{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 2000,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0002,
      "grad_norm": 0.3626861274242401,
      "learning_rate": 2e-05,
      "loss": 2.8391363620758057,
      "step": 2
    },
    {
      "epoch": 0.0004,
      "grad_norm": 0.31211480498313904,
      "learning_rate": 6e-05,
      "loss": 2.78165864944458,
      "step": 4
    },
    {
      "epoch": 0.0006,
      "grad_norm": 0.30627524852752686,
      "learning_rate": 0.0001,
      "loss": 2.821012258529663,
      "step": 6
    },
    {
      "epoch": 0.0008,
      "grad_norm": 0.35109856724739075,
      "learning_rate": 0.00014000000000000001,
      "loss": 2.7588582038879395,
      "step": 8
    },
    {
      "epoch": 0.001,
      "grad_norm": 0.343110591173172,
      "learning_rate": 0.00017999999999999998,
      "loss": 2.795839786529541,
      "step": 10
    },
    {
      "epoch": 0.0012,
      "grad_norm": 0.3194366991519928,
      "learning_rate": 0.00022,
      "loss": 2.7377989292144775,
      "step": 12
    },
    {
      "epoch": 0.0014,
      "grad_norm": 0.28374719619750977,
      "learning_rate": 0.00026000000000000003,
      "loss": 2.78313946723938,
      "step": 14
    },
    {
      "epoch": 0.0016,
      "grad_norm": 0.2868261933326721,
      "learning_rate": 0.0003,
      "loss": 2.7046492099761963,
      "step": 16
    },
    {
      "epoch": 0.0018,
      "grad_norm": 0.2876185178756714,
      "learning_rate": 0.00034,
      "loss": 2.818948268890381,
      "step": 18
    },
    {
      "epoch": 0.002,
      "grad_norm": 0.2699960768222809,
      "learning_rate": 0.00038,
      "loss": 2.7019805908203125,
      "step": 20
    },
    {
      "epoch": 0.0022,
      "grad_norm": 0.3006955683231354,
      "learning_rate": 0.00042,
      "loss": 2.756977081298828,
      "step": 22
    },
    {
      "epoch": 0.0024,
      "grad_norm": 0.2830796241760254,
      "learning_rate": 0.00046,
      "loss": 2.9192984104156494,
      "step": 24
    },
    {
      "epoch": 0.0026,
      "grad_norm": 0.23309166729450226,
      "learning_rate": 0.0005,
      "loss": 2.6595919132232666,
      "step": 26
    },
    {
      "epoch": 0.0028,
      "grad_norm": 0.26088038086891174,
      "learning_rate": 0.00054,
      "loss": 2.747100353240967,
      "step": 28
    },
    {
      "epoch": 0.003,
      "grad_norm": 0.2334124892950058,
      "learning_rate": 0.00058,
      "loss": 2.655841827392578,
      "step": 30
    },
    {
      "epoch": 0.0032,
      "grad_norm": 0.21991023421287537,
      "learning_rate": 0.00062,
      "loss": 2.6737802028656006,
      "step": 32
    },
    {
      "epoch": 0.0034,
      "grad_norm": 0.24159392714500427,
      "learning_rate": 0.00066,
      "loss": 2.7439167499542236,
      "step": 34
    },
    {
      "epoch": 0.0036,
      "grad_norm": 0.2386929839849472,
      "learning_rate": 0.0007,
      "loss": 2.763815402984619,
      "step": 36
    },
    {
      "epoch": 0.0038,
      "grad_norm": 0.21081526577472687,
      "learning_rate": 0.00074,
      "loss": 2.715547561645508,
      "step": 38
    },
    {
      "epoch": 0.004,
      "grad_norm": 0.19664035737514496,
      "learning_rate": 0.0007800000000000001,
      "loss": 2.7181568145751953,
      "step": 40
    },
    {
      "epoch": 0.0042,
      "grad_norm": 0.18893127143383026,
      "learning_rate": 0.00082,
      "loss": 2.851597547531128,
      "step": 42
    },
    {
      "epoch": 0.0044,
      "grad_norm": 0.17336373031139374,
      "learning_rate": 0.00086,
      "loss": 2.7288200855255127,
      "step": 44
    },
    {
      "epoch": 0.0046,
      "grad_norm": 0.1965402215719223,
      "learning_rate": 0.0009000000000000001,
      "loss": 2.7796378135681152,
      "step": 46
    },
    {
      "epoch": 0.0048,
      "grad_norm": 0.1652621179819107,
      "learning_rate": 0.00094,
      "loss": 2.667043685913086,
      "step": 48
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.1579713523387909,
      "learning_rate": 0.00098,
      "loss": 2.66037917137146,
      "step": 50
    },
    {
      "epoch": 0.0052,
      "grad_norm": 0.16253702342510223,
      "learning_rate": 0.00102,
      "loss": 2.6319007873535156,
      "step": 52
    },
    {
      "epoch": 0.0054,
      "grad_norm": 0.14703965187072754,
      "learning_rate": 0.0010600000000000002,
      "loss": 2.704150438308716,
      "step": 54
    },
    {
      "epoch": 0.0056,
      "grad_norm": 0.13690407574176788,
      "learning_rate": 0.0011,
      "loss": 2.636317253112793,
      "step": 56
    },
    {
      "epoch": 0.0058,
      "grad_norm": 0.1289760321378708,
      "learning_rate": 0.00114,
      "loss": 2.667541027069092,
      "step": 58
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.1337203085422516,
      "learning_rate": 0.00118,
      "loss": 2.5739758014678955,
      "step": 60
    },
    {
      "epoch": 0.0062,
      "grad_norm": 0.12501287460327148,
      "learning_rate": 0.00122,
      "loss": 2.746004581451416,
      "step": 62
    },
    {
      "epoch": 0.0064,
      "grad_norm": 0.10977648943662643,
      "learning_rate": 0.00126,
      "loss": 2.840315341949463,
      "step": 64
    },
    {
      "epoch": 0.0066,
      "grad_norm": 0.1077275276184082,
      "learning_rate": 0.0013000000000000002,
      "loss": 2.4886257648468018,
      "step": 66
    },
    {
      "epoch": 0.0068,
      "grad_norm": 0.10123786330223083,
      "learning_rate": 0.00134,
      "loss": 2.625009298324585,
      "step": 68
    },
    {
      "epoch": 0.007,
      "grad_norm": 0.11452195793390274,
      "learning_rate": 0.00138,
      "loss": 2.6896843910217285,
      "step": 70
    },
    {
      "epoch": 0.0072,
      "grad_norm": 0.13366173207759857,
      "learning_rate": 0.00142,
      "loss": 2.6507492065429688,
      "step": 72
    },
    {
      "epoch": 0.0074,
      "grad_norm": 0.14093396067619324,
      "learning_rate": 0.00146,
      "loss": 2.6120564937591553,
      "step": 74
    },
    {
      "epoch": 0.0076,
      "grad_norm": 0.0924813374876976,
      "learning_rate": 0.0015,
      "loss": 2.6029052734375,
      "step": 76
    },
    {
      "epoch": 0.0078,
      "grad_norm": 0.08463750034570694,
      "learning_rate": 0.0015400000000000001,
      "loss": 2.5443873405456543,
      "step": 78
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.08462584763765335,
      "learning_rate": 0.00158,
      "loss": 2.632230043411255,
      "step": 80
    },
    {
      "epoch": 0.0082,
      "grad_norm": 0.08414039760828018,
      "learning_rate": 0.0016200000000000001,
      "loss": 2.5740256309509277,
      "step": 82
    },
    {
      "epoch": 0.0084,
      "grad_norm": 0.08951207995414734,
      "learning_rate": 0.00166,
      "loss": 2.6690757274627686,
      "step": 84
    },
    {
      "epoch": 0.0086,
      "grad_norm": 0.08569514751434326,
      "learning_rate": 0.0017,
      "loss": 2.547718048095703,
      "step": 86
    },
    {
      "epoch": 0.0088,
      "grad_norm": 0.07638048380613327,
      "learning_rate": 0.00174,
      "loss": 2.638239622116089,
      "step": 88
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.10180540382862091,
      "learning_rate": 0.0017800000000000001,
      "loss": 2.5370523929595947,
      "step": 90
    },
    {
      "epoch": 0.0092,
      "grad_norm": 0.07965986430644989,
      "learning_rate": 0.00182,
      "loss": 2.627887725830078,
      "step": 92
    },
    {
      "epoch": 0.0094,
      "grad_norm": 0.07960250228643417,
      "learning_rate": 0.00186,
      "loss": 2.628718376159668,
      "step": 94
    },
    {
      "epoch": 0.0096,
      "grad_norm": 0.08066410571336746,
      "learning_rate": 0.0019,
      "loss": 2.5595221519470215,
      "step": 96
    },
    {
      "epoch": 0.0098,
      "grad_norm": 0.08023161441087723,
      "learning_rate": 0.0019399999999999999,
      "loss": 2.6549291610717773,
      "step": 98
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.07528366893529892,
      "learning_rate": 0.00198,
      "loss": 2.5612714290618896,
      "step": 100
    },
    {
      "epoch": 0.0102,
      "grad_norm": 0.08141227811574936,
      "learning_rate": 0.002,
      "loss": 2.6233181953430176,
      "step": 102
    },
    {
      "epoch": 0.0104,
      "grad_norm": 0.08059553802013397,
      "learning_rate": 0.002,
      "loss": 2.512939453125,
      "step": 104
    },
    {
      "epoch": 0.0106,
      "grad_norm": 0.08142438530921936,
      "learning_rate": 0.002,
      "loss": 2.6031289100646973,
      "step": 106
    },
    {
      "epoch": 0.0108,
      "grad_norm": 0.07679063826799393,
      "learning_rate": 0.002,
      "loss": 2.5137503147125244,
      "step": 108
    },
    {
      "epoch": 0.011,
      "grad_norm": 0.07538780570030212,
      "learning_rate": 0.002,
      "loss": 2.5250155925750732,
      "step": 110
    },
    {
      "epoch": 0.0112,
      "grad_norm": 0.07419940084218979,
      "learning_rate": 0.002,
      "loss": 2.569641590118408,
      "step": 112
    },
    {
      "epoch": 0.0114,
      "grad_norm": 0.0733867809176445,
      "learning_rate": 0.002,
      "loss": 2.615628242492676,
      "step": 114
    },
    {
      "epoch": 0.0116,
      "grad_norm": 0.0765562430024147,
      "learning_rate": 0.002,
      "loss": 2.505068063735962,
      "step": 116
    },
    {
      "epoch": 0.0118,
      "grad_norm": 0.07612311840057373,
      "learning_rate": 0.002,
      "loss": 2.397339344024658,
      "step": 118
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.0759640783071518,
      "learning_rate": 0.002,
      "loss": 2.60638165473938,
      "step": 120
    },
    {
      "epoch": 0.0122,
      "grad_norm": 0.06887827068567276,
      "learning_rate": 0.002,
      "loss": 2.4595835208892822,
      "step": 122
    },
    {
      "epoch": 0.0124,
      "grad_norm": 0.06663960963487625,
      "learning_rate": 0.002,
      "loss": 2.524352788925171,
      "step": 124
    },
    {
      "epoch": 0.0126,
      "grad_norm": 0.0705389529466629,
      "learning_rate": 0.002,
      "loss": 2.506653308868408,
      "step": 126
    },
    {
      "epoch": 0.0128,
      "grad_norm": 0.07625237107276917,
      "learning_rate": 0.002,
      "loss": 2.537121295928955,
      "step": 128
    },
    {
      "epoch": 0.013,
      "grad_norm": 0.07665827870368958,
      "learning_rate": 0.002,
      "loss": 2.600534200668335,
      "step": 130
    },
    {
      "epoch": 0.0132,
      "grad_norm": 0.06982910633087158,
      "learning_rate": 0.002,
      "loss": 2.4629504680633545,
      "step": 132
    },
    {
      "epoch": 0.0134,
      "grad_norm": 0.08062932640314102,
      "learning_rate": 0.002,
      "loss": 2.4764151573181152,
      "step": 134
    },
    {
      "epoch": 0.0136,
      "grad_norm": 0.07485084235668182,
      "learning_rate": 0.002,
      "loss": 2.504831314086914,
      "step": 136
    },
    {
      "epoch": 0.0138,
      "grad_norm": 0.0775233581662178,
      "learning_rate": 0.002,
      "loss": 2.5274486541748047,
      "step": 138
    },
    {
      "epoch": 0.014,
      "grad_norm": 0.0755736380815506,
      "learning_rate": 0.002,
      "loss": 2.480137348175049,
      "step": 140
    },
    {
      "epoch": 0.0142,
      "grad_norm": 0.0715257078409195,
      "learning_rate": 0.002,
      "loss": 2.5463967323303223,
      "step": 142
    },
    {
      "epoch": 0.0144,
      "grad_norm": 0.07261666655540466,
      "learning_rate": 0.002,
      "loss": 2.3550631999969482,
      "step": 144
    },
    {
      "epoch": 0.0146,
      "grad_norm": 0.07263581454753876,
      "learning_rate": 0.002,
      "loss": 2.3970537185668945,
      "step": 146
    },
    {
      "epoch": 0.0148,
      "grad_norm": 0.07136516273021698,
      "learning_rate": 0.002,
      "loss": 2.4485599994659424,
      "step": 148
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.07643298804759979,
      "learning_rate": 0.002,
      "loss": 2.453134298324585,
      "step": 150
    },
    {
      "epoch": 0.0152,
      "grad_norm": 0.07142458856105804,
      "learning_rate": 0.002,
      "loss": 2.4583747386932373,
      "step": 152
    },
    {
      "epoch": 0.0154,
      "grad_norm": 0.07379414141178131,
      "learning_rate": 0.002,
      "loss": 2.645972728729248,
      "step": 154
    },
    {
      "epoch": 0.0156,
      "grad_norm": 0.06986001878976822,
      "learning_rate": 0.002,
      "loss": 2.435237169265747,
      "step": 156
    },
    {
      "epoch": 0.0158,
      "grad_norm": 0.07270773500204086,
      "learning_rate": 0.002,
      "loss": 2.4673879146575928,
      "step": 158
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.07634732872247696,
      "learning_rate": 0.002,
      "loss": 2.5851078033447266,
      "step": 160
    },
    {
      "epoch": 0.0162,
      "grad_norm": 0.08367273956537247,
      "learning_rate": 0.002,
      "loss": 2.425518035888672,
      "step": 162
    },
    {
      "epoch": 0.0164,
      "grad_norm": 0.07181693613529205,
      "learning_rate": 0.002,
      "loss": 2.4647722244262695,
      "step": 164
    },
    {
      "epoch": 0.0166,
      "grad_norm": 0.07176841050386429,
      "learning_rate": 0.002,
      "loss": 2.3566653728485107,
      "step": 166
    },
    {
      "epoch": 0.0168,
      "grad_norm": 0.0765727162361145,
      "learning_rate": 0.002,
      "loss": 2.509660243988037,
      "step": 168
    },
    {
      "epoch": 0.017,
      "grad_norm": 0.07351703941822052,
      "learning_rate": 0.002,
      "loss": 2.441680431365967,
      "step": 170
    },
    {
      "epoch": 0.0172,
      "grad_norm": 0.06926538050174713,
      "learning_rate": 0.002,
      "loss": 2.4591798782348633,
      "step": 172
    },
    {
      "epoch": 0.0174,
      "grad_norm": 0.07229094207286835,
      "learning_rate": 0.002,
      "loss": 2.4731459617614746,
      "step": 174
    },
    {
      "epoch": 0.0176,
      "grad_norm": 0.07224471867084503,
      "learning_rate": 0.002,
      "loss": 2.5305635929107666,
      "step": 176
    },
    {
      "epoch": 0.0178,
      "grad_norm": 0.08236755430698395,
      "learning_rate": 0.002,
      "loss": 2.3938305377960205,
      "step": 178
    },
    {
      "epoch": 0.018,
      "grad_norm": 0.07654324173927307,
      "learning_rate": 0.002,
      "loss": 2.3549747467041016,
      "step": 180
    },
    {
      "epoch": 0.0182,
      "grad_norm": 0.07224669307470322,
      "learning_rate": 0.002,
      "loss": 2.4166836738586426,
      "step": 182
    },
    {
      "epoch": 0.0184,
      "grad_norm": 0.0869525894522667,
      "learning_rate": 0.002,
      "loss": 2.2796778678894043,
      "step": 184
    },
    {
      "epoch": 0.0186,
      "grad_norm": 0.07824264466762543,
      "learning_rate": 0.002,
      "loss": 2.443204879760742,
      "step": 186
    },
    {
      "epoch": 0.0188,
      "grad_norm": 0.07252666354179382,
      "learning_rate": 0.002,
      "loss": 2.4111971855163574,
      "step": 188
    },
    {
      "epoch": 0.019,
      "grad_norm": 0.0813930332660675,
      "learning_rate": 0.002,
      "loss": 2.3682315349578857,
      "step": 190
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.06918879598379135,
      "learning_rate": 0.002,
      "loss": 2.425147533416748,
      "step": 192
    },
    {
      "epoch": 0.0194,
      "grad_norm": 0.07671184092760086,
      "learning_rate": 0.002,
      "loss": 2.443631410598755,
      "step": 194
    },
    {
      "epoch": 0.0196,
      "grad_norm": 0.07814231514930725,
      "learning_rate": 0.002,
      "loss": 2.4139137268066406,
      "step": 196
    },
    {
      "epoch": 0.0198,
      "grad_norm": 0.08234577625989914,
      "learning_rate": 0.002,
      "loss": 2.409456729888916,
      "step": 198
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.06655089557170868,
      "learning_rate": 0.002,
      "loss": 2.4807326793670654,
      "step": 200
    },
    {
      "epoch": 0.0202,
      "grad_norm": 0.07245142757892609,
      "learning_rate": 0.002,
      "loss": 2.4890124797821045,
      "step": 202
    },
    {
      "epoch": 0.0204,
      "grad_norm": 0.07318440079689026,
      "learning_rate": 0.002,
      "loss": 2.3685643672943115,
      "step": 204
    },
    {
      "epoch": 0.0206,
      "grad_norm": 0.06911696493625641,
      "learning_rate": 0.002,
      "loss": 2.396023988723755,
      "step": 206
    },
    {
      "epoch": 0.0208,
      "grad_norm": 0.0727556124329567,
      "learning_rate": 0.002,
      "loss": 2.398282289505005,
      "step": 208
    },
    {
      "epoch": 0.021,
      "grad_norm": 0.07419173419475555,
      "learning_rate": 0.002,
      "loss": 2.459522008895874,
      "step": 210
    },
    {
      "epoch": 0.0212,
      "grad_norm": 0.07413039356470108,
      "learning_rate": 0.002,
      "loss": 2.3531744480133057,
      "step": 212
    },
    {
      "epoch": 0.0214,
      "grad_norm": 0.07484415918588638,
      "learning_rate": 0.002,
      "loss": 2.26939058303833,
      "step": 214
    },
    {
      "epoch": 0.0216,
      "grad_norm": 0.0806150883436203,
      "learning_rate": 0.002,
      "loss": 2.3727993965148926,
      "step": 216
    },
    {
      "epoch": 0.0218,
      "grad_norm": 0.07254747301340103,
      "learning_rate": 0.002,
      "loss": 2.574730157852173,
      "step": 218
    },
    {
      "epoch": 0.022,
      "grad_norm": 0.08061928302049637,
      "learning_rate": 0.002,
      "loss": 2.4892091751098633,
      "step": 220
    },
    {
      "epoch": 0.0222,
      "grad_norm": 0.06940927356481552,
      "learning_rate": 0.002,
      "loss": 2.344409704208374,
      "step": 222
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.07415833324193954,
      "learning_rate": 0.002,
      "loss": 2.37369441986084,
      "step": 224
    },
    {
      "epoch": 0.0226,
      "grad_norm": 0.07141406834125519,
      "learning_rate": 0.002,
      "loss": 2.3963000774383545,
      "step": 226
    },
    {
      "epoch": 0.0228,
      "grad_norm": 0.075843945145607,
      "learning_rate": 0.002,
      "loss": 2.4342799186706543,
      "step": 228
    },
    {
      "epoch": 0.023,
      "grad_norm": 0.07074150443077087,
      "learning_rate": 0.002,
      "loss": 2.4830009937286377,
      "step": 230
    },
    {
      "epoch": 0.0232,
      "grad_norm": 0.08684448897838593,
      "learning_rate": 0.002,
      "loss": 2.474318027496338,
      "step": 232
    },
    {
      "epoch": 0.0234,
      "grad_norm": 0.06847158819437027,
      "learning_rate": 0.002,
      "loss": 2.5502119064331055,
      "step": 234
    },
    {
      "epoch": 0.0236,
      "grad_norm": 0.07291707396507263,
      "learning_rate": 0.002,
      "loss": 2.361274003982544,
      "step": 236
    },
    {
      "epoch": 0.0238,
      "grad_norm": 0.0737706795334816,
      "learning_rate": 0.002,
      "loss": 2.3580234050750732,
      "step": 238
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.070985347032547,
      "learning_rate": 0.002,
      "loss": 2.4779670238494873,
      "step": 240
    },
    {
      "epoch": 0.0242,
      "grad_norm": 0.07180329412221909,
      "learning_rate": 0.002,
      "loss": 2.4767282009124756,
      "step": 242
    },
    {
      "epoch": 0.0244,
      "grad_norm": 0.06752586364746094,
      "learning_rate": 0.002,
      "loss": 2.378995418548584,
      "step": 244
    },
    {
      "epoch": 0.0246,
      "grad_norm": 0.0811859741806984,
      "learning_rate": 0.002,
      "loss": 2.436291217803955,
      "step": 246
    },
    {
      "epoch": 0.0248,
      "grad_norm": 0.07182802259922028,
      "learning_rate": 0.002,
      "loss": 2.273919105529785,
      "step": 248
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.07898867875337601,
      "learning_rate": 0.002,
      "loss": 2.5020508766174316,
      "step": 250
    },
    {
      "epoch": 0.0252,
      "grad_norm": 0.08066528290510178,
      "learning_rate": 0.002,
      "loss": 2.431199312210083,
      "step": 252
    },
    {
      "epoch": 0.0254,
      "grad_norm": 0.07442006468772888,
      "learning_rate": 0.002,
      "loss": 2.27687668800354,
      "step": 254
    },
    {
      "epoch": 0.0256,
      "grad_norm": 0.07396183162927628,
      "learning_rate": 0.002,
      "loss": 2.3320813179016113,
      "step": 256
    },
    {
      "epoch": 0.0258,
      "grad_norm": 0.07966624945402145,
      "learning_rate": 0.002,
      "loss": 2.4072208404541016,
      "step": 258
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.07383351027965546,
      "learning_rate": 0.002,
      "loss": 2.328458309173584,
      "step": 260
    },
    {
      "epoch": 0.0262,
      "grad_norm": 0.07041355222463608,
      "learning_rate": 0.002,
      "loss": 2.4800467491149902,
      "step": 262
    },
    {
      "epoch": 0.0264,
      "grad_norm": 0.06973565369844437,
      "learning_rate": 0.002,
      "loss": 2.2997543811798096,
      "step": 264
    },
    {
      "epoch": 0.0266,
      "grad_norm": 0.07369516789913177,
      "learning_rate": 0.002,
      "loss": 2.3620338439941406,
      "step": 266
    },
    {
      "epoch": 0.0268,
      "grad_norm": 0.0725259855389595,
      "learning_rate": 0.002,
      "loss": 2.445523977279663,
      "step": 268
    },
    {
      "epoch": 0.027,
      "grad_norm": 0.07374564558267593,
      "learning_rate": 0.002,
      "loss": 2.416564702987671,
      "step": 270
    },
    {
      "epoch": 0.0272,
      "grad_norm": 0.07738863676786423,
      "learning_rate": 0.002,
      "loss": 2.3530383110046387,
      "step": 272
    },
    {
      "epoch": 0.0274,
      "grad_norm": 0.07384581863880157,
      "learning_rate": 0.002,
      "loss": 2.339061975479126,
      "step": 274
    },
    {
      "epoch": 0.0276,
      "grad_norm": 0.07035136222839355,
      "learning_rate": 0.002,
      "loss": 2.447986364364624,
      "step": 276
    },
    {
      "epoch": 0.0278,
      "grad_norm": 0.07077682763338089,
      "learning_rate": 0.002,
      "loss": 2.2173964977264404,
      "step": 278
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.07278954237699509,
      "learning_rate": 0.002,
      "loss": 2.338022470474243,
      "step": 280
    },
    {
      "epoch": 0.0282,
      "grad_norm": 0.4929845631122589,
      "learning_rate": 0.002,
      "loss": 2.456202745437622,
      "step": 282
    },
    {
      "epoch": 0.0284,
      "grad_norm": 0.07330320030450821,
      "learning_rate": 0.002,
      "loss": 2.506326913833618,
      "step": 284
    },
    {
      "epoch": 0.0286,
      "grad_norm": 0.07894188165664673,
      "learning_rate": 0.002,
      "loss": 2.3673806190490723,
      "step": 286
    },
    {
      "epoch": 0.0288,
      "grad_norm": 0.07210055738687515,
      "learning_rate": 0.002,
      "loss": 2.3784022331237793,
      "step": 288
    },
    {
      "epoch": 0.029,
      "grad_norm": 0.07566384971141815,
      "learning_rate": 0.002,
      "loss": 2.2638120651245117,
      "step": 290
    },
    {
      "epoch": 0.0292,
      "grad_norm": 0.074443019926548,
      "learning_rate": 0.002,
      "loss": 2.417232036590576,
      "step": 292
    },
    {
      "epoch": 0.0294,
      "grad_norm": 0.06952570378780365,
      "learning_rate": 0.002,
      "loss": 2.5657687187194824,
      "step": 294
    },
    {
      "epoch": 0.0296,
      "grad_norm": 0.06942253559827805,
      "learning_rate": 0.002,
      "loss": 2.5299112796783447,
      "step": 296
    },
    {
      "epoch": 0.0298,
      "grad_norm": 0.07203349471092224,
      "learning_rate": 0.002,
      "loss": 2.3146815299987793,
      "step": 298
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.07411694526672363,
      "learning_rate": 0.002,
      "loss": 2.2963788509368896,
      "step": 300
    },
    {
      "epoch": 0.0302,
      "grad_norm": 0.08336383104324341,
      "learning_rate": 0.002,
      "loss": 2.377774477005005,
      "step": 302
    },
    {
      "epoch": 0.0304,
      "grad_norm": 0.07448665052652359,
      "learning_rate": 0.002,
      "loss": 2.325913906097412,
      "step": 304
    },
    {
      "epoch": 0.0306,
      "grad_norm": 0.09045958518981934,
      "learning_rate": 0.002,
      "loss": 2.3425965309143066,
      "step": 306
    },
    {
      "epoch": 0.0308,
      "grad_norm": 0.06627941131591797,
      "learning_rate": 0.002,
      "loss": 2.384185314178467,
      "step": 308
    },
    {
      "epoch": 0.031,
      "grad_norm": 0.0740562230348587,
      "learning_rate": 0.002,
      "loss": 2.375741958618164,
      "step": 310
    },
    {
      "epoch": 0.0312,
      "grad_norm": 0.07392635941505432,
      "learning_rate": 0.002,
      "loss": 2.4404001235961914,
      "step": 312
    },
    {
      "epoch": 0.0314,
      "grad_norm": 0.07531018555164337,
      "learning_rate": 0.002,
      "loss": 2.371039867401123,
      "step": 314
    },
    {
      "epoch": 0.0316,
      "grad_norm": 0.07442319393157959,
      "learning_rate": 0.002,
      "loss": 2.4006941318511963,
      "step": 316
    },
    {
      "epoch": 0.0318,
      "grad_norm": 0.08017536252737045,
      "learning_rate": 0.002,
      "loss": 2.466963768005371,
      "step": 318
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.07036197185516357,
      "learning_rate": 0.002,
      "loss": 2.4882895946502686,
      "step": 320
    },
    {
      "epoch": 0.0322,
      "grad_norm": 0.07494204491376877,
      "learning_rate": 0.002,
      "loss": 2.285078287124634,
      "step": 322
    },
    {
      "epoch": 0.0324,
      "grad_norm": 0.077792689204216,
      "learning_rate": 0.002,
      "loss": 2.4089248180389404,
      "step": 324
    },
    {
      "epoch": 0.0326,
      "grad_norm": 0.07506862282752991,
      "learning_rate": 0.002,
      "loss": 2.3210463523864746,
      "step": 326
    },
    {
      "epoch": 0.0328,
      "grad_norm": 0.07173958420753479,
      "learning_rate": 0.002,
      "loss": 2.3689537048339844,
      "step": 328
    },
    {
      "epoch": 0.033,
      "grad_norm": 0.07319604605436325,
      "learning_rate": 0.002,
      "loss": 2.345275402069092,
      "step": 330
    },
    {
      "epoch": 0.0332,
      "grad_norm": 0.07139241695404053,
      "learning_rate": 0.002,
      "loss": 2.2951619625091553,
      "step": 332
    },
    {
      "epoch": 0.0334,
      "grad_norm": 0.07661397010087967,
      "learning_rate": 0.002,
      "loss": 2.4020214080810547,
      "step": 334
    },
    {
      "epoch": 0.0336,
      "grad_norm": 0.0786265879869461,
      "learning_rate": 0.002,
      "loss": 2.403564453125,
      "step": 336
    },
    {
      "epoch": 0.0338,
      "grad_norm": 0.06836353987455368,
      "learning_rate": 0.002,
      "loss": 2.339182138442993,
      "step": 338
    },
    {
      "epoch": 0.034,
      "grad_norm": 0.07299920171499252,
      "learning_rate": 0.002,
      "loss": 2.3555524349212646,
      "step": 340
    },
    {
      "epoch": 0.0342,
      "grad_norm": 0.08625968545675278,
      "learning_rate": 0.002,
      "loss": 2.415715456008911,
      "step": 342
    },
    {
      "epoch": 0.0344,
      "grad_norm": 0.07323940098285675,
      "learning_rate": 0.002,
      "loss": 2.3695898056030273,
      "step": 344
    },
    {
      "epoch": 0.0346,
      "grad_norm": 0.07468345761299133,
      "learning_rate": 0.002,
      "loss": 2.3610243797302246,
      "step": 346
    },
    {
      "epoch": 0.0348,
      "grad_norm": 0.07465499639511108,
      "learning_rate": 0.002,
      "loss": 2.3424766063690186,
      "step": 348
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.07375511527061462,
      "learning_rate": 0.002,
      "loss": 2.3318700790405273,
      "step": 350
    },
    {
      "epoch": 0.0352,
      "grad_norm": 0.0743214562535286,
      "learning_rate": 0.002,
      "loss": 2.2821524143218994,
      "step": 352
    },
    {
      "epoch": 0.0354,
      "grad_norm": 0.07222682982683182,
      "learning_rate": 0.002,
      "loss": 2.27455472946167,
      "step": 354
    },
    {
      "epoch": 0.0356,
      "grad_norm": 0.07145558297634125,
      "learning_rate": 0.002,
      "loss": 2.270270586013794,
      "step": 356
    },
    {
      "epoch": 0.0358,
      "grad_norm": 0.07416308671236038,
      "learning_rate": 0.002,
      "loss": 2.3169808387756348,
      "step": 358
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.07120826095342636,
      "learning_rate": 0.002,
      "loss": 2.365264654159546,
      "step": 360
    },
    {
      "epoch": 0.0362,
      "grad_norm": 0.07509961724281311,
      "learning_rate": 0.002,
      "loss": 2.3909354209899902,
      "step": 362
    },
    {
      "epoch": 0.0364,
      "grad_norm": 0.07583671808242798,
      "learning_rate": 0.002,
      "loss": 2.245464563369751,
      "step": 364
    },
    {
      "epoch": 0.0366,
      "grad_norm": 0.11906367540359497,
      "learning_rate": 0.002,
      "loss": 2.3648908138275146,
      "step": 366
    },
    {
      "epoch": 0.0368,
      "grad_norm": 0.071116603910923,
      "learning_rate": 0.002,
      "loss": 2.2651329040527344,
      "step": 368
    },
    {
      "epoch": 0.037,
      "grad_norm": 0.07210150361061096,
      "learning_rate": 0.002,
      "loss": 2.370694398880005,
      "step": 370
    },
    {
      "epoch": 0.0372,
      "grad_norm": 0.07139323651790619,
      "learning_rate": 0.002,
      "loss": 2.289454936981201,
      "step": 372
    },
    {
      "epoch": 0.0374,
      "grad_norm": 0.06869690865278244,
      "learning_rate": 0.002,
      "loss": 2.2915279865264893,
      "step": 374
    },
    {
      "epoch": 0.0376,
      "grad_norm": 0.07427705824375153,
      "learning_rate": 0.002,
      "loss": 2.4037320613861084,
      "step": 376
    },
    {
      "epoch": 0.0378,
      "grad_norm": 0.07436185330152512,
      "learning_rate": 0.002,
      "loss": 2.364893674850464,
      "step": 378
    },
    {
      "epoch": 0.038,
      "grad_norm": 0.06889235228300095,
      "learning_rate": 0.002,
      "loss": 2.3860666751861572,
      "step": 380
    },
    {
      "epoch": 0.0382,
      "grad_norm": 0.07245805114507675,
      "learning_rate": 0.002,
      "loss": 2.330124616622925,
      "step": 382
    },
    {
      "epoch": 0.0384,
      "grad_norm": 0.07587055116891861,
      "learning_rate": 0.002,
      "loss": 2.426389455795288,
      "step": 384
    },
    {
      "epoch": 0.0386,
      "grad_norm": 0.0699508860707283,
      "learning_rate": 0.002,
      "loss": 2.273881196975708,
      "step": 386
    },
    {
      "epoch": 0.0388,
      "grad_norm": 0.07509642839431763,
      "learning_rate": 0.002,
      "loss": 2.4309396743774414,
      "step": 388
    },
    {
      "epoch": 0.039,
      "grad_norm": 0.0673062726855278,
      "learning_rate": 0.002,
      "loss": 2.3857004642486572,
      "step": 390
    },
    {
      "epoch": 0.0392,
      "grad_norm": 0.07434364408254623,
      "learning_rate": 0.002,
      "loss": 2.215848684310913,
      "step": 392
    },
    {
      "epoch": 0.0394,
      "grad_norm": 0.07386846840381622,
      "learning_rate": 0.002,
      "loss": 2.4128293991088867,
      "step": 394
    },
    {
      "epoch": 0.0396,
      "grad_norm": 0.07630185782909393,
      "learning_rate": 0.002,
      "loss": 2.2950692176818848,
      "step": 396
    },
    {
      "epoch": 0.0398,
      "grad_norm": 0.07454672455787659,
      "learning_rate": 0.002,
      "loss": 2.246983766555786,
      "step": 398
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.07129548490047455,
      "learning_rate": 0.002,
      "loss": 2.29495906829834,
      "step": 400
    },
    {
      "epoch": 0.0402,
      "grad_norm": 0.07600388675928116,
      "learning_rate": 0.002,
      "loss": 2.2774150371551514,
      "step": 402
    },
    {
      "epoch": 0.0404,
      "grad_norm": 0.06913933902978897,
      "learning_rate": 0.002,
      "loss": 2.2019636631011963,
      "step": 404
    },
    {
      "epoch": 0.0406,
      "grad_norm": 0.07273747771978378,
      "learning_rate": 0.002,
      "loss": 2.479687452316284,
      "step": 406
    },
    {
      "epoch": 0.0408,
      "grad_norm": 0.07171725481748581,
      "learning_rate": 0.002,
      "loss": 2.3696889877319336,
      "step": 408
    },
    {
      "epoch": 0.041,
      "grad_norm": 0.07488934695720673,
      "learning_rate": 0.002,
      "loss": 2.4270987510681152,
      "step": 410
    },
    {
      "epoch": 0.0412,
      "grad_norm": 0.07063031196594238,
      "learning_rate": 0.002,
      "loss": 2.211313009262085,
      "step": 412
    },
    {
      "epoch": 0.0414,
      "grad_norm": 0.07202549278736115,
      "learning_rate": 0.002,
      "loss": 2.4381520748138428,
      "step": 414
    },
    {
      "epoch": 0.0416,
      "grad_norm": 0.07619333267211914,
      "learning_rate": 0.002,
      "loss": 2.294527769088745,
      "step": 416
    },
    {
      "epoch": 0.0418,
      "grad_norm": 0.07141901552677155,
      "learning_rate": 0.002,
      "loss": 2.4251363277435303,
      "step": 418
    },
    {
      "epoch": 0.042,
      "grad_norm": 0.07334240525960922,
      "learning_rate": 0.002,
      "loss": 2.4089479446411133,
      "step": 420
    },
    {
      "epoch": 0.0422,
      "grad_norm": 0.07610740512609482,
      "learning_rate": 0.002,
      "loss": 2.2459805011749268,
      "step": 422
    },
    {
      "epoch": 0.0424,
      "grad_norm": 0.07029259949922562,
      "learning_rate": 0.002,
      "loss": 2.2798752784729004,
      "step": 424
    },
    {
      "epoch": 0.0426,
      "grad_norm": 0.07146398723125458,
      "learning_rate": 0.002,
      "loss": 2.280799150466919,
      "step": 426
    },
    {
      "epoch": 0.0428,
      "grad_norm": 0.07202620804309845,
      "learning_rate": 0.002,
      "loss": 2.259249210357666,
      "step": 428
    },
    {
      "epoch": 0.043,
      "grad_norm": 0.07183374464511871,
      "learning_rate": 0.002,
      "loss": 2.286275625228882,
      "step": 430
    },
    {
      "epoch": 0.0432,
      "grad_norm": 0.07452496886253357,
      "learning_rate": 0.002,
      "loss": 2.3629934787750244,
      "step": 432
    },
    {
      "epoch": 0.0434,
      "grad_norm": 0.0716223418712616,
      "learning_rate": 0.002,
      "loss": 2.377845525741577,
      "step": 434
    },
    {
      "epoch": 0.0436,
      "grad_norm": 0.0725131705403328,
      "learning_rate": 0.002,
      "loss": 2.236332654953003,
      "step": 436
    },
    {
      "epoch": 0.0438,
      "grad_norm": 0.07328413426876068,
      "learning_rate": 0.002,
      "loss": 2.241804361343384,
      "step": 438
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.07485053688287735,
      "learning_rate": 0.002,
      "loss": 2.2544145584106445,
      "step": 440
    },
    {
      "epoch": 0.0442,
      "grad_norm": 0.0708366334438324,
      "learning_rate": 0.002,
      "loss": 2.342287063598633,
      "step": 442
    },
    {
      "epoch": 0.0444,
      "grad_norm": 0.07225901633501053,
      "learning_rate": 0.002,
      "loss": 2.4516775608062744,
      "step": 444
    },
    {
      "epoch": 0.0446,
      "grad_norm": 0.07705280184745789,
      "learning_rate": 0.002,
      "loss": 2.1962647438049316,
      "step": 446
    },
    {
      "epoch": 0.0448,
      "grad_norm": 0.07026544958353043,
      "learning_rate": 0.002,
      "loss": 2.3649749755859375,
      "step": 448
    },
    {
      "epoch": 0.045,
      "grad_norm": 0.0726761594414711,
      "learning_rate": 0.002,
      "loss": 2.4410927295684814,
      "step": 450
    },
    {
      "epoch": 0.0452,
      "grad_norm": 0.07539156824350357,
      "learning_rate": 0.002,
      "loss": 2.3823137283325195,
      "step": 452
    },
    {
      "epoch": 0.0454,
      "grad_norm": 0.07577167451381683,
      "learning_rate": 0.002,
      "loss": 2.4145004749298096,
      "step": 454
    },
    {
      "epoch": 0.0456,
      "grad_norm": 0.07489137351512909,
      "learning_rate": 0.002,
      "loss": 2.2637829780578613,
      "step": 456
    },
    {
      "epoch": 0.0458,
      "grad_norm": 0.07107007503509521,
      "learning_rate": 0.002,
      "loss": 2.1722826957702637,
      "step": 458
    },
    {
      "epoch": 0.046,
      "grad_norm": 0.07039529830217361,
      "learning_rate": 0.002,
      "loss": 2.319223165512085,
      "step": 460
    },
    {
      "epoch": 0.0462,
      "grad_norm": 0.07077178359031677,
      "learning_rate": 0.002,
      "loss": 2.283263921737671,
      "step": 462
    },
    {
      "epoch": 0.0464,
      "grad_norm": 0.07469630986452103,
      "learning_rate": 0.002,
      "loss": 2.313534736633301,
      "step": 464
    },
    {
      "epoch": 0.0466,
      "grad_norm": 0.07318571954965591,
      "learning_rate": 0.002,
      "loss": 2.3414745330810547,
      "step": 466
    },
    {
      "epoch": 0.0468,
      "grad_norm": 0.07057257741689682,
      "learning_rate": 0.002,
      "loss": 2.262636184692383,
      "step": 468
    },
    {
      "epoch": 0.047,
      "grad_norm": 0.07172814756631851,
      "learning_rate": 0.002,
      "loss": 2.467683792114258,
      "step": 470
    },
    {
      "epoch": 0.0472,
      "grad_norm": 0.07254748791456223,
      "learning_rate": 0.002,
      "loss": 2.3331737518310547,
      "step": 472
    },
    {
      "epoch": 0.0474,
      "grad_norm": 0.07001584768295288,
      "learning_rate": 0.002,
      "loss": 2.34952712059021,
      "step": 474
    },
    {
      "epoch": 0.0476,
      "grad_norm": 0.07097060978412628,
      "learning_rate": 0.002,
      "loss": 2.1702311038970947,
      "step": 476
    },
    {
      "epoch": 0.0478,
      "grad_norm": 0.07471913844347,
      "learning_rate": 0.002,
      "loss": 2.341153383255005,
      "step": 478
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.07375536113977432,
      "learning_rate": 0.002,
      "loss": 2.1913836002349854,
      "step": 480
    },
    {
      "epoch": 0.0482,
      "grad_norm": 0.0722331702709198,
      "learning_rate": 0.002,
      "loss": 2.346038579940796,
      "step": 482
    },
    {
      "epoch": 0.0484,
      "grad_norm": 0.06913427263498306,
      "learning_rate": 0.002,
      "loss": 2.3323590755462646,
      "step": 484
    },
    {
      "epoch": 0.0486,
      "grad_norm": 0.07356101274490356,
      "learning_rate": 0.002,
      "loss": 2.479645013809204,
      "step": 486
    },
    {
      "epoch": 0.0488,
      "grad_norm": 0.08431404083967209,
      "learning_rate": 0.002,
      "loss": 2.4104349613189697,
      "step": 488
    },
    {
      "epoch": 0.049,
      "grad_norm": 0.07982538640499115,
      "learning_rate": 0.002,
      "loss": 2.3613436222076416,
      "step": 490
    },
    {
      "epoch": 0.0492,
      "grad_norm": 0.07325701415538788,
      "learning_rate": 0.002,
      "loss": 2.2643299102783203,
      "step": 492
    },
    {
      "epoch": 0.0494,
      "grad_norm": 0.07172851264476776,
      "learning_rate": 0.002,
      "loss": 2.3110156059265137,
      "step": 494
    },
    {
      "epoch": 0.0496,
      "grad_norm": 0.07235249876976013,
      "learning_rate": 0.002,
      "loss": 2.1768910884857178,
      "step": 496
    },
    {
      "epoch": 0.0498,
      "grad_norm": 0.0703548938035965,
      "learning_rate": 0.002,
      "loss": 2.331031084060669,
      "step": 498
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.09161035716533661,
      "learning_rate": 0.002,
      "loss": 2.291215181350708,
      "step": 500
    },
    {
      "epoch": 0.0502,
      "grad_norm": 0.07420938462018967,
      "learning_rate": 0.002,
      "loss": 2.2634973526000977,
      "step": 502
    },
    {
      "epoch": 0.0504,
      "grad_norm": 0.07346899807453156,
      "learning_rate": 0.002,
      "loss": 2.2813985347747803,
      "step": 504
    },
    {
      "epoch": 0.0506,
      "grad_norm": 0.0710308626294136,
      "learning_rate": 0.002,
      "loss": 2.325071334838867,
      "step": 506
    },
    {
      "epoch": 0.0508,
      "grad_norm": 0.07305864244699478,
      "learning_rate": 0.002,
      "loss": 2.23004150390625,
      "step": 508
    },
    {
      "epoch": 0.051,
      "grad_norm": 0.07333915680646896,
      "learning_rate": 0.002,
      "loss": 2.208730697631836,
      "step": 510
    },
    {
      "epoch": 0.0512,
      "grad_norm": 0.07314565777778625,
      "learning_rate": 0.002,
      "loss": 2.2246975898742676,
      "step": 512
    },
    {
      "epoch": 0.0514,
      "grad_norm": 0.07092233002185822,
      "learning_rate": 0.002,
      "loss": 2.3782448768615723,
      "step": 514
    },
    {
      "epoch": 0.0516,
      "grad_norm": 0.0718497782945633,
      "learning_rate": 0.002,
      "loss": 2.4314613342285156,
      "step": 516
    },
    {
      "epoch": 0.0518,
      "grad_norm": 0.06949327141046524,
      "learning_rate": 0.002,
      "loss": 2.3504679203033447,
      "step": 518
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.0747607946395874,
      "learning_rate": 0.002,
      "loss": 2.4246621131896973,
      "step": 520
    },
    {
      "epoch": 0.0522,
      "grad_norm": 0.07107537239789963,
      "learning_rate": 0.002,
      "loss": 2.2178404331207275,
      "step": 522
    },
    {
      "epoch": 0.0524,
      "grad_norm": 0.06897857785224915,
      "learning_rate": 0.002,
      "loss": 2.2822928428649902,
      "step": 524
    },
    {
      "epoch": 0.0526,
      "grad_norm": 0.07319613546133041,
      "learning_rate": 0.002,
      "loss": 2.217073678970337,
      "step": 526
    },
    {
      "epoch": 0.0528,
      "grad_norm": 0.07724731415510178,
      "learning_rate": 0.002,
      "loss": 2.249058485031128,
      "step": 528
    },
    {
      "epoch": 0.053,
      "grad_norm": 0.07000409811735153,
      "learning_rate": 0.002,
      "loss": 2.345240831375122,
      "step": 530
    },
    {
      "epoch": 0.0532,
      "grad_norm": 0.07471994310617447,
      "learning_rate": 0.002,
      "loss": 2.273231029510498,
      "step": 532
    },
    {
      "epoch": 0.0534,
      "grad_norm": 0.06968633085489273,
      "learning_rate": 0.002,
      "loss": 2.3197133541107178,
      "step": 534
    },
    {
      "epoch": 0.0536,
      "grad_norm": 0.07194774597883224,
      "learning_rate": 0.002,
      "loss": 2.262237548828125,
      "step": 536
    },
    {
      "epoch": 0.0538,
      "grad_norm": 0.07644084841012955,
      "learning_rate": 0.002,
      "loss": 2.2549495697021484,
      "step": 538
    },
    {
      "epoch": 0.054,
      "grad_norm": 0.07824552059173584,
      "learning_rate": 0.002,
      "loss": 2.2192821502685547,
      "step": 540
    },
    {
      "epoch": 0.0542,
      "grad_norm": 0.07208573073148727,
      "learning_rate": 0.002,
      "loss": 2.2678325176239014,
      "step": 542
    },
    {
      "epoch": 0.0544,
      "grad_norm": 0.0775320827960968,
      "learning_rate": 0.002,
      "loss": 2.228356122970581,
      "step": 544
    },
    {
      "epoch": 0.0546,
      "grad_norm": 0.07124815136194229,
      "learning_rate": 0.002,
      "loss": 2.234506368637085,
      "step": 546
    },
    {
      "epoch": 0.0548,
      "grad_norm": 0.07150725275278091,
      "learning_rate": 0.002,
      "loss": 2.35404896736145,
      "step": 548
    },
    {
      "epoch": 0.055,
      "grad_norm": 0.07144594192504883,
      "learning_rate": 0.002,
      "loss": 2.319481611251831,
      "step": 550
    },
    {
      "epoch": 0.0552,
      "grad_norm": 0.07902579754590988,
      "learning_rate": 0.002,
      "loss": 2.190006732940674,
      "step": 552
    },
    {
      "epoch": 0.0554,
      "grad_norm": 0.0708867534995079,
      "learning_rate": 0.002,
      "loss": 2.3096513748168945,
      "step": 554
    },
    {
      "epoch": 0.0556,
      "grad_norm": 0.07353733479976654,
      "learning_rate": 0.002,
      "loss": 2.326336622238159,
      "step": 556
    },
    {
      "epoch": 0.0558,
      "grad_norm": 0.07306534051895142,
      "learning_rate": 0.002,
      "loss": 2.277613878250122,
      "step": 558
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.07439783215522766,
      "learning_rate": 0.002,
      "loss": 2.167388916015625,
      "step": 560
    },
    {
      "epoch": 0.0562,
      "grad_norm": 0.07338475435972214,
      "learning_rate": 0.002,
      "loss": 2.24794340133667,
      "step": 562
    },
    {
      "epoch": 0.0564,
      "grad_norm": 0.0688517615199089,
      "learning_rate": 0.002,
      "loss": 2.279372215270996,
      "step": 564
    },
    {
      "epoch": 0.0566,
      "grad_norm": 0.0748622938990593,
      "learning_rate": 0.002,
      "loss": 2.3502681255340576,
      "step": 566
    },
    {
      "epoch": 0.0568,
      "grad_norm": 0.07456439733505249,
      "learning_rate": 0.002,
      "loss": 2.2131736278533936,
      "step": 568
    },
    {
      "epoch": 0.057,
      "grad_norm": 0.0729939416050911,
      "learning_rate": 0.002,
      "loss": 2.2854838371276855,
      "step": 570
    },
    {
      "epoch": 0.0572,
      "grad_norm": 0.07482558488845825,
      "learning_rate": 0.002,
      "loss": 2.326373338699341,
      "step": 572
    },
    {
      "epoch": 0.0574,
      "grad_norm": 0.07817833125591278,
      "learning_rate": 0.002,
      "loss": 2.26328182220459,
      "step": 574
    },
    {
      "epoch": 0.0576,
      "grad_norm": 0.07271433621644974,
      "learning_rate": 0.002,
      "loss": 2.2213103771209717,
      "step": 576
    },
    {
      "epoch": 0.0578,
      "grad_norm": 0.07567833364009857,
      "learning_rate": 0.002,
      "loss": 2.2155230045318604,
      "step": 578
    },
    {
      "epoch": 0.058,
      "grad_norm": 0.07319347560405731,
      "learning_rate": 0.002,
      "loss": 2.319739580154419,
      "step": 580
    },
    {
      "epoch": 0.0582,
      "grad_norm": 0.07753495126962662,
      "learning_rate": 0.002,
      "loss": 2.3420681953430176,
      "step": 582
    },
    {
      "epoch": 0.0584,
      "grad_norm": 0.07506182044744492,
      "learning_rate": 0.002,
      "loss": 2.1648738384246826,
      "step": 584
    },
    {
      "epoch": 0.0586,
      "grad_norm": 0.0706704631447792,
      "learning_rate": 0.002,
      "loss": 2.201256036758423,
      "step": 586
    },
    {
      "epoch": 0.0588,
      "grad_norm": 0.07100903242826462,
      "learning_rate": 0.002,
      "loss": 2.2186548709869385,
      "step": 588
    },
    {
      "epoch": 0.059,
      "grad_norm": 0.07426289469003677,
      "learning_rate": 0.002,
      "loss": 2.2193641662597656,
      "step": 590
    },
    {
      "epoch": 0.0592,
      "grad_norm": 0.08388280868530273,
      "learning_rate": 0.002,
      "loss": 2.3096694946289062,
      "step": 592
    },
    {
      "epoch": 0.0594,
      "grad_norm": 0.07024192810058594,
      "learning_rate": 0.002,
      "loss": 2.3697125911712646,
      "step": 594
    },
    {
      "epoch": 0.0596,
      "grad_norm": 0.07914943993091583,
      "learning_rate": 0.002,
      "loss": 2.1751325130462646,
      "step": 596
    },
    {
      "epoch": 0.0598,
      "grad_norm": 0.07455535233020782,
      "learning_rate": 0.002,
      "loss": 2.2722086906433105,
      "step": 598
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.06893296539783478,
      "learning_rate": 0.002,
      "loss": 2.2316699028015137,
      "step": 600
    },
    {
      "epoch": 0.0602,
      "grad_norm": 0.07333877682685852,
      "learning_rate": 0.002,
      "loss": 2.2802155017852783,
      "step": 602
    },
    {
      "epoch": 0.0604,
      "grad_norm": 0.07228221744298935,
      "learning_rate": 0.002,
      "loss": 2.3180289268493652,
      "step": 604
    },
    {
      "epoch": 0.0606,
      "grad_norm": 0.07011230289936066,
      "learning_rate": 0.002,
      "loss": 2.337013006210327,
      "step": 606
    },
    {
      "epoch": 0.0608,
      "grad_norm": 0.07029574364423752,
      "learning_rate": 0.002,
      "loss": 2.344632148742676,
      "step": 608
    },
    {
      "epoch": 0.061,
      "grad_norm": 0.0701148509979248,
      "learning_rate": 0.002,
      "loss": 2.2388885021209717,
      "step": 610
    },
    {
      "epoch": 0.0612,
      "grad_norm": 0.06939805299043655,
      "learning_rate": 0.002,
      "loss": 2.250124454498291,
      "step": 612
    },
    {
      "epoch": 0.0614,
      "grad_norm": 0.07221860438585281,
      "learning_rate": 0.002,
      "loss": 2.1665332317352295,
      "step": 614
    },
    {
      "epoch": 0.0616,
      "grad_norm": 0.07402171939611435,
      "learning_rate": 0.002,
      "loss": 2.2270114421844482,
      "step": 616
    },
    {
      "epoch": 0.0618,
      "grad_norm": 0.07727774232625961,
      "learning_rate": 0.002,
      "loss": 2.2267913818359375,
      "step": 618
    },
    {
      "epoch": 0.062,
      "grad_norm": 0.07065577805042267,
      "learning_rate": 0.002,
      "loss": 2.2791788578033447,
      "step": 620
    },
    {
      "epoch": 0.0622,
      "grad_norm": 0.0749676302075386,
      "learning_rate": 0.002,
      "loss": 2.2102551460266113,
      "step": 622
    },
    {
      "epoch": 0.0624,
      "grad_norm": 0.07280126214027405,
      "learning_rate": 0.002,
      "loss": 2.2585511207580566,
      "step": 624
    },
    {
      "epoch": 0.0626,
      "grad_norm": 0.07930736243724823,
      "learning_rate": 0.002,
      "loss": 2.2406718730926514,
      "step": 626
    },
    {
      "epoch": 0.0628,
      "grad_norm": 0.0734144002199173,
      "learning_rate": 0.002,
      "loss": 2.2946624755859375,
      "step": 628
    },
    {
      "epoch": 0.063,
      "grad_norm": 0.07968856394290924,
      "learning_rate": 0.002,
      "loss": 2.3353049755096436,
      "step": 630
    },
    {
      "epoch": 0.0632,
      "grad_norm": 0.07369498163461685,
      "learning_rate": 0.002,
      "loss": 2.2540104389190674,
      "step": 632
    },
    {
      "epoch": 0.0634,
      "grad_norm": 0.075535349547863,
      "learning_rate": 0.002,
      "loss": 2.3142411708831787,
      "step": 634
    },
    {
      "epoch": 0.0636,
      "grad_norm": 0.07109439373016357,
      "learning_rate": 0.002,
      "loss": 2.2230167388916016,
      "step": 636
    },
    {
      "epoch": 0.0638,
      "grad_norm": 0.0774507150053978,
      "learning_rate": 0.002,
      "loss": 2.2978708744049072,
      "step": 638
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.07279954105615616,
      "learning_rate": 0.002,
      "loss": 2.3673861026763916,
      "step": 640
    },
    {
      "epoch": 0.0642,
      "grad_norm": 0.07239579409360886,
      "learning_rate": 0.002,
      "loss": 2.25063157081604,
      "step": 642
    },
    {
      "epoch": 0.0644,
      "grad_norm": 0.07270967960357666,
      "learning_rate": 0.002,
      "loss": 2.152308225631714,
      "step": 644
    },
    {
      "epoch": 0.0646,
      "grad_norm": 0.0752786174416542,
      "learning_rate": 0.002,
      "loss": 2.419753313064575,
      "step": 646
    },
    {
      "epoch": 0.0648,
      "grad_norm": 0.0713186264038086,
      "learning_rate": 0.002,
      "loss": 2.272944927215576,
      "step": 648
    },
    {
      "epoch": 0.065,
      "grad_norm": 0.07522806525230408,
      "learning_rate": 0.002,
      "loss": 2.2693440914154053,
      "step": 650
    },
    {
      "epoch": 0.0652,
      "grad_norm": 0.07373610883951187,
      "learning_rate": 0.002,
      "loss": 2.224578857421875,
      "step": 652
    },
    {
      "epoch": 0.0654,
      "grad_norm": 0.0745680183172226,
      "learning_rate": 0.002,
      "loss": 2.3322482109069824,
      "step": 654
    },
    {
      "epoch": 0.0656,
      "grad_norm": 0.07777392119169235,
      "learning_rate": 0.002,
      "loss": 2.2732057571411133,
      "step": 656
    },
    {
      "epoch": 0.0658,
      "grad_norm": 0.07151927798986435,
      "learning_rate": 0.002,
      "loss": 2.317502737045288,
      "step": 658
    },
    {
      "epoch": 0.066,
      "grad_norm": 0.07438585162162781,
      "learning_rate": 0.002,
      "loss": 2.2899398803710938,
      "step": 660
    },
    {
      "epoch": 0.0662,
      "grad_norm": 0.0719580128788948,
      "learning_rate": 0.002,
      "loss": 2.2779953479766846,
      "step": 662
    },
    {
      "epoch": 0.0664,
      "grad_norm": 0.07105470448732376,
      "learning_rate": 0.002,
      "loss": 2.3206989765167236,
      "step": 664
    },
    {
      "epoch": 0.0666,
      "grad_norm": 0.07447022944688797,
      "learning_rate": 0.002,
      "loss": 2.3460276126861572,
      "step": 666
    },
    {
      "epoch": 0.0668,
      "grad_norm": 0.07267604768276215,
      "learning_rate": 0.002,
      "loss": 2.3276047706604004,
      "step": 668
    },
    {
      "epoch": 0.067,
      "grad_norm": 0.07084916532039642,
      "learning_rate": 0.002,
      "loss": 2.1856143474578857,
      "step": 670
    },
    {
      "epoch": 0.0672,
      "grad_norm": 0.07767835259437561,
      "learning_rate": 0.002,
      "loss": 2.2237675189971924,
      "step": 672
    },
    {
      "epoch": 0.0674,
      "grad_norm": 0.07320534437894821,
      "learning_rate": 0.002,
      "loss": 2.34051775932312,
      "step": 674
    },
    {
      "epoch": 0.0676,
      "grad_norm": 0.0703076496720314,
      "learning_rate": 0.002,
      "loss": 2.2506422996520996,
      "step": 676
    },
    {
      "epoch": 0.0678,
      "grad_norm": 0.0718763917684555,
      "learning_rate": 0.002,
      "loss": 2.1845664978027344,
      "step": 678
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.0752389207482338,
      "learning_rate": 0.002,
      "loss": 2.2942471504211426,
      "step": 680
    },
    {
      "epoch": 0.0682,
      "grad_norm": 0.07255885004997253,
      "learning_rate": 0.002,
      "loss": 2.2754502296447754,
      "step": 682
    },
    {
      "epoch": 0.0684,
      "grad_norm": 0.07524032890796661,
      "learning_rate": 0.002,
      "loss": 2.255314350128174,
      "step": 684
    },
    {
      "epoch": 0.0686,
      "grad_norm": 0.07546036690473557,
      "learning_rate": 0.002,
      "loss": 2.180896520614624,
      "step": 686
    },
    {
      "epoch": 0.0688,
      "grad_norm": 0.07169272005558014,
      "learning_rate": 0.002,
      "loss": 2.0486462116241455,
      "step": 688
    },
    {
      "epoch": 0.069,
      "grad_norm": 0.07340013980865479,
      "learning_rate": 0.002,
      "loss": 2.247344732284546,
      "step": 690
    },
    {
      "epoch": 0.0692,
      "grad_norm": 0.07301296293735504,
      "learning_rate": 0.002,
      "loss": 2.310363531112671,
      "step": 692
    },
    {
      "epoch": 0.0694,
      "grad_norm": 0.07676922529935837,
      "learning_rate": 0.002,
      "loss": 2.289912462234497,
      "step": 694
    },
    {
      "epoch": 0.0696,
      "grad_norm": 0.07127033919095993,
      "learning_rate": 0.002,
      "loss": 2.208883285522461,
      "step": 696
    },
    {
      "epoch": 0.0698,
      "grad_norm": 0.08236896991729736,
      "learning_rate": 0.002,
      "loss": 2.25746488571167,
      "step": 698
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.0688832551240921,
      "learning_rate": 0.002,
      "loss": 2.148230791091919,
      "step": 700
    },
    {
      "epoch": 0.0702,
      "grad_norm": 0.07343868911266327,
      "learning_rate": 0.002,
      "loss": 2.3926994800567627,
      "step": 702
    },
    {
      "epoch": 0.0704,
      "grad_norm": 0.07114210724830627,
      "learning_rate": 0.002,
      "loss": 2.302194118499756,
      "step": 704
    },
    {
      "epoch": 0.0706,
      "grad_norm": 0.07199382036924362,
      "learning_rate": 0.002,
      "loss": 2.213334321975708,
      "step": 706
    },
    {
      "epoch": 0.0708,
      "grad_norm": 0.07205311208963394,
      "learning_rate": 0.002,
      "loss": 2.1924545764923096,
      "step": 708
    },
    {
      "epoch": 0.071,
      "grad_norm": 0.07413750886917114,
      "learning_rate": 0.002,
      "loss": 2.2514028549194336,
      "step": 710
    },
    {
      "epoch": 0.0712,
      "grad_norm": 0.07480225712060928,
      "learning_rate": 0.002,
      "loss": 2.285456657409668,
      "step": 712
    },
    {
      "epoch": 0.0714,
      "grad_norm": 0.070404052734375,
      "learning_rate": 0.002,
      "loss": 2.3142588138580322,
      "step": 714
    },
    {
      "epoch": 0.0716,
      "grad_norm": 0.08025650680065155,
      "learning_rate": 0.002,
      "loss": 2.1633384227752686,
      "step": 716
    },
    {
      "epoch": 0.0718,
      "grad_norm": 0.07180091738700867,
      "learning_rate": 0.002,
      "loss": 2.2465360164642334,
      "step": 718
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.0749167799949646,
      "learning_rate": 0.002,
      "loss": 2.1551146507263184,
      "step": 720
    },
    {
      "epoch": 0.0722,
      "grad_norm": 0.07380841672420502,
      "learning_rate": 0.002,
      "loss": 2.130033016204834,
      "step": 722
    },
    {
      "epoch": 0.0724,
      "grad_norm": 0.07614101469516754,
      "learning_rate": 0.002,
      "loss": 2.284179210662842,
      "step": 724
    },
    {
      "epoch": 0.0726,
      "grad_norm": 0.07090556621551514,
      "learning_rate": 0.002,
      "loss": 2.136704921722412,
      "step": 726
    },
    {
      "epoch": 0.0728,
      "grad_norm": 0.07137584686279297,
      "learning_rate": 0.002,
      "loss": 2.137085437774658,
      "step": 728
    },
    {
      "epoch": 0.073,
      "grad_norm": 0.07333864271640778,
      "learning_rate": 0.002,
      "loss": 2.0440621376037598,
      "step": 730
    },
    {
      "epoch": 0.0732,
      "grad_norm": 0.07598543912172318,
      "learning_rate": 0.002,
      "loss": 2.2394542694091797,
      "step": 732
    },
    {
      "epoch": 0.0734,
      "grad_norm": 0.07136410474777222,
      "learning_rate": 0.002,
      "loss": 2.2622618675231934,
      "step": 734
    },
    {
      "epoch": 0.0736,
      "grad_norm": 0.0702420026063919,
      "learning_rate": 0.002,
      "loss": 2.274116277694702,
      "step": 736
    },
    {
      "epoch": 0.0738,
      "grad_norm": 0.08503563702106476,
      "learning_rate": 0.002,
      "loss": 2.1803767681121826,
      "step": 738
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.07403557747602463,
      "learning_rate": 0.002,
      "loss": 2.207209348678589,
      "step": 740
    },
    {
      "epoch": 0.0742,
      "grad_norm": 0.07537230849266052,
      "learning_rate": 0.002,
      "loss": 2.267427921295166,
      "step": 742
    },
    {
      "epoch": 0.0744,
      "grad_norm": 0.09657429158687592,
      "learning_rate": 0.002,
      "loss": 2.2643229961395264,
      "step": 744
    },
    {
      "epoch": 0.0746,
      "grad_norm": 0.07308726757764816,
      "learning_rate": 0.002,
      "loss": 2.270045042037964,
      "step": 746
    },
    {
      "epoch": 0.0748,
      "grad_norm": 0.06949540972709656,
      "learning_rate": 0.002,
      "loss": 2.2665822505950928,
      "step": 748
    },
    {
      "epoch": 0.075,
      "grad_norm": 0.06928335130214691,
      "learning_rate": 0.002,
      "loss": 2.2277493476867676,
      "step": 750
    },
    {
      "epoch": 0.0752,
      "grad_norm": 0.07039999216794968,
      "learning_rate": 0.002,
      "loss": 2.2953014373779297,
      "step": 752
    },
    {
      "epoch": 0.0754,
      "grad_norm": 0.07101280242204666,
      "learning_rate": 0.002,
      "loss": 2.276066303253174,
      "step": 754
    },
    {
      "epoch": 0.0756,
      "grad_norm": 0.0785365179181099,
      "learning_rate": 0.002,
      "loss": 2.1980574131011963,
      "step": 756
    },
    {
      "epoch": 0.0758,
      "grad_norm": 0.070271797478199,
      "learning_rate": 0.002,
      "loss": 2.2403957843780518,
      "step": 758
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.07686503976583481,
      "learning_rate": 0.002,
      "loss": 2.2867789268493652,
      "step": 760
    },
    {
      "epoch": 0.0762,
      "grad_norm": 0.07873029261827469,
      "learning_rate": 0.002,
      "loss": 2.319978713989258,
      "step": 762
    },
    {
      "epoch": 0.0764,
      "grad_norm": 0.07217779755592346,
      "learning_rate": 0.002,
      "loss": 2.102283000946045,
      "step": 764
    },
    {
      "epoch": 0.0766,
      "grad_norm": 0.0785188302397728,
      "learning_rate": 0.002,
      "loss": 2.2239789962768555,
      "step": 766
    },
    {
      "epoch": 0.0768,
      "grad_norm": 0.07748454809188843,
      "learning_rate": 0.002,
      "loss": 2.113046646118164,
      "step": 768
    },
    {
      "epoch": 0.077,
      "grad_norm": 0.07211107760667801,
      "learning_rate": 0.002,
      "loss": 2.2169735431671143,
      "step": 770
    },
    {
      "epoch": 0.0772,
      "grad_norm": 0.07038187235593796,
      "learning_rate": 0.002,
      "loss": 2.188551425933838,
      "step": 772
    },
    {
      "epoch": 0.0774,
      "grad_norm": 0.07528981566429138,
      "learning_rate": 0.002,
      "loss": 2.2348170280456543,
      "step": 774
    },
    {
      "epoch": 0.0776,
      "grad_norm": 0.07083573192358017,
      "learning_rate": 0.002,
      "loss": 2.240328550338745,
      "step": 776
    },
    {
      "epoch": 0.0778,
      "grad_norm": 0.08164046704769135,
      "learning_rate": 0.002,
      "loss": 2.211606740951538,
      "step": 778
    },
    {
      "epoch": 0.078,
      "grad_norm": 0.07621562480926514,
      "learning_rate": 0.002,
      "loss": 2.277646541595459,
      "step": 780
    },
    {
      "epoch": 0.0782,
      "grad_norm": 0.07361312955617905,
      "learning_rate": 0.002,
      "loss": 2.2900424003601074,
      "step": 782
    },
    {
      "epoch": 0.0784,
      "grad_norm": 0.07287199795246124,
      "learning_rate": 0.002,
      "loss": 2.232506036758423,
      "step": 784
    },
    {
      "epoch": 0.0786,
      "grad_norm": 0.07188855111598969,
      "learning_rate": 0.002,
      "loss": 2.185377836227417,
      "step": 786
    },
    {
      "epoch": 0.0788,
      "grad_norm": 0.0707741230726242,
      "learning_rate": 0.002,
      "loss": 2.1703908443450928,
      "step": 788
    },
    {
      "epoch": 0.079,
      "grad_norm": 0.07555016130208969,
      "learning_rate": 0.002,
      "loss": 2.140191078186035,
      "step": 790
    },
    {
      "epoch": 0.0792,
      "grad_norm": 0.07370918989181519,
      "learning_rate": 0.002,
      "loss": 2.0460965633392334,
      "step": 792
    },
    {
      "epoch": 0.0794,
      "grad_norm": 0.07398246973752975,
      "learning_rate": 0.002,
      "loss": 2.18871808052063,
      "step": 794
    },
    {
      "epoch": 0.0796,
      "grad_norm": 0.07762464880943298,
      "learning_rate": 0.002,
      "loss": 2.2658934593200684,
      "step": 796
    },
    {
      "epoch": 0.0798,
      "grad_norm": 0.07634468376636505,
      "learning_rate": 0.002,
      "loss": 2.2865793704986572,
      "step": 798
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.07308990508317947,
      "learning_rate": 0.002,
      "loss": 2.160325765609741,
      "step": 800
    },
    {
      "epoch": 0.0802,
      "grad_norm": 0.08453764021396637,
      "learning_rate": 0.002,
      "loss": 2.213151454925537,
      "step": 802
    },
    {
      "epoch": 0.0804,
      "grad_norm": 0.07261340320110321,
      "learning_rate": 0.002,
      "loss": 2.111349582672119,
      "step": 804
    },
    {
      "epoch": 0.0806,
      "grad_norm": 0.07345470786094666,
      "learning_rate": 0.002,
      "loss": 2.129185438156128,
      "step": 806
    },
    {
      "epoch": 0.0808,
      "grad_norm": 0.07541390508413315,
      "learning_rate": 0.002,
      "loss": 2.208940029144287,
      "step": 808
    },
    {
      "epoch": 0.081,
      "grad_norm": 0.07372540235519409,
      "learning_rate": 0.002,
      "loss": 2.222878932952881,
      "step": 810
    },
    {
      "epoch": 0.0812,
      "grad_norm": 0.06998228281736374,
      "learning_rate": 0.002,
      "loss": 2.226853132247925,
      "step": 812
    },
    {
      "epoch": 0.0814,
      "grad_norm": 0.07062869518995285,
      "learning_rate": 0.002,
      "loss": 2.1583786010742188,
      "step": 814
    },
    {
      "epoch": 0.0816,
      "grad_norm": 0.06911607831716537,
      "learning_rate": 0.002,
      "loss": 2.315582752227783,
      "step": 816
    },
    {
      "epoch": 0.0818,
      "grad_norm": 0.07161376625299454,
      "learning_rate": 0.002,
      "loss": 2.0797927379608154,
      "step": 818
    },
    {
      "epoch": 0.082,
      "grad_norm": 0.07752925902605057,
      "learning_rate": 0.002,
      "loss": 2.133472442626953,
      "step": 820
    },
    {
      "epoch": 0.0822,
      "grad_norm": 0.08120434731245041,
      "learning_rate": 0.002,
      "loss": 2.205599308013916,
      "step": 822
    },
    {
      "epoch": 0.0824,
      "grad_norm": 0.07119914144277573,
      "learning_rate": 0.002,
      "loss": 2.2620389461517334,
      "step": 824
    },
    {
      "epoch": 0.0826,
      "grad_norm": 0.07078907638788223,
      "learning_rate": 0.002,
      "loss": 2.1297965049743652,
      "step": 826
    },
    {
      "epoch": 0.0828,
      "grad_norm": 0.07739049941301346,
      "learning_rate": 0.002,
      "loss": 2.062345027923584,
      "step": 828
    },
    {
      "epoch": 0.083,
      "grad_norm": 0.07349462807178497,
      "learning_rate": 0.002,
      "loss": 2.082428216934204,
      "step": 830
    },
    {
      "epoch": 0.0832,
      "grad_norm": 0.07543961703777313,
      "learning_rate": 0.002,
      "loss": 2.1199893951416016,
      "step": 832
    },
    {
      "epoch": 0.0834,
      "grad_norm": 0.07260638475418091,
      "learning_rate": 0.002,
      "loss": 2.2648003101348877,
      "step": 834
    },
    {
      "epoch": 0.0836,
      "grad_norm": 0.07462988048791885,
      "learning_rate": 0.002,
      "loss": 2.183993101119995,
      "step": 836
    },
    {
      "epoch": 0.0838,
      "grad_norm": 0.07780813425779343,
      "learning_rate": 0.002,
      "loss": 2.2332801818847656,
      "step": 838
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.07524517178535461,
      "learning_rate": 0.002,
      "loss": 2.157496690750122,
      "step": 840
    },
    {
      "epoch": 0.0842,
      "grad_norm": 0.07251094281673431,
      "learning_rate": 0.002,
      "loss": 2.0967729091644287,
      "step": 842
    },
    {
      "epoch": 0.0844,
      "grad_norm": 0.07608260214328766,
      "learning_rate": 0.002,
      "loss": 2.250211715698242,
      "step": 844
    },
    {
      "epoch": 0.0846,
      "grad_norm": 0.0735645443201065,
      "learning_rate": 0.002,
      "loss": 2.174501895904541,
      "step": 846
    },
    {
      "epoch": 0.0848,
      "grad_norm": 0.07014891505241394,
      "learning_rate": 0.002,
      "loss": 2.244321584701538,
      "step": 848
    },
    {
      "epoch": 0.085,
      "grad_norm": 0.077987901866436,
      "learning_rate": 0.002,
      "loss": 2.0997698307037354,
      "step": 850
    },
    {
      "epoch": 0.0852,
      "grad_norm": 0.07205221056938171,
      "learning_rate": 0.002,
      "loss": 2.1326282024383545,
      "step": 852
    },
    {
      "epoch": 0.0854,
      "grad_norm": 0.07838629186153412,
      "learning_rate": 0.002,
      "loss": 2.166931390762329,
      "step": 854
    },
    {
      "epoch": 0.0856,
      "grad_norm": 0.07645858079195023,
      "learning_rate": 0.002,
      "loss": 2.1296794414520264,
      "step": 856
    },
    {
      "epoch": 0.0858,
      "grad_norm": 0.07235304266214371,
      "learning_rate": 0.002,
      "loss": 2.2111241817474365,
      "step": 858
    },
    {
      "epoch": 0.086,
      "grad_norm": 0.0837651863694191,
      "learning_rate": 0.002,
      "loss": 2.167710065841675,
      "step": 860
    },
    {
      "epoch": 0.0862,
      "grad_norm": 0.07638353854417801,
      "learning_rate": 0.002,
      "loss": 2.118605613708496,
      "step": 862
    },
    {
      "epoch": 0.0864,
      "grad_norm": 0.07286670804023743,
      "learning_rate": 0.002,
      "loss": 2.2549030780792236,
      "step": 864
    },
    {
      "epoch": 0.0866,
      "grad_norm": 0.07611484825611115,
      "learning_rate": 0.002,
      "loss": 2.2853569984436035,
      "step": 866
    },
    {
      "epoch": 0.0868,
      "grad_norm": 0.07528790086507797,
      "learning_rate": 0.002,
      "loss": 2.190622091293335,
      "step": 868
    },
    {
      "epoch": 0.087,
      "grad_norm": 0.07928916811943054,
      "learning_rate": 0.002,
      "loss": 2.224142074584961,
      "step": 870
    },
    {
      "epoch": 0.0872,
      "grad_norm": 0.07761674374341965,
      "learning_rate": 0.002,
      "loss": 2.1696512699127197,
      "step": 872
    },
    {
      "epoch": 0.0874,
      "grad_norm": 0.08167286962270737,
      "learning_rate": 0.002,
      "loss": 2.151870012283325,
      "step": 874
    },
    {
      "epoch": 0.0876,
      "grad_norm": 0.07754576951265335,
      "learning_rate": 0.002,
      "loss": 2.22660756111145,
      "step": 876
    },
    {
      "epoch": 0.0878,
      "grad_norm": 0.07292933762073517,
      "learning_rate": 0.002,
      "loss": 2.168794870376587,
      "step": 878
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.07828113436698914,
      "learning_rate": 0.002,
      "loss": 2.224976062774658,
      "step": 880
    },
    {
      "epoch": 0.0882,
      "grad_norm": 0.0726737380027771,
      "learning_rate": 0.002,
      "loss": 2.2155354022979736,
      "step": 882
    },
    {
      "epoch": 0.0884,
      "grad_norm": 0.08198512345552444,
      "learning_rate": 0.002,
      "loss": 2.095844030380249,
      "step": 884
    },
    {
      "epoch": 0.0886,
      "grad_norm": 0.07030656188726425,
      "learning_rate": 0.002,
      "loss": 2.1966006755828857,
      "step": 886
    },
    {
      "epoch": 0.0888,
      "grad_norm": 0.0729076936841011,
      "learning_rate": 0.002,
      "loss": 2.163865566253662,
      "step": 888
    },
    {
      "epoch": 0.089,
      "grad_norm": 0.07960838824510574,
      "learning_rate": 0.002,
      "loss": 2.1911630630493164,
      "step": 890
    },
    {
      "epoch": 0.0892,
      "grad_norm": 0.07703646272420883,
      "learning_rate": 0.002,
      "loss": 2.1365089416503906,
      "step": 892
    },
    {
      "epoch": 0.0894,
      "grad_norm": 0.07154688239097595,
      "learning_rate": 0.002,
      "loss": 2.230971574783325,
      "step": 894
    },
    {
      "epoch": 0.0896,
      "grad_norm": 0.07950998842716217,
      "learning_rate": 0.002,
      "loss": 2.1386072635650635,
      "step": 896
    },
    {
      "epoch": 0.0898,
      "grad_norm": 0.07266457378864288,
      "learning_rate": 0.002,
      "loss": 2.123721122741699,
      "step": 898
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.07779520004987717,
      "learning_rate": 0.002,
      "loss": 2.1806583404541016,
      "step": 900
    },
    {
      "epoch": 0.0902,
      "grad_norm": 0.07444702833890915,
      "learning_rate": 0.002,
      "loss": 2.099088668823242,
      "step": 902
    },
    {
      "epoch": 0.0904,
      "grad_norm": 0.07581638544797897,
      "learning_rate": 0.002,
      "loss": 2.176795482635498,
      "step": 904
    },
    {
      "epoch": 0.0906,
      "grad_norm": 0.07643268257379532,
      "learning_rate": 0.002,
      "loss": 2.1350550651550293,
      "step": 906
    },
    {
      "epoch": 0.0908,
      "grad_norm": 0.07969532161951065,
      "learning_rate": 0.002,
      "loss": 2.057523012161255,
      "step": 908
    },
    {
      "epoch": 0.091,
      "grad_norm": 0.07287921011447906,
      "learning_rate": 0.002,
      "loss": 1.9980111122131348,
      "step": 910
    },
    {
      "epoch": 0.0912,
      "grad_norm": 0.07304186373949051,
      "learning_rate": 0.002,
      "loss": 2.2014057636260986,
      "step": 912
    },
    {
      "epoch": 0.0914,
      "grad_norm": 0.07331191748380661,
      "learning_rate": 0.002,
      "loss": 2.2062792778015137,
      "step": 914
    },
    {
      "epoch": 0.0916,
      "grad_norm": 0.07019214332103729,
      "learning_rate": 0.002,
      "loss": 2.037095785140991,
      "step": 916
    },
    {
      "epoch": 0.0918,
      "grad_norm": 0.07438710331916809,
      "learning_rate": 0.002,
      "loss": 2.2597415447235107,
      "step": 918
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.07565884292125702,
      "learning_rate": 0.002,
      "loss": 2.2872583866119385,
      "step": 920
    },
    {
      "epoch": 0.0922,
      "grad_norm": 0.0788540169596672,
      "learning_rate": 0.002,
      "loss": 2.1922593116760254,
      "step": 922
    },
    {
      "epoch": 0.0924,
      "grad_norm": 0.08020783960819244,
      "learning_rate": 0.002,
      "loss": 2.1460044384002686,
      "step": 924
    },
    {
      "epoch": 0.0926,
      "grad_norm": 0.07896710187196732,
      "learning_rate": 0.002,
      "loss": 1.9879710674285889,
      "step": 926
    },
    {
      "epoch": 0.0928,
      "grad_norm": 0.0781594067811966,
      "learning_rate": 0.002,
      "loss": 2.0701892375946045,
      "step": 928
    },
    {
      "epoch": 0.093,
      "grad_norm": 0.07467583566904068,
      "learning_rate": 0.002,
      "loss": 2.145289897918701,
      "step": 930
    },
    {
      "epoch": 0.0932,
      "grad_norm": 0.07271774113178253,
      "learning_rate": 0.002,
      "loss": 2.250311851501465,
      "step": 932
    },
    {
      "epoch": 0.0934,
      "grad_norm": 0.07633378356695175,
      "learning_rate": 0.002,
      "loss": 2.136322021484375,
      "step": 934
    },
    {
      "epoch": 0.0936,
      "grad_norm": 0.07973302155733109,
      "learning_rate": 0.002,
      "loss": 2.222205877304077,
      "step": 936
    },
    {
      "epoch": 0.0938,
      "grad_norm": 0.0762925073504448,
      "learning_rate": 0.002,
      "loss": 2.041592597961426,
      "step": 938
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.07892216742038727,
      "learning_rate": 0.002,
      "loss": 2.044208288192749,
      "step": 940
    },
    {
      "epoch": 0.0942,
      "grad_norm": 0.07398161292076111,
      "learning_rate": 0.002,
      "loss": 2.0489468574523926,
      "step": 942
    },
    {
      "epoch": 0.0944,
      "grad_norm": 0.08096542209386826,
      "learning_rate": 0.002,
      "loss": 2.231882095336914,
      "step": 944
    },
    {
      "epoch": 0.0946,
      "grad_norm": 0.07238005846738815,
      "learning_rate": 0.002,
      "loss": 2.114208459854126,
      "step": 946
    },
    {
      "epoch": 0.0948,
      "grad_norm": 0.0702667385339737,
      "learning_rate": 0.002,
      "loss": 2.1004698276519775,
      "step": 948
    },
    {
      "epoch": 0.095,
      "grad_norm": 0.06968449801206589,
      "learning_rate": 0.002,
      "loss": 2.1844141483306885,
      "step": 950
    },
    {
      "epoch": 0.0952,
      "grad_norm": 0.07529022544622421,
      "learning_rate": 0.002,
      "loss": 2.161083221435547,
      "step": 952
    },
    {
      "epoch": 0.0954,
      "grad_norm": 0.07257848232984543,
      "learning_rate": 0.002,
      "loss": 2.1482396125793457,
      "step": 954
    },
    {
      "epoch": 0.0956,
      "grad_norm": 0.07533086091279984,
      "learning_rate": 0.002,
      "loss": 2.1445465087890625,
      "step": 956
    },
    {
      "epoch": 0.0958,
      "grad_norm": 0.0761774331331253,
      "learning_rate": 0.002,
      "loss": 2.1782426834106445,
      "step": 958
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.07364509999752045,
      "learning_rate": 0.002,
      "loss": 2.0680530071258545,
      "step": 960
    },
    {
      "epoch": 0.0962,
      "grad_norm": 0.079878069460392,
      "learning_rate": 0.002,
      "loss": 2.237175464630127,
      "step": 962
    },
    {
      "epoch": 0.0964,
      "grad_norm": 0.06845875829458237,
      "learning_rate": 0.002,
      "loss": 2.185602903366089,
      "step": 964
    },
    {
      "epoch": 0.0966,
      "grad_norm": 0.07009263336658478,
      "learning_rate": 0.002,
      "loss": 2.212388753890991,
      "step": 966
    },
    {
      "epoch": 0.0968,
      "grad_norm": 0.07105608284473419,
      "learning_rate": 0.002,
      "loss": 2.1743319034576416,
      "step": 968
    },
    {
      "epoch": 0.097,
      "grad_norm": 0.07293667644262314,
      "learning_rate": 0.002,
      "loss": 2.1729116439819336,
      "step": 970
    },
    {
      "epoch": 0.0972,
      "grad_norm": 0.07323897629976273,
      "learning_rate": 0.002,
      "loss": 2.1292495727539062,
      "step": 972
    },
    {
      "epoch": 0.0974,
      "grad_norm": 0.09951165318489075,
      "learning_rate": 0.002,
      "loss": 2.0387752056121826,
      "step": 974
    },
    {
      "epoch": 0.0976,
      "grad_norm": 0.07220254838466644,
      "learning_rate": 0.002,
      "loss": 2.0899648666381836,
      "step": 976
    },
    {
      "epoch": 0.0978,
      "grad_norm": 0.07627923786640167,
      "learning_rate": 0.002,
      "loss": 2.1366374492645264,
      "step": 978
    },
    {
      "epoch": 0.098,
      "grad_norm": 0.07620730251073837,
      "learning_rate": 0.002,
      "loss": 2.1590631008148193,
      "step": 980
    },
    {
      "epoch": 0.0982,
      "grad_norm": 0.07129782438278198,
      "learning_rate": 0.002,
      "loss": 2.1353065967559814,
      "step": 982
    },
    {
      "epoch": 0.0984,
      "grad_norm": 0.07142531871795654,
      "learning_rate": 0.002,
      "loss": 2.074662446975708,
      "step": 984
    },
    {
      "epoch": 0.0986,
      "grad_norm": 0.08024367690086365,
      "learning_rate": 0.002,
      "loss": 2.214207887649536,
      "step": 986
    },
    {
      "epoch": 0.0988,
      "grad_norm": 0.0746978297829628,
      "learning_rate": 0.002,
      "loss": 2.099921226501465,
      "step": 988
    },
    {
      "epoch": 0.099,
      "grad_norm": 0.07755237817764282,
      "learning_rate": 0.002,
      "loss": 2.260185956954956,
      "step": 990
    },
    {
      "epoch": 0.0992,
      "grad_norm": 0.07463093101978302,
      "learning_rate": 0.002,
      "loss": 2.1920037269592285,
      "step": 992
    },
    {
      "epoch": 0.0994,
      "grad_norm": 0.07097990810871124,
      "learning_rate": 0.002,
      "loss": 2.239779472351074,
      "step": 994
    },
    {
      "epoch": 0.0996,
      "grad_norm": 0.07190890610218048,
      "learning_rate": 0.002,
      "loss": 2.178900718688965,
      "step": 996
    },
    {
      "epoch": 0.0998,
      "grad_norm": 0.07009754329919815,
      "learning_rate": 0.002,
      "loss": 2.1559388637542725,
      "step": 998
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.07546056807041168,
      "learning_rate": 0.002,
      "loss": 2.1190061569213867,
      "step": 1000
    },
    {
      "epoch": 0.1002,
      "grad_norm": 0.0721445381641388,
      "learning_rate": 0.002,
      "loss": 2.172283172607422,
      "step": 1002
    },
    {
      "epoch": 0.1004,
      "grad_norm": 0.07409537583589554,
      "learning_rate": 0.002,
      "loss": 2.03796648979187,
      "step": 1004
    },
    {
      "epoch": 0.1006,
      "grad_norm": 0.07212292402982712,
      "learning_rate": 0.002,
      "loss": 2.1891837120056152,
      "step": 1006
    },
    {
      "epoch": 0.1008,
      "grad_norm": 0.07398109883069992,
      "learning_rate": 0.002,
      "loss": 2.336242914199829,
      "step": 1008
    },
    {
      "epoch": 0.101,
      "grad_norm": 0.07375752925872803,
      "learning_rate": 0.002,
      "loss": 2.1167736053466797,
      "step": 1010
    },
    {
      "epoch": 0.1012,
      "grad_norm": 0.07798315584659576,
      "learning_rate": 0.002,
      "loss": 2.2498393058776855,
      "step": 1012
    },
    {
      "epoch": 0.1014,
      "grad_norm": 0.07301631569862366,
      "learning_rate": 0.002,
      "loss": 2.134582042694092,
      "step": 1014
    },
    {
      "epoch": 0.1016,
      "grad_norm": 0.07903663069009781,
      "learning_rate": 0.002,
      "loss": 2.1524388790130615,
      "step": 1016
    },
    {
      "epoch": 0.1018,
      "grad_norm": 0.0734393298625946,
      "learning_rate": 0.002,
      "loss": 2.2745954990386963,
      "step": 1018
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.07571080327033997,
      "learning_rate": 0.002,
      "loss": 2.1404924392700195,
      "step": 1020
    },
    {
      "epoch": 0.1022,
      "grad_norm": 0.07368601113557816,
      "learning_rate": 0.002,
      "loss": 2.110748529434204,
      "step": 1022
    },
    {
      "epoch": 0.1024,
      "grad_norm": 0.06855492293834686,
      "learning_rate": 0.002,
      "loss": 2.2289109230041504,
      "step": 1024
    },
    {
      "epoch": 0.1026,
      "grad_norm": 0.07603514194488525,
      "learning_rate": 0.002,
      "loss": 2.1090359687805176,
      "step": 1026
    },
    {
      "epoch": 0.1028,
      "grad_norm": 0.0808582529425621,
      "learning_rate": 0.002,
      "loss": 2.0453224182128906,
      "step": 1028
    },
    {
      "epoch": 0.103,
      "grad_norm": 0.07189541310071945,
      "learning_rate": 0.002,
      "loss": 2.174919605255127,
      "step": 1030
    },
    {
      "epoch": 0.1032,
      "grad_norm": 0.0745692327618599,
      "learning_rate": 0.002,
      "loss": 2.2082841396331787,
      "step": 1032
    },
    {
      "epoch": 0.1034,
      "grad_norm": 0.07547178119421005,
      "learning_rate": 0.002,
      "loss": 2.1058688163757324,
      "step": 1034
    },
    {
      "epoch": 0.1036,
      "grad_norm": 0.07315513491630554,
      "learning_rate": 0.002,
      "loss": 2.2061972618103027,
      "step": 1036
    },
    {
      "epoch": 0.1038,
      "grad_norm": 0.0768812745809555,
      "learning_rate": 0.002,
      "loss": 2.2342112064361572,
      "step": 1038
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.07388103008270264,
      "learning_rate": 0.002,
      "loss": 2.134312629699707,
      "step": 1040
    },
    {
      "epoch": 0.1042,
      "grad_norm": 0.0759049579501152,
      "learning_rate": 0.002,
      "loss": 2.2925689220428467,
      "step": 1042
    },
    {
      "epoch": 0.1044,
      "grad_norm": 0.07359053194522858,
      "learning_rate": 0.002,
      "loss": 2.0979182720184326,
      "step": 1044
    },
    {
      "epoch": 0.1046,
      "grad_norm": 0.0745314359664917,
      "learning_rate": 0.002,
      "loss": 2.1159324645996094,
      "step": 1046
    },
    {
      "epoch": 0.1048,
      "grad_norm": 0.0748969167470932,
      "learning_rate": 0.002,
      "loss": 2.2266550064086914,
      "step": 1048
    },
    {
      "epoch": 0.105,
      "grad_norm": 0.0749678984284401,
      "learning_rate": 0.002,
      "loss": 2.1067774295806885,
      "step": 1050
    },
    {
      "epoch": 0.1052,
      "grad_norm": 0.06975509971380234,
      "learning_rate": 0.002,
      "loss": 2.149784803390503,
      "step": 1052
    },
    {
      "epoch": 0.1054,
      "grad_norm": 0.07229141145944595,
      "learning_rate": 0.002,
      "loss": 2.2860937118530273,
      "step": 1054
    },
    {
      "epoch": 0.1056,
      "grad_norm": 0.07149404287338257,
      "learning_rate": 0.002,
      "loss": 2.3322577476501465,
      "step": 1056
    },
    {
      "epoch": 0.1058,
      "grad_norm": 0.0729236975312233,
      "learning_rate": 0.002,
      "loss": 2.0751750469207764,
      "step": 1058
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.07678281515836716,
      "learning_rate": 0.002,
      "loss": 2.12359619140625,
      "step": 1060
    },
    {
      "epoch": 0.1062,
      "grad_norm": 0.07419778406620026,
      "learning_rate": 0.002,
      "loss": 2.043370485305786,
      "step": 1062
    },
    {
      "epoch": 0.1064,
      "grad_norm": 0.06995157152414322,
      "learning_rate": 0.002,
      "loss": 2.1271049976348877,
      "step": 1064
    },
    {
      "epoch": 0.1066,
      "grad_norm": 0.0794706717133522,
      "learning_rate": 0.002,
      "loss": 2.2551066875457764,
      "step": 1066
    },
    {
      "epoch": 0.1068,
      "grad_norm": 0.0730832889676094,
      "learning_rate": 0.002,
      "loss": 2.0715439319610596,
      "step": 1068
    },
    {
      "epoch": 0.107,
      "grad_norm": 0.07194572687149048,
      "learning_rate": 0.002,
      "loss": 2.192267894744873,
      "step": 1070
    },
    {
      "epoch": 0.1072,
      "grad_norm": 0.07492946833372116,
      "learning_rate": 0.002,
      "loss": 2.198681116104126,
      "step": 1072
    },
    {
      "epoch": 0.1074,
      "grad_norm": 0.07487789541482925,
      "learning_rate": 0.002,
      "loss": 2.2635657787323,
      "step": 1074
    },
    {
      "epoch": 0.1076,
      "grad_norm": 0.07483327388763428,
      "learning_rate": 0.002,
      "loss": 2.0767524242401123,
      "step": 1076
    },
    {
      "epoch": 0.1078,
      "grad_norm": 0.07829601317644119,
      "learning_rate": 0.002,
      "loss": 1.9860057830810547,
      "step": 1078
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.0763917788863182,
      "learning_rate": 0.002,
      "loss": 2.141301155090332,
      "step": 1080
    },
    {
      "epoch": 0.1082,
      "grad_norm": 0.07612070441246033,
      "learning_rate": 0.002,
      "loss": 2.119746685028076,
      "step": 1082
    },
    {
      "epoch": 0.1084,
      "grad_norm": 0.0760447308421135,
      "learning_rate": 0.002,
      "loss": 2.20205020904541,
      "step": 1084
    },
    {
      "epoch": 0.1086,
      "grad_norm": 0.07326960563659668,
      "learning_rate": 0.002,
      "loss": 2.1251564025878906,
      "step": 1086
    },
    {
      "epoch": 0.1088,
      "grad_norm": 0.07328005135059357,
      "learning_rate": 0.002,
      "loss": 2.2636373043060303,
      "step": 1088
    },
    {
      "epoch": 0.109,
      "grad_norm": 0.07502838224172592,
      "learning_rate": 0.002,
      "loss": 2.1950507164001465,
      "step": 1090
    },
    {
      "epoch": 0.1092,
      "grad_norm": 0.0707181990146637,
      "learning_rate": 0.002,
      "loss": 2.144662618637085,
      "step": 1092
    },
    {
      "epoch": 0.1094,
      "grad_norm": 0.07617582380771637,
      "learning_rate": 0.002,
      "loss": 2.1324892044067383,
      "step": 1094
    },
    {
      "epoch": 0.1096,
      "grad_norm": 0.07278620451688766,
      "learning_rate": 0.002,
      "loss": 2.082261800765991,
      "step": 1096
    },
    {
      "epoch": 0.1098,
      "grad_norm": 0.07565721869468689,
      "learning_rate": 0.002,
      "loss": 2.1475939750671387,
      "step": 1098
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.08005080372095108,
      "learning_rate": 0.002,
      "loss": 2.1007742881774902,
      "step": 1100
    },
    {
      "epoch": 0.1102,
      "grad_norm": 0.07498965412378311,
      "learning_rate": 0.002,
      "loss": 2.073636531829834,
      "step": 1102
    },
    {
      "epoch": 0.1104,
      "grad_norm": 0.07267436385154724,
      "learning_rate": 0.002,
      "loss": 2.178069591522217,
      "step": 1104
    },
    {
      "epoch": 0.1106,
      "grad_norm": 0.07365086674690247,
      "learning_rate": 0.002,
      "loss": 2.066964626312256,
      "step": 1106
    },
    {
      "epoch": 0.1108,
      "grad_norm": 0.07104740291833878,
      "learning_rate": 0.002,
      "loss": 2.274270534515381,
      "step": 1108
    },
    {
      "epoch": 0.111,
      "grad_norm": 0.07345458120107651,
      "learning_rate": 0.002,
      "loss": 2.277094841003418,
      "step": 1110
    },
    {
      "epoch": 0.1112,
      "grad_norm": 0.07415177673101425,
      "learning_rate": 0.002,
      "loss": 2.1100540161132812,
      "step": 1112
    },
    {
      "epoch": 0.1114,
      "grad_norm": 0.07272006571292877,
      "learning_rate": 0.002,
      "loss": 1.9805363416671753,
      "step": 1114
    },
    {
      "epoch": 0.1116,
      "grad_norm": 0.07305069267749786,
      "learning_rate": 0.002,
      "loss": 2.0581307411193848,
      "step": 1116
    },
    {
      "epoch": 0.1118,
      "grad_norm": 0.07427258044481277,
      "learning_rate": 0.002,
      "loss": 2.1613399982452393,
      "step": 1118
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.07428186386823654,
      "learning_rate": 0.002,
      "loss": 2.3126909732818604,
      "step": 1120
    },
    {
      "epoch": 0.1122,
      "grad_norm": 0.0752764642238617,
      "learning_rate": 0.002,
      "loss": 2.004159927368164,
      "step": 1122
    },
    {
      "epoch": 0.1124,
      "grad_norm": 0.08465594053268433,
      "learning_rate": 0.002,
      "loss": 2.199995756149292,
      "step": 1124
    },
    {
      "epoch": 0.1126,
      "grad_norm": 0.07358529418706894,
      "learning_rate": 0.002,
      "loss": 1.9912794828414917,
      "step": 1126
    },
    {
      "epoch": 0.1128,
      "grad_norm": 0.0733361691236496,
      "learning_rate": 0.002,
      "loss": 2.0023438930511475,
      "step": 1128
    },
    {
      "epoch": 0.113,
      "grad_norm": 0.08205096423625946,
      "learning_rate": 0.002,
      "loss": 2.1227903366088867,
      "step": 1130
    },
    {
      "epoch": 0.1132,
      "grad_norm": 0.06866461783647537,
      "learning_rate": 0.002,
      "loss": 2.139390230178833,
      "step": 1132
    },
    {
      "epoch": 0.1134,
      "grad_norm": 0.0740932896733284,
      "learning_rate": 0.002,
      "loss": 2.272157669067383,
      "step": 1134
    },
    {
      "epoch": 0.1136,
      "grad_norm": 0.08060478419065475,
      "learning_rate": 0.002,
      "loss": 2.099661111831665,
      "step": 1136
    },
    {
      "epoch": 0.1138,
      "grad_norm": 0.07331594079732895,
      "learning_rate": 0.002,
      "loss": 2.2203280925750732,
      "step": 1138
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.07731856405735016,
      "learning_rate": 0.002,
      "loss": 2.125272035598755,
      "step": 1140
    },
    {
      "epoch": 0.1142,
      "grad_norm": 0.07412073761224747,
      "learning_rate": 0.002,
      "loss": 2.1330039501190186,
      "step": 1142
    },
    {
      "epoch": 0.1144,
      "grad_norm": 0.07352765649557114,
      "learning_rate": 0.002,
      "loss": 2.164257287979126,
      "step": 1144
    },
    {
      "epoch": 0.1146,
      "grad_norm": 0.07029831409454346,
      "learning_rate": 0.002,
      "loss": 2.2174627780914307,
      "step": 1146
    },
    {
      "epoch": 0.1148,
      "grad_norm": 0.07499506324529648,
      "learning_rate": 0.002,
      "loss": 2.2317168712615967,
      "step": 1148
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.07939469814300537,
      "learning_rate": 0.002,
      "loss": 2.2263972759246826,
      "step": 1150
    },
    {
      "epoch": 0.1152,
      "grad_norm": 0.0757032111287117,
      "learning_rate": 0.002,
      "loss": 2.0671920776367188,
      "step": 1152
    },
    {
      "epoch": 0.1154,
      "grad_norm": 0.0747682973742485,
      "learning_rate": 0.002,
      "loss": 2.2007083892822266,
      "step": 1154
    },
    {
      "epoch": 0.1156,
      "grad_norm": 0.07464288175106049,
      "learning_rate": 0.002,
      "loss": 2.1639177799224854,
      "step": 1156
    },
    {
      "epoch": 0.1158,
      "grad_norm": 0.07429301738739014,
      "learning_rate": 0.002,
      "loss": 1.967406153678894,
      "step": 1158
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.0751800537109375,
      "learning_rate": 0.002,
      "loss": 2.04834246635437,
      "step": 1160
    },
    {
      "epoch": 0.1162,
      "grad_norm": 0.08047569543123245,
      "learning_rate": 0.002,
      "loss": 2.1337246894836426,
      "step": 1162
    },
    {
      "epoch": 0.1164,
      "grad_norm": 0.07371657341718674,
      "learning_rate": 0.002,
      "loss": 2.017942428588867,
      "step": 1164
    },
    {
      "epoch": 0.1166,
      "grad_norm": 0.07549389451742172,
      "learning_rate": 0.002,
      "loss": 2.113239049911499,
      "step": 1166
    },
    {
      "epoch": 0.1168,
      "grad_norm": 0.07099882513284683,
      "learning_rate": 0.002,
      "loss": 1.96674382686615,
      "step": 1168
    },
    {
      "epoch": 0.117,
      "grad_norm": 0.069034643471241,
      "learning_rate": 0.002,
      "loss": 2.0233230590820312,
      "step": 1170
    },
    {
      "epoch": 0.1172,
      "grad_norm": 0.07497158646583557,
      "learning_rate": 0.002,
      "loss": 1.9830009937286377,
      "step": 1172
    },
    {
      "epoch": 0.1174,
      "grad_norm": 0.07264664024114609,
      "learning_rate": 0.002,
      "loss": 2.2477784156799316,
      "step": 1174
    },
    {
      "epoch": 0.1176,
      "grad_norm": 0.07496596872806549,
      "learning_rate": 0.002,
      "loss": 2.1375670433044434,
      "step": 1176
    },
    {
      "epoch": 0.1178,
      "grad_norm": 0.07663583010435104,
      "learning_rate": 0.002,
      "loss": 2.1026036739349365,
      "step": 1178
    },
    {
      "epoch": 0.118,
      "grad_norm": 0.07101371139287949,
      "learning_rate": 0.002,
      "loss": 2.0782814025878906,
      "step": 1180
    },
    {
      "epoch": 0.1182,
      "grad_norm": 0.07188084721565247,
      "learning_rate": 0.002,
      "loss": 2.103645086288452,
      "step": 1182
    },
    {
      "epoch": 0.1184,
      "grad_norm": 0.0709480345249176,
      "learning_rate": 0.002,
      "loss": 2.153522253036499,
      "step": 1184
    },
    {
      "epoch": 0.1186,
      "grad_norm": 0.07403906434774399,
      "learning_rate": 0.002,
      "loss": 2.1187493801116943,
      "step": 1186
    },
    {
      "epoch": 0.1188,
      "grad_norm": 0.07533378899097443,
      "learning_rate": 0.002,
      "loss": 2.139988660812378,
      "step": 1188
    },
    {
      "epoch": 0.119,
      "grad_norm": 0.0702051892876625,
      "learning_rate": 0.002,
      "loss": 2.0612099170684814,
      "step": 1190
    },
    {
      "epoch": 0.1192,
      "grad_norm": 0.07475084066390991,
      "learning_rate": 0.002,
      "loss": 2.0514333248138428,
      "step": 1192
    },
    {
      "epoch": 0.1194,
      "grad_norm": 0.0737379640340805,
      "learning_rate": 0.002,
      "loss": 2.1701278686523438,
      "step": 1194
    },
    {
      "epoch": 0.1196,
      "grad_norm": 0.08116551488637924,
      "learning_rate": 0.002,
      "loss": 2.158406972885132,
      "step": 1196
    },
    {
      "epoch": 0.1198,
      "grad_norm": 0.07483541965484619,
      "learning_rate": 0.002,
      "loss": 2.1717021465301514,
      "step": 1198
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.07554017007350922,
      "learning_rate": 0.002,
      "loss": 2.086056709289551,
      "step": 1200
    },
    {
      "epoch": 0.1202,
      "grad_norm": 0.0791301503777504,
      "learning_rate": 0.002,
      "loss": 2.0326573848724365,
      "step": 1202
    },
    {
      "epoch": 0.1204,
      "grad_norm": 0.0760003849864006,
      "learning_rate": 0.002,
      "loss": 2.1414124965667725,
      "step": 1204
    },
    {
      "epoch": 0.1206,
      "grad_norm": 0.07325771450996399,
      "learning_rate": 0.002,
      "loss": 2.0560693740844727,
      "step": 1206
    },
    {
      "epoch": 0.1208,
      "grad_norm": 0.07368165254592896,
      "learning_rate": 0.002,
      "loss": 2.194891929626465,
      "step": 1208
    },
    {
      "epoch": 0.121,
      "grad_norm": 0.0809967964887619,
      "learning_rate": 0.002,
      "loss": 2.075925350189209,
      "step": 1210
    },
    {
      "epoch": 0.1212,
      "grad_norm": 0.07469695061445236,
      "learning_rate": 0.002,
      "loss": 2.1630022525787354,
      "step": 1212
    },
    {
      "epoch": 0.1214,
      "grad_norm": 0.07636883854866028,
      "learning_rate": 0.002,
      "loss": 2.142662525177002,
      "step": 1214
    },
    {
      "epoch": 0.1216,
      "grad_norm": 0.0737556740641594,
      "learning_rate": 0.002,
      "loss": 2.1493847370147705,
      "step": 1216
    },
    {
      "epoch": 0.1218,
      "grad_norm": 0.08186283707618713,
      "learning_rate": 0.002,
      "loss": 2.0885581970214844,
      "step": 1218
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.07902347296476364,
      "learning_rate": 0.002,
      "loss": 2.1004858016967773,
      "step": 1220
    },
    {
      "epoch": 0.1222,
      "grad_norm": 0.07315637171268463,
      "learning_rate": 0.002,
      "loss": 2.2506372928619385,
      "step": 1222
    },
    {
      "epoch": 0.1224,
      "grad_norm": 0.07557561248540878,
      "learning_rate": 0.002,
      "loss": 2.046790838241577,
      "step": 1224
    },
    {
      "epoch": 0.1226,
      "grad_norm": 0.07439936697483063,
      "learning_rate": 0.002,
      "loss": 2.1162328720092773,
      "step": 1226
    },
    {
      "epoch": 0.1228,
      "grad_norm": 0.0746796578168869,
      "learning_rate": 0.002,
      "loss": 2.0996177196502686,
      "step": 1228
    },
    {
      "epoch": 0.123,
      "grad_norm": 0.07287529855966568,
      "learning_rate": 0.002,
      "loss": 1.9575451612472534,
      "step": 1230
    },
    {
      "epoch": 0.1232,
      "grad_norm": 0.07638369500637054,
      "learning_rate": 0.002,
      "loss": 2.2129688262939453,
      "step": 1232
    },
    {
      "epoch": 0.1234,
      "grad_norm": 0.07679726183414459,
      "learning_rate": 0.002,
      "loss": 2.1270339488983154,
      "step": 1234
    },
    {
      "epoch": 0.1236,
      "grad_norm": 0.0753592848777771,
      "learning_rate": 0.002,
      "loss": 2.2584049701690674,
      "step": 1236
    },
    {
      "epoch": 0.1238,
      "grad_norm": 0.07483881711959839,
      "learning_rate": 0.002,
      "loss": 2.0660622119903564,
      "step": 1238
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.07298348844051361,
      "learning_rate": 0.002,
      "loss": 2.238267183303833,
      "step": 1240
    },
    {
      "epoch": 0.1242,
      "grad_norm": 0.07839833199977875,
      "learning_rate": 0.002,
      "loss": 2.10762095451355,
      "step": 1242
    },
    {
      "epoch": 0.1244,
      "grad_norm": 0.07315675169229507,
      "learning_rate": 0.002,
      "loss": 2.087207317352295,
      "step": 1244
    },
    {
      "epoch": 0.1246,
      "grad_norm": 0.07588371634483337,
      "learning_rate": 0.002,
      "loss": 2.0736515522003174,
      "step": 1246
    },
    {
      "epoch": 0.1248,
      "grad_norm": 0.07554105669260025,
      "learning_rate": 0.002,
      "loss": 2.188538074493408,
      "step": 1248
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.07133322209119797,
      "learning_rate": 0.002,
      "loss": 2.1506593227386475,
      "step": 1250
    },
    {
      "epoch": 0.1252,
      "grad_norm": 0.07444269210100174,
      "learning_rate": 0.002,
      "loss": 2.207246780395508,
      "step": 1252
    },
    {
      "epoch": 0.1254,
      "grad_norm": 0.07421612739562988,
      "learning_rate": 0.002,
      "loss": 2.0821216106414795,
      "step": 1254
    },
    {
      "epoch": 0.1256,
      "grad_norm": 0.07575299590826035,
      "learning_rate": 0.002,
      "loss": 2.0959625244140625,
      "step": 1256
    },
    {
      "epoch": 0.1258,
      "grad_norm": 0.07432271540164948,
      "learning_rate": 0.002,
      "loss": 2.0701475143432617,
      "step": 1258
    },
    {
      "epoch": 0.126,
      "grad_norm": 0.07448964565992355,
      "learning_rate": 0.002,
      "loss": 1.9905776977539062,
      "step": 1260
    },
    {
      "epoch": 0.1262,
      "grad_norm": 0.0720694363117218,
      "learning_rate": 0.002,
      "loss": 2.040846586227417,
      "step": 1262
    },
    {
      "epoch": 0.1264,
      "grad_norm": 0.07592764496803284,
      "learning_rate": 0.002,
      "loss": 2.076841115951538,
      "step": 1264
    },
    {
      "epoch": 0.1266,
      "grad_norm": 0.07607028633356094,
      "learning_rate": 0.002,
      "loss": 2.0503110885620117,
      "step": 1266
    },
    {
      "epoch": 0.1268,
      "grad_norm": 0.0738333910703659,
      "learning_rate": 0.002,
      "loss": 2.1880764961242676,
      "step": 1268
    },
    {
      "epoch": 0.127,
      "grad_norm": 0.07521246373653412,
      "learning_rate": 0.002,
      "loss": 2.214982271194458,
      "step": 1270
    },
    {
      "epoch": 0.1272,
      "grad_norm": 0.07788985967636108,
      "learning_rate": 0.002,
      "loss": 2.0677390098571777,
      "step": 1272
    },
    {
      "epoch": 0.1274,
      "grad_norm": 0.07489370554685593,
      "learning_rate": 0.002,
      "loss": 2.080449342727661,
      "step": 1274
    },
    {
      "epoch": 0.1276,
      "grad_norm": 0.07221408188343048,
      "learning_rate": 0.002,
      "loss": 2.0182976722717285,
      "step": 1276
    },
    {
      "epoch": 0.1278,
      "grad_norm": 0.07412543892860413,
      "learning_rate": 0.002,
      "loss": 2.0048165321350098,
      "step": 1278
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.07315009087324142,
      "learning_rate": 0.002,
      "loss": 2.1076204776763916,
      "step": 1280
    },
    {
      "epoch": 0.1282,
      "grad_norm": 0.0710095539689064,
      "learning_rate": 0.002,
      "loss": 2.0729317665100098,
      "step": 1282
    },
    {
      "epoch": 0.1284,
      "grad_norm": 0.07210604101419449,
      "learning_rate": 0.002,
      "loss": 2.1375925540924072,
      "step": 1284
    },
    {
      "epoch": 0.1286,
      "grad_norm": 0.07607733458280563,
      "learning_rate": 0.002,
      "loss": 2.2397677898406982,
      "step": 1286
    },
    {
      "epoch": 0.1288,
      "grad_norm": 0.08006145060062408,
      "learning_rate": 0.002,
      "loss": 2.197540760040283,
      "step": 1288
    },
    {
      "epoch": 0.129,
      "grad_norm": 0.07209553569555283,
      "learning_rate": 0.002,
      "loss": 2.125035047531128,
      "step": 1290
    },
    {
      "epoch": 0.1292,
      "grad_norm": 0.07754766196012497,
      "learning_rate": 0.002,
      "loss": 2.036334753036499,
