{
  "best_global_step": 32436,
  "best_metric": 0.47668150067329407,
  "best_model_checkpoint": "saves_multiple/lntuning/llama-3-8b-instruct/train_codealpacapy_789_1767686248/checkpoint-32436",
  "epoch": 20.0,
  "eval_steps": 1908,
  "global_step": 38160,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002620545073375262,
      "grad_norm": 1.1479791402816772,
      "learning_rate": 5.241090146750525e-08,
      "loss": 1.2396,
      "num_input_tokens_seen": 3168,
      "step": 5
    },
    {
      "epoch": 0.005241090146750524,
      "grad_norm": 1.8654550313949585,
      "learning_rate": 1.1792452830188679e-07,
      "loss": 1.4164,
      "num_input_tokens_seen": 6048,
      "step": 10
    },
    {
      "epoch": 0.007861635220125786,
      "grad_norm": 1.1817377805709839,
      "learning_rate": 1.8343815513626837e-07,
      "loss": 1.1676,
      "num_input_tokens_seen": 9024,
      "step": 15
    },
    {
      "epoch": 0.010482180293501049,
      "grad_norm": 1.30488121509552,
      "learning_rate": 2.489517819706499e-07,
      "loss": 1.077,
      "num_input_tokens_seen": 12224,
      "step": 20
    },
    {
      "epoch": 0.01310272536687631,
      "grad_norm": 1.328601360321045,
      "learning_rate": 3.144654088050315e-07,
      "loss": 1.4581,
      "num_input_tokens_seen": 15200,
      "step": 25
    },
    {
      "epoch": 0.015723270440251572,
      "grad_norm": 0.982725203037262,
      "learning_rate": 3.79979035639413e-07,
      "loss": 0.9944,
      "num_input_tokens_seen": 18784,
      "step": 30
    },
    {
      "epoch": 0.018343815513626835,
      "grad_norm": 1.4781028032302856,
      "learning_rate": 4.454926624737945e-07,
      "loss": 1.3855,
      "num_input_tokens_seen": 21792,
      "step": 35
    },
    {
      "epoch": 0.020964360587002098,
      "grad_norm": 1.915358543395996,
      "learning_rate": 5.110062893081761e-07,
      "loss": 1.0866,
      "num_input_tokens_seen": 24928,
      "step": 40
    },
    {
      "epoch": 0.02358490566037736,
      "grad_norm": 1.6765177249908447,
      "learning_rate": 5.765199161425577e-07,
      "loss": 1.1609,
      "num_input_tokens_seen": 27872,
      "step": 45
    },
    {
      "epoch": 0.02620545073375262,
      "grad_norm": 1.4349006414413452,
      "learning_rate": 6.420335429769392e-07,
      "loss": 1.2127,
      "num_input_tokens_seen": 30496,
      "step": 50
    },
    {
      "epoch": 0.028825995807127882,
      "grad_norm": 1.6217701435089111,
      "learning_rate": 7.075471698113208e-07,
      "loss": 1.0636,
      "num_input_tokens_seen": 33728,
      "step": 55
    },
    {
      "epoch": 0.031446540880503145,
      "grad_norm": 1.3211981058120728,
      "learning_rate": 7.730607966457024e-07,
      "loss": 1.1391,
      "num_input_tokens_seen": 37344,
      "step": 60
    },
    {
      "epoch": 0.034067085953878404,
      "grad_norm": 1.4012514352798462,
      "learning_rate": 8.38574423480084e-07,
      "loss": 1.3271,
      "num_input_tokens_seen": 40096,
      "step": 65
    },
    {
      "epoch": 0.03668763102725367,
      "grad_norm": 1.431108832359314,
      "learning_rate": 9.040880503144654e-07,
      "loss": 0.9915,
      "num_input_tokens_seen": 44736,
      "step": 70
    },
    {
      "epoch": 0.03930817610062893,
      "grad_norm": 1.0054680109024048,
      "learning_rate": 9.69601677148847e-07,
      "loss": 1.2877,
      "num_input_tokens_seen": 48896,
      "step": 75
    },
    {
      "epoch": 0.041928721174004195,
      "grad_norm": 1.6504813432693481,
      "learning_rate": 1.0351153039832286e-06,
      "loss": 1.427,
      "num_input_tokens_seen": 51744,
      "step": 80
    },
    {
      "epoch": 0.044549266247379454,
      "grad_norm": 0.9307228922843933,
      "learning_rate": 1.10062893081761e-06,
      "loss": 0.775,
      "num_input_tokens_seen": 58880,
      "step": 85
    },
    {
      "epoch": 0.04716981132075472,
      "grad_norm": 0.95051109790802,
      "learning_rate": 1.1661425576519918e-06,
      "loss": 1.2105,
      "num_input_tokens_seen": 61536,
      "step": 90
    },
    {
      "epoch": 0.04979035639412998,
      "grad_norm": 1.1340010166168213,
      "learning_rate": 1.2316561844863732e-06,
      "loss": 1.3035,
      "num_input_tokens_seen": 65248,
      "step": 95
    },
    {
      "epoch": 0.05241090146750524,
      "grad_norm": 1.221006155014038,
      "learning_rate": 1.2971698113207548e-06,
      "loss": 1.3976,
      "num_input_tokens_seen": 68096,
      "step": 100
    },
    {
      "epoch": 0.055031446540880505,
      "grad_norm": 1.5738166570663452,
      "learning_rate": 1.3626834381551364e-06,
      "loss": 1.2392,
      "num_input_tokens_seen": 70848,
      "step": 105
    },
    {
      "epoch": 0.057651991614255764,
      "grad_norm": 1.238604187965393,
      "learning_rate": 1.4281970649895177e-06,
      "loss": 1.1893,
      "num_input_tokens_seen": 73376,
      "step": 110
    },
    {
      "epoch": 0.06027253668763103,
      "grad_norm": 1.5418628454208374,
      "learning_rate": 1.4937106918238995e-06,
      "loss": 0.9749,
      "num_input_tokens_seen": 77184,
      "step": 115
    },
    {
      "epoch": 0.06289308176100629,
      "grad_norm": 2.180368661880493,
      "learning_rate": 1.559224318658281e-06,
      "loss": 1.4837,
      "num_input_tokens_seen": 79904,
      "step": 120
    },
    {
      "epoch": 0.06551362683438156,
      "grad_norm": 0.7668371796607971,
      "learning_rate": 1.6247379454926625e-06,
      "loss": 1.2966,
      "num_input_tokens_seen": 82784,
      "step": 125
    },
    {
      "epoch": 0.06813417190775681,
      "grad_norm": 1.0710119009017944,
      "learning_rate": 1.690251572327044e-06,
      "loss": 1.0977,
      "num_input_tokens_seen": 86112,
      "step": 130
    },
    {
      "epoch": 0.07075471698113207,
      "grad_norm": 1.2614489793777466,
      "learning_rate": 1.7557651991614257e-06,
      "loss": 1.1484,
      "num_input_tokens_seen": 89600,
      "step": 135
    },
    {
      "epoch": 0.07337526205450734,
      "grad_norm": 1.056548833847046,
      "learning_rate": 1.821278825995807e-06,
      "loss": 1.2153,
      "num_input_tokens_seen": 92928,
      "step": 140
    },
    {
      "epoch": 0.0759958071278826,
      "grad_norm": 1.8094673156738281,
      "learning_rate": 1.8867924528301887e-06,
      "loss": 1.2635,
      "num_input_tokens_seen": 96384,
      "step": 145
    },
    {
      "epoch": 0.07861635220125786,
      "grad_norm": 1.8480890989303589,
      "learning_rate": 1.9523060796645703e-06,
      "loss": 1.0724,
      "num_input_tokens_seen": 99104,
      "step": 150
    },
    {
      "epoch": 0.08123689727463312,
      "grad_norm": 0.9703503847122192,
      "learning_rate": 2.017819706498952e-06,
      "loss": 1.5446,
      "num_input_tokens_seen": 101920,
      "step": 155
    },
    {
      "epoch": 0.08385744234800839,
      "grad_norm": 1.167481541633606,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 1.146,
      "num_input_tokens_seen": 104800,
      "step": 160
    },
    {
      "epoch": 0.08647798742138364,
      "grad_norm": 1.4266393184661865,
      "learning_rate": 2.148846960167715e-06,
      "loss": 1.5634,
      "num_input_tokens_seen": 107168,
      "step": 165
    },
    {
      "epoch": 0.08909853249475891,
      "grad_norm": 0.9710317850112915,
      "learning_rate": 2.2143605870020966e-06,
      "loss": 1.0968,
      "num_input_tokens_seen": 110336,
      "step": 170
    },
    {
      "epoch": 0.09171907756813417,
      "grad_norm": 1.4122989177703857,
      "learning_rate": 2.279874213836478e-06,
      "loss": 1.0908,
      "num_input_tokens_seen": 113568,
      "step": 175
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 1.3630129098892212,
      "learning_rate": 2.34538784067086e-06,
      "loss": 1.0575,
      "num_input_tokens_seen": 117120,
      "step": 180
    },
    {
      "epoch": 0.09696016771488469,
      "grad_norm": 0.9468575716018677,
      "learning_rate": 2.410901467505241e-06,
      "loss": 1.2821,
      "num_input_tokens_seen": 120320,
      "step": 185
    },
    {
      "epoch": 0.09958071278825996,
      "grad_norm": 1.0569820404052734,
      "learning_rate": 2.476415094339623e-06,
      "loss": 1.1171,
      "num_input_tokens_seen": 123392,
      "step": 190
    },
    {
      "epoch": 0.10220125786163523,
      "grad_norm": 1.226874589920044,
      "learning_rate": 2.5419287211740044e-06,
      "loss": 1.1045,
      "num_input_tokens_seen": 126464,
      "step": 195
    },
    {
      "epoch": 0.10482180293501048,
      "grad_norm": 1.3569964170455933,
      "learning_rate": 2.6074423480083857e-06,
      "loss": 1.2033,
      "num_input_tokens_seen": 129504,
      "step": 200
    },
    {
      "epoch": 0.10744234800838574,
      "grad_norm": 1.337145209312439,
      "learning_rate": 2.6729559748427675e-06,
      "loss": 1.0416,
      "num_input_tokens_seen": 132608,
      "step": 205
    },
    {
      "epoch": 0.11006289308176101,
      "grad_norm": 1.6442594528198242,
      "learning_rate": 2.738469601677149e-06,
      "loss": 1.1792,
      "num_input_tokens_seen": 135264,
      "step": 210
    },
    {
      "epoch": 0.11268343815513626,
      "grad_norm": 2.2357373237609863,
      "learning_rate": 2.8039832285115303e-06,
      "loss": 1.1067,
      "num_input_tokens_seen": 138208,
      "step": 215
    },
    {
      "epoch": 0.11530398322851153,
      "grad_norm": 0.7529154419898987,
      "learning_rate": 2.869496855345912e-06,
      "loss": 1.1347,
      "num_input_tokens_seen": 141696,
      "step": 220
    },
    {
      "epoch": 0.1179245283018868,
      "grad_norm": 1.1493631601333618,
      "learning_rate": 2.935010482180294e-06,
      "loss": 1.3208,
      "num_input_tokens_seen": 144256,
      "step": 225
    },
    {
      "epoch": 0.12054507337526206,
      "grad_norm": 1.3320499658584595,
      "learning_rate": 3.0005241090146753e-06,
      "loss": 0.9491,
      "num_input_tokens_seen": 147904,
      "step": 230
    },
    {
      "epoch": 0.12316561844863731,
      "grad_norm": 1.060980200767517,
      "learning_rate": 3.0660377358490567e-06,
      "loss": 1.0128,
      "num_input_tokens_seen": 151616,
      "step": 235
    },
    {
      "epoch": 0.12578616352201258,
      "grad_norm": 2.3706953525543213,
      "learning_rate": 3.131551362683438e-06,
      "loss": 1.3354,
      "num_input_tokens_seen": 154048,
      "step": 240
    },
    {
      "epoch": 0.12840670859538783,
      "grad_norm": 1.2924612760543823,
      "learning_rate": 3.1970649895178203e-06,
      "loss": 1.0722,
      "num_input_tokens_seen": 156992,
      "step": 245
    },
    {
      "epoch": 0.1310272536687631,
      "grad_norm": 1.3629494905471802,
      "learning_rate": 3.2625786163522012e-06,
      "loss": 1.2822,
      "num_input_tokens_seen": 160544,
      "step": 250
    },
    {
      "epoch": 0.13364779874213836,
      "grad_norm": 0.808962881565094,
      "learning_rate": 3.3280922431865826e-06,
      "loss": 1.0967,
      "num_input_tokens_seen": 164000,
      "step": 255
    },
    {
      "epoch": 0.13626834381551362,
      "grad_norm": 1.1733357906341553,
      "learning_rate": 3.393605870020965e-06,
      "loss": 1.1386,
      "num_input_tokens_seen": 167200,
      "step": 260
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 1.5040850639343262,
      "learning_rate": 3.459119496855346e-06,
      "loss": 1.145,
      "num_input_tokens_seen": 169888,
      "step": 265
    },
    {
      "epoch": 0.14150943396226415,
      "grad_norm": 1.6307668685913086,
      "learning_rate": 3.524633123689727e-06,
      "loss": 1.1548,
      "num_input_tokens_seen": 172832,
      "step": 270
    },
    {
      "epoch": 0.1441299790356394,
      "grad_norm": 1.1821324825286865,
      "learning_rate": 3.5901467505241094e-06,
      "loss": 1.218,
      "num_input_tokens_seen": 176672,
      "step": 275
    },
    {
      "epoch": 0.14675052410901468,
      "grad_norm": 2.042625665664673,
      "learning_rate": 3.6556603773584908e-06,
      "loss": 1.2733,
      "num_input_tokens_seen": 180000,
      "step": 280
    },
    {
      "epoch": 0.14937106918238993,
      "grad_norm": 1.5842503309249878,
      "learning_rate": 3.7211740041928726e-06,
      "loss": 1.387,
      "num_input_tokens_seen": 182880,
      "step": 285
    },
    {
      "epoch": 0.1519916142557652,
      "grad_norm": 1.328354835510254,
      "learning_rate": 3.786687631027254e-06,
      "loss": 1.2718,
      "num_input_tokens_seen": 185760,
      "step": 290
    },
    {
      "epoch": 0.15461215932914046,
      "grad_norm": 1.4790843725204468,
      "learning_rate": 3.852201257861635e-06,
      "loss": 1.0621,
      "num_input_tokens_seen": 188960,
      "step": 295
    },
    {
      "epoch": 0.15723270440251572,
      "grad_norm": 1.5432802438735962,
      "learning_rate": 3.917714884696017e-06,
      "loss": 1.0525,
      "num_input_tokens_seen": 191488,
      "step": 300
    },
    {
      "epoch": 0.159853249475891,
      "grad_norm": 2.1004273891448975,
      "learning_rate": 3.9832285115303985e-06,
      "loss": 1.1667,
      "num_input_tokens_seen": 194240,
      "step": 305
    },
    {
      "epoch": 0.16247379454926625,
      "grad_norm": 1.9383958578109741,
      "learning_rate": 4.04874213836478e-06,
      "loss": 1.1059,
      "num_input_tokens_seen": 197408,
      "step": 310
    },
    {
      "epoch": 0.1650943396226415,
      "grad_norm": 1.1305123567581177,
      "learning_rate": 4.114255765199162e-06,
      "loss": 1.1533,
      "num_input_tokens_seen": 200384,
      "step": 315
    },
    {
      "epoch": 0.16771488469601678,
      "grad_norm": 0.818388044834137,
      "learning_rate": 4.1797693920335435e-06,
      "loss": 1.0632,
      "num_input_tokens_seen": 204480,
      "step": 320
    },
    {
      "epoch": 0.17033542976939203,
      "grad_norm": 1.3565993309020996,
      "learning_rate": 4.245283018867925e-06,
      "loss": 1.0918,
      "num_input_tokens_seen": 207264,
      "step": 325
    },
    {
      "epoch": 0.17295597484276728,
      "grad_norm": 0.7678110003471375,
      "learning_rate": 4.310796645702306e-06,
      "loss": 0.9057,
      "num_input_tokens_seen": 211360,
      "step": 330
    },
    {
      "epoch": 0.17557651991614256,
      "grad_norm": 0.7663017511367798,
      "learning_rate": 4.376310272536688e-06,
      "loss": 0.919,
      "num_input_tokens_seen": 215712,
      "step": 335
    },
    {
      "epoch": 0.17819706498951782,
      "grad_norm": 0.8756281733512878,
      "learning_rate": 4.441823899371069e-06,
      "loss": 1.0134,
      "num_input_tokens_seen": 219680,
      "step": 340
    },
    {
      "epoch": 0.18081761006289307,
      "grad_norm": 1.0025907754898071,
      "learning_rate": 4.507337526205451e-06,
      "loss": 1.2942,
      "num_input_tokens_seen": 223232,
      "step": 345
    },
    {
      "epoch": 0.18343815513626835,
      "grad_norm": 1.5143290758132935,
      "learning_rate": 4.572851153039833e-06,
      "loss": 1.289,
      "num_input_tokens_seen": 226528,
      "step": 350
    },
    {
      "epoch": 0.1860587002096436,
      "grad_norm": 1.0516499280929565,
      "learning_rate": 4.638364779874214e-06,
      "loss": 0.9162,
      "num_input_tokens_seen": 229600,
      "step": 355
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 0.6681299805641174,
      "learning_rate": 4.703878406708595e-06,
      "loss": 1.2946,
      "num_input_tokens_seen": 232512,
      "step": 360
    },
    {
      "epoch": 0.19129979035639413,
      "grad_norm": 1.3105086088180542,
      "learning_rate": 4.769392033542977e-06,
      "loss": 1.4451,
      "num_input_tokens_seen": 234784,
      "step": 365
    },
    {
      "epoch": 0.19392033542976939,
      "grad_norm": 1.106291651725769,
      "learning_rate": 4.834905660377359e-06,
      "loss": 1.2831,
      "num_input_tokens_seen": 237440,
      "step": 370
    },
    {
      "epoch": 0.19654088050314467,
      "grad_norm": 2.3086161613464355,
      "learning_rate": 4.90041928721174e-06,
      "loss": 1.1946,
      "num_input_tokens_seen": 240064,
      "step": 375
    },
    {
      "epoch": 0.19916142557651992,
      "grad_norm": 1.2394379377365112,
      "learning_rate": 4.965932914046122e-06,
      "loss": 1.2443,
      "num_input_tokens_seen": 243168,
      "step": 380
    },
    {
      "epoch": 0.20178197064989517,
      "grad_norm": 1.2500741481781006,
      "learning_rate": 5.031446540880504e-06,
      "loss": 1.2176,
      "num_input_tokens_seen": 246496,
      "step": 385
    },
    {
      "epoch": 0.20440251572327045,
      "grad_norm": 0.9927883148193359,
      "learning_rate": 5.096960167714885e-06,
      "loss": 1.053,
      "num_input_tokens_seen": 249408,
      "step": 390
    },
    {
      "epoch": 0.2070230607966457,
      "grad_norm": 0.7438414692878723,
      "learning_rate": 5.162473794549266e-06,
      "loss": 1.0393,
      "num_input_tokens_seen": 253248,
      "step": 395
    },
    {
      "epoch": 0.20964360587002095,
      "grad_norm": 1.5015522241592407,
      "learning_rate": 5.227987421383648e-06,
      "loss": 1.2808,
      "num_input_tokens_seen": 256352,
      "step": 400
    },
    {
      "epoch": 0.21226415094339623,
      "grad_norm": 1.0207701921463013,
      "learning_rate": 5.2935010482180295e-06,
      "loss": 1.0288,
      "num_input_tokens_seen": 259648,
      "step": 405
    },
    {
      "epoch": 0.2148846960167715,
      "grad_norm": 0.8713642358779907,
      "learning_rate": 5.359014675052411e-06,
      "loss": 0.9951,
      "num_input_tokens_seen": 262912,
      "step": 410
    },
    {
      "epoch": 0.21750524109014674,
      "grad_norm": 1.2778135538101196,
      "learning_rate": 5.424528301886793e-06,
      "loss": 1.2497,
      "num_input_tokens_seen": 265824,
      "step": 415
    },
    {
      "epoch": 0.22012578616352202,
      "grad_norm": 1.096075415611267,
      "learning_rate": 5.4900419287211744e-06,
      "loss": 1.259,
      "num_input_tokens_seen": 268032,
      "step": 420
    },
    {
      "epoch": 0.22274633123689727,
      "grad_norm": 1.528388500213623,
      "learning_rate": 5.555555555555556e-06,
      "loss": 1.1122,
      "num_input_tokens_seen": 271744,
      "step": 425
    },
    {
      "epoch": 0.22536687631027252,
      "grad_norm": 1.3035497665405273,
      "learning_rate": 5.621069182389937e-06,
      "loss": 1.116,
      "num_input_tokens_seen": 274944,
      "step": 430
    },
    {
      "epoch": 0.2279874213836478,
      "grad_norm": 1.7384201288223267,
      "learning_rate": 5.686582809224319e-06,
      "loss": 1.2531,
      "num_input_tokens_seen": 278080,
      "step": 435
    },
    {
      "epoch": 0.23060796645702306,
      "grad_norm": 1.252299427986145,
      "learning_rate": 5.752096436058701e-06,
      "loss": 0.9664,
      "num_input_tokens_seen": 281280,
      "step": 440
    },
    {
      "epoch": 0.23322851153039834,
      "grad_norm": 1.5272600650787354,
      "learning_rate": 5.817610062893082e-06,
      "loss": 0.9933,
      "num_input_tokens_seen": 285312,
      "step": 445
    },
    {
      "epoch": 0.2358490566037736,
      "grad_norm": 1.772614598274231,
      "learning_rate": 5.8831236897274636e-06,
      "loss": 1.1279,
      "num_input_tokens_seen": 288064,
      "step": 450
    },
    {
      "epoch": 0.23846960167714884,
      "grad_norm": 1.0684884786605835,
      "learning_rate": 5.948637316561846e-06,
      "loss": 0.8073,
      "num_input_tokens_seen": 291648,
      "step": 455
    },
    {
      "epoch": 0.24109014675052412,
      "grad_norm": 1.8625969886779785,
      "learning_rate": 6.014150943396227e-06,
      "loss": 1.1703,
      "num_input_tokens_seen": 294368,
      "step": 460
    },
    {
      "epoch": 0.24371069182389937,
      "grad_norm": 2.254037857055664,
      "learning_rate": 6.079664570230608e-06,
      "loss": 1.3073,
      "num_input_tokens_seen": 296864,
      "step": 465
    },
    {
      "epoch": 0.24633123689727462,
      "grad_norm": 1.7698010206222534,
      "learning_rate": 6.14517819706499e-06,
      "loss": 1.0753,
      "num_input_tokens_seen": 299968,
      "step": 470
    },
    {
      "epoch": 0.2489517819706499,
      "grad_norm": 1.1834242343902588,
      "learning_rate": 6.210691823899371e-06,
      "loss": 1.164,
      "num_input_tokens_seen": 302624,
      "step": 475
    },
    {
      "epoch": 0.25157232704402516,
      "grad_norm": 2.4810643196105957,
      "learning_rate": 6.2762054507337535e-06,
      "loss": 1.1043,
      "num_input_tokens_seen": 305216,
      "step": 480
    },
    {
      "epoch": 0.25419287211740044,
      "grad_norm": 2.092313528060913,
      "learning_rate": 6.341719077568134e-06,
      "loss": 1.4929,
      "num_input_tokens_seen": 307648,
      "step": 485
    },
    {
      "epoch": 0.25681341719077566,
      "grad_norm": 1.004888653755188,
      "learning_rate": 6.407232704402516e-06,
      "loss": 1.2271,
      "num_input_tokens_seen": 310464,
      "step": 490
    },
    {
      "epoch": 0.25943396226415094,
      "grad_norm": 0.9034015536308289,
      "learning_rate": 6.472746331236898e-06,
      "loss": 0.8909,
      "num_input_tokens_seen": 314912,
      "step": 495
    },
    {
      "epoch": 0.2620545073375262,
      "grad_norm": 1.6619904041290283,
      "learning_rate": 6.538259958071278e-06,
      "loss": 1.1037,
      "num_input_tokens_seen": 317792,
      "step": 500
    },
    {
      "epoch": 0.26467505241090145,
      "grad_norm": 1.5716919898986816,
      "learning_rate": 6.60377358490566e-06,
      "loss": 1.0317,
      "num_input_tokens_seen": 321024,
      "step": 505
    },
    {
      "epoch": 0.2672955974842767,
      "grad_norm": 1.155630350112915,
      "learning_rate": 6.669287211740043e-06,
      "loss": 1.1932,
      "num_input_tokens_seen": 323168,
      "step": 510
    },
    {
      "epoch": 0.269916142557652,
      "grad_norm": 1.0812463760375977,
      "learning_rate": 6.734800838574423e-06,
      "loss": 1.1408,
      "num_input_tokens_seen": 326496,
      "step": 515
    },
    {
      "epoch": 0.27253668763102723,
      "grad_norm": 0.8527852892875671,
      "learning_rate": 6.800314465408805e-06,
      "loss": 0.9794,
      "num_input_tokens_seen": 331072,
      "step": 520
    },
    {
      "epoch": 0.2751572327044025,
      "grad_norm": 1.431134819984436,
      "learning_rate": 6.865828092243188e-06,
      "loss": 1.1172,
      "num_input_tokens_seen": 334976,
      "step": 525
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 1.3294955492019653,
      "learning_rate": 6.931341719077568e-06,
      "loss": 1.1559,
      "num_input_tokens_seen": 337568,
      "step": 530
    },
    {
      "epoch": 0.280398322851153,
      "grad_norm": 1.3230212926864624,
      "learning_rate": 6.9968553459119495e-06,
      "loss": 1.0784,
      "num_input_tokens_seen": 340256,
      "step": 535
    },
    {
      "epoch": 0.2830188679245283,
      "grad_norm": 1.6227290630340576,
      "learning_rate": 7.062368972746332e-06,
      "loss": 1.7389,
      "num_input_tokens_seen": 342688,
      "step": 540
    },
    {
      "epoch": 0.2856394129979036,
      "grad_norm": 1.0893802642822266,
      "learning_rate": 7.127882599580712e-06,
      "loss": 0.9851,
      "num_input_tokens_seen": 345824,
      "step": 545
    },
    {
      "epoch": 0.2882599580712788,
      "grad_norm": 1.7350273132324219,
      "learning_rate": 7.1933962264150945e-06,
      "loss": 0.8773,
      "num_input_tokens_seen": 349408,
      "step": 550
    },
    {
      "epoch": 0.2908805031446541,
      "grad_norm": 1.3345091342926025,
      "learning_rate": 7.258909853249477e-06,
      "loss": 0.9308,
      "num_input_tokens_seen": 352544,
      "step": 555
    },
    {
      "epoch": 0.29350104821802936,
      "grad_norm": 1.1998487710952759,
      "learning_rate": 7.324423480083858e-06,
      "loss": 1.2108,
      "num_input_tokens_seen": 355232,
      "step": 560
    },
    {
      "epoch": 0.29612159329140464,
      "grad_norm": 1.1440997123718262,
      "learning_rate": 7.3899371069182395e-06,
      "loss": 1.1632,
      "num_input_tokens_seen": 358720,
      "step": 565
    },
    {
      "epoch": 0.29874213836477986,
      "grad_norm": 1.7511945962905884,
      "learning_rate": 7.455450733752621e-06,
      "loss": 1.1364,
      "num_input_tokens_seen": 361344,
      "step": 570
    },
    {
      "epoch": 0.30136268343815514,
      "grad_norm": 0.8281915187835693,
      "learning_rate": 7.520964360587003e-06,
      "loss": 1.1571,
      "num_input_tokens_seen": 364288,
      "step": 575
    },
    {
      "epoch": 0.3039832285115304,
      "grad_norm": 2.1254031658172607,
      "learning_rate": 7.586477987421384e-06,
      "loss": 1.0578,
      "num_input_tokens_seen": 366784,
      "step": 580
    },
    {
      "epoch": 0.30660377358490565,
      "grad_norm": 1.613650918006897,
      "learning_rate": 7.651991614255765e-06,
      "loss": 1.1301,
      "num_input_tokens_seen": 369952,
      "step": 585
    },
    {
      "epoch": 0.30922431865828093,
      "grad_norm": 1.2786773443222046,
      "learning_rate": 7.717505241090148e-06,
      "loss": 0.9981,
      "num_input_tokens_seen": 373120,
      "step": 590
    },
    {
      "epoch": 0.3118448637316562,
      "grad_norm": 0.6460239887237549,
      "learning_rate": 7.783018867924528e-06,
      "loss": 0.9016,
      "num_input_tokens_seen": 376416,
      "step": 595
    },
    {
      "epoch": 0.31446540880503143,
      "grad_norm": 0.8473554849624634,
      "learning_rate": 7.848532494758911e-06,
      "loss": 1.0454,
      "num_input_tokens_seen": 379232,
      "step": 600
    },
    {
      "epoch": 0.3170859538784067,
      "grad_norm": 1.1436009407043457,
      "learning_rate": 7.914046121593292e-06,
      "loss": 0.8445,
      "num_input_tokens_seen": 383072,
      "step": 605
    },
    {
      "epoch": 0.319706498951782,
      "grad_norm": 0.994705319404602,
      "learning_rate": 7.979559748427674e-06,
      "loss": 1.0124,
      "num_input_tokens_seen": 386944,
      "step": 610
    },
    {
      "epoch": 0.3223270440251572,
      "grad_norm": 0.5618464946746826,
      "learning_rate": 8.045073375262055e-06,
      "loss": 0.9546,
      "num_input_tokens_seen": 390944,
      "step": 615
    },
    {
      "epoch": 0.3249475890985325,
      "grad_norm": 1.4821349382400513,
      "learning_rate": 8.110587002096436e-06,
      "loss": 0.9566,
      "num_input_tokens_seen": 393856,
      "step": 620
    },
    {
      "epoch": 0.3275681341719078,
      "grad_norm": 1.8436481952667236,
      "learning_rate": 8.176100628930818e-06,
      "loss": 1.0948,
      "num_input_tokens_seen": 397280,
      "step": 625
    },
    {
      "epoch": 0.330188679245283,
      "grad_norm": 2.887361526489258,
      "learning_rate": 8.241614255765199e-06,
      "loss": 1.1031,
      "num_input_tokens_seen": 400448,
      "step": 630
    },
    {
      "epoch": 0.3328092243186583,
      "grad_norm": 1.9585963487625122,
      "learning_rate": 8.307127882599582e-06,
      "loss": 1.0229,
      "num_input_tokens_seen": 404192,
      "step": 635
    },
    {
      "epoch": 0.33542976939203356,
      "grad_norm": 1.2262060642242432,
      "learning_rate": 8.372641509433962e-06,
      "loss": 0.9708,
      "num_input_tokens_seen": 407744,
      "step": 640
    },
    {
      "epoch": 0.3380503144654088,
      "grad_norm": 1.8002655506134033,
      "learning_rate": 8.438155136268345e-06,
      "loss": 0.9346,
      "num_input_tokens_seen": 410592,
      "step": 645
    },
    {
      "epoch": 0.34067085953878407,
      "grad_norm": 1.3083292245864868,
      "learning_rate": 8.503668763102726e-06,
      "loss": 0.8646,
      "num_input_tokens_seen": 414624,
      "step": 650
    },
    {
      "epoch": 0.34329140461215935,
      "grad_norm": 1.0280052423477173,
      "learning_rate": 8.569182389937106e-06,
      "loss": 0.8933,
      "num_input_tokens_seen": 417600,
      "step": 655
    },
    {
      "epoch": 0.34591194968553457,
      "grad_norm": 1.1652308702468872,
      "learning_rate": 8.634696016771489e-06,
      "loss": 0.9705,
      "num_input_tokens_seen": 421024,
      "step": 660
    },
    {
      "epoch": 0.34853249475890985,
      "grad_norm": 1.256777286529541,
      "learning_rate": 8.70020964360587e-06,
      "loss": 0.9202,
      "num_input_tokens_seen": 424256,
      "step": 665
    },
    {
      "epoch": 0.35115303983228513,
      "grad_norm": 1.2865931987762451,
      "learning_rate": 8.765723270440252e-06,
      "loss": 0.904,
      "num_input_tokens_seen": 427808,
      "step": 670
    },
    {
      "epoch": 0.35377358490566035,
      "grad_norm": 1.011330485343933,
      "learning_rate": 8.831236897274633e-06,
      "loss": 0.9192,
      "num_input_tokens_seen": 430400,
      "step": 675
    },
    {
      "epoch": 0.35639412997903563,
      "grad_norm": 1.1321011781692505,
      "learning_rate": 8.896750524109015e-06,
      "loss": 0.8401,
      "num_input_tokens_seen": 434176,
      "step": 680
    },
    {
      "epoch": 0.3590146750524109,
      "grad_norm": 1.830630898475647,
      "learning_rate": 8.962264150943396e-06,
      "loss": 0.9899,
      "num_input_tokens_seen": 437536,
      "step": 685
    },
    {
      "epoch": 0.36163522012578614,
      "grad_norm": 0.9190878868103027,
      "learning_rate": 9.027777777777777e-06,
      "loss": 0.9247,
      "num_input_tokens_seen": 440384,
      "step": 690
    },
    {
      "epoch": 0.3642557651991614,
      "grad_norm": 0.6780405044555664,
      "learning_rate": 9.09329140461216e-06,
      "loss": 0.9443,
      "num_input_tokens_seen": 443808,
      "step": 695
    },
    {
      "epoch": 0.3668763102725367,
      "grad_norm": 2.895373821258545,
      "learning_rate": 9.15880503144654e-06,
      "loss": 1.0433,
      "num_input_tokens_seen": 446880,
      "step": 700
    },
    {
      "epoch": 0.3694968553459119,
      "grad_norm": 1.0361524820327759,
      "learning_rate": 9.224318658280923e-06,
      "loss": 0.743,
      "num_input_tokens_seen": 450368,
      "step": 705
    },
    {
      "epoch": 0.3721174004192872,
      "grad_norm": 1.3437695503234863,
      "learning_rate": 9.289832285115305e-06,
      "loss": 0.8993,
      "num_input_tokens_seen": 453344,
      "step": 710
    },
    {
      "epoch": 0.3747379454926625,
      "grad_norm": 1.2733701467514038,
      "learning_rate": 9.355345911949686e-06,
      "loss": 0.9977,
      "num_input_tokens_seen": 456768,
      "step": 715
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 1.0573630332946777,
      "learning_rate": 9.420859538784067e-06,
      "loss": 0.8158,
      "num_input_tokens_seen": 459936,
      "step": 720
    },
    {
      "epoch": 0.379979035639413,
      "grad_norm": 0.6502221822738647,
      "learning_rate": 9.486373165618449e-06,
      "loss": 0.9579,
      "num_input_tokens_seen": 463040,
      "step": 725
    },
    {
      "epoch": 0.38259958071278827,
      "grad_norm": 1.149724006652832,
      "learning_rate": 9.551886792452832e-06,
      "loss": 0.9068,
      "num_input_tokens_seen": 466432,
      "step": 730
    },
    {
      "epoch": 0.38522012578616355,
      "grad_norm": 1.527093529701233,
      "learning_rate": 9.617400419287211e-06,
      "loss": 0.8459,
      "num_input_tokens_seen": 469440,
      "step": 735
    },
    {
      "epoch": 0.38784067085953877,
      "grad_norm": 1.7061209678649902,
      "learning_rate": 9.682914046121595e-06,
      "loss": 0.9586,
      "num_input_tokens_seen": 474400,
      "step": 740
    },
    {
      "epoch": 0.39046121593291405,
      "grad_norm": 0.991736888885498,
      "learning_rate": 9.748427672955976e-06,
      "loss": 1.0267,
      "num_input_tokens_seen": 478144,
      "step": 745
    },
    {
      "epoch": 0.39308176100628933,
      "grad_norm": 0.7642011642456055,
      "learning_rate": 9.813941299790357e-06,
      "loss": 0.7839,
      "num_input_tokens_seen": 481248,
      "step": 750
    },
    {
      "epoch": 0.39570230607966456,
      "grad_norm": 0.7987241148948669,
      "learning_rate": 9.879454926624739e-06,
      "loss": 0.8189,
      "num_input_tokens_seen": 484576,
      "step": 755
    },
    {
      "epoch": 0.39832285115303984,
      "grad_norm": 0.946847140789032,
      "learning_rate": 9.94496855345912e-06,
      "loss": 0.712,
      "num_input_tokens_seen": 488288,
      "step": 760
    },
    {
      "epoch": 0.4009433962264151,
      "grad_norm": 1.022737979888916,
      "learning_rate": 1.0010482180293501e-05,
      "loss": 1.1696,
      "num_input_tokens_seen": 490688,
      "step": 765
    },
    {
      "epoch": 0.40356394129979034,
      "grad_norm": 0.6939501762390137,
      "learning_rate": 1.0075995807127883e-05,
      "loss": 0.7391,
      "num_input_tokens_seen": 493728,
      "step": 770
    },
    {
      "epoch": 0.4061844863731656,
      "grad_norm": 1.6825511455535889,
      "learning_rate": 1.0141509433962266e-05,
      "loss": 0.9197,
      "num_input_tokens_seen": 496512,
      "step": 775
    },
    {
      "epoch": 0.4088050314465409,
      "grad_norm": 0.830901026725769,
      "learning_rate": 1.0207023060796646e-05,
      "loss": 0.8877,
      "num_input_tokens_seen": 499616,
      "step": 780
    },
    {
      "epoch": 0.4114255765199161,
      "grad_norm": 0.8901025056838989,
      "learning_rate": 1.0272536687631027e-05,
      "loss": 0.7916,
      "num_input_tokens_seen": 502080,
      "step": 785
    },
    {
      "epoch": 0.4140461215932914,
      "grad_norm": 0.6117364168167114,
      "learning_rate": 1.033805031446541e-05,
      "loss": 0.8536,
      "num_input_tokens_seen": 505248,
      "step": 790
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.9237949848175049,
      "learning_rate": 1.040356394129979e-05,
      "loss": 0.9699,
      "num_input_tokens_seen": 508128,
      "step": 795
    },
    {
      "epoch": 0.4192872117400419,
      "grad_norm": 0.8899444937705994,
      "learning_rate": 1.0469077568134173e-05,
      "loss": 0.8238,
      "num_input_tokens_seen": 511392,
      "step": 800
    },
    {
      "epoch": 0.4219077568134172,
      "grad_norm": 0.3952719569206238,
      "learning_rate": 1.0534591194968554e-05,
      "loss": 0.7689,
      "num_input_tokens_seen": 516288,
      "step": 805
    },
    {
      "epoch": 0.42452830188679247,
      "grad_norm": 0.7852498888969421,
      "learning_rate": 1.0600104821802936e-05,
      "loss": 0.8616,
      "num_input_tokens_seen": 519360,
      "step": 810
    },
    {
      "epoch": 0.4271488469601677,
      "grad_norm": 0.7102294564247131,
      "learning_rate": 1.0665618448637317e-05,
      "loss": 0.7787,
      "num_input_tokens_seen": 522368,
      "step": 815
    },
    {
      "epoch": 0.429769392033543,
      "grad_norm": 1.069189190864563,
      "learning_rate": 1.0731132075471698e-05,
      "loss": 0.7677,
      "num_input_tokens_seen": 526336,
      "step": 820
    },
    {
      "epoch": 0.43238993710691825,
      "grad_norm": 1.2221101522445679,
      "learning_rate": 1.079664570230608e-05,
      "loss": 0.7223,
      "num_input_tokens_seen": 529920,
      "step": 825
    },
    {
      "epoch": 0.4350104821802935,
      "grad_norm": 0.6926294565200806,
      "learning_rate": 1.0862159329140461e-05,
      "loss": 0.8463,
      "num_input_tokens_seen": 536256,
      "step": 830
    },
    {
      "epoch": 0.43763102725366876,
      "grad_norm": 0.9846360087394714,
      "learning_rate": 1.0927672955974844e-05,
      "loss": 0.8228,
      "num_input_tokens_seen": 539616,
      "step": 835
    },
    {
      "epoch": 0.44025157232704404,
      "grad_norm": 1.049879789352417,
      "learning_rate": 1.0993186582809224e-05,
      "loss": 0.7275,
      "num_input_tokens_seen": 543168,
      "step": 840
    },
    {
      "epoch": 0.44287211740041926,
      "grad_norm": 0.7242798805236816,
      "learning_rate": 1.1058700209643607e-05,
      "loss": 0.7174,
      "num_input_tokens_seen": 546528,
      "step": 845
    },
    {
      "epoch": 0.44549266247379454,
      "grad_norm": 1.1593117713928223,
      "learning_rate": 1.1124213836477988e-05,
      "loss": 0.7,
      "num_input_tokens_seen": 549216,
      "step": 850
    },
    {
      "epoch": 0.4481132075471698,
      "grad_norm": 0.9161275625228882,
      "learning_rate": 1.118972746331237e-05,
      "loss": 0.8362,
      "num_input_tokens_seen": 552224,
      "step": 855
    },
    {
      "epoch": 0.45073375262054505,
      "grad_norm": 1.2199952602386475,
      "learning_rate": 1.1255241090146751e-05,
      "loss": 0.7515,
      "num_input_tokens_seen": 555424,
      "step": 860
    },
    {
      "epoch": 0.4533542976939203,
      "grad_norm": 1.4655280113220215,
      "learning_rate": 1.1320754716981132e-05,
      "loss": 0.8984,
      "num_input_tokens_seen": 557696,
      "step": 865
    },
    {
      "epoch": 0.4559748427672956,
      "grad_norm": 0.9997352361679077,
      "learning_rate": 1.1386268343815514e-05,
      "loss": 0.6857,
      "num_input_tokens_seen": 560768,
      "step": 870
    },
    {
      "epoch": 0.4585953878406709,
      "grad_norm": 0.8739400506019592,
      "learning_rate": 1.1451781970649895e-05,
      "loss": 0.8635,
      "num_input_tokens_seen": 563616,
      "step": 875
    },
    {
      "epoch": 0.4612159329140461,
      "grad_norm": 0.8887079358100891,
      "learning_rate": 1.1517295597484278e-05,
      "loss": 0.7798,
      "num_input_tokens_seen": 567072,
      "step": 880
    },
    {
      "epoch": 0.4638364779874214,
      "grad_norm": 1.5959123373031616,
      "learning_rate": 1.158280922431866e-05,
      "loss": 0.5976,
      "num_input_tokens_seen": 571232,
      "step": 885
    },
    {
      "epoch": 0.46645702306079667,
      "grad_norm": 1.1102871894836426,
      "learning_rate": 1.164832285115304e-05,
      "loss": 0.641,
      "num_input_tokens_seen": 575232,
      "step": 890
    },
    {
      "epoch": 0.4690775681341719,
      "grad_norm": 1.134380578994751,
      "learning_rate": 1.1713836477987422e-05,
      "loss": 0.7065,
      "num_input_tokens_seen": 578784,
      "step": 895
    },
    {
      "epoch": 0.4716981132075472,
      "grad_norm": 0.925142765045166,
      "learning_rate": 1.1779350104821804e-05,
      "loss": 0.7684,
      "num_input_tokens_seen": 582176,
      "step": 900
    },
    {
      "epoch": 0.47431865828092246,
      "grad_norm": 1.0942577123641968,
      "learning_rate": 1.1844863731656185e-05,
      "loss": 0.8767,
      "num_input_tokens_seen": 584512,
      "step": 905
    },
    {
      "epoch": 0.4769392033542977,
      "grad_norm": 0.8740180730819702,
      "learning_rate": 1.1910377358490566e-05,
      "loss": 0.6299,
      "num_input_tokens_seen": 587104,
      "step": 910
    },
    {
      "epoch": 0.47955974842767296,
      "grad_norm": 0.7230186462402344,
      "learning_rate": 1.197589098532495e-05,
      "loss": 0.8966,
      "num_input_tokens_seen": 590848,
      "step": 915
    },
    {
      "epoch": 0.48218029350104824,
      "grad_norm": 0.7237220406532288,
      "learning_rate": 1.204140461215933e-05,
      "loss": 0.8967,
      "num_input_tokens_seen": 593536,
      "step": 920
    },
    {
      "epoch": 0.48480083857442346,
      "grad_norm": 0.6954053044319153,
      "learning_rate": 1.210691823899371e-05,
      "loss": 0.6845,
      "num_input_tokens_seen": 600256,
      "step": 925
    },
    {
      "epoch": 0.48742138364779874,
      "grad_norm": 1.0264605283737183,
      "learning_rate": 1.2172431865828094e-05,
      "loss": 0.6429,
      "num_input_tokens_seen": 604960,
      "step": 930
    },
    {
      "epoch": 0.490041928721174,
      "grad_norm": 0.8662064671516418,
      "learning_rate": 1.2237945492662473e-05,
      "loss": 0.8106,
      "num_input_tokens_seen": 607520,
      "step": 935
    },
    {
      "epoch": 0.49266247379454925,
      "grad_norm": 1.0290753841400146,
      "learning_rate": 1.2303459119496856e-05,
      "loss": 0.6408,
      "num_input_tokens_seen": 610368,
      "step": 940
    },
    {
      "epoch": 0.49528301886792453,
      "grad_norm": 0.6103917360305786,
      "learning_rate": 1.2368972746331238e-05,
      "loss": 0.6588,
      "num_input_tokens_seen": 614080,
      "step": 945
    },
    {
      "epoch": 0.4979035639412998,
      "grad_norm": 0.9035815000534058,
      "learning_rate": 1.243448637316562e-05,
      "loss": 0.767,
      "num_input_tokens_seen": 617056,
      "step": 950
    },
    {
      "epoch": 0.500524109014675,
      "grad_norm": 1.669739842414856,
      "learning_rate": 1.25e-05,
      "loss": 0.9155,
      "num_input_tokens_seen": 620800,
      "step": 955
    },
    {
      "epoch": 0.5031446540880503,
      "grad_norm": 0.695198118686676,
      "learning_rate": 1.2565513626834382e-05,
      "loss": 0.7314,
      "num_input_tokens_seen": 623872,
      "step": 960
    },
    {
      "epoch": 0.5057651991614256,
      "grad_norm": 0.6967965960502625,
      "learning_rate": 1.2631027253668765e-05,
      "loss": 0.8346,
      "num_input_tokens_seen": 626752,
      "step": 965
    },
    {
      "epoch": 0.5083857442348009,
      "grad_norm": 0.7541130185127258,
      "learning_rate": 1.2696540880503146e-05,
      "loss": 0.6533,
      "num_input_tokens_seen": 630304,
      "step": 970
    },
    {
      "epoch": 0.5110062893081762,
      "grad_norm": 0.4898817241191864,
      "learning_rate": 1.2762054507337526e-05,
      "loss": 0.7242,
      "num_input_tokens_seen": 634048,
      "step": 975
    },
    {
      "epoch": 0.5136268343815513,
      "grad_norm": 1.2283053398132324,
      "learning_rate": 1.2827568134171907e-05,
      "loss": 0.6257,
      "num_input_tokens_seen": 637728,
      "step": 980
    },
    {
      "epoch": 0.5162473794549266,
      "grad_norm": 0.5999546051025391,
      "learning_rate": 1.289308176100629e-05,
      "loss": 0.7189,
      "num_input_tokens_seen": 641024,
      "step": 985
    },
    {
      "epoch": 0.5188679245283019,
      "grad_norm": 0.5880163311958313,
      "learning_rate": 1.2958595387840672e-05,
      "loss": 0.766,
      "num_input_tokens_seen": 643936,
      "step": 990
    },
    {
      "epoch": 0.5214884696016772,
      "grad_norm": 0.9242783188819885,
      "learning_rate": 1.3024109014675053e-05,
      "loss": 0.7639,
      "num_input_tokens_seen": 647744,
      "step": 995
    },
    {
      "epoch": 0.5241090146750524,
      "grad_norm": 1.4268386363983154,
      "learning_rate": 1.3089622641509436e-05,
      "loss": 0.7734,
      "num_input_tokens_seen": 650560,
      "step": 1000
    },
    {
      "epoch": 0.5267295597484277,
      "grad_norm": 0.8678954243659973,
      "learning_rate": 1.3155136268343814e-05,
      "loss": 0.7733,
      "num_input_tokens_seen": 652896,
      "step": 1005
    },
    {
      "epoch": 0.5293501048218029,
      "grad_norm": 0.6175528764724731,
      "learning_rate": 1.3220649895178197e-05,
      "loss": 0.9232,
      "num_input_tokens_seen": 655456,
      "step": 1010
    },
    {
      "epoch": 0.5319706498951782,
      "grad_norm": 0.9663990139961243,
      "learning_rate": 1.3286163522012579e-05,
      "loss": 0.7858,
      "num_input_tokens_seen": 658912,
      "step": 1015
    },
    {
      "epoch": 0.5345911949685535,
      "grad_norm": 2.2702572345733643,
      "learning_rate": 1.3351677148846962e-05,
      "loss": 0.6795,
      "num_input_tokens_seen": 661728,
      "step": 1020
    },
    {
      "epoch": 0.5372117400419287,
      "grad_norm": 0.6384498476982117,
      "learning_rate": 1.3417190775681343e-05,
      "loss": 0.7393,
      "num_input_tokens_seen": 665280,
      "step": 1025
    },
    {
      "epoch": 0.539832285115304,
      "grad_norm": 0.60023432970047,
      "learning_rate": 1.3482704402515725e-05,
      "loss": 0.6431,
      "num_input_tokens_seen": 668512,
      "step": 1030
    },
    {
      "epoch": 0.5424528301886793,
      "grad_norm": 1.482827067375183,
      "learning_rate": 1.3548218029350104e-05,
      "loss": 0.8231,
      "num_input_tokens_seen": 670400,
      "step": 1035
    },
    {
      "epoch": 0.5450733752620545,
      "grad_norm": 1.0700109004974365,
      "learning_rate": 1.3613731656184486e-05,
      "loss": 0.8013,
      "num_input_tokens_seen": 672864,
      "step": 1040
    },
    {
      "epoch": 0.5476939203354297,
      "grad_norm": 0.6908577680587769,
      "learning_rate": 1.3679245283018869e-05,
      "loss": 0.7736,
      "num_input_tokens_seen": 676256,
      "step": 1045
    },
    {
      "epoch": 0.550314465408805,
      "grad_norm": 1.122469186782837,
      "learning_rate": 1.374475890985325e-05,
      "loss": 0.8523,
      "num_input_tokens_seen": 678816,
      "step": 1050
    },
    {
      "epoch": 0.5529350104821803,
      "grad_norm": 0.7409030199050903,
      "learning_rate": 1.3810272536687632e-05,
      "loss": 0.7713,
      "num_input_tokens_seen": 682560,
      "step": 1055
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.5258896946907043,
      "learning_rate": 1.3875786163522015e-05,
      "loss": 0.5929,
      "num_input_tokens_seen": 686592,
      "step": 1060
    },
    {
      "epoch": 0.5581761006289309,
      "grad_norm": 1.4558638334274292,
      "learning_rate": 1.3941299790356394e-05,
      "loss": 0.7546,
      "num_input_tokens_seen": 689248,
      "step": 1065
    },
    {
      "epoch": 0.560796645702306,
      "grad_norm": 0.8756601810455322,
      "learning_rate": 1.4006813417190776e-05,
      "loss": 0.6113,
      "num_input_tokens_seen": 692576,
      "step": 1070
    },
    {
      "epoch": 0.5634171907756813,
      "grad_norm": 2.6592483520507812,
      "learning_rate": 1.4072327044025157e-05,
      "loss": 0.7095,
      "num_input_tokens_seen": 695872,
      "step": 1075
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 0.9924477338790894,
      "learning_rate": 1.413784067085954e-05,
      "loss": 0.8327,
      "num_input_tokens_seen": 699264,
      "step": 1080
    },
    {
      "epoch": 0.5686582809224319,
      "grad_norm": 0.6818705201148987,
      "learning_rate": 1.4203354297693922e-05,
      "loss": 0.7208,
      "num_input_tokens_seen": 702016,
      "step": 1085
    },
    {
      "epoch": 0.5712788259958071,
      "grad_norm": 0.37946847081184387,
      "learning_rate": 1.4268867924528303e-05,
      "loss": 0.644,
      "num_input_tokens_seen": 706688,
      "step": 1090
    },
    {
      "epoch": 0.5738993710691824,
      "grad_norm": 0.703409731388092,
      "learning_rate": 1.4334381551362683e-05,
      "loss": 0.8037,
      "num_input_tokens_seen": 709312,
      "step": 1095
    },
    {
      "epoch": 0.5765199161425576,
      "grad_norm": 0.9165586829185486,
      "learning_rate": 1.4399895178197064e-05,
      "loss": 0.5237,
      "num_input_tokens_seen": 712032,
      "step": 1100
    },
    {
      "epoch": 0.5791404612159329,
      "grad_norm": 1.196414828300476,
      "learning_rate": 1.4465408805031447e-05,
      "loss": 0.6071,
      "num_input_tokens_seen": 715040,
      "step": 1105
    },
    {
      "epoch": 0.5817610062893082,
      "grad_norm": 0.5982649922370911,
      "learning_rate": 1.4530922431865828e-05,
      "loss": 0.5949,
      "num_input_tokens_seen": 718752,
      "step": 1110
    },
    {
      "epoch": 0.5843815513626834,
      "grad_norm": 1.0060839653015137,
      "learning_rate": 1.4596436058700211e-05,
      "loss": 0.6999,
      "num_input_tokens_seen": 721984,
      "step": 1115
    },
    {
      "epoch": 0.5870020964360587,
      "grad_norm": 1.9557948112487793,
      "learning_rate": 1.4661949685534593e-05,
      "loss": 0.5954,
      "num_input_tokens_seen": 724128,
      "step": 1120
    },
    {
      "epoch": 0.589622641509434,
      "grad_norm": 1.2036139965057373,
      "learning_rate": 1.4727463312368974e-05,
      "loss": 0.7578,
      "num_input_tokens_seen": 726944,
      "step": 1125
    },
    {
      "epoch": 0.5922431865828093,
      "grad_norm": 0.4343279302120209,
      "learning_rate": 1.4792976939203354e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 730688,
      "step": 1130
    },
    {
      "epoch": 0.5948637316561844,
      "grad_norm": 0.6154919266700745,
      "learning_rate": 1.4858490566037735e-05,
      "loss": 0.6967,
      "num_input_tokens_seen": 734464,
      "step": 1135
    },
    {
      "epoch": 0.5974842767295597,
      "grad_norm": 0.6523181796073914,
      "learning_rate": 1.4924004192872118e-05,
      "loss": 0.7887,
      "num_input_tokens_seen": 737856,
      "step": 1140
    },
    {
      "epoch": 0.600104821802935,
      "grad_norm": 0.617931067943573,
      "learning_rate": 1.49895178197065e-05,
      "loss": 0.7626,
      "num_input_tokens_seen": 740832,
      "step": 1145
    },
    {
      "epoch": 0.6027253668763103,
      "grad_norm": 1.4804531335830688,
      "learning_rate": 1.5055031446540883e-05,
      "loss": 0.712,
      "num_input_tokens_seen": 743712,
      "step": 1150
    },
    {
      "epoch": 0.6053459119496856,
      "grad_norm": 0.6655548810958862,
      "learning_rate": 1.5120545073375264e-05,
      "loss": 0.8289,
      "num_input_tokens_seen": 746144,
      "step": 1155
    },
    {
      "epoch": 0.6079664570230608,
      "grad_norm": 0.883868396282196,
      "learning_rate": 1.5186058700209644e-05,
      "loss": 0.6158,
      "num_input_tokens_seen": 749312,
      "step": 1160
    },
    {
      "epoch": 0.610587002096436,
      "grad_norm": 0.5131903290748596,
      "learning_rate": 1.5251572327044025e-05,
      "loss": 0.6922,
      "num_input_tokens_seen": 752928,
      "step": 1165
    },
    {
      "epoch": 0.6132075471698113,
      "grad_norm": 1.1576027870178223,
      "learning_rate": 1.5317085953878407e-05,
      "loss": 0.6589,
      "num_input_tokens_seen": 755520,
      "step": 1170
    },
    {
      "epoch": 0.6158280922431866,
      "grad_norm": 0.8635493516921997,
      "learning_rate": 1.538259958071279e-05,
      "loss": 0.5974,
      "num_input_tokens_seen": 758272,
      "step": 1175
    },
    {
      "epoch": 0.6184486373165619,
      "grad_norm": 0.4943788945674896,
      "learning_rate": 1.5448113207547173e-05,
      "loss": 0.6506,
      "num_input_tokens_seen": 761728,
      "step": 1180
    },
    {
      "epoch": 0.6210691823899371,
      "grad_norm": 0.6860030293464661,
      "learning_rate": 1.5513626834381552e-05,
      "loss": 0.6382,
      "num_input_tokens_seen": 765344,
      "step": 1185
    },
    {
      "epoch": 0.6236897274633124,
      "grad_norm": 0.8834742307662964,
      "learning_rate": 1.5579140461215932e-05,
      "loss": 0.6724,
      "num_input_tokens_seen": 768352,
      "step": 1190
    },
    {
      "epoch": 0.6263102725366876,
      "grad_norm": 0.9074076414108276,
      "learning_rate": 1.5644654088050315e-05,
      "loss": 0.7521,
      "num_input_tokens_seen": 771424,
      "step": 1195
    },
    {
      "epoch": 0.6289308176100629,
      "grad_norm": 0.8703955411911011,
      "learning_rate": 1.5710167714884695e-05,
      "loss": 0.5367,
      "num_input_tokens_seen": 776000,
      "step": 1200
    },
    {
      "epoch": 0.6315513626834381,
      "grad_norm": 0.4681857228279114,
      "learning_rate": 1.5775681341719078e-05,
      "loss": 0.6862,
      "num_input_tokens_seen": 779264,
      "step": 1205
    },
    {
      "epoch": 0.6341719077568134,
      "grad_norm": 0.6027076244354248,
      "learning_rate": 1.584119496855346e-05,
      "loss": 0.5302,
      "num_input_tokens_seen": 782656,
      "step": 1210
    },
    {
      "epoch": 0.6367924528301887,
      "grad_norm": 1.2271872758865356,
      "learning_rate": 1.5906708595387844e-05,
      "loss": 0.6609,
      "num_input_tokens_seen": 786144,
      "step": 1215
    },
    {
      "epoch": 0.639412997903564,
      "grad_norm": 1.4019187688827515,
      "learning_rate": 1.597222222222222e-05,
      "loss": 0.6663,
      "num_input_tokens_seen": 789856,
      "step": 1220
    },
    {
      "epoch": 0.6420335429769392,
      "grad_norm": 0.6132024526596069,
      "learning_rate": 1.6037735849056604e-05,
      "loss": 0.7234,
      "num_input_tokens_seen": 792768,
      "step": 1225
    },
    {
      "epoch": 0.6446540880503144,
      "grad_norm": 1.1024788618087769,
      "learning_rate": 1.6103249475890987e-05,
      "loss": 0.5935,
      "num_input_tokens_seen": 795456,
      "step": 1230
    },
    {
      "epoch": 0.6472746331236897,
      "grad_norm": 0.4812246561050415,
      "learning_rate": 1.6168763102725366e-05,
      "loss": 0.5871,
      "num_input_tokens_seen": 799648,
      "step": 1235
    },
    {
      "epoch": 0.649895178197065,
      "grad_norm": 0.36749911308288574,
      "learning_rate": 1.623427672955975e-05,
      "loss": 0.711,
      "num_input_tokens_seen": 803776,
      "step": 1240
    },
    {
      "epoch": 0.6525157232704403,
      "grad_norm": 0.5557461380958557,
      "learning_rate": 1.6299790356394132e-05,
      "loss": 0.5217,
      "num_input_tokens_seen": 807296,
      "step": 1245
    },
    {
      "epoch": 0.6551362683438156,
      "grad_norm": 0.6282304525375366,
      "learning_rate": 1.6365303983228512e-05,
      "loss": 0.617,
      "num_input_tokens_seen": 810176,
      "step": 1250
    },
    {
      "epoch": 0.6577568134171907,
      "grad_norm": 0.5200085043907166,
      "learning_rate": 1.6430817610062892e-05,
      "loss": 0.4234,
      "num_input_tokens_seen": 815680,
      "step": 1255
    },
    {
      "epoch": 0.660377358490566,
      "grad_norm": 0.6530146598815918,
      "learning_rate": 1.6496331236897275e-05,
      "loss": 0.6029,
      "num_input_tokens_seen": 818688,
      "step": 1260
    },
    {
      "epoch": 0.6629979035639413,
      "grad_norm": 0.40726855397224426,
      "learning_rate": 1.6561844863731658e-05,
      "loss": 0.6919,
      "num_input_tokens_seen": 823744,
      "step": 1265
    },
    {
      "epoch": 0.6656184486373166,
      "grad_norm": 1.0209078788757324,
      "learning_rate": 1.6627358490566038e-05,
      "loss": 0.5612,
      "num_input_tokens_seen": 826944,
      "step": 1270
    },
    {
      "epoch": 0.6682389937106918,
      "grad_norm": 0.5726354122161865,
      "learning_rate": 1.669287211740042e-05,
      "loss": 0.5478,
      "num_input_tokens_seen": 830304,
      "step": 1275
    },
    {
      "epoch": 0.6708595387840671,
      "grad_norm": 0.5796958208084106,
      "learning_rate": 1.6758385744234804e-05,
      "loss": 0.7062,
      "num_input_tokens_seen": 833440,
      "step": 1280
    },
    {
      "epoch": 0.6734800838574424,
      "grad_norm": 0.984205424785614,
      "learning_rate": 1.6823899371069183e-05,
      "loss": 0.6922,
      "num_input_tokens_seen": 835968,
      "step": 1285
    },
    {
      "epoch": 0.6761006289308176,
      "grad_norm": 0.9387558102607727,
      "learning_rate": 1.6889412997903563e-05,
      "loss": 0.7213,
      "num_input_tokens_seen": 838656,
      "step": 1290
    },
    {
      "epoch": 0.6787211740041929,
      "grad_norm": 0.4562090039253235,
      "learning_rate": 1.6954926624737946e-05,
      "loss": 0.5741,
      "num_input_tokens_seen": 841888,
      "step": 1295
    },
    {
      "epoch": 0.6813417190775681,
      "grad_norm": 0.33940449357032776,
      "learning_rate": 1.702044025157233e-05,
      "loss": 0.5399,
      "num_input_tokens_seen": 847232,
      "step": 1300
    },
    {
      "epoch": 0.6839622641509434,
      "grad_norm": 0.5394147038459778,
      "learning_rate": 1.708595387840671e-05,
      "loss": 0.6725,
      "num_input_tokens_seen": 850688,
      "step": 1305
    },
    {
      "epoch": 0.6865828092243187,
      "grad_norm": 0.7547639012336731,
      "learning_rate": 1.7151467505241092e-05,
      "loss": 0.8016,
      "num_input_tokens_seen": 853568,
      "step": 1310
    },
    {
      "epoch": 0.689203354297694,
      "grad_norm": 0.6436670422554016,
      "learning_rate": 1.7216981132075472e-05,
      "loss": 0.7021,
      "num_input_tokens_seen": 856832,
      "step": 1315
    },
    {
      "epoch": 0.6918238993710691,
      "grad_norm": 0.6070843935012817,
      "learning_rate": 1.7282494758909855e-05,
      "loss": 0.647,
      "num_input_tokens_seen": 859744,
      "step": 1320
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 0.5064834952354431,
      "learning_rate": 1.7348008385744234e-05,
      "loss": 0.6717,
      "num_input_tokens_seen": 862432,
      "step": 1325
    },
    {
      "epoch": 0.6970649895178197,
      "grad_norm": 0.8743461966514587,
      "learning_rate": 1.7413522012578618e-05,
      "loss": 0.5476,
      "num_input_tokens_seen": 865632,
      "step": 1330
    },
    {
      "epoch": 0.699685534591195,
      "grad_norm": 1.4016116857528687,
      "learning_rate": 1.7479035639413e-05,
      "loss": 0.5964,
      "num_input_tokens_seen": 869248,
      "step": 1335
    },
    {
      "epoch": 0.7023060796645703,
      "grad_norm": 0.7901773452758789,
      "learning_rate": 1.754454926624738e-05,
      "loss": 0.5867,
      "num_input_tokens_seen": 873504,
      "step": 1340
    },
    {
      "epoch": 0.7049266247379455,
      "grad_norm": 0.860251247882843,
      "learning_rate": 1.761006289308176e-05,
      "loss": 0.6425,
      "num_input_tokens_seen": 876032,
      "step": 1345
    },
    {
      "epoch": 0.7075471698113207,
      "grad_norm": 0.7910761833190918,
      "learning_rate": 1.7675576519916143e-05,
      "loss": 0.7156,
      "num_input_tokens_seen": 879616,
      "step": 1350
    },
    {
      "epoch": 0.710167714884696,
      "grad_norm": 0.6380378007888794,
      "learning_rate": 1.7741090146750526e-05,
      "loss": 0.6966,
      "num_input_tokens_seen": 881984,
      "step": 1355
    },
    {
      "epoch": 0.7127882599580713,
      "grad_norm": 0.6264626383781433,
      "learning_rate": 1.7806603773584906e-05,
      "loss": 0.6368,
      "num_input_tokens_seen": 886016,
      "step": 1360
    },
    {
      "epoch": 0.7154088050314465,
      "grad_norm": 0.8961204886436462,
      "learning_rate": 1.787211740041929e-05,
      "loss": 0.6513,
      "num_input_tokens_seen": 889024,
      "step": 1365
    },
    {
      "epoch": 0.7180293501048218,
