384857089776e-05,
      "loss": 3.3034,
      "step": 2830
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8286839127540588,
      "learning_rate": 3.748091012756915e-05,
      "loss": 3.3475,
      "step": 2835
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.0059750080108643,
      "learning_rate": 3.731957458095467e-05,
      "loss": 3.399,
      "step": 2840
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9001671671867371,
      "learning_rate": 3.71583800109191e-05,
      "loss": 3.368,
      "step": 2845
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9101215600967407,
      "learning_rate": 3.699732820956784e-05,
      "loss": 3.4451,
      "step": 2850
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.885056734085083,
      "learning_rate": 3.6836420967419057e-05,
      "loss": 3.4028,
      "step": 2855
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8770620226860046,
      "learning_rate": 3.6675660073383745e-05,
      "loss": 3.2998,
      "step": 2860
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9198933839797974,
      "learning_rate": 3.6515047314745856e-05,
      "loss": 3.4097,
      "step": 2865
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8237200975418091,
      "learning_rate": 3.6354584477142437e-05,
      "loss": 3.4916,
      "step": 2870
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8838125467300415,
      "learning_rate": 3.6194273344543736e-05,
      "loss": 3.3887,
      "step": 2875
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8582887053489685,
      "learning_rate": 3.6034115699233425e-05,
      "loss": 3.4103,
      "step": 2880
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.83761066198349,
      "learning_rate": 3.5874113321788736e-05,
      "loss": 3.3979,
      "step": 2885
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9017011523246765,
      "learning_rate": 3.571426799106071e-05,
      "loss": 3.4158,
      "step": 2890
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8480706810951233,
      "learning_rate": 3.555458148415437e-05,
      "loss": 3.332,
      "step": 2895
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8610851168632507,
      "learning_rate": 3.539505557640901e-05,
      "loss": 3.36,
      "step": 2900
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8271325826644897,
      "learning_rate": 3.523569204137843e-05,
      "loss": 3.3321,
      "step": 2905
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9846967458724976,
      "learning_rate": 3.5076492650811246e-05,
      "loss": 3.3163,
      "step": 2910
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9517627358436584,
      "learning_rate": 3.491745917463113e-05,
      "loss": 3.437,
      "step": 2915
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9230513572692871,
      "learning_rate": 3.475859338091721e-05,
      "loss": 3.354,
      "step": 2920
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9414743185043335,
      "learning_rate": 3.4599897035884374e-05,
      "loss": 3.3626,
      "step": 2925
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8482127785682678,
      "learning_rate": 3.444137190386363e-05,
      "loss": 3.3172,
      "step": 2930
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8926299214363098,
      "learning_rate": 3.4283019747282514e-05,
      "loss": 3.4264,
      "step": 2935
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8302354216575623,
      "learning_rate": 3.412484232664545e-05,
      "loss": 3.3653,
      "step": 2940
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.8593830466270447,
      "learning_rate": 3.396684140051424e-05,
      "loss": 3.353,
      "step": 2945
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9003075361251831,
      "learning_rate": 3.3809018725488466e-05,
      "loss": 3.3689,
      "step": 2950
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9126880168914795,
      "learning_rate": 3.365137605618598e-05,
      "loss": 3.3892,
      "step": 2955
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8957822918891907,
      "learning_rate": 3.3493915145223395e-05,
      "loss": 3.4007,
      "step": 2960
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8402200937271118,
      "learning_rate": 3.3336637743196584e-05,
      "loss": 3.2873,
      "step": 2965
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9142640829086304,
      "learning_rate": 3.317954559866126e-05,
      "loss": 3.2816,
      "step": 2970
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9357993602752686,
      "learning_rate": 3.302264045811344e-05,
      "loss": 3.2826,
      "step": 2975
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9008080363273621,
      "learning_rate": 3.286592406597021e-05,
      "loss": 3.3283,
      "step": 2980
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8646254539489746,
      "learning_rate": 3.270939816455012e-05,
      "loss": 3.3901,
      "step": 2985
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.0222582817077637,
      "learning_rate": 3.255306449405395e-05,
      "loss": 3.3172,
      "step": 2990
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8637706637382507,
      "learning_rate": 3.2396924792545304e-05,
      "loss": 3.3113,
      "step": 2995
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.0547915697097778,