      "step": 1292
    },
    {
      "epoch": 0.1294,
      "grad_norm": 0.07770873606204987,
      "learning_rate": 0.002,
      "loss": 1.9982722997665405,
      "step": 1294
    },
    {
      "epoch": 0.1296,
      "grad_norm": 0.071790911257267,
      "learning_rate": 0.002,
      "loss": 2.172300100326538,
      "step": 1296
    },
    {
      "epoch": 0.1298,
      "grad_norm": 0.0765722393989563,
      "learning_rate": 0.002,
      "loss": 2.10780668258667,
      "step": 1298
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.07190870493650436,
      "learning_rate": 0.002,
      "loss": 1.997384786605835,
      "step": 1300
    },
    {
      "epoch": 0.1302,
      "grad_norm": 0.07562543451786041,
      "learning_rate": 0.002,
      "loss": 2.0349936485290527,
      "step": 1302
    },
    {
      "epoch": 0.1304,
      "grad_norm": 0.07203975319862366,
      "learning_rate": 0.002,
      "loss": 2.0420520305633545,
      "step": 1304
    },
    {
      "epoch": 0.1306,
      "grad_norm": 0.07388690859079361,
      "learning_rate": 0.002,
      "loss": 2.199613094329834,
      "step": 1306
    },
    {
      "epoch": 0.1308,
      "grad_norm": 0.07435441017150879,
      "learning_rate": 0.002,
      "loss": 2.056870460510254,
      "step": 1308
    },
    {
      "epoch": 0.131,
      "grad_norm": 0.0699809268116951,
      "learning_rate": 0.002,
      "loss": 1.9636199474334717,
      "step": 1310
    },
    {
      "epoch": 0.1312,
      "grad_norm": 0.07822288572788239,
      "learning_rate": 0.002,
      "loss": 2.112325668334961,
      "step": 1312
    },
    {
      "epoch": 0.1314,
      "grad_norm": 0.07557473331689835,
      "learning_rate": 0.002,
      "loss": 2.083538293838501,
      "step": 1314
    },
    {
      "epoch": 0.1316,
      "grad_norm": 0.07774205505847931,
      "learning_rate": 0.002,
      "loss": 2.210442304611206,
      "step": 1316
    },
    {
      "epoch": 0.1318,
      "grad_norm": 0.07238688319921494,
      "learning_rate": 0.002,
      "loss": 2.1419994831085205,
      "step": 1318
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.07282508909702301,
      "learning_rate": 0.002,
      "loss": 2.077404737472534,
      "step": 1320
    },
    {
      "epoch": 0.1322,
      "grad_norm": 0.07106118649244308,
      "learning_rate": 0.002,
      "loss": 2.2272658348083496,
      "step": 1322
    },
    {
      "epoch": 0.1324,
      "grad_norm": 0.0719086304306984,
      "learning_rate": 0.002,
      "loss": 2.0603206157684326,
      "step": 1324
    },
    {
      "epoch": 0.1326,
      "grad_norm": 0.07280543446540833,
      "learning_rate": 0.002,
      "loss": 2.101006031036377,
      "step": 1326
    },
    {
      "epoch": 0.1328,
      "grad_norm": 0.07681945711374283,
      "learning_rate": 0.002,
      "loss": 2.1512069702148438,
      "step": 1328
    },
    {
      "epoch": 0.133,
      "grad_norm": 0.07596462965011597,
      "learning_rate": 0.002,
      "loss": 2.129323959350586,
      "step": 1330
    },
    {
      "epoch": 0.1332,
      "grad_norm": 0.07781364768743515,
      "learning_rate": 0.002,
      "loss": 2.205763339996338,
      "step": 1332
    },
    {
      "epoch": 0.1334,
      "grad_norm": 0.07397348433732986,
      "learning_rate": 0.002,
      "loss": 2.0765039920806885,
      "step": 1334
    },
    {
      "epoch": 0.1336,
      "grad_norm": 0.07495831698179245,
      "learning_rate": 0.002,
      "loss": 2.0954713821411133,
      "step": 1336
    },
    {
      "epoch": 0.1338,
      "grad_norm": 0.07160273939371109,
      "learning_rate": 0.002,
      "loss": 2.151064395904541,
      "step": 1338
    },
    {
      "epoch": 0.134,
      "grad_norm": 0.07355043292045593,
      "learning_rate": 0.002,
      "loss": 2.1685070991516113,
      "step": 1340
    },
    {
      "epoch": 0.1342,
      "grad_norm": 0.08039357513189316,
      "learning_rate": 0.002,
      "loss": 2.1560189723968506,
      "step": 1342
    },
    {
      "epoch": 0.1344,
      "grad_norm": 0.07675959914922714,
      "learning_rate": 0.002,
      "loss": 1.9618535041809082,
      "step": 1344
    },
    {
      "epoch": 0.1346,
      "grad_norm": 0.07204312831163406,
      "learning_rate": 0.002,
      "loss": 2.0685036182403564,
      "step": 1346
    },
    {
      "epoch": 0.1348,
      "grad_norm": 0.07283423095941544,
      "learning_rate": 0.002,
      "loss": 2.0789856910705566,
      "step": 1348
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.07497082650661469,
      "learning_rate": 0.002,
      "loss": 2.130349636077881,
      "step": 1350
    },
    {
      "epoch": 0.1352,
      "grad_norm": 0.07317103445529938,
      "learning_rate": 0.002,
      "loss": 2.1243836879730225,
      "step": 1352
    },
    {
      "epoch": 0.1354,
      "grad_norm": 0.07483819127082825,
      "learning_rate": 0.002,
      "loss": 2.05188250541687,
      "step": 1354
    },
    {
      "epoch": 0.1356,
      "grad_norm": 0.07903642952442169,
      "learning_rate": 0.002,
      "loss": 2.045804977416992,
      "step": 1356
    },
    {
      "epoch": 0.1358,
      "grad_norm": 0.07433786988258362,
      "learning_rate": 0.002,
      "loss": 1.9893256425857544,
      "step": 1358
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.07189954072237015,
      "learning_rate": 0.002,
      "loss": 2.1197075843811035,
      "step": 1360
    },
    {
      "epoch": 0.1362,
      "grad_norm": 0.07306112349033356,
      "learning_rate": 0.002,
      "loss": 2.1252212524414062,
      "step": 1362
    },
    {
      "epoch": 0.1364,
      "grad_norm": 0.07411345839500427,
      "learning_rate": 0.002,
      "loss": 2.104438304901123,
      "step": 1364
    },
    {
      "epoch": 0.1366,
      "grad_norm": 0.07158102840185165,
      "learning_rate": 0.002,
      "loss": 2.0209038257598877,
      "step": 1366
    },
    {
      "epoch": 0.1368,
      "grad_norm": 0.07576750218868256,
      "learning_rate": 0.002,
      "loss": 2.0809884071350098,
      "step": 1368
    },
    {
      "epoch": 0.137,
      "grad_norm": 0.07767020165920258,
      "learning_rate": 0.002,
      "loss": 2.1134068965911865,
      "step": 1370
    },
    {
      "epoch": 0.1372,
      "grad_norm": 0.07013819366693497,
      "learning_rate": 0.002,
      "loss": 2.1186466217041016,
      "step": 1372
    },
    {
      "epoch": 0.1374,
      "grad_norm": 0.07400870323181152,
      "learning_rate": 0.002,
      "loss": 2.0877649784088135,
      "step": 1374
    },
    {
      "epoch": 0.1376,
      "grad_norm": 0.06925424933433533,
      "learning_rate": 0.002,
      "loss": 2.087359666824341,
      "step": 1376
    },
    {
      "epoch": 0.1378,
      "grad_norm": 0.07414641231298447,
      "learning_rate": 0.002,
      "loss": 2.1010630130767822,
      "step": 1378
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.0767003670334816,
      "learning_rate": 0.002,
      "loss": 2.1526339054107666,
      "step": 1380
    },
    {
      "epoch": 0.1382,
      "grad_norm": 0.07344821095466614,
      "learning_rate": 0.002,
      "loss": 2.0420312881469727,
      "step": 1382
    },
    {
      "epoch": 0.1384,
      "grad_norm": 0.07301363348960876,
      "learning_rate": 0.002,
      "loss": 2.099546432495117,
      "step": 1384
    },
    {
      "epoch": 0.1386,
      "grad_norm": 0.07663039118051529,
      "learning_rate": 0.002,
      "loss": 2.0541231632232666,
      "step": 1386
    },
    {
      "epoch": 0.1388,
      "grad_norm": 0.0751442089676857,
      "learning_rate": 0.002,
      "loss": 2.0829429626464844,
      "step": 1388
    },
    {
      "epoch": 0.139,
      "grad_norm": 0.0771271362900734,
      "learning_rate": 0.002,
      "loss": 2.1359546184539795,
      "step": 1390
    },
    {
      "epoch": 0.1392,
      "grad_norm": 0.0833253487944603,
      "learning_rate": 0.002,
      "loss": 2.0948920249938965,
      "step": 1392
    },
    {
      "epoch": 0.1394,
      "grad_norm": 0.0736006423830986,
      "learning_rate": 0.002,
      "loss": 1.9986518621444702,
      "step": 1394
    },
    {
      "epoch": 0.1396,
      "grad_norm": 0.07763192802667618,
      "learning_rate": 0.002,
      "loss": 2.1851320266723633,
      "step": 1396
    },
    {
      "epoch": 0.1398,
      "grad_norm": 0.07578496634960175,
      "learning_rate": 0.002,
      "loss": 2.0996432304382324,
      "step": 1398
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.07257342338562012,
      "learning_rate": 0.002,
      "loss": 2.061452627182007,
      "step": 1400
    },
    {
      "epoch": 0.1402,
      "grad_norm": 0.0738401710987091,
      "learning_rate": 0.002,
      "loss": 2.1607351303100586,
      "step": 1402
    },
    {
      "epoch": 0.1404,
      "grad_norm": 0.07418519258499146,
      "learning_rate": 0.002,
      "loss": 2.136280059814453,
      "step": 1404
    },
    {
      "epoch": 0.1406,
      "grad_norm": 0.07844164222478867,
      "learning_rate": 0.002,
      "loss": 2.0440738201141357,
      "step": 1406
    },
    {
      "epoch": 0.1408,
      "grad_norm": 0.07174115628004074,
      "learning_rate": 0.002,
      "loss": 2.127737522125244,
      "step": 1408
    },
    {
      "epoch": 0.141,
      "grad_norm": 0.07693090289831161,
      "learning_rate": 0.002,
      "loss": 2.0323946475982666,
      "step": 1410
    },
    {
      "epoch": 0.1412,
      "grad_norm": 0.07711713016033173,
      "learning_rate": 0.002,
      "loss": 2.0991463661193848,
      "step": 1412
    },
    {
      "epoch": 0.1414,
      "grad_norm": 0.07523197680711746,
      "learning_rate": 0.002,
      "loss": 1.9388043880462646,
      "step": 1414
    },
    {
      "epoch": 0.1416,
      "grad_norm": 0.07716935873031616,
      "learning_rate": 0.002,
      "loss": 2.110644578933716,
      "step": 1416
    },
    {
      "epoch": 0.1418,
      "grad_norm": 0.07524771243333817,
      "learning_rate": 0.002,
      "loss": 2.099801778793335,
      "step": 1418
    },
    {
      "epoch": 0.142,
      "grad_norm": 0.07205197960138321,
      "learning_rate": 0.002,
      "loss": 2.039616107940674,
      "step": 1420
    },
    {
      "epoch": 0.1422,
      "grad_norm": 0.07068921625614166,
      "learning_rate": 0.002,
      "loss": 2.0113296508789062,
      "step": 1422
    },
    {
      "epoch": 0.1424,
      "grad_norm": 0.08478952944278717,
      "learning_rate": 0.002,
      "loss": 2.068065643310547,
      "step": 1424
    },
    {
      "epoch": 0.1426,
      "grad_norm": 0.07446826249361038,
      "learning_rate": 0.002,
      "loss": 2.064588785171509,
      "step": 1426
    },
    {
      "epoch": 0.1428,
      "grad_norm": 0.07332564890384674,
      "learning_rate": 0.002,
      "loss": 2.1091630458831787,
      "step": 1428
    },
    {
      "epoch": 0.143,
      "grad_norm": 0.07425611466169357,
      "learning_rate": 0.002,
      "loss": 2.0847034454345703,
      "step": 1430
    },
    {
      "epoch": 0.1432,
      "grad_norm": 0.07217207551002502,
      "learning_rate": 0.002,
      "loss": 2.057948112487793,
      "step": 1432
    },
    {
      "epoch": 0.1434,
      "grad_norm": 0.07432244718074799,
      "learning_rate": 0.002,
      "loss": 2.0595812797546387,
      "step": 1434
    },
    {
      "epoch": 0.1436,
      "grad_norm": 0.0717838704586029,
      "learning_rate": 0.002,
      "loss": 2.0152392387390137,
      "step": 1436
    },
    {
      "epoch": 0.1438,
      "grad_norm": 0.07548312097787857,
      "learning_rate": 0.002,
      "loss": 2.040618658065796,
      "step": 1438
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.07060568034648895,
      "learning_rate": 0.002,
      "loss": 2.1086573600769043,
      "step": 1440
    },
    {
      "epoch": 0.1442,
      "grad_norm": 0.0751064196228981,
      "learning_rate": 0.002,
      "loss": 2.178683280944824,
      "step": 1442
    },
    {
      "epoch": 0.1444,
      "grad_norm": 0.07254419475793839,
      "learning_rate": 0.002,
      "loss": 2.062152862548828,
      "step": 1444
    },
    {
      "epoch": 0.1446,
      "grad_norm": 0.07547087967395782,
      "learning_rate": 0.002,
      "loss": 2.105653762817383,
      "step": 1446
    },
    {
      "epoch": 0.1448,
      "grad_norm": 0.07600472122430801,
      "learning_rate": 0.002,
      "loss": 2.0811767578125,
      "step": 1448
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.07655759900808334,
      "learning_rate": 0.002,
      "loss": 2.057117462158203,
      "step": 1450
    },
    {
      "epoch": 0.1452,
      "grad_norm": 0.07337045669555664,
      "learning_rate": 0.002,
      "loss": 2.0979163646698,
      "step": 1452
    },
    {
      "epoch": 0.1454,
      "grad_norm": 0.07062342762947083,
      "learning_rate": 0.002,
      "loss": 2.1257376670837402,
      "step": 1454
    },
    {
      "epoch": 0.1456,
      "grad_norm": 0.07141361385583878,
      "learning_rate": 0.002,
      "loss": 2.1165003776550293,
      "step": 1456
    },
    {
      "epoch": 0.1458,
      "grad_norm": 0.07973118871450424,
      "learning_rate": 0.002,
      "loss": 2.1623384952545166,
      "step": 1458
    },
    {
      "epoch": 0.146,
      "grad_norm": 0.07220372557640076,
      "learning_rate": 0.002,
      "loss": 2.010887622833252,
      "step": 1460
    },
    {
      "epoch": 0.1462,
      "grad_norm": 0.07381699234247208,
      "learning_rate": 0.002,
      "loss": 2.11995792388916,
      "step": 1462
    },
    {
      "epoch": 0.1464,
      "grad_norm": 0.07707764953374863,
      "learning_rate": 0.002,
      "loss": 2.0592098236083984,
      "step": 1464
    },
    {
      "epoch": 0.1466,
      "grad_norm": 0.07372082769870758,
      "learning_rate": 0.002,
      "loss": 2.112884759902954,
      "step": 1466
    },
    {
      "epoch": 0.1468,
      "grad_norm": 0.07582217454910278,
      "learning_rate": 0.002,
      "loss": 2.094902753829956,
      "step": 1468
    },
    {
      "epoch": 0.147,
      "grad_norm": 0.07246437668800354,
      "learning_rate": 0.002,
      "loss": 2.0258281230926514,
      "step": 1470
    },
    {
      "epoch": 0.1472,
      "grad_norm": 0.07027522474527359,
      "learning_rate": 0.002,
      "loss": 1.995959997177124,
      "step": 1472
    },
    {
      "epoch": 0.1474,
      "grad_norm": 0.07412761449813843,
      "learning_rate": 0.002,
      "loss": 1.967400312423706,
      "step": 1474
    },
    {
      "epoch": 0.1476,
      "grad_norm": 0.0732477605342865,
      "learning_rate": 0.002,
      "loss": 2.0308661460876465,
      "step": 1476
    },
    {
      "epoch": 0.1478,
      "grad_norm": 0.0706886500120163,
      "learning_rate": 0.002,
      "loss": 2.193591594696045,
      "step": 1478
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.07896345853805542,
      "learning_rate": 0.002,
      "loss": 2.118762493133545,
      "step": 1480
    },
    {
      "epoch": 0.1482,
      "grad_norm": 0.0713469535112381,
      "learning_rate": 0.002,
      "loss": 2.006922721862793,
      "step": 1482
    },
    {
      "epoch": 0.1484,
      "grad_norm": 0.07564645260572433,
      "learning_rate": 0.002,
      "loss": 2.124854326248169,
      "step": 1484
    },
    {
      "epoch": 0.1486,
      "grad_norm": 0.07450724393129349,
      "learning_rate": 0.002,
      "loss": 2.088937759399414,
      "step": 1486
    },
    {
      "epoch": 0.1488,
      "grad_norm": 0.07264949381351471,
      "learning_rate": 0.002,
      "loss": 2.0539491176605225,
      "step": 1488
    },
    {
      "epoch": 0.149,
      "grad_norm": 0.07765015959739685,
      "learning_rate": 0.002,
      "loss": 2.1706202030181885,
      "step": 1490
    },
    {
      "epoch": 0.1492,
      "grad_norm": 0.07687120884656906,
      "learning_rate": 0.002,
      "loss": 2.0416269302368164,
      "step": 1492
    },
    {
      "epoch": 0.1494,
      "grad_norm": 0.07305614650249481,
      "learning_rate": 0.002,
      "loss": 2.046542167663574,
      "step": 1494
    },
    {
      "epoch": 0.1496,
      "grad_norm": 0.07372792810201645,
      "learning_rate": 0.002,
      "loss": 1.9286292791366577,
      "step": 1496
    },
    {
      "epoch": 0.1498,
      "grad_norm": 0.07961343228816986,
      "learning_rate": 0.002,
      "loss": 2.0458920001983643,
      "step": 1498
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.0820629671216011,
      "learning_rate": 0.002,
      "loss": 2.1374316215515137,
      "step": 1500
    },
    {
      "epoch": 0.1502,
      "grad_norm": 0.07368797063827515,
      "learning_rate": 0.002,
      "loss": 2.0823757648468018,
      "step": 1502
    },
    {
      "epoch": 0.1504,
      "grad_norm": 0.07460813224315643,
      "learning_rate": 0.002,
      "loss": 1.8870255947113037,
      "step": 1504
    },
    {
      "epoch": 0.1506,
      "grad_norm": 0.07642543315887451,
      "learning_rate": 0.002,
      "loss": 2.0043563842773438,
      "step": 1506
    },
    {
      "epoch": 0.1508,
      "grad_norm": 0.07093997299671173,
      "learning_rate": 0.002,
      "loss": 2.136366844177246,
      "step": 1508
    },
    {
      "epoch": 0.151,
      "grad_norm": 0.07244401425123215,
      "learning_rate": 0.002,
      "loss": 2.113682985305786,
      "step": 1510
    },
    {
      "epoch": 0.1512,
      "grad_norm": 0.08035255968570709,
      "learning_rate": 0.002,
      "loss": 2.11247181892395,
      "step": 1512
    },
    {
      "epoch": 0.1514,
      "grad_norm": 0.07973160594701767,
      "learning_rate": 0.002,
      "loss": 2.0302140712738037,
      "step": 1514
    },
    {
      "epoch": 0.1516,
      "grad_norm": 0.07703500986099243,
      "learning_rate": 0.002,
      "loss": 2.0129308700561523,
      "step": 1516
    },
    {
      "epoch": 0.1518,
      "grad_norm": 0.07648824900388718,
      "learning_rate": 0.002,
      "loss": 2.1478829383850098,
      "step": 1518
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.08575855940580368,
      "learning_rate": 0.002,
      "loss": 2.0756638050079346,
      "step": 1520
    },
    {
      "epoch": 0.1522,
      "grad_norm": 0.07385808974504471,
      "learning_rate": 0.002,
      "loss": 2.0634255409240723,
      "step": 1522
    },
    {
      "epoch": 0.1524,
      "grad_norm": 0.07510539144277573,
      "learning_rate": 0.002,
      "loss": 1.9820526838302612,
      "step": 1524
    },
    {
      "epoch": 0.1526,
      "grad_norm": 0.08474548906087875,
      "learning_rate": 0.002,
      "loss": 1.9759695529937744,
      "step": 1526
    },
    {
      "epoch": 0.1528,
      "grad_norm": 0.07325828075408936,
      "learning_rate": 0.002,
      "loss": 2.0810132026672363,
      "step": 1528
    },
    {
      "epoch": 0.153,
      "grad_norm": 0.07181402295827866,
      "learning_rate": 0.002,
      "loss": 2.1337506771087646,
      "step": 1530
    },
    {
      "epoch": 0.1532,
      "grad_norm": 0.07261587679386139,
      "learning_rate": 0.002,
      "loss": 2.0383944511413574,
      "step": 1532
    },
    {
      "epoch": 0.1534,
      "grad_norm": 0.07591050863265991,
      "learning_rate": 0.002,
      "loss": 1.8666166067123413,
      "step": 1534
    },
    {
      "epoch": 0.1536,
      "grad_norm": 0.07050628215074539,
      "learning_rate": 0.002,
      "loss": 2.0281646251678467,
      "step": 1536
    },
    {
      "epoch": 0.1538,
      "grad_norm": 0.07352118194103241,
      "learning_rate": 0.002,
      "loss": 2.116265296936035,
      "step": 1538
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.07260631024837494,
      "learning_rate": 0.002,
      "loss": 2.036597728729248,
      "step": 1540
    },
    {
      "epoch": 0.1542,
      "grad_norm": 0.07308587431907654,
      "learning_rate": 0.002,
      "loss": 2.105753183364868,
      "step": 1542
    },
    {
      "epoch": 0.1544,
      "grad_norm": 0.07279038429260254,
      "learning_rate": 0.002,
      "loss": 2.097508192062378,
      "step": 1544
    },
    {
      "epoch": 0.1546,
      "grad_norm": 0.0749455988407135,
      "learning_rate": 0.002,
      "loss": 2.0140602588653564,
      "step": 1546
    },
    {
      "epoch": 0.1548,
      "grad_norm": 0.0754837617278099,
      "learning_rate": 0.002,
      "loss": 1.8678150177001953,
      "step": 1548
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.07635439187288284,
      "learning_rate": 0.002,
      "loss": 2.1300814151763916,
      "step": 1550
    },
    {
      "epoch": 0.1552,
      "grad_norm": 0.07521235197782516,
      "learning_rate": 0.002,
      "loss": 2.0812828540802,
      "step": 1552
    },
    {
      "epoch": 0.1554,
      "grad_norm": 0.07857983559370041,
      "learning_rate": 0.002,
      "loss": 2.1078896522521973,
      "step": 1554
    },
    {
      "epoch": 0.1556,
      "grad_norm": 0.07463716715574265,
      "learning_rate": 0.002,
      "loss": 1.9632741212844849,
      "step": 1556
    },
    {
      "epoch": 0.1558,
      "grad_norm": 0.07418062537908554,
      "learning_rate": 0.002,
      "loss": 1.952651023864746,
      "step": 1558
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.07721658051013947,
      "learning_rate": 0.002,
      "loss": 1.9926362037658691,
      "step": 1560
    },
    {
      "epoch": 0.1562,
      "grad_norm": 0.07594937831163406,
      "learning_rate": 0.002,
      "loss": 2.142509698867798,
      "step": 1562
    },
    {
      "epoch": 0.1564,
      "grad_norm": 0.07187453657388687,
      "learning_rate": 0.002,
      "loss": 2.082077741622925,
      "step": 1564
    },
    {
      "epoch": 0.1566,
      "grad_norm": 0.07446449995040894,
      "learning_rate": 0.002,
      "loss": 2.0862817764282227,
      "step": 1566
    },
    {
      "epoch": 0.1568,
      "grad_norm": 0.07260862737894058,
      "learning_rate": 0.002,
      "loss": 2.1520578861236572,
      "step": 1568
    },
    {
      "epoch": 0.157,
      "grad_norm": 0.07474535703659058,
      "learning_rate": 0.002,
      "loss": 2.0227062702178955,
      "step": 1570
    },
    {
      "epoch": 0.1572,
      "grad_norm": 0.07134494930505753,
      "learning_rate": 0.002,
      "loss": 2.0814313888549805,
      "step": 1572
    },
    {
      "epoch": 0.1574,
      "grad_norm": 0.07805604487657547,
      "learning_rate": 0.002,
      "loss": 2.15790057182312,
      "step": 1574
    },
    {
      "epoch": 0.1576,
      "grad_norm": 0.07595029473304749,
      "learning_rate": 0.002,
      "loss": 2.0175621509552,
      "step": 1576
    },
    {
      "epoch": 0.1578,
      "grad_norm": 0.07194190472364426,
      "learning_rate": 0.002,
      "loss": 2.0004515647888184,
      "step": 1578
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.07226980477571487,
      "learning_rate": 0.002,
      "loss": 1.8962677717208862,
      "step": 1580
    },
    {
      "epoch": 0.1582,
      "grad_norm": 0.07918702065944672,
      "learning_rate": 0.002,
      "loss": 2.0338785648345947,
      "step": 1582
    },
    {
      "epoch": 0.1584,
      "grad_norm": 0.07334207743406296,
      "learning_rate": 0.002,
      "loss": 1.9894874095916748,
      "step": 1584
    },
    {
      "epoch": 0.1586,
      "grad_norm": 0.0737851932644844,
      "learning_rate": 0.002,
      "loss": 2.132560968399048,
      "step": 1586
    },
    {
      "epoch": 0.1588,
      "grad_norm": 0.07402051240205765,
      "learning_rate": 0.002,
      "loss": 2.073052167892456,
      "step": 1588
    },
    {
      "epoch": 0.159,
      "grad_norm": 0.07295355945825577,
      "learning_rate": 0.002,
      "loss": 1.9998468160629272,
      "step": 1590
    },
    {
      "epoch": 0.1592,
      "grad_norm": 0.08218786865472794,
      "learning_rate": 0.002,
      "loss": 2.1164300441741943,
      "step": 1592
    },
    {
      "epoch": 0.1594,
      "grad_norm": 0.07137202471494675,
      "learning_rate": 0.002,
      "loss": 2.0453524589538574,
      "step": 1594
    },
    {
      "epoch": 0.1596,
      "grad_norm": 0.07293243706226349,
      "learning_rate": 0.002,
      "loss": 2.126450300216675,
      "step": 1596
    },
    {
      "epoch": 0.1598,
      "grad_norm": 0.07506708055734634,
      "learning_rate": 0.002,
      "loss": 2.1242542266845703,
      "step": 1598
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.07331743836402893,
      "learning_rate": 0.002,
      "loss": 2.026481866836548,
      "step": 1600
    },
    {
      "epoch": 0.1602,
      "grad_norm": 0.074144147336483,
      "learning_rate": 0.002,
      "loss": 2.1167285442352295,
      "step": 1602
    },
    {
      "epoch": 0.1604,
      "grad_norm": 0.07606136798858643,
      "learning_rate": 0.002,
      "loss": 2.1567087173461914,
      "step": 1604
    },
    {
      "epoch": 0.1606,
      "grad_norm": 0.07727319002151489,
      "learning_rate": 0.002,
      "loss": 2.008620023727417,
      "step": 1606
    },
    {
      "epoch": 0.1608,
      "grad_norm": 0.07652614265680313,
      "learning_rate": 0.002,
      "loss": 2.023723602294922,
      "step": 1608
    },
    {
      "epoch": 0.161,
      "grad_norm": 0.07944142818450928,
      "learning_rate": 0.002,
      "loss": 2.0489375591278076,
      "step": 1610
    },
    {
      "epoch": 0.1612,
      "grad_norm": 0.08112798631191254,
      "learning_rate": 0.002,
      "loss": 1.9857752323150635,
      "step": 1612
    },
    {
      "epoch": 0.1614,
      "grad_norm": 0.07324112206697464,
      "learning_rate": 0.002,
      "loss": 1.9679566621780396,
      "step": 1614
    },
    {
      "epoch": 0.1616,
      "grad_norm": 0.08540373295545578,
      "learning_rate": 0.002,
      "loss": 2.0271008014678955,
      "step": 1616
    },
    {
      "epoch": 0.1618,
      "grad_norm": 0.07922551035881042,
      "learning_rate": 0.002,
      "loss": 1.9334667921066284,
      "step": 1618
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.07764491438865662,
      "learning_rate": 0.002,
      "loss": 2.0488972663879395,
      "step": 1620
    },
    {
      "epoch": 0.1622,
      "grad_norm": 0.07708816975355148,
      "learning_rate": 0.002,
      "loss": 2.1202070713043213,
      "step": 1622
    },
    {
      "epoch": 0.1624,
      "grad_norm": 0.0747041329741478,
      "learning_rate": 0.002,
      "loss": 2.0174412727355957,
      "step": 1624
    },
    {
      "epoch": 0.1626,
      "grad_norm": 0.07188914716243744,
      "learning_rate": 0.002,
      "loss": 2.0339152812957764,
      "step": 1626
    },
    {
      "epoch": 0.1628,
      "grad_norm": 0.0745021402835846,
      "learning_rate": 0.002,
      "loss": 2.1272027492523193,
      "step": 1628
    },
    {
      "epoch": 0.163,
      "grad_norm": 0.07572438567876816,
      "learning_rate": 0.002,
      "loss": 1.974536657333374,
      "step": 1630
    },
    {
      "epoch": 0.1632,
      "grad_norm": 0.07948266714811325,
      "learning_rate": 0.002,
      "loss": 1.9441884756088257,
      "step": 1632
    },
    {
      "epoch": 0.1634,
      "grad_norm": 0.08547366410493851,
      "learning_rate": 0.002,
      "loss": 2.0210535526275635,
      "step": 1634
    },
    {
      "epoch": 0.1636,
      "grad_norm": 0.07394084334373474,
      "learning_rate": 0.002,
      "loss": 2.054553508758545,
      "step": 1636
    },
    {
      "epoch": 0.1638,
      "grad_norm": 0.07616014033555984,
      "learning_rate": 0.002,
      "loss": 2.0444605350494385,
      "step": 1638
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.075071781873703,
      "learning_rate": 0.002,
      "loss": 2.157501220703125,
      "step": 1640
    },
    {
      "epoch": 0.1642,
      "grad_norm": 0.07492058724164963,
      "learning_rate": 0.002,
      "loss": 2.229353904724121,
      "step": 1642
    },
    {
      "epoch": 0.1644,
      "grad_norm": 0.0739431381225586,
      "learning_rate": 0.002,
      "loss": 2.0768918991088867,
      "step": 1644
    },
    {
      "epoch": 0.1646,
      "grad_norm": 0.07818491011857986,
      "learning_rate": 0.002,
      "loss": 1.978445053100586,
      "step": 1646
    },
    {
      "epoch": 0.1648,
      "grad_norm": 0.07702851295471191,
      "learning_rate": 0.002,
      "loss": 2.0013813972473145,
      "step": 1648
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.0778750628232956,
      "learning_rate": 0.002,
      "loss": 2.1169087886810303,
      "step": 1650
    },
    {
      "epoch": 0.1652,
      "grad_norm": 0.07372111827135086,
      "learning_rate": 0.002,
      "loss": 1.9516445398330688,
      "step": 1652
    },
    {
      "epoch": 0.1654,
      "grad_norm": 0.07664504647254944,
      "learning_rate": 0.002,
      "loss": 2.0571932792663574,
      "step": 1654
    },
    {
      "epoch": 0.1656,
      "grad_norm": 0.07747307419776917,
      "learning_rate": 0.002,
      "loss": 2.112342357635498,
      "step": 1656
    },
    {
      "epoch": 0.1658,
      "grad_norm": 0.07394912838935852,
      "learning_rate": 0.002,
      "loss": 1.9988903999328613,
      "step": 1658
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.07434363663196564,
      "learning_rate": 0.002,
      "loss": 2.127570867538452,
      "step": 1660
    },
    {
      "epoch": 0.1662,
      "grad_norm": 0.07162563502788544,
      "learning_rate": 0.002,
      "loss": 2.1028223037719727,
      "step": 1662
    },
    {
      "epoch": 0.1664,
      "grad_norm": 0.08235505223274231,
      "learning_rate": 0.002,
      "loss": 2.0046074390411377,
      "step": 1664
    },
    {
      "epoch": 0.1666,
      "grad_norm": 0.0764487236738205,
      "learning_rate": 0.002,
      "loss": 2.0746984481811523,
      "step": 1666
    },
    {
      "epoch": 0.1668,
      "grad_norm": 0.08113086968660355,
      "learning_rate": 0.002,
      "loss": 1.9035871028900146,
      "step": 1668
    },
    {
      "epoch": 0.167,
      "grad_norm": 0.08087792992591858,
      "learning_rate": 0.002,
      "loss": 2.029357433319092,
      "step": 1670
    },
    {
      "epoch": 0.1672,
      "grad_norm": 0.0802219808101654,
      "learning_rate": 0.002,
      "loss": 2.0138373374938965,
      "step": 1672
    },
    {
      "epoch": 0.1674,
      "grad_norm": 0.07401550561189651,
      "learning_rate": 0.002,
      "loss": 2.0571253299713135,
      "step": 1674
    },
    {
      "epoch": 0.1676,
      "grad_norm": 0.07494015991687775,
      "learning_rate": 0.002,
      "loss": 2.064505100250244,
      "step": 1676
    },
    {
      "epoch": 0.1678,
      "grad_norm": 0.07123389095067978,
      "learning_rate": 0.002,
      "loss": 1.9455479383468628,
      "step": 1678
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.07775776833295822,
      "learning_rate": 0.002,
      "loss": 2.0193943977355957,
      "step": 1680
    },
    {
      "epoch": 0.1682,
      "grad_norm": 0.07578908652067184,
      "learning_rate": 0.002,
      "loss": 2.022446393966675,
      "step": 1682
    },
    {
      "epoch": 0.1684,
      "grad_norm": 0.07389418035745621,
      "learning_rate": 0.002,
      "loss": 2.026526689529419,
      "step": 1684
    },
    {
      "epoch": 0.1686,
      "grad_norm": 0.07718732953071594,
      "learning_rate": 0.002,
      "loss": 2.0270955562591553,
      "step": 1686
    },
    {
      "epoch": 0.1688,
      "grad_norm": 0.07536289095878601,
      "learning_rate": 0.002,
      "loss": 2.1234540939331055,
      "step": 1688
    },
    {
      "epoch": 0.169,
      "grad_norm": 0.07100782543420792,
      "learning_rate": 0.002,
      "loss": 2.156771659851074,
      "step": 1690
    },
    {
      "epoch": 0.1692,
      "grad_norm": 0.07666651904582977,
      "learning_rate": 0.002,
      "loss": 1.9003492593765259,
      "step": 1692
    },
    {
      "epoch": 0.1694,
      "grad_norm": 0.07424253225326538,
      "learning_rate": 0.002,
      "loss": 2.0253427028656006,
      "step": 1694
    },
    {
      "epoch": 0.1696,
      "grad_norm": 0.08001311123371124,
      "learning_rate": 0.002,
      "loss": 2.1323394775390625,
      "step": 1696
    },
    {
      "epoch": 0.1698,
      "grad_norm": 0.07312895357608795,
      "learning_rate": 0.002,
      "loss": 1.9378201961517334,
      "step": 1698
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.07589532434940338,
      "learning_rate": 0.002,
      "loss": 2.1028289794921875,
      "step": 1700
    },
    {
      "epoch": 0.1702,
      "grad_norm": 0.07242558151483536,
      "learning_rate": 0.002,
      "loss": 2.03623628616333,
      "step": 1702
    },
    {
      "epoch": 0.1704,
      "grad_norm": 0.07230544835329056,
      "learning_rate": 0.002,
      "loss": 2.069361448287964,
      "step": 1704
    },
    {
      "epoch": 0.1706,
      "grad_norm": 0.08096840977668762,
      "learning_rate": 0.002,
      "loss": 2.0677847862243652,
      "step": 1706
    },
    {
      "epoch": 0.1708,
      "grad_norm": 0.07448755949735641,
      "learning_rate": 0.002,
      "loss": 1.9019745588302612,
      "step": 1708
    },
    {
      "epoch": 0.171,
      "grad_norm": 0.07541760057210922,
      "learning_rate": 0.002,
      "loss": 1.98647141456604,
      "step": 1710
    },
    {
      "epoch": 0.1712,
      "grad_norm": 0.07599081844091415,
      "learning_rate": 0.002,
      "loss": 2.056687831878662,
      "step": 1712
    },
    {
      "epoch": 0.1714,
      "grad_norm": 0.07685410976409912,
      "learning_rate": 0.002,
      "loss": 1.9951837062835693,
      "step": 1714
    },
    {
      "epoch": 0.1716,
      "grad_norm": 0.07613454759120941,
      "learning_rate": 0.002,
      "loss": 1.950670599937439,
      "step": 1716
    },
    {
      "epoch": 0.1718,
      "grad_norm": 0.07113487273454666,
      "learning_rate": 0.002,
      "loss": 1.9866174459457397,
      "step": 1718
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.07698488980531693,
      "learning_rate": 0.002,
      "loss": 1.9908841848373413,
      "step": 1720
    },
    {
      "epoch": 0.1722,
      "grad_norm": 0.07847699522972107,
      "learning_rate": 0.002,
      "loss": 1.9502193927764893,
      "step": 1722
    },
    {
      "epoch": 0.1724,
      "grad_norm": 0.07652683556079865,
      "learning_rate": 0.002,
      "loss": 1.9898251295089722,
      "step": 1724
    },
    {
      "epoch": 0.1726,
      "grad_norm": 0.08788255602121353,
      "learning_rate": 0.002,
      "loss": 2.1220459938049316,
      "step": 1726
    },
    {
      "epoch": 0.1728,
      "grad_norm": 0.07747672498226166,
      "learning_rate": 0.002,
      "loss": 2.058579921722412,
      "step": 1728
    },
    {
      "epoch": 0.173,
      "grad_norm": 0.07136350870132446,
      "learning_rate": 0.002,
      "loss": 1.9220625162124634,
      "step": 1730
    },
    {
      "epoch": 0.1732,
      "grad_norm": 0.07820304483175278,
      "learning_rate": 0.002,
      "loss": 2.1043810844421387,
      "step": 1732
    },
    {
      "epoch": 0.1734,
      "grad_norm": 0.07268442213535309,
      "learning_rate": 0.002,
      "loss": 2.1724164485931396,
      "step": 1734
    },
    {
      "epoch": 0.1736,
      "grad_norm": 0.07410701364278793,
      "learning_rate": 0.002,
      "loss": 1.9956454038619995,
      "step": 1736
    },
    {
      "epoch": 0.1738,
      "grad_norm": 0.08288054913282394,
      "learning_rate": 0.002,
      "loss": 1.9310275316238403,
      "step": 1738
    },
    {
      "epoch": 0.174,
      "grad_norm": 0.07379169017076492,
      "learning_rate": 0.002,
      "loss": 2.158606767654419,
      "step": 1740
    },
    {
      "epoch": 0.1742,
      "grad_norm": 0.07867053896188736,
      "learning_rate": 0.002,
      "loss": 2.049553632736206,
      "step": 1742
    },
    {
      "epoch": 0.1744,
      "grad_norm": 0.07448194921016693,
      "learning_rate": 0.002,
      "loss": 2.1119422912597656,
      "step": 1744
    },
    {
      "epoch": 0.1746,
      "grad_norm": 0.07308252155780792,
      "learning_rate": 0.002,
      "loss": 2.1226775646209717,
      "step": 1746
    },
    {
      "epoch": 0.1748,
      "grad_norm": 0.07961224019527435,
      "learning_rate": 0.002,
      "loss": 2.097341537475586,
      "step": 1748
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.07117109000682831,
      "learning_rate": 0.002,
      "loss": 1.8197592496871948,
      "step": 1750
    },
    {
      "epoch": 0.1752,
      "grad_norm": 0.07547926157712936,
      "learning_rate": 0.002,
      "loss": 1.973343014717102,
      "step": 1752
    },
    {
      "epoch": 0.1754,
      "grad_norm": 0.07283801585435867,
      "learning_rate": 0.002,
      "loss": 2.0424771308898926,
      "step": 1754
    },
    {
      "epoch": 0.1756,
      "grad_norm": 0.07660172134637833,
      "learning_rate": 0.002,
      "loss": 2.168020486831665,
      "step": 1756
    },
    {
      "epoch": 0.1758,
      "grad_norm": 0.07084596157073975,
      "learning_rate": 0.002,
      "loss": 2.0135645866394043,
      "step": 1758
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.08032834529876709,
      "learning_rate": 0.002,
      "loss": 1.8673094511032104,
      "step": 1760
    },
    {
      "epoch": 0.1762,
      "grad_norm": 0.07184142619371414,
      "learning_rate": 0.002,
      "loss": 2.0526771545410156,
      "step": 1762
    },
    {
      "epoch": 0.1764,
      "grad_norm": 0.07069835066795349,
      "learning_rate": 0.002,
      "loss": 2.148001194000244,
      "step": 1764
    },
    {
      "epoch": 0.1766,
      "grad_norm": 0.07474255561828613,
      "learning_rate": 0.002,
      "loss": 2.152073860168457,
      "step": 1766
    },
    {
      "epoch": 0.1768,
      "grad_norm": 0.07535561919212341,
      "learning_rate": 0.002,
      "loss": 2.0829999446868896,
      "step": 1768
    },
    {
      "epoch": 0.177,
      "grad_norm": 0.07707375288009644,
      "learning_rate": 0.002,
      "loss": 2.0681803226470947,
      "step": 1770
    },
    {
      "epoch": 0.1772,
      "grad_norm": 0.07198737561702728,
      "learning_rate": 0.002,
      "loss": 2.0520801544189453,
      "step": 1772
    },
    {
      "epoch": 0.1774,
      "grad_norm": 0.07704342156648636,
      "learning_rate": 0.002,
      "loss": 1.9790741205215454,
      "step": 1774
    },
    {
      "epoch": 0.1776,
      "grad_norm": 0.07308527827262878,
      "learning_rate": 0.002,
      "loss": 1.8921363353729248,
      "step": 1776
    },
    {
      "epoch": 0.1778,
      "grad_norm": 0.07369029521942139,
      "learning_rate": 0.002,
      "loss": 2.1478757858276367,
      "step": 1778
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.08679457753896713,
      "learning_rate": 0.002,
      "loss": 1.9624991416931152,
      "step": 1780
    },
    {
      "epoch": 0.1782,
      "grad_norm": 0.07663458585739136,
      "learning_rate": 0.002,
      "loss": 2.1142423152923584,
      "step": 1782
    },
    {
      "epoch": 0.1784,
      "grad_norm": 0.07674773782491684,
      "learning_rate": 0.002,
      "loss": 2.0569021701812744,
      "step": 1784
    },
    {
      "epoch": 0.1786,
      "grad_norm": 0.07379690557718277,
      "learning_rate": 0.002,
      "loss": 1.9466909170150757,
      "step": 1786
    },
    {
      "epoch": 0.1788,
      "grad_norm": 0.07840987294912338,
      "learning_rate": 0.002,
      "loss": 2.107332468032837,
      "step": 1788
    },
    {
      "epoch": 0.179,
      "grad_norm": 0.07417541742324829,
      "learning_rate": 0.002,
      "loss": 2.0074548721313477,
      "step": 1790
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.07527904212474823,
      "learning_rate": 0.002,
      "loss": 2.1119160652160645,
      "step": 1792
    },
    {
      "epoch": 0.1794,
      "grad_norm": 0.07240865379571915,
      "learning_rate": 0.002,
      "loss": 2.0384199619293213,
      "step": 1794
    },
    {
      "epoch": 0.1796,
      "grad_norm": 0.07644591480493546,
      "learning_rate": 0.002,
      "loss": 1.9770536422729492,
      "step": 1796
    },
    {
      "epoch": 0.1798,
      "grad_norm": 0.0756530612707138,
      "learning_rate": 0.002,
      "loss": 2.024066925048828,
      "step": 1798
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.07706550508737564,
      "learning_rate": 0.002,
      "loss": 2.0181081295013428,
      "step": 1800
    },
    {
      "epoch": 0.1802,
      "grad_norm": 0.07209282368421555,
      "learning_rate": 0.002,
      "loss": 1.9840136766433716,
      "step": 1802
    },
    {
      "epoch": 0.1804,
      "grad_norm": 0.07570994645357132,
      "learning_rate": 0.002,
      "loss": 1.8836725950241089,
      "step": 1804
    },
    {
      "epoch": 0.1806,
      "grad_norm": 0.07015760987997055,
      "learning_rate": 0.002,
      "loss": 2.0593113899230957,
      "step": 1806
    },
    {
      "epoch": 0.1808,
      "grad_norm": 0.07654726505279541,
      "learning_rate": 0.002,
      "loss": 2.0773136615753174,
      "step": 1808
    },
    {
      "epoch": 0.181,
      "grad_norm": 0.07446228712797165,
      "learning_rate": 0.002,
      "loss": 2.0103225708007812,
      "step": 1810
    },
    {
      "epoch": 0.1812,
      "grad_norm": 0.07884091883897781,
      "learning_rate": 0.002,
      "loss": 1.8713654279708862,
      "step": 1812
    },
    {
      "epoch": 0.1814,
      "grad_norm": 0.07967796176671982,
      "learning_rate": 0.002,
      "loss": 1.9373207092285156,
      "step": 1814
    },
    {
      "epoch": 0.1816,
      "grad_norm": 0.0806257426738739,
      "learning_rate": 0.002,
      "loss": 1.9032498598098755,
      "step": 1816
    },
    {
      "epoch": 0.1818,
      "grad_norm": 0.07404523342847824,
      "learning_rate": 0.002,
      "loss": 2.0094704627990723,
      "step": 1818
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.07477964460849762,
      "learning_rate": 0.002,
      "loss": 2.024261951446533,
      "step": 1820
    },
    {
      "epoch": 0.1822,
      "grad_norm": 0.07229091972112656,
      "learning_rate": 0.002,
      "loss": 1.9480940103530884,
      "step": 1822
    },
    {
      "epoch": 0.1824,
      "grad_norm": 0.08243381977081299,
      "learning_rate": 0.002,
      "loss": 2.14764142036438,
      "step": 1824
    },
    {
      "epoch": 0.1826,
      "grad_norm": 0.07408121228218079,
      "learning_rate": 0.002,
      "loss": 2.0946781635284424,
      "step": 1826
    },
    {
      "epoch": 0.1828,
      "grad_norm": 0.07438065111637115,
      "learning_rate": 0.002,
      "loss": 2.069563627243042,
      "step": 1828
    },
    {
      "epoch": 0.183,
      "grad_norm": 0.07345041632652283,
      "learning_rate": 0.002,
      "loss": 1.961599588394165,
      "step": 1830
    },
    {
      "epoch": 0.1832,
      "grad_norm": 0.07595646381378174,
      "learning_rate": 0.002,
      "loss": 1.9593992233276367,
      "step": 1832
    },
    {
      "epoch": 0.1834,
      "grad_norm": 0.0715472549200058,
      "learning_rate": 0.002,
      "loss": 1.9448604583740234,
      "step": 1834
    },
    {
      "epoch": 0.1836,
      "grad_norm": 0.07439612597227097,
      "learning_rate": 0.002,
      "loss": 1.8884066343307495,
      "step": 1836
    },
    {
      "epoch": 0.1838,
      "grad_norm": 0.08087381720542908,
      "learning_rate": 0.002,
      "loss": 2.1303839683532715,
      "step": 1838
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.07926774024963379,
      "learning_rate": 0.002,
      "loss": 2.06335711479187,
      "step": 1840
    },
    {
      "epoch": 0.1842,
      "grad_norm": 0.07622934132814407,
      "learning_rate": 0.002,
      "loss": 2.016192674636841,
      "step": 1842
    },
    {
      "epoch": 0.1844,
      "grad_norm": 0.07247833907604218,
      "learning_rate": 0.002,
      "loss": 1.8543353080749512,
      "step": 1844
    },
    {
      "epoch": 0.1846,
      "grad_norm": 0.07460576295852661,
      "learning_rate": 0.002,
      "loss": 2.0608930587768555,
      "step": 1846
    },
    {
      "epoch": 0.1848,
      "grad_norm": 0.07780230790376663,
      "learning_rate": 0.002,
      "loss": 2.085209846496582,
      "step": 1848
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.07585702836513519,
      "learning_rate": 0.002,
      "loss": 1.9441457986831665,
      "step": 1850
    },
    {
      "epoch": 0.1852,
      "grad_norm": 0.07667966187000275,
      "learning_rate": 0.002,
      "loss": 1.9441406726837158,
      "step": 1852
    },
    {
      "epoch": 0.1854,
      "grad_norm": 0.0768464133143425,
      "learning_rate": 0.002,
      "loss": 1.9836654663085938,
      "step": 1854
    },
    {
      "epoch": 0.1856,
      "grad_norm": 0.07123462110757828,
      "learning_rate": 0.002,
      "loss": 1.904010534286499,
      "step": 1856
    },
    {
      "epoch": 0.1858,
      "grad_norm": 0.07811989635229111,
      "learning_rate": 0.002,
      "loss": 2.0558528900146484,
      "step": 1858
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.08032301068305969,
      "learning_rate": 0.002,
      "loss": 2.1188182830810547,
      "step": 1860
    },
    {
      "epoch": 0.1862,
      "grad_norm": 0.07770902663469315,
      "learning_rate": 0.002,
      "loss": 2.01084566116333,
      "step": 1862
    },
    {
      "epoch": 0.1864,
      "grad_norm": 0.07392118126153946,
      "learning_rate": 0.002,
      "loss": 2.0825531482696533,
      "step": 1864
    },
    {
      "epoch": 0.1866,
      "grad_norm": 0.07660733163356781,
      "learning_rate": 0.002,
      "loss": 2.021610975265503,
      "step": 1866
    },
    {
      "epoch": 0.1868,
      "grad_norm": 0.07611685246229172,
      "learning_rate": 0.002,
      "loss": 2.014993667602539,
      "step": 1868
    },
    {
      "epoch": 0.187,
      "grad_norm": 0.07406169921159744,
      "learning_rate": 0.002,
      "loss": 2.0350613594055176,
      "step": 1870
    },
    {
      "epoch": 0.1872,
      "grad_norm": 0.0726403146982193,
      "learning_rate": 0.002,
      "loss": 2.031841993331909,
      "step": 1872
    },
    {
      "epoch": 0.1874,
      "grad_norm": 0.07360552996397018,
      "learning_rate": 0.002,
      "loss": 2.0633163452148438,
      "step": 1874
    },
    {
      "epoch": 0.1876,
      "grad_norm": 0.07667297124862671,
      "learning_rate": 0.002,
      "loss": 1.9026193618774414,
      "step": 1876
    },
    {
      "epoch": 0.1878,
      "grad_norm": 0.07344851642847061,
      "learning_rate": 0.002,
      "loss": 1.978562355041504,
      "step": 1878
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.07437446713447571,
      "learning_rate": 0.002,
      "loss": 2.034649133682251,
      "step": 1880
    },
    {
      "epoch": 0.1882,
      "grad_norm": 0.07825596630573273,
      "learning_rate": 0.002,
      "loss": 1.9774253368377686,
      "step": 1882
    },
    {
      "epoch": 0.1884,
      "grad_norm": 0.07880914211273193,
      "learning_rate": 0.002,
      "loss": 1.9826339483261108,
      "step": 1884
    },
    {
      "epoch": 0.1886,
      "grad_norm": 0.07801936566829681,
      "learning_rate": 0.002,
      "loss": 1.8488340377807617,
      "step": 1886
    },
    {
      "epoch": 0.1888,
      "grad_norm": 0.07853774726390839,
      "learning_rate": 0.002,
      "loss": 1.9275364875793457,
      "step": 1888
    },
    {
      "epoch": 0.189,
      "grad_norm": 0.07792159914970398,
      "learning_rate": 0.002,
      "loss": 2.1060097217559814,
      "step": 1890
    },
    {
      "epoch": 0.1892,
      "grad_norm": 0.07848257571458817,
      "learning_rate": 0.002,
      "loss": 2.0415639877319336,
      "step": 1892
    },
    {
      "epoch": 0.1894,
      "grad_norm": 0.07632971554994583,
      "learning_rate": 0.002,
      "loss": 1.9622911214828491,
      "step": 1894
    },
    {
      "epoch": 0.1896,
      "grad_norm": 0.07394014298915863,
      "learning_rate": 0.002,
      "loss": 1.8918399810791016,
      "step": 1896
    },
    {
      "epoch": 0.1898,
      "grad_norm": 0.07834768295288086,
      "learning_rate": 0.002,
      "loss": 1.9275469779968262,
      "step": 1898
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.07229535281658173,
      "learning_rate": 0.002,
      "loss": 2.064926862716675,
      "step": 1900
    },
    {
      "epoch": 0.1902,
      "grad_norm": 0.07392053306102753,
      "learning_rate": 0.002,
      "loss": 2.035594940185547,
      "step": 1902
    },
    {
      "epoch": 0.1904,
      "grad_norm": 0.07954246550798416,
      "learning_rate": 0.002,
      "loss": 1.9765669107437134,
      "step": 1904
    },
    {
      "epoch": 0.1906,
      "grad_norm": 0.07570556551218033,
      "learning_rate": 0.002,
      "loss": 1.9965779781341553,
      "step": 1906
    },
    {
      "epoch": 0.1908,
      "grad_norm": 0.07304073125123978,
      "learning_rate": 0.002,
      "loss": 1.9209762811660767,
      "step": 1908
    },
    {
      "epoch": 0.191,
      "grad_norm": 0.07383333891630173,
      "learning_rate": 0.002,
      "loss": 2.108366012573242,
      "step": 1910
    },
    {
      "epoch": 0.1912,
      "grad_norm": 0.07539542019367218,
      "learning_rate": 0.002,
      "loss": 2.0551326274871826,
      "step": 1912
    },
    {
      "epoch": 0.1914,
      "grad_norm": 0.07704858481884003,
      "learning_rate": 0.002,
      "loss": 2.077397108078003,
      "step": 1914
    },
    {
      "epoch": 0.1916,
      "grad_norm": 0.07218719273805618,
      "learning_rate": 0.002,
      "loss": 2.1298329830169678,
      "step": 1916
    },
    {
      "epoch": 0.1918,
      "grad_norm": 0.07486242800951004,
      "learning_rate": 0.002,
      "loss": 2.0625686645507812,
      "step": 1918
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.07390429824590683,
      "learning_rate": 0.002,
      "loss": 1.9201794862747192,
      "step": 1920
    },
    {
      "epoch": 0.1922,
      "grad_norm": 0.07180503010749817,
      "learning_rate": 0.002,
      "loss": 1.895329236984253,
      "step": 1922
    },
    {
      "epoch": 0.1924,
      "grad_norm": 0.0733010470867157,
      "learning_rate": 0.002,
      "loss": 2.012186288833618,
      "step": 1924
    },
    {
      "epoch": 0.1926,
      "grad_norm": 0.0757322758436203,
      "learning_rate": 0.002,
      "loss": 1.9342530965805054,
      "step": 1926
    },
    {
      "epoch": 0.1928,
      "grad_norm": 0.07954363524913788,
      "learning_rate": 0.002,
      "loss": 2.0458085536956787,
      "step": 1928
    },
    {
      "epoch": 0.193,
      "grad_norm": 0.08379711955785751,
      "learning_rate": 0.002,
      "loss": 1.9098501205444336,
      "step": 1930
    },
    {
      "epoch": 0.1932,
      "grad_norm": 0.0746661052107811,
      "learning_rate": 0.002,
      "loss": 1.9700909852981567,
      "step": 1932
    },
    {
      "epoch": 0.1934,
      "grad_norm": 0.07700544595718384,
      "learning_rate": 0.002,
      "loss": 2.0355162620544434,
      "step": 1934
    },
    {
      "epoch": 0.1936,
      "grad_norm": 0.08141306787729263,
      "learning_rate": 0.002,
      "loss": 2.131028652191162,
      "step": 1936
    },
    {
      "epoch": 0.1938,
      "grad_norm": 0.074761301279068,
      "learning_rate": 0.002,
      "loss": 1.9921268224716187,
      "step": 1938
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.07471061497926712,
      "learning_rate": 0.002,
      "loss": 2.14154052734375,
      "step": 1940
    },
    {
      "epoch": 0.1942,
      "grad_norm": 0.07418189197778702,
      "learning_rate": 0.002,
      "loss": 2.0176968574523926,
      "step": 1942
    },
    {
      "epoch": 0.1944,
      "grad_norm": 0.07647217810153961,
      "learning_rate": 0.002,
      "loss": 2.0676517486572266,
      "step": 1944
    },
    {
      "epoch": 0.1946,
      "grad_norm": 0.07374267280101776,
      "learning_rate": 0.002,
      "loss": 1.9794468879699707,
      "step": 1946
    },
    {
      "epoch": 0.1948,
      "grad_norm": 0.08480340987443924,
      "learning_rate": 0.002,
      "loss": 2.0614590644836426,
      "step": 1948
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.07298968732357025,
      "learning_rate": 0.002,
      "loss": 2.0372188091278076,
      "step": 1950
    },
    {
      "epoch": 0.1952,
      "grad_norm": 0.0747591182589531,
      "learning_rate": 0.002,
      "loss": 2.002885580062866,
      "step": 1952
    },
    {
      "epoch": 0.1954,
      "grad_norm": 0.07806036621332169,
      "learning_rate": 0.002,
      "loss": 2.019678831100464,
      "step": 1954
    },
    {
      "epoch": 0.1956,
      "grad_norm": 0.08047091960906982,
      "learning_rate": 0.002,
      "loss": 2.0275285243988037,
      "step": 1956
    },
    {
      "epoch": 0.1958,
      "grad_norm": 0.07517793029546738,
      "learning_rate": 0.002,
      "loss": 2.021721839904785,
      "step": 1958
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.07783924788236618,
      "learning_rate": 0.002,
      "loss": 2.012787342071533,
      "step": 1960
    },
    {
      "epoch": 0.1962,
      "grad_norm": 0.07707510888576508,
      "learning_rate": 0.002,
      "loss": 1.921724557876587,
      "step": 1962
    },
    {
      "epoch": 0.1964,
      "grad_norm": 0.07263901084661484,
      "learning_rate": 0.002,
      "loss": 2.0467591285705566,
      "step": 1964
    },
    {
      "epoch": 0.1966,
      "grad_norm": 0.07490444928407669,
      "learning_rate": 0.002,
      "loss": 2.0225110054016113,
      "step": 1966
    },
    {
      "epoch": 0.1968,
      "grad_norm": 0.08013395965099335,
      "learning_rate": 0.002,
      "loss": 2.068084716796875,
      "step": 1968
    },
    {
      "epoch": 0.197,
      "grad_norm": 0.07863252609968185,
      "learning_rate": 0.002,
      "loss": 2.0182416439056396,
      "step": 1970
    },
    {
      "epoch": 0.1972,
      "grad_norm": 0.07227256894111633,
      "learning_rate": 0.002,
      "loss": 1.9799894094467163,
      "step": 1972
    },
    {
      "epoch": 0.1974,
      "grad_norm": 0.07425327599048615,
      "learning_rate": 0.002,
      "loss": 1.965545892715454,
      "step": 1974
    },
    {
      "epoch": 0.1976,
      "grad_norm": 0.07363994419574738,
      "learning_rate": 0.002,
      "loss": 1.9761097431182861,
      "step": 1976
    },
    {
      "epoch": 0.1978,
      "grad_norm": 0.07401183247566223,
      "learning_rate": 0.002,
      "loss": 2.0797972679138184,
      "step": 1978
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.07827579975128174,
      "learning_rate": 0.002,
      "loss": 1.9820281267166138,
      "step": 1980
    },
    {
      "epoch": 0.1982,
      "grad_norm": 0.07368913292884827,
      "learning_rate": 0.002,
      "loss": 1.9975016117095947,
      "step": 1982
    },
    {
      "epoch": 0.1984,
      "grad_norm": 0.07454439997673035,
      "learning_rate": 0.002,
      "loss": 1.9389212131500244,
      "step": 1984
    },
    {
      "epoch": 0.1986,
      "grad_norm": 0.09229312092065811,
      "learning_rate": 0.002,
      "loss": 1.9356615543365479,
      "step": 1986
    },
    {
      "epoch": 0.1988,
      "grad_norm": 0.07815497368574142,
      "learning_rate": 0.002,
      "loss": 2.0105371475219727,
      "step": 1988
    },
    {
      "epoch": 0.199,
      "grad_norm": 0.07200359553098679,
      "learning_rate": 0.002,
      "loss": 2.014775514602661,
      "step": 1990
    },
    {
      "epoch": 0.1992,
      "grad_norm": 0.07343850284814835,
      "learning_rate": 0.002,
      "loss": 2.0139901638031006,
      "step": 1992
    },
    {
      "epoch": 0.1994,
      "grad_norm": 0.07420206069946289,
      "learning_rate": 0.002,
      "loss": 2.1029412746429443,
      "step": 1994
    },
    {
      "epoch": 0.1996,
      "grad_norm": 0.08003529161214828,
      "learning_rate": 0.002,
      "loss": 1.8853633403778076,
      "step": 1996
    },
    {
      "epoch": 0.1998,
      "grad_norm": 0.07636172324419022,
      "learning_rate": 0.002,
      "loss": 2.0144386291503906,
      "step": 1998
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.07471030950546265,
      "learning_rate": 0.002,
      "loss": 1.9643774032592773,
      "step": 2000
    },
    {
      "epoch": 0.2,
      "eval_loss": 2.186317205429077,
      "eval_runtime": 247.8595,
      "eval_samples_per_second": 2.017,
      "eval_steps_per_second": 0.202,
      "step": 2000
    },
    {
      "epoch": 0.2002,
      "grad_norm": 0.07861803472042084,
      "learning_rate": 0.002,
      "loss": 2.0866823196411133,
      "step": 2002
    },
    {
      "epoch": 0.2004,
      "grad_norm": 0.07268927991390228,
      "learning_rate": 0.002,
      "loss": 1.9871134757995605,
      "step": 2004
    },
    {
      "epoch": 0.2006,
      "grad_norm": 0.07251483201980591,
      "learning_rate": 0.002,
      "loss": 2.0375821590423584,
      "step": 2006
    },
    {
      "epoch": 0.2008,
      "grad_norm": 0.0814223363995552,
      "learning_rate": 0.002,
      "loss": 2.124399423599243,
      "step": 2008
    },
    {
      "epoch": 0.201,
      "grad_norm": 0.0764809250831604,
      "learning_rate": 0.002,
      "loss": 1.9967992305755615,
      "step": 2010
    },
    {
      "epoch": 0.2012,
      "grad_norm": 0.08138150721788406,
      "learning_rate": 0.002,
      "loss": 1.9746674299240112,
      "step": 2012
    },
    {
      "epoch": 0.2014,
      "grad_norm": 0.07659666240215302,
      "learning_rate": 0.002,
      "loss": 2.035062789916992,
      "step": 2014
    },
    {
      "epoch": 0.2016,
      "grad_norm": 0.07549983263015747,
      "learning_rate": 0.002,
      "loss": 1.9982187747955322,
      "step": 2016
    },
    {
      "epoch": 0.2018,
      "grad_norm": 0.06919535249471664,
      "learning_rate": 0.002,
      "loss": 1.7906666994094849,
      "step": 2018
    },
    {
      "epoch": 0.202,
      "grad_norm": 0.07565700262784958,
      "learning_rate": 0.002,
      "loss": 2.0848805904388428,
      "step": 2020
    },
    {
      "epoch": 0.2022,
      "grad_norm": 0.08306537568569183,
      "learning_rate": 0.002,
      "loss": 1.9999656677246094,
      "step": 2022
    },
    {
      "epoch": 0.2024,
      "grad_norm": 0.07358775287866592,
      "learning_rate": 0.002,
      "loss": 1.8815218210220337,
      "step": 2024
    },
    {
      "epoch": 0.2026,
      "grad_norm": 0.07578912377357483,
      "learning_rate": 0.002,
      "loss": 2.056061267852783,
      "step": 2026
    },
    {
      "epoch": 0.2028,
      "grad_norm": 0.0763857513666153,
      "learning_rate": 0.002,
      "loss": 2.0558993816375732,
      "step": 2028
    },
    {
      "epoch": 0.203,
      "grad_norm": 0.08184877038002014,
      "learning_rate": 0.002,
      "loss": 2.094294548034668,
      "step": 2030
    },
    {
      "epoch": 0.2032,
      "grad_norm": 0.07571811974048615,
      "learning_rate": 0.002,
      "loss": 1.9090760946273804,
      "step": 2032
    },
    {
      "epoch": 0.2034,
      "grad_norm": 0.0753084123134613,
      "learning_rate": 0.002,
      "loss": 2.026407480239868,
      "step": 2034
    },
    {
      "epoch": 0.2036,
      "grad_norm": 0.07484840601682663,
      "learning_rate": 0.002,
      "loss": 1.9780300855636597,
      "step": 2036
    },
    {
      "epoch": 0.2038,
      "grad_norm": 0.07525637000799179,
      "learning_rate": 0.002,
      "loss": 2.023068428039551,
      "step": 2038
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.07598625123500824,
      "learning_rate": 0.002,
      "loss": 1.9718419313430786,
      "step": 2040
    },
    {
      "epoch": 0.2042,
      "grad_norm": 0.07291676104068756,
      "learning_rate": 0.002,
      "loss": 2.0320305824279785,
      "step": 2042
    },
    {
      "epoch": 0.2044,
      "grad_norm": 0.07537215948104858,
      "learning_rate": 0.002,
      "loss": 2.076399803161621,
      "step": 2044
    },
    {
      "epoch": 0.2046,
      "grad_norm": 0.07255725562572479,
      "learning_rate": 0.002,
      "loss": 2.0759201049804688,
      "step": 2046
    },
    {
      "epoch": 0.2048,
      "grad_norm": 0.07773645222187042,
      "learning_rate": 0.002,
      "loss": 1.8930901288986206,
      "step": 2048
    },
    {
      "epoch": 0.205,
      "grad_norm": 0.07848163694143295,
      "learning_rate": 0.002,
      "loss": 2.029914379119873,
      "step": 2050
    },
    {
      "epoch": 0.2052,
      "grad_norm": 0.0765446200966835,
      "learning_rate": 0.002,
      "loss": 1.949304461479187,
      "step": 2052
    },
    {
      "epoch": 0.2054,
      "grad_norm": 0.07454271614551544,
      "learning_rate": 0.002,
      "loss": 2.0348479747772217,
      "step": 2054
    },
    {
      "epoch": 0.2056,
      "grad_norm": 0.07629027962684631,
      "learning_rate": 0.002,
      "loss": 1.987667202949524,
      "step": 2056
    },
    {
      "epoch": 0.2058,
      "grad_norm": 0.07569247484207153,
      "learning_rate": 0.002,
      "loss": 2.0650038719177246,
      "step": 2058
    },
    {
      "epoch": 0.206,
      "grad_norm": 0.07213758677244186,
      "learning_rate": 0.002,
      "loss": 1.9508464336395264,
      "step": 2060
    },
    {
      "epoch": 0.2062,
      "grad_norm": 0.07172496616840363,
      "learning_rate": 0.002,
      "loss": 1.842262625694275,
      "step": 2062
    },
    {
      "epoch": 0.2064,
      "grad_norm": 0.07516039907932281,
      "learning_rate": 0.002,
      "loss": 1.9482581615447998,
      "step": 2064
    },
    {
      "epoch": 0.2066,
      "grad_norm": 0.0802142545580864,
      "learning_rate": 0.002,
      "loss": 1.939031720161438,
      "step": 2066
    },
    {
      "epoch": 0.2068,
      "grad_norm": 0.0735301747918129,
      "learning_rate": 0.002,
      "loss": 2.1084163188934326,
      "step": 2068
    },
    {
      "epoch": 0.207,
      "grad_norm": 0.07555841654539108,
      "learning_rate": 0.002,
      "loss": 2.0543370246887207,
      "step": 2070
    },
    {
      "epoch": 0.2072,
      "grad_norm": 0.07104679197072983,
      "learning_rate": 0.002,
      "loss": 1.9315857887268066,
      "step": 2072
    },
    {
      "epoch": 0.2074,
      "grad_norm": 0.07217837870121002,
      "learning_rate": 0.002,
      "loss": 1.9535218477249146,
      "step": 2074
    },
    {
      "epoch": 0.2076,
      "grad_norm": 0.07526152580976486,
      "learning_rate": 0.002,
      "loss": 1.7873817682266235,
      "step": 2076
    },
    {
      "epoch": 0.2078,
      "grad_norm": 0.07162679731845856,
      "learning_rate": 0.002,
      "loss": 1.9246410131454468,
      "step": 2078
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.07787878066301346,
      "learning_rate": 0.002,
      "loss": 2.0484447479248047,
      "step": 2080
    },
    {
      "epoch": 0.2082,
      "grad_norm": 0.07335934042930603,
      "learning_rate": 0.002,
      "loss": 1.9963961839675903,
      "step": 2082
    },
    {
      "epoch": 0.2084,
      "grad_norm": 0.0763859748840332,
      "learning_rate": 0.002,
      "loss": 2.079028606414795,
      "step": 2084
    },
    {
      "epoch": 0.2086,
      "grad_norm": 0.07674536108970642,
      "learning_rate": 0.002,
      "loss": 1.9238855838775635,
      "step": 2086
    },
    {
      "epoch": 0.2088,
      "grad_norm": 0.07329664379358292,
      "learning_rate": 0.002,
      "loss": 1.8596769571304321,
      "step": 2088
    },
    {
      "epoch": 0.209,
      "grad_norm": 0.07449709624052048,
      "learning_rate": 0.002,
      "loss": 1.9207953214645386,
      "step": 2090
    },
    {
      "epoch": 0.2092,
      "grad_norm": 0.07537084817886353,
      "learning_rate": 0.002,
      "loss": 2.042569398880005,
      "step": 2092
    },
    {
      "epoch": 0.2094,
      "grad_norm": 0.07401484996080399,
      "learning_rate": 0.002,
      "loss": 1.9976391792297363,
      "step": 2094
    },
    {
      "epoch": 0.2096,
      "grad_norm": 0.074497751891613,
      "learning_rate": 0.002,
      "loss": 2.037662982940674,
      "step": 2096
    },
    {
      "epoch": 0.2098,
      "grad_norm": 0.07660408318042755,
      "learning_rate": 0.002,
      "loss": 1.9321916103363037,
      "step": 2098
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.07092493772506714,
      "learning_rate": 0.002,
      "loss": 2.0278117656707764,
      "step": 2100
    },
    {
      "epoch": 0.2102,
      "grad_norm": 0.07291091978549957,
      "learning_rate": 0.002,
      "loss": 2.0196456909179688,
      "step": 2102
    },
    {
      "epoch": 0.2104,
      "grad_norm": 0.0762559100985527,
      "learning_rate": 0.002,
      "loss": 2.095963478088379,
      "step": 2104
    },
    {
      "epoch": 0.2106,
      "grad_norm": 0.08740567415952682,
      "learning_rate": 0.002,
      "loss": 1.862408995628357,
      "step": 2106
    },
    {
      "epoch": 0.2108,
      "grad_norm": 0.07596364617347717,
      "learning_rate": 0.002,
      "loss": 1.8178695440292358,
      "step": 2108
    },
    {
      "epoch": 0.211,
      "grad_norm": 0.0868581235408783,
      "learning_rate": 0.002,
      "loss": 1.9396828413009644,
      "step": 2110
    },
    {
      "epoch": 0.2112,
      "grad_norm": 0.07573510706424713,
      "learning_rate": 0.002,
      "loss": 1.8935412168502808,
      "step": 2112
    },
    {
      "epoch": 0.2114,
      "grad_norm": 0.07743333280086517,
      "learning_rate": 0.002,
      "loss": 1.9438068866729736,
      "step": 2114
    },
    {
      "epoch": 0.2116,
      "grad_norm": 0.07681100070476532,
      "learning_rate": 0.002,
      "loss": 1.7895441055297852,
      "step": 2116
    },
    {
      "epoch": 0.2118,
      "grad_norm": 0.07565531879663467,
      "learning_rate": 0.002,
      "loss": 2.186378002166748,
      "step": 2118
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.07924189418554306,
      "learning_rate": 0.002,
      "loss": 2.069779872894287,
      "step": 2120
    },
    {
      "epoch": 0.2122,
      "grad_norm": 0.07607977837324142,
      "learning_rate": 0.002,
      "loss": 2.001291513442993,
      "step": 2122
    },
    {
      "epoch": 0.2124,
      "grad_norm": 0.07786546647548676,
      "learning_rate": 0.002,
      "loss": 1.956085205078125,
      "step": 2124
    },
    {
      "epoch": 0.2126,
      "grad_norm": 0.07097860425710678,
      "learning_rate": 0.002,
      "loss": 1.8607815504074097,
      "step": 2126
    },
    {
      "epoch": 0.2128,
      "grad_norm": 0.06988345831632614,
      "learning_rate": 0.002,
      "loss": 2.0171728134155273,
      "step": 2128
    },
    {
      "epoch": 0.213,
      "grad_norm": 0.07246838510036469,
      "learning_rate": 0.002,
      "loss": 1.973251461982727,
      "step": 2130
    },
    {
      "epoch": 0.2132,
      "grad_norm": 0.07507634162902832,
      "learning_rate": 0.002,
      "loss": 2.0482473373413086,
      "step": 2132
    },
    {
      "epoch": 0.2134,
      "grad_norm": 0.07542074471712112,
      "learning_rate": 0.002,
      "loss": 1.9785593748092651,
      "step": 2134
    },
    {
      "epoch": 0.2136,
      "grad_norm": 0.08195478469133377,
      "learning_rate": 0.002,
      "loss": 1.898648977279663,
      "step": 2136
    },
    {
      "epoch": 0.2138,
      "grad_norm": 0.07409955561161041,
      "learning_rate": 0.002,
      "loss": 2.0151731967926025,
      "step": 2138
    },
    {
      "epoch": 0.214,
      "grad_norm": 0.07747111469507217,
      "learning_rate": 0.002,
      "loss": 2.0404038429260254,
      "step": 2140
    },
    {
      "epoch": 0.2142,
      "grad_norm": 0.08119469881057739,
      "learning_rate": 0.002,
      "loss": 1.9117904901504517,
      "step": 2142
    },
    {
      "epoch": 0.2144,
      "grad_norm": 0.07472380995750427,
      "learning_rate": 0.002,
      "loss": 1.910109043121338,
      "step": 2144
    },
    {
      "epoch": 0.2146,
      "grad_norm": 0.07511576265096664,
      "learning_rate": 0.002,
      "loss": 2.0228190422058105,
      "step": 2146
    },
    {
      "epoch": 0.2148,
      "grad_norm": 0.07471606880426407,
      "learning_rate": 0.002,
      "loss": 1.9324160814285278,
      "step": 2148
    },
    {
      "epoch": 0.215,
      "grad_norm": 0.09552984684705734,
      "learning_rate": 0.002,
      "loss": 1.9881231784820557,
      "step": 2150
    },
    {
      "epoch": 0.2152,
      "grad_norm": 0.07472539693117142,
      "learning_rate": 0.002,
      "loss": 2.0094480514526367,
      "step": 2152
    },
    {
      "epoch": 0.2154,
      "grad_norm": 0.07474126666784286,
      "learning_rate": 0.002,
      "loss": 2.0636322498321533,
      "step": 2154
    },
    {
      "epoch": 0.2156,
      "grad_norm": 0.07406353950500488,
      "learning_rate": 0.002,
      "loss": 1.9452698230743408,
      "step": 2156
    },
    {
      "epoch": 0.2158,
      "grad_norm": 0.07532574236392975,
      "learning_rate": 0.002,
      "loss": 2.052245855331421,
      "step": 2158
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.077007956802845,
      "learning_rate": 0.002,
      "loss": 2.1642234325408936,
      "step": 2160
    },
    {
      "epoch": 0.2162,
      "grad_norm": 0.07090206444263458,
      "learning_rate": 0.002,
      "loss": 2.0291690826416016,
      "step": 2162
    },
    {
      "epoch": 0.2164,
      "grad_norm": 0.07478799670934677,
      "learning_rate": 0.002,
      "loss": 2.0103042125701904,
      "step": 2164
    },
    {
      "epoch": 0.2166,
      "grad_norm": 0.07721041887998581,
      "learning_rate": 0.002,
      "loss": 1.9661140441894531,
      "step": 2166
    },
    {
      "epoch": 0.2168,
      "grad_norm": 0.07537022233009338,
      "learning_rate": 0.002,
      "loss": 1.9182689189910889,
      "step": 2168
    },
    {
      "epoch": 0.217,
      "grad_norm": 0.0733344703912735,
      "learning_rate": 0.002,
      "loss": 1.918317198753357,
      "step": 2170
    },
    {
      "epoch": 0.2172,
      "grad_norm": 0.07329059392213821,
      "learning_rate": 0.002,
      "loss": 1.9735743999481201,
      "step": 2172
    },
    {
      "epoch": 0.2174,
      "grad_norm": 0.07898969948291779,
      "learning_rate": 0.002,
      "loss": 2.0076992511749268,
      "step": 2174
    },
    {
      "epoch": 0.2176,
      "grad_norm": 0.0751541405916214,
      "learning_rate": 0.002,
      "loss": 2.104896306991577,
      "step": 2176
    },
    {
      "epoch": 0.2178,
      "grad_norm": 0.07676255702972412,
      "learning_rate": 0.002,
      "loss": 1.9784101247787476,
      "step": 2178
    },
    {
      "epoch": 0.218,
      "grad_norm": 0.07694292813539505,
      "learning_rate": 0.002,
      "loss": 1.9321246147155762,
      "step": 2180
    },
    {
      "epoch": 0.2182,
      "grad_norm": 0.07703717052936554,
      "learning_rate": 0.002,
      "loss": 2.063851833343506,
      "step": 2182
    },
    {
      "epoch": 0.2184,
      "grad_norm": 0.07561987638473511,
      "learning_rate": 0.002,
      "loss": 1.9831727743148804,
      "step": 2184
    },
    {
      "epoch": 0.2186,
      "grad_norm": 0.07164116203784943,
      "learning_rate": 0.002,
      "loss": 1.951820969581604,
      "step": 2186
    },
    {
      "epoch": 0.2188,
      "grad_norm": 0.0757484957575798,
      "learning_rate": 0.002,
      "loss": 2.0753118991851807,
      "step": 2188
    },
    {
      "epoch": 0.219,
      "grad_norm": 0.0737743005156517,
      "learning_rate": 0.002,
      "loss": 1.9435093402862549,
      "step": 2190
    },
    {
      "epoch": 0.2192,
      "grad_norm": 0.07735036313533783,
      "learning_rate": 0.002,
      "loss": 1.9307893514633179,
      "step": 2192
    },
    {
      "epoch": 0.2194,
      "grad_norm": 0.0754513368010521,
      "learning_rate": 0.002,
      "loss": 1.880855679512024,
      "step": 2194
    },
    {
      "epoch": 0.2196,
      "grad_norm": 0.0711849182844162,
      "learning_rate": 0.002,
      "loss": 1.8617995977401733,
      "step": 2196
    },
    {
      "epoch": 0.2198,
      "grad_norm": 0.07447977364063263,
      "learning_rate": 0.002,
      "loss": 1.8668371438980103,
      "step": 2198
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.07517198473215103,
      "learning_rate": 0.002,
      "loss": 1.8173487186431885,
      "step": 2200
    },
    {
      "epoch": 0.2202,
      "grad_norm": 0.07738427072763443,
      "learning_rate": 0.002,
      "loss": 2.067300796508789,
      "step": 2202
    },
    {
      "epoch": 0.2204,
      "grad_norm": 0.07278750836849213,
      "learning_rate": 0.002,
      "loss": 1.9065617322921753,
      "step": 2204
    },
    {
      "epoch": 0.2206,
      "grad_norm": 0.07601485401391983,
      "learning_rate": 0.002,
      "loss": 2.0243122577667236,
      "step": 2206
    },
    {
      "epoch": 0.2208,
      "grad_norm": 0.07119227945804596,
      "learning_rate": 0.002,
      "loss": 1.8461320400238037,
      "step": 2208
    },
    {
      "epoch": 0.221,
      "grad_norm": 0.07678936421871185,
      "learning_rate": 0.002,
      "loss": 2.0037336349487305,
      "step": 2210
    },
    {
      "epoch": 0.2212,
      "grad_norm": 0.08074525743722916,
      "learning_rate": 0.002,
      "loss": 1.8309258222579956,
      "step": 2212
    },
    {
      "epoch": 0.2214,
      "grad_norm": 0.07949401438236237,
      "learning_rate": 0.002,
      "loss": 1.8736461400985718,
      "step": 2214
    },
    {
      "epoch": 0.2216,
      "grad_norm": 0.07744450122117996,
      "learning_rate": 0.002,
      "loss": 1.9620195627212524,
      "step": 2216
    },
    {
      "epoch": 0.2218,
      "grad_norm": 0.0773572027683258,
      "learning_rate": 0.002,
      "loss": 2.0777947902679443,
      "step": 2218
    },
    {
      "epoch": 0.222,
      "grad_norm": 0.07511971890926361,
      "learning_rate": 0.002,
      "loss": 1.8780888319015503,
      "step": 2220
    },
    {
      "epoch": 0.2222,
      "grad_norm": 0.08079022914171219,
      "learning_rate": 0.002,
      "loss": 1.9882444143295288,
      "step": 2222
    },
    {
      "epoch": 0.2224,
      "grad_norm": 0.07889550924301147,
      "learning_rate": 0.002,
      "loss": 2.0892727375030518,
      "step": 2224
    },
    {
      "epoch": 0.2226,
      "grad_norm": 0.0785905122756958,
      "learning_rate": 0.002,
      "loss": 1.9932130575180054,
      "step": 2226
    },
    {
      "epoch": 0.2228,
      "grad_norm": 0.0837470144033432,
      "learning_rate": 0.002,
      "loss": 1.9433176517486572,
      "step": 2228
    },
    {
      "epoch": 0.223,
      "grad_norm": 0.07522817701101303,
      "learning_rate": 0.002,
      "loss": 1.913541555404663,
      "step": 2230
    },
    {
      "epoch": 0.2232,
      "grad_norm": 0.07642986625432968,
      "learning_rate": 0.002,
      "loss": 2.026679515838623,
      "step": 2232
    },
    {
      "epoch": 0.2234,
      "grad_norm": 0.07749960571527481,
      "learning_rate": 0.002,
      "loss": 2.054225444793701,
      "step": 2234
    },
    {
      "epoch": 0.2236,
      "grad_norm": 0.07664389908313751,
      "learning_rate": 0.002,
      "loss": 1.9671754837036133,
      "step": 2236
    },
    {
      "epoch": 0.2238,
      "grad_norm": 0.07577154785394669,
      "learning_rate": 0.002,
      "loss": 2.0116381645202637,
      "step": 2238
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.07775341719388962,
      "learning_rate": 0.002,
      "loss": 2.0453379154205322,
      "step": 2240
    },
    {
      "epoch": 0.2242,
      "grad_norm": 0.07799158245325089,
      "learning_rate": 0.002,
      "loss": 2.0062379837036133,
      "step": 2242
    },
    {
      "epoch": 0.2244,
      "grad_norm": 0.08256048709154129,
      "learning_rate": 0.002,
      "loss": 2.0883309841156006,
      "step": 2244
    },
    {
      "epoch": 0.2246,
      "grad_norm": 0.07970818132162094,
      "learning_rate": 0.002,
      "loss": 2.069889545440674,
      "step": 2246
    },
    {
      "epoch": 0.2248,
      "grad_norm": 0.08042725175619125,
      "learning_rate": 0.002,
      "loss": 2.0600576400756836,
      "step": 2248
    },
    {
      "epoch": 0.225,
      "grad_norm": 0.07454901188611984,
      "learning_rate": 0.002,
      "loss": 1.9596612453460693,
      "step": 2250
    },
    {
      "epoch": 0.2252,
      "grad_norm": 0.07726215571165085,
      "learning_rate": 0.002,
      "loss": 2.005075216293335,
      "step": 2252
    },
    {
      "epoch": 0.2254,
      "grad_norm": 0.07490281760692596,
      "learning_rate": 0.002,
      "loss": 1.9650015830993652,
      "step": 2254
    },
    {
      "epoch": 0.2256,
      "grad_norm": 0.07915124297142029,
      "learning_rate": 0.002,
      "loss": 1.9398077726364136,
      "step": 2256
    },
    {
      "epoch": 0.2258,
      "grad_norm": 0.07449433952569962,
      "learning_rate": 0.002,
      "loss": 1.9781391620635986,
      "step": 2258
    },
    {
      "epoch": 0.226,
      "grad_norm": 0.07282926887273788,
      "learning_rate": 0.002,
      "loss": 1.9601349830627441,
      "step": 2260
    },
    {
      "epoch": 0.2262,
      "grad_norm": 0.07406370341777802,
      "learning_rate": 0.002,
      "loss": 1.9218815565109253,
      "step": 2262
    },
    {
      "epoch": 0.2264,
      "grad_norm": 0.07565250247716904,
      "learning_rate": 0.002,
      "loss": 1.9847464561462402,
      "step": 2264
    },
    {
      "epoch": 0.2266,
      "grad_norm": 0.07813519984483719,
      "learning_rate": 0.002,
      "loss": 1.918882966041565,
      "step": 2266
    },
    {
      "epoch": 0.2268,
      "grad_norm": 0.07558415830135345,
      "learning_rate": 0.002,
      "loss": 1.9172977209091187,
      "step": 2268
    },
    {
      "epoch": 0.227,
      "grad_norm": 0.07095683366060257,
      "learning_rate": 0.002,
      "loss": 1.9101675748825073,
      "step": 2270
    },
    {
      "epoch": 0.2272,
      "grad_norm": 0.07344143837690353,
      "learning_rate": 0.002,
      "loss": 1.970721960067749,
      "step": 2272
    },
    {
      "epoch": 0.2274,
      "grad_norm": 0.07081401348114014,
      "learning_rate": 0.002,
      "loss": 1.9405829906463623,
      "step": 2274
    },
    {
      "epoch": 0.2276,
      "grad_norm": 0.08006518334150314,
      "learning_rate": 0.002,
      "loss": 1.9865427017211914,
      "step": 2276
    },
    {
      "epoch": 0.2278,
      "grad_norm": 0.07569582760334015,
      "learning_rate": 0.002,
      "loss": 1.8278820514678955,
      "step": 2278
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.07692670077085495,
      "learning_rate": 0.002,
      "loss": 2.0667433738708496,
      "step": 2280
    },
    {
      "epoch": 0.2282,
      "grad_norm": 0.07923755049705505,
      "learning_rate": 0.002,
      "loss": 1.9467494487762451,
      "step": 2282
    },
    {
      "epoch": 0.2284,
      "grad_norm": 0.07373586297035217,
      "learning_rate": 0.002,
      "loss": 1.8843790292739868,
      "step": 2284
    },
    {
      "epoch": 0.2286,
      "grad_norm": 0.07519227266311646,
      "learning_rate": 0.002,
      "loss": 1.8739937543869019,
      "step": 2286
    },
    {
      "epoch": 0.2288,
      "grad_norm": 0.08486291021108627,
      "learning_rate": 0.002,
      "loss": 1.8771511316299438,
      "step": 2288
    },
    {
      "epoch": 0.229,
      "grad_norm": 0.07666758447885513,
      "learning_rate": 0.002,
      "loss": 2.115143060684204,
      "step": 2290
    },
    {
      "epoch": 0.2292,
      "grad_norm": 0.07398266345262527,
      "learning_rate": 0.002,
      "loss": 1.8903745412826538,
      "step": 2292
    },
    {
      "epoch": 0.2294,
      "grad_norm": 0.07641573250293732,
      "learning_rate": 0.002,
      "loss": 2.0381741523742676,
      "step": 2294
    },
    {
      "epoch": 0.2296,
      "grad_norm": 0.07673580944538116,
      "learning_rate": 0.002,
      "loss": 1.8391193151474,
      "step": 2296
    },
    {
      "epoch": 0.2298,
      "grad_norm": 0.07697977125644684,
      "learning_rate": 0.002,
      "loss": 2.0015251636505127,
      "step": 2298
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.07695791125297546,
      "learning_rate": 0.002,
      "loss": 1.8888438940048218,
      "step": 2300
    },
    {
      "epoch": 0.2302,
      "grad_norm": 0.07621645927429199,
      "learning_rate": 0.002,
      "loss": 2.0480406284332275,
      "step": 2302
    },
    {
      "epoch": 0.2304,
      "grad_norm": 0.07893702387809753,
      "learning_rate": 0.002,
      "loss": 2.003004312515259,
      "step": 2304
    },
    {
      "epoch": 0.2306,
      "grad_norm": 0.07317601144313812,
      "learning_rate": 0.002,
      "loss": 1.88134765625,
      "step": 2306
    },
    {
      "epoch": 0.2308,
      "grad_norm": 0.07790163159370422,
      "learning_rate": 0.002,
      "loss": 1.9083185195922852,
      "step": 2308
    },
    {
      "epoch": 0.231,
      "grad_norm": 0.07870550453662872,
      "learning_rate": 0.002,
      "loss": 2.0158581733703613,
      "step": 2310
    },
    {
      "epoch": 0.2312,
      "grad_norm": 0.07873748242855072,
      "learning_rate": 0.002,
      "loss": 1.8908116817474365,
      "step": 2312
    },
    {
      "epoch": 0.2314,
      "grad_norm": 0.07400235533714294,
      "learning_rate": 0.002,
      "loss": 1.8785400390625,
      "step": 2314
    },
    {
      "epoch": 0.2316,
      "grad_norm": 0.0786917582154274,
      "learning_rate": 0.002,
      "loss": 2.0570855140686035,
      "step": 2316
    },
    {
      "epoch": 0.2318,
      "grad_norm": 0.07423048466444016,
      "learning_rate": 0.002,
      "loss": 2.031418800354004,
      "step": 2318
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.07535126060247421,
      "learning_rate": 0.002,
      "loss": 2.0107831954956055,
      "step": 2320
    },
    {
      "epoch": 0.2322,
      "grad_norm": 0.0773010328412056,
      "learning_rate": 0.002,
      "loss": 1.9957424402236938,
      "step": 2322
    },
    {
      "epoch": 0.2324,
      "grad_norm": 0.07451770454645157,
      "learning_rate": 0.002,
      "loss": 1.9823613166809082,
      "step": 2324
    },
    {
      "epoch": 0.2326,
      "grad_norm": 0.07670950144529343,
      "learning_rate": 0.002,
      "loss": 2.1238534450531006,
      "step": 2326
    },
    {
      "epoch": 0.2328,
      "grad_norm": 0.07594184577465057,
      "learning_rate": 0.002,
      "loss": 1.9498236179351807,
      "step": 2328
    },
    {
      "epoch": 0.233,
      "grad_norm": 0.0736960619688034,
      "learning_rate": 0.002,
      "loss": 1.9953234195709229,
      "step": 2330
    },
    {
      "epoch": 0.2332,
      "grad_norm": 0.07324323058128357,
      "learning_rate": 0.002,
      "loss": 1.9514856338500977,
      "step": 2332
    },
    {
      "epoch": 0.2334,
      "grad_norm": 0.07536971569061279,
      "learning_rate": 0.002,
      "loss": 1.8996328115463257,
      "step": 2334
    },
    {
      "epoch": 0.2336,
      "grad_norm": 0.07896479219198227,
      "learning_rate": 0.002,
      "loss": 2.0236830711364746,
      "step": 2336
    },
    {
      "epoch": 0.2338,
      "grad_norm": 0.07607408612966537,
      "learning_rate": 0.002,
      "loss": 1.9519118070602417,
      "step": 2338
    },
    {
      "epoch": 0.234,
      "grad_norm": 0.07971969246864319,
      "learning_rate": 0.002,
      "loss": 2.0647504329681396,
      "step": 2340
    },
    {
      "epoch": 0.2342,
      "grad_norm": 0.07806158065795898,
      "learning_rate": 0.002,
      "loss": 1.8749405145645142,
      "step": 2342
    },
    {
      "epoch": 0.2344,
      "grad_norm": 0.07596602290868759,
      "learning_rate": 0.002,
      "loss": 1.9997894763946533,
      "step": 2344
    },
    {
      "epoch": 0.2346,
      "grad_norm": 0.07636073976755142,
      "learning_rate": 0.002,
      "loss": 1.9572925567626953,
      "step": 2346
    },
    {
      "epoch": 0.2348,
      "grad_norm": 0.08100183308124542,
      "learning_rate": 0.002,
      "loss": 1.9684869050979614,
      "step": 2348
    },
    {
      "epoch": 0.235,
      "grad_norm": 0.07924142479896545,
      "learning_rate": 0.002,
      "loss": 1.9957492351531982,
      "step": 2350
    },
    {
      "epoch": 0.2352,
      "grad_norm": 0.07816534489393234,
      "learning_rate": 0.002,
      "loss": 1.8162741661071777,
      "step": 2352
    },
    {
      "epoch": 0.2354,
      "grad_norm": 0.07682987302541733,
      "learning_rate": 0.002,
      "loss": 2.046464204788208,
      "step": 2354
    },
    {
      "epoch": 0.2356,
      "grad_norm": 0.07797881960868835,
      "learning_rate": 0.002,
      "loss": 1.995376467704773,
      "step": 2356
    },
    {
      "epoch": 0.2358,
      "grad_norm": 0.07472646236419678,
      "learning_rate": 0.002,
      "loss": 1.889329195022583,
      "step": 2358
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.08463708311319351,
      "learning_rate": 0.002,
      "loss": 2.004965305328369,
      "step": 2360
    },
    {
      "epoch": 0.2362,
      "grad_norm": 0.07404356449842453,
      "learning_rate": 0.002,
      "loss": 1.9515604972839355,
      "step": 2362
    },
    {
      "epoch": 0.2364,
      "grad_norm": 0.0743633583188057,
      "learning_rate": 0.002,
      "loss": 1.8605901002883911,
      "step": 2364
    },
    {
      "epoch": 0.2366,
      "grad_norm": 0.07192595303058624,
      "learning_rate": 0.002,
      "loss": 1.930428147315979,
      "step": 2366
    },
    {
      "epoch": 0.2368,
      "grad_norm": 0.07365553826093674,
      "learning_rate": 0.002,
      "loss": 1.9820444583892822,
      "step": 2368
    },
    {
      "epoch": 0.237,
      "grad_norm": 0.07493475824594498,
      "learning_rate": 0.002,
      "loss": 1.9584970474243164,
      "step": 2370
    },
    {
      "epoch": 0.2372,
      "grad_norm": 0.08070436120033264,
      "learning_rate": 0.002,
      "loss": 2.0095720291137695,
      "step": 2372
    },
    {
      "epoch": 0.2374,
      "grad_norm": 0.07556866854429245,
      "learning_rate": 0.002,
      "loss": 2.0122299194335938,
      "step": 2374
    },
    {
      "epoch": 0.2376,
      "grad_norm": 0.07540794461965561,
      "learning_rate": 0.002,
      "loss": 2.026533603668213,
      "step": 2376
    },
    {
      "epoch": 0.2378,
      "grad_norm": 0.07580967992544174,
      "learning_rate": 0.002,
      "loss": 1.99920654296875,
      "step": 2378
    },
    {
      "epoch": 0.238,
      "grad_norm": 0.07537280023097992,
      "learning_rate": 0.002,
      "loss": 1.9328385591506958,
      "step": 2380
    },
    {
      "epoch": 0.2382,
      "grad_norm": 0.07138173282146454,
      "learning_rate": 0.002,
      "loss": 1.8456979990005493,
      "step": 2382
    },
    {
      "epoch": 0.2384,
      "grad_norm": 0.07857310771942139,
      "learning_rate": 0.002,
      "loss": 2.150524854660034,
      "step": 2384
    },
    {
      "epoch": 0.2386,
      "grad_norm": 0.07670405507087708,
      "learning_rate": 0.002,
      "loss": 1.9228503704071045,
      "step": 2386
    },
    {
      "epoch": 0.2388,
      "grad_norm": 0.0753355547785759,
      "learning_rate": 0.002,
      "loss": 1.995058536529541,
      "step": 2388
    },
    {
      "epoch": 0.239,
      "grad_norm": 0.07544506341218948,
      "learning_rate": 0.002,
      "loss": 1.8527469635009766,
      "step": 2390
    },
    {
      "epoch": 0.2392,
      "grad_norm": 0.08012618869543076,
      "learning_rate": 0.002,
      "loss": 2.0482277870178223,
      "step": 2392
    },
    {
      "epoch": 0.2394,
      "grad_norm": 0.08105237036943436,
      "learning_rate": 0.002,
      "loss": 1.9528847932815552,
      "step": 2394
    },
    {
      "epoch": 0.2396,
      "grad_norm": 0.08020584285259247,
      "learning_rate": 0.002,
      "loss": 2.0109567642211914,
      "step": 2396
    },
    {
      "epoch": 0.2398,
      "grad_norm": 0.07487299293279648,
      "learning_rate": 0.002,
      "loss": 1.8939224481582642,
      "step": 2398
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.07333816587924957,
      "learning_rate": 0.002,
      "loss": 1.9111549854278564,
      "step": 2400
    },
    {
      "epoch": 0.2402,
      "grad_norm": 0.07585359364748001,
      "learning_rate": 0.002,
      "loss": 1.8746062517166138,
      "step": 2402
    },
    {
      "epoch": 0.2404,
      "grad_norm": 0.07331408560276031,
      "learning_rate": 0.002,
      "loss": 1.9288665056228638,
      "step": 2404
    },
    {
      "epoch": 0.2406,
      "grad_norm": 0.07722676545381546,
      "learning_rate": 0.002,
      "loss": 1.845998764038086,
      "step": 2406
    },
    {
      "epoch": 0.2408,
      "grad_norm": 0.07934151589870453,
      "learning_rate": 0.002,
      "loss": 1.9836207628250122,
      "step": 2408
    },
    {
      "epoch": 0.241,
      "grad_norm": 0.07946431636810303,
      "learning_rate": 0.002,
      "loss": 1.9711799621582031,
      "step": 2410
    },
    {
      "epoch": 0.2412,
      "grad_norm": 0.0740002915263176,
      "learning_rate": 0.002,
      "loss": 2.0758330821990967,
      "step": 2412
    },
    {
      "epoch": 0.2414,
      "grad_norm": 0.08159994333982468,
      "learning_rate": 0.002,
      "loss": 1.796172857284546,
      "step": 2414
    },
    {
      "epoch": 0.2416,
      "grad_norm": 0.07702072709798813,
      "learning_rate": 0.002,
      "loss": 1.920156717300415,
      "step": 2416
    },
    {
      "epoch": 0.2418,
      "grad_norm": 0.07163342088460922,
      "learning_rate": 0.002,
      "loss": 1.933868169784546,
      "step": 2418
    },
    {
      "epoch": 0.242,
      "grad_norm": 0.07677235454320908,
      "learning_rate": 0.002,
      "loss": 1.9769153594970703,
      "step": 2420
    },
    {
      "epoch": 0.2422,
      "grad_norm": 0.0758829191327095,
      "learning_rate": 0.002,
      "loss": 1.97017502784729,
      "step": 2422
    },
    {
      "epoch": 0.2424,
      "grad_norm": 0.07877138257026672,
      "learning_rate": 0.002,
      "loss": 1.9894181489944458,
      "step": 2424
    },
    {
      "epoch": 0.2426,
      "grad_norm": 0.07422901690006256,
      "learning_rate": 0.002,
      "loss": 1.8675025701522827,
      "step": 2426
    },
    {
      "epoch": 0.2428,
      "grad_norm": 0.07765959203243256,
      "learning_rate": 0.002,
      "loss": 1.9923760890960693,
      "step": 2428
    },
    {
      "epoch": 0.243,
      "grad_norm": 0.07582318782806396,
      "learning_rate": 0.002,
      "loss": 1.93315851688385,
      "step": 2430
    },
    {
      "epoch": 0.2432,
      "grad_norm": 0.07441096007823944,
      "learning_rate": 0.002,
      "loss": 1.825900912284851,
      "step": 2432
    },
    {
      "epoch": 0.2434,
      "grad_norm": 0.07618384808301926,
      "learning_rate": 0.002,
      "loss": 1.9646142721176147,
      "step": 2434
    },
    {
      "epoch": 0.2436,
      "grad_norm": 0.07354626804590225,
      "learning_rate": 0.002,
      "loss": 1.9382864236831665,
      "step": 2436
    },
    {
      "epoch": 0.2438,
      "grad_norm": 0.07644262164831161,
      "learning_rate": 0.002,
      "loss": 2.078176259994507,
      "step": 2438
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.07793998718261719,
      "learning_rate": 0.002,
      "loss": 2.0721383094787598,
      "step": 2440
    },
    {
      "epoch": 0.2442,
      "grad_norm": 0.07198240607976913,
      "learning_rate": 0.002,
      "loss": 1.8596522808074951,
      "step": 2442
    },
    {
      "epoch": 0.2444,
      "grad_norm": 0.07486219704151154,
      "learning_rate": 0.002,
      "loss": 1.9349654912948608,
      "step": 2444
    },
    {
      "epoch": 0.2446,
      "grad_norm": 0.07519341260194778,
      "learning_rate": 0.002,
      "loss": 1.8300349712371826,
      "step": 2446
    },
    {
      "epoch": 0.2448,
      "grad_norm": 0.07672305405139923,
      "learning_rate": 0.002,
      "loss": 1.9450334310531616,
      "step": 2448
    },
    {
      "epoch": 0.245,
      "grad_norm": 0.0770457312464714,
      "learning_rate": 0.002,
      "loss": 1.9611669778823853,
      "step": 2450
    },
    {
      "epoch": 0.2452,
      "grad_norm": 0.07360785454511642,
      "learning_rate": 0.002,
      "loss": 1.8740695714950562,
      "step": 2452
    },
    {
      "epoch": 0.2454,
      "grad_norm": 0.07318465411663055,
      "learning_rate": 0.002,
      "loss": 1.9022009372711182,
      "step": 2454
    },
    {
      "epoch": 0.2456,
      "grad_norm": 0.0763900950551033,
      "learning_rate": 0.002,
      "loss": 1.9721137285232544,
      "step": 2456
    },
    {
      "epoch": 0.2458,
      "grad_norm": 0.07576487958431244,
      "learning_rate": 0.002,
      "loss": 2.0040574073791504,
      "step": 2458
    },
    {
      "epoch": 0.246,
      "grad_norm": 0.07128343731164932,
      "learning_rate": 0.002,
      "loss": 1.9414198398590088,
      "step": 2460
    },
    {
      "epoch": 0.2462,
      "grad_norm": 0.0813019648194313,
      "learning_rate": 0.002,
      "loss": 2.0144479274749756,
      "step": 2462
    },
    {
      "epoch": 0.2464,
      "grad_norm": 0.07753829658031464,
      "learning_rate": 0.002,
      "loss": 1.933452844619751,
      "step": 2464
    },
    {
      "epoch": 0.2466,
      "grad_norm": 0.0783180445432663,
      "learning_rate": 0.002,
      "loss": 1.799918293952942,
      "step": 2466
    },
    {
      "epoch": 0.2468,
      "grad_norm": 0.07888945937156677,
      "learning_rate": 0.002,
      "loss": 1.933791995048523,
      "step": 2468
    },
    {
      "epoch": 0.247,
      "grad_norm": 0.07549284398555756,
      "learning_rate": 0.002,
      "loss": 2.046027660369873,
      "step": 2470
    },
    {
      "epoch": 0.2472,
      "grad_norm": 0.07902402430772781,
      "learning_rate": 0.002,
      "loss": 1.9850497245788574,
      "step": 2472
    },
    {
      "epoch": 0.2474,
      "grad_norm": 0.07752101868391037,
      "learning_rate": 0.002,
      "loss": 1.9964934587478638,
      "step": 2474
    },
    {
      "epoch": 0.2476,
      "grad_norm": 0.07719989120960236,
      "learning_rate": 0.002,
      "loss": 1.839791178703308,
      "step": 2476
    },
    {
      "epoch": 0.2478,
      "grad_norm": 0.07598260045051575,
      "learning_rate": 0.002,
      "loss": 1.9751819372177124,
      "step": 2478
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.07329141348600388,
      "learning_rate": 0.002,
      "loss": 1.7499403953552246,
      "step": 2480
    },
    {
      "epoch": 0.2482,
      "grad_norm": 0.07821210473775864,
      "learning_rate": 0.002,
      "loss": 1.9873369932174683,
      "step": 2482
    },
    {
      "epoch": 0.2484,
      "grad_norm": 0.07488925009965897,
      "learning_rate": 0.002,
      "loss": 2.0306224822998047,
      "step": 2484
    },
    {
      "epoch": 0.2486,
      "grad_norm": 0.07172843813896179,
      "learning_rate": 0.002,
      "loss": 1.992779016494751,
      "step": 2486
    },
    {
      "epoch": 0.2488,
      "grad_norm": 0.07944095134735107,
      "learning_rate": 0.002,
      "loss": 1.923856258392334,
      "step": 2488
    },
    {
      "epoch": 0.249,
      "grad_norm": 0.07678613066673279,
      "learning_rate": 0.002,
      "loss": 2.0256011486053467,
      "step": 2490
    },
    {
      "epoch": 0.2492,
      "grad_norm": 0.07931242138147354,
      "learning_rate": 0.002,
      "loss": 2.015678644180298,
      "step": 2492
    },
    {
      "epoch": 0.2494,
      "grad_norm": 0.07490148395299911,
      "learning_rate": 0.002,
      "loss": 2.0094642639160156,
      "step": 2494
    },
    {
      "epoch": 0.2496,
      "grad_norm": 0.07554859668016434,
      "learning_rate": 0.002,
      "loss": 2.0898613929748535,
      "step": 2496
    },
    {
      "epoch": 0.2498,
      "grad_norm": 0.07594160735607147,
      "learning_rate": 0.002,
      "loss": 2.018193244934082,
      "step": 2498
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.07530094683170319,
      "learning_rate": 0.002,
      "loss": 2.031240224838257,
      "step": 2500
    },
    {
      "epoch": 0.2502,
      "grad_norm": 0.07500261813402176,
      "learning_rate": 0.002,
      "loss": 1.955904483795166,
      "step": 2502
    },
    {
      "epoch": 0.2504,
      "grad_norm": 0.07728112488985062,
      "learning_rate": 0.002,
      "loss": 2.0274341106414795,
      "step": 2504
    },
    {
      "epoch": 0.2506,
      "grad_norm": 0.076529860496521,
      "learning_rate": 0.002,
      "loss": 2.025641441345215,
      "step": 2506
    },
    {
      "epoch": 0.2508,
      "grad_norm": 0.07727022469043732,
      "learning_rate": 0.002,
      "loss": 2.052780866622925,
      "step": 2508
    },
    {
      "epoch": 0.251,
      "grad_norm": 0.07415511459112167,
      "learning_rate": 0.002,
      "loss": 2.058546781539917,
      "step": 2510
    },
    {
      "epoch": 0.2512,
      "grad_norm": 0.07624007761478424,
      "learning_rate": 0.002,
      "loss": 1.9039034843444824,
      "step": 2512
    },
    {
      "epoch": 0.2514,
      "grad_norm": 0.07801520824432373,
      "learning_rate": 0.002,
      "loss": 2.050907850265503,
      "step": 2514
    },
    {
      "epoch": 0.2516,
      "grad_norm": 0.07822613418102264,
      "learning_rate": 0.002,
      "loss": 1.95448899269104,
      "step": 2516
    },
    {
      "epoch": 0.2518,
      "grad_norm": 0.07908643037080765,
      "learning_rate": 0.002,
      "loss": 2.0220415592193604,
      "step": 2518
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.07656794786453247,
      "learning_rate": 0.002,
      "loss": 1.9839717149734497,
      "step": 2520
    },
    {
      "epoch": 0.2522,
      "grad_norm": 0.08474338054656982,
      "learning_rate": 0.002,
      "loss": 1.9430696964263916,
      "step": 2522
    },
    {
      "epoch": 0.2524,
      "grad_norm": 0.0738358274102211,
      "learning_rate": 0.002,
      "loss": 1.9766037464141846,
      "step": 2524
    },
    {
      "epoch": 0.2526,
      "grad_norm": 0.07816296815872192,
      "learning_rate": 0.002,
      "loss": 1.9712445735931396,
      "step": 2526
    },
    {
      "epoch": 0.2528,
      "grad_norm": 0.07215908169746399,
      "learning_rate": 0.002,
      "loss": 1.8631515502929688,
      "step": 2528
    },
    {
      "epoch": 0.253,
      "grad_norm": 0.077146515250206,
      "learning_rate": 0.002,
      "loss": 2.0486419200897217,
      "step": 2530
    },
    {
      "epoch": 0.2532,
      "grad_norm": 0.0795506089925766,
      "learning_rate": 0.002,
      "loss": 1.997110366821289,
      "step": 2532
    },
    {
      "epoch": 0.2534,
      "grad_norm": 0.07846347987651825,
      "learning_rate": 0.002,
      "loss": 1.8598124980926514,
      "step": 2534
    },
    {
      "epoch": 0.2536,
      "grad_norm": 0.07510930299758911,
      "learning_rate": 0.002,
      "loss": 1.9342654943466187,
      "step": 2536
    },
    {
      "epoch": 0.2538,
      "grad_norm": 0.07638435810804367,
      "learning_rate": 0.002,
      "loss": 2.0640227794647217,
      "step": 2538
    },
    {
      "epoch": 0.254,
      "grad_norm": 0.07365086674690247,
      "learning_rate": 0.002,
      "loss": 1.9159626960754395,
      "step": 2540
    },
    {
      "epoch": 0.2542,
      "grad_norm": 0.07775711268186569,
      "learning_rate": 0.002,
      "loss": 1.9979543685913086,
      "step": 2542
    },
    {
      "epoch": 0.2544,
      "grad_norm": 0.07566583156585693,
      "learning_rate": 0.002,
      "loss": 1.9465508460998535,
      "step": 2544
    },
    {
      "epoch": 0.2546,
      "grad_norm": 0.07504589110612869,
      "learning_rate": 0.002,
      "loss": 1.8733385801315308,
      "step": 2546
    },
    {
      "epoch": 0.2548,
      "grad_norm": 0.0759311318397522,
      "learning_rate": 0.002,
      "loss": 1.9534825086593628,
      "step": 2548
    },
    {
      "epoch": 0.255,
      "grad_norm": 0.08165276795625687,
      "learning_rate": 0.002,
      "loss": 2.0505313873291016,
      "step": 2550
    },
    {
      "epoch": 0.2552,
      "grad_norm": 0.07777179777622223,
      "learning_rate": 0.002,
      "loss": 1.9229892492294312,
      "step": 2552
    },
    {
      "epoch": 0.2554,
      "grad_norm": 0.07632949948310852,
      "learning_rate": 0.002,
      "loss": 2.064345121383667,
      "step": 2554
    },
    {
      "epoch": 0.2556,
      "grad_norm": 0.07547987997531891,
      "learning_rate": 0.002,
      "loss": 2.039017915725708,
      "step": 2556
    },
    {
      "epoch": 0.2558,
      "grad_norm": 0.07939358055591583,
      "learning_rate": 0.002,
      "loss": 1.9279406070709229,
      "step": 2558
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.08081863075494766,
      "learning_rate": 0.002,
      "loss": 1.9838733673095703,
      "step": 2560
    },
    {
      "epoch": 0.2562,
      "grad_norm": 0.08163375407457352,
      "learning_rate": 0.002,
      "loss": 1.999534249305725,
      "step": 2562
    },
    {
      "epoch": 0.2564,
      "grad_norm": 0.07981166988611221,
      "learning_rate": 0.002,
      "loss": 1.8858057260513306,
      "step": 2564
    },
    {
      "epoch": 0.2566,
      "grad_norm": 0.07140326499938965,
      "learning_rate": 0.002,
      "loss": 1.928013563156128,
      "step": 2566
    },
    {
      "epoch": 0.2568,
      "grad_norm": 0.07786645740270615,
      "learning_rate": 0.002,
      "loss": 2.0049424171447754,
      "step": 2568
    },
    {
      "epoch": 0.257,
      "grad_norm": 0.07086050510406494,
      "learning_rate": 0.002,
      "loss": 1.9308507442474365,
      "step": 2570
    },
    {
      "epoch": 0.2572,
      "grad_norm": 0.07360795885324478,
      "learning_rate": 0.002,
      "loss": 2.0009732246398926,
      "step": 2572
    },
    {
      "epoch": 0.2574,
      "grad_norm": 0.07549197226762772,
      "learning_rate": 0.002,
      "loss": 1.8344440460205078,
      "step": 2574
    },
    {
      "epoch": 0.2576,
      "grad_norm": 0.06965584307909012,
      "learning_rate": 0.002,
      "loss": 2.0119261741638184,
      "step": 2576
    },
    {
      "epoch": 0.2578,