      "grad_norm": 0.5881710052490234,
      "learning_rate": 1.7937631027253672e-05,
      "loss": 0.568,
      "num_input_tokens_seen": 891776,
      "step": 1370
    },
    {
      "epoch": 0.7206498951781971,
      "grad_norm": 0.632819414138794,
      "learning_rate": 1.8003144654088048e-05,
      "loss": 0.6276,
      "num_input_tokens_seen": 894624,
      "step": 1375
    },
    {
      "epoch": 0.7232704402515723,
      "grad_norm": 0.5862419605255127,
      "learning_rate": 1.806865828092243e-05,
      "loss": 0.5981,
      "num_input_tokens_seen": 897888,
      "step": 1380
    },
    {
      "epoch": 0.7258909853249476,
      "grad_norm": 0.5159581899642944,
      "learning_rate": 1.8134171907756814e-05,
      "loss": 0.6473,
      "num_input_tokens_seen": 903296,
      "step": 1385
    },
    {
      "epoch": 0.7285115303983228,
      "grad_norm": 0.6237426400184631,
      "learning_rate": 1.8199685534591197e-05,
      "loss": 0.9292,
      "num_input_tokens_seen": 905696,
      "step": 1390
    },
    {
      "epoch": 0.7311320754716981,
      "grad_norm": 0.6184260249137878,
      "learning_rate": 1.8265199161425577e-05,
      "loss": 0.8798,
      "num_input_tokens_seen": 909088,
      "step": 1395
    },
    {
      "epoch": 0.7337526205450734,
      "grad_norm": 1.0778169631958008,
      "learning_rate": 1.833071278825996e-05,
      "loss": 0.6245,
      "num_input_tokens_seen": 912096,
      "step": 1400
    },
    {
      "epoch": 0.7363731656184487,
      "grad_norm": 0.5778596997261047,
      "learning_rate": 1.839622641509434e-05,
      "loss": 0.6169,
      "num_input_tokens_seen": 915328,
      "step": 1405
    },
    {
      "epoch": 0.7389937106918238,
      "grad_norm": 0.4349699318408966,
      "learning_rate": 1.846174004192872e-05,
      "loss": 0.6396,
      "num_input_tokens_seen": 918080,
      "step": 1410
    },
    {
      "epoch": 0.7416142557651991,
      "grad_norm": 0.7611979842185974,
      "learning_rate": 1.8527253668763103e-05,
      "loss": 0.7331,
      "num_input_tokens_seen": 920928,
      "step": 1415
    },
    {
      "epoch": 0.7442348008385744,
      "grad_norm": 0.4905611574649811,
      "learning_rate": 1.8592767295597486e-05,
      "loss": 0.5721,
      "num_input_tokens_seen": 923808,
      "step": 1420
    },
    {
      "epoch": 0.7468553459119497,
      "grad_norm": 0.8107996582984924,
      "learning_rate": 1.865828092243187e-05,
      "loss": 0.5732,
      "num_input_tokens_seen": 926720,
      "step": 1425
    },
    {
      "epoch": 0.749475890985325,
      "grad_norm": 0.7209732532501221,
      "learning_rate": 1.872379454926625e-05,
      "loss": 0.5974,
      "num_input_tokens_seen": 930432,
      "step": 1430
    },
    {
      "epoch": 0.7520964360587002,
      "grad_norm": 0.5339469909667969,
      "learning_rate": 1.878930817610063e-05,
      "loss": 0.5804,
      "num_input_tokens_seen": 933504,
      "step": 1435
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 0.456218957901001,
      "learning_rate": 1.885482180293501e-05,
      "loss": 0.5809,
      "num_input_tokens_seen": 937120,
      "step": 1440
    },
    {
      "epoch": 0.7573375262054507,
      "grad_norm": 0.3926619589328766,
      "learning_rate": 1.892033542976939e-05,
      "loss": 0.5549,
      "num_input_tokens_seen": 939776,
      "step": 1445
    },
    {
      "epoch": 0.759958071278826,
      "grad_norm": 0.9909297227859497,
      "learning_rate": 1.8985849056603774e-05,
      "loss": 0.5979,
      "num_input_tokens_seen": 945120,
      "step": 1450
    },
    {
      "epoch": 0.7625786163522013,
      "grad_norm": 0.5796177387237549,
      "learning_rate": 1.9051362683438157e-05,
      "loss": 0.5749,
      "num_input_tokens_seen": 948064,
      "step": 1455
    },
    {
      "epoch": 0.7651991614255765,
      "grad_norm": 0.6503685116767883,
      "learning_rate": 1.9116876310272537e-05,
      "loss": 0.5294,
      "num_input_tokens_seen": 951968,
      "step": 1460
    },
    {
      "epoch": 0.7678197064989518,
      "grad_norm": 0.3883914053440094,
      "learning_rate": 1.918238993710692e-05,
      "loss": 0.5654,
      "num_input_tokens_seen": 955200,
      "step": 1465
    },
    {
      "epoch": 0.7704402515723271,
      "grad_norm": 0.8573329448699951,
      "learning_rate": 1.92479035639413e-05,
      "loss": 0.5823,
      "num_input_tokens_seen": 958112,
      "step": 1470
    },
    {
      "epoch": 0.7730607966457023,
      "grad_norm": 3.754142999649048,
      "learning_rate": 1.9313417190775683e-05,
      "loss": 0.7989,
      "num_input_tokens_seen": 960832,
      "step": 1475
    },
    {
      "epoch": 0.7756813417190775,
      "grad_norm": 0.7910865545272827,
      "learning_rate": 1.9378930817610062e-05,
      "loss": 0.4768,
      "num_input_tokens_seen": 964704,
      "step": 1480
    },
    {
      "epoch": 0.7783018867924528,
      "grad_norm": 0.4787701964378357,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 0.6152,
      "num_input_tokens_seen": 967552,
      "step": 1485
    },
    {
      "epoch": 0.7809224318658281,
      "grad_norm": 0.61073899269104,
      "learning_rate": 1.950995807127883e-05,
      "loss": 0.6248,
      "num_input_tokens_seen": 970496,
      "step": 1490
    },
    {
      "epoch": 0.7835429769392034,
      "grad_norm": 0.6898142099380493,
      "learning_rate": 1.9575471698113208e-05,
      "loss": 0.5753,
      "num_input_tokens_seen": 973952,
      "step": 1495
    },
    {
      "epoch": 0.7861635220125787,
      "grad_norm": 0.6848020553588867,
      "learning_rate": 1.9640985324947588e-05,
      "loss": 0.4859,
      "num_input_tokens_seen": 981536,
      "step": 1500
    },
    {
      "epoch": 0.7887840670859538,
      "grad_norm": 0.422893226146698,
      "learning_rate": 1.970649895178197e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 985152,
      "step": 1505
    },
    {
      "epoch": 0.7914046121593291,
      "grad_norm": 0.47893455624580383,
      "learning_rate": 1.9772012578616354e-05,
      "loss": 0.5479,
      "num_input_tokens_seen": 988896,
      "step": 1510
    },
    {
      "epoch": 0.7940251572327044,
      "grad_norm": 1.1559182405471802,
      "learning_rate": 1.9837526205450734e-05,
      "loss": 0.5042,
      "num_input_tokens_seen": 991936,
      "step": 1515
    },
    {
      "epoch": 0.7966457023060797,
      "grad_norm": 0.7239714860916138,
      "learning_rate": 1.9903039832285117e-05,
      "loss": 0.6688,
      "num_input_tokens_seen": 995776,
      "step": 1520
    },
    {
      "epoch": 0.799266247379455,
      "grad_norm": 0.8098926544189453,
      "learning_rate": 1.99685534591195e-05,
      "loss": 0.5571,
      "num_input_tokens_seen": 998336,
      "step": 1525
    },
    {
      "epoch": 0.8018867924528302,
      "grad_norm": 0.6282207369804382,
      "learning_rate": 2.003406708595388e-05,
      "loss": 0.5706,
      "num_input_tokens_seen": 1001248,
      "step": 1530
    },
    {
      "epoch": 0.8045073375262054,
      "grad_norm": 0.33340588212013245,
      "learning_rate": 2.009958071278826e-05,
      "loss": 0.5019,
      "num_input_tokens_seen": 1005152,
      "step": 1535
    },
    {
      "epoch": 0.8071278825995807,
      "grad_norm": 0.5584927201271057,
      "learning_rate": 2.0165094339622642e-05,
      "loss": 0.5686,
      "num_input_tokens_seen": 1008736,
      "step": 1540
    },
    {
      "epoch": 0.809748427672956,
      "grad_norm": 0.6922944784164429,
      "learning_rate": 2.0230607966457025e-05,
      "loss": 0.4861,
      "num_input_tokens_seen": 1011520,
      "step": 1545
    },
    {
      "epoch": 0.8123689727463312,
      "grad_norm": 0.3950366675853729,
      "learning_rate": 2.0296121593291405e-05,
      "loss": 0.5065,
      "num_input_tokens_seen": 1014176,
      "step": 1550
    },
    {
      "epoch": 0.8149895178197065,
      "grad_norm": 1.738277792930603,
      "learning_rate": 2.0361635220125788e-05,
      "loss": 1.0162,
      "num_input_tokens_seen": 1017408,
      "step": 1555
    },
    {
      "epoch": 0.8176100628930818,
      "grad_norm": 0.5990610718727112,
      "learning_rate": 2.0427148846960168e-05,
      "loss": 1.4388,
      "num_input_tokens_seen": 1019808,
      "step": 1560
    },
    {
      "epoch": 0.820230607966457,
      "grad_norm": 0.9375083446502686,
      "learning_rate": 2.049266247379455e-05,
      "loss": 0.5912,
      "num_input_tokens_seen": 1022496,
      "step": 1565
    },
    {
      "epoch": 0.8228511530398323,
      "grad_norm": 0.634951651096344,
      "learning_rate": 2.055817610062893e-05,
      "loss": 0.6919,
      "num_input_tokens_seen": 1026272,
      "step": 1570
    },
    {
      "epoch": 0.8254716981132075,
      "grad_norm": 0.841395914554596,
      "learning_rate": 2.0623689727463314e-05,
      "loss": 0.5659,
      "num_input_tokens_seen": 1029280,
      "step": 1575
    },
    {
      "epoch": 0.8280922431865828,
      "grad_norm": 1.2462862730026245,
      "learning_rate": 2.0689203354297697e-05,
      "loss": 0.6666,
      "num_input_tokens_seen": 1031488,
      "step": 1580
    },
    {
      "epoch": 0.8307127882599581,
      "grad_norm": 0.5505123734474182,
      "learning_rate": 2.0754716981132076e-05,
      "loss": 0.7772,
      "num_input_tokens_seen": 1034144,
      "step": 1585
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.7328715920448303,
      "learning_rate": 2.0820230607966456e-05,
      "loss": 0.5406,
      "num_input_tokens_seen": 1036960,
      "step": 1590
    },
    {
      "epoch": 0.8359538784067087,
      "grad_norm": 0.674143373966217,
      "learning_rate": 2.088574423480084e-05,
      "loss": 0.6454,
      "num_input_tokens_seen": 1041568,
      "step": 1595
    },
    {
      "epoch": 0.8385744234800838,
      "grad_norm": 0.7263755202293396,
      "learning_rate": 2.0951257861635222e-05,
      "loss": 0.666,
      "num_input_tokens_seen": 1044768,
      "step": 1600
    },
    {
      "epoch": 0.8411949685534591,
      "grad_norm": 0.5555549263954163,
      "learning_rate": 2.1016771488469602e-05,
      "loss": 0.6441,
      "num_input_tokens_seen": 1048704,
      "step": 1605
    },
    {
      "epoch": 0.8438155136268344,
      "grad_norm": 0.6896020770072937,
      "learning_rate": 2.1082285115303985e-05,
      "loss": 0.3294,
      "num_input_tokens_seen": 1051424,
      "step": 1610
    },
    {
      "epoch": 0.8464360587002097,
      "grad_norm": 0.5721293091773987,
      "learning_rate": 2.1147798742138368e-05,
      "loss": 0.5952,
      "num_input_tokens_seen": 1054784,
      "step": 1615
    },
    {
      "epoch": 0.8490566037735849,
      "grad_norm": 0.7262080311775208,
      "learning_rate": 2.1213312368972748e-05,
      "loss": 0.6641,
      "num_input_tokens_seen": 1057568,
      "step": 1620
    },
    {
      "epoch": 0.8516771488469602,
      "grad_norm": 0.8170379400253296,
      "learning_rate": 2.1278825995807127e-05,
      "loss": 0.575,
      "num_input_tokens_seen": 1059936,
      "step": 1625
    },
    {
      "epoch": 0.8542976939203354,
      "grad_norm": 0.541822075843811,
      "learning_rate": 2.134433962264151e-05,
      "loss": 0.5411,
      "num_input_tokens_seen": 1063168,
      "step": 1630
    },
    {
      "epoch": 0.8569182389937107,
      "grad_norm": 1.3195314407348633,
      "learning_rate": 2.1409853249475894e-05,
      "loss": 0.6216,
      "num_input_tokens_seen": 1065664,
      "step": 1635
    },
    {
      "epoch": 0.859538784067086,
      "grad_norm": 1.0201014280319214,
      "learning_rate": 2.1475366876310273e-05,
      "loss": 0.7231,
      "num_input_tokens_seen": 1069376,
      "step": 1640
    },
    {
      "epoch": 0.8621593291404612,
      "grad_norm": 0.5003520250320435,
      "learning_rate": 2.1540880503144656e-05,
      "loss": 0.5899,
      "num_input_tokens_seen": 1072736,
      "step": 1645
    },
    {
      "epoch": 0.8647798742138365,
      "grad_norm": 0.5794950723648071,
      "learning_rate": 2.160639412997904e-05,
      "loss": 0.5131,
      "num_input_tokens_seen": 1076608,
      "step": 1650
    },
    {
      "epoch": 0.8674004192872118,
      "grad_norm": 0.4041418135166168,
      "learning_rate": 2.1671907756813416e-05,
      "loss": 0.6227,
      "num_input_tokens_seen": 1080640,
      "step": 1655
    },
    {
      "epoch": 0.870020964360587,
      "grad_norm": 0.47745081782341003,
      "learning_rate": 2.17374213836478e-05,
      "loss": 0.6043,
      "num_input_tokens_seen": 1083808,
      "step": 1660
    },
    {
      "epoch": 0.8726415094339622,
      "grad_norm": 0.742652416229248,
      "learning_rate": 2.1802935010482182e-05,
      "loss": 0.6571,
      "num_input_tokens_seen": 1086400,
      "step": 1665
    },
    {
      "epoch": 0.8752620545073375,
      "grad_norm": 0.5901350975036621,
      "learning_rate": 2.186844863731656e-05,
      "loss": 0.6125,
      "num_input_tokens_seen": 1089504,
      "step": 1670
    },
    {
      "epoch": 0.8778825995807128,
      "grad_norm": 0.5809617042541504,
      "learning_rate": 2.1933962264150945e-05,
      "loss": 0.6181,
      "num_input_tokens_seen": 1092160,
      "step": 1675
    },
    {
      "epoch": 0.8805031446540881,
      "grad_norm": 0.37598398327827454,
      "learning_rate": 2.1999475890985328e-05,
      "loss": 0.4244,
      "num_input_tokens_seen": 1095328,
      "step": 1680
    },
    {
      "epoch": 0.8831236897274634,
      "grad_norm": 0.7794609665870667,
      "learning_rate": 2.2064989517819707e-05,
      "loss": 0.7453,
      "num_input_tokens_seen": 1098560,
      "step": 1685
    },
    {
      "epoch": 0.8857442348008385,
      "grad_norm": 1.0715869665145874,
      "learning_rate": 2.2130503144654087e-05,
      "loss": 0.6501,
      "num_input_tokens_seen": 1101056,
      "step": 1690
    },
    {
      "epoch": 0.8883647798742138,
      "grad_norm": 0.7445118427276611,
      "learning_rate": 2.219601677148847e-05,
      "loss": 0.5642,
      "num_input_tokens_seen": 1104160,
      "step": 1695
    },
    {
      "epoch": 0.8909853249475891,
      "grad_norm": 0.3598496913909912,
      "learning_rate": 2.2261530398322853e-05,
      "loss": 0.7167,
      "num_input_tokens_seen": 1107232,
      "step": 1700
    },
    {
      "epoch": 0.8936058700209644,
      "grad_norm": 0.9485227465629578,
      "learning_rate": 2.2327044025157233e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 1112416,
      "step": 1705
    },
    {
      "epoch": 0.8962264150943396,
      "grad_norm": 0.5247765183448792,
      "learning_rate": 2.2392557651991616e-05,
      "loss": 0.6043,
      "num_input_tokens_seen": 1115296,
      "step": 1710
    },
    {
      "epoch": 0.8988469601677149,
      "grad_norm": 0.4724593758583069,
      "learning_rate": 2.2458071278825996e-05,
      "loss": 0.5683,
      "num_input_tokens_seen": 1119008,
      "step": 1715
    },
    {
      "epoch": 0.9014675052410901,
      "grad_norm": 2.1759462356567383,
      "learning_rate": 2.252358490566038e-05,
      "loss": 0.729,
      "num_input_tokens_seen": 1122592,
      "step": 1720
    },
    {
      "epoch": 0.9040880503144654,
      "grad_norm": 0.44263458251953125,
      "learning_rate": 2.258909853249476e-05,
      "loss": 0.7085,
      "num_input_tokens_seen": 1127520,
      "step": 1725
    },
    {
      "epoch": 0.9067085953878407,
      "grad_norm": 0.30781617760658264,
      "learning_rate": 2.265461215932914e-05,
      "loss": 0.5525,
      "num_input_tokens_seen": 1130592,
      "step": 1730
    },
    {
      "epoch": 0.9093291404612159,
      "grad_norm": 1.9412899017333984,
      "learning_rate": 2.2720125786163524e-05,
      "loss": 0.588,
      "num_input_tokens_seen": 1134112,
      "step": 1735
    },
    {
      "epoch": 0.9119496855345912,
      "grad_norm": 0.7958773970603943,
      "learning_rate": 2.2785639412997904e-05,
      "loss": 0.5936,
      "num_input_tokens_seen": 1137856,
      "step": 1740
    },
    {
      "epoch": 0.9145702306079665,
      "grad_norm": 1.6381877660751343,
      "learning_rate": 2.2851153039832284e-05,
      "loss": 0.6002,
      "num_input_tokens_seen": 1140032,
      "step": 1745
    },
    {
      "epoch": 0.9171907756813418,
      "grad_norm": 0.370511919260025,
      "learning_rate": 2.2916666666666667e-05,
      "loss": 0.6424,
      "num_input_tokens_seen": 1143872,
      "step": 1750
    },
    {
      "epoch": 0.9198113207547169,
      "grad_norm": 0.6733387112617493,
      "learning_rate": 2.298218029350105e-05,
      "loss": 0.614,
      "num_input_tokens_seen": 1147232,
      "step": 1755
    },
    {
      "epoch": 0.9224318658280922,
      "grad_norm": 0.4406244158744812,
      "learning_rate": 2.304769392033543e-05,
      "loss": 0.5733,
      "num_input_tokens_seen": 1149792,
      "step": 1760
    },
    {
      "epoch": 0.9250524109014675,
      "grad_norm": 1.2498224973678589,
      "learning_rate": 2.3113207547169813e-05,
      "loss": 0.7889,
      "num_input_tokens_seen": 1152832,
      "step": 1765
    },
    {
      "epoch": 0.9276729559748428,
      "grad_norm": 0.9262797832489014,
      "learning_rate": 2.3178721174004196e-05,
      "loss": 0.733,
      "num_input_tokens_seen": 1155424,
      "step": 1770
    },
    {
      "epoch": 0.9302935010482181,
      "grad_norm": 0.809870183467865,
      "learning_rate": 2.3244234800838576e-05,
      "loss": 0.5324,
      "num_input_tokens_seen": 1158656,
      "step": 1775
    },
    {
      "epoch": 0.9329140461215933,
      "grad_norm": 1.1248623132705688,
      "learning_rate": 2.3309748427672955e-05,
      "loss": 0.705,
      "num_input_tokens_seen": 1161504,
      "step": 1780
    },
    {
      "epoch": 0.9355345911949685,
      "grad_norm": 0.4592053294181824,
      "learning_rate": 2.3375262054507338e-05,
      "loss": 0.5746,
      "num_input_tokens_seen": 1164480,
      "step": 1785
    },
    {
      "epoch": 0.9381551362683438,
      "grad_norm": 0.6326702833175659,
      "learning_rate": 2.344077568134172e-05,
      "loss": 0.5352,
      "num_input_tokens_seen": 1168128,
      "step": 1790
    },
    {
      "epoch": 0.9407756813417191,
      "grad_norm": 0.602389395236969,
      "learning_rate": 2.35062893081761e-05,
      "loss": 0.6326,
      "num_input_tokens_seen": 1171648,
      "step": 1795
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 0.27822527289390564,
      "learning_rate": 2.3571802935010484e-05,
      "loss": 0.6489,
      "num_input_tokens_seen": 1174656,
      "step": 1800
    },
    {
      "epoch": 0.9460167714884696,
      "grad_norm": 0.8491811156272888,
      "learning_rate": 2.3637316561844867e-05,
      "loss": 0.7523,
      "num_input_tokens_seen": 1177728,
      "step": 1805
    },
    {
      "epoch": 0.9486373165618449,
      "grad_norm": 0.3746519982814789,
      "learning_rate": 2.3702830188679247e-05,
      "loss": 0.5762,
      "num_input_tokens_seen": 1182208,
      "step": 1810
    },
    {
      "epoch": 0.9512578616352201,
      "grad_norm": 0.5306198000907898,
      "learning_rate": 2.3768343815513627e-05,
      "loss": 0.5332,
      "num_input_tokens_seen": 1184832,
      "step": 1815
    },
    {
      "epoch": 0.9538784067085954,
      "grad_norm": 0.36500802636146545,
      "learning_rate": 2.383385744234801e-05,
      "loss": 0.5865,
      "num_input_tokens_seen": 1187808,
      "step": 1820
    },
    {
      "epoch": 0.9564989517819706,
      "grad_norm": 0.5377168655395508,
      "learning_rate": 2.3899371069182393e-05,
      "loss": 0.5615,
      "num_input_tokens_seen": 1191136,
      "step": 1825
    },
    {
      "epoch": 0.9591194968553459,
      "grad_norm": 0.5205085873603821,
      "learning_rate": 2.3964884696016772e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 1194944,
      "step": 1830
    },
    {
      "epoch": 0.9617400419287212,
      "grad_norm": 0.8028524518013,
      "learning_rate": 2.4030398322851155e-05,
      "loss": 0.595,
      "num_input_tokens_seen": 1197504,
      "step": 1835
    },
    {
      "epoch": 0.9643605870020965,
      "grad_norm": 0.3692125380039215,
      "learning_rate": 2.4095911949685535e-05,
      "loss": 0.5285,
      "num_input_tokens_seen": 1201056,
      "step": 1840
    },
    {
      "epoch": 0.9669811320754716,
      "grad_norm": 0.3146451413631439,
      "learning_rate": 2.4161425576519918e-05,
      "loss": 0.5721,
      "num_input_tokens_seen": 1203968,
      "step": 1845
    },
    {
      "epoch": 0.9696016771488469,
      "grad_norm": 1.6181031465530396,
      "learning_rate": 2.4226939203354298e-05,
      "loss": 0.759,
      "num_input_tokens_seen": 1207712,
      "step": 1850
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 0.5237797498703003,
      "learning_rate": 2.429245283018868e-05,
      "loss": 0.5208,
      "num_input_tokens_seen": 1210912,
      "step": 1855
    },
    {
      "epoch": 0.9748427672955975,
      "grad_norm": 0.5423349738121033,
      "learning_rate": 2.4357966457023064e-05,
      "loss": 0.6451,
      "num_input_tokens_seen": 1213408,
      "step": 1860
    },
    {
      "epoch": 0.9774633123689728,
      "grad_norm": 0.3520890474319458,
      "learning_rate": 2.4423480083857444e-05,
      "loss": 0.7603,
      "num_input_tokens_seen": 1217184,
      "step": 1865
    },
    {
      "epoch": 0.980083857442348,
      "grad_norm": 0.4013573229312897,
      "learning_rate": 2.4488993710691823e-05,
      "loss": 0.6534,
      "num_input_tokens_seen": 1220576,
      "step": 1870
    },
    {
      "epoch": 0.9827044025157232,
      "grad_norm": 0.9330726861953735,
      "learning_rate": 2.4554507337526206e-05,
      "loss": 0.6676,
      "num_input_tokens_seen": 1223552,
      "step": 1875
    },
    {
      "epoch": 0.9853249475890985,
      "grad_norm": 0.6566826105117798,
      "learning_rate": 2.4620020964360586e-05,
      "loss": 0.7152,
      "num_input_tokens_seen": 1226592,
      "step": 1880
    },
    {
      "epoch": 0.9879454926624738,
      "grad_norm": 0.4300500154495239,
      "learning_rate": 2.468553459119497e-05,
      "loss": 0.6142,
      "num_input_tokens_seen": 1229504,
      "step": 1885
    },
    {
      "epoch": 0.9905660377358491,
      "grad_norm": 0.8413358926773071,
      "learning_rate": 2.4751048218029352e-05,
      "loss": 0.5731,
      "num_input_tokens_seen": 1232896,
      "step": 1890
    },
    {
      "epoch": 0.9931865828092243,
      "grad_norm": 0.7106512784957886,
      "learning_rate": 2.4816561844863735e-05,
      "loss": 0.5099,
      "num_input_tokens_seen": 1235840,
      "step": 1895
    },
    {
      "epoch": 0.9958071278825996,
      "grad_norm": 0.5236139297485352,
      "learning_rate": 2.488207547169811e-05,
      "loss": 0.5991,
      "num_input_tokens_seen": 1239040,
      "step": 1900
    },
    {
      "epoch": 0.9984276729559748,
      "grad_norm": 0.5320724248886108,
      "learning_rate": 2.4947589098532495e-05,
      "loss": 0.6243,
      "num_input_tokens_seen": 1242112,
      "step": 1905
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.580083429813385,
      "eval_runtime": 9.6947,
      "eval_samples_per_second": 87.471,
      "eval_steps_per_second": 21.868,
      "num_input_tokens_seen": 1246616,
      "step": 1908
    },
    {
      "epoch": 1.00104821802935,
      "grad_norm": 0.5779690742492676,
      "learning_rate": 2.5013102725366878e-05,
      "loss": 0.504,
      "num_input_tokens_seen": 1248024,
      "step": 1910
    },
    {
      "epoch": 1.0036687631027255,
      "grad_norm": 0.5185136198997498,
      "learning_rate": 2.507861635220126e-05,
      "loss": 0.5035,
      "num_input_tokens_seen": 1250232,
      "step": 1915
    },
    {
      "epoch": 1.0062893081761006,
      "grad_norm": 0.9217631816864014,
      "learning_rate": 2.514412997903564e-05,
      "loss": 0.7817,
      "num_input_tokens_seen": 1253272,
      "step": 1920
    },
    {
      "epoch": 1.0089098532494758,
      "grad_norm": 0.8817874789237976,
      "learning_rate": 2.520964360587002e-05,
      "loss": 0.5316,
      "num_input_tokens_seen": 1255928,
      "step": 1925
    },
    {
      "epoch": 1.0115303983228512,
      "grad_norm": 0.7305052280426025,
      "learning_rate": 2.5275157232704407e-05,
      "loss": 0.5763,
      "num_input_tokens_seen": 1260376,
      "step": 1930
    },
    {
      "epoch": 1.0141509433962264,
      "grad_norm": 0.37958380579948425,
      "learning_rate": 2.5340670859538783e-05,
      "loss": 0.6573,
      "num_input_tokens_seen": 1263864,
      "step": 1935
    },
    {
      "epoch": 1.0167714884696017,
      "grad_norm": 0.7676419019699097,
      "learning_rate": 2.540618448637317e-05,
      "loss": 0.6415,
      "num_input_tokens_seen": 1267224,
      "step": 1940
    },
    {
      "epoch": 1.019392033542977,
      "grad_norm": 0.30925124883651733,
      "learning_rate": 2.547169811320755e-05,
      "loss": 0.5956,
      "num_input_tokens_seen": 1270520,
      "step": 1945
    },
    {
      "epoch": 1.0220125786163523,
      "grad_norm": 0.6236841678619385,
      "learning_rate": 2.553721174004193e-05,
      "loss": 0.4737,
      "num_input_tokens_seen": 1273144,
      "step": 1950
    },
    {
      "epoch": 1.0246331236897275,
      "grad_norm": 0.5318701267242432,
      "learning_rate": 2.5602725366876312e-05,
      "loss": 0.6786,
      "num_input_tokens_seen": 1275608,
      "step": 1955
    },
    {
      "epoch": 1.0272536687631026,
      "grad_norm": 1.0985840559005737,
      "learning_rate": 2.566823899371069e-05,
      "loss": 0.6657,
      "num_input_tokens_seen": 1278360,
      "step": 1960
    },
    {
      "epoch": 1.029874213836478,
      "grad_norm": 0.334078311920166,
      "learning_rate": 2.5733752620545075e-05,
      "loss": 0.6011,
      "num_input_tokens_seen": 1281720,
      "step": 1965
    },
    {
      "epoch": 1.0324947589098532,
      "grad_norm": 0.4789329171180725,
      "learning_rate": 2.5799266247379454e-05,
      "loss": 0.6526,
      "num_input_tokens_seen": 1284216,
      "step": 1970
    },
    {
      "epoch": 1.0351153039832286,
      "grad_norm": 0.5000085234642029,
      "learning_rate": 2.586477987421384e-05,
      "loss": 0.4763,
      "num_input_tokens_seen": 1287576,
      "step": 1975
    },
    {
      "epoch": 1.0377358490566038,
      "grad_norm": 0.593473494052887,
      "learning_rate": 2.593029350104822e-05,
      "loss": 0.5968,
      "num_input_tokens_seen": 1290456,
      "step": 1980
    },
    {
      "epoch": 1.040356394129979,
      "grad_norm": 0.4495288133621216,
      "learning_rate": 2.59958071278826e-05,
      "loss": 0.6838,
      "num_input_tokens_seen": 1293752,
      "step": 1985
    },
    {
      "epoch": 1.0429769392033543,
      "grad_norm": 0.669816792011261,
      "learning_rate": 2.6061320754716983e-05,
      "loss": 0.4954,
      "num_input_tokens_seen": 1297368,
      "step": 1990
    },
    {
      "epoch": 1.0455974842767295,
      "grad_norm": 0.557707667350769,
      "learning_rate": 2.6126834381551363e-05,
      "loss": 0.585,
      "num_input_tokens_seen": 1300280,
      "step": 1995
    },
    {
      "epoch": 1.0482180293501049,
      "grad_norm": 0.7507773041725159,
      "learning_rate": 2.6192348008385746e-05,
      "loss": 0.6054,
      "num_input_tokens_seen": 1304504,
      "step": 2000
    },
    {
      "epoch": 1.05083857442348,
      "grad_norm": 0.6109243631362915,
      "learning_rate": 2.6257861635220126e-05,
      "loss": 0.6075,
      "num_input_tokens_seen": 1308120,
      "step": 2005
    },
    {
      "epoch": 1.0534591194968554,
      "grad_norm": 0.45269179344177246,
      "learning_rate": 2.6323375262054505e-05,
      "loss": 0.605,
      "num_input_tokens_seen": 1311160,
      "step": 2010
    },
    {
      "epoch": 1.0560796645702306,
      "grad_norm": 0.5031194686889648,
      "learning_rate": 2.6388888888888892e-05,
      "loss": 0.6763,
      "num_input_tokens_seen": 1313784,
      "step": 2015
    },
    {
      "epoch": 1.0587002096436058,
      "grad_norm": 0.5019195079803467,
      "learning_rate": 2.645440251572327e-05,
      "loss": 0.6244,
      "num_input_tokens_seen": 1317528,
      "step": 2020
    },
    {
      "epoch": 1.0613207547169812,
      "grad_norm": 0.43806663155555725,
      "learning_rate": 2.6519916142557655e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 1321400,
      "step": 2025
    },
    {
      "epoch": 1.0639412997903563,
      "grad_norm": 0.5735514760017395,
      "learning_rate": 2.6585429769392034e-05,
      "loss": 0.5605,
      "num_input_tokens_seen": 1325304,
      "step": 2030
    },
    {
      "epoch": 1.0665618448637317,
      "grad_norm": 0.5005062818527222,
      "learning_rate": 2.6650943396226417e-05,
      "loss": 0.5354,
      "num_input_tokens_seen": 1327928,
      "step": 2035
    },
    {
      "epoch": 1.069182389937107,
      "grad_norm": 0.6603549718856812,
      "learning_rate": 2.6716457023060797e-05,
      "loss": 0.5847,
      "num_input_tokens_seen": 1331224,
      "step": 2040
    },
    {
      "epoch": 1.0718029350104823,
      "grad_norm": 0.5071974992752075,
      "learning_rate": 2.6781970649895177e-05,
      "loss": 0.5494,
      "num_input_tokens_seen": 1334456,
      "step": 2045
    },
    {
      "epoch": 1.0744234800838575,
      "grad_norm": 0.6550766229629517,
      "learning_rate": 2.6847484276729563e-05,
      "loss": 0.6027,
      "num_input_tokens_seen": 1337592,
      "step": 2050
    },
    {
      "epoch": 1.0770440251572326,
      "grad_norm": 1.471375823020935,
      "learning_rate": 2.6912997903563943e-05,
      "loss": 0.6298,
      "num_input_tokens_seen": 1339960,
      "step": 2055
    },
    {
      "epoch": 1.079664570230608,
      "grad_norm": 0.6025872230529785,
      "learning_rate": 2.6978511530398326e-05,
      "loss": 0.654,
      "num_input_tokens_seen": 1343864,
      "step": 2060
    },
    {
      "epoch": 1.0822851153039832,
      "grad_norm": 0.34765902161598206,
      "learning_rate": 2.7044025157232706e-05,
      "loss": 0.5718,
      "num_input_tokens_seen": 1347032,
      "step": 2065
    },
    {
      "epoch": 1.0849056603773586,
      "grad_norm": 0.4139513075351715,
      "learning_rate": 2.710953878406709e-05,
      "loss": 0.6091,
      "num_input_tokens_seen": 1350072,
      "step": 2070
    },
    {
      "epoch": 1.0875262054507338,
      "grad_norm": 0.651174008846283,
      "learning_rate": 2.717505241090147e-05,
      "loss": 0.5671,
      "num_input_tokens_seen": 1353496,
      "step": 2075
    },
    {
      "epoch": 1.090146750524109,
      "grad_norm": 0.5352675914764404,
      "learning_rate": 2.7240566037735848e-05,
      "loss": 0.6144,
      "num_input_tokens_seen": 1355832,
      "step": 2080
    },
    {
      "epoch": 1.0927672955974843,
      "grad_norm": 0.4492761492729187,
      "learning_rate": 2.7306079664570235e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 1360312,
      "step": 2085
    },
    {
      "epoch": 1.0953878406708595,
      "grad_norm": 0.502896785736084,
      "learning_rate": 2.737159329140461e-05,
      "loss": 0.571,
      "num_input_tokens_seen": 1363992,
      "step": 2090
    },
    {
      "epoch": 1.0980083857442349,
      "grad_norm": 0.5257335901260376,
      "learning_rate": 2.7437106918238997e-05,
      "loss": 0.4545,
      "num_input_tokens_seen": 1367352,
      "step": 2095
    },
    {
      "epoch": 1.10062893081761,
      "grad_norm": 0.6132773160934448,
      "learning_rate": 2.7502620545073377e-05,
      "loss": 0.671,
      "num_input_tokens_seen": 1370232,
      "step": 2100
    },
    {
      "epoch": 1.1032494758909852,
      "grad_norm": 0.7401959896087646,
      "learning_rate": 2.7568134171907757e-05,
      "loss": 0.6112,
      "num_input_tokens_seen": 1375448,
      "step": 2105
    },
    {
      "epoch": 1.1058700209643606,
      "grad_norm": 0.4497572183609009,
      "learning_rate": 2.763364779874214e-05,
      "loss": 0.52,
      "num_input_tokens_seen": 1378840,
      "step": 2110
    },
    {
      "epoch": 1.1084905660377358,
      "grad_norm": 0.5010729432106018,
      "learning_rate": 2.769916142557652e-05,
      "loss": 0.539,
      "num_input_tokens_seen": 1382616,
      "step": 2115
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.8764589428901672,
      "learning_rate": 2.7764675052410906e-05,
      "loss": 0.5745,
      "num_input_tokens_seen": 1385016,
      "step": 2120
    },
    {
      "epoch": 1.1137316561844863,
      "grad_norm": 0.5166593790054321,
      "learning_rate": 2.7830188679245282e-05,
      "loss": 0.5685,
      "num_input_tokens_seen": 1389368,
      "step": 2125
    },
    {
      "epoch": 1.1163522012578617,
      "grad_norm": 0.33987998962402344,
      "learning_rate": 2.789570230607967e-05,
      "loss": 0.4333,
      "num_input_tokens_seen": 1393176,
      "step": 2130
    },
    {
      "epoch": 1.118972746331237,
      "grad_norm": 0.8017078042030334,
      "learning_rate": 2.796121593291405e-05,
      "loss": 0.5461,
      "num_input_tokens_seen": 1396504,
      "step": 2135
    },
    {
      "epoch": 1.121593291404612,
      "grad_norm": 0.600069522857666,
      "learning_rate": 2.8026729559748428e-05,
      "loss": 0.5694,
      "num_input_tokens_seen": 1400088,
      "step": 2140
    },
    {
      "epoch": 1.1242138364779874,
      "grad_norm": 0.5930898785591125,
      "learning_rate": 2.809224318658281e-05,
      "loss": 0.5504,
      "num_input_tokens_seen": 1402456,
      "step": 2145
    },
    {
      "epoch": 1.1268343815513626,
      "grad_norm": 0.5308684706687927,
      "learning_rate": 2.815775681341719e-05,
      "loss": 0.5394,
      "num_input_tokens_seen": 1405464,
      "step": 2150
    },
    {
      "epoch": 1.129454926624738,
      "grad_norm": 0.7842283844947815,
      "learning_rate": 2.8223270440251577e-05,
      "loss": 0.5076,
      "num_input_tokens_seen": 1407800,
      "step": 2155
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 0.4485495090484619,
      "learning_rate": 2.8288784067085954e-05,
      "loss": 0.4766,
      "num_input_tokens_seen": 1410520,
      "step": 2160
    },
    {
      "epoch": 1.1346960167714886,
      "grad_norm": 0.4265313148498535,
      "learning_rate": 2.8354297693920333e-05,
      "loss": 0.5823,
      "num_input_tokens_seen": 1414808,
      "step": 2165
    },
    {
      "epoch": 1.1373165618448637,
      "grad_norm": 0.7797902226448059,
      "learning_rate": 2.841981132075472e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 1417944,
      "step": 2170
    },
    {
      "epoch": 1.139937106918239,
      "grad_norm": 0.659886360168457,
      "learning_rate": 2.84853249475891e-05,
      "loss": 0.4394,
      "num_input_tokens_seen": 1421272,
      "step": 2175
    },
    {
      "epoch": 1.1425576519916143,
      "grad_norm": 0.43917426466941833,
      "learning_rate": 2.8550838574423482e-05,
      "loss": 0.7446,
      "num_input_tokens_seen": 1423704,
      "step": 2180
    },
    {
      "epoch": 1.1451781970649895,
      "grad_norm": 0.48971083760261536,
      "learning_rate": 2.8616352201257862e-05,
      "loss": 0.505,
      "num_input_tokens_seen": 1427000,
      "step": 2185
    },
    {
      "epoch": 1.1477987421383649,
      "grad_norm": 0.5635573267936707,
      "learning_rate": 2.868186582809225e-05,
      "loss": 0.6402,
      "num_input_tokens_seen": 1429688,
      "step": 2190
    },
    {
      "epoch": 1.15041928721174,
      "grad_norm": 0.5188326239585876,
      "learning_rate": 2.8747379454926625e-05,
      "loss": 0.5283,
      "num_input_tokens_seen": 1433496,
      "step": 2195
    },
    {
      "epoch": 1.1530398322851152,
      "grad_norm": 0.43896758556365967,
      "learning_rate": 2.8812893081761005e-05,
      "loss": 0.4728,
      "num_input_tokens_seen": 1436440,
      "step": 2200
    },
    {
      "epoch": 1.1556603773584906,
      "grad_norm": 0.45821648836135864,
      "learning_rate": 2.887840670859539e-05,
      "loss": 0.5192,
      "num_input_tokens_seen": 1439064,
      "step": 2205
    },
    {
      "epoch": 1.1582809224318658,
      "grad_norm": 0.6556364893913269,
      "learning_rate": 2.894392033542977e-05,
      "loss": 0.4082,
      "num_input_tokens_seen": 1444600,
      "step": 2210
    },
    {
      "epoch": 1.1609014675052411,
      "grad_norm": 1.1144293546676636,
      "learning_rate": 2.9009433962264154e-05,
      "loss": 0.6844,
      "num_input_tokens_seen": 1447160,
      "step": 2215
    },
    {
      "epoch": 1.1635220125786163,
      "grad_norm": 0.38703978061676025,
      "learning_rate": 2.9074947589098533e-05,
      "loss": 0.5704,
      "num_input_tokens_seen": 1450840,
      "step": 2220
    },
    {
      "epoch": 1.1661425576519917,
      "grad_norm": 0.23840254545211792,
      "learning_rate": 2.9140461215932913e-05,
      "loss": 0.4536,
      "num_input_tokens_seen": 1454552,
      "step": 2225
    },
    {
      "epoch": 1.1687631027253669,
      "grad_norm": 0.7482790946960449,
      "learning_rate": 2.9205974842767296e-05,
      "loss": 0.4847,
      "num_input_tokens_seen": 1457240,
      "step": 2230
    },
    {
      "epoch": 1.171383647798742,
      "grad_norm": 0.46493926644325256,
      "learning_rate": 2.9271488469601676e-05,
      "loss": 0.5752,
      "num_input_tokens_seen": 1460472,
      "step": 2235
    },
    {
      "epoch": 1.1740041928721174,
      "grad_norm": 0.39177677035331726,
      "learning_rate": 2.9337002096436062e-05,
      "loss": 0.5368,
      "num_input_tokens_seen": 1463832,
      "step": 2240
    },
    {
      "epoch": 1.1766247379454926,
      "grad_norm": 0.560732901096344,
      "learning_rate": 2.9402515723270442e-05,
      "loss": 0.5848,
      "num_input_tokens_seen": 1466552,
      "step": 2245
    },
    {
      "epoch": 1.179245283018868,
      "grad_norm": 0.8950576186180115,
      "learning_rate": 2.9468029350104825e-05,
      "loss": 0.6607,
      "num_input_tokens_seen": 1469144,
      "step": 2250
    },
    {
      "epoch": 1.1818658280922432,
      "grad_norm": 0.7951366901397705,
      "learning_rate": 2.9533542976939205e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 1471992,
      "step": 2255
    },
    {
      "epoch": 1.1844863731656186,
      "grad_norm": 1.0079824924468994,
      "learning_rate": 2.9599056603773584e-05,
      "loss": 0.5715,
      "num_input_tokens_seen": 1475544,
      "step": 2260
    },
    {
      "epoch": 1.1871069182389937,
      "grad_norm": 1.3613295555114746,
      "learning_rate": 2.9664570230607968e-05,
      "loss": 0.5322,
      "num_input_tokens_seen": 1478456,
      "step": 2265
    },
    {
      "epoch": 1.189727463312369,
      "grad_norm": 0.5588165521621704,
      "learning_rate": 2.9730083857442347e-05,
      "loss": 0.4584,
      "num_input_tokens_seen": 1481208,
      "step": 2270
    },
    {
      "epoch": 1.1923480083857443,
      "grad_norm": 0.5738652348518372,
      "learning_rate": 2.9795597484276734e-05,
      "loss": 0.4515,
      "num_input_tokens_seen": 1486776,
      "step": 2275
    },
    {
      "epoch": 1.1949685534591195,
      "grad_norm": 0.38806790113449097,
      "learning_rate": 2.9861111111111113e-05,
      "loss": 0.602,
      "num_input_tokens_seen": 1489176,
      "step": 2280
    },
    {
      "epoch": 1.1975890985324948,
      "grad_norm": 0.4297347962856293,
      "learning_rate": 2.9926624737945496e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 1492312,
      "step": 2285
    },
    {
      "epoch": 1.20020964360587,
      "grad_norm": 0.4591110348701477,
      "learning_rate": 2.9992138364779876e-05,
      "loss": 0.5391,
      "num_input_tokens_seen": 1495576,
      "step": 2290
    },
    {
      "epoch": 1.2028301886792452,
      "grad_norm": 1.0901964902877808,
      "learning_rate": 3.0057651991614256e-05,
      "loss": 0.5258,
      "num_input_tokens_seen": 1498680,
      "step": 2295
    },
    {
      "epoch": 1.2054507337526206,
      "grad_norm": 0.6574754118919373,
      "learning_rate": 3.012316561844864e-05,
      "loss": 0.761,
      "num_input_tokens_seen": 1501944,
      "step": 2300
    },
    {
      "epoch": 1.2080712788259957,
      "grad_norm": 0.7644612193107605,
      "learning_rate": 3.018867924528302e-05,
      "loss": 0.5893,
      "num_input_tokens_seen": 1506104,
      "step": 2305
    },
    {
      "epoch": 1.2106918238993711,
      "grad_norm": 0.4440242052078247,
      "learning_rate": 3.0254192872117405e-05,
      "loss": 0.5086,
      "num_input_tokens_seen": 1509720,
      "step": 2310
    },
    {
      "epoch": 1.2133123689727463,
      "grad_norm": 0.5312535762786865,
      "learning_rate": 3.0319706498951785e-05,
      "loss": 0.4819,
      "num_input_tokens_seen": 1513592,
      "step": 2315
    },
    {
      "epoch": 1.2159329140461215,
      "grad_norm": 0.5760791301727295,
      "learning_rate": 3.038522012578616e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 1516472,
      "step": 2320
    },
    {
      "epoch": 1.2185534591194969,
      "grad_norm": 0.2823583483695984,
      "learning_rate": 3.0450733752620547e-05,
      "loss": 0.583,
      "num_input_tokens_seen": 1520856,
      "step": 2325
    },
    {
      "epoch": 1.221174004192872,
      "grad_norm": 0.899773120880127,
      "learning_rate": 3.0516247379454927e-05,
      "loss": 0.6044,
      "num_input_tokens_seen": 1524248,
      "step": 2330
    },
    {
      "epoch": 1.2237945492662474,
      "grad_norm": 0.6032740473747253,
      "learning_rate": 3.0581761006289314e-05,
      "loss": 0.5765,
      "num_input_tokens_seen": 1527224,
      "step": 2335
    },
    {
      "epoch": 1.2264150943396226,
      "grad_norm": 0.5387024879455566,
      "learning_rate": 3.064727463312369e-05,
      "loss": 0.5084,
      "num_input_tokens_seen": 1530808,
      "step": 2340
    },
    {
      "epoch": 1.229035639412998,
      "grad_norm": 0.5861983895301819,
      "learning_rate": 3.071278825995807e-05,
      "loss": 0.5182,
      "num_input_tokens_seen": 1533848,
      "step": 2345
    },
    {
      "epoch": 1.2316561844863732,
      "grad_norm": 0.26886922121047974,
      "learning_rate": 3.0778301886792456e-05,
      "loss": 0.491,
      "num_input_tokens_seen": 1537016,
      "step": 2350
    },
    {
      "epoch": 1.2342767295597485,
      "grad_norm": 0.3899213671684265,
      "learning_rate": 3.084381551362683e-05,
      "loss": 0.5937,
      "num_input_tokens_seen": 1540472,
      "step": 2355
    },
    {
      "epoch": 1.2368972746331237,
      "grad_norm": 0.3073633015155792,
      "learning_rate": 3.0909329140461215e-05,
      "loss": 0.5022,
      "num_input_tokens_seen": 1544504,
      "step": 2360
    },
    {
      "epoch": 1.2395178197064989,
      "grad_norm": 0.4332829713821411,
      "learning_rate": 3.09748427672956e-05,
      "loss": 0.5096,
      "num_input_tokens_seen": 1547608,
      "step": 2365
    },
    {
      "epoch": 1.2421383647798743,
      "grad_norm": 0.7224923968315125,
      "learning_rate": 3.104035639412998e-05,
      "loss": 0.5463,
      "num_input_tokens_seen": 1550392,
      "step": 2370
    },
    {
      "epoch": 1.2447589098532494,
      "grad_norm": 0.6858993768692017,
      "learning_rate": 3.1105870020964365e-05,
      "loss": 0.6166,
      "num_input_tokens_seen": 1553048,
      "step": 2375
    },
    {
      "epoch": 1.2473794549266248,
      "grad_norm": 0.797988772392273,
      "learning_rate": 3.117138364779874e-05,
      "loss": 0.6038,
      "num_input_tokens_seen": 1556216,
      "step": 2380
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.43104299902915955,
      "learning_rate": 3.1236897274633124e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 1558872,
      "step": 2385
    },
    {
      "epoch": 1.2526205450733752,
      "grad_norm": 0.5543548464775085,
      "learning_rate": 3.130241090146751e-05,
      "loss": 0.6316,
      "num_input_tokens_seen": 1562040,
      "step": 2390
    },
    {
      "epoch": 1.2552410901467506,
      "grad_norm": 0.24662692844867706,
      "learning_rate": 3.136792452830189e-05,
      "loss": 0.431,
      "num_input_tokens_seen": 1564760,
      "step": 2395
    },
    {
      "epoch": 1.2578616352201257,
      "grad_norm": 0.38884979486465454,
      "learning_rate": 3.1433438155136266e-05,
      "loss": 0.5682,
      "num_input_tokens_seen": 1567416,
      "step": 2400
    },
    {
      "epoch": 1.2604821802935011,
      "grad_norm": 0.47814831137657166,
      "learning_rate": 3.1498951781970656e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 1571256,
      "step": 2405
    },
    {
      "epoch": 1.2631027253668763,
      "grad_norm": 0.47196531295776367,
      "learning_rate": 3.156446540880503e-05,
      "loss": 0.5812,
      "num_input_tokens_seen": 1574488,
      "step": 2410
    },
    {
      "epoch": 1.2657232704402515,
      "grad_norm": 0.39511486887931824,
      "learning_rate": 3.162997903563941e-05,
      "loss": 0.6549,
      "num_input_tokens_seen": 1578616,
      "step": 2415
    },
    {
      "epoch": 1.2683438155136268,
      "grad_norm": 0.4217047691345215,
      "learning_rate": 3.16954926624738e-05,
      "loss": 0.677,
      "num_input_tokens_seen": 1581688,
      "step": 2420
    },
    {
      "epoch": 1.270964360587002,
      "grad_norm": 0.460693895816803,
      "learning_rate": 3.1761006289308175e-05,
      "loss": 0.4754,
      "num_input_tokens_seen": 1584600,
      "step": 2425
    },
    {
      "epoch": 1.2735849056603774,
      "grad_norm": 0.3616626560688019,
      "learning_rate": 3.182651991614256e-05,
      "loss": 0.4475,
      "num_input_tokens_seen": 1588600,
      "step": 2430
    },
    {
      "epoch": 1.2762054507337526,
      "grad_norm": 0.8825986981391907,
      "learning_rate": 3.189203354297694e-05,
      "loss": 0.5353,
      "num_input_tokens_seen": 1591128,
      "step": 2435
    },
    {
      "epoch": 1.2788259958071277,
      "grad_norm": 0.7514281868934631,
      "learning_rate": 3.1957547169811324e-05,
      "loss": 0.5343,
      "num_input_tokens_seen": 1594552,
      "step": 2440
    },
    {
      "epoch": 1.2814465408805031,
      "grad_norm": 1.2874741554260254,
      "learning_rate": 3.202306079664571e-05,
      "loss": 0.565,
      "num_input_tokens_seen": 1597496,
      "step": 2445
    },
    {
      "epoch": 1.2840670859538785,
      "grad_norm": 0.6459832787513733,
      "learning_rate": 3.2088574423480084e-05,
      "loss": 0.5563,
      "num_input_tokens_seen": 1600856,
      "step": 2450
    },
    {
      "epoch": 1.2866876310272537,
      "grad_norm": 0.5344322919845581,
      "learning_rate": 3.215408805031447e-05,
      "loss": 0.4766,
      "num_input_tokens_seen": 1603896,
      "step": 2455
    },
    {
      "epoch": 1.2893081761006289,
      "grad_norm": 0.7566539645195007,
      "learning_rate": 3.221960167714885e-05,
      "loss": 0.6133,
      "num_input_tokens_seen": 1606840,
      "step": 2460
    },
    {
      "epoch": 1.2919287211740043,
      "grad_norm": 0.7940161824226379,
      "learning_rate": 3.228511530398323e-05,
      "loss": 0.6399,
      "num_input_tokens_seen": 1610392,
      "step": 2465
    },
    {
      "epoch": 1.2945492662473794,
      "grad_norm": 0.5592581629753113,
      "learning_rate": 3.235062893081761e-05,
      "loss": 0.6299,
      "num_input_tokens_seen": 1613976,
      "step": 2470
    },
    {
      "epoch": 1.2971698113207548,
      "grad_norm": 0.8227682709693909,
      "learning_rate": 3.241614255765199e-05,
      "loss": 0.658,
      "num_input_tokens_seen": 1617048,
      "step": 2475
    },
    {
      "epoch": 1.29979035639413,
      "grad_norm": 0.49541613459587097,
      "learning_rate": 3.2481656184486375e-05,
      "loss": 0.5919,
      "num_input_tokens_seen": 1620216,
      "step": 2480
    },
    {
      "epoch": 1.3024109014675052,
      "grad_norm": 0.4754926860332489,
      "learning_rate": 3.254716981132075e-05,
      "loss": 0.3575,
      "num_input_tokens_seen": 1623448,
      "step": 2485
    },
    {
      "epoch": 1.3050314465408805,
      "grad_norm": 0.9303324818611145,
      "learning_rate": 3.261268343815514e-05,
      "loss": 0.8487,
      "num_input_tokens_seen": 1627160,
      "step": 2490
    },
    {
      "epoch": 1.3076519916142557,
      "grad_norm": 0.577176034450531,
      "learning_rate": 3.267819706498952e-05,
      "loss": 0.6875,
      "num_input_tokens_seen": 1630488,
      "step": 2495
    },
    {
      "epoch": 1.310272536687631,
      "grad_norm": 0.8548619747161865,
      "learning_rate": 3.27437106918239e-05,
      "loss": 0.4342,
      "num_input_tokens_seen": 1633720,
      "step": 2500
    },
    {
      "epoch": 1.3128930817610063,
      "grad_norm": 1.4681223630905151,
      "learning_rate": 3.2809224318658284e-05,
      "loss": 0.6227,
      "num_input_tokens_seen": 1636824,
      "step": 2505
    },
    {
      "epoch": 1.3155136268343814,
      "grad_norm": 0.41534966230392456,
      "learning_rate": 3.287473794549266e-05,
      "loss": 0.5382,
      "num_input_tokens_seen": 1640312,
      "step": 2510
    },
    {
      "epoch": 1.3181341719077568,
      "grad_norm": 0.4104596674442291,
      "learning_rate": 3.294025157232704e-05,
      "loss": 0.4011,
      "num_input_tokens_seen": 1642904,
      "step": 2515
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 1.307713270187378,
      "learning_rate": 3.3005765199161426e-05,
      "loss": 0.4812,
      "num_input_tokens_seen": 1645624,
      "step": 2520
    },
    {
      "epoch": 1.3233752620545074,
      "grad_norm": 0.6084907650947571,
      "learning_rate": 3.307127882599581e-05,
      "loss": 0.6028,
      "num_input_tokens_seen": 1649432,
      "step": 2525
    },
    {
      "epoch": 1.3259958071278826,
      "grad_norm": 0.9883345365524292,
      "learning_rate": 3.313679245283019e-05,
      "loss": 0.7375,
      "num_input_tokens_seen": 1651864,
      "step": 2530
    },
    {
      "epoch": 1.3286163522012577,
      "grad_norm": 0.7073197960853577,
      "learning_rate": 3.320230607966457e-05,
      "loss": 0.4501,
      "num_input_tokens_seen": 1654904,
      "step": 2535
    },
    {
      "epoch": 1.3312368972746331,
      "grad_norm": 0.49588441848754883,
      "learning_rate": 3.326781970649895e-05,
      "loss": 0.6093,
      "num_input_tokens_seen": 1658488,
      "step": 2540
    },
    {
      "epoch": 1.3338574423480085,
      "grad_norm": 0.6380655169487,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.5261,
      "num_input_tokens_seen": 1661336,
      "step": 2545
    },
    {
      "epoch": 1.3364779874213837,
      "grad_norm": 0.5156749486923218,
      "learning_rate": 3.339884696016772e-05,
      "loss": 0.5246,
      "num_input_tokens_seen": 1664024,
      "step": 2550
    },
    {
      "epoch": 1.3390985324947589,
      "grad_norm": 0.5785089135169983,
      "learning_rate": 3.3464360587002094e-05,
      "loss": 0.6485,
      "num_input_tokens_seen": 1666936,
      "step": 2555
    },
    {
      "epoch": 1.3417190775681342,
      "grad_norm": 0.5079804062843323,
      "learning_rate": 3.3529874213836484e-05,
      "loss": 0.4023,
      "num_input_tokens_seen": 1670648,
      "step": 2560
    },
    {
      "epoch": 1.3443396226415094,
      "grad_norm": 0.695698082447052,
      "learning_rate": 3.359538784067086e-05,
      "loss": 0.4287,
      "num_input_tokens_seen": 1673464,
      "step": 2565
    },
    {
      "epoch": 1.3469601677148848,
      "grad_norm": 0.6341899633407593,
      "learning_rate": 3.3660901467505244e-05,
      "loss": 0.6027,
      "num_input_tokens_seen": 1676024,
      "step": 2570
    },
    {
      "epoch": 1.34958071278826,
      "grad_norm": 0.6711342334747314,
      "learning_rate": 3.3726415094339627e-05,
      "loss": 0.6223,
      "num_input_tokens_seen": 1680024,
      "step": 2575
    },
    {
      "epoch": 1.3522012578616351,
      "grad_norm": 0.3773452937602997,
      "learning_rate": 3.3791928721174e-05,
      "loss": 0.5052,
      "num_input_tokens_seen": 1683384,
      "step": 2580
    },
    {
      "epoch": 1.3548218029350105,
      "grad_norm": 0.3204648792743683,
      "learning_rate": 3.3857442348008386e-05,
      "loss": 0.6449,
      "num_input_tokens_seen": 1686456,
      "step": 2585
    },
    {
      "epoch": 1.3574423480083857,
      "grad_norm": 0.3872120976448059,
      "learning_rate": 3.392295597484277e-05,
      "loss": 0.4231,
      "num_input_tokens_seen": 1690072,
      "step": 2590
    },
    {
      "epoch": 1.360062893081761,
      "grad_norm": 0.3637029528617859,
      "learning_rate": 3.398846960167715e-05,
      "loss": 0.5397,
      "num_input_tokens_seen": 1692664,
      "step": 2595
    },
    {
      "epoch": 1.3626834381551363,
      "grad_norm": 0.517933189868927,
      "learning_rate": 3.4053983228511535e-05,
      "loss": 0.5918,
      "num_input_tokens_seen": 1695288,
      "step": 2600
    },
    {
      "epoch": 1.3653039832285114,
      "grad_norm": 0.386006623506546,
      "learning_rate": 3.411949685534591e-05,
      "loss": 0.4749,
      "num_input_tokens_seen": 1699768,
      "step": 2605
    },
    {
      "epoch": 1.3679245283018868,
      "grad_norm": 0.46657586097717285,
      "learning_rate": 3.4185010482180295e-05,
      "loss": 0.6392,
      "num_input_tokens_seen": 1702264,
      "step": 2610
    },
    {
      "epoch": 1.370545073375262,
      "grad_norm": 1.1491572856903076,
      "learning_rate": 3.425052410901468e-05,
      "loss": 0.6222,
      "num_input_tokens_seen": 1704920,
      "step": 2615
    },
    {
      "epoch": 1.3731656184486374,
      "grad_norm": 0.362641841173172,
      "learning_rate": 3.431603773584906e-05,
      "loss": 0.5863,
      "num_input_tokens_seen": 1708024,
      "step": 2620
    },
    {
      "epoch": 1.3757861635220126,
      "grad_norm": 0.7173078656196594,
      "learning_rate": 3.438155136268344e-05,
      "loss": 0.7152,
      "num_input_tokens_seen": 1709880,
      "step": 2625
    },
    {
      "epoch": 1.3784067085953877,
      "grad_norm": 0.4574228823184967,
      "learning_rate": 3.444706498951782e-05,
      "loss": 0.5888,
      "num_input_tokens_seen": 1714040,
      "step": 2630
    },
    {
      "epoch": 1.381027253668763,
      "grad_norm": 0.2512785792350769,
      "learning_rate": 3.45125786163522e-05,
      "loss": 0.6822,
      "num_input_tokens_seen": 1718712,
      "step": 2635
    },
    {
      "epoch": 1.3836477987421385,
      "grad_norm": 0.6228510737419128,
      "learning_rate": 3.457809224318658e-05,
      "loss": 0.5069,
      "num_input_tokens_seen": 1721304,
      "step": 2640
    },
    {
      "epoch": 1.3862683438155137,
      "grad_norm": 0.3337365388870239,
      "learning_rate": 3.464360587002097e-05,
      "loss": 0.5867,
      "num_input_tokens_seen": 1723896,
      "step": 2645
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 0.655147910118103,
      "learning_rate": 3.4709119496855346e-05,
      "loss": 0.4598,
      "num_input_tokens_seen": 1727864,
      "step": 2650
    },
    {
      "epoch": 1.3915094339622642,
      "grad_norm": 0.3206739127635956,
      "learning_rate": 3.477463312368973e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 1731064,
      "step": 2655
    },
    {
      "epoch": 1.3941299790356394,
      "grad_norm": 1.0238263607025146,
      "learning_rate": 3.484014675052411e-05,
      "loss": 0.6336,
      "num_input_tokens_seen": 1733912,
      "step": 2660
    },
    {
      "epoch": 1.3967505241090148,
      "grad_norm": 0.4107344448566437,
      "learning_rate": 3.490566037735849e-05,
      "loss": 0.4489,
      "num_input_tokens_seen": 1736952,
      "step": 2665
    },
    {
      "epoch": 1.39937106918239,
      "grad_norm": 0.47412118315696716,
      "learning_rate": 3.497117400419288e-05,
      "loss": 0.5283,
      "num_input_tokens_seen": 1740120,
      "step": 2670
    },
    {
      "epoch": 1.4019916142557651,
      "grad_norm": 0.40940263867378235,
      "learning_rate": 3.5036687631027254e-05,
      "loss": 0.5072,
      "num_input_tokens_seen": 1742712,
      "step": 2675
    },
    {
      "epoch": 1.4046121593291405,
      "grad_norm": 0.3864918351173401,
      "learning_rate": 3.510220125786164e-05,
      "loss": 0.6007,
      "num_input_tokens_seen": 1745112,
      "step": 2680
    },
    {
      "epoch": 1.4072327044025157,
      "grad_norm": 0.27079564332962036,
      "learning_rate": 3.516771488469602e-05,
      "loss": 0.4485,
      "num_input_tokens_seen": 1748856,
      "step": 2685
    },
    {
      "epoch": 1.409853249475891,
      "grad_norm": 0.6417648792266846,
      "learning_rate": 3.5233228511530397e-05,
      "loss": 0.5099,
      "num_input_tokens_seen": 1753656,
      "step": 2690
    },
    {
      "epoch": 1.4124737945492662,
      "grad_norm": 1.2417958974838257,
      "learning_rate": 3.529874213836478e-05,
      "loss": 0.5431,
      "num_input_tokens_seen": 1757336,
      "step": 2695
    },
    {
      "epoch": 1.4150943396226414,
      "grad_norm": 0.35728734731674194,
      "learning_rate": 3.536425576519916e-05,
      "loss": 0.5034,
      "num_input_tokens_seen": 1761208,
      "step": 2700
    },
    {
      "epoch": 1.4177148846960168,
      "grad_norm": 0.3614325225353241,
      "learning_rate": 3.5429769392033546e-05,
      "loss": 0.5795,
      "num_input_tokens_seen": 1764792,
      "step": 2705
    },
    {
      "epoch": 1.420335429769392,
      "grad_norm": 0.3991718292236328,
      "learning_rate": 3.549528301886792e-05,
      "loss": 0.4607,
      "num_input_tokens_seen": 1768888,
      "step": 2710
    },
    {
      "epoch": 1.4229559748427674,
      "grad_norm": 0.5224614143371582,
      "learning_rate": 3.556079664570231e-05,
      "loss": 0.6045,
      "num_input_tokens_seen": 1771864,
      "step": 2715
    },
    {
      "epoch": 1.4255765199161425,
      "grad_norm": 0.6308770179748535,
      "learning_rate": 3.562631027253669e-05,
      "loss": 0.5485,
      "num_input_tokens_seen": 1775128,
      "step": 2720
    },
    {
      "epoch": 1.4281970649895177,
      "grad_norm": 0.7845183610916138,
      "learning_rate": 3.569182389937107e-05,
      "loss": 0.5875,