      "learning_rate": 3.224098079593132e-05,
      "loss": 3.3446,
      "step": 3000
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.0070993900299072,
      "learning_rate": 3.2085234237943354e-05,
      "loss": 3.4087,
      "step": 3005
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8321356773376465,
      "learning_rate": 3.19296868501177e-05,
      "loss": 3.3127,
      "step": 3010
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8560150861740112,
      "learning_rate": 3.177434036177636e-05,
      "loss": 3.3001,
      "step": 3015
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9279139041900635,
      "learning_rate": 3.1619196500007804e-05,
      "loss": 3.2725,
      "step": 3020
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.896747887134552,
      "learning_rate": 3.146425698964776e-05,
      "loss": 3.3268,
      "step": 3025
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9544872045516968,
      "learning_rate": 3.1309523553260046e-05,
      "loss": 3.2712,
      "step": 3030
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9028177261352539,
      "learning_rate": 3.115499791111743e-05,
      "loss": 3.3019,
      "step": 3035
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9091874957084656,
      "learning_rate": 3.10006817811825e-05,
      "loss": 3.3588,
      "step": 3040
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8359854817390442,
      "learning_rate": 3.084657687908855e-05,
      "loss": 3.2409,
      "step": 3045
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.070448875427246,
      "learning_rate": 3.069268491812052e-05,
      "loss": 3.317,
      "step": 3050
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8524084091186523,
      "learning_rate": 3.0539007609195934e-05,
      "loss": 3.3603,
      "step": 3055
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.8775138258934021,
      "learning_rate": 3.0385546660845908e-05,
      "loss": 3.3665,
      "step": 3060
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.8480380773544312,
      "learning_rate": 3.0232303779196132e-05,
      "loss": 3.254,
      "step": 3065
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.938522219657898,
      "learning_rate": 3.0079280667947885e-05,
      "loss": 3.4022,
      "step": 3070
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.827492356300354,
      "learning_rate": 2.9926479028359132e-05,
      "loss": 3.2898,
      "step": 3075
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.8760508298873901,
      "learning_rate": 2.97739005592256e-05,
      "loss": 3.4141,
      "step": 3080
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.8683328032493591,
      "learning_rate": 2.962154695686187e-05,
      "loss": 3.3617,
      "step": 3085
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.8718876838684082,
      "learning_rate": 2.9469419915082536e-05,
      "loss": 3.3372,
      "step": 3090
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.8450455665588379,
      "learning_rate": 2.9317521125183368e-05,
      "loss": 3.3932,
      "step": 3095
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.8650111556053162,
      "learning_rate": 2.9165852275922524e-05,
      "loss": 3.3839,
      "step": 3100
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.054282307624817,
      "learning_rate": 2.901441505350174e-05,
      "loss": 3.369,
      "step": 3105
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9530842900276184,
      "learning_rate": 2.886321114154762e-05,
      "loss": 3.3099,
      "step": 3110
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9338812828063965,
      "learning_rate": 2.87122422210929e-05,
      "loss": 3.4111,
      "step": 3115
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9333266615867615,
      "learning_rate": 2.8561509970557736e-05,
      "loss": 3.3992,
      "step": 3120
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8644818067550659,
      "learning_rate": 2.8411016065731146e-05,
      "loss": 3.3228,
      "step": 3125
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9469544887542725,
      "learning_rate": 2.826076217975222e-05,
      "loss": 3.3703,
      "step": 3130
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8587395548820496,
      "learning_rate": 2.8110749983091632e-05,
      "loss": 3.3823,
      "step": 3135
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8624106049537659,
      "learning_rate": 2.7960981143533053e-05,
      "loss": 3.2556,
      "step": 3140
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.8285986185073853,
      "learning_rate": 2.781145732615457e-05,
      "loss": 3.3055,
      "step": 3145
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.897678554058075,
      "learning_rate": 2.7662180193310218e-05,
      "loss": 3.244,
      "step": 3150
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8292384743690491,
      "learning_rate": 2.751315140461145e-05,
      "loss": 3.3844,
      "step": 3155
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8837412595748901,
      "learning_rate": 2.7364372616908744e-05,
      "loss": 3.3891,
      "step": 3160
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8819624781608582,
      "learning_rate": 2.7215845484273152e-05,
      "loss": 3.3456,
      "step": 3165