      "grad_norm": 0.0724850594997406,
      "learning_rate": 0.002,
      "loss": 1.9491331577301025,
      "step": 2578
    },
    {
      "epoch": 0.258,
      "grad_norm": 0.07475418597459793,
      "learning_rate": 0.002,
      "loss": 1.979975938796997,
      "step": 2580
    },
    {
      "epoch": 0.2582,
      "grad_norm": 0.07555396109819412,
      "learning_rate": 0.002,
      "loss": 1.8271989822387695,
      "step": 2582
    },
    {
      "epoch": 0.2584,
      "grad_norm": 0.07971849292516708,
      "learning_rate": 0.002,
      "loss": 2.040419340133667,
      "step": 2584
    },
    {
      "epoch": 0.2586,
      "grad_norm": 0.0836615189909935,
      "learning_rate": 0.002,
      "loss": 1.9001878499984741,
      "step": 2586
    },
    {
      "epoch": 0.2588,
      "grad_norm": 0.07739779353141785,
      "learning_rate": 0.002,
      "loss": 2.055602788925171,
      "step": 2588
    },
    {
      "epoch": 0.259,
      "grad_norm": 0.07386799156665802,
      "learning_rate": 0.002,
      "loss": 1.9171220064163208,
      "step": 2590
    },
    {
      "epoch": 0.2592,
      "grad_norm": 0.07466378062963486,
      "learning_rate": 0.002,
      "loss": 1.9005855321884155,
      "step": 2592
    },
    {
      "epoch": 0.2594,
      "grad_norm": 0.07425230741500854,
      "learning_rate": 0.002,
      "loss": 1.9108275175094604,
      "step": 2594
    },
    {
      "epoch": 0.2596,
      "grad_norm": 0.08254028111696243,
      "learning_rate": 0.002,
      "loss": 2.029829263687134,
      "step": 2596
    },
    {
      "epoch": 0.2598,
      "grad_norm": 0.07304789870977402,
      "learning_rate": 0.002,
      "loss": 2.0457093715667725,
      "step": 2598
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.07582638412714005,
      "learning_rate": 0.002,
      "loss": 1.937441110610962,
      "step": 2600
    },
    {
      "epoch": 0.2602,
      "grad_norm": 0.07460539788007736,
      "learning_rate": 0.002,
      "loss": 1.9596800804138184,
      "step": 2602
    },
    {
      "epoch": 0.2604,
      "grad_norm": 0.07599665969610214,
      "learning_rate": 0.002,
      "loss": 1.9893494844436646,
      "step": 2604
    },
    {
      "epoch": 0.2606,
      "grad_norm": 0.07727498561143875,
      "learning_rate": 0.002,
      "loss": 1.933333396911621,
      "step": 2606
    },
    {
      "epoch": 0.2608,
      "grad_norm": 0.0752563402056694,
      "learning_rate": 0.002,
      "loss": 1.794803500175476,
      "step": 2608
    },
    {
      "epoch": 0.261,
      "grad_norm": 0.07415994256734848,
      "learning_rate": 0.002,
      "loss": 1.8792747259140015,
      "step": 2610
    },
    {
      "epoch": 0.2612,
      "grad_norm": 0.07692742347717285,
      "learning_rate": 0.002,
      "loss": 1.9100780487060547,
      "step": 2612
    },
    {
      "epoch": 0.2614,
      "grad_norm": 0.07951962202787399,
      "learning_rate": 0.002,
      "loss": 2.0127182006835938,
      "step": 2614
    },
    {
      "epoch": 0.2616,
      "grad_norm": 0.07499701529741287,
      "learning_rate": 0.002,
      "loss": 1.9151159524917603,
      "step": 2616
    },
    {
      "epoch": 0.2618,
      "grad_norm": 0.0861487090587616,
      "learning_rate": 0.002,
      "loss": 1.9496415853500366,
      "step": 2618
    },
    {
      "epoch": 0.262,
      "grad_norm": 0.07528205960988998,
      "learning_rate": 0.002,
      "loss": 1.78060781955719,
      "step": 2620
    },
    {
      "epoch": 0.2622,
      "grad_norm": 0.07497971504926682,
      "learning_rate": 0.002,
      "loss": 1.9008928537368774,
      "step": 2622
    },
    {
      "epoch": 0.2624,
      "grad_norm": 0.07492546737194061,
      "learning_rate": 0.002,
      "loss": 1.9909002780914307,
      "step": 2624
    },
    {
      "epoch": 0.2626,
      "grad_norm": 0.07300236821174622,
      "learning_rate": 0.002,
      "loss": 1.9033476114273071,
      "step": 2626
    },
    {
      "epoch": 0.2628,
      "grad_norm": 0.07944224774837494,
      "learning_rate": 0.002,
      "loss": 1.9602298736572266,
      "step": 2628
    },
    {
      "epoch": 0.263,
      "grad_norm": 0.07807758450508118,
      "learning_rate": 0.002,
      "loss": 2.004810333251953,
      "step": 2630
    },
    {
      "epoch": 0.2632,
      "grad_norm": 0.07407421618700027,
      "learning_rate": 0.002,
      "loss": 1.9176597595214844,
      "step": 2632
    },
    {
      "epoch": 0.2634,
      "grad_norm": 0.07552170008420944,
      "learning_rate": 0.002,
      "loss": 1.886703372001648,
      "step": 2634
    },
    {
      "epoch": 0.2636,
      "grad_norm": 0.07771562784910202,
      "learning_rate": 0.002,
      "loss": 2.0769777297973633,
      "step": 2636
    },
    {
      "epoch": 0.2638,
      "grad_norm": 0.07675088942050934,
      "learning_rate": 0.002,
      "loss": 2.016735553741455,
      "step": 2638
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.07750120013952255,
      "learning_rate": 0.002,
      "loss": 1.941680908203125,
      "step": 2640
    },
    {
      "epoch": 0.2642,
      "grad_norm": 0.08077846467494965,
      "learning_rate": 0.002,
      "loss": 1.889835000038147,
      "step": 2642
    },
    {
      "epoch": 0.2644,
      "grad_norm": 0.07880892604589462,
      "learning_rate": 0.002,
      "loss": 1.8660777807235718,
      "step": 2644
    },
    {
      "epoch": 0.2646,
      "grad_norm": 0.08110782504081726,
      "learning_rate": 0.002,
      "loss": 1.951617956161499,
      "step": 2646
    },
    {
      "epoch": 0.2648,
      "grad_norm": 0.0742243081331253,
      "learning_rate": 0.002,
      "loss": 1.9761569499969482,
      "step": 2648
    },
    {
      "epoch": 0.265,
      "grad_norm": 0.07118087261915207,
      "learning_rate": 0.002,
      "loss": 1.8874567747116089,
      "step": 2650
    },
    {
      "epoch": 0.2652,
      "grad_norm": 0.07653947919607162,
      "learning_rate": 0.002,
      "loss": 1.8246474266052246,
      "step": 2652
    },
    {
      "epoch": 0.2654,
      "grad_norm": 0.07701896131038666,
      "learning_rate": 0.002,
      "loss": 1.983399748802185,
      "step": 2654
    },
    {
      "epoch": 0.2656,
      "grad_norm": 0.08037595450878143,
      "learning_rate": 0.002,
      "loss": 2.0417559146881104,
      "step": 2656
    },
    {
      "epoch": 0.2658,
      "grad_norm": 0.07483422756195068,
      "learning_rate": 0.002,
      "loss": 1.8761413097381592,
      "step": 2658
    },
    {
      "epoch": 0.266,
      "grad_norm": 0.08027554303407669,
      "learning_rate": 0.002,
      "loss": 1.9287828207015991,
      "step": 2660
    },
    {
      "epoch": 0.2662,
      "grad_norm": 0.07315602898597717,
      "learning_rate": 0.002,
      "loss": 1.7781846523284912,
      "step": 2662
    },
    {
      "epoch": 0.2664,
      "grad_norm": 0.07764074951410294,
      "learning_rate": 0.002,
      "loss": 1.8195322751998901,
      "step": 2664
    },
    {
      "epoch": 0.2666,
      "grad_norm": 0.07696542143821716,
      "learning_rate": 0.002,
      "loss": 1.8946833610534668,
      "step": 2666
    },
    {
      "epoch": 0.2668,
      "grad_norm": 0.07807137072086334,
      "learning_rate": 0.002,
      "loss": 2.0089361667633057,
      "step": 2668
    },
    {
      "epoch": 0.267,
      "grad_norm": 0.07684317976236343,
      "learning_rate": 0.002,
      "loss": 1.9602919816970825,
      "step": 2670
    },
    {
      "epoch": 0.2672,
      "grad_norm": 0.0734330490231514,
      "learning_rate": 0.002,
      "loss": 1.9607833623886108,
      "step": 2672
    },
    {
      "epoch": 0.2674,
      "grad_norm": 0.07515490055084229,
      "learning_rate": 0.002,
      "loss": 1.954183578491211,
      "step": 2674
    },
    {
      "epoch": 0.2676,
      "grad_norm": 0.07528668642044067,
      "learning_rate": 0.002,
      "loss": 1.9955158233642578,
      "step": 2676
    },
    {
      "epoch": 0.2678,
      "grad_norm": 0.07280945032835007,
      "learning_rate": 0.002,
      "loss": 1.94607675075531,
      "step": 2678
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.0744769349694252,
      "learning_rate": 0.002,
      "loss": 1.9630106687545776,
      "step": 2680
    },
    {
      "epoch": 0.2682,
      "grad_norm": 0.07699563354253769,
      "learning_rate": 0.002,
      "loss": 1.9230986833572388,
      "step": 2682
    },
    {
      "epoch": 0.2684,
      "grad_norm": 0.07636984437704086,
      "learning_rate": 0.002,
      "loss": 2.090508460998535,
      "step": 2684
    },
    {
      "epoch": 0.2686,
      "grad_norm": 0.07423525303602219,
      "learning_rate": 0.002,
      "loss": 1.8247759342193604,
      "step": 2686
    },
    {
      "epoch": 0.2688,
      "grad_norm": 0.07117210328578949,
      "learning_rate": 0.002,
      "loss": 1.8654415607452393,
      "step": 2688
    },
    {
      "epoch": 0.269,
      "grad_norm": 0.07174254953861237,
      "learning_rate": 0.002,
      "loss": 1.9047259092330933,
      "step": 2690
    },
    {
      "epoch": 0.2692,
      "grad_norm": 0.07391062378883362,
      "learning_rate": 0.002,
      "loss": 1.8404324054718018,
      "step": 2692
    },
    {
      "epoch": 0.2694,
      "grad_norm": 0.07765576988458633,
      "learning_rate": 0.002,
      "loss": 2.035804271697998,
      "step": 2694
    },
    {
      "epoch": 0.2696,
      "grad_norm": 0.07642754912376404,
      "learning_rate": 0.002,
      "loss": 2.0214478969573975,
      "step": 2696
    },
    {
      "epoch": 0.2698,
      "grad_norm": 0.07398176938295364,
      "learning_rate": 0.002,
      "loss": 1.8883534669876099,
      "step": 2698
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.07130539417266846,
      "learning_rate": 0.002,
      "loss": 1.9112334251403809,
      "step": 2700
    },
    {
      "epoch": 0.2702,
      "grad_norm": 0.07233530282974243,
      "learning_rate": 0.002,
      "loss": 2.0372231006622314,
      "step": 2702
    },
    {
      "epoch": 0.2704,
      "grad_norm": 0.07697498798370361,
      "learning_rate": 0.002,
      "loss": 1.9719194173812866,
      "step": 2704
    },
    {
      "epoch": 0.2706,
      "grad_norm": 0.07554778456687927,
      "learning_rate": 0.002,
      "loss": 1.9074852466583252,
      "step": 2706
    },
    {
      "epoch": 0.2708,
      "grad_norm": 0.07718528062105179,
      "learning_rate": 0.002,
      "loss": 1.973092794418335,
      "step": 2708
    },
    {
      "epoch": 0.271,
      "grad_norm": 0.07645302265882492,
      "learning_rate": 0.002,
      "loss": 1.9071097373962402,
      "step": 2710
    },
    {
      "epoch": 0.2712,
      "grad_norm": 0.07592093199491501,
      "learning_rate": 0.002,
      "loss": 1.9110814332962036,
      "step": 2712
    },
    {
      "epoch": 0.2714,
      "grad_norm": 0.07124105095863342,
      "learning_rate": 0.002,
      "loss": 1.7199546098709106,
      "step": 2714
    },
    {
      "epoch": 0.2716,
      "grad_norm": 0.08037260919809341,
      "learning_rate": 0.002,
      "loss": 1.9326587915420532,
      "step": 2716
    },
    {
      "epoch": 0.2718,
      "grad_norm": 0.07361160963773727,
      "learning_rate": 0.002,
      "loss": 1.804139256477356,
      "step": 2718
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.0763820931315422,
      "learning_rate": 0.002,
      "loss": 2.015197992324829,
      "step": 2720
    },
    {
      "epoch": 0.2722,
      "grad_norm": 0.07630789279937744,
      "learning_rate": 0.002,
      "loss": 1.964240550994873,
      "step": 2722
    },
    {
      "epoch": 0.2724,
      "grad_norm": 0.07657467573881149,
      "learning_rate": 0.002,
      "loss": 2.0317556858062744,
      "step": 2724
    },
    {
      "epoch": 0.2726,
      "grad_norm": 0.07460824400186539,
      "learning_rate": 0.002,
      "loss": 1.797733187675476,
      "step": 2726
    },
    {
      "epoch": 0.2728,
      "grad_norm": 0.07726311683654785,
      "learning_rate": 0.002,
      "loss": 1.8332479000091553,
      "step": 2728
    },
    {
      "epoch": 0.273,
      "grad_norm": 0.08029306679964066,
      "learning_rate": 0.002,
      "loss": 1.8852342367172241,
      "step": 2730
    },
    {
      "epoch": 0.2732,
      "grad_norm": 0.08435230702161789,
      "learning_rate": 0.002,
      "loss": 2.0342795848846436,
      "step": 2732
    },
    {
      "epoch": 0.2734,
      "grad_norm": 0.07679324597120285,
      "learning_rate": 0.002,
      "loss": 1.9279725551605225,
      "step": 2734
    },
    {
      "epoch": 0.2736,
      "grad_norm": 0.07276396453380585,
      "learning_rate": 0.002,
      "loss": 1.8457417488098145,
      "step": 2736
    },
    {
      "epoch": 0.2738,
      "grad_norm": 0.08078870177268982,
      "learning_rate": 0.002,
      "loss": 1.9229936599731445,
      "step": 2738
    },
    {
      "epoch": 0.274,
      "grad_norm": 0.0761222243309021,
      "learning_rate": 0.002,
      "loss": 1.855264663696289,
      "step": 2740
    },
    {
      "epoch": 0.2742,
      "grad_norm": 0.07716292887926102,
      "learning_rate": 0.002,
      "loss": 1.9951012134552002,
      "step": 2742
    },
    {
      "epoch": 0.2744,
      "grad_norm": 0.07537540048360825,
      "learning_rate": 0.002,
      "loss": 1.9687503576278687,
      "step": 2744
    },
    {
      "epoch": 0.2746,
      "grad_norm": 0.0759337842464447,
      "learning_rate": 0.002,
      "loss": 1.9792839288711548,
      "step": 2746
    },
    {
      "epoch": 0.2748,
      "grad_norm": 0.08525499701499939,
      "learning_rate": 0.002,
      "loss": 1.9378670454025269,
      "step": 2748
    },
    {
      "epoch": 0.275,
      "grad_norm": 0.07662256807088852,
      "learning_rate": 0.002,
      "loss": 1.962996244430542,
      "step": 2750
    },
    {
      "epoch": 0.2752,
      "grad_norm": 0.07155496627092361,
      "learning_rate": 0.002,
      "loss": 1.92281174659729,
      "step": 2752
    },
    {
      "epoch": 0.2754,
      "grad_norm": 0.08044813573360443,
      "learning_rate": 0.002,
      "loss": 1.9837371110916138,
      "step": 2754
    },
    {
      "epoch": 0.2756,
      "grad_norm": 0.07926972955465317,
      "learning_rate": 0.002,
      "loss": 2.031611919403076,
      "step": 2756
    },
    {
      "epoch": 0.2758,
      "grad_norm": 0.08741366118192673,
      "learning_rate": 0.002,
      "loss": 2.0194694995880127,
      "step": 2758
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.07730919122695923,
      "learning_rate": 0.002,
      "loss": 1.9738985300064087,
      "step": 2760
    },
    {
      "epoch": 0.2762,
      "grad_norm": 0.07645008713006973,
      "learning_rate": 0.002,
      "loss": 1.9455294609069824,
      "step": 2762
    },
    {
      "epoch": 0.2764,
      "grad_norm": 0.07672157883644104,
      "learning_rate": 0.002,
      "loss": 1.9267302751541138,
      "step": 2764
    },
    {
      "epoch": 0.2766,
      "grad_norm": 0.07710259407758713,
      "learning_rate": 0.002,
      "loss": 1.9446241855621338,
      "step": 2766
    },
    {
      "epoch": 0.2768,
      "grad_norm": 0.07633387297391891,
      "learning_rate": 0.002,
      "loss": 1.913986325263977,
      "step": 2768
    },
    {
      "epoch": 0.277,
      "grad_norm": 0.08082033693790436,
      "learning_rate": 0.002,
      "loss": 1.9333535432815552,
      "step": 2770
    },
    {
      "epoch": 0.2772,
      "grad_norm": 0.0834222212433815,
      "learning_rate": 0.002,
      "loss": 1.9848682880401611,
      "step": 2772
    },
    {
      "epoch": 0.2774,
      "grad_norm": 0.07374601811170578,
      "learning_rate": 0.002,
      "loss": 1.8428055047988892,
      "step": 2774
    },
    {
      "epoch": 0.2776,
      "grad_norm": 0.07865054160356522,
      "learning_rate": 0.002,
      "loss": 2.0380442142486572,
      "step": 2776
    },
    {
      "epoch": 0.2778,
      "grad_norm": 0.07560556381940842,
      "learning_rate": 0.002,
      "loss": 1.879299283027649,
      "step": 2778
    },
    {
      "epoch": 0.278,
      "grad_norm": 0.07594578713178635,
      "learning_rate": 0.002,
      "loss": 1.9369103908538818,
      "step": 2780
    },
    {
      "epoch": 0.2782,
      "grad_norm": 0.07932909578084946,
      "learning_rate": 0.002,
      "loss": 2.0141189098358154,
      "step": 2782
    },
    {
      "epoch": 0.2784,
      "grad_norm": 0.07658612728118896,
      "learning_rate": 0.002,
      "loss": 1.9492542743682861,
      "step": 2784
    },
    {
      "epoch": 0.2786,
      "grad_norm": 0.07242601364850998,
      "learning_rate": 0.002,
      "loss": 1.9494893550872803,
      "step": 2786
    },
    {
      "epoch": 0.2788,
      "grad_norm": 0.0801464170217514,
      "learning_rate": 0.002,
      "loss": 2.024137020111084,
      "step": 2788
    },
    {
      "epoch": 0.279,
      "grad_norm": 0.07535567879676819,
      "learning_rate": 0.002,
      "loss": 1.7925182580947876,
      "step": 2790
    },
    {
      "epoch": 0.2792,
      "grad_norm": 0.07416930049657822,
      "learning_rate": 0.002,
      "loss": 2.0185821056365967,
      "step": 2792
    },
    {
      "epoch": 0.2794,
      "grad_norm": 0.07223834097385406,
      "learning_rate": 0.002,
      "loss": 1.7672780752182007,
      "step": 2794
    },
    {
      "epoch": 0.2796,
      "grad_norm": 0.07405047118663788,
      "learning_rate": 0.002,
      "loss": 1.864173412322998,
      "step": 2796
    },
    {
      "epoch": 0.2798,
      "grad_norm": 0.07446594536304474,
      "learning_rate": 0.002,
      "loss": 1.9231663942337036,
      "step": 2798
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.07666236907243729,
      "learning_rate": 0.002,
      "loss": 1.9877043962478638,
      "step": 2800
    },
    {
      "epoch": 0.2802,
      "grad_norm": 0.07447870820760727,
      "learning_rate": 0.002,
      "loss": 1.8705964088439941,
      "step": 2802
    },
    {
      "epoch": 0.2804,
      "grad_norm": 0.07724396884441376,
      "learning_rate": 0.002,
      "loss": 1.905720591545105,
      "step": 2804
    },
    {
      "epoch": 0.2806,
      "grad_norm": 0.07699690014123917,
      "learning_rate": 0.002,
      "loss": 1.9583276510238647,
      "step": 2806
    },
    {
      "epoch": 0.2808,
      "grad_norm": 0.07373912632465363,
      "learning_rate": 0.002,
      "loss": 2.040966033935547,
      "step": 2808
    },
    {
      "epoch": 0.281,
      "grad_norm": 0.07842797785997391,
      "learning_rate": 0.002,
      "loss": 2.0029685497283936,
      "step": 2810
    },
    {
      "epoch": 0.2812,
      "grad_norm": 0.0734037384390831,
      "learning_rate": 0.002,
      "loss": 1.926543116569519,
      "step": 2812
    },
    {
      "epoch": 0.2814,
      "grad_norm": 0.07742153853178024,
      "learning_rate": 0.002,
      "loss": 1.928130865097046,
      "step": 2814
    },
    {
      "epoch": 0.2816,
      "grad_norm": 0.07278797030448914,
      "learning_rate": 0.002,
      "loss": 2.015315055847168,
      "step": 2816
    },
    {
      "epoch": 0.2818,
      "grad_norm": 0.07786688953638077,
      "learning_rate": 0.002,
      "loss": 1.8292397260665894,
      "step": 2818
    },
    {
      "epoch": 0.282,
      "grad_norm": 0.07375305891036987,
      "learning_rate": 0.002,
      "loss": 1.9398375749588013,
      "step": 2820
    },
    {
      "epoch": 0.2822,
      "grad_norm": 0.07407531887292862,
      "learning_rate": 0.002,
      "loss": 1.811747431755066,
      "step": 2822
    },
    {
      "epoch": 0.2824,
      "grad_norm": 0.07820440083742142,
      "learning_rate": 0.002,
      "loss": 1.9319361448287964,
      "step": 2824
    },
    {
      "epoch": 0.2826,
      "grad_norm": 0.08054108172655106,
      "learning_rate": 0.002,
      "loss": 1.941037893295288,
      "step": 2826
    },
    {
      "epoch": 0.2828,
      "grad_norm": 0.07716015726327896,
      "learning_rate": 0.002,
      "loss": 1.8489446640014648,
      "step": 2828
    },
    {
      "epoch": 0.283,
      "grad_norm": 0.07855558395385742,
      "learning_rate": 0.002,
      "loss": 1.9728834629058838,
      "step": 2830
    },
    {
      "epoch": 0.2832,
      "grad_norm": 0.07836096733808517,
      "learning_rate": 0.002,
      "loss": 1.9243695735931396,
      "step": 2832
    },
    {
      "epoch": 0.2834,
      "grad_norm": 0.07474042475223541,
      "learning_rate": 0.002,
      "loss": 1.9433419704437256,
      "step": 2834
    },
    {
      "epoch": 0.2836,
      "grad_norm": 0.07406354695558548,
      "learning_rate": 0.002,
      "loss": 1.8454148769378662,
      "step": 2836
    },
    {
      "epoch": 0.2838,
      "grad_norm": 0.07657510787248611,
      "learning_rate": 0.002,
      "loss": 2.0184531211853027,
      "step": 2838
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.08110678941011429,
      "learning_rate": 0.002,
      "loss": 1.7892892360687256,
      "step": 2840
    },
    {
      "epoch": 0.2842,
      "grad_norm": 0.07563204318284988,
      "learning_rate": 0.002,
      "loss": 2.0066659450531006,
      "step": 2842
    },
    {
      "epoch": 0.2844,
      "grad_norm": 0.07308538258075714,
      "learning_rate": 0.002,
      "loss": 1.993926763534546,
      "step": 2844
    },
    {
      "epoch": 0.2846,
      "grad_norm": 0.0733911395072937,
      "learning_rate": 0.002,
      "loss": 1.9582160711288452,
      "step": 2846
    },
    {
      "epoch": 0.2848,
      "grad_norm": 0.07483308762311935,
      "learning_rate": 0.002,
      "loss": 1.8571827411651611,
      "step": 2848
    },
    {
      "epoch": 0.285,
      "grad_norm": 0.07510905712842941,
      "learning_rate": 0.002,
      "loss": 2.002802610397339,
      "step": 2850
    },
    {
      "epoch": 0.2852,
      "grad_norm": 0.07670042663812637,
      "learning_rate": 0.002,
      "loss": 1.8364801406860352,
      "step": 2852
    },
    {
      "epoch": 0.2854,
      "grad_norm": 0.07830021530389786,
      "learning_rate": 0.002,
      "loss": 1.9490923881530762,
      "step": 2854
    },
    {
      "epoch": 0.2856,
      "grad_norm": 0.07448120415210724,
      "learning_rate": 0.002,
      "loss": 1.8776239156723022,
      "step": 2856
    },
    {
      "epoch": 0.2858,
      "grad_norm": 0.07852816581726074,
      "learning_rate": 0.002,
      "loss": 1.8351566791534424,
      "step": 2858
    },
    {
      "epoch": 0.286,
      "grad_norm": 0.07787445187568665,
      "learning_rate": 0.002,
      "loss": 1.7515239715576172,
      "step": 2860
    },
    {
      "epoch": 0.2862,
      "grad_norm": 0.07677292823791504,
      "learning_rate": 0.002,
      "loss": 2.0540904998779297,
      "step": 2862
    },
    {
      "epoch": 0.2864,
      "grad_norm": 0.07623957842588425,
      "learning_rate": 0.002,
      "loss": 2.013598918914795,
      "step": 2864
    },
    {
      "epoch": 0.2866,
      "grad_norm": 0.07868707925081253,
      "learning_rate": 0.002,
      "loss": 1.9770617485046387,
      "step": 2866
    },
    {
      "epoch": 0.2868,
      "grad_norm": 0.07720731943845749,
      "learning_rate": 0.002,
      "loss": 2.0556044578552246,
      "step": 2868
    },
    {
      "epoch": 0.287,
      "grad_norm": 0.08363821357488632,
      "learning_rate": 0.002,
      "loss": 1.9909380674362183,
      "step": 2870
    },
    {
      "epoch": 0.2872,
      "grad_norm": 0.07919886708259583,
      "learning_rate": 0.002,
      "loss": 1.8675583600997925,
      "step": 2872
    },
    {
      "epoch": 0.2874,
      "grad_norm": 0.0829833522439003,
      "learning_rate": 0.002,
      "loss": 2.0488719940185547,
      "step": 2874
    },
    {
      "epoch": 0.2876,
      "grad_norm": 0.08194614201784134,
      "learning_rate": 0.002,
      "loss": 1.9918599128723145,
      "step": 2876
    },
    {
      "epoch": 0.2878,
      "grad_norm": 0.07118509709835052,
      "learning_rate": 0.002,
      "loss": 1.8739105463027954,
      "step": 2878
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.07750877737998962,
      "learning_rate": 0.002,
      "loss": 1.8570948839187622,
      "step": 2880
    },
    {
      "epoch": 0.2882,
      "grad_norm": 0.07656653225421906,
      "learning_rate": 0.002,
      "loss": 1.9429340362548828,
      "step": 2882
    },
    {
      "epoch": 0.2884,
      "grad_norm": 0.07754821330308914,
      "learning_rate": 0.002,
      "loss": 1.8224358558654785,
      "step": 2884
    },
    {
      "epoch": 0.2886,
      "grad_norm": 0.08149085938930511,
      "learning_rate": 0.002,
      "loss": 1.9627740383148193,
      "step": 2886
    },
    {
      "epoch": 0.2888,
      "grad_norm": 0.07260768860578537,
      "learning_rate": 0.002,
      "loss": 1.9511487483978271,
      "step": 2888
    },
    {
      "epoch": 0.289,
      "grad_norm": 0.07239068299531937,
      "learning_rate": 0.002,
      "loss": 1.8933616876602173,
      "step": 2890
    },
    {
      "epoch": 0.2892,
      "grad_norm": 0.07796280831098557,
      "learning_rate": 0.002,
      "loss": 1.9235060214996338,
      "step": 2892
    },
    {
      "epoch": 0.2894,
      "grad_norm": 0.07505206018686295,
      "learning_rate": 0.002,
      "loss": 2.1014316082000732,
      "step": 2894
    },
    {
      "epoch": 0.2896,
      "grad_norm": 0.08731655776500702,
      "learning_rate": 0.002,
      "loss": 1.9209859371185303,
      "step": 2896
    },
    {
      "epoch": 0.2898,
      "grad_norm": 0.07432211190462112,
      "learning_rate": 0.002,
      "loss": 1.9186869859695435,
      "step": 2898
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.0745650976896286,
      "learning_rate": 0.002,
      "loss": 1.8831392526626587,
      "step": 2900
    },
    {
      "epoch": 0.2902,
      "grad_norm": 0.07899170368909836,
      "learning_rate": 0.002,
      "loss": 1.7601778507232666,
      "step": 2902
    },
    {
      "epoch": 0.2904,
      "grad_norm": 0.07498981803655624,
      "learning_rate": 0.002,
      "loss": 2.0300357341766357,
      "step": 2904
    },
    {
      "epoch": 0.2906,
      "grad_norm": 0.07352370768785477,
      "learning_rate": 0.002,
      "loss": 1.9598129987716675,
      "step": 2906
    },
    {
      "epoch": 0.2908,
      "grad_norm": 0.08027636259794235,
      "learning_rate": 0.002,
      "loss": 2.0153167247772217,
      "step": 2908
    },
    {
      "epoch": 0.291,
      "grad_norm": 0.07491970807313919,
      "learning_rate": 0.002,
      "loss": 1.9729491472244263,
      "step": 2910
    },
    {
      "epoch": 0.2912,
      "grad_norm": 0.07740698009729385,
      "learning_rate": 0.002,
      "loss": 1.8526904582977295,
      "step": 2912
    },
    {
      "epoch": 0.2914,
      "grad_norm": 0.07734980434179306,
      "learning_rate": 0.002,
      "loss": 1.9110004901885986,
      "step": 2914
    },
    {
      "epoch": 0.2916,
      "grad_norm": 0.0738893523812294,
      "learning_rate": 0.002,
      "loss": 1.9223397970199585,
      "step": 2916
    },
    {
      "epoch": 0.2918,
      "grad_norm": 0.07332754880189896,
      "learning_rate": 0.002,
      "loss": 1.793445348739624,
      "step": 2918
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.07485819607973099,
      "learning_rate": 0.002,
      "loss": 1.9488404989242554,
      "step": 2920
    },
    {
      "epoch": 0.2922,
      "grad_norm": 0.07355786114931107,
      "learning_rate": 0.002,
      "loss": 1.879209280014038,
      "step": 2922
    },
    {
      "epoch": 0.2924,
      "grad_norm": 0.07534211874008179,
      "learning_rate": 0.002,
      "loss": 1.910578727722168,
      "step": 2924
    },
    {
      "epoch": 0.2926,
      "grad_norm": 0.07705911993980408,
      "learning_rate": 0.002,
      "loss": 2.0061142444610596,
      "step": 2926
    },
    {
      "epoch": 0.2928,
      "grad_norm": 0.07320701330900192,
      "learning_rate": 0.002,
      "loss": 1.7718427181243896,
      "step": 2928
    },
    {
      "epoch": 0.293,
      "grad_norm": 0.07583452761173248,
      "learning_rate": 0.002,
      "loss": 1.9010738134384155,
      "step": 2930
    },
    {
      "epoch": 0.2932,
      "grad_norm": 0.07260576635599136,
      "learning_rate": 0.002,
      "loss": 1.8146823644638062,
      "step": 2932
    },
    {
      "epoch": 0.2934,
      "grad_norm": 0.08015233278274536,
      "learning_rate": 0.002,
      "loss": 1.9116954803466797,
      "step": 2934
    },
    {
      "epoch": 0.2936,
      "grad_norm": 0.07816638797521591,
      "learning_rate": 0.002,
      "loss": 1.7840588092803955,
      "step": 2936
    },
    {
      "epoch": 0.2938,
      "grad_norm": 0.07579055428504944,
      "learning_rate": 0.002,
      "loss": 1.8917311429977417,
      "step": 2938
    },
    {
      "epoch": 0.294,
      "grad_norm": 0.08254476636648178,
      "learning_rate": 0.002,
      "loss": 1.956274390220642,
      "step": 2940
    },
    {
      "epoch": 0.2942,
      "grad_norm": 0.07264906913042068,
      "learning_rate": 0.002,
      "loss": 1.7964205741882324,
      "step": 2942
    },
    {
      "epoch": 0.2944,
      "grad_norm": 0.07300266623497009,
      "learning_rate": 0.002,
      "loss": 1.8886570930480957,
      "step": 2944
    },
    {
      "epoch": 0.2946,
      "grad_norm": 0.07698764652013779,
      "learning_rate": 0.002,
      "loss": 1.9245014190673828,
      "step": 2946
    },
    {
      "epoch": 0.2948,
      "grad_norm": 0.07654164731502533,
      "learning_rate": 0.002,
      "loss": 1.873954176902771,
      "step": 2948
    },
    {
      "epoch": 0.295,
      "grad_norm": 0.07511891424655914,
      "learning_rate": 0.002,
      "loss": 1.9830503463745117,
      "step": 2950
    },
    {
      "epoch": 0.2952,
      "grad_norm": 0.07719919830560684,
      "learning_rate": 0.002,
      "loss": 1.9496874809265137,
      "step": 2952
    },
    {
      "epoch": 0.2954,
      "grad_norm": 0.07752838730812073,
      "learning_rate": 0.002,
      "loss": 1.9158222675323486,
      "step": 2954
    },
    {
      "epoch": 0.2956,
      "grad_norm": 0.07471409440040588,
      "learning_rate": 0.002,
      "loss": 1.8874256610870361,
      "step": 2956
    },
    {
      "epoch": 0.2958,
      "grad_norm": 0.07485508173704147,
      "learning_rate": 0.002,
      "loss": 1.9469550848007202,
      "step": 2958
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.07555102556943893,
      "learning_rate": 0.002,
      "loss": 1.9700407981872559,
      "step": 2960
    },
    {
      "epoch": 0.2962,
      "grad_norm": 0.07959511131048203,
      "learning_rate": 0.002,
      "loss": 1.961776852607727,
      "step": 2962
    },
    {
      "epoch": 0.2964,
      "grad_norm": 0.07360249757766724,
      "learning_rate": 0.002,
      "loss": 1.9538118839263916,
      "step": 2964
    },
    {
      "epoch": 0.2966,
      "grad_norm": 0.07642817497253418,
      "learning_rate": 0.002,
      "loss": 1.8219969272613525,
      "step": 2966
    },
    {
      "epoch": 0.2968,
      "grad_norm": 0.07933645695447922,
      "learning_rate": 0.002,
      "loss": 1.9186493158340454,
      "step": 2968
    },
    {
      "epoch": 0.297,
      "grad_norm": 0.07606649398803711,
      "learning_rate": 0.002,
      "loss": 2.058760404586792,
      "step": 2970
    },
    {
      "epoch": 0.2972,
      "grad_norm": 0.07512557506561279,
      "learning_rate": 0.002,
      "loss": 1.9306681156158447,
      "step": 2972
    },
    {
      "epoch": 0.2974,
      "grad_norm": 0.07431517541408539,
      "learning_rate": 0.002,
      "loss": 1.880419373512268,
      "step": 2974
    },
    {
      "epoch": 0.2976,
      "grad_norm": 0.08510194718837738,
      "learning_rate": 0.002,
      "loss": 1.963371753692627,
      "step": 2976
    },
    {
      "epoch": 0.2978,
      "grad_norm": 0.07280287891626358,
      "learning_rate": 0.002,
      "loss": 1.8742859363555908,
      "step": 2978
    },
    {
      "epoch": 0.298,
      "grad_norm": 0.07997643947601318,
      "learning_rate": 0.002,
      "loss": 1.8870775699615479,
      "step": 2980
    },
    {
      "epoch": 0.2982,
      "grad_norm": 0.08231616765260696,
      "learning_rate": 0.002,
      "loss": 1.831709384918213,
      "step": 2982
    },
    {
      "epoch": 0.2984,
      "grad_norm": 0.0754353329539299,
      "learning_rate": 0.002,
      "loss": 1.9233675003051758,
      "step": 2984
    },
    {
      "epoch": 0.2986,
      "grad_norm": 0.07707013934850693,
      "learning_rate": 0.002,
      "loss": 1.8049591779708862,
      "step": 2986
    },
    {
      "epoch": 0.2988,
      "grad_norm": 0.07766082137823105,
      "learning_rate": 0.002,
      "loss": 1.9054222106933594,
      "step": 2988
    },
    {
      "epoch": 0.299,
      "grad_norm": 0.07823671400547028,
      "learning_rate": 0.002,
      "loss": 1.9719511270523071,
      "step": 2990
    },
    {
      "epoch": 0.2992,
      "grad_norm": 0.07880055159330368,
      "learning_rate": 0.002,
      "loss": 2.097489595413208,
      "step": 2992
    },
    {
      "epoch": 0.2994,
      "grad_norm": 0.0771835669875145,
      "learning_rate": 0.002,
      "loss": 1.9743378162384033,
      "step": 2994
    },
    {
      "epoch": 0.2996,
      "grad_norm": 0.07670241594314575,
      "learning_rate": 0.002,
      "loss": 1.89044189453125,
      "step": 2996
    },
    {
      "epoch": 0.2998,
      "grad_norm": 0.07632274180650711,
      "learning_rate": 0.002,
      "loss": 1.8729441165924072,
      "step": 2998
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.07683467119932175,
      "learning_rate": 0.002,
      "loss": 1.9167380332946777,
      "step": 3000
    },
    {
      "epoch": 0.3002,
      "grad_norm": 0.0798904076218605,
      "learning_rate": 0.002,
      "loss": 1.781358242034912,
      "step": 3002
    },
    {
      "epoch": 0.3004,
      "grad_norm": 0.07926058769226074,
      "learning_rate": 0.002,
      "loss": 1.9474997520446777,
      "step": 3004
    },
    {
      "epoch": 0.3006,
      "grad_norm": 0.07467661052942276,
      "learning_rate": 0.002,
      "loss": 1.9832974672317505,
      "step": 3006
    },
    {
      "epoch": 0.3008,
      "grad_norm": 0.07458962500095367,
      "learning_rate": 0.002,
      "loss": 1.888282060623169,
      "step": 3008
    },
    {
      "epoch": 0.301,
      "grad_norm": 0.0752609372138977,
      "learning_rate": 0.002,
      "loss": 1.8609284162521362,
      "step": 3010
    },
    {
      "epoch": 0.3012,
      "grad_norm": 0.07670833170413971,
      "learning_rate": 0.002,
      "loss": 1.8265644311904907,
      "step": 3012
    },
    {
      "epoch": 0.3014,
      "grad_norm": 0.0742899626493454,
      "learning_rate": 0.002,
      "loss": 1.968847393989563,
      "step": 3014
    },
    {
      "epoch": 0.3016,
      "grad_norm": 0.07967586070299149,
      "learning_rate": 0.002,
      "loss": 1.8724147081375122,
      "step": 3016
    },
    {
      "epoch": 0.3018,
      "grad_norm": 0.07368689775466919,
      "learning_rate": 0.002,
      "loss": 1.8584001064300537,
      "step": 3018
    },
    {
      "epoch": 0.302,
      "grad_norm": 0.08419319242238998,
      "learning_rate": 0.002,
      "loss": 1.9384797811508179,
      "step": 3020
    },
    {
      "epoch": 0.3022,
      "grad_norm": 0.07812511920928955,
      "learning_rate": 0.002,
      "loss": 1.9089970588684082,
      "step": 3022
    },
    {
      "epoch": 0.3024,
      "grad_norm": 0.0751846581697464,
      "learning_rate": 0.002,
      "loss": 1.9349353313446045,
      "step": 3024
    },
    {
      "epoch": 0.3026,
      "grad_norm": 0.07581477612257004,
      "learning_rate": 0.002,
      "loss": 2.0207600593566895,
      "step": 3026
    },
    {
      "epoch": 0.3028,
      "grad_norm": 0.07652150839567184,
      "learning_rate": 0.002,
      "loss": 1.8676458597183228,
      "step": 3028
    },
    {
      "epoch": 0.303,
      "grad_norm": 0.07458852231502533,
      "learning_rate": 0.002,
      "loss": 1.7784605026245117,
      "step": 3030
    },
    {
      "epoch": 0.3032,
      "grad_norm": 0.07359815388917923,
      "learning_rate": 0.002,
      "loss": 1.983904242515564,
      "step": 3032
    },
    {
      "epoch": 0.3034,
      "grad_norm": 0.08367784321308136,
      "learning_rate": 0.002,
      "loss": 1.8834031820297241,
      "step": 3034
    },
    {
      "epoch": 0.3036,
      "grad_norm": 0.07777772098779678,
      "learning_rate": 0.002,
      "loss": 2.003781795501709,
      "step": 3036
    },
    {
      "epoch": 0.3038,
      "grad_norm": 0.08079299330711365,
      "learning_rate": 0.002,
      "loss": 1.9625518321990967,
      "step": 3038
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.07455991953611374,
      "learning_rate": 0.002,
      "loss": 1.898703694343567,
      "step": 3040
    },
    {
      "epoch": 0.3042,
      "grad_norm": 0.07694865018129349,
      "learning_rate": 0.002,
      "loss": 1.9537911415100098,
      "step": 3042
    },
    {
      "epoch": 0.3044,
      "grad_norm": 0.07266010344028473,
      "learning_rate": 0.002,
      "loss": 2.0086374282836914,
      "step": 3044
    },
    {
      "epoch": 0.3046,
      "grad_norm": 0.07769840210676193,
      "learning_rate": 0.002,
      "loss": 1.9524153470993042,
      "step": 3046
    },
    {
      "epoch": 0.3048,
      "grad_norm": 0.07620533555746078,
      "learning_rate": 0.002,
      "loss": 1.8888816833496094,
      "step": 3048
    },
    {
      "epoch": 0.305,
      "grad_norm": 0.07878359407186508,
      "learning_rate": 0.002,
      "loss": 1.8178718090057373,
      "step": 3050
    },
    {
      "epoch": 0.3052,
      "grad_norm": 0.07455829530954361,
      "learning_rate": 0.002,
      "loss": 1.9701063632965088,
      "step": 3052
    },
    {
      "epoch": 0.3054,
      "grad_norm": 0.07732651382684708,
      "learning_rate": 0.002,
      "loss": 1.9621189832687378,
      "step": 3054
    },
    {
      "epoch": 0.3056,
      "grad_norm": 0.07514136284589767,
      "learning_rate": 0.002,
      "loss": 1.9390573501586914,
      "step": 3056
    },
    {
      "epoch": 0.3058,
      "grad_norm": 0.07669403403997421,
      "learning_rate": 0.002,
      "loss": 1.8689807653427124,
      "step": 3058
    },
    {
      "epoch": 0.306,
      "grad_norm": 0.08058839291334152,
      "learning_rate": 0.002,
      "loss": 1.8602112531661987,
      "step": 3060
    },
    {
      "epoch": 0.3062,
      "grad_norm": 0.08016939461231232,
      "learning_rate": 0.002,
      "loss": 1.8787204027175903,
      "step": 3062
    },
    {
      "epoch": 0.3064,
      "grad_norm": 0.07821451127529144,
      "learning_rate": 0.002,
      "loss": 1.8528647422790527,
      "step": 3064
    },
    {
      "epoch": 0.3066,
      "grad_norm": 0.0826551616191864,
      "learning_rate": 0.002,
      "loss": 2.0013554096221924,
      "step": 3066
    },
    {
      "epoch": 0.3068,
      "grad_norm": 0.07207199186086655,
      "learning_rate": 0.002,
      "loss": 1.8164827823638916,
      "step": 3068
    },
    {
      "epoch": 0.307,
      "grad_norm": 0.075330950319767,
      "learning_rate": 0.002,
      "loss": 1.9723820686340332,
      "step": 3070
    },
    {
      "epoch": 0.3072,
      "grad_norm": 0.07763063162565231,
      "learning_rate": 0.002,
      "loss": 1.917581558227539,
      "step": 3072
    },
    {
      "epoch": 0.3074,
      "grad_norm": 0.07879910618066788,
      "learning_rate": 0.002,
      "loss": 1.868463158607483,
      "step": 3074
    },
    {
      "epoch": 0.3076,
      "grad_norm": 0.0817985013127327,
      "learning_rate": 0.002,
      "loss": 1.9404250383377075,
      "step": 3076
    },
    {
      "epoch": 0.3078,
      "grad_norm": 0.0794169083237648,
      "learning_rate": 0.002,
      "loss": 1.9624876976013184,
      "step": 3078
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.07899218052625656,
      "learning_rate": 0.002,
      "loss": 1.947484016418457,
      "step": 3080
    },
    {
      "epoch": 0.3082,
      "grad_norm": 0.0724065825343132,
      "learning_rate": 0.002,
      "loss": 1.9301636219024658,
      "step": 3082
    },
    {
      "epoch": 0.3084,
      "grad_norm": 0.07383622229099274,
      "learning_rate": 0.002,
      "loss": 1.9668265581130981,
      "step": 3084
    },
    {
      "epoch": 0.3086,
      "grad_norm": 0.07591451704502106,
      "learning_rate": 0.002,
      "loss": 1.8801486492156982,
      "step": 3086
    },
    {
      "epoch": 0.3088,
      "grad_norm": 0.07665316760540009,
      "learning_rate": 0.002,
      "loss": 1.9299181699752808,
      "step": 3088
    },
    {
      "epoch": 0.309,
      "grad_norm": 0.07808520644903183,
      "learning_rate": 0.002,
      "loss": 1.817266821861267,
      "step": 3090
    },
    {
      "epoch": 0.3092,
      "grad_norm": 0.07434350997209549,
      "learning_rate": 0.002,
      "loss": 1.8185352087020874,
      "step": 3092
    },
    {
      "epoch": 0.3094,
      "grad_norm": 0.07310652732849121,
      "learning_rate": 0.002,
      "loss": 1.7840343713760376,
      "step": 3094
    },
    {
      "epoch": 0.3096,
      "grad_norm": 0.0757940486073494,
      "learning_rate": 0.002,
      "loss": 1.913875937461853,
      "step": 3096
    },
    {
      "epoch": 0.3098,
      "grad_norm": 0.07390191406011581,
      "learning_rate": 0.002,
      "loss": 1.8846038579940796,
      "step": 3098
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.07946799695491791,
      "learning_rate": 0.002,
      "loss": 1.96548330783844,
      "step": 3100
    },
    {
      "epoch": 0.3102,
      "grad_norm": 0.07205086946487427,
      "learning_rate": 0.002,
      "loss": 1.9565844535827637,
      "step": 3102
    },
    {
      "epoch": 0.3104,
      "grad_norm": 0.07541175931692123,
      "learning_rate": 0.002,
      "loss": 1.8396351337432861,
      "step": 3104
    },
    {
      "epoch": 0.3106,
      "grad_norm": 0.07740567624568939,
      "learning_rate": 0.002,
      "loss": 1.8430784940719604,
      "step": 3106
    },
    {
      "epoch": 0.3108,
      "grad_norm": 0.07248397171497345,
      "learning_rate": 0.002,
      "loss": 2.006537914276123,
      "step": 3108
    },
    {
      "epoch": 0.311,
      "grad_norm": 0.07417640089988708,
      "learning_rate": 0.002,
      "loss": 1.9486442804336548,
      "step": 3110
    },
    {
      "epoch": 0.3112,
      "grad_norm": 0.07525241374969482,
      "learning_rate": 0.002,
      "loss": 1.831403374671936,
      "step": 3112
    },
    {
      "epoch": 0.3114,
      "grad_norm": 0.07479021698236465,
      "learning_rate": 0.002,
      "loss": 1.9738919734954834,
      "step": 3114
    },
    {
      "epoch": 0.3116,
      "grad_norm": 0.07460878789424896,
      "learning_rate": 0.002,
      "loss": 1.8967833518981934,
      "step": 3116
    },
    {
      "epoch": 0.3118,
      "grad_norm": 0.08069875836372375,
      "learning_rate": 0.002,
      "loss": 1.9096249341964722,
      "step": 3118
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.07790526747703552,
      "learning_rate": 0.002,
      "loss": 1.8123494386672974,
      "step": 3120
    },
    {
      "epoch": 0.3122,
      "grad_norm": 0.0753735899925232,
      "learning_rate": 0.002,
      "loss": 1.8249025344848633,
      "step": 3122
    },
    {
      "epoch": 0.3124,
      "grad_norm": 0.0765572190284729,
      "learning_rate": 0.002,
      "loss": 1.9944027662277222,
      "step": 3124
    },
    {
      "epoch": 0.3126,
      "grad_norm": 0.0734083354473114,
      "learning_rate": 0.002,
      "loss": 1.8731932640075684,
      "step": 3126
    },
    {
      "epoch": 0.3128,
      "grad_norm": 0.07741888612508774,
      "learning_rate": 0.002,
      "loss": 1.8853271007537842,
      "step": 3128
    },
    {
      "epoch": 0.313,
      "grad_norm": 0.07886173576116562,
      "learning_rate": 0.002,
      "loss": 1.8358485698699951,
      "step": 3130
    },
    {
      "epoch": 0.3132,
      "grad_norm": 0.07613502442836761,
      "learning_rate": 0.002,
      "loss": 1.7883868217468262,
      "step": 3132
    },
    {
      "epoch": 0.3134,
      "grad_norm": 0.07563106715679169,
      "learning_rate": 0.002,
      "loss": 2.01314377784729,
      "step": 3134
    },
    {
      "epoch": 0.3136,
      "grad_norm": 0.07666026800870895,
      "learning_rate": 0.002,
      "loss": 1.9545036554336548,
      "step": 3136
    },
    {
      "epoch": 0.3138,
      "grad_norm": 0.07890927791595459,
      "learning_rate": 0.002,
      "loss": 1.8676661252975464,
      "step": 3138
    },
    {
      "epoch": 0.314,
      "grad_norm": 0.07849795371294022,
      "learning_rate": 0.002,
      "loss": 1.8602464199066162,
      "step": 3140
    },
    {
      "epoch": 0.3142,
      "grad_norm": 0.07607519626617432,
      "learning_rate": 0.002,
      "loss": 1.9892072677612305,
      "step": 3142
    },
    {
      "epoch": 0.3144,
      "grad_norm": 0.07571230083703995,
      "learning_rate": 0.002,
      "loss": 1.8143649101257324,
      "step": 3144
    },
    {
      "epoch": 0.3146,
      "grad_norm": 0.07368621230125427,
      "learning_rate": 0.002,
      "loss": 1.9062687158584595,
      "step": 3146
    },
    {
      "epoch": 0.3148,
      "grad_norm": 0.07638990134000778,
      "learning_rate": 0.002,
      "loss": 1.9724677801132202,
      "step": 3148
    },
    {
      "epoch": 0.315,
      "grad_norm": 0.07682925462722778,
      "learning_rate": 0.002,
      "loss": 1.8819034099578857,
      "step": 3150
    },
    {
      "epoch": 0.3152,
      "grad_norm": 0.07444681972265244,
      "learning_rate": 0.002,
      "loss": 1.921888828277588,
      "step": 3152
    },
    {
      "epoch": 0.3154,
      "grad_norm": 0.07829263806343079,
      "learning_rate": 0.002,
      "loss": 1.898511290550232,
      "step": 3154
    },
    {
      "epoch": 0.3156,
      "grad_norm": 0.08068505674600601,
      "learning_rate": 0.002,
      "loss": 2.0541322231292725,
      "step": 3156
    },
    {
      "epoch": 0.3158,
      "grad_norm": 0.07389215379953384,
      "learning_rate": 0.002,
      "loss": 2.0057106018066406,
      "step": 3158
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.07520435005426407,
      "learning_rate": 0.002,
      "loss": 1.9146822690963745,
      "step": 3160
    },
    {
      "epoch": 0.3162,
      "grad_norm": 0.0740780308842659,
      "learning_rate": 0.002,
      "loss": 1.9016098976135254,
      "step": 3162
    },
    {
      "epoch": 0.3164,
      "grad_norm": 0.07809039205312729,
      "learning_rate": 0.002,
      "loss": 1.8177156448364258,
      "step": 3164
    },
    {
      "epoch": 0.3166,
      "grad_norm": 0.08293816447257996,
      "learning_rate": 0.002,
      "loss": 1.8428207635879517,
      "step": 3166
    },
    {
      "epoch": 0.3168,
      "grad_norm": 0.07506537437438965,
      "learning_rate": 0.002,
      "loss": 1.9099422693252563,
      "step": 3168
    },
    {
      "epoch": 0.317,
      "grad_norm": 0.0829731896519661,
      "learning_rate": 0.002,
      "loss": 1.9508088827133179,
      "step": 3170
    },
    {
      "epoch": 0.3172,
      "grad_norm": 0.08077346533536911,
      "learning_rate": 0.002,
      "loss": 1.9007725715637207,
      "step": 3172
    },
    {
      "epoch": 0.3174,
      "grad_norm": 0.07320143282413483,
      "learning_rate": 0.002,
      "loss": 2.01872181892395,
      "step": 3174
    },
    {
      "epoch": 0.3176,
      "grad_norm": 0.07783108949661255,
      "learning_rate": 0.002,
      "loss": 1.9276478290557861,
      "step": 3176
    },
    {
      "epoch": 0.3178,
      "grad_norm": 0.07973723113536835,
      "learning_rate": 0.002,
      "loss": 1.9631808996200562,
      "step": 3178
    },
    {
      "epoch": 0.318,
      "grad_norm": 0.08726296573877335,
      "learning_rate": 0.002,
      "loss": 1.9722362756729126,
      "step": 3180
    },
    {
      "epoch": 0.3182,
      "grad_norm": 0.07586264610290527,
      "learning_rate": 0.002,
      "loss": 2.0189192295074463,
      "step": 3182
    },
    {
      "epoch": 0.3184,
      "grad_norm": 0.07563851773738861,
      "learning_rate": 0.002,
      "loss": 1.9564708471298218,
      "step": 3184
    },
    {
      "epoch": 0.3186,
      "grad_norm": 0.07359036058187485,
      "learning_rate": 0.002,
      "loss": 1.991436243057251,
      "step": 3186
    },
    {
      "epoch": 0.3188,
      "grad_norm": 0.07420211285352707,
      "learning_rate": 0.002,
      "loss": 1.8151285648345947,
      "step": 3188
    },
    {
      "epoch": 0.319,
      "grad_norm": 0.08289330452680588,
      "learning_rate": 0.002,
      "loss": 1.9046801328659058,
      "step": 3190
    },
    {
      "epoch": 0.3192,
      "grad_norm": 0.07943344861268997,
      "learning_rate": 0.002,
      "loss": 1.9175231456756592,
      "step": 3192
    },
    {
      "epoch": 0.3194,
      "grad_norm": 0.07164470851421356,
      "learning_rate": 0.002,
      "loss": 1.8837794065475464,
      "step": 3194
    },
    {
      "epoch": 0.3196,
      "grad_norm": 0.0735173374414444,
      "learning_rate": 0.002,
      "loss": 1.8956973552703857,
      "step": 3196
    },
    {
      "epoch": 0.3198,
      "grad_norm": 0.08028364926576614,
      "learning_rate": 0.002,
      "loss": 1.859825611114502,
      "step": 3198
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.07518185675144196,
      "learning_rate": 0.002,
      "loss": 1.9789729118347168,
      "step": 3200
    },
    {
      "epoch": 0.3202,
      "grad_norm": 0.07408410310745239,
      "learning_rate": 0.002,
      "loss": 1.8435853719711304,
      "step": 3202
    },
    {
      "epoch": 0.3204,
      "grad_norm": 0.07469511032104492,
      "learning_rate": 0.002,
      "loss": 1.941458821296692,
      "step": 3204
    },
    {
      "epoch": 0.3206,
      "grad_norm": 0.07771731913089752,
      "learning_rate": 0.002,
      "loss": 2.008272647857666,
      "step": 3206
    },
    {
      "epoch": 0.3208,
      "grad_norm": 0.07569152861833572,
      "learning_rate": 0.002,
      "loss": 1.9496546983718872,
      "step": 3208
    },
    {
      "epoch": 0.321,
      "grad_norm": 0.0777445137500763,
      "learning_rate": 0.002,
      "loss": 1.833438754081726,
      "step": 3210
    },
    {
      "epoch": 0.3212,
      "grad_norm": 0.07313323020935059,
      "learning_rate": 0.002,
      "loss": 1.7675764560699463,
      "step": 3212
    },
    {
      "epoch": 0.3214,
      "grad_norm": 0.07628345489501953,
      "learning_rate": 0.002,
      "loss": 1.7947347164154053,
      "step": 3214
    },
    {
      "epoch": 0.3216,
      "grad_norm": 0.0810098946094513,
      "learning_rate": 0.002,
      "loss": 1.8292778730392456,
      "step": 3216
    },
    {
      "epoch": 0.3218,
      "grad_norm": 0.07760146260261536,
      "learning_rate": 0.002,
      "loss": 1.7980077266693115,
      "step": 3218
    },
    {
      "epoch": 0.322,
      "grad_norm": 0.0809379443526268,
      "learning_rate": 0.002,
      "loss": 1.8243210315704346,
      "step": 3220
    },
    {
      "epoch": 0.3222,
      "grad_norm": 0.07351391762495041,
      "learning_rate": 0.002,
      "loss": 1.8834315538406372,
      "step": 3222
    },
    {
      "epoch": 0.3224,
      "grad_norm": 0.07483334839344025,
      "learning_rate": 0.002,
      "loss": 1.774949312210083,
      "step": 3224
    },
    {
      "epoch": 0.3226,
      "grad_norm": 0.07556086778640747,
      "learning_rate": 0.002,
      "loss": 1.9882946014404297,
      "step": 3226
    },
    {
      "epoch": 0.3228,
      "grad_norm": 0.0749208927154541,
      "learning_rate": 0.002,
      "loss": 2.0168073177337646,
      "step": 3228
    },
    {
      "epoch": 0.323,
      "grad_norm": 0.07624028623104095,
      "learning_rate": 0.002,
      "loss": 1.8837870359420776,
      "step": 3230
    },
    {
      "epoch": 0.3232,
      "grad_norm": 0.076922707259655,
      "learning_rate": 0.002,
      "loss": 1.8674709796905518,
      "step": 3232
    },
    {
      "epoch": 0.3234,
      "grad_norm": 0.07699599862098694,
      "learning_rate": 0.002,
      "loss": 1.86445951461792,
      "step": 3234
    },
    {
      "epoch": 0.3236,
      "grad_norm": 0.07592649012804031,
      "learning_rate": 0.002,
      "loss": 1.941184639930725,
      "step": 3236
    },
    {
      "epoch": 0.3238,
      "grad_norm": 0.07442115247249603,
      "learning_rate": 0.002,
      "loss": 1.8273184299468994,
      "step": 3238
    },
    {
      "epoch": 0.324,
      "grad_norm": 0.0712459608912468,
      "learning_rate": 0.002,
      "loss": 1.9270199537277222,
      "step": 3240
    },
    {
      "epoch": 0.3242,
      "grad_norm": 0.07598423212766647,
      "learning_rate": 0.002,
      "loss": 1.8048747777938843,
      "step": 3242
    },
    {
      "epoch": 0.3244,
      "grad_norm": 0.08778064697980881,
      "learning_rate": 0.002,
      "loss": 1.9278223514556885,
      "step": 3244
    },
    {
      "epoch": 0.3246,
      "grad_norm": 0.07610313594341278,
      "learning_rate": 0.002,
      "loss": 1.8640797138214111,
      "step": 3246
    },
    {
      "epoch": 0.3248,
      "grad_norm": 0.07557915896177292,
      "learning_rate": 0.002,
      "loss": 1.832666277885437,
      "step": 3248
    },
    {
      "epoch": 0.325,
      "grad_norm": 0.07674980908632278,
      "learning_rate": 0.002,
      "loss": 1.9029488563537598,
      "step": 3250
    },
    {
      "epoch": 0.3252,
      "grad_norm": 0.07636220753192902,
      "learning_rate": 0.002,
      "loss": 1.7563024759292603,
      "step": 3252
    },
    {
      "epoch": 0.3254,
      "grad_norm": 0.07522439956665039,
      "learning_rate": 0.002,
      "loss": 1.8830978870391846,
      "step": 3254
    },
    {
      "epoch": 0.3256,
      "grad_norm": 0.07312452793121338,
      "learning_rate": 0.002,
      "loss": 2.0454447269439697,
      "step": 3256
    },
    {
      "epoch": 0.3258,
      "grad_norm": 0.07595562934875488,
      "learning_rate": 0.002,
      "loss": 1.894032597541809,
      "step": 3258
    },
    {
      "epoch": 0.326,
      "grad_norm": 0.07686661928892136,
      "learning_rate": 0.002,
      "loss": 1.8560539484024048,
      "step": 3260
    },
    {
      "epoch": 0.3262,
      "grad_norm": 0.07630996406078339,
      "learning_rate": 0.002,
      "loss": 1.9488720893859863,
      "step": 3262
    },
    {
      "epoch": 0.3264,
      "grad_norm": 0.07552891224622726,
      "learning_rate": 0.002,
      "loss": 1.9772744178771973,
      "step": 3264
    },
    {
      "epoch": 0.3266,
      "grad_norm": 0.07642145454883575,
      "learning_rate": 0.002,
      "loss": 1.8309482336044312,
      "step": 3266
    },
    {
      "epoch": 0.3268,
      "grad_norm": 0.08088517189025879,
      "learning_rate": 0.002,
      "loss": 1.833850383758545,
      "step": 3268
    },
    {
      "epoch": 0.327,
      "grad_norm": 0.07299275696277618,
      "learning_rate": 0.002,
      "loss": 1.9215556383132935,
      "step": 3270
    },
    {
      "epoch": 0.3272,
      "grad_norm": 0.07559346407651901,
      "learning_rate": 0.002,
      "loss": 1.9367531538009644,
      "step": 3272
    },
    {
      "epoch": 0.3274,
      "grad_norm": 0.07678866386413574,
      "learning_rate": 0.002,
      "loss": 1.9505507946014404,
      "step": 3274
    },
    {
      "epoch": 0.3276,
      "grad_norm": 0.07834633439779282,
      "learning_rate": 0.002,
      "loss": 1.8313273191452026,
      "step": 3276
    },
    {
      "epoch": 0.3278,
      "grad_norm": 0.08116936683654785,
      "learning_rate": 0.002,
      "loss": 1.8182188272476196,
      "step": 3278
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.07356537878513336,
      "learning_rate": 0.002,
      "loss": 1.9725091457366943,
      "step": 3280
    },
    {
      "epoch": 0.3282,
      "grad_norm": 0.07769383490085602,
      "learning_rate": 0.002,
      "loss": 2.028529167175293,
      "step": 3282
    },
    {
      "epoch": 0.3284,
      "grad_norm": 0.08050869405269623,
      "learning_rate": 0.002,
      "loss": 1.8236894607543945,
      "step": 3284
    },
    {
      "epoch": 0.3286,
      "grad_norm": 0.07695252448320389,
      "learning_rate": 0.002,
      "loss": 1.7754242420196533,
      "step": 3286
    },
    {
      "epoch": 0.3288,
      "grad_norm": 0.085331030189991,
      "learning_rate": 0.002,
      "loss": 1.8566300868988037,
      "step": 3288
    },
    {
      "epoch": 0.329,
      "grad_norm": 0.07717300206422806,
      "learning_rate": 0.002,
      "loss": 1.960071086883545,
      "step": 3290
    },
    {
      "epoch": 0.3292,
      "grad_norm": 0.0768945962190628,
      "learning_rate": 0.002,
      "loss": 1.93677818775177,
      "step": 3292
    },
    {
      "epoch": 0.3294,
      "grad_norm": 0.086123026907444,
      "learning_rate": 0.002,
      "loss": 2.0089359283447266,
      "step": 3294
    },
    {
      "epoch": 0.3296,
      "grad_norm": 0.07348354905843735,
      "learning_rate": 0.002,
      "loss": 1.862537145614624,
      "step": 3296
    },
    {
      "epoch": 0.3298,
      "grad_norm": 0.07827258110046387,
      "learning_rate": 0.002,
      "loss": 1.9634668827056885,
      "step": 3298
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.07319421321153641,
      "learning_rate": 0.002,
      "loss": 1.7778385877609253,
      "step": 3300
    },
    {
      "epoch": 0.3302,
      "grad_norm": 0.07813398540019989,
      "learning_rate": 0.002,
      "loss": 1.9175084829330444,
      "step": 3302
    },
    {
      "epoch": 0.3304,
      "grad_norm": 0.07616899162530899,
      "learning_rate": 0.002,
      "loss": 1.9376736879348755,
      "step": 3304
    },
    {
      "epoch": 0.3306,
      "grad_norm": 0.07601091265678406,
      "learning_rate": 0.002,
      "loss": 1.7474920749664307,
      "step": 3306
    },
    {
      "epoch": 0.3308,
      "grad_norm": 0.07941628992557526,
      "learning_rate": 0.002,
      "loss": 1.9102580547332764,
      "step": 3308
    },
    {
      "epoch": 0.331,
      "grad_norm": 0.07244042307138443,
      "learning_rate": 0.002,
      "loss": 1.8216360807418823,
      "step": 3310
    },
    {
      "epoch": 0.3312,
      "grad_norm": 0.07703053206205368,
      "learning_rate": 0.002,
      "loss": 1.7760800123214722,
      "step": 3312
    },
    {
      "epoch": 0.3314,
      "grad_norm": 0.07783530652523041,
      "learning_rate": 0.002,
      "loss": 1.9111908674240112,
      "step": 3314
    },
    {
      "epoch": 0.3316,
      "grad_norm": 0.07911171019077301,
      "learning_rate": 0.002,
      "loss": 1.9037055969238281,
      "step": 3316
    },
    {
      "epoch": 0.3318,
      "grad_norm": 0.07504582405090332,
      "learning_rate": 0.002,
      "loss": 1.8711320161819458,
      "step": 3318
    },
    {
      "epoch": 0.332,
      "grad_norm": 0.07258350402116776,
      "learning_rate": 0.002,
      "loss": 1.9042210578918457,
      "step": 3320
    },
    {
      "epoch": 0.3322,
      "grad_norm": 0.07594647258520126,
      "learning_rate": 0.002,
      "loss": 1.8098195791244507,
      "step": 3322
    },
    {
      "epoch": 0.3324,
      "grad_norm": 0.07726648449897766,
      "learning_rate": 0.002,
      "loss": 1.8114030361175537,
      "step": 3324
    },
    {
      "epoch": 0.3326,
      "grad_norm": 0.0761055126786232,
      "learning_rate": 0.002,
      "loss": 1.9696091413497925,
      "step": 3326
    },
    {
      "epoch": 0.3328,
      "grad_norm": 0.07757647335529327,
      "learning_rate": 0.002,
      "loss": 2.0265796184539795,
      "step": 3328
    },
    {
      "epoch": 0.333,
      "grad_norm": 0.07726210355758667,
      "learning_rate": 0.002,
      "loss": 1.9676856994628906,
      "step": 3330
    },
    {
      "epoch": 0.3332,
      "grad_norm": 0.07505819946527481,
      "learning_rate": 0.002,
      "loss": 1.831403374671936,
      "step": 3332
    },
    {
      "epoch": 0.3334,
      "grad_norm": 0.07664559036493301,
      "learning_rate": 0.002,
      "loss": 2.0923662185668945,
      "step": 3334
    },
    {
      "epoch": 0.3336,
      "grad_norm": 0.08089740574359894,
      "learning_rate": 0.002,
      "loss": 1.970381736755371,
      "step": 3336
    },
    {
      "epoch": 0.3338,
      "grad_norm": 0.0764792189002037,
      "learning_rate": 0.002,
      "loss": 1.9361944198608398,
      "step": 3338
    },
    {
      "epoch": 0.334,
      "grad_norm": 0.0727880671620369,
      "learning_rate": 0.002,
      "loss": 1.9340596199035645,
      "step": 3340
    },
    {
      "epoch": 0.3342,
      "grad_norm": 0.07873543351888657,
      "learning_rate": 0.002,
      "loss": 1.8810175657272339,
      "step": 3342
    },
    {
      "epoch": 0.3344,
      "grad_norm": 0.0760052502155304,
      "learning_rate": 0.002,
      "loss": 1.8208961486816406,
      "step": 3344
    },
    {
      "epoch": 0.3346,
      "grad_norm": 0.07719369977712631,
      "learning_rate": 0.002,
      "loss": 1.8286465406417847,
      "step": 3346
    },
    {
      "epoch": 0.3348,
      "grad_norm": 0.07322357594966888,
      "learning_rate": 0.002,
      "loss": 1.7764289379119873,
      "step": 3348
    },
    {
      "epoch": 0.335,
      "grad_norm": 0.07308059185743332,
      "learning_rate": 0.002,
      "loss": 1.9339851140975952,
      "step": 3350
    },
    {
      "epoch": 0.3352,
      "grad_norm": 0.07738848030567169,
      "learning_rate": 0.002,
      "loss": 1.793708324432373,
      "step": 3352
    },
    {
      "epoch": 0.3354,
      "grad_norm": 0.0726936087012291,
      "learning_rate": 0.002,
      "loss": 2.016817808151245,
      "step": 3354
    },
    {
      "epoch": 0.3356,
      "grad_norm": 0.07419881969690323,
      "learning_rate": 0.002,
      "loss": 1.8748533725738525,
      "step": 3356
    },
    {
      "epoch": 0.3358,
      "grad_norm": 0.08168778568506241,
      "learning_rate": 0.002,
      "loss": 1.84581458568573,
      "step": 3358
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.07467389106750488,
      "learning_rate": 0.002,
      "loss": 1.9205442667007446,
      "step": 3360
    },
    {
      "epoch": 0.3362,
      "grad_norm": 0.07521827518939972,
      "learning_rate": 0.002,
      "loss": 1.8237206935882568,
      "step": 3362
    },
    {
      "epoch": 0.3364,
      "grad_norm": 0.08208627998828888,
      "learning_rate": 0.002,
      "loss": 1.858108401298523,
      "step": 3364
    },
    {
      "epoch": 0.3366,
      "grad_norm": 0.07404500246047974,
      "learning_rate": 0.002,
      "loss": 1.9273269176483154,
      "step": 3366
    },
    {
      "epoch": 0.3368,
      "grad_norm": 0.07493554800748825,
      "learning_rate": 0.002,
      "loss": 1.8147317171096802,
      "step": 3368
    },
    {
      "epoch": 0.337,
      "grad_norm": 0.07343003898859024,
      "learning_rate": 0.002,
      "loss": 1.8375153541564941,
      "step": 3370
    },
    {
      "epoch": 0.3372,
      "grad_norm": 0.07511568814516068,
      "learning_rate": 0.002,
      "loss": 1.990877628326416,
      "step": 3372
    },
    {
      "epoch": 0.3374,
      "grad_norm": 0.07723553478717804,
      "learning_rate": 0.002,
      "loss": 2.0192558765411377,
      "step": 3374
    },
    {
      "epoch": 0.3376,
      "grad_norm": 0.07483410090208054,
      "learning_rate": 0.002,
      "loss": 1.9290932416915894,
      "step": 3376
    },
    {
      "epoch": 0.3378,
      "grad_norm": 0.07584553211927414,
      "learning_rate": 0.002,
      "loss": 1.910201907157898,
      "step": 3378
    },
    {
      "epoch": 0.338,
      "grad_norm": 0.07294777035713196,
      "learning_rate": 0.002,
      "loss": 1.9517730474472046,
      "step": 3380
    },
    {
      "epoch": 0.3382,
      "grad_norm": 0.07346012443304062,
      "learning_rate": 0.002,
      "loss": 1.812038779258728,
      "step": 3382
    },
    {
      "epoch": 0.3384,
      "grad_norm": 0.07314706593751907,
      "learning_rate": 0.002,
      "loss": 1.8150686025619507,
      "step": 3384
    },
    {
      "epoch": 0.3386,
      "grad_norm": 0.076694056391716,
      "learning_rate": 0.002,
      "loss": 1.8716307878494263,
      "step": 3386
    },
    {
      "epoch": 0.3388,
      "grad_norm": 0.07694721221923828,
      "learning_rate": 0.002,
      "loss": 2.0095765590667725,
      "step": 3388
    },
    {
      "epoch": 0.339,
      "grad_norm": 0.07940646260976791,
      "learning_rate": 0.002,
      "loss": 1.9510937929153442,
      "step": 3390
    },
    {
      "epoch": 0.3392,
      "grad_norm": 0.07678674161434174,
      "learning_rate": 0.002,
      "loss": 1.8469610214233398,
      "step": 3392
    },
    {
      "epoch": 0.3394,
      "grad_norm": 0.07457257807254791,
      "learning_rate": 0.002,
      "loss": 1.9866886138916016,
      "step": 3394
    },
    {
      "epoch": 0.3396,
      "grad_norm": 0.07715252041816711,
      "learning_rate": 0.002,
      "loss": 1.9017338752746582,
      "step": 3396
    },
    {
      "epoch": 0.3398,
      "grad_norm": 0.07309486716985703,
      "learning_rate": 0.002,
      "loss": 1.7388402223587036,
      "step": 3398
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.07684233784675598,
      "learning_rate": 0.002,
      "loss": 1.9150781631469727,
      "step": 3400
    },
    {
      "epoch": 0.3402,
      "grad_norm": 0.07382813096046448,
      "learning_rate": 0.002,
      "loss": 1.8986775875091553,
      "step": 3402
    },
    {
      "epoch": 0.3404,
      "grad_norm": 0.07401100546121597,
      "learning_rate": 0.002,
      "loss": 1.9583542346954346,
      "step": 3404
    },
    {
      "epoch": 0.3406,
      "grad_norm": 0.07305650413036346,
      "learning_rate": 0.002,
      "loss": 1.760957956314087,
      "step": 3406
    },
    {
      "epoch": 0.3408,
      "grad_norm": 0.0761677548289299,
      "learning_rate": 0.002,
      "loss": 1.9138352870941162,
      "step": 3408
    },
    {
      "epoch": 0.341,
      "grad_norm": 0.0761890634894371,
      "learning_rate": 0.002,
      "loss": 1.8779222965240479,
      "step": 3410
    },
    {
      "epoch": 0.3412,
      "grad_norm": 0.07562009990215302,
      "learning_rate": 0.002,
      "loss": 1.9403308629989624,
      "step": 3412
    },
    {
      "epoch": 0.3414,
      "grad_norm": 0.0803682804107666,
      "learning_rate": 0.002,
      "loss": 1.8801218271255493,
      "step": 3414
    },
    {
      "epoch": 0.3416,
      "grad_norm": 0.07718956470489502,
      "learning_rate": 0.002,
      "loss": 1.8235403299331665,
      "step": 3416
    },
    {
      "epoch": 0.3418,
      "grad_norm": 0.07743823528289795,
      "learning_rate": 0.002,
      "loss": 1.9759082794189453,
      "step": 3418
    },
    {
      "epoch": 0.342,
      "grad_norm": 0.08018504083156586,
      "learning_rate": 0.002,
      "loss": 1.7529040575027466,
      "step": 3420
    },
    {
      "epoch": 0.3422,
      "grad_norm": 0.07375405728816986,
      "learning_rate": 0.002,
      "loss": 1.7630765438079834,
      "step": 3422
    },
    {
      "epoch": 0.3424,
      "grad_norm": 0.07684863358736038,
      "learning_rate": 0.002,
      "loss": 1.909066081047058,
      "step": 3424
    },
    {
      "epoch": 0.3426,
      "grad_norm": 0.09051002562046051,
      "learning_rate": 0.002,
      "loss": 1.961614966392517,
      "step": 3426
    },
    {
      "epoch": 0.3428,
      "grad_norm": 0.0736667737364769,
      "learning_rate": 0.002,
      "loss": 1.975677490234375,
      "step": 3428
    },
    {
      "epoch": 0.343,
      "grad_norm": 0.07793472707271576,
      "learning_rate": 0.002,
      "loss": 2.0257883071899414,
      "step": 3430
    },
    {
      "epoch": 0.3432,
      "grad_norm": 0.07652001082897186,
      "learning_rate": 0.002,
      "loss": 1.9584827423095703,
      "step": 3432
    },
    {
      "epoch": 0.3434,
      "grad_norm": 0.07539180666208267,
      "learning_rate": 0.002,
      "loss": 1.7983207702636719,
      "step": 3434
    },
    {
      "epoch": 0.3436,
      "grad_norm": 0.07759294658899307,
      "learning_rate": 0.002,
      "loss": 1.9375067949295044,
      "step": 3436
    },
    {
      "epoch": 0.3438,
      "grad_norm": 0.0740964412689209,
      "learning_rate": 0.002,
      "loss": 1.9231829643249512,
      "step": 3438
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.07794846594333649,
      "learning_rate": 0.002,
      "loss": 1.9163390398025513,
      "step": 3440
    },
    {
      "epoch": 0.3442,
      "grad_norm": 0.07349395006895065,
      "learning_rate": 0.002,
      "loss": 1.9610203504562378,
      "step": 3442
    },
    {
      "epoch": 0.3444,
      "grad_norm": 0.0722380056977272,
      "learning_rate": 0.002,
      "loss": 1.7511911392211914,
      "step": 3444
    },
    {
      "epoch": 0.3446,
      "grad_norm": 0.07510639727115631,
      "learning_rate": 0.002,
      "loss": 1.7864331007003784,
      "step": 3446
    },
    {
      "epoch": 0.3448,
      "grad_norm": 0.0781417265534401,
      "learning_rate": 0.002,
      "loss": 1.8092836141586304,
      "step": 3448
    },
    {
      "epoch": 0.345,
      "grad_norm": 0.07531540840864182,
      "learning_rate": 0.002,
      "loss": 1.7954148054122925,
      "step": 3450
    },
    {
      "epoch": 0.3452,
      "grad_norm": 0.0753559097647667,
      "learning_rate": 0.002,
      "loss": 1.7577824592590332,
      "step": 3452
    },
    {
      "epoch": 0.3454,
      "grad_norm": 0.07799051702022552,
      "learning_rate": 0.002,
      "loss": 1.8407647609710693,
      "step": 3454
    },
    {
      "epoch": 0.3456,
      "grad_norm": 0.07585075497627258,
      "learning_rate": 0.002,
      "loss": 1.898616909980774,
      "step": 3456
    },
    {
      "epoch": 0.3458,
      "grad_norm": 0.073715440928936,
      "learning_rate": 0.002,
      "loss": 1.8044922351837158,
      "step": 3458
    },
    {
      "epoch": 0.346,
      "grad_norm": 0.07604110985994339,
      "learning_rate": 0.002,
      "loss": 1.9341946840286255,
      "step": 3460
    },
    {
      "epoch": 0.3462,
      "grad_norm": 0.07405763119459152,
      "learning_rate": 0.002,
      "loss": 1.7815486192703247,
      "step": 3462
    },
    {
      "epoch": 0.3464,
      "grad_norm": 0.07333449274301529,
      "learning_rate": 0.002,
      "loss": 1.9667179584503174,
      "step": 3464
    },
    {
      "epoch": 0.3466,
      "grad_norm": 0.0759732574224472,
      "learning_rate": 0.002,
      "loss": 1.8634480237960815,
      "step": 3466
    },
    {
      "epoch": 0.3468,
      "grad_norm": 0.07337398082017899,
      "learning_rate": 0.002,
      "loss": 1.9014747142791748,
      "step": 3468
    },
    {
      "epoch": 0.347,
      "grad_norm": 0.07958661764860153,
      "learning_rate": 0.002,
      "loss": 1.932357907295227,
      "step": 3470
    },
    {
      "epoch": 0.3472,
      "grad_norm": 0.07395008951425552,
      "learning_rate": 0.002,
      "loss": 1.9656627178192139,
      "step": 3472
    },
    {
      "epoch": 0.3474,
      "grad_norm": 0.07776951044797897,
      "learning_rate": 0.002,
      "loss": 1.8677692413330078,
      "step": 3474
    },
    {
      "epoch": 0.3476,
      "grad_norm": 0.07668683677911758,
      "learning_rate": 0.002,
      "loss": 1.9925886392593384,
      "step": 3476
    },
    {
      "epoch": 0.3478,
      "grad_norm": 0.08748235553503036,
      "learning_rate": 0.002,
      "loss": 1.8483613729476929,
      "step": 3478
    },
    {
      "epoch": 0.348,
      "grad_norm": 0.07136409729719162,
      "learning_rate": 0.002,
      "loss": 1.83815598487854,
      "step": 3480
    },
    {
      "epoch": 0.3482,
      "grad_norm": 0.08374970406293869,
      "learning_rate": 0.002,
      "loss": 1.8330681324005127,
      "step": 3482
    },
    {
      "epoch": 0.3484,
      "grad_norm": 0.08480751514434814,
      "learning_rate": 0.002,
      "loss": 2.036363363265991,
      "step": 3484
    },
    {
      "epoch": 0.3486,
      "grad_norm": 0.07811633497476578,
      "learning_rate": 0.002,
      "loss": 1.981468915939331,
      "step": 3486
    },
    {
      "epoch": 0.3488,
      "grad_norm": 0.07695946842432022,
      "learning_rate": 0.002,
      "loss": 1.8358709812164307,
      "step": 3488
    },
    {
      "epoch": 0.349,
      "grad_norm": 0.07570044696331024,
      "learning_rate": 0.002,
      "loss": 1.8021857738494873,
      "step": 3490
    },
    {
      "epoch": 0.3492,
      "grad_norm": 0.07773613929748535,
      "learning_rate": 0.002,
      "loss": 1.867611289024353,
      "step": 3492
    },
    {
      "epoch": 0.3494,
      "grad_norm": 0.07604554295539856,
      "learning_rate": 0.002,
      "loss": 1.9359266757965088,
      "step": 3494
    },
    {
      "epoch": 0.3496,
      "grad_norm": 0.08080974221229553,
      "learning_rate": 0.002,
      "loss": 1.8534070253372192,
      "step": 3496
    },
    {
      "epoch": 0.3498,
      "grad_norm": 0.07892777770757675,
      "learning_rate": 0.002,
      "loss": 1.9366590976715088,
      "step": 3498
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.07488248497247696,
      "learning_rate": 0.002,
      "loss": 1.8171104192733765,
      "step": 3500
    },
    {
      "epoch": 0.3502,
      "grad_norm": 0.07802204042673111,
      "learning_rate": 0.002,
      "loss": 1.9493532180786133,
      "step": 3502
    },
    {
      "epoch": 0.3504,
      "grad_norm": 0.08743824809789658,
      "learning_rate": 0.002,
      "loss": 1.7907127141952515,
      "step": 3504
    },
    {
      "epoch": 0.3506,
      "grad_norm": 0.08003289252519608,
      "learning_rate": 0.002,
      "loss": 1.870506763458252,
      "step": 3506
    },
    {
      "epoch": 0.3508,
      "grad_norm": 0.0768069177865982,
      "learning_rate": 0.002,
      "loss": 1.934154748916626,
      "step": 3508
    },
    {
      "epoch": 0.351,
      "grad_norm": 0.07850606739521027,
      "learning_rate": 0.002,
      "loss": 1.8731114864349365,
      "step": 3510
    },
    {
      "epoch": 0.3512,
      "grad_norm": 0.07760598510503769,
      "learning_rate": 0.002,
      "loss": 1.878669023513794,
      "step": 3512
    },
    {
      "epoch": 0.3514,
      "grad_norm": 0.0784652978181839,
      "learning_rate": 0.002,
      "loss": 1.8762985467910767,
      "step": 3514
    },
    {
      "epoch": 0.3516,
      "grad_norm": 0.07590148597955704,
      "learning_rate": 0.002,
      "loss": 1.7939192056655884,
      "step": 3516
    },
    {
      "epoch": 0.3518,
      "grad_norm": 0.07830556482076645,
      "learning_rate": 0.002,
      "loss": 1.9122284650802612,
      "step": 3518
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.0782352164387703,
      "learning_rate": 0.002,
      "loss": 1.8781335353851318,
      "step": 3520
    },
    {
      "epoch": 0.3522,
      "grad_norm": 0.07482749968767166,
      "learning_rate": 0.002,
      "loss": 2.0248730182647705,
      "step": 3522
    },
    {
      "epoch": 0.3524,
      "grad_norm": 0.07477255910634995,
      "learning_rate": 0.002,
      "loss": 1.9689396619796753,
      "step": 3524
    },
    {
      "epoch": 0.3526,
      "grad_norm": 0.07457265257835388,
      "learning_rate": 0.002,
      "loss": 1.957615613937378,
      "step": 3526
    },
    {
      "epoch": 0.3528,
      "grad_norm": 0.07856965065002441,
      "learning_rate": 0.002,
      "loss": 1.9843051433563232,
      "step": 3528
    },
    {
      "epoch": 0.353,
      "grad_norm": 0.07721417397260666,
      "learning_rate": 0.002,
      "loss": 1.9233367443084717,
      "step": 3530
    },
    {
      "epoch": 0.3532,
      "grad_norm": 0.08766228705644608,
      "learning_rate": 0.002,
      "loss": 1.8555785417556763,
      "step": 3532
    },
    {
      "epoch": 0.3534,
      "grad_norm": 0.07900475710630417,
      "learning_rate": 0.002,
      "loss": 1.9299495220184326,
      "step": 3534
    },
    {
      "epoch": 0.3536,
      "grad_norm": 0.0793154239654541,
      "learning_rate": 0.002,
      "loss": 1.9132239818572998,
      "step": 3536
    },
    {
      "epoch": 0.3538,
      "grad_norm": 0.07481586933135986,
      "learning_rate": 0.002,
      "loss": 1.8572911024093628,
      "step": 3538
    },
    {
      "epoch": 0.354,
      "grad_norm": 0.07758516073226929,
      "learning_rate": 0.002,
      "loss": 1.8385486602783203,
      "step": 3540
    },
    {
      "epoch": 0.3542,
      "grad_norm": 0.0746159702539444,
      "learning_rate": 0.002,
      "loss": 1.9528422355651855,
      "step": 3542
    },
    {
      "epoch": 0.3544,
      "grad_norm": 0.0743945837020874,
      "learning_rate": 0.002,
      "loss": 1.8504002094268799,
      "step": 3544
    },
    {
      "epoch": 0.3546,
      "grad_norm": 0.07991514354944229,
      "learning_rate": 0.002,
      "loss": 1.8021398782730103,
      "step": 3546
    },
    {
      "epoch": 0.3548,
      "grad_norm": 0.0771934986114502,
      "learning_rate": 0.002,
      "loss": 1.9799325466156006,
      "step": 3548
    },
    {
      "epoch": 0.355,
      "grad_norm": 0.07649275660514832,
      "learning_rate": 0.002,
      "loss": 1.950627088546753,
      "step": 3550
    },
    {
      "epoch": 0.3552,
      "grad_norm": 0.07802007347345352,
      "learning_rate": 0.002,
      "loss": 1.891371726989746,
      "step": 3552
    },
    {
      "epoch": 0.3554,
      "grad_norm": 0.07554791867733002,
      "learning_rate": 0.002,
      "loss": 2.0269687175750732,
      "step": 3554
    },
    {
      "epoch": 0.3556,
      "grad_norm": 0.07768017053604126,
      "learning_rate": 0.002,
      "loss": 1.9512943029403687,
      "step": 3556
    },
    {
      "epoch": 0.3558,
      "grad_norm": 0.08097606897354126,
      "learning_rate": 0.002,
      "loss": 1.96127450466156,
      "step": 3558
    },
    {
      "epoch": 0.356,
      "grad_norm": 0.07538260519504547,
      "learning_rate": 0.002,
      "loss": 1.97255539894104,
      "step": 3560
    },
    {
      "epoch": 0.3562,
      "grad_norm": 0.07355134934186935,
      "learning_rate": 0.002,
      "loss": 1.8676389455795288,
      "step": 3562
    },
    {
      "epoch": 0.3564,
      "grad_norm": 0.07421302795410156,
      "learning_rate": 0.002,
      "loss": 1.7499287128448486,
      "step": 3564
    },
    {
      "epoch": 0.3566,
      "grad_norm": 0.07265985757112503,
      "learning_rate": 0.002,
      "loss": 1.8817510604858398,
      "step": 3566
    },
    {
      "epoch": 0.3568,
      "grad_norm": 0.0770273432135582,
      "learning_rate": 0.002,
      "loss": 1.8025826215744019,
      "step": 3568
    },
    {
      "epoch": 0.357,
      "grad_norm": 0.07646295428276062,
      "learning_rate": 0.002,
      "loss": 1.8557876348495483,
      "step": 3570
    },
    {
      "epoch": 0.3572,
      "grad_norm": 0.07299309968948364,
      "learning_rate": 0.002,
      "loss": 1.8870325088500977,
      "step": 3572
    },
    {
      "epoch": 0.3574,
      "grad_norm": 0.07363660633563995,
      "learning_rate": 0.002,
      "loss": 1.9088881015777588,
      "step": 3574
    },
    {
      "epoch": 0.3576,
      "grad_norm": 0.0805443748831749,
      "learning_rate": 0.002,
      "loss": 1.9804097414016724,
      "step": 3576
    },
    {
      "epoch": 0.3578,
      "grad_norm": 0.07430991530418396,
      "learning_rate": 0.002,
      "loss": 1.939389944076538,
      "step": 3578
    },
    {
      "epoch": 0.358,
      "grad_norm": 0.07618822902441025,
      "learning_rate": 0.002,
      "loss": 1.8700499534606934,
      "step": 3580
    },
    {
      "epoch": 0.3582,
      "grad_norm": 0.08192665874958038,
      "learning_rate": 0.002,
      "loss": 1.9001400470733643,
      "step": 3582
    },
    {
      "epoch": 0.3584,
      "grad_norm": 0.07590702176094055,
      "learning_rate": 0.002,
      "loss": 1.8537917137145996,
      "step": 3584
    },
    {
      "epoch": 0.3586,
      "grad_norm": 0.07686740905046463,
      "learning_rate": 0.002,
      "loss": 1.8113338947296143,
      "step": 3586
    },
    {
      "epoch": 0.3588,
      "grad_norm": 0.07825061678886414,
      "learning_rate": 0.002,
      "loss": 1.8659025430679321,
      "step": 3588
    },
    {
      "epoch": 0.359,
      "grad_norm": 0.07414184510707855,
      "learning_rate": 0.002,
      "loss": 1.982071042060852,
      "step": 3590
    },
    {
      "epoch": 0.3592,
      "grad_norm": 0.07633881270885468,
      "learning_rate": 0.002,
      "loss": 1.918049931526184,
      "step": 3592
    },
    {
      "epoch": 0.3594,
      "grad_norm": 0.07198624312877655,
      "learning_rate": 0.002,
      "loss": 1.8706130981445312,
      "step": 3594
    },
    {
      "epoch": 0.3596,
      "grad_norm": 0.07314635068178177,
      "learning_rate": 0.002,
      "loss": 1.9254491329193115,
      "step": 3596
    },
    {
      "epoch": 0.3598,
      "grad_norm": 0.0796109065413475,
      "learning_rate": 0.002,
      "loss": 1.9545797109603882,
      "step": 3598
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.07849898934364319,
      "learning_rate": 0.002,
      "loss": 1.8556289672851562,
      "step": 3600
    },
    {
      "epoch": 0.3602,
      "grad_norm": 0.07312922924757004,
      "learning_rate": 0.002,
      "loss": 1.7145304679870605,
      "step": 3602
    },
    {
      "epoch": 0.3604,
      "grad_norm": 0.0749719887971878,
      "learning_rate": 0.002,
      "loss": 1.8941466808319092,
      "step": 3604
    },
    {
      "epoch": 0.3606,
      "grad_norm": 0.0786941722035408,
      "learning_rate": 0.002,
      "loss": 1.949841022491455,
      "step": 3606
    },
    {
      "epoch": 0.3608,
      "grad_norm": 0.07864448428153992,
      "learning_rate": 0.002,
      "loss": 1.9501556158065796,
      "step": 3608
    },
    {
      "epoch": 0.361,
      "grad_norm": 0.07621503621339798,
      "learning_rate": 0.002,
      "loss": 1.9628987312316895,
      "step": 3610
    },
    {
      "epoch": 0.3612,
      "grad_norm": 0.07476962357759476,
      "learning_rate": 0.002,
      "loss": 1.9307681322097778,
      "step": 3612
    },
    {
      "epoch": 0.3614,
      "grad_norm": 0.07531963288784027,
      "learning_rate": 0.002,
      "loss": 1.8445500135421753,
      "step": 3614
    },
    {
      "epoch": 0.3616,
      "grad_norm": 0.07630260288715363,
      "learning_rate": 0.002,
      "loss": 1.8434884548187256,
      "step": 3616
    },
    {
      "epoch": 0.3618,
      "grad_norm": 0.07521006464958191,
      "learning_rate": 0.002,
      "loss": 1.9191631078720093,
      "step": 3618
    },
    {
      "epoch": 0.362,
      "grad_norm": 0.07619635760784149,
      "learning_rate": 0.002,
      "loss": 1.8085963726043701,
      "step": 3620
    },
    {
      "epoch": 0.3622,
      "grad_norm": 0.07814514636993408,
      "learning_rate": 0.002,
      "loss": 1.8103314638137817,
      "step": 3622
    },
    {
      "epoch": 0.3624,
      "grad_norm": 0.07341709733009338,
      "learning_rate": 0.002,
      "loss": 1.9344607591629028,
      "step": 3624
    },
    {
      "epoch": 0.3626,
      "grad_norm": 0.07697868347167969,
      "learning_rate": 0.002,
      "loss": 1.9474862813949585,
      "step": 3626
    },
    {
      "epoch": 0.3628,
      "grad_norm": 0.08019787818193436,
      "learning_rate": 0.002,
      "loss": 1.8229756355285645,
      "step": 3628
    },
    {
      "epoch": 0.363,
      "grad_norm": 0.07676352560520172,
      "learning_rate": 0.002,
      "loss": 1.8258572816848755,
      "step": 3630
    },
    {
      "epoch": 0.3632,
      "grad_norm": 0.07181580364704132,
      "learning_rate": 0.002,
      "loss": 1.9110095500946045,
      "step": 3632
    },
    {
      "epoch": 0.3634,
      "grad_norm": 0.07514549791812897,
      "learning_rate": 0.002,
      "loss": 1.9150770902633667,
      "step": 3634
    },
    {
      "epoch": 0.3636,
      "grad_norm": 0.07647165656089783,
      "learning_rate": 0.002,
      "loss": 1.9291491508483887,
      "step": 3636
    },
    {
      "epoch": 0.3638,
      "grad_norm": 0.0763157308101654,
      "learning_rate": 0.002,
      "loss": 1.943799376487732,
      "step": 3638
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.0845099464058876,
      "learning_rate": 0.002,
      "loss": 1.796173334121704,
      "step": 3640
    },
    {
      "epoch": 0.3642,
      "grad_norm": 0.07672269642353058,
      "learning_rate": 0.002,
      "loss": 1.930159568786621,
      "step": 3642
    },
    {
      "epoch": 0.3644,
      "grad_norm": 0.07521148025989532,
      "learning_rate": 0.002,
      "loss": 1.895439863204956,
      "step": 3644
    },
    {
      "epoch": 0.3646,
      "grad_norm": 0.08141639083623886,
      "learning_rate": 0.002,
      "loss": 1.898674726486206,
      "step": 3646
    },
    {
      "epoch": 0.3648,
      "grad_norm": 0.07797038555145264,
      "learning_rate": 0.002,
      "loss": 1.9745128154754639,
      "step": 3648
    },
    {
      "epoch": 0.365,
      "grad_norm": 0.07423697412014008,
      "learning_rate": 0.002,
      "loss": 1.8893548250198364,
      "step": 3650
    },
    {
      "epoch": 0.3652,
      "grad_norm": 0.07703891396522522,
      "learning_rate": 0.002,
      "loss": 2.017723560333252,
      "step": 3652
    },
    {
      "epoch": 0.3654,
      "grad_norm": 0.0723886787891388,
      "learning_rate": 0.002,
      "loss": 1.8047925233840942,
      "step": 3654
    },
    {
      "epoch": 0.3656,
      "grad_norm": 0.07898402214050293,
      "learning_rate": 0.002,
      "loss": 1.8885740041732788,
      "step": 3656
    },
    {
      "epoch": 0.3658,
      "grad_norm": 0.07769224047660828,
      "learning_rate": 0.002,
      "loss": 1.7729603052139282,
      "step": 3658
    },
    {
      "epoch": 0.366,
      "grad_norm": 0.07699308544397354,
      "learning_rate": 0.002,
      "loss": 1.923966884613037,
      "step": 3660
    },
    {
      "epoch": 0.3662,
      "grad_norm": 0.07605497539043427,
      "learning_rate": 0.002,
      "loss": 1.9066624641418457,
      "step": 3662
    },
    {
      "epoch": 0.3664,
      "grad_norm": 0.0751192718744278,
      "learning_rate": 0.002,
      "loss": 1.7375801801681519,
      "step": 3664
    },
    {
      "epoch": 0.3666,
      "grad_norm": 0.07455801963806152,
      "learning_rate": 0.002,
      "loss": 1.8641417026519775,
      "step": 3666
    },
    {
      "epoch": 0.3668,
      "grad_norm": 0.07722664624452591,
      "learning_rate": 0.002,
      "loss": 2.0617659091949463,
      "step": 3668
    },
    {
      "epoch": 0.367,
      "grad_norm": 0.0835382491350174,
      "learning_rate": 0.002,
      "loss": 1.9540032148361206,
      "step": 3670
    },
    {
      "epoch": 0.3672,
      "grad_norm": 0.07129634916782379,
      "learning_rate": 0.002,
      "loss": 2.038173198699951,
      "step": 3672
    },
    {
      "epoch": 0.3674,
      "grad_norm": 0.07656315714120865,
      "learning_rate": 0.002,
      "loss": 1.8645180463790894,
      "step": 3674
    },
    {
      "epoch": 0.3676,
      "grad_norm": 0.0806804820895195,
      "learning_rate": 0.002,
      "loss": 1.9266833066940308,
      "step": 3676
    },
    {
      "epoch": 0.3678,
      "grad_norm": 0.0722256526350975,
      "learning_rate": 0.002,
      "loss": 1.792579174041748,
      "step": 3678
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.07823967933654785,
      "learning_rate": 0.002,
      "loss": 1.8987270593643188,
      "step": 3680
    },
    {
      "epoch": 0.3682,
      "grad_norm": 0.08028064668178558,
      "learning_rate": 0.002,
      "loss": 1.8091484308242798,
      "step": 3682
    },
    {
      "epoch": 0.3684,
      "grad_norm": 0.07667369395494461,
      "learning_rate": 0.002,
      "loss": 1.870315670967102,
      "step": 3684
    },
    {
      "epoch": 0.3686,
      "grad_norm": 0.082509845495224,
      "learning_rate": 0.002,
      "loss": 1.7806155681610107,
      "step": 3686
    },
    {
      "epoch": 0.3688,
      "grad_norm": 0.07264959067106247,
      "learning_rate": 0.002,
      "loss": 1.9883729219436646,
      "step": 3688
    },
    {
      "epoch": 0.369,
      "grad_norm": 0.07740284502506256,
      "learning_rate": 0.002,
      "loss": 1.9607720375061035,
      "step": 3690
    },
    {
      "epoch": 0.3692,
      "grad_norm": 0.08035602420568466,
      "learning_rate": 0.002,
      "loss": 1.9259860515594482,
      "step": 3692
    },
    {
      "epoch": 0.3694,
      "grad_norm": 0.07530100643634796,
      "learning_rate": 0.002,
      "loss": 1.8164478540420532,
      "step": 3694
    },
    {
      "epoch": 0.3696,
      "grad_norm": 0.07914891093969345,
      "learning_rate": 0.002,
      "loss": 1.9563355445861816,
      "step": 3696
    },
    {
      "epoch": 0.3698,
      "grad_norm": 0.0742560625076294,
      "learning_rate": 0.002,
      "loss": 1.8365286588668823,
      "step": 3698
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.07518824934959412,
      "learning_rate": 0.002,
      "loss": 1.8583500385284424,
      "step": 3700
    },
    {
      "epoch": 0.3702,
      "grad_norm": 0.07884156703948975,
      "learning_rate": 0.002,
      "loss": 1.8358502388000488,
      "step": 3702
    },
    {
      "epoch": 0.3704,
      "grad_norm": 0.07558273524045944,
      "learning_rate": 0.002,
      "loss": 1.8728877305984497,
      "step": 3704
    },
    {
      "epoch": 0.3706,
      "grad_norm": 0.07476159185171127,
      "learning_rate": 0.002,
      "loss": 1.9256298542022705,
      "step": 3706
    },
    {
      "epoch": 0.3708,
      "grad_norm": 0.07767853140830994,
      "learning_rate": 0.002,
      "loss": 1.889451503753662,
      "step": 3708
    },
    {
      "epoch": 0.371,
      "grad_norm": 0.07584236562252045,
      "learning_rate": 0.002,
      "loss": 1.9503493309020996,
      "step": 3710
    },
    {
      "epoch": 0.3712,
      "grad_norm": 0.07738467305898666,
      "learning_rate": 0.002,
      "loss": 1.8716189861297607,
      "step": 3712
    },
    {
      "epoch": 0.3714,
      "grad_norm": 0.08234751224517822,
      "learning_rate": 0.002,
      "loss": 1.8933093547821045,
      "step": 3714
    },
    {
      "epoch": 0.3716,
      "grad_norm": 0.07703767716884613,
      "learning_rate": 0.002,
      "loss": 1.9241031408309937,
      "step": 3716
    },
    {
      "epoch": 0.3718,
      "grad_norm": 0.07229023426771164,
      "learning_rate": 0.002,
      "loss": 1.974159836769104,
      "step": 3718
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.07968708872795105,
      "learning_rate": 0.002,
      "loss": 1.857396125793457,
      "step": 3720
    },
    {
      "epoch": 0.3722,
      "grad_norm": 0.07678307592868805,
      "learning_rate": 0.002,
      "loss": 1.9408193826675415,
      "step": 3722
    },
    {
      "epoch": 0.3724,
      "grad_norm": 0.07597918808460236,
      "learning_rate": 0.002,
      "loss": 1.8608862161636353,
      "step": 3724
    },
    {
      "epoch": 0.3726,
      "grad_norm": 0.0756707638502121,
      "learning_rate": 0.002,
      "loss": 1.9049042463302612,
      "step": 3726
    },
    {
      "epoch": 0.3728,
      "grad_norm": 0.07919974625110626,
      "learning_rate": 0.002,
      "loss": 1.9156627655029297,
      "step": 3728
    },
    {
      "epoch": 0.373,
      "grad_norm": 0.08001970499753952,
      "learning_rate": 0.002,
      "loss": 1.9354339838027954,
      "step": 3730
    },
    {
      "epoch": 0.3732,
      "grad_norm": 0.07884741574525833,
      "learning_rate": 0.002,
      "loss": 1.8547817468643188,
      "step": 3732
    },
    {
      "epoch": 0.3734,
      "grad_norm": 0.08191996067762375,
      "learning_rate": 0.002,
      "loss": 1.9197579622268677,
      "step": 3734
    },
    {
      "epoch": 0.3736,
      "grad_norm": 0.08238981664180756,
      "learning_rate": 0.002,
      "loss": 2.0622692108154297,
      "step": 3736
    },
    {
      "epoch": 0.3738,
      "grad_norm": 0.07578539103269577,
      "learning_rate": 0.002,
      "loss": 1.9987757205963135,
      "step": 3738
    },
    {
      "epoch": 0.374,
      "grad_norm": 0.07833431661128998,
      "learning_rate": 0.002,
      "loss": 1.968658685684204,
      "step": 3740
    },
    {
      "epoch": 0.3742,
      "grad_norm": 0.07417851686477661,
      "learning_rate": 0.002,
      "loss": 1.8539870977401733,
      "step": 3742
    },
    {
      "epoch": 0.3744,
      "grad_norm": 0.07481931895017624,
      "learning_rate": 0.002,
      "loss": 1.8673652410507202,
      "step": 3744
    },
    {
      "epoch": 0.3746,
      "grad_norm": 0.07641009241342545,
      "learning_rate": 0.002,
      "loss": 1.8864761590957642,
      "step": 3746
    },
    {
      "epoch": 0.3748,
      "grad_norm": 0.07645934820175171,
      "learning_rate": 0.002,
      "loss": 1.865876317024231,
      "step": 3748
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.07857610285282135,
      "learning_rate": 0.002,
      "loss": 2.1402676105499268,
      "step": 3750
    },
    {
      "epoch": 0.3752,
      "grad_norm": 0.07557015866041183,
      "learning_rate": 0.002,
      "loss": 1.8448972702026367,
      "step": 3752
    },
    {
      "epoch": 0.3754,
      "grad_norm": 0.07534940540790558,
      "learning_rate": 0.002,
      "loss": 1.7901828289031982,
      "step": 3754
    },
    {
      "epoch": 0.3756,
      "grad_norm": 0.07879865914583206,
      "learning_rate": 0.002,
      "loss": 2.0152993202209473,
      "step": 3756
    },
    {
      "epoch": 0.3758,
      "grad_norm": 0.07804311066865921,
      "learning_rate": 0.002,
      "loss": 1.8522676229476929,
      "step": 3758
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.07653983682394028,
      "learning_rate": 0.002,
      "loss": 1.909687876701355,
      "step": 3760
    },
    {
      "epoch": 0.3762,
      "grad_norm": 0.07828226685523987,
      "learning_rate": 0.002,
      "loss": 1.8164770603179932,
      "step": 3762
    },
    {
      "epoch": 0.3764,
      "grad_norm": 0.0762704536318779,
      "learning_rate": 0.002,
      "loss": 1.9475700855255127,
      "step": 3764
    },
    {
      "epoch": 0.3766,
      "grad_norm": 0.07391384243965149,
      "learning_rate": 0.002,
      "loss": 1.910660982131958,
      "step": 3766
    },
    {
      "epoch": 0.3768,
      "grad_norm": 0.08187048137187958,
      "learning_rate": 0.002,
      "loss": 1.9028013944625854,
      "step": 3768
    },
    {
      "epoch": 0.377,
      "grad_norm": 0.07986662536859512,
      "learning_rate": 0.002,
      "loss": 1.9831196069717407,
      "step": 3770
    },
    {
      "epoch": 0.3772,
      "grad_norm": 0.08017430454492569,
      "learning_rate": 0.002,
      "loss": 1.821402907371521,
      "step": 3772
    },
    {
      "epoch": 0.3774,
      "grad_norm": 0.07389948517084122,
      "learning_rate": 0.002,
      "loss": 1.7769744396209717,
      "step": 3774
    },
    {
      "epoch": 0.3776,
      "grad_norm": 0.07720602303743362,
      "learning_rate": 0.002,
      "loss": 1.797814130783081,
      "step": 3776
    },
    {
      "epoch": 0.3778,
      "grad_norm": 0.07621373236179352,
      "learning_rate": 0.002,
      "loss": 2.0079987049102783,
      "step": 3778
    },
    {
      "epoch": 0.378,
      "grad_norm": 0.07422790676355362,
      "learning_rate": 0.002,
      "loss": 1.9049131870269775,
      "step": 3780
    },
    {
      "epoch": 0.3782,
      "grad_norm": 0.07719302177429199,
      "learning_rate": 0.002,
      "loss": 1.8266865015029907,
      "step": 3782
    },
    {
      "epoch": 0.3784,
      "grad_norm": 0.07475008815526962,
      "learning_rate": 0.002,
      "loss": 1.9923032522201538,
      "step": 3784
    },
    {
      "epoch": 0.3786,
      "grad_norm": 0.07947646081447601,
      "learning_rate": 0.002,
      "loss": 1.8965623378753662,
      "step": 3786
    },
    {
      "epoch": 0.3788,
      "grad_norm": 0.07607763260602951,
      "learning_rate": 0.002,
      "loss": 1.9254378080368042,
      "step": 3788
    },
    {
      "epoch": 0.379,
      "grad_norm": 0.08092699199914932,
      "learning_rate": 0.002,
      "loss": 1.938356876373291,
      "step": 3790
    },
    {
      "epoch": 0.3792,
      "grad_norm": 0.0804581567645073,
      "learning_rate": 0.002,
      "loss": 1.7963590621948242,
      "step": 3792
    },
    {
      "epoch": 0.3794,
      "grad_norm": 0.07363013178110123,
      "learning_rate": 0.002,
      "loss": 1.8473052978515625,
      "step": 3794
    },
    {
      "epoch": 0.3796,
      "grad_norm": 0.07614168524742126,
      "learning_rate": 0.002,
      "loss": 1.8434139490127563,
      "step": 3796
    },
    {
      "epoch": 0.3798,
      "grad_norm": 0.07975179702043533,
      "learning_rate": 0.002,
      "loss": 2.0017895698547363,
      "step": 3798
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.07706401497125626,
      "learning_rate": 0.002,
      "loss": 1.8562955856323242,
      "step": 3800
    },
    {
      "epoch": 0.3802,
      "grad_norm": 0.07503646612167358,
      "learning_rate": 0.002,
      "loss": 1.7249102592468262,
      "step": 3802
    },
    {
      "epoch": 0.3804,
      "grad_norm": 0.07530051469802856,
      "learning_rate": 0.002,
      "loss": 1.8521058559417725,
      "step": 3804
    },
    {
      "epoch": 0.3806,
      "grad_norm": 0.07733409106731415,
      "learning_rate": 0.002,
      "loss": 1.874802589416504,
      "step": 3806
    },
    {
      "epoch": 0.3808,
      "grad_norm": 0.08134853839874268,
      "learning_rate": 0.002,
      "loss": 1.824118971824646,
      "step": 3808
    },
    {
      "epoch": 0.381,
      "grad_norm": 0.07637467235326767,
      "learning_rate": 0.002,
      "loss": 1.8636279106140137,
      "step": 3810
    },
    {
      "epoch": 0.3812,
      "grad_norm": 0.07861248403787613,
      "learning_rate": 0.002,
      "loss": 1.9252758026123047,
      "step": 3812
    },
    {
      "epoch": 0.3814,
      "grad_norm": 0.07576914876699448,
      "learning_rate": 0.002,
      "loss": 1.7496867179870605,
      "step": 3814
    },
    {
      "epoch": 0.3816,
      "grad_norm": 0.07910390198230743,
      "learning_rate": 0.002,
      "loss": 1.9636446237564087,
      "step": 3816
    },
    {
      "epoch": 0.3818,
      "grad_norm": 0.07147835940122604,
      "learning_rate": 0.002,
      "loss": 2.0125303268432617,
      "step": 3818
    },
    {
      "epoch": 0.382,
      "grad_norm": 0.07528392970561981,
      "learning_rate": 0.002,
      "loss": 1.918552041053772,
      "step": 3820
    },
    {
      "epoch": 0.3822,
      "grad_norm": 0.07304639369249344,
      "learning_rate": 0.002,
      "loss": 1.9651622772216797,
      "step": 3822
    },
    {
      "epoch": 0.3824,
      "grad_norm": 0.07535813748836517,
      "learning_rate": 0.002,
      "loss": 1.9451298713684082,
      "step": 3824
    },
    {
      "epoch": 0.3826,
      "grad_norm": 0.07491259276866913,
      "learning_rate": 0.002,
      "loss": 1.813269853591919,
      "step": 3826
    },
    {
      "epoch": 0.3828,
      "grad_norm": 0.07398955523967743,
      "learning_rate": 0.002,
      "loss": 1.8732731342315674,
      "step": 3828
    },
    {
      "epoch": 0.383,
      "grad_norm": 0.0713084489107132,
      "learning_rate": 0.002,
      "loss": 1.734319806098938,
      "step": 3830
    },
    {
      "epoch": 0.3832,
      "grad_norm": 0.0926905944943428,
      "learning_rate": 0.002,
      "loss": 1.9061832427978516,
      "step": 3832
    },
    {
      "epoch": 0.3834,
      "grad_norm": 0.08392517268657684,
      "learning_rate": 0.002,
      "loss": 1.8231638669967651,
      "step": 3834
    },
    {
      "epoch": 0.3836,
      "grad_norm": 0.09092401713132858,
      "learning_rate": 0.002,
      "loss": 1.8220235109329224,
      "step": 3836
    },
    {
      "epoch": 0.3838,
      "grad_norm": 0.08237649500370026,
      "learning_rate": 0.002,
      "loss": 1.8942757844924927,
      "step": 3838
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.07766924798488617,
      "learning_rate": 0.002,
      "loss": 1.947446584701538,
      "step": 3840
    },
    {
      "epoch": 0.3842,
      "grad_norm": 0.08212791383266449,
      "learning_rate": 0.002,
      "loss": 1.8299658298492432,
      "step": 3842
    },
    {
      "epoch": 0.3844,
      "grad_norm": 0.08325431495904922,
      "learning_rate": 0.002,
      "loss": 1.8817064762115479,
      "step": 3844
    },
    {
      "epoch": 0.3846,
      "grad_norm": 0.07531513273715973,
      "learning_rate": 0.002,
      "loss": 1.91996169090271,
      "step": 3846
    },
    {
      "epoch": 0.3848,
      "grad_norm": 0.07307408004999161,
      "learning_rate": 0.002,
      "loss": 1.9642865657806396,
      "step": 3848
    },
    {
      "epoch": 0.385,
      "grad_norm": 0.08027838915586472,
      "learning_rate": 0.002,
      "loss": 1.9383424520492554,
      "step": 3850
    },
    {
      "epoch": 0.3852,
      "grad_norm": 0.07424161583185196,
      "learning_rate": 0.002,
      "loss": 1.9383633136749268,
      "step": 3852
    },
    {
      "epoch": 0.3854,
      "grad_norm": 0.07552167773246765,
      "learning_rate": 0.002,
      "loss": 1.962444543838501,
      "step": 3854
    },
    {
      "epoch": 0.3856,
      "grad_norm": 0.07907506078481674,
      "learning_rate": 0.002,
      "loss": 1.8308818340301514,
      "step": 3856
    },
    {
      "epoch": 0.3858,
      "grad_norm": 0.07614236325025558,
      "learning_rate": 0.002,
      "loss": 1.8993725776672363,
      "step": 3858
    },
    {
      "epoch": 0.386,
      "grad_norm": 0.07864109426736832,
      "learning_rate": 0.002,
      "loss": 2.0852789878845215,
      "step": 3860
    },
    {
      "epoch": 0.3862,
      "grad_norm": 0.07835591584444046,
      "learning_rate": 0.002,
      "loss": 2.0443670749664307,
      "step": 3862
    },
    {
      "epoch": 0.3864,