      "num_input_tokens_seen": 1777976,
      "step": 2725
    },
    {
      "epoch": 1.430817610062893,
      "grad_norm": 0.6444384455680847,
      "learning_rate": 3.5757337526205454e-05,
      "loss": 0.4483,
      "num_input_tokens_seen": 1780824,
      "step": 2730
    },
    {
      "epoch": 1.4334381551362683,
      "grad_norm": 1.6016490459442139,
      "learning_rate": 3.582285115303983e-05,
      "loss": 0.6736,
      "num_input_tokens_seen": 1783896,
      "step": 2735
    },
    {
      "epoch": 1.4360587002096437,
      "grad_norm": 0.46415644884109497,
      "learning_rate": 3.588836477987422e-05,
      "loss": 0.462,
      "num_input_tokens_seen": 1787288,
      "step": 2740
    },
    {
      "epoch": 1.4386792452830188,
      "grad_norm": 0.45747271180152893,
      "learning_rate": 3.59538784067086e-05,
      "loss": 0.7452,
      "num_input_tokens_seen": 1790904,
      "step": 2745
    },
    {
      "epoch": 1.441299790356394,
      "grad_norm": 0.5281717777252197,
      "learning_rate": 3.601939203354298e-05,
      "loss": 0.5233,
      "num_input_tokens_seen": 1794392,
      "step": 2750
    },
    {
      "epoch": 1.4439203354297694,
      "grad_norm": 0.49977174401283264,
      "learning_rate": 3.608490566037736e-05,
      "loss": 0.4232,
      "num_input_tokens_seen": 1797656,
      "step": 2755
    },
    {
      "epoch": 1.4465408805031448,
      "grad_norm": 0.7086086273193359,
      "learning_rate": 3.615041928721174e-05,
      "loss": 0.6406,
      "num_input_tokens_seen": 1801944,
      "step": 2760
    },
    {
      "epoch": 1.44916142557652,
      "grad_norm": 0.2889890670776367,
      "learning_rate": 3.621593291404612e-05,
      "loss": 0.5021,
      "num_input_tokens_seen": 1804888,
      "step": 2765
    },
    {
      "epoch": 1.4517819706498951,
      "grad_norm": 0.5579442381858826,
      "learning_rate": 3.6281446540880505e-05,
      "loss": 0.5269,
      "num_input_tokens_seen": 1808120,
      "step": 2770
    },
    {
      "epoch": 1.4544025157232705,
      "grad_norm": 0.31058281660079956,
      "learning_rate": 3.634696016771489e-05,
      "loss": 0.4992,
      "num_input_tokens_seen": 1812408,
      "step": 2775
    },
    {
      "epoch": 1.4570230607966457,
      "grad_norm": 0.5565035343170166,
      "learning_rate": 3.6412473794549265e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 1815352,
      "step": 2780
    },
    {
      "epoch": 1.459643605870021,
      "grad_norm": 0.6492956280708313,
      "learning_rate": 3.647798742138365e-05,
      "loss": 0.5937,
      "num_input_tokens_seen": 1817720,
      "step": 2785
    },
    {
      "epoch": 1.4622641509433962,
      "grad_norm": 0.5672410130500793,
      "learning_rate": 3.654350104821803e-05,
      "loss": 0.5917,
      "num_input_tokens_seen": 1821400,
      "step": 2790
    },
    {
      "epoch": 1.4648846960167714,
      "grad_norm": 1.18295156955719,
      "learning_rate": 3.6609014675052414e-05,
      "loss": 0.5113,
      "num_input_tokens_seen": 1824152,
      "step": 2795
    },
    {
      "epoch": 1.4675052410901468,
      "grad_norm": 0.6491718292236328,
      "learning_rate": 3.66745283018868e-05,
      "loss": 0.7118,
      "num_input_tokens_seen": 1827704,
      "step": 2800
    },
    {
      "epoch": 1.470125786163522,
      "grad_norm": 0.5988360643386841,
      "learning_rate": 3.6740041928721173e-05,
      "loss": 0.6585,
      "num_input_tokens_seen": 1830776,
      "step": 2805
    },
    {
      "epoch": 1.4727463312368974,
      "grad_norm": 0.42234092950820923,
      "learning_rate": 3.6805555555555556e-05,
      "loss": 0.399,
      "num_input_tokens_seen": 1834136,
      "step": 2810
    },
    {
      "epoch": 1.4753668763102725,
      "grad_norm": 0.4752788245677948,
      "learning_rate": 3.687106918238994e-05,
      "loss": 0.6109,
      "num_input_tokens_seen": 1837784,
      "step": 2815
    },
    {
      "epoch": 1.4779874213836477,
      "grad_norm": 0.5241593718528748,
      "learning_rate": 3.6936582809224316e-05,
      "loss": 0.5872,
      "num_input_tokens_seen": 1841656,
      "step": 2820
    },
    {
      "epoch": 1.480607966457023,
      "grad_norm": 0.5408892631530762,
      "learning_rate": 3.7002096436058706e-05,
      "loss": 0.582,
      "num_input_tokens_seen": 1844792,
      "step": 2825
    },
    {
      "epoch": 1.4832285115303983,
      "grad_norm": 0.6076368093490601,
      "learning_rate": 3.706761006289308e-05,
      "loss": 0.6285,
      "num_input_tokens_seen": 1848312,
      "step": 2830
    },
    {
      "epoch": 1.4858490566037736,
      "grad_norm": 0.5897635817527771,
      "learning_rate": 3.7133123689727465e-05,
      "loss": 0.5569,
      "num_input_tokens_seen": 1850968,
      "step": 2835
    },
    {
      "epoch": 1.4884696016771488,
      "grad_norm": 0.6253833770751953,
      "learning_rate": 3.719863731656185e-05,
      "loss": 0.5419,
      "num_input_tokens_seen": 1853528,
      "step": 2840
    },
    {
      "epoch": 1.491090146750524,
      "grad_norm": 0.5587652921676636,
      "learning_rate": 3.7264150943396224e-05,
      "loss": 0.5173,
      "num_input_tokens_seen": 1856376,
      "step": 2845
    },
    {
      "epoch": 1.4937106918238994,
      "grad_norm": 0.5385523438453674,
      "learning_rate": 3.732966457023061e-05,
      "loss": 0.5507,
      "num_input_tokens_seen": 1859672,
      "step": 2850
    },
    {
      "epoch": 1.4963312368972748,
      "grad_norm": 0.5993587970733643,
      "learning_rate": 3.739517819706499e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 1862680,
      "step": 2855
    },
    {
      "epoch": 1.49895178197065,
      "grad_norm": 0.8058831095695496,
      "learning_rate": 3.7460691823899374e-05,
      "loss": 0.6015,
      "num_input_tokens_seen": 1865304,
      "step": 2860
    },
    {
      "epoch": 1.501572327044025,
      "grad_norm": 0.8399238586425781,
      "learning_rate": 3.752620545073376e-05,
      "loss": 0.718,
      "num_input_tokens_seen": 1868152,
      "step": 2865
    },
    {
      "epoch": 1.5041928721174003,
      "grad_norm": 0.5981653928756714,
      "learning_rate": 3.759171907756814e-05,
      "loss": 0.7101,
      "num_input_tokens_seen": 1871192,
      "step": 2870
    },
    {
      "epoch": 1.5068134171907757,
      "grad_norm": 0.5943511128425598,
      "learning_rate": 3.7657232704402516e-05,
      "loss": 0.6278,
      "num_input_tokens_seen": 1874360,
      "step": 2875
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 0.3546455502510071,
      "learning_rate": 3.77227463312369e-05,
      "loss": 0.5596,
      "num_input_tokens_seen": 1877400,
      "step": 2880
    },
    {
      "epoch": 1.5120545073375262,
      "grad_norm": 0.598412275314331,
      "learning_rate": 3.778825995807128e-05,
      "loss": 0.556,
      "num_input_tokens_seen": 1880440,
      "step": 2885
    },
    {
      "epoch": 1.5146750524109014,
      "grad_norm": 1.1852914094924927,
      "learning_rate": 3.785377358490566e-05,
      "loss": 0.536,
      "num_input_tokens_seen": 1883352,
      "step": 2890
    },
    {
      "epoch": 1.5172955974842768,
      "grad_norm": 0.9353783130645752,
      "learning_rate": 3.791928721174005e-05,
      "loss": 0.4793,
      "num_input_tokens_seen": 1886264,
      "step": 2895
    },
    {
      "epoch": 1.519916142557652,
      "grad_norm": 0.5552292466163635,
      "learning_rate": 3.7984800838574425e-05,
      "loss": 0.5116,
      "num_input_tokens_seen": 1889144,
      "step": 2900
    },
    {
      "epoch": 1.5225366876310273,
      "grad_norm": 0.5678912997245789,
      "learning_rate": 3.805031446540881e-05,
      "loss": 0.4744,
      "num_input_tokens_seen": 1894168,
      "step": 2905
    },
    {
      "epoch": 1.5251572327044025,
      "grad_norm": 0.5760968923568726,
      "learning_rate": 3.811582809224319e-05,
      "loss": 0.4484,
      "num_input_tokens_seen": 1898072,
      "step": 2910
    },
    {
      "epoch": 1.5277777777777777,
      "grad_norm": 0.608201265335083,
      "learning_rate": 3.818134171907757e-05,
      "loss": 0.4362,
      "num_input_tokens_seen": 1901464,
      "step": 2915
    },
    {
      "epoch": 1.530398322851153,
      "grad_norm": 0.4103991389274597,
      "learning_rate": 3.824685534591195e-05,
      "loss": 0.4312,
      "num_input_tokens_seen": 1904888,
      "step": 2920
    },
    {
      "epoch": 1.5330188679245285,
      "grad_norm": 0.7111467123031616,
      "learning_rate": 3.831236897274633e-05,
      "loss": 0.5993,
      "num_input_tokens_seen": 1908088,
      "step": 2925
    },
    {
      "epoch": 1.5356394129979036,
      "grad_norm": 0.9036352038383484,
      "learning_rate": 3.8377882599580716e-05,
      "loss": 0.6244,
      "num_input_tokens_seen": 1911288,
      "step": 2930
    },
    {
      "epoch": 1.5382599580712788,
      "grad_norm": 0.3432755768299103,
      "learning_rate": 3.844339622641509e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 1914264,
      "step": 2935
    },
    {
      "epoch": 1.540880503144654,
      "grad_norm": 0.2700664699077606,
      "learning_rate": 3.8508909853249476e-05,
      "loss": 0.494,
      "num_input_tokens_seen": 1918296,
      "step": 2940
    },
    {
      "epoch": 1.5435010482180294,
      "grad_norm": 0.4062797427177429,
      "learning_rate": 3.857442348008386e-05,
      "loss": 0.7891,
      "num_input_tokens_seen": 1922136,
      "step": 2945
    },
    {
      "epoch": 1.5461215932914047,
      "grad_norm": 0.5338535904884338,
      "learning_rate": 3.863993710691824e-05,
      "loss": 0.5757,
      "num_input_tokens_seen": 1925240,
      "step": 2950
    },
    {
      "epoch": 1.54874213836478,
      "grad_norm": 0.5383860468864441,
      "learning_rate": 3.8705450733752625e-05,
      "loss": 0.5344,
      "num_input_tokens_seen": 1927864,
      "step": 2955
    },
    {
      "epoch": 1.551362683438155,
      "grad_norm": 0.3868836462497711,
      "learning_rate": 3.8770964360587e-05,
      "loss": 0.5323,
      "num_input_tokens_seen": 1930360,
      "step": 2960
    },
    {
      "epoch": 1.5539832285115303,
      "grad_norm": 0.3844742476940155,
      "learning_rate": 3.883647798742139e-05,
      "loss": 0.6391,
      "num_input_tokens_seen": 1933016,
      "step": 2965
    },
    {
      "epoch": 1.5566037735849056,
      "grad_norm": 0.8026425838470459,
      "learning_rate": 3.890199161425577e-05,
      "loss": 0.4658,
      "num_input_tokens_seen": 1936184,
      "step": 2970
    },
    {
      "epoch": 1.559224318658281,
      "grad_norm": 0.4246020019054413,
      "learning_rate": 3.8967505241090144e-05,
      "loss": 0.518,
      "num_input_tokens_seen": 1939480,
      "step": 2975
    },
    {
      "epoch": 1.5618448637316562,
      "grad_norm": 0.45765870809555054,
      "learning_rate": 3.9033018867924534e-05,
      "loss": 0.5957,
      "num_input_tokens_seen": 1943384,
      "step": 2980
    },
    {
      "epoch": 1.5644654088050314,
      "grad_norm": 0.7186525464057922,
      "learning_rate": 3.909853249475891e-05,
      "loss": 0.6466,
      "num_input_tokens_seen": 1946264,
      "step": 2985
    },
    {
      "epoch": 1.5670859538784065,
      "grad_norm": 0.5312483310699463,
      "learning_rate": 3.916404612159329e-05,
      "loss": 0.5733,
      "num_input_tokens_seen": 1949336,
      "step": 2990
    },
    {
      "epoch": 1.569706498951782,
      "grad_norm": 0.5283254981040955,
      "learning_rate": 3.9229559748427676e-05,
      "loss": 0.5487,
      "num_input_tokens_seen": 1951928,
      "step": 2995
    },
    {
      "epoch": 1.5723270440251573,
      "grad_norm": 0.6651094555854797,
      "learning_rate": 3.929507337526205e-05,
      "loss": 0.4916,
      "num_input_tokens_seen": 1954680,
      "step": 3000
    },
    {
      "epoch": 1.5749475890985325,
      "grad_norm": 0.37610605359077454,
      "learning_rate": 3.9360587002096435e-05,
      "loss": 0.5648,
      "num_input_tokens_seen": 1957784,
      "step": 3005
    },
    {
      "epoch": 1.5775681341719077,
      "grad_norm": 0.43640419840812683,
      "learning_rate": 3.942610062893082e-05,
      "loss": 0.5172,
      "num_input_tokens_seen": 1960984,
      "step": 3010
    },
    {
      "epoch": 1.580188679245283,
      "grad_norm": 0.37910714745521545,
      "learning_rate": 3.94916142557652e-05,
      "loss": 0.3995,
      "num_input_tokens_seen": 1965240,
      "step": 3015
    },
    {
      "epoch": 1.5828092243186582,
      "grad_norm": 0.2406591922044754,
      "learning_rate": 3.9557127882599585e-05,
      "loss": 0.4829,
      "num_input_tokens_seen": 1969080,
      "step": 3020
    },
    {
      "epoch": 1.5854297693920336,
      "grad_norm": 0.4885497987270355,
      "learning_rate": 3.962264150943397e-05,
      "loss": 0.6976,
      "num_input_tokens_seen": 1972952,
      "step": 3025
    },
    {
      "epoch": 1.5880503144654088,
      "grad_norm": 0.45780450105667114,
      "learning_rate": 3.9688155136268344e-05,
      "loss": 0.8319,
      "num_input_tokens_seen": 1975736,
      "step": 3030
    },
    {
      "epoch": 1.590670859538784,
      "grad_norm": 0.44172316789627075,
      "learning_rate": 3.975366876310273e-05,
      "loss": 0.5479,
      "num_input_tokens_seen": 1979832,
      "step": 3035
    },
    {
      "epoch": 1.5932914046121593,
      "grad_norm": 0.6871852278709412,
      "learning_rate": 3.981918238993711e-05,
      "loss": 0.4488,
      "num_input_tokens_seen": 1982104,
      "step": 3040
    },
    {
      "epoch": 1.5959119496855347,
      "grad_norm": 0.6359638571739197,
      "learning_rate": 3.9884696016771486e-05,
      "loss": 0.5141,
      "num_input_tokens_seen": 1984920,
      "step": 3045
    },
    {
      "epoch": 1.59853249475891,
      "grad_norm": 0.5884039402008057,
      "learning_rate": 3.9950209643605876e-05,
      "loss": 0.5234,
      "num_input_tokens_seen": 1987448,
      "step": 3050
    },
    {
      "epoch": 1.601153039832285,
      "grad_norm": 0.33832070231437683,
      "learning_rate": 4.001572327044025e-05,
      "loss": 0.7838,
      "num_input_tokens_seen": 1991960,
      "step": 3055
    },
    {
      "epoch": 1.6037735849056602,
      "grad_norm": 0.3432486355304718,
      "learning_rate": 4.0081236897274636e-05,
      "loss": 0.5128,
      "num_input_tokens_seen": 1995416,
      "step": 3060
    },
    {
      "epoch": 1.6063941299790356,
      "grad_norm": 0.5184414386749268,
      "learning_rate": 4.014675052410902e-05,
      "loss": 0.4423,
      "num_input_tokens_seen": 1999064,
      "step": 3065
    },
    {
      "epoch": 1.609014675052411,
      "grad_norm": 0.40664607286453247,
      "learning_rate": 4.0212264150943395e-05,
      "loss": 0.5163,
      "num_input_tokens_seen": 2002328,
      "step": 3070
    },
    {
      "epoch": 1.6116352201257862,
      "grad_norm": 0.44490042328834534,
      "learning_rate": 4.027777777777778e-05,
      "loss": 0.5794,
      "num_input_tokens_seen": 2005240,
      "step": 3075
    },
    {
      "epoch": 1.6142557651991614,
      "grad_norm": 0.5582923889160156,
      "learning_rate": 4.034329140461216e-05,
      "loss": 0.567,
      "num_input_tokens_seen": 2009592,
      "step": 3080
    },
    {
      "epoch": 1.6168763102725365,
      "grad_norm": 0.6358270049095154,
      "learning_rate": 4.0408805031446544e-05,
      "loss": 0.5529,
      "num_input_tokens_seen": 2012056,
      "step": 3085
    },
    {
      "epoch": 1.619496855345912,
      "grad_norm": 0.4640117287635803,
      "learning_rate": 4.047431865828093e-05,
      "loss": 0.5898,
      "num_input_tokens_seen": 2015608,
      "step": 3090
    },
    {
      "epoch": 1.6221174004192873,
      "grad_norm": 0.7850682735443115,
      "learning_rate": 4.0539832285115304e-05,
      "loss": 0.58,
      "num_input_tokens_seen": 2018456,
      "step": 3095
    },
    {
      "epoch": 1.6247379454926625,
      "grad_norm": 0.4665972590446472,
      "learning_rate": 4.0605345911949687e-05,
      "loss": 0.5104,
      "num_input_tokens_seen": 2021720,
      "step": 3100
    },
    {
      "epoch": 1.6273584905660377,
      "grad_norm": 0.49629560112953186,
      "learning_rate": 4.067085953878407e-05,
      "loss": 0.5412,
      "num_input_tokens_seen": 2024376,
      "step": 3105
    },
    {
      "epoch": 1.629979035639413,
      "grad_norm": 0.9328791499137878,
      "learning_rate": 4.073637316561845e-05,
      "loss": 0.6461,
      "num_input_tokens_seen": 2028440,
      "step": 3110
    },
    {
      "epoch": 1.6325995807127882,
      "grad_norm": 0.6896419525146484,
      "learning_rate": 4.080188679245283e-05,
      "loss": 0.5717,
      "num_input_tokens_seen": 2031352,
      "step": 3115
    },
    {
      "epoch": 1.6352201257861636,
      "grad_norm": 0.45232319831848145,
      "learning_rate": 4.086740041928722e-05,
      "loss": 0.6055,
      "num_input_tokens_seen": 2034680,
      "step": 3120
    },
    {
      "epoch": 1.6378406708595388,
      "grad_norm": 1.2703324556350708,
      "learning_rate": 4.0932914046121595e-05,
      "loss": 0.5952,
      "num_input_tokens_seen": 2038104,
      "step": 3125
    },
    {
      "epoch": 1.640461215932914,
      "grad_norm": 0.46080756187438965,
      "learning_rate": 4.099842767295597e-05,
      "loss": 0.5057,
      "num_input_tokens_seen": 2041624,
      "step": 3130
    },
    {
      "epoch": 1.6430817610062893,
      "grad_norm": 0.3218070864677429,
      "learning_rate": 4.106394129979036e-05,
      "loss": 0.6815,
      "num_input_tokens_seen": 2048056,
      "step": 3135
    },
    {
      "epoch": 1.6457023060796647,
      "grad_norm": 0.9881914258003235,
      "learning_rate": 4.112945492662474e-05,
      "loss": 0.5,
      "num_input_tokens_seen": 2050968,
      "step": 3140
    },
    {
      "epoch": 1.64832285115304,
      "grad_norm": 0.44052693247795105,
      "learning_rate": 4.119496855345912e-05,
      "loss": 0.6112,
      "num_input_tokens_seen": 2054008,
      "step": 3145
    },
    {
      "epoch": 1.650943396226415,
      "grad_norm": 0.24061554670333862,
      "learning_rate": 4.1260482180293504e-05,
      "loss": 0.5262,
      "num_input_tokens_seen": 2057496,
      "step": 3150
    },
    {
      "epoch": 1.6535639412997902,
      "grad_norm": 0.8729889988899231,
      "learning_rate": 4.132599580712788e-05,
      "loss": 0.5682,
      "num_input_tokens_seen": 2060984,
      "step": 3155
    },
    {
      "epoch": 1.6561844863731656,
      "grad_norm": 0.5761516094207764,
      "learning_rate": 4.139150943396227e-05,
      "loss": 0.5327,
      "num_input_tokens_seen": 2064056,
      "step": 3160
    },
    {
      "epoch": 1.658805031446541,
      "grad_norm": 0.43883225321769714,
      "learning_rate": 4.1457023060796646e-05,
      "loss": 0.5683,
      "num_input_tokens_seen": 2067160,
      "step": 3165
    },
    {
      "epoch": 1.6614255765199162,
      "grad_norm": 0.42488792538642883,
      "learning_rate": 4.152253668763103e-05,
      "loss": 0.5753,
      "num_input_tokens_seen": 2069656,
      "step": 3170
    },
    {
      "epoch": 1.6640461215932913,
      "grad_norm": 0.33425870537757874,
      "learning_rate": 4.158805031446541e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 2074552,
      "step": 3175
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.7388004064559937,
      "learning_rate": 4.1653563941299795e-05,
      "loss": 0.5853,
      "num_input_tokens_seen": 2077176,
      "step": 3180
    },
    {
      "epoch": 1.669287211740042,
      "grad_norm": 0.33924543857574463,
      "learning_rate": 4.171907756813417e-05,
      "loss": 0.4125,
      "num_input_tokens_seen": 2080728,
      "step": 3185
    },
    {
      "epoch": 1.6719077568134173,
      "grad_norm": 0.6891064047813416,
      "learning_rate": 4.1784591194968555e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 2083672,
      "step": 3190
    },
    {
      "epoch": 1.6745283018867925,
      "grad_norm": 0.8006487488746643,
      "learning_rate": 4.185010482180294e-05,
      "loss": 0.639,
      "num_input_tokens_seen": 2086872,
      "step": 3195
    },
    {
      "epoch": 1.6771488469601676,
      "grad_norm": 0.7308323383331299,
      "learning_rate": 4.1915618448637314e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 2090072,
      "step": 3200
    },
    {
      "epoch": 1.679769392033543,
      "grad_norm": 0.956451952457428,
      "learning_rate": 4.1981132075471704e-05,
      "loss": 0.5203,
      "num_input_tokens_seen": 2093112,
      "step": 3205
    },
    {
      "epoch": 1.6823899371069182,
      "grad_norm": 0.7438709735870361,
      "learning_rate": 4.204664570230608e-05,
      "loss": 0.5693,
      "num_input_tokens_seen": 2096408,
      "step": 3210
    },
    {
      "epoch": 1.6850104821802936,
      "grad_norm": 0.5464678406715393,
      "learning_rate": 4.2112159329140463e-05,
      "loss": 0.6194,
      "num_input_tokens_seen": 2099448,
      "step": 3215
    },
    {
      "epoch": 1.6876310272536688,
      "grad_norm": 1.1175469160079956,
      "learning_rate": 4.2177672955974846e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 2102520,
      "step": 3220
    },
    {
      "epoch": 1.690251572327044,
      "grad_norm": 0.5505601763725281,
      "learning_rate": 4.224318658280922e-05,
      "loss": 0.4527,
      "num_input_tokens_seen": 2105496,
      "step": 3225
    },
    {
      "epoch": 1.6928721174004193,
      "grad_norm": 0.528885543346405,
      "learning_rate": 4.2308700209643606e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 2108536,
      "step": 3230
    },
    {
      "epoch": 1.6954926624737947,
      "grad_norm": 0.4143374562263489,
      "learning_rate": 4.237421383647799e-05,
      "loss": 0.4349,
      "num_input_tokens_seen": 2112248,
      "step": 3235
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 0.6721603274345398,
      "learning_rate": 4.243972746331237e-05,
      "loss": 0.5841,
      "num_input_tokens_seen": 2114456,
      "step": 3240
    },
    {
      "epoch": 1.700733752620545,
      "grad_norm": 0.3702182173728943,
      "learning_rate": 4.2505241090146755e-05,
      "loss": 0.5152,
      "num_input_tokens_seen": 2117592,
      "step": 3245
    },
    {
      "epoch": 1.7033542976939202,
      "grad_norm": 0.494463175535202,
      "learning_rate": 4.257075471698113e-05,
      "loss": 0.5616,
      "num_input_tokens_seen": 2120376,
      "step": 3250
    },
    {
      "epoch": 1.7059748427672956,
      "grad_norm": 0.5483316779136658,
      "learning_rate": 4.2636268343815514e-05,
      "loss": 0.4607,
      "num_input_tokens_seen": 2124376,
      "step": 3255
    },
    {
      "epoch": 1.708595387840671,
      "grad_norm": 0.7869540452957153,
      "learning_rate": 4.27017819706499e-05,
      "loss": 0.5569,
      "num_input_tokens_seen": 2127384,
      "step": 3260
    },
    {
      "epoch": 1.7112159329140462,
      "grad_norm": 0.38165533542633057,
      "learning_rate": 4.276729559748428e-05,
      "loss": 0.5693,
      "num_input_tokens_seen": 2130616,
      "step": 3265
    },
    {
      "epoch": 1.7138364779874213,
      "grad_norm": 0.5710911750793457,
      "learning_rate": 4.283280922431866e-05,
      "loss": 0.5795,
      "num_input_tokens_seen": 2134168,
      "step": 3270
    },
    {
      "epoch": 1.7164570230607965,
      "grad_norm": 0.8757425546646118,
      "learning_rate": 4.289832285115305e-05,
      "loss": 0.5547,
      "num_input_tokens_seen": 2138360,
      "step": 3275
    },
    {
      "epoch": 1.719077568134172,
      "grad_norm": 0.5192080736160278,
      "learning_rate": 4.296383647798742e-05,
      "loss": 0.5276,
      "num_input_tokens_seen": 2141016,
      "step": 3280
    },
    {
      "epoch": 1.7216981132075473,
      "grad_norm": 0.5243817567825317,
      "learning_rate": 4.3029350104821806e-05,
      "loss": 0.6049,
      "num_input_tokens_seen": 2143448,
      "step": 3285
    },
    {
      "epoch": 1.7243186582809225,
      "grad_norm": 1.03437077999115,
      "learning_rate": 4.309486373165619e-05,
      "loss": 0.3771,
      "num_input_tokens_seen": 2146552,
      "step": 3290
    },
    {
      "epoch": 1.7269392033542976,
      "grad_norm": 0.5492845773696899,
      "learning_rate": 4.3160377358490565e-05,
      "loss": 0.5133,
      "num_input_tokens_seen": 2150424,
      "step": 3295
    },
    {
      "epoch": 1.7295597484276728,
      "grad_norm": 0.38959741592407227,
      "learning_rate": 4.322589098532495e-05,
      "loss": 0.4312,
      "num_input_tokens_seen": 2153912,
      "step": 3300
    },
    {
      "epoch": 1.7321802935010482,
      "grad_norm": 0.3663407564163208,
      "learning_rate": 4.329140461215933e-05,
      "loss": 0.561,
      "num_input_tokens_seen": 2156920,
      "step": 3305
    },
    {
      "epoch": 1.7348008385744236,
      "grad_norm": 0.3205285966396332,
      "learning_rate": 4.335691823899371e-05,
      "loss": 0.5903,
      "num_input_tokens_seen": 2159704,
      "step": 3310
    },
    {
      "epoch": 1.7374213836477987,
      "grad_norm": 0.6982426643371582,
      "learning_rate": 4.34224318658281e-05,
      "loss": 0.5563,
      "num_input_tokens_seen": 2162680,
      "step": 3315
    },
    {
      "epoch": 1.740041928721174,
      "grad_norm": 0.35760587453842163,
      "learning_rate": 4.3487945492662474e-05,
      "loss": 0.6011,
      "num_input_tokens_seen": 2166488,
      "step": 3320
    },
    {
      "epoch": 1.7426624737945493,
      "grad_norm": 0.5659242868423462,
      "learning_rate": 4.355345911949686e-05,
      "loss": 0.4908,
      "num_input_tokens_seen": 2170360,
      "step": 3325
    },
    {
      "epoch": 1.7452830188679245,
      "grad_norm": 0.3076644539833069,
      "learning_rate": 4.361897274633124e-05,
      "loss": 0.6442,
      "num_input_tokens_seen": 2173144,
      "step": 3330
    },
    {
      "epoch": 1.7479035639412999,
      "grad_norm": 0.4478340148925781,
      "learning_rate": 4.368448637316562e-05,
      "loss": 0.5543,
      "num_input_tokens_seen": 2176792,
      "step": 3335
    },
    {
      "epoch": 1.750524109014675,
      "grad_norm": 0.5510303378105164,
      "learning_rate": 4.375e-05,
      "loss": 0.5517,
      "num_input_tokens_seen": 2180408,
      "step": 3340
    },
    {
      "epoch": 1.7531446540880502,
      "grad_norm": 0.662065863609314,
      "learning_rate": 4.381551362683438e-05,
      "loss": 0.5787,
      "num_input_tokens_seen": 2183320,
      "step": 3345
    },
    {
      "epoch": 1.7557651991614256,
      "grad_norm": 0.7733682990074158,
      "learning_rate": 4.3881027253668766e-05,
      "loss": 0.5879,
      "num_input_tokens_seen": 2186040,
      "step": 3350
    },
    {
      "epoch": 1.758385744234801,
      "grad_norm": 0.8747007250785828,
      "learning_rate": 4.394654088050314e-05,
      "loss": 0.5554,
      "num_input_tokens_seen": 2188632,
      "step": 3355
    },
    {
      "epoch": 1.7610062893081762,
      "grad_norm": 0.43728509545326233,
      "learning_rate": 4.401205450733753e-05,
      "loss": 0.5109,
      "num_input_tokens_seen": 2192440,
      "step": 3360
    },
    {
      "epoch": 1.7636268343815513,
      "grad_norm": 0.3876517117023468,
      "learning_rate": 4.407756813417191e-05,
      "loss": 0.4198,
      "num_input_tokens_seen": 2196536,
      "step": 3365
    },
    {
      "epoch": 1.7662473794549265,
      "grad_norm": 0.3623616099357605,
      "learning_rate": 4.414308176100629e-05,
      "loss": 0.5541,
      "num_input_tokens_seen": 2202872,
      "step": 3370
    },
    {
      "epoch": 1.7688679245283019,
      "grad_norm": 1.04331374168396,
      "learning_rate": 4.4208595387840674e-05,
      "loss": 0.5394,
      "num_input_tokens_seen": 2205464,
      "step": 3375
    },
    {
      "epoch": 1.7714884696016773,
      "grad_norm": 0.44641828536987305,
      "learning_rate": 4.427410901467505e-05,
      "loss": 0.4966,
      "num_input_tokens_seen": 2208888,
      "step": 3380
    },
    {
      "epoch": 1.7741090146750524,
      "grad_norm": 0.6330898404121399,
      "learning_rate": 4.433962264150944e-05,
      "loss": 0.4899,
      "num_input_tokens_seen": 2212568,
      "step": 3385
    },
    {
      "epoch": 1.7767295597484276,
      "grad_norm": 0.3000768721103668,
      "learning_rate": 4.440513626834382e-05,
      "loss": 0.5056,
      "num_input_tokens_seen": 2216152,
      "step": 3390
    },
    {
      "epoch": 1.7793501048218028,
      "grad_norm": 0.5869066119194031,
      "learning_rate": 4.44706498951782e-05,
      "loss": 0.5117,
      "num_input_tokens_seen": 2218712,
      "step": 3395
    },
    {
      "epoch": 1.7819706498951782,
      "grad_norm": 0.34565454721450806,
      "learning_rate": 4.453616352201258e-05,
      "loss": 0.4969,
      "num_input_tokens_seen": 2222360,
      "step": 3400
    },
    {
      "epoch": 1.7845911949685536,
      "grad_norm": 0.8373364806175232,
      "learning_rate": 4.460167714884696e-05,
      "loss": 0.5031,
      "num_input_tokens_seen": 2224664,
      "step": 3405
    },
    {
      "epoch": 1.7872117400419287,
      "grad_norm": 0.4023176431655884,
      "learning_rate": 4.466719077568134e-05,
      "loss": 0.5729,
      "num_input_tokens_seen": 2227576,
      "step": 3410
    },
    {
      "epoch": 1.789832285115304,
      "grad_norm": 0.3760538399219513,
      "learning_rate": 4.4732704402515725e-05,
      "loss": 0.6054,
      "num_input_tokens_seen": 2230968,
      "step": 3415
    },
    {
      "epoch": 1.7924528301886793,
      "grad_norm": 0.5205124616622925,
      "learning_rate": 4.479821802935011e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 2234584,
      "step": 3420
    },
    {
      "epoch": 1.7950733752620545,
      "grad_norm": 0.7074105143547058,
      "learning_rate": 4.4863731656184485e-05,
      "loss": 0.4126,
      "num_input_tokens_seen": 2237400,
      "step": 3425
    },
    {
      "epoch": 1.7976939203354299,
      "grad_norm": 0.46054109930992126,
      "learning_rate": 4.4929245283018875e-05,
      "loss": 0.5586,
      "num_input_tokens_seen": 2239960,
      "step": 3430
    },
    {
      "epoch": 1.800314465408805,
      "grad_norm": 0.5416988134384155,
      "learning_rate": 4.499475890985325e-05,
      "loss": 0.4981,
      "num_input_tokens_seen": 2243192,
      "step": 3435
    },
    {
      "epoch": 1.8029350104821802,
      "grad_norm": 0.3993215560913086,
      "learning_rate": 4.5060272536687634e-05,
      "loss": 0.4226,
      "num_input_tokens_seen": 2246328,
      "step": 3440
    },
    {
      "epoch": 1.8055555555555556,
      "grad_norm": 0.38230982422828674,
      "learning_rate": 4.512578616352202e-05,
      "loss": 0.5255,
      "num_input_tokens_seen": 2249432,
      "step": 3445
    },
    {
      "epoch": 1.808176100628931,
      "grad_norm": 0.6288992166519165,
      "learning_rate": 4.519129979035639e-05,
      "loss": 0.4199,
      "num_input_tokens_seen": 2252184,
      "step": 3450
    },
    {
      "epoch": 1.8107966457023061,
      "grad_norm": 0.6012132167816162,
      "learning_rate": 4.525681341719078e-05,
      "loss": 0.5231,
      "num_input_tokens_seen": 2254808,
      "step": 3455
    },
    {
      "epoch": 1.8134171907756813,
      "grad_norm": 0.4171452820301056,
      "learning_rate": 4.532232704402516e-05,
      "loss": 0.4791,
      "num_input_tokens_seen": 2258552,
      "step": 3460
    },
    {
      "epoch": 1.8160377358490565,
      "grad_norm": 0.7967525720596313,
      "learning_rate": 4.5387840670859536e-05,
      "loss": 0.639,
      "num_input_tokens_seen": 2261208,
      "step": 3465
    },
    {
      "epoch": 1.8186582809224319,
      "grad_norm": 0.4459402561187744,
      "learning_rate": 4.5453354297693926e-05,
      "loss": 0.5562,
      "num_input_tokens_seen": 2264280,
      "step": 3470
    },
    {
      "epoch": 1.8212788259958073,
      "grad_norm": 0.6559849381446838,
      "learning_rate": 4.55188679245283e-05,
      "loss": 0.707,
      "num_input_tokens_seen": 2267672,
      "step": 3475
    },
    {
      "epoch": 1.8238993710691824,
      "grad_norm": 0.4562144875526428,
      "learning_rate": 4.5584381551362685e-05,
      "loss": 0.5314,
      "num_input_tokens_seen": 2270136,
      "step": 3480
    },
    {
      "epoch": 1.8265199161425576,
      "grad_norm": 0.899574339389801,
      "learning_rate": 4.564989517819707e-05,
      "loss": 0.5987,
      "num_input_tokens_seen": 2272888,
      "step": 3485
    },
    {
      "epoch": 1.8291404612159328,
      "grad_norm": 0.46711769700050354,
      "learning_rate": 4.571540880503145e-05,
      "loss": 0.3714,
      "num_input_tokens_seen": 2276600,
      "step": 3490
    },
    {
      "epoch": 1.8317610062893082,
      "grad_norm": 0.6603565812110901,
      "learning_rate": 4.578092243186583e-05,
      "loss": 0.6129,
      "num_input_tokens_seen": 2279320,
      "step": 3495
    },
    {
      "epoch": 1.8343815513626835,
      "grad_norm": 0.9465827345848083,
      "learning_rate": 4.584643605870021e-05,
      "loss": 0.5892,
      "num_input_tokens_seen": 2283512,
      "step": 3500
    },
    {
      "epoch": 1.8370020964360587,
      "grad_norm": 0.35687974095344543,
      "learning_rate": 4.5911949685534594e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 2286072,
      "step": 3505
    },
    {
      "epoch": 1.8396226415094339,
      "grad_norm": 0.4031253159046173,
      "learning_rate": 4.5977463312368977e-05,
      "loss": 0.5874,
      "num_input_tokens_seen": 2288824,
      "step": 3510
    },
    {
      "epoch": 1.8422431865828093,
      "grad_norm": 0.3238833546638489,
      "learning_rate": 4.604297693920336e-05,
      "loss": 0.5049,
      "num_input_tokens_seen": 2291992,
      "step": 3515
    },
    {
      "epoch": 1.8448637316561844,
      "grad_norm": 0.9428720474243164,
      "learning_rate": 4.6108490566037736e-05,
      "loss": 0.6334,
      "num_input_tokens_seen": 2295192,
      "step": 3520
    },
    {
      "epoch": 1.8474842767295598,
      "grad_norm": 0.46217650175094604,
      "learning_rate": 4.617400419287212e-05,
      "loss": 0.4811,
      "num_input_tokens_seen": 2298008,
      "step": 3525
    },
    {
      "epoch": 1.850104821802935,
      "grad_norm": 0.3184749186038971,
      "learning_rate": 4.62395178197065e-05,
      "loss": 0.5654,
      "num_input_tokens_seen": 2301432,
      "step": 3530
    },
    {
      "epoch": 1.8527253668763102,
      "grad_norm": 0.40300458669662476,
      "learning_rate": 4.630503144654088e-05,
      "loss": 0.473,
      "num_input_tokens_seen": 2304280,
      "step": 3535
    },
    {
      "epoch": 1.8553459119496856,
      "grad_norm": 0.3869777023792267,
      "learning_rate": 4.637054507337527e-05,
      "loss": 0.6017,
      "num_input_tokens_seen": 2307864,
      "step": 3540
    },
    {
      "epoch": 1.857966457023061,
      "grad_norm": 1.1070582866668701,
      "learning_rate": 4.6436058700209645e-05,
      "loss": 0.4826,
      "num_input_tokens_seen": 2311352,
      "step": 3545
    },
    {
      "epoch": 1.8605870020964361,
      "grad_norm": 0.3829704225063324,
      "learning_rate": 4.650157232704403e-05,
      "loss": 0.5316,
      "num_input_tokens_seen": 2315128,
      "step": 3550
    },
    {
      "epoch": 1.8632075471698113,
      "grad_norm": 0.3771264851093292,
      "learning_rate": 4.656708595387841e-05,
      "loss": 0.5501,
      "num_input_tokens_seen": 2318968,
      "step": 3555
    },
    {
      "epoch": 1.8658280922431865,
      "grad_norm": 0.38315507769584656,
      "learning_rate": 4.663259958071279e-05,
      "loss": 0.6393,
      "num_input_tokens_seen": 2321464,
      "step": 3560
    },
    {
      "epoch": 1.8684486373165619,
      "grad_norm": 0.463652104139328,
      "learning_rate": 4.669811320754717e-05,
      "loss": 0.4098,
      "num_input_tokens_seen": 2324312,
      "step": 3565
    },
    {
      "epoch": 1.8710691823899372,
      "grad_norm": 0.7069509029388428,
      "learning_rate": 4.676362683438155e-05,
      "loss": 0.4107,
      "num_input_tokens_seen": 2327384,
      "step": 3570
    },
    {
      "epoch": 1.8736897274633124,
      "grad_norm": 0.4640647768974304,
      "learning_rate": 4.6829140461215936e-05,
      "loss": 0.5304,
      "num_input_tokens_seen": 2330072,
      "step": 3575
    },
    {
      "epoch": 1.8763102725366876,
      "grad_norm": 0.9400482177734375,
      "learning_rate": 4.689465408805032e-05,
      "loss": 0.5746,
      "num_input_tokens_seen": 2332920,
      "step": 3580
    },
    {
      "epoch": 1.8789308176100628,
      "grad_norm": 0.4965188801288605,
      "learning_rate": 4.69601677148847e-05,
      "loss": 0.4664,
      "num_input_tokens_seen": 2335608,
      "step": 3585
    },
    {
      "epoch": 1.8815513626834381,
      "grad_norm": 0.514259934425354,
      "learning_rate": 4.702568134171908e-05,
      "loss": 0.4193,
      "num_input_tokens_seen": 2338360,
      "step": 3590
    },
    {
      "epoch": 1.8841719077568135,
      "grad_norm": 1.5675740242004395,
      "learning_rate": 4.709119496855346e-05,
      "loss": 0.6297,
      "num_input_tokens_seen": 2341592,
      "step": 3595
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 0.39940837025642395,
      "learning_rate": 4.7156708595387845e-05,
      "loss": 0.3096,
      "num_input_tokens_seen": 2344504,
      "step": 3600
    },
    {
      "epoch": 1.8894129979035639,
      "grad_norm": 0.3411221504211426,
      "learning_rate": 4.722222222222222e-05,
      "loss": 0.5702,
      "num_input_tokens_seen": 2347800,
      "step": 3605
    },
    {
      "epoch": 1.892033542976939,
      "grad_norm": 0.5294316411018372,
      "learning_rate": 4.728773584905661e-05,
      "loss": 0.5747,
      "num_input_tokens_seen": 2351320,
      "step": 3610
    },
    {
      "epoch": 1.8946540880503144,
      "grad_norm": 0.7676883935928345,
      "learning_rate": 4.735324947589099e-05,
      "loss": 0.5324,
      "num_input_tokens_seen": 2354296,
      "step": 3615
    },
    {
      "epoch": 1.8972746331236898,
      "grad_norm": 0.3678743839263916,
      "learning_rate": 4.7418763102725364e-05,
      "loss": 0.5953,
      "num_input_tokens_seen": 2357592,
      "step": 3620
    },
    {
      "epoch": 1.899895178197065,
      "grad_norm": 0.7132207155227661,
      "learning_rate": 4.7484276729559753e-05,
      "loss": 0.5417,
      "num_input_tokens_seen": 2360824,
      "step": 3625
    },
    {
      "epoch": 1.9025157232704402,
      "grad_norm": 0.4047158360481262,
      "learning_rate": 4.754979035639413e-05,
      "loss": 0.4918,
      "num_input_tokens_seen": 2364280,
      "step": 3630
    },
    {
      "epoch": 1.9051362683438156,
      "grad_norm": 0.5307246446609497,
      "learning_rate": 4.761530398322851e-05,
      "loss": 0.5774,
      "num_input_tokens_seen": 2367032,
      "step": 3635
    },
    {
      "epoch": 1.9077568134171907,
      "grad_norm": 0.7392288446426392,
      "learning_rate": 4.7680817610062896e-05,
      "loss": 0.549,
      "num_input_tokens_seen": 2369816,
      "step": 3640
    },
    {
      "epoch": 1.9103773584905661,
      "grad_norm": 0.29069793224334717,
      "learning_rate": 4.774633123689728e-05,
      "loss": 0.6359,
      "num_input_tokens_seen": 2373272,
      "step": 3645
    },
    {
      "epoch": 1.9129979035639413,
      "grad_norm": 0.4782866835594177,
      "learning_rate": 4.7811844863731655e-05,
      "loss": 0.4408,
      "num_input_tokens_seen": 2376184,
      "step": 3650
    },
    {
      "epoch": 1.9156184486373165,
      "grad_norm": 0.6489821672439575,
      "learning_rate": 4.787735849056604e-05,
      "loss": 0.5674,
      "num_input_tokens_seen": 2379672,
      "step": 3655
    },
    {
      "epoch": 1.9182389937106918,
      "grad_norm": 0.38305118680000305,
      "learning_rate": 4.794287211740042e-05,
      "loss": 0.6945,
      "num_input_tokens_seen": 2383032,
      "step": 3660
    },
    {
      "epoch": 1.9208595387840672,
      "grad_norm": 0.39953237771987915,
      "learning_rate": 4.8008385744234804e-05,
      "loss": 0.552,
      "num_input_tokens_seen": 2385592,
      "step": 3665
    },
    {
      "epoch": 1.9234800838574424,
      "grad_norm": 0.38890591263771057,
      "learning_rate": 4.807389937106919e-05,
      "loss": 0.4464,
      "num_input_tokens_seen": 2389176,
      "step": 3670
    },
    {
      "epoch": 1.9261006289308176,
      "grad_norm": 0.9633838534355164,
      "learning_rate": 4.8139412997903564e-05,
      "loss": 0.5178,
      "num_input_tokens_seen": 2392216,
      "step": 3675
    },
    {
      "epoch": 1.9287211740041927,
      "grad_norm": 0.6336709260940552,
      "learning_rate": 4.8204926624737954e-05,
      "loss": 0.6338,
      "num_input_tokens_seen": 2395608,
      "step": 3680
    },
    {
      "epoch": 1.9313417190775681,
      "grad_norm": 0.8016367554664612,
      "learning_rate": 4.827044025157233e-05,
      "loss": 0.6554,
      "num_input_tokens_seen": 2398744,
      "step": 3685
    },
    {
      "epoch": 1.9339622641509435,
      "grad_norm": 0.5422064661979675,
      "learning_rate": 4.8335953878406706e-05,
      "loss": 0.4018,
      "num_input_tokens_seen": 2401720,
      "step": 3690
    },
    {
      "epoch": 1.9365828092243187,
      "grad_norm": 0.7762633562088013,
      "learning_rate": 4.8401467505241096e-05,
      "loss": 0.4325,
      "num_input_tokens_seen": 2404760,
      "step": 3695
    },
    {
      "epoch": 1.9392033542976939,
      "grad_norm": 1.024160623550415,
      "learning_rate": 4.846698113207547e-05,
      "loss": 0.6919,
      "num_input_tokens_seen": 2408024,
      "step": 3700
    },
    {
      "epoch": 1.941823899371069,
      "grad_norm": 0.7347580194473267,
      "learning_rate": 4.8532494758909855e-05,
      "loss": 0.5878,
      "num_input_tokens_seen": 2411160,
      "step": 3705
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 0.6652771234512329,
      "learning_rate": 4.859800838574424e-05,
      "loss": 0.5861,
      "num_input_tokens_seen": 2414296,
      "step": 3710
    },
    {
      "epoch": 1.9470649895178198,
      "grad_norm": 0.6553481817245483,
      "learning_rate": 4.8663522012578615e-05,
      "loss": 0.5468,
      "num_input_tokens_seen": 2418456,
      "step": 3715
    },
    {
      "epoch": 1.949685534591195,
      "grad_norm": 0.8688281774520874,
      "learning_rate": 4.8729035639413e-05,
      "loss": 0.6089,
      "num_input_tokens_seen": 2420952,
      "step": 3720
    },
    {
      "epoch": 1.9523060796645701,
      "grad_norm": 0.4201144576072693,
      "learning_rate": 4.879454926624738e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 2424472,
      "step": 3725
    },
    {
      "epoch": 1.9549266247379455,
      "grad_norm": 0.7682141065597534,
      "learning_rate": 4.8860062893081764e-05,
      "loss": 0.67,
      "num_input_tokens_seen": 2427064,
      "step": 3730
    },
    {
      "epoch": 1.9575471698113207,
      "grad_norm": 0.8016714453697205,
      "learning_rate": 4.892557651991615e-05,
      "loss": 0.5774,
      "num_input_tokens_seen": 2430072,
      "step": 3735
    },
    {
      "epoch": 1.960167714884696,
      "grad_norm": 0.5208331942558289,
      "learning_rate": 4.899109014675053e-05,
      "loss": 0.3919,
      "num_input_tokens_seen": 2433016,
      "step": 3740
    },
    {
      "epoch": 1.9627882599580713,
      "grad_norm": 0.3646230101585388,
      "learning_rate": 4.9056603773584906e-05,
      "loss": 0.6115,
      "num_input_tokens_seen": 2436536,
      "step": 3745
    },
    {
      "epoch": 1.9654088050314464,
      "grad_norm": 0.4526829719543457,
      "learning_rate": 4.912211740041929e-05,
      "loss": 0.5014,
      "num_input_tokens_seen": 2439736,
      "step": 3750
    },
    {
      "epoch": 1.9680293501048218,
      "grad_norm": 0.4572622776031494,
      "learning_rate": 4.918763102725367e-05,
      "loss": 0.5685,
      "num_input_tokens_seen": 2443576,
      "step": 3755
    },
    {
      "epoch": 1.9706498951781972,
      "grad_norm": 0.415002703666687,
      "learning_rate": 4.925314465408805e-05,
      "loss": 0.4648,
      "num_input_tokens_seen": 2447128,
      "step": 3760
    },
    {
      "epoch": 1.9732704402515724,
      "grad_norm": 0.40333011746406555,
      "learning_rate": 4.931865828092244e-05,
      "loss": 0.5066,
      "num_input_tokens_seen": 2449944,
      "step": 3765
    },
    {
      "epoch": 1.9758909853249476,
      "grad_norm": 0.24664585292339325,
      "learning_rate": 4.9384171907756815e-05,
      "loss": 0.4156,
      "num_input_tokens_seen": 2455384,
      "step": 3770
    },
    {
      "epoch": 1.9785115303983227,
      "grad_norm": 0.5215503573417664,
      "learning_rate": 4.944968553459119e-05,
      "loss": 0.6974,
      "num_input_tokens_seen": 2459512,
      "step": 3775
    },
    {
      "epoch": 1.9811320754716981,
      "grad_norm": 0.34024202823638916,
      "learning_rate": 4.951519916142558e-05,
      "loss": 0.6634,
      "num_input_tokens_seen": 2461688,
      "step": 3780
    },
    {
      "epoch": 1.9837526205450735,
      "grad_norm": 0.41047102212905884,
      "learning_rate": 4.958071278825996e-05,
      "loss": 0.4915,
      "num_input_tokens_seen": 2468504,
      "step": 3785
    },
    {
      "epoch": 1.9863731656184487,
      "grad_norm": 0.44424372911453247,
      "learning_rate": 4.964622641509434e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 2471384,
      "step": 3790
    },
    {
      "epoch": 1.9889937106918238,
      "grad_norm": 0.4132085144519806,
      "learning_rate": 4.9711740041928724e-05,
      "loss": 0.4236,
      "num_input_tokens_seen": 2474168,
      "step": 3795
    },
    {
      "epoch": 1.991614255765199,
      "grad_norm": 0.44106680154800415,
      "learning_rate": 4.977725366876311e-05,
      "loss": 0.4003,
      "num_input_tokens_seen": 2481240,
      "step": 3800
    },
    {
      "epoch": 1.9942348008385744,
      "grad_norm": 0.4457460641860962,
      "learning_rate": 4.984276729559749e-05,
      "loss": 0.6052,
      "num_input_tokens_seen": 2485048,
      "step": 3805
    },
    {
      "epoch": 1.9968553459119498,
      "grad_norm": 0.2677628993988037,
      "learning_rate": 4.9908280922431866e-05,
      "loss": 0.5239,
      "num_input_tokens_seen": 2487896,
      "step": 3810
    },
    {
      "epoch": 1.999475890985325,
      "grad_norm": 0.35182440280914307,
      "learning_rate": 4.997379454926625e-05,
      "loss": 0.5014,
      "num_input_tokens_seen": 2495960,
      "step": 3815
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5206274390220642,
      "eval_runtime": 9.7561,
      "eval_samples_per_second": 86.92,
      "eval_steps_per_second": 21.73,
      "num_input_tokens_seen": 2496088,
      "step": 3816
    },
    {
      "epoch": 2.0020964360587,
      "grad_norm": 0.7511436343193054,
      "learning_rate": 4.999999905865133e-05,
      "loss": 0.54,
      "num_input_tokens_seen": 2498744,
      "step": 3820
    },
    {
      "epoch": 2.0047169811320753,
      "grad_norm": 0.6964349150657654,
      "learning_rate": 4.999999330596526e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 2501560,
      "step": 3825
    },
    {
      "epoch": 2.007337526205451,
      "grad_norm": 0.5859907269477844,
      "learning_rate": 4.9999982323565806e-05,
      "loss": 0.5341,
      "num_input_tokens_seen": 2505528,
      "step": 3830
    },
    {
      "epoch": 2.009958071278826,
      "grad_norm": 1.0420702695846558,
      "learning_rate": 4.999996611145527e-05,
      "loss": 0.6803,
      "num_input_tokens_seen": 2507992,
      "step": 3835
    },
    {
      "epoch": 2.0125786163522013,
      "grad_norm": 0.37191256880760193,
      "learning_rate": 4.9999944669637024e-05,
      "loss": 0.5693,
      "num_input_tokens_seen": 2511416,
      "step": 3840
    },
    {
      "epoch": 2.0151991614255764,
      "grad_norm": 0.59185391664505,
      "learning_rate": 4.999991799811557e-05,
      "loss": 0.4314,
      "num_input_tokens_seen": 2514264,
      "step": 3845
    },
    {
      "epoch": 2.0178197064989516,
      "grad_norm": 0.8174664974212646,
      "learning_rate": 4.99998860968965e-05,
      "loss": 0.5463,
      "num_input_tokens_seen": 2518424,
      "step": 3850
    },
    {
      "epoch": 2.020440251572327,
      "grad_norm": 0.46444445848464966,
      "learning_rate": 4.999984896598646e-05,
      "loss": 0.5529,
      "num_input_tokens_seen": 2521048,
      "step": 3855
    },
    {
      "epoch": 2.0230607966457024,
      "grad_norm": 0.7258176803588867,
      "learning_rate": 4.999980660539323e-05,
      "loss": 0.5389,
      "num_input_tokens_seen": 2525176,
      "step": 3860
    },
    {
      "epoch": 2.0256813417190775,
      "grad_norm": 0.40193524956703186,
      "learning_rate": 4.999975901512567e-05,
      "loss": 0.4955,
      "num_input_tokens_seen": 2528568,
      "step": 3865
    },
    {
      "epoch": 2.0283018867924527,
      "grad_norm": 0.6566658020019531,
      "learning_rate": 4.999970619519374e-05,
      "loss": 0.5365,
      "num_input_tokens_seen": 2531832,
      "step": 3870
    },
    {
      "epoch": 2.030922431865828,
      "grad_norm": 0.5370175242424011,
      "learning_rate": 4.999964814560849e-05,
      "loss": 0.4837,
      "num_input_tokens_seen": 2535224,
      "step": 3875
    },
    {
      "epoch": 2.0335429769392035,
      "grad_norm": 0.3648800551891327,
      "learning_rate": 4.999958486638205e-05,
      "loss": 0.5541,
      "num_input_tokens_seen": 2538456,
      "step": 3880
    },
    {
      "epoch": 2.0361635220125787,
      "grad_norm": 0.33327749371528625,
      "learning_rate": 4.999951635752767e-05,
      "loss": 0.6136,
      "num_input_tokens_seen": 2542040,
      "step": 3885
    },
    {
      "epoch": 2.038784067085954,
      "grad_norm": 0.3474299907684326,
      "learning_rate": 4.999944261905968e-05,
      "loss": 0.4271,
      "num_input_tokens_seen": 2544920,
      "step": 3890
    },
    {
      "epoch": 2.041404612159329,
      "grad_norm": 0.39758527278900146,
      "learning_rate": 4.99993636509935e-05,
      "loss": 0.5317,
      "num_input_tokens_seen": 2548088,
      "step": 3895
    },
    {
      "epoch": 2.0440251572327046,
      "grad_norm": 0.39773112535476685,
      "learning_rate": 4.999927945334565e-05,
      "loss": 0.5647,
      "num_input_tokens_seen": 2551704,
      "step": 3900
    },
    {
      "epoch": 2.04664570230608,
      "grad_norm": 0.46096205711364746,
      "learning_rate": 4.9999190026133745e-05,
      "loss": 0.5165,
      "num_input_tokens_seen": 2555320,
      "step": 3905
    },
    {
      "epoch": 2.049266247379455,
      "grad_norm": 0.7689683437347412,
      "learning_rate": 4.9999095369376506e-05,
      "loss": 0.5521,
      "num_input_tokens_seen": 2558808,
      "step": 3910
    },
    {
      "epoch": 2.05188679245283,
      "grad_norm": 2.071016788482666,
      "learning_rate": 4.99989954830937e-05,
      "loss": 0.6305,
      "num_input_tokens_seen": 2561976,
      "step": 3915
    },
    {
      "epoch": 2.0545073375262053,
      "grad_norm": 0.4992186427116394,
      "learning_rate": 4.9998890367306255e-05,
      "loss": 0.5605,
      "num_input_tokens_seen": 2564696,
      "step": 3920
    },
    {
      "epoch": 2.057127882599581,
      "grad_norm": 1.4659568071365356,
      "learning_rate": 4.999878002203614e-05,
      "loss": 0.5265,
      "num_input_tokens_seen": 2567160,
      "step": 3925
    },
    {
      "epoch": 2.059748427672956,
      "grad_norm": 0.5348338484764099,
      "learning_rate": 4.999866444730645e-05,
      "loss": 0.4322,
      "num_input_tokens_seen": 2570520,
      "step": 3930
    },
    {
      "epoch": 2.0623689727463312,
      "grad_norm": 0.5751182436943054,
      "learning_rate": 4.999854364314136e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 2573528,
      "step": 3935
    },
    {
      "epoch": 2.0649895178197064,
      "grad_norm": 0.5186047554016113,
      "learning_rate": 4.999841760956614e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 2575864,
      "step": 3940
    },
    {
      "epoch": 2.0676100628930816,
      "grad_norm": 0.43637773394584656,
      "learning_rate": 4.999828634660715e-05,
      "loss": 0.4781,
      "num_input_tokens_seen": 2578776,
      "step": 3945
    },
    {
      "epoch": 2.070230607966457,
      "grad_norm": 0.4172484576702118,
      "learning_rate": 4.9998149854291856e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 2581240,
      "step": 3950
    },
    {
      "epoch": 2.0728511530398324,
      "grad_norm": 0.6722705960273743,
      "learning_rate": 4.99980081326488e-05,
      "loss": 0.7482,
      "num_input_tokens_seen": 2584056,
      "step": 3955
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 0.8101542592048645,
      "learning_rate": 4.999786118170763e-05,
      "loss": 0.6257,
      "num_input_tokens_seen": 2587032,
      "step": 3960
    },
    {
      "epoch": 2.0780922431865827,
      "grad_norm": 0.43424010276794434,
      "learning_rate": 4.99977090014991e-05,
      "loss": 0.4848,
      "num_input_tokens_seen": 2590424,
      "step": 3965
    },
    {
      "epoch": 2.080712788259958,
      "grad_norm": 0.7417204976081848,
      "learning_rate": 4.9997551592055034e-05,
      "loss": 0.7069,
      "num_input_tokens_seen": 2594040,
      "step": 3970
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 0.8839415907859802,
      "learning_rate": 4.999738895340836e-05,
      "loss": 0.5251,
      "num_input_tokens_seen": 2596856,
      "step": 3975
    },
    {
      "epoch": 2.0859538784067087,
      "grad_norm": 1.2637219429016113,
      "learning_rate": 4.999722108559311e-05,
      "loss": 0.5906,
      "num_input_tokens_seen": 2599736,
      "step": 3980
    },
    {
      "epoch": 2.088574423480084,
      "grad_norm": 0.3105624318122864,
      "learning_rate": 4.9997047988644385e-05,
      "loss": 0.3688,
      "num_input_tokens_seen": 2602392,
      "step": 3985
    },
    {
      "epoch": 2.091194968553459,
      "grad_norm": 1.1363563537597656,
      "learning_rate": 4.99968696625984e-05,
      "loss": 0.5177,
      "num_input_tokens_seen": 2605592,
      "step": 3990
    },
    {
      "epoch": 2.0938155136268346,
      "grad_norm": 0.6050761938095093,
      "learning_rate": 4.999668610749246e-05,
      "loss": 0.4383,
      "num_input_tokens_seen": 2608760,
      "step": 3995
    },
    {
      "epoch": 2.0964360587002098,
      "grad_norm": 0.42769044637680054,
      "learning_rate": 4.999649732336498e-05,
      "loss": 0.5397,
      "num_input_tokens_seen": 2611672,
      "step": 4000
    },
    {
      "epoch": 2.099056603773585,
      "grad_norm": 0.933816134929657,
      "learning_rate": 4.999630331025542e-05,
      "loss": 0.5174,
      "num_input_tokens_seen": 2616504,
      "step": 4005
    },
    {
      "epoch": 2.10167714884696,
      "grad_norm": 0.4704139232635498,
      "learning_rate": 4.9996104068204385e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 2619320,
      "step": 4010
    },
    {
      "epoch": 2.1042976939203353,
      "grad_norm": 0.81191486120224,
      "learning_rate": 4.999589959725355e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 2621784,
      "step": 4015
    },
    {
      "epoch": 2.106918238993711,
      "grad_norm": 0.44585710763931274,
      "learning_rate": 4.999568989744568e-05,
      "loss": 0.6406,
      "num_input_tokens_seen": 2624792,
      "step": 4020
    },
    {
      "epoch": 2.109538784067086,
      "grad_norm": 0.5617390275001526,
      "learning_rate": 4.999547496882466e-05,
      "loss": 0.5409,
      "num_input_tokens_seen": 2628152,
      "step": 4025
    },
    {
      "epoch": 2.1121593291404612,
      "grad_norm": 0.9310269355773926,
      "learning_rate": 4.999525481143543e-05,
      "loss": 0.539,
      "num_input_tokens_seen": 2630648,
      "step": 4030
    },
    {
      "epoch": 2.1147798742138364,
      "grad_norm": 0.41864627599716187,
      "learning_rate": 4.999502942532406e-05,
      "loss": 0.5206,
      "num_input_tokens_seen": 2634040,
      "step": 4035
    },
    {
      "epoch": 2.1174004192872116,
      "grad_norm": 0.7571933269500732,
      "learning_rate": 4.999479881053769e-05,
      "loss": 0.4784,
      "num_input_tokens_seen": 2637304,
      "step": 4040
    },
    {
      "epoch": 2.120020964360587,
      "grad_norm": 0.5895537734031677,
      "learning_rate": 4.999456296712457e-05,
      "loss": 0.688,
      "num_input_tokens_seen": 2640024,
      "step": 4045
    },
    {
      "epoch": 2.1226415094339623,
      "grad_norm": 0.3248578608036041,
      "learning_rate": 4.9994321895134025e-05,
      "loss": 0.5412,
      "num_input_tokens_seen": 2642968,
      "step": 4050
    },
    {
      "epoch": 2.1252620545073375,
      "grad_norm": 0.5004737377166748,
      "learning_rate": 4.999407559461649e-05,
      "loss": 0.4592,
      "num_input_tokens_seen": 2646584,
      "step": 4055
    },
    {
      "epoch": 2.1278825995807127,
      "grad_norm": 0.4956149756908417,
      "learning_rate": 4.999382406562349e-05,
      "loss": 0.5951,
      "num_input_tokens_seen": 2649080,
      "step": 4060
    },
    {
      "epoch": 2.130503144654088,
      "grad_norm": 0.4603669047355652,
      "learning_rate": 4.999356730820765e-05,
      "loss": 0.5719,
      "num_input_tokens_seen": 2652600,
      "step": 4065
    },
    {
      "epoch": 2.1331236897274635,
      "grad_norm": 0.7570340037345886,
      "learning_rate": 4.999330532242266e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 2659192,
      "step": 4070
    },
    {
      "epoch": 2.1357442348008386,
      "grad_norm": 0.5477859377861023,
      "learning_rate": 4.999303810832333e-05,
      "loss": 0.6091,
      "num_input_tokens_seen": 2662968,
      "step": 4075
    },
    {
      "epoch": 2.138364779874214,
      "grad_norm": 0.60578852891922,
      "learning_rate": 4.999276566596557e-05,
      "loss": 0.7302,
      "num_input_tokens_seen": 2666776,
      "step": 4080
    },
    {
      "epoch": 2.140985324947589,
      "grad_norm": 0.645183265209198,
      "learning_rate": 4.9992487995406364e-05,