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9516217708587646,
      "learning_rate": 2.7067571657977893e-05,
      "loss": 3.317,
      "step": 3170
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8784796595573425,
      "learning_rate": 2.691955278648003e-05,
      "loss": 3.3318,
      "step": 3175
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8507090210914612,
      "learning_rate": 2.6771790515402112e-05,
      "loss": 3.4276,
      "step": 3180
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8158341646194458,
      "learning_rate": 2.6624286487513916e-05,
      "loss": 3.2964,
      "step": 3185
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8636460304260254,
      "learning_rate": 2.6477042342714137e-05,
      "loss": 3.3046,
      "step": 3190
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8573281168937683,
      "learning_rate": 2.633005971801219e-05,
      "loss": 3.3494,
      "step": 3195
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8206995725631714,
      "learning_rate": 2.6183340247510013e-05,
      "loss": 3.2589,
      "step": 3200
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.7886756658554077,
      "learning_rate": 2.6036885562383856e-05,
      "loss": 3.3254,
      "step": 3205
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8958569765090942,
      "learning_rate": 2.5890697290866206e-05,
      "loss": 3.349,
      "step": 3210
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8987053036689758,
      "learning_rate": 2.5744777058227642e-05,
      "loss": 3.3414,
      "step": 3215
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9279022812843323,
      "learning_rate": 2.5599126486758777e-05,
      "loss": 3.2571,
      "step": 3220
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.979684054851532,
      "learning_rate": 2.5453747195752243e-05,
      "loss": 3.3664,
      "step": 3225
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8469237685203552,
      "learning_rate": 2.530864080148464e-05,
      "loss": 3.3882,
      "step": 3230
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.938224196434021,
      "learning_rate": 2.5163808917198615e-05,
      "loss": 3.313,
      "step": 3235
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8933103680610657,
      "learning_rate": 2.501925315308492e-05,
      "loss": 3.2644,
      "step": 3240
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9098454117774963,
      "learning_rate": 2.4874975116264477e-05,
      "loss": 3.2601,
      "step": 3245
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.832898736000061,
      "learning_rate": 2.4730976410770534e-05,
      "loss": 3.3678,
      "step": 3250
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8103572726249695,
      "learning_rate": 2.458725863753084e-05,
      "loss": 3.3714,
      "step": 3255
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8854662775993347,
      "learning_rate": 2.4443823394349834e-05,
      "loss": 3.3424,
      "step": 3260
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8286263942718506,
      "learning_rate": 2.430067227589088e-05,
      "loss": 3.3551,
      "step": 3265
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9545296430587769,
      "learning_rate": 2.4157806873658517e-05,
      "loss": 3.3823,
      "step": 3270
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9493871927261353,
      "learning_rate": 2.401522877598087e-05,
      "loss": 3.396,
      "step": 3275
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8324505686759949,
      "learning_rate": 2.3872939567991827e-05,
      "loss": 3.3239,
      "step": 3280
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.857708215713501,
      "learning_rate": 2.373094083161353e-05,
      "loss": 3.3066,
      "step": 3285
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8161062002182007,
      "learning_rate": 2.358923414553877e-05,
      "loss": 3.3485,
      "step": 3290
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8786885738372803,
      "learning_rate": 2.3447821085213405e-05,
      "loss": 3.3984,
      "step": 3295
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.870347797870636,
      "learning_rate": 2.3306703222818878e-05,
      "loss": 3.3084,
      "step": 3300
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.833512544631958,
      "learning_rate": 2.3165882127254705e-05,
      "loss": 3.3442,
      "step": 3305
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8745336532592773,
      "learning_rate": 2.302535936412108e-05,
      "loss": 3.3367,
      "step": 3310
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8376188278198242,
      "learning_rate": 2.2885136495701415e-05,
      "loss": 3.3021,
      "step": 3315
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8755170106887817,
      "learning_rate": 2.274521508094501e-05,
      "loss": 3.3799,
      "step": 3320
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8672605752944946,
      "learning_rate": 2.2605596675449698e-05,
      "loss": 3.3517,
      "step": 3325
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8204872608184814,
      "learning_rate": 2.246628283144457e-05,
      "loss": 3.2989,
      "step": 3330
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8279518485069275,
      "learning_rate": 2.232727509777269e-05,