      "grad_norm": 0.08079689741134644,
      "learning_rate": 0.002,
      "loss": 1.869042158126831,
      "step": 3864
    },
    {
      "epoch": 0.3866,
      "grad_norm": 0.07854266464710236,
      "learning_rate": 0.002,
      "loss": 2.0158684253692627,
      "step": 3866
    },
    {
      "epoch": 0.3868,
      "grad_norm": 0.08078118413686752,
      "learning_rate": 0.002,
      "loss": 1.9094102382659912,
      "step": 3868
    },
    {
      "epoch": 0.387,
      "grad_norm": 0.0757889375090599,
      "learning_rate": 0.002,
      "loss": 1.8817167282104492,
      "step": 3870
    },
    {
      "epoch": 0.3872,
      "grad_norm": 0.07666004449129105,
      "learning_rate": 0.002,
      "loss": 1.851369023323059,
      "step": 3872
    },
    {
      "epoch": 0.3874,
      "grad_norm": 0.07776094228029251,
      "learning_rate": 0.002,
      "loss": 2.0065674781799316,
      "step": 3874
    },
    {
      "epoch": 0.3876,
      "grad_norm": 0.0758264884352684,
      "learning_rate": 0.002,
      "loss": 1.8304331302642822,
      "step": 3876
    },
    {
      "epoch": 0.3878,
      "grad_norm": 0.07681552320718765,
      "learning_rate": 0.002,
      "loss": 1.7814550399780273,
      "step": 3878
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.07376690208911896,
      "learning_rate": 0.002,
      "loss": 1.9095120429992676,
      "step": 3880
    },
    {
      "epoch": 0.3882,
      "grad_norm": 0.07599755376577377,
      "learning_rate": 0.002,
      "loss": 1.8884520530700684,
      "step": 3882
    },
    {
      "epoch": 0.3884,
      "grad_norm": 0.07582483440637589,
      "learning_rate": 0.002,
      "loss": 1.9375197887420654,
      "step": 3884
    },
    {
      "epoch": 0.3886,
      "grad_norm": 0.07509557902812958,
      "learning_rate": 0.002,
      "loss": 1.9558533430099487,
      "step": 3886
    },
    {
      "epoch": 0.3888,
      "grad_norm": 0.07705897837877274,
      "learning_rate": 0.002,
      "loss": 1.8355813026428223,
      "step": 3888
    },
    {
      "epoch": 0.389,
      "grad_norm": 0.07831571996212006,
      "learning_rate": 0.002,
      "loss": 2.025024652481079,
      "step": 3890
    },
    {
      "epoch": 0.3892,
      "grad_norm": 0.07545682042837143,
      "learning_rate": 0.002,
      "loss": 1.9042625427246094,
      "step": 3892
    },
    {
      "epoch": 0.3894,
      "grad_norm": 0.08503580093383789,
      "learning_rate": 0.002,
      "loss": 1.9528634548187256,
      "step": 3894
    },
    {
      "epoch": 0.3896,
      "grad_norm": 0.08000177145004272,
      "learning_rate": 0.002,
      "loss": 1.9591327905654907,
      "step": 3896
    },
    {
      "epoch": 0.3898,
      "grad_norm": 0.08108843863010406,
      "learning_rate": 0.002,
      "loss": 1.891493320465088,
      "step": 3898
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.08007476478815079,
      "learning_rate": 0.002,
      "loss": 1.9338650703430176,
      "step": 3900
    },
    {
      "epoch": 0.3902,
      "grad_norm": 0.08067160844802856,
      "learning_rate": 0.002,
      "loss": 1.9518085718154907,
      "step": 3902
    },
    {
      "epoch": 0.3904,
      "grad_norm": 0.07834915071725845,
      "learning_rate": 0.002,
      "loss": 1.8885083198547363,
      "step": 3904
    },
    {
      "epoch": 0.3906,
      "grad_norm": 0.07749299705028534,
      "learning_rate": 0.002,
      "loss": 1.8747612237930298,
      "step": 3906
    },
    {
      "epoch": 0.3908,
      "grad_norm": 0.07230233401060104,
      "learning_rate": 0.002,
      "loss": 1.7834268808364868,
      "step": 3908
    },
    {
      "epoch": 0.391,
      "grad_norm": 0.07688970118761063,
      "learning_rate": 0.002,
      "loss": 1.8375571966171265,
      "step": 3910
    },
    {
      "epoch": 0.3912,
      "grad_norm": 0.07850463688373566,
      "learning_rate": 0.002,
      "loss": 2.043393611907959,
      "step": 3912
    },
    {
      "epoch": 0.3914,
      "grad_norm": 0.07630472630262375,
      "learning_rate": 0.002,
      "loss": 2.000812292098999,
      "step": 3914
    },
    {
      "epoch": 0.3916,
      "grad_norm": 0.07181942462921143,
      "learning_rate": 0.002,
      "loss": 1.9570146799087524,
      "step": 3916
    },
    {
      "epoch": 0.3918,
      "grad_norm": 0.08207793533802032,
      "learning_rate": 0.002,
      "loss": 1.8826371431350708,
      "step": 3918
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.07761338353157043,
      "learning_rate": 0.002,
      "loss": 1.7892943620681763,
      "step": 3920
    },
    {
      "epoch": 0.3922,
      "grad_norm": 0.07558319717645645,
      "learning_rate": 0.002,
      "loss": 1.7663850784301758,
      "step": 3922
    },
    {
      "epoch": 0.3924,
      "grad_norm": 0.07731781154870987,
      "learning_rate": 0.002,
      "loss": 1.8642579317092896,
      "step": 3924
    },
    {
      "epoch": 0.3926,
      "grad_norm": 0.0827215164899826,
      "learning_rate": 0.002,
      "loss": 1.861351490020752,
      "step": 3926
    },
    {
      "epoch": 0.3928,
      "grad_norm": 0.07730540633201599,
      "learning_rate": 0.002,
      "loss": 2.1014087200164795,
      "step": 3928
    },
    {
      "epoch": 0.393,
      "grad_norm": 0.07731523364782333,
      "learning_rate": 0.002,
      "loss": 2.1434504985809326,
      "step": 3930
    },
    {
      "epoch": 0.3932,
      "grad_norm": 0.07598098367452621,
      "learning_rate": 0.002,
      "loss": 1.8110675811767578,
      "step": 3932
    },
    {
      "epoch": 0.3934,
      "grad_norm": 0.07571492344141006,
      "learning_rate": 0.002,
      "loss": 1.866537094116211,
      "step": 3934
    },
    {
      "epoch": 0.3936,
      "grad_norm": 0.07323433458805084,
      "learning_rate": 0.002,
      "loss": 1.9532673358917236,
      "step": 3936
    },
    {
      "epoch": 0.3938,
      "grad_norm": 0.07565046846866608,
      "learning_rate": 0.002,
      "loss": 1.920181393623352,
      "step": 3938
    },
    {
      "epoch": 0.394,
      "grad_norm": 0.08196945488452911,
      "learning_rate": 0.002,
      "loss": 1.8956708908081055,
      "step": 3940
    },
    {
      "epoch": 0.3942,
      "grad_norm": 0.07952551543712616,
      "learning_rate": 0.002,
      "loss": 1.9793634414672852,
      "step": 3942
    },
    {
      "epoch": 0.3944,
      "grad_norm": 0.07661064714193344,
      "learning_rate": 0.002,
      "loss": 1.899196743965149,
      "step": 3944
    },
    {
      "epoch": 0.3946,
      "grad_norm": 0.08222769945859909,
      "learning_rate": 0.002,
      "loss": 1.9897619485855103,
      "step": 3946
    },
    {
      "epoch": 0.3948,
      "grad_norm": 0.07776714116334915,
      "learning_rate": 0.002,
      "loss": 1.7915970087051392,
      "step": 3948
    },
    {
      "epoch": 0.395,
      "grad_norm": 0.07631386816501617,
      "learning_rate": 0.002,
      "loss": 1.9782952070236206,
      "step": 3950
    },
    {
      "epoch": 0.3952,
      "grad_norm": 0.07778548449277878,
      "learning_rate": 0.002,
      "loss": 1.8509025573730469,
      "step": 3952
    },
    {
      "epoch": 0.3954,
      "grad_norm": 0.07842205464839935,
      "learning_rate": 0.002,
      "loss": 1.9012858867645264,
      "step": 3954
    },
    {
      "epoch": 0.3956,
      "grad_norm": 0.07904726266860962,
      "learning_rate": 0.002,
      "loss": 1.8550790548324585,
      "step": 3956
    },
    {
      "epoch": 0.3958,
      "grad_norm": 0.07592861354351044,
      "learning_rate": 0.002,
      "loss": 1.8888773918151855,
      "step": 3958
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.08065808564424515,
      "learning_rate": 0.002,
      "loss": 1.8746908903121948,
      "step": 3960
    },
    {
      "epoch": 0.3962,
      "grad_norm": 0.0764981135725975,
      "learning_rate": 0.002,
      "loss": 1.9003902673721313,
      "step": 3962
    },
    {
      "epoch": 0.3964,
      "grad_norm": 0.07312619686126709,
      "learning_rate": 0.002,
      "loss": 1.8711589574813843,
      "step": 3964
    },
    {
      "epoch": 0.3966,
      "grad_norm": 0.07409442216157913,
      "learning_rate": 0.002,
      "loss": 1.957722544670105,
      "step": 3966
    },
    {
      "epoch": 0.3968,
      "grad_norm": 0.07939273864030838,
      "learning_rate": 0.002,
      "loss": 1.9728569984436035,
      "step": 3968
    },
    {
      "epoch": 0.397,
      "grad_norm": 0.07407698035240173,
      "learning_rate": 0.002,
      "loss": 1.8882869482040405,
      "step": 3970
    },
    {
      "epoch": 0.3972,
      "grad_norm": 0.07514256983995438,
      "learning_rate": 0.002,
      "loss": 1.9089418649673462,
      "step": 3972
    },
    {
      "epoch": 0.3974,
      "grad_norm": 0.07867216318845749,
      "learning_rate": 0.002,
      "loss": 1.8900182247161865,
      "step": 3974
    },
    {
      "epoch": 0.3976,
      "grad_norm": 0.07348164170980453,
      "learning_rate": 0.002,
      "loss": 1.9357459545135498,
      "step": 3976
    },
    {
      "epoch": 0.3978,
      "grad_norm": 0.07649075984954834,
      "learning_rate": 0.002,
      "loss": 1.836266040802002,
      "step": 3978
    },
    {
      "epoch": 0.398,
      "grad_norm": 0.07343616336584091,
      "learning_rate": 0.002,
      "loss": 1.8552354574203491,
      "step": 3980
    },
    {
      "epoch": 0.3982,
      "grad_norm": 0.07769015431404114,
      "learning_rate": 0.002,
      "loss": 1.963325023651123,
      "step": 3982
    },
    {
      "epoch": 0.3984,
      "grad_norm": 0.07663120329380035,
      "learning_rate": 0.002,
      "loss": 1.8721203804016113,
      "step": 3984
    },
    {
      "epoch": 0.3986,
      "grad_norm": 0.07610958069562912,
      "learning_rate": 0.002,
      "loss": 1.944555640220642,
      "step": 3986
    },
    {
      "epoch": 0.3988,
      "grad_norm": 0.0752977654337883,
      "learning_rate": 0.002,
      "loss": 1.9089962244033813,
      "step": 3988
    },
    {
      "epoch": 0.399,
      "grad_norm": 0.07334261387586594,
      "learning_rate": 0.002,
      "loss": 1.850709080696106,
      "step": 3990
    },
    {
      "epoch": 0.3992,
      "grad_norm": 0.07294167578220367,
      "learning_rate": 0.002,
      "loss": 1.8317985534667969,
      "step": 3992
    },
    {
      "epoch": 0.3994,
      "grad_norm": 0.078300341963768,
      "learning_rate": 0.002,
      "loss": 1.9397162199020386,
      "step": 3994
    },
    {
      "epoch": 0.3996,
      "grad_norm": 0.07677441835403442,
      "learning_rate": 0.002,
      "loss": 1.8911528587341309,
      "step": 3996
    },
    {
      "epoch": 0.3998,
      "grad_norm": 0.08062278479337692,
      "learning_rate": 0.002,
      "loss": 1.9361690282821655,
      "step": 3998
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.07373557984828949,
      "learning_rate": 0.002,
      "loss": 2.006596803665161,
      "step": 4000
    },
    {
      "epoch": 0.4,
      "eval_loss": 2.146089553833008,
      "eval_runtime": 223.3573,
      "eval_samples_per_second": 2.239,
      "eval_steps_per_second": 0.224,
      "step": 4000
    },
    {
      "epoch": 0.4002,
      "grad_norm": 0.08078253269195557,
      "learning_rate": 0.002,
      "loss": 1.9482641220092773,
      "step": 4002
    },
    {
      "epoch": 0.4004,
      "grad_norm": 0.07420505583286285,
      "learning_rate": 0.002,
      "loss": 1.906592607498169,
      "step": 4004
    },
    {
      "epoch": 0.4006,
      "grad_norm": 0.07562243193387985,
      "learning_rate": 0.002,
      "loss": 1.9249175786972046,
      "step": 4006
    },
    {
      "epoch": 0.4008,
      "grad_norm": 0.08016744256019592,
      "learning_rate": 0.002,
      "loss": 1.8716185092926025,
      "step": 4008
    },
    {
      "epoch": 0.401,
      "grad_norm": 0.07457074522972107,
      "learning_rate": 0.002,
      "loss": 1.9031033515930176,
      "step": 4010
    },
    {
      "epoch": 0.4012,
      "grad_norm": 0.07312285155057907,
      "learning_rate": 0.002,
      "loss": 1.7489268779754639,
      "step": 4012
    },
    {
      "epoch": 0.4014,
      "grad_norm": 0.07584507763385773,
      "learning_rate": 0.002,
      "loss": 1.9106251001358032,
      "step": 4014
    },
    {
      "epoch": 0.4016,
      "grad_norm": 0.07464927434921265,
      "learning_rate": 0.002,
      "loss": 1.9689737558364868,
      "step": 4016
    },
    {
      "epoch": 0.4018,
      "grad_norm": 0.07750211656093597,
      "learning_rate": 0.002,
      "loss": 1.9641990661621094,
      "step": 4018
    },
    {
      "epoch": 0.402,
      "grad_norm": 0.0753922387957573,
      "learning_rate": 0.002,
      "loss": 1.945327877998352,
      "step": 4020
    },
    {
      "epoch": 0.4022,
      "grad_norm": 0.07748948782682419,
      "learning_rate": 0.002,
      "loss": 2.0075173377990723,
      "step": 4022
    },
    {
      "epoch": 0.4024,
      "grad_norm": 0.0744083970785141,
      "learning_rate": 0.002,
      "loss": 1.7759054899215698,
      "step": 4024
    },
    {
      "epoch": 0.4026,
      "grad_norm": 0.07544244080781937,
      "learning_rate": 0.002,
      "loss": 1.820534348487854,
      "step": 4026
    },
    {
      "epoch": 0.4028,
      "grad_norm": 0.07498213648796082,
      "learning_rate": 0.002,
      "loss": 1.9721184968948364,
      "step": 4028
    },
    {
      "epoch": 0.403,
      "grad_norm": 0.07315094023942947,
      "learning_rate": 0.002,
      "loss": 1.8144896030426025,
      "step": 4030
    },
    {
      "epoch": 0.4032,
      "grad_norm": 0.07697001844644547,
      "learning_rate": 0.002,
      "loss": 1.8866560459136963,
      "step": 4032
    },
    {
      "epoch": 0.4034,
      "grad_norm": 0.07449573278427124,
      "learning_rate": 0.002,
      "loss": 1.8210413455963135,
      "step": 4034
    },
    {
      "epoch": 0.4036,
      "grad_norm": 0.07471410185098648,
      "learning_rate": 0.002,
      "loss": 1.8776957988739014,
      "step": 4036
    },
    {
      "epoch": 0.4038,
      "grad_norm": 0.07672756165266037,
      "learning_rate": 0.002,
      "loss": 1.8717650175094604,
      "step": 4038
    },
    {
      "epoch": 0.404,
      "grad_norm": 0.0722528025507927,
      "learning_rate": 0.002,
      "loss": 1.9288685321807861,
      "step": 4040
    },
    {
      "epoch": 0.4042,
      "grad_norm": 0.08617658168077469,
      "learning_rate": 0.002,
      "loss": 1.992219090461731,
      "step": 4042
    },
    {
      "epoch": 0.4044,
      "grad_norm": 0.07969435304403305,
      "learning_rate": 0.002,
      "loss": 1.9593027830123901,
      "step": 4044
    },
    {
      "epoch": 0.4046,
      "grad_norm": 0.07855483144521713,
      "learning_rate": 0.002,
      "loss": 1.9692578315734863,
      "step": 4046
    },
    {
      "epoch": 0.4048,
      "grad_norm": 0.07563787698745728,
      "learning_rate": 0.002,
      "loss": 1.84761381149292,
      "step": 4048
    },
    {
      "epoch": 0.405,
      "grad_norm": 0.07120487093925476,
      "learning_rate": 0.002,
      "loss": 1.9043619632720947,
      "step": 4050
    },
    {
      "epoch": 0.4052,
      "grad_norm": 0.0840064138174057,
      "learning_rate": 0.002,
      "loss": 2.025479316711426,
      "step": 4052
    },
    {
      "epoch": 0.4054,
      "grad_norm": 0.07469344884157181,
      "learning_rate": 0.002,
      "loss": 1.9413737058639526,
      "step": 4054
    },
    {
      "epoch": 0.4056,
      "grad_norm": 0.0743858590722084,
      "learning_rate": 0.002,
      "loss": 1.8964217901229858,
      "step": 4056
    },
    {
      "epoch": 0.4058,
      "grad_norm": 0.0764516294002533,
      "learning_rate": 0.002,
      "loss": 1.8687353134155273,
      "step": 4058
    },
    {
      "epoch": 0.406,
      "grad_norm": 0.07638052850961685,
      "learning_rate": 0.002,
      "loss": 1.8728022575378418,
      "step": 4060
    },
    {
      "epoch": 0.4062,
      "grad_norm": 0.07337995618581772,
      "learning_rate": 0.002,
      "loss": 1.84538733959198,
      "step": 4062
    },
    {
      "epoch": 0.4064,
      "grad_norm": 0.07831481099128723,
      "learning_rate": 0.002,
      "loss": 1.8058686256408691,
      "step": 4064
    },
    {
      "epoch": 0.4066,
      "grad_norm": 0.07668601721525192,
      "learning_rate": 0.002,
      "loss": 1.9441038370132446,
      "step": 4066
    },
    {
      "epoch": 0.4068,
      "grad_norm": 0.07644247263669968,
      "learning_rate": 0.002,
      "loss": 1.8525892496109009,
      "step": 4068
    },
    {
      "epoch": 0.407,
      "grad_norm": 0.0754496157169342,
      "learning_rate": 0.002,
      "loss": 1.8819143772125244,
      "step": 4070
    },
    {
      "epoch": 0.4072,
      "grad_norm": 0.07455544918775558,
      "learning_rate": 0.002,
      "loss": 1.889320969581604,
      "step": 4072
    },
    {
      "epoch": 0.4074,
      "grad_norm": 0.0769689530134201,
      "learning_rate": 0.002,
      "loss": 1.8966617584228516,
      "step": 4074
    },
    {
      "epoch": 0.4076,
      "grad_norm": 0.07910874485969543,
      "learning_rate": 0.002,
      "loss": 1.9548118114471436,
      "step": 4076
    },
    {
      "epoch": 0.4078,
      "grad_norm": 0.07125923782587051,
      "learning_rate": 0.002,
      "loss": 1.779754400253296,
      "step": 4078
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.07667715102434158,
      "learning_rate": 0.002,
      "loss": 1.8582401275634766,
      "step": 4080
    },
    {
      "epoch": 0.4082,
      "grad_norm": 0.07288666069507599,
      "learning_rate": 0.002,
      "loss": 1.851564884185791,
      "step": 4082
    },
    {
      "epoch": 0.4084,
      "grad_norm": 0.07896990329027176,
      "learning_rate": 0.002,
      "loss": 1.9532413482666016,
      "step": 4084
    },
    {
      "epoch": 0.4086,
      "grad_norm": 0.07743237912654877,
      "learning_rate": 0.002,
      "loss": 1.9173991680145264,
      "step": 4086
    },
    {
      "epoch": 0.4088,
      "grad_norm": 0.07496917992830276,
      "learning_rate": 0.002,
      "loss": 1.9553017616271973,
      "step": 4088
    },
    {
      "epoch": 0.409,
      "grad_norm": 0.0751170739531517,
      "learning_rate": 0.002,
      "loss": 1.834702491760254,
      "step": 4090
    },
    {
      "epoch": 0.4092,
      "grad_norm": 0.0754900723695755,
      "learning_rate": 0.002,
      "loss": 1.85604989528656,
      "step": 4092
    },
    {
      "epoch": 0.4094,
      "grad_norm": 0.08124259859323502,
      "learning_rate": 0.002,
      "loss": 1.8804482221603394,
      "step": 4094
    },
    {
      "epoch": 0.4096,
      "grad_norm": 0.07965625822544098,
      "learning_rate": 0.002,
      "loss": 1.9332491159439087,
      "step": 4096
    },
    {
      "epoch": 0.4098,
      "grad_norm": 0.07725631445646286,
      "learning_rate": 0.002,
      "loss": 1.8404574394226074,
      "step": 4098
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.0763261690735817,
      "learning_rate": 0.002,
      "loss": 1.9444373846054077,
      "step": 4100
    },
    {
      "epoch": 0.4102,
      "grad_norm": 0.07817362248897552,
      "learning_rate": 0.002,
      "loss": 1.9313228130340576,
      "step": 4102
    },
    {
      "epoch": 0.4104,
      "grad_norm": 0.0791681706905365,
      "learning_rate": 0.002,
      "loss": 1.8920150995254517,
      "step": 4104
    },
    {
      "epoch": 0.4106,
      "grad_norm": 0.07556883245706558,
      "learning_rate": 0.002,
      "loss": 2.014460325241089,
      "step": 4106
    },
    {
      "epoch": 0.4108,
      "grad_norm": 0.07777917385101318,
      "learning_rate": 0.002,
      "loss": 1.9098635911941528,
      "step": 4108
    },
    {
      "epoch": 0.411,
      "grad_norm": 0.0728716030716896,
      "learning_rate": 0.002,
      "loss": 1.958244800567627,
      "step": 4110
    },
    {
      "epoch": 0.4112,
      "grad_norm": 0.07841340452432632,
      "learning_rate": 0.002,
      "loss": 1.8737155199050903,
      "step": 4112
    },
    {
      "epoch": 0.4114,
      "grad_norm": 0.07660096138715744,
      "learning_rate": 0.002,
      "loss": 1.972198247909546,
      "step": 4114
    },
    {
      "epoch": 0.4116,
      "grad_norm": 0.07508470118045807,
      "learning_rate": 0.002,
      "loss": 1.9167850017547607,
      "step": 4116
    },
    {
      "epoch": 0.4118,
      "grad_norm": 0.07344569265842438,
      "learning_rate": 0.002,
      "loss": 1.9125460386276245,
      "step": 4118
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.08051607012748718,
      "learning_rate": 0.002,
      "loss": 1.9042103290557861,
      "step": 4120
    },
    {
      "epoch": 0.4122,
      "grad_norm": 0.07543040812015533,
      "learning_rate": 0.002,
      "loss": 1.9174129962921143,
      "step": 4122
    },
    {
      "epoch": 0.4124,
      "grad_norm": 0.0780162364244461,
      "learning_rate": 0.002,
      "loss": 1.7990913391113281,
      "step": 4124
    },
    {
      "epoch": 0.4126,
      "grad_norm": 0.07747749239206314,
      "learning_rate": 0.002,
      "loss": 1.8591270446777344,
      "step": 4126
    },
    {
      "epoch": 0.4128,
      "grad_norm": 0.07563339173793793,
      "learning_rate": 0.002,
      "loss": 1.957175612449646,
      "step": 4128
    },
    {
      "epoch": 0.413,
      "grad_norm": 0.07503668963909149,
      "learning_rate": 0.002,
      "loss": 1.8080925941467285,
      "step": 4130
    },
    {
      "epoch": 0.4132,
      "grad_norm": 0.07525282353162766,
      "learning_rate": 0.002,
      "loss": 1.7916691303253174,
      "step": 4132
    },
    {
      "epoch": 0.4134,
      "grad_norm": 0.07550451904535294,
      "learning_rate": 0.002,
      "loss": 1.906698226928711,
      "step": 4134
    },
    {
      "epoch": 0.4136,
      "grad_norm": 0.07365169376134872,
      "learning_rate": 0.002,
      "loss": 1.851055383682251,
      "step": 4136
    },
    {
      "epoch": 0.4138,
      "grad_norm": 0.07400266081094742,
      "learning_rate": 0.002,
      "loss": 1.9490554332733154,
      "step": 4138
    },
    {
      "epoch": 0.414,
      "grad_norm": 0.08334493637084961,
      "learning_rate": 0.002,
      "loss": 1.9375752210617065,
      "step": 4140
    },
    {
      "epoch": 0.4142,
      "grad_norm": 0.0734611451625824,
      "learning_rate": 0.002,
      "loss": 1.9636461734771729,
      "step": 4142
    },
    {
      "epoch": 0.4144,
      "grad_norm": 0.07630740851163864,
      "learning_rate": 0.002,
      "loss": 1.938332438468933,
      "step": 4144
    },
    {
      "epoch": 0.4146,
      "grad_norm": 0.07620089501142502,
      "learning_rate": 0.002,
      "loss": 1.842044472694397,
      "step": 4146
    },
    {
      "epoch": 0.4148,
      "grad_norm": 0.0818135142326355,
      "learning_rate": 0.002,
      "loss": 1.8954861164093018,
      "step": 4148
    },
    {
      "epoch": 0.415,
      "grad_norm": 0.08522411435842514,
      "learning_rate": 0.002,
      "loss": 1.8790068626403809,
      "step": 4150
    },
    {
      "epoch": 0.4152,
      "grad_norm": 0.07840808480978012,
      "learning_rate": 0.002,
      "loss": 1.8200887441635132,
      "step": 4152
    },
    {
      "epoch": 0.4154,
      "grad_norm": 0.07767242193222046,
      "learning_rate": 0.002,
      "loss": 1.9227131605148315,
      "step": 4154
    },
    {
      "epoch": 0.4156,
      "grad_norm": 0.07787889242172241,
      "learning_rate": 0.002,
      "loss": 1.7908965349197388,
      "step": 4156
    },
    {
      "epoch": 0.4158,
      "grad_norm": 0.07447278499603271,
      "learning_rate": 0.002,
      "loss": 1.9550516605377197,
      "step": 4158
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.07322821021080017,
      "learning_rate": 0.002,
      "loss": 1.8805122375488281,
      "step": 4160
    },
    {
      "epoch": 0.4162,
      "grad_norm": 0.07706347107887268,
      "learning_rate": 0.002,
      "loss": 1.8739769458770752,
      "step": 4162
    },
    {
      "epoch": 0.4164,
      "grad_norm": 0.08174683153629303,
      "learning_rate": 0.002,
      "loss": 1.9334070682525635,
      "step": 4164
    },
    {
      "epoch": 0.4166,
      "grad_norm": 0.07526180148124695,
      "learning_rate": 0.002,
      "loss": 1.9139032363891602,
      "step": 4166
    },
    {
      "epoch": 0.4168,
      "grad_norm": 0.07744515687227249,
      "learning_rate": 0.002,
      "loss": 1.8811485767364502,
      "step": 4168
    },
    {
      "epoch": 0.417,
      "grad_norm": 0.08158347755670547,
      "learning_rate": 0.002,
      "loss": 1.9678006172180176,
      "step": 4170
    },
    {
      "epoch": 0.4172,
      "grad_norm": 0.08034583181142807,
      "learning_rate": 0.002,
      "loss": 1.9728056192398071,
      "step": 4172
    },
    {
      "epoch": 0.4174,
      "grad_norm": 0.07714955508708954,
      "learning_rate": 0.002,
      "loss": 1.9231047630310059,
      "step": 4174
    },
    {
      "epoch": 0.4176,
      "grad_norm": 0.07719308137893677,
      "learning_rate": 0.002,
      "loss": 1.8943336009979248,
      "step": 4176
    },
    {
      "epoch": 0.4178,
      "grad_norm": 0.07827340811491013,
      "learning_rate": 0.002,
      "loss": 1.9759410619735718,
      "step": 4178
    },
    {
      "epoch": 0.418,
      "grad_norm": 0.07464239001274109,
      "learning_rate": 0.002,
      "loss": 1.906068205833435,
      "step": 4180
    },
    {
      "epoch": 0.4182,
      "grad_norm": 0.07590453326702118,
      "learning_rate": 0.002,
      "loss": 1.814756155014038,
      "step": 4182
    },
    {
      "epoch": 0.4184,
      "grad_norm": 0.0739801898598671,
      "learning_rate": 0.002,
      "loss": 1.8328384160995483,
      "step": 4184
    },
    {
      "epoch": 0.4186,
      "grad_norm": 0.07748102396726608,
      "learning_rate": 0.002,
      "loss": 1.8462421894073486,
      "step": 4186
    },
    {
      "epoch": 0.4188,
      "grad_norm": 0.07544176280498505,
      "learning_rate": 0.002,
      "loss": 1.816696286201477,
      "step": 4188
    },
    {
      "epoch": 0.419,
      "grad_norm": 0.07586260885000229,
      "learning_rate": 0.002,
      "loss": 1.900833249092102,
      "step": 4190
    },
    {
      "epoch": 0.4192,
      "grad_norm": 0.07963413745164871,
      "learning_rate": 0.002,
      "loss": 2.0167784690856934,
      "step": 4192
    },
    {
      "epoch": 0.4194,
      "grad_norm": 0.07902710139751434,
      "learning_rate": 0.002,
      "loss": 2.04274845123291,
      "step": 4194
    },
    {
      "epoch": 0.4196,
      "grad_norm": 0.07515408843755722,
      "learning_rate": 0.002,
      "loss": 2.0080924034118652,
      "step": 4196
    },
    {
      "epoch": 0.4198,
      "grad_norm": 0.07464880496263504,
      "learning_rate": 0.002,
      "loss": 1.8520005941390991,
      "step": 4198
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.08735612034797668,
      "learning_rate": 0.002,
      "loss": 1.9508332014083862,
      "step": 4200
    },
    {
      "epoch": 0.4202,
      "grad_norm": 0.07612677663564682,
      "learning_rate": 0.002,
      "loss": 1.9919184446334839,
      "step": 4202
    },
    {
      "epoch": 0.4204,
      "grad_norm": 0.0749637708067894,
      "learning_rate": 0.002,
      "loss": 1.9709899425506592,
      "step": 4204
    },
    {
      "epoch": 0.4206,
      "grad_norm": 0.0800078958272934,
      "learning_rate": 0.002,
      "loss": 1.863309621810913,
      "step": 4206
    },
    {
      "epoch": 0.4208,
      "grad_norm": 0.07968925684690475,
      "learning_rate": 0.002,
      "loss": 1.9482617378234863,
      "step": 4208
    },
    {
      "epoch": 0.421,
      "grad_norm": 0.0786309540271759,
      "learning_rate": 0.002,
      "loss": 1.8526126146316528,
      "step": 4210
    },
    {
      "epoch": 0.4212,
      "grad_norm": 0.08781051635742188,
      "learning_rate": 0.002,
      "loss": 1.8717793226242065,
      "step": 4212
    },
    {
      "epoch": 0.4214,
      "grad_norm": 0.07510731369256973,
      "learning_rate": 0.002,
      "loss": 1.9371273517608643,
      "step": 4214
    },
    {
      "epoch": 0.4216,
      "grad_norm": 0.07608207315206528,
      "learning_rate": 0.002,
      "loss": 1.8909275531768799,
      "step": 4216
    },
    {
      "epoch": 0.4218,
      "grad_norm": 0.0775182694196701,
      "learning_rate": 0.002,
      "loss": 1.8661470413208008,
      "step": 4218
    },
    {
      "epoch": 0.422,
      "grad_norm": 0.07561466842889786,
      "learning_rate": 0.002,
      "loss": 2.0030064582824707,
      "step": 4220
    },
    {
      "epoch": 0.4222,
      "grad_norm": 0.07785036414861679,
      "learning_rate": 0.002,
      "loss": 1.9054127931594849,
      "step": 4222
    },
    {
      "epoch": 0.4224,
      "grad_norm": 0.07512462139129639,
      "learning_rate": 0.002,
      "loss": 1.9159350395202637,
      "step": 4224
    },
    {
      "epoch": 0.4226,
      "grad_norm": 0.07731577754020691,
      "learning_rate": 0.002,
      "loss": 1.9434722661972046,
      "step": 4226
    },
    {
      "epoch": 0.4228,
      "grad_norm": 0.07428988069295883,
      "learning_rate": 0.002,
      "loss": 1.8471741676330566,
      "step": 4228
    },
    {
      "epoch": 0.423,
      "grad_norm": 0.07670367509126663,
      "learning_rate": 0.002,
      "loss": 1.9739584922790527,
      "step": 4230
    },
    {
      "epoch": 0.4232,
      "grad_norm": 0.07939007878303528,
      "learning_rate": 0.002,
      "loss": 1.9217320680618286,
      "step": 4232
    },
    {
      "epoch": 0.4234,
      "grad_norm": 0.07776238024234772,
      "learning_rate": 0.002,
      "loss": 1.8222310543060303,
      "step": 4234
    },
    {
      "epoch": 0.4236,
      "grad_norm": 0.07365307211875916,
      "learning_rate": 0.002,
      "loss": 1.9656591415405273,
      "step": 4236
    },
    {
      "epoch": 0.4238,
      "grad_norm": 0.0753554180264473,
      "learning_rate": 0.002,
      "loss": 1.9936803579330444,
      "step": 4238
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.07631412893533707,
      "learning_rate": 0.002,
      "loss": 2.0885379314422607,
      "step": 4240
    },
    {
      "epoch": 0.4242,
      "grad_norm": 0.08017933368682861,
      "learning_rate": 0.002,
      "loss": 1.8670634031295776,
      "step": 4242
    },
    {
      "epoch": 0.4244,
      "grad_norm": 0.0759596973657608,
      "learning_rate": 0.002,
      "loss": 1.9169341325759888,
      "step": 4244
    },
    {
      "epoch": 0.4246,
      "grad_norm": 0.07591453194618225,
      "learning_rate": 0.002,
      "loss": 1.9331457614898682,
      "step": 4246
    },
    {
      "epoch": 0.4248,
      "grad_norm": 0.07942825555801392,
      "learning_rate": 0.002,
      "loss": 1.8803160190582275,
      "step": 4248
    },
    {
      "epoch": 0.425,
      "grad_norm": 0.0769944116473198,
      "learning_rate": 0.002,
      "loss": 2.040729522705078,
      "step": 4250
    },
    {
      "epoch": 0.4252,
      "grad_norm": 0.07445128262042999,
      "learning_rate": 0.002,
      "loss": 1.8724312782287598,
      "step": 4252
    },
    {
      "epoch": 0.4254,
      "grad_norm": 0.07758878916501999,
      "learning_rate": 0.002,
      "loss": 1.9528216123580933,
      "step": 4254
    },
    {
      "epoch": 0.4256,
      "grad_norm": 0.0760437473654747,
      "learning_rate": 0.002,
      "loss": 1.8931469917297363,
      "step": 4256
    },
    {
      "epoch": 0.4258,
      "grad_norm": 0.07449397444725037,
      "learning_rate": 0.002,
      "loss": 1.8895163536071777,
      "step": 4258
    },
    {
      "epoch": 0.426,
      "grad_norm": 0.07357186824083328,
      "learning_rate": 0.002,
      "loss": 1.8861020803451538,
      "step": 4260
    },
    {
      "epoch": 0.4262,
      "grad_norm": 0.08066032081842422,
      "learning_rate": 0.002,
      "loss": 1.7865500450134277,
      "step": 4262
    },
    {
      "epoch": 0.4264,
      "grad_norm": 0.0791747123003006,
      "learning_rate": 0.002,
      "loss": 1.9068416357040405,
      "step": 4264
    },
    {
      "epoch": 0.4266,
      "grad_norm": 0.07638660818338394,
      "learning_rate": 0.002,
      "loss": 1.8659659624099731,
      "step": 4266
    },
    {
      "epoch": 0.4268,
      "grad_norm": 0.07593495398759842,
      "learning_rate": 0.002,
      "loss": 1.8468682765960693,
      "step": 4268
    },
    {
      "epoch": 0.427,
      "grad_norm": 0.07607666403055191,
      "learning_rate": 0.002,
      "loss": 1.9797258377075195,
      "step": 4270
    },
    {
      "epoch": 0.4272,
      "grad_norm": 0.084589384496212,
      "learning_rate": 0.002,
      "loss": 1.966013789176941,
      "step": 4272
    },
    {
      "epoch": 0.4274,
      "grad_norm": 0.0794810950756073,
      "learning_rate": 0.002,
      "loss": 1.854353904724121,
      "step": 4274
    },
    {
      "epoch": 0.4276,
      "grad_norm": 0.0726780816912651,
      "learning_rate": 0.002,
      "loss": 1.9455147981643677,
      "step": 4276
    },
    {
      "epoch": 0.4278,
      "grad_norm": 0.0875878781080246,
      "learning_rate": 0.002,
      "loss": 1.953208088874817,
      "step": 4278
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.07376430183649063,
      "learning_rate": 0.002,
      "loss": 1.8785291910171509,
      "step": 4280
    },
    {
      "epoch": 0.4282,
      "grad_norm": 0.0756203755736351,
      "learning_rate": 0.002,
      "loss": 1.9342656135559082,
      "step": 4282
    },
    {
      "epoch": 0.4284,
      "grad_norm": 0.07411914318799973,
      "learning_rate": 0.002,
      "loss": 1.931947112083435,
      "step": 4284
    },
    {
      "epoch": 0.4286,
      "grad_norm": 0.07569687813520432,
      "learning_rate": 0.002,
      "loss": 1.8791124820709229,
      "step": 4286
    },
    {
      "epoch": 0.4288,
      "grad_norm": 0.08020618557929993,
      "learning_rate": 0.002,
      "loss": 1.9966621398925781,
      "step": 4288
    },
    {
      "epoch": 0.429,
      "grad_norm": 0.07593727111816406,
      "learning_rate": 0.002,
      "loss": 1.9194267988204956,
      "step": 4290
    },
    {
      "epoch": 0.4292,
      "grad_norm": 0.07846052944660187,
      "learning_rate": 0.002,
      "loss": 1.8824501037597656,
      "step": 4292
    },
    {
      "epoch": 0.4294,
      "grad_norm": 0.07860667258501053,
      "learning_rate": 0.002,
      "loss": 1.9319835901260376,
      "step": 4294
    },
    {
      "epoch": 0.4296,
      "grad_norm": 0.07908262312412262,
      "learning_rate": 0.002,
      "loss": 1.769196629524231,
      "step": 4296
    },
    {
      "epoch": 0.4298,
      "grad_norm": 0.0730125829577446,
      "learning_rate": 0.002,
      "loss": 1.8330373764038086,
      "step": 4298
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.08126205205917358,
      "learning_rate": 0.002,
      "loss": 1.9101344347000122,
      "step": 4300
    },
    {
      "epoch": 0.4302,
      "grad_norm": 0.07408862560987473,
      "learning_rate": 0.002,
      "loss": 1.9107120037078857,
      "step": 4302
    },
    {
      "epoch": 0.4304,
      "grad_norm": 0.07773074507713318,
      "learning_rate": 0.002,
      "loss": 2.0041873455047607,
      "step": 4304
    },
    {
      "epoch": 0.4306,
      "grad_norm": 0.07806869596242905,
      "learning_rate": 0.002,
      "loss": 1.9989985227584839,
      "step": 4306
    },
    {
      "epoch": 0.4308,
      "grad_norm": 0.0720582976937294,
      "learning_rate": 0.002,
      "loss": 1.879107117652893,
      "step": 4308
    },
    {
      "epoch": 0.431,
      "grad_norm": 0.07913489639759064,
      "learning_rate": 0.002,
      "loss": 1.8451093435287476,
      "step": 4310
    },
    {
      "epoch": 0.4312,
      "grad_norm": 0.0748002901673317,
      "learning_rate": 0.002,
      "loss": 1.808273196220398,
      "step": 4312
    },
    {
      "epoch": 0.4314,
      "grad_norm": 0.07794218510389328,
      "learning_rate": 0.002,
      "loss": 1.950249433517456,
      "step": 4314
    },
    {
      "epoch": 0.4316,
      "grad_norm": 0.07556872069835663,
      "learning_rate": 0.002,
      "loss": 1.8850520849227905,
      "step": 4316
    },
    {
      "epoch": 0.4318,
      "grad_norm": 0.07674817740917206,
      "learning_rate": 0.002,
      "loss": 1.944439172744751,
      "step": 4318
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.07530413568019867,
      "learning_rate": 0.002,
      "loss": 1.8383493423461914,
      "step": 4320
    },
    {
      "epoch": 0.4322,
      "grad_norm": 0.07598013430833817,
      "learning_rate": 0.002,
      "loss": 1.9621741771697998,
      "step": 4322
    },
    {
      "epoch": 0.4324,
      "grad_norm": 0.07271574437618256,
      "learning_rate": 0.002,
      "loss": 1.9589512348175049,
      "step": 4324
    },
    {
      "epoch": 0.4326,
      "grad_norm": 0.07355896383523941,
      "learning_rate": 0.002,
      "loss": 1.949769377708435,
      "step": 4326
    },
    {
      "epoch": 0.4328,
      "grad_norm": 0.07558101415634155,
      "learning_rate": 0.002,
      "loss": 2.0214569568634033,
      "step": 4328
    },
    {
      "epoch": 0.433,
      "grad_norm": 0.07666043192148209,
      "learning_rate": 0.002,
      "loss": 1.87637460231781,
      "step": 4330
    },
    {
      "epoch": 0.4332,
      "grad_norm": 0.07712709158658981,
      "learning_rate": 0.002,
      "loss": 1.947717547416687,
      "step": 4332
    },
    {
      "epoch": 0.4334,
      "grad_norm": 0.07523921877145767,
      "learning_rate": 0.002,
      "loss": 1.8709204196929932,
      "step": 4334
    },
    {
      "epoch": 0.4336,
      "grad_norm": 0.0756927952170372,
      "learning_rate": 0.002,
      "loss": 1.8964067697525024,
      "step": 4336
    },
    {
      "epoch": 0.4338,
      "grad_norm": 0.07831448316574097,
      "learning_rate": 0.002,
      "loss": 1.9592169523239136,
      "step": 4338
    },
    {
      "epoch": 0.434,
      "grad_norm": 0.07411221414804459,
      "learning_rate": 0.002,
      "loss": 1.8350437879562378,
      "step": 4340
    },
    {
      "epoch": 0.4342,
      "grad_norm": 0.07754648476839066,
      "learning_rate": 0.002,
      "loss": 2.014805316925049,
      "step": 4342
    },
    {
      "epoch": 0.4344,
      "grad_norm": 0.08711760491132736,
      "learning_rate": 0.002,
      "loss": 1.8718562126159668,
      "step": 4344
    },
    {
      "epoch": 0.4346,
      "grad_norm": 0.07618580013513565,
      "learning_rate": 0.002,
      "loss": 1.9253380298614502,
      "step": 4346
    },
    {
      "epoch": 0.4348,
      "grad_norm": 0.07572635263204575,
      "learning_rate": 0.002,
      "loss": 1.8813097476959229,
      "step": 4348
    },
    {
      "epoch": 0.435,
      "grad_norm": 0.0751434713602066,
      "learning_rate": 0.002,
      "loss": 2.0261709690093994,
      "step": 4350
    },
    {
      "epoch": 0.4352,
      "grad_norm": 0.08035343140363693,
      "learning_rate": 0.002,
      "loss": 1.9659714698791504,
      "step": 4352
    },
    {
      "epoch": 0.4354,
      "grad_norm": 0.0749044343829155,
      "learning_rate": 0.002,
      "loss": 1.8361821174621582,
      "step": 4354
    },
    {
      "epoch": 0.4356,
      "grad_norm": 0.07542683184146881,
      "learning_rate": 0.002,
      "loss": 1.832962155342102,
      "step": 4356
    },
    {
      "epoch": 0.4358,
      "grad_norm": 0.07598323374986649,
      "learning_rate": 0.002,
      "loss": 1.8491196632385254,
      "step": 4358
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.08707521110773087,
      "learning_rate": 0.002,
      "loss": 1.9270144701004028,
      "step": 4360
    },
    {
      "epoch": 0.4362,
      "grad_norm": 0.07467620819807053,
      "learning_rate": 0.002,
      "loss": 1.8182520866394043,
      "step": 4362
    },
    {
      "epoch": 0.4364,
      "grad_norm": 0.08100887387990952,
      "learning_rate": 0.002,
      "loss": 1.9119635820388794,
      "step": 4364
    },
    {
      "epoch": 0.4366,
      "grad_norm": 0.07570109516382217,
      "learning_rate": 0.002,
      "loss": 1.9592124223709106,
      "step": 4366
    },
    {
      "epoch": 0.4368,
      "grad_norm": 0.07436879724264145,
      "learning_rate": 0.002,
      "loss": 1.999469518661499,
      "step": 4368
    },
    {
      "epoch": 0.437,
      "grad_norm": 0.07690920680761337,
      "learning_rate": 0.002,
      "loss": 1.8985843658447266,
      "step": 4370
    },
    {
      "epoch": 0.4372,
      "grad_norm": 0.07608062773942947,
      "learning_rate": 0.002,
      "loss": 1.8459922075271606,
      "step": 4372
    },
    {
      "epoch": 0.4374,
      "grad_norm": 0.07592877745628357,
      "learning_rate": 0.002,
      "loss": 1.8663642406463623,
      "step": 4374
    },
    {
      "epoch": 0.4376,
      "grad_norm": 0.0764259397983551,
      "learning_rate": 0.002,
      "loss": 2.0394747257232666,
      "step": 4376
    },
    {
      "epoch": 0.4378,
      "grad_norm": 0.07447104156017303,
      "learning_rate": 0.002,
      "loss": 1.982466697692871,
      "step": 4378
    },
    {
      "epoch": 0.438,
      "grad_norm": 0.07552336901426315,
      "learning_rate": 0.002,
      "loss": 1.8637127876281738,
      "step": 4380
    },
    {
      "epoch": 0.4382,
      "grad_norm": 0.07694466412067413,
      "learning_rate": 0.002,
      "loss": 1.9145708084106445,
      "step": 4382
    },
    {
      "epoch": 0.4384,
      "grad_norm": 0.08034700155258179,
      "learning_rate": 0.002,
      "loss": 1.7948272228240967,
      "step": 4384
    },
    {
      "epoch": 0.4386,
      "grad_norm": 0.0777938961982727,
      "learning_rate": 0.002,
      "loss": 1.9506664276123047,
      "step": 4386
    },
    {
      "epoch": 0.4388,
      "grad_norm": 0.07934419810771942,
      "learning_rate": 0.002,
      "loss": 2.0953855514526367,
      "step": 4388
    },
    {
      "epoch": 0.439,
      "grad_norm": 0.08605864644050598,
      "learning_rate": 0.002,
      "loss": 1.802490472793579,
      "step": 4390
    },
    {
      "epoch": 0.4392,
      "grad_norm": 0.07758272439241409,
      "learning_rate": 0.002,
      "loss": 2.0451157093048096,
      "step": 4392
    },
    {
      "epoch": 0.4394,
      "grad_norm": 0.07612921297550201,
      "learning_rate": 0.002,
      "loss": 1.9528419971466064,
      "step": 4394
    },
    {
      "epoch": 0.4396,
      "grad_norm": 0.083491250872612,
      "learning_rate": 0.002,
      "loss": 2.0484418869018555,
      "step": 4396
    },
    {
      "epoch": 0.4398,
      "grad_norm": 0.07763595879077911,
      "learning_rate": 0.002,
      "loss": 1.984311580657959,
      "step": 4398
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.0791102945804596,
      "learning_rate": 0.002,
      "loss": 1.9786157608032227,
      "step": 4400
    },
    {
      "epoch": 0.4402,
      "grad_norm": 0.07638436555862427,
      "learning_rate": 0.002,
      "loss": 1.9122146368026733,
      "step": 4402
    },
    {
      "epoch": 0.4404,
      "grad_norm": 0.07634909451007843,
      "learning_rate": 0.002,
      "loss": 1.904516577720642,
      "step": 4404
    },
    {
      "epoch": 0.4406,
      "grad_norm": 0.07712217420339584,
      "learning_rate": 0.002,
      "loss": 1.7982704639434814,
      "step": 4406
    },
    {
      "epoch": 0.4408,
      "grad_norm": 0.07951488345861435,
      "learning_rate": 0.002,
      "loss": 1.94503915309906,
      "step": 4408
    },
    {
      "epoch": 0.441,
      "grad_norm": 0.08116553723812103,
      "learning_rate": 0.002,
      "loss": 1.8954395055770874,
      "step": 4410
    },
    {
      "epoch": 0.4412,
      "grad_norm": 0.09295563399791718,
      "learning_rate": 0.002,
      "loss": 1.9488602876663208,
      "step": 4412
    },
    {
      "epoch": 0.4414,
      "grad_norm": 0.08214160054922104,
      "learning_rate": 0.002,
      "loss": 1.8433921337127686,
      "step": 4414
    },
    {
      "epoch": 0.4416,
      "grad_norm": 0.07691451907157898,
      "learning_rate": 0.002,
      "loss": 1.9080727100372314,
      "step": 4416
    },
    {
      "epoch": 0.4418,
      "grad_norm": 0.07867610454559326,
      "learning_rate": 0.002,
      "loss": 1.9388656616210938,
      "step": 4418
    },
    {
      "epoch": 0.442,
      "grad_norm": 0.07817283272743225,
      "learning_rate": 0.002,
      "loss": 1.916428565979004,
      "step": 4420
    },
    {
      "epoch": 0.4422,
      "grad_norm": 0.07565802335739136,
      "learning_rate": 0.002,
      "loss": 1.9548544883728027,
      "step": 4422
    },
    {
      "epoch": 0.4424,
      "grad_norm": 0.0769091248512268,
      "learning_rate": 0.002,
      "loss": 1.9108988046646118,
      "step": 4424
    },
    {
      "epoch": 0.4426,
      "grad_norm": 0.08099309355020523,
      "learning_rate": 0.002,
      "loss": 1.9864474534988403,
      "step": 4426
    },
    {
      "epoch": 0.4428,
      "grad_norm": 0.07217597216367722,
      "learning_rate": 0.002,
      "loss": 1.93940007686615,
      "step": 4428
    },
    {
      "epoch": 0.443,
      "grad_norm": 0.07883615791797638,
      "learning_rate": 0.002,
      "loss": 1.8673101663589478,
      "step": 4430
    },
    {
      "epoch": 0.4432,
      "grad_norm": 0.07334518432617188,
      "learning_rate": 0.002,
      "loss": 1.750142216682434,
      "step": 4432
    },
    {
      "epoch": 0.4434,
      "grad_norm": 0.07515864074230194,
      "learning_rate": 0.002,
      "loss": 1.989821195602417,
      "step": 4434
    },
    {
      "epoch": 0.4436,
      "grad_norm": 0.07558707892894745,
      "learning_rate": 0.002,
      "loss": 2.042022228240967,
      "step": 4436
    },
    {
      "epoch": 0.4438,
      "grad_norm": 0.07716289162635803,
      "learning_rate": 0.002,
      "loss": 1.8329424858093262,
      "step": 4438
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.07675910741090775,
      "learning_rate": 0.002,
      "loss": 1.9081138372421265,
      "step": 4440
    },
    {
      "epoch": 0.4442,
      "grad_norm": 0.07461361587047577,
      "learning_rate": 0.002,
      "loss": 2.0277440547943115,
      "step": 4442
    },
    {
      "epoch": 0.4444,
      "grad_norm": 0.07568936794996262,
      "learning_rate": 0.002,
      "loss": 1.9995198249816895,
      "step": 4444
    },
    {
      "epoch": 0.4446,
      "grad_norm": 0.08073969930410385,
      "learning_rate": 0.002,
      "loss": 1.9222761392593384,
      "step": 4446
    },
    {
      "epoch": 0.4448,
      "grad_norm": 0.07419782876968384,
      "learning_rate": 0.002,
      "loss": 1.9171452522277832,
      "step": 4448
    },
    {
      "epoch": 0.445,
      "grad_norm": 0.07464037090539932,
      "learning_rate": 0.002,
      "loss": 1.8200064897537231,
      "step": 4450
    },
    {
      "epoch": 0.4452,
      "grad_norm": 0.07842778414487839,
      "learning_rate": 0.002,
      "loss": 1.945543646812439,
      "step": 4452
    },
    {
      "epoch": 0.4454,
      "grad_norm": 0.07946744561195374,
      "learning_rate": 0.002,
      "loss": 2.0392794609069824,
      "step": 4454
    },
    {
      "epoch": 0.4456,
      "grad_norm": 0.07547920942306519,
      "learning_rate": 0.002,
      "loss": 1.9224576950073242,
      "step": 4456
    },
    {
      "epoch": 0.4458,
      "grad_norm": 0.07485529035329819,
      "learning_rate": 0.002,
      "loss": 1.9203962087631226,
      "step": 4458
    },
    {
      "epoch": 0.446,
      "grad_norm": 0.07680600881576538,
      "learning_rate": 0.002,
      "loss": 1.9757267236709595,
      "step": 4460
    },
    {
      "epoch": 0.4462,
      "grad_norm": 0.07710090279579163,
      "learning_rate": 0.002,
      "loss": 1.9077199697494507,
      "step": 4462
    },
    {
      "epoch": 0.4464,
      "grad_norm": 0.07553689926862717,
      "learning_rate": 0.002,
      "loss": 1.9404401779174805,
      "step": 4464
    },
    {
      "epoch": 0.4466,
      "grad_norm": 0.07356130331754684,
      "learning_rate": 0.002,
      "loss": 1.8897314071655273,
      "step": 4466
    },
    {
      "epoch": 0.4468,
      "grad_norm": 0.07682061195373535,
      "learning_rate": 0.002,
      "loss": 1.855183720588684,
      "step": 4468
    },
    {
      "epoch": 0.447,
      "grad_norm": 0.07832589745521545,
      "learning_rate": 0.002,
      "loss": 2.0080978870391846,
      "step": 4470
    },
    {
      "epoch": 0.4472,
      "grad_norm": 0.0767950788140297,
      "learning_rate": 0.002,
      "loss": 1.846803069114685,
      "step": 4472
    },
    {
      "epoch": 0.4474,
      "grad_norm": 0.07722268253564835,
      "learning_rate": 0.002,
      "loss": 1.9280329942703247,
      "step": 4474
    },
    {
      "epoch": 0.4476,
      "grad_norm": 0.0756676197052002,
      "learning_rate": 0.002,
      "loss": 1.8951389789581299,
      "step": 4476
    },
    {
      "epoch": 0.4478,
      "grad_norm": 0.08184629678726196,
      "learning_rate": 0.002,
      "loss": 1.8642023801803589,
      "step": 4478
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.07699218392372131,
      "learning_rate": 0.002,
      "loss": 1.9975289106369019,
      "step": 4480
    },
    {
      "epoch": 0.4482,
      "grad_norm": 0.07823902368545532,
      "learning_rate": 0.002,
      "loss": 1.9961580038070679,
      "step": 4482
    },
    {
      "epoch": 0.4484,
      "grad_norm": 0.07319702208042145,
      "learning_rate": 0.002,
      "loss": 1.9073787927627563,
      "step": 4484
    },
    {
      "epoch": 0.4486,
      "grad_norm": 0.07331398129463196,
      "learning_rate": 0.002,
      "loss": 1.8133389949798584,
      "step": 4486
    },
    {
      "epoch": 0.4488,
      "grad_norm": 0.07497964054346085,
      "learning_rate": 0.002,
      "loss": 1.8885499238967896,
      "step": 4488
    },
    {
      "epoch": 0.449,
      "grad_norm": 0.08179699629545212,
      "learning_rate": 0.002,
      "loss": 2.000436544418335,
      "step": 4490
    },
    {
      "epoch": 0.4492,
      "grad_norm": 0.07817253470420837,
      "learning_rate": 0.002,
      "loss": 1.9456244707107544,
      "step": 4492
    },
    {
      "epoch": 0.4494,
      "grad_norm": 0.07585041224956512,
      "learning_rate": 0.002,
      "loss": 1.8229479789733887,
      "step": 4494
    },
    {
      "epoch": 0.4496,
      "grad_norm": 0.0758819654583931,
      "learning_rate": 0.002,
      "loss": 1.9011741876602173,
      "step": 4496
    },
    {
      "epoch": 0.4498,
      "grad_norm": 0.07940684258937836,
      "learning_rate": 0.002,
      "loss": 1.967094898223877,
      "step": 4498
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.07653756439685822,
      "learning_rate": 0.002,
      "loss": 1.881679892539978,
      "step": 4500
    },
    {
      "epoch": 0.4502,
      "grad_norm": 0.07750954478979111,
      "learning_rate": 0.002,
      "loss": 1.814648151397705,
      "step": 4502
    },
    {
      "epoch": 0.4504,
      "grad_norm": 0.08117345720529556,
      "learning_rate": 0.002,
      "loss": 1.9430146217346191,
      "step": 4504
    },
    {
      "epoch": 0.4506,
      "grad_norm": 0.07604934275150299,
      "learning_rate": 0.002,
      "loss": 1.8993536233901978,
      "step": 4506
    },
    {
      "epoch": 0.4508,
      "grad_norm": 0.07624521851539612,
      "learning_rate": 0.002,
      "loss": 1.870461106300354,
      "step": 4508
    },
    {
      "epoch": 0.451,
      "grad_norm": 0.0805351659655571,
      "learning_rate": 0.002,
      "loss": 1.8940660953521729,
      "step": 4510
    },
    {
      "epoch": 0.4512,
      "grad_norm": 0.07419658452272415,
      "learning_rate": 0.002,
      "loss": 1.7942047119140625,
      "step": 4512
    },
    {
      "epoch": 0.4514,
      "grad_norm": 0.0773969516158104,
      "learning_rate": 0.002,
      "loss": 1.8044962882995605,
      "step": 4514
    },
    {
      "epoch": 0.4516,
      "grad_norm": 0.07546841353178024,
      "learning_rate": 0.002,
      "loss": 1.8688353300094604,
      "step": 4516
    },
    {
      "epoch": 0.4518,
      "grad_norm": 0.07659590989351273,
      "learning_rate": 0.002,
      "loss": 1.9362694025039673,
      "step": 4518
    },
    {
      "epoch": 0.452,
      "grad_norm": 0.07669554650783539,
      "learning_rate": 0.002,
      "loss": 2.0105865001678467,
      "step": 4520
    },
    {
      "epoch": 0.4522,
      "grad_norm": 0.07856843620538712,
      "learning_rate": 0.002,
      "loss": 1.9837236404418945,
      "step": 4522
    },
    {
      "epoch": 0.4524,
      "grad_norm": 0.07527710497379303,
      "learning_rate": 0.002,
      "loss": 1.9381033182144165,
      "step": 4524
    },
    {
      "epoch": 0.4526,
      "grad_norm": 0.07974345982074738,
      "learning_rate": 0.002,
      "loss": 1.8816139698028564,
      "step": 4526
    },
    {
      "epoch": 0.4528,
      "grad_norm": 0.0734303742647171,
      "learning_rate": 0.002,
      "loss": 1.9985204935073853,
      "step": 4528
    },
    {
      "epoch": 0.453,
      "grad_norm": 0.08445390313863754,
      "learning_rate": 0.002,
      "loss": 1.859186053276062,
      "step": 4530
    },
    {
      "epoch": 0.4532,
      "grad_norm": 0.08036835491657257,
      "learning_rate": 0.002,
      "loss": 1.841417908668518,
      "step": 4532
    },
    {
      "epoch": 0.4534,
      "grad_norm": 0.0748956948518753,
      "learning_rate": 0.002,
      "loss": 1.8956902027130127,
      "step": 4534
    },
    {
      "epoch": 0.4536,
      "grad_norm": 0.07836831361055374,
      "learning_rate": 0.002,
      "loss": 1.825932502746582,
      "step": 4536
    },
    {
      "epoch": 0.4538,
      "grad_norm": 0.07777626812458038,
      "learning_rate": 0.002,
      "loss": 1.86238694190979,
      "step": 4538
    },
    {
      "epoch": 0.454,
      "grad_norm": 0.07480144500732422,
      "learning_rate": 0.002,
      "loss": 1.9913572072982788,
      "step": 4540
    },
    {
      "epoch": 0.4542,
      "grad_norm": 0.07819059491157532,
      "learning_rate": 0.002,
      "loss": 1.9350841045379639,
      "step": 4542
    },
    {
      "epoch": 0.4544,
      "grad_norm": 0.08274917304515839,
      "learning_rate": 0.002,
      "loss": 1.9642534255981445,
      "step": 4544
    },
    {
      "epoch": 0.4546,
      "grad_norm": 0.0751175731420517,
      "learning_rate": 0.002,
      "loss": 1.907556414604187,
      "step": 4546
    },
    {
      "epoch": 0.4548,
      "grad_norm": 0.07898165285587311,
      "learning_rate": 0.002,
      "loss": 1.8027478456497192,
      "step": 4548
    },
    {
      "epoch": 0.455,
      "grad_norm": 0.07397587597370148,
      "learning_rate": 0.002,
      "loss": 1.862371563911438,
      "step": 4550
    },
    {
      "epoch": 0.4552,
      "grad_norm": 0.07550656795501709,
      "learning_rate": 0.002,
      "loss": 1.8224050998687744,
      "step": 4552
    },
    {
      "epoch": 0.4554,
      "grad_norm": 0.07698109745979309,
      "learning_rate": 0.002,
      "loss": 1.8893905878067017,
      "step": 4554
    },
    {
      "epoch": 0.4556,
      "grad_norm": 0.07509088516235352,
      "learning_rate": 0.002,
      "loss": 1.8712774515151978,
      "step": 4556
    },
    {
      "epoch": 0.4558,
      "grad_norm": 0.07845640927553177,
      "learning_rate": 0.002,
      "loss": 1.8949352502822876,
      "step": 4558
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.07710308581590652,
      "learning_rate": 0.002,
      "loss": 1.8873896598815918,
      "step": 4560
    },
    {
      "epoch": 0.4562,
      "grad_norm": 0.07981869578361511,
      "learning_rate": 0.002,
      "loss": 1.8482447862625122,
      "step": 4562
    },
    {
      "epoch": 0.4564,
      "grad_norm": 0.07421505451202393,
      "learning_rate": 0.002,
      "loss": 1.8956412076950073,
      "step": 4564
    },
    {
      "epoch": 0.4566,
      "grad_norm": 0.07779022306203842,
      "learning_rate": 0.002,
      "loss": 1.8980358839035034,
      "step": 4566
    },
    {
      "epoch": 0.4568,
      "grad_norm": 0.07917476445436478,
      "learning_rate": 0.002,
      "loss": 1.9414198398590088,
      "step": 4568
    },
    {
      "epoch": 0.457,
      "grad_norm": 0.07971633225679398,
      "learning_rate": 0.002,
      "loss": 1.8665781021118164,
      "step": 4570
    },
    {
      "epoch": 0.4572,
      "grad_norm": 0.07188338786363602,
      "learning_rate": 0.002,
      "loss": 1.928192138671875,
      "step": 4572
    },
    {
      "epoch": 0.4574,
      "grad_norm": 0.07669255882501602,
      "learning_rate": 0.002,
      "loss": 1.9775776863098145,
      "step": 4574
    },
    {
      "epoch": 0.4576,
      "grad_norm": 0.07645385712385178,
      "learning_rate": 0.002,
      "loss": 1.8731781244277954,
      "step": 4576
    },
    {
      "epoch": 0.4578,
      "grad_norm": 0.07640064507722855,
      "learning_rate": 0.002,
      "loss": 1.95632004737854,
      "step": 4578
    },
    {
      "epoch": 0.458,
      "grad_norm": 0.07402151077985764,
      "learning_rate": 0.002,
      "loss": 1.9244754314422607,
      "step": 4580
    },
    {
      "epoch": 0.4582,
      "grad_norm": 0.07611144334077835,
      "learning_rate": 0.002,
      "loss": 1.9091871976852417,
      "step": 4582
    },
    {
      "epoch": 0.4584,
      "grad_norm": 0.07756911963224411,
      "learning_rate": 0.002,
      "loss": 1.8687697649002075,
      "step": 4584
    },
    {
      "epoch": 0.4586,
      "grad_norm": 0.07415728271007538,
      "learning_rate": 0.002,
      "loss": 1.9561268091201782,
      "step": 4586
    },
    {
      "epoch": 0.4588,
      "grad_norm": 0.08233364671468735,
      "learning_rate": 0.002,
      "loss": 1.9128543138504028,
      "step": 4588
    },
    {
      "epoch": 0.459,
      "grad_norm": 0.07377967238426208,
      "learning_rate": 0.002,
      "loss": 1.8130148649215698,
      "step": 4590
    },
    {
      "epoch": 0.4592,
      "grad_norm": 0.07528968900442123,
      "learning_rate": 0.002,
      "loss": 1.936729073524475,
      "step": 4592
    },
    {
      "epoch": 0.4594,
      "grad_norm": 0.08138202875852585,
      "learning_rate": 0.002,
      "loss": 1.9555259943008423,
      "step": 4594
    },
    {
      "epoch": 0.4596,
      "grad_norm": 0.07286155223846436,
      "learning_rate": 0.002,
      "loss": 2.00209641456604,
      "step": 4596
    },
    {
      "epoch": 0.4598,
      "grad_norm": 0.07846204936504364,
      "learning_rate": 0.002,
      "loss": 1.8868814706802368,
      "step": 4598
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.07967796921730042,
      "learning_rate": 0.002,
      "loss": 1.9617364406585693,
      "step": 4600
    },
    {
      "epoch": 0.4602,
      "grad_norm": 0.07430858910083771,
      "learning_rate": 0.002,
      "loss": 1.9069019556045532,
      "step": 4602
    },
    {
      "epoch": 0.4604,
      "grad_norm": 0.08361660689115524,
      "learning_rate": 0.002,
      "loss": 2.0301084518432617,
      "step": 4604
    },
    {
      "epoch": 0.4606,
      "grad_norm": 0.07656025886535645,
      "learning_rate": 0.002,
      "loss": 1.8396046161651611,
      "step": 4606
    },
    {
      "epoch": 0.4608,
      "grad_norm": 0.08147291839122772,
      "learning_rate": 0.002,
      "loss": 2.0369067192077637,
      "step": 4608
    },
    {
      "epoch": 0.461,
      "grad_norm": 0.08734532445669174,
      "learning_rate": 0.002,
      "loss": 1.9723149538040161,
      "step": 4610
    },
    {
      "epoch": 0.4612,
      "grad_norm": 0.082717664539814,
      "learning_rate": 0.002,
      "loss": 1.9132580757141113,
      "step": 4612
    },
    {
      "epoch": 0.4614,
      "grad_norm": 0.07726579904556274,
      "learning_rate": 0.002,
      "loss": 1.9110057353973389,
      "step": 4614
    },
    {
      "epoch": 0.4616,
      "grad_norm": 0.07526474446058273,
      "learning_rate": 0.002,
      "loss": 1.9605709314346313,
      "step": 4616
    },
    {
      "epoch": 0.4618,
      "grad_norm": 0.07667647302150726,
      "learning_rate": 0.002,
      "loss": 1.8231209516525269,
      "step": 4618
    },
    {
      "epoch": 0.462,
      "grad_norm": 0.07885342836380005,
      "learning_rate": 0.002,
      "loss": 1.9372972249984741,
      "step": 4620
    },
    {
      "epoch": 0.4622,
      "grad_norm": 0.07632264494895935,
      "learning_rate": 0.002,
      "loss": 1.902121901512146,
      "step": 4622
    },
    {
      "epoch": 0.4624,
      "grad_norm": 0.07621581107378006,
      "learning_rate": 0.002,
      "loss": 1.905409336090088,
      "step": 4624
    },
    {
      "epoch": 0.4626,
      "grad_norm": 0.07466264069080353,
      "learning_rate": 0.002,
      "loss": 1.8778988122940063,
      "step": 4626
    },
    {
      "epoch": 0.4628,
      "grad_norm": 0.07600996643304825,
      "learning_rate": 0.002,
      "loss": 1.9414368867874146,
      "step": 4628
    },
    {
      "epoch": 0.463,
      "grad_norm": 0.07603158801794052,
      "learning_rate": 0.002,
      "loss": 1.9728502035140991,
      "step": 4630
    },
    {
      "epoch": 0.4632,
      "grad_norm": 0.07601157575845718,
      "learning_rate": 0.002,
      "loss": 2.003896713256836,
      "step": 4632
    },
    {
      "epoch": 0.4634,
      "grad_norm": 0.0737384632229805,
      "learning_rate": 0.002,
      "loss": 1.9144420623779297,
      "step": 4634
    },
    {
      "epoch": 0.4636,
      "grad_norm": 0.07685447484254837,
      "learning_rate": 0.002,
      "loss": 1.9235494136810303,
      "step": 4636
    },
    {
      "epoch": 0.4638,
      "grad_norm": 0.07489930093288422,
      "learning_rate": 0.002,
      "loss": 1.9681390523910522,
      "step": 4638
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.07560908049345016,
      "learning_rate": 0.002,
      "loss": 1.8456811904907227,
      "step": 4640
    },
    {
      "epoch": 0.4642,
      "grad_norm": 0.07514713704586029,
      "learning_rate": 0.002,
      "loss": 1.9637227058410645,
      "step": 4642
    },
    {
      "epoch": 0.4644,
      "grad_norm": 0.07777653634548187,
      "learning_rate": 0.002,
      "loss": 1.9169467687606812,
      "step": 4644
    },
    {
      "epoch": 0.4646,
      "grad_norm": 0.07749330252408981,
      "learning_rate": 0.002,
      "loss": 1.784414291381836,
      "step": 4646
    },
    {
      "epoch": 0.4648,
      "grad_norm": 0.07174395769834518,
      "learning_rate": 0.002,
      "loss": 1.981076717376709,
      "step": 4648
    },
    {
      "epoch": 0.465,
      "grad_norm": 0.07256866246461868,
      "learning_rate": 0.002,
      "loss": 1.8354450464248657,
      "step": 4650
    },
    {
      "epoch": 0.4652,
      "grad_norm": 0.07869680970907211,
      "learning_rate": 0.002,
      "loss": 1.9138424396514893,
      "step": 4652
    },
    {
      "epoch": 0.4654,
      "grad_norm": 0.07865259051322937,
      "learning_rate": 0.002,
      "loss": 1.8351348638534546,
      "step": 4654
    },
    {
      "epoch": 0.4656,
      "grad_norm": 0.07548552751541138,
      "learning_rate": 0.002,
      "loss": 1.8511457443237305,
      "step": 4656
    },
    {
      "epoch": 0.4658,
      "grad_norm": 0.07721854746341705,
      "learning_rate": 0.002,
      "loss": 1.8904929161071777,
      "step": 4658
    },
    {
      "epoch": 0.466,
      "grad_norm": 0.08108288794755936,
      "learning_rate": 0.002,
      "loss": 1.9396981000900269,
      "step": 4660
    },
    {
      "epoch": 0.4662,
      "grad_norm": 0.07977964729070663,
      "learning_rate": 0.002,
      "loss": 1.9784750938415527,
      "step": 4662
    },
    {
      "epoch": 0.4664,
      "grad_norm": 0.07868921011686325,
      "learning_rate": 0.002,
      "loss": 1.9139889478683472,
      "step": 4664
    },
    {
      "epoch": 0.4666,
      "grad_norm": 0.07331487536430359,
      "learning_rate": 0.002,
      "loss": 1.8636809587478638,
      "step": 4666
    },
    {
      "epoch": 0.4668,
      "grad_norm": 0.07579874992370605,
      "learning_rate": 0.002,
      "loss": 1.9328420162200928,
      "step": 4668
    },
    {
      "epoch": 0.467,
      "grad_norm": 0.0770130529999733,
      "learning_rate": 0.002,
      "loss": 1.8904999494552612,
      "step": 4670
    },
    {
      "epoch": 0.4672,
      "grad_norm": 0.07800159603357315,
      "learning_rate": 0.002,
      "loss": 1.8191505670547485,
      "step": 4672
    },
    {
      "epoch": 0.4674,
      "grad_norm": 0.0747978538274765,
      "learning_rate": 0.002,
      "loss": 1.8864425420761108,
      "step": 4674
    },
    {
      "epoch": 0.4676,
      "grad_norm": 0.08291503041982651,
      "learning_rate": 0.002,
      "loss": 2.0076372623443604,
      "step": 4676
    },
    {
      "epoch": 0.4678,
      "grad_norm": 0.07752741873264313,
      "learning_rate": 0.002,
      "loss": 1.9571179151535034,
      "step": 4678
    },
    {
      "epoch": 0.468,
      "grad_norm": 0.07727371156215668,
      "learning_rate": 0.002,
      "loss": 1.9225423336029053,
      "step": 4680
    },
    {
      "epoch": 0.4682,
      "grad_norm": 0.07288603484630585,
      "learning_rate": 0.002,
      "loss": 1.8352162837982178,
      "step": 4682
    },
    {
      "epoch": 0.4684,
      "grad_norm": 0.07666505873203278,
      "learning_rate": 0.002,
      "loss": 2.033914566040039,
      "step": 4684
    },
    {
      "epoch": 0.4686,
      "grad_norm": 0.07902664691209793,
      "learning_rate": 0.002,
      "loss": 1.9038366079330444,
      "step": 4686
    },
    {
      "epoch": 0.4688,
      "grad_norm": 0.0789552628993988,
      "learning_rate": 0.002,
      "loss": 1.9633876085281372,
      "step": 4688
    },
    {
      "epoch": 0.469,
      "grad_norm": 0.0835079476237297,
      "learning_rate": 0.002,
      "loss": 1.9947394132614136,
      "step": 4690
    },
    {
      "epoch": 0.4692,
      "grad_norm": 0.07937032729387283,
      "learning_rate": 0.002,
      "loss": 2.0458548069000244,
      "step": 4692
    },
    {
      "epoch": 0.4694,
      "grad_norm": 0.07432875782251358,
      "learning_rate": 0.002,
      "loss": 2.0382583141326904,
      "step": 4694
    },
    {
      "epoch": 0.4696,
      "grad_norm": 0.07617273926734924,
      "learning_rate": 0.002,
      "loss": 1.9269047975540161,
      "step": 4696
    },
    {
      "epoch": 0.4698,
      "grad_norm": 0.0784926563501358,
      "learning_rate": 0.002,
      "loss": 2.0084354877471924,
      "step": 4698
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.0750475823879242,
      "learning_rate": 0.002,
      "loss": 1.8045767545700073,
      "step": 4700
    },
    {
      "epoch": 0.4702,
      "grad_norm": 0.07789719849824905,
      "learning_rate": 0.002,
      "loss": 1.9716475009918213,
      "step": 4702
    },
    {
      "epoch": 0.4704,
      "grad_norm": 0.07764974236488342,
      "learning_rate": 0.002,
      "loss": 1.8761680126190186,
      "step": 4704
    },
    {
      "epoch": 0.4706,
      "grad_norm": 0.07644437998533249,
      "learning_rate": 0.002,
      "loss": 1.8406964540481567,
      "step": 4706
    },
    {
      "epoch": 0.4708,
      "grad_norm": 0.08073212951421738,
      "learning_rate": 0.002,
      "loss": 2.0253984928131104,
      "step": 4708
    },
    {
      "epoch": 0.471,
      "grad_norm": 0.07572949677705765,
      "learning_rate": 0.002,
      "loss": 1.8979188203811646,
      "step": 4710
    },
    {
      "epoch": 0.4712,
      "grad_norm": 0.07566970586776733,
      "learning_rate": 0.002,
      "loss": 1.9204180240631104,
      "step": 4712
    },
    {
      "epoch": 0.4714,
      "grad_norm": 0.07663935422897339,
      "learning_rate": 0.002,
      "loss": 1.9192816019058228,
      "step": 4714
    },
    {
      "epoch": 0.4716,
      "grad_norm": 0.07693604379892349,
      "learning_rate": 0.002,
      "loss": 2.017857551574707,
      "step": 4716
    },
    {
      "epoch": 0.4718,
      "grad_norm": 0.07713306695222855,
      "learning_rate": 0.002,
      "loss": 1.7905046939849854,
      "step": 4718
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.07763632386922836,
      "learning_rate": 0.002,
      "loss": 1.8461984395980835,
      "step": 4720
    },
    {
      "epoch": 0.4722,
      "grad_norm": 0.07668328285217285,
      "learning_rate": 0.002,
      "loss": 1.8679635524749756,
      "step": 4722
    },
    {
      "epoch": 0.4724,
      "grad_norm": 0.07909642159938812,
      "learning_rate": 0.002,
      "loss": 1.9264851808547974,
      "step": 4724
    },
    {
      "epoch": 0.4726,
      "grad_norm": 0.07663606852293015,
      "learning_rate": 0.002,
      "loss": 1.8158987760543823,
      "step": 4726
    },
    {
      "epoch": 0.4728,
      "grad_norm": 0.07747005671262741,
      "learning_rate": 0.002,
      "loss": 1.794020175933838,
      "step": 4728
    },
    {
      "epoch": 0.473,
      "grad_norm": 0.07363753765821457,
      "learning_rate": 0.002,
      "loss": 1.890429973602295,
      "step": 4730
    },
    {
      "epoch": 0.4732,
      "grad_norm": 0.07768632471561432,
      "learning_rate": 0.002,
      "loss": 1.9538636207580566,
      "step": 4732
    },
    {
      "epoch": 0.4734,
      "grad_norm": 0.07520966231822968,
      "learning_rate": 0.002,
      "loss": 1.850694179534912,
      "step": 4734
    },
    {
      "epoch": 0.4736,
      "grad_norm": 0.07730185985565186,
      "learning_rate": 0.002,
      "loss": 1.9345868825912476,
      "step": 4736
    },
    {
      "epoch": 0.4738,
      "grad_norm": 0.07741639763116837,
      "learning_rate": 0.002,
      "loss": 1.97011137008667,
      "step": 4738
    },
    {
      "epoch": 0.474,
      "grad_norm": 0.07200247049331665,
      "learning_rate": 0.002,
      "loss": 1.9062471389770508,
      "step": 4740
    },
    {
      "epoch": 0.4742,
      "grad_norm": 0.07758310437202454,
      "learning_rate": 0.002,
      "loss": 2.0186941623687744,
      "step": 4742
    },
    {
      "epoch": 0.4744,
      "grad_norm": 0.07586324214935303,
      "learning_rate": 0.002,
      "loss": 1.8399120569229126,
      "step": 4744
    },
    {
      "epoch": 0.4746,
      "grad_norm": 0.07630598545074463,
      "learning_rate": 0.002,
      "loss": 1.973314881324768,
      "step": 4746
    },
    {
      "epoch": 0.4748,
      "grad_norm": 0.08194039762020111,
      "learning_rate": 0.002,
      "loss": 1.8798596858978271,
      "step": 4748
    },
    {
      "epoch": 0.475,
      "grad_norm": 0.07479400187730789,
      "learning_rate": 0.002,
      "loss": 1.825071096420288,
      "step": 4750
    },
    {
      "epoch": 0.4752,
      "grad_norm": 0.07654538005590439,
      "learning_rate": 0.002,
      "loss": 1.8827930688858032,
      "step": 4752
    },
    {
      "epoch": 0.4754,
      "grad_norm": 0.07659152895212173,
      "learning_rate": 0.002,
      "loss": 1.9078583717346191,
      "step": 4754
    },
    {
      "epoch": 0.4756,
      "grad_norm": 0.07481306046247482,
      "learning_rate": 0.002,
      "loss": 1.8636754751205444,
      "step": 4756
    },
    {
      "epoch": 0.4758,
      "grad_norm": 0.0761183351278305,
      "learning_rate": 0.002,
      "loss": 1.924257755279541,
      "step": 4758
    },
    {
      "epoch": 0.476,
      "grad_norm": 0.0770338699221611,
      "learning_rate": 0.002,
      "loss": 1.913481593132019,
      "step": 4760
    },
    {
      "epoch": 0.4762,
      "grad_norm": 0.07345000654459,
      "learning_rate": 0.002,
      "loss": 1.9214493036270142,
      "step": 4762
    },
    {
      "epoch": 0.4764,
      "grad_norm": 0.07696391642093658,
      "learning_rate": 0.002,
      "loss": 1.8494043350219727,
      "step": 4764
    },
    {
      "epoch": 0.4766,
      "grad_norm": 0.0777350440621376,
      "learning_rate": 0.002,
      "loss": 1.9322575330734253,
      "step": 4766
    },
    {
      "epoch": 0.4768,
      "grad_norm": 0.07826673239469528,
      "learning_rate": 0.002,
      "loss": 1.8740935325622559,
      "step": 4768
    },
    {
      "epoch": 0.477,
      "grad_norm": 0.07503682374954224,
      "learning_rate": 0.002,
      "loss": 1.812285304069519,
      "step": 4770
    },
    {
      "epoch": 0.4772,
      "grad_norm": 0.07542384415864944,
      "learning_rate": 0.002,
      "loss": 1.8322913646697998,
      "step": 4772
    },
    {
      "epoch": 0.4774,
      "grad_norm": 0.0784786194562912,
      "learning_rate": 0.002,
      "loss": 2.024500608444214,
      "step": 4774
    },
    {
      "epoch": 0.4776,
      "grad_norm": 0.10391594469547272,
      "learning_rate": 0.002,
      "loss": 1.9053958654403687,
      "step": 4776
    },
    {
      "epoch": 0.4778,
      "grad_norm": 0.07730287313461304,
      "learning_rate": 0.002,
      "loss": 1.882632851600647,
      "step": 4778
    },
    {
      "epoch": 0.478,
      "grad_norm": 0.07634396851062775,
      "learning_rate": 0.002,
      "loss": 1.8886483907699585,
      "step": 4780
    },
    {
      "epoch": 0.4782,
      "grad_norm": 0.07878262549638748,
      "learning_rate": 0.002,
      "loss": 1.943835735321045,
      "step": 4782
    },
    {
      "epoch": 0.4784,
      "grad_norm": 0.07813434302806854,
      "learning_rate": 0.002,
      "loss": 1.8831496238708496,
      "step": 4784
    },
    {
      "epoch": 0.4786,
      "grad_norm": 0.08609431982040405,
      "learning_rate": 0.002,
      "loss": 1.963183879852295,
      "step": 4786
    },
    {
      "epoch": 0.4788,
      "grad_norm": 0.07599510997533798,
      "learning_rate": 0.002,
      "loss": 1.933810830116272,
      "step": 4788
    },
    {
      "epoch": 0.479,
      "grad_norm": 0.07493213564157486,
      "learning_rate": 0.002,
      "loss": 1.8319814205169678,
      "step": 4790
    },
    {
      "epoch": 0.4792,
      "grad_norm": 0.07820174843072891,
      "learning_rate": 0.002,
      "loss": 1.924299955368042,
      "step": 4792
    },
    {
      "epoch": 0.4794,
      "grad_norm": 0.07866837084293365,
      "learning_rate": 0.002,
      "loss": 1.9374932050704956,
      "step": 4794
    },
    {
      "epoch": 0.4796,
      "grad_norm": 0.07307623326778412,
      "learning_rate": 0.002,
      "loss": 1.844019889831543,
      "step": 4796
    },
    {
      "epoch": 0.4798,
      "grad_norm": 0.07412610203027725,
      "learning_rate": 0.002,
      "loss": 1.9692238569259644,
      "step": 4798
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.07561323791742325,
      "learning_rate": 0.002,
      "loss": 1.915809154510498,
      "step": 4800
    },
    {
      "epoch": 0.4802,
      "grad_norm": 0.07382841408252716,
      "learning_rate": 0.002,
      "loss": 1.8909051418304443,
      "step": 4802
    },
    {
      "epoch": 0.4804,
      "grad_norm": 0.08334486931562424,
      "learning_rate": 0.002,
      "loss": 1.9549474716186523,
      "step": 4804
    },
    {
      "epoch": 0.4806,
      "grad_norm": 0.08007567375898361,
      "learning_rate": 0.002,
      "loss": 1.852167010307312,
      "step": 4806
    },
    {
      "epoch": 0.4808,
      "grad_norm": 0.07787317782640457,
      "learning_rate": 0.002,
      "loss": 1.9649176597595215,
      "step": 4808
    },
    {
      "epoch": 0.481,
      "grad_norm": 0.07743795216083527,
      "learning_rate": 0.002,
      "loss": 1.9184291362762451,
      "step": 4810
    },
    {
      "epoch": 0.4812,
      "grad_norm": 0.0806642696261406,
      "learning_rate": 0.002,
      "loss": 1.9885300397872925,
      "step": 4812
    },
    {
      "epoch": 0.4814,
      "grad_norm": 0.07526041567325592,
      "learning_rate": 0.002,
      "loss": 1.9253368377685547,
      "step": 4814
    },
    {
      "epoch": 0.4816,
      "grad_norm": 0.07983514666557312,
      "learning_rate": 0.002,
      "loss": 1.8332823514938354,
      "step": 4816
    },
    {
      "epoch": 0.4818,
      "grad_norm": 0.08002350479364395,
      "learning_rate": 0.002,
      "loss": 1.8506144285202026,
      "step": 4818
    },
    {
      "epoch": 0.482,
      "grad_norm": 0.08066806942224503,
      "learning_rate": 0.002,
      "loss": 1.8576843738555908,
      "step": 4820
    },
    {
      "epoch": 0.4822,
      "grad_norm": 0.07428060472011566,
      "learning_rate": 0.002,
      "loss": 1.928370475769043,
      "step": 4822
    },
    {
      "epoch": 0.4824,
      "grad_norm": 0.07778865098953247,
      "learning_rate": 0.002,
      "loss": 1.8861929178237915,
      "step": 4824
    },
    {
      "epoch": 0.4826,
      "grad_norm": 0.07518257945775986,
      "learning_rate": 0.002,
      "loss": 1.9131715297698975,
      "step": 4826
    },
    {
      "epoch": 0.4828,
      "grad_norm": 0.07906034588813782,
      "learning_rate": 0.002,
      "loss": 1.8528467416763306,
      "step": 4828
    },
    {
      "epoch": 0.483,
      "grad_norm": 0.0774356946349144,
      "learning_rate": 0.002,
      "loss": 1.818013310432434,
      "step": 4830
    },
    {
      "epoch": 0.4832,
      "grad_norm": 0.07714931666851044,
      "learning_rate": 0.002,
      "loss": 1.9585936069488525,
      "step": 4832
    },
    {
      "epoch": 0.4834,
      "grad_norm": 0.07589243352413177,
      "learning_rate": 0.002,
      "loss": 1.9058692455291748,
      "step": 4834
    },
    {
      "epoch": 0.4836,
      "grad_norm": 0.07692918926477432,
      "learning_rate": 0.002,
      "loss": 1.8647947311401367,
      "step": 4836
    },
    {
      "epoch": 0.4838,
      "grad_norm": 0.07463713735342026,
      "learning_rate": 0.002,
      "loss": 1.94510817527771,
      "step": 4838
    },
    {
      "epoch": 0.484,
      "grad_norm": 0.09202795475721359,
      "learning_rate": 0.002,
      "loss": 2.0534725189208984,
      "step": 4840
    },
    {
      "epoch": 0.4842,
      "grad_norm": 0.08207589387893677,
      "learning_rate": 0.002,
      "loss": 1.9933029413223267,
      "step": 4842
    },
    {
      "epoch": 0.4844,
      "grad_norm": 0.073893703520298,
      "learning_rate": 0.002,
      "loss": 1.952620267868042,
      "step": 4844
    },
    {
      "epoch": 0.4846,
      "grad_norm": 0.07167269289493561,
      "learning_rate": 0.002,
      "loss": 1.9098035097122192,
      "step": 4846
    },
    {
      "epoch": 0.4848,
      "grad_norm": 0.07455581426620483,
      "learning_rate": 0.002,
      "loss": 1.8664599657058716,
      "step": 4848
    },
    {
      "epoch": 0.485,
      "grad_norm": 0.07533092051744461,
      "learning_rate": 0.002,
      "loss": 1.834463119506836,
      "step": 4850
    },
    {
      "epoch": 0.4852,
      "grad_norm": 0.08034082502126694,
      "learning_rate": 0.002,
      "loss": 1.805594801902771,
      "step": 4852
    },
    {
      "epoch": 0.4854,
      "grad_norm": 0.08065011352300644,
      "learning_rate": 0.002,
      "loss": 1.9593884944915771,
      "step": 4854
    },
    {
      "epoch": 0.4856,
      "grad_norm": 0.07778632640838623,
      "learning_rate": 0.002,
      "loss": 1.8672219514846802,
      "step": 4856
    },
    {
      "epoch": 0.4858,
      "grad_norm": 0.07897785305976868,
      "learning_rate": 0.002,
      "loss": 1.8247095346450806,
      "step": 4858
    },
    {
      "epoch": 0.486,
      "grad_norm": 0.07917468249797821,
      "learning_rate": 0.002,
      "loss": 2.090236186981201,
      "step": 4860
    },
    {
      "epoch": 0.4862,
      "grad_norm": 0.07715284824371338,
      "learning_rate": 0.002,
      "loss": 1.8762091398239136,
      "step": 4862
    },
    {
      "epoch": 0.4864,
      "grad_norm": 0.08164190500974655,
      "learning_rate": 0.002,
      "loss": 2.001370668411255,
      "step": 4864
    },
    {
      "epoch": 0.4866,
      "grad_norm": 0.0748026967048645,
      "learning_rate": 0.002,
      "loss": 1.8363094329833984,
      "step": 4866
    },
    {
      "epoch": 0.4868,
      "grad_norm": 0.07714849710464478,
      "learning_rate": 0.002,
      "loss": 1.985836386680603,
      "step": 4868
    },
    {
      "epoch": 0.487,
      "grad_norm": 0.07694429159164429,
      "learning_rate": 0.002,
      "loss": 1.8325716257095337,
      "step": 4870
    },
    {
      "epoch": 0.4872,
      "grad_norm": 0.0753120556473732,
      "learning_rate": 0.002,
      "loss": 1.9109128713607788,
      "step": 4872
    },
    {
      "epoch": 0.4874,
      "grad_norm": 0.07350266724824905,
      "learning_rate": 0.002,
      "loss": 1.8235780000686646,
      "step": 4874
    },
    {
      "epoch": 0.4876,
      "grad_norm": 0.08188395947217941,
      "learning_rate": 0.002,
      "loss": 1.81047523021698,
      "step": 4876
    },
    {
      "epoch": 0.4878,
      "grad_norm": 0.07481710612773895,
      "learning_rate": 0.002,
      "loss": 1.8376716375350952,
      "step": 4878
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.07356520742177963,
      "learning_rate": 0.002,
      "loss": 1.8815280199050903,
      "step": 4880
    },
    {
      "epoch": 0.4882,
      "grad_norm": 0.07942367345094681,
      "learning_rate": 0.002,
      "loss": 1.8801779747009277,
      "step": 4882
    },
    {
      "epoch": 0.4884,
      "grad_norm": 0.075577013194561,
      "learning_rate": 0.002,
      "loss": 1.8638858795166016,
      "step": 4884
    },
    {
      "epoch": 0.4886,
      "grad_norm": 0.07552029192447662,
      "learning_rate": 0.002,
      "loss": 1.8475561141967773,
      "step": 4886
    },
    {
      "epoch": 0.4888,
      "grad_norm": 0.0779888778924942,
      "learning_rate": 0.002,
      "loss": 1.9262275695800781,
      "step": 4888
    },
    {
      "epoch": 0.489,
      "grad_norm": 0.07683084160089493,
      "learning_rate": 0.002,
      "loss": 1.8393968343734741,
      "step": 4890
    },
    {
      "epoch": 0.4892,
      "grad_norm": 0.07581464946269989,
      "learning_rate": 0.002,
      "loss": 1.8984711170196533,
      "step": 4892
    },
    {
      "epoch": 0.4894,
      "grad_norm": 0.07783406972885132,
      "learning_rate": 0.002,
      "loss": 1.9602982997894287,
      "step": 4894
    },
    {
      "epoch": 0.4896,
      "grad_norm": 0.07966019213199615,
      "learning_rate": 0.002,
      "loss": 1.8246008157730103,
      "step": 4896
    },
    {
      "epoch": 0.4898,
      "grad_norm": 0.07424872368574142,
      "learning_rate": 0.002,
      "loss": 1.918647289276123,
      "step": 4898
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.07709755003452301,
      "learning_rate": 0.002,
      "loss": 1.8377209901809692,
      "step": 4900
    },
    {
      "epoch": 0.4902,
      "grad_norm": 0.07895360141992569,
      "learning_rate": 0.002,
      "loss": 1.9338321685791016,
      "step": 4902
    },
    {
      "epoch": 0.4904,
      "grad_norm": 0.07957840710878372,
      "learning_rate": 0.002,
      "loss": 1.8476760387420654,
      "step": 4904
    },
    {
      "epoch": 0.4906,
      "grad_norm": 0.08054302632808685,
      "learning_rate": 0.002,
      "loss": 1.9263107776641846,
      "step": 4906
    },
    {
      "epoch": 0.4908,
      "grad_norm": 0.0837155282497406,
      "learning_rate": 0.002,
      "loss": 2.023362398147583,
      "step": 4908
    },
    {
      "epoch": 0.491,
      "grad_norm": 0.07585028558969498,
      "learning_rate": 0.002,
      "loss": 1.8775887489318848,
      "step": 4910
    },
    {
      "epoch": 0.4912,
      "grad_norm": 0.07576946169137955,
      "learning_rate": 0.002,
      "loss": 1.96893310546875,
      "step": 4912
    },
    {
      "epoch": 0.4914,
      "grad_norm": 0.08085428923368454,
      "learning_rate": 0.002,
      "loss": 1.8107705116271973,
      "step": 4914
    },
    {
      "epoch": 0.4916,
      "grad_norm": 0.07595674693584442,
      "learning_rate": 0.002,
      "loss": 1.9148123264312744,
      "step": 4916
    },
    {
      "epoch": 0.4918,
      "grad_norm": 0.0735955536365509,
      "learning_rate": 0.002,
      "loss": 1.9818700551986694,
      "step": 4918
    },
    {
      "epoch": 0.492,
      "grad_norm": 0.07626427710056305,
      "learning_rate": 0.002,
      "loss": 1.8123599290847778,
      "step": 4920
    },
    {
      "epoch": 0.4922,
      "grad_norm": 0.07677875459194183,
      "learning_rate": 0.002,
      "loss": 1.8991621732711792,
      "step": 4922
    },
    {
      "epoch": 0.4924,
      "grad_norm": 0.07982435822486877,
      "learning_rate": 0.002,
      "loss": 1.8996753692626953,
      "step": 4924
    },
    {
      "epoch": 0.4926,
      "grad_norm": 0.0748787522315979,
      "learning_rate": 0.002,
      "loss": 2.017789602279663,
      "step": 4926
    },
    {
      "epoch": 0.4928,
      "grad_norm": 0.07427710294723511,
      "learning_rate": 0.002,
      "loss": 1.8368968963623047,
      "step": 4928
    },
    {
      "epoch": 0.493,
      "grad_norm": 0.07763507962226868,
      "learning_rate": 0.002,
      "loss": 1.8646732568740845,
      "step": 4930
    },
    {
      "epoch": 0.4932,
      "grad_norm": 0.07974207401275635,
      "learning_rate": 0.002,
      "loss": 1.9171581268310547,
      "step": 4932
    },
    {
      "epoch": 0.4934,
      "grad_norm": 0.07491708546876907,
      "learning_rate": 0.002,
      "loss": 1.894774079322815,
      "step": 4934
    },
    {
      "epoch": 0.4936,
      "grad_norm": 0.07721830159425735,
      "learning_rate": 0.002,
      "loss": 1.7978003025054932,
      "step": 4936
    },
    {
      "epoch": 0.4938,
      "grad_norm": 0.07563165575265884,
      "learning_rate": 0.002,
      "loss": 1.793831706047058,
      "step": 4938
    },
    {
      "epoch": 0.494,
      "grad_norm": 0.07253357768058777,
      "learning_rate": 0.002,
      "loss": 1.8788586854934692,
      "step": 4940
    },
    {
      "epoch": 0.4942,
      "grad_norm": 0.07743583619594574,
      "learning_rate": 0.002,
      "loss": 1.9890260696411133,
      "step": 4942
    },
    {
      "epoch": 0.4944,
      "grad_norm": 0.07775986939668655,
      "learning_rate": 0.002,
      "loss": 1.8367117643356323,
      "step": 4944
    },
    {
      "epoch": 0.4946,
      "grad_norm": 0.07449087500572205,
      "learning_rate": 0.002,
      "loss": 1.8067561388015747,
      "step": 4946
    },
    {
      "epoch": 0.4948,
      "grad_norm": 0.07632667571306229,
      "learning_rate": 0.002,
      "loss": 1.9029849767684937,
      "step": 4948
    },
    {
      "epoch": 0.495,
      "grad_norm": 0.0780206099152565,
      "learning_rate": 0.002,
      "loss": 1.8837454319000244,
      "step": 4950
    },
    {
      "epoch": 0.4952,
      "grad_norm": 0.07611632347106934,
      "learning_rate": 0.002,
      "loss": 1.9619537591934204,
      "step": 4952
    },
    {
      "epoch": 0.4954,
      "grad_norm": 0.08308790624141693,
      "learning_rate": 0.002,
      "loss": 2.035999298095703,
      "step": 4954
    },
    {
      "epoch": 0.4956,
      "grad_norm": 0.07843099534511566,
      "learning_rate": 0.002,
      "loss": 1.8731293678283691,
      "step": 4956
    },
    {
      "epoch": 0.4958,
      "grad_norm": 0.07453984767198563,
      "learning_rate": 0.002,
      "loss": 1.8806856870651245,
      "step": 4958
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.073385089635849,
      "learning_rate": 0.002,
      "loss": 1.8784050941467285,
      "step": 4960
    },
    {
      "epoch": 0.4962,
      "grad_norm": 0.07442124933004379,
      "learning_rate": 0.002,
      "loss": 1.9233325719833374,
      "step": 4962
    },
    {
      "epoch": 0.4964,
      "grad_norm": 0.07996006309986115,
      "learning_rate": 0.002,
      "loss": 1.9295052289962769,
      "step": 4964
    },
    {
      "epoch": 0.4966,
      "grad_norm": 0.07920581847429276,
      "learning_rate": 0.002,
      "loss": 1.8843786716461182,
      "step": 4966
    },
    {
      "epoch": 0.4968,
      "grad_norm": 0.07264573872089386,
      "learning_rate": 0.002,
      "loss": 1.7712863683700562,
      "step": 4968
    },
    {
      "epoch": 0.497,
      "grad_norm": 0.07456866651773453,
      "learning_rate": 0.002,
      "loss": 1.921347975730896,
      "step": 4970
    },
    {
      "epoch": 0.4972,
      "grad_norm": 0.0737089067697525,
      "learning_rate": 0.002,
      "loss": 2.0018229484558105,
      "step": 4972
    },
    {
      "epoch": 0.4974,
      "grad_norm": 0.07591192424297333,
      "learning_rate": 0.002,
      "loss": 1.8762205839157104,
      "step": 4974
    },
    {
      "epoch": 0.4976,
      "grad_norm": 0.07879926264286041,
      "learning_rate": 0.002,
      "loss": 1.932007908821106,
      "step": 4976
    },
    {
      "epoch": 0.4978,
      "grad_norm": 0.07478541880846024,
      "learning_rate": 0.002,
      "loss": 1.8184970617294312,
      "step": 4978
    },
    {
      "epoch": 0.498,
      "grad_norm": 0.08026406913995743,
      "learning_rate": 0.002,
      "loss": 2.036841630935669,
      "step": 4980
    },
    {
      "epoch": 0.4982,
      "grad_norm": 0.08200377970933914,
      "learning_rate": 0.002,
      "loss": 1.8547558784484863,
      "step": 4982
    },
    {
      "epoch": 0.4984,
      "grad_norm": 0.07829662412405014,
      "learning_rate": 0.002,
      "loss": 1.8324763774871826,
      "step": 4984
    },
    {
      "epoch": 0.4986,
      "grad_norm": 0.07853958755731583,
      "learning_rate": 0.002,
      "loss": 1.950388789176941,
      "step": 4986
    },
    {
      "epoch": 0.4988,
      "grad_norm": 0.07966439425945282,
      "learning_rate": 0.002,
      "loss": 1.9807337522506714,
      "step": 4988
    },
    {
      "epoch": 0.499,
      "grad_norm": 0.0745282992720604,
      "learning_rate": 0.002,
      "loss": 1.8589468002319336,
      "step": 4990
    },
    {
      "epoch": 0.4992,
      "grad_norm": 0.07370910793542862,
      "learning_rate": 0.002,
      "loss": 1.813887119293213,
      "step": 4992
    },
    {
      "epoch": 0.4994,
      "grad_norm": 0.080593541264534,
      "learning_rate": 0.002,
      "loss": 2.0139708518981934,
      "step": 4994
    },
    {
      "epoch": 0.4996,
      "grad_norm": 0.08283080905675888,
      "learning_rate": 0.002,
      "loss": 1.8465596437454224,
      "step": 4996
    },
    {
      "epoch": 0.4998,
      "grad_norm": 0.07392429560422897,
      "learning_rate": 0.002,
      "loss": 1.992016315460205,
      "step": 4998
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.07739411294460297,
      "learning_rate": 0.002,
      "loss": 1.9476388692855835,
      "step": 5000
    },
    {
      "epoch": 0.5002,
      "grad_norm": 0.07594774663448334,
      "learning_rate": 0.002,
      "loss": 2.0409107208251953,
      "step": 5002
    },
    {
      "epoch": 0.5004,
      "grad_norm": 0.07916322350502014,
      "learning_rate": 0.002,
      "loss": 1.9696218967437744,
      "step": 5004
    },
    {
      "epoch": 0.5006,
      "grad_norm": 0.07460252940654755,
      "learning_rate": 0.002,
      "loss": 2.0490009784698486,
      "step": 5006
    },
    {
      "epoch": 0.5008,
      "grad_norm": 0.07346449792385101,
      "learning_rate": 0.002,
      "loss": 1.898690938949585,
      "step": 5008
    },
    {
      "epoch": 0.501,
      "grad_norm": 0.07465500384569168,
      "learning_rate": 0.002,
      "loss": 1.7941608428955078,
      "step": 5010
    },
    {
      "epoch": 0.5012,
      "grad_norm": 0.07248670607805252,
      "learning_rate": 0.002,
      "loss": 1.9617249965667725,
      "step": 5012
    },
    {
      "epoch": 0.5014,
      "grad_norm": 0.07392016798257828,
      "learning_rate": 0.002,
      "loss": 1.8954261541366577,
      "step": 5014
    },
    {
      "epoch": 0.5016,
      "grad_norm": 0.0829000174999237,
      "learning_rate": 0.002,
      "loss": 1.975329875946045,
      "step": 5016
    },
    {
      "epoch": 0.5018,
      "grad_norm": 0.07955685257911682,
      "learning_rate": 0.002,
      "loss": 1.9159717559814453,
      "step": 5018
    },
    {
      "epoch": 0.502,
      "grad_norm": 0.0793507769703865,
      "learning_rate": 0.002,
      "loss": 1.9395978450775146,
      "step": 5020
    },
    {
      "epoch": 0.5022,
      "grad_norm": 0.07498736679553986,
      "learning_rate": 0.002,
      "loss": 1.7243741750717163,
      "step": 5022
    },
    {
      "epoch": 0.5024,
      "grad_norm": 0.07303804904222488,
      "learning_rate": 0.002,
      "loss": 1.8006255626678467,
      "step": 5024
    },
    {
      "epoch": 0.5026,
      "grad_norm": 0.07791190594434738,
      "learning_rate": 0.002,
      "loss": 1.8934156894683838,
      "step": 5026
    },
    {
      "epoch": 0.5028,
      "grad_norm": 0.0763891190290451,
      "learning_rate": 0.002,
      "loss": 2.0099637508392334,
      "step": 5028
    },
    {
      "epoch": 0.503,
      "grad_norm": 0.07665219902992249,
      "learning_rate": 0.002,
      "loss": 1.920555591583252,
      "step": 5030
    },
    {
      "epoch": 0.5032,
      "grad_norm": 0.07862132787704468,
      "learning_rate": 0.002,
      "loss": 1.806519865989685,
      "step": 5032
    },
    {
      "epoch": 0.5034,
      "grad_norm": 0.07611983269453049,
      "learning_rate": 0.002,
      "loss": 1.9145935773849487,
      "step": 5034
    },
    {
      "epoch": 0.5036,
      "grad_norm": 0.07283962517976761,
      "learning_rate": 0.002,
      "loss": 1.8958399295806885,
      "step": 5036
    },
    {
      "epoch": 0.5038,
      "grad_norm": 0.07288657128810883,
      "learning_rate": 0.002,
      "loss": 1.7204465866088867,
      "step": 5038
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.07579738646745682,
      "learning_rate": 0.002,
      "loss": 1.8043689727783203,
      "step": 5040
    },
    {
      "epoch": 0.5042,
      "grad_norm": 0.08039271086454391,
      "learning_rate": 0.002,
      "loss": 1.826141357421875,
      "step": 5042
    },
    {
      "epoch": 0.5044,
      "grad_norm": 0.077018141746521,
      "learning_rate": 0.002,
      "loss": 1.8760178089141846,
      "step": 5044
    },
    {
      "epoch": 0.5046,
      "grad_norm": 0.07945241779088974,
      "learning_rate": 0.002,
      "loss": 1.959104061126709,
      "step": 5046
    },
    {
      "epoch": 0.5048,
      "grad_norm": 0.07172520458698273,
      "learning_rate": 0.002,
      "loss": 1.8944891691207886,
      "step": 5048
    },
    {
      "epoch": 0.505,
      "grad_norm": 0.07647925615310669,
      "learning_rate": 0.002,
      "loss": 1.8338801860809326,
      "step": 5050
    },
    {
      "epoch": 0.5052,
      "grad_norm": 0.07393758744001389,
      "learning_rate": 0.002,
      "loss": 1.914129376411438,
      "step": 5052
    },
    {
      "epoch": 0.5054,
      "grad_norm": 0.0776066705584526,
      "learning_rate": 0.002,
      "loss": 1.8342747688293457,
      "step": 5054
    },
    {
      "epoch": 0.5056,
      "grad_norm": 0.07847458124160767,
      "learning_rate": 0.002,
      "loss": 2.009162187576294,
      "step": 5056
    },
    {
      "epoch": 0.5058,
      "grad_norm": 0.07723797112703323,
      "learning_rate": 0.002,
      "loss": 1.9371421337127686,
      "step": 5058
    },
    {
      "epoch": 0.506,
      "grad_norm": 0.07898452132940292,
      "learning_rate": 0.002,
      "loss": 1.8751329183578491,
      "step": 5060
    },
    {
      "epoch": 0.5062,
      "grad_norm": 0.07772908359766006,
      "learning_rate": 0.002,
      "loss": 1.9930739402770996,
      "step": 5062
    },
    {
      "epoch": 0.5064,
      "grad_norm": 0.07557539641857147,
      "learning_rate": 0.002,
      "loss": 1.8920923471450806,
      "step": 5064
    },
    {
      "epoch": 0.5066,
      "grad_norm": 0.07549525797367096,
      "learning_rate": 0.002,
      "loss": 1.8945205211639404,
      "step": 5066
    },
    {
      "epoch": 0.5068,
      "grad_norm": 0.07647896558046341,
      "learning_rate": 0.002,
      "loss": 2.0142900943756104,
      "step": 5068
    },
    {
      "epoch": 0.507,
      "grad_norm": 0.0849747434258461,
      "learning_rate": 0.002,
      "loss": 1.870362401008606,
      "step": 5070
    },
    {
      "epoch": 0.5072,
      "grad_norm": 0.0741705447435379,
      "learning_rate": 0.002,
      "loss": 1.8379937410354614,
      "step": 5072
    },
    {
      "epoch": 0.5074,
      "grad_norm": 0.0838652029633522,
      "learning_rate": 0.002,
      "loss": 1.9150141477584839,
      "step": 5074
    },
    {
      "epoch": 0.5076,
      "grad_norm": 0.07777337729930878,
      "learning_rate": 0.002,
      "loss": 1.9697108268737793,
      "step": 5076
    },
    {
      "epoch": 0.5078,
      "grad_norm": 0.08154033124446869,
      "learning_rate": 0.002,
      "loss": 1.7857279777526855,
      "step": 5078
    },
    {
      "epoch": 0.508,
      "grad_norm": 0.08361705392599106,
      "learning_rate": 0.002,
      "loss": 1.9448461532592773,
      "step": 5080
    },
    {
      "epoch": 0.5082,
      "grad_norm": 0.07721496373414993,
      "learning_rate": 0.002,
      "loss": 1.8750019073486328,
      "step": 5082
    },
    {
      "epoch": 0.5084,
      "grad_norm": 0.07685309648513794,
      "learning_rate": 0.002,
      "loss": 1.912903904914856,
      "step": 5084
    },
    {
      "epoch": 0.5086,
      "grad_norm": 0.0816967785358429,
      "learning_rate": 0.002,
      "loss": 1.9582080841064453,
      "step": 5086
    },
    {
      "epoch": 0.5088,
      "grad_norm": 0.0760994553565979,
      "learning_rate": 0.002,
      "loss": 1.8707983493804932,
      "step": 5088
    },
    {
      "epoch": 0.509,
      "grad_norm": 0.08128777146339417,
      "learning_rate": 0.002,
      "loss": 1.8136075735092163,
      "step": 5090
    },
    {
      "epoch": 0.5092,
      "grad_norm": 0.07756774127483368,
      "learning_rate": 0.002,
      "loss": 1.9161361455917358,
      "step": 5092
    },
    {
      "epoch": 0.5094,
      "grad_norm": 0.07670318335294724,
      "learning_rate": 0.002,
      "loss": 1.831620216369629,
      "step": 5094
    },
    {
      "epoch": 0.5096,
      "grad_norm": 0.0776263102889061,
      "learning_rate": 0.002,
      "loss": 1.9522898197174072,
      "step": 5096
    },
    {
      "epoch": 0.5098,
      "grad_norm": 0.0823109820485115,
      "learning_rate": 0.002,
      "loss": 1.9194687604904175,
      "step": 5098
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.07668840140104294,
      "learning_rate": 0.002,
      "loss": 1.8564108610153198,
      "step": 5100
    },
    {
      "epoch": 0.5102,
      "grad_norm": 0.07813489437103271,
      "learning_rate": 0.002,
      "loss": 1.8699147701263428,
      "step": 5102
    },
    {
      "epoch": 0.5104,
      "grad_norm": 0.07608754932880402,
      "learning_rate": 0.002,
      "loss": 1.7895469665527344,
      "step": 5104
    },
    {
      "epoch": 0.5106,
      "grad_norm": 0.07630978524684906,
      "learning_rate": 0.002,
      "loss": 1.8849613666534424,
      "step": 5106
    },
    {
      "epoch": 0.5108,
      "grad_norm": 0.07648218423128128,
      "learning_rate": 0.002,
      "loss": 1.7621816396713257,
      "step": 5108
    },
    {
      "epoch": 0.511,
      "grad_norm": 0.0772683396935463,
      "learning_rate": 0.002,
      "loss": 1.870362639427185,
      "step": 5110
    },
    {
      "epoch": 0.5112,
      "grad_norm": 0.08088277280330658,
      "learning_rate": 0.002,
      "loss": 1.761869192123413,
      "step": 5112
    },
    {
      "epoch": 0.5114,
      "grad_norm": 0.07760428637266159,
      "learning_rate": 0.002,
      "loss": 2.0166001319885254,
      "step": 5114
    },
    {
      "epoch": 0.5116,
      "grad_norm": 0.07449308037757874,
      "learning_rate": 0.002,
      "loss": 1.7532470226287842,
      "step": 5116
    },
    {
      "epoch": 0.5118,
      "grad_norm": 0.08060857653617859,
      "learning_rate": 0.002,
      "loss": 1.9972913265228271,
      "step": 5118
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.08016487210988998,
      "learning_rate": 0.002,
      "loss": 1.8265923261642456,
      "step": 5120
    },
    {
      "epoch": 0.5122,
      "grad_norm": 0.07921528071165085,
      "learning_rate": 0.002,
      "loss": 1.9419389963150024,
      "step": 5122
    },
    {
      "epoch": 0.5124,
      "grad_norm": 0.07685460895299911,
      "learning_rate": 0.002,
      "loss": 1.8314576148986816,
      "step": 5124
    },
    {
      "epoch": 0.5126,
      "grad_norm": 0.08251681178808212,
      "learning_rate": 0.002,
      "loss": 1.9242486953735352,
      "step": 5126
    },
    {
      "epoch": 0.5128,
      "grad_norm": 0.07392717897891998,
      "learning_rate": 0.002,
      "loss": 1.8970091342926025,
      "step": 5128
    },
    {
      "epoch": 0.513,
      "grad_norm": 0.0764135867357254,
      "learning_rate": 0.002,
      "loss": 1.942941665649414,
      "step": 5130
    },
    {
      "epoch": 0.5132,
      "grad_norm": 0.0731266587972641,
      "learning_rate": 0.002,
      "loss": 1.918443202972412,
      "step": 5132
    },
    {
      "epoch": 0.5134,
      "grad_norm": 0.07673751562833786,
      "learning_rate": 0.002,
      "loss": 1.944442868232727,
      "step": 5134
    },
    {
      "epoch": 0.5136,
      "grad_norm": 0.07794837653636932,
      "learning_rate": 0.002,
      "loss": 1.929308533668518,
      "step": 5136
    },
    {
      "epoch": 0.5138,
      "grad_norm": 0.07390327751636505,
      "learning_rate": 0.002,
      "loss": 1.9844017028808594,
      "step": 5138
    },
    {
      "epoch": 0.514,
      "grad_norm": 0.0778968557715416,
      "learning_rate": 0.002,
      "loss": 1.8240013122558594,
      "step": 5140
    },
    {
      "epoch": 0.5142,
      "grad_norm": 0.07498517632484436,
      "learning_rate": 0.002,
      "loss": 1.8871710300445557,
      "step": 5142
    },
    {
      "epoch": 0.5144,
      "grad_norm": 0.07639095932245255,
      "learning_rate": 0.002,
      "loss": 1.8863147497177124,
      "step": 5144
    },
    {
      "epoch": 0.5146,
      "grad_norm": 0.07725038379430771,
      "learning_rate": 0.002,
      "loss": 1.8795701265335083,
      "step": 5146