      "loss": 0.5483,
      "num_input_tokens_seen": 2669656,
      "step": 4085
    },
    {
      "epoch": 2.1436058700209646,
      "grad_norm": 0.4728403687477112,
      "learning_rate": 4.9992205096703806e-05,
      "loss": 0.5412,
      "num_input_tokens_seen": 2672984,
      "step": 4090
    },
    {
      "epoch": 2.1462264150943398,
      "grad_norm": 0.66242516040802,
      "learning_rate": 4.9991916969917054e-05,
      "loss": 0.6337,
      "num_input_tokens_seen": 2676792,
      "step": 4095
    },
    {
      "epoch": 2.148846960167715,
      "grad_norm": 0.6719081401824951,
      "learning_rate": 4.99916236151064e-05,
      "loss": 0.4684,
      "num_input_tokens_seen": 2679640,
      "step": 4100
    },
    {
      "epoch": 2.15146750524109,
      "grad_norm": 0.2834463119506836,
      "learning_rate": 4.999132503233321e-05,
      "loss": 0.5419,
      "num_input_tokens_seen": 2687448,
      "step": 4105
    },
    {
      "epoch": 2.1540880503144653,
      "grad_norm": 0.4085135757923126,
      "learning_rate": 4.999102122165994e-05,
      "loss": 0.6127,
      "num_input_tokens_seen": 2690104,
      "step": 4110
    },
    {
      "epoch": 2.156708595387841,
      "grad_norm": 0.3710445463657379,
      "learning_rate": 4.9990712183150135e-05,
      "loss": 0.5447,
      "num_input_tokens_seen": 2693400,
      "step": 4115
    },
    {
      "epoch": 2.159329140461216,
      "grad_norm": 0.668628454208374,
      "learning_rate": 4.9990397916868456e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 2696504,
      "step": 4120
    },
    {
      "epoch": 2.161949685534591,
      "grad_norm": 0.5318310856819153,
      "learning_rate": 4.999007842288064e-05,
      "loss": 0.4195,
      "num_input_tokens_seen": 2699768,
      "step": 4125
    },
    {
      "epoch": 2.1645702306079664,
      "grad_norm": 0.46130672097206116,
      "learning_rate": 4.9989753701253514e-05,
      "loss": 0.5846,
      "num_input_tokens_seen": 2703320,
      "step": 4130
    },
    {
      "epoch": 2.1671907756813416,
      "grad_norm": 2.608541250228882,
      "learning_rate": 4.998942375205502e-05,
      "loss": 0.5169,
      "num_input_tokens_seen": 2706264,
      "step": 4135
    },
    {
      "epoch": 2.169811320754717,
      "grad_norm": 0.3166894316673279,
      "learning_rate": 4.998908857535416e-05,
      "loss": 0.6247,
      "num_input_tokens_seen": 2713464,
      "step": 4140
    },
    {
      "epoch": 2.1724318658280923,
      "grad_norm": 0.3074808418750763,
      "learning_rate": 4.9988748171221065e-05,
      "loss": 0.5408,
      "num_input_tokens_seen": 2717912,
      "step": 4145
    },
    {
      "epoch": 2.1750524109014675,
      "grad_norm": 0.5366907715797424,
      "learning_rate": 4.998840253972695e-05,
      "loss": 0.5022,
      "num_input_tokens_seen": 2720504,
      "step": 4150
    },
    {
      "epoch": 2.1776729559748427,
      "grad_norm": 0.4556225836277008,
      "learning_rate": 4.9988051680944094e-05,
      "loss": 0.4573,
      "num_input_tokens_seen": 2723256,
      "step": 4155
    },
    {
      "epoch": 2.180293501048218,
      "grad_norm": 0.29861488938331604,
      "learning_rate": 4.998769559494591e-05,
      "loss": 0.4363,
      "num_input_tokens_seen": 2726488,
      "step": 4160
    },
    {
      "epoch": 2.1829140461215935,
      "grad_norm": 0.44566604495048523,
      "learning_rate": 4.998733428180689e-05,
      "loss": 0.5281,
      "num_input_tokens_seen": 2729880,
      "step": 4165
    },
    {
      "epoch": 2.1855345911949686,
      "grad_norm": 0.6150112152099609,
      "learning_rate": 4.998696774160259e-05,
      "loss": 0.5724,
      "num_input_tokens_seen": 2733080,
      "step": 4170
    },
    {
      "epoch": 2.188155136268344,
      "grad_norm": 0.24950844049453735,
      "learning_rate": 4.9986595974409714e-05,
      "loss": 0.491,
      "num_input_tokens_seen": 2736792,
      "step": 4175
    },
    {
      "epoch": 2.190775681341719,
      "grad_norm": 0.47618019580841064,
      "learning_rate": 4.998621898030603e-05,
      "loss": 0.6953,
      "num_input_tokens_seen": 2739480,
      "step": 4180
    },
    {
      "epoch": 2.1933962264150946,
      "grad_norm": 0.45911943912506104,
      "learning_rate": 4.998583675937039e-05,
      "loss": 0.5307,
      "num_input_tokens_seen": 2743512,
      "step": 4185
    },
    {
      "epoch": 2.1960167714884697,
      "grad_norm": 0.7535449266433716,
      "learning_rate": 4.998544931168274e-05,
      "loss": 0.5438,
      "num_input_tokens_seen": 2746744,
      "step": 4190
    },
    {
      "epoch": 2.198637316561845,
      "grad_norm": 0.4680337905883789,
      "learning_rate": 4.9985056637324154e-05,
      "loss": 0.5094,
      "num_input_tokens_seen": 2749496,
      "step": 4195
    },
    {
      "epoch": 2.20125786163522,
      "grad_norm": 0.40374091267585754,
      "learning_rate": 4.998465873637677e-05,
      "loss": 0.7912,
      "num_input_tokens_seen": 2752792,
      "step": 4200
    },
    {
      "epoch": 2.2038784067085953,
      "grad_norm": 0.393686980009079,
      "learning_rate": 4.9984255608923806e-05,
      "loss": 0.5853,
      "num_input_tokens_seen": 2755832,
      "step": 4205
    },
    {
      "epoch": 2.2064989517819704,
      "grad_norm": 0.30216431617736816,
      "learning_rate": 4.998384725504961e-05,
      "loss": 0.6287,
      "num_input_tokens_seen": 2758744,
      "step": 4210
    },
    {
      "epoch": 2.209119496855346,
      "grad_norm": 0.7025323510169983,
      "learning_rate": 4.9983433674839596e-05,
      "loss": 0.4638,
      "num_input_tokens_seen": 2762104,
      "step": 4215
    },
    {
      "epoch": 2.211740041928721,
      "grad_norm": 0.28503045439720154,
      "learning_rate": 4.998301486838028e-05,
      "loss": 0.4848,
      "num_input_tokens_seen": 2765144,
      "step": 4220
    },
    {
      "epoch": 2.2143605870020964,
      "grad_norm": 0.4115592837333679,
      "learning_rate": 4.998259083575929e-05,
      "loss": 0.4981,
      "num_input_tokens_seen": 2768760,
      "step": 4225
    },
    {
      "epoch": 2.2169811320754715,
      "grad_norm": 0.42924630641937256,
      "learning_rate": 4.99821615770653e-05,
      "loss": 0.5248,
      "num_input_tokens_seen": 2772056,
      "step": 4230
    },
    {
      "epoch": 2.219601677148847,
      "grad_norm": 0.4883922338485718,
      "learning_rate": 4.998172709238813e-05,
      "loss": 0.5385,
      "num_input_tokens_seen": 2775352,
      "step": 4235
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.23148590326309204,
      "learning_rate": 4.9981287381818654e-05,
      "loss": 0.4642,
      "num_input_tokens_seen": 2778840,
      "step": 4240
    },
    {
      "epoch": 2.2248427672955975,
      "grad_norm": 0.6354508399963379,
      "learning_rate": 4.9980842445448856e-05,
      "loss": 0.6381,
      "num_input_tokens_seen": 2781656,
      "step": 4245
    },
    {
      "epoch": 2.2274633123689727,
      "grad_norm": 1.726343035697937,
      "learning_rate": 4.9980392283371815e-05,
      "loss": 0.4473,
      "num_input_tokens_seen": 2784888,
      "step": 4250
    },
    {
      "epoch": 2.230083857442348,
      "grad_norm": 1.5061469078063965,
      "learning_rate": 4.99799368956817e-05,
      "loss": 0.6041,
      "num_input_tokens_seen": 2788024,
      "step": 4255
    },
    {
      "epoch": 2.2327044025157234,
      "grad_norm": 0.6899846196174622,
      "learning_rate": 4.997947628247378e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 2791928,
      "step": 4260
    },
    {
      "epoch": 2.2353249475890986,
      "grad_norm": 0.6667555570602417,
      "learning_rate": 4.9979010443844395e-05,
      "loss": 0.4577,
      "num_input_tokens_seen": 2794648,
      "step": 4265
    },
    {
      "epoch": 2.237945492662474,
      "grad_norm": 0.29128366708755493,
      "learning_rate": 4.9978539379891e-05,
      "loss": 0.2609,
      "num_input_tokens_seen": 2798648,
      "step": 4270
    },
    {
      "epoch": 2.240566037735849,
      "grad_norm": 0.25352323055267334,
      "learning_rate": 4.997806309071214e-05,
      "loss": 0.4563,
      "num_input_tokens_seen": 2803064,
      "step": 4275
    },
    {
      "epoch": 2.243186582809224,
      "grad_norm": 0.3880961537361145,
      "learning_rate": 4.997758157640745e-05,
      "loss": 0.4294,
      "num_input_tokens_seen": 2806200,
      "step": 4280
    },
    {
      "epoch": 2.2458071278825997,
      "grad_norm": 0.37594830989837646,
      "learning_rate": 4.997709483707765e-05,
      "loss": 0.6105,
      "num_input_tokens_seen": 2809816,
      "step": 4285
    },
    {
      "epoch": 2.248427672955975,
      "grad_norm": 0.5933057069778442,
      "learning_rate": 4.997660287282456e-05,
      "loss": 0.538,
      "num_input_tokens_seen": 2812696,
      "step": 4290
    },
    {
      "epoch": 2.25104821802935,
      "grad_norm": 0.4574577212333679,
      "learning_rate": 4.99761056837511e-05,
      "loss": 0.4706,
      "num_input_tokens_seen": 2815992,
      "step": 4295
    },
    {
      "epoch": 2.2536687631027252,
      "grad_norm": 0.37928304076194763,
      "learning_rate": 4.997560326996128e-05,
      "loss": 0.5503,
      "num_input_tokens_seen": 2818712,
      "step": 4300
    },
    {
      "epoch": 2.2562893081761004,
      "grad_norm": 0.3996679484844208,
      "learning_rate": 4.997509563156018e-05,
      "loss": 0.4563,
      "num_input_tokens_seen": 2822008,
      "step": 4305
    },
    {
      "epoch": 2.258909853249476,
      "grad_norm": 1.2197256088256836,
      "learning_rate": 4.997458276865401e-05,
      "loss": 0.6152,
      "num_input_tokens_seen": 2825176,
      "step": 4310
    },
    {
      "epoch": 2.261530398322851,
      "grad_norm": 0.5185303688049316,
      "learning_rate": 4.997406468135005e-05,
      "loss": 0.5629,
      "num_input_tokens_seen": 2827576,
      "step": 4315
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 0.649246096611023,
      "learning_rate": 4.9973541369756686e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 2830872,
      "step": 4320
    },
    {
      "epoch": 2.2667714884696015,
      "grad_norm": 1.144702434539795,
      "learning_rate": 4.9973012833983365e-05,
      "loss": 0.533,
      "num_input_tokens_seen": 2834392,
      "step": 4325
    },
    {
      "epoch": 2.269392033542977,
      "grad_norm": 0.4230138659477234,
      "learning_rate": 4.9972479074140684e-05,
      "loss": 0.4977,
      "num_input_tokens_seen": 2838424,
      "step": 4330
    },
    {
      "epoch": 2.2720125786163523,
      "grad_norm": 0.548458456993103,
      "learning_rate": 4.9971940090340276e-05,
      "loss": 0.557,
      "num_input_tokens_seen": 2841144,
      "step": 4335
    },
    {
      "epoch": 2.2746331236897275,
      "grad_norm": 0.8104798197746277,
      "learning_rate": 4.997139588269489e-05,
      "loss": 0.5624,
      "num_input_tokens_seen": 2843992,
      "step": 4340
    },
    {
      "epoch": 2.2772536687631026,
      "grad_norm": 0.8014077544212341,
      "learning_rate": 4.997084645131838e-05,
      "loss": 0.5462,
      "num_input_tokens_seen": 2846200,
      "step": 4345
    },
    {
      "epoch": 2.279874213836478,
      "grad_norm": 0.3458859324455261,
      "learning_rate": 4.9970291796325666e-05,
      "loss": 0.6649,
      "num_input_tokens_seen": 2851736,
      "step": 4350
    },
    {
      "epoch": 2.2824947589098534,
      "grad_norm": 0.3311760723590851,
      "learning_rate": 4.996973191783279e-05,
      "loss": 0.4198,
      "num_input_tokens_seen": 2855320,
      "step": 4355
    },
    {
      "epoch": 2.2851153039832286,
      "grad_norm": 0.32998061180114746,
      "learning_rate": 4.996916681595687e-05,
      "loss": 0.4952,
      "num_input_tokens_seen": 2857976,
      "step": 4360
    },
    {
      "epoch": 2.2877358490566038,
      "grad_norm": 0.43134158849716187,
      "learning_rate": 4.996859649081611e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 2860792,
      "step": 4365
    },
    {
      "epoch": 2.290356394129979,
      "grad_norm": 0.4469950199127197,
      "learning_rate": 4.996802094252983e-05,
      "loss": 0.6489,
      "num_input_tokens_seen": 2864120,
      "step": 4370
    },
    {
      "epoch": 2.2929769392033545,
      "grad_norm": 0.3707418441772461,
      "learning_rate": 4.996744017121841e-05,
      "loss": 0.4333,
      "num_input_tokens_seen": 2871288,
      "step": 4375
    },
    {
      "epoch": 2.2955974842767297,
      "grad_norm": 0.26568499207496643,
      "learning_rate": 4.9966854177003355e-05,
      "loss": 0.4187,
      "num_input_tokens_seen": 2876056,
      "step": 4380
    },
    {
      "epoch": 2.298218029350105,
      "grad_norm": 0.4062865972518921,
      "learning_rate": 4.996626296000724e-05,
      "loss": 0.6648,
      "num_input_tokens_seen": 2879544,
      "step": 4385
    },
    {
      "epoch": 2.30083857442348,
      "grad_norm": 0.5014866590499878,
      "learning_rate": 4.996566652035374e-05,
      "loss": 0.5418,
      "num_input_tokens_seen": 2883384,
      "step": 4390
    },
    {
      "epoch": 2.3034591194968552,
      "grad_norm": 1.1835707426071167,
      "learning_rate": 4.996506485816763e-05,
      "loss": 0.5579,
      "num_input_tokens_seen": 2886360,
      "step": 4395
    },
    {
      "epoch": 2.3060796645702304,
      "grad_norm": 0.4409770667552948,
      "learning_rate": 4.996445797357477e-05,
      "loss": 0.5004,
      "num_input_tokens_seen": 2889688,
      "step": 4400
    },
    {
      "epoch": 2.308700209643606,
      "grad_norm": 0.4658597409725189,
      "learning_rate": 4.996384586670211e-05,
      "loss": 0.5097,
      "num_input_tokens_seen": 2892536,
      "step": 4405
    },
    {
      "epoch": 2.311320754716981,
      "grad_norm": 0.6187301278114319,
      "learning_rate": 4.99632285376777e-05,
      "loss": 0.4111,
      "num_input_tokens_seen": 2895928,
      "step": 4410
    },
    {
      "epoch": 2.3139412997903563,
      "grad_norm": 0.289286732673645,
      "learning_rate": 4.9962605986630674e-05,
      "loss": 0.4315,
      "num_input_tokens_seen": 2899640,
      "step": 4415
    },
    {
      "epoch": 2.3165618448637315,
      "grad_norm": 0.35269540548324585,
      "learning_rate": 4.996197821369126e-05,
      "loss": 0.4455,
      "num_input_tokens_seen": 2902648,
      "step": 4420
    },
    {
      "epoch": 2.319182389937107,
      "grad_norm": 0.5078797340393066,
      "learning_rate": 4.9961345218990796e-05,
      "loss": 0.7057,
      "num_input_tokens_seen": 2905784,
      "step": 4425
    },
    {
      "epoch": 2.3218029350104823,
      "grad_norm": 0.4081362187862396,
      "learning_rate": 4.996070700266168e-05,
      "loss": 0.441,
      "num_input_tokens_seen": 2909560,
      "step": 4430
    },
    {
      "epoch": 2.3244234800838575,
      "grad_norm": 0.6392662525177002,
      "learning_rate": 4.9960063564837424e-05,
      "loss": 0.4506,
      "num_input_tokens_seen": 2911992,
      "step": 4435
    },
    {
      "epoch": 2.3270440251572326,
      "grad_norm": 0.4972396492958069,
      "learning_rate": 4.995941490565262e-05,
      "loss": 0.4916,
      "num_input_tokens_seen": 2914744,
      "step": 4440
    },
    {
      "epoch": 2.329664570230608,
      "grad_norm": 0.6579731702804565,
      "learning_rate": 4.995876102524299e-05,
      "loss": 0.5187,
      "num_input_tokens_seen": 2917272,
      "step": 4445
    },
    {
      "epoch": 2.3322851153039834,
      "grad_norm": 0.4633757472038269,
      "learning_rate": 4.995810192374529e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 2920280,
      "step": 4450
    },
    {
      "epoch": 2.3349056603773586,
      "grad_norm": 0.6633248925209045,
      "learning_rate": 4.99574376012974e-05,
      "loss": 0.6079,
      "num_input_tokens_seen": 2922840,
      "step": 4455
    },
    {
      "epoch": 2.3375262054507338,
      "grad_norm": 0.4691452980041504,
      "learning_rate": 4.995676805803829e-05,
      "loss": 0.4555,
      "num_input_tokens_seen": 2925720,
      "step": 4460
    },
    {
      "epoch": 2.340146750524109,
      "grad_norm": 0.3186036944389343,
      "learning_rate": 4.995609329410804e-05,
      "loss": 0.7282,
      "num_input_tokens_seen": 2929656,
      "step": 4465
    },
    {
      "epoch": 2.342767295597484,
      "grad_norm": 0.27361705899238586,
      "learning_rate": 4.9955413309647776e-05,
      "loss": 0.3704,
      "num_input_tokens_seen": 2932568,
      "step": 4470
    },
    {
      "epoch": 2.3453878406708597,
      "grad_norm": 0.5090540051460266,
      "learning_rate": 4.9954728104799756e-05,
      "loss": 0.4646,
      "num_input_tokens_seen": 2935384,
      "step": 4475
    },
    {
      "epoch": 2.348008385744235,
      "grad_norm": 0.475938618183136,
      "learning_rate": 4.995403767970731e-05,
      "loss": 0.5581,
      "num_input_tokens_seen": 2938712,
      "step": 4480
    },
    {
      "epoch": 2.35062893081761,
      "grad_norm": 0.251338928937912,
      "learning_rate": 4.995334203451488e-05,
      "loss": 0.6251,
      "num_input_tokens_seen": 2943960,
      "step": 4485
    },
    {
      "epoch": 2.353249475890985,
      "grad_norm": 0.5829347372055054,
      "learning_rate": 4.995264116936798e-05,
      "loss": 0.4981,
      "num_input_tokens_seen": 2946840,
      "step": 4490
    },
    {
      "epoch": 2.3558700209643604,
      "grad_norm": 0.48063093423843384,
      "learning_rate": 4.995193508441322e-05,
      "loss": 0.5158,
      "num_input_tokens_seen": 2950488,
      "step": 4495
    },
    {
      "epoch": 2.358490566037736,
      "grad_norm": 0.31426766514778137,
      "learning_rate": 4.99512237797983e-05,
      "loss": 0.5458,
      "num_input_tokens_seen": 2954072,
      "step": 4500
    },
    {
      "epoch": 2.361111111111111,
      "grad_norm": 0.39509421586990356,
      "learning_rate": 4.995050725567203e-05,
      "loss": 0.6154,
      "num_input_tokens_seen": 2957048,
      "step": 4505
    },
    {
      "epoch": 2.3637316561844863,
      "grad_norm": 0.6882928609848022,
      "learning_rate": 4.994978551218429e-05,
      "loss": 0.5144,
      "num_input_tokens_seen": 2960760,
      "step": 4510
    },
    {
      "epoch": 2.3663522012578615,
      "grad_norm": 0.4214285612106323,
      "learning_rate": 4.9949058549486066e-05,
      "loss": 0.4298,
      "num_input_tokens_seen": 2965368,
      "step": 4515
    },
    {
      "epoch": 2.368972746331237,
      "grad_norm": 0.445468008518219,
      "learning_rate": 4.994832636772943e-05,
      "loss": 0.4657,
      "num_input_tokens_seen": 2968728,
      "step": 4520
    },
    {
      "epoch": 2.3715932914046123,
      "grad_norm": 0.5150705575942993,
      "learning_rate": 4.994758896706754e-05,
      "loss": 0.4976,
      "num_input_tokens_seen": 2972312,
      "step": 4525
    },
    {
      "epoch": 2.3742138364779874,
      "grad_norm": 0.6001101732254028,
      "learning_rate": 4.994684634765465e-05,
      "loss": 0.3711,
      "num_input_tokens_seen": 2975608,
      "step": 4530
    },
    {
      "epoch": 2.3768343815513626,
      "grad_norm": 1.2575628757476807,
      "learning_rate": 4.9946098509646125e-05,
      "loss": 0.4835,
      "num_input_tokens_seen": 2978872,
      "step": 4535
    },
    {
      "epoch": 2.379454926624738,
      "grad_norm": 0.2863723635673523,
      "learning_rate": 4.994534545319838e-05,
      "loss": 0.4415,
      "num_input_tokens_seen": 2982136,
      "step": 4540
    },
    {
      "epoch": 2.3820754716981134,
      "grad_norm": 0.4473058879375458,
      "learning_rate": 4.994458717846896e-05,
      "loss": 0.6047,
      "num_input_tokens_seen": 2985624,
      "step": 4545
    },
    {
      "epoch": 2.3846960167714886,
      "grad_norm": 0.539219856262207,
      "learning_rate": 4.9943823685616485e-05,
      "loss": 0.5013,
      "num_input_tokens_seen": 2988856,
      "step": 4550
    },
    {
      "epoch": 2.3873165618448637,
      "grad_norm": 0.6163763403892517,
      "learning_rate": 4.994305497480067e-05,
      "loss": 0.4512,
      "num_input_tokens_seen": 2991608,
      "step": 4555
    },
    {
      "epoch": 2.389937106918239,
      "grad_norm": 0.4424918591976166,
      "learning_rate": 4.994228104618233e-05,
      "loss": 0.601,
      "num_input_tokens_seen": 2995064,
      "step": 4560
    },
    {
      "epoch": 2.392557651991614,
      "grad_norm": 0.6305143237113953,
      "learning_rate": 4.994150189992334e-05,
      "loss": 0.6399,
      "num_input_tokens_seen": 2998488,
      "step": 4565
    },
    {
      "epoch": 2.3951781970649897,
      "grad_norm": 0.3525293469429016,
      "learning_rate": 4.99407175361867e-05,
      "loss": 0.4563,
      "num_input_tokens_seen": 3001656,
      "step": 4570
    },
    {
      "epoch": 2.397798742138365,
      "grad_norm": 0.5423989295959473,
      "learning_rate": 4.99399279551365e-05,
      "loss": 0.6865,
      "num_input_tokens_seen": 3005688,
      "step": 4575
    },
    {
      "epoch": 2.40041928721174,
      "grad_norm": 0.4803548753261566,
      "learning_rate": 4.9939133156937886e-05,
      "loss": 0.5511,
      "num_input_tokens_seen": 3009176,
      "step": 4580
    },
    {
      "epoch": 2.403039832285115,
      "grad_norm": 0.6746698021888733,
      "learning_rate": 4.993833314175716e-05,
      "loss": 0.5707,
      "num_input_tokens_seen": 3012344,
      "step": 4585
    },
    {
      "epoch": 2.4056603773584904,
      "grad_norm": 0.47236168384552,
      "learning_rate": 4.993752790976163e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 3015640,
      "step": 4590
    },
    {
      "epoch": 2.408280922431866,
      "grad_norm": 0.43844524025917053,
      "learning_rate": 4.993671746111977e-05,
      "loss": 0.5163,
      "num_input_tokens_seen": 3019224,
      "step": 4595
    },
    {
      "epoch": 2.410901467505241,
      "grad_norm": 0.6316875219345093,
      "learning_rate": 4.993590179600111e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 3021944,
      "step": 4600
    },
    {
      "epoch": 2.4135220125786163,
      "grad_norm": 0.4604470729827881,
      "learning_rate": 4.9935080914576284e-05,
      "loss": 0.5381,
      "num_input_tokens_seen": 3025368,
      "step": 4605
    },
    {
      "epoch": 2.4161425576519915,
      "grad_norm": 0.5769137740135193,
      "learning_rate": 4.9934254817016995e-05,
      "loss": 0.5057,
      "num_input_tokens_seen": 3027928,
      "step": 4610
    },
    {
      "epoch": 2.418763102725367,
      "grad_norm": 0.306438684463501,
      "learning_rate": 4.993342350349607e-05,
      "loss": 0.5855,
      "num_input_tokens_seen": 3031928,
      "step": 4615
    },
    {
      "epoch": 2.4213836477987423,
      "grad_norm": 0.3300345242023468,
      "learning_rate": 4.9932586974187404e-05,
      "loss": 0.4234,
      "num_input_tokens_seen": 3035288,
      "step": 4620
    },
    {
      "epoch": 2.4240041928721174,
      "grad_norm": 0.608485221862793,
      "learning_rate": 4.9931745229265986e-05,
      "loss": 0.5343,
      "num_input_tokens_seen": 3038136,
      "step": 4625
    },
    {
      "epoch": 2.4266247379454926,
      "grad_norm": 0.6572824716567993,
      "learning_rate": 4.99308982689079e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 3040600,
      "step": 4630
    },
    {
      "epoch": 2.4292452830188678,
      "grad_norm": 0.5923836827278137,
      "learning_rate": 4.993004609329033e-05,
      "loss": 0.5311,
      "num_input_tokens_seen": 3044088,
      "step": 4635
    },
    {
      "epoch": 2.431865828092243,
      "grad_norm": 0.3279767632484436,
      "learning_rate": 4.992918870259153e-05,
      "loss": 0.4135,
      "num_input_tokens_seen": 3047160,
      "step": 4640
    },
    {
      "epoch": 2.4344863731656186,
      "grad_norm": 0.27602240443229675,
      "learning_rate": 4.992832609699086e-05,
      "loss": 0.4992,
      "num_input_tokens_seen": 3050616,
      "step": 4645
    },
    {
      "epoch": 2.4371069182389937,
      "grad_norm": 0.3361873924732208,
      "learning_rate": 4.992745827666877e-05,
      "loss": 0.4613,
      "num_input_tokens_seen": 3054328,
      "step": 4650
    },
    {
      "epoch": 2.439727463312369,
      "grad_norm": 0.4004492461681366,
      "learning_rate": 4.99265852418068e-05,
      "loss": 0.4407,
      "num_input_tokens_seen": 3057336,
      "step": 4655
    },
    {
      "epoch": 2.442348008385744,
      "grad_norm": 0.7317053079605103,
      "learning_rate": 4.992570699258757e-05,
      "loss": 0.6131,
      "num_input_tokens_seen": 3060824,
      "step": 4660
    },
    {
      "epoch": 2.4449685534591197,
      "grad_norm": 0.6030654311180115,
      "learning_rate": 4.99248235291948e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 3064152,
      "step": 4665
    },
    {
      "epoch": 2.447589098532495,
      "grad_norm": 0.3939392864704132,
      "learning_rate": 4.9923934851813314e-05,
      "loss": 0.462,
      "num_input_tokens_seen": 3068504,
      "step": 4670
    },
    {
      "epoch": 2.45020964360587,
      "grad_norm": 0.37222859263420105,
      "learning_rate": 4.9923040960628996e-05,
      "loss": 0.6034,
      "num_input_tokens_seen": 3071608,
      "step": 4675
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 0.3415868580341339,
      "learning_rate": 4.9922141855828856e-05,
      "loss": 0.478,
      "num_input_tokens_seen": 3074264,
      "step": 4680
    },
    {
      "epoch": 2.4554507337526204,
      "grad_norm": 0.4347119629383087,
      "learning_rate": 4.992123753760096e-05,
      "loss": 0.4287,
      "num_input_tokens_seen": 3076376,
      "step": 4685
    },
    {
      "epoch": 2.458071278825996,
      "grad_norm": 0.5479987263679504,
      "learning_rate": 4.9920328006134484e-05,
      "loss": 0.4774,
      "num_input_tokens_seen": 3079224,
      "step": 4690
    },
    {
      "epoch": 2.460691823899371,
      "grad_norm": 0.3539942800998688,
      "learning_rate": 4.991941326161971e-05,
      "loss": 0.3492,
      "num_input_tokens_seen": 3082616,
      "step": 4695
    },
    {
      "epoch": 2.4633123689727463,
      "grad_norm": 0.3156946003437042,
      "learning_rate": 4.991849330424796e-05,
      "loss": 0.4962,
      "num_input_tokens_seen": 3085784,
      "step": 4700
    },
    {
      "epoch": 2.4659329140461215,
      "grad_norm": 0.6595343947410583,
      "learning_rate": 4.9917568134211703e-05,
      "loss": 0.5645,
      "num_input_tokens_seen": 3089336,
      "step": 4705
    },
    {
      "epoch": 2.468553459119497,
      "grad_norm": 0.5484135150909424,
      "learning_rate": 4.991663775170447e-05,
      "loss": 0.4984,
      "num_input_tokens_seen": 3092152,
      "step": 4710
    },
    {
      "epoch": 2.4711740041928723,
      "grad_norm": 0.616834282875061,
      "learning_rate": 4.991570215692088e-05,
      "loss": 0.5133,
      "num_input_tokens_seen": 3095576,
      "step": 4715
    },
    {
      "epoch": 2.4737945492662474,
      "grad_norm": 0.47173479199409485,
      "learning_rate": 4.991476135005665e-05,
      "loss": 0.4346,
      "num_input_tokens_seen": 3098680,
      "step": 4720
    },
    {
      "epoch": 2.4764150943396226,
      "grad_norm": 0.7801305055618286,
      "learning_rate": 4.991381533130859e-05,
      "loss": 0.4231,
      "num_input_tokens_seen": 3101560,
      "step": 4725
    },
    {
      "epoch": 2.4790356394129978,
      "grad_norm": 0.33910897374153137,
      "learning_rate": 4.99128641008746e-05,
      "loss": 0.4509,
      "num_input_tokens_seen": 3105240,
      "step": 4730
    },
    {
      "epoch": 2.481656184486373,
      "grad_norm": 0.36373171210289,
      "learning_rate": 4.9911907658953664e-05,
      "loss": 0.4514,
      "num_input_tokens_seen": 3108376,
      "step": 4735
    },
    {
      "epoch": 2.4842767295597485,
      "grad_norm": 0.41317853331565857,
      "learning_rate": 4.991094600574584e-05,
      "loss": 0.5339,
      "num_input_tokens_seen": 3111992,
      "step": 4740
    },
    {
      "epoch": 2.4868972746331237,
      "grad_norm": 0.14774569869041443,
      "learning_rate": 4.9909979141452324e-05,
      "loss": 0.387,
      "num_input_tokens_seen": 3117496,
      "step": 4745
    },
    {
      "epoch": 2.489517819706499,
      "grad_norm": 0.4140394926071167,
      "learning_rate": 4.9909007066275346e-05,
      "loss": 0.3784,
      "num_input_tokens_seen": 3120568,
      "step": 4750
    },
    {
      "epoch": 2.492138364779874,
      "grad_norm": 0.6760610938072205,
      "learning_rate": 4.990802978041828e-05,
      "loss": 0.5843,
      "num_input_tokens_seen": 3123320,
      "step": 4755
    },
    {
      "epoch": 2.4947589098532497,
      "grad_norm": 0.4264778792858124,
      "learning_rate": 4.990704728408555e-05,
      "loss": 0.4663,
      "num_input_tokens_seen": 3126072,
      "step": 4760
    },
    {
      "epoch": 2.497379454926625,
      "grad_norm": 0.3292146325111389,
      "learning_rate": 4.990605957748268e-05,
      "loss": 0.5073,
      "num_input_tokens_seen": 3129528,
      "step": 4765
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.3812471330165863,
      "learning_rate": 4.990506666081628e-05,
      "loss": 0.5968,
      "num_input_tokens_seen": 3132600,
      "step": 4770
    },
    {
      "epoch": 2.502620545073375,
      "grad_norm": 0.31291159987449646,
      "learning_rate": 4.990406853429408e-05,
      "loss": 0.3889,
      "num_input_tokens_seen": 3135384,
      "step": 4775
    },
    {
      "epoch": 2.5052410901467503,
      "grad_norm": 0.44568389654159546,
      "learning_rate": 4.990306519812485e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 3139000,
      "step": 4780
    },
    {
      "epoch": 2.507861635220126,
      "grad_norm": 0.463652640581131,
      "learning_rate": 4.99020566525185e-05,
      "loss": 0.6218,
      "num_input_tokens_seen": 3141368,
      "step": 4785
    },
    {
      "epoch": 2.510482180293501,
      "grad_norm": 0.3522050380706787,
      "learning_rate": 4.9901042897685996e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 3144888,
      "step": 4790
    },
    {
      "epoch": 2.5131027253668763,
      "grad_norm": 0.3867899179458618,
      "learning_rate": 4.99000239338394e-05,
      "loss": 0.5312,
      "num_input_tokens_seen": 3147704,
      "step": 4795
    },
    {
      "epoch": 2.5157232704402515,
      "grad_norm": 0.4412108361721039,
      "learning_rate": 4.989899976119187e-05,
      "loss": 0.5244,
      "num_input_tokens_seen": 3150616,
      "step": 4800
    },
    {
      "epoch": 2.518343815513627,
      "grad_norm": 0.2738785445690155,
      "learning_rate": 4.989797037995766e-05,
      "loss": 0.5326,
      "num_input_tokens_seen": 3153240,
      "step": 4805
    },
    {
      "epoch": 2.5209643605870022,
      "grad_norm": 0.3859086036682129,
      "learning_rate": 4.989693579035209e-05,
      "loss": 0.5312,
      "num_input_tokens_seen": 3156216,
      "step": 4810
    },
    {
      "epoch": 2.5235849056603774,
      "grad_norm": 0.3298909366130829,
      "learning_rate": 4.989589599259159e-05,
      "loss": 0.4544,
      "num_input_tokens_seen": 3158936,
      "step": 4815
    },
    {
      "epoch": 2.5262054507337526,
      "grad_norm": 0.5595592260360718,
      "learning_rate": 4.9894850986893685e-05,
      "loss": 0.5686,
      "num_input_tokens_seen": 3162168,
      "step": 4820
    },
    {
      "epoch": 2.5288259958071277,
      "grad_norm": 0.5197535753250122,
      "learning_rate": 4.9893800773476965e-05,
      "loss": 0.4871,
      "num_input_tokens_seen": 3165560,
      "step": 4825
    },
    {
      "epoch": 2.531446540880503,
      "grad_norm": 0.28397050499916077,
      "learning_rate": 4.9892745352561134e-05,
      "loss": 0.5297,
      "num_input_tokens_seen": 3169400,
      "step": 4830
    },
    {
      "epoch": 2.5340670859538785,
      "grad_norm": 0.47889575362205505,
      "learning_rate": 4.9891684724366955e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 3172696,
      "step": 4835
    },
    {
      "epoch": 2.5366876310272537,
      "grad_norm": 0.5740084648132324,
      "learning_rate": 4.989061888911631e-05,
      "loss": 0.4872,
      "num_input_tokens_seen": 3175384,
      "step": 4840
    },
    {
      "epoch": 2.539308176100629,
      "grad_norm": 0.35065388679504395,
      "learning_rate": 4.988954784703217e-05,
      "loss": 0.3934,
      "num_input_tokens_seen": 3178232,
      "step": 4845
    },
    {
      "epoch": 2.541928721174004,
      "grad_norm": 0.39127108454704285,
      "learning_rate": 4.9888471598338574e-05,
      "loss": 0.4577,
      "num_input_tokens_seen": 3181048,
      "step": 4850
    },
    {
      "epoch": 2.5445492662473796,
      "grad_norm": 0.36796972155570984,
      "learning_rate": 4.988739014326066e-05,
      "loss": 0.4758,
      "num_input_tokens_seen": 3184504,
      "step": 4855
    },
    {
      "epoch": 2.547169811320755,
      "grad_norm": 0.371848464012146,
      "learning_rate": 4.9886303482024656e-05,
      "loss": 0.4284,
      "num_input_tokens_seen": 3187480,
      "step": 4860
    },
    {
      "epoch": 2.54979035639413,
      "grad_norm": 0.3533948063850403,
      "learning_rate": 4.988521161485789e-05,
      "loss": 0.5328,
      "num_input_tokens_seen": 3191064,
      "step": 4865
    },
    {
      "epoch": 2.552410901467505,
      "grad_norm": 0.5441691875457764,
      "learning_rate": 4.9884114541988747e-05,
      "loss": 0.4904,
      "num_input_tokens_seen": 3194296,
      "step": 4870
    },
    {
      "epoch": 2.5550314465408803,
      "grad_norm": 0.6921298503875732,
      "learning_rate": 4.988301226364674e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 3197784,
      "step": 4875
    },
    {
      "epoch": 2.5576519916142555,
      "grad_norm": 0.5372940301895142,
      "learning_rate": 4.988190478006245e-05,
      "loss": 0.5437,
      "num_input_tokens_seen": 3200856,
      "step": 4880
    },
    {
      "epoch": 2.560272536687631,
      "grad_norm": 0.889894425868988,
      "learning_rate": 4.988079209146754e-05,
      "loss": 0.3659,
      "num_input_tokens_seen": 3203192,
      "step": 4885
    },
    {
      "epoch": 2.5628930817610063,
      "grad_norm": 0.28228938579559326,
      "learning_rate": 4.9879674198094794e-05,
      "loss": 0.5424,
      "num_input_tokens_seen": 3206712,
      "step": 4890
    },
    {
      "epoch": 2.5655136268343814,
      "grad_norm": 0.30740439891815186,
      "learning_rate": 4.987855110017804e-05,
      "loss": 0.3529,
      "num_input_tokens_seen": 3210168,
      "step": 4895
    },
    {
      "epoch": 2.568134171907757,
      "grad_norm": 0.5007047057151794,
      "learning_rate": 4.9877422797952224e-05,
      "loss": 0.5334,
      "num_input_tokens_seen": 3213208,
      "step": 4900
    },
    {
      "epoch": 2.5707547169811322,
      "grad_norm": 0.4108736217021942,
      "learning_rate": 4.987628929165338e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 3215608,
      "step": 4905
    },
    {
      "epoch": 2.5733752620545074,
      "grad_norm": 0.9760721921920776,
      "learning_rate": 4.9875150581518615e-05,
      "loss": 0.6714,
      "num_input_tokens_seen": 3218360,
      "step": 4910
    },
    {
      "epoch": 2.5759958071278826,
      "grad_norm": 0.5765912532806396,
      "learning_rate": 4.987400666778614e-05,
      "loss": 0.4823,
      "num_input_tokens_seen": 3221656,
      "step": 4915
    },
    {
      "epoch": 2.5786163522012577,
      "grad_norm": 0.45068788528442383,
      "learning_rate": 4.987285755069524e-05,
      "loss": 0.37,
      "num_input_tokens_seen": 3225464,
      "step": 4920
    },
    {
      "epoch": 2.581236897274633,
      "grad_norm": 0.6500669121742249,
      "learning_rate": 4.987170323048632e-05,
      "loss": 0.5152,
      "num_input_tokens_seen": 3228824,
      "step": 4925
    },
    {
      "epoch": 2.5838574423480085,
      "grad_norm": 0.8816668391227722,
      "learning_rate": 4.987054370740083e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 3231160,
      "step": 4930
    },
    {
      "epoch": 2.5864779874213837,
      "grad_norm": 0.508108377456665,
      "learning_rate": 4.986937898168134e-05,
      "loss": 0.5738,
      "num_input_tokens_seen": 3233912,
      "step": 4935
    },
    {
      "epoch": 2.589098532494759,
      "grad_norm": 0.26352277398109436,
      "learning_rate": 4.986820905357148e-05,
      "loss": 0.5008,
      "num_input_tokens_seen": 3237144,
      "step": 4940
    },
    {
      "epoch": 2.591719077568134,
      "grad_norm": 1.2174339294433594,
      "learning_rate": 4.9867033923315995e-05,
      "loss": 0.4921,
      "num_input_tokens_seen": 3239896,
      "step": 4945
    },
    {
      "epoch": 2.5943396226415096,
      "grad_norm": 0.512902557849884,
      "learning_rate": 4.9865853591160724e-05,
      "loss": 0.5054,
      "num_input_tokens_seen": 3242744,
      "step": 4950
    },
    {
      "epoch": 2.596960167714885,
      "grad_norm": 0.5143552422523499,
      "learning_rate": 4.9864668057352556e-05,
      "loss": 0.5789,
      "num_input_tokens_seen": 3246264,
      "step": 4955
    },
    {
      "epoch": 2.59958071278826,
      "grad_norm": 0.5460503697395325,
      "learning_rate": 4.9863477322139517e-05,
      "loss": 0.4404,
      "num_input_tokens_seen": 3250040,
      "step": 4960
    },
    {
      "epoch": 2.602201257861635,
      "grad_norm": 0.5087428092956543,
      "learning_rate": 4.986228138577067e-05,
      "loss": 0.5528,
      "num_input_tokens_seen": 3252920,
      "step": 4965
    },
    {
      "epoch": 2.6048218029350103,
      "grad_norm": 0.4930756688117981,
      "learning_rate": 4.986108024849619e-05,
      "loss": 0.5076,
      "num_input_tokens_seen": 3256184,
      "step": 4970
    },
    {
      "epoch": 2.6074423480083855,
      "grad_norm": 1.1776424646377563,
      "learning_rate": 4.985987391056737e-05,
      "loss": 0.5199,
      "num_input_tokens_seen": 3259416,
      "step": 4975
    },
    {
      "epoch": 2.610062893081761,
      "grad_norm": 0.4339406490325928,
      "learning_rate": 4.9858662372236544e-05,
      "loss": 0.4955,
      "num_input_tokens_seen": 3263224,
      "step": 4980
    },
    {
      "epoch": 2.6126834381551363,
      "grad_norm": 0.46599069237709045,
      "learning_rate": 4.985744563375714e-05,
      "loss": 0.4095,
      "num_input_tokens_seen": 3265720,
      "step": 4985
    },
    {
      "epoch": 2.6153039832285114,
      "grad_norm": 0.4136963486671448,
      "learning_rate": 4.985622369538371e-05,
      "loss": 0.3955,
      "num_input_tokens_seen": 3270936,
      "step": 4990
    },
    {
      "epoch": 2.617924528301887,
      "grad_norm": 0.31941676139831543,
      "learning_rate": 4.985499655737185e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 3274136,
      "step": 4995
    },
    {
      "epoch": 2.620545073375262,
      "grad_norm": 0.4268212616443634,
      "learning_rate": 4.985376421997827e-05,
      "loss": 0.5564,
      "num_input_tokens_seen": 3277784,
      "step": 5000
    },
    {
      "epoch": 2.6231656184486374,
      "grad_norm": 0.5485895276069641,
      "learning_rate": 4.985252668346076e-05,
      "loss": 0.5994,
      "num_input_tokens_seen": 3280408,
      "step": 5005
    },
    {
      "epoch": 2.6257861635220126,
      "grad_norm": 0.6296403408050537,
      "learning_rate": 4.9851283948078206e-05,
      "loss": 0.4992,
      "num_input_tokens_seen": 3283512,
      "step": 5010
    },
    {
      "epoch": 2.6284067085953877,
      "grad_norm": 0.29974162578582764,
      "learning_rate": 4.985003601409057e-05,
      "loss": 0.5723,
      "num_input_tokens_seen": 3287096,
      "step": 5015
    },
    {
      "epoch": 2.631027253668763,
      "grad_norm": 0.4705175757408142,
      "learning_rate": 4.9848782881758906e-05,
      "loss": 0.4522,
      "num_input_tokens_seen": 3290008,
      "step": 5020
    },
    {
      "epoch": 2.6336477987421385,
      "grad_norm": 0.3649095892906189,
      "learning_rate": 4.9847524551345345e-05,
      "loss": 0.3997,
      "num_input_tokens_seen": 3293048,
      "step": 5025
    },
    {
      "epoch": 2.6362683438155137,
      "grad_norm": 0.4698164761066437,
      "learning_rate": 4.984626102311313e-05,
      "loss": 0.3348,
      "num_input_tokens_seen": 3295928,
      "step": 5030
    },
    {
      "epoch": 2.638888888888889,
      "grad_norm": 0.500487744808197,
      "learning_rate": 4.984499229732656e-05,
      "loss": 0.4122,
      "num_input_tokens_seen": 3299480,
      "step": 5035
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 0.44345149397850037,
      "learning_rate": 4.984371837425106e-05,
      "loss": 0.3949,
      "num_input_tokens_seen": 3302008,
      "step": 5040
    },
    {
      "epoch": 2.6441299790356396,
      "grad_norm": 0.4705497622489929,
      "learning_rate": 4.9842439254153106e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 3304632,
      "step": 5045
    },
    {
      "epoch": 2.646750524109015,
      "grad_norm": 0.47457200288772583,
      "learning_rate": 4.984115493730027e-05,
      "loss": 0.3618,
      "num_input_tokens_seen": 3307576,
      "step": 5050
    },
    {
      "epoch": 2.64937106918239,
      "grad_norm": 0.6661854386329651,
      "learning_rate": 4.983986542396123e-05,
      "loss": 0.4443,
      "num_input_tokens_seen": 3310136,
      "step": 5055
    },
    {
      "epoch": 2.651991614255765,
      "grad_norm": 0.3818792998790741,
      "learning_rate": 4.983857071440573e-05,
      "loss": 0.8396,
      "num_input_tokens_seen": 3313080,
      "step": 5060
    },
    {
      "epoch": 2.6546121593291403,
      "grad_norm": 0.3690452575683594,
      "learning_rate": 4.9837270808904616e-05,
      "loss": 0.5332,
      "num_input_tokens_seen": 3315512,
      "step": 5065
    },
    {
      "epoch": 2.6572327044025155,
      "grad_norm": 0.5218209028244019,
      "learning_rate": 4.9835965707729807e-05,
      "loss": 0.5682,
      "num_input_tokens_seen": 3318552,
      "step": 5070
    },
    {
      "epoch": 2.659853249475891,
      "grad_norm": 0.5685155391693115,
      "learning_rate": 4.983465541115431e-05,
      "loss": 0.4404,
      "num_input_tokens_seen": 3321560,
      "step": 5075
    },
    {
      "epoch": 2.6624737945492662,
      "grad_norm": 0.30668115615844727,
      "learning_rate": 4.983333991945223e-05,
      "loss": 0.4437,
      "num_input_tokens_seen": 3326136,
      "step": 5080
    },
    {
      "epoch": 2.6650943396226414,
      "grad_norm": 0.23579207062721252,
      "learning_rate": 4.983201923289875e-05,
      "loss": 0.4339,
      "num_input_tokens_seen": 3329816,
      "step": 5085
    },
    {
      "epoch": 2.667714884696017,
      "grad_norm": 0.692589521408081,
      "learning_rate": 4.983069335177014e-05,
      "loss": 0.4765,
      "num_input_tokens_seen": 3332600,
      "step": 5090
    },
    {
      "epoch": 2.670335429769392,
      "grad_norm": 0.45901936292648315,
      "learning_rate": 4.982936227634378e-05,
      "loss": 0.5762,
      "num_input_tokens_seen": 3335832,
      "step": 5095
    },
    {
      "epoch": 2.6729559748427674,
      "grad_norm": 0.4068652391433716,
      "learning_rate": 4.982802600689809e-05,
      "loss": 0.5074,
      "num_input_tokens_seen": 3338680,
      "step": 5100
    },
    {
      "epoch": 2.6755765199161425,
      "grad_norm": 0.6256952881813049,
      "learning_rate": 4.982668454371261e-05,
      "loss": 0.5503,
      "num_input_tokens_seen": 3341752,
      "step": 5105
    },
    {
      "epoch": 2.6781970649895177,
      "grad_norm": 0.4283396899700165,
      "learning_rate": 4.9825337887067966e-05,
      "loss": 0.5424,
      "num_input_tokens_seen": 3345016,
      "step": 5110
    },
    {
      "epoch": 2.680817610062893,
      "grad_norm": 0.3728429973125458,
      "learning_rate": 4.982398603724586e-05,
      "loss": 0.5291,
      "num_input_tokens_seen": 3348056,
      "step": 5115
    },
    {
      "epoch": 2.6834381551362685,
      "grad_norm": 0.71095871925354,
      "learning_rate": 4.982262899452909e-05,
      "loss": 0.5813,
      "num_input_tokens_seen": 3350904,
      "step": 5120
    },
    {
      "epoch": 2.6860587002096437,
      "grad_norm": 0.5631696581840515,
      "learning_rate": 4.982126675920151e-05,
      "loss": 0.4134,
      "num_input_tokens_seen": 3353816,
      "step": 5125
    },
    {
      "epoch": 2.688679245283019,
      "grad_norm": 0.3447795510292053,
      "learning_rate": 4.98198993315481e-05,
      "loss": 0.4111,
      "num_input_tokens_seen": 3357080,
      "step": 5130
    },
    {
      "epoch": 2.691299790356394,
      "grad_norm": 0.3489712178707123,
      "learning_rate": 4.981852671185492e-05,
      "loss": 0.5567,
      "num_input_tokens_seen": 3360088,
      "step": 5135
    },
    {
      "epoch": 2.6939203354297696,
      "grad_norm": 0.9554756879806519,
      "learning_rate": 4.9817148900409093e-05,
      "loss": 0.5585,
      "num_input_tokens_seen": 3362520,
      "step": 5140
    },
    {
      "epoch": 2.6965408805031448,
      "grad_norm": 0.4298619031906128,
      "learning_rate": 4.981576589749885e-05,
      "loss": 0.4642,
      "num_input_tokens_seen": 3365240,
      "step": 5145
    },
    {
      "epoch": 2.69916142557652,
      "grad_norm": 0.47813937067985535,
      "learning_rate": 4.981437770341348e-05,
      "loss": 0.5724,
      "num_input_tokens_seen": 3369048,
      "step": 5150
    },
    {
      "epoch": 2.701781970649895,
      "grad_norm": 0.32927176356315613,
      "learning_rate": 4.9812984318443414e-05,
      "loss": 0.4629,
      "num_input_tokens_seen": 3371576,
      "step": 5155
    },
    {
      "epoch": 2.7044025157232703,
      "grad_norm": 0.5377748608589172,
      "learning_rate": 4.98115857428801e-05,
      "loss": 0.5182,
      "num_input_tokens_seen": 3374648,
      "step": 5160
    },
    {
      "epoch": 2.7070230607966455,
      "grad_norm": 0.5411098599433899,
      "learning_rate": 4.9810181977016104e-05,
      "loss": 0.4837,
      "num_input_tokens_seen": 3377464,
      "step": 5165
    },
    {
      "epoch": 2.709643605870021,
      "grad_norm": 0.447170615196228,
      "learning_rate": 4.9808773021145096e-05,
      "loss": 0.4381,
      "num_input_tokens_seen": 3380632,
      "step": 5170
    },
    {
      "epoch": 2.7122641509433962,
      "grad_norm": 0.5058719515800476,
      "learning_rate": 4.98073588755618e-05,
      "loss": 0.5082,
      "num_input_tokens_seen": 3384216,
      "step": 5175
    },
    {
      "epoch": 2.7148846960167714,
      "grad_norm": 0.7473098039627075,
      "learning_rate": 4.980593954056205e-05,
      "loss": 0.4709,
      "num_input_tokens_seen": 3386520,
      "step": 5180
    },
    {
      "epoch": 2.717505241090147,
      "grad_norm": 0.5357159376144409,
      "learning_rate": 4.9804515016442754e-05,
      "loss": 0.5049,
      "num_input_tokens_seen": 3389560,
      "step": 5185
    },
    {
      "epoch": 2.720125786163522,
      "grad_norm": 0.8110059499740601,
      "learning_rate": 4.9803085303501896e-05,
      "loss": 0.6166,
      "num_input_tokens_seen": 3392408,
      "step": 5190
    },
    {
      "epoch": 2.7227463312368974,
      "grad_norm": 0.6413405537605286,
      "learning_rate": 4.9801650402038555e-05,
      "loss": 0.3977,
      "num_input_tokens_seen": 3395672,
      "step": 5195
    },
    {
      "epoch": 2.7253668763102725,
      "grad_norm": 0.3276548683643341,
      "learning_rate": 4.980021031235291e-05,
      "loss": 0.3844,
      "num_input_tokens_seen": 3399608,
      "step": 5200
    },
    {
      "epoch": 2.7279874213836477,
      "grad_norm": 0.37580370903015137,
      "learning_rate": 4.9798765034746206e-05,
      "loss": 0.5435,
      "num_input_tokens_seen": 3402776,
      "step": 5205
    },
    {
      "epoch": 2.730607966457023,
      "grad_norm": 0.5060549974441528,
      "learning_rate": 4.979731456952077e-05,
      "loss": 0.5534,
      "num_input_tokens_seen": 3405688,
      "step": 5210
    },
    {
      "epoch": 2.7332285115303985,
      "grad_norm": 0.3301379680633545,
      "learning_rate": 4.9795858916980034e-05,
      "loss": 0.6162,
      "num_input_tokens_seen": 3409048,
      "step": 5215
    },
    {
      "epoch": 2.7358490566037736,
      "grad_norm": 0.6534701585769653,
      "learning_rate": 4.9794398077428495e-05,
      "loss": 0.5142,
      "num_input_tokens_seen": 3411416,
      "step": 5220
    },
    {
      "epoch": 2.738469601677149,
      "grad_norm": 0.24836501479148865,
      "learning_rate": 4.979293205117175e-05,
      "loss": 0.4247,
      "num_input_tokens_seen": 3414392,
      "step": 5225
    },
    {
      "epoch": 2.741090146750524,
      "grad_norm": 0.38694143295288086,
      "learning_rate": 4.979146083851647e-05,
      "loss": 0.4289,
      "num_input_tokens_seen": 3417048,
      "step": 5230
    },
    {
      "epoch": 2.7437106918238996,
      "grad_norm": 0.5740531086921692,
      "learning_rate": 4.978998443977042e-05,
      "loss": 0.613,
      "num_input_tokens_seen": 3419832,
      "step": 5235
    },
    {
      "epoch": 2.7463312368972748,
      "grad_norm": 0.3991996645927429,
      "learning_rate": 4.9788502855242436e-05,
      "loss": 0.4783,
      "num_input_tokens_seen": 3422840,
      "step": 5240
    },
    {
      "epoch": 2.74895178197065,
      "grad_norm": 0.44479119777679443,
      "learning_rate": 4.978701608524247e-05,
      "loss": 0.5973,
      "num_input_tokens_seen": 3426136,
      "step": 5245
    },
    {
      "epoch": 2.751572327044025,
      "grad_norm": 0.5807952284812927,
      "learning_rate": 4.978552413008152e-05,
      "loss": 0.5511,
      "num_input_tokens_seen": 3429848,
      "step": 5250
    },
    {
      "epoch": 2.7541928721174003,
      "grad_norm": 0.5813663601875305,
      "learning_rate": 4.9784026990071687e-05,
      "loss": 0.5131,
      "num_input_tokens_seen": 3432632,
      "step": 5255
    },
    {
      "epoch": 2.7568134171907754,
      "grad_norm": 0.44974204897880554,
      "learning_rate": 4.978252466552615e-05,
      "loss": 0.5044,
      "num_input_tokens_seen": 3435384,
      "step": 5260
    },
    {
      "epoch": 2.759433962264151,
      "grad_norm": 0.3940238058567047,
      "learning_rate": 4.9781017156759205e-05,
      "loss": 0.495,
      "num_input_tokens_seen": 3439608,
      "step": 5265
    },
    {
      "epoch": 2.762054507337526,
      "grad_norm": 0.33523187041282654,
      "learning_rate": 4.977950446408618e-05,
      "loss": 0.5043,
      "num_input_tokens_seen": 3443192,
      "step": 5270
    },
    {
      "epoch": 2.7646750524109014,
      "grad_norm": 0.22713591158390045,
      "learning_rate": 4.9777986587823523e-05,
      "loss": 0.4016,
      "num_input_tokens_seen": 3447096,
      "step": 5275
    },
    {
      "epoch": 2.767295597484277,
      "grad_norm": 0.5287631154060364,
      "learning_rate": 4.977646352828875e-05,
      "loss": 0.5596,
      "num_input_tokens_seen": 3450648,
      "step": 5280
    },
    {
      "epoch": 2.769916142557652,
      "grad_norm": 0.40138059854507446,
      "learning_rate": 4.9774935285800476e-05,
      "loss": 0.5791,
      "num_input_tokens_seen": 3453976,
      "step": 5285
    },
    {
      "epoch": 2.7725366876310273,
      "grad_norm": 0.48606398701667786,
      "learning_rate": 4.977340186067838e-05,
      "loss": 0.4262,
      "num_input_tokens_seen": 3457080,
      "step": 5290
    },
    {
      "epoch": 2.7751572327044025,
      "grad_norm": 0.37415337562561035,
      "learning_rate": 4.977186325324325e-05,
      "loss": 0.524,
      "num_input_tokens_seen": 3461784,
      "step": 5295
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 0.4998932182788849,
      "learning_rate": 4.977031946381694e-05,
      "loss": 0.6749,
      "num_input_tokens_seen": 3465624,
      "step": 5300
    },
    {
      "epoch": 2.780398322851153,
      "grad_norm": 0.5633404850959778,
      "learning_rate": 4.9768770492722394e-05,
      "loss": 0.655,
      "num_input_tokens_seen": 3468632,
      "step": 5305
    },
    {
      "epoch": 2.7830188679245285,
      "grad_norm": 0.7296350002288818,
      "learning_rate": 4.976721634028363e-05,
      "loss": 0.4589,
      "num_input_tokens_seen": 3472376,
      "step": 5310
    },
    {
      "epoch": 2.7856394129979036,
      "grad_norm": 0.30670997500419617,
      "learning_rate": 4.976565700682578e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 3475256,
      "step": 5315
    },
    {
      "epoch": 2.788259958071279,
      "grad_norm": 0.40606480836868286,
      "learning_rate": 4.976409249267502e-05,
      "loss": 0.4127,
      "num_input_tokens_seen": 3478168,
      "step": 5320
    },
    {
      "epoch": 2.790880503144654,
      "grad_norm": 0.7173824310302734,
      "learning_rate": 4.976252279815863e-05,
      "loss": 0.578,
      "num_input_tokens_seen": 3482040,
      "step": 5325
    },
    {
      "epoch": 2.7935010482180296,
      "grad_norm": 0.34373795986175537,
      "learning_rate": 4.976094792360498e-05,
      "loss": 0.5724,
      "num_input_tokens_seen": 3486168,
      "step": 5330
    },
    {
      "epoch": 2.7961215932914047,
      "grad_norm": 0.4430333971977234,
      "learning_rate": 4.975936786934351e-05,
      "loss": 0.4597,
      "num_input_tokens_seen": 3490136,
      "step": 5335
    },
    {
      "epoch": 2.79874213836478,
      "grad_norm": 0.44413211941719055,
      "learning_rate": 4.975778263570475e-05,
      "loss": 0.6999,
      "num_input_tokens_seen": 3493368,
      "step": 5340
    },
    {
      "epoch": 2.801362683438155,
      "grad_norm": 0.4175981283187866,
      "learning_rate": 4.975619222302032e-05,
      "loss": 0.4313,
      "num_input_tokens_seen": 3497272,
      "step": 5345
    },
    {
      "epoch": 2.8039832285115303,
      "grad_norm": 0.4459098279476166,
      "learning_rate": 4.975459663162291e-05,
      "loss": 0.5091,
      "num_input_tokens_seen": 3500152,
      "step": 5350
    },
    {
      "epoch": 2.8066037735849054,
      "grad_norm": 0.4133112132549286,
      "learning_rate": 4.9752995861846296e-05,
      "loss": 0.4331,
      "num_input_tokens_seen": 3506328,
      "step": 5355
    },
    {
      "epoch": 2.809224318658281,
      "grad_norm": 0.571759045124054,
      "learning_rate": 4.975138991402535e-05,
      "loss": 0.4973,
      "num_input_tokens_seen": 3508824,
      "step": 5360
    },
    {
      "epoch": 2.811844863731656,
      "grad_norm": 0.4022855758666992,
      "learning_rate": 4.974977878849601e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 3512920,
      "step": 5365
    },
    {
      "epoch": 2.8144654088050314,
      "grad_norm": 0.282320111989975,
      "learning_rate": 4.9748162485595303e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 3516984,
      "step": 5370
    },
    {
      "epoch": 2.8170859538784065,
      "grad_norm": 0.6248133182525635,
      "learning_rate": 4.974654100566135e-05,
      "loss": 0.5089,
      "num_input_tokens_seen": 3520664,
      "step": 5375
    },
    {
      "epoch": 2.819706498951782,
      "grad_norm": 0.5427289009094238,
      "learning_rate": 4.974491434903335e-05,
      "loss": 0.7331,
      "num_input_tokens_seen": 3523864,
      "step": 5380
    },
    {
      "epoch": 2.8223270440251573,
      "grad_norm": 0.5714131593704224,
      "learning_rate": 4.9743282516051556e-05,
      "loss": 0.6558,
      "num_input_tokens_seen": 3526264,
      "step": 5385
    },
    {
      "epoch": 2.8249475890985325,
      "grad_norm": 0.3358728885650635,
      "learning_rate": 4.974164550705736e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 3528952,
      "step": 5390
    },
    {
      "epoch": 2.8275681341719077,
      "grad_norm": 0.3806276321411133,
      "learning_rate": 4.974000332239319e-05,
      "loss": 0.628,
      "num_input_tokens_seen": 3531384,
      "step": 5395
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 0.5681930184364319,
      "learning_rate": 4.973835596240257e-05,
      "loss": 0.4115,
      "num_input_tokens_seen": 3534488,
      "step": 5400
    },
    {
      "epoch": 2.832809224318658,
      "grad_norm": 0.6198461055755615,
      "learning_rate": 4.9736703427430114e-05,
      "loss": 0.5214,
      "num_input_tokens_seen": 3537880,
      "step": 5405
    },
    {
      "epoch": 2.8354297693920336,
      "grad_norm": 0.6591648459434509,
      "learning_rate": 4.9735045717821514e-05,
      "loss": 0.3959,
      "num_input_tokens_seen": 3541272,
      "step": 5410
    },
    {
      "epoch": 2.838050314465409,
      "grad_norm": 0.3759389817714691,
      "learning_rate": 4.973338283392354e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 3544280,
      "step": 5415
    },
    {
      "epoch": 2.840670859538784,
      "grad_norm": 0.4711204767227173,
      "learning_rate": 4.9731714776084046e-05,
      "loss": 0.6034,
      "num_input_tokens_seen": 3548024,
      "step": 5420
    },
    {
      "epoch": 2.8432914046121596,
      "grad_norm": 0.3894682824611664,
      "learning_rate": 4.973004154465198e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 3550680,
      "step": 5425
    },
    {
      "epoch": 2.8459119496855347,
      "grad_norm": 0.46923142671585083,
      "learning_rate": 4.972836313997736e-05,
      "loss": 0.5289,
      "num_input_tokens_seen": 3553560,
      "step": 5430
    },
    {
      "epoch": 2.84853249475891,
      "grad_norm": 0.40470707416534424,
      "learning_rate": 4.9726679562411285e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 3556600,
      "step": 5435
    },
    {
      "epoch": 2.851153039832285,
      "grad_norm": 0.34212571382522583,
      "learning_rate": 4.972499081230595e-05,
      "loss": 0.428,
      "num_input_tokens_seen": 3560184,
      "step": 5440
    },
    {
      "epoch": 2.8537735849056602,
      "grad_norm": 0.5154420137405396,
      "learning_rate": 4.9723296890014605e-05,
      "loss": 0.5921,
      "num_input_tokens_seen": 3563928,
      "step": 5445
    },
    {
      "epoch": 2.8563941299790354,
      "grad_norm": 0.4206479787826538,
      "learning_rate": 4.9721597795891605e-05,
      "loss": 0.4511,