      "loss": 3.35,
      "step": 3335
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8202655911445618,
      "learning_rate": 2.2188575019873932e-05,
      "loss": 3.2457,
      "step": 3340
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9008398056030273,
      "learning_rate": 2.2050184139767704e-05,
      "loss": 3.4337,
      "step": 3345
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.840510368347168,
      "learning_rate": 2.191210399603591e-05,
      "loss": 3.3695,
      "step": 3350
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8364140391349792,
      "learning_rate": 2.1774336123805772e-05,
      "loss": 3.3407,
      "step": 3355
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.881549060344696,
      "learning_rate": 2.1636882054732776e-05,
      "loss": 3.3053,
      "step": 3360
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.827682375907898,
      "learning_rate": 2.1499743316983684e-05,
      "loss": 3.2035,
      "step": 3365
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8124437928199768,
      "learning_rate": 2.1362921435219473e-05,
      "loss": 3.272,
      "step": 3370
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.5780527591705322,
      "learning_rate": 2.1226417930578464e-05,
      "loss": 3.304,
      "step": 3375
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8243139982223511,
      "learning_rate": 2.109023432065935e-05,
      "loss": 3.3359,
      "step": 3380
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8204466700553894,
      "learning_rate": 2.095437211950434e-05,
      "loss": 3.2912,
      "step": 3385
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9210676550865173,
      "learning_rate": 2.0818832837582352e-05,
      "loss": 3.3311,
      "step": 3390
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8979406952857971,
      "learning_rate": 2.068361798177218e-05,
      "loss": 3.2767,
      "step": 3395
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8867124319076538,
      "learning_rate": 2.0548729055345778e-05,
      "loss": 3.3362,
      "step": 3400
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8429162502288818,
      "learning_rate": 2.0414167557951514e-05,
      "loss": 3.2875,
      "step": 3405
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8652162551879883,
      "learning_rate": 2.0279934985597527e-05,
      "loss": 3.3248,
      "step": 3410
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.830157995223999,
      "learning_rate": 2.0146032830635054e-05,
      "loss": 3.3039,
      "step": 3415
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.851249098777771,
      "learning_rate": 2.001246258174192e-05,
      "loss": 3.228,
      "step": 3420
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8676759004592896,
      "learning_rate": 1.9879225723905886e-05,
      "loss": 3.2683,
      "step": 3425
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8468626737594604,
      "learning_rate": 1.9746323738408203e-05,
      "loss": 3.2965,
      "step": 3430
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8902848362922668,
      "learning_rate": 1.9613758102807117e-05,
      "loss": 3.3187,
      "step": 3435
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8139768242835999,
      "learning_rate": 1.9481530290921474e-05,
      "loss": 3.2649,
      "step": 3440
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8436658978462219,
      "learning_rate": 1.934964177281428e-05,
      "loss": 3.2641,
      "step": 3445
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8619456887245178,
      "learning_rate": 1.9218094014776434e-05,
      "loss": 3.2942,
      "step": 3450
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8965754508972168,
      "learning_rate": 1.9086888479310333e-05,
      "loss": 3.2437,
      "step": 3455
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8351267576217651,
      "learning_rate": 1.895602662511371e-05,
      "loss": 3.2898,
      "step": 3460
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8266618847846985,
      "learning_rate": 1.8825509907063327e-05,
      "loss": 3.3272,
      "step": 3465
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9140317440032959,
      "learning_rate": 1.8695339776198872e-05,
      "loss": 3.3165,
      "step": 3470
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8285201191902161,
      "learning_rate": 1.8565517679706783e-05,
      "loss": 3.2977,
      "step": 3475
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8304650783538818,
      "learning_rate": 1.8436045060904174e-05,
      "loss": 3.2944,
      "step": 3480
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8460079431533813,
      "learning_rate": 1.830692335922279e-05,
      "loss": 3.2678,
      "step": 3485
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8778470754623413,
      "learning_rate": 1.8178154010192994e-05,
      "loss": 3.3551,
      "step": 3490
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8502268195152283,
      "learning_rate": 1.8049738445427822e-05,
      "loss": 3.3525,
      "step": 3495
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8323951363563538,
      "learning_rate": 1.7921678092607052e-05,
      "loss": 3.3441,
      "step": 3500
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8833833336830139,