    },
    {
      "epoch": 0.5148,
      "grad_norm": 0.08082045614719391,
      "learning_rate": 0.002,
      "loss": 1.9137427806854248,
      "step": 5148
    },
    {
      "epoch": 0.515,
      "grad_norm": 0.07402458786964417,
      "learning_rate": 0.002,
      "loss": 1.8900240659713745,
      "step": 5150
    },
    {
      "epoch": 0.5152,
      "grad_norm": 0.0807914063334465,
      "learning_rate": 0.002,
      "loss": 1.8350580930709839,
      "step": 5152
    },
    {
      "epoch": 0.5154,
      "grad_norm": 0.08162681013345718,
      "learning_rate": 0.002,
      "loss": 1.824307918548584,
      "step": 5154
    },
    {
      "epoch": 0.5156,
      "grad_norm": 0.07692142575979233,
      "learning_rate": 0.002,
      "loss": 1.8351041078567505,
      "step": 5156
    },
    {
      "epoch": 0.5158,
      "grad_norm": 0.07873734086751938,
      "learning_rate": 0.002,
      "loss": 1.9198652505874634,
      "step": 5158
    },
    {
      "epoch": 0.516,
      "grad_norm": 0.07547885179519653,
      "learning_rate": 0.002,
      "loss": 1.8699356317520142,
      "step": 5160
    },
    {
      "epoch": 0.5162,
      "grad_norm": 0.07362046837806702,
      "learning_rate": 0.002,
      "loss": 1.9499273300170898,
      "step": 5162
    },
    {
      "epoch": 0.5164,
      "grad_norm": 0.07638122886419296,
      "learning_rate": 0.002,
      "loss": 1.7958338260650635,
      "step": 5164
    },
    {
      "epoch": 0.5166,
      "grad_norm": 0.07879696041345596,
      "learning_rate": 0.002,
      "loss": 1.8312219381332397,
      "step": 5166
    },
    {
      "epoch": 0.5168,
      "grad_norm": 0.07410363852977753,
      "learning_rate": 0.002,
      "loss": 1.8741885423660278,
      "step": 5168
    },
    {
      "epoch": 0.517,
      "grad_norm": 0.08037342131137848,
      "learning_rate": 0.002,
      "loss": 1.8642127513885498,
      "step": 5170
    },
    {
      "epoch": 0.5172,
      "grad_norm": 0.08542365580797195,
      "learning_rate": 0.002,
      "loss": 1.8695824146270752,
      "step": 5172
    },
    {
      "epoch": 0.5174,
      "grad_norm": 0.0754755288362503,
      "learning_rate": 0.002,
      "loss": 1.8518729209899902,
      "step": 5174
    },
    {
      "epoch": 0.5176,
      "grad_norm": 0.07608392834663391,
      "learning_rate": 0.002,
      "loss": 1.8589109182357788,
      "step": 5176
    },
    {
      "epoch": 0.5178,
      "grad_norm": 0.07700195908546448,
      "learning_rate": 0.002,
      "loss": 1.8292295932769775,
      "step": 5178
    },
    {
      "epoch": 0.518,
      "grad_norm": 0.0791751891374588,
      "learning_rate": 0.002,
      "loss": 1.8342182636260986,
      "step": 5180
    },
    {
      "epoch": 0.5182,
      "grad_norm": 0.07704221457242966,
      "learning_rate": 0.002,
      "loss": 1.833548903465271,
      "step": 5182
    },
    {
      "epoch": 0.5184,
      "grad_norm": 0.07493855804204941,
      "learning_rate": 0.002,
      "loss": 1.8084256649017334,
      "step": 5184
    },
    {
      "epoch": 0.5186,
      "grad_norm": 0.07918963581323624,
      "learning_rate": 0.002,
      "loss": 1.898005485534668,
      "step": 5186
    },
    {
      "epoch": 0.5188,
      "grad_norm": 0.07263720035552979,
      "learning_rate": 0.002,
      "loss": 1.8033756017684937,
      "step": 5188
    },
    {
      "epoch": 0.519,
      "grad_norm": 0.08026127517223358,
      "learning_rate": 0.002,
      "loss": 1.9711040258407593,
      "step": 5190
    },
    {
      "epoch": 0.5192,
      "grad_norm": 0.07555105537176132,
      "learning_rate": 0.002,
      "loss": 1.8234866857528687,
      "step": 5192
    },
    {
      "epoch": 0.5194,
      "grad_norm": 0.0735904723405838,
      "learning_rate": 0.002,
      "loss": 1.8204383850097656,
      "step": 5194
    },
    {
      "epoch": 0.5196,
      "grad_norm": 0.07828403264284134,
      "learning_rate": 0.002,
      "loss": 1.9361250400543213,
      "step": 5196
    },
    {
      "epoch": 0.5198,
      "grad_norm": 0.07370290160179138,
      "learning_rate": 0.002,
      "loss": 1.7350815534591675,
      "step": 5198
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.07771139591932297,
      "learning_rate": 0.002,
      "loss": 1.9164432287216187,
      "step": 5200
    },
    {
      "epoch": 0.5202,
      "grad_norm": 0.07952547818422318,
      "learning_rate": 0.002,
      "loss": 1.8448803424835205,
      "step": 5202
    },
    {
      "epoch": 0.5204,
      "grad_norm": 0.07486733049154282,
      "learning_rate": 0.002,
      "loss": 1.8840383291244507,
      "step": 5204
    },
    {
      "epoch": 0.5206,
      "grad_norm": 0.07573020458221436,
      "learning_rate": 0.002,
      "loss": 1.7942256927490234,
      "step": 5206
    },
    {
      "epoch": 0.5208,
      "grad_norm": 0.07550449669361115,
      "learning_rate": 0.002,
      "loss": 2.0034115314483643,
      "step": 5208
    },
    {
      "epoch": 0.521,
      "grad_norm": 0.07674907892942429,
      "learning_rate": 0.002,
      "loss": 1.8732675313949585,
      "step": 5210
    },
    {
      "epoch": 0.5212,
      "grad_norm": 0.07994276285171509,
      "learning_rate": 0.002,
      "loss": 1.8649296760559082,
      "step": 5212
    },
    {
      "epoch": 0.5214,
      "grad_norm": 0.07827724516391754,
      "learning_rate": 0.002,
      "loss": 1.8282057046890259,
      "step": 5214
    },
    {
      "epoch": 0.5216,
      "grad_norm": 0.08453748375177383,
      "learning_rate": 0.002,
      "loss": 1.8458802700042725,
      "step": 5216
    },
    {
      "epoch": 0.5218,
      "grad_norm": 0.07947976887226105,
      "learning_rate": 0.002,
      "loss": 2.0157620906829834,
      "step": 5218
    },
    {
      "epoch": 0.522,
      "grad_norm": 0.0797232985496521,
      "learning_rate": 0.002,
      "loss": 1.8798168897628784,
      "step": 5220
    },
    {
      "epoch": 0.5222,
      "grad_norm": 0.07695663720369339,
      "learning_rate": 0.002,
      "loss": 1.9025025367736816,
      "step": 5222
    },
    {
      "epoch": 0.5224,
      "grad_norm": 0.08021770417690277,
      "learning_rate": 0.002,
      "loss": 1.7922742366790771,
      "step": 5224
    },
    {
      "epoch": 0.5226,
      "grad_norm": 0.07538264989852905,
      "learning_rate": 0.002,
      "loss": 1.9202117919921875,
      "step": 5226
    },
    {
      "epoch": 0.5228,
      "grad_norm": 0.07457560300827026,
      "learning_rate": 0.002,
      "loss": 1.7810386419296265,
      "step": 5228
    },
    {
      "epoch": 0.523,
      "grad_norm": 0.07858188450336456,
      "learning_rate": 0.002,
      "loss": 1.9455301761627197,
      "step": 5230
    },
    {
      "epoch": 0.5232,
      "grad_norm": 0.07546094059944153,
      "learning_rate": 0.002,
      "loss": 1.8648796081542969,
      "step": 5232
    },
    {
      "epoch": 0.5234,
      "grad_norm": 0.07659821212291718,
      "learning_rate": 0.002,
      "loss": 1.8118470907211304,
      "step": 5234
    },
    {
      "epoch": 0.5236,
      "grad_norm": 0.07298414409160614,
      "learning_rate": 0.002,
      "loss": 1.886103868484497,
      "step": 5236
    },
    {
      "epoch": 0.5238,
      "grad_norm": 0.07685582339763641,
      "learning_rate": 0.002,
      "loss": 1.9016804695129395,
      "step": 5238
    },
    {
      "epoch": 0.524,
      "grad_norm": 0.08488738536834717,
      "learning_rate": 0.002,
      "loss": 1.874596118927002,
      "step": 5240
    },
    {
      "epoch": 0.5242,
      "grad_norm": 0.07662788033485413,
      "learning_rate": 0.002,
      "loss": 1.845860481262207,
      "step": 5242
    },
    {
      "epoch": 0.5244,
      "grad_norm": 0.08092677593231201,
      "learning_rate": 0.002,
      "loss": 1.9367355108261108,
      "step": 5244
    },
    {
      "epoch": 0.5246,
      "grad_norm": 0.07751931995153427,
      "learning_rate": 0.002,
      "loss": 1.912753939628601,
      "step": 5246
    },
    {
      "epoch": 0.5248,
      "grad_norm": 0.08290396630764008,
      "learning_rate": 0.002,
      "loss": 1.8307981491088867,
      "step": 5248
    },
    {
      "epoch": 0.525,
      "grad_norm": 0.07485955953598022,
      "learning_rate": 0.002,
      "loss": 2.0869743824005127,
      "step": 5250
    },
    {
      "epoch": 0.5252,
      "grad_norm": 0.07572956383228302,
      "learning_rate": 0.002,
      "loss": 1.8750519752502441,
      "step": 5252
    },
    {
      "epoch": 0.5254,
      "grad_norm": 0.0774465873837471,
      "learning_rate": 0.002,
      "loss": 1.8033535480499268,
      "step": 5254
    },
    {
      "epoch": 0.5256,
      "grad_norm": 0.07423629611730576,
      "learning_rate": 0.002,
      "loss": 1.8835421800613403,
      "step": 5256
    },
    {
      "epoch": 0.5258,
      "grad_norm": 0.07377532869577408,
      "learning_rate": 0.002,
      "loss": 1.8701756000518799,
      "step": 5258
    },
    {
      "epoch": 0.526,
      "grad_norm": 0.0829099491238594,
      "learning_rate": 0.002,
      "loss": 1.8990751504898071,
      "step": 5260
    },
    {
      "epoch": 0.5262,
      "grad_norm": 0.08044442534446716,
      "learning_rate": 0.002,
      "loss": 1.9566335678100586,
      "step": 5262
    },
    {
      "epoch": 0.5264,
      "grad_norm": 0.0733838677406311,
      "learning_rate": 0.002,
      "loss": 1.7449414730072021,
      "step": 5264
    },
    {
      "epoch": 0.5266,
      "grad_norm": 0.07456961274147034,
      "learning_rate": 0.002,
      "loss": 1.8696593046188354,
      "step": 5266
    },
    {
      "epoch": 0.5268,
      "grad_norm": 0.07472024112939835,
      "learning_rate": 0.002,
      "loss": 1.899727463722229,
      "step": 5268
    },
    {
      "epoch": 0.527,
      "grad_norm": 0.08353771269321442,
      "learning_rate": 0.002,
      "loss": 1.942381501197815,
      "step": 5270
    },
    {
      "epoch": 0.5272,
      "grad_norm": 0.07745543867349625,
      "learning_rate": 0.002,
      "loss": 1.7804327011108398,
      "step": 5272
    },
    {
      "epoch": 0.5274,
      "grad_norm": 0.07834707200527191,
      "learning_rate": 0.002,
      "loss": 1.7932460308074951,
      "step": 5274
    },
    {
      "epoch": 0.5276,
      "grad_norm": 0.08128416538238525,
      "learning_rate": 0.002,
      "loss": 1.8386563062667847,
      "step": 5276
    },
    {
      "epoch": 0.5278,
      "grad_norm": 0.0748409703373909,
      "learning_rate": 0.002,
      "loss": 2.0435991287231445,
      "step": 5278
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.07765509188175201,
      "learning_rate": 0.002,
      "loss": 1.9305529594421387,
      "step": 5280
    },
    {
      "epoch": 0.5282,
      "grad_norm": 0.07527963072061539,
      "learning_rate": 0.002,
      "loss": 1.7233651876449585,
      "step": 5282
    },
    {
      "epoch": 0.5284,
      "grad_norm": 0.08035358041524887,
      "learning_rate": 0.002,
      "loss": 1.9181550741195679,
      "step": 5284
    },
    {
      "epoch": 0.5286,
      "grad_norm": 0.07726377993822098,
      "learning_rate": 0.002,
      "loss": 1.9593099355697632,
      "step": 5286
    },
    {
      "epoch": 0.5288,
      "grad_norm": 0.0778605043888092,
      "learning_rate": 0.002,
      "loss": 1.848912000656128,
      "step": 5288
    },
    {
      "epoch": 0.529,
      "grad_norm": 0.07729844748973846,
      "learning_rate": 0.002,
      "loss": 1.873283863067627,
      "step": 5290
    },
    {
      "epoch": 0.5292,
      "grad_norm": 0.07875926792621613,
      "learning_rate": 0.002,
      "loss": 1.814029574394226,
      "step": 5292
    },
    {
      "epoch": 0.5294,
      "grad_norm": 0.07670070976018906,
      "learning_rate": 0.002,
      "loss": 1.9682183265686035,
      "step": 5294
    },
    {
      "epoch": 0.5296,
      "grad_norm": 0.07401247322559357,
      "learning_rate": 0.002,
      "loss": 1.9392893314361572,
      "step": 5296
    },
    {
      "epoch": 0.5298,
      "grad_norm": 0.07485025376081467,
      "learning_rate": 0.002,
      "loss": 1.935969352722168,
      "step": 5298
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.10448147356510162,
      "learning_rate": 0.002,
      "loss": 1.8127135038375854,
      "step": 5300
    },
    {
      "epoch": 0.5302,
      "grad_norm": 0.07613164931535721,
      "learning_rate": 0.002,
      "loss": 1.9390932321548462,
      "step": 5302
    },
    {
      "epoch": 0.5304,
      "grad_norm": 0.07596370577812195,
      "learning_rate": 0.002,
      "loss": 1.8748060464859009,
      "step": 5304
    },
    {
      "epoch": 0.5306,
      "grad_norm": 0.07627006620168686,
      "learning_rate": 0.002,
      "loss": 1.939496397972107,
      "step": 5306
    },
    {
      "epoch": 0.5308,
      "grad_norm": 0.07646030932664871,
      "learning_rate": 0.002,
      "loss": 1.9313780069351196,
      "step": 5308
    },
    {
      "epoch": 0.531,
      "grad_norm": 0.07657410949468613,
      "learning_rate": 0.002,
      "loss": 1.9144728183746338,
      "step": 5310
    },
    {
      "epoch": 0.5312,
      "grad_norm": 0.07746975123882294,
      "learning_rate": 0.002,
      "loss": 1.855944037437439,
      "step": 5312
    },
    {
      "epoch": 0.5314,
      "grad_norm": 0.07417240738868713,
      "learning_rate": 0.002,
      "loss": 1.8833484649658203,
      "step": 5314
    },
    {
      "epoch": 0.5316,
      "grad_norm": 0.07843102514743805,
      "learning_rate": 0.002,
      "loss": 1.8690749406814575,
      "step": 5316
    },
    {
      "epoch": 0.5318,
      "grad_norm": 0.07484722137451172,
      "learning_rate": 0.002,
      "loss": 1.836608648300171,
      "step": 5318
    },
    {
      "epoch": 0.532,
      "grad_norm": 0.07851185649633408,
      "learning_rate": 0.002,
      "loss": 1.9157403707504272,
      "step": 5320
    },
    {
      "epoch": 0.5322,
      "grad_norm": 0.07637521624565125,
      "learning_rate": 0.002,
      "loss": 1.921626329421997,
      "step": 5322
    },
    {
      "epoch": 0.5324,
      "grad_norm": 0.07805584371089935,
      "learning_rate": 0.002,
      "loss": 1.7760878801345825,
      "step": 5324
    },
    {
      "epoch": 0.5326,
      "grad_norm": 0.07690722495317459,
      "learning_rate": 0.002,
      "loss": 1.7294896841049194,
      "step": 5326
    },
    {
      "epoch": 0.5328,
      "grad_norm": 0.08228043466806412,
      "learning_rate": 0.002,
      "loss": 1.8263497352600098,
      "step": 5328
    },
    {
      "epoch": 0.533,
      "grad_norm": 0.07627937942743301,
      "learning_rate": 0.002,
      "loss": 1.8619251251220703,
      "step": 5330
    },
    {
      "epoch": 0.5332,
      "grad_norm": 0.0801718607544899,
      "learning_rate": 0.002,
      "loss": 1.8611221313476562,
      "step": 5332
    },
    {
      "epoch": 0.5334,
      "grad_norm": 0.08051115274429321,
      "learning_rate": 0.002,
      "loss": 1.9104714393615723,
      "step": 5334
    },
    {
      "epoch": 0.5336,
      "grad_norm": 0.075651153922081,
      "learning_rate": 0.002,
      "loss": 1.90333890914917,
      "step": 5336
    },
    {
      "epoch": 0.5338,
      "grad_norm": 0.07489682734012604,
      "learning_rate": 0.002,
      "loss": 1.8632383346557617,
      "step": 5338
    },
    {
      "epoch": 0.534,
      "grad_norm": 0.07908770442008972,
      "learning_rate": 0.002,
      "loss": 1.82367742061615,
      "step": 5340
    },
    {
      "epoch": 0.5342,
      "grad_norm": 0.08004000782966614,
      "learning_rate": 0.002,
      "loss": 1.879951000213623,
      "step": 5342
    },
    {
      "epoch": 0.5344,
      "grad_norm": 0.07676207274198532,
      "learning_rate": 0.002,
      "loss": 1.9231170415878296,
      "step": 5344
    },
    {
      "epoch": 0.5346,
      "grad_norm": 0.0831858366727829,
      "learning_rate": 0.002,
      "loss": 1.7388079166412354,
      "step": 5346
    },
    {
      "epoch": 0.5348,
      "grad_norm": 0.08434810489416122,
      "learning_rate": 0.002,
      "loss": 1.9308407306671143,
      "step": 5348
    },
    {
      "epoch": 0.535,
      "grad_norm": 0.07718940079212189,
      "learning_rate": 0.002,
      "loss": 1.855771780014038,
      "step": 5350
    },
    {
      "epoch": 0.5352,
      "grad_norm": 0.07986968755722046,
      "learning_rate": 0.002,
      "loss": 1.9391182661056519,
      "step": 5352
    },
    {
      "epoch": 0.5354,
      "grad_norm": 0.07567508518695831,
      "learning_rate": 0.002,
      "loss": 1.9564179182052612,
      "step": 5354
    },
    {
      "epoch": 0.5356,
      "grad_norm": 0.07904330641031265,
      "learning_rate": 0.002,
      "loss": 1.789055347442627,
      "step": 5356
    },
    {
      "epoch": 0.5358,
      "grad_norm": 0.07765140384435654,
      "learning_rate": 0.002,
      "loss": 1.8490198850631714,
      "step": 5358
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.07637099176645279,
      "learning_rate": 0.002,
      "loss": 1.9757835865020752,
      "step": 5360
    },
    {
      "epoch": 0.5362,
      "grad_norm": 0.07582999020814896,
      "learning_rate": 0.002,
      "loss": 1.96444833278656,
      "step": 5362
    },
    {
      "epoch": 0.5364,
      "grad_norm": 0.07483256608247757,
      "learning_rate": 0.002,
      "loss": 1.8726392984390259,
      "step": 5364
    },
    {
      "epoch": 0.5366,
      "grad_norm": 0.07780462503433228,
      "learning_rate": 0.002,
      "loss": 1.9561338424682617,
      "step": 5366
    },
    {
      "epoch": 0.5368,
      "grad_norm": 0.0807398334145546,
      "learning_rate": 0.002,
      "loss": 1.878947377204895,
      "step": 5368
    },
    {
      "epoch": 0.537,
      "grad_norm": 0.07602346688508987,
      "learning_rate": 0.002,
      "loss": 1.86123788356781,
      "step": 5370
    },
    {
      "epoch": 0.5372,
      "grad_norm": 0.07932428270578384,
      "learning_rate": 0.002,
      "loss": 1.8985503911972046,
      "step": 5372
    },
    {
      "epoch": 0.5374,
      "grad_norm": 0.07526547461748123,
      "learning_rate": 0.002,
      "loss": 1.798912763595581,
      "step": 5374
    },
    {
      "epoch": 0.5376,
      "grad_norm": 0.07971195131540298,
      "learning_rate": 0.002,
      "loss": 1.923749566078186,
      "step": 5376
    },
    {
      "epoch": 0.5378,
      "grad_norm": 0.09021654725074768,
      "learning_rate": 0.002,
      "loss": 1.872841238975525,
      "step": 5378
    },
    {
      "epoch": 0.538,
      "grad_norm": 0.08231820911169052,
      "learning_rate": 0.002,
      "loss": 1.7932724952697754,
      "step": 5380
    },
    {
      "epoch": 0.5382,
      "grad_norm": 0.07551641017198563,
      "learning_rate": 0.002,
      "loss": 1.76743745803833,
      "step": 5382
    },
    {
      "epoch": 0.5384,
      "grad_norm": 0.07362394034862518,
      "learning_rate": 0.002,
      "loss": 1.775719165802002,
      "step": 5384
    },
    {
      "epoch": 0.5386,
      "grad_norm": 0.07808487862348557,
      "learning_rate": 0.002,
      "loss": 1.8641765117645264,
      "step": 5386
    },
    {
      "epoch": 0.5388,
      "grad_norm": 0.07806996256113052,
      "learning_rate": 0.002,
      "loss": 1.922606110572815,
      "step": 5388
    },
    {
      "epoch": 0.539,
      "grad_norm": 0.07596611976623535,
      "learning_rate": 0.002,
      "loss": 1.8929851055145264,
      "step": 5390
    },
    {
      "epoch": 0.5392,
      "grad_norm": 0.07640298455953598,
      "learning_rate": 0.002,
      "loss": 1.982085108757019,
      "step": 5392
    },
    {
      "epoch": 0.5394,
      "grad_norm": 0.0767398253083229,
      "learning_rate": 0.002,
      "loss": 1.7832551002502441,
      "step": 5394
    },
    {
      "epoch": 0.5396,
      "grad_norm": 0.07064790278673172,
      "learning_rate": 0.002,
      "loss": 1.8212254047393799,
      "step": 5396
    },
    {
      "epoch": 0.5398,
      "grad_norm": 0.07621841877698898,
      "learning_rate": 0.002,
      "loss": 1.8636693954467773,
      "step": 5398
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.07888048142194748,
      "learning_rate": 0.002,
      "loss": 1.8853226900100708,
      "step": 5400
    },
    {
      "epoch": 0.5402,
      "grad_norm": 0.07347539812326431,
      "learning_rate": 0.002,
      "loss": 1.6915736198425293,
      "step": 5402
    },
    {
      "epoch": 0.5404,
      "grad_norm": 0.07866611331701279,
      "learning_rate": 0.002,
      "loss": 1.9924774169921875,
      "step": 5404
    },
    {
      "epoch": 0.5406,
      "grad_norm": 0.07315598428249359,
      "learning_rate": 0.002,
      "loss": 1.866955041885376,
      "step": 5406
    },
    {
      "epoch": 0.5408,
      "grad_norm": 0.08214186877012253,
      "learning_rate": 0.002,
      "loss": 1.8730112314224243,
      "step": 5408
    },
    {
      "epoch": 0.541,
      "grad_norm": 0.07667940109968185,
      "learning_rate": 0.002,
      "loss": 2.0231688022613525,
      "step": 5410
    },
    {
      "epoch": 0.5412,
      "grad_norm": 0.08105332404375076,
      "learning_rate": 0.002,
      "loss": 1.901677131652832,
      "step": 5412
    },
    {
      "epoch": 0.5414,
      "grad_norm": 0.07824359089136124,
      "learning_rate": 0.002,
      "loss": 2.0046603679656982,
      "step": 5414
    },
    {
      "epoch": 0.5416,
      "grad_norm": 0.08314904570579529,
      "learning_rate": 0.002,
      "loss": 1.9073660373687744,
      "step": 5416
    },
    {
      "epoch": 0.5418,
      "grad_norm": 0.07456666231155396,
      "learning_rate": 0.002,
      "loss": 1.8555278778076172,
      "step": 5418
    },
    {
      "epoch": 0.542,
      "grad_norm": 0.08020937442779541,
      "learning_rate": 0.002,
      "loss": 1.9359958171844482,
      "step": 5420
    },
    {
      "epoch": 0.5422,
      "grad_norm": 0.07463578879833221,
      "learning_rate": 0.002,
      "loss": 1.7625683546066284,
      "step": 5422
    },
    {
      "epoch": 0.5424,
      "grad_norm": 0.07155022770166397,
      "learning_rate": 0.002,
      "loss": 1.786858320236206,
      "step": 5424
    },
    {
      "epoch": 0.5426,
      "grad_norm": 0.07765116542577744,
      "learning_rate": 0.002,
      "loss": 1.8851401805877686,
      "step": 5426
    },
    {
      "epoch": 0.5428,
      "grad_norm": 0.07986035197973251,
      "learning_rate": 0.002,
      "loss": 1.9497251510620117,
      "step": 5428
    },
    {
      "epoch": 0.543,
      "grad_norm": 0.07369924336671829,
      "learning_rate": 0.002,
      "loss": 1.9412814378738403,
      "step": 5430
    },
    {
      "epoch": 0.5432,
      "grad_norm": 0.07479347288608551,
      "learning_rate": 0.002,
      "loss": 1.835795521736145,
      "step": 5432
    },
    {
      "epoch": 0.5434,
      "grad_norm": 0.07460984587669373,
      "learning_rate": 0.002,
      "loss": 1.7993561029434204,
      "step": 5434
    },
    {
      "epoch": 0.5436,
      "grad_norm": 0.07706576585769653,
      "learning_rate": 0.002,
      "loss": 1.8326581716537476,
      "step": 5436
    },
    {
      "epoch": 0.5438,
      "grad_norm": 0.08218351006507874,
      "learning_rate": 0.002,
      "loss": 1.9171096086502075,
      "step": 5438
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.07827456295490265,
      "learning_rate": 0.002,
      "loss": 1.992058277130127,
      "step": 5440
    },
    {
      "epoch": 0.5442,
      "grad_norm": 0.0753851905465126,
      "learning_rate": 0.002,
      "loss": 1.8773654699325562,
      "step": 5442
    },
    {
      "epoch": 0.5444,
      "grad_norm": 0.0743233859539032,
      "learning_rate": 0.002,
      "loss": 1.8932212591171265,
      "step": 5444
    },
    {
      "epoch": 0.5446,
      "grad_norm": 0.07445847243070602,
      "learning_rate": 0.002,
      "loss": 1.7764934301376343,
      "step": 5446
    },
    {
      "epoch": 0.5448,
      "grad_norm": 0.07824169844388962,
      "learning_rate": 0.002,
      "loss": 1.815353274345398,
      "step": 5448
    },
    {
      "epoch": 0.545,
      "grad_norm": 0.0766414925456047,
      "learning_rate": 0.002,
      "loss": 1.7908250093460083,
      "step": 5450
    },
    {
      "epoch": 0.5452,
      "grad_norm": 0.07343129068613052,
      "learning_rate": 0.002,
      "loss": 1.8441188335418701,
      "step": 5452
    },
    {
      "epoch": 0.5454,
      "grad_norm": 0.07367821782827377,
      "learning_rate": 0.002,
      "loss": 1.869140625,
      "step": 5454
    },
    {
      "epoch": 0.5456,
      "grad_norm": 0.0762803703546524,
      "learning_rate": 0.002,
      "loss": 1.8323713541030884,
      "step": 5456
    },
    {
      "epoch": 0.5458,
      "grad_norm": 0.07583177834749222,
      "learning_rate": 0.002,
      "loss": 1.8390858173370361,
      "step": 5458
    },
    {
      "epoch": 0.546,
      "grad_norm": 0.08275330811738968,
      "learning_rate": 0.002,
      "loss": 1.875418782234192,
      "step": 5460
    },
    {
      "epoch": 0.5462,
      "grad_norm": 0.07593163102865219,
      "learning_rate": 0.002,
      "loss": 1.905677318572998,
      "step": 5462
    },
    {
      "epoch": 0.5464,
      "grad_norm": 0.08709431439638138,
      "learning_rate": 0.002,
      "loss": 1.9181008338928223,
      "step": 5464
    },
    {
      "epoch": 0.5466,
      "grad_norm": 0.07647739350795746,
      "learning_rate": 0.002,
      "loss": 1.8474078178405762,
      "step": 5466
    },
    {
      "epoch": 0.5468,
      "grad_norm": 0.08807466924190521,
      "learning_rate": 0.002,
      "loss": 1.8557348251342773,
      "step": 5468
    },
    {
      "epoch": 0.547,
      "grad_norm": 0.07635874301195145,
      "learning_rate": 0.002,
      "loss": 1.8481892347335815,
      "step": 5470
    },
    {
      "epoch": 0.5472,
      "grad_norm": 0.07559250295162201,
      "learning_rate": 0.002,
      "loss": 1.964597463607788,
      "step": 5472
    },
    {
      "epoch": 0.5474,
      "grad_norm": 0.07722021639347076,
      "learning_rate": 0.002,
      "loss": 1.925666332244873,
      "step": 5474
    },
    {
      "epoch": 0.5476,
      "grad_norm": 0.07510940730571747,
      "learning_rate": 0.002,
      "loss": 1.8722916841506958,
      "step": 5476
    },
    {
      "epoch": 0.5478,
      "grad_norm": 0.07499680668115616,
      "learning_rate": 0.002,
      "loss": 1.8303123712539673,
      "step": 5478
    },
    {
      "epoch": 0.548,
      "grad_norm": 0.07761944085359573,
      "learning_rate": 0.002,
      "loss": 1.97859787940979,
      "step": 5480
    },
    {
      "epoch": 0.5482,
      "grad_norm": 0.0775679349899292,
      "learning_rate": 0.002,
      "loss": 2.004718542098999,
      "step": 5482
    },
    {
      "epoch": 0.5484,
      "grad_norm": 0.07681342959403992,
      "learning_rate": 0.002,
      "loss": 2.0413620471954346,
      "step": 5484
    },
    {
      "epoch": 0.5486,
      "grad_norm": 0.07737144827842712,
      "learning_rate": 0.002,
      "loss": 1.6867049932479858,
      "step": 5486
    },
    {
      "epoch": 0.5488,
      "grad_norm": 0.07797475159168243,
      "learning_rate": 0.002,
      "loss": 1.8933348655700684,
      "step": 5488
    },
    {
      "epoch": 0.549,
      "grad_norm": 0.0750206932425499,
      "learning_rate": 0.002,
      "loss": 1.946869969367981,
      "step": 5490
    },
    {
      "epoch": 0.5492,
      "grad_norm": 0.07588593661785126,
      "learning_rate": 0.002,
      "loss": 2.0188915729522705,
      "step": 5492
    },
    {
      "epoch": 0.5494,
      "grad_norm": 0.07771960645914078,
      "learning_rate": 0.002,
      "loss": 1.9159858226776123,
      "step": 5494
    },
    {
      "epoch": 0.5496,
      "grad_norm": 0.07969757914543152,
      "learning_rate": 0.002,
      "loss": 1.9361910820007324,
      "step": 5496
    },
    {
      "epoch": 0.5498,
      "grad_norm": 0.07622114568948746,
      "learning_rate": 0.002,
      "loss": 1.9306578636169434,
      "step": 5498
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.08019876480102539,
      "learning_rate": 0.002,
      "loss": 1.869563102722168,
      "step": 5500
    },
    {
      "epoch": 0.5502,
      "grad_norm": 0.07412100583314896,
      "learning_rate": 0.002,
      "loss": 1.8142645359039307,
      "step": 5502
    },
    {
      "epoch": 0.5504,
      "grad_norm": 0.08425413072109222,
      "learning_rate": 0.002,
      "loss": 1.9233673810958862,
      "step": 5504
    },
    {
      "epoch": 0.5506,
      "grad_norm": 0.07430815696716309,
      "learning_rate": 0.002,
      "loss": 2.0038070678710938,
      "step": 5506
    },
    {
      "epoch": 0.5508,
      "grad_norm": 0.07671178877353668,
      "learning_rate": 0.002,
      "loss": 1.9786094427108765,
      "step": 5508
    },
    {
      "epoch": 0.551,
      "grad_norm": 0.07414177060127258,
      "learning_rate": 0.002,
      "loss": 1.9453308582305908,
      "step": 5510
    },
    {
      "epoch": 0.5512,
      "grad_norm": 0.07374726980924606,
      "learning_rate": 0.002,
      "loss": 2.019514560699463,
      "step": 5512
    },
    {
      "epoch": 0.5514,
      "grad_norm": 0.07391533255577087,
      "learning_rate": 0.002,
      "loss": 1.9087951183319092,
      "step": 5514
    },
    {
      "epoch": 0.5516,
      "grad_norm": 0.07460352778434753,
      "learning_rate": 0.002,
      "loss": 1.8544456958770752,
      "step": 5516
    },
    {
      "epoch": 0.5518,
      "grad_norm": 0.08074525743722916,
      "learning_rate": 0.002,
      "loss": 2.010685682296753,
      "step": 5518
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.07810112088918686,
      "learning_rate": 0.002,
      "loss": 1.8811429738998413,
      "step": 5520
    },
    {
      "epoch": 0.5522,
      "grad_norm": 0.07522027939558029,
      "learning_rate": 0.002,
      "loss": 1.895832896232605,
      "step": 5522
    },
    {
      "epoch": 0.5524,
      "grad_norm": 0.07503741979598999,
      "learning_rate": 0.002,
      "loss": 2.0145914554595947,
      "step": 5524
    },
    {
      "epoch": 0.5526,
      "grad_norm": 0.07838806509971619,
      "learning_rate": 0.002,
      "loss": 1.9971139430999756,
      "step": 5526
    },
    {
      "epoch": 0.5528,
      "grad_norm": 0.07469892501831055,
      "learning_rate": 0.002,
      "loss": 1.9175406694412231,
      "step": 5528
    },
    {
      "epoch": 0.553,
      "grad_norm": 0.07661428302526474,
      "learning_rate": 0.002,
      "loss": 1.8692083358764648,
      "step": 5530
    },
    {
      "epoch": 0.5532,
      "grad_norm": 0.07836861908435822,
      "learning_rate": 0.002,
      "loss": 1.9861955642700195,
      "step": 5532
    },
    {
      "epoch": 0.5534,
      "grad_norm": 0.0793938860297203,
      "learning_rate": 0.002,
      "loss": 1.80211341381073,
      "step": 5534
    },
    {
      "epoch": 0.5536,
      "grad_norm": 0.07667416334152222,
      "learning_rate": 0.002,
      "loss": 1.950399398803711,
      "step": 5536
    },
    {
      "epoch": 0.5538,
      "grad_norm": 0.07593931257724762,
      "learning_rate": 0.002,
      "loss": 1.8796535730361938,
      "step": 5538
    },
    {
      "epoch": 0.554,
      "grad_norm": 0.07545366883277893,
      "learning_rate": 0.002,
      "loss": 2.025834798812866,
      "step": 5540
    },
    {
      "epoch": 0.5542,
      "grad_norm": 0.08704961091279984,
      "learning_rate": 0.002,
      "loss": 1.9322092533111572,
      "step": 5542
    },
    {
      "epoch": 0.5544,
      "grad_norm": 0.074212446808815,
      "learning_rate": 0.002,
      "loss": 1.9569259881973267,
      "step": 5544
    },
    {
      "epoch": 0.5546,
      "grad_norm": 0.07910770922899246,
      "learning_rate": 0.002,
      "loss": 1.9224964380264282,
      "step": 5546
    },
    {
      "epoch": 0.5548,
      "grad_norm": 0.08258166909217834,
      "learning_rate": 0.002,
      "loss": 1.8851237297058105,
      "step": 5548
    },
    {
      "epoch": 0.555,
      "grad_norm": 0.07724165916442871,
      "learning_rate": 0.002,
      "loss": 1.8085724115371704,
      "step": 5550
    },
    {
      "epoch": 0.5552,
      "grad_norm": 0.08023381233215332,
      "learning_rate": 0.002,
      "loss": 2.009922742843628,
      "step": 5552
    },
    {
      "epoch": 0.5554,
      "grad_norm": 0.07663775980472565,
      "learning_rate": 0.002,
      "loss": 1.7875182628631592,
      "step": 5554
    },
    {
      "epoch": 0.5556,
      "grad_norm": 0.07809925824403763,
      "learning_rate": 0.002,
      "loss": 1.9483747482299805,
      "step": 5556
    },
    {
      "epoch": 0.5558,
      "grad_norm": 0.08806604146957397,
      "learning_rate": 0.002,
      "loss": 1.991146206855774,
      "step": 5558
    },
    {
      "epoch": 0.556,
      "grad_norm": 0.07655586302280426,
      "learning_rate": 0.002,
      "loss": 1.816723346710205,
      "step": 5560
    },
    {
      "epoch": 0.5562,
      "grad_norm": 0.07416453957557678,
      "learning_rate": 0.002,
      "loss": 1.7992420196533203,
      "step": 5562
    },
    {
      "epoch": 0.5564,
      "grad_norm": 0.07335084676742554,
      "learning_rate": 0.002,
      "loss": 1.9189651012420654,
      "step": 5564
    },
    {
      "epoch": 0.5566,
      "grad_norm": 0.07715223729610443,
      "learning_rate": 0.002,
      "loss": 1.9071760177612305,
      "step": 5566
    },
    {
      "epoch": 0.5568,
      "grad_norm": 0.08032162487506866,
      "learning_rate": 0.002,
      "loss": 1.9436423778533936,
      "step": 5568
    },
    {
      "epoch": 0.557,
      "grad_norm": 0.0757770910859108,
      "learning_rate": 0.002,
      "loss": 1.9344741106033325,
      "step": 5570
    },
    {
      "epoch": 0.5572,
      "grad_norm": 0.07703933119773865,
      "learning_rate": 0.002,
      "loss": 1.9102532863616943,
      "step": 5572
    },
    {
      "epoch": 0.5574,
      "grad_norm": 0.08383007347583771,
      "learning_rate": 0.002,
      "loss": 1.956328272819519,
      "step": 5574
    },
    {
      "epoch": 0.5576,
      "grad_norm": 0.07873156666755676,
      "learning_rate": 0.002,
      "loss": 1.9972081184387207,
      "step": 5576
    },
    {
      "epoch": 0.5578,
      "grad_norm": 0.08129816502332687,
      "learning_rate": 0.002,
      "loss": 1.7948200702667236,
      "step": 5578
    },
    {
      "epoch": 0.558,
      "grad_norm": 0.08651072531938553,
      "learning_rate": 0.002,
      "loss": 1.989911437034607,
      "step": 5580
    },
    {
      "epoch": 0.5582,
      "grad_norm": 0.07929384708404541,
      "learning_rate": 0.002,
      "loss": 1.8918012380599976,
      "step": 5582
    },
    {
      "epoch": 0.5584,
      "grad_norm": 0.07567258179187775,
      "learning_rate": 0.002,
      "loss": 1.9725462198257446,
      "step": 5584
    },
    {
      "epoch": 0.5586,
      "grad_norm": 0.0767667293548584,
      "learning_rate": 0.002,
      "loss": 1.8949285745620728,
      "step": 5586
    },
    {
      "epoch": 0.5588,
      "grad_norm": 0.07697207480669022,
      "learning_rate": 0.002,
      "loss": 1.9352049827575684,
      "step": 5588
    },
    {
      "epoch": 0.559,
      "grad_norm": 0.07601137459278107,
      "learning_rate": 0.002,
      "loss": 1.842820167541504,
      "step": 5590
    },
    {
      "epoch": 0.5592,
      "grad_norm": 0.08094388991594315,
      "learning_rate": 0.002,
      "loss": 1.9148374795913696,
      "step": 5592
    },
    {
      "epoch": 0.5594,
      "grad_norm": 0.08334986120462418,
      "learning_rate": 0.002,
      "loss": 1.9037617444992065,
      "step": 5594
    },
    {
      "epoch": 0.5596,
      "grad_norm": 0.08717408776283264,
      "learning_rate": 0.002,
      "loss": 1.9563738107681274,
      "step": 5596
    },
    {
      "epoch": 0.5598,
      "grad_norm": 0.07888885587453842,
      "learning_rate": 0.002,
      "loss": 1.988650918006897,
      "step": 5598
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.07781656831502914,
      "learning_rate": 0.002,
      "loss": 1.9284470081329346,
      "step": 5600
    },
    {
      "epoch": 0.5602,
      "grad_norm": 0.08003741502761841,
      "learning_rate": 0.002,
      "loss": 1.9064650535583496,
      "step": 5602
    },
    {
      "epoch": 0.5604,
      "grad_norm": 0.07881873100996017,
      "learning_rate": 0.002,
      "loss": 2.0596139430999756,
      "step": 5604
    },
    {
      "epoch": 0.5606,
      "grad_norm": 0.07505381852388382,
      "learning_rate": 0.002,
      "loss": 1.87384033203125,
      "step": 5606
    },
    {
      "epoch": 0.5608,
      "grad_norm": 0.07418455928564072,
      "learning_rate": 0.002,
      "loss": 2.0352988243103027,
      "step": 5608
    },
    {
      "epoch": 0.561,
      "grad_norm": 0.07939650118350983,
      "learning_rate": 0.002,
      "loss": 1.8930234909057617,
      "step": 5610
    },
    {
      "epoch": 0.5612,
      "grad_norm": 0.07495488971471786,
      "learning_rate": 0.002,
      "loss": 1.9713009595870972,
      "step": 5612
    },
    {
      "epoch": 0.5614,
      "grad_norm": 0.0762350782752037,
      "learning_rate": 0.002,
      "loss": 1.8231005668640137,
      "step": 5614
    },
    {
      "epoch": 0.5616,
      "grad_norm": 0.07656882703304291,
      "learning_rate": 0.002,
      "loss": 1.9346281290054321,
      "step": 5616
    },
    {
      "epoch": 0.5618,
      "grad_norm": 0.07837933301925659,
      "learning_rate": 0.002,
      "loss": 1.9184848070144653,
      "step": 5618
    },
    {
      "epoch": 0.562,
      "grad_norm": 0.07556180655956268,
      "learning_rate": 0.002,
      "loss": 1.893092155456543,
      "step": 5620
    },
    {
      "epoch": 0.5622,
      "grad_norm": 0.08093723654747009,
      "learning_rate": 0.002,
      "loss": 1.9709957838058472,
      "step": 5622
    },
    {
      "epoch": 0.5624,
      "grad_norm": 0.07930196076631546,
      "learning_rate": 0.002,
      "loss": 1.798247218132019,
      "step": 5624
    },
    {
      "epoch": 0.5626,
      "grad_norm": 0.07671930640935898,
      "learning_rate": 0.002,
      "loss": 1.978737473487854,
      "step": 5626
    },
    {
      "epoch": 0.5628,
      "grad_norm": 0.08000637590885162,
      "learning_rate": 0.002,
      "loss": 1.9157158136367798,
      "step": 5628
    },
    {
      "epoch": 0.563,
      "grad_norm": 0.08384983986616135,
      "learning_rate": 0.002,
      "loss": 1.90665602684021,
      "step": 5630
    },
    {
      "epoch": 0.5632,
      "grad_norm": 0.07552701979875565,
      "learning_rate": 0.002,
      "loss": 1.814539909362793,
      "step": 5632
    },
    {
      "epoch": 0.5634,
      "grad_norm": 0.07923998683691025,
      "learning_rate": 0.002,
      "loss": 1.914002776145935,
      "step": 5634
    },
    {
      "epoch": 0.5636,
      "grad_norm": 0.080339215695858,
      "learning_rate": 0.002,
      "loss": 2.067397356033325,
      "step": 5636
    },
    {
      "epoch": 0.5638,
      "grad_norm": 0.07846985012292862,
      "learning_rate": 0.002,
      "loss": 1.8834997415542603,
      "step": 5638
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.07526293396949768,
      "learning_rate": 0.002,
      "loss": 1.9654651880264282,
      "step": 5640
    },
    {
      "epoch": 0.5642,
      "grad_norm": 0.07541612535715103,
      "learning_rate": 0.002,
      "loss": 1.9247241020202637,
      "step": 5642
    },
    {
      "epoch": 0.5644,
      "grad_norm": 0.0773259699344635,
      "learning_rate": 0.002,
      "loss": 1.863250494003296,
      "step": 5644
    },
    {
      "epoch": 0.5646,
      "grad_norm": 0.08372116088867188,
      "learning_rate": 0.002,
      "loss": 2.079369068145752,
      "step": 5646
    },
    {
      "epoch": 0.5648,
      "grad_norm": 0.07644826173782349,
      "learning_rate": 0.002,
      "loss": 2.064924716949463,
      "step": 5648
    },
    {
      "epoch": 0.565,
      "grad_norm": 0.07839616388082504,
      "learning_rate": 0.002,
      "loss": 1.9017109870910645,
      "step": 5650
    },
    {
      "epoch": 0.5652,
      "grad_norm": 0.08159967511892319,
      "learning_rate": 0.002,
      "loss": 1.888301968574524,
      "step": 5652
    },
    {
      "epoch": 0.5654,
      "grad_norm": 0.07778169959783554,
      "learning_rate": 0.002,
      "loss": 1.9601709842681885,
      "step": 5654
    },
    {
      "epoch": 0.5656,
      "grad_norm": 0.07816203683614731,
      "learning_rate": 0.002,
      "loss": 1.940409541130066,
      "step": 5656
    },
    {
      "epoch": 0.5658,
      "grad_norm": 0.07862038910388947,
      "learning_rate": 0.002,
      "loss": 1.9089500904083252,
      "step": 5658
    },
    {
      "epoch": 0.566,
      "grad_norm": 0.08293874561786652,
      "learning_rate": 0.002,
      "loss": 1.9872462749481201,
      "step": 5660
    },
    {
      "epoch": 0.5662,
      "grad_norm": 0.07953059673309326,
      "learning_rate": 0.002,
      "loss": 1.9550377130508423,
      "step": 5662
    },
    {
      "epoch": 0.5664,
      "grad_norm": 0.07458725571632385,
      "learning_rate": 0.002,
      "loss": 1.8831186294555664,
      "step": 5664
    },
    {
      "epoch": 0.5666,
      "grad_norm": 0.07476375997066498,
      "learning_rate": 0.002,
      "loss": 1.9715633392333984,
      "step": 5666
    },
    {
      "epoch": 0.5668,
      "grad_norm": 0.07885619252920151,
      "learning_rate": 0.002,
      "loss": 1.8866256475448608,
      "step": 5668
    },
    {
      "epoch": 0.567,
      "grad_norm": 0.08068262040615082,
      "learning_rate": 0.002,
      "loss": 1.9630290269851685,
      "step": 5670
    },
    {
      "epoch": 0.5672,
      "grad_norm": 0.07429569214582443,
      "learning_rate": 0.002,
      "loss": 1.8952049016952515,
      "step": 5672
    },
    {
      "epoch": 0.5674,
      "grad_norm": 0.07647793740034103,
      "learning_rate": 0.002,
      "loss": 1.997688889503479,
      "step": 5674
    },
    {
      "epoch": 0.5676,
      "grad_norm": 0.07835842669010162,
      "learning_rate": 0.002,
      "loss": 2.0263519287109375,
      "step": 5676
    },
    {
      "epoch": 0.5678,
      "grad_norm": 0.07568638771772385,
      "learning_rate": 0.002,
      "loss": 1.8746857643127441,
      "step": 5678
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.07755614072084427,
      "learning_rate": 0.002,
      "loss": 1.8997060060501099,
      "step": 5680
    },
    {
      "epoch": 0.5682,
      "grad_norm": 0.07535624504089355,
      "learning_rate": 0.002,
      "loss": 1.912558913230896,
      "step": 5682
    },
    {
      "epoch": 0.5684,
      "grad_norm": 0.07714831084012985,
      "learning_rate": 0.002,
      "loss": 2.0615265369415283,
      "step": 5684
    },
    {
      "epoch": 0.5686,
      "grad_norm": 0.0751330703496933,
      "learning_rate": 0.002,
      "loss": 1.8673163652420044,
      "step": 5686
    },
    {
      "epoch": 0.5688,
      "grad_norm": 0.07960078865289688,
      "learning_rate": 0.002,
      "loss": 1.924634337425232,
      "step": 5688
    },
    {
      "epoch": 0.569,
      "grad_norm": 0.07454539835453033,
      "learning_rate": 0.002,
      "loss": 1.962032437324524,
      "step": 5690
    },
    {
      "epoch": 0.5692,
      "grad_norm": 0.07394523918628693,
      "learning_rate": 0.002,
      "loss": 1.9047205448150635,
      "step": 5692
    },
    {
      "epoch": 0.5694,
      "grad_norm": 0.07806015014648438,
      "learning_rate": 0.002,
      "loss": 1.9510114192962646,
      "step": 5694
    },
    {
      "epoch": 0.5696,
      "grad_norm": 0.07229483872652054,
      "learning_rate": 0.002,
      "loss": 1.8815155029296875,
      "step": 5696
    },
    {
      "epoch": 0.5698,
      "grad_norm": 0.07337135821580887,
      "learning_rate": 0.002,
      "loss": 1.8452109098434448,
      "step": 5698
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.07223459333181381,
      "learning_rate": 0.002,
      "loss": 1.975262999534607,
      "step": 5700
    },
    {
      "epoch": 0.5702,
      "grad_norm": 0.07533871382474899,
      "learning_rate": 0.002,
      "loss": 1.9807052612304688,
      "step": 5702
    },
    {
      "epoch": 0.5704,
      "grad_norm": 0.0782478079199791,
      "learning_rate": 0.002,
      "loss": 1.9637356996536255,
      "step": 5704
    },
    {
      "epoch": 0.5706,
      "grad_norm": 0.07759701460599899,
      "learning_rate": 0.002,
      "loss": 1.8976234197616577,
      "step": 5706
    },
    {
      "epoch": 0.5708,
      "grad_norm": 0.07608474045991898,
      "learning_rate": 0.002,
      "loss": 2.0019564628601074,
      "step": 5708
    },
    {
      "epoch": 0.571,
      "grad_norm": 0.0767127275466919,
      "learning_rate": 0.002,
      "loss": 1.8958090543746948,
      "step": 5710
    },
    {
      "epoch": 0.5712,
      "grad_norm": 0.07655338943004608,
      "learning_rate": 0.002,
      "loss": 1.7894985675811768,
      "step": 5712
    },
    {
      "epoch": 0.5714,
      "grad_norm": 0.07453480362892151,
      "learning_rate": 0.002,
      "loss": 1.934304118156433,
      "step": 5714
    },
    {
      "epoch": 0.5716,
      "grad_norm": 0.07398398220539093,
      "learning_rate": 0.002,
      "loss": 1.9268661737442017,
      "step": 5716
    },
    {
      "epoch": 0.5718,
      "grad_norm": 0.08108047395944595,
      "learning_rate": 0.002,
      "loss": 1.970738172531128,
      "step": 5718
    },
    {
      "epoch": 0.572,
      "grad_norm": 0.07666958123445511,
      "learning_rate": 0.002,
      "loss": 1.9643765687942505,
      "step": 5720
    },
    {
      "epoch": 0.5722,
      "grad_norm": 0.07620645314455032,
      "learning_rate": 0.002,
      "loss": 1.9512121677398682,
      "step": 5722
    },
    {
      "epoch": 0.5724,
      "grad_norm": 0.08008616417646408,
      "learning_rate": 0.002,
      "loss": 1.898937463760376,
      "step": 5724
    },
    {
      "epoch": 0.5726,
      "grad_norm": 0.07819913327693939,
      "learning_rate": 0.002,
      "loss": 1.9393359422683716,
      "step": 5726
    },
    {
      "epoch": 0.5728,
      "grad_norm": 0.08039375394582748,
      "learning_rate": 0.002,
      "loss": 1.991085410118103,
      "step": 5728
    },
    {
      "epoch": 0.573,
      "grad_norm": 0.07507607340812683,
      "learning_rate": 0.002,
      "loss": 1.9809010028839111,
      "step": 5730
    },
    {
      "epoch": 0.5732,
      "grad_norm": 0.07807087153196335,
      "learning_rate": 0.002,
      "loss": 1.923509120941162,
      "step": 5732
    },
    {
      "epoch": 0.5734,
      "grad_norm": 0.0750097781419754,
      "learning_rate": 0.002,
      "loss": 2.1095104217529297,
      "step": 5734
    },
    {
      "epoch": 0.5736,
      "grad_norm": 0.08191845566034317,
      "learning_rate": 0.002,
      "loss": 1.9116714000701904,
      "step": 5736
    },
    {
      "epoch": 0.5738,
      "grad_norm": 0.07808974385261536,
      "learning_rate": 0.002,
      "loss": 1.9301105737686157,
      "step": 5738
    },
    {
      "epoch": 0.574,
      "grad_norm": 0.0821467712521553,
      "learning_rate": 0.002,
      "loss": 2.0358035564422607,
      "step": 5740
    },
    {
      "epoch": 0.5742,
      "grad_norm": 0.07791148126125336,
      "learning_rate": 0.002,
      "loss": 1.944560170173645,
      "step": 5742
    },
    {
      "epoch": 0.5744,
      "grad_norm": 0.07718358933925629,
      "learning_rate": 0.002,
      "loss": 2.002582311630249,
      "step": 5744
    },
    {
      "epoch": 0.5746,
      "grad_norm": 0.09201342612504959,
      "learning_rate": 0.002,
      "loss": 2.029015302658081,
      "step": 5746
    },
    {
      "epoch": 0.5748,
      "grad_norm": 0.07608085125684738,
      "learning_rate": 0.002,
      "loss": 1.9627946615219116,
      "step": 5748
    },
    {
      "epoch": 0.575,
      "grad_norm": 0.07724042236804962,
      "learning_rate": 0.002,
      "loss": 1.8587732315063477,
      "step": 5750
    },
    {
      "epoch": 0.5752,
      "grad_norm": 0.07777610421180725,
      "learning_rate": 0.002,
      "loss": 2.0050406455993652,
      "step": 5752
    },
    {
      "epoch": 0.5754,
      "grad_norm": 0.07652690261602402,
      "learning_rate": 0.002,
      "loss": 1.952979564666748,
      "step": 5754
    },
    {
      "epoch": 0.5756,
      "grad_norm": 0.07511743903160095,
      "learning_rate": 0.002,
      "loss": 1.9473222494125366,
      "step": 5756
    },
    {
      "epoch": 0.5758,
      "grad_norm": 0.0809517577290535,
      "learning_rate": 0.002,
      "loss": 1.9111155271530151,
      "step": 5758
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.08461012691259384,
      "learning_rate": 0.002,
      "loss": 1.942346215248108,
      "step": 5760
    },
    {
      "epoch": 0.5762,
      "grad_norm": 0.08058876544237137,
      "learning_rate": 0.002,
      "loss": 1.9618330001831055,
      "step": 5762
    },
    {
      "epoch": 0.5764,
      "grad_norm": 0.0769747942686081,
      "learning_rate": 0.002,
      "loss": 1.8777179718017578,
      "step": 5764
    },
    {
      "epoch": 0.5766,
      "grad_norm": 0.08171173930168152,
      "learning_rate": 0.002,
      "loss": 1.9705621004104614,
      "step": 5766
    },
    {
      "epoch": 0.5768,
      "grad_norm": 0.07805880159139633,
      "learning_rate": 0.002,
      "loss": 1.9838393926620483,
      "step": 5768
    },
    {
      "epoch": 0.577,
      "grad_norm": 0.07680511474609375,
      "learning_rate": 0.002,
      "loss": 1.9911702871322632,
      "step": 5770
    },
    {
      "epoch": 0.5772,
      "grad_norm": 0.08390338718891144,
      "learning_rate": 0.002,
      "loss": 2.134838581085205,
      "step": 5772
    },
    {
      "epoch": 0.5774,
      "grad_norm": 0.07949217408895493,
      "learning_rate": 0.002,
      "loss": 2.0477631092071533,
      "step": 5774
    },
    {
      "epoch": 0.5776,
      "grad_norm": 0.07902462780475616,
      "learning_rate": 0.002,
      "loss": 1.913058876991272,
      "step": 5776
    },
    {
      "epoch": 0.5778,
      "grad_norm": 0.07731308043003082,
      "learning_rate": 0.002,
      "loss": 2.0687520503997803,
      "step": 5778
    },
    {
      "epoch": 0.578,
      "grad_norm": 0.0779237300157547,
      "learning_rate": 0.002,
      "loss": 2.0239365100860596,
      "step": 5780
    },
    {
      "epoch": 0.5782,
      "grad_norm": 0.08086138963699341,
      "learning_rate": 0.002,
      "loss": 2.0689213275909424,
      "step": 5782
    },
    {
      "epoch": 0.5784,
      "grad_norm": 0.07655498385429382,
      "learning_rate": 0.002,
      "loss": 2.0706839561462402,
      "step": 5784
    },
    {
      "epoch": 0.5786,
      "grad_norm": 0.07651811838150024,
      "learning_rate": 0.002,
      "loss": 1.810100793838501,
      "step": 5786
    },
    {
      "epoch": 0.5788,
      "grad_norm": 0.07894332706928253,
      "learning_rate": 0.002,
      "loss": 1.9305320978164673,
      "step": 5788
    },
    {
      "epoch": 0.579,
      "grad_norm": 0.0810433030128479,
      "learning_rate": 0.002,
      "loss": 1.8941279649734497,
      "step": 5790
    },
    {
      "epoch": 0.5792,
      "grad_norm": 0.07732107490301132,
      "learning_rate": 0.002,
      "loss": 1.9366049766540527,
      "step": 5792
    },
    {
      "epoch": 0.5794,
      "grad_norm": 0.07768285274505615,
      "learning_rate": 0.002,
      "loss": 2.05240797996521,
      "step": 5794
    },
    {
      "epoch": 0.5796,
      "grad_norm": 0.0749460905790329,
      "learning_rate": 0.002,
      "loss": 1.9875437021255493,
      "step": 5796
    },
    {
      "epoch": 0.5798,
      "grad_norm": 0.08036568015813828,
      "learning_rate": 0.002,
      "loss": 2.1871516704559326,
      "step": 5798
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.07736418396234512,
      "learning_rate": 0.002,
      "loss": 2.038949489593506,
      "step": 5800
    },
    {
      "epoch": 0.5802,
      "grad_norm": 0.08712290227413177,
      "learning_rate": 0.002,
      "loss": 1.956246256828308,
      "step": 5802
    },
    {
      "epoch": 0.5804,
      "grad_norm": 0.07727576047182083,
      "learning_rate": 0.002,
      "loss": 2.003352165222168,
      "step": 5804
    },
    {
      "epoch": 0.5806,
      "grad_norm": 0.07900335639715195,
      "learning_rate": 0.002,
      "loss": 1.8459343910217285,
      "step": 5806
    },
    {
      "epoch": 0.5808,
      "grad_norm": 0.09016159176826477,
      "learning_rate": 0.002,
      "loss": 2.0143651962280273,
      "step": 5808
    },
    {
      "epoch": 0.581,
      "grad_norm": 0.07722342759370804,
      "learning_rate": 0.002,
      "loss": 2.0052199363708496,
      "step": 5810
    },
    {
      "epoch": 0.5812,
      "grad_norm": 0.07400644570589066,
      "learning_rate": 0.002,
      "loss": 1.9821670055389404,
      "step": 5812
    },
    {
      "epoch": 0.5814,
      "grad_norm": 0.07703004032373428,
      "learning_rate": 0.002,
      "loss": 1.978453278541565,
      "step": 5814
    },
    {
      "epoch": 0.5816,
      "grad_norm": 0.07967765629291534,
      "learning_rate": 0.002,
      "loss": 2.033041000366211,
      "step": 5816
    },
    {
      "epoch": 0.5818,
      "grad_norm": 0.07617972791194916,
      "learning_rate": 0.002,
      "loss": 1.9196749925613403,
      "step": 5818
    },
    {
      "epoch": 0.582,
      "grad_norm": 0.07685504108667374,
      "learning_rate": 0.002,
      "loss": 1.8556381464004517,
      "step": 5820
    },
    {
      "epoch": 0.5822,
      "grad_norm": 0.08020743727684021,
      "learning_rate": 0.002,
      "loss": 2.0004658699035645,
      "step": 5822
    },
    {
      "epoch": 0.5824,
      "grad_norm": 0.07747578620910645,
      "learning_rate": 0.002,
      "loss": 1.9395153522491455,
      "step": 5824
    },
    {
      "epoch": 0.5826,
      "grad_norm": 0.08172491192817688,
      "learning_rate": 0.002,
      "loss": 2.105172872543335,
      "step": 5826
    },
    {
      "epoch": 0.5828,
      "grad_norm": 0.08161337673664093,
      "learning_rate": 0.002,
      "loss": 2.129234790802002,
      "step": 5828
    },
    {
      "epoch": 0.583,
      "grad_norm": 0.07553619146347046,
      "learning_rate": 0.002,
      "loss": 1.980723261833191,
      "step": 5830
    },
    {
      "epoch": 0.5832,
      "grad_norm": 0.0838640108704567,
      "learning_rate": 0.002,
      "loss": 1.9058669805526733,
      "step": 5832
    },
    {
      "epoch": 0.5834,
      "grad_norm": 0.07774312794208527,
      "learning_rate": 0.002,
      "loss": 1.9995192289352417,
      "step": 5834
    },
    {
      "epoch": 0.5836,
      "grad_norm": 0.07520958036184311,
      "learning_rate": 0.002,
      "loss": 1.9167461395263672,
      "step": 5836
    },
    {
      "epoch": 0.5838,
      "grad_norm": 0.0723535344004631,
      "learning_rate": 0.002,
      "loss": 1.9101070165634155,
      "step": 5838
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.07869106531143188,
      "learning_rate": 0.002,
      "loss": 1.9464014768600464,
      "step": 5840
    },
    {
      "epoch": 0.5842,
      "grad_norm": 0.07701065391302109,
      "learning_rate": 0.002,
      "loss": 2.0355064868927,
      "step": 5842
    },
    {
      "epoch": 0.5844,
      "grad_norm": 0.07505746930837631,
      "learning_rate": 0.002,
      "loss": 1.9334088563919067,
      "step": 5844
    },
    {
      "epoch": 0.5846,
      "grad_norm": 0.08051802963018417,
      "learning_rate": 0.002,
      "loss": 2.010833501815796,
      "step": 5846
    },
    {
      "epoch": 0.5848,
      "grad_norm": 0.07463803887367249,
      "learning_rate": 0.002,
      "loss": 1.8709665536880493,
      "step": 5848
    },
    {
      "epoch": 0.585,
      "grad_norm": 0.07327635586261749,
      "learning_rate": 0.002,
      "loss": 1.9854792356491089,
      "step": 5850
    },
    {
      "epoch": 0.5852,
      "grad_norm": 0.07895408570766449,
      "learning_rate": 0.002,
      "loss": 1.8809282779693604,
      "step": 5852
    },
    {
      "epoch": 0.5854,
      "grad_norm": 0.07616578042507172,
      "learning_rate": 0.002,
      "loss": 1.8868435621261597,
      "step": 5854
    },
    {
      "epoch": 0.5856,
      "grad_norm": 0.07775107026100159,
      "learning_rate": 0.002,
      "loss": 2.0350594520568848,
      "step": 5856
    },
    {
      "epoch": 0.5858,
      "grad_norm": 0.07604984194040298,
      "learning_rate": 0.002,
      "loss": 2.0754826068878174,
      "step": 5858
    },
    {
      "epoch": 0.586,
      "grad_norm": 0.0785481408238411,
      "learning_rate": 0.002,
      "loss": 2.1002023220062256,
      "step": 5860
    },
    {
      "epoch": 0.5862,
      "grad_norm": 0.0771908387541771,
      "learning_rate": 0.002,
      "loss": 1.9023587703704834,
      "step": 5862
    },
    {
      "epoch": 0.5864,
      "grad_norm": 0.07751934975385666,
      "learning_rate": 0.002,
      "loss": 2.0453643798828125,
      "step": 5864
    },
    {
      "epoch": 0.5866,
      "grad_norm": 0.07951347529888153,
      "learning_rate": 0.002,
      "loss": 2.1505494117736816,
      "step": 5866
    },
    {
      "epoch": 0.5868,
      "grad_norm": 0.07559913396835327,
      "learning_rate": 0.002,
      "loss": 1.964423656463623,
      "step": 5868
    },
    {
      "epoch": 0.587,
      "grad_norm": 0.07663024216890335,
      "learning_rate": 0.002,
      "loss": 1.970767617225647,
      "step": 5870
    },
    {
      "epoch": 0.5872,
      "grad_norm": 0.07997342199087143,
      "learning_rate": 0.002,
      "loss": 1.9898357391357422,
      "step": 5872
    },
    {
      "epoch": 0.5874,
      "grad_norm": 0.0813855454325676,
      "learning_rate": 0.002,
      "loss": 1.9416239261627197,
      "step": 5874
    },
    {
      "epoch": 0.5876,
      "grad_norm": 0.07647668570280075,
      "learning_rate": 0.002,
      "loss": 1.9605637788772583,
      "step": 5876
    },
    {
      "epoch": 0.5878,
      "grad_norm": 0.07916806638240814,
      "learning_rate": 0.002,
      "loss": 1.9469356536865234,
      "step": 5878
    },
    {
      "epoch": 0.588,
      "grad_norm": 0.07628824561834335,
      "learning_rate": 0.002,
      "loss": 1.9821738004684448,
      "step": 5880
    },
    {
      "epoch": 0.5882,
      "grad_norm": 0.08715467900037766,
      "learning_rate": 0.002,
      "loss": 1.9444355964660645,
      "step": 5882
    },
    {
      "epoch": 0.5884,
      "grad_norm": 0.07778462767601013,
      "learning_rate": 0.002,
      "loss": 1.9858460426330566,
      "step": 5884
    },
    {
      "epoch": 0.5886,
      "grad_norm": 0.07784069329500198,
      "learning_rate": 0.002,
      "loss": 1.9611915349960327,
      "step": 5886
    },
    {
      "epoch": 0.5888,
      "grad_norm": 0.07826340943574905,
      "learning_rate": 0.002,
      "loss": 2.0066752433776855,
      "step": 5888
    },
    {
      "epoch": 0.589,
      "grad_norm": 0.07841828465461731,
      "learning_rate": 0.002,
      "loss": 1.9805666208267212,
      "step": 5890
    },
    {
      "epoch": 0.5892,
      "grad_norm": 0.08190486580133438,
      "learning_rate": 0.002,
      "loss": 1.9674503803253174,
      "step": 5892
    },
    {
      "epoch": 0.5894,
      "grad_norm": 0.07779610902070999,
      "learning_rate": 0.002,
      "loss": 2.0369784832000732,
      "step": 5894
    },
    {
      "epoch": 0.5896,
      "grad_norm": 0.07707847654819489,
      "learning_rate": 0.002,
      "loss": 1.9778743982315063,
      "step": 5896
    },
    {
      "epoch": 0.5898,
      "grad_norm": 0.07429515570402145,
      "learning_rate": 0.002,
      "loss": 1.907243013381958,
      "step": 5898
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.07814276963472366,
      "learning_rate": 0.002,
      "loss": 2.0151188373565674,
      "step": 5900
    },
    {
      "epoch": 0.5902,
      "grad_norm": 0.08453090488910675,
      "learning_rate": 0.002,
      "loss": 2.0013461112976074,
      "step": 5902
    },
    {
      "epoch": 0.5904,
      "grad_norm": 0.07581886649131775,
      "learning_rate": 0.002,
      "loss": 1.8977627754211426,
      "step": 5904
    },
    {
      "epoch": 0.5906,
      "grad_norm": 0.09055624157190323,
      "learning_rate": 0.002,
      "loss": 2.052175521850586,
      "step": 5906
    },
    {
      "epoch": 0.5908,
      "grad_norm": 0.07786256074905396,
      "learning_rate": 0.002,
      "loss": 1.9621212482452393,
      "step": 5908
    },
    {
      "epoch": 0.591,
      "grad_norm": 0.07611776888370514,
      "learning_rate": 0.002,
      "loss": 1.8923684358596802,
      "step": 5910
    },
    {
      "epoch": 0.5912,
      "grad_norm": 0.0842597484588623,
      "learning_rate": 0.002,
      "loss": 2.041764259338379,
      "step": 5912
    },
    {
      "epoch": 0.5914,
      "grad_norm": 0.076950304210186,
      "learning_rate": 0.002,
      "loss": 1.871627926826477,
      "step": 5914
    },
    {
      "epoch": 0.5916,
      "grad_norm": 0.07688014209270477,
      "learning_rate": 0.002,
      "loss": 1.936655044555664,
      "step": 5916
    },
    {
      "epoch": 0.5918,
      "grad_norm": 0.07720822095870972,
      "learning_rate": 0.002,
      "loss": 1.9422593116760254,
      "step": 5918
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.07753828167915344,
      "learning_rate": 0.002,
      "loss": 1.9308003187179565,
      "step": 5920
    },
    {
      "epoch": 0.5922,
      "grad_norm": 0.07850523293018341,
      "learning_rate": 0.002,
      "loss": 2.0499253273010254,
      "step": 5922
    },
    {
      "epoch": 0.5924,
      "grad_norm": 0.0784951001405716,
      "learning_rate": 0.002,
      "loss": 1.9579402208328247,
      "step": 5924
    },
    {
      "epoch": 0.5926,
      "grad_norm": 0.07639297097921371,
      "learning_rate": 0.002,
      "loss": 1.9659534692764282,
      "step": 5926
    },
    {
      "epoch": 0.5928,
      "grad_norm": 0.07565931975841522,
      "learning_rate": 0.002,
      "loss": 2.0255422592163086,
      "step": 5928
    },
    {
      "epoch": 0.593,
      "grad_norm": 0.07482057064771652,
      "learning_rate": 0.002,
      "loss": 1.837839126586914,
      "step": 5930
    },
    {
      "epoch": 0.5932,
      "grad_norm": 0.0781397596001625,
      "learning_rate": 0.002,
      "loss": 1.965792179107666,
      "step": 5932
    },
    {
      "epoch": 0.5934,
      "grad_norm": 0.07431638985872269,
      "learning_rate": 0.002,
      "loss": 1.9439351558685303,
      "step": 5934
    },
    {
      "epoch": 0.5936,
      "grad_norm": 0.07510343939065933,
      "learning_rate": 0.002,
      "loss": 2.072211503982544,
      "step": 5936
    },
    {
      "epoch": 0.5938,
      "grad_norm": 0.07851964235305786,
      "learning_rate": 0.002,
      "loss": 2.0595216751098633,
      "step": 5938
    },
    {
      "epoch": 0.594,
      "grad_norm": 0.07924996316432953,
      "learning_rate": 0.002,
      "loss": 2.0498404502868652,
      "step": 5940
    },
    {
      "epoch": 0.5942,
      "grad_norm": 0.07696901261806488,
      "learning_rate": 0.002,
      "loss": 2.066027879714966,
      "step": 5942
    },
    {
      "epoch": 0.5944,
      "grad_norm": 0.07499818503856659,
      "learning_rate": 0.002,
      "loss": 2.0135490894317627,
      "step": 5944
    },
    {
      "epoch": 0.5946,
      "grad_norm": 0.0800037533044815,
      "learning_rate": 0.002,
      "loss": 1.876547932624817,
      "step": 5946
    },
    {
      "epoch": 0.5948,
      "grad_norm": 0.0780591145157814,
      "learning_rate": 0.002,
      "loss": 2.0900697708129883,
      "step": 5948
    },
    {
      "epoch": 0.595,
      "grad_norm": 0.0779573917388916,
      "learning_rate": 0.002,
      "loss": 1.9464911222457886,
      "step": 5950
    },
    {
      "epoch": 0.5952,
      "grad_norm": 0.08457666635513306,
      "learning_rate": 0.002,
      "loss": 2.0775094032287598,
      "step": 5952
    },
    {
      "epoch": 0.5954,
      "grad_norm": 0.0775676965713501,
      "learning_rate": 0.002,
      "loss": 2.0103986263275146,
      "step": 5954
    },
    {
      "epoch": 0.5956,
      "grad_norm": 0.07535495609045029,
      "learning_rate": 0.002,
      "loss": 2.0462234020233154,
      "step": 5956
    },
    {
      "epoch": 0.5958,
      "grad_norm": 0.07679515331983566,
      "learning_rate": 0.002,
      "loss": 1.9842476844787598,
      "step": 5958
    },
    {
      "epoch": 0.596,
      "grad_norm": 0.08431209623813629,
      "learning_rate": 0.002,
      "loss": 1.9616801738739014,
      "step": 5960
    },
    {
      "epoch": 0.5962,
      "grad_norm": 0.07812201976776123,
      "learning_rate": 0.002,
      "loss": 2.0276222229003906,
      "step": 5962
    },
    {
      "epoch": 0.5964,
      "grad_norm": 0.07452664524316788,
      "learning_rate": 0.002,
      "loss": 2.0061228275299072,
      "step": 5964
    },
    {
      "epoch": 0.5966,
      "grad_norm": 0.07594982534646988,
      "learning_rate": 0.002,
      "loss": 2.0013980865478516,
      "step": 5966
    },
    {
      "epoch": 0.5968,
      "grad_norm": 0.07974756509065628,
      "learning_rate": 0.002,
      "loss": 2.094162702560425,
      "step": 5968
    },
    {
      "epoch": 0.597,
      "grad_norm": 0.07780814915895462,
      "learning_rate": 0.002,
      "loss": 2.0979723930358887,
      "step": 5970
    },
    {
      "epoch": 0.5972,
      "grad_norm": 0.0786043033003807,
      "learning_rate": 0.002,
      "loss": 1.997055172920227,
      "step": 5972
    },
    {
      "epoch": 0.5974,
      "grad_norm": 0.0764949768781662,
      "learning_rate": 0.002,
      "loss": 1.9603490829467773,
      "step": 5974
    },
    {
      "epoch": 0.5976,
      "grad_norm": 0.08069033920764923,
      "learning_rate": 0.002,
      "loss": 2.0768637657165527,
      "step": 5976
    },
    {
      "epoch": 0.5978,
      "grad_norm": 0.07547701150178909,
      "learning_rate": 0.002,
      "loss": 2.0351877212524414,
      "step": 5978
    },
    {
      "epoch": 0.598,
      "grad_norm": 0.07448173314332962,
      "learning_rate": 0.002,
      "loss": 1.97465980052948,
      "step": 5980
    },
    {
      "epoch": 0.5982,
      "grad_norm": 0.07879731059074402,
      "learning_rate": 0.002,
      "loss": 2.004188060760498,
      "step": 5982
    },
    {
      "epoch": 0.5984,
      "grad_norm": 0.07817785441875458,
      "learning_rate": 0.002,
      "loss": 2.02921724319458,
      "step": 5984
    },
    {
      "epoch": 0.5986,
      "grad_norm": 0.07980640977621078,
      "learning_rate": 0.002,
      "loss": 1.9530928134918213,
      "step": 5986
    },
    {
      "epoch": 0.5988,
      "grad_norm": 0.07530110329389572,
      "learning_rate": 0.002,
      "loss": 2.077218532562256,
      "step": 5988
    },
    {
      "epoch": 0.599,
      "grad_norm": 0.0750943198800087,
      "learning_rate": 0.002,
      "loss": 2.175501585006714,
      "step": 5990
    },
    {
      "epoch": 0.5992,
      "grad_norm": 0.0788198933005333,
      "learning_rate": 0.002,
      "loss": 2.0092644691467285,
      "step": 5992
    },
    {
      "epoch": 0.5994,
      "grad_norm": 0.07636000216007233,
      "learning_rate": 0.002,
      "loss": 2.0372519493103027,
      "step": 5994
    },
    {
      "epoch": 0.5996,
      "grad_norm": 0.07690858095884323,
      "learning_rate": 0.002,
      "loss": 1.9217990636825562,
      "step": 5996
    },
    {
      "epoch": 0.5998,
      "grad_norm": 0.07347720861434937,
      "learning_rate": 0.002,
      "loss": 1.9622722864151,
      "step": 5998
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.08011019229888916,
      "learning_rate": 0.002,
      "loss": 2.07912015914917,
      "step": 6000
    },
    {
      "epoch": 0.6,
      "eval_loss": 2.1336655616760254,
      "eval_runtime": 232.2834,
      "eval_samples_per_second": 2.153,
      "eval_steps_per_second": 0.215,
      "step": 6000
    },
    {
      "epoch": 0.6002,
      "grad_norm": 0.07604338973760605,
      "learning_rate": 0.002,
      "loss": 1.9753174781799316,
      "step": 6002
    },
    {
      "epoch": 0.6004,
      "grad_norm": 0.08399670571088791,
      "learning_rate": 0.002,
      "loss": 1.9375505447387695,
      "step": 6004
    },
    {
      "epoch": 0.6006,
      "grad_norm": 0.07638192176818848,
      "learning_rate": 0.002,
      "loss": 2.056913375854492,
      "step": 6006
    },
    {
      "epoch": 0.6008,
      "grad_norm": 0.07881305366754532,
      "learning_rate": 0.002,
      "loss": 2.036707639694214,
      "step": 6008
    },
    {
      "epoch": 0.601,
      "grad_norm": 0.07959351688623428,
      "learning_rate": 0.002,
      "loss": 2.08945369720459,
      "step": 6010
    },
    {
      "epoch": 0.6012,
      "grad_norm": 0.07405447214841843,
      "learning_rate": 0.002,
      "loss": 2.0885884761810303,
      "step": 6012
    },
    {
      "epoch": 0.6014,
      "grad_norm": 0.08402067422866821,
      "learning_rate": 0.002,
      "loss": 2.034414052963257,
      "step": 6014
    },
    {
      "epoch": 0.6016,
      "grad_norm": 0.0887426882982254,
      "learning_rate": 0.002,
      "loss": 2.0866012573242188,
      "step": 6016
    },
    {
      "epoch": 0.6018,
      "grad_norm": 0.07548937201499939,
      "learning_rate": 0.002,
      "loss": 1.998429536819458,
      "step": 6018
    },
    {
      "epoch": 0.602,
      "grad_norm": 0.07626893371343613,
      "learning_rate": 0.002,
      "loss": 1.989201545715332,
      "step": 6020
    },
    {
      "epoch": 0.6022,
      "grad_norm": 0.07573816925287247,
      "learning_rate": 0.002,
      "loss": 2.013731002807617,
      "step": 6022
    },
    {
      "epoch": 0.6024,
      "grad_norm": 0.08309171348810196,
      "learning_rate": 0.002,
      "loss": 2.156155586242676,
      "step": 6024
    },
    {
      "epoch": 0.6026,
      "grad_norm": 0.07680265605449677,
      "learning_rate": 0.002,
      "loss": 2.008910894393921,
      "step": 6026
    },
    {
      "epoch": 0.6028,
      "grad_norm": 0.08088570833206177,
      "learning_rate": 0.002,
      "loss": 1.9811259508132935,
      "step": 6028
    },
    {
      "epoch": 0.603,
      "grad_norm": 0.07755240797996521,
      "learning_rate": 0.002,
      "loss": 2.0791914463043213,
      "step": 6030
    },
    {
      "epoch": 0.6032,
      "grad_norm": 0.07661478221416473,
      "learning_rate": 0.002,
      "loss": 2.0059404373168945,
      "step": 6032
    },
    {
      "epoch": 0.6034,
      "grad_norm": 0.08144840598106384,
      "learning_rate": 0.002,
      "loss": 2.0289976596832275,
      "step": 6034
    },
    {
      "epoch": 0.6036,
      "grad_norm": 0.07995229214429855,
      "learning_rate": 0.002,
      "loss": 1.9940117597579956,
      "step": 6036
    },
    {
      "epoch": 0.6038,
      "grad_norm": 0.07423283159732819,
      "learning_rate": 0.002,
      "loss": 2.0145978927612305,
      "step": 6038
    },
    {
      "epoch": 0.604,
      "grad_norm": 0.07667619735002518,
      "learning_rate": 0.002,
      "loss": 1.9503633975982666,
      "step": 6040
    },
    {
      "epoch": 0.6042,
      "grad_norm": 0.07422744482755661,
      "learning_rate": 0.002,
      "loss": 2.0690624713897705,
      "step": 6042
    },
    {
      "epoch": 0.6044,
      "grad_norm": 0.08365897089242935,
      "learning_rate": 0.002,
      "loss": 2.0511350631713867,
      "step": 6044
    },
    {
      "epoch": 0.6046,
      "grad_norm": 0.07661974430084229,
      "learning_rate": 0.002,
      "loss": 2.060992479324341,
      "step": 6046
    },
    {
      "epoch": 0.6048,
      "grad_norm": 0.07761109620332718,
      "learning_rate": 0.002,
      "loss": 2.0939273834228516,
      "step": 6048
    },
    {
      "epoch": 0.605,
      "grad_norm": 0.07712405174970627,
      "learning_rate": 0.002,
      "loss": 2.0195374488830566,
      "step": 6050
    },
    {
      "epoch": 0.6052,
      "grad_norm": 0.0804099440574646,
      "learning_rate": 0.002,
      "loss": 1.963290810585022,
      "step": 6052
    },
    {
      "epoch": 0.6054,
      "grad_norm": 0.08128418773412704,
      "learning_rate": 0.002,
      "loss": 2.1995763778686523,
      "step": 6054
    },
    {
      "epoch": 0.6056,
      "grad_norm": 0.07456329464912415,
      "learning_rate": 0.002,
      "loss": 1.9971987009048462,
      "step": 6056
    },
    {
      "epoch": 0.6058,
      "grad_norm": 0.07757089287042618,
      "learning_rate": 0.002,
      "loss": 2.035733938217163,
      "step": 6058
    },
    {
      "epoch": 0.606,
      "grad_norm": 0.08252770453691483,
      "learning_rate": 0.002,
      "loss": 2.040443181991577,
      "step": 6060
    },
    {
      "epoch": 0.6062,
      "grad_norm": 0.07323038578033447,
      "learning_rate": 0.002,
      "loss": 1.9960448741912842,
      "step": 6062
    },
    {
      "epoch": 0.6064,
      "grad_norm": 0.07789022475481033,
      "learning_rate": 0.002,
      "loss": 1.9978294372558594,
      "step": 6064
    },
    {
      "epoch": 0.6066,
      "grad_norm": 0.08683673292398453,
      "learning_rate": 0.002,
      "loss": 2.0169835090637207,
      "step": 6066
    },
    {
      "epoch": 0.6068,
      "grad_norm": 0.07740672677755356,
      "learning_rate": 0.002,
      "loss": 1.9716196060180664,
      "step": 6068
    },
    {
      "epoch": 0.607,
      "grad_norm": 0.0786198228597641,
      "learning_rate": 0.002,
      "loss": 2.074467420578003,
      "step": 6070
    },
    {
      "epoch": 0.6072,
      "grad_norm": 0.07505638152360916,
      "learning_rate": 0.002,
      "loss": 2.0615406036376953,
      "step": 6072
    },
    {
      "epoch": 0.6074,
      "grad_norm": 0.07937132567167282,
      "learning_rate": 0.002,
      "loss": 1.9581818580627441,
      "step": 6074
    },
    {
      "epoch": 0.6076,
      "grad_norm": 0.07556100934743881,
      "learning_rate": 0.002,
      "loss": 1.990000605583191,
      "step": 6076
    },
    {
      "epoch": 0.6078,
      "grad_norm": 0.07891649752855301,
      "learning_rate": 0.002,
      "loss": 2.067547082901001,
      "step": 6078
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.0806807279586792,
      "learning_rate": 0.002,
      "loss": 1.9739726781845093,
      "step": 6080
    },
    {
      "epoch": 0.6082,
      "grad_norm": 0.07922226935625076,
      "learning_rate": 0.002,
      "loss": 1.988715648651123,
      "step": 6082
    },
    {
      "epoch": 0.6084,
      "grad_norm": 0.07592223584651947,
      "learning_rate": 0.002,
      "loss": 2.012101650238037,
      "step": 6084
    },
    {
      "epoch": 0.6086,
      "grad_norm": 0.07602278143167496,
      "learning_rate": 0.002,
      "loss": 2.0907483100891113,
      "step": 6086
    },
    {
      "epoch": 0.6088,
      "grad_norm": 0.07723119109869003,
      "learning_rate": 0.002,
      "loss": 1.9872479438781738,
      "step": 6088
    },
    {
      "epoch": 0.609,
      "grad_norm": 0.07906521111726761,
      "learning_rate": 0.002,
      "loss": 2.018954277038574,
      "step": 6090
    },
    {
      "epoch": 0.6092,
      "grad_norm": 0.08043338358402252,
      "learning_rate": 0.002,
      "loss": 2.132708787918091,
      "step": 6092
    },
    {
      "epoch": 0.6094,
      "grad_norm": 0.08126092702150345,
      "learning_rate": 0.002,
      "loss": 1.9395338296890259,
      "step": 6094
    },
    {
      "epoch": 0.6096,
      "grad_norm": 0.07839156687259674,
      "learning_rate": 0.002,
      "loss": 2.0532302856445312,
      "step": 6096
    },
    {
      "epoch": 0.6098,
      "grad_norm": 0.0808495283126831,
      "learning_rate": 0.002,
      "loss": 1.9830503463745117,
      "step": 6098
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.08279605954885483,
      "learning_rate": 0.002,
      "loss": 2.13291597366333,
      "step": 6100
    },
    {
      "epoch": 0.6102,
      "grad_norm": 0.08135608583688736,
      "learning_rate": 0.002,
      "loss": 2.003235340118408,
      "step": 6102
    },
    {
      "epoch": 0.6104,
      "grad_norm": 0.08470860123634338,
      "learning_rate": 0.002,
      "loss": 2.043872117996216,
      "step": 6104
    },
    {
      "epoch": 0.6106,
      "grad_norm": 0.07927827537059784,
      "learning_rate": 0.002,
      "loss": 1.9969956874847412,
      "step": 6106
    },
    {
      "epoch": 0.6108,
      "grad_norm": 0.07654355466365814,
      "learning_rate": 0.002,
      "loss": 2.0150930881500244,
      "step": 6108
    },
    {
      "epoch": 0.611,
      "grad_norm": 0.07660844922065735,
      "learning_rate": 0.002,
      "loss": 1.9911346435546875,
      "step": 6110
    },
    {
      "epoch": 0.6112,
      "grad_norm": 0.08040948957204819,
      "learning_rate": 0.002,
      "loss": 2.0009872913360596,
      "step": 6112
    },
    {
      "epoch": 0.6114,
      "grad_norm": 0.08138270676136017,
      "learning_rate": 0.002,
      "loss": 2.049833059310913,
      "step": 6114
    },
    {
      "epoch": 0.6116,
      "grad_norm": 0.08012638986110687,
      "learning_rate": 0.002,
      "loss": 2.151442766189575,
      "step": 6116
    },
    {
      "epoch": 0.6118,
      "grad_norm": 0.08007648587226868,
      "learning_rate": 0.002,
      "loss": 2.0135657787323,
      "step": 6118
    },
    {
      "epoch": 0.612,
      "grad_norm": 0.08208825439214706,
      "learning_rate": 0.002,
      "loss": 2.0915684700012207,
      "step": 6120
    },
    {
      "epoch": 0.6122,
      "grad_norm": 0.07867388427257538,
      "learning_rate": 0.002,
      "loss": 2.019890069961548,
      "step": 6122
    },
    {
      "epoch": 0.6124,
      "grad_norm": 0.07224659621715546,
      "learning_rate": 0.002,
      "loss": 1.98256516456604,
      "step": 6124
    },
    {
      "epoch": 0.6126,
      "grad_norm": 0.07927681505680084,
      "learning_rate": 0.002,
      "loss": 2.0406739711761475,
      "step": 6126
    },
    {
      "epoch": 0.6128,
      "grad_norm": 0.07556013762950897,
      "learning_rate": 0.002,
      "loss": 2.0572402477264404,
      "step": 6128
    },
    {
      "epoch": 0.613,
      "grad_norm": 0.08009274303913116,
      "learning_rate": 0.002,
      "loss": 2.1565897464752197,
      "step": 6130
    },
    {
      "epoch": 0.6132,
      "grad_norm": 0.0768369808793068,
      "learning_rate": 0.002,
      "loss": 2.03334641456604,
      "step": 6132
    },
    {
      "epoch": 0.6134,
      "grad_norm": 0.08003520220518112,
      "learning_rate": 0.002,
      "loss": 2.016633987426758,
      "step": 6134
    },
    {
      "epoch": 0.6136,
      "grad_norm": 0.07531091570854187,
      "learning_rate": 0.002,
      "loss": 1.966112732887268,
      "step": 6136
    },
    {
      "epoch": 0.6138,
      "grad_norm": 0.07906363904476166,
      "learning_rate": 0.002,
      "loss": 1.9127998352050781,
      "step": 6138
    },
    {
      "epoch": 0.614,
      "grad_norm": 0.0786106288433075,
      "learning_rate": 0.002,
      "loss": 2.0388870239257812,
      "step": 6140
    },
    {
      "epoch": 0.6142,
      "grad_norm": 0.07961447536945343,
      "learning_rate": 0.002,
      "loss": 2.1206138134002686,
      "step": 6142
    },
    {
      "epoch": 0.6144,
      "grad_norm": 0.07806965708732605,
      "learning_rate": 0.002,
      "loss": 1.9214831590652466,
      "step": 6144
    },
    {
      "epoch": 0.6146,
      "grad_norm": 0.07325156033039093,
      "learning_rate": 0.002,
      "loss": 2.039318323135376,
      "step": 6146
    },
    {
      "epoch": 0.6148,
      "grad_norm": 0.0786585733294487,
      "learning_rate": 0.002,
      "loss": 2.091599225997925,
      "step": 6148
    },
    {
      "epoch": 0.615,
      "grad_norm": 0.07802821695804596,
      "learning_rate": 0.002,
      "loss": 1.921869158744812,
      "step": 6150
    },
    {
      "epoch": 0.6152,
      "grad_norm": 0.0732760950922966,
      "learning_rate": 0.002,
      "loss": 2.033945322036743,
      "step": 6152
    },
    {
      "epoch": 0.6154,
      "grad_norm": 0.07584500312805176,
      "learning_rate": 0.002,
      "loss": 2.013887405395508,
      "step": 6154
    },
    {
      "epoch": 0.6156,
      "grad_norm": 0.07789415866136551,
      "learning_rate": 0.002,
      "loss": 2.0268356800079346,
      "step": 6156
    },
    {
      "epoch": 0.6158,
      "grad_norm": 0.07910613715648651,
      "learning_rate": 0.002,
      "loss": 2.1789262294769287,
      "step": 6158
    },
    {
      "epoch": 0.616,
      "grad_norm": 0.08226309716701508,
      "learning_rate": 0.002,
      "loss": 1.9973793029785156,
      "step": 6160
    },
    {
      "epoch": 0.6162,
      "grad_norm": 0.08274933695793152,
      "learning_rate": 0.002,
      "loss": 2.0675742626190186,
      "step": 6162
    },
    {
      "epoch": 0.6164,
      "grad_norm": 0.08000992983579636,
      "learning_rate": 0.002,
      "loss": 2.010561227798462,
      "step": 6164
    },
    {
      "epoch": 0.6166,
      "grad_norm": 0.077796071767807,
      "learning_rate": 0.002,
      "loss": 2.070453405380249,
      "step": 6166
    },
    {
      "epoch": 0.6168,
      "grad_norm": 0.07824783027172089,
      "learning_rate": 0.002,
      "loss": 1.9655159711837769,
      "step": 6168
    },
    {
      "epoch": 0.617,
      "grad_norm": 0.0752086192369461,
      "learning_rate": 0.002,
      "loss": 2.089231014251709,
      "step": 6170
    },
    {
      "epoch": 0.6172,
      "grad_norm": 0.07897814363241196,
      "learning_rate": 0.002,
      "loss": 2.0504791736602783,
      "step": 6172
    },
    {
      "epoch": 0.6174,
      "grad_norm": 0.0764254778623581,
      "learning_rate": 0.002,
      "loss": 2.041217088699341,
      "step": 6174
    },
    {
      "epoch": 0.6176,
      "grad_norm": 0.07763165235519409,
      "learning_rate": 0.002,
      "loss": 2.0164217948913574,
      "step": 6176
    },
    {
      "epoch": 0.6178,
      "grad_norm": 0.0753839910030365,
      "learning_rate": 0.002,
      "loss": 2.15096378326416,
      "step": 6178
    },
    {
      "epoch": 0.618,
      "grad_norm": 0.08047185093164444,
      "learning_rate": 0.002,
      "loss": 1.9343676567077637,
      "step": 6180
    },
    {
      "epoch": 0.6182,
      "grad_norm": 0.07759809494018555,
      "learning_rate": 0.002,
      "loss": 2.1437861919403076,
      "step": 6182
    },
    {
      "epoch": 0.6184,
      "grad_norm": 0.0854780375957489,
      "learning_rate": 0.002,
      "loss": 2.0681564807891846,
      "step": 6184
    },
    {
      "epoch": 0.6186,
      "grad_norm": 0.07746338844299316,
      "learning_rate": 0.002,
      "loss": 2.0366694927215576,
      "step": 6186
    },
    {
      "epoch": 0.6188,
      "grad_norm": 0.07863811403512955,
      "learning_rate": 0.002,
      "loss": 2.012580156326294,
      "step": 6188
    },
    {
      "epoch": 0.619,
      "grad_norm": 0.08576951920986176,
      "learning_rate": 0.002,
      "loss": 2.096017599105835,
      "step": 6190
    },
    {
      "epoch": 0.6192,
      "grad_norm": 0.07853306084871292,
      "learning_rate": 0.002,
      "loss": 2.1028079986572266,
      "step": 6192
    },
    {
      "epoch": 0.6194,
      "grad_norm": 0.07796917855739594,
      "learning_rate": 0.002,
      "loss": 1.9686118364334106,
      "step": 6194
    },
    {
      "epoch": 0.6196,
      "grad_norm": 0.07582274824380875,
      "learning_rate": 0.002,
      "loss": 2.006455421447754,
      "step": 6196
    },
    {
      "epoch": 0.6198,
      "grad_norm": 0.07602935284376144,
      "learning_rate": 0.002,
      "loss": 1.9841501712799072,
      "step": 6198
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.08168172836303711,
      "learning_rate": 0.002,
      "loss": 2.0705020427703857,
      "step": 6200
    },
    {
      "epoch": 0.6202,
      "grad_norm": 0.08011497557163239,
      "learning_rate": 0.002,
      "loss": 2.0669233798980713,
      "step": 6202
    },
    {
      "epoch": 0.6204,
      "grad_norm": 0.0828249454498291,
      "learning_rate": 0.002,
      "loss": 2.1275932788848877,
      "step": 6204
    },
    {
      "epoch": 0.6206,
      "grad_norm": 0.08094964921474457,
      "learning_rate": 0.002,
      "loss": 2.0331339836120605,
      "step": 6206
    },
    {
      "epoch": 0.6208,
      "grad_norm": 0.08295127749443054,
      "learning_rate": 0.002,
      "loss": 2.1500918865203857,
      "step": 6208
    },
    {
      "epoch": 0.621,
      "grad_norm": 0.0811968520283699,
      "learning_rate": 0.002,
      "loss": 1.9624569416046143,
      "step": 6210
    },
    {
      "epoch": 0.6212,
      "grad_norm": 0.07673266530036926,
      "learning_rate": 0.002,
      "loss": 2.1567211151123047,
      "step": 6212
    },
    {
      "epoch": 0.6214,
      "grad_norm": 0.07878787815570831,
      "learning_rate": 0.002,
      "loss": 2.1363391876220703,
      "step": 6214
    },
    {
      "epoch": 0.6216,
      "grad_norm": 0.07475794851779938,
      "learning_rate": 0.002,
      "loss": 2.050685167312622,
      "step": 6216
    },
    {
      "epoch": 0.6218,
      "grad_norm": 0.07973384857177734,
      "learning_rate": 0.002,
      "loss": 2.0383663177490234,
      "step": 6218
    },
    {
      "epoch": 0.622,
      "grad_norm": 0.07493112236261368,
      "learning_rate": 0.002,
      "loss": 1.9964189529418945,
      "step": 6220
    },
    {
      "epoch": 0.6222,
      "grad_norm": 0.07875581830739975,
      "learning_rate": 0.002,
      "loss": 2.1266348361968994,
      "step": 6222
    },
    {
      "epoch": 0.6224,
      "grad_norm": 0.07953702658414841,
      "learning_rate": 0.002,
      "loss": 2.124748468399048,
      "step": 6224
    },
    {
      "epoch": 0.6226,
      "grad_norm": 0.07551873475313187,
      "learning_rate": 0.002,
      "loss": 2.017103910446167,
      "step": 6226
    },
    {
      "epoch": 0.6228,
      "grad_norm": 0.0783180370926857,
      "learning_rate": 0.002,
      "loss": 2.0000784397125244,
      "step": 6228
    },
    {
      "epoch": 0.623,
      "grad_norm": 0.07795071601867676,
      "learning_rate": 0.002,
      "loss": 1.989213228225708,
      "step": 6230
    },
    {
      "epoch": 0.6232,
      "grad_norm": 0.0812332034111023,
      "learning_rate": 0.002,
      "loss": 2.0915675163269043,
      "step": 6232
    },
    {
      "epoch": 0.6234,
      "grad_norm": 0.07623962312936783,
      "learning_rate": 0.002,
      "loss": 1.927616000175476,
      "step": 6234
    },
    {
      "epoch": 0.6236,
      "grad_norm": 0.07972859591245651,
      "learning_rate": 0.002,
      "loss": 2.0064520835876465,
      "step": 6236
    },
    {
      "epoch": 0.6238,
      "grad_norm": 0.08073878288269043,
      "learning_rate": 0.002,
      "loss": 2.0170295238494873,
      "step": 6238
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.07713044434785843,
      "learning_rate": 0.002,
      "loss": 1.9298557043075562,
      "step": 6240
    },
    {
      "epoch": 0.6242,
      "grad_norm": 0.07838242501020432,
      "learning_rate": 0.002,
      "loss": 1.9940468072891235,
      "step": 6242
    },
    {
      "epoch": 0.6244,
      "grad_norm": 0.07592581957578659,
      "learning_rate": 0.002,
      "loss": 1.9824814796447754,
      "step": 6244
    },
    {
      "epoch": 0.6246,
      "grad_norm": 0.07884396612644196,
      "learning_rate": 0.002,
      "loss": 1.9896868467330933,
      "step": 6246
    },
    {
      "epoch": 0.6248,
      "grad_norm": 0.07731294631958008,
      "learning_rate": 0.002,
      "loss": 1.9719157218933105,
      "step": 6248
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.08256322145462036,
      "learning_rate": 0.002,
      "loss": 2.077684164047241,
      "step": 6250
    },
    {
      "epoch": 0.6252,
      "grad_norm": 0.08179715275764465,
      "learning_rate": 0.002,
      "loss": 2.0982534885406494,
      "step": 6252
    },
    {
      "epoch": 0.6254,
      "grad_norm": 0.08428557217121124,
      "learning_rate": 0.002,
      "loss": 2.1607601642608643,
      "step": 6254
    },
    {
      "epoch": 0.6256,
      "grad_norm": 0.07624182850122452,
      "learning_rate": 0.002,
      "loss": 2.08620023727417,
      "step": 6256
    },
    {
      "epoch": 0.6258,
      "grad_norm": 0.08871359378099442,
      "learning_rate": 0.002,
      "loss": 2.1059207916259766,
      "step": 6258
    },
    {
      "epoch": 0.626,
      "grad_norm": 0.08266192674636841,
      "learning_rate": 0.002,
      "loss": 2.0928092002868652,
      "step": 6260
    },
    {
      "epoch": 0.6262,
      "grad_norm": 0.0831819698214531,
      "learning_rate": 0.002,
      "loss": 2.093472719192505,
      "step": 6262
    },
    {
      "epoch": 0.6264,
      "grad_norm": 0.08511131256818771,
      "learning_rate": 0.002,
      "loss": 2.092061996459961,
      "step": 6264
    },
    {
      "epoch": 0.6266,
      "grad_norm": 0.08151786029338837,
      "learning_rate": 0.002,
      "loss": 2.142213821411133,
      "step": 6266
    },
    {
      "epoch": 0.6268,
      "grad_norm": 0.07915618270635605,
      "learning_rate": 0.002,
      "loss": 2.0570943355560303,
      "step": 6268
    },
    {
      "epoch": 0.627,
      "grad_norm": 0.07871002703905106,
      "learning_rate": 0.002,
      "loss": 1.9148147106170654,
      "step": 6270
    },
    {
      "epoch": 0.6272,
      "grad_norm": 0.08040455728769302,
      "learning_rate": 0.002,
      "loss": 2.145176410675049,
      "step": 6272
    },
    {
      "epoch": 0.6274,
      "grad_norm": 0.08053966611623764,
      "learning_rate": 0.002,
      "loss": 2.2168784141540527,
      "step": 6274
    },
    {
      "epoch": 0.6276,
      "grad_norm": 0.08029678463935852,
      "learning_rate": 0.002,
      "loss": 2.1299266815185547,
      "step": 6276
    },
    {
      "epoch": 0.6278,
      "grad_norm": 0.07770596444606781,
      "learning_rate": 0.002,
      "loss": 2.1855742931365967,
      "step": 6278
    },
    {
      "epoch": 0.628,
      "grad_norm": 0.07823921740055084,
      "learning_rate": 0.002,
      "loss": 2.0959413051605225,
      "step": 6280
    },
    {
      "epoch": 0.6282,
      "grad_norm": 0.07557258754968643,
      "learning_rate": 0.002,
      "loss": 2.035186290740967,
      "step": 6282
    },
    {
      "epoch": 0.6284,
      "grad_norm": 0.08072342723608017,
      "learning_rate": 0.002,
      "loss": 2.0713508129119873,
      "step": 6284
    },
    {
      "epoch": 0.6286,
      "grad_norm": 0.07892139256000519,
      "learning_rate": 0.002,
      "loss": 2.1937222480773926,
      "step": 6286
    },
    {
      "epoch": 0.6288,
      "grad_norm": 0.07815294712781906,
      "learning_rate": 0.002,
      "loss": 2.0739877223968506,
      "step": 6288
    },
    {
      "epoch": 0.629,
      "grad_norm": 0.0758562684059143,
      "learning_rate": 0.002,
      "loss": 2.044987678527832,
      "step": 6290
    },
    {
      "epoch": 0.6292,
      "grad_norm": 0.07958438992500305,
      "learning_rate": 0.002,
      "loss": 2.078300714492798,
      "step": 6292
    },
    {
      "epoch": 0.6294,
      "grad_norm": 0.07902704179286957,
      "learning_rate": 0.002,
      "loss": 2.000159740447998,
      "step": 6294
    },
    {
      "epoch": 0.6296,
      "grad_norm": 0.07725685089826584,
      "learning_rate": 0.002,
      "loss": 2.0814850330352783,
      "step": 6296
    },
    {
      "epoch": 0.6298,
      "grad_norm": 0.08224157243967056,
      "learning_rate": 0.002,
      "loss": 2.0989797115325928,
      "step": 6298
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.07767102867364883,
      "learning_rate": 0.002,
      "loss": 2.0834062099456787,
      "step": 6300
    },
    {
      "epoch": 0.6302,
      "grad_norm": 0.08379923552274704,
      "learning_rate": 0.002,
      "loss": 2.0215916633605957,
      "step": 6302
    },
    {
      "epoch": 0.6304,
      "grad_norm": 0.08269500732421875,
      "learning_rate": 0.002,
      "loss": 2.036106824874878,
      "step": 6304
    },
    {
      "epoch": 0.6306,
      "grad_norm": 0.08234849572181702,
      "learning_rate": 0.002,
      "loss": 1.9654628038406372,
      "step": 6306
    },
    {
      "epoch": 0.6308,
      "grad_norm": 0.07531068474054337,
      "learning_rate": 0.002,
      "loss": 2.042896032333374,
      "step": 6308
    },
    {
      "epoch": 0.631,
      "grad_norm": 0.07776293158531189,
      "learning_rate": 0.002,
      "loss": 2.1002118587493896,
      "step": 6310
    },
    {
      "epoch": 0.6312,
      "grad_norm": 0.07352367043495178,
      "learning_rate": 0.002,
      "loss": 1.9143092632293701,
      "step": 6312
    },
    {
      "epoch": 0.6314,
      "grad_norm": 0.08077633380889893,
      "learning_rate": 0.002,
      "loss": 2.0349574089050293,
      "step": 6314
    },
    {
      "epoch": 0.6316,
      "grad_norm": 0.0741092786192894,
      "learning_rate": 0.002,
      "loss": 1.9761992692947388,
      "step": 6316
    },
    {
      "epoch": 0.6318,
      "grad_norm": 0.078939288854599,
      "learning_rate": 0.002,
      "loss": 1.9875186681747437,
      "step": 6318
    },
    {
      "epoch": 0.632,
      "grad_norm": 0.07605218887329102,
      "learning_rate": 0.002,
      "loss": 2.156099796295166,
      "step": 6320
    },
    {
      "epoch": 0.6322,
      "grad_norm": 0.0793258473277092,
      "learning_rate": 0.002,
      "loss": 2.0915277004241943,
      "step": 6322
    },
    {
      "epoch": 0.6324,
      "grad_norm": 0.08433623611927032,
      "learning_rate": 0.002,
      "loss": 2.1180410385131836,
      "step": 6324
    },
    {
      "epoch": 0.6326,
      "grad_norm": 0.07935789972543716,
      "learning_rate": 0.002,
      "loss": 2.101377487182617,
      "step": 6326
    },
    {
      "epoch": 0.6328,
      "grad_norm": 0.07912053912878036,
      "learning_rate": 0.002,
      "loss": 2.000880002975464,
      "step": 6328
    },
    {
      "epoch": 0.633,
      "grad_norm": 0.08137912303209305,
      "learning_rate": 0.002,
      "loss": 2.125798225402832,
      "step": 6330
    },
    {
      "epoch": 0.6332,
      "grad_norm": 0.07988686859607697,
      "learning_rate": 0.002,
      "loss": 2.168882369995117,
      "step": 6332
    },
    {
      "epoch": 0.6334,
      "grad_norm": 0.07591064274311066,
      "learning_rate": 0.002,
      "loss": 2.038820266723633,
      "step": 6334
    },
    {
      "epoch": 0.6336,
      "grad_norm": 0.07940693199634552,
      "learning_rate": 0.002,
      "loss": 2.0537495613098145,
      "step": 6336
    },
    {
      "epoch": 0.6338,
      "grad_norm": 0.08221717178821564,
      "learning_rate": 0.002,
      "loss": 2.1510097980499268,
      "step": 6338
    },
    {
      "epoch": 0.634,
      "grad_norm": 0.07833975553512573,
      "learning_rate": 0.002,
      "loss": 2.048053741455078,
      "step": 6340
    },
    {
      "epoch": 0.6342,
      "grad_norm": 0.07710389047861099,
      "learning_rate": 0.002,
      "loss": 1.9866348505020142,
      "step": 6342
    },
    {
      "epoch": 0.6344,
      "grad_norm": 0.07424788922071457,
      "learning_rate": 0.002,
      "loss": 2.0938608646392822,
      "step": 6344
    },
    {
      "epoch": 0.6346,
      "grad_norm": 0.077277272939682,
      "learning_rate": 0.002,
      "loss": 2.0192911624908447,
      "step": 6346
    },
    {
      "epoch": 0.6348,
      "grad_norm": 0.08336544036865234,
      "learning_rate": 0.002,
      "loss": 2.0141971111297607,
      "step": 6348
    },
    {
      "epoch": 0.635,
      "grad_norm": 0.07780896127223969,
      "learning_rate": 0.002,
      "loss": 2.063798666000366,
      "step": 6350
    },
    {
      "epoch": 0.6352,
      "grad_norm": 0.0795053243637085,
      "learning_rate": 0.002,
      "loss": 2.075533151626587,
      "step": 6352
    },
    {
      "epoch": 0.6354,
      "grad_norm": 0.07965632528066635,
      "learning_rate": 0.002,
      "loss": 2.040285348892212,
      "step": 6354
    },
    {
      "epoch": 0.6356,
      "grad_norm": 0.07910115271806717,
      "learning_rate": 0.002,
      "loss": 2.174776792526245,
      "step": 6356
    },
    {
      "epoch": 0.6358,
      "grad_norm": 0.07868518680334091,
      "learning_rate": 0.002,
      "loss": 2.0381977558135986,
      "step": 6358
    },
    {
      "epoch": 0.636,
      "grad_norm": 0.07606080919504166,
      "learning_rate": 0.002,
      "loss": 2.071441173553467,
      "step": 6360
    },
    {
      "epoch": 0.6362,
      "grad_norm": 0.08462180197238922,
      "learning_rate": 0.002,
      "loss": 2.095954656600952,
      "step": 6362
    },
    {
      "epoch": 0.6364,
      "grad_norm": 0.0784149318933487,
      "learning_rate": 0.002,
      "loss": 1.9789042472839355,
      "step": 6364
    },
    {
      "epoch": 0.6366,
      "grad_norm": 0.0797652080655098,
      "learning_rate": 0.002,
      "loss": 2.098750352859497,
      "step": 6366
    },
    {
      "epoch": 0.6368,
      "grad_norm": 0.08523876965045929,
      "learning_rate": 0.002,
      "loss": 2.129258871078491,
      "step": 6368
    },
    {
      "epoch": 0.637,
      "grad_norm": 0.08182325959205627,
      "learning_rate": 0.002,
      "loss": 2.137357234954834,
      "step": 6370
    },
    {
      "epoch": 0.6372,
      "grad_norm": 0.07672186195850372,
      "learning_rate": 0.002,
      "loss": 2.077967405319214,
      "step": 6372
    },
    {
      "epoch": 0.6374,
      "grad_norm": 0.08270055800676346,
      "learning_rate": 0.002,
      "loss": 2.107825517654419,
      "step": 6374
    },
    {
      "epoch": 0.6376,
      "grad_norm": 0.07827715575695038,
      "learning_rate": 0.002,
      "loss": 2.2157142162323,
      "step": 6376
    },
    {
      "epoch": 0.6378,
      "grad_norm": 0.08016137033700943,
      "learning_rate": 0.002,
      "loss": 2.1667792797088623,
      "step": 6378
    },
    {
      "epoch": 0.638,
      "grad_norm": 0.07836662977933884,
      "learning_rate": 0.002,
      "loss": 2.1165359020233154,
      "step": 6380
    },
    {
      "epoch": 0.6382,
      "grad_norm": 0.07586970925331116,
      "learning_rate": 0.002,
      "loss": 2.1539933681488037,
      "step": 6382
    },
    {
      "epoch": 0.6384,
      "grad_norm": 0.07776647806167603,
      "learning_rate": 0.002,
      "loss": 2.138148784637451,
      "step": 6384
    },
    {
      "epoch": 0.6386,
      "grad_norm": 0.07965588569641113,
      "learning_rate": 0.002,
      "loss": 2.053962230682373,
      "step": 6386
    },
    {
      "epoch": 0.6388,
      "grad_norm": 0.07974972575902939,
      "learning_rate": 0.002,
      "loss": 2.0518341064453125,
      "step": 6388
    },
    {
      "epoch": 0.639,
      "grad_norm": 0.07535842806100845,
      "learning_rate": 0.002,
      "loss": 2.013820171356201,
      "step": 6390
    },
    {
      "epoch": 0.6392,
      "grad_norm": 0.08230482786893845,
      "learning_rate": 0.002,
      "loss": 2.0927724838256836,
      "step": 6392
    },
    {
      "epoch": 0.6394,
      "grad_norm": 0.07586695998907089,
      "learning_rate": 0.002,
      "loss": 2.0706088542938232,
      "step": 6394
    },
    {
      "epoch": 0.6396,
      "grad_norm": 0.07515781372785568,
      "learning_rate": 0.002,
      "loss": 2.115168333053589,
      "step": 6396
    },
    {
      "epoch": 0.6398,
      "grad_norm": 0.07670266926288605,
      "learning_rate": 0.002,
      "loss": 2.1552343368530273,
      "step": 6398
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.07923460006713867,
      "learning_rate": 0.002,
      "loss": 2.092106342315674,
      "step": 6400
    },
    {
      "epoch": 0.6402,
      "grad_norm": 0.07980695366859436,
      "learning_rate": 0.002,
      "loss": 2.078113079071045,
      "step": 6402
    },
    {
      "epoch": 0.6404,
      "grad_norm": 0.07607570290565491,
      "learning_rate": 0.002,
      "loss": 2.085761308670044,
      "step": 6404
    },
    {
      "epoch": 0.6406,
      "grad_norm": 0.07860725373029709,
      "learning_rate": 0.002,
      "loss": 2.123821973800659,
      "step": 6406
    },
    {
      "epoch": 0.6408,
      "grad_norm": 0.07913851737976074,
      "learning_rate": 0.002,
      "loss": 2.182516098022461,
      "step": 6408
    },
    {
      "epoch": 0.641,
      "grad_norm": 0.07836799323558807,
      "learning_rate": 0.002,
      "loss": 2.0663161277770996,
      "step": 6410
    },
    {
      "epoch": 0.6412,
      "grad_norm": 0.0834578424692154,
      "learning_rate": 0.002,
      "loss": 2.0186104774475098,
      "step": 6412
    },
    {
      "epoch": 0.6414,
      "grad_norm": 0.08262518793344498,
      "learning_rate": 0.002,
      "loss": 2.23996639251709,
      "step": 6414
    },
    {
      "epoch": 0.6416,
      "grad_norm": 0.08002250641584396,
      "learning_rate": 0.002,
      "loss": 2.0523629188537598,
      "step": 6416
    },
    {
      "epoch": 0.6418,
      "grad_norm": 0.07926962524652481,
      "learning_rate": 0.002,
      "loss": 2.075770616531372,
      "step": 6418
    },
    {
      "epoch": 0.642,
      "grad_norm": 0.08266501128673553,
      "learning_rate": 0.002,
      "loss": 2.092263698577881,
      "step": 6420
    },
    {
      "epoch": 0.6422,
      "grad_norm": 0.07709616422653198,
      "learning_rate": 0.002,
      "loss": 2.134772300720215,
      "step": 6422
    },
    {
      "epoch": 0.6424,
      "grad_norm": 0.07990540564060211,
      "learning_rate": 0.002,
      "loss": 2.120889663696289,
      "step": 6424
    },
    {
      "epoch": 0.6426,
      "grad_norm": 0.07253920286893845,
      "learning_rate": 0.002,
      "loss": 2.017014503479004,
      "step": 6426
    },
    {
      "epoch": 0.6428,
      "grad_norm": 0.07695186883211136,
      "learning_rate": 0.002,
      "loss": 2.106471300125122,
      "step": 6428
    },
    {
      "epoch": 0.643,
      "grad_norm": 0.07851019501686096,
      "learning_rate": 0.002,
      "loss": 2.1396830081939697,
      "step": 6430
    },
    {