      "num_input_tokens_seen": 3568024,
      "step": 5450
    },
    {
      "epoch": 2.859014675052411,
      "grad_norm": 0.6291370987892151,
      "learning_rate": 4.9719893530292396e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 3570904,
      "step": 5455
    },
    {
      "epoch": 2.861635220125786,
      "grad_norm": 0.3677840232849121,
      "learning_rate": 4.9718184093573475e-05,
      "loss": 0.4997,
      "num_input_tokens_seen": 3574232,
      "step": 5460
    },
    {
      "epoch": 2.8642557651991614,
      "grad_norm": 0.8366602063179016,
      "learning_rate": 4.9716469486092435e-05,
      "loss": 0.4925,
      "num_input_tokens_seen": 3576856,
      "step": 5465
    },
    {
      "epoch": 2.8668763102725365,
      "grad_norm": 0.5198577046394348,
      "learning_rate": 4.971474970820797e-05,
      "loss": 0.78,
      "num_input_tokens_seen": 3580024,
      "step": 5470
    },
    {
      "epoch": 2.869496855345912,
      "grad_norm": 0.3530573546886444,
      "learning_rate": 4.971302476027982e-05,
      "loss": 0.5274,
      "num_input_tokens_seen": 3583352,
      "step": 5475
    },
    {
      "epoch": 2.8721174004192873,
      "grad_norm": 0.7347520589828491,
      "learning_rate": 4.971129464266884e-05,
      "loss": 0.417,
      "num_input_tokens_seen": 3586104,
      "step": 5480
    },
    {
      "epoch": 2.8747379454926625,
      "grad_norm": 0.3625946342945099,
      "learning_rate": 4.970955935573693e-05,
      "loss": 0.3378,
      "num_input_tokens_seen": 3588984,
      "step": 5485
    },
    {
      "epoch": 2.8773584905660377,
      "grad_norm": 0.38790249824523926,
      "learning_rate": 4.9707818899847106e-05,
      "loss": 0.4907,
      "num_input_tokens_seen": 3592312,
      "step": 5490
    },
    {
      "epoch": 2.879979035639413,
      "grad_norm": 0.3461206257343292,
      "learning_rate": 4.970607327536346e-05,
      "loss": 0.3666,
      "num_input_tokens_seen": 3595256,
      "step": 5495
    },
    {
      "epoch": 2.882599580712788,
      "grad_norm": 1.0178192853927612,
      "learning_rate": 4.970432248265114e-05,
      "loss": 0.5858,
      "num_input_tokens_seen": 3597784,
      "step": 5500
    },
    {
      "epoch": 2.8852201257861636,
      "grad_norm": 0.3014279901981354,
      "learning_rate": 4.97025665220764e-05,
      "loss": 0.4026,
      "num_input_tokens_seen": 3601464,
      "step": 5505
    },
    {
      "epoch": 2.8878406708595388,
      "grad_norm": 0.6887813806533813,
      "learning_rate": 4.970080539400656e-05,
      "loss": 0.4866,
      "num_input_tokens_seen": 3604088,
      "step": 5510
    },
    {
      "epoch": 2.890461215932914,
      "grad_norm": 0.6384339332580566,
      "learning_rate": 4.969903909881004e-05,
      "loss": 0.5947,
      "num_input_tokens_seen": 3607544,
      "step": 5515
    },
    {
      "epoch": 2.8930817610062896,
      "grad_norm": 0.3234609365463257,
      "learning_rate": 4.9697267636856317e-05,
      "loss": 0.4663,
      "num_input_tokens_seen": 3611480,
      "step": 5520
    },
    {
      "epoch": 2.8957023060796647,
      "grad_norm": 0.49230819940567017,
      "learning_rate": 4.969549100851597e-05,
      "loss": 0.6893,
      "num_input_tokens_seen": 3614808,
      "step": 5525
    },
    {
      "epoch": 2.89832285115304,
      "grad_norm": 0.6021590828895569,
      "learning_rate": 4.9693709214160645e-05,
      "loss": 0.6288,
      "num_input_tokens_seen": 3617624,
      "step": 5530
    },
    {
      "epoch": 2.900943396226415,
      "grad_norm": 0.4577595591545105,
      "learning_rate": 4.969192225416307e-05,
      "loss": 0.5304,
      "num_input_tokens_seen": 3620312,
      "step": 5535
    },
    {
      "epoch": 2.9035639412997902,
      "grad_norm": 0.49995654821395874,
      "learning_rate": 4.969013012889706e-05,
      "loss": 0.4961,
      "num_input_tokens_seen": 3625272,
      "step": 5540
    },
    {
      "epoch": 2.9061844863731654,
      "grad_norm": 0.6527161598205566,
      "learning_rate": 4.9688332838737504e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 3629208,
      "step": 5545
    },
    {
      "epoch": 2.908805031446541,
      "grad_norm": 0.48894935846328735,
      "learning_rate": 4.9686530384060386e-05,
      "loss": 0.472,
      "num_input_tokens_seen": 3632696,
      "step": 5550
    },
    {
      "epoch": 2.911425576519916,
      "grad_norm": 0.43484240770339966,
      "learning_rate": 4.968472276524275e-05,
      "loss": 0.4802,
      "num_input_tokens_seen": 3635640,
      "step": 5555
    },
    {
      "epoch": 2.9140461215932913,
      "grad_norm": 0.6105705499649048,
      "learning_rate": 4.9682909982662725e-05,
      "loss": 0.3627,
      "num_input_tokens_seen": 3639672,
      "step": 5560
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 0.35027775168418884,
      "learning_rate": 4.968109203669953e-05,
      "loss": 0.4661,
      "num_input_tokens_seen": 3642264,
      "step": 5565
    },
    {
      "epoch": 2.919287211740042,
      "grad_norm": 0.5196674466133118,
      "learning_rate": 4.967926892773346e-05,
      "loss": 0.531,
      "num_input_tokens_seen": 3646200,
      "step": 5570
    },
    {
      "epoch": 2.9219077568134173,
      "grad_norm": 0.5981946587562561,
      "learning_rate": 4.967744065614588e-05,
      "loss": 0.4596,
      "num_input_tokens_seen": 3649240,
      "step": 5575
    },
    {
      "epoch": 2.9245283018867925,
      "grad_norm": 0.6585130095481873,
      "learning_rate": 4.9675607222319255e-05,
      "loss": 0.603,
      "num_input_tokens_seen": 3653016,
      "step": 5580
    },
    {
      "epoch": 2.9271488469601676,
      "grad_norm": 0.2821461260318756,
      "learning_rate": 4.967376862663711e-05,
      "loss": 0.4048,
      "num_input_tokens_seen": 3656792,
      "step": 5585
    },
    {
      "epoch": 2.929769392033543,
      "grad_norm": 0.3906414806842804,
      "learning_rate": 4.9671924869484064e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 3659672,
      "step": 5590
    },
    {
      "epoch": 2.932389937106918,
      "grad_norm": 0.5857247114181519,
      "learning_rate": 4.967007595124581e-05,
      "loss": 0.4166,
      "num_input_tokens_seen": 3663000,
      "step": 5595
    },
    {
      "epoch": 2.9350104821802936,
      "grad_norm": 0.28740444779396057,
      "learning_rate": 4.966822187230911e-05,
      "loss": 0.4309,
      "num_input_tokens_seen": 3666040,
      "step": 5600
    },
    {
      "epoch": 2.9376310272536688,
      "grad_norm": 0.33182740211486816,
      "learning_rate": 4.966636263306183e-05,
      "loss": 0.4862,
      "num_input_tokens_seen": 3670808,
      "step": 5605
    },
    {
      "epoch": 2.940251572327044,
      "grad_norm": 0.5459467768669128,
      "learning_rate": 4.9664498233892895e-05,
      "loss": 0.4459,
      "num_input_tokens_seen": 3676184,
      "step": 5610
    },
    {
      "epoch": 2.9428721174004195,
      "grad_norm": 0.25203436613082886,
      "learning_rate": 4.966262867519232e-05,
      "loss": 0.5238,
      "num_input_tokens_seen": 3679768,
      "step": 5615
    },
    {
      "epoch": 2.9454926624737947,
      "grad_norm": 0.4070933759212494,
      "learning_rate": 4.966075395735119e-05,
      "loss": 0.6517,
      "num_input_tokens_seen": 3683000,
      "step": 5620
    },
    {
      "epoch": 2.94811320754717,
      "grad_norm": 0.7091167569160461,
      "learning_rate": 4.965887408076167e-05,
      "loss": 0.3449,
      "num_input_tokens_seen": 3686584,
      "step": 5625
    },
    {
      "epoch": 2.950733752620545,
      "grad_norm": 0.3705088496208191,
      "learning_rate": 4.965698904581703e-05,
      "loss": 0.4149,
      "num_input_tokens_seen": 3689912,
      "step": 5630
    },
    {
      "epoch": 2.95335429769392,
      "grad_norm": 0.7191604375839233,
      "learning_rate": 4.965509885291157e-05,
      "loss": 0.5146,
      "num_input_tokens_seen": 3692824,
      "step": 5635
    },
    {
      "epoch": 2.9559748427672954,
      "grad_norm": 0.31888270378112793,
      "learning_rate": 4.9653203502440716e-05,
      "loss": 0.4964,
      "num_input_tokens_seen": 3695736,
      "step": 5640
    },
    {
      "epoch": 2.958595387840671,
      "grad_norm": 0.43789923191070557,
      "learning_rate": 4.965130299480095e-05,
      "loss": 0.4994,
      "num_input_tokens_seen": 3698648,
      "step": 5645
    },
    {
      "epoch": 2.961215932914046,
      "grad_norm": 0.5129582285881042,
      "learning_rate": 4.9649397330389834e-05,
      "loss": 0.4431,
      "num_input_tokens_seen": 3701176,
      "step": 5650
    },
    {
      "epoch": 2.9638364779874213,
      "grad_norm": 0.7868735790252686,
      "learning_rate": 4.964748650960601e-05,
      "loss": 0.6419,
      "num_input_tokens_seen": 3703960,
      "step": 5655
    },
    {
      "epoch": 2.9664570230607965,
      "grad_norm": 0.3470224142074585,
      "learning_rate": 4.96455705328492e-05,
      "loss": 0.4425,
      "num_input_tokens_seen": 3707352,
      "step": 5660
    },
    {
      "epoch": 2.969077568134172,
      "grad_norm": 0.569758951663971,
      "learning_rate": 4.964364940052021e-05,
      "loss": 0.7098,
      "num_input_tokens_seen": 3710264,
      "step": 5665
    },
    {
      "epoch": 2.9716981132075473,
      "grad_norm": 0.6302332878112793,
      "learning_rate": 4.964172311302091e-05,
      "loss": 0.6192,
      "num_input_tokens_seen": 3713144,
      "step": 5670
    },
    {
      "epoch": 2.9743186582809225,
      "grad_norm": 0.429218053817749,
      "learning_rate": 4.9639791670754265e-05,
      "loss": 0.3493,
      "num_input_tokens_seen": 3715640,
      "step": 5675
    },
    {
      "epoch": 2.9769392033542976,
      "grad_norm": 0.4170079827308655,
      "learning_rate": 4.9637855074124305e-05,
      "loss": 0.446,
      "num_input_tokens_seen": 3718616,
      "step": 5680
    },
    {
      "epoch": 2.979559748427673,
      "grad_norm": 0.5156463384628296,
      "learning_rate": 4.963591332353615e-05,
      "loss": 0.3791,
      "num_input_tokens_seen": 3721560,
      "step": 5685
    },
    {
      "epoch": 2.982180293501048,
      "grad_norm": 0.305982768535614,
      "learning_rate": 4.9633966419396e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 3724920,
      "step": 5690
    },
    {
      "epoch": 2.9848008385744236,
      "grad_norm": 0.3405451476573944,
      "learning_rate": 4.96320143621111e-05,
      "loss": 0.4361,
      "num_input_tokens_seen": 3727576,
      "step": 5695
    },
    {
      "epoch": 2.9874213836477987,
      "grad_norm": 0.37566012144088745,
      "learning_rate": 4.963005715208981e-05,
      "loss": 0.5346,
      "num_input_tokens_seen": 3731224,
      "step": 5700
    },
    {
      "epoch": 2.990041928721174,
      "grad_norm": 0.4335268437862396,
      "learning_rate": 4.962809478974157e-05,
      "loss": 0.4689,
      "num_input_tokens_seen": 3734392,
      "step": 5705
    },
    {
      "epoch": 2.9926624737945495,
      "grad_norm": 0.47620856761932373,
      "learning_rate": 4.962612727547686e-05,
      "loss": 0.606,
      "num_input_tokens_seen": 3737848,
      "step": 5710
    },
    {
      "epoch": 2.9952830188679247,
      "grad_norm": 0.3142266273498535,
      "learning_rate": 4.962415460970728e-05,
      "loss": 0.3797,
      "num_input_tokens_seen": 3742264,
      "step": 5715
    },
    {
      "epoch": 2.9979035639413,
      "grad_norm": 0.4803931713104248,
      "learning_rate": 4.962217679284548e-05,
      "loss": 0.4748,
      "num_input_tokens_seen": 3744792,
      "step": 5720
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.5001356601715088,
      "eval_runtime": 9.5979,
      "eval_samples_per_second": 88.353,
      "eval_steps_per_second": 22.088,
      "num_input_tokens_seen": 3746728,
      "step": 5724
    },
    {
      "epoch": 3.000524109014675,
      "grad_norm": 0.3834708333015442,
      "learning_rate": 4.962019382530521e-05,
      "loss": 0.7762,
      "num_input_tokens_seen": 3747560,
      "step": 5725
    },
    {
      "epoch": 3.00314465408805,
      "grad_norm": 0.6551494002342224,
      "learning_rate": 4.9618205707501256e-05,
      "loss": 0.4701,
      "num_input_tokens_seen": 3750440,
      "step": 5730
    },
    {
      "epoch": 3.0057651991614254,
      "grad_norm": 0.5001507997512817,
      "learning_rate": 4.961621243984954e-05,
      "loss": 0.4883,
      "num_input_tokens_seen": 3753544,
      "step": 5735
    },
    {
      "epoch": 3.008385744234801,
      "grad_norm": 0.37910300493240356,
      "learning_rate": 4.9614214022767014e-05,
      "loss": 0.5327,
      "num_input_tokens_seen": 3757768,
      "step": 5740
    },
    {
      "epoch": 3.011006289308176,
      "grad_norm": 0.34988337755203247,
      "learning_rate": 4.9612210456671735e-05,
      "loss": 0.5248,
      "num_input_tokens_seen": 3761032,
      "step": 5745
    },
    {
      "epoch": 3.0136268343815513,
      "grad_norm": 0.7432518601417542,
      "learning_rate": 4.961020174198281e-05,
      "loss": 0.4613,
      "num_input_tokens_seen": 3763560,
      "step": 5750
    },
    {
      "epoch": 3.0162473794549265,
      "grad_norm": 0.3903621435165405,
      "learning_rate": 4.9608187879120454e-05,
      "loss": 0.4976,
      "num_input_tokens_seen": 3766376,
      "step": 5755
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 0.39276498556137085,
      "learning_rate": 4.9606168868505936e-05,
      "loss": 0.7409,
      "num_input_tokens_seen": 3769544,
      "step": 5760
    },
    {
      "epoch": 3.0214884696016773,
      "grad_norm": 0.5546068549156189,
      "learning_rate": 4.9604144710561616e-05,
      "loss": 0.4919,
      "num_input_tokens_seen": 3772616,
      "step": 5765
    },
    {
      "epoch": 3.0241090146750524,
      "grad_norm": 0.6837950348854065,
      "learning_rate": 4.960211540571092e-05,
      "loss": 0.4931,
      "num_input_tokens_seen": 3776712,
      "step": 5770
    },
    {
      "epoch": 3.0267295597484276,
      "grad_norm": 0.3973993957042694,
      "learning_rate": 4.960008095437836e-05,
      "loss": 0.516,
      "num_input_tokens_seen": 3779560,
      "step": 5775
    },
    {
      "epoch": 3.029350104821803,
      "grad_norm": 0.4177587032318115,
      "learning_rate": 4.959804135698951e-05,
      "loss": 0.6193,
      "num_input_tokens_seen": 3782056,
      "step": 5780
    },
    {
      "epoch": 3.0319706498951784,
      "grad_norm": 0.47817763686180115,
      "learning_rate": 4.959599661397104e-05,
      "loss": 0.5696,
      "num_input_tokens_seen": 3784616,
      "step": 5785
    },
    {
      "epoch": 3.0345911949685536,
      "grad_norm": 0.4996304512023926,
      "learning_rate": 4.959394672575068e-05,
      "loss": 0.4312,
      "num_input_tokens_seen": 3787368,
      "step": 5790
    },
    {
      "epoch": 3.0372117400419287,
      "grad_norm": 0.45566365122795105,
      "learning_rate": 4.959189169275725e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 3790184,
      "step": 5795
    },
    {
      "epoch": 3.039832285115304,
      "grad_norm": 0.5378088355064392,
      "learning_rate": 4.958983151542065e-05,
      "loss": 0.5198,
      "num_input_tokens_seen": 3793256,
      "step": 5800
    },
    {
      "epoch": 3.042452830188679,
      "grad_norm": 0.45484763383865356,
      "learning_rate": 4.958776619417182e-05,
      "loss": 0.568,
      "num_input_tokens_seen": 3796360,
      "step": 5805
    },
    {
      "epoch": 3.0450733752620547,
      "grad_norm": 0.44137391448020935,
      "learning_rate": 4.9585695729442814e-05,
      "loss": 0.492,
      "num_input_tokens_seen": 3799144,
      "step": 5810
    },
    {
      "epoch": 3.04769392033543,
      "grad_norm": 0.46135368943214417,
      "learning_rate": 4.9583620121666754e-05,
      "loss": 0.502,
      "num_input_tokens_seen": 3802280,
      "step": 5815
    },
    {
      "epoch": 3.050314465408805,
      "grad_norm": 0.3480454385280609,
      "learning_rate": 4.958153937127783e-05,
      "loss": 0.3577,
      "num_input_tokens_seen": 3807144,
      "step": 5820
    },
    {
      "epoch": 3.05293501048218,
      "grad_norm": 0.4541211724281311,
      "learning_rate": 4.95794534787113e-05,
      "loss": 0.4108,
      "num_input_tokens_seen": 3810088,
      "step": 5825
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 0.3192178010940552,
      "learning_rate": 4.957736244440353e-05,
      "loss": 0.4758,
      "num_input_tokens_seen": 3813000,
      "step": 5830
    },
    {
      "epoch": 3.058176100628931,
      "grad_norm": 0.6137679815292358,
      "learning_rate": 4.957526626879193e-05,
      "loss": 0.5109,
      "num_input_tokens_seen": 3815976,
      "step": 5835
    },
    {
      "epoch": 3.060796645702306,
      "grad_norm": 0.4701014459133148,
      "learning_rate": 4.957316495231499e-05,
      "loss": 0.5388,
      "num_input_tokens_seen": 3818824,
      "step": 5840
    },
    {
      "epoch": 3.0634171907756813,
      "grad_norm": 0.6289803385734558,
      "learning_rate": 4.9571058495412294e-05,
      "loss": 0.5144,
      "num_input_tokens_seen": 3823112,
      "step": 5845
    },
    {
      "epoch": 3.0660377358490565,
      "grad_norm": 0.4606979489326477,
      "learning_rate": 4.9568946898524485e-05,
      "loss": 0.6312,
      "num_input_tokens_seen": 3826984,
      "step": 5850
    },
    {
      "epoch": 3.068658280922432,
      "grad_norm": 0.9917166829109192,
      "learning_rate": 4.956683016209327e-05,
      "loss": 0.5124,
      "num_input_tokens_seen": 3829384,
      "step": 5855
    },
    {
      "epoch": 3.0712788259958073,
      "grad_norm": 0.9890066981315613,
      "learning_rate": 4.956470828656147e-05,
      "loss": 0.5913,
      "num_input_tokens_seen": 3831976,
      "step": 5860
    },
    {
      "epoch": 3.0738993710691824,
      "grad_norm": 0.5465759038925171,
      "learning_rate": 4.956258127237294e-05,
      "loss": 0.4623,
      "num_input_tokens_seen": 3834728,
      "step": 5865
    },
    {
      "epoch": 3.0765199161425576,
      "grad_norm": 0.8676737546920776,
      "learning_rate": 4.956044911997263e-05,
      "loss": 0.4028,
      "num_input_tokens_seen": 3837064,
      "step": 5870
    },
    {
      "epoch": 3.0791404612159328,
      "grad_norm": 0.480347603559494,
      "learning_rate": 4.955831182980657e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 3840456,
      "step": 5875
    },
    {
      "epoch": 3.0817610062893084,
      "grad_norm": 0.7474439740180969,
      "learning_rate": 4.955616940232185e-05,
      "loss": 0.5917,
      "num_input_tokens_seen": 3843528,
      "step": 5880
    },
    {
      "epoch": 3.0843815513626835,
      "grad_norm": 0.8100748658180237,
      "learning_rate": 4.9554021837966636e-05,
      "loss": 0.3809,
      "num_input_tokens_seen": 3845960,
      "step": 5885
    },
    {
      "epoch": 3.0870020964360587,
      "grad_norm": 0.4429990351200104,
      "learning_rate": 4.955186913719019e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 3851656,
      "step": 5890
    },
    {
      "epoch": 3.089622641509434,
      "grad_norm": 0.5497224926948547,
      "learning_rate": 4.954971130044281e-05,
      "loss": 0.4211,
      "num_input_tokens_seen": 3855496,
      "step": 5895
    },
    {
      "epoch": 3.092243186582809,
      "grad_norm": 0.30883538722991943,
      "learning_rate": 4.954754832817592e-05,
      "loss": 0.4845,
      "num_input_tokens_seen": 3858760,
      "step": 5900
    },
    {
      "epoch": 3.0948637316561847,
      "grad_norm": 0.486576110124588,
      "learning_rate": 4.954538022084196e-05,
      "loss": 0.6302,
      "num_input_tokens_seen": 3861640,
      "step": 5905
    },
    {
      "epoch": 3.09748427672956,
      "grad_norm": 0.4820285737514496,
      "learning_rate": 4.9543206978894496e-05,
      "loss": 0.437,
      "num_input_tokens_seen": 3864392,
      "step": 5910
    },
    {
      "epoch": 3.100104821802935,
      "grad_norm": 0.3375043570995331,
      "learning_rate": 4.954102860278813e-05,
      "loss": 0.6278,
      "num_input_tokens_seen": 3867880,
      "step": 5915
    },
    {
      "epoch": 3.10272536687631,
      "grad_norm": 0.6273105144500732,
      "learning_rate": 4.9538845092978567e-05,
      "loss": 0.6239,
      "num_input_tokens_seen": 3871080,
      "step": 5920
    },
    {
      "epoch": 3.1053459119496853,
      "grad_norm": 0.48640406131744385,
      "learning_rate": 4.953665644992256e-05,
      "loss": 0.5112,
      "num_input_tokens_seen": 3875112,
      "step": 5925
    },
    {
      "epoch": 3.107966457023061,
      "grad_norm": 0.41752132773399353,
      "learning_rate": 4.953446267407795e-05,
      "loss": 0.38,
      "num_input_tokens_seen": 3877704,
      "step": 5930
    },
    {
      "epoch": 3.110587002096436,
      "grad_norm": 0.4022509753704071,
      "learning_rate": 4.953226376590365e-05,
      "loss": 0.5714,
      "num_input_tokens_seen": 3881064,
      "step": 5935
    },
    {
      "epoch": 3.1132075471698113,
      "grad_norm": 0.6197505593299866,
      "learning_rate": 4.953005972585965e-05,
      "loss": 0.689,
      "num_input_tokens_seen": 3883816,
      "step": 5940
    },
    {
      "epoch": 3.1158280922431865,
      "grad_norm": 0.32613444328308105,
      "learning_rate": 4.952785055440701e-05,
      "loss": 0.4698,
      "num_input_tokens_seen": 3887720,
      "step": 5945
    },
    {
      "epoch": 3.1184486373165616,
      "grad_norm": 0.45736604928970337,
      "learning_rate": 4.9525636252007864e-05,
      "loss": 0.4087,
      "num_input_tokens_seen": 3890440,
      "step": 5950
    },
    {
      "epoch": 3.1210691823899372,
      "grad_norm": 0.4683159589767456,
      "learning_rate": 4.9523416819125425e-05,
      "loss": 0.4693,
      "num_input_tokens_seen": 3895784,
      "step": 5955
    },
    {
      "epoch": 3.1236897274633124,
      "grad_norm": 0.39507338404655457,
      "learning_rate": 4.952119225622396e-05,
      "loss": 0.4744,
      "num_input_tokens_seen": 3900296,
      "step": 5960
    },
    {
      "epoch": 3.1263102725366876,
      "grad_norm": 0.6023291945457458,
      "learning_rate": 4.951896256376882e-05,
      "loss": 0.3789,
      "num_input_tokens_seen": 3902952,
      "step": 5965
    },
    {
      "epoch": 3.1289308176100628,
      "grad_norm": 0.6501587629318237,
      "learning_rate": 4.9516727742226444e-05,
      "loss": 0.6886,
      "num_input_tokens_seen": 3905576,
      "step": 5970
    },
    {
      "epoch": 3.131551362683438,
      "grad_norm": 0.3388316035270691,
      "learning_rate": 4.9514487792064323e-05,
      "loss": 0.3931,
      "num_input_tokens_seen": 3911368,
      "step": 5975
    },
    {
      "epoch": 3.1341719077568135,
      "grad_norm": 0.2652741074562073,
      "learning_rate": 4.951224271375104e-05,
      "loss": 0.5828,
      "num_input_tokens_seen": 3915560,
      "step": 5980
    },
    {
      "epoch": 3.1367924528301887,
      "grad_norm": 0.3971351087093353,
      "learning_rate": 4.950999250775623e-05,
      "loss": 0.511,
      "num_input_tokens_seen": 3918376,
      "step": 5985
    },
    {
      "epoch": 3.139412997903564,
      "grad_norm": 0.47374626994132996,
      "learning_rate": 4.950773717455061e-05,
      "loss": 0.4691,
      "num_input_tokens_seen": 3921608,
      "step": 5990
    },
    {
      "epoch": 3.142033542976939,
      "grad_norm": 0.5396822690963745,
      "learning_rate": 4.950547671460596e-05,
      "loss": 0.5115,
      "num_input_tokens_seen": 3924072,
      "step": 5995
    },
    {
      "epoch": 3.1446540880503147,
      "grad_norm": 0.6659433841705322,
      "learning_rate": 4.950321112839517e-05,
      "loss": 0.5101,
      "num_input_tokens_seen": 3927080,
      "step": 6000
    },
    {
      "epoch": 3.14727463312369,
      "grad_norm": 0.734555721282959,
      "learning_rate": 4.9500940416392156e-05,
      "loss": 0.5594,
      "num_input_tokens_seen": 3929768,
      "step": 6005
    },
    {
      "epoch": 3.149895178197065,
      "grad_norm": 0.45670631527900696,
      "learning_rate": 4.949866457907193e-05,
      "loss": 0.5651,
      "num_input_tokens_seen": 3933512,
      "step": 6010
    },
    {
      "epoch": 3.15251572327044,
      "grad_norm": 0.44757476449012756,
      "learning_rate": 4.9496383616910565e-05,
      "loss": 0.5969,
      "num_input_tokens_seen": 3937032,
      "step": 6015
    },
    {
      "epoch": 3.1551362683438153,
      "grad_norm": 0.3458891212940216,
      "learning_rate": 4.949409753038522e-05,
      "loss": 0.4939,
      "num_input_tokens_seen": 3940616,
      "step": 6020
    },
    {
      "epoch": 3.157756813417191,
      "grad_norm": 0.5741514563560486,
      "learning_rate": 4.949180631997411e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 3944296,
      "step": 6025
    },
    {
      "epoch": 3.160377358490566,
      "grad_norm": 0.533291757106781,
      "learning_rate": 4.948950998615654e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 3946760,
      "step": 6030
    },
    {
      "epoch": 3.1629979035639413,
      "grad_norm": 0.5971566438674927,
      "learning_rate": 4.9487208529412876e-05,
      "loss": 0.5089,
      "num_input_tokens_seen": 3950984,
      "step": 6035
    },
    {
      "epoch": 3.1656184486373165,
      "grad_norm": 0.5262508392333984,
      "learning_rate": 4.9484901950224546e-05,
      "loss": 0.4801,
      "num_input_tokens_seen": 3953576,
      "step": 6040
    },
    {
      "epoch": 3.1682389937106916,
      "grad_norm": 0.31992313265800476,
      "learning_rate": 4.948259024907407e-05,
      "loss": 0.3847,
      "num_input_tokens_seen": 3956552,
      "step": 6045
    },
    {
      "epoch": 3.1708595387840672,
      "grad_norm": 0.5094180107116699,
      "learning_rate": 4.948027342644502e-05,
      "loss": 0.5484,
      "num_input_tokens_seen": 3960584,
      "step": 6050
    },
    {
      "epoch": 3.1734800838574424,
      "grad_norm": 0.4054414629936218,
      "learning_rate": 4.947795148282206e-05,
      "loss": 0.4691,
      "num_input_tokens_seen": 3963496,
      "step": 6055
    },
    {
      "epoch": 3.1761006289308176,
      "grad_norm": 0.716635525226593,
      "learning_rate": 4.947562441869091e-05,
      "loss": 0.4682,
      "num_input_tokens_seen": 3970312,
      "step": 6060
    },
    {
      "epoch": 3.1787211740041927,
      "grad_norm": 0.4738166928291321,
      "learning_rate": 4.947329223453836e-05,
      "loss": 0.4599,
      "num_input_tokens_seen": 3973288,
      "step": 6065
    },
    {
      "epoch": 3.181341719077568,
      "grad_norm": 0.98904949426651,
      "learning_rate": 4.947095493085229e-05,
      "loss": 0.5504,
      "num_input_tokens_seen": 3976456,
      "step": 6070
    },
    {
      "epoch": 3.1839622641509435,
      "grad_norm": 0.4045560359954834,
      "learning_rate": 4.9468612508121616e-05,
      "loss": 0.4785,
      "num_input_tokens_seen": 3979528,
      "step": 6075
    },
    {
      "epoch": 3.1865828092243187,
      "grad_norm": 0.42357251048088074,
      "learning_rate": 4.9466264966836365e-05,
      "loss": 0.5014,
      "num_input_tokens_seen": 3982536,
      "step": 6080
    },
    {
      "epoch": 3.189203354297694,
      "grad_norm": 0.23171833157539368,
      "learning_rate": 4.94639123074876e-05,
      "loss": 0.4076,
      "num_input_tokens_seen": 3988904,
      "step": 6085
    },
    {
      "epoch": 3.191823899371069,
      "grad_norm": 0.4270811080932617,
      "learning_rate": 4.946155453056749e-05,
      "loss": 0.486,
      "num_input_tokens_seen": 3991944,
      "step": 6090
    },
    {
      "epoch": 3.1944444444444446,
      "grad_norm": 0.5925570130348206,
      "learning_rate": 4.945919163656923e-05,
      "loss": 0.5095,
      "num_input_tokens_seen": 3995688,
      "step": 6095
    },
    {
      "epoch": 3.19706498951782,
      "grad_norm": 0.6018973588943481,
      "learning_rate": 4.945682362598715e-05,
      "loss": 0.4802,
      "num_input_tokens_seen": 3999016,
      "step": 6100
    },
    {
      "epoch": 3.199685534591195,
      "grad_norm": 0.6141414642333984,
      "learning_rate": 4.945445049931656e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 4002120,
      "step": 6105
    },
    {
      "epoch": 3.20230607966457,
      "grad_norm": 0.39087846875190735,
      "learning_rate": 4.945207225705393e-05,
      "loss": 0.6055,
      "num_input_tokens_seen": 4005640,
      "step": 6110
    },
    {
      "epoch": 3.2049266247379453,
      "grad_norm": 0.3523576855659485,
      "learning_rate": 4.944968889969674e-05,
      "loss": 0.4575,
      "num_input_tokens_seen": 4008968,
      "step": 6115
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 0.2830754220485687,
      "learning_rate": 4.944730042774358e-05,
      "loss": 0.3348,
      "num_input_tokens_seen": 4011528,
      "step": 6120
    },
    {
      "epoch": 3.210167714884696,
      "grad_norm": 0.42360684275627136,
      "learning_rate": 4.944490684169408e-05,
      "loss": 0.4927,
      "num_input_tokens_seen": 4014696,
      "step": 6125
    },
    {
      "epoch": 3.2127882599580713,
      "grad_norm": 0.3641110062599182,
      "learning_rate": 4.9442508142048945e-05,
      "loss": 0.3994,
      "num_input_tokens_seen": 4017384,
      "step": 6130
    },
    {
      "epoch": 3.2154088050314464,
      "grad_norm": 0.6748549342155457,
      "learning_rate": 4.9440104329309956e-05,
      "loss": 0.4506,
      "num_input_tokens_seen": 4020520,
      "step": 6135
    },
    {
      "epoch": 3.2180293501048216,
      "grad_norm": 0.6086339354515076,
      "learning_rate": 4.943769540397998e-05,
      "loss": 0.5724,
      "num_input_tokens_seen": 4023624,
      "step": 6140
    },
    {
      "epoch": 3.220649895178197,
      "grad_norm": 0.48034337162971497,
      "learning_rate": 4.943528136656292e-05,
      "loss": 0.5577,
      "num_input_tokens_seen": 4027304,
      "step": 6145
    },
    {
      "epoch": 3.2232704402515724,
      "grad_norm": 0.5294297933578491,
      "learning_rate": 4.943286221756377e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 4030536,
      "step": 6150
    },
    {
      "epoch": 3.2258909853249476,
      "grad_norm": 0.6584643125534058,
      "learning_rate": 4.9430437957488586e-05,
      "loss": 0.4749,
      "num_input_tokens_seen": 4033448,
      "step": 6155
    },
    {
      "epoch": 3.2285115303983227,
      "grad_norm": 0.5392314195632935,
      "learning_rate": 4.94280085868445e-05,
      "loss": 0.4734,
      "num_input_tokens_seen": 4036264,
      "step": 6160
    },
    {
      "epoch": 3.231132075471698,
      "grad_norm": 0.3699185252189636,
      "learning_rate": 4.942557410613972e-05,
      "loss": 0.4384,
      "num_input_tokens_seen": 4040040,
      "step": 6165
    },
    {
      "epoch": 3.2337526205450735,
      "grad_norm": 0.5373073220252991,
      "learning_rate": 4.942313451588348e-05,
      "loss": 0.4581,
      "num_input_tokens_seen": 4042856,
      "step": 6170
    },
    {
      "epoch": 3.2363731656184487,
      "grad_norm": 0.3053615689277649,
      "learning_rate": 4.942068981658614e-05,
      "loss": 0.5089,
      "num_input_tokens_seen": 4046312,
      "step": 6175
    },
    {
      "epoch": 3.238993710691824,
      "grad_norm": 0.4133860170841217,
      "learning_rate": 4.9418240008759097e-05,
      "loss": 0.4794,
      "num_input_tokens_seen": 4049128,
      "step": 6180
    },
    {
      "epoch": 3.241614255765199,
      "grad_norm": 0.43272218108177185,
      "learning_rate": 4.9415785092914826e-05,
      "loss": 0.4203,
      "num_input_tokens_seen": 4052776,
      "step": 6185
    },
    {
      "epoch": 3.2442348008385746,
      "grad_norm": 0.435716837644577,
      "learning_rate": 4.941332506956686e-05,
      "loss": 0.5348,
      "num_input_tokens_seen": 4055720,
      "step": 6190
    },
    {
      "epoch": 3.24685534591195,
      "grad_norm": 0.33757975697517395,
      "learning_rate": 4.941085993922981e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 4060968,
      "step": 6195
    },
    {
      "epoch": 3.249475890985325,
      "grad_norm": 0.37824952602386475,
      "learning_rate": 4.9408389702419354e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 4063784,
      "step": 6200
    },
    {
      "epoch": 3.2520964360587,
      "grad_norm": 0.3352862298488617,
      "learning_rate": 4.940591435965224e-05,
      "loss": 0.5566,
      "num_input_tokens_seen": 4066728,
      "step": 6205
    },
    {
      "epoch": 3.2547169811320753,
      "grad_norm": 0.6511121392250061,
      "learning_rate": 4.9403433911446285e-05,
      "loss": 0.3745,
      "num_input_tokens_seen": 4071368,
      "step": 6210
    },
    {
      "epoch": 3.257337526205451,
      "grad_norm": 0.4825773239135742,
      "learning_rate": 4.9400948358320356e-05,
      "loss": 0.6218,
      "num_input_tokens_seen": 4074152,
      "step": 6215
    },
    {
      "epoch": 3.259958071278826,
      "grad_norm": 0.3040751516819,
      "learning_rate": 4.939845770079442e-05,
      "loss": 0.4114,
      "num_input_tokens_seen": 4077672,
      "step": 6220
    },
    {
      "epoch": 3.2625786163522013,
      "grad_norm": 0.6837431788444519,
      "learning_rate": 4.9395961939389475e-05,
      "loss": 0.6951,
      "num_input_tokens_seen": 4081128,
      "step": 6225
    },
    {
      "epoch": 3.2651991614255764,
      "grad_norm": 0.6284922361373901,
      "learning_rate": 4.939346107462762e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 4083976,
      "step": 6230
    },
    {
      "epoch": 3.2678197064989516,
      "grad_norm": 0.4923025965690613,
      "learning_rate": 4.9390955107032e-05,
      "loss": 0.522,
      "num_input_tokens_seen": 4087432,
      "step": 6235
    },
    {
      "epoch": 3.270440251572327,
      "grad_norm": 0.28879010677337646,
      "learning_rate": 4.938844403712685e-05,
      "loss": 0.4906,
      "num_input_tokens_seen": 4093096,
      "step": 6240
    },
    {
      "epoch": 3.2730607966457024,
      "grad_norm": 1.0398098230361938,
      "learning_rate": 4.938592786543743e-05,
      "loss": 0.4975,
      "num_input_tokens_seen": 4096424,
      "step": 6245
    },
    {
      "epoch": 3.2756813417190775,
      "grad_norm": 0.3969496190547943,
      "learning_rate": 4.938340659249012e-05,
      "loss": 0.5464,
      "num_input_tokens_seen": 4099944,
      "step": 6250
    },
    {
      "epoch": 3.2783018867924527,
      "grad_norm": 0.629136860370636,
      "learning_rate": 4.938088021881233e-05,
      "loss": 0.4358,
      "num_input_tokens_seen": 4103368,
      "step": 6255
    },
    {
      "epoch": 3.280922431865828,
      "grad_norm": 0.875653088092804,
      "learning_rate": 4.937834874493255e-05,
      "loss": 0.4374,
      "num_input_tokens_seen": 4106248,
      "step": 6260
    },
    {
      "epoch": 3.2835429769392035,
      "grad_norm": 0.35878103971481323,
      "learning_rate": 4.9375812171380334e-05,
      "loss": 0.3771,
      "num_input_tokens_seen": 4109128,
      "step": 6265
    },
    {
      "epoch": 3.2861635220125787,
      "grad_norm": 0.6071704030036926,
      "learning_rate": 4.9373270498686306e-05,
      "loss": 0.4927,
      "num_input_tokens_seen": 4111880,
      "step": 6270
    },
    {
      "epoch": 3.288784067085954,
      "grad_norm": 0.7656713724136353,
      "learning_rate": 4.937072372738216e-05,
      "loss": 0.4587,
      "num_input_tokens_seen": 4115208,
      "step": 6275
    },
    {
      "epoch": 3.291404612159329,
      "grad_norm": 0.40729522705078125,
      "learning_rate": 4.9368171858000635e-05,
      "loss": 0.6372,
      "num_input_tokens_seen": 4118440,
      "step": 6280
    },
    {
      "epoch": 3.2940251572327046,
      "grad_norm": 0.6987012624740601,
      "learning_rate": 4.9365614891075576e-05,
      "loss": 0.4927,
      "num_input_tokens_seen": 4121608,
      "step": 6285
    },
    {
      "epoch": 3.29664570230608,
      "grad_norm": 0.5977954864501953,
      "learning_rate": 4.936305282714185e-05,
      "loss": 0.5799,
      "num_input_tokens_seen": 4124648,
      "step": 6290
    },
    {
      "epoch": 3.299266247379455,
      "grad_norm": 0.3578675389289856,
      "learning_rate": 4.936048566673542e-05,
      "loss": 0.508,
      "num_input_tokens_seen": 4127688,
      "step": 6295
    },
    {
      "epoch": 3.30188679245283,
      "grad_norm": 0.594246506690979,
      "learning_rate": 4.935791341039332e-05,
      "loss": 0.4441,
      "num_input_tokens_seen": 4131528,
      "step": 6300
    },
    {
      "epoch": 3.3045073375262053,
      "grad_norm": 0.48347461223602295,
      "learning_rate": 4.935533605865361e-05,
      "loss": 0.4919,
      "num_input_tokens_seen": 4134216,
      "step": 6305
    },
    {
      "epoch": 3.307127882599581,
      "grad_norm": 0.8595330119132996,
      "learning_rate": 4.935275361205547e-05,
      "loss": 0.4298,
      "num_input_tokens_seen": 4137544,
      "step": 6310
    },
    {
      "epoch": 3.309748427672956,
      "grad_norm": 0.30133089423179626,
      "learning_rate": 4.93501660711391e-05,
      "loss": 0.5774,
      "num_input_tokens_seen": 4140424,
      "step": 6315
    },
    {
      "epoch": 3.3123689727463312,
      "grad_norm": 2.2484123706817627,
      "learning_rate": 4.934757343644578e-05,
      "loss": 0.5332,
      "num_input_tokens_seen": 4142984,
      "step": 6320
    },
    {
      "epoch": 3.3149895178197064,
      "grad_norm": 0.868732213973999,
      "learning_rate": 4.934497570851788e-05,
      "loss": 0.4729,
      "num_input_tokens_seen": 4145768,
      "step": 6325
    },
    {
      "epoch": 3.3176100628930816,
      "grad_norm": 0.4200952351093292,
      "learning_rate": 4.9342372887898804e-05,
      "loss": 0.3904,
      "num_input_tokens_seen": 4148840,
      "step": 6330
    },
    {
      "epoch": 3.320230607966457,
      "grad_norm": 0.2766086757183075,
      "learning_rate": 4.9339764975133026e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 4152840,
      "step": 6335
    },
    {
      "epoch": 3.3228511530398324,
      "grad_norm": 0.3844527304172516,
      "learning_rate": 4.93371519707661e-05,
      "loss": 0.5999,
      "num_input_tokens_seen": 4155400,
      "step": 6340
    },
    {
      "epoch": 3.3254716981132075,
      "grad_norm": 0.43256211280822754,
      "learning_rate": 4.933453387534464e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 4158312,
      "step": 6345
    },
    {
      "epoch": 3.3280922431865827,
      "grad_norm": 0.6308853626251221,
      "learning_rate": 4.933191068941631e-05,
      "loss": 0.4955,
      "num_input_tokens_seen": 4161096,
      "step": 6350
    },
    {
      "epoch": 3.330712788259958,
      "grad_norm": 0.40943679213523865,
      "learning_rate": 4.932928241352987e-05,
      "loss": 0.4024,
      "num_input_tokens_seen": 4164520,
      "step": 6355
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.4086556136608124,
      "learning_rate": 4.9326649048235095e-05,
      "loss": 0.5786,
      "num_input_tokens_seen": 4168264,
      "step": 6360
    },
    {
      "epoch": 3.3359538784067087,
      "grad_norm": 0.3518812656402588,
      "learning_rate": 4.932401059408287e-05,
      "loss": 0.4589,
      "num_input_tokens_seen": 4171112,
      "step": 6365
    },
    {
      "epoch": 3.338574423480084,
      "grad_norm": 0.6022894382476807,
      "learning_rate": 4.932136705162514e-05,
      "loss": 0.5318,
      "num_input_tokens_seen": 4174120,
      "step": 6370
    },
    {
      "epoch": 3.341194968553459,
      "grad_norm": 0.5377420783042908,
      "learning_rate": 4.93187184214149e-05,
      "loss": 0.6441,
      "num_input_tokens_seen": 4177096,
      "step": 6375
    },
    {
      "epoch": 3.3438155136268346,
      "grad_norm": 0.3724648654460907,
      "learning_rate": 4.9316064704006205e-05,
      "loss": 0.5378,
      "num_input_tokens_seen": 4180104,
      "step": 6380
    },
    {
      "epoch": 3.3464360587002098,
      "grad_norm": 0.4467386305332184,
      "learning_rate": 4.931340589995419e-05,
      "loss": 0.5387,
      "num_input_tokens_seen": 4184008,
      "step": 6385
    },
    {
      "epoch": 3.349056603773585,
      "grad_norm": 0.327923446893692,
      "learning_rate": 4.9310742009815035e-05,
      "loss": 0.4118,
      "num_input_tokens_seen": 4186984,
      "step": 6390
    },
    {
      "epoch": 3.35167714884696,
      "grad_norm": 0.382425457239151,
      "learning_rate": 4.930807303414601e-05,
      "loss": 0.3777,
      "num_input_tokens_seen": 4190216,
      "step": 6395
    },
    {
      "epoch": 3.3542976939203353,
      "grad_norm": 0.4800722599029541,
      "learning_rate": 4.930539897350542e-05,
      "loss": 0.5231,
      "num_input_tokens_seen": 4193256,
      "step": 6400
    },
    {
      "epoch": 3.3569182389937104,
      "grad_norm": 0.3366715610027313,
      "learning_rate": 4.930271982845266e-05,
      "loss": 0.5497,
      "num_input_tokens_seen": 4197256,
      "step": 6405
    },
    {
      "epoch": 3.359538784067086,
      "grad_norm": 0.5580301880836487,
      "learning_rate": 4.930003559954817e-05,
      "loss": 0.458,
      "num_input_tokens_seen": 4200168,
      "step": 6410
    },
    {
      "epoch": 3.3621593291404612,
      "grad_norm": 0.39149245619773865,
      "learning_rate": 4.929734628735346e-05,
      "loss": 0.3565,
      "num_input_tokens_seen": 4203848,
      "step": 6415
    },
    {
      "epoch": 3.3647798742138364,
      "grad_norm": 0.4357253313064575,
      "learning_rate": 4.9294651892431115e-05,
      "loss": 0.4326,
      "num_input_tokens_seen": 4207592,
      "step": 6420
    },
    {
      "epoch": 3.3674004192872116,
      "grad_norm": 0.4253406226634979,
      "learning_rate": 4.929195241534476e-05,
      "loss": 0.4938,
      "num_input_tokens_seen": 4211912,
      "step": 6425
    },
    {
      "epoch": 3.370020964360587,
      "grad_norm": 0.26312586665153503,
      "learning_rate": 4.9289247856659095e-05,
      "loss": 0.565,
      "num_input_tokens_seen": 4214408,
      "step": 6430
    },
    {
      "epoch": 3.3726415094339623,
      "grad_norm": 0.7449379563331604,
      "learning_rate": 4.9286538216939883e-05,
      "loss": 0.5121,
      "num_input_tokens_seen": 4216712,
      "step": 6435
    },
    {
      "epoch": 3.3752620545073375,
      "grad_norm": 0.5595530271530151,
      "learning_rate": 4.9283823496753944e-05,
      "loss": 0.4727,
      "num_input_tokens_seen": 4220168,
      "step": 6440
    },
    {
      "epoch": 3.3778825995807127,
      "grad_norm": 0.5882972478866577,
      "learning_rate": 4.928110369666918e-05,
      "loss": 0.6992,
      "num_input_tokens_seen": 4223208,
      "step": 6445
    },
    {
      "epoch": 3.380503144654088,
      "grad_norm": 0.42681604623794556,
      "learning_rate": 4.927837881725455e-05,
      "loss": 0.4185,
      "num_input_tokens_seen": 4226760,
      "step": 6450
    },
    {
      "epoch": 3.3831236897274635,
      "grad_norm": 0.40926340222358704,
      "learning_rate": 4.927564885908003e-05,
      "loss": 0.4819,
      "num_input_tokens_seen": 4230152,
      "step": 6455
    },
    {
      "epoch": 3.3857442348008386,
      "grad_norm": 0.3249514102935791,
      "learning_rate": 4.9272913822716735e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 4233416,
      "step": 6460
    },
    {
      "epoch": 3.388364779874214,
      "grad_norm": 0.5752225518226624,
      "learning_rate": 4.927017370873678e-05,
      "loss": 0.5223,
      "num_input_tokens_seen": 4236328,
      "step": 6465
    },
    {
      "epoch": 3.390985324947589,
      "grad_norm": 0.760959267616272,
      "learning_rate": 4.926742851771338e-05,
      "loss": 0.4673,
      "num_input_tokens_seen": 4238760,
      "step": 6470
    },
    {
      "epoch": 3.3936058700209646,
      "grad_norm": 0.5185478329658508,
      "learning_rate": 4.926467825022079e-05,
      "loss": 0.4286,
      "num_input_tokens_seen": 4241992,
      "step": 6475
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 0.5997231602668762,
      "learning_rate": 4.926192290683434e-05,
      "loss": 0.4693,
      "num_input_tokens_seen": 4244584,
      "step": 6480
    },
    {
      "epoch": 3.398846960167715,
      "grad_norm": 0.6270609498023987,
      "learning_rate": 4.9259162488130404e-05,
      "loss": 0.4122,
      "num_input_tokens_seen": 4247496,
      "step": 6485
    },
    {
      "epoch": 3.40146750524109,
      "grad_norm": 0.5783266425132751,
      "learning_rate": 4.925639699468645e-05,
      "loss": 0.4006,
      "num_input_tokens_seen": 4250760,
      "step": 6490
    },
    {
      "epoch": 3.4040880503144653,
      "grad_norm": 0.5670232772827148,
      "learning_rate": 4.925362642708097e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 4253256,
      "step": 6495
    },
    {
      "epoch": 3.4067085953878404,
      "grad_norm": 0.40283554792404175,
      "learning_rate": 4.9250850785893536e-05,
      "loss": 0.4696,
      "num_input_tokens_seen": 4257320,
      "step": 6500
    },
    {
      "epoch": 3.409329140461216,
      "grad_norm": 0.23968206346035004,
      "learning_rate": 4.924807007170479e-05,
      "loss": 0.6417,
      "num_input_tokens_seen": 4261864,
      "step": 6505
    },
    {
      "epoch": 3.411949685534591,
      "grad_norm": 0.58461594581604,
      "learning_rate": 4.924528428509643e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 4264744,
      "step": 6510
    },
    {
      "epoch": 3.4145702306079664,
      "grad_norm": 0.48601919412612915,
      "learning_rate": 4.9242493426651196e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 4267080,
      "step": 6515
    },
    {
      "epoch": 3.4171907756813416,
      "grad_norm": 0.3978184163570404,
      "learning_rate": 4.92396974969529e-05,
      "loss": 0.5283,
      "num_input_tokens_seen": 4269928,
      "step": 6520
    },
    {
      "epoch": 3.419811320754717,
      "grad_norm": 0.5257227420806885,
      "learning_rate": 4.923689649658644e-05,
      "loss": 0.3952,
      "num_input_tokens_seen": 4272872,
      "step": 6525
    },
    {
      "epoch": 3.4224318658280923,
      "grad_norm": 0.6462187170982361,
      "learning_rate": 4.923409042613774e-05,
      "loss": 0.4939,
      "num_input_tokens_seen": 4275912,
      "step": 6530
    },
    {
      "epoch": 3.4250524109014675,
      "grad_norm": 0.36906933784484863,
      "learning_rate": 4.92312792861938e-05,
      "loss": 0.5179,
      "num_input_tokens_seen": 4280104,
      "step": 6535
    },
    {
      "epoch": 3.4276729559748427,
      "grad_norm": 0.3802386522293091,
      "learning_rate": 4.922846307734268e-05,
      "loss": 0.3649,
      "num_input_tokens_seen": 4283080,
      "step": 6540
    },
    {
      "epoch": 3.430293501048218,
      "grad_norm": 0.6438649296760559,
      "learning_rate": 4.9225641800173496e-05,
      "loss": 0.4447,
      "num_input_tokens_seen": 4286024,
      "step": 6545
    },
    {
      "epoch": 3.4329140461215935,
      "grad_norm": 0.3857273459434509,
      "learning_rate": 4.922281545527643e-05,
      "loss": 0.5561,
      "num_input_tokens_seen": 4289672,
      "step": 6550
    },
    {
      "epoch": 3.4355345911949686,
      "grad_norm": 0.5049158930778503,
      "learning_rate": 4.9219984043242715e-05,
      "loss": 0.4058,
      "num_input_tokens_seen": 4292840,
      "step": 6555
    },
    {
      "epoch": 3.438155136268344,
      "grad_norm": 0.28825172781944275,
      "learning_rate": 4.921714756466466e-05,
      "loss": 0.4163,
      "num_input_tokens_seen": 4296072,
      "step": 6560
    },
    {
      "epoch": 3.440775681341719,
      "grad_norm": 0.5793781876564026,
      "learning_rate": 4.921430602013561e-05,
      "loss": 0.5103,
      "num_input_tokens_seen": 4299720,
      "step": 6565
    },
    {
      "epoch": 3.4433962264150946,
      "grad_norm": 0.6361990571022034,
      "learning_rate": 4.9211459410249995e-05,
      "loss": 0.4809,
      "num_input_tokens_seen": 4302856,
      "step": 6570
    },
    {
      "epoch": 3.4460167714884697,
      "grad_norm": 0.3399712145328522,
      "learning_rate": 4.92086077356033e-05,
      "loss": 0.4418,
      "num_input_tokens_seen": 4305864,
      "step": 6575
    },
    {
      "epoch": 3.448637316561845,
      "grad_norm": 0.43119287490844727,
      "learning_rate": 4.920575099679204e-05,
      "loss": 0.378,
      "num_input_tokens_seen": 4309352,
      "step": 6580
    },
    {
      "epoch": 3.45125786163522,
      "grad_norm": 0.28900146484375,
      "learning_rate": 4.920288919441383e-05,
      "loss": 0.3817,
      "num_input_tokens_seen": 4313192,
      "step": 6585
    },
    {
      "epoch": 3.4538784067085953,
      "grad_norm": 0.5215328335762024,
      "learning_rate": 4.9200022329067316e-05,
      "loss": 0.5381,
      "num_input_tokens_seen": 4315976,
      "step": 6590
    },
    {
      "epoch": 3.4564989517819704,
      "grad_norm": 0.30627793073654175,
      "learning_rate": 4.9197150401352224e-05,
      "loss": 0.5021,
      "num_input_tokens_seen": 4319176,
      "step": 6595
    },
    {
      "epoch": 3.459119496855346,
      "grad_norm": 0.4358447194099426,
      "learning_rate": 4.919427341186933e-05,
      "loss": 0.7331,
      "num_input_tokens_seen": 4321960,
      "step": 6600
    },
    {
      "epoch": 3.461740041928721,
      "grad_norm": 0.7102927565574646,
      "learning_rate": 4.919139136122045e-05,
      "loss": 0.4463,
      "num_input_tokens_seen": 4324488,
      "step": 6605
    },
    {
      "epoch": 3.4643605870020964,
      "grad_norm": 0.2867971360683441,
      "learning_rate": 4.918850425000849e-05,
      "loss": 0.535,
      "num_input_tokens_seen": 4328264,
      "step": 6610
    },
    {
      "epoch": 3.4669811320754715,
      "grad_norm": 0.5778099298477173,
      "learning_rate": 4.918561207883741e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 4335528,
      "step": 6615
    },
    {
      "epoch": 3.469601677148847,
      "grad_norm": 0.5258058905601501,
      "learning_rate": 4.9182714848312195e-05,
      "loss": 0.6092,
      "num_input_tokens_seen": 4338408,
      "step": 6620
    },
    {
      "epoch": 3.4722222222222223,
      "grad_norm": 0.4493969678878784,
      "learning_rate": 4.917981255903893e-05,
      "loss": 0.7453,
      "num_input_tokens_seen": 4341512,
      "step": 6625
    },
    {
      "epoch": 3.4748427672955975,
      "grad_norm": 0.4610320031642914,
      "learning_rate": 4.917690521162474e-05,
      "loss": 0.5992,
      "num_input_tokens_seen": 4345960,
      "step": 6630
    },
    {
      "epoch": 3.4774633123689727,
      "grad_norm": 0.2995556592941284,
      "learning_rate": 4.9173992806677796e-05,
      "loss": 0.4767,
      "num_input_tokens_seen": 4349384,
      "step": 6635
    },
    {
      "epoch": 3.480083857442348,
      "grad_norm": 0.26790666580200195,
      "learning_rate": 4.917107534480735e-05,
      "loss": 0.4201,
      "num_input_tokens_seen": 4352712,
      "step": 6640
    },
    {
      "epoch": 3.4827044025157234,
      "grad_norm": 0.4566367268562317,
      "learning_rate": 4.916815282662369e-05,
      "loss": 0.4851,
      "num_input_tokens_seen": 4356680,
      "step": 6645
    },
    {
      "epoch": 3.4853249475890986,
      "grad_norm": 0.3385845720767975,
      "learning_rate": 4.91652252527382e-05,
      "loss": 0.4923,
      "num_input_tokens_seen": 4360040,
      "step": 6650
    },
    {
      "epoch": 3.487945492662474,
      "grad_norm": 0.882286012172699,
      "learning_rate": 4.9162292623763275e-05,
      "loss": 0.522,
      "num_input_tokens_seen": 4362504,
      "step": 6655
    },
    {
      "epoch": 3.490566037735849,
      "grad_norm": 0.3296046853065491,
      "learning_rate": 4.9159354940312386e-05,
      "loss": 0.5634,
      "num_input_tokens_seen": 4366280,
      "step": 6660
    },
    {
      "epoch": 3.4931865828092246,
      "grad_norm": 0.6075438857078552,
      "learning_rate": 4.915641220300007e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 4369320,
      "step": 6665
    },
    {
      "epoch": 3.4958071278825997,
      "grad_norm": 0.5863070487976074,
      "learning_rate": 4.9153464412441915e-05,
      "loss": 0.3429,
      "num_input_tokens_seen": 4372776,
      "step": 6670
    },
    {
      "epoch": 3.498427672955975,
      "grad_norm": 0.43080493807792664,
      "learning_rate": 4.915051156925456e-05,
      "loss": 0.5424,
      "num_input_tokens_seen": 4375048,
      "step": 6675
    },
    {
      "epoch": 3.50104821802935,
      "grad_norm": 0.24965620040893555,
      "learning_rate": 4.91475536740557e-05,
      "loss": 0.5548,
      "num_input_tokens_seen": 4378248,
      "step": 6680
    },
    {
      "epoch": 3.5036687631027252,
      "grad_norm": 0.666090190410614,
      "learning_rate": 4.914459072746411e-05,
      "loss": 0.6379,
      "num_input_tokens_seen": 4381352,
      "step": 6685
    },
    {
      "epoch": 3.5062893081761004,
      "grad_norm": 0.5594822764396667,
      "learning_rate": 4.91416227300996e-05,
      "loss": 0.4651,
      "num_input_tokens_seen": 4384424,
      "step": 6690
    },
    {
      "epoch": 3.508909853249476,
      "grad_norm": 0.5853273272514343,
      "learning_rate": 4.913864968258303e-05,
      "loss": 0.6729,
      "num_input_tokens_seen": 4387592,
      "step": 6695
    },
    {
      "epoch": 3.511530398322851,
      "grad_norm": 0.4407907724380493,
      "learning_rate": 4.913567158553634e-05,
      "loss": 0.5454,
      "num_input_tokens_seen": 4390504,
      "step": 6700
    },
    {
      "epoch": 3.5141509433962264,
      "grad_norm": 0.6113634705543518,
      "learning_rate": 4.9132688439582505e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 4393864,
      "step": 6705
    },
    {
      "epoch": 3.5167714884696015,
      "grad_norm": 0.4260737895965576,
      "learning_rate": 4.9129700245345565e-05,
      "loss": 0.4073,
      "num_input_tokens_seen": 4396488,
      "step": 6710
    },
    {
      "epoch": 3.519392033542977,
      "grad_norm": 0.35498955845832825,
      "learning_rate": 4.912670700345063e-05,
      "loss": 0.4347,
      "num_input_tokens_seen": 4399624,
      "step": 6715
    },
    {
      "epoch": 3.5220125786163523,
      "grad_norm": 1.030295968055725,
      "learning_rate": 4.912370871452383e-05,
      "loss": 0.6,
      "num_input_tokens_seen": 4403528,
      "step": 6720
    },
    {
      "epoch": 3.5246331236897275,
      "grad_norm": 0.28840455412864685,
      "learning_rate": 4.91207053791924e-05,
      "loss": 0.3981,
      "num_input_tokens_seen": 4406184,
      "step": 6725
    },
    {
      "epoch": 3.5272536687631026,
      "grad_norm": 0.28302884101867676,
      "learning_rate": 4.911769699808457e-05,
      "loss": 0.4025,
      "num_input_tokens_seen": 4409640,
      "step": 6730
    },
    {
      "epoch": 3.529874213836478,
      "grad_norm": 0.5028561353683472,
      "learning_rate": 4.911468357182969e-05,
      "loss": 0.3487,
      "num_input_tokens_seen": 4412680,
      "step": 6735
    },
    {
      "epoch": 3.532494758909853,
      "grad_norm": 0.3120138347148895,
      "learning_rate": 4.9111665101058116e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 4415880,
      "step": 6740
    },
    {
      "epoch": 3.5351153039832286,
      "grad_norm": 0.5071484446525574,
      "learning_rate": 4.910864158640128e-05,
      "loss": 0.469,
      "num_input_tokens_seen": 4418376,
      "step": 6745
    },
    {
      "epoch": 3.5377358490566038,
      "grad_norm": 0.5017330646514893,
      "learning_rate": 4.910561302849168e-05,
      "loss": 0.4704,
      "num_input_tokens_seen": 4420936,
      "step": 6750
    },
    {
      "epoch": 3.540356394129979,
      "grad_norm": 0.514702320098877,
      "learning_rate": 4.910257942796284e-05,
      "loss": 0.551,
      "num_input_tokens_seen": 4424616,
      "step": 6755
    },
    {
      "epoch": 3.5429769392033545,
      "grad_norm": 0.2509782314300537,
      "learning_rate": 4.909954078544936e-05,
      "loss": 0.4341,
      "num_input_tokens_seen": 4427816,
      "step": 6760
    },
    {
      "epoch": 3.5455974842767297,
      "grad_norm": 0.30926796793937683,
      "learning_rate": 4.909649710158689e-05,
      "loss": 0.3828,
      "num_input_tokens_seen": 4431112,
      "step": 6765
    },
    {
      "epoch": 3.548218029350105,
      "grad_norm": 0.4106766879558563,
      "learning_rate": 4.909344837701214e-05,
      "loss": 0.4969,
      "num_input_tokens_seen": 4434056,
      "step": 6770
    },
    {
      "epoch": 3.55083857442348,
      "grad_norm": 0.8918957114219666,
      "learning_rate": 4.909039461236285e-05,
      "loss": 0.4934,
      "num_input_tokens_seen": 4437480,
      "step": 6775
    },
    {
      "epoch": 3.5534591194968552,
      "grad_norm": 0.2940579354763031,
      "learning_rate": 4.908733580827786e-05,
      "loss": 0.4469,
      "num_input_tokens_seen": 4441224,
      "step": 6780
    },
    {
      "epoch": 3.5560796645702304,
      "grad_norm": 0.40551695227622986,
      "learning_rate": 4.9084271965397014e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 4444200,
      "step": 6785
    },
    {
      "epoch": 3.558700209643606,
      "grad_norm": 0.8418428301811218,
      "learning_rate": 4.908120308436124e-05,
      "loss": 0.7089,
      "num_input_tokens_seen": 4446920,
      "step": 6790
    },
    {
      "epoch": 3.561320754716981,
      "grad_norm": 0.38575029373168945,
      "learning_rate": 4.907812916581251e-05,
      "loss": 0.588,
      "num_input_tokens_seen": 4450536,
      "step": 6795
    },
    {
      "epoch": 3.5639412997903563,
      "grad_norm": 0.48143270611763,
      "learning_rate": 4.907505021039386e-05,
      "loss": 0.4978,
      "num_input_tokens_seen": 4453608,
      "step": 6800
    },
    {
      "epoch": 3.5665618448637315,
      "grad_norm": 0.3921072483062744,
      "learning_rate": 4.907196621874937e-05,
      "loss": 0.4541,
      "num_input_tokens_seen": 4456520,
      "step": 6805
    },
    {
      "epoch": 3.569182389937107,
      "grad_norm": 0.35518908500671387,
      "learning_rate": 4.9068877191524176e-05,
      "loss": 0.6474,