      "learning_rate": 1.7793974375461352e-05,
      "loss": 3.349,
      "step": 3505
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8423852920532227,
      "learning_rate": 1.7666628713756417e-05,
      "loss": 3.2798,
      "step": 3510
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8736124634742737,
      "learning_rate": 1.7539642523277228e-05,
      "loss": 3.2949,
      "step": 3515
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8281068801879883,
      "learning_rate": 1.7413017215812273e-05,
      "loss": 3.2283,
      "step": 3520
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8401610851287842,
      "learning_rate": 1.728675419913788e-05,
      "loss": 3.4024,
      "step": 3525
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.7921164035797119,
      "learning_rate": 1.716085487700253e-05,
      "loss": 3.2893,
      "step": 3530
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8460646867752075,
      "learning_rate": 1.703532064911131e-05,
      "loss": 3.288,
      "step": 3535
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8160696029663086,
      "learning_rate": 1.6910152911110283e-05,
      "loss": 3.2195,
      "step": 3540
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8647582530975342,
      "learning_rate": 1.6785353054571024e-05,
      "loss": 3.3154,
      "step": 3545
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8196638822555542,
      "learning_rate": 1.666092246697512e-05,
      "loss": 3.3802,
      "step": 3550
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8931661248207092,
      "learning_rate": 1.6536862531698766e-05,
      "loss": 3.2419,
      "step": 3555
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8836420774459839,
      "learning_rate": 1.6413174627997328e-05,
      "loss": 3.3038,
      "step": 3560
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8200202584266663,
      "learning_rate": 1.6289860130990147e-05,
      "loss": 3.336,
      "step": 3565
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8285385370254517,
      "learning_rate": 1.6166920411645064e-05,
      "loss": 3.3422,
      "step": 3570
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.863246738910675,
      "learning_rate": 1.6044356836763315e-05,
      "loss": 3.3437,
      "step": 3575
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.81345534324646,
      "learning_rate": 1.5922170768964285e-05,
      "loss": 3.4152,
      "step": 3580
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8124732375144958,
      "learning_rate": 1.5800363566670362e-05,
      "loss": 3.3005,
      "step": 3585
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8287694454193115,
      "learning_rate": 1.5678936584091852e-05,
      "loss": 3.2783,
      "step": 3590
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8340721726417542,
      "learning_rate": 1.5557891171211892e-05,
      "loss": 3.2003,
      "step": 3595
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8558523058891296,
      "learning_rate": 1.5437228673771465e-05,
      "loss": 3.3051,
      "step": 3600
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8867878317832947,
      "learning_rate": 1.5316950433254445e-05,
      "loss": 3.3998,
      "step": 3605
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8162564039230347,
      "learning_rate": 1.5197057786872649e-05,
      "loss": 3.3834,
      "step": 3610
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8151681423187256,
      "learning_rate": 1.5077552067551015e-05,
      "loss": 3.247,
      "step": 3615
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.825150191783905,
      "learning_rate": 1.4958434603912747e-05,
      "loss": 3.3278,
      "step": 3620
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8217668533325195,
      "learning_rate": 1.4839706720264546e-05,
      "loss": 3.345,
      "step": 3625
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8449623584747314,
      "learning_rate": 1.4721369736581924e-05,
      "loss": 3.2981,
      "step": 3630
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8323595523834229,
      "learning_rate": 1.4603424968494484e-05,
      "loss": 3.3461,
      "step": 3635
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8136734366416931,
      "learning_rate": 1.448587372727132e-05,
      "loss": 3.3347,
      "step": 3640
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8722835779190063,
      "learning_rate": 1.4368717319806419e-05,
      "loss": 3.2841,
      "step": 3645
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.897956907749176,
      "learning_rate": 1.4251957048604152e-05,
      "loss": 3.3148,
      "step": 3650
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9544664621353149,
      "learning_rate": 1.413559421176479e-05,
      "loss": 3.3538,
      "step": 3655
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8626210689544678,
      "learning_rate": 1.4019630102970056e-05,
      "loss": 3.3596,
      "step": 3660
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8736276030540466,
      "learning_rate": 1.3904066011468753e-05,
      "loss": 3.2998,
      "step": 3665
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8064122796058655,
      "learning_rate": 1.3788903222062433e-05,
      "loss": 3.273,