      "epoch": 0.6432,
      "grad_norm": 0.08103150874376297,
      "learning_rate": 0.002,
      "loss": 2.1599340438842773,
      "step": 6432
    },
    {
      "epoch": 0.6434,
      "grad_norm": 0.07607518136501312,
      "learning_rate": 0.002,
      "loss": 2.104358196258545,
      "step": 6434
    },
    {
      "epoch": 0.6436,
      "grad_norm": 0.07754198461771011,
      "learning_rate": 0.002,
      "loss": 2.010542392730713,
      "step": 6436
    },
    {
      "epoch": 0.6438,
      "grad_norm": 0.07896433770656586,
      "learning_rate": 0.002,
      "loss": 2.077421188354492,
      "step": 6438
    },
    {
      "epoch": 0.644,
      "grad_norm": 0.0790986567735672,
      "learning_rate": 0.002,
      "loss": 2.0161118507385254,
      "step": 6440
    },
    {
      "epoch": 0.6442,
      "grad_norm": 0.07831224799156189,
      "learning_rate": 0.002,
      "loss": 2.0093750953674316,
      "step": 6442
    },
    {
      "epoch": 0.6444,
      "grad_norm": 0.08881847560405731,
      "learning_rate": 0.002,
      "loss": 2.16497540473938,
      "step": 6444
    },
    {
      "epoch": 0.6446,
      "grad_norm": 0.07842765003442764,
      "learning_rate": 0.002,
      "loss": 2.139080286026001,
      "step": 6446
    },
    {
      "epoch": 0.6448,
      "grad_norm": 0.08505600690841675,
      "learning_rate": 0.002,
      "loss": 2.0738472938537598,
      "step": 6448
    },
    {
      "epoch": 0.645,
      "grad_norm": 0.07446635514497757,
      "learning_rate": 0.002,
      "loss": 2.0326597690582275,
      "step": 6450
    },
    {
      "epoch": 0.6452,
      "grad_norm": 0.08120739459991455,
      "learning_rate": 0.002,
      "loss": 2.097571611404419,
      "step": 6452
    },
    {
      "epoch": 0.6454,
      "grad_norm": 0.07566303759813309,
      "learning_rate": 0.002,
      "loss": 2.033074378967285,
      "step": 6454
    },
    {
      "epoch": 0.6456,
      "grad_norm": 0.07899384945631027,
      "learning_rate": 0.002,
      "loss": 2.1237990856170654,
      "step": 6456
    },
    {
      "epoch": 0.6458,
      "grad_norm": 0.0750228762626648,
      "learning_rate": 0.002,
      "loss": 2.0201642513275146,
      "step": 6458
    },
    {
      "epoch": 0.646,
      "grad_norm": 0.07521556317806244,
      "learning_rate": 0.002,
      "loss": 2.0939197540283203,
      "step": 6460
    },
    {
      "epoch": 0.6462,
      "grad_norm": 0.08067812025547028,
      "learning_rate": 0.002,
      "loss": 2.1319520473480225,
      "step": 6462
    },
    {
      "epoch": 0.6464,
      "grad_norm": 0.08662962168455124,
      "learning_rate": 0.002,
      "loss": 2.1739039421081543,
      "step": 6464
    },
    {
      "epoch": 0.6466,
      "grad_norm": 0.07496922463178635,
      "learning_rate": 0.002,
      "loss": 2.077181577682495,
      "step": 6466
    },
    {
      "epoch": 0.6468,
      "grad_norm": 0.07686156779527664,
      "learning_rate": 0.002,
      "loss": 2.006990909576416,
      "step": 6468
    },
    {
      "epoch": 0.647,
      "grad_norm": 0.08061195909976959,
      "learning_rate": 0.002,
      "loss": 2.1158692836761475,
      "step": 6470
    },
    {
      "epoch": 0.6472,
      "grad_norm": 0.07470306009054184,
      "learning_rate": 0.002,
      "loss": 2.1185874938964844,
      "step": 6472
    },
    {
      "epoch": 0.6474,
      "grad_norm": 0.08287925273180008,
      "learning_rate": 0.002,
      "loss": 2.075343608856201,
      "step": 6474
    },
    {
      "epoch": 0.6476,
      "grad_norm": 0.08257737755775452,
      "learning_rate": 0.002,
      "loss": 2.1517367362976074,
      "step": 6476
    },
    {
      "epoch": 0.6478,
      "grad_norm": 0.07866178452968597,
      "learning_rate": 0.002,
      "loss": 2.0972976684570312,
      "step": 6478
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.08271392434835434,
      "learning_rate": 0.002,
      "loss": 2.0156025886535645,
      "step": 6480
    },
    {
      "epoch": 0.6482,
      "grad_norm": 0.07995027303695679,
      "learning_rate": 0.002,
      "loss": 1.972216010093689,
      "step": 6482
    },
    {
      "epoch": 0.6484,
      "grad_norm": 0.07695799320936203,
      "learning_rate": 0.002,
      "loss": 2.056699752807617,
      "step": 6484
    },
    {
      "epoch": 0.6486,
      "grad_norm": 0.08049394935369492,
      "learning_rate": 0.002,
      "loss": 2.1031994819641113,
      "step": 6486
    },
    {
      "epoch": 0.6488,
      "grad_norm": 0.07892326265573502,
      "learning_rate": 0.002,
      "loss": 2.0755345821380615,
      "step": 6488
    },
    {
      "epoch": 0.649,
      "grad_norm": 0.07889028638601303,
      "learning_rate": 0.002,
      "loss": 2.205625295639038,
      "step": 6490
    },
    {
      "epoch": 0.6492,
      "grad_norm": 0.07959353923797607,
      "learning_rate": 0.002,
      "loss": 2.203972101211548,
      "step": 6492
    },
    {
      "epoch": 0.6494,
      "grad_norm": 0.082419753074646,
      "learning_rate": 0.002,
      "loss": 2.149407386779785,
      "step": 6494
    },
    {
      "epoch": 0.6496,
      "grad_norm": 0.08376852422952652,
      "learning_rate": 0.002,
      "loss": 2.272080421447754,
      "step": 6496
    },
    {
      "epoch": 0.6498,
      "grad_norm": 0.0781112089753151,
      "learning_rate": 0.002,
      "loss": 2.095327377319336,
      "step": 6498
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.07797998189926147,
      "learning_rate": 0.002,
      "loss": 2.072518825531006,
      "step": 6500
    },
    {
      "epoch": 0.6502,
      "grad_norm": 0.08415030688047409,
      "learning_rate": 0.002,
      "loss": 2.065596103668213,
      "step": 6502
    },
    {
      "epoch": 0.6504,
      "grad_norm": 0.08138847351074219,
      "learning_rate": 0.002,
      "loss": 2.18147611618042,
      "step": 6504
    },
    {
      "epoch": 0.6506,
      "grad_norm": 0.07876664400100708,
      "learning_rate": 0.002,
      "loss": 2.047252655029297,
      "step": 6506
    },
    {
      "epoch": 0.6508,
      "grad_norm": 0.07782850414514542,
      "learning_rate": 0.002,
      "loss": 2.08225154876709,
      "step": 6508
    },
    {
      "epoch": 0.651,
      "grad_norm": 0.07858437299728394,
      "learning_rate": 0.002,
      "loss": 2.1112160682678223,
      "step": 6510
    },
    {
      "epoch": 0.6512,
      "grad_norm": 0.07937879860401154,
      "learning_rate": 0.002,
      "loss": 2.03151535987854,
      "step": 6512
    },
    {
      "epoch": 0.6514,
      "grad_norm": 0.07974623143672943,
      "learning_rate": 0.002,
      "loss": 2.111650228500366,
      "step": 6514
    },
    {
      "epoch": 0.6516,
      "grad_norm": 0.07538212090730667,
      "learning_rate": 0.002,
      "loss": 2.0029265880584717,
      "step": 6516
    },
    {
      "epoch": 0.6518,
      "grad_norm": 0.08014639467000961,
      "learning_rate": 0.002,
      "loss": 2.1368460655212402,
      "step": 6518
    },
    {
      "epoch": 0.652,
      "grad_norm": 0.07945128530263901,
      "learning_rate": 0.002,
      "loss": 2.024014949798584,
      "step": 6520
    },
    {
      "epoch": 0.6522,
      "grad_norm": 0.07821159809827805,
      "learning_rate": 0.002,
      "loss": 1.990460991859436,
      "step": 6522
    },
    {
      "epoch": 0.6524,
      "grad_norm": 0.07692032307386398,
      "learning_rate": 0.002,
      "loss": 2.133155345916748,
      "step": 6524
    },
    {
      "epoch": 0.6526,
      "grad_norm": 0.07701855897903442,
      "learning_rate": 0.002,
      "loss": 2.0780484676361084,
      "step": 6526
    },
    {
      "epoch": 0.6528,
      "grad_norm": 0.07764878123998642,
      "learning_rate": 0.002,
      "loss": 2.0968637466430664,
      "step": 6528
    },
    {
      "epoch": 0.653,
      "grad_norm": 0.08185577392578125,
      "learning_rate": 0.002,
      "loss": 2.1542704105377197,
      "step": 6530
    },
    {
      "epoch": 0.6532,
      "grad_norm": 0.07941040396690369,
      "learning_rate": 0.002,
      "loss": 2.0647494792938232,
      "step": 6532
    },
    {
      "epoch": 0.6534,
      "grad_norm": 0.08077459037303925,
      "learning_rate": 0.002,
      "loss": 2.127591133117676,
      "step": 6534
    },
    {
      "epoch": 0.6536,
      "grad_norm": 0.0807715579867363,
      "learning_rate": 0.002,
      "loss": 2.13081955909729,
      "step": 6536
    },
    {
      "epoch": 0.6538,
      "grad_norm": 0.07980871945619583,
      "learning_rate": 0.002,
      "loss": 2.0260775089263916,
      "step": 6538
    },
    {
      "epoch": 0.654,
      "grad_norm": 0.08008255064487457,
      "learning_rate": 0.002,
      "loss": 2.0623464584350586,
      "step": 6540
    },
    {
      "epoch": 0.6542,
      "grad_norm": 0.08051558583974838,
      "learning_rate": 0.002,
      "loss": 2.1031434535980225,
      "step": 6542
    },
    {
      "epoch": 0.6544,
      "grad_norm": 0.07545653730630875,
      "learning_rate": 0.002,
      "loss": 2.1418874263763428,
      "step": 6544
    },
    {
      "epoch": 0.6546,
      "grad_norm": 0.08669177442789078,
      "learning_rate": 0.002,
      "loss": 2.080371856689453,
      "step": 6546
    },
    {
      "epoch": 0.6548,
      "grad_norm": 0.08440108597278595,
      "learning_rate": 0.002,
      "loss": 2.120692014694214,
      "step": 6548
    },
    {
      "epoch": 0.655,
      "grad_norm": 0.07966786623001099,
      "learning_rate": 0.002,
      "loss": 2.024085760116577,
      "step": 6550
    },
    {
      "epoch": 0.6552,
      "grad_norm": 0.08144734054803848,
      "learning_rate": 0.002,
      "loss": 2.1305243968963623,
      "step": 6552
    },
    {
      "epoch": 0.6554,
      "grad_norm": 0.08255793154239655,
      "learning_rate": 0.002,
      "loss": 2.116543769836426,
      "step": 6554
    },
    {
      "epoch": 0.6556,
      "grad_norm": 0.07750437408685684,
      "learning_rate": 0.002,
      "loss": 2.1149990558624268,
      "step": 6556
    },
    {
      "epoch": 0.6558,
      "grad_norm": 0.07970361411571503,
      "learning_rate": 0.002,
      "loss": 2.087726593017578,
      "step": 6558
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.08013636618852615,
      "learning_rate": 0.002,
      "loss": 2.0655534267425537,
      "step": 6560
    },
    {
      "epoch": 0.6562,
      "grad_norm": 0.0809660255908966,
      "learning_rate": 0.002,
      "loss": 2.1279449462890625,
      "step": 6562
    },
    {
      "epoch": 0.6564,
      "grad_norm": 0.07750412076711655,
      "learning_rate": 0.002,
      "loss": 2.1799957752227783,
      "step": 6564
    },
    {
      "epoch": 0.6566,
      "grad_norm": 0.0787065327167511,
      "learning_rate": 0.002,
      "loss": 2.0964736938476562,
      "step": 6566
    },
    {
      "epoch": 0.6568,
      "grad_norm": 0.0761215090751648,
      "learning_rate": 0.002,
      "loss": 2.151193857192993,
      "step": 6568
    },
    {
      "epoch": 0.657,
      "grad_norm": 0.08338164538145065,
      "learning_rate": 0.002,
      "loss": 1.9827766418457031,
      "step": 6570
    },
    {
      "epoch": 0.6572,
      "grad_norm": 0.07824510335922241,
      "learning_rate": 0.002,
      "loss": 2.132720708847046,
      "step": 6572
    },
    {
      "epoch": 0.6574,
      "grad_norm": 0.07794688642024994,
      "learning_rate": 0.002,
      "loss": 2.156541347503662,
      "step": 6574
    },
    {
      "epoch": 0.6576,
      "grad_norm": 0.07615967094898224,
      "learning_rate": 0.002,
      "loss": 2.116481304168701,
      "step": 6576
    },
    {
      "epoch": 0.6578,
      "grad_norm": 0.08315353840589523,
      "learning_rate": 0.002,
      "loss": 2.136971950531006,
      "step": 6578
    },
    {
      "epoch": 0.658,
      "grad_norm": 0.0844867154955864,
      "learning_rate": 0.002,
      "loss": 2.0933356285095215,
      "step": 6580
    },
    {
      "epoch": 0.6582,
      "grad_norm": 0.08478633314371109,
      "learning_rate": 0.002,
      "loss": 2.1107678413391113,
      "step": 6582
    },
    {
      "epoch": 0.6584,
      "grad_norm": 0.07703419029712677,
      "learning_rate": 0.002,
      "loss": 2.1739909648895264,
      "step": 6584
    },
    {
      "epoch": 0.6586,
      "grad_norm": 0.08493146300315857,
      "learning_rate": 0.002,
      "loss": 2.1060054302215576,
      "step": 6586
    },
    {
      "epoch": 0.6588,
      "grad_norm": 0.07960671931505203,
      "learning_rate": 0.002,
      "loss": 2.0868608951568604,
      "step": 6588
    },
    {
      "epoch": 0.659,
      "grad_norm": 0.08494608849287033,
      "learning_rate": 0.002,
      "loss": 2.1539337635040283,
      "step": 6590
    },
    {
      "epoch": 0.6592,
      "grad_norm": 0.07570089399814606,
      "learning_rate": 0.002,
      "loss": 2.1089937686920166,
      "step": 6592
    },
    {
      "epoch": 0.6594,
      "grad_norm": 0.07643125206232071,
      "learning_rate": 0.002,
      "loss": 2.175563097000122,
      "step": 6594
    },
    {
      "epoch": 0.6596,
      "grad_norm": 0.07879031449556351,
      "learning_rate": 0.002,
      "loss": 2.1425094604492188,
      "step": 6596
    },
    {
      "epoch": 0.6598,
      "grad_norm": 0.07939176261425018,
      "learning_rate": 0.002,
      "loss": 2.1987900733947754,
      "step": 6598
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.07637582719326019,
      "learning_rate": 0.002,
      "loss": 2.1768345832824707,
      "step": 6600
    },
    {
      "epoch": 0.6602,
      "grad_norm": 0.08722255378961563,
      "learning_rate": 0.002,
      "loss": 2.1757709980010986,
      "step": 6602
    },
    {
      "epoch": 0.6604,
      "grad_norm": 0.07901013642549515,
      "learning_rate": 0.002,
      "loss": 2.153095245361328,
      "step": 6604
    },
    {
      "epoch": 0.6606,
      "grad_norm": 0.07598739117383957,
      "learning_rate": 0.002,
      "loss": 2.0875966548919678,
      "step": 6606
    },
    {
      "epoch": 0.6608,
      "grad_norm": 0.08540836721658707,
      "learning_rate": 0.002,
      "loss": 2.1106789112091064,
      "step": 6608
    },
    {
      "epoch": 0.661,
      "grad_norm": 0.07943204790353775,
      "learning_rate": 0.002,
      "loss": 2.0666470527648926,
      "step": 6610
    },
    {
      "epoch": 0.6612,
      "grad_norm": 0.07779556512832642,
      "learning_rate": 0.002,
      "loss": 2.1327366828918457,
      "step": 6612
    },
    {
      "epoch": 0.6614,
      "grad_norm": 0.07549969106912613,
      "learning_rate": 0.002,
      "loss": 2.108412504196167,
      "step": 6614
    },
    {
      "epoch": 0.6616,
      "grad_norm": 0.07703448086977005,
      "learning_rate": 0.002,
      "loss": 2.1468734741210938,
      "step": 6616
    },
    {
      "epoch": 0.6618,
      "grad_norm": 0.07824315875768661,
      "learning_rate": 0.002,
      "loss": 2.022526264190674,
      "step": 6618
    },
    {
      "epoch": 0.662,
      "grad_norm": 0.07603824883699417,
      "learning_rate": 0.002,
      "loss": 2.052356481552124,
      "step": 6620
    },
    {
      "epoch": 0.6622,
      "grad_norm": 0.09060277044773102,
      "learning_rate": 0.002,
      "loss": 2.0901589393615723,
      "step": 6622
    },
    {
      "epoch": 0.6624,
      "grad_norm": 0.07728119194507599,
      "learning_rate": 0.002,
      "loss": 2.064678192138672,
      "step": 6624
    },
    {
      "epoch": 0.6626,
      "grad_norm": 0.078868068754673,
      "learning_rate": 0.002,
      "loss": 2.1663787364959717,
      "step": 6626
    },
    {
      "epoch": 0.6628,
      "grad_norm": 0.07827866822481155,
      "learning_rate": 0.002,
      "loss": 2.165571689605713,
      "step": 6628
    },
    {
      "epoch": 0.663,
      "grad_norm": 0.08049779385328293,
      "learning_rate": 0.002,
      "loss": 2.0970492362976074,
      "step": 6630
    },
    {
      "epoch": 0.6632,
      "grad_norm": 0.07845597714185715,
      "learning_rate": 0.002,
      "loss": 2.02058482170105,
      "step": 6632
    },
    {
      "epoch": 0.6634,
      "grad_norm": 0.07910637557506561,
      "learning_rate": 0.002,
      "loss": 2.064495086669922,
      "step": 6634
    },
    {
      "epoch": 0.6636,
      "grad_norm": 0.07624828070402145,
      "learning_rate": 0.002,
      "loss": 2.065011501312256,
      "step": 6636
    },
    {
      "epoch": 0.6638,
      "grad_norm": 0.0797739326953888,
      "learning_rate": 0.002,
      "loss": 2.1118757724761963,
      "step": 6638
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.07718904316425323,
      "learning_rate": 0.002,
      "loss": 2.0282068252563477,
      "step": 6640
    },
    {
      "epoch": 0.6642,
      "grad_norm": 0.07780918478965759,
      "learning_rate": 0.002,
      "loss": 2.08194899559021,
      "step": 6642
    },
    {
      "epoch": 0.6644,
      "grad_norm": 0.08049635589122772,
      "learning_rate": 0.002,
      "loss": 2.0526134967803955,
      "step": 6644
    },
    {
      "epoch": 0.6646,
      "grad_norm": 0.07903172820806503,
      "learning_rate": 0.002,
      "loss": 2.1841306686401367,
      "step": 6646
    },
    {
      "epoch": 0.6648,
      "grad_norm": 0.07925070822238922,
      "learning_rate": 0.002,
      "loss": 2.1138362884521484,
      "step": 6648
    },
    {
      "epoch": 0.665,
      "grad_norm": 0.08087515085935593,
      "learning_rate": 0.002,
      "loss": 2.139361619949341,
      "step": 6650
    },
    {
      "epoch": 0.6652,
      "grad_norm": 0.07751813530921936,
      "learning_rate": 0.002,
      "loss": 2.127671480178833,
      "step": 6652
    },
    {
      "epoch": 0.6654,
      "grad_norm": 0.07779194414615631,
      "learning_rate": 0.002,
      "loss": 2.1437084674835205,
      "step": 6654
    },
    {
      "epoch": 0.6656,
      "grad_norm": 0.07840143144130707,
      "learning_rate": 0.002,
      "loss": 2.1765902042388916,
      "step": 6656
    },
    {
      "epoch": 0.6658,
      "grad_norm": 0.07842352241277695,
      "learning_rate": 0.002,
      "loss": 2.005610227584839,
      "step": 6658
    },
    {
      "epoch": 0.666,
      "grad_norm": 0.07944174855947495,
      "learning_rate": 0.002,
      "loss": 2.141171455383301,
      "step": 6660
    },
    {
      "epoch": 0.6662,
      "grad_norm": 0.07824776321649551,
      "learning_rate": 0.002,
      "loss": 2.1186180114746094,
      "step": 6662
    },
    {
      "epoch": 0.6664,
      "grad_norm": 0.07842502743005753,
      "learning_rate": 0.002,
      "loss": 2.0858466625213623,
      "step": 6664
    },
    {
      "epoch": 0.6666,
      "grad_norm": 0.08217931538820267,
      "learning_rate": 0.002,
      "loss": 2.104302167892456,
      "step": 6666
    },
    {
      "epoch": 0.6668,
      "grad_norm": 0.07798385620117188,
      "learning_rate": 0.002,
      "loss": 2.116572856903076,
      "step": 6668
    },
    {
      "epoch": 0.667,
      "grad_norm": 0.0794372633099556,
      "learning_rate": 0.002,
      "loss": 2.0075936317443848,
      "step": 6670
    },
    {
      "epoch": 0.6672,
      "grad_norm": 0.07847985625267029,
      "learning_rate": 0.002,
      "loss": 2.0762343406677246,
      "step": 6672
    },
    {
      "epoch": 0.6674,
      "grad_norm": 0.07526922225952148,
      "learning_rate": 0.002,
      "loss": 2.1034657955169678,
      "step": 6674
    },
    {
      "epoch": 0.6676,
      "grad_norm": 0.08032264560461044,
      "learning_rate": 0.002,
      "loss": 2.1657979488372803,
      "step": 6676
    },
    {
      "epoch": 0.6678,
      "grad_norm": 0.07831133157014847,
      "learning_rate": 0.002,
      "loss": 2.1738650798797607,
      "step": 6678
    },
    {
      "epoch": 0.668,
      "grad_norm": 0.08404078334569931,
      "learning_rate": 0.002,
      "loss": 2.146841049194336,
      "step": 6680
    },
    {
      "epoch": 0.6682,
      "grad_norm": 0.07827918976545334,
      "learning_rate": 0.002,
      "loss": 2.0738189220428467,
      "step": 6682
    },
    {
      "epoch": 0.6684,
      "grad_norm": 0.0790071114897728,
      "learning_rate": 0.002,
      "loss": 2.057255983352661,
      "step": 6684
    },
    {
      "epoch": 0.6686,
      "grad_norm": 0.07858891785144806,
      "learning_rate": 0.002,
      "loss": 2.163315773010254,
      "step": 6686
    },
    {
      "epoch": 0.6688,
      "grad_norm": 0.0750180259346962,
      "learning_rate": 0.002,
      "loss": 2.1382243633270264,
      "step": 6688
    },
    {
      "epoch": 0.669,
      "grad_norm": 0.08001725375652313,
      "learning_rate": 0.002,
      "loss": 2.1780824661254883,
      "step": 6690
    },
    {
      "epoch": 0.6692,
      "grad_norm": 0.07856681942939758,
      "learning_rate": 0.002,
      "loss": 2.0685181617736816,
      "step": 6692
    },
    {
      "epoch": 0.6694,
      "grad_norm": 0.0782242938876152,
      "learning_rate": 0.002,
      "loss": 2.038634777069092,
      "step": 6694
    },
    {
      "epoch": 0.6696,
      "grad_norm": 0.08130025118589401,
      "learning_rate": 0.002,
      "loss": 2.1031668186187744,
      "step": 6696
    },
    {
      "epoch": 0.6698,
      "grad_norm": 0.0845252051949501,
      "learning_rate": 0.002,
      "loss": 2.1156182289123535,
      "step": 6698
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.07879341393709183,
      "learning_rate": 0.002,
      "loss": 2.061433792114258,
      "step": 6700
    },
    {
      "epoch": 0.6702,
      "grad_norm": 0.08086971938610077,
      "learning_rate": 0.002,
      "loss": 2.149674892425537,
      "step": 6702
    },
    {
      "epoch": 0.6704,
      "grad_norm": 0.07791486382484436,
      "learning_rate": 0.002,
      "loss": 2.098773956298828,
      "step": 6704
    },
    {
      "epoch": 0.6706,
      "grad_norm": 0.07951268553733826,
      "learning_rate": 0.002,
      "loss": 2.065009832382202,
      "step": 6706
    },
    {
      "epoch": 0.6708,
      "grad_norm": 0.07674980908632278,
      "learning_rate": 0.002,
      "loss": 2.1336724758148193,
      "step": 6708
    },
    {
      "epoch": 0.671,
      "grad_norm": 0.07922488451004028,
      "learning_rate": 0.002,
      "loss": 2.1241140365600586,
      "step": 6710
    },
    {
      "epoch": 0.6712,
      "grad_norm": 0.07849503308534622,
      "learning_rate": 0.002,
      "loss": 2.112231731414795,
      "step": 6712
    },
    {
      "epoch": 0.6714,
      "grad_norm": 0.07710650563240051,
      "learning_rate": 0.002,
      "loss": 2.0838258266448975,
      "step": 6714
    },
    {
      "epoch": 0.6716,
      "grad_norm": 0.08649444580078125,
      "learning_rate": 0.002,
      "loss": 2.1255881786346436,
      "step": 6716
    },
    {
      "epoch": 0.6718,
      "grad_norm": 0.07987397909164429,
      "learning_rate": 0.002,
      "loss": 2.1667730808258057,
      "step": 6718
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.08292744308710098,
      "learning_rate": 0.002,
      "loss": 2.1402525901794434,
      "step": 6720
    },
    {
      "epoch": 0.6722,
      "grad_norm": 0.08481664955615997,
      "learning_rate": 0.002,
      "loss": 2.133803129196167,
      "step": 6722
    },
    {
      "epoch": 0.6724,
      "grad_norm": 0.07912091910839081,
      "learning_rate": 0.002,
      "loss": 2.1656646728515625,
      "step": 6724
    },
    {
      "epoch": 0.6726,
      "grad_norm": 0.07910934835672379,
      "learning_rate": 0.002,
      "loss": 2.1418509483337402,
      "step": 6726
    },
    {
      "epoch": 0.6728,
      "grad_norm": 0.07801549136638641,
      "learning_rate": 0.002,
      "loss": 2.2024412155151367,
      "step": 6728
    },
    {
      "epoch": 0.673,
      "grad_norm": 0.08280784636735916,
      "learning_rate": 0.002,
      "loss": 2.1019856929779053,
      "step": 6730
    },
    {
      "epoch": 0.6732,
      "grad_norm": 0.07962936162948608,
      "learning_rate": 0.002,
      "loss": 2.153700113296509,
      "step": 6732
    },
    {
      "epoch": 0.6734,
      "grad_norm": 0.077731192111969,
      "learning_rate": 0.002,
      "loss": 2.1404550075531006,
      "step": 6734
    },
    {
      "epoch": 0.6736,
      "grad_norm": 0.08247546851634979,
      "learning_rate": 0.002,
      "loss": 2.1138663291931152,
      "step": 6736
    },
    {
      "epoch": 0.6738,
      "grad_norm": 0.07393250614404678,
      "learning_rate": 0.002,
      "loss": 2.0810272693634033,
      "step": 6738
    },
    {
      "epoch": 0.674,
      "grad_norm": 0.07841020077466965,
      "learning_rate": 0.002,
      "loss": 2.1343798637390137,
      "step": 6740
    },
    {
      "epoch": 0.6742,
      "grad_norm": 0.07979895919561386,
      "learning_rate": 0.002,
      "loss": 2.029613494873047,
      "step": 6742
    },
    {
      "epoch": 0.6744,
      "grad_norm": 0.08056328445672989,
      "learning_rate": 0.002,
      "loss": 2.1792783737182617,
      "step": 6744
    },
    {
      "epoch": 0.6746,
      "grad_norm": 0.08267771452665329,
      "learning_rate": 0.002,
      "loss": 2.086555242538452,
      "step": 6746
    },
    {
      "epoch": 0.6748,
      "grad_norm": 0.07674989104270935,
      "learning_rate": 0.002,
      "loss": 2.0839288234710693,
      "step": 6748
    },
    {
      "epoch": 0.675,
      "grad_norm": 0.07894333451986313,
      "learning_rate": 0.002,
      "loss": 2.200251817703247,
      "step": 6750
    },
    {
      "epoch": 0.6752,
      "grad_norm": 0.08064015209674835,
      "learning_rate": 0.002,
      "loss": 2.1748483180999756,
      "step": 6752
    },
    {
      "epoch": 0.6754,
      "grad_norm": 0.08038888871669769,
      "learning_rate": 0.002,
      "loss": 2.0680992603302,
      "step": 6754
    },
    {
      "epoch": 0.6756,
      "grad_norm": 0.07655466347932816,
      "learning_rate": 0.002,
      "loss": 2.175581455230713,
      "step": 6756
    },
    {
      "epoch": 0.6758,
      "grad_norm": 0.0820341631770134,
      "learning_rate": 0.002,
      "loss": 2.12591290473938,
      "step": 6758
    },
    {
      "epoch": 0.676,
      "grad_norm": 0.07611099630594254,
      "learning_rate": 0.002,
      "loss": 2.100398540496826,
      "step": 6760
    },
    {
      "epoch": 0.6762,
      "grad_norm": 0.07582495361566544,
      "learning_rate": 0.002,
      "loss": 2.029967784881592,
      "step": 6762
    },
    {
      "epoch": 0.6764,
      "grad_norm": 0.07983488589525223,
      "learning_rate": 0.002,
      "loss": 2.2079646587371826,
      "step": 6764
    },
    {
      "epoch": 0.6766,
      "grad_norm": 0.07795756310224533,
      "learning_rate": 0.002,
      "loss": 2.104626417160034,
      "step": 6766
    },
    {
      "epoch": 0.6768,
      "grad_norm": 0.08214631676673889,
      "learning_rate": 0.002,
      "loss": 2.1289432048797607,
      "step": 6768
    },
    {
      "epoch": 0.677,
      "grad_norm": 0.07846441864967346,
      "learning_rate": 0.002,
      "loss": 2.002681255340576,
      "step": 6770
    },
    {
      "epoch": 0.6772,
      "grad_norm": 0.08085110038518906,
      "learning_rate": 0.002,
      "loss": 2.041501522064209,
      "step": 6772
    },
    {
      "epoch": 0.6774,
      "grad_norm": 0.07693544030189514,
      "learning_rate": 0.002,
      "loss": 2.085850238800049,
      "step": 6774
    },
    {
      "epoch": 0.6776,
      "grad_norm": 0.07854898273944855,
      "learning_rate": 0.002,
      "loss": 2.2011220455169678,
      "step": 6776
    },
    {
      "epoch": 0.6778,
      "grad_norm": 0.07713020592927933,
      "learning_rate": 0.002,
      "loss": 2.13450026512146,
      "step": 6778
    },
    {
      "epoch": 0.678,
      "grad_norm": 0.08457133919000626,
      "learning_rate": 0.002,
      "loss": 2.18753981590271,
      "step": 6780
    },
    {
      "epoch": 0.6782,
      "grad_norm": 0.07638809829950333,
      "learning_rate": 0.002,
      "loss": 2.1472365856170654,
      "step": 6782
    },
    {
      "epoch": 0.6784,
      "grad_norm": 0.08244199305772781,
      "learning_rate": 0.002,
      "loss": 2.145637273788452,
      "step": 6784
    },
    {
      "epoch": 0.6786,
      "grad_norm": 0.08261250704526901,
      "learning_rate": 0.002,
      "loss": 2.127143383026123,
      "step": 6786
    },
    {
      "epoch": 0.6788,
      "grad_norm": 0.08170846104621887,
      "learning_rate": 0.002,
      "loss": 2.100860834121704,
      "step": 6788
    },
    {
      "epoch": 0.679,
      "grad_norm": 0.07678824663162231,
      "learning_rate": 0.002,
      "loss": 2.1533100605010986,
      "step": 6790
    },
    {
      "epoch": 0.6792,
      "grad_norm": 0.0754859670996666,
      "learning_rate": 0.002,
      "loss": 2.1637673377990723,
      "step": 6792
    },
    {
      "epoch": 0.6794,
      "grad_norm": 0.08017821609973907,
      "learning_rate": 0.002,
      "loss": 2.151637315750122,
      "step": 6794
    },
    {
      "epoch": 0.6796,
      "grad_norm": 0.0774381160736084,
      "learning_rate": 0.002,
      "loss": 2.18253493309021,
      "step": 6796
    },
    {
      "epoch": 0.6798,
      "grad_norm": 0.07651591300964355,
      "learning_rate": 0.002,
      "loss": 2.148949384689331,
      "step": 6798
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.0777880921959877,
      "learning_rate": 0.002,
      "loss": 2.159592628479004,
      "step": 6800
    },
    {
      "epoch": 0.6802,
      "grad_norm": 0.07918092608451843,
      "learning_rate": 0.002,
      "loss": 2.112189531326294,
      "step": 6802
    },
    {
      "epoch": 0.6804,
      "grad_norm": 0.07968147099018097,
      "learning_rate": 0.002,
      "loss": 2.0612668991088867,
      "step": 6804
    },
    {
      "epoch": 0.6806,
      "grad_norm": 0.08113916963338852,
      "learning_rate": 0.002,
      "loss": 2.1662235260009766,
      "step": 6806
    },
    {
      "epoch": 0.6808,
      "grad_norm": 0.07584865391254425,
      "learning_rate": 0.002,
      "loss": 2.112905502319336,
      "step": 6808
    },
    {
      "epoch": 0.681,
      "grad_norm": 0.08394674956798553,
      "learning_rate": 0.002,
      "loss": 2.154540777206421,
      "step": 6810
    },
    {
      "epoch": 0.6812,
      "grad_norm": 0.08504300564527512,
      "learning_rate": 0.002,
      "loss": 2.2262561321258545,
      "step": 6812
    },
    {
      "epoch": 0.6814,
      "grad_norm": 0.07882723212242126,
      "learning_rate": 0.002,
      "loss": 2.229646921157837,
      "step": 6814
    },
    {
      "epoch": 0.6816,
      "grad_norm": 0.07806238532066345,
      "learning_rate": 0.002,
      "loss": 2.1390786170959473,
      "step": 6816
    },
    {
      "epoch": 0.6818,
      "grad_norm": 0.08438471704721451,
      "learning_rate": 0.002,
      "loss": 2.0546209812164307,
      "step": 6818
    },
    {
      "epoch": 0.682,
      "grad_norm": 0.07747203856706619,
      "learning_rate": 0.002,
      "loss": 2.1355884075164795,
      "step": 6820
    },
    {
      "epoch": 0.6822,
      "grad_norm": 0.0814301148056984,
      "learning_rate": 0.002,
      "loss": 2.1262755393981934,
      "step": 6822
    },
    {
      "epoch": 0.6824,
      "grad_norm": 0.07920104265213013,
      "learning_rate": 0.002,
      "loss": 2.132664442062378,
      "step": 6824
    },
    {
      "epoch": 0.6826,
      "grad_norm": 0.07697220146656036,
      "learning_rate": 0.002,
      "loss": 2.007537841796875,
      "step": 6826
    },
    {
      "epoch": 0.6828,
      "grad_norm": 0.07844442874193192,
      "learning_rate": 0.002,
      "loss": 2.149467706680298,
      "step": 6828
    },
    {
      "epoch": 0.683,
      "grad_norm": 0.08013933897018433,
      "learning_rate": 0.002,
      "loss": 2.120941638946533,
      "step": 6830
    },
    {
      "epoch": 0.6832,
      "grad_norm": 0.07934365421533585,
      "learning_rate": 0.002,
      "loss": 2.1138646602630615,
      "step": 6832
    },
    {
      "epoch": 0.6834,
      "grad_norm": 0.07881626486778259,
      "learning_rate": 0.002,
      "loss": 2.0990657806396484,
      "step": 6834
    },
    {
      "epoch": 0.6836,
      "grad_norm": 0.07856346666812897,
      "learning_rate": 0.002,
      "loss": 2.216526985168457,
      "step": 6836
    },
    {
      "epoch": 0.6838,
      "grad_norm": 0.08227071166038513,
      "learning_rate": 0.002,
      "loss": 2.2685928344726562,
      "step": 6838
    },
    {
      "epoch": 0.684,
      "grad_norm": 0.08027108013629913,
      "learning_rate": 0.002,
      "loss": 2.132488250732422,
      "step": 6840
    },
    {
      "epoch": 0.6842,
      "grad_norm": 0.08158251643180847,
      "learning_rate": 0.002,
      "loss": 2.044736623764038,
      "step": 6842
    },
    {
      "epoch": 0.6844,
      "grad_norm": 0.08107484132051468,
      "learning_rate": 0.002,
      "loss": 2.1591451168060303,
      "step": 6844
    },
    {
      "epoch": 0.6846,
      "grad_norm": 0.07912014424800873,
      "learning_rate": 0.002,
      "loss": 2.0668282508850098,
      "step": 6846
    },
    {
      "epoch": 0.6848,
      "grad_norm": 0.08116241544485092,
      "learning_rate": 0.002,
      "loss": 2.144548177719116,
      "step": 6848
    },
    {
      "epoch": 0.685,
      "grad_norm": 0.07773120701313019,
      "learning_rate": 0.002,
      "loss": 2.066882610321045,
      "step": 6850
    },
    {
      "epoch": 0.6852,
      "grad_norm": 0.08064596354961395,
      "learning_rate": 0.002,
      "loss": 2.0967822074890137,
      "step": 6852
    },
    {
      "epoch": 0.6854,
      "grad_norm": 0.07940958440303802,
      "learning_rate": 0.002,
      "loss": 2.0578513145446777,
      "step": 6854
    },
    {
      "epoch": 0.6856,
      "grad_norm": 0.07885811477899551,
      "learning_rate": 0.002,
      "loss": 2.098431348800659,
      "step": 6856
    },
    {
      "epoch": 0.6858,
      "grad_norm": 0.08184957504272461,
      "learning_rate": 0.002,
      "loss": 2.20243501663208,
      "step": 6858
    },
    {
      "epoch": 0.686,
      "grad_norm": 0.07637583464384079,
      "learning_rate": 0.002,
      "loss": 2.0662388801574707,
      "step": 6860
    },
    {
      "epoch": 0.6862,
      "grad_norm": 0.0801612064242363,
      "learning_rate": 0.002,
      "loss": 2.234480857849121,
      "step": 6862
    },
    {
      "epoch": 0.6864,
      "grad_norm": 0.08009053766727448,
      "learning_rate": 0.002,
      "loss": 2.0932743549346924,
      "step": 6864
    },
    {
      "epoch": 0.6866,
      "grad_norm": 0.08168060332536697,
      "learning_rate": 0.002,
      "loss": 2.049325704574585,
      "step": 6866
    },
    {
      "epoch": 0.6868,
      "grad_norm": 0.08431342244148254,
      "learning_rate": 0.002,
      "loss": 2.1145083904266357,
      "step": 6868
    },
    {
      "epoch": 0.687,
      "grad_norm": 0.08354835212230682,
      "learning_rate": 0.002,
      "loss": 2.139003038406372,
      "step": 6870
    },
    {
      "epoch": 0.6872,
      "grad_norm": 0.0800337865948677,
      "learning_rate": 0.002,
      "loss": 2.1456573009490967,
      "step": 6872
    },
    {
      "epoch": 0.6874,
      "grad_norm": 0.08860030025243759,
      "learning_rate": 0.002,
      "loss": 2.11999773979187,
      "step": 6874
    },
    {
      "epoch": 0.6876,
      "grad_norm": 0.07931220531463623,
      "learning_rate": 0.002,
      "loss": 2.125032901763916,
      "step": 6876
    },
    {
      "epoch": 0.6878,
      "grad_norm": 0.08154179155826569,
      "learning_rate": 0.002,
      "loss": 2.0194430351257324,
      "step": 6878
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.07845328748226166,
      "learning_rate": 0.002,
      "loss": 2.0978522300720215,
      "step": 6880
    },
    {
      "epoch": 0.6882,
      "grad_norm": 0.08059057593345642,
      "learning_rate": 0.002,
      "loss": 2.2360939979553223,
      "step": 6882
    },
    {
      "epoch": 0.6884,
      "grad_norm": 0.07963883876800537,
      "learning_rate": 0.002,
      "loss": 2.244898796081543,
      "step": 6884
    },
    {
      "epoch": 0.6886,
      "grad_norm": 0.07903927564620972,
      "learning_rate": 0.002,
      "loss": 2.1423254013061523,
      "step": 6886
    },
    {
      "epoch": 0.6888,
      "grad_norm": 0.07941485196352005,
      "learning_rate": 0.002,
      "loss": 2.093639373779297,
      "step": 6888
    },
    {
      "epoch": 0.689,
      "grad_norm": 0.08225835859775543,
      "learning_rate": 0.002,
      "loss": 2.0730371475219727,
      "step": 6890
    },
    {
      "epoch": 0.6892,
      "grad_norm": 0.07971034198999405,
      "learning_rate": 0.002,
      "loss": 2.2315855026245117,
      "step": 6892
    },
    {
      "epoch": 0.6894,
      "grad_norm": 0.07850974798202515,
      "learning_rate": 0.002,
      "loss": 2.1969871520996094,
      "step": 6894
    },
    {
      "epoch": 0.6896,
      "grad_norm": 0.08294066041707993,
      "learning_rate": 0.002,
      "loss": 2.1359100341796875,
      "step": 6896
    },
    {
      "epoch": 0.6898,
      "grad_norm": 0.08005452901124954,
      "learning_rate": 0.002,
      "loss": 2.168846368789673,
      "step": 6898
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.08289874345064163,
      "learning_rate": 0.002,
      "loss": 2.15865159034729,
      "step": 6900
    },
    {
      "epoch": 0.6902,
      "grad_norm": 0.09078337997198105,
      "learning_rate": 0.002,
      "loss": 2.245309829711914,
      "step": 6902
    },
    {
      "epoch": 0.6904,
      "grad_norm": 0.08236376196146011,
      "learning_rate": 0.002,
      "loss": 2.198464870452881,
      "step": 6904
    },
    {
      "epoch": 0.6906,
      "grad_norm": 0.07724680006504059,
      "learning_rate": 0.002,
      "loss": 2.110182762145996,
      "step": 6906
    },
    {
      "epoch": 0.6908,
      "grad_norm": 0.0827706977725029,
      "learning_rate": 0.002,
      "loss": 2.157935380935669,
      "step": 6908
    },
    {
      "epoch": 0.691,
      "grad_norm": 0.07777063548564911,
      "learning_rate": 0.002,
      "loss": 2.131395101547241,
      "step": 6910
    },
    {
      "epoch": 0.6912,
      "grad_norm": 0.07645947486162186,
      "learning_rate": 0.002,
      "loss": 2.1665446758270264,
      "step": 6912
    },
    {
      "epoch": 0.6914,
      "grad_norm": 0.08398589491844177,
      "learning_rate": 0.002,
      "loss": 2.18804669380188,
      "step": 6914
    },
    {
      "epoch": 0.6916,
      "grad_norm": 0.07737821340560913,
      "learning_rate": 0.002,
      "loss": 2.060917377471924,
      "step": 6916
    },
    {
      "epoch": 0.6918,
      "grad_norm": 0.08236833661794662,
      "learning_rate": 0.002,
      "loss": 2.1953256130218506,
      "step": 6918
    },
    {
      "epoch": 0.692,
      "grad_norm": 0.08260106295347214,
      "learning_rate": 0.002,
      "loss": 2.1136085987091064,
      "step": 6920
    },
    {
      "epoch": 0.6922,
      "grad_norm": 0.07851315289735794,
      "learning_rate": 0.002,
      "loss": 2.1255764961242676,
      "step": 6922
    },
    {
      "epoch": 0.6924,
      "grad_norm": 0.07727869600057602,
      "learning_rate": 0.002,
      "loss": 2.2082340717315674,
      "step": 6924
    },
    {
      "epoch": 0.6926,
      "grad_norm": 0.08235117793083191,
      "learning_rate": 0.002,
      "loss": 2.2241950035095215,
      "step": 6926
    },
    {
      "epoch": 0.6928,
      "grad_norm": 0.07938013225793839,
      "learning_rate": 0.002,
      "loss": 2.0881755352020264,
      "step": 6928
    },
    {
      "epoch": 0.693,
      "grad_norm": 0.08796534687280655,
      "learning_rate": 0.002,
      "loss": 2.19380784034729,
      "step": 6930
    },
    {
      "epoch": 0.6932,
      "grad_norm": 0.08066044747829437,
      "learning_rate": 0.002,
      "loss": 2.155702829360962,
      "step": 6932
    },
    {
      "epoch": 0.6934,
      "grad_norm": 0.08274912089109421,
      "learning_rate": 0.002,
      "loss": 2.2334823608398438,
      "step": 6934
    },
    {
      "epoch": 0.6936,
      "grad_norm": 0.07765056937932968,
      "learning_rate": 0.002,
      "loss": 2.078707218170166,
      "step": 6936
    },
    {
      "epoch": 0.6938,
      "grad_norm": 0.07886277139186859,
      "learning_rate": 0.002,
      "loss": 2.2151386737823486,
      "step": 6938
    },
    {
      "epoch": 0.694,
      "grad_norm": 0.08267339318990707,
      "learning_rate": 0.002,
      "loss": 2.0940356254577637,
      "step": 6940
    },
    {
      "epoch": 0.6942,
      "grad_norm": 0.08358126878738403,
      "learning_rate": 0.002,
      "loss": 2.18345308303833,
      "step": 6942
    },
    {
      "epoch": 0.6944,
      "grad_norm": 0.08010749518871307,
      "learning_rate": 0.002,
      "loss": 2.135159492492676,
      "step": 6944
    },
    {
      "epoch": 0.6946,
      "grad_norm": 0.07743362337350845,
      "learning_rate": 0.002,
      "loss": 2.1915488243103027,
      "step": 6946
    },
    {
      "epoch": 0.6948,
      "grad_norm": 0.0816405713558197,
      "learning_rate": 0.002,
      "loss": 2.101032018661499,
      "step": 6948
    },
    {
      "epoch": 0.695,
      "grad_norm": 0.07998359203338623,
      "learning_rate": 0.002,
      "loss": 2.1050450801849365,
      "step": 6950
    },
    {
      "epoch": 0.6952,
      "grad_norm": 0.08107997477054596,
      "learning_rate": 0.002,
      "loss": 2.222290515899658,
      "step": 6952
    },
    {
      "epoch": 0.6954,
      "grad_norm": 0.08093603700399399,
      "learning_rate": 0.002,
      "loss": 2.1022162437438965,
      "step": 6954
    },
    {
      "epoch": 0.6956,
      "grad_norm": 0.07658195495605469,
      "learning_rate": 0.002,
      "loss": 2.14501690864563,
      "step": 6956
    },
    {
      "epoch": 0.6958,
      "grad_norm": 0.07889202982187271,
      "learning_rate": 0.002,
      "loss": 2.084648609161377,
      "step": 6958
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.07662421464920044,
      "learning_rate": 0.002,
      "loss": 2.1019599437713623,
      "step": 6960
    },
    {
      "epoch": 0.6962,
      "grad_norm": 0.07786141335964203,
      "learning_rate": 0.002,
      "loss": 2.168714761734009,
      "step": 6962
    },
    {
      "epoch": 0.6964,
      "grad_norm": 0.0806601345539093,
      "learning_rate": 0.002,
      "loss": 2.152282476425171,
      "step": 6964
    },
    {
      "epoch": 0.6966,
      "grad_norm": 0.08429265767335892,
      "learning_rate": 0.002,
      "loss": 2.1284687519073486,
      "step": 6966
    },
    {
      "epoch": 0.6968,
      "grad_norm": 0.07845107465982437,
      "learning_rate": 0.002,
      "loss": 2.0628507137298584,
      "step": 6968
    },
    {
      "epoch": 0.697,
      "grad_norm": 0.08645320683717728,
      "learning_rate": 0.002,
      "loss": 2.13240385055542,
      "step": 6970
    },
    {
      "epoch": 0.6972,
      "grad_norm": 0.08323147892951965,
      "learning_rate": 0.002,
      "loss": 2.302461624145508,
      "step": 6972
    },
    {
      "epoch": 0.6974,
      "grad_norm": 0.076296865940094,
      "learning_rate": 0.002,
      "loss": 2.010124444961548,
      "step": 6974
    },
    {
      "epoch": 0.6976,
      "grad_norm": 0.08046647161245346,
      "learning_rate": 0.002,
      "loss": 2.2502896785736084,
      "step": 6976
    },
    {
      "epoch": 0.6978,
      "grad_norm": 0.07469715923070908,
      "learning_rate": 0.002,
      "loss": 2.106027364730835,
      "step": 6978
    },
    {
      "epoch": 0.698,
      "grad_norm": 0.08191102743148804,
      "learning_rate": 0.002,
      "loss": 2.097848892211914,
      "step": 6980
    },
    {
      "epoch": 0.6982,
      "grad_norm": 0.08125919848680496,
      "learning_rate": 0.002,
      "loss": 2.114117383956909,
      "step": 6982
    },
    {
      "epoch": 0.6984,
      "grad_norm": 0.0815996527671814,
      "learning_rate": 0.002,
      "loss": 2.0276060104370117,
      "step": 6984
    },
    {
      "epoch": 0.6986,
      "grad_norm": 0.08162117004394531,
      "learning_rate": 0.002,
      "loss": 2.1430234909057617,
      "step": 6986
    },
    {
      "epoch": 0.6988,
      "grad_norm": 0.07850626111030579,
      "learning_rate": 0.002,
      "loss": 2.1531295776367188,
      "step": 6988
    },
    {
      "epoch": 0.699,
      "grad_norm": 0.0737154558300972,
      "learning_rate": 0.002,
      "loss": 2.1212265491485596,
      "step": 6990
    },
    {
      "epoch": 0.6992,
      "grad_norm": 0.0811629369854927,
      "learning_rate": 0.002,
      "loss": 2.1172971725463867,
      "step": 6992
    },
    {
      "epoch": 0.6994,
      "grad_norm": 0.0775432288646698,
      "learning_rate": 0.002,
      "loss": 2.196016311645508,
      "step": 6994
    },
    {
      "epoch": 0.6996,
      "grad_norm": 0.08011272549629211,
      "learning_rate": 0.002,
      "loss": 2.0605461597442627,
      "step": 6996
    },
    {
      "epoch": 0.6998,
      "grad_norm": 0.0757266953587532,
      "learning_rate": 0.002,
      "loss": 2.0985448360443115,
      "step": 6998
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.07808296382427216,
      "learning_rate": 0.002,
      "loss": 2.160062789916992,
      "step": 7000
    },
    {
      "epoch": 0.7002,
      "grad_norm": 0.08386440575122833,
      "learning_rate": 0.002,
      "loss": 2.1554160118103027,
      "step": 7002
    },
    {
      "epoch": 0.7004,
      "grad_norm": 0.07784059643745422,
      "learning_rate": 0.002,
      "loss": 2.1353821754455566,
      "step": 7004
    },
    {
      "epoch": 0.7006,
      "grad_norm": 0.08467942476272583,
      "learning_rate": 0.002,
      "loss": 2.0796096324920654,
      "step": 7006
    },
    {
      "epoch": 0.7008,
      "grad_norm": 0.07742491364479065,
      "learning_rate": 0.002,
      "loss": 2.049973249435425,
      "step": 7008
    },
    {
      "epoch": 0.701,
      "grad_norm": 0.08130952715873718,
      "learning_rate": 0.002,
      "loss": 2.1547884941101074,
      "step": 7010
    },
    {
      "epoch": 0.7012,
      "grad_norm": 0.0833539366722107,
      "learning_rate": 0.002,
      "loss": 2.1749768257141113,
      "step": 7012
    },
    {
      "epoch": 0.7014,
      "grad_norm": 0.08293142914772034,
      "learning_rate": 0.002,
      "loss": 2.1575496196746826,
      "step": 7014
    },
    {
      "epoch": 0.7016,
      "grad_norm": 0.08062107115983963,
      "learning_rate": 0.002,
      "loss": 2.1891634464263916,
      "step": 7016
    },
    {
      "epoch": 0.7018,
      "grad_norm": 0.07892117649316788,
      "learning_rate": 0.002,
      "loss": 2.0521388053894043,
      "step": 7018
    },
    {
      "epoch": 0.702,
      "grad_norm": 0.08486044406890869,
      "learning_rate": 0.002,
      "loss": 2.1575045585632324,
      "step": 7020
    },
    {
      "epoch": 0.7022,
      "grad_norm": 0.0810265839099884,
      "learning_rate": 0.002,
      "loss": 2.1682910919189453,
      "step": 7022
    },
    {
      "epoch": 0.7024,
      "grad_norm": 0.07904266566038132,
      "learning_rate": 0.002,
      "loss": 2.133307933807373,
      "step": 7024
    },
    {
      "epoch": 0.7026,
      "grad_norm": 0.08303944021463394,
      "learning_rate": 0.002,
      "loss": 2.240293502807617,
      "step": 7026
    },
    {
      "epoch": 0.7028,
      "grad_norm": 0.07879125326871872,
      "learning_rate": 0.002,
      "loss": 2.0837531089782715,
      "step": 7028
    },
    {
      "epoch": 0.703,
      "grad_norm": 0.08199089765548706,
      "learning_rate": 0.002,
      "loss": 2.1306169033050537,
      "step": 7030
    },
    {
      "epoch": 0.7032,
      "grad_norm": 0.07980598509311676,
      "learning_rate": 0.002,
      "loss": 2.0750041007995605,
      "step": 7032
    },
    {
      "epoch": 0.7034,
      "grad_norm": 0.07846025377511978,
      "learning_rate": 0.002,
      "loss": 2.1687381267547607,
      "step": 7034
    },
    {
      "epoch": 0.7036,
      "grad_norm": 0.07902868092060089,
      "learning_rate": 0.002,
      "loss": 2.046247959136963,
      "step": 7036
    },
    {
      "epoch": 0.7038,
      "grad_norm": 0.0822536051273346,
      "learning_rate": 0.002,
      "loss": 2.0866968631744385,
      "step": 7038
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.07765032351016998,
      "learning_rate": 0.002,
      "loss": 2.230395793914795,
      "step": 7040
    },
    {
      "epoch": 0.7042,
      "grad_norm": 0.07492192834615707,
      "learning_rate": 0.002,
      "loss": 2.1303653717041016,
      "step": 7042
    },
    {
      "epoch": 0.7044,
      "grad_norm": 0.08632918447256088,
      "learning_rate": 0.002,
      "loss": 2.170811653137207,
      "step": 7044
    },
    {
      "epoch": 0.7046,
      "grad_norm": 0.07709383964538574,
      "learning_rate": 0.002,
      "loss": 2.2126386165618896,
      "step": 7046
    },
    {
      "epoch": 0.7048,
      "grad_norm": 0.07999474555253983,
      "learning_rate": 0.002,
      "loss": 2.0963215827941895,
      "step": 7048
    },
    {
      "epoch": 0.705,
      "grad_norm": 0.079810231924057,
      "learning_rate": 0.002,
      "loss": 2.123276472091675,
      "step": 7050
    },
    {
      "epoch": 0.7052,
      "grad_norm": 0.08258571475744247,
      "learning_rate": 0.002,
      "loss": 2.1675877571105957,
      "step": 7052
    },
    {
      "epoch": 0.7054,
      "grad_norm": 0.07946793735027313,
      "learning_rate": 0.002,
      "loss": 2.061309337615967,
      "step": 7054
    },
    {
      "epoch": 0.7056,
      "grad_norm": 0.08138535171747208,
      "learning_rate": 0.002,
      "loss": 2.2229220867156982,
      "step": 7056
    },
    {
      "epoch": 0.7058,
      "grad_norm": 0.09142571687698364,
      "learning_rate": 0.002,
      "loss": 2.04245662689209,
      "step": 7058
    },
    {
      "epoch": 0.706,
      "grad_norm": 0.08878492563962936,
      "learning_rate": 0.002,
      "loss": 2.1428415775299072,
      "step": 7060
    },
    {
      "epoch": 0.7062,
      "grad_norm": 0.07811098545789719,
      "learning_rate": 0.002,
      "loss": 2.105480432510376,
      "step": 7062
    },
    {
      "epoch": 0.7064,
      "grad_norm": 0.07917789369821548,
      "learning_rate": 0.002,
      "loss": 2.1760270595550537,
      "step": 7064
    },
    {
      "epoch": 0.7066,
      "grad_norm": 0.0848713293671608,
      "learning_rate": 0.002,
      "loss": 2.116480827331543,
      "step": 7066
    },
    {
      "epoch": 0.7068,
      "grad_norm": 0.08168463408946991,
      "learning_rate": 0.002,
      "loss": 2.2476067543029785,
      "step": 7068
    },
    {
      "epoch": 0.707,
      "grad_norm": 0.07683828473091125,
      "learning_rate": 0.002,
      "loss": 2.217120409011841,
      "step": 7070
    },
    {
      "epoch": 0.7072,
      "grad_norm": 0.0825890302658081,
      "learning_rate": 0.002,
      "loss": 2.228317975997925,
      "step": 7072
    },
    {
      "epoch": 0.7074,
      "grad_norm": 0.07892631739377975,
      "learning_rate": 0.002,
      "loss": 2.163893938064575,
      "step": 7074
    },
    {
      "epoch": 0.7076,
      "grad_norm": 0.08350306749343872,
      "learning_rate": 0.002,
      "loss": 2.1939220428466797,
      "step": 7076
    },
    {
      "epoch": 0.7078,
      "grad_norm": 0.08085200935602188,
      "learning_rate": 0.002,
      "loss": 2.149044990539551,
      "step": 7078
    },
    {
      "epoch": 0.708,
      "grad_norm": 0.08068107813596725,
      "learning_rate": 0.002,
      "loss": 2.164863348007202,
      "step": 7080
    },
    {
      "epoch": 0.7082,
      "grad_norm": 0.08026311546564102,
      "learning_rate": 0.002,
      "loss": 2.2922613620758057,
      "step": 7082
    },
    {
      "epoch": 0.7084,
      "grad_norm": 0.08471262454986572,
      "learning_rate": 0.002,
      "loss": 2.179973602294922,
      "step": 7084
    },
    {
      "epoch": 0.7086,
      "grad_norm": 0.08346312493085861,
      "learning_rate": 0.002,
      "loss": 2.1120047569274902,
      "step": 7086
    },
    {
      "epoch": 0.7088,
      "grad_norm": 0.0757518857717514,
      "learning_rate": 0.002,
      "loss": 2.1206114292144775,
      "step": 7088
    },
    {
      "epoch": 0.709,
      "grad_norm": 0.07829666137695312,
      "learning_rate": 0.002,
      "loss": 2.1622588634490967,
      "step": 7090
    },
    {
      "epoch": 0.7092,
      "grad_norm": 0.07704702764749527,
      "learning_rate": 0.002,
      "loss": 2.1885218620300293,
      "step": 7092
    },
    {
      "epoch": 0.7094,
      "grad_norm": 0.07974343001842499,
      "learning_rate": 0.002,
      "loss": 2.134626865386963,
      "step": 7094
    },
    {
      "epoch": 0.7096,
      "grad_norm": 0.08152250945568085,
      "learning_rate": 0.002,
      "loss": 2.196444034576416,
      "step": 7096
    },
    {
      "epoch": 0.7098,
      "grad_norm": 0.07920412719249725,
      "learning_rate": 0.002,
      "loss": 2.2125015258789062,
      "step": 7098
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.08056474477052689,
      "learning_rate": 0.002,
      "loss": 2.1969263553619385,
      "step": 7100
    },
    {
      "epoch": 0.7102,
      "grad_norm": 0.07792741805315018,
      "learning_rate": 0.0019993103448275864,
      "loss": 2.2327687740325928,
      "step": 7102
    },
    {
      "epoch": 0.7104,
      "grad_norm": 0.07602299004793167,
      "learning_rate": 0.0019979310344827588,
      "loss": 2.0763442516326904,
      "step": 7104
    },
    {
      "epoch": 0.7106,
      "grad_norm": 0.07921279966831207,
      "learning_rate": 0.001996551724137931,
      "loss": 2.153966188430786,
      "step": 7106
    },
    {
      "epoch": 0.7108,
      "grad_norm": 0.08701226860284805,
      "learning_rate": 0.0019951724137931034,
      "loss": 2.1568336486816406,
      "step": 7108
    },
    {
      "epoch": 0.711,
      "grad_norm": 0.07567008584737778,
      "learning_rate": 0.0019937931034482758,
      "loss": 2.125420093536377,
      "step": 7110
    },
    {
      "epoch": 0.7112,
      "grad_norm": 0.0804176777601242,
      "learning_rate": 0.001992413793103448,
      "loss": 2.241528034210205,
      "step": 7112
    },
    {
      "epoch": 0.7114,
      "grad_norm": 0.07921349257230759,
      "learning_rate": 0.0019910344827586204,
      "loss": 2.1719422340393066,
      "step": 7114
    },
    {
      "epoch": 0.7116,
      "grad_norm": 0.08291216194629669,
      "learning_rate": 0.001989655172413793,
      "loss": 2.1436102390289307,
      "step": 7116
    },
    {
      "epoch": 0.7118,
      "grad_norm": 0.08487723767757416,
      "learning_rate": 0.0019882758620689655,
      "loss": 2.0737314224243164,
      "step": 7118
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.07753840833902359,
      "learning_rate": 0.001986896551724138,
      "loss": 2.1960158348083496,
      "step": 7120
    },
    {
      "epoch": 0.7122,
      "grad_norm": 0.08197183907032013,
      "learning_rate": 0.00198551724137931,
      "loss": 2.1560771465301514,
      "step": 7122
    },
    {
      "epoch": 0.7124,
      "grad_norm": 0.07650943100452423,
      "learning_rate": 0.0019841379310344825,
      "loss": 2.080005407333374,
      "step": 7124
    },
    {
      "epoch": 0.7126,
      "grad_norm": 0.08194252103567123,
      "learning_rate": 0.0019827586206896553,
      "loss": 2.225090742111206,
      "step": 7126
    },
    {
      "epoch": 0.7128,
      "grad_norm": 0.07933368533849716,
      "learning_rate": 0.0019813793103448277,
      "loss": 2.122527599334717,
      "step": 7128
    },
    {
      "epoch": 0.713,
      "grad_norm": 0.08032652735710144,
      "learning_rate": 0.00198,
      "loss": 2.2629292011260986,
      "step": 7130
    },
    {
      "epoch": 0.7132,
      "grad_norm": 0.08151187002658844,
      "learning_rate": 0.0019786206896551723,
      "loss": 2.2281484603881836,
      "step": 7132
    },
    {
      "epoch": 0.7134,
      "grad_norm": 0.07949984818696976,
      "learning_rate": 0.001977241379310345,
      "loss": 2.1688308715820312,
      "step": 7134
    },
    {
      "epoch": 0.7136,
      "grad_norm": 0.08174599707126617,
      "learning_rate": 0.0019758620689655174,
      "loss": 2.187023162841797,
      "step": 7136
    },
    {
      "epoch": 0.7138,
      "grad_norm": 0.0768292024731636,
      "learning_rate": 0.0019744827586206898,
      "loss": 2.192894220352173,
      "step": 7138
    },
    {
      "epoch": 0.714,
      "grad_norm": 0.07450198382139206,
      "learning_rate": 0.001973103448275862,
      "loss": 2.1211628913879395,
      "step": 7140
    },
    {
      "epoch": 0.7142,
      "grad_norm": 0.08157164603471756,
      "learning_rate": 0.0019717241379310344,
      "loss": 2.1135337352752686,
      "step": 7142
    },
    {
      "epoch": 0.7144,
      "grad_norm": 0.07876700907945633,
      "learning_rate": 0.001970344827586207,
      "loss": 2.143261671066284,
      "step": 7144
    },
    {
      "epoch": 0.7146,
      "grad_norm": 0.08044891804456711,
      "learning_rate": 0.0019689655172413795,
      "loss": 2.090092182159424,
      "step": 7146
    },
    {
      "epoch": 0.7148,
      "grad_norm": 0.08099083602428436,
      "learning_rate": 0.001967586206896552,
      "loss": 2.173614501953125,
      "step": 7148
    },
    {
      "epoch": 0.715,
      "grad_norm": 0.07867464423179626,
      "learning_rate": 0.001966206896551724,
      "loss": 2.048114061355591,
      "step": 7150
    },
    {
      "epoch": 0.7152,
      "grad_norm": 0.08297286927700043,
      "learning_rate": 0.0019648275862068965,
      "loss": 2.1290981769561768,
      "step": 7152
    },
    {
      "epoch": 0.7154,
      "grad_norm": 0.0811758041381836,
      "learning_rate": 0.001963448275862069,
      "loss": 2.1847755908966064,
      "step": 7154
    },
    {
      "epoch": 0.7156,
      "grad_norm": 0.07557240128517151,
      "learning_rate": 0.0019620689655172417,
      "loss": 2.135352373123169,
      "step": 7156
    },
    {
      "epoch": 0.7158,
      "grad_norm": 0.07719894498586655,
      "learning_rate": 0.001960689655172414,
      "loss": 2.1288952827453613,
      "step": 7158
    },
    {
      "epoch": 0.716,
      "grad_norm": 0.08037374168634415,
      "learning_rate": 0.0019593103448275863,
      "loss": 2.209567070007324,
      "step": 7160
    },
    {
      "epoch": 0.7162,
      "grad_norm": 0.08390284329652786,
      "learning_rate": 0.0019579310344827587,
      "loss": 2.2271876335144043,
      "step": 7162
    },
    {
      "epoch": 0.7164,
      "grad_norm": 0.07609125971794128,
      "learning_rate": 0.001956551724137931,
      "loss": 2.0979866981506348,
      "step": 7164
    },
    {
      "epoch": 0.7166,
      "grad_norm": 0.08271478861570358,
      "learning_rate": 0.0019551724137931033,
      "loss": 2.2137327194213867,
      "step": 7166
    },
    {
      "epoch": 0.7168,
      "grad_norm": 0.0837763100862503,
      "learning_rate": 0.0019537931034482757,
      "loss": 2.197678804397583,
      "step": 7168
    },
    {
      "epoch": 0.717,
      "grad_norm": 0.08037419617176056,
      "learning_rate": 0.0019524137931034482,
      "loss": 2.1388726234436035,
      "step": 7170
    },
    {
      "epoch": 0.7172,
      "grad_norm": 0.0814507007598877,
      "learning_rate": 0.0019510344827586208,
      "loss": 2.280663013458252,
      "step": 7172
    },
    {
      "epoch": 0.7174,
      "grad_norm": 0.07771873474121094,
      "learning_rate": 0.001949655172413793,
      "loss": 2.1352572441101074,
      "step": 7174
    },
    {
      "epoch": 0.7176,
      "grad_norm": 0.0800943523645401,
      "learning_rate": 0.0019482758620689657,
      "loss": 2.2624571323394775,
      "step": 7176
    },
    {
      "epoch": 0.7178,
      "grad_norm": 0.07409847527742386,
      "learning_rate": 0.001946896551724138,
      "loss": 2.102511167526245,
      "step": 7178
    },
    {
      "epoch": 0.718,
      "grad_norm": 0.08009517937898636,
      "learning_rate": 0.0019455172413793103,
      "loss": 2.129549026489258,
      "step": 7180
    },
    {
      "epoch": 0.7182,
      "grad_norm": 0.08009310811758041,
      "learning_rate": 0.0019441379310344829,
      "loss": 2.1652095317840576,
      "step": 7182
    },
    {
      "epoch": 0.7184,
      "grad_norm": 0.08342896401882172,
      "learning_rate": 0.0019427586206896552,
      "loss": 2.217256546020508,
      "step": 7184
    },
    {
      "epoch": 0.7186,
      "grad_norm": 0.08397495746612549,
      "learning_rate": 0.0019413793103448275,
      "loss": 2.1921465396881104,
      "step": 7186
    },
    {
      "epoch": 0.7188,
      "grad_norm": 0.0834931880235672,
      "learning_rate": 0.0019399999999999999,
      "loss": 2.1635921001434326,
      "step": 7188
    },
    {
      "epoch": 0.719,
      "grad_norm": 0.0749993845820427,
      "learning_rate": 0.0019386206896551724,
      "loss": 2.2057313919067383,
      "step": 7190
    },
    {
      "epoch": 0.7192,
      "grad_norm": 0.08271288871765137,
      "learning_rate": 0.001937241379310345,
      "loss": 2.1481034755706787,
      "step": 7192
    },
    {
      "epoch": 0.7194,
      "grad_norm": 0.07578112185001373,
      "learning_rate": 0.0019358620689655173,
      "loss": 2.115204095840454,
      "step": 7194
    },
    {
      "epoch": 0.7196,
      "grad_norm": 0.0818672850728035,
      "learning_rate": 0.0019344827586206897,
      "loss": 2.318047523498535,
      "step": 7196
    },
    {
      "epoch": 0.7198,
      "grad_norm": 0.08138971775770187,
      "learning_rate": 0.001933103448275862,
      "loss": 2.149446487426758,
      "step": 7198
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.07932395488023758,
      "learning_rate": 0.0019317241379310345,
      "loss": 2.144550085067749,
      "step": 7200
    },
    {
      "epoch": 0.7202,
      "grad_norm": 0.08066140860319138,
      "learning_rate": 0.001930344827586207,
      "loss": 2.0274200439453125,
      "step": 7202
    },
    {
      "epoch": 0.7204,
      "grad_norm": 0.07447659969329834,
      "learning_rate": 0.0019289655172413794,
      "loss": 2.145231008529663,
      "step": 7204
    },
    {
      "epoch": 0.7206,
      "grad_norm": 0.08241638541221619,
      "learning_rate": 0.0019275862068965518,
      "loss": 2.1422786712646484,
      "step": 7206
    },
    {
      "epoch": 0.7208,
      "grad_norm": 0.0868416503071785,
      "learning_rate": 0.001926206896551724,
      "loss": 2.2598893642425537,
      "step": 7208
    },
    {
      "epoch": 0.721,
      "grad_norm": 0.08047399669885635,
      "learning_rate": 0.0019248275862068967,
      "loss": 2.119394540786743,
      "step": 7210
    },
    {
      "epoch": 0.7212,
      "grad_norm": 0.07579444348812103,
      "learning_rate": 0.001923448275862069,
      "loss": 2.0860509872436523,
      "step": 7212
    },
    {
      "epoch": 0.7214,
      "grad_norm": 0.07916457206010818,
      "learning_rate": 0.0019220689655172413,
      "loss": 2.1729400157928467,
      "step": 7214
    },
    {
      "epoch": 0.7216,
      "grad_norm": 0.07587218284606934,
      "learning_rate": 0.0019206896551724139,
      "loss": 2.2344038486480713,
      "step": 7216
    },
    {
      "epoch": 0.7218,
      "grad_norm": 0.0797668918967247,
      "learning_rate": 0.0019193103448275862,
      "loss": 2.108630418777466,
      "step": 7218
    },
    {
      "epoch": 0.722,
      "grad_norm": 0.07887205481529236,
      "learning_rate": 0.0019179310344827588,
      "loss": 2.283581018447876,
      "step": 7220
    },
    {
      "epoch": 0.7222,
      "grad_norm": 0.0780055969953537,
      "learning_rate": 0.001916551724137931,
      "loss": 2.1609933376312256,
      "step": 7222
    },
    {
      "epoch": 0.7224,
      "grad_norm": 0.0776340663433075,
      "learning_rate": 0.0019151724137931034,
      "loss": 2.158816337585449,
      "step": 7224
    },
    {
      "epoch": 0.7226,
      "grad_norm": 0.07749968767166138,
      "learning_rate": 0.0019137931034482758,
      "loss": 2.0506792068481445,
      "step": 7226
    },
    {
      "epoch": 0.7228,
      "grad_norm": 0.07908536493778229,
      "learning_rate": 0.0019124137931034481,
      "loss": 2.1486408710479736,
      "step": 7228
    },
    {
      "epoch": 0.723,
      "grad_norm": 0.07804745435714722,
      "learning_rate": 0.0019110344827586209,
      "loss": 2.1671767234802246,
      "step": 7230
    },
    {
      "epoch": 0.7232,
      "grad_norm": 0.07687795162200928,
      "learning_rate": 0.0019096551724137932,
      "loss": 2.1494503021240234,
      "step": 7232
    },
    {
      "epoch": 0.7234,
      "grad_norm": 0.08517304062843323,
      "learning_rate": 0.0019082758620689656,
      "loss": 2.250797986984253,
      "step": 7234
    },
    {
      "epoch": 0.7236,
      "grad_norm": 0.08212630450725555,
      "learning_rate": 0.0019068965517241379,
      "loss": 2.0983381271362305,
      "step": 7236
    },
    {
      "epoch": 0.7238,
      "grad_norm": 0.0816820040345192,
      "learning_rate": 0.0019055172413793104,
      "loss": 2.1506922245025635,
      "step": 7238
    },
    {
      "epoch": 0.724,
      "grad_norm": 0.08227749168872833,
      "learning_rate": 0.0019041379310344828,
      "loss": 2.113798141479492,
      "step": 7240
    },
    {
      "epoch": 0.7242,
      "grad_norm": 0.08178123086690903,
      "learning_rate": 0.0019027586206896553,
      "loss": 2.2808997631073,
      "step": 7242
    },
    {
      "epoch": 0.7244,
      "grad_norm": 0.08913150429725647,
      "learning_rate": 0.0019013793103448277,
      "loss": 2.116638422012329,
      "step": 7244
    },
    {
      "epoch": 0.7246,
      "grad_norm": 0.08170966058969498,
      "learning_rate": 0.0019,
      "loss": 2.2469711303710938,
      "step": 7246
    },
    {
      "epoch": 0.7248,
      "grad_norm": 0.09158118069171906,
      "learning_rate": 0.0018986206896551725,
      "loss": 2.168942928314209,
      "step": 7248
    },
    {
      "epoch": 0.725,
      "grad_norm": 0.07939305156469345,
      "learning_rate": 0.0018972413793103449,
      "loss": 2.1472015380859375,
      "step": 7250
    },
    {
      "epoch": 0.7252,
      "grad_norm": 0.07841555029153824,
      "learning_rate": 0.0018958620689655172,
      "loss": 2.155827522277832,
      "step": 7252
    },
    {
      "epoch": 0.7254,
      "grad_norm": 0.0806606262922287,
      "learning_rate": 0.0018944827586206896,
      "loss": 2.1014788150787354,
      "step": 7254
    },
    {
      "epoch": 0.7256,
      "grad_norm": 0.07687100023031235,
      "learning_rate": 0.001893103448275862,
      "loss": 2.1593997478485107,
      "step": 7256
    },
    {
      "epoch": 0.7258,
      "grad_norm": 0.08045358210802078,
      "learning_rate": 0.0018917241379310347,
      "loss": 2.25496768951416,
      "step": 7258
    },
    {
      "epoch": 0.726,
      "grad_norm": 0.08081984519958496,
      "learning_rate": 0.001890344827586207,
      "loss": 2.1842544078826904,
      "step": 7260
    },
    {
      "epoch": 0.7262,
      "grad_norm": 0.07540112733840942,
      "learning_rate": 0.0018889655172413793,
      "loss": 2.200930118560791,
      "step": 7262
    },
    {
      "epoch": 0.7264,
      "grad_norm": 0.07813423871994019,
      "learning_rate": 0.0018875862068965517,
      "loss": 2.145798444747925,
      "step": 7264
    },
    {
      "epoch": 0.7266,
      "grad_norm": 0.07813139259815216,
      "learning_rate": 0.0018862068965517242,
      "loss": 2.262037992477417,
      "step": 7266
    },
    {
      "epoch": 0.7268,
      "grad_norm": 0.07882880419492722,
      "learning_rate": 0.0018848275862068966,
      "loss": 2.136561632156372,
      "step": 7268
    },
    {
      "epoch": 0.727,
      "grad_norm": 0.07865049690008163,
      "learning_rate": 0.001883448275862069,
      "loss": 2.151733636856079,
      "step": 7270
    },
    {
      "epoch": 0.7272,
      "grad_norm": 0.07860574871301651,
      "learning_rate": 0.0018820689655172414,
      "loss": 2.1687560081481934,
      "step": 7272
    },
    {
      "epoch": 0.7274,
      "grad_norm": 0.07622741162776947,
      "learning_rate": 0.0018806896551724138,
      "loss": 2.246670722961426,
      "step": 7274
    },
    {
      "epoch": 0.7276,
      "grad_norm": 0.08956118673086166,
      "learning_rate": 0.0018793103448275863,
      "loss": 2.2313475608825684,
      "step": 7276
    },
    {
      "epoch": 0.7278,
      "grad_norm": 0.08334733545780182,
      "learning_rate": 0.0018779310344827587,
      "loss": 2.1508371829986572,
      "step": 7278
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.08154570311307907,
      "learning_rate": 0.001876551724137931,
      "loss": 2.140082597732544,
      "step": 7280
    },
    {
      "epoch": 0.7282,
      "grad_norm": 0.08382118493318558,
      "learning_rate": 0.0018751724137931033,
      "loss": 2.2305431365966797,
      "step": 7282
    },
    {
      "epoch": 0.7284,
      "grad_norm": 0.08029402792453766,
      "learning_rate": 0.0018737931034482759,
      "loss": 2.082244634628296,
      "step": 7284
    },
    {
      "epoch": 0.7286,
      "grad_norm": 0.07952895760536194,
      "learning_rate": 0.0018724137931034484,
      "loss": 2.1433920860290527,
      "step": 7286
    },
    {
      "epoch": 0.7288,
      "grad_norm": 0.08096469193696976,
      "learning_rate": 0.0018710344827586208,
      "loss": 2.103776693344116,
      "step": 7288
    },
    {
      "epoch": 0.729,
      "grad_norm": 0.09733892232179642,
      "learning_rate": 0.0018696551724137931,
      "loss": 2.230036497116089,
      "step": 7290
    },
    {
      "epoch": 0.7292,
      "grad_norm": 0.08626320213079453,
      "learning_rate": 0.0018682758620689654,
      "loss": 2.216620445251465,
      "step": 7292
    },
    {
      "epoch": 0.7294,
      "grad_norm": 0.07694953680038452,
      "learning_rate": 0.0018668965517241378,
      "loss": 2.083921194076538,
      "step": 7294
    },
    {
      "epoch": 0.7296,
      "grad_norm": 0.08180315792560577,
      "learning_rate": 0.0018655172413793105,
      "loss": 2.2150347232818604,
      "step": 7296
    },
    {
      "epoch": 0.7298,
      "grad_norm": 0.08367721736431122,
      "learning_rate": 0.0018641379310344829,
      "loss": 2.1169254779815674,
      "step": 7298
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.07892346382141113,
      "learning_rate": 0.0018627586206896552,
      "loss": 2.186040163040161,
      "step": 7300
    },
    {
      "epoch": 0.7302,
      "grad_norm": 0.07967325299978256,
      "learning_rate": 0.0018613793103448276,
      "loss": 2.181168556213379,
      "step": 7302
    },
    {
      "epoch": 0.7304,
      "grad_norm": 0.07639475911855698,
      "learning_rate": 0.00186,
      "loss": 2.215930223464966,
      "step": 7304
    },
    {
      "epoch": 0.7306,
      "grad_norm": 0.08138453215360641,
      "learning_rate": 0.0018586206896551724,
      "loss": 2.141669511795044,
      "step": 7306
    },
    {
      "epoch": 0.7308,
      "grad_norm": 0.08204241842031479,
      "learning_rate": 0.0018572413793103448,
      "loss": 2.1423532962799072,
      "step": 7308
    },
    {
      "epoch": 0.731,
      "grad_norm": 0.07975761592388153,
      "learning_rate": 0.0018558620689655173,
      "loss": 2.2379488945007324,
      "step": 7310
    },
    {
      "epoch": 0.7312,
      "grad_norm": 0.07931425422430038,
      "learning_rate": 0.0018544827586206897,
      "loss": 2.1669507026672363,
      "step": 7312
    },
    {
      "epoch": 0.7314,
      "grad_norm": 0.07829828560352325,
      "learning_rate": 0.0018531034482758622,
      "loss": 2.108837842941284,
      "step": 7314
    },
    {
      "epoch": 0.7316,
      "grad_norm": 0.07791906595230103,
      "learning_rate": 0.0018517241379310346,
      "loss": 2.16524338722229,
      "step": 7316
    },
    {
      "epoch": 0.7318,
      "grad_norm": 0.08197811990976334,
      "learning_rate": 0.0018503448275862069,
      "loss": 2.222151756286621,
      "step": 7318
    },
    {
      "epoch": 0.732,
      "grad_norm": 0.08007356524467468,
      "learning_rate": 0.0018489655172413792,
      "loss": 2.211862564086914,
      "step": 7320
    },
    {
      "epoch": 0.7322,
      "grad_norm": 0.07760484516620636,
      "learning_rate": 0.0018475862068965516,
      "loss": 2.139594793319702,
      "step": 7322
    },
    {
      "epoch": 0.7324,
      "grad_norm": 0.07874445617198944,
      "learning_rate": 0.0018462068965517243,
      "loss": 2.177776336669922,
      "step": 7324
    },
    {
      "epoch": 0.7326,
      "grad_norm": 0.08031252771615982,
      "learning_rate": 0.0018448275862068967,
      "loss": 2.149054527282715,
      "step": 7326
    },
    {
      "epoch": 0.7328,
      "grad_norm": 0.08348140120506287,
      "learning_rate": 0.001843448275862069,
      "loss": 2.242650032043457,
      "step": 7328
    },
    {
      "epoch": 0.733,
      "grad_norm": 0.07987255603075027,
      "learning_rate": 0.0018420689655172413,
      "loss": 2.2017102241516113,
      "step": 7330
    },
    {
      "epoch": 0.7332,
      "grad_norm": 0.08613470941781998,
      "learning_rate": 0.0018406896551724139,
      "loss": 2.1715197563171387,
      "step": 7332
    },
    {
      "epoch": 0.7334,
      "grad_norm": 0.08582089841365814,
      "learning_rate": 0.0018393103448275862,
      "loss": 2.313183307647705,
      "step": 7334
    },
    {
      "epoch": 0.7336,
      "grad_norm": 0.08157283812761307,
      "learning_rate": 0.0018379310344827588,
      "loss": 2.191145181655884,
      "step": 7336
    },
    {
      "epoch": 0.7338,
      "grad_norm": 0.07934538275003433,
      "learning_rate": 0.0018365517241379311,
      "loss": 2.187695026397705,
      "step": 7338
    },
    {
      "epoch": 0.734,
      "grad_norm": 0.07877888530492783,
      "learning_rate": 0.0018351724137931034,
      "loss": 2.2358040809631348,
      "step": 7340
    },
    {
      "epoch": 0.7342,
      "grad_norm": 0.08227518200874329,
      "learning_rate": 0.001833793103448276,
      "loss": 2.1596953868865967,
      "step": 7342
    },
    {
      "epoch": 0.7344,
      "grad_norm": 0.0786016583442688,
      "learning_rate": 0.0018324137931034483,
      "loss": 2.2699155807495117,
      "step": 7344
    },
    {
      "epoch": 0.7346,
      "grad_norm": 0.0796242207288742,
      "learning_rate": 0.0018310344827586207,
      "loss": 2.166008234024048,
      "step": 7346
    },
    {
      "epoch": 0.7348,
      "grad_norm": 0.07979603856801987,
      "learning_rate": 0.001829655172413793,
      "loss": 2.19624662399292,
      "step": 7348
    },
    {
      "epoch": 0.735,
      "grad_norm": 0.0754130631685257,
      "learning_rate": 0.0018282758620689656,
      "loss": 2.121434211730957,
      "step": 7350
    },
    {
      "epoch": 0.7352,
      "grad_norm": 0.07952281087636948,
      "learning_rate": 0.0018268965517241381,
      "loss": 2.149099111557007,
      "step": 7352
    },
    {
      "epoch": 0.7354,
      "grad_norm": 0.0775696262717247,
      "learning_rate": 0.0018255172413793104,
      "loss": 2.228285074234009,
      "step": 7354
    },
    {
      "epoch": 0.7356,
      "grad_norm": 0.08097950369119644,
      "learning_rate": 0.0018241379310344828,
      "loss": 2.164606809616089,
      "step": 7356
    },
    {
      "epoch": 0.7358,
      "grad_norm": 0.08089274913072586,
      "learning_rate": 0.0018227586206896551,
      "loss": 2.1325793266296387,
      "step": 7358
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.07678491622209549,
      "learning_rate": 0.0018213793103448277,
      "loss": 2.151829957962036,
      "step": 7360
    },
    {
      "epoch": 0.7362,
      "grad_norm": 0.07850159704685211,
      "learning_rate": 0.00182,
      "loss": 2.1631150245666504,
      "step": 7362
    },
    {
      "epoch": 0.7364,
      "grad_norm": 0.07595252990722656,
      "learning_rate": 0.0018186206896551726,
      "loss": 2.161785364151001,
      "step": 7364
    },
    {
      "epoch": 0.7366,
      "grad_norm": 0.08001919835805893,
      "learning_rate": 0.0018172413793103449,
      "loss": 2.2471418380737305,
      "step": 7366
    },
    {
      "epoch": 0.7368,
      "grad_norm": 0.07959180325269699,
      "learning_rate": 0.0018158620689655172,
      "loss": 2.1968212127685547,
      "step": 7368
    },
    {
      "epoch": 0.737,
      "grad_norm": 0.07893354445695877,
      "learning_rate": 0.0018144827586206898,
      "loss": 2.157308578491211,
      "step": 7370
    },
    {
      "epoch": 0.7372,
      "grad_norm": 0.08402976393699646,
      "learning_rate": 0.0018131034482758621,
      "loss": 2.1993300914764404,
      "step": 7372
    },
    {
      "epoch": 0.7374,
      "grad_norm": 0.07613751292228699,
      "learning_rate": 0.0018117241379310344,
      "loss": 2.1904666423797607,
      "step": 7374
    },
    {
      "epoch": 0.7376,
      "grad_norm": 0.08077151328325272,
      "learning_rate": 0.0018103448275862068,
      "loss": 2.210062026977539,
      "step": 7376
    },
    {
      "epoch": 0.7378,
      "grad_norm": 0.07953599840402603,
      "learning_rate": 0.0018089655172413793,
      "loss": 2.2666826248168945,
      "step": 7378
    },
    {
      "epoch": 0.738,
      "grad_norm": 0.07641123235225677,
      "learning_rate": 0.0018075862068965519,
      "loss": 2.1847424507141113,
      "step": 7380
    },
    {
      "epoch": 0.7382,
      "grad_norm": 0.08090943098068237,
      "learning_rate": 0.0018062068965517242,
      "loss": 2.2258315086364746,
      "step": 7382
    },
    {
      "epoch": 0.7384,
      "grad_norm": 0.07623188197612762,
      "learning_rate": 0.0018048275862068966,
      "loss": 2.1771621704101562,
      "step": 7384
    },
    {
      "epoch": 0.7386,
      "grad_norm": 0.07901765406131744,
      "learning_rate": 0.001803448275862069,
      "loss": 2.1371798515319824,
      "step": 7386
    },
    {
      "epoch": 0.7388,
      "grad_norm": 0.07721264660358429,
      "learning_rate": 0.0018020689655172412,
      "loss": 2.121018648147583,
      "step": 7388
    },
    {
      "epoch": 0.739,
      "grad_norm": 0.08379843086004257,
      "learning_rate": 0.001800689655172414,
      "loss": 2.2511937618255615,
      "step": 7390
    },
    {
      "epoch": 0.7392,
      "grad_norm": 0.0816497653722763,
      "learning_rate": 0.0017993103448275863,
      "loss": 2.2749359607696533,
      "step": 7392
    },
    {
      "epoch": 0.7394,
      "grad_norm": 0.09098511934280396,
      "learning_rate": 0.0017979310344827587,
      "loss": 2.2972497940063477,
      "step": 7394
    },
    {
      "epoch": 0.7396,
      "grad_norm": 0.08813531696796417,
      "learning_rate": 0.001796551724137931,
      "loss": 2.18528413772583,
      "step": 7396
    },
    {
      "epoch": 0.7398,
      "grad_norm": 0.07556232064962387,
      "learning_rate": 0.0017951724137931036,
      "loss": 2.1684913635253906,
      "step": 7398
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.0775287002325058,
      "learning_rate": 0.001793793103448276,
      "loss": 2.1837031841278076,
      "step": 7400
    },
    {
      "epoch": 0.7402,
      "grad_norm": 0.07807229459285736,
      "learning_rate": 0.0017924137931034482,
      "loss": 2.2423906326293945,
      "step": 7402
    },
    {
      "epoch": 0.7404,
      "grad_norm": 0.09154165536165237,
      "learning_rate": 0.0017910344827586208,
      "loss": 2.1780452728271484,
      "step": 7404
    },
    {
      "epoch": 0.7406,
      "grad_norm": 0.08222435414791107,
      "learning_rate": 0.0017896551724137931,
      "loss": 2.19199538230896,
      "step": 7406
    },
    {
      "epoch": 0.7408,
      "grad_norm": 0.07858211547136307,
      "learning_rate": 0.0017882758620689657,
      "loss": 2.1951069831848145,
      "step": 7408
    },
    {
      "epoch": 0.741,
      "grad_norm": 0.08132569491863251,
      "learning_rate": 0.001786896551724138,
      "loss": 2.185429573059082,
      "step": 7410
    },
    {
      "epoch": 0.7412,
      "grad_norm": 0.0792522206902504,
      "learning_rate": 0.0017855172413793103,
      "loss": 2.229841709136963,
      "step": 7412
    },
    {
      "epoch": 0.7414,
      "grad_norm": 0.08077333867549896,
      "learning_rate": 0.0017841379310344827,
      "loss": 2.177556037902832,
      "step": 7414
    },
    {
      "epoch": 0.7416,
      "grad_norm": 0.07736355066299438,
      "learning_rate": 0.001782758620689655,
      "loss": 2.159531593322754,
      "step": 7416
    },
    {
      "epoch": 0.7418,
      "grad_norm": 0.08323058485984802,
      "learning_rate": 0.0017813793103448278,
      "loss": 2.1865341663360596,
      "step": 7418
    },
    {
      "epoch": 0.742,
      "grad_norm": 0.08025029301643372,
      "learning_rate": 0.0017800000000000001,
      "loss": 2.2299106121063232,
      "step": 7420
    },
    {
      "epoch": 0.7422,
      "grad_norm": 0.0831066370010376,
      "learning_rate": 0.0017786206896551725,
      "loss": 2.2008056640625,
      "step": 7422
    },
    {
      "epoch": 0.7424,
      "grad_norm": 0.08331754058599472,
      "learning_rate": 0.0017772413793103448,
      "loss": 2.2460854053497314,
      "step": 7424
    },
    {
      "epoch": 0.7426,
      "grad_norm": 0.07705157995223999,
      "learning_rate": 0.0017758620689655173,
      "loss": 2.2941064834594727,
      "step": 7426
    },
    {
      "epoch": 0.7428,
      "grad_norm": 0.0811794176697731,
      "learning_rate": 0.0017744827586206897,
      "loss": 2.28470516204834,
      "step": 7428
    },
    {
      "epoch": 0.743,
      "grad_norm": 0.07709278911352158,
      "learning_rate": 0.0017731034482758622,
      "loss": 2.202225685119629,
      "step": 7430
    },
    {
      "epoch": 0.7432,
      "grad_norm": 0.08334296941757202,
      "learning_rate": 0.0017717241379310346,
      "loss": 2.163623094558716,
      "step": 7432
    },
    {
      "epoch": 0.7434,
      "grad_norm": 0.08690624684095383,
      "learning_rate": 0.001770344827586207,
      "loss": 2.1461105346679688,
      "step": 7434
    },
    {
      "epoch": 0.7436,
      "grad_norm": 0.08195517212152481,
      "learning_rate": 0.0017689655172413794,
      "loss": 2.2455496788024902,
      "step": 7436
    },
    {
      "epoch": 0.7438,
      "grad_norm": 0.07917207479476929,
      "learning_rate": 0.0017675862068965518,
      "loss": 2.2623798847198486,
      "step": 7438
    },
    {
      "epoch": 0.744,
      "grad_norm": 0.0790497213602066,
      "learning_rate": 0.0017662068965517241,
      "loss": 2.054762363433838,
      "step": 7440
    },
    {
      "epoch": 0.7442,
      "grad_norm": 0.07682965695858002,
      "learning_rate": 0.0017648275862068965,
      "loss": 2.180366039276123,
      "step": 7442
    },
    {
      "epoch": 0.7444,
      "grad_norm": 0.08216055482625961,
      "learning_rate": 0.001763448275862069,
      "loss": 2.2702462673187256,
      "step": 7444
    },
    {
      "epoch": 0.7446,
      "grad_norm": 0.07747960835695267,
      "learning_rate": 0.0017620689655172416,
      "loss": 2.248725652694702,
      "step": 7446
    },
    {
      "epoch": 0.7448,
      "grad_norm": 0.07707945257425308,
      "learning_rate": 0.001760689655172414,
      "loss": 2.1633899211883545,
      "step": 7448
    },
    {
      "epoch": 0.745,
      "grad_norm": 0.08368521183729172,
      "learning_rate": 0.0017593103448275862,
      "loss": 2.1458919048309326,
      "step": 7450
    },
    {
      "epoch": 0.7452,
      "grad_norm": 0.07771165668964386,
      "learning_rate": 0.0017579310344827586,
      "loss": 2.197068929672241,
      "step": 7452
    },
    {
      "epoch": 0.7454,
      "grad_norm": 0.07746849954128265,
      "learning_rate": 0.001756551724137931,
      "loss": 2.1545188426971436,
      "step": 7454
    },
    {
      "epoch": 0.7456,
      "grad_norm": 0.07802563160657883,
      "learning_rate": 0.0017551724137931035,
      "loss": 2.0985612869262695,
      "step": 7456
    },
    {
      "epoch": 0.7458,
      "grad_norm": 0.07800067961215973,
      "learning_rate": 0.001753793103448276,
      "loss": 2.2019426822662354,
      "step": 7458
    },
    {
      "epoch": 0.746,
      "grad_norm": 0.08202242106199265,
      "learning_rate": 0.0017524137931034483,
      "loss": 2.196044445037842,
      "step": 7460
    },
    {
      "epoch": 0.7462,
      "grad_norm": 0.08340528607368469,
      "learning_rate": 0.0017510344827586207,
      "loss": 2.3035547733306885,
      "step": 7462
    },
    {
      "epoch": 0.7464,
      "grad_norm": 0.07818252593278885,
      "learning_rate": 0.0017496551724137932,
      "loss": 2.2251386642456055,
      "step": 7464
    },
    {
      "epoch": 0.7466,
      "grad_norm": 0.0827045887708664,
      "learning_rate": 0.0017482758620689656,
      "loss": 2.2105207443237305,
      "step": 7466
    },
    {
      "epoch": 0.7468,
      "grad_norm": 0.07596834748983383,
      "learning_rate": 0.001746896551724138,
      "loss": 2.1919732093811035,
      "step": 7468
    },
    {
      "epoch": 0.747,
      "grad_norm": 0.07854852080345154,
      "learning_rate": 0.0017455172413793102,
      "loss": 2.165837526321411,
      "step": 7470
    },
    {
      "epoch": 0.7472,
      "grad_norm": 0.08142662793397903,
      "learning_rate": 0.0017441379310344828,
      "loss": 2.2175121307373047,
      "step": 7472
    },
    {
      "epoch": 0.7474,
      "grad_norm": 0.07995925843715668,
      "learning_rate": 0.0017427586206896553,
      "loss": 2.1464896202087402,
      "step": 7474
    },
    {
      "epoch": 0.7476,
      "grad_norm": 0.07482891529798508,
      "learning_rate": 0.0017413793103448277,
      "loss": 2.18076229095459,
      "step": 7476
    },
    {
      "epoch": 0.7478,
      "grad_norm": 0.08265764266252518,
      "learning_rate": 0.00174,
      "loss": 2.2098982334136963,
      "step": 7478
    },
    {
      "epoch": 0.748,
      "grad_norm": 0.08524929732084274,
      "learning_rate": 0.0017386206896551723,
      "loss": 2.2458105087280273,
      "step": 7480
    },
    {
      "epoch": 0.7482,
      "grad_norm": 0.08265812695026398,
      "learning_rate": 0.0017372413793103447,
      "loss": 2.290919065475464,
      "step": 7482
    },
    {
      "epoch": 0.7484,
      "grad_norm": 0.08349166810512543,
      "learning_rate": 0.0017358620689655175,
      "loss": 2.325352430343628,
      "step": 7484
    },
    {
      "epoch": 0.7486,
      "grad_norm": 0.0823255330324173,
      "learning_rate": 0.0017344827586206898,
      "loss": 2.1259093284606934,
      "step": 7486
    },
    {
      "epoch": 0.7488,
      "grad_norm": 0.07883036881685257,
      "learning_rate": 0.0017331034482758621,
      "loss": 2.239353895187378,
      "step": 7488
    },
    {
      "epoch": 0.749,
      "grad_norm": 0.08627406507730484,
      "learning_rate": 0.0017317241379310345,
      "loss": 2.07407283782959,
      "step": 7490
    },
    {
      "epoch": 0.7492,
      "grad_norm": 0.08116484433412552,
      "learning_rate": 0.001730344827586207,
      "loss": 2.1857073307037354,
      "step": 7492
    },
    {
      "epoch": 0.7494,
      "grad_norm": 0.07805582880973816,
      "learning_rate": 0.0017289655172413793,
      "loss": 2.2642600536346436,
      "step": 7494
    },
    {
      "epoch": 0.7496,
      "grad_norm": 0.0879882425069809,
      "learning_rate": 0.0017275862068965517,
      "loss": 2.271496057510376,
      "step": 7496
    },
    {
      "epoch": 0.7498,
      "grad_norm": 0.0805080458521843,
      "learning_rate": 0.0017262068965517242,
      "loss": 2.2225005626678467,
      "step": 7498
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.08105958253145218,
      "learning_rate": 0.0017248275862068966,
      "loss": 2.220454692840576,
      "step": 7500
    },
    {
      "epoch": 0.7502,
      "grad_norm": 0.08215310424566269,
      "learning_rate": 0.0017234482758620691,
      "loss": 2.1726293563842773,
      "step": 7502
    },
    {
      "epoch": 0.7504,
      "grad_norm": 0.07755617052316666,
      "learning_rate": 0.0017220689655172415,
      "loss": 2.138507843017578,
      "step": 7504
    },
    {
      "epoch": 0.7506,
      "grad_norm": 0.0794951543211937,
      "learning_rate": 0.0017206896551724138,
      "loss": 2.0330872535705566,
      "step": 7506
    },
    {
      "epoch": 0.7508,
      "grad_norm": 0.07874695211648941,
      "learning_rate": 0.0017193103448275861,
      "loss": 2.2345728874206543,
      "step": 7508
    },
    {
      "epoch": 0.751,
      "grad_norm": 0.08089590817689896,
      "learning_rate": 0.0017179310344827585,
      "loss": 2.1437175273895264,
      "step": 7510
    },
    {
      "epoch": 0.7512,
      "grad_norm": 0.0757904127240181,
      "learning_rate": 0.0017165517241379312,
      "loss": 2.1087188720703125,
      "step": 7512
    },
    {
      "epoch": 0.7514,
      "grad_norm": 0.08175050467252731,
      "learning_rate": 0.0017151724137931036,
      "loss": 2.246318817138672,
      "step": 7514
    },
    {
      "epoch": 0.7516,
      "grad_norm": 0.07591501623392105,
      "learning_rate": 0.001713793103448276,
      "loss": 2.1954143047332764,
      "step": 7516
    },
    {
      "epoch": 0.7518,
      "grad_norm": 0.08076362311840057,
      "learning_rate": 0.0017124137931034482,
      "loss": 2.1898975372314453,
      "step": 7518
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.08178997039794922,
      "learning_rate": 0.0017110344827586208,
      "loss": 2.183227062225342,
      "step": 7520
    },
    {
      "epoch": 0.7522,
      "grad_norm": 0.07470525056123734,
      "learning_rate": 0.0017096551724137931,
      "loss": 2.0809285640716553,
      "step": 7522
    },
    {
      "epoch": 0.7524,
      "grad_norm": 0.0773320198059082,
      "learning_rate": 0.0017082758620689657,
      "loss": 2.151867389678955,
      "step": 7524
    },
    {
      "epoch": 0.7526,
      "grad_norm": 0.08174049854278564,
      "learning_rate": 0.001706896551724138,
      "loss": 2.244729518890381,
      "step": 7526
    },
    {
      "epoch": 0.7528,
      "grad_norm": 0.0797489583492279,
      "learning_rate": 0.0017055172413793103,
      "loss": 2.206587076187134,
      "step": 7528
    },
    {
      "epoch": 0.753,
      "grad_norm": 0.07486908882856369,
      "learning_rate": 0.001704137931034483,
      "loss": 2.204963207244873,
      "step": 7530
    },
    {
      "epoch": 0.7532,
      "grad_norm": 0.07732000201940536,
      "learning_rate": 0.0017027586206896552,
      "loss": 2.2096030712127686,
      "step": 7532
    },
    {
      "epoch": 0.7534,
      "grad_norm": 0.07896760106086731,
      "learning_rate": 0.0017013793103448276,
      "loss": 2.265629529953003,
      "step": 7534
    },
    {
      "epoch": 0.7536,
      "grad_norm": 0.08080091327428818,
      "learning_rate": 0.0017,
      "loss": 2.1796061992645264,
      "step": 7536
    },
    {
      "epoch": 0.7538,
      "grad_norm": 0.07610408216714859,
      "learning_rate": 0.0016986206896551725,
      "loss": 2.1802561283111572,
      "step": 7538
    },
    {
      "epoch": 0.754,
      "grad_norm": 0.08694523572921753,
      "learning_rate": 0.001697241379310345,
      "loss": 2.2714006900787354,
      "step": 7540
    },
    {
      "epoch": 0.7542,
      "grad_norm": 0.07833575457334518,
      "learning_rate": 0.0016958620689655173,
      "loss": 2.2046923637390137,
      "step": 7542
    },
    {
      "epoch": 0.7544,
      "grad_norm": 0.08369411528110504,
      "learning_rate": 0.0016944827586206897,
      "loss": 2.159550666809082,
      "step": 7544
    },
    {
      "epoch": 0.7546,
      "grad_norm": 0.08643598109483719,
      "learning_rate": 0.001693103448275862,
      "loss": 2.1224539279937744,
      "step": 7546
    },
    {
      "epoch": 0.7548,
      "grad_norm": 0.08424069732427597,
      "learning_rate": 0.0016917241379310344,
      "loss": 2.2034997940063477,
      "step": 7548
    },
    {
      "epoch": 0.755,
      "grad_norm": 0.07829577475786209,
      "learning_rate": 0.001690344827586207,
      "loss": 2.168548345565796,
      "step": 7550
    },
    {
      "epoch": 0.7552,
      "grad_norm": 0.0813216045498848,
      "learning_rate": 0.0016889655172413795,
      "loss": 2.271493911743164,
      "step": 7552
    },
    {
      "epoch": 0.7554,
      "grad_norm": 0.08754345029592514,
      "learning_rate": 0.0016875862068965518,
      "loss": 2.2276132106781006,
      "step": 7554
    },
    {
      "epoch": 0.7556,
      "grad_norm": 0.0770009309053421,
      "learning_rate": 0.0016862068965517241,
      "loss": 2.156676769256592,
      "step": 7556
    },
    {
      "epoch": 0.7558,
      "grad_norm": 0.07719790935516357,
      "learning_rate": 0.0016848275862068967,
      "loss": 2.214024543762207,
      "step": 7558
    },
    {
      "epoch": 0.756,
      "grad_norm": 0.08043976128101349,
      "learning_rate": 0.001683448275862069,
      "loss": 2.109283924102783,
      "step": 7560
    },
    {
      "epoch": 0.7562,
      "grad_norm": 0.08468304574489594,
      "learning_rate": 0.0016820689655172413,
      "loss": 2.206822395324707,
      "step": 7562
    },
    {
      "epoch": 0.7564,
      "grad_norm": 0.07843149453401566,
      "learning_rate": 0.0016806896551724137,
      "loss": 2.171013116836548,
      "step": 7564
    },
    {
      "epoch": 0.7566,
      "grad_norm": 0.07655097544193268,
      "learning_rate": 0.0016793103448275862,
      "loss": 2.088608980178833,
      "step": 7566
    },
    {
      "epoch": 0.7568,
      "grad_norm": 0.0778559222817421,
      "learning_rate": 0.0016779310344827588,
      "loss": 2.254950761795044,
      "step": 7568
    },
    {
      "epoch": 0.757,
      "grad_norm": 0.08286519348621368,
      "learning_rate": 0.0016765517241379311,
      "loss": 2.244896173477173,
      "step": 7570
    },
    {
      "epoch": 0.7572,
      "grad_norm": 0.08312572538852692,
      "learning_rate": 0.0016751724137931035,
      "loss": 2.194303512573242,
      "step": 7572
    },
    {
      "epoch": 0.7574,
      "grad_norm": 0.07899332046508789,
      "learning_rate": 0.0016737931034482758,
      "loss": 2.155904769897461,
      "step": 7574
    },
    {
      "epoch": 0.7576,
      "grad_norm": 0.08221065998077393,
      "learning_rate": 0.0016724137931034481,
      "loss": 2.1814255714416504,
      "step": 7576
    },
    {
      "epoch": 0.7578,
      "grad_norm": 0.08236226439476013,
      "learning_rate": 0.001671034482758621,
      "loss": 2.2499003410339355,
      "step": 7578
    },
    {
      "epoch": 0.758,
      "grad_norm": 0.08236458897590637,
      "learning_rate": 0.0016696551724137932,
      "loss": 2.2695226669311523,
      "step": 7580
    },
    {
      "epoch": 0.7582,
      "grad_norm": 0.08215565234422684,
      "learning_rate": 0.0016682758620689656,
      "loss": 2.235947608947754,
      "step": 7582
    },
    {
      "epoch": 0.7584,
      "grad_norm": 0.08375836908817291,
      "learning_rate": 0.001666896551724138,
      "loss": 2.2475364208221436,
      "step": 7584
    },
    {
      "epoch": 0.7586,
      "grad_norm": 0.08162402361631393,
      "learning_rate": 0.0016655172413793105,
      "loss": 2.150113344192505,
      "step": 7586
    },
    {
      "epoch": 0.7588,
      "grad_norm": 0.07861527800559998,
      "learning_rate": 0.0016641379310344828,
      "loss": 2.309682607650757,
      "step": 7588
    },
    {
      "epoch": 0.759,
      "grad_norm": 0.08433007448911667,
      "learning_rate": 0.0016627586206896551,
      "loss": 2.223620891571045,
      "step": 7590
    },
    {
      "epoch": 0.7592,
      "grad_norm": 0.08411461859941483,
      "learning_rate": 0.0016613793103448277,
      "loss": 2.2867414951324463,
      "step": 7592
    },
    {
      "epoch": 0.7594,
      "grad_norm": 0.0768723413348198,
      "learning_rate": 0.00166,
      "loss": 2.2671031951904297,
      "step": 7594
    },
    {
      "epoch": 0.7596,
      "grad_norm": 0.08676881343126297,
      "learning_rate": 0.0016586206896551726,
      "loss": 2.207580089569092,
      "step": 7596
    },
    {
      "epoch": 0.7598,
      "grad_norm": 0.0796148031949997,
      "learning_rate": 0.001657241379310345,
      "loss": 2.193141460418701,
      "step": 7598
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.07776238024234772,
      "learning_rate": 0.0016558620689655172,
      "loss": 2.2861573696136475,
      "step": 7600
    },
    {
      "epoch": 0.7602,
      "grad_norm": 0.07577463984489441,
      "learning_rate": 0.0016544827586206896,
      "loss": 2.227980375289917,
      "step": 7602
    },
    {
      "epoch": 0.7604,
      "grad_norm": 0.08307389914989471,
      "learning_rate": 0.001653103448275862,
      "loss": 2.130309581756592,
      "step": 7604
    },
    {
      "epoch": 0.7606,
      "grad_norm": 0.08791155368089676,
      "learning_rate": 0.0016517241379310347,
      "loss": 2.2398159503936768,
      "step": 7606
    },
    {
      "epoch": 0.7608,
      "grad_norm": 0.07906752824783325,
      "learning_rate": 0.001650344827586207,
      "loss": 2.1096203327178955,
      "step": 7608
    },
    {
      "epoch": 0.761,
      "grad_norm": 0.07902688533067703,
      "learning_rate": 0.0016489655172413794,
      "loss": 2.2648775577545166,
      "step": 7610
    },
    {
      "epoch": 0.7612,
      "grad_norm": 0.08019865304231644,
      "learning_rate": 0.0016475862068965517,
      "loss": 2.1370482444763184,
      "step": 7612
    },
    {
      "epoch": 0.7614,
      "grad_norm": 0.08137810230255127,
      "learning_rate": 0.001646206896551724,
      "loss": 2.2363436222076416,
      "step": 7614
    },
    {
      "epoch": 0.7616,
      "grad_norm": 0.08392900228500366,
      "learning_rate": 0.0016448275862068966,
      "loss": 2.260500192642212,
      "step": 7616
    },
    {
      "epoch": 0.7618,
      "grad_norm": 0.0806586816906929,
      "learning_rate": 0.0016434482758620691,
      "loss": 2.2858192920684814,
      "step": 7618
    },
    {
      "epoch": 0.762,
      "grad_norm": 0.08220323175191879,
      "learning_rate": 0.0016420689655172415,
      "loss": 2.2091150283813477,
      "step": 7620
    },
    {
      "epoch": 0.7622,
      "grad_norm": 0.08193227648735046,
      "learning_rate": 0.0016406896551724138,
      "loss": 2.200533628463745,
      "step": 7622
    },
    {
      "epoch": 0.7624,
      "grad_norm": 0.0784020721912384,
      "learning_rate": 0.0016393103448275863,
      "loss": 2.269362688064575,
      "step": 7624
    },
    {
      "epoch": 0.7626,
      "grad_norm": 0.07883140444755554,
      "learning_rate": 0.0016379310344827587,
      "loss": 2.328253984451294,
      "step": 7626
    },
    {
      "epoch": 0.7628,
      "grad_norm": 0.07938855141401291,
      "learning_rate": 0.001636551724137931,
      "loss": 2.1759026050567627,
      "step": 7628
    },
    {
      "epoch": 0.763,
      "grad_norm": 0.08153939247131348,
      "learning_rate": 0.0016351724137931034,
      "loss": 2.3212337493896484,
      "step": 7630
    },
    {
      "epoch": 0.7632,
      "grad_norm": 0.08014828711748123,
      "learning_rate": 0.001633793103448276,
      "loss": 2.199690580368042,
      "step": 7632
    },
    {
      "epoch": 0.7634,
      "grad_norm": 0.07655312865972519,
      "learning_rate": 0.0016324137931034485,
      "loss": 2.2051587104797363,
      "step": 7634
    },
    {
      "epoch": 0.7636,
      "grad_norm": 0.08530862629413605,
      "learning_rate": 0.0016310344827586208,
      "loss": 2.3041017055511475,
      "step": 7636
    },
    {
      "epoch": 0.7638,
      "grad_norm": 0.07679013162851334,
      "learning_rate": 0.0016296551724137931,
      "loss": 2.13805890083313,
      "step": 7638
    },
    {
      "epoch": 0.764,
      "grad_norm": 0.07486862689256668,
      "learning_rate": 0.0016282758620689655,
      "loss": 2.243576765060425,
      "step": 7640
    },
    {
      "epoch": 0.7642,
      "grad_norm": 0.07755804806947708,
      "learning_rate": 0.0016268965517241378,
      "loss": 2.139188528060913,
      "step": 7642
    },
    {
      "epoch": 0.7644,
      "grad_norm": 0.0863494947552681,
      "learning_rate": 0.0016255172413793104,
      "loss": 2.137500524520874,
      "step": 7644
    },
    {
      "epoch": 0.7646,
      "grad_norm": 0.07637692242860794,
      "learning_rate": 0.001624137931034483,
      "loss": 2.165166139602661,
      "step": 7646
    },
    {
      "epoch": 0.7648,
      "grad_norm": 0.07926864176988602,
      "learning_rate": 0.0016227586206896552,
      "loss": 2.1481494903564453,
      "step": 7648
    },
    {
      "epoch": 0.765,
      "grad_norm": 0.08599768579006195,
      "learning_rate": 0.0016213793103448276,
      "loss": 2.240790605545044,
      "step": 7650
    },
    {
      "epoch": 0.7652,
      "grad_norm": 0.08610418438911438,
      "learning_rate": 0.0016200000000000001,
      "loss": 2.152756690979004,
      "step": 7652
    },
    {
      "epoch": 0.7654,
      "grad_norm": 0.07600505650043488,
      "learning_rate": 0.0016186206896551725,
      "loss": 2.235300064086914,
      "step": 7654
    },
    {
      "epoch": 0.7656,
      "grad_norm": 0.0821981132030487,
      "learning_rate": 0.0016172413793103448,
      "loss": 2.167221784591675,
      "step": 7656
    },
    {
      "epoch": 0.7658,
      "grad_norm": 0.0793757438659668,
      "learning_rate": 0.0016158620689655171,
      "loss": 2.2456166744232178,
      "step": 7658
    },
    {
      "epoch": 0.766,
      "grad_norm": 0.0820632129907608,
      "learning_rate": 0.0016144827586206897,
      "loss": 2.290071964263916,
      "step": 7660
    },
    {
      "epoch": 0.7662,
      "grad_norm": 0.07885101437568665,
      "learning_rate": 0.0016131034482758622,
      "loss": 2.176023244857788,
      "step": 7662
    },
    {
      "epoch": 0.7664,
      "grad_norm": 0.09262693673372269,