      "num_input_tokens_seen": 4460008,
      "step": 6810
    },
    {
      "epoch": 3.5718029350104823,
      "grad_norm": 0.8520781397819519,
      "learning_rate": 4.906578312936446e-05,
      "loss": 0.3944,
      "num_input_tokens_seen": 4462344,
      "step": 6815
    },
    {
      "epoch": 3.5744234800838575,
      "grad_norm": 0.37088167667388916,
      "learning_rate": 4.906268403291747e-05,
      "loss": 0.6559,
      "num_input_tokens_seen": 4466568,
      "step": 6820
    },
    {
      "epoch": 3.5770440251572326,
      "grad_norm": 0.37923485040664673,
      "learning_rate": 4.905957990283151e-05,
      "loss": 0.4493,
      "num_input_tokens_seen": 4470056,
      "step": 6825
    },
    {
      "epoch": 3.579664570230608,
      "grad_norm": 0.3589740991592407,
      "learning_rate": 4.905647073975592e-05,
      "loss": 0.4586,
      "num_input_tokens_seen": 4472712,
      "step": 6830
    },
    {
      "epoch": 3.582285115303983,
      "grad_norm": 0.4877866804599762,
      "learning_rate": 4.9053356544341095e-05,
      "loss": 0.4659,
      "num_input_tokens_seen": 4476456,
      "step": 6835
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 1.0391474962234497,
      "learning_rate": 4.90502373172385e-05,
      "loss": 0.5755,
      "num_input_tokens_seen": 4479816,
      "step": 6840
    },
    {
      "epoch": 3.5875262054507338,
      "grad_norm": 0.34400853514671326,
      "learning_rate": 4.9047113059100635e-05,
      "loss": 0.4248,
      "num_input_tokens_seen": 4484200,
      "step": 6845
    },
    {
      "epoch": 3.590146750524109,
      "grad_norm": 0.6627772450447083,
      "learning_rate": 4.904398377058106e-05,
      "loss": 0.552,
      "num_input_tokens_seen": 4486984,
      "step": 6850
    },
    {
      "epoch": 3.5927672955974845,
      "grad_norm": 0.3382827043533325,
      "learning_rate": 4.90408494523344e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 4490472,
      "step": 6855
    },
    {
      "epoch": 3.5953878406708597,
      "grad_norm": 0.42324209213256836,
      "learning_rate": 4.90377101050163e-05,
      "loss": 0.4151,
      "num_input_tokens_seen": 4493608,
      "step": 6860
    },
    {
      "epoch": 3.598008385744235,
      "grad_norm": 1.1041864156723022,
      "learning_rate": 4.903456572928348e-05,
      "loss": 0.5234,
      "num_input_tokens_seen": 4496840,
      "step": 6865
    },
    {
      "epoch": 3.60062893081761,
      "grad_norm": 0.4473124146461487,
      "learning_rate": 4.903141632579371e-05,
      "loss": 0.794,
      "num_input_tokens_seen": 4499976,
      "step": 6870
    },
    {
      "epoch": 3.603249475890985,
      "grad_norm": 0.36269181966781616,
      "learning_rate": 4.902826189520581e-05,
      "loss": 0.4241,
      "num_input_tokens_seen": 4503592,
      "step": 6875
    },
    {
      "epoch": 3.6058700209643604,
      "grad_norm": 0.3691270053386688,
      "learning_rate": 4.9025102438179656e-05,
      "loss": 0.3247,
      "num_input_tokens_seen": 4506824,
      "step": 6880
    },
    {
      "epoch": 3.608490566037736,
      "grad_norm": 1.1714503765106201,
      "learning_rate": 4.902193795537616e-05,
      "loss": 0.6839,
      "num_input_tokens_seen": 4509096,
      "step": 6885
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 0.24829207360744476,
      "learning_rate": 4.90187684474573e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 4512136,
      "step": 6890
    },
    {
      "epoch": 3.6137316561844863,
      "grad_norm": 0.3642241358757019,
      "learning_rate": 4.901559391508611e-05,
      "loss": 0.4767,
      "num_input_tokens_seen": 4516392,
      "step": 6895
    },
    {
      "epoch": 3.6163522012578615,
      "grad_norm": 1.2744663953781128,
      "learning_rate": 4.9012414358926644e-05,
      "loss": 0.4867,
      "num_input_tokens_seen": 4518984,
      "step": 6900
    },
    {
      "epoch": 3.618972746331237,
      "grad_norm": 0.5182470083236694,
      "learning_rate": 4.900922977964405e-05,
      "loss": 0.4585,
      "num_input_tokens_seen": 4522664,
      "step": 6905
    },
    {
      "epoch": 3.6215932914046123,
      "grad_norm": 0.3424244821071625,
      "learning_rate": 4.90060401779045e-05,
      "loss": 0.4724,
      "num_input_tokens_seen": 4526152,
      "step": 6910
    },
    {
      "epoch": 3.6242138364779874,
      "grad_norm": 0.40713781118392944,
      "learning_rate": 4.900284555437521e-05,
      "loss": 0.5152,
      "num_input_tokens_seen": 4528904,
      "step": 6915
    },
    {
      "epoch": 3.6268343815513626,
      "grad_norm": 0.7822067737579346,
      "learning_rate": 4.899964590972448e-05,
      "loss": 0.5307,
      "num_input_tokens_seen": 4532424,
      "step": 6920
    },
    {
      "epoch": 3.629454926624738,
      "grad_norm": 0.5053097009658813,
      "learning_rate": 4.899644124462163e-05,
      "loss": 0.3423,
      "num_input_tokens_seen": 4538088,
      "step": 6925
    },
    {
      "epoch": 3.632075471698113,
      "grad_norm": 0.45651763677597046,
      "learning_rate": 4.8993231559737026e-05,
      "loss": 0.512,
      "num_input_tokens_seen": 4541480,
      "step": 6930
    },
    {
      "epoch": 3.6346960167714886,
      "grad_norm": 0.7389808297157288,
      "learning_rate": 4.8990016855742114e-05,
      "loss": 0.4248,
      "num_input_tokens_seen": 4544936,
      "step": 6935
    },
    {
      "epoch": 3.6373165618448637,
      "grad_norm": 0.31321460008621216,
      "learning_rate": 4.898679713330937e-05,
      "loss": 0.4388,
      "num_input_tokens_seen": 4549608,
      "step": 6940
    },
    {
      "epoch": 3.639937106918239,
      "grad_norm": 0.37595582008361816,
      "learning_rate": 4.898357239311232e-05,
      "loss": 0.505,
      "num_input_tokens_seen": 4552168,
      "step": 6945
    },
    {
      "epoch": 3.6425576519916145,
      "grad_norm": 0.5076583623886108,
      "learning_rate": 4.898034263582554e-05,
      "loss": 0.5116,
      "num_input_tokens_seen": 4554536,
      "step": 6950
    },
    {
      "epoch": 3.6451781970649897,
      "grad_norm": 0.5025731921195984,
      "learning_rate": 4.897710786212467e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 4557352,
      "step": 6955
    },
    {
      "epoch": 3.647798742138365,
      "grad_norm": 0.46293893456459045,
      "learning_rate": 4.897386807268638e-05,
      "loss": 0.4441,
      "num_input_tokens_seen": 4560168,
      "step": 6960
    },
    {
      "epoch": 3.65041928721174,
      "grad_norm": 0.3650036156177521,
      "learning_rate": 4.897062326818839e-05,
      "loss": 0.4563,
      "num_input_tokens_seen": 4563336,
      "step": 6965
    },
    {
      "epoch": 3.653039832285115,
      "grad_norm": 0.2967894673347473,
      "learning_rate": 4.8967373449309495e-05,
      "loss": 0.5958,
      "num_input_tokens_seen": 4567528,
      "step": 6970
    },
    {
      "epoch": 3.6556603773584904,
      "grad_norm": 0.36564451456069946,
      "learning_rate": 4.89641186167295e-05,
      "loss": 0.4207,
      "num_input_tokens_seen": 4571176,
      "step": 6975
    },
    {
      "epoch": 3.658280922431866,
      "grad_norm": 0.39675799012184143,
      "learning_rate": 4.8960858771129294e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 4574152,
      "step": 6980
    },
    {
      "epoch": 3.660901467505241,
      "grad_norm": 0.31923791766166687,
      "learning_rate": 4.895759391319079e-05,
      "loss": 0.3295,
      "num_input_tokens_seen": 4577576,
      "step": 6985
    },
    {
      "epoch": 3.6635220125786163,
      "grad_norm": 0.4891569912433624,
      "learning_rate": 4.8954324043596954e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 4580936,
      "step": 6990
    },
    {
      "epoch": 3.6661425576519915,
      "grad_norm": 0.42967307567596436,
      "learning_rate": 4.8951049163031826e-05,
      "loss": 0.3214,
      "num_input_tokens_seen": 4583720,
      "step": 6995
    },
    {
      "epoch": 3.668763102725367,
      "grad_norm": 0.9568672180175781,
      "learning_rate": 4.894776927218045e-05,
      "loss": 0.4651,
      "num_input_tokens_seen": 4586856,
      "step": 7000
    },
    {
      "epoch": 3.6713836477987423,
      "grad_norm": 0.7507651448249817,
      "learning_rate": 4.894448437172896e-05,
      "loss": 0.8325,
      "num_input_tokens_seen": 4589160,
      "step": 7005
    },
    {
      "epoch": 3.6740041928721174,
      "grad_norm": 0.30414146184921265,
      "learning_rate": 4.894119446236452e-05,
      "loss": 0.4232,
      "num_input_tokens_seen": 4592008,
      "step": 7010
    },
    {
      "epoch": 3.6766247379454926,
      "grad_norm": 0.7180719375610352,
      "learning_rate": 4.893789954477532e-05,
      "loss": 0.4937,
      "num_input_tokens_seen": 4595336,
      "step": 7015
    },
    {
      "epoch": 3.6792452830188678,
      "grad_norm": 0.720914900302887,
      "learning_rate": 4.893459961965063e-05,
      "loss": 0.4934,
      "num_input_tokens_seen": 4599016,
      "step": 7020
    },
    {
      "epoch": 3.681865828092243,
      "grad_norm": 0.5125278830528259,
      "learning_rate": 4.8931294687680775e-05,
      "loss": 0.5578,
      "num_input_tokens_seen": 4602888,
      "step": 7025
    },
    {
      "epoch": 3.6844863731656186,
      "grad_norm": 0.5206392407417297,
      "learning_rate": 4.8927984749557086e-05,
      "loss": 0.5178,
      "num_input_tokens_seen": 4606024,
      "step": 7030
    },
    {
      "epoch": 3.6871069182389937,
      "grad_norm": 0.4083264470100403,
      "learning_rate": 4.892466980597198e-05,
      "loss": 0.5202,
      "num_input_tokens_seen": 4609288,
      "step": 7035
    },
    {
      "epoch": 3.689727463312369,
      "grad_norm": 0.3858862817287445,
      "learning_rate": 4.892134985761889e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 4613352,
      "step": 7040
    },
    {
      "epoch": 3.6923480083857445,
      "grad_norm": 0.40481704473495483,
      "learning_rate": 4.891802490519232e-05,
      "loss": 0.5662,
      "num_input_tokens_seen": 4616488,
      "step": 7045
    },
    {
      "epoch": 3.6949685534591197,
      "grad_norm": 0.4776070713996887,
      "learning_rate": 4.891469494938781e-05,
      "loss": 0.5321,
      "num_input_tokens_seen": 4620040,
      "step": 7050
    },
    {
      "epoch": 3.697589098532495,
      "grad_norm": 0.3230946958065033,
      "learning_rate": 4.891135999090195e-05,
      "loss": 0.4075,
      "num_input_tokens_seen": 4623464,
      "step": 7055
    },
    {
      "epoch": 3.70020964360587,
      "grad_norm": 0.8591694235801697,
      "learning_rate": 4.890802003043238e-05,
      "loss": 0.7307,
      "num_input_tokens_seen": 4626472,
      "step": 7060
    },
    {
      "epoch": 3.702830188679245,
      "grad_norm": 0.39093920588493347,
      "learning_rate": 4.890467506867777e-05,
      "loss": 0.4658,
      "num_input_tokens_seen": 4629864,
      "step": 7065
    },
    {
      "epoch": 3.7054507337526204,
      "grad_norm": 0.44425132870674133,
      "learning_rate": 4.890132510633786e-05,
      "loss": 0.5764,
      "num_input_tokens_seen": 4633032,
      "step": 7070
    },
    {
      "epoch": 3.708071278825996,
      "grad_norm": 0.38562512397766113,
      "learning_rate": 4.8897970144113416e-05,
      "loss": 0.564,
      "num_input_tokens_seen": 4636392,
      "step": 7075
    },
    {
      "epoch": 3.710691823899371,
      "grad_norm": 1.2556906938552856,
      "learning_rate": 4.889461018270626e-05,
      "loss": 0.558,
      "num_input_tokens_seen": 4639848,
      "step": 7080
    },
    {
      "epoch": 3.7133123689727463,
      "grad_norm": 0.4324202239513397,
      "learning_rate": 4.889124522281926e-05,
      "loss": 0.5106,
      "num_input_tokens_seen": 4642920,
      "step": 7085
    },
    {
      "epoch": 3.7159329140461215,
      "grad_norm": 0.4902162551879883,
      "learning_rate": 4.888787526515632e-05,
      "loss": 0.436,
      "num_input_tokens_seen": 4646536,
      "step": 7090
    },
    {
      "epoch": 3.718553459119497,
      "grad_norm": 0.533514142036438,
      "learning_rate": 4.8884500310422406e-05,
      "loss": 0.5778,
      "num_input_tokens_seen": 4649416,
      "step": 7095
    },
    {
      "epoch": 3.7211740041928723,
      "grad_norm": 0.3220612406730652,
      "learning_rate": 4.8881120359323516e-05,
      "loss": 0.5149,
      "num_input_tokens_seen": 4653064,
      "step": 7100
    },
    {
      "epoch": 3.7237945492662474,
      "grad_norm": 0.43613147735595703,
      "learning_rate": 4.887773541256669e-05,
      "loss": 0.5828,
      "num_input_tokens_seen": 4655752,
      "step": 7105
    },
    {
      "epoch": 3.7264150943396226,
      "grad_norm": 0.4455145001411438,
      "learning_rate": 4.887434547086004e-05,
      "loss": 0.5141,
      "num_input_tokens_seen": 4658248,
      "step": 7110
    },
    {
      "epoch": 3.7290356394129978,
      "grad_norm": 0.3768244981765747,
      "learning_rate": 4.887095053491269e-05,
      "loss": 0.4812,
      "num_input_tokens_seen": 4662056,
      "step": 7115
    },
    {
      "epoch": 3.731656184486373,
      "grad_norm": 0.4080442190170288,
      "learning_rate": 4.8867550605434806e-05,
      "loss": 0.3957,
      "num_input_tokens_seen": 4664968,
      "step": 7120
    },
    {
      "epoch": 3.7342767295597485,
      "grad_norm": 0.48501646518707275,
      "learning_rate": 4.8864145683137644e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 4667560,
      "step": 7125
    },
    {
      "epoch": 3.7368972746331237,
      "grad_norm": 0.45828965306282043,
      "learning_rate": 4.886073576873346e-05,
      "loss": 0.4298,
      "num_input_tokens_seen": 4669864,
      "step": 7130
    },
    {
      "epoch": 3.739517819706499,
      "grad_norm": 0.5346969962120056,
      "learning_rate": 4.885732086293556e-05,
      "loss": 0.5134,
      "num_input_tokens_seen": 4673128,
      "step": 7135
    },
    {
      "epoch": 3.742138364779874,
      "grad_norm": 0.4611679017543793,
      "learning_rate": 4.8853900966458335e-05,
      "loss": 0.5033,
      "num_input_tokens_seen": 4676968,
      "step": 7140
    },
    {
      "epoch": 3.7447589098532497,
      "grad_norm": 0.8240959048271179,
      "learning_rate": 4.8850476080017146e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 4679624,
      "step": 7145
    },
    {
      "epoch": 3.747379454926625,
      "grad_norm": 0.45044663548469543,
      "learning_rate": 4.884704620432848e-05,
      "loss": 0.4851,
      "num_input_tokens_seen": 4682344,
      "step": 7150
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.34252604842185974,
      "learning_rate": 4.8843611340109794e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 4687048,
      "step": 7155
    },
    {
      "epoch": 3.752620545073375,
      "grad_norm": 0.3837355077266693,
      "learning_rate": 4.884017148807965e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 4689768,
      "step": 7160
    },
    {
      "epoch": 3.7552410901467503,
      "grad_norm": 0.5387489795684814,
      "learning_rate": 4.883672664895761e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 4692936,
      "step": 7165
    },
    {
      "epoch": 3.757861635220126,
      "grad_norm": 0.512242317199707,
      "learning_rate": 4.8833276823464296e-05,
      "loss": 0.4491,
      "num_input_tokens_seen": 4695784,
      "step": 7170
    },
    {
      "epoch": 3.760482180293501,
      "grad_norm": 0.5711973905563354,
      "learning_rate": 4.882982201232138e-05,
      "loss": 0.5788,
      "num_input_tokens_seen": 4698760,
      "step": 7175
    },
    {
      "epoch": 3.7631027253668763,
      "grad_norm": 0.37376829981803894,
      "learning_rate": 4.882636221625156e-05,
      "loss": 0.3961,
      "num_input_tokens_seen": 4701640,
      "step": 7180
    },
    {
      "epoch": 3.7657232704402515,
      "grad_norm": 0.3039923310279846,
      "learning_rate": 4.8822897435978586e-05,
      "loss": 0.507,
      "num_input_tokens_seen": 4704232,
      "step": 7185
    },
    {
      "epoch": 3.768343815513627,
      "grad_norm": 0.4264058470726013,
      "learning_rate": 4.8819427672227264e-05,
      "loss": 0.4968,
      "num_input_tokens_seen": 4707048,
      "step": 7190
    },
    {
      "epoch": 3.7709643605870022,
      "grad_norm": 0.38117656111717224,
      "learning_rate": 4.881595292572341e-05,
      "loss": 0.5596,
      "num_input_tokens_seen": 4710184,
      "step": 7195
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 0.47303307056427,
      "learning_rate": 4.881247319719392e-05,
      "loss": 0.5303,
      "num_input_tokens_seen": 4713000,
      "step": 7200
    },
    {
      "epoch": 3.7762054507337526,
      "grad_norm": 0.40698960423469543,
      "learning_rate": 4.88089884873667e-05,
      "loss": 0.5809,
      "num_input_tokens_seen": 4717224,
      "step": 7205
    },
    {
      "epoch": 3.7788259958071277,
      "grad_norm": 0.6187700033187866,
      "learning_rate": 4.880549879697071e-05,
      "loss": 0.5128,
      "num_input_tokens_seen": 4720488,
      "step": 7210
    },
    {
      "epoch": 3.781446540880503,
      "grad_norm": 0.4074189364910126,
      "learning_rate": 4.8802004126735965e-05,
      "loss": 0.5199,
      "num_input_tokens_seen": 4724104,
      "step": 7215
    },
    {
      "epoch": 3.7840670859538785,
      "grad_norm": 0.2667805254459381,
      "learning_rate": 4.87985044773935e-05,
      "loss": 0.5367,
      "num_input_tokens_seen": 4728104,
      "step": 7220
    },
    {
      "epoch": 3.7866876310272537,
      "grad_norm": 0.5563499331474304,
      "learning_rate": 4.879499984967541e-05,
      "loss": 0.4869,
      "num_input_tokens_seen": 4732008,
      "step": 7225
    },
    {
      "epoch": 3.789308176100629,
      "grad_norm": 0.38859912753105164,
      "learning_rate": 4.8791490244314824e-05,
      "loss": 0.4665,
      "num_input_tokens_seen": 4735528,
      "step": 7230
    },
    {
      "epoch": 3.791928721174004,
      "grad_norm": 0.4477139711380005,
      "learning_rate": 4.8787975662045904e-05,
      "loss": 0.5054,
      "num_input_tokens_seen": 4738728,
      "step": 7235
    },
    {
      "epoch": 3.7945492662473796,
      "grad_norm": 0.3630582094192505,
      "learning_rate": 4.878445610360386e-05,
      "loss": 0.531,
      "num_input_tokens_seen": 4742760,
      "step": 7240
    },
    {
      "epoch": 3.797169811320755,
      "grad_norm": 0.3523898124694824,
      "learning_rate": 4.878093156972494e-05,
      "loss": 0.6186,
      "num_input_tokens_seen": 4745544,
      "step": 7245
    },
    {
      "epoch": 3.79979035639413,
      "grad_norm": 0.4804137945175171,
      "learning_rate": 4.877740206114645e-05,
      "loss": 0.5469,
      "num_input_tokens_seen": 4749096,
      "step": 7250
    },
    {
      "epoch": 3.802410901467505,
      "grad_norm": 0.3945666253566742,
      "learning_rate": 4.8773867578606716e-05,
      "loss": 0.4525,
      "num_input_tokens_seen": 4752168,
      "step": 7255
    },
    {
      "epoch": 3.8050314465408803,
      "grad_norm": 0.6520699262619019,
      "learning_rate": 4.877032812284511e-05,
      "loss": 0.3538,
      "num_input_tokens_seen": 4756424,
      "step": 7260
    },
    {
      "epoch": 3.8076519916142555,
      "grad_norm": 0.41715189814567566,
      "learning_rate": 4.876678369460205e-05,
      "loss": 0.4085,
      "num_input_tokens_seen": 4759560,
      "step": 7265
    },
    {
      "epoch": 3.810272536687631,
      "grad_norm": 0.28690746426582336,
      "learning_rate": 4.8763234294618985e-05,
      "loss": 0.3624,
      "num_input_tokens_seen": 4764392,
      "step": 7270
    },
    {
      "epoch": 3.8128930817610063,
      "grad_norm": 0.505184531211853,
      "learning_rate": 4.875967992363841e-05,
      "loss": 0.5101,
      "num_input_tokens_seen": 4767592,
      "step": 7275
    },
    {
      "epoch": 3.8155136268343814,
      "grad_norm": 0.4512024223804474,
      "learning_rate": 4.875612058240386e-05,
      "loss": 0.4793,
      "num_input_tokens_seen": 4770920,
      "step": 7280
    },
    {
      "epoch": 3.818134171907757,
      "grad_norm": 0.7275902628898621,
      "learning_rate": 4.875255627165991e-05,
      "loss": 0.5974,
      "num_input_tokens_seen": 4773736,
      "step": 7285
    },
    {
      "epoch": 3.8207547169811322,
      "grad_norm": 0.5316482186317444,
      "learning_rate": 4.874898699215217e-05,
      "loss": 0.4769,
      "num_input_tokens_seen": 4777480,
      "step": 7290
    },
    {
      "epoch": 3.8233752620545074,
      "grad_norm": 0.5152127742767334,
      "learning_rate": 4.8745412744627296e-05,
      "loss": 0.6568,
      "num_input_tokens_seen": 4779976,
      "step": 7295
    },
    {
      "epoch": 3.8259958071278826,
      "grad_norm": 0.5392559170722961,
      "learning_rate": 4.874183352983297e-05,
      "loss": 0.3923,
      "num_input_tokens_seen": 4782824,
      "step": 7300
    },
    {
      "epoch": 3.8286163522012577,
      "grad_norm": 0.4805198311805725,
      "learning_rate": 4.873824934851794e-05,
      "loss": 0.4962,
      "num_input_tokens_seen": 4785800,
      "step": 7305
    },
    {
      "epoch": 3.831236897274633,
      "grad_norm": 0.6028704643249512,
      "learning_rate": 4.873466020143196e-05,
      "loss": 0.5588,
      "num_input_tokens_seen": 4789576,
      "step": 7310
    },
    {
      "epoch": 3.8338574423480085,
      "grad_norm": 0.4024425745010376,
      "learning_rate": 4.873106608932585e-05,
      "loss": 0.5782,
      "num_input_tokens_seen": 4792360,
      "step": 7315
    },
    {
      "epoch": 3.8364779874213837,
      "grad_norm": 0.4459058344364166,
      "learning_rate": 4.872746701295145e-05,
      "loss": 0.3633,
      "num_input_tokens_seen": 4795112,
      "step": 7320
    },
    {
      "epoch": 3.839098532494759,
      "grad_norm": 0.40201422572135925,
      "learning_rate": 4.872386297306165e-05,
      "loss": 0.4654,
      "num_input_tokens_seen": 4798248,
      "step": 7325
    },
    {
      "epoch": 3.841719077568134,
      "grad_norm": 0.41136953234672546,
      "learning_rate": 4.872025397041037e-05,
      "loss": 0.3912,
      "num_input_tokens_seen": 4801512,
      "step": 7330
    },
    {
      "epoch": 3.8443396226415096,
      "grad_norm": 0.21009036898612976,
      "learning_rate": 4.871664000575258e-05,
      "loss": 0.4669,
      "num_input_tokens_seen": 4805064,
      "step": 7335
    },
    {
      "epoch": 3.846960167714885,
      "grad_norm": 2.7163102626800537,
      "learning_rate": 4.871302107984427e-05,
      "loss": 0.5169,
      "num_input_tokens_seen": 4807400,
      "step": 7340
    },
    {
      "epoch": 3.84958071278826,
      "grad_norm": 0.8544679284095764,
      "learning_rate": 4.8709397193442474e-05,
      "loss": 0.7067,
      "num_input_tokens_seen": 4810760,
      "step": 7345
    },
    {
      "epoch": 3.852201257861635,
      "grad_norm": 0.3529253304004669,
      "learning_rate": 4.870576834730529e-05,
      "loss": 0.601,
      "num_input_tokens_seen": 4813992,
      "step": 7350
    },
    {
      "epoch": 3.8548218029350103,
      "grad_norm": 0.5177277326583862,
      "learning_rate": 4.8702134542191816e-05,
      "loss": 0.4657,
      "num_input_tokens_seen": 4816968,
      "step": 7355
    },
    {
      "epoch": 3.8574423480083855,
      "grad_norm": 0.7566848993301392,
      "learning_rate": 4.86984957788622e-05,
      "loss": 0.4814,
      "num_input_tokens_seen": 4819304,
      "step": 7360
    },
    {
      "epoch": 3.860062893081761,
      "grad_norm": 0.39007624983787537,
      "learning_rate": 4.869485205807763e-05,
      "loss": 0.4077,
      "num_input_tokens_seen": 4822888,
      "step": 7365
    },
    {
      "epoch": 3.8626834381551363,
      "grad_norm": 0.7168406844139099,
      "learning_rate": 4.869120338060034e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 4825544,
      "step": 7370
    },
    {
      "epoch": 3.8653039832285114,
      "grad_norm": 0.7675294280052185,
      "learning_rate": 4.8687549747193585e-05,
      "loss": 0.6476,
      "num_input_tokens_seen": 4829160,
      "step": 7375
    },
    {
      "epoch": 3.867924528301887,
      "grad_norm": 0.4517735540866852,
      "learning_rate": 4.868389115862166e-05,
      "loss": 0.5409,
      "num_input_tokens_seen": 4831624,
      "step": 7380
    },
    {
      "epoch": 3.870545073375262,
      "grad_norm": 0.24680830538272858,
      "learning_rate": 4.868022761564991e-05,
      "loss": 0.3712,
      "num_input_tokens_seen": 4835336,
      "step": 7385
    },
    {
      "epoch": 3.8731656184486374,
      "grad_norm": 0.3981788456439972,
      "learning_rate": 4.8676559119044696e-05,
      "loss": 0.3944,
      "num_input_tokens_seen": 4838184,
      "step": 7390
    },
    {
      "epoch": 3.8757861635220126,
      "grad_norm": 0.43367433547973633,
      "learning_rate": 4.867288566957344e-05,
      "loss": 0.5607,
      "num_input_tokens_seen": 4840776,
      "step": 7395
    },
    {
      "epoch": 3.8784067085953877,
      "grad_norm": 0.3812592923641205,
      "learning_rate": 4.866920726800457e-05,
      "loss": 0.3704,
      "num_input_tokens_seen": 4844456,
      "step": 7400
    },
    {
      "epoch": 3.881027253668763,
      "grad_norm": 0.5541914701461792,
      "learning_rate": 4.8665523915107576e-05,
      "loss": 0.4018,
      "num_input_tokens_seen": 4849832,
      "step": 7405
    },
    {
      "epoch": 3.8836477987421385,
      "grad_norm": 0.21573881804943085,
      "learning_rate": 4.866183561165296e-05,
      "loss": 0.5571,
      "num_input_tokens_seen": 4852936,
      "step": 7410
    },
    {
      "epoch": 3.8862683438155137,
      "grad_norm": 0.3034406900405884,
      "learning_rate": 4.86581423584123e-05,
      "loss": 0.656,
      "num_input_tokens_seen": 4857256,
      "step": 7415
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.26633119583129883,
      "learning_rate": 4.8654444156158154e-05,
      "loss": 0.5452,
      "num_input_tokens_seen": 4861704,
      "step": 7420
    },
    {
      "epoch": 3.891509433962264,
      "grad_norm": 0.6018309593200684,
      "learning_rate": 4.865074100566415e-05,
      "loss": 0.4728,
      "num_input_tokens_seen": 4865416,
      "step": 7425
    },
    {
      "epoch": 3.8941299790356396,
      "grad_norm": 0.5099804997444153,
      "learning_rate": 4.864703290770495e-05,
      "loss": 0.5552,
      "num_input_tokens_seen": 4868584,
      "step": 7430
    },
    {
      "epoch": 3.896750524109015,
      "grad_norm": 0.3497767746448517,
      "learning_rate": 4.864331986305626e-05,
      "loss": 0.4669,
      "num_input_tokens_seen": 4871944,
      "step": 7435
    },
    {
      "epoch": 3.89937106918239,
      "grad_norm": 0.6981555223464966,
      "learning_rate": 4.8639601872494775e-05,
      "loss": 0.5656,
      "num_input_tokens_seen": 4875208,
      "step": 7440
    },
    {
      "epoch": 3.901991614255765,
      "grad_norm": 0.6921773552894592,
      "learning_rate": 4.8635878936798287e-05,
      "loss": 0.5147,
      "num_input_tokens_seen": 4877608,
      "step": 7445
    },
    {
      "epoch": 3.9046121593291403,
      "grad_norm": 0.6802617311477661,
      "learning_rate": 4.863215105674557e-05,
      "loss": 0.4594,
      "num_input_tokens_seen": 4881480,
      "step": 7450
    },
    {
      "epoch": 3.9072327044025155,
      "grad_norm": 0.5989943742752075,
      "learning_rate": 4.8628418233116457e-05,
      "loss": 0.4973,
      "num_input_tokens_seen": 4883880,
      "step": 7455
    },
    {
      "epoch": 3.909853249475891,
      "grad_norm": 0.46566498279571533,
      "learning_rate": 4.8624680466691824e-05,
      "loss": 0.519,
      "num_input_tokens_seen": 4886920,
      "step": 7460
    },
    {
      "epoch": 3.9124737945492662,
      "grad_norm": 0.5806230902671814,
      "learning_rate": 4.862093775825356e-05,
      "loss": 0.5623,
      "num_input_tokens_seen": 4890088,
      "step": 7465
    },
    {
      "epoch": 3.9150943396226414,
      "grad_norm": 0.3403034508228302,
      "learning_rate": 4.86171901085846e-05,
      "loss": 0.3822,
      "num_input_tokens_seen": 4894056,
      "step": 7470
    },
    {
      "epoch": 3.917714884696017,
      "grad_norm": 0.4291621744632721,
      "learning_rate": 4.861343751846891e-05,
      "loss": 0.4117,
      "num_input_tokens_seen": 4897864,
      "step": 7475
    },
    {
      "epoch": 3.920335429769392,
      "grad_norm": 0.4614846110343933,
      "learning_rate": 4.860967998869148e-05,
      "loss": 0.3769,
      "num_input_tokens_seen": 4901512,
      "step": 7480
    },
    {
      "epoch": 3.9229559748427674,
      "grad_norm": 0.4621559679508209,
      "learning_rate": 4.860591752003836e-05,
      "loss": 0.4017,
      "num_input_tokens_seen": 4904456,
      "step": 7485
    },
    {
      "epoch": 3.9255765199161425,
      "grad_norm": 0.3342978060245514,
      "learning_rate": 4.8602150113296596e-05,
      "loss": 0.3966,
      "num_input_tokens_seen": 4907688,
      "step": 7490
    },
    {
      "epoch": 3.9281970649895177,
      "grad_norm": 0.2682761549949646,
      "learning_rate": 4.8598377769254305e-05,
      "loss": 0.6878,
      "num_input_tokens_seen": 4911336,
      "step": 7495
    },
    {
      "epoch": 3.930817610062893,
      "grad_norm": 0.4274529814720154,
      "learning_rate": 4.85946004887006e-05,
      "loss": 0.382,
      "num_input_tokens_seen": 4914600,
      "step": 7500
    },
    {
      "epoch": 3.9334381551362685,
      "grad_norm": 0.35457754135131836,
      "learning_rate": 4.859081827242566e-05,
      "loss": 0.4949,
      "num_input_tokens_seen": 4918312,
      "step": 7505
    },
    {
      "epoch": 3.9360587002096437,
      "grad_norm": 0.3440766930580139,
      "learning_rate": 4.858703112122068e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 4921320,
      "step": 7510
    },
    {
      "epoch": 3.938679245283019,
      "grad_norm": 0.4170224964618683,
      "learning_rate": 4.8583239035877876e-05,
      "loss": 0.4061,
      "num_input_tokens_seen": 4924168,
      "step": 7515
    },
    {
      "epoch": 3.941299790356394,
      "grad_norm": 0.5280313491821289,
      "learning_rate": 4.8579442017190526e-05,
      "loss": 0.514,
      "num_input_tokens_seen": 4926888,
      "step": 7520
    },
    {
      "epoch": 3.9439203354297696,
      "grad_norm": 0.6266222596168518,
      "learning_rate": 4.8575640065952905e-05,
      "loss": 0.3865,
      "num_input_tokens_seen": 4930536,
      "step": 7525
    },
    {
      "epoch": 3.9465408805031448,
      "grad_norm": 0.43129050731658936,
      "learning_rate": 4.857183318296035e-05,
      "loss": 0.5012,
      "num_input_tokens_seen": 4934120,
      "step": 7530
    },
    {
      "epoch": 3.94916142557652,
      "grad_norm": 0.5617517828941345,
      "learning_rate": 4.856802136900922e-05,
      "loss": 0.5309,
      "num_input_tokens_seen": 4936776,
      "step": 7535
    },
    {
      "epoch": 3.951781970649895,
      "grad_norm": 0.5771308541297913,
      "learning_rate": 4.856420462489689e-05,
      "loss": 0.485,
      "num_input_tokens_seen": 4939688,
      "step": 7540
    },
    {
      "epoch": 3.9544025157232703,
      "grad_norm": 0.8226243853569031,
      "learning_rate": 4.856038295142179e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 4942952,
      "step": 7545
    },
    {
      "epoch": 3.9570230607966455,
      "grad_norm": 0.4038497805595398,
      "learning_rate": 4.855655634938337e-05,
      "loss": 0.3829,
      "num_input_tokens_seen": 4946504,
      "step": 7550
    },
    {
      "epoch": 3.959643605870021,
      "grad_norm": 0.6227050423622131,
      "learning_rate": 4.8552724819582106e-05,
      "loss": 0.387,
      "num_input_tokens_seen": 4949384,
      "step": 7555
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 0.8001956343650818,
      "learning_rate": 4.8548888362819515e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 4953000,
      "step": 7560
    },
    {
      "epoch": 3.9648846960167714,
      "grad_norm": 0.3178911805152893,
      "learning_rate": 4.854504697989814e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 4956392,
      "step": 7565
    },
    {
      "epoch": 3.967505241090147,
      "grad_norm": 0.43556568026542664,
      "learning_rate": 4.8541200671621554e-05,
      "loss": 0.5093,
      "num_input_tokens_seen": 4959656,
      "step": 7570
    },
    {
      "epoch": 3.970125786163522,
      "grad_norm": 0.39286887645721436,
      "learning_rate": 4.8537349438794354e-05,
      "loss": 0.4605,
      "num_input_tokens_seen": 4963144,
      "step": 7575
    },
    {
      "epoch": 3.9727463312368974,
      "grad_norm": 0.5152885317802429,
      "learning_rate": 4.853349328222219e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 4965704,
      "step": 7580
    },
    {
      "epoch": 3.9753668763102725,
      "grad_norm": 0.5043932795524597,
      "learning_rate": 4.8529632202711706e-05,
      "loss": 0.4717,
      "num_input_tokens_seen": 4968584,
      "step": 7585
    },
    {
      "epoch": 3.9779874213836477,
      "grad_norm": 0.4466063380241394,
      "learning_rate": 4.852576620107061e-05,
      "loss": 0.4025,
      "num_input_tokens_seen": 4971624,
      "step": 7590
    },
    {
      "epoch": 3.980607966457023,
      "grad_norm": 0.93424391746521,
      "learning_rate": 4.8521895278107615e-05,
      "loss": 0.5513,
      "num_input_tokens_seen": 4974248,
      "step": 7595
    },
    {
      "epoch": 3.9832285115303985,
      "grad_norm": 0.2545543313026428,
      "learning_rate": 4.8518019434632484e-05,
      "loss": 0.6821,
      "num_input_tokens_seen": 4977480,
      "step": 7600
    },
    {
      "epoch": 3.9858490566037736,
      "grad_norm": 0.4818737208843231,
      "learning_rate": 4.8514138671455995e-05,
      "loss": 0.55,
      "num_input_tokens_seen": 4980616,
      "step": 7605
    },
    {
      "epoch": 3.988469601677149,
      "grad_norm": 0.25341859459877014,
      "learning_rate": 4.851025298938996e-05,
      "loss": 0.5501,
      "num_input_tokens_seen": 4985512,
      "step": 7610
    },
    {
      "epoch": 3.991090146750524,
      "grad_norm": 0.4294357895851135,
      "learning_rate": 4.850636238924722e-05,
      "loss": 0.4743,
      "num_input_tokens_seen": 4988904,
      "step": 7615
    },
    {
      "epoch": 3.9937106918238996,
      "grad_norm": 0.7395448684692383,
      "learning_rate": 4.850246687184163e-05,
      "loss": 0.3923,
      "num_input_tokens_seen": 4993032,
      "step": 7620
    },
    {
      "epoch": 3.9963312368972748,
      "grad_norm": 0.5503888726234436,
      "learning_rate": 4.849856643798812e-05,
      "loss": 0.3297,
      "num_input_tokens_seen": 4995656,
      "step": 7625
    },
    {
      "epoch": 3.99895178197065,
      "grad_norm": 0.4330175518989563,
      "learning_rate": 4.849466108850259e-05,
      "loss": 0.6017,
      "num_input_tokens_seen": 4999112,
      "step": 7630
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.49147623777389526,
      "eval_runtime": 9.6824,
      "eval_samples_per_second": 87.582,
      "eval_steps_per_second": 21.895,
      "num_input_tokens_seen": 4999712,
      "step": 7632
    },
    {
      "epoch": 4.001572327044025,
      "grad_norm": 0.40616950392723083,
      "learning_rate": 4.8490750824202e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 5001888,
      "step": 7635
    },
    {
      "epoch": 4.0041928721174,
      "grad_norm": 0.38544005155563354,
      "learning_rate": 4.8486835645904324e-05,
      "loss": 0.423,
      "num_input_tokens_seen": 5005248,
      "step": 7640
    },
    {
      "epoch": 4.006813417190775,
      "grad_norm": 0.4326156973838806,
      "learning_rate": 4.84829155544286e-05,
      "loss": 0.4808,
      "num_input_tokens_seen": 5007648,
      "step": 7645
    },
    {
      "epoch": 4.009433962264151,
      "grad_norm": 0.6223864555358887,
      "learning_rate": 4.847899055059484e-05,
      "loss": 0.5156,
      "num_input_tokens_seen": 5010368,
      "step": 7650
    },
    {
      "epoch": 4.012054507337526,
      "grad_norm": 0.5911266803741455,
      "learning_rate": 4.847506063522412e-05,
      "loss": 0.4628,
      "num_input_tokens_seen": 5013056,
      "step": 7655
    },
    {
      "epoch": 4.014675052410902,
      "grad_norm": 0.3433648347854614,
      "learning_rate": 4.847112580913854e-05,
      "loss": 0.5603,
      "num_input_tokens_seen": 5016608,
      "step": 7660
    },
    {
      "epoch": 4.017295597484277,
      "grad_norm": 0.5061776041984558,
      "learning_rate": 4.846718607316121e-05,
      "loss": 0.4162,
      "num_input_tokens_seen": 5019456,
      "step": 7665
    },
    {
      "epoch": 4.019916142557652,
      "grad_norm": 0.4944215416908264,
      "learning_rate": 4.846324142811628e-05,
      "loss": 0.4697,
      "num_input_tokens_seen": 5022336,
      "step": 7670
    },
    {
      "epoch": 4.022536687631027,
      "grad_norm": 0.6054646968841553,
      "learning_rate": 4.8459291874828924e-05,
      "loss": 0.4498,
      "num_input_tokens_seen": 5025888,
      "step": 7675
    },
    {
      "epoch": 4.0251572327044025,
      "grad_norm": 0.4684426784515381,
      "learning_rate": 4.845533741412534e-05,
      "loss": 0.4844,
      "num_input_tokens_seen": 5029216,
      "step": 7680
    },
    {
      "epoch": 4.027777777777778,
      "grad_norm": 0.4005807936191559,
      "learning_rate": 4.845137804683277e-05,
      "loss": 0.3047,
      "num_input_tokens_seen": 5033280,
      "step": 7685
    },
    {
      "epoch": 4.030398322851153,
      "grad_norm": 0.6839565634727478,
      "learning_rate": 4.844741377377946e-05,
      "loss": 0.3817,
      "num_input_tokens_seen": 5036224,
      "step": 7690
    },
    {
      "epoch": 4.033018867924528,
      "grad_norm": 0.4291682243347168,
      "learning_rate": 4.844344459579469e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 5039168,
      "step": 7695
    },
    {
      "epoch": 4.035639412997903,
      "grad_norm": 0.3400490880012512,
      "learning_rate": 4.8439470513708766e-05,
      "loss": 0.6435,
      "num_input_tokens_seen": 5042944,
      "step": 7700
    },
    {
      "epoch": 4.038259958071279,
      "grad_norm": 0.42151784896850586,
      "learning_rate": 4.8435491528353026e-05,
      "loss": 0.5414,
      "num_input_tokens_seen": 5046240,
      "step": 7705
    },
    {
      "epoch": 4.040880503144654,
      "grad_norm": 0.40540120005607605,
      "learning_rate": 4.843150764055982e-05,
      "loss": 0.7291,
      "num_input_tokens_seen": 5049728,
      "step": 7710
    },
    {
      "epoch": 4.04350104821803,
      "grad_norm": 0.49297457933425903,
      "learning_rate": 4.842751885116254e-05,
      "loss": 0.4767,
      "num_input_tokens_seen": 5052384,
      "step": 7715
    },
    {
      "epoch": 4.046121593291405,
      "grad_norm": 0.7514108419418335,
      "learning_rate": 4.842352516099559e-05,
      "loss": 0.4917,
      "num_input_tokens_seen": 5055872,
      "step": 7720
    },
    {
      "epoch": 4.04874213836478,
      "grad_norm": 0.3762468695640564,
      "learning_rate": 4.84195265708944e-05,
      "loss": 0.4258,
      "num_input_tokens_seen": 5059328,
      "step": 7725
    },
    {
      "epoch": 4.051362683438155,
      "grad_norm": 1.4032588005065918,
      "learning_rate": 4.841552308169545e-05,
      "loss": 0.4269,
      "num_input_tokens_seen": 5061728,
      "step": 7730
    },
    {
      "epoch": 4.05398322851153,
      "grad_norm": 0.5011178255081177,
      "learning_rate": 4.841151469423619e-05,
      "loss": 0.5054,
      "num_input_tokens_seen": 5064608,
      "step": 7735
    },
    {
      "epoch": 4.056603773584905,
      "grad_norm": 0.4852016568183899,
      "learning_rate": 4.840750140935516e-05,
      "loss": 0.5561,
      "num_input_tokens_seen": 5067840,
      "step": 7740
    },
    {
      "epoch": 4.059224318658281,
      "grad_norm": 0.38454872369766235,
      "learning_rate": 4.840348322789189e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 5071328,
      "step": 7745
    },
    {
      "epoch": 4.061844863731656,
      "grad_norm": 1.3386591672897339,
      "learning_rate": 4.839946015068692e-05,
      "loss": 0.5122,
      "num_input_tokens_seen": 5074208,
      "step": 7750
    },
    {
      "epoch": 4.064465408805032,
      "grad_norm": 0.24183990061283112,
      "learning_rate": 4.839543217858184e-05,
      "loss": 0.5369,
      "num_input_tokens_seen": 5078304,
      "step": 7755
    },
    {
      "epoch": 4.067085953878407,
      "grad_norm": 0.7456451058387756,
      "learning_rate": 4.839139931241926e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 5081856,
      "step": 7760
    },
    {
      "epoch": 4.069706498951782,
      "grad_norm": 0.7524151802062988,
      "learning_rate": 4.838736155304281e-05,
      "loss": 0.5034,
      "num_input_tokens_seen": 5085504,
      "step": 7765
    },
    {
      "epoch": 4.072327044025157,
      "grad_norm": 1.2014920711517334,
      "learning_rate": 4.8383318901297135e-05,
      "loss": 0.3481,
      "num_input_tokens_seen": 5089664,
      "step": 7770
    },
    {
      "epoch": 4.0749475890985325,
      "grad_norm": 0.7149369120597839,
      "learning_rate": 4.837927135802792e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 5092256,
      "step": 7775
    },
    {
      "epoch": 4.077568134171908,
      "grad_norm": 0.5354017019271851,
      "learning_rate": 4.8375218924081856e-05,
      "loss": 0.4605,
      "num_input_tokens_seen": 5095456,
      "step": 7780
    },
    {
      "epoch": 4.080188679245283,
      "grad_norm": 0.39420732855796814,
      "learning_rate": 4.837116160030668e-05,
      "loss": 0.4491,
      "num_input_tokens_seen": 5098432,
      "step": 7785
    },
    {
      "epoch": 4.082809224318658,
      "grad_norm": 0.4769972860813141,
      "learning_rate": 4.8367099387551126e-05,
      "loss": 0.7305,
      "num_input_tokens_seen": 5102528,
      "step": 7790
    },
    {
      "epoch": 4.085429769392033,
      "grad_norm": 0.3098193407058716,
      "learning_rate": 4.8363032286664964e-05,
      "loss": 0.4644,
      "num_input_tokens_seen": 5106080,
      "step": 7795
    },
    {
      "epoch": 4.088050314465409,
      "grad_norm": 2.1114695072174072,
      "learning_rate": 4.835896029849899e-05,
      "loss": 0.4722,
      "num_input_tokens_seen": 5108416,
      "step": 7800
    },
    {
      "epoch": 4.090670859538784,
      "grad_norm": 0.5428920984268188,
      "learning_rate": 4.835488342390502e-05,
      "loss": 0.4,
      "num_input_tokens_seen": 5111584,
      "step": 7805
    },
    {
      "epoch": 4.09329140461216,
      "grad_norm": 0.5198719501495361,
      "learning_rate": 4.835080166373587e-05,
      "loss": 0.509,
      "num_input_tokens_seen": 5114176,
      "step": 7810
    },
    {
      "epoch": 4.095911949685535,
      "grad_norm": 0.44363167881965637,
      "learning_rate": 4.834671501884542e-05,
      "loss": 0.3775,
      "num_input_tokens_seen": 5116768,
      "step": 7815
    },
    {
      "epoch": 4.09853249475891,
      "grad_norm": 0.7974680662155151,
      "learning_rate": 4.8342623490088543e-05,
      "loss": 0.3861,
      "num_input_tokens_seen": 5120992,
      "step": 7820
    },
    {
      "epoch": 4.101153039832285,
      "grad_norm": 0.3965749442577362,
      "learning_rate": 4.833852707832114e-05,
      "loss": 0.5325,
      "num_input_tokens_seen": 5124448,
      "step": 7825
    },
    {
      "epoch": 4.10377358490566,
      "grad_norm": 0.38139188289642334,
      "learning_rate": 4.833442578440013e-05,
      "loss": 0.4954,
      "num_input_tokens_seen": 5127648,
      "step": 7830
    },
    {
      "epoch": 4.106394129979035,
      "grad_norm": 0.3952175974845886,
      "learning_rate": 4.833031960918346e-05,
      "loss": 0.453,
      "num_input_tokens_seen": 5130688,
      "step": 7835
    },
    {
      "epoch": 4.109014675052411,
      "grad_norm": 0.29467570781707764,
      "learning_rate": 4.8326208553530095e-05,
      "loss": 0.6,
      "num_input_tokens_seen": 5133664,
      "step": 7840
    },
    {
      "epoch": 4.111635220125786,
      "grad_norm": 0.3528907895088196,
      "learning_rate": 4.832209261830002e-05,
      "loss": 0.382,
      "num_input_tokens_seen": 5136512,
      "step": 7845
    },
    {
      "epoch": 4.114255765199162,
      "grad_norm": 0.37304583191871643,
      "learning_rate": 4.831797180435423e-05,
      "loss": 0.5394,
      "num_input_tokens_seen": 5140480,
      "step": 7850
    },
    {
      "epoch": 4.116876310272537,
      "grad_norm": 0.5478489995002747,
      "learning_rate": 4.831384611255478e-05,
      "loss": 0.3834,
      "num_input_tokens_seen": 5144704,
      "step": 7855
    },
    {
      "epoch": 4.119496855345912,
      "grad_norm": 0.4633520841598511,
      "learning_rate": 4.8309715543764697e-05,
      "loss": 0.4727,
      "num_input_tokens_seen": 5148256,
      "step": 7860
    },
    {
      "epoch": 4.122117400419287,
      "grad_norm": 0.4599428176879883,
      "learning_rate": 4.8305580098848046e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 5151136,
      "step": 7865
    },
    {
      "epoch": 4.1247379454926625,
      "grad_norm": 0.30573925375938416,
      "learning_rate": 4.830143977866993e-05,
      "loss": 0.6505,
      "num_input_tokens_seen": 5154368,
      "step": 7870
    },
    {
      "epoch": 4.127358490566038,
      "grad_norm": 0.7736834287643433,
      "learning_rate": 4.829729458409645e-05,
      "loss": 0.4876,
      "num_input_tokens_seen": 5157216,
      "step": 7875
    },
    {
      "epoch": 4.129979035639413,
      "grad_norm": 0.3738466799259186,
      "learning_rate": 4.8293144515994725e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 5160608,
      "step": 7880
    },
    {
      "epoch": 4.132599580712788,
      "grad_norm": 0.6304634213447571,
      "learning_rate": 4.828898957523292e-05,
      "loss": 0.5337,
      "num_input_tokens_seen": 5163552,
      "step": 7885
    },
    {
      "epoch": 4.135220125786163,
      "grad_norm": 0.61283940076828,
      "learning_rate": 4.8284829762680183e-05,
      "loss": 0.4849,
      "num_input_tokens_seen": 5166496,
      "step": 7890
    },
    {
      "epoch": 4.137840670859539,
      "grad_norm": 0.5431502461433411,
      "learning_rate": 4.828066507920671e-05,
      "loss": 0.4367,
      "num_input_tokens_seen": 5169088,
      "step": 7895
    },
    {
      "epoch": 4.140461215932914,
      "grad_norm": 0.39270609617233276,
      "learning_rate": 4.82764955256837e-05,
      "loss": 0.3634,
      "num_input_tokens_seen": 5171456,
      "step": 7900
    },
    {
      "epoch": 4.1430817610062896,
      "grad_norm": 0.44587284326553345,
      "learning_rate": 4.827232110298338e-05,
      "loss": 0.4749,
      "num_input_tokens_seen": 5175360,
      "step": 7905
    },
    {
      "epoch": 4.145702306079665,
      "grad_norm": 0.8435508608818054,
      "learning_rate": 4.8268141811978995e-05,
      "loss": 0.4421,
      "num_input_tokens_seen": 5177760,
      "step": 7910
    },
    {
      "epoch": 4.14832285115304,
      "grad_norm": 0.37348005175590515,
      "learning_rate": 4.826395765354479e-05,
      "loss": 0.4332,
      "num_input_tokens_seen": 5180128,
      "step": 7915
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 0.23261027038097382,
      "learning_rate": 4.825976862855606e-05,
      "loss": 0.5881,
      "num_input_tokens_seen": 5183296,
      "step": 7920
    },
    {
      "epoch": 4.15356394129979,
      "grad_norm": 0.6814179420471191,
      "learning_rate": 4.8255574737889094e-05,
      "loss": 0.4829,
      "num_input_tokens_seen": 5185632,
      "step": 7925
    },
    {
      "epoch": 4.156184486373165,
      "grad_norm": 0.38718387484550476,
      "learning_rate": 4.82513759824212e-05,
      "loss": 0.5563,
      "num_input_tokens_seen": 5188640,
      "step": 7930
    },
    {
      "epoch": 4.158805031446541,
      "grad_norm": 0.5090497136116028,
      "learning_rate": 4.824717236303072e-05,
      "loss": 0.4841,
      "num_input_tokens_seen": 5191808,
      "step": 7935
    },
    {
      "epoch": 4.161425576519916,
      "grad_norm": 0.5895650386810303,
      "learning_rate": 4.8242963880597005e-05,
      "loss": 0.576,
      "num_input_tokens_seen": 5194560,
      "step": 7940
    },
    {
      "epoch": 4.164046121593292,
      "grad_norm": 0.48301395773887634,
      "learning_rate": 4.823875053600041e-05,
      "loss": 0.5487,
      "num_input_tokens_seen": 5197472,
      "step": 7945
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 0.44844067096710205,
      "learning_rate": 4.823453233012233e-05,
      "loss": 0.5114,
      "num_input_tokens_seen": 5201152,
      "step": 7950
    },
    {
      "epoch": 4.169287211740042,
      "grad_norm": 1.0509008169174194,
      "learning_rate": 4.8230309263845155e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 5205920,
      "step": 7955
    },
    {
      "epoch": 4.171907756813417,
      "grad_norm": 0.5808376669883728,
      "learning_rate": 4.822608133805231e-05,
      "loss": 0.6803,
      "num_input_tokens_seen": 5208768,
      "step": 7960
    },
    {
      "epoch": 4.1745283018867925,
      "grad_norm": 0.4159703552722931,
      "learning_rate": 4.8221848553628226e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 5212064,
      "step": 7965
    },
    {
      "epoch": 4.177148846960168,
      "grad_norm": 0.3338267505168915,
      "learning_rate": 4.8217610911458346e-05,
      "loss": 0.5741,
      "num_input_tokens_seen": 5215712,
      "step": 7970
    },
    {
      "epoch": 4.179769392033543,
      "grad_norm": 0.5694443583488464,
      "learning_rate": 4.821336841242915e-05,
      "loss": 0.4081,
      "num_input_tokens_seen": 5218944,
      "step": 7975
    },
    {
      "epoch": 4.182389937106918,
      "grad_norm": 0.36977243423461914,
      "learning_rate": 4.820912105742811e-05,
      "loss": 0.5997,
      "num_input_tokens_seen": 5222080,
      "step": 7980
    },
    {
      "epoch": 4.185010482180293,
      "grad_norm": 0.5068930983543396,
      "learning_rate": 4.8204868847343736e-05,
      "loss": 0.4891,
      "num_input_tokens_seen": 5224896,
      "step": 7985
    },
    {
      "epoch": 4.187631027253669,
      "grad_norm": 0.5492900013923645,
      "learning_rate": 4.820061178306553e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 5228160,
      "step": 7990
    },
    {
      "epoch": 4.190251572327044,
      "grad_norm": 0.4816339910030365,
      "learning_rate": 4.819634986548403e-05,
      "loss": 0.41,
      "num_input_tokens_seen": 5231232,
      "step": 7995
    },
    {
      "epoch": 4.1928721174004195,
      "grad_norm": 0.33102163672447205,
      "learning_rate": 4.819208309549077e-05,
      "loss": 0.4451,
      "num_input_tokens_seen": 5234432,
      "step": 8000
    },
    {
      "epoch": 4.195492662473795,
      "grad_norm": 0.6413337588310242,
      "learning_rate": 4.8187811473978316e-05,
      "loss": 0.3083,
      "num_input_tokens_seen": 5237728,
      "step": 8005
    },
    {
      "epoch": 4.19811320754717,
      "grad_norm": 0.8391722440719604,
      "learning_rate": 4.818353500184025e-05,
      "loss": 0.4065,
      "num_input_tokens_seen": 5240480,
      "step": 8010
    },
    {
      "epoch": 4.200733752620545,
      "grad_norm": 0.413167268037796,
      "learning_rate": 4.817925367997115e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 5244384,
      "step": 8015
    },
    {
      "epoch": 4.20335429769392,
      "grad_norm": 0.8130810856819153,
      "learning_rate": 4.8174967509266615e-05,
      "loss": 0.3466,
      "num_input_tokens_seen": 5246752,
      "step": 8020
    },
    {
      "epoch": 4.205974842767295,
      "grad_norm": 0.35898733139038086,
      "learning_rate": 4.817067649062328e-05,
      "loss": 0.5282,
      "num_input_tokens_seen": 5249760,
      "step": 8025
    },
    {
      "epoch": 4.2085953878406706,
      "grad_norm": 0.2907782196998596,
      "learning_rate": 4.816638062493876e-05,
      "loss": 0.4505,
      "num_input_tokens_seen": 5252608,
      "step": 8030
    },
    {
      "epoch": 4.211215932914046,
      "grad_norm": 0.34021705389022827,
      "learning_rate": 4.8162079913111716e-05,
      "loss": 0.435,
      "num_input_tokens_seen": 5256288,
      "step": 8035
    },
    {
      "epoch": 4.213836477987422,
      "grad_norm": 0.6674535870552063,
      "learning_rate": 4.81577743560418e-05,
      "loss": 0.3928,
      "num_input_tokens_seen": 5259840,
      "step": 8040
    },
    {
      "epoch": 4.216457023060797,
      "grad_norm": 0.3202607035636902,
      "learning_rate": 4.8153463954629684e-05,
      "loss": 0.456,
      "num_input_tokens_seen": 5263808,
      "step": 8045
    },
    {
      "epoch": 4.219077568134172,
      "grad_norm": 0.5611643195152283,
      "learning_rate": 4.814914870977706e-05,
      "loss": 0.5285,
      "num_input_tokens_seen": 5266816,
      "step": 8050
    },
    {
      "epoch": 4.221698113207547,
      "grad_norm": 0.48621708154678345,
      "learning_rate": 4.8144828622386627e-05,
      "loss": 0.4553,
      "num_input_tokens_seen": 5270528,
      "step": 8055
    },
    {
      "epoch": 4.2243186582809225,
      "grad_norm": 0.34258410334587097,
      "learning_rate": 4.8140503693362095e-05,
      "loss": 0.4602,
      "num_input_tokens_seen": 5273856,
      "step": 8060
    },
    {
      "epoch": 4.226939203354298,
      "grad_norm": 0.5082047581672668,
      "learning_rate": 4.813617392360818e-05,
      "loss": 0.6101,
      "num_input_tokens_seen": 5276288,
      "step": 8065
    },
    {
      "epoch": 4.229559748427673,
      "grad_norm": 0.36677491664886475,
      "learning_rate": 4.813183931403065e-05,
      "loss": 0.6224,
      "num_input_tokens_seen": 5279808,
      "step": 8070
    },
    {
      "epoch": 4.232180293501048,
      "grad_norm": 0.5795240998268127,
      "learning_rate": 4.8127499865536226e-05,
      "loss": 0.4733,
      "num_input_tokens_seen": 5282528,
      "step": 8075
    },
    {
      "epoch": 4.234800838574423,
      "grad_norm": 0.39742112159729004,
      "learning_rate": 4.812315557903269e-05,
      "loss": 0.4689,
      "num_input_tokens_seen": 5285824,
      "step": 8080
    },
    {
      "epoch": 4.237421383647799,
      "grad_norm": 0.6426422595977783,
      "learning_rate": 4.8118806455428806e-05,
      "loss": 0.4529,
      "num_input_tokens_seen": 5288672,
      "step": 8085
    },
    {
      "epoch": 4.240041928721174,
      "grad_norm": 0.6537584662437439,
      "learning_rate": 4.811445249563437e-05,
      "loss": 0.4808,
      "num_input_tokens_seen": 5291808,
      "step": 8090
    },
    {
      "epoch": 4.2426624737945495,
      "grad_norm": 0.2734171450138092,
      "learning_rate": 4.811009370056017e-05,
      "loss": 0.4395,
      "num_input_tokens_seen": 5295840,
      "step": 8095
    },
    {
      "epoch": 4.245283018867925,
      "grad_norm": 0.6922221183776855,
      "learning_rate": 4.8105730071118016e-05,
      "loss": 0.4488,
      "num_input_tokens_seen": 5298720,
      "step": 8100
    },
    {
      "epoch": 4.2479035639413,
      "grad_norm": 0.8792716860771179,
      "learning_rate": 4.810136160822074e-05,
      "loss": 0.5257,
      "num_input_tokens_seen": 5302048,
      "step": 8105
    },
    {
      "epoch": 4.250524109014675,
      "grad_norm": 0.31749069690704346,
      "learning_rate": 4.8096988312782174e-05,
      "loss": 0.3716,
      "num_input_tokens_seen": 5305568,
      "step": 8110
    },
    {
      "epoch": 4.25314465408805,
      "grad_norm": 0.49795180559158325,
      "learning_rate": 4.809261018571715e-05,
      "loss": 0.4914,
      "num_input_tokens_seen": 5308192,
      "step": 8115
    },
    {
      "epoch": 4.255765199161425,
      "grad_norm": 0.4867281913757324,
      "learning_rate": 4.808822722794153e-05,
      "loss": 0.4164,
      "num_input_tokens_seen": 5311200,
      "step": 8120
    },
    {
      "epoch": 4.2583857442348005,
      "grad_norm": 0.450368732213974,
      "learning_rate": 4.8083839440372194e-05,
      "loss": 0.549,
      "num_input_tokens_seen": 5314560,
      "step": 8125
    },
    {
      "epoch": 4.261006289308176,
      "grad_norm": 0.3627893030643463,
      "learning_rate": 4.8079446823926984e-05,
      "loss": 0.5452,
      "num_input_tokens_seen": 5318240,
      "step": 8130
    },
    {
      "epoch": 4.263626834381552,
      "grad_norm": 0.47902944684028625,
      "learning_rate": 4.8075049379524805e-05,
      "loss": 0.5002,
      "num_input_tokens_seen": 5321376,
      "step": 8135
    },
    {
      "epoch": 4.266247379454927,
      "grad_norm": 0.3038124442100525,
      "learning_rate": 4.807064710808555e-05,
      "loss": 0.4616,
      "num_input_tokens_seen": 5325248,
      "step": 8140
    },
    {
      "epoch": 4.268867924528302,
      "grad_norm": 0.5413493514060974,
      "learning_rate": 4.806624001053013e-05,
      "loss": 0.4851,
      "num_input_tokens_seen": 5328256,
      "step": 8145
    },
    {
      "epoch": 4.271488469601677,
      "grad_norm": 0.6888342499732971,
      "learning_rate": 4.8061828087780455e-05,
      "loss": 0.498,
      "num_input_tokens_seen": 5331392,
      "step": 8150
    },
    {
      "epoch": 4.274109014675052,
      "grad_norm": 0.33616137504577637,
      "learning_rate": 4.805741134075944e-05,
      "loss": 0.4625,
      "num_input_tokens_seen": 5335328,
      "step": 8155
    },
    {
      "epoch": 4.276729559748428,
      "grad_norm": 0.28218740224838257,
      "learning_rate": 4.8052989770391026e-05,
      "loss": 0.4188,
      "num_input_tokens_seen": 5340288,
      "step": 8160
    },
    {
      "epoch": 4.279350104821803,
      "grad_norm": 0.4848461151123047,
      "learning_rate": 4.804856337760016e-05,
      "loss": 0.3605,
      "num_input_tokens_seen": 5343360,
      "step": 8165
    },
    {
      "epoch": 4.281970649895178,
      "grad_norm": 0.5510286688804626,
      "learning_rate": 4.804413216331278e-05,
      "loss": 0.5413,
      "num_input_tokens_seen": 5346112,
      "step": 8170
    },
    {