      "step": 3670
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8083548545837402,
      "learning_rate": 1.3674143015091118e-05,
      "loss": 3.251,
      "step": 3675
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8138768672943115,
      "learning_rate": 1.355978666641905e-05,
      "loss": 3.2628,
      "step": 3680
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9377182722091675,
      "learning_rate": 1.3445835447420507e-05,
      "loss": 3.235,
      "step": 3685
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.7928673624992371,
      "learning_rate": 1.3332290624965688e-05,
      "loss": 3.2746,
      "step": 3690
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8481169939041138,
      "learning_rate": 1.3219153461406609e-05,
      "loss": 3.2695,
      "step": 3695
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9184083938598633,
      "learning_rate": 1.3106425214563078e-05,
      "loss": 3.2278,
      "step": 3700
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8539621233940125,
      "learning_rate": 1.2994107137708716e-05,
      "loss": 3.2818,
      "step": 3705
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8199257850646973,
      "learning_rate": 1.2882200479556988e-05,
      "loss": 3.2726,
      "step": 3710
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.810304582118988,
      "learning_rate": 1.2770706484247397e-05,
      "loss": 3.2467,
      "step": 3715
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8342577219009399,
      "learning_rate": 1.2659626391331564e-05,
      "loss": 3.3297,
      "step": 3720
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8215441107749939,
      "learning_rate": 1.2548961435759493e-05,
      "loss": 3.3136,
      "step": 3725
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8240970969200134,
      "learning_rate": 1.2438712847865846e-05,
      "loss": 3.2634,
      "step": 3730
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8182936906814575,
      "learning_rate": 1.2328881853356244e-05,
      "loss": 3.3069,
      "step": 3735
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8966035842895508,
      "learning_rate": 1.221946967329365e-05,
      "loss": 3.3569,
      "step": 3740
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.876420795917511,
      "learning_rate": 1.2110477524084796e-05,
      "loss": 3.2909,
      "step": 3745
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8695542216300964,
      "learning_rate": 1.2001906617466657e-05,
      "loss": 3.3204,
      "step": 3750
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8503195643424988,
      "learning_rate": 1.1893758160492978e-05,
      "loss": 3.1899,
      "step": 3755
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.807569146156311,
      "learning_rate": 1.1786033355520859e-05,
      "loss": 3.3462,
      "step": 3760
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8479645252227783,
      "learning_rate": 1.1678733400197373e-05,
      "loss": 3.3846,
      "step": 3765
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.7932925224304199,
      "learning_rate": 1.1571859487446263e-05,
      "loss": 3.2313,
      "step": 3770
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8063516616821289,
      "learning_rate": 1.1465412805454695e-05,
      "loss": 3.2578,
      "step": 3775
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8032195568084717,
      "learning_rate": 1.1359394537660011e-05,
      "loss": 3.3719,
      "step": 3780
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8684897422790527,
      "learning_rate": 1.125380586273661e-05,
      "loss": 3.2291,
      "step": 3785
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8473917841911316,
      "learning_rate": 1.1148647954582808e-05,
      "loss": 3.2538,
      "step": 3790
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.9202986359596252,
      "learning_rate": 1.1043921982307819e-05,
      "loss": 3.3475,
      "step": 3795
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8170850276947021,
      "learning_rate": 1.0939629110218735e-05,
      "loss": 3.3528,
      "step": 3800
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.7920132875442505,
      "learning_rate": 1.0835770497807596e-05,
      "loss": 3.1723,
      "step": 3805
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8625448942184448,
      "learning_rate": 1.0732347299738493e-05,
      "loss": 3.2547,
      "step": 3810
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8212171196937561,
      "learning_rate": 1.0629360665834732e-05,
      "loss": 3.3117,
      "step": 3815
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8107876777648926,
      "learning_rate": 1.052681174106604e-05,
      "loss": 3.2923,
      "step": 3820
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8409060835838318,
      "learning_rate": 1.0424701665535852e-05,
      "loss": 3.2746,
      "step": 3825
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8198840022087097,
      "learning_rate": 1.0323031574468638e-05,
      "loss": 3.287,
      "step": 3830
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8378563523292542,
      "learning_rate": 1.0221802598197261e-05,
      "loss": 3.3435,
      "step": 3835
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8228487372398376,