      "learning_rate": 0.0016117241379310346,
      "loss": 2.217634916305542,
      "step": 7664
    },
    {
      "epoch": 0.7666,
      "grad_norm": 0.08269712328910828,
      "learning_rate": 0.001610344827586207,
      "loss": 2.259113311767578,
      "step": 7666
    },
    {
      "epoch": 0.7668,
      "grad_norm": 0.07573617994785309,
      "learning_rate": 0.0016089655172413792,
      "loss": 2.2237911224365234,
      "step": 7668
    },
    {
      "epoch": 0.767,
      "grad_norm": 0.08028333634138107,
      "learning_rate": 0.0016075862068965516,
      "loss": 2.2020506858825684,
      "step": 7670
    },
    {
      "epoch": 0.7672,
      "grad_norm": 0.07983676344156265,
      "learning_rate": 0.0016062068965517244,
      "loss": 2.2861478328704834,
      "step": 7672
    },
    {
      "epoch": 0.7674,
      "grad_norm": 0.0774623304605484,
      "learning_rate": 0.0016048275862068967,
      "loss": 2.2170650959014893,
      "step": 7674
    },
    {
      "epoch": 0.7676,
      "grad_norm": 0.08171001821756363,
      "learning_rate": 0.001603448275862069,
      "loss": 2.145623207092285,
      "step": 7676
    },
    {
      "epoch": 0.7678,
      "grad_norm": 0.07777678966522217,
      "learning_rate": 0.0016020689655172414,
      "loss": 2.250150442123413,
      "step": 7678
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.07568249106407166,
      "learning_rate": 0.0016006896551724137,
      "loss": 2.1812191009521484,
      "step": 7680
    },
    {
      "epoch": 0.7682,
      "grad_norm": 0.08414747565984726,
      "learning_rate": 0.0015993103448275862,
      "loss": 2.197181224822998,
      "step": 7682
    },
    {
      "epoch": 0.7684,
      "grad_norm": 0.07522237300872803,
      "learning_rate": 0.0015979310344827586,
      "loss": 2.12785005569458,
      "step": 7684
    },
    {
      "epoch": 0.7686,
      "grad_norm": 0.07389989495277405,
      "learning_rate": 0.0015965517241379311,
      "loss": 2.2091407775878906,
      "step": 7686
    },
    {
      "epoch": 0.7688,
      "grad_norm": 0.0774855837225914,
      "learning_rate": 0.0015951724137931035,
      "loss": 2.1910736560821533,
      "step": 7688
    },
    {
      "epoch": 0.769,
      "grad_norm": 0.08214222639799118,
      "learning_rate": 0.001593793103448276,
      "loss": 2.3104286193847656,
      "step": 7690
    },
    {
      "epoch": 0.7692,
      "grad_norm": 0.07611072808504105,
      "learning_rate": 0.0015924137931034484,
      "loss": 2.189105272293091,
      "step": 7692
    },
    {
      "epoch": 0.7694,
      "grad_norm": 0.08163014054298401,
      "learning_rate": 0.0015910344827586207,
      "loss": 2.277026414871216,
      "step": 7694
    },
    {
      "epoch": 0.7696,
      "grad_norm": 0.08092281222343445,
      "learning_rate": 0.001589655172413793,
      "loss": 2.193686008453369,
      "step": 7696
    },
    {
      "epoch": 0.7698,
      "grad_norm": 0.08122935891151428,
      "learning_rate": 0.0015882758620689654,
      "loss": 2.232574462890625,
      "step": 7698
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.08177176862955093,
      "learning_rate": 0.0015868965517241381,
      "loss": 2.1833932399749756,
      "step": 7700
    },
    {
      "epoch": 0.7702,
      "grad_norm": 0.081122487783432,
      "learning_rate": 0.0015855172413793105,
      "loss": 2.165330171585083,
      "step": 7702
    },
    {
      "epoch": 0.7704,
      "grad_norm": 0.08104776591062546,
      "learning_rate": 0.0015841379310344828,
      "loss": 2.277902603149414,
      "step": 7704
    },
    {
      "epoch": 0.7706,
      "grad_norm": 0.07691297680139542,
      "learning_rate": 0.0015827586206896551,
      "loss": 2.182734727859497,
      "step": 7706
    },
    {
      "epoch": 0.7708,
      "grad_norm": 0.08338617533445358,
      "learning_rate": 0.0015813793103448275,
      "loss": 2.189148187637329,
      "step": 7708
    },
    {
      "epoch": 0.771,
      "grad_norm": 0.08908209949731827,
      "learning_rate": 0.00158,
      "loss": 2.3185372352600098,
      "step": 7710
    },
    {
      "epoch": 0.7712,
      "grad_norm": 0.08135868608951569,
      "learning_rate": 0.0015786206896551726,
      "loss": 2.2290539741516113,
      "step": 7712
    },
    {
      "epoch": 0.7714,
      "grad_norm": 0.07908999174833298,
      "learning_rate": 0.001577241379310345,
      "loss": 2.2680745124816895,
      "step": 7714
    },
    {
      "epoch": 0.7716,
      "grad_norm": 0.08183372020721436,
      "learning_rate": 0.0015758620689655172,
      "loss": 2.1116976737976074,
      "step": 7716
    },
    {
      "epoch": 0.7718,
      "grad_norm": 0.07743075489997864,
      "learning_rate": 0.0015744827586206898,
      "loss": 2.2214179039001465,
      "step": 7718
    },
    {
      "epoch": 0.772,
      "grad_norm": 0.07661928236484528,
      "learning_rate": 0.0015731034482758621,
      "loss": 2.244431495666504,
      "step": 7720
    },
    {
      "epoch": 0.7722,
      "grad_norm": 0.0796574056148529,
      "learning_rate": 0.0015717241379310345,
      "loss": 2.2716102600097656,
      "step": 7722
    },
    {
      "epoch": 0.7724,
      "grad_norm": 0.07824878394603729,
      "learning_rate": 0.0015703448275862068,
      "loss": 2.166890859603882,
      "step": 7724
    },
    {
      "epoch": 0.7726,
      "grad_norm": 0.07597117871046066,
      "learning_rate": 0.0015689655172413794,
      "loss": 2.2487411499023438,
      "step": 7726
    },
    {
      "epoch": 0.7728,
      "grad_norm": 0.08078086376190186,
      "learning_rate": 0.001567586206896552,
      "loss": 2.1265320777893066,
      "step": 7728
    },
    {
      "epoch": 0.773,
      "grad_norm": 0.08665042370557785,
      "learning_rate": 0.0015662068965517242,
      "loss": 2.216029405593872,
      "step": 7730
    },
    {
      "epoch": 0.7732,
      "grad_norm": 0.08460024744272232,
      "learning_rate": 0.0015648275862068966,
      "loss": 2.257350444793701,
      "step": 7732
    },
    {
      "epoch": 0.7734,
      "grad_norm": 0.08325197547674179,
      "learning_rate": 0.001563448275862069,
      "loss": 2.267395496368408,
      "step": 7734
    },
    {
      "epoch": 0.7736,
      "grad_norm": 0.0773928090929985,
      "learning_rate": 0.0015620689655172413,
      "loss": 2.2248926162719727,
      "step": 7736
    },
    {
      "epoch": 0.7738,
      "grad_norm": 0.07996456325054169,
      "learning_rate": 0.0015606896551724138,
      "loss": 2.2627217769622803,
      "step": 7738
    },
    {
      "epoch": 0.774,
      "grad_norm": 0.077995166182518,
      "learning_rate": 0.0015593103448275864,
      "loss": 2.2200565338134766,
      "step": 7740
    },
    {
      "epoch": 0.7742,
      "grad_norm": 0.0829506441950798,
      "learning_rate": 0.0015579310344827587,
      "loss": 2.2110931873321533,
      "step": 7742
    },
    {
      "epoch": 0.7744,
      "grad_norm": 0.07982036471366882,
      "learning_rate": 0.001556551724137931,
      "loss": 2.195258855819702,
      "step": 7744
    },
    {
      "epoch": 0.7746,
      "grad_norm": 0.07938794791698456,
      "learning_rate": 0.0015551724137931036,
      "loss": 2.2228333950042725,
      "step": 7746
    },
    {
      "epoch": 0.7748,
      "grad_norm": 0.08098748326301575,
      "learning_rate": 0.001553793103448276,
      "loss": 2.273986339569092,
      "step": 7748
    },
    {
      "epoch": 0.775,
      "grad_norm": 0.07884461432695389,
      "learning_rate": 0.0015524137931034483,
      "loss": 2.1566576957702637,
      "step": 7750
    },
    {
      "epoch": 0.7752,
      "grad_norm": 0.08365359902381897,
      "learning_rate": 0.0015510344827586206,
      "loss": 2.2068138122558594,
      "step": 7752
    },
    {
      "epoch": 0.7754,
      "grad_norm": 0.07908908277750015,
      "learning_rate": 0.0015496551724137931,
      "loss": 2.2721149921417236,
      "step": 7754
    },
    {
      "epoch": 0.7756,
      "grad_norm": 0.0737784206867218,
      "learning_rate": 0.0015482758620689657,
      "loss": 2.088411331176758,
      "step": 7756
    },
    {
      "epoch": 0.7758,
      "grad_norm": 0.07676439732313156,
      "learning_rate": 0.001546896551724138,
      "loss": 2.216423988342285,
      "step": 7758
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.07919684797525406,
      "learning_rate": 0.0015455172413793104,
      "loss": 2.2188527584075928,
      "step": 7760
    },
    {
      "epoch": 0.7762,
      "grad_norm": 0.07634162902832031,
      "learning_rate": 0.0015441379310344827,
      "loss": 2.204359531402588,
      "step": 7762
    },
    {
      "epoch": 0.7764,
      "grad_norm": 0.08151911944150925,
      "learning_rate": 0.001542758620689655,
      "loss": 2.2040491104125977,
      "step": 7764
    },
    {
      "epoch": 0.7766,
      "grad_norm": 0.08093851059675217,
      "learning_rate": 0.0015413793103448278,
      "loss": 2.207434892654419,
      "step": 7766
    },
    {
      "epoch": 0.7768,
      "grad_norm": 0.07635492831468582,
      "learning_rate": 0.0015400000000000001,
      "loss": 2.0984370708465576,
      "step": 7768
    },
    {
      "epoch": 0.777,
      "grad_norm": 0.08482514321804047,
      "learning_rate": 0.0015386206896551725,
      "loss": 2.1654391288757324,
      "step": 7770
    },
    {
      "epoch": 0.7772,
      "grad_norm": 0.08070098608732224,
      "learning_rate": 0.0015372413793103448,
      "loss": 2.2290940284729004,
      "step": 7772
    },
    {
      "epoch": 0.7774,
      "grad_norm": 0.0796789675951004,
      "learning_rate": 0.0015358620689655171,
      "loss": 2.1555838584899902,
      "step": 7774
    },
    {
      "epoch": 0.7776,
      "grad_norm": 0.08157386630773544,
      "learning_rate": 0.0015344827586206897,
      "loss": 2.223238468170166,
      "step": 7776
    },
    {
      "epoch": 0.7778,
      "grad_norm": 0.08163181692361832,
      "learning_rate": 0.001533103448275862,
      "loss": 2.305711507797241,
      "step": 7778
    },
    {
      "epoch": 0.778,
      "grad_norm": 0.07750622928142548,
      "learning_rate": 0.0015317241379310346,
      "loss": 2.2366135120391846,
      "step": 7780
    },
    {
      "epoch": 0.7782,
      "grad_norm": 0.07922708243131638,
      "learning_rate": 0.001530344827586207,
      "loss": 2.2158379554748535,
      "step": 7782
    },
    {
      "epoch": 0.7784,
      "grad_norm": 0.07896183431148529,
      "learning_rate": 0.0015289655172413795,
      "loss": 2.253439426422119,
      "step": 7784
    },
    {
      "epoch": 0.7786,
      "grad_norm": 0.0845339447259903,
      "learning_rate": 0.0015275862068965518,
      "loss": 2.1493031978607178,
      "step": 7786
    },
    {
      "epoch": 0.7788,
      "grad_norm": 0.08011936396360397,
      "learning_rate": 0.0015262068965517241,
      "loss": 2.203407049179077,
      "step": 7788
    },
    {
      "epoch": 0.779,
      "grad_norm": 0.08285480737686157,
      "learning_rate": 0.0015248275862068965,
      "loss": 2.252565622329712,
      "step": 7790
    },
    {
      "epoch": 0.7792,
      "grad_norm": 0.0790906772017479,
      "learning_rate": 0.0015234482758620688,
      "loss": 2.2130942344665527,
      "step": 7792
    },
    {
      "epoch": 0.7794,
      "grad_norm": 0.08131363987922668,
      "learning_rate": 0.0015220689655172416,
      "loss": 2.181152820587158,
      "step": 7794
    },
    {
      "epoch": 0.7796,
      "grad_norm": 0.07991079241037369,
      "learning_rate": 0.001520689655172414,
      "loss": 2.0841281414031982,
      "step": 7796
    },
    {
      "epoch": 0.7798,
      "grad_norm": 0.08005955070257187,
      "learning_rate": 0.0015193103448275863,
      "loss": 2.257326126098633,
      "step": 7798
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.08165935426950455,
      "learning_rate": 0.0015179310344827586,
      "loss": 2.3287129402160645,
      "step": 7800
    },
    {
      "epoch": 0.7802,
      "grad_norm": 0.0795530453324318,
      "learning_rate": 0.001516551724137931,
      "loss": 2.187593936920166,
      "step": 7802
    },
    {
      "epoch": 0.7804,
      "grad_norm": 0.08230245858430862,
      "learning_rate": 0.0015151724137931035,
      "loss": 2.219924211502075,
      "step": 7804
    },
    {
      "epoch": 0.7806,
      "grad_norm": 0.07812695950269699,
      "learning_rate": 0.001513793103448276,
      "loss": 2.260209321975708,
      "step": 7806
    },
    {
      "epoch": 0.7808,
      "grad_norm": 0.0811026468873024,
      "learning_rate": 0.0015124137931034484,
      "loss": 2.1912331581115723,
      "step": 7808
    },
    {
      "epoch": 0.781,
      "grad_norm": 0.08302277326583862,
      "learning_rate": 0.0015110344827586207,
      "loss": 2.225034475326538,
      "step": 7810
    },
    {
      "epoch": 0.7812,
      "grad_norm": 0.07955874502658844,
      "learning_rate": 0.0015096551724137932,
      "loss": 2.1965091228485107,
      "step": 7812
    },
    {
      "epoch": 0.7814,
      "grad_norm": 0.08275934308767319,
      "learning_rate": 0.0015082758620689656,
      "loss": 2.2075765132904053,
      "step": 7814
    },
    {
      "epoch": 0.7816,
      "grad_norm": 0.08100152015686035,
      "learning_rate": 0.001506896551724138,
      "loss": 2.1831817626953125,
      "step": 7816
    },
    {
      "epoch": 0.7818,
      "grad_norm": 0.08107851445674896,
      "learning_rate": 0.0015055172413793103,
      "loss": 2.2056024074554443,
      "step": 7818
    },
    {
      "epoch": 0.782,
      "grad_norm": 0.07852602005004883,
      "learning_rate": 0.0015041379310344828,
      "loss": 2.199537515640259,
      "step": 7820
    },
    {
      "epoch": 0.7822,
      "grad_norm": 0.08177196234464645,
      "learning_rate": 0.0015027586206896554,
      "loss": 2.271919012069702,
      "step": 7822
    },
    {
      "epoch": 0.7824,
      "grad_norm": 0.07757262140512466,
      "learning_rate": 0.0015013793103448277,
      "loss": 2.137545108795166,
      "step": 7824
    },
    {
      "epoch": 0.7826,
      "grad_norm": 0.0899883508682251,
      "learning_rate": 0.0015,
      "loss": 2.15350079536438,
      "step": 7826
    },
    {
      "epoch": 0.7828,
      "grad_norm": 0.0837680771946907,
      "learning_rate": 0.0014986206896551724,
      "loss": 2.2263128757476807,
      "step": 7828
    },
    {
      "epoch": 0.783,
      "grad_norm": 0.0811859518289566,
      "learning_rate": 0.0014972413793103447,
      "loss": 2.3136508464813232,
      "step": 7830
    },
    {
      "epoch": 0.7832,
      "grad_norm": 0.07782556861639023,
      "learning_rate": 0.0014958620689655173,
      "loss": 2.1318235397338867,
      "step": 7832
    },
    {
      "epoch": 0.7834,
      "grad_norm": 0.0857536643743515,
      "learning_rate": 0.0014944827586206898,
      "loss": 2.27470326423645,
      "step": 7834
    },
    {
      "epoch": 0.7836,
      "grad_norm": 0.08059096336364746,
      "learning_rate": 0.0014931034482758621,
      "loss": 2.244060516357422,
      "step": 7836
    },
    {
      "epoch": 0.7838,
      "grad_norm": 0.07558286190032959,
      "learning_rate": 0.0014917241379310345,
      "loss": 2.2349414825439453,
      "step": 7838
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.08088254183530807,
      "learning_rate": 0.0014903448275862068,
      "loss": 2.1793036460876465,
      "step": 7840
    },
    {
      "epoch": 0.7842,
      "grad_norm": 0.08410267531871796,
      "learning_rate": 0.0014889655172413794,
      "loss": 2.2907772064208984,
      "step": 7842
    },
    {
      "epoch": 0.7844,
      "grad_norm": 0.08167484402656555,
      "learning_rate": 0.0014875862068965517,
      "loss": 2.193671464920044,
      "step": 7844
    },
    {
      "epoch": 0.7846,
      "grad_norm": 0.07896294444799423,
      "learning_rate": 0.001486206896551724,
      "loss": 2.225841999053955,
      "step": 7846
    },
    {
      "epoch": 0.7848,
      "grad_norm": 0.09054478257894516,
      "learning_rate": 0.0014848275862068966,
      "loss": 2.2377490997314453,
      "step": 7848
    },
    {
      "epoch": 0.785,
      "grad_norm": 0.07774683088064194,
      "learning_rate": 0.0014834482758620691,
      "loss": 2.1688718795776367,
      "step": 7850
    },
    {
      "epoch": 0.7852,
      "grad_norm": 0.07824282348155975,
      "learning_rate": 0.0014820689655172415,
      "loss": 2.169656753540039,
      "step": 7852
    },
    {
      "epoch": 0.7854,
      "grad_norm": 0.08310148864984512,
      "learning_rate": 0.0014806896551724138,
      "loss": 2.2481210231781006,
      "step": 7854
    },
    {
      "epoch": 0.7856,
      "grad_norm": 0.08506441116333008,
      "learning_rate": 0.0014793103448275861,
      "loss": 2.163787364959717,
      "step": 7856
    },
    {
      "epoch": 0.7858,
      "grad_norm": 0.07818873971700668,
      "learning_rate": 0.0014779310344827585,
      "loss": 2.2556653022766113,
      "step": 7858
    },
    {
      "epoch": 0.786,
      "grad_norm": 0.07918113470077515,
      "learning_rate": 0.0014765517241379313,
      "loss": 2.268378734588623,
      "step": 7860
    },
    {
      "epoch": 0.7862,
      "grad_norm": 0.07894071936607361,
      "learning_rate": 0.0014751724137931036,
      "loss": 2.1745471954345703,
      "step": 7862
    },
    {
      "epoch": 0.7864,
      "grad_norm": 0.07930309325456619,
      "learning_rate": 0.001473793103448276,
      "loss": 2.1774344444274902,
      "step": 7864
    },
    {
      "epoch": 0.7866,
      "grad_norm": 0.07695753127336502,
      "learning_rate": 0.0014724137931034483,
      "loss": 2.247217893600464,
      "step": 7866
    },
    {
      "epoch": 0.7868,
      "grad_norm": 0.08415736258029938,
      "learning_rate": 0.0014710344827586206,
      "loss": 2.2290849685668945,
      "step": 7868
    },
    {
      "epoch": 0.787,
      "grad_norm": 0.07719022035598755,
      "learning_rate": 0.0014696551724137931,
      "loss": 2.189026117324829,
      "step": 7870
    },
    {
      "epoch": 0.7872,
      "grad_norm": 0.07827731221914291,
      "learning_rate": 0.0014682758620689655,
      "loss": 2.1517279148101807,
      "step": 7872
    },
    {
      "epoch": 0.7874,
      "grad_norm": 0.08329004794359207,
      "learning_rate": 0.001466896551724138,
      "loss": 2.202204465866089,
      "step": 7874
    },
    {
      "epoch": 0.7876,
      "grad_norm": 0.08134055882692337,
      "learning_rate": 0.0014655172413793104,
      "loss": 2.276446580886841,
      "step": 7876
    },
    {
      "epoch": 0.7878,
      "grad_norm": 0.08014576882123947,
      "learning_rate": 0.001464137931034483,
      "loss": 2.3130693435668945,
      "step": 7878
    },
    {
      "epoch": 0.788,
      "grad_norm": 0.07941871136426926,
      "learning_rate": 0.0014627586206896553,
      "loss": 2.1559910774230957,
      "step": 7880
    },
    {
      "epoch": 0.7882,
      "grad_norm": 0.0776563212275505,
      "learning_rate": 0.0014613793103448276,
      "loss": 2.2143423557281494,
      "step": 7882
    },
    {
      "epoch": 0.7884,
      "grad_norm": 0.08085295557975769,
      "learning_rate": 0.00146,
      "loss": 2.362873077392578,
      "step": 7884
    },
    {
      "epoch": 0.7886,
      "grad_norm": 0.08153900504112244,
      "learning_rate": 0.0014586206896551723,
      "loss": 2.2564127445220947,
      "step": 7886
    },
    {
      "epoch": 0.7888,
      "grad_norm": 0.07762106508016586,
      "learning_rate": 0.001457241379310345,
      "loss": 2.209336757659912,
      "step": 7888
    },
    {
      "epoch": 0.789,
      "grad_norm": 0.08201394230127335,
      "learning_rate": 0.0014558620689655174,
      "loss": 2.2395007610321045,
      "step": 7890
    },
    {
      "epoch": 0.7892,
      "grad_norm": 0.0739077627658844,
      "learning_rate": 0.0014544827586206897,
      "loss": 2.25726318359375,
      "step": 7892
    },
    {
      "epoch": 0.7894,
      "grad_norm": 0.08010048419237137,
      "learning_rate": 0.001453103448275862,
      "loss": 2.1300058364868164,
      "step": 7894
    },
    {
      "epoch": 0.7896,
      "grad_norm": 0.08377201855182648,
      "learning_rate": 0.0014517241379310344,
      "loss": 2.192377805709839,
      "step": 7896
    },
    {
      "epoch": 0.7898,
      "grad_norm": 0.0813211053609848,
      "learning_rate": 0.001450344827586207,
      "loss": 2.2913525104522705,
      "step": 7898
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.08125517517328262,
      "learning_rate": 0.0014489655172413795,
      "loss": 2.1377487182617188,
      "step": 7900
    },
    {
      "epoch": 0.7902,
      "grad_norm": 0.07978387176990509,
      "learning_rate": 0.0014475862068965518,
      "loss": 2.1467201709747314,
      "step": 7902
    },
    {
      "epoch": 0.7904,
      "grad_norm": 0.07762601226568222,
      "learning_rate": 0.0014462068965517241,
      "loss": 2.226804733276367,
      "step": 7904
    },
    {
      "epoch": 0.7906,
      "grad_norm": 0.07741605490446091,
      "learning_rate": 0.0014448275862068965,
      "loss": 2.2240636348724365,
      "step": 7906
    },
    {
      "epoch": 0.7908,
      "grad_norm": 0.08180098235607147,
      "learning_rate": 0.001443448275862069,
      "loss": 2.2942981719970703,
      "step": 7908
    },
    {
      "epoch": 0.791,
      "grad_norm": 0.0784553661942482,
      "learning_rate": 0.0014420689655172414,
      "loss": 2.167675018310547,
      "step": 7910
    },
    {
      "epoch": 0.7912,
      "grad_norm": 0.08142534643411636,
      "learning_rate": 0.0014406896551724137,
      "loss": 2.2151663303375244,
      "step": 7912
    },
    {
      "epoch": 0.7914,
      "grad_norm": 0.08076682686805725,
      "learning_rate": 0.0014393103448275863,
      "loss": 2.303284168243408,
      "step": 7914
    },
    {
      "epoch": 0.7916,
      "grad_norm": 0.07734204083681107,
      "learning_rate": 0.0014379310344827588,
      "loss": 2.243464469909668,
      "step": 7916
    },
    {
      "epoch": 0.7918,
      "grad_norm": 0.07821843773126602,
      "learning_rate": 0.0014365517241379311,
      "loss": 2.2503838539123535,
      "step": 7918
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.0842091292142868,
      "learning_rate": 0.0014351724137931035,
      "loss": 2.192127227783203,
      "step": 7920
    },
    {
      "epoch": 0.7922,
      "grad_norm": 0.0817052349448204,
      "learning_rate": 0.0014337931034482758,
      "loss": 2.262002468109131,
      "step": 7922
    },
    {
      "epoch": 0.7924,
      "grad_norm": 0.07967431843280792,
      "learning_rate": 0.0014324137931034482,
      "loss": 2.257807493209839,
      "step": 7924
    },
    {
      "epoch": 0.7926,
      "grad_norm": 0.07909440249204636,
      "learning_rate": 0.0014310344827586207,
      "loss": 2.0845887660980225,
      "step": 7926
    },
    {
      "epoch": 0.7928,
      "grad_norm": 0.07973084598779678,
      "learning_rate": 0.0014296551724137933,
      "loss": 2.228534460067749,
      "step": 7928
    },
    {
      "epoch": 0.793,
      "grad_norm": 0.0823158249258995,
      "learning_rate": 0.0014282758620689656,
      "loss": 2.242006301879883,
      "step": 7930
    },
    {
      "epoch": 0.7932,
      "grad_norm": 0.08106878399848938,
      "learning_rate": 0.001426896551724138,
      "loss": 2.2398111820220947,
      "step": 7932
    },
    {
      "epoch": 0.7934,
      "grad_norm": 0.0800270214676857,
      "learning_rate": 0.0014255172413793103,
      "loss": 2.2258222103118896,
      "step": 7934
    },
    {
      "epoch": 0.7936,
      "grad_norm": 0.07815147191286087,
      "learning_rate": 0.0014241379310344828,
      "loss": 2.247581720352173,
      "step": 7936
    },
    {
      "epoch": 0.7938,
      "grad_norm": 0.08537198603153229,
      "learning_rate": 0.0014227586206896552,
      "loss": 2.1839447021484375,
      "step": 7938
    },
    {
      "epoch": 0.794,
      "grad_norm": 0.08210943639278412,
      "learning_rate": 0.0014213793103448275,
      "loss": 2.3281304836273193,
      "step": 7940
    },
    {
      "epoch": 0.7942,
      "grad_norm": 0.08043132722377777,
      "learning_rate": 0.00142,
      "loss": 2.1857237815856934,
      "step": 7942
    },
    {
      "epoch": 0.7944,
      "grad_norm": 0.0818597748875618,
      "learning_rate": 0.0014186206896551726,
      "loss": 2.1833581924438477,
      "step": 7944
    },
    {
      "epoch": 0.7946,
      "grad_norm": 0.08058155328035355,
      "learning_rate": 0.001417241379310345,
      "loss": 2.2738025188446045,
      "step": 7946
    },
    {
      "epoch": 0.7948,
      "grad_norm": 0.07749499380588531,
      "learning_rate": 0.0014158620689655173,
      "loss": 2.204679012298584,
      "step": 7948
    },
    {
      "epoch": 0.795,
      "grad_norm": 0.07994908094406128,
      "learning_rate": 0.0014144827586206896,
      "loss": 2.246385335922241,
      "step": 7950
    },
    {
      "epoch": 0.7952,
      "grad_norm": 0.08074181526899338,
      "learning_rate": 0.001413103448275862,
      "loss": 2.187699794769287,
      "step": 7952
    },
    {
      "epoch": 0.7954,
      "grad_norm": 0.0843653753399849,
      "learning_rate": 0.0014117241379310347,
      "loss": 2.2270431518554688,
      "step": 7954
    },
    {
      "epoch": 0.7956,
      "grad_norm": 0.09045952558517456,
      "learning_rate": 0.001410344827586207,
      "loss": 2.19736909866333,
      "step": 7956
    },
    {
      "epoch": 0.7958,
      "grad_norm": 0.08220089226961136,
      "learning_rate": 0.0014089655172413794,
      "loss": 2.1904895305633545,
      "step": 7958
    },
    {
      "epoch": 0.796,
      "grad_norm": 0.08148471266031265,
      "learning_rate": 0.0014075862068965517,
      "loss": 2.224782705307007,
      "step": 7960
    },
    {
      "epoch": 0.7962,
      "grad_norm": 0.07662423700094223,
      "learning_rate": 0.001406206896551724,
      "loss": 2.2995688915252686,
      "step": 7962
    },
    {
      "epoch": 0.7964,
      "grad_norm": 0.08024570345878601,
      "learning_rate": 0.0014048275862068966,
      "loss": 2.23532772064209,
      "step": 7964
    },
    {
      "epoch": 0.7966,
      "grad_norm": 0.07719099521636963,
      "learning_rate": 0.001403448275862069,
      "loss": 2.2909491062164307,
      "step": 7966
    },
    {
      "epoch": 0.7968,
      "grad_norm": 0.08117994666099548,
      "learning_rate": 0.0014020689655172415,
      "loss": 2.1733195781707764,
      "step": 7968
    },
    {
      "epoch": 0.797,
      "grad_norm": 0.08103813230991364,
      "learning_rate": 0.0014006896551724138,
      "loss": 2.2893147468566895,
      "step": 7970
    },
    {
      "epoch": 0.7972,
      "grad_norm": 0.08117412775754929,
      "learning_rate": 0.0013993103448275864,
      "loss": 2.251180648803711,
      "step": 7972
    },
    {
      "epoch": 0.7974,
      "grad_norm": 0.08348733931779861,
      "learning_rate": 0.0013979310344827587,
      "loss": 2.290250062942505,
      "step": 7974
    },
    {
      "epoch": 0.7976,
      "grad_norm": 0.07757361978292465,
      "learning_rate": 0.001396551724137931,
      "loss": 2.3180391788482666,
      "step": 7976
    },
    {
      "epoch": 0.7978,
      "grad_norm": 0.07956240326166153,
      "learning_rate": 0.0013951724137931034,
      "loss": 2.329864263534546,
      "step": 7978
    },
    {
      "epoch": 0.798,
      "grad_norm": 0.08171912282705307,
      "learning_rate": 0.0013937931034482757,
      "loss": 2.2012884616851807,
      "step": 7980
    },
    {
      "epoch": 0.7982,
      "grad_norm": 0.07814043760299683,
      "learning_rate": 0.0013924137931034485,
      "loss": 2.241257429122925,
      "step": 7982
    },
    {
      "epoch": 0.7984,
      "grad_norm": 0.08455490320920944,
      "learning_rate": 0.0013910344827586208,
      "loss": 2.247533082962036,
      "step": 7984
    },
    {
      "epoch": 0.7986,
      "grad_norm": 0.07962596416473389,
      "learning_rate": 0.0013896551724137932,
      "loss": 2.254328966140747,
      "step": 7986
    },
    {
      "epoch": 0.7988,
      "grad_norm": 0.0790083259344101,
      "learning_rate": 0.0013882758620689655,
      "loss": 2.353860378265381,
      "step": 7988
    },
    {
      "epoch": 0.799,
      "grad_norm": 0.07588019222021103,
      "learning_rate": 0.0013868965517241378,
      "loss": 2.2494142055511475,
      "step": 7990
    },
    {
      "epoch": 0.7992,
      "grad_norm": 0.08224412798881531,
      "learning_rate": 0.0013855172413793104,
      "loss": 2.24320125579834,
      "step": 7992
    },
    {
      "epoch": 0.7994,
      "grad_norm": 0.07825282961130142,
      "learning_rate": 0.001384137931034483,
      "loss": 2.2986481189727783,
      "step": 7994
    },
    {
      "epoch": 0.7996,
      "grad_norm": 0.07898140698671341,
      "learning_rate": 0.0013827586206896553,
      "loss": 2.288921594619751,
      "step": 7996
    },
    {
      "epoch": 0.7998,
      "grad_norm": 0.08499843627214432,
      "learning_rate": 0.0013813793103448276,
      "loss": 2.1960766315460205,
      "step": 7998
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.0854124128818512,
      "learning_rate": 0.00138,
      "loss": 2.268019676208496,
      "step": 8000
    },
    {
      "epoch": 0.8,
      "eval_loss": 2.1292669773101807,
      "eval_runtime": 232.4556,
      "eval_samples_per_second": 2.151,
      "eval_steps_per_second": 0.215,
      "step": 8000
    },
    {
      "epoch": 0.8002,
      "grad_norm": 0.0822964683175087,
      "learning_rate": 0.0013786206896551725,
      "loss": 2.231445074081421,
      "step": 8002
    },
    {
      "epoch": 0.8004,
      "grad_norm": 0.07994810491800308,
      "learning_rate": 0.0013772413793103448,
      "loss": 2.275146722793579,
      "step": 8004
    },
    {
      "epoch": 0.8006,
      "grad_norm": 0.07795769721269608,
      "learning_rate": 0.0013758620689655172,
      "loss": 2.243046998977661,
      "step": 8006
    },
    {
      "epoch": 0.8008,
      "grad_norm": 0.07842851430177689,
      "learning_rate": 0.0013744827586206897,
      "loss": 2.2656383514404297,
      "step": 8008
    },
    {
      "epoch": 0.801,
      "grad_norm": 0.07895290851593018,
      "learning_rate": 0.0013731034482758623,
      "loss": 2.179978370666504,
      "step": 8010
    },
    {
      "epoch": 0.8012,
      "grad_norm": 0.08065899461507797,
      "learning_rate": 0.0013717241379310346,
      "loss": 2.287886142730713,
      "step": 8012
    },
    {
      "epoch": 0.8014,
      "grad_norm": 0.08087752014398575,
      "learning_rate": 0.001370344827586207,
      "loss": 2.247004747390747,
      "step": 8014
    },
    {
      "epoch": 0.8016,
      "grad_norm": 0.08383286744356155,
      "learning_rate": 0.0013689655172413793,
      "loss": 2.1660237312316895,
      "step": 8016
    },
    {
      "epoch": 0.8018,
      "grad_norm": 0.08190280199050903,
      "learning_rate": 0.0013675862068965516,
      "loss": 2.2273449897766113,
      "step": 8018
    },
    {
      "epoch": 0.802,
      "grad_norm": 0.07853353768587112,
      "learning_rate": 0.0013662068965517242,
      "loss": 2.2553036212921143,
      "step": 8020
    },
    {
      "epoch": 0.8022,
      "grad_norm": 0.07591172307729721,
      "learning_rate": 0.0013648275862068967,
      "loss": 2.224483013153076,
      "step": 8022
    },
    {
      "epoch": 0.8024,
      "grad_norm": 0.08745883405208588,
      "learning_rate": 0.001363448275862069,
      "loss": 2.2607054710388184,
      "step": 8024
    },
    {
      "epoch": 0.8026,
      "grad_norm": 0.07654345035552979,
      "learning_rate": 0.0013620689655172414,
      "loss": 2.346867322921753,
      "step": 8026
    },
    {
      "epoch": 0.8028,
      "grad_norm": 0.08455179631710052,
      "learning_rate": 0.0013606896551724137,
      "loss": 2.2552545070648193,
      "step": 8028
    },
    {
      "epoch": 0.803,
      "grad_norm": 0.07936019450426102,
      "learning_rate": 0.0013593103448275863,
      "loss": 2.264779567718506,
      "step": 8030
    },
    {
      "epoch": 0.8032,
      "grad_norm": 0.07887570559978485,
      "learning_rate": 0.0013579310344827586,
      "loss": 2.1520934104919434,
      "step": 8032
    },
    {
      "epoch": 0.8034,
      "grad_norm": 0.08063124120235443,
      "learning_rate": 0.001356551724137931,
      "loss": 2.360383987426758,
      "step": 8034
    },
    {
      "epoch": 0.8036,
      "grad_norm": 0.07420628517866135,
      "learning_rate": 0.0013551724137931035,
      "loss": 2.1888554096221924,
      "step": 8036
    },
    {
      "epoch": 0.8038,
      "grad_norm": 0.08026766777038574,
      "learning_rate": 0.001353793103448276,
      "loss": 2.2509076595306396,
      "step": 8038
    },
    {
      "epoch": 0.804,
      "grad_norm": 0.0805048942565918,
      "learning_rate": 0.0013524137931034484,
      "loss": 2.2850091457366943,
      "step": 8040
    },
    {
      "epoch": 0.8042,
      "grad_norm": 0.08338328450918198,
      "learning_rate": 0.0013510344827586207,
      "loss": 2.2811877727508545,
      "step": 8042
    },
    {
      "epoch": 0.8044,
      "grad_norm": 0.08355988562107086,
      "learning_rate": 0.001349655172413793,
      "loss": 2.225057601928711,
      "step": 8044
    },
    {
      "epoch": 0.8046,
      "grad_norm": 0.07885811477899551,
      "learning_rate": 0.0013482758620689654,
      "loss": 2.2058634757995605,
      "step": 8046
    },
    {
      "epoch": 0.8048,
      "grad_norm": 0.07670096307992935,
      "learning_rate": 0.0013468965517241382,
      "loss": 2.295389175415039,
      "step": 8048
    },
    {
      "epoch": 0.805,
      "grad_norm": 0.07632070034742355,
      "learning_rate": 0.0013455172413793105,
      "loss": 2.267148494720459,
      "step": 8050
    },
    {
      "epoch": 0.8052,
      "grad_norm": 0.08372079581022263,
      "learning_rate": 0.0013441379310344828,
      "loss": 2.247232675552368,
      "step": 8052
    },
    {
      "epoch": 0.8054,
      "grad_norm": 0.08219657093286514,
      "learning_rate": 0.0013427586206896552,
      "loss": 2.223492383956909,
      "step": 8054
    },
    {
      "epoch": 0.8056,
      "grad_norm": 0.08138206601142883,
      "learning_rate": 0.0013413793103448275,
      "loss": 2.276371717453003,
      "step": 8056
    },
    {
      "epoch": 0.8058,
      "grad_norm": 0.07471200823783875,
      "learning_rate": 0.00134,
      "loss": 2.229717493057251,
      "step": 8058
    },
    {
      "epoch": 0.806,
      "grad_norm": 0.08029186725616455,
      "learning_rate": 0.0013386206896551724,
      "loss": 2.2910876274108887,
      "step": 8060
    },
    {
      "epoch": 0.8062,
      "grad_norm": 0.07774993777275085,
      "learning_rate": 0.001337241379310345,
      "loss": 2.2465457916259766,
      "step": 8062
    },
    {
      "epoch": 0.8064,
      "grad_norm": 0.0753326341509819,
      "learning_rate": 0.0013358620689655173,
      "loss": 2.2378439903259277,
      "step": 8064
    },
    {
      "epoch": 0.8066,
      "grad_norm": 0.0814567431807518,
      "learning_rate": 0.0013344827586206896,
      "loss": 2.2043001651763916,
      "step": 8066
    },
    {
      "epoch": 0.8068,
      "grad_norm": 0.07852637767791748,
      "learning_rate": 0.0013331034482758622,
      "loss": 2.2775046825408936,
      "step": 8068
    },
    {
      "epoch": 0.807,
      "grad_norm": 0.08524804562330246,
      "learning_rate": 0.0013317241379310345,
      "loss": 2.2517943382263184,
      "step": 8070
    },
    {
      "epoch": 0.8072,
      "grad_norm": 0.0793514996767044,
      "learning_rate": 0.0013303448275862068,
      "loss": 2.209904670715332,
      "step": 8072
    },
    {
      "epoch": 0.8074,
      "grad_norm": 0.07987002283334732,
      "learning_rate": 0.0013289655172413792,
      "loss": 2.3282217979431152,
      "step": 8074
    },
    {
      "epoch": 0.8076,
      "grad_norm": 0.07683899998664856,
      "learning_rate": 0.001327586206896552,
      "loss": 2.2219276428222656,
      "step": 8076
    },
    {
      "epoch": 0.8078,
      "grad_norm": 0.0752185806632042,
      "learning_rate": 0.0013262068965517243,
      "loss": 2.261983633041382,
      "step": 8078
    },
    {
      "epoch": 0.808,
      "grad_norm": 0.07665837556123734,
      "learning_rate": 0.0013248275862068966,
      "loss": 2.160799503326416,
      "step": 8080
    },
    {
      "epoch": 0.8082,
      "grad_norm": 0.07604717463254929,
      "learning_rate": 0.001323448275862069,
      "loss": 2.280207395553589,
      "step": 8082
    },
    {
      "epoch": 0.8084,
      "grad_norm": 0.07835229486227036,
      "learning_rate": 0.0013220689655172413,
      "loss": 2.198436737060547,
      "step": 8084
    },
    {
      "epoch": 0.8086,
      "grad_norm": 0.07892300188541412,
      "learning_rate": 0.0013206896551724138,
      "loss": 2.1588375568389893,
      "step": 8086
    },
    {
      "epoch": 0.8088,
      "grad_norm": 0.08333923667669296,
      "learning_rate": 0.0013193103448275864,
      "loss": 2.262472152709961,
      "step": 8088
    },
    {
      "epoch": 0.809,
      "grad_norm": 0.08136303722858429,
      "learning_rate": 0.0013179310344827587,
      "loss": 2.332111358642578,
      "step": 8090
    },
    {
      "epoch": 0.8092,
      "grad_norm": 0.07782427966594696,
      "learning_rate": 0.001316551724137931,
      "loss": 2.2990317344665527,
      "step": 8092
    },
    {
      "epoch": 0.8094,
      "grad_norm": 0.08197788894176483,
      "learning_rate": 0.0013151724137931034,
      "loss": 2.3127524852752686,
      "step": 8094
    },
    {
      "epoch": 0.8096,
      "grad_norm": 0.07825304567813873,
      "learning_rate": 0.001313793103448276,
      "loss": 2.223666191101074,
      "step": 8096
    },
    {
      "epoch": 0.8098,
      "grad_norm": 0.08128377795219421,
      "learning_rate": 0.0013124137931034483,
      "loss": 2.3054027557373047,
      "step": 8098
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.07950598001480103,
      "learning_rate": 0.0013110344827586206,
      "loss": 2.265711784362793,
      "step": 8100
    },
    {
      "epoch": 0.8102,
      "grad_norm": 0.08748220652341843,
      "learning_rate": 0.0013096551724137932,
      "loss": 2.179694652557373,
      "step": 8102
    },
    {
      "epoch": 0.8104,
      "grad_norm": 0.0818571075797081,
      "learning_rate": 0.0013082758620689657,
      "loss": 2.3121464252471924,
      "step": 8104
    },
    {
      "epoch": 0.8106,
      "grad_norm": 0.08197981864213943,
      "learning_rate": 0.001306896551724138,
      "loss": 2.273561477661133,
      "step": 8106
    },
    {
      "epoch": 0.8108,
      "grad_norm": 0.07965533435344696,
      "learning_rate": 0.0013055172413793104,
      "loss": 2.126613140106201,
      "step": 8108
    },
    {
      "epoch": 0.811,
      "grad_norm": 0.07972276955842972,
      "learning_rate": 0.0013041379310344827,
      "loss": 2.315452814102173,
      "step": 8110
    },
    {
      "epoch": 0.8112,
      "grad_norm": 0.083716481924057,
      "learning_rate": 0.001302758620689655,
      "loss": 2.1757516860961914,
      "step": 8112
    },
    {
      "epoch": 0.8114,
      "grad_norm": 0.0793827474117279,
      "learning_rate": 0.0013013793103448276,
      "loss": 2.2028648853302,
      "step": 8114
    },
    {
      "epoch": 0.8116,
      "grad_norm": 0.08462642878293991,
      "learning_rate": 0.0013000000000000002,
      "loss": 2.290154218673706,
      "step": 8116
    },
    {
      "epoch": 0.8118,
      "grad_norm": 0.07887853682041168,
      "learning_rate": 0.0012986206896551725,
      "loss": 2.187685012817383,
      "step": 8118
    },
    {
      "epoch": 0.812,
      "grad_norm": 0.07769525796175003,
      "learning_rate": 0.0012972413793103448,
      "loss": 2.2753586769104004,
      "step": 8120
    },
    {
      "epoch": 0.8122,
      "grad_norm": 0.08299819380044937,
      "learning_rate": 0.0012958620689655172,
      "loss": 2.2671637535095215,
      "step": 8122
    },
    {
      "epoch": 0.8124,
      "grad_norm": 0.07896696031093597,
      "learning_rate": 0.0012944827586206897,
      "loss": 2.235421657562256,
      "step": 8124
    },
    {
      "epoch": 0.8126,
      "grad_norm": 0.08020088821649551,
      "learning_rate": 0.001293103448275862,
      "loss": 2.2425601482391357,
      "step": 8126
    },
    {
      "epoch": 0.8128,
      "grad_norm": 0.07824566960334778,
      "learning_rate": 0.0012917241379310344,
      "loss": 2.1971001625061035,
      "step": 8128
    },
    {
      "epoch": 0.813,
      "grad_norm": 0.07733284682035446,
      "learning_rate": 0.001290344827586207,
      "loss": 2.276799201965332,
      "step": 8130
    },
    {
      "epoch": 0.8132,
      "grad_norm": 0.07960283756256104,
      "learning_rate": 0.0012889655172413793,
      "loss": 2.283719062805176,
      "step": 8132
    },
    {
      "epoch": 0.8134,
      "grad_norm": 0.08190801739692688,
      "learning_rate": 0.0012875862068965518,
      "loss": 2.209043502807617,
      "step": 8134
    },
    {
      "epoch": 0.8136,
      "grad_norm": 0.0810534730553627,
      "learning_rate": 0.0012862068965517242,
      "loss": 2.266322374343872,
      "step": 8136
    },
    {
      "epoch": 0.8138,
      "grad_norm": 0.07858620584011078,
      "learning_rate": 0.0012848275862068965,
      "loss": 2.142792224884033,
      "step": 8138
    },
    {
      "epoch": 0.814,
      "grad_norm": 0.08134564757347107,
      "learning_rate": 0.0012834482758620688,
      "loss": 2.2346105575561523,
      "step": 8140
    },
    {
      "epoch": 0.8142,
      "grad_norm": 0.08399301022291183,
      "learning_rate": 0.0012820689655172416,
      "loss": 2.2570035457611084,
      "step": 8142
    },
    {
      "epoch": 0.8144,
      "grad_norm": 0.07985822856426239,
      "learning_rate": 0.001280689655172414,
      "loss": 2.213646650314331,
      "step": 8144
    },
    {
      "epoch": 0.8146,
      "grad_norm": 0.08244934678077698,
      "learning_rate": 0.0012793103448275863,
      "loss": 2.241314649581909,
      "step": 8146
    },
    {
      "epoch": 0.8148,
      "grad_norm": 0.0834265723824501,
      "learning_rate": 0.0012779310344827586,
      "loss": 2.167147636413574,
      "step": 8148
    },
    {
      "epoch": 0.815,
      "grad_norm": 0.0769837349653244,
      "learning_rate": 0.001276551724137931,
      "loss": 2.2877955436706543,
      "step": 8150
    },
    {
      "epoch": 0.8152,
      "grad_norm": 0.08006204664707184,
      "learning_rate": 0.0012751724137931035,
      "loss": 2.155499219894409,
      "step": 8152
    },
    {
      "epoch": 0.8154,
      "grad_norm": 0.07841562479734421,
      "learning_rate": 0.0012737931034482758,
      "loss": 2.244776964187622,
      "step": 8154
    },
    {
      "epoch": 0.8156,
      "grad_norm": 0.07936537265777588,
      "learning_rate": 0.0012724137931034484,
      "loss": 2.2221662998199463,
      "step": 8156
    },
    {
      "epoch": 0.8158,
      "grad_norm": 0.07961289584636688,
      "learning_rate": 0.0012710344827586207,
      "loss": 2.235325336456299,
      "step": 8158
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.07622797042131424,
      "learning_rate": 0.001269655172413793,
      "loss": 2.25036358833313,
      "step": 8160
    },
    {
      "epoch": 0.8162,
      "grad_norm": 0.07772400230169296,
      "learning_rate": 0.0012682758620689656,
      "loss": 2.3544504642486572,
      "step": 8162
    },
    {
      "epoch": 0.8164,
      "grad_norm": 0.07563842833042145,
      "learning_rate": 0.001266896551724138,
      "loss": 2.283782720565796,
      "step": 8164
    },
    {
      "epoch": 0.8166,
      "grad_norm": 0.08470213413238525,
      "learning_rate": 0.0012655172413793103,
      "loss": 2.3285224437713623,
      "step": 8166
    },
    {
      "epoch": 0.8168,
      "grad_norm": 0.07647120207548141,
      "learning_rate": 0.0012641379310344826,
      "loss": 2.300931930541992,
      "step": 8168
    },
    {
      "epoch": 0.817,
      "grad_norm": 0.09246749430894852,
      "learning_rate": 0.0012627586206896554,
      "loss": 2.203627824783325,
      "step": 8170
    },
    {
      "epoch": 0.8172,
      "grad_norm": 0.07770388573408127,
      "learning_rate": 0.0012613793103448277,
      "loss": 2.266324758529663,
      "step": 8172
    },
    {
      "epoch": 0.8174,
      "grad_norm": 0.07687217742204666,
      "learning_rate": 0.00126,
      "loss": 2.251319169998169,
      "step": 8174
    },
    {
      "epoch": 0.8176,
      "grad_norm": 0.08421613276004791,
      "learning_rate": 0.0012586206896551724,
      "loss": 2.286240816116333,
      "step": 8176
    },
    {
      "epoch": 0.8178,
      "grad_norm": 0.08317701518535614,
      "learning_rate": 0.0012572413793103447,
      "loss": 2.2832653522491455,
      "step": 8178
    },
    {
      "epoch": 0.818,
      "grad_norm": 0.07697004824876785,
      "learning_rate": 0.0012558620689655173,
      "loss": 2.173602819442749,
      "step": 8180
    },
    {
      "epoch": 0.8182,
      "grad_norm": 0.08189192414283752,
      "learning_rate": 0.0012544827586206898,
      "loss": 2.226740837097168,
      "step": 8182
    },
    {
      "epoch": 0.8184,
      "grad_norm": 0.08013266324996948,
      "learning_rate": 0.0012531034482758622,
      "loss": 2.2594079971313477,
      "step": 8184
    },
    {
      "epoch": 0.8186,
      "grad_norm": 0.0833071917295456,
      "learning_rate": 0.0012517241379310345,
      "loss": 2.3045592308044434,
      "step": 8186
    },
    {
      "epoch": 0.8188,
      "grad_norm": 0.07781028002500534,
      "learning_rate": 0.0012503448275862068,
      "loss": 2.1100564002990723,
      "step": 8188
    },
    {
      "epoch": 0.819,
      "grad_norm": 0.07896687090396881,
      "learning_rate": 0.0012489655172413794,
      "loss": 2.2547695636749268,
      "step": 8190
    },
    {
      "epoch": 0.8192,
      "grad_norm": 0.08070356398820877,
      "learning_rate": 0.0012475862068965517,
      "loss": 2.2141294479370117,
      "step": 8192
    },
    {
      "epoch": 0.8194,
      "grad_norm": 0.08155851811170578,
      "learning_rate": 0.001246206896551724,
      "loss": 2.2618868350982666,
      "step": 8194
    },
    {
      "epoch": 0.8196,
      "grad_norm": 0.08035022020339966,
      "learning_rate": 0.0012448275862068966,
      "loss": 2.3025057315826416,
      "step": 8196
    },
    {
      "epoch": 0.8198,
      "grad_norm": 0.07711077481508255,
      "learning_rate": 0.0012434482758620692,
      "loss": 2.27168607711792,
      "step": 8198
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.08002863079309464,
      "learning_rate": 0.0012420689655172415,
      "loss": 2.265057325363159,
      "step": 8200
    },
    {
      "epoch": 0.8202,
      "grad_norm": 0.0832831859588623,
      "learning_rate": 0.0012406896551724138,
      "loss": 2.21736741065979,
      "step": 8202
    },
    {
      "epoch": 0.8204,
      "grad_norm": 0.0791119933128357,
      "learning_rate": 0.0012393103448275862,
      "loss": 2.2394304275512695,
      "step": 8204
    },
    {
      "epoch": 0.8206,
      "grad_norm": 0.08052811026573181,
      "learning_rate": 0.0012379310344827585,
      "loss": 2.2172484397888184,
      "step": 8206
    },
    {
      "epoch": 0.8208,
      "grad_norm": 0.07862995564937592,
      "learning_rate": 0.001236551724137931,
      "loss": 2.222612142562866,
      "step": 8208
    },
    {
      "epoch": 0.821,
      "grad_norm": 0.08151400834321976,
      "learning_rate": 0.0012351724137931036,
      "loss": 2.164564371109009,
      "step": 8210
    },
    {
      "epoch": 0.8212,
      "grad_norm": 0.08037521690130234,
      "learning_rate": 0.001233793103448276,
      "loss": 2.2081451416015625,
      "step": 8212
    },
    {
      "epoch": 0.8214,
      "grad_norm": 0.08150716871023178,
      "learning_rate": 0.0012324137931034483,
      "loss": 2.262235641479492,
      "step": 8214
    },
    {
      "epoch": 0.8216,
      "grad_norm": 0.07760751247406006,
      "learning_rate": 0.0012310344827586206,
      "loss": 2.2042465209960938,
      "step": 8216
    },
    {
      "epoch": 0.8218,
      "grad_norm": 0.08038149029016495,
      "learning_rate": 0.0012296551724137932,
      "loss": 2.289816379547119,
      "step": 8218
    },
    {
      "epoch": 0.822,
      "grad_norm": 0.08036130666732788,
      "learning_rate": 0.0012282758620689655,
      "loss": 2.262690305709839,
      "step": 8220
    },
    {
      "epoch": 0.8222,
      "grad_norm": 0.07825655490159988,
      "learning_rate": 0.0012268965517241378,
      "loss": 2.2802767753601074,
      "step": 8222
    },
    {
      "epoch": 0.8224,
      "grad_norm": 0.08123378455638885,
      "learning_rate": 0.0012255172413793104,
      "loss": 2.333155870437622,
      "step": 8224
    },
    {
      "epoch": 0.8226,
      "grad_norm": 0.08171836286783218,
      "learning_rate": 0.0012241379310344827,
      "loss": 2.307440996170044,
      "step": 8226
    },
    {
      "epoch": 0.8228,
      "grad_norm": 0.0790727362036705,
      "learning_rate": 0.0012227586206896553,
      "loss": 2.2021076679229736,
      "step": 8228
    },
    {
      "epoch": 0.823,
      "grad_norm": 0.07972564548254013,
      "learning_rate": 0.0012213793103448276,
      "loss": 2.176894187927246,
      "step": 8230
    },
    {
      "epoch": 0.8232,
      "grad_norm": 0.0799635499715805,
      "learning_rate": 0.00122,
      "loss": 2.25286865234375,
      "step": 8232
    },
    {
      "epoch": 0.8234,
      "grad_norm": 0.07873166352510452,
      "learning_rate": 0.0012186206896551723,
      "loss": 2.319401979446411,
      "step": 8234
    },
    {
      "epoch": 0.8236,
      "grad_norm": 0.07813207805156708,
      "learning_rate": 0.001217241379310345,
      "loss": 2.2149863243103027,
      "step": 8236
    },
    {
      "epoch": 0.8238,
      "grad_norm": 0.07913829386234283,
      "learning_rate": 0.0012158620689655174,
      "loss": 2.22713303565979,
      "step": 8238
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.07510267198085785,
      "learning_rate": 0.0012144827586206897,
      "loss": 2.2263176441192627,
      "step": 8240
    },
    {
      "epoch": 0.8242,
      "grad_norm": 0.0804116427898407,
      "learning_rate": 0.001213103448275862,
      "loss": 2.2201497554779053,
      "step": 8242
    },
    {
      "epoch": 0.8244,
      "grad_norm": 0.0791761577129364,
      "learning_rate": 0.0012117241379310344,
      "loss": 2.215348243713379,
      "step": 8244
    },
    {
      "epoch": 0.8246,
      "grad_norm": 0.0782235711812973,
      "learning_rate": 0.001210344827586207,
      "loss": 2.265529155731201,
      "step": 8246
    },
    {
      "epoch": 0.8248,
      "grad_norm": 0.08005128800868988,
      "learning_rate": 0.0012089655172413793,
      "loss": 2.22890305519104,
      "step": 8248
    },
    {
      "epoch": 0.825,
      "grad_norm": 0.08349703252315521,
      "learning_rate": 0.0012075862068965518,
      "loss": 2.3492298126220703,
      "step": 8250
    },
    {
      "epoch": 0.8252,
      "grad_norm": 0.08281335234642029,
      "learning_rate": 0.0012062068965517242,
      "loss": 2.2582502365112305,
      "step": 8252
    },
    {
      "epoch": 0.8254,
      "grad_norm": 0.07966741174459457,
      "learning_rate": 0.0012048275862068965,
      "loss": 2.195415496826172,
      "step": 8254
    },
    {
      "epoch": 0.8256,
      "grad_norm": 0.08138450235128403,
      "learning_rate": 0.001203448275862069,
      "loss": 2.269470453262329,
      "step": 8256
    },
    {
      "epoch": 0.8258,
      "grad_norm": 0.07850107550621033,
      "learning_rate": 0.0012020689655172414,
      "loss": 2.239987373352051,
      "step": 8258
    },
    {
      "epoch": 0.826,
      "grad_norm": 0.07602666318416595,
      "learning_rate": 0.0012006896551724137,
      "loss": 2.266747236251831,
      "step": 8260
    },
    {
      "epoch": 0.8262,
      "grad_norm": 0.07767262309789658,
      "learning_rate": 0.001199310344827586,
      "loss": 2.226301431655884,
      "step": 8262
    },
    {
      "epoch": 0.8264,
      "grad_norm": 0.08522136509418488,
      "learning_rate": 0.0011979310344827588,
      "loss": 2.279536247253418,
      "step": 8264
    },
    {
      "epoch": 0.8266,
      "grad_norm": 0.26142820715904236,
      "learning_rate": 0.0011965517241379312,
      "loss": 2.279280424118042,
      "step": 8266
    },
    {
      "epoch": 0.8268,
      "grad_norm": 0.08278573304414749,
      "learning_rate": 0.0011951724137931035,
      "loss": 2.2241225242614746,
      "step": 8268
    },
    {
      "epoch": 0.827,
      "grad_norm": 0.08848052471876144,
      "learning_rate": 0.0011937931034482758,
      "loss": 2.3196468353271484,
      "step": 8270
    },
    {
      "epoch": 0.8272,
      "grad_norm": 0.0813036635518074,
      "learning_rate": 0.0011924137931034482,
      "loss": 2.2600340843200684,
      "step": 8272
    },
    {
      "epoch": 0.8274,
      "grad_norm": 0.0802016481757164,
      "learning_rate": 0.0011910344827586207,
      "loss": 2.2205848693847656,
      "step": 8274
    },
    {
      "epoch": 0.8276,
      "grad_norm": 0.0775938406586647,
      "learning_rate": 0.0011896551724137933,
      "loss": 2.2318036556243896,
      "step": 8276
    },
    {
      "epoch": 0.8278,
      "grad_norm": 0.0788973942399025,
      "learning_rate": 0.0011882758620689656,
      "loss": 2.28296160697937,
      "step": 8278
    },
    {
      "epoch": 0.828,
      "grad_norm": 0.08003505319356918,
      "learning_rate": 0.001186896551724138,
      "loss": 2.2350189685821533,
      "step": 8280
    },
    {
      "epoch": 0.8282,
      "grad_norm": 0.08100344240665436,
      "learning_rate": 0.0011855172413793103,
      "loss": 2.207869291305542,
      "step": 8282
    },
    {
      "epoch": 0.8284,
      "grad_norm": 0.07845170795917511,
      "learning_rate": 0.0011841379310344828,
      "loss": 2.228118658065796,
      "step": 8284
    },
    {
      "epoch": 0.8286,
      "grad_norm": 0.08850737661123276,
      "learning_rate": 0.0011827586206896552,
      "loss": 2.140287160873413,
      "step": 8286
    },
    {
      "epoch": 0.8288,
      "grad_norm": 0.08282245695590973,
      "learning_rate": 0.0011813793103448275,
      "loss": 2.213566780090332,
      "step": 8288
    },
    {
      "epoch": 0.829,
      "grad_norm": 0.07928934693336487,
      "learning_rate": 0.00118,
      "loss": 2.2883315086364746,
      "step": 8290
    },
    {
      "epoch": 0.8292,
      "grad_norm": 0.08545828610658646,
      "learning_rate": 0.0011786206896551724,
      "loss": 2.241973638534546,
      "step": 8292
    },
    {
      "epoch": 0.8294,
      "grad_norm": 0.08122190833091736,
      "learning_rate": 0.001177241379310345,
      "loss": 2.2567219734191895,
      "step": 8294
    },
    {
      "epoch": 0.8296,
      "grad_norm": 0.08476492017507553,
      "learning_rate": 0.0011758620689655173,
      "loss": 2.265932321548462,
      "step": 8296
    },
    {
      "epoch": 0.8298,
      "grad_norm": 0.07942639291286469,
      "learning_rate": 0.0011744827586206896,
      "loss": 2.3538546562194824,
      "step": 8298
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.08652371168136597,
      "learning_rate": 0.001173103448275862,
      "loss": 2.232773542404175,
      "step": 8300
    },
    {
      "epoch": 0.8302,
      "grad_norm": 0.07672926038503647,
      "learning_rate": 0.0011717241379310345,
      "loss": 2.2646825313568115,
      "step": 8302
    },
    {
      "epoch": 0.8304,
      "grad_norm": 0.08045296370983124,
      "learning_rate": 0.001170344827586207,
      "loss": 2.3227756023406982,
      "step": 8304
    },
    {
      "epoch": 0.8306,
      "grad_norm": 0.07748139649629593,
      "learning_rate": 0.0011689655172413794,
      "loss": 2.2304341793060303,
      "step": 8306
    },
    {
      "epoch": 0.8308,
      "grad_norm": 0.07768302410840988,
      "learning_rate": 0.0011675862068965517,
      "loss": 2.2259466648101807,
      "step": 8308
    },
    {
      "epoch": 0.831,
      "grad_norm": 0.08765775710344315,
      "learning_rate": 0.001166206896551724,
      "loss": 2.2978668212890625,
      "step": 8310
    },
    {
      "epoch": 0.8312,
      "grad_norm": 0.07871970534324646,
      "learning_rate": 0.0011648275862068966,
      "loss": 2.3637287616729736,
      "step": 8312
    },
    {
      "epoch": 0.8314,
      "grad_norm": 0.08069834113121033,
      "learning_rate": 0.001163448275862069,
      "loss": 2.265223741531372,
      "step": 8314
    },
    {
      "epoch": 0.8316,
      "grad_norm": 0.08335893601179123,
      "learning_rate": 0.0011620689655172413,
      "loss": 2.2736430168151855,
      "step": 8316
    },
    {
      "epoch": 0.8318,
      "grad_norm": 0.0817859098315239,
      "learning_rate": 0.0011606896551724138,
      "loss": 2.2830843925476074,
      "step": 8318
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.07991389185190201,
      "learning_rate": 0.0011593103448275862,
      "loss": 2.2987842559814453,
      "step": 8320
    },
    {
      "epoch": 0.8322,
      "grad_norm": 0.0827486515045166,
      "learning_rate": 0.0011579310344827587,
      "loss": 2.203901767730713,
      "step": 8322
    },
    {
      "epoch": 0.8324,
      "grad_norm": 0.07941918075084686,
      "learning_rate": 0.001156551724137931,
      "loss": 2.268815279006958,
      "step": 8324
    },
    {
      "epoch": 0.8326,
      "grad_norm": 0.08113035559654236,
      "learning_rate": 0.0011551724137931034,
      "loss": 2.250140905380249,
      "step": 8326
    },
    {
      "epoch": 0.8328,
      "grad_norm": 0.07972580194473267,
      "learning_rate": 0.0011537931034482757,
      "loss": 2.218916893005371,
      "step": 8328
    },
    {
      "epoch": 0.833,
      "grad_norm": 0.08105263859033585,
      "learning_rate": 0.0011524137931034485,
      "loss": 2.316554546356201,
      "step": 8330
    },
    {
      "epoch": 0.8332,
      "grad_norm": 0.08157651126384735,
      "learning_rate": 0.0011510344827586208,
      "loss": 2.199338674545288,
      "step": 8332
    },
    {
      "epoch": 0.8334,
      "grad_norm": 0.09644006937742233,
      "learning_rate": 0.0011496551724137932,
      "loss": 2.287956714630127,
      "step": 8334
    },
    {
      "epoch": 0.8336,
      "grad_norm": 0.07748743146657944,
      "learning_rate": 0.0011482758620689655,
      "loss": 2.268951416015625,
      "step": 8336
    },
    {
      "epoch": 0.8338,
      "grad_norm": 0.08142866939306259,
      "learning_rate": 0.0011468965517241378,
      "loss": 2.2009823322296143,
      "step": 8338
    },
    {
      "epoch": 0.834,
      "grad_norm": 0.07910612225532532,
      "learning_rate": 0.0011455172413793104,
      "loss": 2.2419958114624023,
      "step": 8340
    },
    {
      "epoch": 0.8342,
      "grad_norm": 0.08341453224420547,
      "learning_rate": 0.0011441379310344827,
      "loss": 2.3334736824035645,
      "step": 8342
    },
    {
      "epoch": 0.8344,
      "grad_norm": 0.0862821638584137,
      "learning_rate": 0.0011427586206896553,
      "loss": 2.4052226543426514,
      "step": 8344
    },
    {
      "epoch": 0.8346,
      "grad_norm": 0.08013910800218582,
      "learning_rate": 0.0011413793103448276,
      "loss": 2.246532440185547,
      "step": 8346
    },
    {
      "epoch": 0.8348,
      "grad_norm": 0.07806544005870819,
      "learning_rate": 0.00114,
      "loss": 2.27583909034729,
      "step": 8348
    },
    {
      "epoch": 0.835,
      "grad_norm": 0.08055407553911209,
      "learning_rate": 0.0011386206896551725,
      "loss": 2.2122137546539307,
      "step": 8350
    },
    {
      "epoch": 0.8352,
      "grad_norm": 0.08062446117401123,
      "learning_rate": 0.0011372413793103448,
      "loss": 2.3142173290252686,
      "step": 8352
    },
    {
      "epoch": 0.8354,
      "grad_norm": 0.08001694083213806,
      "learning_rate": 0.0011358620689655172,
      "loss": 2.24208664894104,
      "step": 8354
    },
    {
      "epoch": 0.8356,
      "grad_norm": 0.08063651621341705,
      "learning_rate": 0.0011344827586206895,
      "loss": 2.235323190689087,
      "step": 8356
    },
    {
      "epoch": 0.8358,
      "grad_norm": 0.07819624245166779,
      "learning_rate": 0.0011331034482758623,
      "loss": 2.2994794845581055,
      "step": 8358
    },
    {
      "epoch": 0.836,
      "grad_norm": 0.07856627553701401,
      "learning_rate": 0.0011317241379310346,
      "loss": 2.2474117279052734,
      "step": 8360
    },
    {
      "epoch": 0.8362,
      "grad_norm": 0.08174595981836319,
      "learning_rate": 0.001130344827586207,
      "loss": 2.351923942565918,
      "step": 8362
    },
    {
      "epoch": 0.8364,
      "grad_norm": 0.08435530215501785,
      "learning_rate": 0.0011289655172413793,
      "loss": 2.278897523880005,
      "step": 8364
    },
    {
      "epoch": 0.8366,
      "grad_norm": 0.07849057763814926,
      "learning_rate": 0.0011275862068965516,
      "loss": 2.2932190895080566,
      "step": 8366
    },
    {
      "epoch": 0.8368,
      "grad_norm": 0.08343887329101562,
      "learning_rate": 0.0011262068965517242,
      "loss": 2.232908010482788,
      "step": 8368
    },
    {
      "epoch": 0.837,
      "grad_norm": 0.0792740061879158,
      "learning_rate": 0.0011248275862068967,
      "loss": 2.2873318195343018,
      "step": 8370
    },
    {
      "epoch": 0.8372,
      "grad_norm": 0.07867293804883957,
      "learning_rate": 0.001123448275862069,
      "loss": 2.3666303157806396,
      "step": 8372
    },
    {
      "epoch": 0.8374,
      "grad_norm": 0.07864101976156235,
      "learning_rate": 0.0011220689655172414,
      "loss": 2.2374267578125,
      "step": 8374
    },
    {
      "epoch": 0.8376,
      "grad_norm": 0.07864057272672653,
      "learning_rate": 0.0011206896551724137,
      "loss": 2.1991355419158936,
      "step": 8376
    },
    {
      "epoch": 0.8378,
      "grad_norm": 0.08170250803232193,
      "learning_rate": 0.0011193103448275863,
      "loss": 2.2609634399414062,
      "step": 8378
    },
    {
      "epoch": 0.838,
      "grad_norm": 0.07920057326555252,
      "learning_rate": 0.0011179310344827586,
      "loss": 2.148421287536621,
      "step": 8380
    },
    {
      "epoch": 0.8382,
      "grad_norm": 0.08171901106834412,
      "learning_rate": 0.001116551724137931,
      "loss": 2.2704949378967285,
      "step": 8382
    },
    {
      "epoch": 0.8384,
      "grad_norm": 0.07938747853040695,
      "learning_rate": 0.0011151724137931035,
      "loss": 2.2884647846221924,
      "step": 8384
    },
    {
      "epoch": 0.8386,
      "grad_norm": 0.07905926555395126,
      "learning_rate": 0.0011137931034482758,
      "loss": 2.212705135345459,
      "step": 8386
    },
    {
      "epoch": 0.8388,
      "grad_norm": 0.09468917548656464,
      "learning_rate": 0.0011124137931034484,
      "loss": 2.2418534755706787,
      "step": 8388
    },
    {
      "epoch": 0.839,
      "grad_norm": 0.07988791167736053,
      "learning_rate": 0.0011110344827586207,
      "loss": 2.2194035053253174,
      "step": 8390
    },
    {
      "epoch": 0.8392,
      "grad_norm": 0.08062989264726639,
      "learning_rate": 0.001109655172413793,
      "loss": 2.330963611602783,
      "step": 8392
    },
    {
      "epoch": 0.8394,
      "grad_norm": 0.07870481163263321,
      "learning_rate": 0.0011082758620689654,
      "loss": 2.2648873329162598,
      "step": 8394
    },
    {
      "epoch": 0.8396,
      "grad_norm": 0.07957063615322113,
      "learning_rate": 0.001106896551724138,
      "loss": 2.2303788661956787,
      "step": 8396
    },
    {
      "epoch": 0.8398,
      "grad_norm": 0.07708758860826492,
      "learning_rate": 0.0011055172413793105,
      "loss": 2.2236263751983643,
      "step": 8398
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.08035912364721298,
      "learning_rate": 0.0011041379310344828,
      "loss": 2.323809862136841,
      "step": 8400
    },
    {
      "epoch": 0.8402,
      "grad_norm": 0.07771014422178268,
      "learning_rate": 0.0011027586206896552,
      "loss": 2.276885509490967,
      "step": 8402
    },
    {
      "epoch": 0.8404,
      "grad_norm": 0.09305450320243835,
      "learning_rate": 0.0011013793103448275,
      "loss": 2.2150862216949463,
      "step": 8404
    },
    {
      "epoch": 0.8406,
      "grad_norm": 0.07814213633537292,
      "learning_rate": 0.0011,
      "loss": 2.295365571975708,
      "step": 8406
    },
    {
      "epoch": 0.8408,
      "grad_norm": 0.07784608006477356,
      "learning_rate": 0.0010986206896551724,
      "loss": 2.289787769317627,
      "step": 8408
    },
    {
      "epoch": 0.841,
      "grad_norm": 0.08526397496461868,
      "learning_rate": 0.0010972413793103447,
      "loss": 2.2160043716430664,
      "step": 8410
    },
    {
      "epoch": 0.8412,
      "grad_norm": 0.07639072835445404,
      "learning_rate": 0.0010958620689655173,
      "loss": 2.2254042625427246,
      "step": 8412
    },
    {
      "epoch": 0.8414,
      "grad_norm": 0.08222838491201401,
      "learning_rate": 0.0010944827586206896,
      "loss": 2.307478427886963,
      "step": 8414
    },
    {
      "epoch": 0.8416,
      "grad_norm": 0.07856354117393494,
      "learning_rate": 0.0010931034482758622,
      "loss": 2.227937698364258,
      "step": 8416
    },
    {
      "epoch": 0.8418,
      "grad_norm": 0.08053700625896454,
      "learning_rate": 0.0010917241379310345,
      "loss": 2.291863203048706,
      "step": 8418
    },
    {
      "epoch": 0.842,
      "grad_norm": 0.08108150213956833,
      "learning_rate": 0.0010903448275862068,
      "loss": 2.270310640335083,
      "step": 8420
    },
    {
      "epoch": 0.8422,
      "grad_norm": 0.07640858739614487,
      "learning_rate": 0.0010889655172413792,
      "loss": 2.3426918983459473,
      "step": 8422
    },
    {
      "epoch": 0.8424,
      "grad_norm": 0.07913515716791153,
      "learning_rate": 0.001087586206896552,
      "loss": 2.303283929824829,
      "step": 8424
    },
    {
      "epoch": 0.8426,
      "grad_norm": 0.08712572604417801,
      "learning_rate": 0.0010862068965517243,
      "loss": 2.1257755756378174,
      "step": 8426
    },
    {
      "epoch": 0.8428,
      "grad_norm": 0.08138784766197205,
      "learning_rate": 0.0010848275862068966,
      "loss": 2.3012871742248535,
      "step": 8428
    },
    {
      "epoch": 0.843,
      "grad_norm": 0.08183243125677109,
      "learning_rate": 0.001083448275862069,
      "loss": 2.3693134784698486,
      "step": 8430
    },
    {
      "epoch": 0.8432,
      "grad_norm": 0.07891813665628433,
      "learning_rate": 0.0010820689655172413,
      "loss": 2.2698147296905518,
      "step": 8432
    },
    {
      "epoch": 0.8434,
      "grad_norm": 0.08681058883666992,
      "learning_rate": 0.0010806896551724138,
      "loss": 2.284398078918457,
      "step": 8434
    },
    {
      "epoch": 0.8436,
      "grad_norm": 0.09006673842668533,
      "learning_rate": 0.0010793103448275862,
      "loss": 2.371614694595337,
      "step": 8436
    },
    {
      "epoch": 0.8438,
      "grad_norm": 0.08397416025400162,
      "learning_rate": 0.0010779310344827587,
      "loss": 2.265425205230713,
      "step": 8438
    },
    {
      "epoch": 0.844,
      "grad_norm": 0.07820859551429749,
      "learning_rate": 0.001076551724137931,
      "loss": 2.234297037124634,
      "step": 8440
    },
    {
      "epoch": 0.8442,
      "grad_norm": 0.07933638244867325,
      "learning_rate": 0.0010751724137931034,
      "loss": 2.2366058826446533,
      "step": 8442
    },
    {
      "epoch": 0.8444,
      "grad_norm": 0.08222569525241852,
      "learning_rate": 0.001073793103448276,
      "loss": 2.305586814880371,
      "step": 8444
    },
    {
      "epoch": 0.8446,
      "grad_norm": 0.079996258020401,
      "learning_rate": 0.0010724137931034483,
      "loss": 2.2077059745788574,
      "step": 8446
    },
    {
      "epoch": 0.8448,
      "grad_norm": 0.07903905212879181,
      "learning_rate": 0.0010710344827586206,
      "loss": 2.2052319049835205,
      "step": 8448
    },
    {
      "epoch": 0.845,
      "grad_norm": 0.07715138047933578,
      "learning_rate": 0.001069655172413793,
      "loss": 2.2682318687438965,
      "step": 8450
    },
    {
      "epoch": 0.8452,
      "grad_norm": 0.07855002582073212,
      "learning_rate": 0.0010682758620689655,
      "loss": 2.2072181701660156,
      "step": 8452
    },
    {
      "epoch": 0.8454,
      "grad_norm": 0.08151520788669586,
      "learning_rate": 0.001066896551724138,
      "loss": 2.1616244316101074,
      "step": 8454
    },
    {
      "epoch": 0.8456,
      "grad_norm": 0.09819554537534714,
      "learning_rate": 0.0010655172413793104,
      "loss": 2.367929458618164,
      "step": 8456
    },
    {
      "epoch": 0.8458,
      "grad_norm": 0.0819675624370575,
      "learning_rate": 0.0010641379310344827,
      "loss": 2.3177664279937744,
      "step": 8458
    },
    {
      "epoch": 0.846,
      "grad_norm": 0.08499930799007416,
      "learning_rate": 0.001062758620689655,
      "loss": 2.2780604362487793,
      "step": 8460
    },
    {
      "epoch": 0.8462,
      "grad_norm": 0.07803680002689362,
      "learning_rate": 0.0010613793103448276,
      "loss": 2.2398059368133545,
      "step": 8462
    },
    {
      "epoch": 0.8464,
      "grad_norm": 0.08061318099498749,
      "learning_rate": 0.0010600000000000002,
      "loss": 2.278542995452881,
      "step": 8464
    },
    {
      "epoch": 0.8466,
      "grad_norm": 0.08036350458860397,
      "learning_rate": 0.0010586206896551725,
      "loss": 2.1838178634643555,
      "step": 8466
    },
    {
      "epoch": 0.8468,
      "grad_norm": 0.08037454634904861,
      "learning_rate": 0.0010572413793103448,
      "loss": 2.179788589477539,
      "step": 8468
    },
    {
      "epoch": 0.847,
      "grad_norm": 0.08704026788473129,
      "learning_rate": 0.0010558620689655172,
      "loss": 2.2596988677978516,
      "step": 8470
    },
    {
      "epoch": 0.8472,
      "grad_norm": 0.08014703541994095,
      "learning_rate": 0.0010544827586206897,
      "loss": 2.1778085231781006,
      "step": 8472
    },
    {
      "epoch": 0.8474,
      "grad_norm": 0.07946956157684326,
      "learning_rate": 0.001053103448275862,
      "loss": 2.3498706817626953,
      "step": 8474
    },
    {
      "epoch": 0.8476,
      "grad_norm": 0.07788874208927155,
      "learning_rate": 0.0010517241379310344,
      "loss": 2.3071911334991455,
      "step": 8476
    },
    {
      "epoch": 0.8478,
      "grad_norm": 0.08569888025522232,
      "learning_rate": 0.001050344827586207,
      "loss": 2.335707902908325,
      "step": 8478
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.07911849766969681,
      "learning_rate": 0.0010489655172413793,
      "loss": 2.293299913406372,
      "step": 8480
    },
    {
      "epoch": 0.8482,
      "grad_norm": 0.08010365813970566,
      "learning_rate": 0.0010475862068965518,
      "loss": 2.347456693649292,
      "step": 8482
    },
    {
      "epoch": 0.8484,
      "grad_norm": 0.08557455986738205,
      "learning_rate": 0.0010462068965517242,
      "loss": 2.274707317352295,
      "step": 8484
    },
    {
      "epoch": 0.8486,
      "grad_norm": 0.08040586113929749,
      "learning_rate": 0.0010448275862068965,
      "loss": 2.2992701530456543,
      "step": 8486
    },
    {
      "epoch": 0.8488,
      "grad_norm": 0.07554929703474045,
      "learning_rate": 0.0010434482758620689,
      "loss": 2.2346394062042236,
      "step": 8488
    },
    {
      "epoch": 0.849,
      "grad_norm": 0.08111128211021423,
      "learning_rate": 0.0010420689655172414,
      "loss": 2.28090763092041,
      "step": 8490
    },
    {
      "epoch": 0.8492,
      "grad_norm": 0.08407483994960785,
      "learning_rate": 0.001040689655172414,
      "loss": 2.3013203144073486,
      "step": 8492
    },
    {
      "epoch": 0.8494,
      "grad_norm": 0.07686182111501694,
      "learning_rate": 0.0010393103448275863,
      "loss": 2.3645498752593994,
      "step": 8494
    },
    {
      "epoch": 0.8496,
      "grad_norm": 0.08433957397937775,
      "learning_rate": 0.0010379310344827586,
      "loss": 2.271799325942993,
      "step": 8496
    },
    {
      "epoch": 0.8498,
      "grad_norm": 0.07991495728492737,
      "learning_rate": 0.001036551724137931,
      "loss": 2.2723617553710938,
      "step": 8498
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.07721377164125443,
      "learning_rate": 0.0010351724137931035,
      "loss": 2.316878318786621,
      "step": 8500
    },
    {
      "epoch": 0.8502,
      "grad_norm": 0.07941431552171707,
      "learning_rate": 0.0010337931034482759,
      "loss": 2.26816725730896,
      "step": 8502
    },
    {
      "epoch": 0.8504,
      "grad_norm": 0.08324822038412094,
      "learning_rate": 0.0010324137931034482,
      "loss": 2.2445690631866455,
      "step": 8504
    },
    {
      "epoch": 0.8506,
      "grad_norm": 0.075588159263134,
      "learning_rate": 0.0010310344827586207,
      "loss": 2.3083415031433105,
      "step": 8506
    },
    {
      "epoch": 0.8508,
      "grad_norm": 0.08025657385587692,
      "learning_rate": 0.001029655172413793,
      "loss": 2.184675931930542,
      "step": 8508
    },
    {
      "epoch": 0.851,
      "grad_norm": 0.08005322515964508,
      "learning_rate": 0.0010282758620689656,
      "loss": 2.1728014945983887,
      "step": 8510
    },
    {
      "epoch": 0.8512,
      "grad_norm": 0.0832725465297699,
      "learning_rate": 0.001026896551724138,
      "loss": 2.249559164047241,
      "step": 8512
    },
    {
      "epoch": 0.8514,
      "grad_norm": 0.07927492260932922,
      "learning_rate": 0.0010255172413793103,
      "loss": 2.2746551036834717,
      "step": 8514
    },
    {
      "epoch": 0.8516,
      "grad_norm": 0.08412516862154007,
      "learning_rate": 0.0010241379310344826,
      "loss": 2.2230982780456543,
      "step": 8516
    },
    {
      "epoch": 0.8518,
      "grad_norm": 0.07827918231487274,
      "learning_rate": 0.0010227586206896552,
      "loss": 2.3406338691711426,
      "step": 8518
    },
    {
      "epoch": 0.852,
      "grad_norm": 0.08477301895618439,
      "learning_rate": 0.0010213793103448277,
      "loss": 2.2794954776763916,
      "step": 8520
    },
    {
      "epoch": 0.8522,
      "grad_norm": 0.07723060995340347,
      "learning_rate": 0.00102,
      "loss": 2.289487361907959,
      "step": 8522
    },
    {
      "epoch": 0.8524,
      "grad_norm": 0.07953017950057983,
      "learning_rate": 0.0010186206896551724,
      "loss": 2.2626686096191406,
      "step": 8524
    },
    {
      "epoch": 0.8526,
      "grad_norm": 0.0776280090212822,
      "learning_rate": 0.0010172413793103447,
      "loss": 2.1471543312072754,
      "step": 8526
    },
    {
      "epoch": 0.8528,
      "grad_norm": 0.07997231185436249,
      "learning_rate": 0.0010158620689655173,
      "loss": 2.2927310466766357,
      "step": 8528
    },
    {
      "epoch": 0.853,
      "grad_norm": 0.07992935180664062,
      "learning_rate": 0.0010144827586206896,
      "loss": 2.275836229324341,
      "step": 8530
    },
    {
      "epoch": 0.8532,
      "grad_norm": 0.08183673769235611,
      "learning_rate": 0.0010131034482758622,
      "loss": 2.252934217453003,
      "step": 8532
    },
    {
      "epoch": 0.8534,
      "grad_norm": 0.0839201882481575,
      "learning_rate": 0.0010117241379310345,
      "loss": 2.2627503871917725,
      "step": 8534
    },
    {
      "epoch": 0.8536,
      "grad_norm": 0.09208835661411285,
      "learning_rate": 0.0010103448275862069,
      "loss": 2.428154706954956,
      "step": 8536
    },
    {
      "epoch": 0.8538,
      "grad_norm": 0.08317186683416367,
      "learning_rate": 0.0010089655172413794,
      "loss": 2.2766835689544678,
      "step": 8538
    },
    {
      "epoch": 0.854,
      "grad_norm": 0.07625557482242584,
      "learning_rate": 0.0010075862068965517,
      "loss": 2.2494404315948486,
      "step": 8540
    },
    {
      "epoch": 0.8542,
      "grad_norm": 0.08511292934417725,
      "learning_rate": 0.001006206896551724,
      "loss": 2.298069477081299,
      "step": 8542
    },
    {
      "epoch": 0.8544,
      "grad_norm": 0.07897746562957764,
      "learning_rate": 0.0010048275862068964,
      "loss": 2.324054718017578,
      "step": 8544
    },
    {
      "epoch": 0.8546,
      "grad_norm": 0.07949665933847427,
      "learning_rate": 0.001003448275862069,
      "loss": 2.179680585861206,
      "step": 8546
    },
    {
      "epoch": 0.8548,
      "grad_norm": 0.08102643489837646,
      "learning_rate": 0.0010020689655172415,
      "loss": 2.2837345600128174,
      "step": 8548
    },
    {
      "epoch": 0.855,
      "grad_norm": 0.08311939984560013,
      "learning_rate": 0.0010006896551724139,
      "loss": 2.205777168273926,
      "step": 8550
    },
    {
      "epoch": 0.8552,
      "grad_norm": 0.07935646176338196,
      "learning_rate": 0.0009993103448275862,
      "loss": 2.2789578437805176,
      "step": 8552
    },
    {
      "epoch": 0.8554,
      "grad_norm": 0.07759179174900055,
      "learning_rate": 0.0009979310344827587,
      "loss": 2.2556095123291016,
      "step": 8554
    },
    {
      "epoch": 0.8556,
      "grad_norm": 0.0790112242102623,
      "learning_rate": 0.000996551724137931,
      "loss": 2.283576250076294,
      "step": 8556
    },
    {
      "epoch": 0.8558,
      "grad_norm": 0.08340583741664886,
      "learning_rate": 0.0009951724137931034,
      "loss": 2.2510921955108643,
      "step": 8558
    },
    {
      "epoch": 0.856,
      "grad_norm": 0.07772726565599442,
      "learning_rate": 0.000993793103448276,
      "loss": 2.2974469661712646,
      "step": 8560
    },
    {
      "epoch": 0.8562,
      "grad_norm": 0.07840023934841156,
      "learning_rate": 0.0009924137931034483,
      "loss": 2.320466995239258,
      "step": 8562
    },
    {
      "epoch": 0.8564,
      "grad_norm": 0.08804876357316971,
      "learning_rate": 0.0009910344827586206,
      "loss": 2.334584951400757,
      "step": 8564
    },
    {
      "epoch": 0.8566,
      "grad_norm": 0.0775604397058487,
      "learning_rate": 0.0009896551724137932,
      "loss": 2.3144004344940186,
      "step": 8566
    },
    {
      "epoch": 0.8568,
      "grad_norm": 0.07828965038061142,
      "learning_rate": 0.0009882758620689655,
      "loss": 2.3215625286102295,
      "step": 8568
    },
    {
      "epoch": 0.857,
      "grad_norm": 0.0813099592924118,
      "learning_rate": 0.0009868965517241379,
      "loss": 2.334340810775757,
      "step": 8570
    },
    {
      "epoch": 0.8572,
      "grad_norm": 0.08303730934858322,
      "learning_rate": 0.0009855172413793104,
      "loss": 2.267648935317993,
      "step": 8572
    },
    {
      "epoch": 0.8574,
      "grad_norm": 0.08594066649675369,
      "learning_rate": 0.0009841379310344827,
      "loss": 2.450516700744629,
      "step": 8574
    },
    {
      "epoch": 0.8576,
      "grad_norm": 0.08761117607355118,
      "learning_rate": 0.000982758620689655,
      "loss": 2.2722878456115723,
      "step": 8576
    },
    {
      "epoch": 0.8578,
      "grad_norm": 0.07826398313045502,
      "learning_rate": 0.0009813793103448276,
      "loss": 2.2779598236083984,
      "step": 8578
    },
    {
      "epoch": 0.858,
      "grad_norm": 0.07986059784889221,
      "learning_rate": 0.00098,
      "loss": 2.2544846534729004,
      "step": 8580
    },
    {
      "epoch": 0.8582,
      "grad_norm": 0.0801604688167572,
      "learning_rate": 0.0009786206896551725,
      "loss": 2.2512314319610596,
      "step": 8582
    },
    {
      "epoch": 0.8584,
      "grad_norm": 0.08207270503044128,
      "learning_rate": 0.0009772413793103449,
      "loss": 2.3137502670288086,
      "step": 8584
    },
    {
      "epoch": 0.8586,
      "grad_norm": 0.07999543845653534,
      "learning_rate": 0.0009758620689655172,
      "loss": 2.3038482666015625,
      "step": 8586
    },
    {
      "epoch": 0.8588,
      "grad_norm": 0.08005564659833908,
      "learning_rate": 0.0009744827586206897,
      "loss": 2.276292085647583,
      "step": 8588
    },
    {
      "epoch": 0.859,
      "grad_norm": 0.08111951500177383,
      "learning_rate": 0.0009731034482758621,
      "loss": 2.286299467086792,
      "step": 8590
    },
    {
      "epoch": 0.8592,
      "grad_norm": 0.080774687230587,
      "learning_rate": 0.0009717241379310345,
      "loss": 2.316005229949951,
      "step": 8592
    },
    {
      "epoch": 0.8594,
      "grad_norm": 0.07518048584461212,
      "learning_rate": 0.0009703448275862069,
      "loss": 2.2799794673919678,
      "step": 8594
    },
    {
      "epoch": 0.8596,
      "grad_norm": 0.08311876654624939,
      "learning_rate": 0.0009689655172413794,
      "loss": 2.195293426513672,
      "step": 8596
    },
    {
      "epoch": 0.8598,
      "grad_norm": 0.07987449318170547,
      "learning_rate": 0.0009675862068965517,
      "loss": 2.2811222076416016,
      "step": 8598
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.0801374688744545,
      "learning_rate": 0.0009662068965517241,
      "loss": 2.2671027183532715,
      "step": 8600
    },
    {
      "epoch": 0.8602,
      "grad_norm": 0.08442121744155884,
      "learning_rate": 0.0009648275862068966,
      "loss": 2.257389783859253,
      "step": 8602
    },
    {
      "epoch": 0.8604,
      "grad_norm": 0.08201821148395538,
      "learning_rate": 0.000963448275862069,
      "loss": 2.255882740020752,
      "step": 8604
    },
    {
      "epoch": 0.8606,
      "grad_norm": 0.08076457679271698,
      "learning_rate": 0.0009620689655172414,
      "loss": 2.2641968727111816,
      "step": 8606
    },
    {
      "epoch": 0.8608,
      "grad_norm": 0.07790926098823547,
      "learning_rate": 0.0009606896551724137,
      "loss": 2.2560150623321533,
      "step": 8608
    },
    {
      "epoch": 0.861,
      "grad_norm": 0.08727674186229706,
      "learning_rate": 0.0009593103448275863,
      "loss": 2.2928946018218994,
      "step": 8610
    },
    {
      "epoch": 0.8612,
      "grad_norm": 0.08241625875234604,
      "learning_rate": 0.0009579310344827586,
      "loss": 2.3033857345581055,
      "step": 8612
    },
    {
      "epoch": 0.8614,
      "grad_norm": 0.0773894190788269,
      "learning_rate": 0.000956551724137931,
      "loss": 2.2240395545959473,
      "step": 8614
    },
    {
      "epoch": 0.8616,
      "grad_norm": 0.07759292423725128,
      "learning_rate": 0.0009551724137931035,
      "loss": 2.229985475540161,
      "step": 8616
    },
    {
      "epoch": 0.8618,
      "grad_norm": 0.08198855072259903,
      "learning_rate": 0.0009537931034482759,
      "loss": 2.345729112625122,
      "step": 8618
    },
    {
      "epoch": 0.862,
      "grad_norm": 0.08628855645656586,
      "learning_rate": 0.0009524137931034483,
      "loss": 2.2819759845733643,
      "step": 8620
    },
    {
      "epoch": 0.8622,
      "grad_norm": 0.09010352939367294,
      "learning_rate": 0.0009510344827586207,
      "loss": 2.2653207778930664,
      "step": 8622
    },
    {
      "epoch": 0.8624,
      "grad_norm": 0.08809956908226013,
      "learning_rate": 0.0009496551724137932,
      "loss": 2.3447279930114746,
      "step": 8624
    },
    {
      "epoch": 0.8626,
      "grad_norm": 0.08207309246063232,
      "learning_rate": 0.0009482758620689655,
      "loss": 2.3068671226501465,
      "step": 8626
    },
    {
      "epoch": 0.8628,
      "grad_norm": 0.08302458375692368,
      "learning_rate": 0.0009468965517241379,
      "loss": 2.362335205078125,
      "step": 8628
    },
    {
      "epoch": 0.863,
      "grad_norm": 0.0845847874879837,
      "learning_rate": 0.0009455172413793104,
      "loss": 2.2732367515563965,
      "step": 8630
    },
    {
      "epoch": 0.8632,
      "grad_norm": 0.11834915727376938,
      "learning_rate": 0.0009441379310344827,
      "loss": 2.301239252090454,
      "step": 8632
    },
    {
      "epoch": 0.8634,
      "grad_norm": 0.07841874659061432,
      "learning_rate": 0.0009427586206896552,
      "loss": 2.3169946670532227,
      "step": 8634
    },
    {
      "epoch": 0.8636,
      "grad_norm": 0.08119630813598633,
      "learning_rate": 0.0009413793103448276,
      "loss": 2.283738136291504,
      "step": 8636
    },
    {
      "epoch": 0.8638,
      "grad_norm": 0.07856898754835129,
      "learning_rate": 0.00094,
      "loss": 2.3117473125457764,
      "step": 8638
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.08046866953372955,
      "learning_rate": 0.0009386206896551724,
      "loss": 2.2985072135925293,
      "step": 8640
    },
    {
      "epoch": 0.8642,
      "grad_norm": 0.07943491637706757,
      "learning_rate": 0.0009372413793103449,
      "loss": 2.301666021347046,
      "step": 8642
    },
    {
      "epoch": 0.8644,
      "grad_norm": 0.08408946543931961,
      "learning_rate": 0.0009358620689655173,
      "loss": 2.2672770023345947,
      "step": 8644
    },
    {
      "epoch": 0.8646,
      "grad_norm": 0.07945779711008072,
      "learning_rate": 0.0009344827586206896,
      "loss": 2.281032085418701,
      "step": 8646
    },
    {
      "epoch": 0.8648,
      "grad_norm": 0.07856013625860214,
      "learning_rate": 0.0009331034482758621,
      "loss": 2.263155221939087,
      "step": 8648
    },
    {
      "epoch": 0.865,
      "grad_norm": 0.07946278899908066,
      "learning_rate": 0.0009317241379310345,
      "loss": 2.2876360416412354,
      "step": 8650
    },
    {
      "epoch": 0.8652,
      "grad_norm": 0.08000805974006653,
      "learning_rate": 0.0009303448275862069,
      "loss": 2.2339723110198975,
      "step": 8652
    },
    {
      "epoch": 0.8654,
      "grad_norm": 0.07669277489185333,
      "learning_rate": 0.0009289655172413793,
      "loss": 2.260051965713501,
      "step": 8654
    },
    {
      "epoch": 0.8656,
      "grad_norm": 0.07928746938705444,
      "learning_rate": 0.0009275862068965517,
      "loss": 2.283933162689209,
      "step": 8656
    },
    {
      "epoch": 0.8658,
      "grad_norm": 0.08113522827625275,
      "learning_rate": 0.0009262068965517242,
      "loss": 2.2504947185516357,
      "step": 8658
    },
    {
      "epoch": 0.866,
      "grad_norm": 0.07996169477701187,
      "learning_rate": 0.0009248275862068965,
      "loss": 2.270969867706299,
      "step": 8660
    },
    {
      "epoch": 0.8662,
      "grad_norm": 0.07757966220378876,
      "learning_rate": 0.0009234482758620691,
      "loss": 2.3390772342681885,
      "step": 8662
    },
    {
      "epoch": 0.8664,
      "grad_norm": 0.07998756319284439,
      "learning_rate": 0.0009220689655172414,
      "loss": 2.3327107429504395,
      "step": 8664
    },
    {
      "epoch": 0.8666,
      "grad_norm": 0.0810607448220253,
      "learning_rate": 0.0009206896551724138,
      "loss": 2.1815431118011475,
      "step": 8666
    },
    {
      "epoch": 0.8668,
      "grad_norm": 0.07765872776508331,
      "learning_rate": 0.0009193103448275862,
      "loss": 2.3846206665039062,
      "step": 8668
    },
    {
      "epoch": 0.867,
      "grad_norm": 0.08455213159322739,
      "learning_rate": 0.0009179310344827586,
      "loss": 2.2688701152801514,
      "step": 8670
    },
    {
      "epoch": 0.8672,
      "grad_norm": 0.0823834091424942,
      "learning_rate": 0.0009165517241379311,
      "loss": 2.2223544120788574,
      "step": 8672
    },
    {
      "epoch": 0.8674,
      "grad_norm": 0.07827700674533844,
      "learning_rate": 0.0009151724137931034,
      "loss": 2.2175161838531494,
      "step": 8674
    },
    {
      "epoch": 0.8676,
      "grad_norm": 0.07852313667535782,
      "learning_rate": 0.000913793103448276,
      "loss": 2.2208902835845947,
      "step": 8676
    },
    {
      "epoch": 0.8678,
      "grad_norm": 0.07911550998687744,
      "learning_rate": 0.0009124137931034483,
      "loss": 2.284764051437378,
      "step": 8678
    },
    {
      "epoch": 0.868,
      "grad_norm": 0.07841844111680984,
      "learning_rate": 0.0009110344827586206,
      "loss": 2.247361898422241,
      "step": 8680
    },
    {
      "epoch": 0.8682,
      "grad_norm": 0.07501046359539032,
      "learning_rate": 0.0009096551724137932,
      "loss": 2.2834889888763428,
      "step": 8682
    },
    {
      "epoch": 0.8684,
      "grad_norm": 0.07996895164251328,
      "learning_rate": 0.0009082758620689655,
      "loss": 2.3749282360076904,
      "step": 8684
    },
    {
      "epoch": 0.8686,
      "grad_norm": 0.07964544743299484,
      "learning_rate": 0.000906896551724138,
      "loss": 2.321061849594116,
      "step": 8686
    },
    {
      "epoch": 0.8688,
      "grad_norm": 0.07832322269678116,
      "learning_rate": 0.0009055172413793103,
      "loss": 2.2718708515167236,
      "step": 8688
    },
    {
      "epoch": 0.869,
      "grad_norm": 0.07862125337123871,
      "learning_rate": 0.0009041379310344829,
      "loss": 2.3481085300445557,
      "step": 8690
    },
    {
      "epoch": 0.8692,
      "grad_norm": 0.08325941115617752,
      "learning_rate": 0.0009027586206896552,
      "loss": 2.2867960929870605,
      "step": 8692
    },
    {
      "epoch": 0.8694,
      "grad_norm": 0.08251561224460602,
      "learning_rate": 0.0009013793103448275,
      "loss": 2.2971150875091553,
      "step": 8694
    },
    {
      "epoch": 0.8696,
      "grad_norm": 0.080591581761837,
      "learning_rate": 0.0009000000000000001,
      "loss": 2.272239923477173,
      "step": 8696
    },
    {
      "epoch": 0.8698,
      "grad_norm": 0.0785486102104187,
      "learning_rate": 0.0008986206896551724,
      "loss": 2.33854079246521,
      "step": 8698
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.078777976334095,
      "learning_rate": 0.0008972413793103449,
      "loss": 2.239391565322876,
      "step": 8700
    },
    {
      "epoch": 0.8702,
      "grad_norm": 0.07969389855861664,
      "learning_rate": 0.0008958620689655172,
      "loss": 2.29776668548584,
      "step": 8702
    },
    {
      "epoch": 0.8704,
      "grad_norm": 0.08138518035411835,
      "learning_rate": 0.0008944827586206896,
      "loss": 2.3531436920166016,
      "step": 8704
    },
    {
      "epoch": 0.8706,
      "grad_norm": 0.07734669744968414,
      "learning_rate": 0.0008931034482758621,
      "loss": 2.299590826034546,
      "step": 8706
    },
    {
      "epoch": 0.8708,
      "grad_norm": 0.0801185816526413,
      "learning_rate": 0.0008917241379310344,
      "loss": 2.343576669692993,
      "step": 8708
    },
    {
      "epoch": 0.871,
      "grad_norm": 0.07593826949596405,
      "learning_rate": 0.000890344827586207,
      "loss": 2.3183062076568604,
      "step": 8710
    },
    {
      "epoch": 0.8712,
      "grad_norm": 0.07844239473342896,
      "learning_rate": 0.0008889655172413793,
      "loss": 2.302121162414551,
      "step": 8712
    },
    {
      "epoch": 0.8714,
      "grad_norm": 0.08352925628423691,
      "learning_rate": 0.0008875862068965518,
      "loss": 2.328165292739868,
      "step": 8714
    },
    {
      "epoch": 0.8716,
      "grad_norm": 0.07798857241868973,
      "learning_rate": 0.0008862068965517242,
      "loss": 2.294914722442627,
      "step": 8716
    },
    {
      "epoch": 0.8718,
      "grad_norm": 0.0804460346698761,
      "learning_rate": 0.0008848275862068965,
      "loss": 2.323951244354248,
      "step": 8718
    },
    {
      "epoch": 0.872,
      "grad_norm": 0.07743699103593826,
      "learning_rate": 0.000883448275862069,
      "loss": 2.372144937515259,
      "step": 8720
    },
    {
      "epoch": 0.8722,
      "grad_norm": 0.07801779359579086,
      "learning_rate": 0.0008820689655172413,
      "loss": 2.186762571334839,
      "step": 8722
    },
    {
      "epoch": 0.8724,
      "grad_norm": 0.0829208642244339,
      "learning_rate": 0.0008806896551724139,
      "loss": 2.241305351257324,
      "step": 8724
    },
    {
      "epoch": 0.8726,
      "grad_norm": 0.0782838985323906,
      "learning_rate": 0.0008793103448275862,
      "loss": 2.2925667762756348,
      "step": 8726
    },
    {
      "epoch": 0.8728,
      "grad_norm": 0.07731062918901443,
      "learning_rate": 0.0008779310344827586,
      "loss": 2.3399479389190674,
      "step": 8728
    },
    {
      "epoch": 0.873,
      "grad_norm": 0.08337566256523132,
      "learning_rate": 0.0008765517241379311,
      "loss": 2.2835285663604736,
      "step": 8730
    },
    {
      "epoch": 0.8732,
      "grad_norm": 0.08527350425720215,
      "learning_rate": 0.0008751724137931034,
      "loss": 2.2989156246185303,
      "step": 8732
    },
    {
      "epoch": 0.8734,
      "grad_norm": 0.08176404237747192,
      "learning_rate": 0.0008737931034482759,
      "loss": 2.22485613822937,
      "step": 8734
    },
    {
      "epoch": 0.8736,
      "grad_norm": 0.07761428505182266,
      "learning_rate": 0.0008724137931034483,
      "loss": 2.3240463733673096,
      "step": 8736
    },
    {
      "epoch": 0.8738,
      "grad_norm": 0.07939528673887253,
      "learning_rate": 0.0008710344827586208,
      "loss": 2.308175802230835,
      "step": 8738
    },
    {
      "epoch": 0.874,
      "grad_norm": 0.08088484406471252,
      "learning_rate": 0.0008696551724137931,
      "loss": 2.287008762359619,
      "step": 8740
    },
    {
      "epoch": 0.8742,
      "grad_norm": 0.08017789572477341,
      "learning_rate": 0.0008682758620689655,
      "loss": 2.2566885948181152,
      "step": 8742
    },
    {
      "epoch": 0.8744,
      "grad_norm": 0.07965514808893204,
      "learning_rate": 0.000866896551724138,
      "loss": 2.3042211532592773,
      "step": 8744
    },
    {
      "epoch": 0.8746,
      "grad_norm": 0.08019448816776276,
      "learning_rate": 0.0008655172413793103,
      "loss": 2.26267671585083,
      "step": 8746
    },
    {
      "epoch": 0.8748,
      "grad_norm": 0.0809718668460846,
      "learning_rate": 0.0008641379310344828,
      "loss": 2.347611427307129,
      "step": 8748
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.08060459792613983,
      "learning_rate": 0.0008627586206896552,
      "loss": 2.332327127456665,
      "step": 8750
    },
    {
      "epoch": 0.8752,
      "grad_norm": 0.07809365540742874,
      "learning_rate": 0.0008613793103448276,
      "loss": 2.2253851890563965,
      "step": 8752
    },
    {
      "epoch": 0.8754,
      "grad_norm": 0.08266472071409225,
      "learning_rate": 0.00086,
      "loss": 2.2801599502563477,
      "step": 8754
    },
    {
      "epoch": 0.8756,
      "grad_norm": 0.08021393418312073,
      "learning_rate": 0.0008586206896551725,
      "loss": 2.2699289321899414,
      "step": 8756
    },
    {
      "epoch": 0.8758,
      "grad_norm": 0.08290563523769379,
      "learning_rate": 0.0008572413793103449,
      "loss": 2.356593608856201,
      "step": 8758
    },
    {
      "epoch": 0.876,
      "grad_norm": 0.08407673984766006,
      "learning_rate": 0.0008558620689655172,
      "loss": 2.276451587677002,
      "step": 8760
    },
    {
      "epoch": 0.8762,
      "grad_norm": 0.07763871550559998,
      "learning_rate": 0.0008544827586206896,
      "loss": 2.2361178398132324,
      "step": 8762
    },
    {
      "epoch": 0.8764,
      "grad_norm": 0.08346902579069138,
      "learning_rate": 0.0008531034482758621,
      "loss": 2.236997604370117,
      "step": 8764
    },
    {
      "epoch": 0.8766,
      "grad_norm": 0.08042485266923904,
      "learning_rate": 0.0008517241379310345,
      "loss": 2.3373947143554688,
      "step": 8766
    },
    {
      "epoch": 0.8768,
      "grad_norm": 0.08331667631864548,
      "learning_rate": 0.0008503448275862069,
      "loss": 2.2938339710235596,
      "step": 8768
    },
    {
      "epoch": 0.877,
      "grad_norm": 0.07893479615449905,
      "learning_rate": 0.0008489655172413794,
      "loss": 2.3386049270629883,
      "step": 8770
    },
    {
      "epoch": 0.8772,
      "grad_norm": 0.07886341214179993,
      "learning_rate": 0.0008475862068965518,
      "loss": 2.428621292114258,
      "step": 8772
    },
    {
      "epoch": 0.8774,
      "grad_norm": 0.08211275190114975,
      "learning_rate": 0.0008462068965517241,
      "loss": 2.328303337097168,
      "step": 8774
    },
    {
      "epoch": 0.8776,
      "grad_norm": 0.08115790039300919,
      "learning_rate": 0.0008448275862068966,
      "loss": 2.3940672874450684,
      "step": 8776
    },
    {
      "epoch": 0.8778,
      "grad_norm": 0.0827493816614151,
      "learning_rate": 0.000843448275862069,
      "loss": 2.363318681716919,
      "step": 8778
    },
    {
      "epoch": 0.878,
      "grad_norm": 0.08345569670200348,
      "learning_rate": 0.0008420689655172414,
      "loss": 2.2900261878967285,
      "step": 8780
    },
    {
      "epoch": 0.8782,
      "grad_norm": 0.08085129410028458,
      "learning_rate": 0.0008406896551724138,
      "loss": 2.3910210132598877,
      "step": 8782
    },
    {
      "epoch": 0.8784,
      "grad_norm": 0.08174458891153336,
      "learning_rate": 0.0008393103448275862,
      "loss": 2.388145685195923,
      "step": 8784
    },
    {
      "epoch": 0.8786,
      "grad_norm": 0.0823163390159607,
      "learning_rate": 0.0008379310344827586,
      "loss": 2.3041484355926514,
      "step": 8786
    },
    {
      "epoch": 0.8788,
      "grad_norm": 0.08093644678592682,
      "learning_rate": 0.000836551724137931,
      "loss": 2.309612989425659,
      "step": 8788
    },
    {
      "epoch": 0.879,
      "grad_norm": 0.08140049874782562,
      "learning_rate": 0.0008351724137931035,
      "loss": 2.292426347732544,
      "step": 8790
    },
    {
      "epoch": 0.8792,
      "grad_norm": 0.08166998624801636,
      "learning_rate": 0.0008337931034482759,
      "loss": 2.3470187187194824,
      "step": 8792
    },
    {
      "epoch": 0.8794,
      "grad_norm": 0.08710470050573349,
      "learning_rate": 0.0008324137931034483,
      "loss": 2.320646047592163,
      "step": 8794
    },
    {
      "epoch": 0.8796,
      "grad_norm": 0.08035363256931305,
      "learning_rate": 0.0008310344827586206,
      "loss": 2.2880215644836426,
      "step": 8796
    },
    {
      "epoch": 0.8798,
      "grad_norm": 0.0995086282491684,
      "learning_rate": 0.0008296551724137931,
      "loss": 2.2799265384674072,
      "step": 8798
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.0767660140991211,
      "learning_rate": 0.0008282758620689655,
      "loss": 2.298517942428589,
      "step": 8800
    },
    {
      "epoch": 0.8802,
      "grad_norm": 0.07821908593177795,
      "learning_rate": 0.0008268965517241379,
      "loss": 2.3066296577453613,
      "step": 8802
    },
    {
      "epoch": 0.8804,
      "grad_norm": 0.08115953207015991,
      "learning_rate": 0.0008255172413793104,
      "loss": 2.269801139831543,
      "step": 8804
    },
    {
      "epoch": 0.8806,
      "grad_norm": 0.07764173299074173,
      "learning_rate": 0.0008241379310344828,
      "loss": 2.2893152236938477,
      "step": 8806
    },
    {
      "epoch": 0.8808,
      "grad_norm": 0.08190420269966125,
      "learning_rate": 0.0008227586206896552,
      "loss": 2.349179744720459,
      "step": 8808
    },
    {
      "epoch": 0.881,
      "grad_norm": 0.08266187459230423,
      "learning_rate": 0.0008213793103448276,
      "loss": 2.284303903579712,
      "step": 8810
    },
    {
      "epoch": 0.8812,
      "grad_norm": 0.07960504293441772,
      "learning_rate": 0.00082,
      "loss": 2.261338710784912,
      "step": 8812
    },
    {
      "epoch": 0.8814,
      "grad_norm": 0.07685971260070801,
      "learning_rate": 0.0008186206896551724,
      "loss": 2.2498955726623535,
      "step": 8814
    },
    {
      "epoch": 0.8816,
      "grad_norm": 0.08038755506277084,
      "learning_rate": 0.0008172413793103448,
      "loss": 2.2475669384002686,
      "step": 8816
    },
    {
      "epoch": 0.8818,
      "grad_norm": 0.0782775953412056,
      "learning_rate": 0.0008158620689655173,
      "loss": 2.2646842002868652,
      "step": 8818
    },
    {
      "epoch": 0.882,
      "grad_norm": 0.08139389753341675,
      "learning_rate": 0.0008144827586206896,
      "loss": 2.3637888431549072,
      "step": 8820
    },
    {
      "epoch": 0.8822,
      "grad_norm": 0.07714074105024338,
      "learning_rate": 0.0008131034482758621,
      "loss": 2.242300271987915,
      "step": 8822
    },
    {
      "epoch": 0.8824,
      "grad_norm": 0.080418162047863,
      "learning_rate": 0.0008117241379310345,
      "loss": 2.2892367839813232,
      "step": 8824
    },
    {
      "epoch": 0.8826,
      "grad_norm": 0.0776117593050003,
      "learning_rate": 0.0008103448275862069,
      "loss": 2.2207908630371094,
      "step": 8826
    },
    {
      "epoch": 0.8828,
      "grad_norm": 0.08139217644929886,
      "learning_rate": 0.0008089655172413793,
      "loss": 2.344057321548462,
      "step": 8828
    },
    {
      "epoch": 0.883,
      "grad_norm": 0.09064886718988419,
      "learning_rate": 0.0008075862068965518,
      "loss": 2.3596725463867188,
      "step": 8830
    },
    {
      "epoch": 0.8832,
      "grad_norm": 0.07830341160297394,
      "learning_rate": 0.0008062068965517242,
      "loss": 2.294125556945801,
      "step": 8832
    },
    {
      "epoch": 0.8834,
      "grad_norm": 0.07961355894804001,
      "learning_rate": 0.0008048275862068965,
      "loss": 2.3126392364501953,
      "step": 8834
    },