      "epoch": 4.284591194968553,
      "grad_norm": 0.4445549249649048,
      "learning_rate": 4.8039696128455856e-05,
      "loss": 0.403,
      "num_input_tokens_seen": 5348544,
      "step": 8175
    },
    {
      "epoch": 4.287211740041929,
      "grad_norm": 1.0891895294189453,
      "learning_rate": 4.803525527395736e-05,
      "loss": 0.3782,
      "num_input_tokens_seen": 5351648,
      "step": 8180
    },
    {
      "epoch": 4.289832285115304,
      "grad_norm": 0.5030607581138611,
      "learning_rate": 4.803080960074625e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 5354976,
      "step": 8185
    },
    {
      "epoch": 4.2924528301886795,
      "grad_norm": 0.8014975190162659,
      "learning_rate": 4.8026359109752525e-05,
      "loss": 0.3906,
      "num_input_tokens_seen": 5357984,
      "step": 8190
    },
    {
      "epoch": 4.295073375262055,
      "grad_norm": 0.5354403853416443,
      "learning_rate": 4.8021903801907175e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 5361728,
      "step": 8195
    },
    {
      "epoch": 4.29769392033543,
      "grad_norm": 0.5123254656791687,
      "learning_rate": 4.801744367814219e-05,
      "loss": 0.4844,
      "num_input_tokens_seen": 5364512,
      "step": 8200
    },
    {
      "epoch": 4.300314465408805,
      "grad_norm": 0.7681249380111694,
      "learning_rate": 4.801297873939059e-05,
      "loss": 0.4968,
      "num_input_tokens_seen": 5366656,
      "step": 8205
    },
    {
      "epoch": 4.30293501048218,
      "grad_norm": 0.7299628257751465,
      "learning_rate": 4.8008508986586384e-05,
      "loss": 0.3961,
      "num_input_tokens_seen": 5369440,
      "step": 8210
    },
    {
      "epoch": 4.305555555555555,
      "grad_norm": 0.37472525238990784,
      "learning_rate": 4.800403442066459e-05,
      "loss": 0.4031,
      "num_input_tokens_seen": 5373408,
      "step": 8215
    },
    {
      "epoch": 4.3081761006289305,
      "grad_norm": 0.4489189684391022,
      "learning_rate": 4.799955504256125e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 5376224,
      "step": 8220
    },
    {
      "epoch": 4.310796645702306,
      "grad_norm": 0.4565981924533844,
      "learning_rate": 4.7995070853213376e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 5378720,
      "step": 8225
    },
    {
      "epoch": 4.313417190775682,
      "grad_norm": 0.46504607796669006,
      "learning_rate": 4.799058185355903e-05,
      "loss": 0.3389,
      "num_input_tokens_seen": 5382848,
      "step": 8230
    },
    {
      "epoch": 4.316037735849057,
      "grad_norm": 0.5695677995681763,
      "learning_rate": 4.7986088044537235e-05,
      "loss": 0.5378,
      "num_input_tokens_seen": 5387456,
      "step": 8235
    },
    {
      "epoch": 4.318658280922432,
      "grad_norm": 0.42723774909973145,
      "learning_rate": 4.798158942708807e-05,
      "loss": 0.3948,
      "num_input_tokens_seen": 5390592,
      "step": 8240
    },
    {
      "epoch": 4.321278825995807,
      "grad_norm": 0.4043360948562622,
      "learning_rate": 4.797708600215258e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 5393760,
      "step": 8245
    },
    {
      "epoch": 4.323899371069182,
      "grad_norm": 0.654577374458313,
      "learning_rate": 4.7972577770672845e-05,
      "loss": 0.5762,
      "num_input_tokens_seen": 5396320,
      "step": 8250
    },
    {
      "epoch": 4.326519916142558,
      "grad_norm": 0.3945055603981018,
      "learning_rate": 4.796806473359192e-05,
      "loss": 0.4111,
      "num_input_tokens_seen": 5399360,
      "step": 8255
    },
    {
      "epoch": 4.329140461215933,
      "grad_norm": 0.4039689004421234,
      "learning_rate": 4.796354689185389e-05,
      "loss": 0.4689,
      "num_input_tokens_seen": 5401824,
      "step": 8260
    },
    {
      "epoch": 4.331761006289308,
      "grad_norm": 0.40883007645606995,
      "learning_rate": 4.795902424640383e-05,
      "loss": 0.5494,
      "num_input_tokens_seen": 5405088,
      "step": 8265
    },
    {
      "epoch": 4.334381551362683,
      "grad_norm": 0.42197269201278687,
      "learning_rate": 4.795449679818782e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 5408032,
      "step": 8270
    },
    {
      "epoch": 4.337002096436059,
      "grad_norm": 0.28562700748443604,
      "learning_rate": 4.794996454815297e-05,
      "loss": 0.4704,
      "num_input_tokens_seen": 5411328,
      "step": 8275
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 0.4277392327785492,
      "learning_rate": 4.7945427497247355e-05,
      "loss": 0.4433,
      "num_input_tokens_seen": 5415648,
      "step": 8280
    },
    {
      "epoch": 4.3422431865828095,
      "grad_norm": 0.34394538402557373,
      "learning_rate": 4.7940885646420084e-05,
      "loss": 0.4062,
      "num_input_tokens_seen": 5418784,
      "step": 8285
    },
    {
      "epoch": 4.344863731656185,
      "grad_norm": 0.3970257043838501,
      "learning_rate": 4.7936338996621264e-05,
      "loss": 0.5337,
      "num_input_tokens_seen": 5422272,
      "step": 8290
    },
    {
      "epoch": 4.34748427672956,
      "grad_norm": 0.7159919738769531,
      "learning_rate": 4.7931787548801996e-05,
      "loss": 0.439,
      "num_input_tokens_seen": 5425984,
      "step": 8295
    },
    {
      "epoch": 4.350104821802935,
      "grad_norm": 0.5874181985855103,
      "learning_rate": 4.7927231303914394e-05,
      "loss": 0.6057,
      "num_input_tokens_seen": 5428832,
      "step": 8300
    },
    {
      "epoch": 4.35272536687631,
      "grad_norm": 0.2775159478187561,
      "learning_rate": 4.7922670262911565e-05,
      "loss": 0.4963,
      "num_input_tokens_seen": 5431648,
      "step": 8305
    },
    {
      "epoch": 4.355345911949685,
      "grad_norm": 0.4436167776584625,
      "learning_rate": 4.791810442674764e-05,
      "loss": 0.4484,
      "num_input_tokens_seen": 5435136,
      "step": 8310
    },
    {
      "epoch": 4.3579664570230605,
      "grad_norm": 0.708445131778717,
      "learning_rate": 4.791353379637773e-05,
      "loss": 0.5377,
      "num_input_tokens_seen": 5438368,
      "step": 8315
    },
    {
      "epoch": 4.360587002096436,
      "grad_norm": 0.4513969421386719,
      "learning_rate": 4.790895837275796e-05,
      "loss": 0.3631,
      "num_input_tokens_seen": 5441664,
      "step": 8320
    },
    {
      "epoch": 4.363207547169811,
      "grad_norm": 0.5402042269706726,
      "learning_rate": 4.790437815684545e-05,
      "loss": 0.4926,
      "num_input_tokens_seen": 5444672,
      "step": 8325
    },
    {
      "epoch": 4.365828092243187,
      "grad_norm": 0.45468249917030334,
      "learning_rate": 4.789979314959835e-05,
      "loss": 0.4736,
      "num_input_tokens_seen": 5448032,
      "step": 8330
    },
    {
      "epoch": 4.368448637316562,
      "grad_norm": 0.5018969178199768,
      "learning_rate": 4.789520335197577e-05,
      "loss": 0.4823,
      "num_input_tokens_seen": 5450656,
      "step": 8335
    },
    {
      "epoch": 4.371069182389937,
      "grad_norm": 0.4266291856765747,
      "learning_rate": 4.789060876493785e-05,
      "loss": 0.4612,
      "num_input_tokens_seen": 5453696,
      "step": 8340
    },
    {
      "epoch": 4.373689727463312,
      "grad_norm": 0.5663183331489563,
      "learning_rate": 4.788600938944573e-05,
      "loss": 0.4957,
      "num_input_tokens_seen": 5456288,
      "step": 8345
    },
    {
      "epoch": 4.376310272536688,
      "grad_norm": 0.636272132396698,
      "learning_rate": 4.788140522646155e-05,
      "loss": 0.5271,
      "num_input_tokens_seen": 5459648,
      "step": 8350
    },
    {
      "epoch": 4.378930817610063,
      "grad_norm": 0.6549190282821655,
      "learning_rate": 4.787679627694843e-05,
      "loss": 0.4527,
      "num_input_tokens_seen": 5463040,
      "step": 8355
    },
    {
      "epoch": 4.381551362683438,
      "grad_norm": 0.36554065346717834,
      "learning_rate": 4.787218254187053e-05,
      "loss": 0.4063,
      "num_input_tokens_seen": 5465728,
      "step": 8360
    },
    {
      "epoch": 4.384171907756813,
      "grad_norm": 0.4047152101993561,
      "learning_rate": 4.7867564022192965e-05,
      "loss": 0.5453,
      "num_input_tokens_seen": 5469536,
      "step": 8365
    },
    {
      "epoch": 4.386792452830189,
      "grad_norm": 0.318441778421402,
      "learning_rate": 4.786294071888191e-05,
      "loss": 0.3895,
      "num_input_tokens_seen": 5473024,
      "step": 8370
    },
    {
      "epoch": 4.389412997903564,
      "grad_norm": 0.5812760591506958,
      "learning_rate": 4.785831263290449e-05,
      "loss": 0.4088,
      "num_input_tokens_seen": 5476704,
      "step": 8375
    },
    {
      "epoch": 4.3920335429769395,
      "grad_norm": 0.4980023503303528,
      "learning_rate": 4.7853679765228845e-05,
      "loss": 0.5065,
      "num_input_tokens_seen": 5480160,
      "step": 8380
    },
    {
      "epoch": 4.394654088050315,
      "grad_norm": 0.7315605878829956,
      "learning_rate": 4.7849042116824116e-05,
      "loss": 0.441,
      "num_input_tokens_seen": 5483008,
      "step": 8385
    },
    {
      "epoch": 4.39727463312369,
      "grad_norm": 0.5638500452041626,
      "learning_rate": 4.7844399688660456e-05,
      "loss": 0.5027,
      "num_input_tokens_seen": 5487392,
      "step": 8390
    },
    {
      "epoch": 4.399895178197065,
      "grad_norm": 0.3177517056465149,
      "learning_rate": 4.783975248170901e-05,
      "loss": 0.643,
      "num_input_tokens_seen": 5490432,
      "step": 8395
    },
    {
      "epoch": 4.40251572327044,
      "grad_norm": 0.3619779646396637,
      "learning_rate": 4.783510049694191e-05,
      "loss": 0.5086,
      "num_input_tokens_seen": 5494976,
      "step": 8400
    },
    {
      "epoch": 4.405136268343815,
      "grad_norm": 0.30211567878723145,
      "learning_rate": 4.78304437353323e-05,
      "loss": 0.4388,
      "num_input_tokens_seen": 5498400,
      "step": 8405
    },
    {
      "epoch": 4.4077568134171905,
      "grad_norm": 0.4651917517185211,
      "learning_rate": 4.782578219785433e-05,
      "loss": 0.4226,
      "num_input_tokens_seen": 5502368,
      "step": 8410
    },
    {
      "epoch": 4.410377358490566,
      "grad_norm": 0.3775535523891449,
      "learning_rate": 4.782111588548312e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 5505408,
      "step": 8415
    },
    {
      "epoch": 4.412997903563941,
      "grad_norm": 0.350635826587677,
      "learning_rate": 4.781644479919484e-05,
      "loss": 0.5081,
      "num_input_tokens_seen": 5508480,
      "step": 8420
    },
    {
      "epoch": 4.415618448637317,
      "grad_norm": 0.4475116729736328,
      "learning_rate": 4.7811768939966605e-05,
      "loss": 0.5886,
      "num_input_tokens_seen": 5512128,
      "step": 8425
    },
    {
      "epoch": 4.418238993710692,
      "grad_norm": 0.3793335556983948,
      "learning_rate": 4.780708830877656e-05,
      "loss": 0.456,
      "num_input_tokens_seen": 5515744,
      "step": 8430
    },
    {
      "epoch": 4.420859538784067,
      "grad_norm": 0.6951034069061279,
      "learning_rate": 4.780240290660384e-05,
      "loss": 0.5369,
      "num_input_tokens_seen": 5519616,
      "step": 8435
    },
    {
      "epoch": 4.423480083857442,
      "grad_norm": 0.3662271797657013,
      "learning_rate": 4.7797712734428575e-05,
      "loss": 0.5111,
      "num_input_tokens_seen": 5522400,
      "step": 8440
    },
    {
      "epoch": 4.426100628930818,
      "grad_norm": 0.521782398223877,
      "learning_rate": 4.779301779323189e-05,
      "loss": 0.3658,
      "num_input_tokens_seen": 5524992,
      "step": 8445
    },
    {
      "epoch": 4.428721174004193,
      "grad_norm": 0.2280702441930771,
      "learning_rate": 4.778831808399593e-05,
      "loss": 0.3669,
      "num_input_tokens_seen": 5528256,
      "step": 8450
    },
    {
      "epoch": 4.431341719077568,
      "grad_norm": 0.43551144003868103,
      "learning_rate": 4.77836136077038e-05,
      "loss": 0.42,
      "num_input_tokens_seen": 5531200,
      "step": 8455
    },
    {
      "epoch": 4.433962264150943,
      "grad_norm": 0.4090980589389801,
      "learning_rate": 4.777890436533963e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 5533984,
      "step": 8460
    },
    {
      "epoch": 4.436582809224318,
      "grad_norm": 0.5491631031036377,
      "learning_rate": 4.777419035788855e-05,
      "loss": 0.4843,
      "num_input_tokens_seen": 5537184,
      "step": 8465
    },
    {
      "epoch": 4.439203354297694,
      "grad_norm": 0.30693745613098145,
      "learning_rate": 4.776947158633667e-05,
      "loss": 0.3633,
      "num_input_tokens_seen": 5539776,
      "step": 8470
    },
    {
      "epoch": 4.4418238993710695,
      "grad_norm": 0.47976890206336975,
      "learning_rate": 4.7764748051671095e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 5542816,
      "step": 8475
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.6829315423965454,
      "learning_rate": 4.7760019754879946e-05,
      "loss": 0.4883,
      "num_input_tokens_seen": 5545824,
      "step": 8480
    },
    {
      "epoch": 4.44706498951782,
      "grad_norm": 0.20896834135055542,
      "learning_rate": 4.7755286696952324e-05,
      "loss": 0.4964,
      "num_input_tokens_seen": 5548992,
      "step": 8485
    },
    {
      "epoch": 4.449685534591195,
      "grad_norm": 0.5676760077476501,
      "learning_rate": 4.775054887887833e-05,
      "loss": 0.4521,
      "num_input_tokens_seen": 5552384,
      "step": 8490
    },
    {
      "epoch": 4.45230607966457,
      "grad_norm": 0.48271727561950684,
      "learning_rate": 4.774580630164906e-05,
      "loss": 0.5973,
      "num_input_tokens_seen": 5555744,
      "step": 8495
    },
    {
      "epoch": 4.454926624737945,
      "grad_norm": 0.315462589263916,
      "learning_rate": 4.774105896625661e-05,
      "loss": 0.4478,
      "num_input_tokens_seen": 5559328,
      "step": 8500
    },
    {
      "epoch": 4.4575471698113205,
      "grad_norm": 0.2792726159095764,
      "learning_rate": 4.773630687369407e-05,
      "loss": 0.6213,
      "num_input_tokens_seen": 5563200,
      "step": 8505
    },
    {
      "epoch": 4.460167714884696,
      "grad_norm": 0.5984863042831421,
      "learning_rate": 4.773155002495552e-05,
      "loss": 0.5899,
      "num_input_tokens_seen": 5566752,
      "step": 8510
    },
    {
      "epoch": 4.462788259958071,
      "grad_norm": 0.320135235786438,
      "learning_rate": 4.772678842103605e-05,
      "loss": 0.41,
      "num_input_tokens_seen": 5570176,
      "step": 8515
    },
    {
      "epoch": 4.465408805031447,
      "grad_norm": 0.5526943802833557,
      "learning_rate": 4.772202206293172e-05,
      "loss": 0.4601,
      "num_input_tokens_seen": 5573504,
      "step": 8520
    },
    {
      "epoch": 4.468029350104822,
      "grad_norm": 0.3437677025794983,
      "learning_rate": 4.7717250951639586e-05,
      "loss": 0.3651,
      "num_input_tokens_seen": 5580128,
      "step": 8525
    },
    {
      "epoch": 4.470649895178197,
      "grad_norm": 0.3843805193901062,
      "learning_rate": 4.771247508815774e-05,
      "loss": 0.638,
      "num_input_tokens_seen": 5583680,
      "step": 8530
    },
    {
      "epoch": 4.473270440251572,
      "grad_norm": 0.46952521800994873,
      "learning_rate": 4.770769447348522e-05,
      "loss": 0.4803,
      "num_input_tokens_seen": 5587072,
      "step": 8535
    },
    {
      "epoch": 4.475890985324948,
      "grad_norm": 0.2808758020401001,
      "learning_rate": 4.7702909108622074e-05,
      "loss": 0.49,
      "num_input_tokens_seen": 5591072,
      "step": 8540
    },
    {
      "epoch": 4.478511530398323,
      "grad_norm": 0.36997750401496887,
      "learning_rate": 4.7698118994569355e-05,
      "loss": 0.5203,
      "num_input_tokens_seen": 5594528,
      "step": 8545
    },
    {
      "epoch": 4.481132075471698,
      "grad_norm": 0.9469361305236816,
      "learning_rate": 4.769332413232909e-05,
      "loss": 0.5788,
      "num_input_tokens_seen": 5597184,
      "step": 8550
    },
    {
      "epoch": 4.483752620545073,
      "grad_norm": 0.43440380692481995,
      "learning_rate": 4.768852452290432e-05,
      "loss": 0.5455,
      "num_input_tokens_seen": 5600512,
      "step": 8555
    },
    {
      "epoch": 4.486373165618448,
      "grad_norm": 0.6556311249732971,
      "learning_rate": 4.768372016729907e-05,
      "loss": 0.5564,
      "num_input_tokens_seen": 5603072,
      "step": 8560
    },
    {
      "epoch": 4.488993710691824,
      "grad_norm": 0.2846789062023163,
      "learning_rate": 4.767891106651834e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 5607040,
      "step": 8565
    },
    {
      "epoch": 4.4916142557651995,
      "grad_norm": 0.4658925235271454,
      "learning_rate": 4.767409722156816e-05,
      "loss": 0.5639,
      "num_input_tokens_seen": 5611136,
      "step": 8570
    },
    {
      "epoch": 4.494234800838575,
      "grad_norm": 0.7394786477088928,
      "learning_rate": 4.7669278633455515e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 5613504,
      "step": 8575
    },
    {
      "epoch": 4.49685534591195,
      "grad_norm": 0.3272421956062317,
      "learning_rate": 4.7664455303188404e-05,
      "loss": 0.5622,
      "num_input_tokens_seen": 5617056,
      "step": 8580
    },
    {
      "epoch": 4.499475890985325,
      "grad_norm": 0.5386461019515991,
      "learning_rate": 4.7659627231775815e-05,
      "loss": 0.5421,
      "num_input_tokens_seen": 5620800,
      "step": 8585
    },
    {
      "epoch": 4.5020964360587,
      "grad_norm": 0.27534791827201843,
      "learning_rate": 4.765479442022773e-05,
      "loss": 0.4932,
      "num_input_tokens_seen": 5624864,
      "step": 8590
    },
    {
      "epoch": 4.504716981132075,
      "grad_norm": 0.9905965924263,
      "learning_rate": 4.76499568695551e-05,
      "loss": 0.6423,
      "num_input_tokens_seen": 5627552,
      "step": 8595
    },
    {
      "epoch": 4.5073375262054505,
      "grad_norm": 0.6363106369972229,
      "learning_rate": 4.764511458076991e-05,
      "loss": 0.5327,
      "num_input_tokens_seen": 5629888,
      "step": 8600
    },
    {
      "epoch": 4.509958071278826,
      "grad_norm": 0.4832497835159302,
      "learning_rate": 4.7640267554885085e-05,
      "loss": 0.4661,
      "num_input_tokens_seen": 5632512,
      "step": 8605
    },
    {
      "epoch": 4.512578616352201,
      "grad_norm": 0.6417698264122009,
      "learning_rate": 4.763541579291459e-05,
      "loss": 0.6068,
      "num_input_tokens_seen": 5638880,
      "step": 8610
    },
    {
      "epoch": 4.515199161425577,
      "grad_norm": 0.6216455101966858,
      "learning_rate": 4.763055929587336e-05,
      "loss": 0.3929,
      "num_input_tokens_seen": 5641856,
      "step": 8615
    },
    {
      "epoch": 4.517819706498952,
      "grad_norm": 0.6123128533363342,
      "learning_rate": 4.7625698064777294e-05,
      "loss": 0.5556,
      "num_input_tokens_seen": 5644256,
      "step": 8620
    },
    {
      "epoch": 4.520440251572327,
      "grad_norm": 0.7330706119537354,
      "learning_rate": 4.762083210064332e-05,
      "loss": 0.4269,
      "num_input_tokens_seen": 5647008,
      "step": 8625
    },
    {
      "epoch": 4.523060796645702,
      "grad_norm": 0.3982251286506653,
      "learning_rate": 4.761596140448934e-05,
      "loss": 0.5954,
      "num_input_tokens_seen": 5650208,
      "step": 8630
    },
    {
      "epoch": 4.5256813417190775,
      "grad_norm": 0.7982743382453918,
      "learning_rate": 4.7611085977334256e-05,
      "loss": 0.4871,
      "num_input_tokens_seen": 5652640,
      "step": 8635
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 0.46379947662353516,
      "learning_rate": 4.7606205820197946e-05,
      "loss": 0.5819,
      "num_input_tokens_seen": 5656896,
      "step": 8640
    },
    {
      "epoch": 4.530922431865828,
      "grad_norm": 0.5177033543586731,
      "learning_rate": 4.760132093410129e-05,
      "loss": 0.3606,
      "num_input_tokens_seen": 5660576,
      "step": 8645
    },
    {
      "epoch": 4.533542976939203,
      "grad_norm": 0.3164561092853546,
      "learning_rate": 4.759643132006614e-05,
      "loss": 0.455,
      "num_input_tokens_seen": 5663456,
      "step": 8650
    },
    {
      "epoch": 4.536163522012579,
      "grad_norm": 0.39617919921875,
      "learning_rate": 4.759153697911535e-05,
      "loss": 0.2826,
      "num_input_tokens_seen": 5666752,
      "step": 8655
    },
    {
      "epoch": 4.538784067085954,
      "grad_norm": 0.6081299185752869,
      "learning_rate": 4.758663791227276e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 5670272,
      "step": 8660
    },
    {
      "epoch": 4.5414046121593294,
      "grad_norm": 0.3984132707118988,
      "learning_rate": 4.7581734120563206e-05,
      "loss": 0.6951,
      "num_input_tokens_seen": 5673120,
      "step": 8665
    },
    {
      "epoch": 4.544025157232705,
      "grad_norm": 0.5069027543067932,
      "learning_rate": 4.757682560501249e-05,
      "loss": 0.3782,
      "num_input_tokens_seen": 5676032,
      "step": 8670
    },
    {
      "epoch": 4.54664570230608,
      "grad_norm": 0.5793702006340027,
      "learning_rate": 4.757191236664744e-05,
      "loss": 0.4016,
      "num_input_tokens_seen": 5678784,
      "step": 8675
    },
    {
      "epoch": 4.549266247379455,
      "grad_norm": 0.5102466344833374,
      "learning_rate": 4.7566994406495825e-05,
      "loss": 0.489,
      "num_input_tokens_seen": 5682976,
      "step": 8680
    },
    {
      "epoch": 4.55188679245283,
      "grad_norm": 0.27285388112068176,
      "learning_rate": 4.756207172558645e-05,
      "loss": 0.4898,
      "num_input_tokens_seen": 5686752,
      "step": 8685
    },
    {
      "epoch": 4.554507337526205,
      "grad_norm": 0.29896458983421326,
      "learning_rate": 4.7557144324949065e-05,
      "loss": 0.3877,
      "num_input_tokens_seen": 5690432,
      "step": 8690
    },
    {
      "epoch": 4.5571278825995805,
      "grad_norm": 0.5385617017745972,
      "learning_rate": 4.755221220561443e-05,
      "loss": 0.5066,
      "num_input_tokens_seen": 5694112,
      "step": 8695
    },
    {
      "epoch": 4.559748427672956,
      "grad_norm": 0.3351892828941345,
      "learning_rate": 4.7547275368614295e-05,
      "loss": 0.5375,
      "num_input_tokens_seen": 5696800,
      "step": 8700
    },
    {
      "epoch": 4.562368972746331,
      "grad_norm": 0.594815731048584,
      "learning_rate": 4.754233381498139e-05,
      "loss": 0.4026,
      "num_input_tokens_seen": 5699328,
      "step": 8705
    },
    {
      "epoch": 4.564989517819707,
      "grad_norm": 0.42388206720352173,
      "learning_rate": 4.753738754574942e-05,
      "loss": 0.4423,
      "num_input_tokens_seen": 5702400,
      "step": 8710
    },
    {
      "epoch": 4.567610062893082,
      "grad_norm": 0.4552631974220276,
      "learning_rate": 4.75324365619531e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 5706496,
      "step": 8715
    },
    {
      "epoch": 4.570230607966457,
      "grad_norm": 0.7494230270385742,
      "learning_rate": 4.752748086462811e-05,
      "loss": 0.5596,
      "num_input_tokens_seen": 5711680,
      "step": 8720
    },
    {
      "epoch": 4.572851153039832,
      "grad_norm": 0.6084966659545898,
      "learning_rate": 4.752252045481113e-05,
      "loss": 0.5871,
      "num_input_tokens_seen": 5714560,
      "step": 8725
    },
    {
      "epoch": 4.5754716981132075,
      "grad_norm": 0.3746282458305359,
      "learning_rate": 4.7517555333539834e-05,
      "loss": 0.3819,
      "num_input_tokens_seen": 5717632,
      "step": 8730
    },
    {
      "epoch": 4.578092243186583,
      "grad_norm": 0.32860422134399414,
      "learning_rate": 4.751258550185284e-05,
      "loss": 0.3843,
      "num_input_tokens_seen": 5720928,
      "step": 8735
    },
    {
      "epoch": 4.580712788259958,
      "grad_norm": 0.4548417925834656,
      "learning_rate": 4.7507610960789804e-05,
      "loss": 0.4551,
      "num_input_tokens_seen": 5723840,
      "step": 8740
    },
    {
      "epoch": 4.583333333333333,
      "grad_norm": 0.3788869082927704,
      "learning_rate": 4.7502631711391334e-05,
      "loss": 0.372,
      "num_input_tokens_seen": 5729408,
      "step": 8745
    },
    {
      "epoch": 4.585953878406709,
      "grad_norm": 0.38195109367370605,
      "learning_rate": 4.749764775469904e-05,
      "loss": 0.4862,
      "num_input_tokens_seen": 5731840,
      "step": 8750
    },
    {
      "epoch": 4.588574423480084,
      "grad_norm": 0.8363075852394104,
      "learning_rate": 4.749265909175549e-05,
      "loss": 0.3561,
      "num_input_tokens_seen": 5739008,
      "step": 8755
    },
    {
      "epoch": 4.591194968553459,
      "grad_norm": 0.3910010755062103,
      "learning_rate": 4.7487665723604275e-05,
      "loss": 0.4107,
      "num_input_tokens_seen": 5742784,
      "step": 8760
    },
    {
      "epoch": 4.593815513626835,
      "grad_norm": 0.34057751297950745,
      "learning_rate": 4.7482667651289944e-05,
      "loss": 0.4528,
      "num_input_tokens_seen": 5745184,
      "step": 8765
    },
    {
      "epoch": 4.59643605870021,
      "grad_norm": 0.33328741788864136,
      "learning_rate": 4.747766487585803e-05,
      "loss": 0.3753,
      "num_input_tokens_seen": 5748256,
      "step": 8770
    },
    {
      "epoch": 4.599056603773585,
      "grad_norm": 0.3754218816757202,
      "learning_rate": 4.747265739835507e-05,
      "loss": 0.5486,
      "num_input_tokens_seen": 5752032,
      "step": 8775
    },
    {
      "epoch": 4.60167714884696,
      "grad_norm": 0.6031083464622498,
      "learning_rate": 4.746764521982856e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 5754592,
      "step": 8780
    },
    {
      "epoch": 4.604297693920335,
      "grad_norm": 0.27590301632881165,
      "learning_rate": 4.7462628341326995e-05,
      "loss": 0.5541,
      "num_input_tokens_seen": 5757184,
      "step": 8785
    },
    {
      "epoch": 4.6069182389937104,
      "grad_norm": 0.5131482481956482,
      "learning_rate": 4.7457606763899856e-05,
      "loss": 0.4212,
      "num_input_tokens_seen": 5761024,
      "step": 8790
    },
    {
      "epoch": 4.609538784067086,
      "grad_norm": 0.8091534376144409,
      "learning_rate": 4.745258048859758e-05,
      "loss": 0.5034,
      "num_input_tokens_seen": 5763776,
      "step": 8795
    },
    {
      "epoch": 4.612159329140461,
      "grad_norm": 0.5582616925239563,
      "learning_rate": 4.744754951647162e-05,
      "loss": 0.4051,
      "num_input_tokens_seen": 5766336,
      "step": 8800
    },
    {
      "epoch": 4.614779874213837,
      "grad_norm": 0.3154968321323395,
      "learning_rate": 4.74425138485744e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 5769184,
      "step": 8805
    },
    {
      "epoch": 4.617400419287212,
      "grad_norm": 0.2862713932991028,
      "learning_rate": 4.743747348595932e-05,
      "loss": 0.4548,
      "num_input_tokens_seen": 5772416,
      "step": 8810
    },
    {
      "epoch": 4.620020964360587,
      "grad_norm": 0.449144184589386,
      "learning_rate": 4.743242842968077e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 5776192,
      "step": 8815
    },
    {
      "epoch": 4.622641509433962,
      "grad_norm": 0.32955116033554077,
      "learning_rate": 4.742737868079411e-05,
      "loss": 0.4568,
      "num_input_tokens_seen": 5780160,
      "step": 8820
    },
    {
      "epoch": 4.6252620545073375,
      "grad_norm": 0.5325759649276733,
      "learning_rate": 4.7422324240355696e-05,
      "loss": 0.4463,
      "num_input_tokens_seen": 5783040,
      "step": 8825
    },
    {
      "epoch": 4.627882599580713,
      "grad_norm": 1.89590585231781,
      "learning_rate": 4.741726510942286e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 5785952,
      "step": 8830
    },
    {
      "epoch": 4.630503144654088,
      "grad_norm": 0.3759578466415405,
      "learning_rate": 4.741220128905391e-05,
      "loss": 0.4484,
      "num_input_tokens_seen": 5789536,
      "step": 8835
    },
    {
      "epoch": 4.633123689727463,
      "grad_norm": 0.8780340552330017,
      "learning_rate": 4.740713278030815e-05,
      "loss": 0.4596,
      "num_input_tokens_seen": 5792672,
      "step": 8840
    },
    {
      "epoch": 4.635744234800838,
      "grad_norm": 0.1752854734659195,
      "learning_rate": 4.740205958424584e-05,
      "loss": 0.4588,
      "num_input_tokens_seen": 5797152,
      "step": 8845
    },
    {
      "epoch": 4.638364779874214,
      "grad_norm": 0.30922433733940125,
      "learning_rate": 4.739698170192824e-05,
      "loss": 0.4297,
      "num_input_tokens_seen": 5800032,
      "step": 8850
    },
    {
      "epoch": 4.640985324947589,
      "grad_norm": 0.3998490273952484,
      "learning_rate": 4.7391899134417596e-05,
      "loss": 0.5291,
      "num_input_tokens_seen": 5802848,
      "step": 8855
    },
    {
      "epoch": 4.643605870020965,
      "grad_norm": 0.7860298156738281,
      "learning_rate": 4.73868118827771e-05,
      "loss": 0.6255,
      "num_input_tokens_seen": 5805472,
      "step": 8860
    },
    {
      "epoch": 4.64622641509434,
      "grad_norm": 0.44213685393333435,
      "learning_rate": 4.738171994807097e-05,
      "loss": 0.4109,
      "num_input_tokens_seen": 5808512,
      "step": 8865
    },
    {
      "epoch": 4.648846960167715,
      "grad_norm": 0.5750396847724915,
      "learning_rate": 4.7376623331364364e-05,
      "loss": 0.3913,
      "num_input_tokens_seen": 5811488,
      "step": 8870
    },
    {
      "epoch": 4.65146750524109,
      "grad_norm": 0.8556397557258606,
      "learning_rate": 4.737152203372345e-05,
      "loss": 0.5691,
      "num_input_tokens_seen": 5814208,
      "step": 8875
    },
    {
      "epoch": 4.654088050314465,
      "grad_norm": 0.5511450171470642,
      "learning_rate": 4.7366416056215345e-05,
      "loss": 0.4753,
      "num_input_tokens_seen": 5816992,
      "step": 8880
    },
    {
      "epoch": 4.65670859538784,
      "grad_norm": 0.49786821007728577,
      "learning_rate": 4.7361305399908185e-05,
      "loss": 0.5026,
      "num_input_tokens_seen": 5819840,
      "step": 8885
    },
    {
      "epoch": 4.659329140461216,
      "grad_norm": 0.41653141379356384,
      "learning_rate": 4.735619006587103e-05,
      "loss": 0.497,
      "num_input_tokens_seen": 5822880,
      "step": 8890
    },
    {
      "epoch": 4.661949685534591,
      "grad_norm": 0.42679131031036377,
      "learning_rate": 4.735107005517398e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 5826080,
      "step": 8895
    },
    {
      "epoch": 4.664570230607967,
      "grad_norm": 0.612026035785675,
      "learning_rate": 4.734594536888806e-05,
      "loss": 0.8862,
      "num_input_tokens_seen": 5829280,
      "step": 8900
    },
    {
      "epoch": 4.667190775681342,
      "grad_norm": 0.3579176366329193,
      "learning_rate": 4.734081600808531e-05,
      "loss": 0.622,
      "num_input_tokens_seen": 5832320,
      "step": 8905
    },
    {
      "epoch": 4.669811320754717,
      "grad_norm": 0.3822415769100189,
      "learning_rate": 4.733568197383872e-05,
      "loss": 0.4014,
      "num_input_tokens_seen": 5835488,
      "step": 8910
    },
    {
      "epoch": 4.672431865828092,
      "grad_norm": 0.335801899433136,
      "learning_rate": 4.7330543267222294e-05,
      "loss": 0.3566,
      "num_input_tokens_seen": 5838208,
      "step": 8915
    },
    {
      "epoch": 4.6750524109014675,
      "grad_norm": 0.38022580742836,
      "learning_rate": 4.732539988931097e-05,
      "loss": 0.4201,
      "num_input_tokens_seen": 5841792,
      "step": 8920
    },
    {
      "epoch": 4.677672955974843,
      "grad_norm": 0.5795624256134033,
      "learning_rate": 4.732025184118068e-05,
      "loss": 0.5124,
      "num_input_tokens_seen": 5845344,
      "step": 8925
    },
    {
      "epoch": 4.680293501048218,
      "grad_norm": 0.4303871989250183,
      "learning_rate": 4.731509912390837e-05,
      "loss": 0.6169,
      "num_input_tokens_seen": 5847712,
      "step": 8930
    },
    {
      "epoch": 4.682914046121593,
      "grad_norm": 0.3809991478919983,
      "learning_rate": 4.730994173857189e-05,
      "loss": 0.4951,
      "num_input_tokens_seen": 5850720,
      "step": 8935
    },
    {
      "epoch": 4.685534591194968,
      "grad_norm": 0.4866112172603607,
      "learning_rate": 4.7304779686250126e-05,
      "loss": 0.6116,
      "num_input_tokens_seen": 5853760,
      "step": 8940
    },
    {
      "epoch": 4.688155136268344,
      "grad_norm": 0.3952195346355438,
      "learning_rate": 4.729961296802292e-05,
      "loss": 0.6027,
      "num_input_tokens_seen": 5857856,
      "step": 8945
    },
    {
      "epoch": 4.690775681341719,
      "grad_norm": 0.5593749284744263,
      "learning_rate": 4.7294441584971095e-05,
      "loss": 0.6514,
      "num_input_tokens_seen": 5860800,
      "step": 8950
    },
    {
      "epoch": 4.693396226415095,
      "grad_norm": 0.3308628797531128,
      "learning_rate": 4.728926553817643e-05,
      "loss": 0.4181,
      "num_input_tokens_seen": 5865184,
      "step": 8955
    },
    {
      "epoch": 4.69601677148847,
      "grad_norm": 0.6932462453842163,
      "learning_rate": 4.728408482872171e-05,
      "loss": 0.4642,
      "num_input_tokens_seen": 5868704,
      "step": 8960
    },
    {
      "epoch": 4.698637316561845,
      "grad_norm": 0.3071248531341553,
      "learning_rate": 4.727889945769067e-05,
      "loss": 0.5308,
      "num_input_tokens_seen": 5872896,
      "step": 8965
    },
    {
      "epoch": 4.70125786163522,
      "grad_norm": 0.6125551462173462,
      "learning_rate": 4.727370942616803e-05,
      "loss": 0.5353,
      "num_input_tokens_seen": 5875776,
      "step": 8970
    },
    {
      "epoch": 4.703878406708595,
      "grad_norm": 0.3618026077747345,
      "learning_rate": 4.72685147352395e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 5878656,
      "step": 8975
    },
    {
      "epoch": 4.70649895178197,
      "grad_norm": 0.4485103189945221,
      "learning_rate": 4.7263315385991735e-05,
      "loss": 0.5691,
      "num_input_tokens_seen": 5882400,
      "step": 8980
    },
    {
      "epoch": 4.709119496855346,
      "grad_norm": 0.5830134153366089,
      "learning_rate": 4.7258111379512384e-05,
      "loss": 0.5547,
      "num_input_tokens_seen": 5885664,
      "step": 8985
    },
    {
      "epoch": 4.711740041928721,
      "grad_norm": 0.5755674242973328,
      "learning_rate": 4.725290271689007e-05,
      "loss": 0.4197,
      "num_input_tokens_seen": 5888576,
      "step": 8990
    },
    {
      "epoch": 4.714360587002097,
      "grad_norm": 0.5222364068031311,
      "learning_rate": 4.724768939921438e-05,
      "loss": 0.4331,
      "num_input_tokens_seen": 5891552,
      "step": 8995
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 0.4661974012851715,
      "learning_rate": 4.7242471427575884e-05,
      "loss": 0.584,
      "num_input_tokens_seen": 5894560,
      "step": 9000
    },
    {
      "epoch": 4.719601677148847,
      "grad_norm": 0.35963475704193115,
      "learning_rate": 4.723724880306612e-05,
      "loss": 0.4634,
      "num_input_tokens_seen": 5897088,
      "step": 9005
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 0.3821851313114166,
      "learning_rate": 4.72320215267776e-05,
      "loss": 0.464,
      "num_input_tokens_seen": 5901664,
      "step": 9010
    },
    {
      "epoch": 4.7248427672955975,
      "grad_norm": 0.43390583992004395,
      "learning_rate": 4.722678959980381e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 5905472,
      "step": 9015
    },
    {
      "epoch": 4.727463312368973,
      "grad_norm": 0.4731997549533844,
      "learning_rate": 4.7221553023239216e-05,
      "loss": 0.3593,
      "num_input_tokens_seen": 5908416,
      "step": 9020
    },
    {
      "epoch": 4.730083857442348,
      "grad_norm": 0.32073327898979187,
      "learning_rate": 4.721631179817925e-05,
      "loss": 0.4346,
      "num_input_tokens_seen": 5911552,
      "step": 9025
    },
    {
      "epoch": 4.732704402515723,
      "grad_norm": 0.5166798830032349,
      "learning_rate": 4.721106592572031e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 5914752,
      "step": 9030
    },
    {
      "epoch": 4.735324947589098,
      "grad_norm": 0.4229084253311157,
      "learning_rate": 4.7205815406959774e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 5917312,
      "step": 9035
    },
    {
      "epoch": 4.737945492662474,
      "grad_norm": 0.29460814595222473,
      "learning_rate": 4.7200560242996e-05,
      "loss": 0.5094,
      "num_input_tokens_seen": 5921088,
      "step": 9040
    },
    {
      "epoch": 4.740566037735849,
      "grad_norm": 0.5178367495536804,
      "learning_rate": 4.719530043492829e-05,
      "loss": 0.5242,
      "num_input_tokens_seen": 5924032,
      "step": 9045
    },
    {
      "epoch": 4.743186582809225,
      "grad_norm": 0.5424672365188599,
      "learning_rate": 4.7190035983856954e-05,
      "loss": 0.4438,
      "num_input_tokens_seen": 5928192,
      "step": 9050
    },
    {
      "epoch": 4.7458071278826,
      "grad_norm": 0.6380351781845093,
      "learning_rate": 4.718476689088325e-05,
      "loss": 0.4818,
      "num_input_tokens_seen": 5931200,
      "step": 9055
    },
    {
      "epoch": 4.748427672955975,
      "grad_norm": 0.4152025282382965,
      "learning_rate": 4.71794931571094e-05,
      "loss": 0.5244,
      "num_input_tokens_seen": 5934944,
      "step": 9060
    },
    {
      "epoch": 4.75104821802935,
      "grad_norm": 0.5168023109436035,
      "learning_rate": 4.717421478363862e-05,
      "loss": 0.3934,
      "num_input_tokens_seen": 5938688,
      "step": 9065
    },
    {
      "epoch": 4.753668763102725,
      "grad_norm": 0.46114835143089294,
      "learning_rate": 4.7168931771575085e-05,
      "loss": 0.439,
      "num_input_tokens_seen": 5941248,
      "step": 9070
    },
    {
      "epoch": 4.7562893081761,
      "grad_norm": 0.47314849495887756,
      "learning_rate": 4.716364412202394e-05,
      "loss": 0.548,
      "num_input_tokens_seen": 5944064,
      "step": 9075
    },
    {
      "epoch": 4.758909853249476,
      "grad_norm": 0.4347383677959442,
      "learning_rate": 4.7158351836091306e-05,
      "loss": 0.4487,
      "num_input_tokens_seen": 5947232,
      "step": 9080
    },
    {
      "epoch": 4.761530398322851,
      "grad_norm": 0.4169091582298279,
      "learning_rate": 4.7153054914884266e-05,
      "loss": 0.6122,
      "num_input_tokens_seen": 5950784,
      "step": 9085
    },
    {
      "epoch": 4.764150943396227,
      "grad_norm": 0.5281362533569336,
      "learning_rate": 4.714775335951086e-05,
      "loss": 0.5806,
      "num_input_tokens_seen": 5954016,
      "step": 9090
    },
    {
      "epoch": 4.766771488469602,
      "grad_norm": 0.480430006980896,
      "learning_rate": 4.7142447171080137e-05,
      "loss": 0.395,
      "num_input_tokens_seen": 5957056,
      "step": 9095
    },
    {
      "epoch": 4.769392033542977,
      "grad_norm": 0.5627025365829468,
      "learning_rate": 4.713713635070208e-05,
      "loss": 0.5145,
      "num_input_tokens_seen": 5960352,
      "step": 9100
    },
    {
      "epoch": 4.772012578616352,
      "grad_norm": 0.1980087161064148,
      "learning_rate": 4.713182089948765e-05,
      "loss": 0.3776,
      "num_input_tokens_seen": 5963808,
      "step": 9105
    },
    {
      "epoch": 4.7746331236897275,
      "grad_norm": 0.3578706383705139,
      "learning_rate": 4.712650081854878e-05,
      "loss": 0.4997,
      "num_input_tokens_seen": 5967648,
      "step": 9110
    },
    {
      "epoch": 4.777253668763103,
      "grad_norm": 1.9592618942260742,
      "learning_rate": 4.712117610899837e-05,
      "loss": 0.5272,
      "num_input_tokens_seen": 5969696,
      "step": 9115
    },
    {
      "epoch": 4.779874213836478,
      "grad_norm": 0.432587593793869,
      "learning_rate": 4.7115846771950286e-05,
      "loss": 0.5427,
      "num_input_tokens_seen": 5972864,
      "step": 9120
    },
    {
      "epoch": 4.782494758909853,
      "grad_norm": 0.4063643217086792,
      "learning_rate": 4.711051280851938e-05,
      "loss": 0.4887,
      "num_input_tokens_seen": 5975488,
      "step": 9125
    },
    {
      "epoch": 4.785115303983228,
      "grad_norm": 0.535304844379425,
      "learning_rate": 4.7105174219821435e-05,
      "loss": 0.6702,
      "num_input_tokens_seen": 5978368,
      "step": 9130
    },
    {
      "epoch": 4.787735849056604,
      "grad_norm": 0.3437677323818207,
      "learning_rate": 4.709983100697323e-05,
      "loss": 0.4082,
      "num_input_tokens_seen": 5981792,
      "step": 9135
    },
    {
      "epoch": 4.790356394129979,
      "grad_norm": 1.8330838680267334,
      "learning_rate": 4.709448317109252e-05,
      "loss": 0.4958,
      "num_input_tokens_seen": 5984384,
      "step": 9140
    },
    {
      "epoch": 4.7929769392033545,
      "grad_norm": 0.6257678866386414,
      "learning_rate": 4.708913071329798e-05,
      "loss": 0.6135,
      "num_input_tokens_seen": 5986624,
      "step": 9145
    },
    {
      "epoch": 4.79559748427673,
      "grad_norm": 0.5836499929428101,
      "learning_rate": 4.708377363470932e-05,
      "loss": 0.4571,
      "num_input_tokens_seen": 5989856,
      "step": 9150
    },
    {
      "epoch": 4.798218029350105,
      "grad_norm": 0.49848803877830505,
      "learning_rate": 4.7078411936447144e-05,
      "loss": 0.4663,
      "num_input_tokens_seen": 5992608,
      "step": 9155
    },
    {
      "epoch": 4.80083857442348,
      "grad_norm": 1.0253286361694336,
      "learning_rate": 4.707304561963308e-05,
      "loss": 0.6924,
      "num_input_tokens_seen": 5994976,
      "step": 9160
    },
    {
      "epoch": 4.803459119496855,
      "grad_norm": 0.6019527912139893,
      "learning_rate": 4.70676746853897e-05,
      "loss": 0.3871,
      "num_input_tokens_seen": 5997792,
      "step": 9165
    },
    {
      "epoch": 4.80607966457023,
      "grad_norm": 0.3356931507587433,
      "learning_rate": 4.706229913484054e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 6001472,
      "step": 9170
    },
    {
      "epoch": 4.808700209643606,
      "grad_norm": 0.6817938089370728,
      "learning_rate": 4.705691896911009e-05,
      "loss": 0.5429,
      "num_input_tokens_seen": 6004448,
      "step": 9175
    },
    {
      "epoch": 4.811320754716981,
      "grad_norm": 0.5839144587516785,
      "learning_rate": 4.705153418932383e-05,
      "loss": 0.4118,
      "num_input_tokens_seen": 6007776,
      "step": 9180
    },
    {
      "epoch": 4.813941299790356,
      "grad_norm": 0.46318551898002625,
      "learning_rate": 4.7046144796608206e-05,
      "loss": 0.5733,
      "num_input_tokens_seen": 6011360,
      "step": 9185
    },
    {
      "epoch": 4.816561844863732,
      "grad_norm": 0.5946261286735535,
      "learning_rate": 4.70407507920906e-05,
      "loss": 0.5637,
      "num_input_tokens_seen": 6014464,
      "step": 9190
    },
    {
      "epoch": 4.819182389937107,
      "grad_norm": 0.3658270835876465,
      "learning_rate": 4.7035352176899384e-05,
      "loss": 0.496,
      "num_input_tokens_seen": 6017088,
      "step": 9195
    },
    {
      "epoch": 4.821802935010482,
      "grad_norm": 0.29273751378059387,
      "learning_rate": 4.7029948952163885e-05,
      "loss": 0.5393,
      "num_input_tokens_seen": 6020256,
      "step": 9200
    },
    {
      "epoch": 4.8244234800838575,
      "grad_norm": 0.38432687520980835,
      "learning_rate": 4.7024541119014386e-05,
      "loss": 0.407,
      "num_input_tokens_seen": 6022592,
      "step": 9205
    },
    {
      "epoch": 4.827044025157233,
      "grad_norm": 0.45373937487602234,
      "learning_rate": 4.701912867858217e-05,
      "loss": 0.4403,
      "num_input_tokens_seen": 6025344,
      "step": 9210
    },
    {
      "epoch": 4.829664570230608,
      "grad_norm": 0.6093267798423767,
      "learning_rate": 4.7013711631999434e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 6028128,
      "step": 9215
    },
    {
      "epoch": 4.832285115303983,
      "grad_norm": 0.5001159310340881,
      "learning_rate": 4.700828998039937e-05,
      "loss": 0.4273,
      "num_input_tokens_seen": 6032256,
      "step": 9220
    },
    {
      "epoch": 4.834905660377358,
      "grad_norm": 0.5805115699768066,
      "learning_rate": 4.700286372491612e-05,
      "loss": 0.4861,
      "num_input_tokens_seen": 6035200,
      "step": 9225
    },
    {
      "epoch": 4.837526205450734,
      "grad_norm": 0.41246941685676575,
      "learning_rate": 4.6997432866684816e-05,
      "loss": 0.5466,
      "num_input_tokens_seen": 6037536,
      "step": 9230
    },
    {
      "epoch": 4.840146750524109,
      "grad_norm": 0.3940020799636841,
      "learning_rate": 4.6991997406841504e-05,
      "loss": 0.4369,
      "num_input_tokens_seen": 6040736,
      "step": 9235
    },
    {
      "epoch": 4.8427672955974845,
      "grad_norm": 0.5645831823348999,
      "learning_rate": 4.698655734652324e-05,
      "loss": 0.424,
      "num_input_tokens_seen": 6043904,
      "step": 9240
    },
    {
      "epoch": 4.84538784067086,
      "grad_norm": 0.45393258333206177,
      "learning_rate": 4.698111268686801e-05,
      "loss": 0.4803,
      "num_input_tokens_seen": 6046336,
      "step": 9245
    },
    {
      "epoch": 4.848008385744235,
      "grad_norm": 0.276987224817276,
      "learning_rate": 4.6975663429014785e-05,
      "loss": 0.5203,
      "num_input_tokens_seen": 6049824,
      "step": 9250
    },
    {
      "epoch": 4.85062893081761,
      "grad_norm": 0.3831877112388611,
      "learning_rate": 4.697020957410347e-05,
      "loss": 0.3828,
      "num_input_tokens_seen": 6053600,
      "step": 9255
    },
    {
      "epoch": 4.853249475890985,
      "grad_norm": 0.41437190771102905,
      "learning_rate": 4.6964751123274974e-05,
      "loss": 0.4372,
      "num_input_tokens_seen": 6056192,
      "step": 9260
    },
    {
      "epoch": 4.85587002096436,
      "grad_norm": 0.5558605790138245,
      "learning_rate": 4.695928807767113e-05,
      "loss": 0.5143,
      "num_input_tokens_seen": 6060352,
      "step": 9265
    },
    {
      "epoch": 4.8584905660377355,
      "grad_norm": 0.7298892140388489,
      "learning_rate": 4.695382043843474e-05,
      "loss": 0.3712,
      "num_input_tokens_seen": 6063488,
      "step": 9270
    },
    {
      "epoch": 4.861111111111111,
      "grad_norm": 0.23948480188846588,
      "learning_rate": 4.694834820670958e-05,
      "loss": 0.4775,
      "num_input_tokens_seen": 6066112,
      "step": 9275
    },
    {
      "epoch": 4.863731656184486,
      "grad_norm": 0.42286208271980286,
      "learning_rate": 4.694287138364037e-05,
      "loss": 0.4408,
      "num_input_tokens_seen": 6068928,
      "step": 9280
    },
    {
      "epoch": 4.866352201257862,
      "grad_norm": 0.28198012709617615,
      "learning_rate": 4.6937389970372815e-05,
      "loss": 0.4051,
      "num_input_tokens_seen": 6072256,
      "step": 9285
    },
    {
      "epoch": 4.868972746331237,
      "grad_norm": 0.5376607775688171,
      "learning_rate": 4.693190396805354e-05,
      "loss": 0.4416,
      "num_input_tokens_seen": 6076640,
      "step": 9290
    },
    {
      "epoch": 4.871593291404612,
      "grad_norm": 0.48775532841682434,
      "learning_rate": 4.692641337783017e-05,
      "loss": 0.4298,
      "num_input_tokens_seen": 6081504,
      "step": 9295
    },
    {
      "epoch": 4.8742138364779874,
      "grad_norm": 0.4283694624900818,
      "learning_rate": 4.6920918200851274e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 6084224,
      "step": 9300
    },
    {
      "epoch": 4.876834381551363,
      "grad_norm": 0.5313525199890137,
      "learning_rate": 4.691541843826638e-05,
      "loss": 0.4274,
      "num_input_tokens_seen": 6087040,
      "step": 9305
    },
    {
      "epoch": 4.879454926624738,
      "grad_norm": 0.5009474158287048,
      "learning_rate": 4.690991409122597e-05,
      "loss": 0.5008,
      "num_input_tokens_seen": 6089664,
      "step": 9310
    },
    {
      "epoch": 4.882075471698113,
      "grad_norm": 0.5934856534004211,
      "learning_rate": 4.690440516088149e-05,
      "loss": 0.5372,
      "num_input_tokens_seen": 6092352,
      "step": 9315
    },
    {
      "epoch": 4.884696016771488,
      "grad_norm": 0.6716631650924683,
      "learning_rate": 4.689889164838536e-05,
      "loss": 0.5715,
      "num_input_tokens_seen": 6096000,
      "step": 9320
    },
    {
      "epoch": 4.887316561844864,
      "grad_norm": 0.6613292694091797,
      "learning_rate": 4.6893373554890925e-05,
      "loss": 0.3757,
      "num_input_tokens_seen": 6101920,
      "step": 9325
    },
    {
      "epoch": 4.889937106918239,
      "grad_norm": 0.41409626603126526,
      "learning_rate": 4.688785088155251e-05,
      "loss": 0.3194,
      "num_input_tokens_seen": 6104128,
      "step": 9330
    },
    {
      "epoch": 4.8925576519916145,
      "grad_norm": 0.9148011207580566,
      "learning_rate": 4.688232362952541e-05,
      "loss": 0.6798,
      "num_input_tokens_seen": 6108096,
      "step": 9335
    },
    {
      "epoch": 4.89517819706499,
      "grad_norm": 0.31359246373176575,
      "learning_rate": 4.687679179996585e-05,
      "loss": 0.4653,
      "num_input_tokens_seen": 6112704,
      "step": 9340
    },
    {
      "epoch": 4.897798742138365,
      "grad_norm": 0.5196114182472229,
      "learning_rate": 4.687125539403103e-05,
      "loss": 0.4303,
      "num_input_tokens_seen": 6118816,
      "step": 9345
    },
    {
      "epoch": 4.90041928721174,
      "grad_norm": 0.5029844641685486,
      "learning_rate": 4.68657144128791e-05,
      "loss": 0.4794,
      "num_input_tokens_seen": 6122304,
      "step": 9350
    },
    {
      "epoch": 4.903039832285115,
      "grad_norm": 0.9534074664115906,
      "learning_rate": 4.6860168857669175e-05,
      "loss": 0.3788,
      "num_input_tokens_seen": 6125888,
      "step": 9355
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 0.6732410192489624,
      "learning_rate": 4.6854618729561315e-05,
      "loss": 0.4585,
      "num_input_tokens_seen": 6129056,
      "step": 9360
    },
    {
      "epoch": 4.9082809224318655,
      "grad_norm": 0.4872576594352722,
      "learning_rate": 4.684906402971655e-05,
      "loss": 0.4806,
      "num_input_tokens_seen": 6132672,
      "step": 9365
    },
    {
      "epoch": 4.910901467505241,
      "grad_norm": 0.6056691408157349,
      "learning_rate": 4.684350475929685e-05,
      "loss": 0.5267,
      "num_input_tokens_seen": 6135168,
      "step": 9370
    },
    {
      "epoch": 4.913522012578616,
      "grad_norm": 0.5679202675819397,
      "learning_rate": 4.683794091946517e-05,
      "loss": 0.5962,
      "num_input_tokens_seen": 6139296,
      "step": 9375
    },
    {
      "epoch": 4.916142557651992,
      "grad_norm": 0.4200024902820587,
      "learning_rate": 4.683237251138538e-05,
      "loss": 0.5656,
      "num_input_tokens_seen": 6141824,
      "step": 9380
    },
    {
      "epoch": 4.918763102725367,
      "grad_norm": 0.44931739568710327,
      "learning_rate": 4.6826799536222346e-05,
      "loss": 0.4719,
      "num_input_tokens_seen": 6145248,
      "step": 9385
    },
    {
      "epoch": 4.921383647798742,
      "grad_norm": 0.6132237911224365,
      "learning_rate": 4.6821221995141865e-05,
      "loss": 0.443,
      "num_input_tokens_seen": 6148000,
      "step": 9390
    },
    {
      "epoch": 4.924004192872117,
      "grad_norm": 0.4706277847290039,
      "learning_rate": 4.681563988931068e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 6151392,
      "step": 9395
    },
    {
      "epoch": 4.926624737945493,
      "grad_norm": 0.44898977875709534,
      "learning_rate": 4.681005321989652e-05,
      "loss": 0.5299,
      "num_input_tokens_seen": 6155360,
      "step": 9400
    },
    {
      "epoch": 4.929245283018868,
      "grad_norm": 0.563819944858551,
      "learning_rate": 4.6804461988068055e-05,
      "loss": 0.4316,
      "num_input_tokens_seen": 6159168,
      "step": 9405
    },
    {
      "epoch": 4.931865828092243,
      "grad_norm": 0.7522189617156982,
      "learning_rate": 4.6798866194994896e-05,
      "loss": 0.6087,
      "num_input_tokens_seen": 6162752,
      "step": 9410
    },
    {
      "epoch": 4.934486373165618,
      "grad_norm": 0.575835645198822,
      "learning_rate": 4.6793265841847624e-05,
      "loss": 0.4871,
      "num_input_tokens_seen": 6165536,
      "step": 9415
    },
    {
      "epoch": 4.937106918238994,
      "grad_norm": 0.4054315686225891,
      "learning_rate": 4.6787660929797766e-05,
      "loss": 0.5103,
      "num_input_tokens_seen": 6168768,
      "step": 9420
    },
    {
      "epoch": 4.939727463312369,
      "grad_norm": 0.22794297337532043,
      "learning_rate": 4.678205146001781e-05,
      "loss": 0.4047,
      "num_input_tokens_seen": 6171584,
      "step": 9425
    },
    {
      "epoch": 4.9423480083857445,
      "grad_norm": 0.5045681595802307,
      "learning_rate": 4.67764374336812e-05,
      "loss": 0.4759,
      "num_input_tokens_seen": 6174944,
      "step": 9430
    },
    {
      "epoch": 4.94496855345912,
      "grad_norm": 0.4920738935470581,
      "learning_rate": 4.6770818851962305e-05,
      "loss": 0.4018,
      "num_input_tokens_seen": 6177536,
      "step": 9435
    },
    {
      "epoch": 4.947589098532495,
      "grad_norm": 0.3479418456554413,
      "learning_rate": 4.676519571603648e-05,
      "loss": 0.5795,
      "num_input_tokens_seen": 6181280,
      "step": 9440
    },
    {
      "epoch": 4.95020964360587,
      "grad_norm": 0.29529640078544617,
      "learning_rate": 4.6759568027080034e-05,
      "loss": 0.5588,
      "num_input_tokens_seen": 6184576,
      "step": 9445
    },
    {
      "epoch": 4.952830188679245,
      "grad_norm": 0.38627010583877563,
      "learning_rate": 4.67539357862702e-05,
      "loss": 0.4039,
      "num_input_tokens_seen": 6188288,
      "step": 9450
    },
    {
      "epoch": 4.95545073375262,
      "grad_norm": 0.25129392743110657,
      "learning_rate": 4.6748298994785175e-05,
      "loss": 0.4119,
      "num_input_tokens_seen": 6191680,
      "step": 9455
    },
    {
      "epoch": 4.9580712788259955,
      "grad_norm": 0.6255735754966736,
      "learning_rate": 4.6742657653804116e-05,
      "loss": 0.5011,
      "num_input_tokens_seen": 6194624,
      "step": 9460
    },
    {
      "epoch": 4.960691823899371,
      "grad_norm": 0.3700569272041321,
      "learning_rate": 4.673701176450713e-05,
      "loss": 0.448,
      "num_input_tokens_seen": 6198528,
      "step": 9465
    },
    {
      "epoch": 4.963312368972746,
      "grad_norm": 0.33901527523994446,
      "learning_rate": 4.6731361328075286e-05,
      "loss": 0.4092,
      "num_input_tokens_seen": 6202720,
      "step": 9470
    },
    {
      "epoch": 4.965932914046122,
      "grad_norm": 0.4027758240699768,
      "learning_rate": 4.672570634569056e-05,
      "loss": 0.5042,
      "num_input_tokens_seen": 6205280,
      "step": 9475
    },
    {
      "epoch": 4.968553459119497,
      "grad_norm": 0.41228175163269043,
      "learning_rate": 4.672004681853593e-05,
      "loss": 0.4766,
      "num_input_tokens_seen": 6208032,
      "step": 9480
    },
    {
      "epoch": 4.971174004192872,
      "grad_norm": 0.32158875465393066,
      "learning_rate": 4.6714382747795304e-05,
      "loss": 0.4464,
      "num_input_tokens_seen": 6211552,
      "step": 9485
    },
    {
      "epoch": 4.973794549266247,
      "grad_norm": 0.42655277252197266,
      "learning_rate": 4.670871413465354e-05,
      "loss": 0.4377,
      "num_input_tokens_seen": 6214560,
      "step": 9490
    },
    {
      "epoch": 4.976415094339623,
      "grad_norm": 0.5264145731925964,
      "learning_rate": 4.670304098029645e-05,
      "loss": 0.5633,
      "num_input_tokens_seen": 6217344,
      "step": 9495
    },
    {
      "epoch": 4.979035639412998,
      "grad_norm": 0.37881913781166077,
      "learning_rate": 4.669736328591078e-05,
      "loss": 0.5019,
      "num_input_tokens_seen": 6221152,
      "step": 9500
    },
    {
      "epoch": 4.981656184486373,
      "grad_norm": 0.525000274181366,
      "learning_rate": 4.669168105268424e-05,
      "loss": 0.5756,
      "num_input_tokens_seen": 6223680,
      "step": 9505
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.3402053713798523,
      "learning_rate": 4.6685994281805514e-05,
      "loss": 0.5361,
      "num_input_tokens_seen": 6226688,
      "step": 9510
    },
    {
      "epoch": 4.986897274633124,
      "grad_norm": 0.5196491479873657,
      "learning_rate": 4.6680302974464175e-05,
      "loss": 0.6564,
      "num_input_tokens_seen": 6230016,
      "step": 9515
    },
    {
      "epoch": 4.989517819706499,
      "grad_norm": 0.35979655385017395,
      "learning_rate": 4.66746071318508e-05,
      "loss": 0.5367,
      "num_input_tokens_seen": 6233376,
      "step": 9520
    },
    {
      "epoch": 4.9921383647798745,
      "grad_norm": 0.4340919852256775,
      "learning_rate": 4.6668906755156896e-05,
      "loss": 0.3983,
      "num_input_tokens_seen": 6236832,
      "step": 9525
    },
    {
      "epoch": 4.99475890985325,
      "grad_norm": 0.476689875125885,
      "learning_rate": 4.6663201845574906e-05,
      "loss": 0.495,
      "num_input_tokens_seen": 6239840,
      "step": 9530
    },
    {
      "epoch": 4.997379454926625,
      "grad_norm": 0.5754254460334778,
      "learning_rate": 4.6657492404298234e-05,
      "loss": 0.5237,
      "num_input_tokens_seen": 6243008,
      "step": 9535
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.8035189509391785,
      "learning_rate": 4.665177843252124e-05,
      "loss": 0.4635,
      "num_input_tokens_seen": 6245072,
      "step": 9540