      "learning_rate": 1.0121015862150423e-05,
      "loss": 3.2075,
      "step": 3840
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8124101161956787,
      "learning_rate": 1.0020672486840154e-05,
      "loss": 3.3079,
      "step": 3845
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8262704014778137,
      "learning_rate": 9.920773587849364e-06,
      "loss": 3.3592,
      "step": 3850
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8362215161323547,
      "learning_rate": 9.821320275819401e-06,
      "loss": 3.2244,
      "step": 3855
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8361976146697998,
      "learning_rate": 9.72231365643777e-06,
      "loss": 3.3835,
      "step": 3860
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8179517388343811,
      "learning_rate": 9.623754830425779e-06,
      "loss": 3.4081,
      "step": 3865
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.816852867603302,
      "learning_rate": 9.52564489352632e-06,
      "loss": 3.3189,
      "step": 3870
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8509985208511353,
      "learning_rate": 9.427984936491702e-06,
      "loss": 3.2415,
      "step": 3875
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8221267461776733,
      "learning_rate": 9.330776045071509e-06,
      "loss": 3.2113,
      "step": 3880
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8320128321647644,
      "learning_rate": 9.23401930000054e-06,
      "loss": 3.3388,
      "step": 3885
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8314358592033386,
      "learning_rate": 9.137715776986772e-06,
      "loss": 3.3267,
      "step": 3890
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.822598397731781,
      "learning_rate": 9.041866546699434e-06,
      "loss": 3.351,
      "step": 3895
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8570499420166016,
      "learning_rate": 8.946472674757078e-06,
      "loss": 3.2219,
      "step": 3900
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.816481351852417,
      "learning_rate": 8.851535221715735e-06,
      "loss": 3.2443,
      "step": 3905
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8919932842254639,
      "learning_rate": 8.757055243057132e-06,
      "loss": 3.2184,
      "step": 3910
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8713335990905762,
      "learning_rate": 8.663033789176967e-06,
      "loss": 3.3185,
      "step": 3915
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.4435936212539673,
      "learning_rate": 8.5694719053732e-06,
      "loss": 3.2908,
      "step": 3920
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8382534980773926,
      "learning_rate": 8.476370631834458e-06,
      "loss": 3.2048,
      "step": 3925
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8237469792366028,
      "learning_rate": 8.383731003628452e-06,
      "loss": 3.2839,
      "step": 3930
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8565986156463623,
      "learning_rate": 8.291554050690508e-06,
      "loss": 3.2783,
      "step": 3935
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8280335068702698,
      "learning_rate": 8.199840797812058e-06,
      "loss": 3.2714,
      "step": 3940
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8038430213928223,
      "learning_rate": 8.108592264629295e-06,
      "loss": 3.3384,
      "step": 3945
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.7964035868644714,
      "learning_rate": 8.017809465611803e-06,
      "loss": 3.1829,
      "step": 3950
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8033803701400757,
      "learning_rate": 7.927493410051324e-06,
      "loss": 3.3429,
      "step": 3955
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8323132395744324,
      "learning_rate": 7.837645102050473e-06,
      "loss": 3.3593,
      "step": 3960
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.7978505492210388,
      "learning_rate": 7.748265540511635e-06,
      "loss": 3.2938,
      "step": 3965
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8222649097442627,
      "learning_rate": 7.65935571912582e-06,
      "loss": 3.2629,
      "step": 3970
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8070308566093445,
      "learning_rate": 7.5709166263616405e-06,
      "loss": 3.2682,
      "step": 3975
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8659495115280151,
      "learning_rate": 7.482949245454302e-06,
      "loss": 3.283,
      "step": 3980
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.817457377910614,
      "learning_rate": 7.3954545543946876e-06,
      "loss": 3.2556,
      "step": 3985
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8457687497138977,
      "learning_rate": 7.308433525918468e-06,
      "loss": 3.3078,
      "step": 3990
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8201070427894592,
      "learning_rate": 7.221887127495313e-06,
      "loss": 3.2487,
      "step": 3995
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8550896644592285,
      "learning_rate": 7.1358163213181114e-06,
      "loss": 3.3523,
      "step": 4000
    }
  ],
  "logging_steps": 5,
  "max_steps": 4811,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "total_flos": 1.0865380348133376e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}