    {
      "epoch": 0.8836,
      "grad_norm": 0.0825362354516983,
      "learning_rate": 0.000803448275862069,
      "loss": 2.3614115715026855,
      "step": 8836
    },
    {
      "epoch": 0.8838,
      "grad_norm": 0.08578339964151382,
      "learning_rate": 0.0008020689655172414,
      "loss": 2.326136589050293,
      "step": 8838
    },
    {
      "epoch": 0.884,
      "grad_norm": 0.08104421198368073,
      "learning_rate": 0.0008006896551724138,
      "loss": 2.3402721881866455,
      "step": 8840
    },
    {
      "epoch": 0.8842,
      "grad_norm": 0.07844129949808121,
      "learning_rate": 0.0007993103448275862,
      "loss": 2.2741825580596924,
      "step": 8842
    },
    {
      "epoch": 0.8844,
      "grad_norm": 0.07941241562366486,
      "learning_rate": 0.0007979310344827586,
      "loss": 2.3379435539245605,
      "step": 8844
    },
    {
      "epoch": 0.8846,
      "grad_norm": 0.0820191353559494,
      "learning_rate": 0.0007965517241379311,
      "loss": 2.243190288543701,
      "step": 8846
    },
    {
      "epoch": 0.8848,
      "grad_norm": 0.08025038242340088,
      "learning_rate": 0.0007951724137931034,
      "loss": 2.2528483867645264,
      "step": 8848
    },
    {
      "epoch": 0.885,
      "grad_norm": 0.08074972778558731,
      "learning_rate": 0.000793793103448276,
      "loss": 2.2906124591827393,
      "step": 8850
    },
    {
      "epoch": 0.8852,
      "grad_norm": 0.08053814619779587,
      "learning_rate": 0.0007924137931034483,
      "loss": 2.320384979248047,
      "step": 8852
    },
    {
      "epoch": 0.8854,
      "grad_norm": 0.07449758797883987,
      "learning_rate": 0.0007910344827586207,
      "loss": 2.2920877933502197,
      "step": 8854
    },
    {
      "epoch": 0.8856,
      "grad_norm": 0.0772114023566246,
      "learning_rate": 0.0007896551724137931,
      "loss": 2.390002965927124,
      "step": 8856
    },
    {
      "epoch": 0.8858,
      "grad_norm": 0.07874036580324173,
      "learning_rate": 0.0007882758620689655,
      "loss": 2.2738747596740723,
      "step": 8858
    },
    {
      "epoch": 0.886,
      "grad_norm": 0.0810052901506424,
      "learning_rate": 0.000786896551724138,
      "loss": 2.2974441051483154,
      "step": 8860
    },
    {
      "epoch": 0.8862,
      "grad_norm": 0.0806816965341568,
      "learning_rate": 0.0007855172413793103,
      "loss": 2.258357048034668,
      "step": 8862
    },
    {
      "epoch": 0.8864,
      "grad_norm": 0.0788218155503273,
      "learning_rate": 0.0007841379310344828,
      "loss": 2.2856249809265137,
      "step": 8864
    },
    {
      "epoch": 0.8866,
      "grad_norm": 0.08351460099220276,
      "learning_rate": 0.0007827586206896552,
      "loss": 2.383493185043335,
      "step": 8866
    },
    {
      "epoch": 0.8868,
      "grad_norm": 0.07769401371479034,
      "learning_rate": 0.0007813793103448275,
      "loss": 2.248908519744873,
      "step": 8868
    },
    {
      "epoch": 0.887,
      "grad_norm": 0.08189140260219574,
      "learning_rate": 0.0007800000000000001,
      "loss": 2.356705904006958,
      "step": 8870
    },
    {
      "epoch": 0.8872,
      "grad_norm": 0.08326449245214462,
      "learning_rate": 0.0007786206896551724,
      "loss": 2.3165509700775146,
      "step": 8872
    },
    {
      "epoch": 0.8874,
      "grad_norm": 0.0828094482421875,
      "learning_rate": 0.0007772413793103449,
      "loss": 2.3919436931610107,
      "step": 8874
    },
    {
      "epoch": 0.8876,
      "grad_norm": 0.08130276948213577,
      "learning_rate": 0.0007758620689655172,
      "loss": 2.3130457401275635,
      "step": 8876
    },
    {
      "epoch": 0.8878,
      "grad_norm": 0.08160854876041412,
      "learning_rate": 0.0007744827586206897,
      "loss": 2.3099241256713867,
      "step": 8878
    },
    {
      "epoch": 0.888,
      "grad_norm": 0.07970861345529556,
      "learning_rate": 0.0007731034482758621,
      "loss": 2.3449008464813232,
      "step": 8880
    },
    {
      "epoch": 0.8882,
      "grad_norm": 0.0925278440117836,
      "learning_rate": 0.0007717241379310344,
      "loss": 2.238924264907837,
      "step": 8882
    },
    {
      "epoch": 0.8884,
      "grad_norm": 0.07858434319496155,
      "learning_rate": 0.000770344827586207,
      "loss": 2.2962331771850586,
      "step": 8884
    },
    {
      "epoch": 0.8886,
      "grad_norm": 0.08134745061397552,
      "learning_rate": 0.0007689655172413793,
      "loss": 2.300368070602417,
      "step": 8886
    },
    {
      "epoch": 0.8888,
      "grad_norm": 0.0899118185043335,
      "learning_rate": 0.0007675862068965518,
      "loss": 2.24780535697937,
      "step": 8888
    },
    {
      "epoch": 0.889,
      "grad_norm": 0.0792151466012001,
      "learning_rate": 0.0007662068965517241,
      "loss": 2.3725481033325195,
      "step": 8890
    },
    {
      "epoch": 0.8892,
      "grad_norm": 0.08708511292934418,
      "learning_rate": 0.0007648275862068965,
      "loss": 2.2852632999420166,
      "step": 8892
    },
    {
      "epoch": 0.8894,
      "grad_norm": 0.08259522914886475,
      "learning_rate": 0.000763448275862069,
      "loss": 2.233151912689209,
      "step": 8894
    },
    {
      "epoch": 0.8896,
      "grad_norm": 0.0800049751996994,
      "learning_rate": 0.0007620689655172413,
      "loss": 2.3139615058898926,
      "step": 8896
    },
    {
      "epoch": 0.8898,
      "grad_norm": 0.08159743994474411,
      "learning_rate": 0.0007606896551724139,
      "loss": 2.3025410175323486,
      "step": 8898
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.08073541522026062,
      "learning_rate": 0.0007593103448275862,
      "loss": 2.2551398277282715,
      "step": 8900
    },
    {
      "epoch": 0.8902,
      "grad_norm": 0.07756489515304565,
      "learning_rate": 0.0007579310344827587,
      "loss": 2.316403388977051,
      "step": 8902
    },
    {
      "epoch": 0.8904,
      "grad_norm": 0.08459212630987167,
      "learning_rate": 0.0007565517241379311,
      "loss": 2.237489700317383,
      "step": 8904
    },
    {
      "epoch": 0.8906,
      "grad_norm": 0.0781688243150711,
      "learning_rate": 0.0007551724137931034,
      "loss": 2.2909927368164062,
      "step": 8906
    },
    {
      "epoch": 0.8908,
      "grad_norm": 0.0778971016407013,
      "learning_rate": 0.0007537931034482759,
      "loss": 2.2962117195129395,
      "step": 8908
    },
    {
      "epoch": 0.891,
      "grad_norm": 0.08095421642065048,
      "learning_rate": 0.0007524137931034482,
      "loss": 2.3364973068237305,
      "step": 8910
    },
    {
      "epoch": 0.8912,
      "grad_norm": 0.07966719567775726,
      "learning_rate": 0.0007510344827586208,
      "loss": 2.348832130432129,
      "step": 8912
    },
    {
      "epoch": 0.8914,
      "grad_norm": 0.08108032494783401,
      "learning_rate": 0.0007496551724137931,
      "loss": 2.221893072128296,
      "step": 8914
    },
    {
      "epoch": 0.8916,
      "grad_norm": 0.08291000127792358,
      "learning_rate": 0.0007482758620689655,
      "loss": 2.275341749191284,
      "step": 8916
    },
    {
      "epoch": 0.8918,
      "grad_norm": 0.07943326234817505,
      "learning_rate": 0.000746896551724138,
      "loss": 2.269195318222046,
      "step": 8918
    },
    {
      "epoch": 0.892,
      "grad_norm": 0.07809513062238693,
      "learning_rate": 0.0007455172413793103,
      "loss": 2.2643837928771973,
      "step": 8920
    },
    {
      "epoch": 0.8922,
      "grad_norm": 0.07485774159431458,
      "learning_rate": 0.0007441379310344828,
      "loss": 2.200526714324951,
      "step": 8922
    },
    {
      "epoch": 0.8924,
      "grad_norm": 0.08419068902730942,
      "learning_rate": 0.0007427586206896552,
      "loss": 2.4302968978881836,
      "step": 8924
    },
    {
      "epoch": 0.8926,
      "grad_norm": 0.08195514231920242,
      "learning_rate": 0.0007413793103448277,
      "loss": 2.455188035964966,
      "step": 8926
    },
    {
      "epoch": 0.8928,
      "grad_norm": 0.08044300973415375,
      "learning_rate": 0.00074,
      "loss": 2.253384828567505,
      "step": 8928
    },
    {
      "epoch": 0.893,
      "grad_norm": 0.08068639785051346,
      "learning_rate": 0.0007386206896551724,
      "loss": 2.31198787689209,
      "step": 8930
    },
    {
      "epoch": 0.8932,
      "grad_norm": 0.07958860695362091,
      "learning_rate": 0.0007372413793103449,
      "loss": 2.301419973373413,
      "step": 8932
    },
    {
      "epoch": 0.8934,
      "grad_norm": 0.08093810826539993,
      "learning_rate": 0.0007358620689655172,
      "loss": 2.3577301502227783,
      "step": 8934
    },
    {
      "epoch": 0.8936,
      "grad_norm": 0.08171824365854263,
      "learning_rate": 0.0007344827586206897,
      "loss": 2.4090120792388916,
      "step": 8936
    },
    {
      "epoch": 0.8938,
      "grad_norm": 0.07946425676345825,
      "learning_rate": 0.0007331034482758621,
      "loss": 2.222733974456787,
      "step": 8938
    },
    {
      "epoch": 0.894,
      "grad_norm": 0.08603188395500183,
      "learning_rate": 0.0007317241379310345,
      "loss": 2.35015869140625,
      "step": 8940
    },
    {
      "epoch": 0.8942,
      "grad_norm": 0.08952043205499649,
      "learning_rate": 0.0007303448275862069,
      "loss": 2.306037187576294,
      "step": 8942
    },
    {
      "epoch": 0.8944,
      "grad_norm": 0.0833539366722107,
      "learning_rate": 0.0007289655172413793,
      "loss": 2.3458659648895264,
      "step": 8944
    },
    {
      "epoch": 0.8946,
      "grad_norm": 0.08163369446992874,
      "learning_rate": 0.0007275862068965518,
      "loss": 2.264064073562622,
      "step": 8946
    },
    {
      "epoch": 0.8948,
      "grad_norm": 0.08209577947854996,
      "learning_rate": 0.0007262068965517241,
      "loss": 2.23891019821167,
      "step": 8948
    },
    {
      "epoch": 0.895,
      "grad_norm": 0.08143900334835052,
      "learning_rate": 0.0007248275862068965,
      "loss": 2.3186774253845215,
      "step": 8950
    },
    {
      "epoch": 0.8952,
      "grad_norm": 0.07763932645320892,
      "learning_rate": 0.000723448275862069,
      "loss": 2.3788554668426514,
      "step": 8952
    },
    {
      "epoch": 0.8954,
      "grad_norm": 0.08084189891815186,
      "learning_rate": 0.0007220689655172414,
      "loss": 2.2625205516815186,
      "step": 8954
    },
    {
      "epoch": 0.8956,
      "grad_norm": 0.0797935426235199,
      "learning_rate": 0.0007206896551724138,
      "loss": 2.2948412895202637,
      "step": 8956
    },
    {
      "epoch": 0.8958,
      "grad_norm": 0.08172356337308884,
      "learning_rate": 0.0007193103448275862,
      "loss": 2.229840040206909,
      "step": 8958
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.07949266582727432,
      "learning_rate": 0.0007179310344827587,
      "loss": 2.3092100620269775,
      "step": 8960
    },
    {
      "epoch": 0.8962,
      "grad_norm": 0.0845036432147026,
      "learning_rate": 0.000716551724137931,
      "loss": 2.3561315536499023,
      "step": 8962
    },
    {
      "epoch": 0.8964,
      "grad_norm": 0.07649905979633331,
      "learning_rate": 0.0007151724137931035,
      "loss": 2.3151662349700928,
      "step": 8964
    },
    {
      "epoch": 0.8966,
      "grad_norm": 0.0826631486415863,
      "learning_rate": 0.0007137931034482759,
      "loss": 2.351572036743164,
      "step": 8966
    },
    {
      "epoch": 0.8968,
      "grad_norm": 0.08226320147514343,
      "learning_rate": 0.0007124137931034483,
      "loss": 2.3213634490966797,
      "step": 8968
    },
    {
      "epoch": 0.897,
      "grad_norm": 0.07987559586763382,
      "learning_rate": 0.0007110344827586207,
      "loss": 2.2649989128112793,
      "step": 8970
    },
    {
      "epoch": 0.8972,
      "grad_norm": 0.0811890959739685,
      "learning_rate": 0.0007096551724137931,
      "loss": 2.3694233894348145,
      "step": 8972
    },
    {
      "epoch": 0.8974,
      "grad_norm": 0.08170734345912933,
      "learning_rate": 0.0007082758620689655,
      "loss": 2.24364972114563,
      "step": 8974
    },
    {
      "epoch": 0.8976,
      "grad_norm": 0.08104456961154938,
      "learning_rate": 0.0007068965517241379,
      "loss": 2.295954465866089,
      "step": 8976
    },
    {
      "epoch": 0.8978,
      "grad_norm": 0.08008222281932831,
      "learning_rate": 0.0007055172413793104,
      "loss": 2.2986152172088623,
      "step": 8978
    },
    {
      "epoch": 0.898,
      "grad_norm": 0.08258602023124695,
      "learning_rate": 0.0007041379310344828,
      "loss": 2.355663776397705,
      "step": 8980
    },
    {
      "epoch": 0.8982,
      "grad_norm": 0.09541033953428268,
      "learning_rate": 0.0007027586206896552,
      "loss": 2.3288562297821045,
      "step": 8982
    },
    {
      "epoch": 0.8984,
      "grad_norm": 0.0790795087814331,
      "learning_rate": 0.0007013793103448277,
      "loss": 2.288428783416748,
      "step": 8984
    },
    {
      "epoch": 0.8986,
      "grad_norm": 0.08495095372200012,
      "learning_rate": 0.0007,
      "loss": 2.3323540687561035,
      "step": 8986
    },
    {
      "epoch": 0.8988,
      "grad_norm": 0.08215225487947464,
      "learning_rate": 0.0006986206896551724,
      "loss": 2.328554153442383,
      "step": 8988
    },
    {
      "epoch": 0.899,
      "grad_norm": 0.08415501564741135,
      "learning_rate": 0.0006972413793103448,
      "loss": 2.317833423614502,
      "step": 8990
    },
    {
      "epoch": 0.8992,
      "grad_norm": 0.07736361026763916,
      "learning_rate": 0.0006958620689655173,
      "loss": 2.2733359336853027,
      "step": 8992
    },
    {
      "epoch": 0.8994,
      "grad_norm": 0.0829680860042572,
      "learning_rate": 0.0006944827586206897,
      "loss": 2.2712819576263428,
      "step": 8994
    },
    {
      "epoch": 0.8996,
      "grad_norm": 0.08202574402093887,
      "learning_rate": 0.0006931034482758621,
      "loss": 2.294654607772827,
      "step": 8996
    },
    {
      "epoch": 0.8998,
      "grad_norm": 0.07706158608198166,
      "learning_rate": 0.0006917241379310345,
      "loss": 2.3103506565093994,
      "step": 8998
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.08145575225353241,
      "learning_rate": 0.0006903448275862069,
      "loss": 2.3749165534973145,
      "step": 9000
    },
    {
      "epoch": 0.9002,
      "grad_norm": 0.07983099669218063,
      "learning_rate": 0.0006889655172413793,
      "loss": 2.321338176727295,
      "step": 9002
    },
    {
      "epoch": 0.9004,
      "grad_norm": 0.08485081791877747,
      "learning_rate": 0.0006875862068965517,
      "loss": 2.2465991973876953,
      "step": 9004
    },
    {
      "epoch": 0.9006,
      "grad_norm": 0.07836580276489258,
      "learning_rate": 0.0006862068965517242,
      "loss": 2.2640273571014404,
      "step": 9006
    },
    {
      "epoch": 0.9008,
      "grad_norm": 0.07672987133264542,
      "learning_rate": 0.0006848275862068965,
      "loss": 2.3338422775268555,
      "step": 9008
    },
    {
      "epoch": 0.901,
      "grad_norm": 0.08094470202922821,
      "learning_rate": 0.0006834482758620689,
      "loss": 2.300093173980713,
      "step": 9010
    },
    {
      "epoch": 0.9012,
      "grad_norm": 0.07684634625911713,
      "learning_rate": 0.0006820689655172414,
      "loss": 2.3220574855804443,
      "step": 9012
    },
    {
      "epoch": 0.9014,
      "grad_norm": 0.08274059742689133,
      "learning_rate": 0.0006806896551724138,
      "loss": 2.2708382606506348,
      "step": 9014
    },
    {
      "epoch": 0.9016,
      "grad_norm": 0.0811045914888382,
      "learning_rate": 0.0006793103448275862,
      "loss": 2.3458282947540283,
      "step": 9016
    },
    {
      "epoch": 0.9018,
      "grad_norm": 0.08330173045396805,
      "learning_rate": 0.0006779310344827587,
      "loss": 2.2482361793518066,
      "step": 9018
    },
    {
      "epoch": 0.902,
      "grad_norm": 0.08005283772945404,
      "learning_rate": 0.0006765517241379311,
      "loss": 2.2900304794311523,
      "step": 9020
    },
    {
      "epoch": 0.9022,
      "grad_norm": 0.08553680032491684,
      "learning_rate": 0.0006751724137931034,
      "loss": 2.281824827194214,
      "step": 9022
    },
    {
      "epoch": 0.9024,
      "grad_norm": 0.07718579471111298,
      "learning_rate": 0.0006737931034482758,
      "loss": 2.3696227073669434,
      "step": 9024
    },
    {
      "epoch": 0.9026,
      "grad_norm": 0.07831374555826187,
      "learning_rate": 0.0006724137931034483,
      "loss": 2.3242626190185547,
      "step": 9026
    },
    {
      "epoch": 0.9028,
      "grad_norm": 0.07969929277896881,
      "learning_rate": 0.0006710344827586207,
      "loss": 2.248699188232422,
      "step": 9028
    },
    {
      "epoch": 0.903,
      "grad_norm": 0.08007797598838806,
      "learning_rate": 0.0006696551724137931,
      "loss": 2.377521276473999,
      "step": 9030
    },
    {
      "epoch": 0.9032,
      "grad_norm": 0.07794005423784256,
      "learning_rate": 0.0006682758620689656,
      "loss": 2.4001967906951904,
      "step": 9032
    },
    {
      "epoch": 0.9034,
      "grad_norm": 0.07959753274917603,
      "learning_rate": 0.000666896551724138,
      "loss": 2.2768309116363525,
      "step": 9034
    },
    {
      "epoch": 0.9036,
      "grad_norm": 0.08245349675416946,
      "learning_rate": 0.0006655172413793103,
      "loss": 2.2513513565063477,
      "step": 9036
    },
    {
      "epoch": 0.9038,
      "grad_norm": 0.08291910588741302,
      "learning_rate": 0.0006641379310344828,
      "loss": 2.3709776401519775,
      "step": 9038
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.08515916019678116,
      "learning_rate": 0.0006627586206896552,
      "loss": 2.2924816608428955,
      "step": 9040
    },
    {
      "epoch": 0.9042,
      "grad_norm": 0.08018539100885391,
      "learning_rate": 0.0006613793103448276,
      "loss": 2.338871717453003,
      "step": 9042
    },
    {
      "epoch": 0.9044,
      "grad_norm": 0.08126635104417801,
      "learning_rate": 0.00066,
      "loss": 2.372223138809204,
      "step": 9044
    },
    {
      "epoch": 0.9046,
      "grad_norm": 0.0838804841041565,
      "learning_rate": 0.0006586206896551724,
      "loss": 2.376492977142334,
      "step": 9046
    },
    {
      "epoch": 0.9048,
      "grad_norm": 0.07832759618759155,
      "learning_rate": 0.0006572413793103449,
      "loss": 2.380758047103882,
      "step": 9048
    },
    {
      "epoch": 0.905,
      "grad_norm": 0.08398386090993881,
      "learning_rate": 0.0006558620689655172,
      "loss": 2.252264976501465,
      "step": 9050
    },
    {
      "epoch": 0.9052,
      "grad_norm": 0.07939556241035461,
      "learning_rate": 0.0006544827586206897,
      "loss": 2.330033779144287,
      "step": 9052
    },
    {
      "epoch": 0.9054,
      "grad_norm": 0.08452840894460678,
      "learning_rate": 0.0006531034482758621,
      "loss": 2.357936143875122,
      "step": 9054
    },
    {
      "epoch": 0.9056,
      "grad_norm": 0.07861179113388062,
      "learning_rate": 0.0006517241379310344,
      "loss": 2.2528257369995117,
      "step": 9056
    },
    {
      "epoch": 0.9058,
      "grad_norm": 0.08050607144832611,
      "learning_rate": 0.000650344827586207,
      "loss": 2.2785098552703857,
      "step": 9058
    },
    {
      "epoch": 0.906,
      "grad_norm": 0.07919915020465851,
      "learning_rate": 0.0006489655172413793,
      "loss": 2.3206872940063477,
      "step": 9060
    },
    {
      "epoch": 0.9062,
      "grad_norm": 0.08116759359836578,
      "learning_rate": 0.0006475862068965518,
      "loss": 2.3044028282165527,
      "step": 9062
    },
    {
      "epoch": 0.9064,
      "grad_norm": 0.08287791907787323,
      "learning_rate": 0.0006462068965517241,
      "loss": 2.312238931655884,
      "step": 9064
    },
    {
      "epoch": 0.9066,
      "grad_norm": 0.07818850874900818,
      "learning_rate": 0.0006448275862068966,
      "loss": 2.3000333309173584,
      "step": 9066
    },
    {
      "epoch": 0.9068,
      "grad_norm": 0.08858446776866913,
      "learning_rate": 0.000643448275862069,
      "loss": 2.3898026943206787,
      "step": 9068
    },
    {
      "epoch": 0.907,
      "grad_norm": 0.08354880660772324,
      "learning_rate": 0.0006420689655172413,
      "loss": 2.2834010124206543,
      "step": 9070
    },
    {
      "epoch": 0.9072,
      "grad_norm": 0.08100327849388123,
      "learning_rate": 0.0006406896551724139,
      "loss": 2.3130345344543457,
      "step": 9072
    },
    {
      "epoch": 0.9074,
      "grad_norm": 0.08035283535718918,
      "learning_rate": 0.0006393103448275862,
      "loss": 2.210296154022217,
      "step": 9074
    },
    {
      "epoch": 0.9076,
      "grad_norm": 0.08047741651535034,
      "learning_rate": 0.0006379310344827587,
      "loss": 2.3291633129119873,
      "step": 9076
    },
    {
      "epoch": 0.9078,
      "grad_norm": 0.07862217724323273,
      "learning_rate": 0.0006365517241379311,
      "loss": 2.3417446613311768,
      "step": 9078
    },
    {
      "epoch": 0.908,
      "grad_norm": 0.07982630282640457,
      "learning_rate": 0.0006351724137931034,
      "loss": 2.3571395874023438,
      "step": 9080
    },
    {
      "epoch": 0.9082,
      "grad_norm": 0.07628427445888519,
      "learning_rate": 0.0006337931034482759,
      "loss": 2.278214454650879,
      "step": 9082
    },
    {
      "epoch": 0.9084,
      "grad_norm": 0.0846344530582428,
      "learning_rate": 0.0006324137931034482,
      "loss": 2.3704817295074463,
      "step": 9084
    },
    {
      "epoch": 0.9086,
      "grad_norm": 0.08451878279447556,
      "learning_rate": 0.0006310344827586208,
      "loss": 2.3079750537872314,
      "step": 9086
    },
    {
      "epoch": 0.9088,
      "grad_norm": 0.07840487360954285,
      "learning_rate": 0.0006296551724137931,
      "loss": 2.3522043228149414,
      "step": 9088
    },
    {
      "epoch": 0.909,
      "grad_norm": 0.07820161432027817,
      "learning_rate": 0.0006282758620689654,
      "loss": 2.3476531505584717,
      "step": 9090
    },
    {
      "epoch": 0.9092,
      "grad_norm": 0.0813615620136261,
      "learning_rate": 0.000626896551724138,
      "loss": 2.2568986415863037,
      "step": 9092
    },
    {
      "epoch": 0.9094,
      "grad_norm": 0.08074536919593811,
      "learning_rate": 0.0006255172413793103,
      "loss": 2.271224021911621,
      "step": 9094
    },
    {
      "epoch": 0.9096,
      "grad_norm": 0.08967258781194687,
      "learning_rate": 0.0006241379310344828,
      "loss": 2.310100555419922,
      "step": 9096
    },
    {
      "epoch": 0.9098,
      "grad_norm": 0.08368536084890366,
      "learning_rate": 0.0006227586206896551,
      "loss": 2.2827556133270264,
      "step": 9098
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.08090511709451675,
      "learning_rate": 0.0006213793103448277,
      "loss": 2.3341143131256104,
      "step": 9100
    },
    {
      "epoch": 0.9102,
      "grad_norm": 0.079908587038517,
      "learning_rate": 0.00062,
      "loss": 2.291778564453125,
      "step": 9102
    },
    {
      "epoch": 0.9104,
      "grad_norm": 0.08080825209617615,
      "learning_rate": 0.0006186206896551723,
      "loss": 2.3053219318389893,
      "step": 9104
    },
    {
      "epoch": 0.9106,
      "grad_norm": 0.0810495913028717,
      "learning_rate": 0.0006172413793103449,
      "loss": 2.3798770904541016,
      "step": 9106
    },
    {
      "epoch": 0.9108,
      "grad_norm": 0.08151929080486298,
      "learning_rate": 0.0006158620689655172,
      "loss": 2.264303684234619,
      "step": 9108
    },
    {
      "epoch": 0.911,
      "grad_norm": 0.07956381887197495,
      "learning_rate": 0.0006144827586206897,
      "loss": 2.3159310817718506,
      "step": 9110
    },
    {
      "epoch": 0.9112,
      "grad_norm": 0.08235956728458405,
      "learning_rate": 0.0006131034482758621,
      "loss": 2.4286437034606934,
      "step": 9112
    },
    {
      "epoch": 0.9114,
      "grad_norm": 0.07950492948293686,
      "learning_rate": 0.0006117241379310346,
      "loss": 2.2772905826568604,
      "step": 9114
    },
    {
      "epoch": 0.9116,
      "grad_norm": 0.0770396813750267,
      "learning_rate": 0.0006103448275862069,
      "loss": 2.2464072704315186,
      "step": 9116
    },
    {
      "epoch": 0.9118,
      "grad_norm": 0.08214803785085678,
      "learning_rate": 0.0006089655172413792,
      "loss": 2.2810137271881104,
      "step": 9118
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.0780428797006607,
      "learning_rate": 0.0006075862068965518,
      "loss": 2.280090093612671,
      "step": 9120
    },
    {
      "epoch": 0.9122,
      "grad_norm": 0.08088808506727219,
      "learning_rate": 0.0006062068965517241,
      "loss": 2.2888922691345215,
      "step": 9122
    },
    {
      "epoch": 0.9124,
      "grad_norm": 0.08073312044143677,
      "learning_rate": 0.0006048275862068966,
      "loss": 2.321852207183838,
      "step": 9124
    },
    {
      "epoch": 0.9126,
      "grad_norm": 0.07747519761323929,
      "learning_rate": 0.000603448275862069,
      "loss": 2.2999913692474365,
      "step": 9126
    },
    {
      "epoch": 0.9128,
      "grad_norm": 0.07968180626630783,
      "learning_rate": 0.0006020689655172414,
      "loss": 2.389949321746826,
      "step": 9128
    },
    {
      "epoch": 0.913,
      "grad_norm": 0.08092550933361053,
      "learning_rate": 0.0006006896551724138,
      "loss": 2.3255348205566406,
      "step": 9130
    },
    {
      "epoch": 0.9132,
      "grad_norm": 0.07620339840650558,
      "learning_rate": 0.0005993103448275862,
      "loss": 2.359590530395508,
      "step": 9132
    },
    {
      "epoch": 0.9134,
      "grad_norm": 0.08500664681196213,
      "learning_rate": 0.0005979310344827587,
      "loss": 2.325040817260742,
      "step": 9134
    },
    {
      "epoch": 0.9136,
      "grad_norm": 0.0798172578215599,
      "learning_rate": 0.000596551724137931,
      "loss": 2.3262522220611572,
      "step": 9136
    },
    {
      "epoch": 0.9138,
      "grad_norm": 0.0835803896188736,
      "learning_rate": 0.0005951724137931034,
      "loss": 2.2477314472198486,
      "step": 9138
    },
    {
      "epoch": 0.914,
      "grad_norm": 0.07733577489852905,
      "learning_rate": 0.0005937931034482759,
      "loss": 2.309770345687866,
      "step": 9140
    },
    {
      "epoch": 0.9142,
      "grad_norm": 0.08180791139602661,
      "learning_rate": 0.0005924137931034483,
      "loss": 2.3679656982421875,
      "step": 9142
    },
    {
      "epoch": 0.9144,
      "grad_norm": 0.07833100855350494,
      "learning_rate": 0.0005910344827586207,
      "loss": 2.3263814449310303,
      "step": 9144
    },
    {
      "epoch": 0.9146,
      "grad_norm": 0.09069383144378662,
      "learning_rate": 0.0005896551724137931,
      "loss": 2.270331621170044,
      "step": 9146
    },
    {
      "epoch": 0.9148,
      "grad_norm": 0.08209031820297241,
      "learning_rate": 0.0005882758620689656,
      "loss": 2.2137885093688965,
      "step": 9148
    },
    {
      "epoch": 0.915,
      "grad_norm": 0.08096657693386078,
      "learning_rate": 0.0005868965517241379,
      "loss": 2.309213638305664,
      "step": 9150
    },
    {
      "epoch": 0.9152,
      "grad_norm": 0.07927078008651733,
      "learning_rate": 0.0005855172413793104,
      "loss": 2.2825183868408203,
      "step": 9152
    },
    {
      "epoch": 0.9154,
      "grad_norm": 0.07829756289720535,
      "learning_rate": 0.0005841379310344828,
      "loss": 2.2994298934936523,
      "step": 9154
    },
    {
      "epoch": 0.9156,
      "grad_norm": 0.08560130000114441,
      "learning_rate": 0.0005827586206896552,
      "loss": 2.351656436920166,
      "step": 9156
    },
    {
      "epoch": 0.9158,
      "grad_norm": 0.0875101089477539,
      "learning_rate": 0.0005813793103448276,
      "loss": 2.323829174041748,
      "step": 9158
    },
    {
      "epoch": 0.916,
      "grad_norm": 0.07831353694200516,
      "learning_rate": 0.00058,
      "loss": 2.3003475666046143,
      "step": 9160
    },
    {
      "epoch": 0.9162,
      "grad_norm": 0.08028778433799744,
      "learning_rate": 0.0005786206896551724,
      "loss": 2.3475868701934814,
      "step": 9162
    },
    {
      "epoch": 0.9164,
      "grad_norm": 0.08030833303928375,
      "learning_rate": 0.0005772413793103448,
      "loss": 2.3249926567077637,
      "step": 9164
    },
    {
      "epoch": 0.9166,
      "grad_norm": 0.0786430761218071,
      "learning_rate": 0.0005758620689655173,
      "loss": 2.3236727714538574,
      "step": 9166
    },
    {
      "epoch": 0.9168,
      "grad_norm": 0.08178653568029404,
      "learning_rate": 0.0005744827586206897,
      "loss": 2.306929111480713,
      "step": 9168
    },
    {
      "epoch": 0.917,
      "grad_norm": 0.07906122505664825,
      "learning_rate": 0.000573103448275862,
      "loss": 2.3696768283843994,
      "step": 9170
    },
    {
      "epoch": 0.9172,
      "grad_norm": 0.07822833955287933,
      "learning_rate": 0.0005717241379310346,
      "loss": 2.320056915283203,
      "step": 9172
    },
    {
      "epoch": 0.9174,
      "grad_norm": 0.07944116741418839,
      "learning_rate": 0.0005703448275862069,
      "loss": 2.242115020751953,
      "step": 9174
    },
    {
      "epoch": 0.9176,
      "grad_norm": 0.08131048083305359,
      "learning_rate": 0.0005689655172413793,
      "loss": 2.3236260414123535,
      "step": 9176
    },
    {
      "epoch": 0.9178,
      "grad_norm": 0.08419989794492722,
      "learning_rate": 0.0005675862068965517,
      "loss": 2.3626880645751953,
      "step": 9178
    },
    {
      "epoch": 0.918,
      "grad_norm": 0.08120336383581161,
      "learning_rate": 0.0005662068965517242,
      "loss": 2.3249056339263916,
      "step": 9180
    },
    {
      "epoch": 0.9182,
      "grad_norm": 0.07895584404468536,
      "learning_rate": 0.0005648275862068966,
      "loss": 2.3399643898010254,
      "step": 9182
    },
    {
      "epoch": 0.9184,
      "grad_norm": 0.07616822421550751,
      "learning_rate": 0.0005634482758620689,
      "loss": 2.283245325088501,
      "step": 9184
    },
    {
      "epoch": 0.9186,
      "grad_norm": 0.07912542670965195,
      "learning_rate": 0.0005620689655172414,
      "loss": 2.2173428535461426,
      "step": 9186
    },
    {
      "epoch": 0.9188,
      "grad_norm": 0.07951515913009644,
      "learning_rate": 0.0005606896551724138,
      "loss": 2.2191109657287598,
      "step": 9188
    },
    {
      "epoch": 0.919,
      "grad_norm": 0.07882311195135117,
      "learning_rate": 0.0005593103448275862,
      "loss": 2.3075993061065674,
      "step": 9190
    },
    {
      "epoch": 0.9192,
      "grad_norm": 0.08021175116300583,
      "learning_rate": 0.0005579310344827586,
      "loss": 2.2569849491119385,
      "step": 9192
    },
    {
      "epoch": 0.9194,
      "grad_norm": 0.07641041278839111,
      "learning_rate": 0.0005565517241379311,
      "loss": 2.2952992916107178,
      "step": 9194
    },
    {
      "epoch": 0.9196,
      "grad_norm": 0.0782671868801117,
      "learning_rate": 0.0005551724137931035,
      "loss": 2.285463809967041,
      "step": 9196
    },
    {
      "epoch": 0.9198,
      "grad_norm": 0.075486920773983,
      "learning_rate": 0.0005537931034482758,
      "loss": 2.3206214904785156,
      "step": 9198
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.07987932860851288,
      "learning_rate": 0.0005524137931034483,
      "loss": 2.308666706085205,
      "step": 9200
    },
    {
      "epoch": 0.9202,
      "grad_norm": 0.08367325365543365,
      "learning_rate": 0.0005510344827586207,
      "loss": 2.305549383163452,
      "step": 9202
    },
    {
      "epoch": 0.9204,
      "grad_norm": 0.08025375753641129,
      "learning_rate": 0.0005496551724137931,
      "loss": 2.3578200340270996,
      "step": 9204
    },
    {
      "epoch": 0.9206,
      "grad_norm": 0.07905250042676926,
      "learning_rate": 0.0005482758620689656,
      "loss": 2.371040105819702,
      "step": 9206
    },
    {
      "epoch": 0.9208,
      "grad_norm": 0.08128735423088074,
      "learning_rate": 0.000546896551724138,
      "loss": 2.37032413482666,
      "step": 9208
    },
    {
      "epoch": 0.921,
      "grad_norm": 0.08282549679279327,
      "learning_rate": 0.0005455172413793103,
      "loss": 2.2583799362182617,
      "step": 9210
    },
    {
      "epoch": 0.9212,
      "grad_norm": 0.10167878866195679,
      "learning_rate": 0.0005441379310344827,
      "loss": 2.284208297729492,
      "step": 9212
    },
    {
      "epoch": 0.9214,
      "grad_norm": 0.08701951056718826,
      "learning_rate": 0.0005427586206896552,
      "loss": 2.382791519165039,
      "step": 9214
    },
    {
      "epoch": 0.9216,
      "grad_norm": 0.08024299144744873,
      "learning_rate": 0.0005413793103448276,
      "loss": 2.3067502975463867,
      "step": 9216
    },
    {
      "epoch": 0.9218,
      "grad_norm": 0.07950326055288315,
      "learning_rate": 0.00054,
      "loss": 2.4089956283569336,
      "step": 9218
    },
    {
      "epoch": 0.922,
      "grad_norm": 0.0802566409111023,
      "learning_rate": 0.0005386206896551725,
      "loss": 2.3250629901885986,
      "step": 9220
    },
    {
      "epoch": 0.9222,
      "grad_norm": 0.08091439306735992,
      "learning_rate": 0.0005372413793103449,
      "loss": 2.363241672515869,
      "step": 9222
    },
    {
      "epoch": 0.9224,
      "grad_norm": 0.07880213111639023,
      "learning_rate": 0.0005358620689655172,
      "loss": 2.3455312252044678,
      "step": 9224
    },
    {
      "epoch": 0.9226,
      "grad_norm": 0.08048462867736816,
      "learning_rate": 0.0005344827586206897,
      "loss": 2.3283751010894775,
      "step": 9226
    },
    {
      "epoch": 0.9228,
      "grad_norm": 0.07924452424049377,
      "learning_rate": 0.0005331034482758621,
      "loss": 2.285543203353882,
      "step": 9228
    },
    {
      "epoch": 0.923,
      "grad_norm": 0.0805445984005928,
      "learning_rate": 0.0005317241379310345,
      "loss": 2.356816053390503,
      "step": 9230
    },
    {
      "epoch": 0.9232,
      "grad_norm": 0.08062271773815155,
      "learning_rate": 0.0005303448275862069,
      "loss": 2.319025993347168,
      "step": 9232
    },
    {
      "epoch": 0.9234,
      "grad_norm": 0.08059551566839218,
      "learning_rate": 0.0005289655172413793,
      "loss": 2.373528003692627,
      "step": 9234
    },
    {
      "epoch": 0.9236,
      "grad_norm": 0.07965069264173508,
      "learning_rate": 0.0005275862068965518,
      "loss": 2.3252272605895996,
      "step": 9236
    },
    {
      "epoch": 0.9238,
      "grad_norm": 0.07978533208370209,
      "learning_rate": 0.0005262068965517241,
      "loss": 2.3865976333618164,
      "step": 9238
    },
    {
      "epoch": 0.924,
      "grad_norm": 0.07824084162712097,
      "learning_rate": 0.0005248275862068966,
      "loss": 2.3416671752929688,
      "step": 9240
    },
    {
      "epoch": 0.9242,
      "grad_norm": 0.08132484555244446,
      "learning_rate": 0.000523448275862069,
      "loss": 2.3043203353881836,
      "step": 9242
    },
    {
      "epoch": 0.9244,
      "grad_norm": 0.07825306802988052,
      "learning_rate": 0.0005220689655172413,
      "loss": 2.393069267272949,
      "step": 9244
    },
    {
      "epoch": 0.9246,
      "grad_norm": 0.07680784910917282,
      "learning_rate": 0.0005206896551724139,
      "loss": 2.3226301670074463,
      "step": 9246
    },
    {
      "epoch": 0.9248,
      "grad_norm": 0.0797777995467186,
      "learning_rate": 0.0005193103448275862,
      "loss": 2.279766798019409,
      "step": 9248
    },
    {
      "epoch": 0.925,
      "grad_norm": 0.08178012818098068,
      "learning_rate": 0.0005179310344827586,
      "loss": 2.3050594329833984,
      "step": 9250
    },
    {
      "epoch": 0.9252,
      "grad_norm": 0.08338954299688339,
      "learning_rate": 0.000516551724137931,
      "loss": 2.355121374130249,
      "step": 9252
    },
    {
      "epoch": 0.9254,
      "grad_norm": 0.07707846164703369,
      "learning_rate": 0.0005151724137931035,
      "loss": 2.391958713531494,
      "step": 9254
    },
    {
      "epoch": 0.9256,
      "grad_norm": 0.0838618054986,
      "learning_rate": 0.0005137931034482759,
      "loss": 2.332699775695801,
      "step": 9256
    },
    {
      "epoch": 0.9258,
      "grad_norm": 0.08067038655281067,
      "learning_rate": 0.0005124137931034482,
      "loss": 2.3155386447906494,
      "step": 9258
    },
    {
      "epoch": 0.926,
      "grad_norm": 0.07601403445005417,
      "learning_rate": 0.0005110344827586208,
      "loss": 2.2715039253234863,
      "step": 9260
    },
    {
      "epoch": 0.9262,
      "grad_norm": 0.07888073474168777,
      "learning_rate": 0.0005096551724137931,
      "loss": 2.3582777976989746,
      "step": 9262
    },
    {
      "epoch": 0.9264,
      "grad_norm": 0.08189954608678818,
      "learning_rate": 0.0005082758620689655,
      "loss": 2.2767550945281982,
      "step": 9264
    },
    {
      "epoch": 0.9266,
      "grad_norm": 0.0782490000128746,
      "learning_rate": 0.000506896551724138,
      "loss": 2.3489558696746826,
      "step": 9266
    },
    {
      "epoch": 0.9268,
      "grad_norm": 0.08404358476400375,
      "learning_rate": 0.0005055172413793103,
      "loss": 2.373305082321167,
      "step": 9268
    },
    {
      "epoch": 0.927,
      "grad_norm": 0.08019094169139862,
      "learning_rate": 0.0005041379310344828,
      "loss": 2.3255529403686523,
      "step": 9270
    },
    {
      "epoch": 0.9272,
      "grad_norm": 0.08086287975311279,
      "learning_rate": 0.0005027586206896551,
      "loss": 2.3973183631896973,
      "step": 9272
    },
    {
      "epoch": 0.9274,
      "grad_norm": 0.0791512206196785,
      "learning_rate": 0.0005013793103448277,
      "loss": 2.2964699268341064,
      "step": 9274
    },
    {
      "epoch": 0.9276,
      "grad_norm": 0.07904848456382751,
      "learning_rate": 0.0005,
      "loss": 2.334219455718994,
      "step": 9276
    },
    {
      "epoch": 0.9278,
      "grad_norm": 0.0841880589723587,
      "learning_rate": 0.0004986206896551725,
      "loss": 2.276782989501953,
      "step": 9278
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.07858940958976746,
      "learning_rate": 0.0004972413793103449,
      "loss": 2.290588855743408,
      "step": 9280
    },
    {
      "epoch": 0.9282,
      "grad_norm": 0.08773598074913025,
      "learning_rate": 0.0004958620689655172,
      "loss": 2.333935499191284,
      "step": 9282
    },
    {
      "epoch": 0.9284,
      "grad_norm": 0.08233916014432907,
      "learning_rate": 0.0004944827586206897,
      "loss": 2.373599052429199,
      "step": 9284
    },
    {
      "epoch": 0.9286,
      "grad_norm": 0.08290643990039825,
      "learning_rate": 0.000493103448275862,
      "loss": 2.329275131225586,
      "step": 9286
    },
    {
      "epoch": 0.9288,
      "grad_norm": 0.07779265940189362,
      "learning_rate": 0.0004917241379310345,
      "loss": 2.2847580909729004,
      "step": 9288
    },
    {
      "epoch": 0.929,
      "grad_norm": 0.08277283608913422,
      "learning_rate": 0.0004903448275862069,
      "loss": 2.3317511081695557,
      "step": 9290
    },
    {
      "epoch": 0.9292,
      "grad_norm": 0.08392252027988434,
      "learning_rate": 0.0004889655172413793,
      "loss": 2.386775255203247,
      "step": 9292
    },
    {
      "epoch": 0.9294,
      "grad_norm": 0.08098266273736954,
      "learning_rate": 0.00048758620689655173,
      "loss": 2.342334270477295,
      "step": 9294
    },
    {
      "epoch": 0.9296,
      "grad_norm": 0.08113212138414383,
      "learning_rate": 0.0004862068965517241,
      "loss": 2.3174381256103516,
      "step": 9296
    },
    {
      "epoch": 0.9298,
      "grad_norm": 0.08084134757518768,
      "learning_rate": 0.00048482758620689657,
      "loss": 2.290421724319458,
      "step": 9298
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.07912883907556534,
      "learning_rate": 0.00048344827586206896,
      "loss": 2.343431234359741,
      "step": 9300
    },
    {
      "epoch": 0.9302,
      "grad_norm": 0.08213602751493454,
      "learning_rate": 0.0004820689655172414,
      "loss": 2.3119139671325684,
      "step": 9302
    },
    {
      "epoch": 0.9304,
      "grad_norm": 0.08455326408147812,
      "learning_rate": 0.0004806896551724138,
      "loss": 2.292022466659546,
      "step": 9304
    },
    {
      "epoch": 0.9306,
      "grad_norm": 0.07699854671955109,
      "learning_rate": 0.00047931034482758623,
      "loss": 2.284964084625244,
      "step": 9306
    },
    {
      "epoch": 0.9308,
      "grad_norm": 0.07991994917392731,
      "learning_rate": 0.0004779310344827587,
      "loss": 2.338876485824585,
      "step": 9308
    },
    {
      "epoch": 0.931,
      "grad_norm": 0.07919906824827194,
      "learning_rate": 0.000476551724137931,
      "loss": 2.3097803592681885,
      "step": 9310
    },
    {
      "epoch": 0.9312,
      "grad_norm": 0.08032292127609253,
      "learning_rate": 0.00047517241379310346,
      "loss": 2.2821314334869385,
      "step": 9312
    },
    {
      "epoch": 0.9314,
      "grad_norm": 0.08009005337953568,
      "learning_rate": 0.00047379310344827585,
      "loss": 2.3972344398498535,
      "step": 9314
    },
    {
      "epoch": 0.9316,
      "grad_norm": 0.07919152826070786,
      "learning_rate": 0.0004724137931034483,
      "loss": 2.3200786113739014,
      "step": 9316
    },
    {
      "epoch": 0.9318,
      "grad_norm": 0.08129893243312836,
      "learning_rate": 0.00047103448275862073,
      "loss": 2.327148199081421,
      "step": 9318
    },
    {
      "epoch": 0.932,
      "grad_norm": 0.07758411020040512,
      "learning_rate": 0.0004696551724137931,
      "loss": 2.282979726791382,
      "step": 9320
    },
    {
      "epoch": 0.9322,
      "grad_norm": 0.07698608189821243,
      "learning_rate": 0.00046827586206896557,
      "loss": 2.334815263748169,
      "step": 9322
    },
    {
      "epoch": 0.9324,
      "grad_norm": 0.07886682450771332,
      "learning_rate": 0.0004668965517241379,
      "loss": 2.414616346359253,
      "step": 9324
    },
    {
      "epoch": 0.9326,
      "grad_norm": 0.07773898541927338,
      "learning_rate": 0.00046551724137931035,
      "loss": 2.251735210418701,
      "step": 9326
    },
    {
      "epoch": 0.9328,
      "grad_norm": 0.07711044698953629,
      "learning_rate": 0.00046413793103448274,
      "loss": 2.369746208190918,
      "step": 9328
    },
    {
      "epoch": 0.933,
      "grad_norm": 0.08883288502693176,
      "learning_rate": 0.0004627586206896552,
      "loss": 2.325996160507202,
      "step": 9330
    },
    {
      "epoch": 0.9332,
      "grad_norm": 0.07917878031730652,
      "learning_rate": 0.0004613793103448276,
      "loss": 2.2856297492980957,
      "step": 9332
    },
    {
      "epoch": 0.9334,
      "grad_norm": 0.07864882051944733,
      "learning_rate": 0.00046,
      "loss": 2.3300349712371826,
      "step": 9334
    },
    {
      "epoch": 0.9336,
      "grad_norm": 0.07954011112451553,
      "learning_rate": 0.0004586206896551724,
      "loss": 2.267597198486328,
      "step": 9336
    },
    {
      "epoch": 0.9338,
      "grad_norm": 0.07912689447402954,
      "learning_rate": 0.0004572413793103448,
      "loss": 2.300138235092163,
      "step": 9338
    },
    {
      "epoch": 0.934,
      "grad_norm": 0.0863974541425705,
      "learning_rate": 0.00045586206896551724,
      "loss": 2.3356611728668213,
      "step": 9340
    },
    {
      "epoch": 0.9342,
      "grad_norm": 0.08051128685474396,
      "learning_rate": 0.0004544827586206897,
      "loss": 2.298614978790283,
      "step": 9342
    },
    {
      "epoch": 0.9344,
      "grad_norm": 0.079220712184906,
      "learning_rate": 0.00045310344827586207,
      "loss": 2.325369119644165,
      "step": 9344
    },
    {
      "epoch": 0.9346,
      "grad_norm": 0.0800003632903099,
      "learning_rate": 0.0004517241379310345,
      "loss": 2.335437536239624,
      "step": 9346
    },
    {
      "epoch": 0.9348,
      "grad_norm": 0.07500538229942322,
      "learning_rate": 0.0004503448275862069,
      "loss": 2.282726287841797,
      "step": 9348
    },
    {
      "epoch": 0.935,
      "grad_norm": 0.07805955410003662,
      "learning_rate": 0.0004489655172413793,
      "loss": 2.3440310955047607,
      "step": 9350
    },
    {
      "epoch": 0.9352,
      "grad_norm": 0.08114265650510788,
      "learning_rate": 0.00044758620689655174,
      "loss": 2.328864336013794,
      "step": 9352
    },
    {
      "epoch": 0.9354,
      "grad_norm": 0.08233185857534409,
      "learning_rate": 0.0004462068965517241,
      "loss": 2.367729425430298,
      "step": 9354
    },
    {
      "epoch": 0.9356,
      "grad_norm": 0.077791728079319,
      "learning_rate": 0.00044482758620689657,
      "loss": 2.2717702388763428,
      "step": 9356
    },
    {
      "epoch": 0.9358,
      "grad_norm": 0.0788520872592926,
      "learning_rate": 0.00044344827586206896,
      "loss": 2.383053779602051,
      "step": 9358
    },
    {
      "epoch": 0.936,
      "grad_norm": 0.0802096351981163,
      "learning_rate": 0.0004420689655172414,
      "loss": 2.3336832523345947,
      "step": 9360
    },
    {
      "epoch": 0.9362,
      "grad_norm": 0.07962863892316818,
      "learning_rate": 0.00044068965517241385,
      "loss": 2.332352638244629,
      "step": 9362
    },
    {
      "epoch": 0.9364,
      "grad_norm": 0.07901357859373093,
      "learning_rate": 0.0004393103448275862,
      "loss": 2.3047730922698975,
      "step": 9364
    },
    {
      "epoch": 0.9366,
      "grad_norm": 0.08156507462263107,
      "learning_rate": 0.00043793103448275863,
      "loss": 2.353278398513794,
      "step": 9366
    },
    {
      "epoch": 0.9368,
      "grad_norm": 0.07867038249969482,
      "learning_rate": 0.000436551724137931,
      "loss": 2.3351850509643555,
      "step": 9368
    },
    {
      "epoch": 0.937,
      "grad_norm": 0.08329422026872635,
      "learning_rate": 0.00043517241379310346,
      "loss": 2.2400436401367188,
      "step": 9370
    },
    {
      "epoch": 0.9372,
      "grad_norm": 0.0777529627084732,
      "learning_rate": 0.0004337931034482759,
      "loss": 2.2925734519958496,
      "step": 9372
    },
    {
      "epoch": 0.9374,
      "grad_norm": 0.08021242171525955,
      "learning_rate": 0.0004324137931034483,
      "loss": 2.261256217956543,
      "step": 9374
    },
    {
      "epoch": 0.9376,
      "grad_norm": 0.08222951740026474,
      "learning_rate": 0.0004310344827586207,
      "loss": 2.31005859375,
      "step": 9376
    },
    {
      "epoch": 0.9378,
      "grad_norm": 0.0798024907708168,
      "learning_rate": 0.0004296551724137931,
      "loss": 2.245692014694214,
      "step": 9378
    },
    {
      "epoch": 0.938,
      "grad_norm": 0.08035808801651001,
      "learning_rate": 0.0004282758620689655,
      "loss": 2.309943914413452,
      "step": 9380
    },
    {
      "epoch": 0.9382,
      "grad_norm": 0.08066728711128235,
      "learning_rate": 0.00042689655172413796,
      "loss": 2.3276736736297607,
      "step": 9382
    },
    {
      "epoch": 0.9384,
      "grad_norm": 0.07878252863883972,
      "learning_rate": 0.00042551724137931035,
      "loss": 2.3342411518096924,
      "step": 9384
    },
    {
      "epoch": 0.9386,
      "grad_norm": 0.07646515220403671,
      "learning_rate": 0.0004241379310344828,
      "loss": 2.254122018814087,
      "step": 9386
    },
    {
      "epoch": 0.9388,
      "grad_norm": 0.08139613270759583,
      "learning_rate": 0.0004227586206896552,
      "loss": 2.3871066570281982,
      "step": 9388
    },
    {
      "epoch": 0.939,
      "grad_norm": 0.0784415751695633,
      "learning_rate": 0.0004213793103448276,
      "loss": 2.2639026641845703,
      "step": 9390
    },
    {
      "epoch": 0.9392,
      "grad_norm": 0.08214456588029861,
      "learning_rate": 0.00042,
      "loss": 2.4069101810455322,
      "step": 9392
    },
    {
      "epoch": 0.9394,
      "grad_norm": 0.08408313244581223,
      "learning_rate": 0.0004186206896551724,
      "loss": 2.3401870727539062,
      "step": 9394
    },
    {
      "epoch": 0.9396,
      "grad_norm": 0.08221951872110367,
      "learning_rate": 0.00041724137931034485,
      "loss": 2.3978006839752197,
      "step": 9396
    },
    {
      "epoch": 0.9398,
      "grad_norm": 0.08303272724151611,
      "learning_rate": 0.00041586206896551724,
      "loss": 2.33444881439209,
      "step": 9398
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.08755884319543839,
      "learning_rate": 0.0004144827586206897,
      "loss": 2.3574697971343994,
      "step": 9400
    },
    {
      "epoch": 0.9402,
      "grad_norm": 0.07844647020101547,
      "learning_rate": 0.00041310344827586213,
      "loss": 2.3427329063415527,
      "step": 9402
    },
    {
      "epoch": 0.9404,
      "grad_norm": 0.08141172677278519,
      "learning_rate": 0.00041172413793103446,
      "loss": 2.2753894329071045,
      "step": 9404
    },
    {
      "epoch": 0.9406,
      "grad_norm": 0.08616205304861069,
      "learning_rate": 0.0004103448275862069,
      "loss": 2.342247486114502,
      "step": 9406
    },
    {
      "epoch": 0.9408,
      "grad_norm": 0.07867114990949631,
      "learning_rate": 0.0004089655172413793,
      "loss": 2.286829948425293,
      "step": 9408
    },
    {
      "epoch": 0.941,
      "grad_norm": 0.0817803218960762,
      "learning_rate": 0.00040758620689655174,
      "loss": 2.338228702545166,
      "step": 9410
    },
    {
      "epoch": 0.9412,
      "grad_norm": 0.07700518518686295,
      "learning_rate": 0.0004062068965517242,
      "loss": 2.3351898193359375,
      "step": 9412
    },
    {
      "epoch": 0.9414,
      "grad_norm": 0.0791410505771637,
      "learning_rate": 0.0004048275862068966,
      "loss": 2.393407106399536,
      "step": 9414
    },
    {
      "epoch": 0.9416,
      "grad_norm": 0.07984901964664459,
      "learning_rate": 0.00040344827586206896,
      "loss": 2.336146831512451,
      "step": 9416
    },
    {
      "epoch": 0.9418,
      "grad_norm": 0.08341842889785767,
      "learning_rate": 0.00040206896551724135,
      "loss": 2.38725209236145,
      "step": 9418
    },
    {
      "epoch": 0.942,
      "grad_norm": 0.07996757328510284,
      "learning_rate": 0.0004006896551724138,
      "loss": 2.3744406700134277,
      "step": 9420
    },
    {
      "epoch": 0.9422,
      "grad_norm": 0.08057905733585358,
      "learning_rate": 0.0003993103448275862,
      "loss": 2.3074398040771484,
      "step": 9422
    },
    {
      "epoch": 0.9424,
      "grad_norm": 0.08138029277324677,
      "learning_rate": 0.00039793103448275863,
      "loss": 2.2957205772399902,
      "step": 9424
    },
    {
      "epoch": 0.9426,
      "grad_norm": 0.08205711096525192,
      "learning_rate": 0.0003965517241379311,
      "loss": 2.3836302757263184,
      "step": 9426
    },
    {
      "epoch": 0.9428,
      "grad_norm": 0.08113666623830795,
      "learning_rate": 0.00039517241379310346,
      "loss": 2.3148856163024902,
      "step": 9428
    },
    {
      "epoch": 0.943,
      "grad_norm": 0.08345279842615128,
      "learning_rate": 0.00039379310344827585,
      "loss": 2.356283187866211,
      "step": 9430
    },
    {
      "epoch": 0.9432,
      "grad_norm": 0.08036386966705322,
      "learning_rate": 0.00039241379310344824,
      "loss": 2.3600893020629883,
      "step": 9432
    },
    {
      "epoch": 0.9434,
      "grad_norm": 0.07640381902456284,
      "learning_rate": 0.0003910344827586207,
      "loss": 2.3116772174835205,
      "step": 9434
    },
    {
      "epoch": 0.9436,
      "grad_norm": 0.08101827651262283,
      "learning_rate": 0.00038965517241379313,
      "loss": 2.322699785232544,
      "step": 9436
    },
    {
      "epoch": 0.9438,
      "grad_norm": 0.09103842824697495,
      "learning_rate": 0.0003882758620689655,
      "loss": 2.2626795768737793,
      "step": 9438
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.07836393266916275,
      "learning_rate": 0.00038689655172413796,
      "loss": 2.317237377166748,
      "step": 9440
    },
    {
      "epoch": 0.9442,
      "grad_norm": 0.08116529881954193,
      "learning_rate": 0.00038551724137931035,
      "loss": 2.390596628189087,
      "step": 9442
    },
    {
      "epoch": 0.9444,
      "grad_norm": 0.08591926097869873,
      "learning_rate": 0.00038413793103448274,
      "loss": 2.352504014968872,
      "step": 9444
    },
    {
      "epoch": 0.9446,
      "grad_norm": 0.08183090388774872,
      "learning_rate": 0.0003827586206896552,
      "loss": 2.2604758739471436,
      "step": 9446
    },
    {
      "epoch": 0.9448,
      "grad_norm": 0.07740812748670578,
      "learning_rate": 0.0003813793103448276,
      "loss": 2.314781427383423,
      "step": 9448
    },
    {
      "epoch": 0.945,
      "grad_norm": 0.079798623919487,
      "learning_rate": 0.00038,
      "loss": 2.292912244796753,
      "step": 9450
    },
    {
      "epoch": 0.9452,
      "grad_norm": 0.08282376825809479,
      "learning_rate": 0.0003786206896551724,
      "loss": 2.2967605590820312,
      "step": 9452
    },
    {
      "epoch": 0.9454,
      "grad_norm": 0.07858589291572571,
      "learning_rate": 0.00037724137931034485,
      "loss": 2.2455971240997314,
      "step": 9454
    },
    {
      "epoch": 0.9456,
      "grad_norm": 0.08877740055322647,
      "learning_rate": 0.00037586206896551724,
      "loss": 2.3468027114868164,
      "step": 9456
    },
    {
      "epoch": 0.9458,
      "grad_norm": 0.08037698268890381,
      "learning_rate": 0.00037448275862068963,
      "loss": 2.3693318367004395,
      "step": 9458
    },
    {
      "epoch": 0.946,
      "grad_norm": 0.07633772492408752,
      "learning_rate": 0.0003731034482758621,
      "loss": 2.35552978515625,
      "step": 9460
    },
    {
      "epoch": 0.9462,
      "grad_norm": 0.0815320760011673,
      "learning_rate": 0.00037172413793103447,
      "loss": 2.34963059425354,
      "step": 9462
    },
    {
      "epoch": 0.9464,
      "grad_norm": 0.08786502480506897,
      "learning_rate": 0.0003703448275862069,
      "loss": 2.333101511001587,
      "step": 9464
    },
    {
      "epoch": 0.9466,
      "grad_norm": 0.07735063135623932,
      "learning_rate": 0.00036896551724137935,
      "loss": 2.334210157394409,
      "step": 9466
    },
    {
      "epoch": 0.9468,
      "grad_norm": 0.07943792641162872,
      "learning_rate": 0.00036758620689655174,
      "loss": 2.379848003387451,
      "step": 9468
    },
    {
      "epoch": 0.947,
      "grad_norm": 0.07951770722866058,
      "learning_rate": 0.00036620689655172413,
      "loss": 2.322204351425171,
      "step": 9470
    },
    {
      "epoch": 0.9472,
      "grad_norm": 0.08254657685756683,
      "learning_rate": 0.0003648275862068965,
      "loss": 2.301048517227173,
      "step": 9472
    },
    {
      "epoch": 0.9474,
      "grad_norm": 0.08026549965143204,
      "learning_rate": 0.00036344827586206897,
      "loss": 2.3904972076416016,
      "step": 9474
    },
    {
      "epoch": 0.9476,
      "grad_norm": 0.0821588933467865,
      "learning_rate": 0.0003620689655172414,
      "loss": 2.2950119972229004,
      "step": 9476
    },
    {
      "epoch": 0.9478,
      "grad_norm": 0.08354046940803528,
      "learning_rate": 0.0003606896551724138,
      "loss": 2.3467531204223633,
      "step": 9478
    },
    {
      "epoch": 0.948,
      "grad_norm": 0.08471706509590149,
      "learning_rate": 0.00035931034482758624,
      "loss": 2.3270397186279297,
      "step": 9480
    },
    {
      "epoch": 0.9482,
      "grad_norm": 0.08081135898828506,
      "learning_rate": 0.00035793103448275863,
      "loss": 2.3442373275756836,
      "step": 9482
    },
    {
      "epoch": 0.9484,
      "grad_norm": 0.08071006834506989,
      "learning_rate": 0.000356551724137931,
      "loss": 2.338052749633789,
      "step": 9484
    },
    {
      "epoch": 0.9486,
      "grad_norm": 0.08289598673582077,
      "learning_rate": 0.00035517241379310347,
      "loss": 2.3394978046417236,
      "step": 9486
    },
    {
      "epoch": 0.9488,
      "grad_norm": 0.0798049196600914,
      "learning_rate": 0.00035379310344827586,
      "loss": 2.331315517425537,
      "step": 9488
    },
    {
      "epoch": 0.949,
      "grad_norm": 0.08267686516046524,
      "learning_rate": 0.0003524137931034483,
      "loss": 2.3439676761627197,
      "step": 9490
    },
    {
      "epoch": 0.9492,
      "grad_norm": 0.08091051876544952,
      "learning_rate": 0.0003510344827586207,
      "loss": 2.3304238319396973,
      "step": 9492
    },
    {
      "epoch": 0.9494,
      "grad_norm": 0.0765710175037384,
      "learning_rate": 0.00034965517241379313,
      "loss": 2.3339614868164062,
      "step": 9494
    },
    {
      "epoch": 0.9496,
      "grad_norm": 0.07866253703832626,
      "learning_rate": 0.0003482758620689655,
      "loss": 2.2295291423797607,
      "step": 9496
    },
    {
      "epoch": 0.9498,
      "grad_norm": 0.08194364607334137,
      "learning_rate": 0.0003468965517241379,
      "loss": 2.3746020793914795,
      "step": 9498
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.08119377493858337,
      "learning_rate": 0.00034551724137931036,
      "loss": 2.3776602745056152,
      "step": 9500
    },
    {
      "epoch": 0.9502,
      "grad_norm": 0.08652893453836441,
      "learning_rate": 0.00034413793103448275,
      "loss": 2.374191999435425,
      "step": 9502
    },
    {
      "epoch": 0.9504,
      "grad_norm": 0.07797849923372269,
      "learning_rate": 0.0003427586206896552,
      "loss": 2.304413318634033,
      "step": 9504
    },
    {
      "epoch": 0.9506,
      "grad_norm": 0.07907082885503769,
      "learning_rate": 0.00034137931034482763,
      "loss": 2.3674428462982178,
      "step": 9506
    },
    {
      "epoch": 0.9508,
      "grad_norm": 0.07997339963912964,
      "learning_rate": 0.00034,
      "loss": 2.385397434234619,
      "step": 9508
    },
    {
      "epoch": 0.951,
      "grad_norm": 0.07961413264274597,
      "learning_rate": 0.0003386206896551724,
      "loss": 2.3671956062316895,
      "step": 9510
    },
    {
      "epoch": 0.9512,
      "grad_norm": 0.08182445168495178,
      "learning_rate": 0.0003372413793103448,
      "loss": 2.325037956237793,
      "step": 9512
    },
    {
      "epoch": 0.9514,
      "grad_norm": 0.07927090674638748,
      "learning_rate": 0.00033586206896551725,
      "loss": 2.3004848957061768,
      "step": 9514
    },
    {
      "epoch": 0.9516,
      "grad_norm": 0.08427319675683975,
      "learning_rate": 0.0003344827586206897,
      "loss": 2.236727714538574,
      "step": 9516
    },
    {
      "epoch": 0.9518,
      "grad_norm": 0.08212579786777496,
      "learning_rate": 0.0003331034482758621,
      "loss": 2.343320369720459,
      "step": 9518
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.08851324766874313,
      "learning_rate": 0.0003317241379310345,
      "loss": 2.4014670848846436,
      "step": 9520
    },
    {
      "epoch": 0.9522,
      "grad_norm": 0.08058400452136993,
      "learning_rate": 0.0003303448275862069,
      "loss": 2.4325387477874756,
      "step": 9522
    },
    {
      "epoch": 0.9524,
      "grad_norm": 0.08165767043828964,
      "learning_rate": 0.0003289655172413793,
      "loss": 2.390857219696045,
      "step": 9524
    },
    {
      "epoch": 0.9526,
      "grad_norm": 0.08380422741174698,
      "learning_rate": 0.0003275862068965517,
      "loss": 2.3439457416534424,
      "step": 9526
    },
    {
      "epoch": 0.9528,
      "grad_norm": 0.08007734268903732,
      "learning_rate": 0.00032620689655172414,
      "loss": 2.4307122230529785,
      "step": 9528
    },
    {
      "epoch": 0.953,
      "grad_norm": 0.08109787106513977,
      "learning_rate": 0.0003248275862068966,
      "loss": 2.328859567642212,
      "step": 9530
    },
    {
      "epoch": 0.9532,
      "grad_norm": 0.07779102027416229,
      "learning_rate": 0.00032344827586206897,
      "loss": 2.329617738723755,
      "step": 9532
    },
    {
      "epoch": 0.9534,
      "grad_norm": 0.08196359872817993,
      "learning_rate": 0.0003220689655172414,
      "loss": 2.3881208896636963,
      "step": 9534
    },
    {
      "epoch": 0.9536,
      "grad_norm": 0.08217644691467285,
      "learning_rate": 0.00032068965517241375,
      "loss": 2.417518138885498,
      "step": 9536
    },
    {
      "epoch": 0.9538,
      "grad_norm": 0.08541196584701538,
      "learning_rate": 0.0003193103448275862,
      "loss": 2.3189826011657715,
      "step": 9538
    },
    {
      "epoch": 0.954,
      "grad_norm": 0.0828603133559227,
      "learning_rate": 0.00031793103448275864,
      "loss": 2.2529988288879395,
      "step": 9540
    },
    {
      "epoch": 0.9542,
      "grad_norm": 0.08243314921855927,
      "learning_rate": 0.00031655172413793103,
      "loss": 2.3811652660369873,
      "step": 9542
    },
    {
      "epoch": 0.9544,
      "grad_norm": 0.07986893504858017,
      "learning_rate": 0.00031517241379310347,
      "loss": 2.383762836456299,
      "step": 9544
    },
    {
      "epoch": 0.9546,
      "grad_norm": 0.07936201244592667,
      "learning_rate": 0.00031379310344827586,
      "loss": 2.349426746368408,
      "step": 9546
    },
    {
      "epoch": 0.9548,
      "grad_norm": 0.07983409613370895,
      "learning_rate": 0.0003124137931034483,
      "loss": 2.3915064334869385,
      "step": 9548
    },
    {
      "epoch": 0.955,
      "grad_norm": 0.07826077938079834,
      "learning_rate": 0.0003110344827586207,
      "loss": 2.291491746902466,
      "step": 9550
    },
    {
      "epoch": 0.9552,
      "grad_norm": 0.0797303318977356,
      "learning_rate": 0.0003096551724137931,
      "loss": 2.376883029937744,
      "step": 9552
    },
    {
      "epoch": 0.9554,
      "grad_norm": 0.08183964341878891,
      "learning_rate": 0.00030827586206896553,
      "loss": 2.455576181411743,
      "step": 9554
    },
    {
      "epoch": 0.9556,
      "grad_norm": 0.08220323920249939,
      "learning_rate": 0.0003068965517241379,
      "loss": 2.350548267364502,
      "step": 9556
    },
    {
      "epoch": 0.9558,
      "grad_norm": 0.07671164721250534,
      "learning_rate": 0.00030551724137931036,
      "loss": 2.318354606628418,
      "step": 9558
    },
    {
      "epoch": 0.956,
      "grad_norm": 0.08048506826162338,
      "learning_rate": 0.0003041379310344828,
      "loss": 2.3003079891204834,
      "step": 9560
    },
    {
      "epoch": 0.9562,
      "grad_norm": 0.08152240514755249,
      "learning_rate": 0.0003027586206896552,
      "loss": 2.40347957611084,
      "step": 9562
    },
    {
      "epoch": 0.9564,
      "grad_norm": 0.07840800285339355,
      "learning_rate": 0.0003013793103448276,
      "loss": 2.277255058288574,
      "step": 9564
    },
    {
      "epoch": 0.9566,
      "grad_norm": 0.08026247471570969,
      "learning_rate": 0.0003,
      "loss": 2.320704460144043,
      "step": 9566
    },
    {
      "epoch": 0.9568,
      "grad_norm": 0.08455732464790344,
      "learning_rate": 0.0002986206896551724,
      "loss": 2.4301373958587646,
      "step": 9568
    },
    {
      "epoch": 0.957,
      "grad_norm": 0.08245748281478882,
      "learning_rate": 0.00029724137931034486,
      "loss": 2.427987575531006,
      "step": 9570
    },
    {
      "epoch": 0.9572,
      "grad_norm": 0.08109789341688156,
      "learning_rate": 0.00029586206896551725,
      "loss": 2.3691928386688232,
      "step": 9572
    },
    {
      "epoch": 0.9574,
      "grad_norm": 0.08310593664646149,
      "learning_rate": 0.0002944827586206897,
      "loss": 2.32045841217041,
      "step": 9574
    },
    {
      "epoch": 0.9576,
      "grad_norm": 0.08386830985546112,
      "learning_rate": 0.00029310344827586203,
      "loss": 2.4201905727386475,
      "step": 9576
    },
    {
      "epoch": 0.9578,
      "grad_norm": 0.08579981327056885,
      "learning_rate": 0.0002917241379310345,
      "loss": 2.359121322631836,
      "step": 9578
    },
    {
      "epoch": 0.958,
      "grad_norm": 0.08118868619203568,
      "learning_rate": 0.0002903448275862069,
      "loss": 2.3743183612823486,
      "step": 9580
    },
    {
      "epoch": 0.9582,
      "grad_norm": 0.07974836975336075,
      "learning_rate": 0.0002889655172413793,
      "loss": 2.447850227355957,
      "step": 9582
    },
    {
      "epoch": 0.9584,
      "grad_norm": 0.07605169713497162,
      "learning_rate": 0.00028758620689655175,
      "loss": 2.289257287979126,
      "step": 9584
    },
    {
      "epoch": 0.9586,
      "grad_norm": 0.07559153437614441,
      "learning_rate": 0.00028620689655172414,
      "loss": 2.307105302810669,
      "step": 9586
    },
    {
      "epoch": 0.9588,
      "grad_norm": 0.07817770540714264,
      "learning_rate": 0.0002848275862068966,
      "loss": 2.3343393802642822,
      "step": 9588
    },
    {
      "epoch": 0.959,
      "grad_norm": 0.08583378791809082,
      "learning_rate": 0.000283448275862069,
      "loss": 2.3160271644592285,
      "step": 9590
    },
    {
      "epoch": 0.9592,
      "grad_norm": 0.0791434571146965,
      "learning_rate": 0.00028206896551724136,
      "loss": 2.331648588180542,
      "step": 9592
    },
    {
      "epoch": 0.9594,
      "grad_norm": 0.07893548160791397,
      "learning_rate": 0.0002806896551724138,
      "loss": 2.34572696685791,
      "step": 9594
    },
    {
      "epoch": 0.9596,
      "grad_norm": 0.08565984666347504,
      "learning_rate": 0.0002793103448275862,
      "loss": 2.4220433235168457,
      "step": 9596
    },
    {
      "epoch": 0.9598,
      "grad_norm": 0.07926066964864731,
      "learning_rate": 0.00027793103448275864,
      "loss": 2.3227932453155518,
      "step": 9598
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.08507687598466873,
      "learning_rate": 0.0002765517241379311,
      "loss": 2.368541717529297,
      "step": 9600
    },
    {
      "epoch": 0.9602,
      "grad_norm": 0.08046536892652512,
      "learning_rate": 0.0002751724137931034,
      "loss": 2.3408727645874023,
      "step": 9602
    },
    {
      "epoch": 0.9604,
      "grad_norm": 0.0807982087135315,
      "learning_rate": 0.00027379310344827586,
      "loss": 2.2936317920684814,
      "step": 9604
    },
    {
      "epoch": 0.9606,
      "grad_norm": 0.07706589996814728,
      "learning_rate": 0.00027241379310344825,
      "loss": 2.380758047103882,
      "step": 9606
    },
    {
      "epoch": 0.9608,
      "grad_norm": 0.08141545951366425,
      "learning_rate": 0.0002710344827586207,
      "loss": 2.388275384902954,
      "step": 9608
    },
    {
      "epoch": 0.961,
      "grad_norm": 0.08043265342712402,
      "learning_rate": 0.00026965517241379314,
      "loss": 2.250501871109009,
      "step": 9610
    },
    {
      "epoch": 0.9612,
      "grad_norm": 0.07826190441846848,
      "learning_rate": 0.00026827586206896553,
      "loss": 2.4795708656311035,
      "step": 9612
    },
    {
      "epoch": 0.9614,
      "grad_norm": 0.08019635081291199,
      "learning_rate": 0.000266896551724138,
      "loss": 2.3867828845977783,
      "step": 9614
    },
    {
      "epoch": 0.9616,
      "grad_norm": 0.08009565621614456,
      "learning_rate": 0.0002655172413793103,
      "loss": 2.3967769145965576,
      "step": 9616
    },
    {
      "epoch": 0.9618,
      "grad_norm": 0.0827154740691185,
      "learning_rate": 0.00026413793103448275,
      "loss": 2.2890207767486572,
      "step": 9618
    },
    {
      "epoch": 0.962,
      "grad_norm": 0.08196484297513962,
      "learning_rate": 0.00026275862068965514,
      "loss": 2.4252684116363525,
      "step": 9620
    },
    {
      "epoch": 0.9622,
      "grad_norm": 0.07716897130012512,
      "learning_rate": 0.0002613793103448276,
      "loss": 2.3991200923919678,
      "step": 9622
    },
    {
      "epoch": 0.9624,
      "grad_norm": 0.08279022574424744,
      "learning_rate": 0.00026000000000000003,
      "loss": 2.2637791633605957,
      "step": 9624
    },
    {
      "epoch": 0.9626,
      "grad_norm": 0.08642256259918213,
      "learning_rate": 0.0002586206896551724,
      "loss": 2.3897435665130615,
      "step": 9626
    },
    {
      "epoch": 0.9628,
      "grad_norm": 0.08063513040542603,
      "learning_rate": 0.00025724137931034486,
      "loss": 2.4235806465148926,
      "step": 9628
    },
    {
      "epoch": 0.963,
      "grad_norm": 0.081954725086689,
      "learning_rate": 0.0002558620689655172,
      "loss": 2.3142714500427246,
      "step": 9630
    },
    {
      "epoch": 0.9632,
      "grad_norm": 0.08258011937141418,
      "learning_rate": 0.00025448275862068964,
      "loss": 2.3842594623565674,
      "step": 9632
    },
    {
      "epoch": 0.9634,
      "grad_norm": 0.07700935006141663,
      "learning_rate": 0.0002531034482758621,
      "loss": 2.366304874420166,
      "step": 9634
    },
    {
      "epoch": 0.9636,
      "grad_norm": 0.08401020616292953,
      "learning_rate": 0.0002517241379310345,
      "loss": 2.3454675674438477,
      "step": 9636
    },
    {
      "epoch": 0.9638,
      "grad_norm": 0.08533082902431488,
      "learning_rate": 0.0002503448275862069,
      "loss": 2.401991128921509,
      "step": 9638
    },
    {
      "epoch": 0.964,
      "grad_norm": 0.08013403415679932,
      "learning_rate": 0.0002489655172413793,
      "loss": 2.2883336544036865,
      "step": 9640
    },
    {
      "epoch": 0.9642,
      "grad_norm": 0.08050772547721863,
      "learning_rate": 0.00024758620689655175,
      "loss": 2.3334853649139404,
      "step": 9642
    },
    {
      "epoch": 0.9644,
      "grad_norm": 0.08253761380910873,
      "learning_rate": 0.00024620689655172414,
      "loss": 2.4801337718963623,
      "step": 9644
    },
    {
      "epoch": 0.9646,
      "grad_norm": 0.08026421815156937,
      "learning_rate": 0.00024482758620689653,
      "loss": 2.415254592895508,
      "step": 9646
    },
    {
      "epoch": 0.9648,
      "grad_norm": 0.083962582051754,
      "learning_rate": 0.00024344827586206898,
      "loss": 2.323978900909424,
      "step": 9648
    },
    {
      "epoch": 0.965,
      "grad_norm": 0.092951200902462,
      "learning_rate": 0.0002420689655172414,
      "loss": 2.3495569229125977,
      "step": 9650
    },
    {
      "epoch": 0.9652,
      "grad_norm": 0.07913967221975327,
      "learning_rate": 0.00024068965517241378,
      "loss": 2.392270088195801,
      "step": 9652
    },
    {
      "epoch": 0.9654,
      "grad_norm": 0.07916171848773956,
      "learning_rate": 0.0002393103448275862,
      "loss": 2.2950029373168945,
      "step": 9654
    },
    {
      "epoch": 0.9656,
      "grad_norm": 0.08488109707832336,
      "learning_rate": 0.00023793103448275862,
      "loss": 2.357252597808838,
      "step": 9656
    },
    {
      "epoch": 0.9658,
      "grad_norm": 0.08531811833381653,
      "learning_rate": 0.00023655172413793106,
      "loss": 2.3438405990600586,
      "step": 9658
    },
    {
      "epoch": 0.966,
      "grad_norm": 0.08019422739744186,
      "learning_rate": 0.00023517241379310345,
      "loss": 2.404076337814331,
      "step": 9660
    },
    {
      "epoch": 0.9662,
      "grad_norm": 0.08106068521738052,
      "learning_rate": 0.00023379310344827587,
      "loss": 2.2889742851257324,
      "step": 9662
    },
    {
      "epoch": 0.9664,
      "grad_norm": 0.08003227412700653,
      "learning_rate": 0.00023241379310344828,
      "loss": 2.2877399921417236,
      "step": 9664
    },
    {
      "epoch": 0.9666,
      "grad_norm": 0.08504921197891235,
      "learning_rate": 0.00023103448275862067,
      "loss": 2.3687891960144043,
      "step": 9666
    },
    {
      "epoch": 0.9668,
      "grad_norm": 0.079212985932827,
      "learning_rate": 0.00022965517241379312,
      "loss": 2.3161885738372803,
      "step": 9668
    },
    {
      "epoch": 0.967,
      "grad_norm": 0.0808134451508522,
      "learning_rate": 0.00022827586206896553,
      "loss": 2.346780300140381,
      "step": 9670
    },
    {
      "epoch": 0.9672,
      "grad_norm": 0.08100850880146027,
      "learning_rate": 0.00022689655172413792,
      "loss": 2.358175754547119,
      "step": 9672
    },
    {
      "epoch": 0.9674,
      "grad_norm": 0.07927393168210983,
      "learning_rate": 0.00022551724137931034,
      "loss": 2.3466498851776123,
      "step": 9674
    },
    {
      "epoch": 0.9676,
      "grad_norm": 0.08110120892524719,
      "learning_rate": 0.00022413793103448276,
      "loss": 2.396928310394287,
      "step": 9676
    },
    {
      "epoch": 0.9678,
      "grad_norm": 0.08214782178401947,
      "learning_rate": 0.0002227586206896552,
      "loss": 2.3001973628997803,
      "step": 9678
    },
    {
      "epoch": 0.968,
      "grad_norm": 0.08595370501279831,
      "learning_rate": 0.0002213793103448276,
      "loss": 2.2798445224761963,
      "step": 9680
    },
    {
      "epoch": 0.9682,
      "grad_norm": 0.07984452694654465,
      "learning_rate": 0.00022,
      "loss": 2.308748483657837,
      "step": 9682
    },
    {
      "epoch": 0.9684,
      "grad_norm": 0.0779782235622406,
      "learning_rate": 0.00021862068965517242,
      "loss": 2.298013925552368,
      "step": 9684
    },
    {
      "epoch": 0.9686,
      "grad_norm": 0.08283451199531555,
      "learning_rate": 0.00021724137931034481,
      "loss": 2.3957269191741943,
      "step": 9686
    },
    {
      "epoch": 0.9688,
      "grad_norm": 0.07935620844364166,
      "learning_rate": 0.00021586206896551723,
      "loss": 2.3338327407836914,
      "step": 9688
    },
    {
      "epoch": 0.969,
      "grad_norm": 0.08624466508626938,
      "learning_rate": 0.00021448275862068967,
      "loss": 2.3156580924987793,
      "step": 9690
    },
    {
      "epoch": 0.9692,
      "grad_norm": 0.08566388487815857,
      "learning_rate": 0.00021310344827586206,
      "loss": 2.3623924255371094,
      "step": 9692
    },
    {
      "epoch": 0.9694,
      "grad_norm": 0.08012395352125168,
      "learning_rate": 0.00021172413793103448,
      "loss": 2.3309061527252197,
      "step": 9694
    },
    {
      "epoch": 0.9696,
      "grad_norm": 0.0789051204919815,
      "learning_rate": 0.0002103448275862069,
      "loss": 2.3390567302703857,
      "step": 9696
    },
    {
      "epoch": 0.9698,
      "grad_norm": 0.08181953430175781,
      "learning_rate": 0.00020896551724137931,
      "loss": 2.3824808597564697,
      "step": 9698
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.08486181497573853,
      "learning_rate": 0.00020758620689655173,
      "loss": 2.2390336990356445,
      "step": 9700
    },
    {
      "epoch": 0.9702,
      "grad_norm": 0.07844749838113785,
      "learning_rate": 0.00020620689655172415,
      "loss": 2.360351324081421,
      "step": 9702
    },
    {
      "epoch": 0.9704,
      "grad_norm": 0.08251507580280304,
      "learning_rate": 0.00020482758620689656,
      "loss": 2.3300139904022217,
      "step": 9704
    },
    {
      "epoch": 0.9706,
      "grad_norm": 0.0957828164100647,
      "learning_rate": 0.00020344827586206895,
      "loss": 2.3367233276367188,
      "step": 9706
    },
    {
      "epoch": 0.9708,
      "grad_norm": 0.07923724502325058,
      "learning_rate": 0.00020206896551724137,
      "loss": 2.3585917949676514,
      "step": 9708
    },
    {
      "epoch": 0.971,
      "grad_norm": 0.08480320870876312,
      "learning_rate": 0.00020068965517241381,
      "loss": 2.436326503753662,
      "step": 9710
    },
    {
      "epoch": 0.9712,
      "grad_norm": 0.07901017367839813,
      "learning_rate": 0.0001993103448275862,
      "loss": 2.3223609924316406,
      "step": 9712
    },
    {
      "epoch": 0.9714,
      "grad_norm": 0.08036114275455475,
      "learning_rate": 0.00019793103448275862,
      "loss": 2.3130440711975098,
      "step": 9714
    },
    {
      "epoch": 0.9716,
      "grad_norm": 0.08195436000823975,
      "learning_rate": 0.00019655172413793104,
      "loss": 2.26741361618042,
      "step": 9716
    },
    {
      "epoch": 0.9718,
      "grad_norm": 0.08250656723976135,
      "learning_rate": 0.00019517241379310345,
      "loss": 2.3509931564331055,
      "step": 9718
    },
    {
      "epoch": 0.972,
      "grad_norm": 0.07781665027141571,
      "learning_rate": 0.00019379310344827587,
      "loss": 2.329132080078125,
      "step": 9720
    },
    {
      "epoch": 0.9722,
      "grad_norm": 0.08036928623914719,
      "learning_rate": 0.0001924137931034483,
      "loss": 2.2804229259490967,
      "step": 9722
    },
    {
      "epoch": 0.9724,
      "grad_norm": 0.08182676136493683,
      "learning_rate": 0.0001910344827586207,
      "loss": 2.374272108078003,
      "step": 9724
    },
    {
      "epoch": 0.9726,
      "grad_norm": 0.07762088626623154,
      "learning_rate": 0.0001896551724137931,
      "loss": 2.3786609172821045,
      "step": 9726
    },
    {
      "epoch": 0.9728,
      "grad_norm": 0.08010619878768921,
      "learning_rate": 0.0001882758620689655,
      "loss": 2.305896759033203,
      "step": 9728
    },
    {
      "epoch": 0.973,
      "grad_norm": 0.08205893635749817,
      "learning_rate": 0.00018689655172413795,
      "loss": 2.467349052429199,
      "step": 9730
    },
    {
      "epoch": 0.9732,
      "grad_norm": 0.07573944330215454,
      "learning_rate": 0.00018551724137931034,
      "loss": 2.3547019958496094,
      "step": 9732
    },
    {
      "epoch": 0.9734,
      "grad_norm": 0.07767363637685776,
      "learning_rate": 0.00018413793103448276,
      "loss": 2.381455659866333,
      "step": 9734
    },
    {
      "epoch": 0.9736,
      "grad_norm": 0.07611992210149765,
      "learning_rate": 0.00018275862068965518,
      "loss": 2.3231101036071777,
      "step": 9736
    },
    {
      "epoch": 0.9738,
      "grad_norm": 0.07825005799531937,
      "learning_rate": 0.0001813793103448276,
      "loss": 2.3738315105438232,
      "step": 9738
    },
    {
      "epoch": 0.974,
      "grad_norm": 0.07750774919986725,
      "learning_rate": 0.00017999999999999998,
      "loss": 2.3138506412506104,
      "step": 9740
    },
    {
      "epoch": 0.9742,
      "grad_norm": 0.08378524333238602,
      "learning_rate": 0.00017862068965517243,
      "loss": 2.434779644012451,
      "step": 9742
    },
    {
      "epoch": 0.9744,
      "grad_norm": 0.08465835452079773,
      "learning_rate": 0.00017724137931034484,
      "loss": 2.397526264190674,
      "step": 9744
    },
    {
      "epoch": 0.9746,
      "grad_norm": 0.08439000695943832,
      "learning_rate": 0.00017586206896551723,
      "loss": 2.3807663917541504,
      "step": 9746
    },
    {
      "epoch": 0.9748,
      "grad_norm": 0.0789857879281044,
      "learning_rate": 0.00017448275862068965,
      "loss": 2.351566791534424,
      "step": 9748
    },
    {
      "epoch": 0.975,
      "grad_norm": 0.08107300102710724,
      "learning_rate": 0.00017310344827586207,
      "loss": 2.288994073867798,
      "step": 9750
    },
    {
      "epoch": 0.9752,
      "grad_norm": 0.08648376911878586,
      "learning_rate": 0.00017172413793103448,
      "loss": 2.3549280166625977,
      "step": 9752
    },
    {
      "epoch": 0.9754,
      "grad_norm": 0.07840137928724289,
      "learning_rate": 0.0001703448275862069,
      "loss": 2.3397138118743896,
      "step": 9754
    },
    {
      "epoch": 0.9756,
      "grad_norm": 0.08030110597610474,
      "learning_rate": 0.00016896551724137932,
      "loss": 2.289769411087036,
      "step": 9756
    },
    {
      "epoch": 0.9758,
      "grad_norm": 0.08935897052288055,
      "learning_rate": 0.00016758620689655173,
      "loss": 2.34385085105896,
      "step": 9758
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.08107522130012512,
      "learning_rate": 0.00016620689655172412,
      "loss": 2.2685561180114746,
      "step": 9760
    },
    {
      "epoch": 0.9762,
      "grad_norm": 0.08086903393268585,
      "learning_rate": 0.00016482758620689657,
      "loss": 2.252805471420288,
      "step": 9762
    },
    {
      "epoch": 0.9764,
      "grad_norm": 0.0815119743347168,
      "learning_rate": 0.00016344827586206898,
      "loss": 2.361950397491455,
      "step": 9764
    },
    {
      "epoch": 0.9766,
      "grad_norm": 0.08220860362052917,
      "learning_rate": 0.00016206896551724137,
      "loss": 2.3041181564331055,
      "step": 9766
    },
    {
      "epoch": 0.9768,
      "grad_norm": 0.07794570922851562,
      "learning_rate": 0.0001606896551724138,
      "loss": 2.452603340148926,
      "step": 9768
    },
    {
      "epoch": 0.977,
      "grad_norm": 0.08432630449533463,
      "learning_rate": 0.0001593103448275862,
      "loss": 2.3263490200042725,
      "step": 9770
    },
    {
      "epoch": 0.9772,
      "grad_norm": 0.08050213009119034,
      "learning_rate": 0.00015793103448275862,
      "loss": 2.288043260574341,
      "step": 9772
    },
    {
      "epoch": 0.9774,
      "grad_norm": 0.08157087117433548,
      "learning_rate": 0.00015655172413793104,
      "loss": 2.35184383392334,
      "step": 9774
    },
    {
      "epoch": 0.9776,
      "grad_norm": 0.08214464038610458,
      "learning_rate": 0.00015517241379310346,
      "loss": 2.323279619216919,
      "step": 9776
    },
    {
      "epoch": 0.9778,
      "grad_norm": 0.0825587660074234,
      "learning_rate": 0.00015379310344827587,
      "loss": 2.370523452758789,
      "step": 9778
    },
    {
      "epoch": 0.978,
      "grad_norm": 0.07782994955778122,
      "learning_rate": 0.00015241379310344826,
      "loss": 2.3112053871154785,
      "step": 9780
    },
    {
      "epoch": 0.9782,
      "grad_norm": 0.07968544214963913,
      "learning_rate": 0.0001510344827586207,
      "loss": 2.360333204269409,
      "step": 9782
    },
    {
      "epoch": 0.9784,
      "grad_norm": 0.0840863436460495,
      "learning_rate": 0.00014965517241379312,
      "loss": 2.3112878799438477,
      "step": 9784
    },
    {
      "epoch": 0.9786,
      "grad_norm": 0.0784270167350769,
      "learning_rate": 0.00014827586206896551,
      "loss": 2.3632969856262207,
      "step": 9786
    },
    {
      "epoch": 0.9788,
      "grad_norm": 0.07928818464279175,
      "learning_rate": 0.00014689655172413793,
      "loss": 2.381436824798584,
      "step": 9788
    },
    {
      "epoch": 0.979,
      "grad_norm": 0.08057167381048203,
      "learning_rate": 0.00014551724137931035,
      "loss": 2.363534450531006,
      "step": 9790
    },
    {
      "epoch": 0.9792,
      "grad_norm": 0.0842352956533432,
      "learning_rate": 0.00014413793103448274,
      "loss": 2.3742926120758057,
      "step": 9792
    },
    {
      "epoch": 0.9794,
      "grad_norm": 0.08135809749364853,
      "learning_rate": 0.00014275862068965518,
      "loss": 2.3600518703460693,
      "step": 9794
    },
    {
      "epoch": 0.9796,
      "grad_norm": 0.07797564566135406,
      "learning_rate": 0.0001413793103448276,
      "loss": 2.390300989151001,
      "step": 9796
    },
    {
      "epoch": 0.9798,
      "grad_norm": 0.0769975483417511,
      "learning_rate": 0.00014000000000000001,
      "loss": 2.3194234371185303,
      "step": 9798
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.08442690968513489,
      "learning_rate": 0.0001386206896551724,
      "loss": 2.307832956314087,
      "step": 9800
    },
    {
      "epoch": 0.9802,
      "grad_norm": 0.0796445906162262,
      "learning_rate": 0.00013724137931034482,
      "loss": 2.344453811645508,
      "step": 9802
    },
    {
      "epoch": 0.9804,
      "grad_norm": 0.07709208875894547,
      "learning_rate": 0.00013586206896551726,
      "loss": 2.2736620903015137,
      "step": 9804
    },
    {
      "epoch": 0.9806,
      "grad_norm": 0.07947950065135956,
      "learning_rate": 0.00013448275862068965,
      "loss": 2.3601043224334717,
      "step": 9806
    },
    {
      "epoch": 0.9808,
      "grad_norm": 0.07871530205011368,
      "learning_rate": 0.00013310344827586207,
      "loss": 2.3705666065216064,
      "step": 9808
    },
    {
      "epoch": 0.981,
      "grad_norm": 0.08042515814304352,
      "learning_rate": 0.0001317241379310345,
      "loss": 2.358426809310913,
      "step": 9810
    },
    {
      "epoch": 0.9812,
      "grad_norm": 0.07887619733810425,
      "learning_rate": 0.00013034482758620688,
      "loss": 2.3458075523376465,
      "step": 9812
    },
    {
      "epoch": 0.9814,
      "grad_norm": 0.08038248866796494,
      "learning_rate": 0.00012896551724137932,
      "loss": 2.372666835784912,
      "step": 9814
    },
    {
      "epoch": 0.9816,
      "grad_norm": 0.0782962441444397,
      "learning_rate": 0.00012758620689655174,
      "loss": 2.3991291522979736,
      "step": 9816
    },
    {
      "epoch": 0.9818,
      "grad_norm": 0.07555270940065384,
      "learning_rate": 0.00012620689655172415,
      "loss": 2.3389344215393066,
      "step": 9818
    },
    {
      "epoch": 0.982,
      "grad_norm": 0.08380011469125748,
      "learning_rate": 0.00012482758620689654,
      "loss": 2.3874714374542236,
      "step": 9820
    },
    {
      "epoch": 0.9822,
      "grad_norm": 0.08302925527095795,
      "learning_rate": 0.00012344827586206896,
      "loss": 2.3247525691986084,
      "step": 9822
    },
    {
      "epoch": 0.9824,
      "grad_norm": 0.08192239701747894,
      "learning_rate": 0.00012206896551724138,
      "loss": 2.3466873168945312,
      "step": 9824
    },
    {
      "epoch": 0.9826,
      "grad_norm": 0.08537617325782776,
      "learning_rate": 0.0001206896551724138,
      "loss": 2.3764662742614746,
      "step": 9826
    },
    {
      "epoch": 0.9828,
      "grad_norm": 0.07756063342094421,
      "learning_rate": 0.00011931034482758621,
      "loss": 2.347201108932495,
      "step": 9828
    },
    {
      "epoch": 0.983,
      "grad_norm": 0.07540380954742432,
      "learning_rate": 0.00011793103448275861,
      "loss": 2.390536069869995,
      "step": 9830
    },
    {
      "epoch": 0.9832,
      "grad_norm": 0.07818792015314102,
      "learning_rate": 0.00011655172413793104,
      "loss": 2.338031530380249,
      "step": 9832
    },
    {
      "epoch": 0.9834,
      "grad_norm": 0.07933175563812256,
      "learning_rate": 0.00011517241379310345,
      "loss": 2.322932243347168,
      "step": 9834
    },
    {
      "epoch": 0.9836,
      "grad_norm": 0.07875898480415344,
      "learning_rate": 0.00011379310344827586,
      "loss": 2.347003221511841,
      "step": 9836
    },
    {
      "epoch": 0.9838,
      "grad_norm": 0.07851476222276688,
      "learning_rate": 0.00011241379310344828,
      "loss": 2.3656961917877197,
      "step": 9838
    },
    {
      "epoch": 0.984,
      "grad_norm": 0.08181953430175781,
      "learning_rate": 0.00011103448275862068,
      "loss": 2.2634079456329346,
      "step": 9840
    },
    {
      "epoch": 0.9842,
      "grad_norm": 0.08258124440908432,
      "learning_rate": 0.00010965517241379311,
      "loss": 2.3006749153137207,
      "step": 9842
    },
    {
      "epoch": 0.9844,
      "grad_norm": 0.08079875260591507,
      "learning_rate": 0.00010827586206896552,
      "loss": 2.379978656768799,
      "step": 9844
    },
    {
      "epoch": 0.9846,
      "grad_norm": 0.07874976098537445,
      "learning_rate": 0.00010689655172413793,
      "loss": 2.4396228790283203,
      "step": 9846
    },
    {
      "epoch": 0.9848,
      "grad_norm": 0.0800589993596077,
      "learning_rate": 0.00010551724137931035,
      "loss": 2.321579933166504,
      "step": 9848
    },
    {
      "epoch": 0.985,
      "grad_norm": 0.07894232124090195,
      "learning_rate": 0.00010413793103448275,
      "loss": 2.383321762084961,
      "step": 9850
    },
    {
      "epoch": 0.9852,
      "grad_norm": 0.08333214372396469,
      "learning_rate": 0.00010275862068965517,
      "loss": 2.3059754371643066,
      "step": 9852
    },
    {
      "epoch": 0.9854,
      "grad_norm": 0.08336926996707916,
      "learning_rate": 0.00010137931034482759,
      "loss": 2.3123271465301514,
      "step": 9854
    },
    {
      "epoch": 0.9856,
      "grad_norm": 0.08011731505393982,
      "learning_rate": 0.0001,
      "loss": 2.3773326873779297,
      "step": 9856
    },
    {
      "epoch": 0.9858,
      "grad_norm": 0.08305983245372772,
      "learning_rate": 9.862068965517242e-05,
      "loss": 2.2877438068389893,
      "step": 9858
    },
    {
      "epoch": 0.986,
      "grad_norm": 0.08173101395368576,
      "learning_rate": 9.724137931034482e-05,
      "loss": 2.368527412414551,
      "step": 9860
    },
    {
      "epoch": 0.9862,
      "grad_norm": 0.0808301642537117,
      "learning_rate": 9.586206896551724e-05,
      "loss": 2.317934989929199,
      "step": 9862
    },
    {
      "epoch": 0.9864,
      "grad_norm": 0.08173970133066177,
      "learning_rate": 9.448275862068966e-05,
      "loss": 2.41127610206604,
      "step": 9864
    },
    {
      "epoch": 0.9866,
      "grad_norm": 0.09866008907556534,
      "learning_rate": 9.310344827586207e-05,
      "loss": 2.3534419536590576,
      "step": 9866
    },
    {
      "epoch": 0.9868,
      "grad_norm": 0.07849210500717163,
      "learning_rate": 9.172413793103449e-05,
      "loss": 2.310269594192505,
      "step": 9868
    },
    {
      "epoch": 0.987,
      "grad_norm": 0.07629062980413437,
      "learning_rate": 9.03448275862069e-05,
      "loss": 2.2772374153137207,
      "step": 9870
    },
    {
      "epoch": 0.9872,
      "grad_norm": 0.08273548632860184,
      "learning_rate": 8.896551724137931e-05,
      "loss": 2.325639247894287,
      "step": 9872
    },
    {
      "epoch": 0.9874,
      "grad_norm": 0.07988311350345612,
      "learning_rate": 8.758620689655173e-05,
      "loss": 2.354635000228882,
      "step": 9874
    },
    {
      "epoch": 0.9876,
      "grad_norm": 0.07666224241256714,
      "learning_rate": 8.620689655172414e-05,
      "loss": 2.4293415546417236,
      "step": 9876
    },
    {
      "epoch": 0.9878,
      "grad_norm": 0.0793016105890274,
      "learning_rate": 8.482758620689655e-05,
      "loss": 2.3611152172088623,
      "step": 9878
    },
    {
      "epoch": 0.988,
      "grad_norm": 0.08695759624242783,
      "learning_rate": 8.344827586206896e-05,
      "loss": 2.3841450214385986,
      "step": 9880
    },
    {
      "epoch": 0.9882,
      "grad_norm": 0.08243729174137115,
      "learning_rate": 8.206896551724138e-05,
      "loss": 2.4000353813171387,
      "step": 9882
    },
    {
      "epoch": 0.9884,
      "grad_norm": 0.07873734831809998,
      "learning_rate": 8.06896551724138e-05,
      "loss": 2.3718783855438232,
      "step": 9884
    },
    {
      "epoch": 0.9886,
      "grad_norm": 0.07970809936523438,
      "learning_rate": 7.931034482758621e-05,
      "loss": 2.3848721981048584,
      "step": 9886
    },
    {
      "epoch": 0.9888,
      "grad_norm": 0.08186116069555283,
      "learning_rate": 7.793103448275862e-05,
      "loss": 2.394646167755127,
      "step": 9888
    },
    {
      "epoch": 0.989,
      "grad_norm": 0.08045443147420883,
      "learning_rate": 7.655172413793103e-05,
      "loss": 2.3348228931427,
      "step": 9890
    },
    {
      "epoch": 0.9892,
      "grad_norm": 0.08206921815872192,
      "learning_rate": 7.517241379310345e-05,
      "loss": 2.3270657062530518,
      "step": 9892
    },
    {
      "epoch": 0.9894,
      "grad_norm": 0.0818614810705185,
      "learning_rate": 7.379310344827585e-05,
      "loss": 2.3717634677886963,
      "step": 9894
    },
    {
      "epoch": 0.9896,
      "grad_norm": 0.07991863787174225,
      "learning_rate": 7.241379310344828e-05,
      "loss": 2.3548693656921387,
      "step": 9896
    },
    {
      "epoch": 0.9898,
      "grad_norm": 0.08408034592866898,
      "learning_rate": 7.103448275862069e-05,
      "loss": 2.2981839179992676,
      "step": 9898
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.07852338254451752,
      "learning_rate": 6.96551724137931e-05,
      "loss": 2.3758533000946045,
      "step": 9900
    },
    {
      "epoch": 0.9902,
      "grad_norm": 0.08794118463993073,
      "learning_rate": 6.827586206896552e-05,
      "loss": 2.418628454208374,
      "step": 9902
    },
    {
      "epoch": 0.9904,
      "grad_norm": 0.0800185352563858,
      "learning_rate": 6.689655172413792e-05,
      "loss": 2.385288715362549,
      "step": 9904
    },
    {
      "epoch": 0.9906,
      "grad_norm": 0.08253008127212524,
      "learning_rate": 6.551724137931035e-05,
      "loss": 2.4523580074310303,
      "step": 9906
    },
    {
      "epoch": 0.9908,
      "grad_norm": 0.08206792175769806,
      "learning_rate": 6.413793103448276e-05,
      "loss": 2.4083080291748047,
      "step": 9908
    },
    {
      "epoch": 0.991,
      "grad_norm": 0.08537327498197556,
      "learning_rate": 6.275862068965517e-05,
      "loss": 2.3637688159942627,
      "step": 9910
    },
    {
      "epoch": 0.9912,
      "grad_norm": 0.08164107799530029,
      "learning_rate": 6.137931034482759e-05,
      "loss": 2.381415605545044,
      "step": 9912
    },
    {
      "epoch": 0.9914,
      "grad_norm": 0.08063610643148422,
      "learning_rate": 6e-05,
      "loss": 2.3693549633026123,
      "step": 9914
    },
    {
      "epoch": 0.9916,
      "grad_norm": 0.08052631467580795,
      "learning_rate": 5.862068965517241e-05,
      "loss": 2.3752665519714355,
      "step": 9916
    },
    {
      "epoch": 0.9918,
      "grad_norm": 0.08687854558229446,
      "learning_rate": 5.724137931034483e-05,
      "loss": 2.357262134552002,
      "step": 9918
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.08150909096002579,
      "learning_rate": 5.5862068965517245e-05,
      "loss": 2.3708667755126953,
      "step": 9920
    },
    {
      "epoch": 0.9922,
      "grad_norm": 0.07943639904260635,
      "learning_rate": 5.448275862068966e-05,
      "loss": 2.4035654067993164,
      "step": 9922
    },
    {
      "epoch": 0.9924,
      "grad_norm": 0.0803312435746193,
      "learning_rate": 5.3103448275862065e-05,
      "loss": 2.403261184692383,
      "step": 9924
    },
    {
      "epoch": 0.9926,
      "grad_norm": 0.07658237963914871,
      "learning_rate": 5.172413793103448e-05,
      "loss": 2.3510024547576904,
      "step": 9926
    },
    {
      "epoch": 0.9928,
      "grad_norm": 0.08047880977392197,
      "learning_rate": 5.03448275862069e-05,
      "loss": 2.3390095233917236,
      "step": 9928
    },
    {
      "epoch": 0.993,
      "grad_norm": 0.08098559826612473,
      "learning_rate": 4.8965517241379315e-05,
      "loss": 2.3504722118377686,
      "step": 9930
    },
    {
      "epoch": 0.9932,
      "grad_norm": 0.08468184620141983,
      "learning_rate": 4.7586206896551725e-05,
      "loss": 2.41129207611084,
      "step": 9932
    },
    {
      "epoch": 0.9934,
      "grad_norm": 0.0805194303393364,
      "learning_rate": 4.6206896551724135e-05,
      "loss": 2.379081964492798,
      "step": 9934
    },
    {
      "epoch": 0.9936,
      "grad_norm": 0.08066009730100632,
      "learning_rate": 4.482758620689655e-05,
      "loss": 2.416440963745117,
      "step": 9936
    },
    {
      "epoch": 0.9938,
      "grad_norm": 0.08129583299160004,
      "learning_rate": 4.344827586206897e-05,
      "loss": 2.423678159713745,
      "step": 9938
    },
    {
      "epoch": 0.994,
      "grad_norm": 0.08004023134708405,
      "learning_rate": 4.2068965517241385e-05,
      "loss": 2.3439249992370605,
      "step": 9940
    },
    {
      "epoch": 0.9942,
      "grad_norm": 0.08346652239561081,
      "learning_rate": 4.068965517241379e-05,
      "loss": 2.3347222805023193,
      "step": 9942
    },
    {
      "epoch": 0.9944,
      "grad_norm": 0.07767464965581894,
      "learning_rate": 3.9310344827586205e-05,
      "loss": 2.3127822875976562,
      "step": 9944
    },
    {
      "epoch": 0.9946,
      "grad_norm": 0.08130376785993576,
      "learning_rate": 3.793103448275862e-05,
      "loss": 2.285306215286255,
      "step": 9946
    },
    {
      "epoch": 0.9948,
      "grad_norm": 0.07888536900281906,
      "learning_rate": 3.655172413793104e-05,
      "loss": 2.34043288230896,
      "step": 9948
    },
    {
      "epoch": 0.995,
      "grad_norm": 0.08326157182455063,
      "learning_rate": 3.517241379310345e-05,
      "loss": 2.350534439086914,
      "step": 9950
    },
    {
      "epoch": 0.9952,
      "grad_norm": 0.0797511488199234,
      "learning_rate": 3.379310344827586e-05,
      "loss": 2.4353268146514893,
      "step": 9952
    },
    {
      "epoch": 0.9954,
      "grad_norm": 0.08057495951652527,
      "learning_rate": 3.2413793103448275e-05,
      "loss": 2.354050636291504,
      "step": 9954
    },
    {
      "epoch": 0.9956,
      "grad_norm": 0.0811692327260971,
      "learning_rate": 3.103448275862069e-05,
      "loss": 2.3672752380371094,
      "step": 9956
    },
    {
      "epoch": 0.9958,
      "grad_norm": 0.0787394791841507,
      "learning_rate": 2.9655172413793105e-05,
      "loss": 2.3535141944885254,
      "step": 9958
    },
    {
      "epoch": 0.996,
      "grad_norm": 0.08226253092288971,
      "learning_rate": 2.8275862068965518e-05,
      "loss": 2.3988888263702393,
      "step": 9960
    },
    {
      "epoch": 0.9962,
      "grad_norm": 0.07919766753911972,
      "learning_rate": 2.689655172413793e-05,
      "loss": 2.303706407546997,
      "step": 9962
    },
    {
      "epoch": 0.9964,
      "grad_norm": 0.08856409043073654,
      "learning_rate": 2.551724137931035e-05,
      "loss": 2.3884658813476562,
      "step": 9964
    },
    {
      "epoch": 0.9966,
      "grad_norm": 0.08525054156780243,
      "learning_rate": 2.4137931034482758e-05,
      "loss": 2.3274543285369873,
      "step": 9966
    },
    {
      "epoch": 0.9968,
      "grad_norm": 0.08125246316194534,
      "learning_rate": 2.2758620689655175e-05,
      "loss": 2.4319777488708496,
      "step": 9968
    },
    {
      "epoch": 0.997,
      "grad_norm": 0.07981839776039124,
      "learning_rate": 2.1379310344827585e-05,
      "loss": 2.3553271293640137,
      "step": 9970
    },
    {
      "epoch": 0.9972,
      "grad_norm": 0.08588403463363647,
      "learning_rate": 2e-05,
      "loss": 2.3589415550231934,
      "step": 9972
    },
    {
      "epoch": 0.9974,
      "grad_norm": 0.07860904932022095,
      "learning_rate": 1.862068965517241e-05,
      "loss": 2.3531320095062256,
      "step": 9974
    },
    {
      "epoch": 0.9976,
      "grad_norm": 0.07722477614879608,
      "learning_rate": 1.7241379310344828e-05,
      "loss": 2.3386991024017334,
      "step": 9976
    },
    {
      "epoch": 0.9978,
      "grad_norm": 0.07741781324148178,
      "learning_rate": 1.586206896551724e-05,
      "loss": 2.331669569015503,
      "step": 9978
    },
    {
      "epoch": 0.998,
      "grad_norm": 0.0779128149151802,
      "learning_rate": 1.4482758620689657e-05,
      "loss": 2.3889827728271484,
      "step": 9980
    },
    {
      "epoch": 0.9982,
      "grad_norm": 0.0798075944185257,
      "learning_rate": 1.310344827586207e-05,
      "loss": 2.431318521499634,
      "step": 9982
    },
    {
      "epoch": 0.9984,
      "grad_norm": 0.08234637975692749,
      "learning_rate": 1.1724137931034483e-05,
      "loss": 2.415722131729126,
      "step": 9984
    },
    {
      "epoch": 0.9986,
      "grad_norm": 0.08344531059265137,
      "learning_rate": 1.0344827586206897e-05,
      "loss": 2.3518848419189453,
      "step": 9986
    },
    {
      "epoch": 0.9988,
      "grad_norm": 0.0806950181722641,
      "learning_rate": 8.96551724137931e-06,
      "loss": 2.3289496898651123,
      "step": 9988
    },
    {
      "epoch": 0.999,
      "grad_norm": 0.08386988192796707,
      "learning_rate": 7.586206896551724e-06,
      "loss": 2.4109370708465576,
      "step": 9990
    },
    {
      "epoch": 0.9992,
      "grad_norm": 0.07844487577676773,
      "learning_rate": 6.2068965517241375e-06,
      "loss": 2.3820836544036865,
      "step": 9992
    },
    {
      "epoch": 0.9994,
      "grad_norm": 0.07977902889251709,
      "learning_rate": 4.8275862068965525e-06,
      "loss": 2.3160905838012695,
      "step": 9994
    },
    {
      "epoch": 0.9996,
      "grad_norm": 0.0817895382642746,
      "learning_rate": 3.4482758620689654e-06,
      "loss": 2.428175449371338,
      "step": 9996
    },
    {
      "epoch": 0.9998,
      "grad_norm": 0.08118752390146255,
      "learning_rate": 2.0689655172413796e-06,
      "loss": 2.39603590965271,
      "step": 9998
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.07649844139814377,
      "learning_rate": 6.89655172413793e-07,
      "loss": 2.37339448928833,
      "step": 10000
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.1258270740509033,
      "eval_runtime": 235.9816,
      "eval_samples_per_second": 2.119,
      "eval_steps_per_second": 0.212,
      "step": 10000
    }
  ],
  "logging_steps": 2,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1271215411517952e+19,
  "train_batch_size": 10,
  "trial_name": null,
  "trial_params": null
}