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.4868377149105072,
      "eval_runtime": 9.7564,
      "eval_samples_per_second": 86.917,
      "eval_steps_per_second": 21.729,
      "num_input_tokens_seen": 6245072,
      "step": 9540
    },
    {
      "epoch": 5.002620545073375,
      "grad_norm": 0.5138049721717834,
      "learning_rate": 4.664605993143921e-05,
      "loss": 0.5127,
      "num_input_tokens_seen": 6249424,
      "step": 9545
    },
    {
      "epoch": 5.00524109014675,
      "grad_norm": 0.5067752003669739,
      "learning_rate": 4.6640336902248396e-05,
      "loss": 0.4887,
      "num_input_tokens_seen": 6251952,
      "step": 9550
    },
    {
      "epoch": 5.0078616352201255,
      "grad_norm": 0.3374130427837372,
      "learning_rate": 4.6634609346145984e-05,
      "loss": 0.5043,
      "num_input_tokens_seen": 6254928,
      "step": 9555
    },
    {
      "epoch": 5.010482180293501,
      "grad_norm": 0.43282222747802734,
      "learning_rate": 4.6628877264330116e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 6258000,
      "step": 9560
    },
    {
      "epoch": 5.013102725366877,
      "grad_norm": 0.27740851044654846,
      "learning_rate": 4.6623140657999886e-05,
      "loss": 0.3598,
      "num_input_tokens_seen": 6261200,
      "step": 9565
    },
    {
      "epoch": 5.015723270440252,
      "grad_norm": 0.4857799708843231,
      "learning_rate": 4.661739952835532e-05,
      "loss": 0.4561,
      "num_input_tokens_seen": 6264112,
      "step": 9570
    },
    {
      "epoch": 5.018343815513627,
      "grad_norm": 0.4430839717388153,
      "learning_rate": 4.66116538765974e-05,
      "loss": 0.4136,
      "num_input_tokens_seen": 6268784,
      "step": 9575
    },
    {
      "epoch": 5.020964360587002,
      "grad_norm": 0.5893736481666565,
      "learning_rate": 4.6605903703928034e-05,
      "loss": 0.6237,
      "num_input_tokens_seen": 6271984,
      "step": 9580
    },
    {
      "epoch": 5.023584905660377,
      "grad_norm": 0.48379871249198914,
      "learning_rate": 4.6600149011550116e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 6275568,
      "step": 9585
    },
    {
      "epoch": 5.026205450733753,
      "grad_norm": 0.32327696681022644,
      "learning_rate": 4.659438980066745e-05,
      "loss": 0.4949,
      "num_input_tokens_seen": 6278448,
      "step": 9590
    },
    {
      "epoch": 5.028825995807128,
      "grad_norm": 0.5520633459091187,
      "learning_rate": 4.65886260724848e-05,
      "loss": 0.6352,
      "num_input_tokens_seen": 6281104,
      "step": 9595
    },
    {
      "epoch": 5.031446540880503,
      "grad_norm": 0.37005162239074707,
      "learning_rate": 4.658285782820787e-05,
      "loss": 0.5412,
      "num_input_tokens_seen": 6284752,
      "step": 9600
    },
    {
      "epoch": 5.034067085953878,
      "grad_norm": 0.4008062183856964,
      "learning_rate": 4.657708506904331e-05,
      "loss": 0.4499,
      "num_input_tokens_seen": 6288048,
      "step": 9605
    },
    {
      "epoch": 5.036687631027253,
      "grad_norm": 0.5911264419555664,
      "learning_rate": 4.6571307796198724e-05,
      "loss": 0.5482,
      "num_input_tokens_seen": 6291088,
      "step": 9610
    },
    {
      "epoch": 5.039308176100629,
      "grad_norm": 0.3848652243614197,
      "learning_rate": 4.656552601088265e-05,
      "loss": 0.6585,
      "num_input_tokens_seen": 6294224,
      "step": 9615
    },
    {
      "epoch": 5.0419287211740045,
      "grad_norm": 0.5244715809822083,
      "learning_rate": 4.655973971430456e-05,
      "loss": 0.4893,
      "num_input_tokens_seen": 6297328,
      "step": 9620
    },
    {
      "epoch": 5.04454926624738,
      "grad_norm": 0.3543325364589691,
      "learning_rate": 4.6553948907674895e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 6300208,
      "step": 9625
    },
    {
      "epoch": 5.047169811320755,
      "grad_norm": 0.5074748992919922,
      "learning_rate": 4.654815359220501e-05,
      "loss": 0.3232,
      "num_input_tokens_seen": 6303856,
      "step": 9630
    },
    {
      "epoch": 5.04979035639413,
      "grad_norm": 0.825731098651886,
      "learning_rate": 4.6542353769107226e-05,
      "loss": 0.507,
      "num_input_tokens_seen": 6306960,
      "step": 9635
    },
    {
      "epoch": 5.052410901467505,
      "grad_norm": 0.396777480840683,
      "learning_rate": 4.653654943959481e-05,
      "loss": 0.6817,
      "num_input_tokens_seen": 6310736,
      "step": 9640
    },
    {
      "epoch": 5.05503144654088,
      "grad_norm": 0.6834322214126587,
      "learning_rate": 4.653074060488195e-05,
      "loss": 0.5574,
      "num_input_tokens_seen": 6313776,
      "step": 9645
    },
    {
      "epoch": 5.0576519916142555,
      "grad_norm": 0.44487348198890686,
      "learning_rate": 4.6524927266183796e-05,
      "loss": 0.3526,
      "num_input_tokens_seen": 6316848,
      "step": 9650
    },
    {
      "epoch": 5.060272536687631,
      "grad_norm": 0.2365301251411438,
      "learning_rate": 4.6519109424716426e-05,
      "loss": 0.5421,
      "num_input_tokens_seen": 6321584,
      "step": 9655
    },
    {
      "epoch": 5.062893081761007,
      "grad_norm": 0.2751290500164032,
      "learning_rate": 4.6513287081696855e-05,
      "loss": 0.4673,
      "num_input_tokens_seen": 6325040,
      "step": 9660
    },
    {
      "epoch": 5.065513626834382,
      "grad_norm": 0.5071536898612976,
      "learning_rate": 4.6507460238343076e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 6327984,
      "step": 9665
    },
    {
      "epoch": 5.068134171907757,
      "grad_norm": 0.46117496490478516,
      "learning_rate": 4.6501628895873984e-05,
      "loss": 0.6252,
      "num_input_tokens_seen": 6330800,
      "step": 9670
    },
    {
      "epoch": 5.070754716981132,
      "grad_norm": 0.3692462146282196,
      "learning_rate": 4.649579305550942e-05,
      "loss": 0.3705,
      "num_input_tokens_seen": 6333776,
      "step": 9675
    },
    {
      "epoch": 5.073375262054507,
      "grad_norm": 0.6377898454666138,
      "learning_rate": 4.64899527184702e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 6336816,
      "step": 9680
    },
    {
      "epoch": 5.075995807127883,
      "grad_norm": 0.35258278250694275,
      "learning_rate": 4.6484107885978035e-05,
      "loss": 0.4661,
      "num_input_tokens_seen": 6339280,
      "step": 9685
    },
    {
      "epoch": 5.078616352201258,
      "grad_norm": 0.34058859944343567,
      "learning_rate": 4.64782585592556e-05,
      "loss": 0.404,
      "num_input_tokens_seen": 6342544,
      "step": 9690
    },
    {
      "epoch": 5.081236897274633,
      "grad_norm": 0.3917861878871918,
      "learning_rate": 4.647240473952652e-05,
      "loss": 0.3841,
      "num_input_tokens_seen": 6345072,
      "step": 9695
    },
    {
      "epoch": 5.083857442348008,
      "grad_norm": 0.3422340750694275,
      "learning_rate": 4.6466546428015336e-05,
      "loss": 0.3999,
      "num_input_tokens_seen": 6348496,
      "step": 9700
    },
    {
      "epoch": 5.086477987421383,
      "grad_norm": 0.4052616059780121,
      "learning_rate": 4.646068362594753e-05,
      "loss": 0.5107,
      "num_input_tokens_seen": 6352560,
      "step": 9705
    },
    {
      "epoch": 5.089098532494759,
      "grad_norm": 0.4171753227710724,
      "learning_rate": 4.645481633454956e-05,
      "loss": 0.6376,
      "num_input_tokens_seen": 6356400,
      "step": 9710
    },
    {
      "epoch": 5.0917190775681345,
      "grad_norm": 0.46420469880104065,
      "learning_rate": 4.644894455504878e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 6359312,
      "step": 9715
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 0.30970755219459534,
      "learning_rate": 4.64430682886735e-05,
      "loss": 0.4391,
      "num_input_tokens_seen": 6362480,
      "step": 9720
    },
    {
      "epoch": 5.096960167714885,
      "grad_norm": 0.3311852216720581,
      "learning_rate": 4.643718753665297e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 6365584,
      "step": 9725
    },
    {
      "epoch": 5.09958071278826,
      "grad_norm": 0.4560285806655884,
      "learning_rate": 4.643130230021737e-05,
      "loss": 0.4876,
      "num_input_tokens_seen": 6370384,
      "step": 9730
    },
    {
      "epoch": 5.102201257861635,
      "grad_norm": 0.5972071886062622,
      "learning_rate": 4.6425412580597836e-05,
      "loss": 0.5624,
      "num_input_tokens_seen": 6373104,
      "step": 9735
    },
    {
      "epoch": 5.10482180293501,
      "grad_norm": 0.38333287835121155,
      "learning_rate": 4.641951837902643e-05,
      "loss": 0.4765,
      "num_input_tokens_seen": 6376848,
      "step": 9740
    },
    {
      "epoch": 5.1074423480083855,
      "grad_norm": 0.3504626154899597,
      "learning_rate": 4.641361969673614e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 6381232,
      "step": 9745
    },
    {
      "epoch": 5.110062893081761,
      "grad_norm": 0.8079971075057983,
      "learning_rate": 4.640771653496091e-05,
      "loss": 0.5094,
      "num_input_tokens_seen": 6383760,
      "step": 9750
    },
    {
      "epoch": 5.112683438155136,
      "grad_norm": 0.3916490972042084,
      "learning_rate": 4.64018088949356e-05,
      "loss": 0.5902,
      "num_input_tokens_seen": 6386864,
      "step": 9755
    },
    {
      "epoch": 5.115303983228512,
      "grad_norm": 0.40869155526161194,
      "learning_rate": 4.6395896777896055e-05,
      "loss": 0.4715,
      "num_input_tokens_seen": 6389264,
      "step": 9760
    },
    {
      "epoch": 5.117924528301887,
      "grad_norm": 0.41791489720344543,
      "learning_rate": 4.638998018507899e-05,
      "loss": 0.4542,
      "num_input_tokens_seen": 6393104,
      "step": 9765
    },
    {
      "epoch": 5.120545073375262,
      "grad_norm": 0.3676176965236664,
      "learning_rate": 4.63840591177221e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 6397040,
      "step": 9770
    },
    {
      "epoch": 5.123165618448637,
      "grad_norm": 0.5942994952201843,
      "learning_rate": 4.637813357706401e-05,
      "loss": 0.2995,
      "num_input_tokens_seen": 6400848,
      "step": 9775
    },
    {
      "epoch": 5.1257861635220126,
      "grad_norm": 0.3321831524372101,
      "learning_rate": 4.6372203564344264e-05,
      "loss": 0.3821,
      "num_input_tokens_seen": 6404272,
      "step": 9780
    },
    {
      "epoch": 5.128406708595388,
      "grad_norm": 0.3531937599182129,
      "learning_rate": 4.6366269080803364e-05,
      "loss": 0.4775,
      "num_input_tokens_seen": 6408304,
      "step": 9785
    },
    {
      "epoch": 5.131027253668763,
      "grad_norm": 0.40462231636047363,
      "learning_rate": 4.636033012768273e-05,
      "loss": 0.4456,
      "num_input_tokens_seen": 6410896,
      "step": 9790
    },
    {
      "epoch": 5.133647798742138,
      "grad_norm": 0.5492240786552429,
      "learning_rate": 4.635438670622473e-05,
      "loss": 0.5606,
      "num_input_tokens_seen": 6413968,
      "step": 9795
    },
    {
      "epoch": 5.136268343815513,
      "grad_norm": 0.42170634865760803,
      "learning_rate": 4.6348438817672646e-05,
      "loss": 0.3595,
      "num_input_tokens_seen": 6416432,
      "step": 9800
    },
    {
      "epoch": 5.138888888888889,
      "grad_norm": 0.6186378002166748,
      "learning_rate": 4.634248646327072e-05,
      "loss": 0.6021,
      "num_input_tokens_seen": 6419664,
      "step": 9805
    },
    {
      "epoch": 5.1415094339622645,
      "grad_norm": 0.3363397419452667,
      "learning_rate": 4.6336529644264114e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 6422192,
      "step": 9810
    },
    {
      "epoch": 5.14412997903564,
      "grad_norm": 0.3571457266807556,
      "learning_rate": 4.633056836189893e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 6425360,
      "step": 9815
    },
    {
      "epoch": 5.146750524109015,
      "grad_norm": 0.3076179623603821,
      "learning_rate": 4.632460261742219e-05,
      "loss": 0.4633,
      "num_input_tokens_seen": 6428720,
      "step": 9820
    },
    {
      "epoch": 5.14937106918239,
      "grad_norm": 0.48899757862091064,
      "learning_rate": 4.631863241208187e-05,
      "loss": 0.3673,
      "num_input_tokens_seen": 6431728,
      "step": 9825
    },
    {
      "epoch": 5.151991614255765,
      "grad_norm": 0.6251500248908997,
      "learning_rate": 4.6312657747126865e-05,
      "loss": 0.5519,
      "num_input_tokens_seen": 6434864,
      "step": 9830
    },
    {
      "epoch": 5.15461215932914,
      "grad_norm": 0.49637851119041443,
      "learning_rate": 4.630667862380701e-05,
      "loss": 0.4566,
      "num_input_tokens_seen": 6437904,
      "step": 9835
    },
    {
      "epoch": 5.1572327044025155,
      "grad_norm": 0.3640033006668091,
      "learning_rate": 4.630069504337306e-05,
      "loss": 0.398,
      "num_input_tokens_seen": 6440816,
      "step": 9840
    },
    {
      "epoch": 5.159853249475891,
      "grad_norm": 0.38560932874679565,
      "learning_rate": 4.629470700707672e-05,
      "loss": 0.5916,
      "num_input_tokens_seen": 6443856,
      "step": 9845
    },
    {
      "epoch": 5.162473794549266,
      "grad_norm": 0.23320935666561127,
      "learning_rate": 4.628871451617062e-05,
      "loss": 0.4043,
      "num_input_tokens_seen": 6450064,
      "step": 9850
    },
    {
      "epoch": 5.165094339622642,
      "grad_norm": 0.5753904581069946,
      "learning_rate": 4.628271757190832e-05,
      "loss": 0.3497,
      "num_input_tokens_seen": 6452688,
      "step": 9855
    },
    {
      "epoch": 5.167714884696017,
      "grad_norm": 0.432046502828598,
      "learning_rate": 4.62767161755443e-05,
      "loss": 0.516,
      "num_input_tokens_seen": 6455216,
      "step": 9860
    },
    {
      "epoch": 5.170335429769392,
      "grad_norm": 0.37730205059051514,
      "learning_rate": 4.6270710328334004e-05,
      "loss": 0.5624,
      "num_input_tokens_seen": 6459472,
      "step": 9865
    },
    {
      "epoch": 5.172955974842767,
      "grad_norm": 0.5043449997901917,
      "learning_rate": 4.626470003153377e-05,
      "loss": 0.5607,
      "num_input_tokens_seen": 6465232,
      "step": 9870
    },
    {
      "epoch": 5.1755765199161425,
      "grad_norm": 2.253726005554199,
      "learning_rate": 4.625868528640089e-05,
      "loss": 0.4786,
      "num_input_tokens_seen": 6468016,
      "step": 9875
    },
    {
      "epoch": 5.178197064989518,
      "grad_norm": 0.6341983079910278,
      "learning_rate": 4.6252666094193586e-05,
      "loss": 0.4825,
      "num_input_tokens_seen": 6472144,
      "step": 9880
    },
    {
      "epoch": 5.180817610062893,
      "grad_norm": 0.49072015285491943,
      "learning_rate": 4.6246642456171e-05,
      "loss": 0.5092,
      "num_input_tokens_seen": 6475184,
      "step": 9885
    },
    {
      "epoch": 5.183438155136268,
      "grad_norm": 0.4463558793067932,
      "learning_rate": 4.6240614373593196e-05,
      "loss": 0.4626,
      "num_input_tokens_seen": 6479248,
      "step": 9890
    },
    {
      "epoch": 5.186058700209643,
      "grad_norm": 0.4313424527645111,
      "learning_rate": 4.623458184772119e-05,
      "loss": 0.5033,
      "num_input_tokens_seen": 6482320,
      "step": 9895
    },
    {
      "epoch": 5.188679245283019,
      "grad_norm": 0.28669050335884094,
      "learning_rate": 4.622854487981693e-05,
      "loss": 0.3777,
      "num_input_tokens_seen": 6485680,
      "step": 9900
    },
    {
      "epoch": 5.191299790356394,
      "grad_norm": 0.4057638943195343,
      "learning_rate": 4.622250347114326e-05,
      "loss": 0.4758,
      "num_input_tokens_seen": 6489648,
      "step": 9905
    },
    {
      "epoch": 5.19392033542977,
      "grad_norm": 0.34343287348747253,
      "learning_rate": 4.621645762296399e-05,
      "loss": 0.4716,
      "num_input_tokens_seen": 6493520,
      "step": 9910
    },
    {
      "epoch": 5.196540880503145,
      "grad_norm": 0.2849849760532379,
      "learning_rate": 4.6210407336543835e-05,
      "loss": 0.476,
      "num_input_tokens_seen": 6496752,
      "step": 9915
    },
    {
      "epoch": 5.19916142557652,
      "grad_norm": 0.32733386754989624,
      "learning_rate": 4.620435261314844e-05,
      "loss": 0.4355,
      "num_input_tokens_seen": 6499920,
      "step": 9920
    },
    {
      "epoch": 5.201781970649895,
      "grad_norm": 0.4575313627719879,
      "learning_rate": 4.61982934540444e-05,
      "loss": 0.4301,
      "num_input_tokens_seen": 6502352,
      "step": 9925
    },
    {
      "epoch": 5.20440251572327,
      "grad_norm": 0.4372417628765106,
      "learning_rate": 4.6192229860499206e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 6505008,
      "step": 9930
    },
    {
      "epoch": 5.2070230607966455,
      "grad_norm": 0.34755009412765503,
      "learning_rate": 4.6186161833781295e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 6507568,
      "step": 9935
    },
    {
      "epoch": 5.209643605870021,
      "grad_norm": 0.7654082179069519,
      "learning_rate": 4.618008937516003e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 6510864,
      "step": 9940
    },
    {
      "epoch": 5.212264150943396,
      "grad_norm": 0.5867680311203003,
      "learning_rate": 4.617401248590572e-05,
      "loss": 0.3635,
      "num_input_tokens_seen": 6514544,
      "step": 9945
    },
    {
      "epoch": 5.214884696016772,
      "grad_norm": 0.4544137418270111,
      "learning_rate": 4.616793116728955e-05,
      "loss": 0.559,
      "num_input_tokens_seen": 6517648,
      "step": 9950
    },
    {
      "epoch": 5.217505241090147,
      "grad_norm": 0.2400515228509903,
      "learning_rate": 4.6161845420583685e-05,
      "loss": 0.5488,
      "num_input_tokens_seen": 6521360,
      "step": 9955
    },
    {
      "epoch": 5.220125786163522,
      "grad_norm": 0.6572339534759521,
      "learning_rate": 4.6155755247061174e-05,
      "loss": 0.686,
      "num_input_tokens_seen": 6524560,
      "step": 9960
    },
    {
      "epoch": 5.222746331236897,
      "grad_norm": 0.39121681451797485,
      "learning_rate": 4.614966064799603e-05,
      "loss": 0.5455,
      "num_input_tokens_seen": 6529232,
      "step": 9965
    },
    {
      "epoch": 5.2253668763102725,
      "grad_norm": 0.5321119427680969,
      "learning_rate": 4.6143561624663166e-05,
      "loss": 0.5518,
      "num_input_tokens_seen": 6532016,
      "step": 9970
    },
    {
      "epoch": 5.227987421383648,
      "grad_norm": 0.4835885763168335,
      "learning_rate": 4.6137458178338425e-05,
      "loss": 0.4771,
      "num_input_tokens_seen": 6534960,
      "step": 9975
    },
    {
      "epoch": 5.230607966457023,
      "grad_norm": 0.7196714878082275,
      "learning_rate": 4.613135031029859e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 6538032,
      "step": 9980
    },
    {
      "epoch": 5.233228511530398,
      "grad_norm": 0.3560824990272522,
      "learning_rate": 4.612523802182134e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 6541168,
      "step": 9985
    },
    {
      "epoch": 5.235849056603773,
      "grad_norm": 0.4085720479488373,
      "learning_rate": 4.611912131418531e-05,
      "loss": 0.4491,
      "num_input_tokens_seen": 6543888,
      "step": 9990
    },
    {
      "epoch": 5.238469601677149,
      "grad_norm": 0.6220126748085022,
      "learning_rate": 4.611300018867004e-05,
      "loss": 0.523,
      "num_input_tokens_seen": 6547440,
      "step": 9995
    },
    {
      "epoch": 5.241090146750524,
      "grad_norm": 0.38889697194099426,
      "learning_rate": 4.6106874646556e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 6550864,
      "step": 10000
    },
    {
      "epoch": 5.2437106918239,
      "grad_norm": 0.6648264527320862,
      "learning_rate": 4.6100744689124584e-05,
      "loss": 0.3779,
      "num_input_tokens_seen": 6553872,
      "step": 10005
    },
    {
      "epoch": 5.246331236897275,
      "grad_norm": 0.2828083038330078,
      "learning_rate": 4.6094610317658116e-05,
      "loss": 0.4147,
      "num_input_tokens_seen": 6557456,
      "step": 10010
    },
    {
      "epoch": 5.24895178197065,
      "grad_norm": 0.41432544589042664,
      "learning_rate": 4.608847153343982e-05,
      "loss": 0.4627,
      "num_input_tokens_seen": 6560592,
      "step": 10015
    },
    {
      "epoch": 5.251572327044025,
      "grad_norm": 0.5100093483924866,
      "learning_rate": 4.6082328337753874e-05,
      "loss": 0.5416,
      "num_input_tokens_seen": 6564016,
      "step": 10020
    },
    {
      "epoch": 5.2541928721174,
      "grad_norm": 0.37676045298576355,
      "learning_rate": 4.6076180731885353e-05,
      "loss": 0.5853,
      "num_input_tokens_seen": 6567440,
      "step": 10025
    },
    {
      "epoch": 5.256813417190775,
      "grad_norm": 0.3394431173801422,
      "learning_rate": 4.6070028717120286e-05,
      "loss": 0.4199,
      "num_input_tokens_seen": 6570512,
      "step": 10030
    },
    {
      "epoch": 5.259433962264151,
      "grad_norm": 0.3996858298778534,
      "learning_rate": 4.606387229474558e-05,
      "loss": 0.4175,
      "num_input_tokens_seen": 6573968,
      "step": 10035
    },
    {
      "epoch": 5.262054507337526,
      "grad_norm": 0.35140460729599,
      "learning_rate": 4.60577114660491e-05,
      "loss": 0.3911,
      "num_input_tokens_seen": 6578672,
      "step": 10040
    },
    {
      "epoch": 5.264675052410902,
      "grad_norm": 0.5338471531867981,
      "learning_rate": 4.605154623231962e-05,
      "loss": 0.4805,
      "num_input_tokens_seen": 6581232,
      "step": 10045
    },
    {
      "epoch": 5.267295597484277,
      "grad_norm": 0.43689459562301636,
      "learning_rate": 4.604537659484684e-05,
      "loss": 0.4664,
      "num_input_tokens_seen": 6588688,
      "step": 10050
    },
    {
      "epoch": 5.269916142557652,
      "grad_norm": 0.4004936218261719,
      "learning_rate": 4.603920255492138e-05,
      "loss": 0.4028,
      "num_input_tokens_seen": 6592048,
      "step": 10055
    },
    {
      "epoch": 5.272536687631027,
      "grad_norm": 0.2577179968357086,
      "learning_rate": 4.603302411383477e-05,
      "loss": 0.4517,
      "num_input_tokens_seen": 6595728,
      "step": 10060
    },
    {
      "epoch": 5.2751572327044025,
      "grad_norm": 0.5205550193786621,
      "learning_rate": 4.602684127287946e-05,
      "loss": 0.4084,
      "num_input_tokens_seen": 6598256,
      "step": 10065
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 0.3239362835884094,
      "learning_rate": 4.602065403334885e-05,
      "loss": 0.3968,
      "num_input_tokens_seen": 6601488,
      "step": 10070
    },
    {
      "epoch": 5.280398322851153,
      "grad_norm": 0.28563231229782104,
      "learning_rate": 4.601446239653724e-05,
      "loss": 0.5733,
      "num_input_tokens_seen": 6605616,
      "step": 10075
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 0.4267690181732178,
      "learning_rate": 4.600826636373984e-05,
      "loss": 0.4778,
      "num_input_tokens_seen": 6609616,
      "step": 10080
    },
    {
      "epoch": 5.285639412997903,
      "grad_norm": 0.4344671666622162,
      "learning_rate": 4.600206593625279e-05,
      "loss": 0.6111,
      "num_input_tokens_seen": 6612816,
      "step": 10085
    },
    {
      "epoch": 5.288259958071279,
      "grad_norm": 0.46021321415901184,
      "learning_rate": 4.599586111537315e-05,
      "loss": 0.5539,
      "num_input_tokens_seen": 6615984,
      "step": 10090
    },
    {
      "epoch": 5.290880503144654,
      "grad_norm": 0.40695685148239136,
      "learning_rate": 4.59896519023989e-05,
      "loss": 0.4656,
      "num_input_tokens_seen": 6618480,
      "step": 10095
    },
    {
      "epoch": 5.29350104821803,
      "grad_norm": 0.42855900526046753,
      "learning_rate": 4.5983438298628935e-05,
      "loss": 0.5054,
      "num_input_tokens_seen": 6622928,
      "step": 10100
    },
    {
      "epoch": 5.296121593291405,
      "grad_norm": 0.5113503932952881,
      "learning_rate": 4.597722030536306e-05,
      "loss": 0.4144,
      "num_input_tokens_seen": 6625872,
      "step": 10105
    },
    {
      "epoch": 5.29874213836478,
      "grad_norm": 0.4130924344062805,
      "learning_rate": 4.597099792390202e-05,
      "loss": 0.5946,
      "num_input_tokens_seen": 6628912,
      "step": 10110
    },
    {
      "epoch": 5.301362683438155,
      "grad_norm": 0.39806196093559265,
      "learning_rate": 4.596477115554748e-05,
      "loss": 0.5847,
      "num_input_tokens_seen": 6632336,
      "step": 10115
    },
    {
      "epoch": 5.30398322851153,
      "grad_norm": 0.5984174013137817,
      "learning_rate": 4.5958540001601976e-05,
      "loss": 0.4866,
      "num_input_tokens_seen": 6635440,
      "step": 10120
    },
    {
      "epoch": 5.306603773584905,
      "grad_norm": 0.46714258193969727,
      "learning_rate": 4.595230446336902e-05,
      "loss": 0.3692,
      "num_input_tokens_seen": 6640304,
      "step": 10125
    },
    {
      "epoch": 5.309224318658281,
      "grad_norm": 0.40191495418548584,
      "learning_rate": 4.5946064542153005e-05,
      "loss": 0.4525,
      "num_input_tokens_seen": 6646064,
      "step": 10130
    },
    {
      "epoch": 5.311844863731656,
      "grad_norm": 0.47679343819618225,
      "learning_rate": 4.593982023925926e-05,
      "loss": 0.4866,
      "num_input_tokens_seen": 6648816,
      "step": 10135
    },
    {
      "epoch": 5.314465408805032,
      "grad_norm": 0.5649965405464172,
      "learning_rate": 4.5933571555994e-05,
      "loss": 0.5348,
      "num_input_tokens_seen": 6652144,
      "step": 10140
    },
    {
      "epoch": 5.317085953878407,
      "grad_norm": 0.38173210620880127,
      "learning_rate": 4.59273184936644e-05,
      "loss": 0.4025,
      "num_input_tokens_seen": 6656080,
      "step": 10145
    },
    {
      "epoch": 5.319706498951782,
      "grad_norm": 1.1495949029922485,
      "learning_rate": 4.5921061053578523e-05,
      "loss": 0.5323,
      "num_input_tokens_seen": 6660048,
      "step": 10150
    },
    {
      "epoch": 5.322327044025157,
      "grad_norm": 0.39833247661590576,
      "learning_rate": 4.5914799237045355e-05,
      "loss": 0.5572,
      "num_input_tokens_seen": 6663344,
      "step": 10155
    },
    {
      "epoch": 5.3249475890985325,
      "grad_norm": 1.0070589780807495,
      "learning_rate": 4.5908533045374796e-05,
      "loss": 0.6107,
      "num_input_tokens_seen": 6666544,
      "step": 10160
    },
    {
      "epoch": 5.327568134171908,
      "grad_norm": 0.36275172233581543,
      "learning_rate": 4.590226247987766e-05,
      "loss": 0.473,
      "num_input_tokens_seen": 6670032,
      "step": 10165
    },
    {
      "epoch": 5.330188679245283,
      "grad_norm": 0.3092229664325714,
      "learning_rate": 4.589598754186568e-05,
      "loss": 0.5965,
      "num_input_tokens_seen": 6672592,
      "step": 10170
    },
    {
      "epoch": 5.332809224318658,
      "grad_norm": 0.5215729475021362,
      "learning_rate": 4.5889708232651494e-05,
      "loss": 0.5271,
      "num_input_tokens_seen": 6675664,
      "step": 10175
    },
    {
      "epoch": 5.335429769392033,
      "grad_norm": 0.3354957699775696,
      "learning_rate": 4.588342455354867e-05,
      "loss": 0.3675,
      "num_input_tokens_seen": 6678512,
      "step": 10180
    },
    {
      "epoch": 5.338050314465409,
      "grad_norm": 0.42925381660461426,
      "learning_rate": 4.5877136505871685e-05,
      "loss": 0.5703,
      "num_input_tokens_seen": 6682352,
      "step": 10185
    },
    {
      "epoch": 5.340670859538784,
      "grad_norm": 0.6919520497322083,
      "learning_rate": 4.5870844090935914e-05,
      "loss": 0.4548,
      "num_input_tokens_seen": 6684720,
      "step": 10190
    },
    {
      "epoch": 5.34329140461216,
      "grad_norm": 0.30942490696907043,
      "learning_rate": 4.586454731005768e-05,
      "loss": 0.4362,
      "num_input_tokens_seen": 6688304,
      "step": 10195
    },
    {
      "epoch": 5.345911949685535,
      "grad_norm": 0.3610786199569702,
      "learning_rate": 4.5858246164554175e-05,
      "loss": 0.4099,
      "num_input_tokens_seen": 6691632,
      "step": 10200
    },
    {
      "epoch": 5.34853249475891,
      "grad_norm": 0.38756445050239563,
      "learning_rate": 4.5851940655743536e-05,
      "loss": 0.503,
      "num_input_tokens_seen": 6694576,
      "step": 10205
    },
    {
      "epoch": 5.351153039832285,
      "grad_norm": 0.6617856025695801,
      "learning_rate": 4.5845630784944804e-05,
      "loss": 0.5731,
      "num_input_tokens_seen": 6697648,
      "step": 10210
    },
    {
      "epoch": 5.35377358490566,
      "grad_norm": 0.44317910075187683,
      "learning_rate": 4.583931655347793e-05,
      "loss": 0.5371,
      "num_input_tokens_seen": 6701296,
      "step": 10215
    },
    {
      "epoch": 5.356394129979035,
      "grad_norm": 0.6184029579162598,
      "learning_rate": 4.5832997962663785e-05,
      "loss": 0.4937,
      "num_input_tokens_seen": 6704240,
      "step": 10220
    },
    {
      "epoch": 5.359014675052411,
      "grad_norm": 0.31336042284965515,
      "learning_rate": 4.582667501382414e-05,
      "loss": 0.4156,
      "num_input_tokens_seen": 6707248,
      "step": 10225
    },
    {
      "epoch": 5.361635220125786,
      "grad_norm": 0.40087008476257324,
      "learning_rate": 4.582034770828169e-05,
      "loss": 0.615,
      "num_input_tokens_seen": 6709712,
      "step": 10230
    },
    {
      "epoch": 5.364255765199162,
      "grad_norm": 0.21725380420684814,
      "learning_rate": 4.581401604736002e-05,
      "loss": 0.5536,
      "num_input_tokens_seen": 6715376,
      "step": 10235
    },
    {
      "epoch": 5.366876310272537,
      "grad_norm": 0.6256313920021057,
      "learning_rate": 4.580768003238365e-05,
      "loss": 0.4068,
      "num_input_tokens_seen": 6717840,
      "step": 10240
    },
    {
      "epoch": 5.369496855345912,
      "grad_norm": 0.45114049315452576,
      "learning_rate": 4.5801339664678014e-05,
      "loss": 0.2878,
      "num_input_tokens_seen": 6720816,
      "step": 10245
    },
    {
      "epoch": 5.372117400419287,
      "grad_norm": 0.6995701789855957,
      "learning_rate": 4.579499494556943e-05,
      "loss": 0.5666,
      "num_input_tokens_seen": 6724176,
      "step": 10250
    },
    {
      "epoch": 5.3747379454926625,
      "grad_norm": 0.45738914608955383,
      "learning_rate": 4.5788645876385147e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 6727280,
      "step": 10255
    },
    {
      "epoch": 5.377358490566038,
      "grad_norm": 0.6435501575469971,
      "learning_rate": 4.578229245845331e-05,
      "loss": 0.4173,
      "num_input_tokens_seen": 6730832,
      "step": 10260
    },
    {
      "epoch": 5.379979035639413,
      "grad_norm": 0.4181099832057953,
      "learning_rate": 4.5775934693102994e-05,
      "loss": 0.5129,
      "num_input_tokens_seen": 6734032,
      "step": 10265
    },
    {
      "epoch": 5.382599580712788,
      "grad_norm": 0.4922373294830322,
      "learning_rate": 4.5769572581664165e-05,
      "loss": 0.502,
      "num_input_tokens_seen": 6737232,
      "step": 10270
    },
    {
      "epoch": 5.385220125786163,
      "grad_norm": 0.465262234210968,
      "learning_rate": 4.57632061254677e-05,
      "loss": 0.4374,
      "num_input_tokens_seen": 6740560,
      "step": 10275
    },
    {
      "epoch": 5.387840670859539,
      "grad_norm": 0.32511958479881287,
      "learning_rate": 4.5756835325845394e-05,
      "loss": 0.433,
      "num_input_tokens_seen": 6743344,
      "step": 10280
    },
    {
      "epoch": 5.390461215932914,
      "grad_norm": 0.8672952055931091,
      "learning_rate": 4.575046018412995e-05,
      "loss": 0.5637,
      "num_input_tokens_seen": 6746448,
      "step": 10285
    },
    {
      "epoch": 5.3930817610062896,
      "grad_norm": 0.35809561610221863,
      "learning_rate": 4.5744080701654955e-05,
      "loss": 0.4084,
      "num_input_tokens_seen": 6749008,
      "step": 10290
    },
    {
      "epoch": 5.395702306079665,
      "grad_norm": 0.4642430245876312,
      "learning_rate": 4.573769687975495e-05,
      "loss": 0.6111,
      "num_input_tokens_seen": 6752048,
      "step": 10295
    },
    {
      "epoch": 5.39832285115304,
      "grad_norm": 0.6508763432502747,
      "learning_rate": 4.573130871976534e-05,
      "loss": 0.537,
      "num_input_tokens_seen": 6754512,
      "step": 10300
    },
    {
      "epoch": 5.400943396226415,
      "grad_norm": 0.36560919880867004,
      "learning_rate": 4.572491622302247e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 6757744,
      "step": 10305
    },
    {
      "epoch": 5.40356394129979,
      "grad_norm": 0.271802693605423,
      "learning_rate": 4.571851939086356e-05,
      "loss": 0.395,
      "num_input_tokens_seen": 6762704,
      "step": 10310
    },
    {
      "epoch": 5.406184486373165,
      "grad_norm": 0.6290729641914368,
      "learning_rate": 4.571211822462676e-05,
      "loss": 0.4812,
      "num_input_tokens_seen": 6765264,
      "step": 10315
    },
    {
      "epoch": 5.408805031446541,
      "grad_norm": 0.48026344180107117,
      "learning_rate": 4.570571272565114e-05,
      "loss": 0.4755,
      "num_input_tokens_seen": 6768976,
      "step": 10320
    },
    {
      "epoch": 5.411425576519916,
      "grad_norm": 0.5269800424575806,
      "learning_rate": 4.569930289527663e-05,
      "loss": 0.4444,
      "num_input_tokens_seen": 6771952,
      "step": 10325
    },
    {
      "epoch": 5.414046121593292,
      "grad_norm": 0.24107438325881958,
      "learning_rate": 4.56928887348441e-05,
      "loss": 0.3619,
      "num_input_tokens_seen": 6776656,
      "step": 10330
    },
    {
      "epoch": 5.416666666666667,
      "grad_norm": 0.7366405725479126,
      "learning_rate": 4.568647024569533e-05,
      "loss": 0.6325,
      "num_input_tokens_seen": 6779440,
      "step": 10335
    },
    {
      "epoch": 5.419287211740042,
      "grad_norm": 0.45189279317855835,
      "learning_rate": 4.568004742917299e-05,
      "loss": 0.4187,
      "num_input_tokens_seen": 6782352,
      "step": 10340
    },
    {
      "epoch": 5.421907756813417,
      "grad_norm": 0.3500976264476776,
      "learning_rate": 4.567362028662065e-05,
      "loss": 0.3962,
      "num_input_tokens_seen": 6785744,
      "step": 10345
    },
    {
      "epoch": 5.4245283018867925,
      "grad_norm": 0.842940092086792,
      "learning_rate": 4.566718881938281e-05,
      "loss": 0.6215,
      "num_input_tokens_seen": 6788688,
      "step": 10350
    },
    {
      "epoch": 5.427148846960168,
      "grad_norm": 0.34458282589912415,
      "learning_rate": 4.5660753028804846e-05,
      "loss": 0.3607,
      "num_input_tokens_seen": 6793072,
      "step": 10355
    },
    {
      "epoch": 5.429769392033543,
      "grad_norm": 0.427960067987442,
      "learning_rate": 4.5654312916233066e-05,
      "loss": 0.4886,
      "num_input_tokens_seen": 6796720,
      "step": 10360
    },
    {
      "epoch": 5.432389937106918,
      "grad_norm": 0.9129108190536499,
      "learning_rate": 4.564786848301466e-05,
      "loss": 0.6248,
      "num_input_tokens_seen": 6799312,
      "step": 10365
    },
    {
      "epoch": 5.435010482180293,
      "grad_norm": 0.40517961978912354,
      "learning_rate": 4.564141973049773e-05,
      "loss": 0.4397,
      "num_input_tokens_seen": 6803696,
      "step": 10370
    },
    {
      "epoch": 5.437631027253669,
      "grad_norm": 0.5112671256065369,
      "learning_rate": 4.563496666003128e-05,
      "loss": 0.4939,
      "num_input_tokens_seen": 6807312,
      "step": 10375
    },
    {
      "epoch": 5.440251572327044,
      "grad_norm": 0.22947467863559723,
      "learning_rate": 4.562850927296523e-05,
      "loss": 0.5247,
      "num_input_tokens_seen": 6810640,
      "step": 10380
    },
    {
      "epoch": 5.4428721174004195,
      "grad_norm": 0.31095799803733826,
      "learning_rate": 4.5622047570650365e-05,
      "loss": 0.3974,
      "num_input_tokens_seen": 6817904,
      "step": 10385
    },
    {
      "epoch": 5.445492662473795,
      "grad_norm": 0.3515860438346863,
      "learning_rate": 4.561558155443843e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 6820400,
      "step": 10390
    },
    {
      "epoch": 5.44811320754717,
      "grad_norm": 0.5632727146148682,
      "learning_rate": 4.560911122568202e-05,
      "loss": 0.5269,
      "num_input_tokens_seen": 6823376,
      "step": 10395
    },
    {
      "epoch": 5.450733752620545,
      "grad_norm": 0.4363687336444855,
      "learning_rate": 4.560263658573467e-05,
      "loss": 0.5393,
      "num_input_tokens_seen": 6826928,
      "step": 10400
    },
    {
      "epoch": 5.45335429769392,
      "grad_norm": 0.3011741042137146,
      "learning_rate": 4.559615763595079e-05,
      "loss": 0.4634,
      "num_input_tokens_seen": 6830064,
      "step": 10405
    },
    {
      "epoch": 5.455974842767295,
      "grad_norm": 0.5075963735580444,
      "learning_rate": 4.558967437768571e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 6833232,
      "step": 10410
    },
    {
      "epoch": 5.4585953878406706,
      "grad_norm": 0.2687356770038605,
      "learning_rate": 4.558318681229565e-05,
      "loss": 0.5899,
      "num_input_tokens_seen": 6837072,
      "step": 10415
    },
    {
      "epoch": 5.461215932914046,
      "grad_norm": 0.5043230056762695,
      "learning_rate": 4.557669494113773e-05,
      "loss": 0.5138,
      "num_input_tokens_seen": 6840080,
      "step": 10420
    },
    {
      "epoch": 5.463836477987422,
      "grad_norm": 0.43380776047706604,
      "learning_rate": 4.557019876556998e-05,
      "loss": 0.5265,
      "num_input_tokens_seen": 6843184,
      "step": 10425
    },
    {
      "epoch": 5.466457023060797,
      "grad_norm": 0.35507169365882874,
      "learning_rate": 4.556369828695133e-05,
      "loss": 0.5504,
      "num_input_tokens_seen": 6845968,
      "step": 10430
    },
    {
      "epoch": 5.469077568134172,
      "grad_norm": 0.5636680126190186,
      "learning_rate": 4.5557193506641596e-05,
      "loss": 0.5182,
      "num_input_tokens_seen": 6848688,
      "step": 10435
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 0.46101123094558716,
      "learning_rate": 4.555068442600151e-05,
      "loss": 0.4507,
      "num_input_tokens_seen": 6851984,
      "step": 10440
    },
    {
      "epoch": 5.4743186582809225,
      "grad_norm": 0.3168731927871704,
      "learning_rate": 4.5544171046392703e-05,
      "loss": 0.4979,
      "num_input_tokens_seen": 6855056,
      "step": 10445
    },
    {
      "epoch": 5.476939203354298,
      "grad_norm": 0.3506469428539276,
      "learning_rate": 4.5537653369177685e-05,
      "loss": 0.5399,
      "num_input_tokens_seen": 6858480,
      "step": 10450
    },
    {
      "epoch": 5.479559748427673,
      "grad_norm": 0.6124463081359863,
      "learning_rate": 4.5531131395719895e-05,
      "loss": 0.5375,
      "num_input_tokens_seen": 6861104,
      "step": 10455
    },
    {
      "epoch": 5.482180293501048,
      "grad_norm": 0.36075857281684875,
      "learning_rate": 4.552460512738364e-05,
      "loss": 0.6341,
      "num_input_tokens_seen": 6864848,
      "step": 10460
    },
    {
      "epoch": 5.484800838574423,
      "grad_norm": 0.349478155374527,
      "learning_rate": 4.551807456553416e-05,
      "loss": 0.556,
      "num_input_tokens_seen": 6867248,
      "step": 10465
    },
    {
      "epoch": 5.487421383647799,
      "grad_norm": 0.2773211896419525,
      "learning_rate": 4.551153971153755e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 6870800,
      "step": 10470
    },
    {
      "epoch": 5.490041928721174,
      "grad_norm": 0.8518502116203308,
      "learning_rate": 4.550500056676085e-05,
      "loss": 0.4853,
      "num_input_tokens_seen": 6874288,
      "step": 10475
    },
    {
      "epoch": 5.4926624737945495,
      "grad_norm": 0.2438548356294632,
      "learning_rate": 4.5498457132571966e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 6877968,
      "step": 10480
    },
    {
      "epoch": 5.495283018867925,
      "grad_norm": 0.29526805877685547,
      "learning_rate": 4.5491909410339706e-05,
      "loss": 0.4613,
      "num_input_tokens_seen": 6882416,
      "step": 10485
    },
    {
      "epoch": 5.4979035639413,
      "grad_norm": 0.5157425999641418,
      "learning_rate": 4.548535740143378e-05,
      "loss": 0.4162,
      "num_input_tokens_seen": 6885232,
      "step": 10490
    },
    {
      "epoch": 5.500524109014675,
      "grad_norm": 0.31875255703926086,
      "learning_rate": 4.54788011072248e-05,
      "loss": 0.6006,
      "num_input_tokens_seen": 6888368,
      "step": 10495
    },
    {
      "epoch": 5.50314465408805,
      "grad_norm": 0.44243356585502625,
      "learning_rate": 4.5472240529084254e-05,
      "loss": 0.4026,
      "num_input_tokens_seen": 6892080,
      "step": 10500
    },
    {
      "epoch": 5.505765199161425,
      "grad_norm": 0.26448121666908264,
      "learning_rate": 4.5465675668384557e-05,
      "loss": 0.4485,
      "num_input_tokens_seen": 6895664,
      "step": 10505
    },
    {
      "epoch": 5.5083857442348005,
      "grad_norm": 0.5857195854187012,
      "learning_rate": 4.545910652649899e-05,
      "loss": 0.4611,
      "num_input_tokens_seen": 6898992,
      "step": 10510
    },
    {
      "epoch": 5.511006289308176,
      "grad_norm": 0.7229539155960083,
      "learning_rate": 4.5452533104801756e-05,
      "loss": 0.4517,
      "num_input_tokens_seen": 6902160,
      "step": 10515
    },
    {
      "epoch": 5.513626834381551,
      "grad_norm": 1.3091329336166382,
      "learning_rate": 4.544595540466792e-05,
      "loss": 0.5128,
      "num_input_tokens_seen": 6904752,
      "step": 10520
    },
    {
      "epoch": 5.516247379454927,
      "grad_norm": 0.35234352946281433,
      "learning_rate": 4.5439373427473476e-05,
      "loss": 0.5493,
      "num_input_tokens_seen": 6907792,
      "step": 10525
    },
    {
      "epoch": 5.518867924528302,
      "grad_norm": 0.33688411116600037,
      "learning_rate": 4.54327871745953e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 6910608,
      "step": 10530
    },
    {
      "epoch": 5.521488469601677,
      "grad_norm": 0.7717386484146118,
      "learning_rate": 4.542619664741116e-05,
      "loss": 0.4855,
      "num_input_tokens_seen": 6913584,
      "step": 10535
    },
    {
      "epoch": 5.524109014675052,
      "grad_norm": 0.462739497423172,
      "learning_rate": 4.5419601847299706e-05,
      "loss": 0.3718,
      "num_input_tokens_seen": 6916432,
      "step": 10540
    },
    {
      "epoch": 5.526729559748428,
      "grad_norm": 0.4581146538257599,
      "learning_rate": 4.5413002775640503e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 6920048,
      "step": 10545
    },
    {
      "epoch": 5.529350104821803,
      "grad_norm": 0.4157060980796814,
      "learning_rate": 4.540639943381401e-05,
      "loss": 0.3874,
      "num_input_tokens_seen": 6922544,
      "step": 10550
    },
    {
      "epoch": 5.531970649895178,
      "grad_norm": 0.5417701601982117,
      "learning_rate": 4.539979182320157e-05,
      "loss": 0.3298,
      "num_input_tokens_seen": 6926576,
      "step": 10555
    },
    {
      "epoch": 5.534591194968553,
      "grad_norm": 0.6609598994255066,
      "learning_rate": 4.5393179945185404e-05,
      "loss": 0.4583,
      "num_input_tokens_seen": 6930288,
      "step": 10560
    },
    {
      "epoch": 5.537211740041929,
      "grad_norm": 0.2683587372303009,
      "learning_rate": 4.538656380114865e-05,
      "loss": 0.4292,
      "num_input_tokens_seen": 6933232,
      "step": 10565
    },
    {
      "epoch": 5.539832285115304,
      "grad_norm": 0.4703189730644226,
      "learning_rate": 4.537994339247533e-05,
      "loss": 0.4354,
      "num_input_tokens_seen": 6936752,
      "step": 10570
    },
    {
      "epoch": 5.5424528301886795,
      "grad_norm": 0.6178542375564575,
      "learning_rate": 4.5373318720550366e-05,
      "loss": 0.4876,
      "num_input_tokens_seen": 6939280,
      "step": 10575
    },
    {
      "epoch": 5.545073375262055,
      "grad_norm": 0.42740949988365173,
      "learning_rate": 4.536668978675956e-05,
      "loss": 0.4554,
      "num_input_tokens_seen": 6941520,
      "step": 10580
    },
    {
      "epoch": 5.54769392033543,
      "grad_norm": 0.9957689642906189,
      "learning_rate": 4.53600565924896e-05,
      "loss": 0.5984,
      "num_input_tokens_seen": 6944048,
      "step": 10585
    },
    {
      "epoch": 5.550314465408805,
      "grad_norm": 1.1129319667816162,
      "learning_rate": 4.535341913912808e-05,
      "loss": 0.5367,
      "num_input_tokens_seen": 6946768,
      "step": 10590
    },
    {
      "epoch": 5.55293501048218,
      "grad_norm": 0.45529088377952576,
      "learning_rate": 4.534677742806348e-05,
      "loss": 0.6748,
      "num_input_tokens_seen": 6950192,
      "step": 10595
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 0.43013691902160645,
      "learning_rate": 4.5340131460685175e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 6953168,
      "step": 10600
    },
    {
      "epoch": 5.5581761006289305,
      "grad_norm": 0.5381441712379456,
      "learning_rate": 4.533348123838341e-05,
      "loss": 0.379,
      "num_input_tokens_seen": 6956560,
      "step": 10605
    },
    {
      "epoch": 5.560796645702306,
      "grad_norm": 0.8704571723937988,
      "learning_rate": 4.532682676254936e-05,
      "loss": 0.462,
      "num_input_tokens_seen": 6958896,
      "step": 10610
    },
    {
      "epoch": 5.563417190775681,
      "grad_norm": 0.8085813522338867,
      "learning_rate": 4.5320168034575044e-05,
      "loss": 0.4206,
      "num_input_tokens_seen": 6962864,
      "step": 10615
    },
    {
      "epoch": 5.566037735849057,
      "grad_norm": 1.0773733854293823,
      "learning_rate": 4.53135050558534e-05,
      "loss": 0.5462,
      "num_input_tokens_seen": 6966448,
      "step": 10620
    },
    {
      "epoch": 5.568658280922432,
      "grad_norm": 0.45465800166130066,
      "learning_rate": 4.530683782777825e-05,
      "loss": 0.3699,
      "num_input_tokens_seen": 6969072,
      "step": 10625
    },
    {
      "epoch": 5.571278825995807,
      "grad_norm": 0.372235506772995,
      "learning_rate": 4.53001663517443e-05,
      "loss": 0.2708,
      "num_input_tokens_seen": 6972048,
      "step": 10630
    },
    {
      "epoch": 5.573899371069182,
      "grad_norm": 0.3003990054130554,
      "learning_rate": 4.529349062914714e-05,
      "loss": 0.4011,
      "num_input_tokens_seen": 6975280,
      "step": 10635
    },
    {
      "epoch": 5.576519916142558,
      "grad_norm": 0.2718542218208313,
      "learning_rate": 4.528681066138326e-05,
      "loss": 0.3393,
      "num_input_tokens_seen": 6978096,
      "step": 10640
    },
    {
      "epoch": 5.579140461215933,
      "grad_norm": 0.3917706310749054,
      "learning_rate": 4.528012644985005e-05,
      "loss": 0.3994,
      "num_input_tokens_seen": 6982896,
      "step": 10645
    },
    {
      "epoch": 5.581761006289308,
      "grad_norm": 0.6297888159751892,
      "learning_rate": 4.527343799594574e-05,
      "loss": 0.4945,
      "num_input_tokens_seen": 6985392,
      "step": 10650
    },
    {
      "epoch": 5.584381551362683,
      "grad_norm": 0.6935398578643799,
      "learning_rate": 4.5266745301069494e-05,
      "loss": 0.4082,
      "num_input_tokens_seen": 6989456,
      "step": 10655
    },
    {
      "epoch": 5.587002096436059,
      "grad_norm": 0.42573362588882446,
      "learning_rate": 4.526004836662135e-05,
      "loss": 0.3195,
      "num_input_tokens_seen": 6991856,
      "step": 10660
    },
    {
      "epoch": 5.589622641509434,
      "grad_norm": 0.44242140650749207,
      "learning_rate": 4.525334719400223e-05,
      "loss": 0.4815,
      "num_input_tokens_seen": 6995440,
      "step": 10665
    },
    {
      "epoch": 5.5922431865828095,
      "grad_norm": 0.38969773054122925,
      "learning_rate": 4.524664178461393e-05,
      "loss": 0.377,
      "num_input_tokens_seen": 6998128,
      "step": 10670
    },
    {
      "epoch": 5.594863731656185,
      "grad_norm": 0.4809652268886566,
      "learning_rate": 4.5239932139859154e-05,
      "loss": 0.4028,
      "num_input_tokens_seen": 7001232,
      "step": 10675
    },
    {
      "epoch": 5.59748427672956,
      "grad_norm": 0.3113069534301758,
      "learning_rate": 4.523321826114149e-05,
      "loss": 0.3974,
      "num_input_tokens_seen": 7005040,
      "step": 10680
    },
    {
      "epoch": 5.600104821802935,
      "grad_norm": 0.8668071627616882,
      "learning_rate": 4.5226500149865395e-05,
      "loss": 0.4169,
      "num_input_tokens_seen": 7007792,
      "step": 10685
    },
    {
      "epoch": 5.60272536687631,
      "grad_norm": 0.3567987084388733,
      "learning_rate": 4.5219777807436224e-05,
      "loss": 0.4545,
      "num_input_tokens_seen": 7010576,
      "step": 10690
    },
    {
      "epoch": 5.605345911949685,
      "grad_norm": 0.9457409977912903,
      "learning_rate": 4.5213051235260225e-05,
      "loss": 0.4053,
      "num_input_tokens_seen": 7013328,
      "step": 10695
    },
    {
      "epoch": 5.6079664570230605,
      "grad_norm": 0.48015493154525757,
      "learning_rate": 4.52063204347445e-05,
      "loss": 0.582,
      "num_input_tokens_seen": 7016976,
      "step": 10700
    },
    {
      "epoch": 5.610587002096436,
      "grad_norm": 0.2945677638053894,
      "learning_rate": 4.519958540729707e-05,
      "loss": 0.4521,
      "num_input_tokens_seen": 7020400,
      "step": 10705
    },
    {
      "epoch": 5.613207547169811,
      "grad_norm": 0.4464269280433655,
      "learning_rate": 4.519284615432682e-05,
      "loss": 0.5314,
      "num_input_tokens_seen": 7023088,
      "step": 10710
    },
    {
      "epoch": 5.615828092243187,
      "grad_norm": 0.25452157855033875,
      "learning_rate": 4.518610267724352e-05,
      "loss": 0.4466,
      "num_input_tokens_seen": 7026960,
      "step": 10715
    },
    {
      "epoch": 5.618448637316562,
      "grad_norm": 0.5247369408607483,
      "learning_rate": 4.517935497745786e-05,
      "loss": 0.3586,
      "num_input_tokens_seen": 7029968,
      "step": 10720
    },
    {
      "epoch": 5.621069182389937,
      "grad_norm": 0.33566129207611084,
      "learning_rate": 4.517260305638133e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 7033520,
      "step": 10725
    },
    {
      "epoch": 5.623689727463312,
      "grad_norm": 0.44591158628463745,
      "learning_rate": 4.516584691542639e-05,
      "loss": 0.4563,
      "num_input_tokens_seen": 7037680,
      "step": 10730
    },
    {
      "epoch": 5.626310272536688,
      "grad_norm": 0.43018049001693726,
      "learning_rate": 4.515908655600634e-05,
      "loss": 0.4052,
      "num_input_tokens_seen": 7040464,
      "step": 10735
    },
    {
      "epoch": 5.628930817610063,
      "grad_norm": 0.3990160822868347,
      "learning_rate": 4.515232197953537e-05,
      "loss": 0.5401,
      "num_input_tokens_seen": 7043024,
      "step": 10740
    },
    {
      "epoch": 5.631551362683438,
      "grad_norm": 0.41264545917510986,
      "learning_rate": 4.514555318742856e-05,
      "loss": 0.5868,
      "num_input_tokens_seen": 7046992,
      "step": 10745
    },
    {
      "epoch": 5.634171907756813,
      "grad_norm": 0.5898436307907104,
      "learning_rate": 4.513878018110184e-05,
      "loss": 0.4956,
      "num_input_tokens_seen": 7050160,
      "step": 10750
    },
    {
      "epoch": 5.636792452830189,
      "grad_norm": 0.3250011205673218,
      "learning_rate": 4.513200296197207e-05,
      "loss": 0.4057,
      "num_input_tokens_seen": 7052880,
      "step": 10755
    },
    {
      "epoch": 5.639412997903564,
      "grad_norm": 0.41512325406074524,
      "learning_rate": 4.512522153145696e-05,
      "loss": 0.3932,
      "num_input_tokens_seen": 7055856,
      "step": 10760
    },
    {
      "epoch": 5.6420335429769395,
      "grad_norm": 0.5889496207237244,
      "learning_rate": 4.5118435890975106e-05,
      "loss": 0.4157,
      "num_input_tokens_seen": 7058352,
      "step": 10765
    },
    {
      "epoch": 5.644654088050315,
      "grad_norm": 0.3924599587917328,
      "learning_rate": 4.511164604194598e-05,
      "loss": 0.4861,
      "num_input_tokens_seen": 7061744,
      "step": 10770
    },
    {
      "epoch": 5.64727463312369,
      "grad_norm": 0.4822172224521637,
      "learning_rate": 4.5104851985789956e-05,
      "loss": 0.3796,
      "num_input_tokens_seen": 7065744,
      "step": 10775
    },
    {
      "epoch": 5.649895178197065,
      "grad_norm": 0.4702568054199219,
      "learning_rate": 4.5098053723928254e-05,
      "loss": 0.3499,
      "num_input_tokens_seen": 7068688,
      "step": 10780
    },
    {
      "epoch": 5.65251572327044,
      "grad_norm": 0.26354798674583435,
      "learning_rate": 4.509125125778302e-05,
      "loss": 0.4111,
      "num_input_tokens_seen": 7071568,
      "step": 10785
    },
    {
      "epoch": 5.655136268343815,
      "grad_norm": 0.5315952897071838,
      "learning_rate": 4.508444458877722e-05,
      "loss": 0.5292,
      "num_input_tokens_seen": 7074384,
      "step": 10790
    },
    {
      "epoch": 5.6577568134171905,
      "grad_norm": 0.3762602210044861,
      "learning_rate": 4.5077633718334755e-05,
      "loss": 0.3729,
      "num_input_tokens_seen": 7077360,
      "step": 10795
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 0.4123159646987915,
      "learning_rate": 4.5070818647880375e-05,
      "loss": 0.5251,
      "num_input_tokens_seen": 7080400,
      "step": 10800
    },
    {
      "epoch": 5.662997903563941,
      "grad_norm": 0.554294228553772,
      "learning_rate": 4.506399937883972e-05,
      "loss": 0.4957,
      "num_input_tokens_seen": 7083344,
      "step": 10805
    },
    {
      "epoch": 5.665618448637317,
      "grad_norm": 0.39218324422836304,
      "learning_rate": 4.5057175912639286e-05,
      "loss": 0.6254,
      "num_input_tokens_seen": 7086352,
      "step": 10810
    },
    {
      "epoch": 5.668238993710692,
      "grad_norm": 0.9872912168502808,
      "learning_rate": 4.5050348250706475e-05,
      "loss": 0.5384,
      "num_input_tokens_seen": 7090480,
      "step": 10815
    },
    {
      "epoch": 5.670859538784067,
      "grad_norm": 0.41872912645339966,
      "learning_rate": 4.5043516394469566e-05,
      "loss": 0.5673,
      "num_input_tokens_seen": 7094768,
      "step": 10820
    },
    {
      "epoch": 5.673480083857442,
      "grad_norm": 0.3625549376010895,
      "learning_rate": 4.503668034535768e-05,
      "loss": 0.4899,
      "num_input_tokens_seen": 7098000,
      "step": 10825
    },
    {
      "epoch": 5.676100628930818,
      "grad_norm": 0.6270972490310669,
      "learning_rate": 4.5029840104800876e-05,
      "loss": 0.4786,
      "num_input_tokens_seen": 7100688,
      "step": 10830
    },
    {
      "epoch": 5.678721174004193,
      "grad_norm": 0.5155867338180542,
      "learning_rate": 4.5022995674230025e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 7103984,
      "step": 10835
    },
    {
      "epoch": 5.681341719077568,
      "grad_norm": 0.3772301971912384,
      "learning_rate": 4.5016147055076914e-05,
      "loss": 0.4783,
      "num_input_tokens_seen": 7107568,
      "step": 10840
    },
    {
      "epoch": 5.683962264150943,
      "grad_norm": 0.3710155189037323,
      "learning_rate": 4.500929424877419e-05,
      "loss": 0.3956,
      "num_input_tokens_seen": 7109680,
      "step": 10845
    },
    {
      "epoch": 5.686582809224319,
      "grad_norm": 0.3995896279811859,
      "learning_rate": 4.5002437256755394e-05,
      "loss": 0.5151,
      "num_input_tokens_seen": 7113232,
      "step": 10850
    },
    {
      "epoch": 5.689203354297694,
      "grad_norm": 1.0415834188461304,
      "learning_rate": 4.4995576080454924e-05,
      "loss": 0.5867,
      "num_input_tokens_seen": 7116176,
      "step": 10855
    },
    {
      "epoch": 5.6918238993710695,
      "grad_norm": 0.6837714910507202,
      "learning_rate": 4.4988710721308056e-05,
      "loss": 0.6389,
      "num_input_tokens_seen": 7118544,
      "step": 10860
    },
    {
      "epoch": 5.694444444444445,
      "grad_norm": 0.35668063163757324,
      "learning_rate": 4.498184118075095e-05,
      "loss": 0.4425,
      "num_input_tokens_seen": 7123152,
      "step": 10865
    },
    {
      "epoch": 5.69706498951782,
      "grad_norm": 0.28324395418167114,
      "learning_rate": 4.497496746022063e-05,
      "loss": 0.4787,
      "num_input_tokens_seen": 7126448,
      "step": 10870
    },
    {
      "epoch": 5.699685534591195,
      "grad_norm": 0.2753552496433258,
      "learning_rate": 4.4968089561155015e-05,
      "loss": 0.4567,
      "num_input_tokens_seen": 7129360,
      "step": 10875
    },
    {
      "epoch": 5.70230607966457,
      "grad_norm": 0.5095311403274536,
      "learning_rate": 4.496120748499286e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 7132432,
      "step": 10880
    },
    {
      "epoch": 5.704926624737945,
      "grad_norm": 0.34217649698257446,
      "learning_rate": 4.4954321233173833e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 7135056,
      "step": 10885
    },
    {
      "epoch": 5.7075471698113205,
      "grad_norm": 0.6248273253440857,
      "learning_rate": 4.494743080713846e-05,
      "loss": 0.5504,
      "num_input_tokens_seen": 7138160,
      "step": 10890
    },
    {
      "epoch": 5.710167714884696,
      "grad_norm": 0.5384497046470642,
      "learning_rate": 4.494053620832812e-05,
      "loss": 0.4718,
      "num_input_tokens_seen": 7141296,
      "step": 10895
    },
    {
      "epoch": 5.712788259958071,
      "grad_norm": 0.5769938826560974,
      "learning_rate": 4.493363743818511e-05,
      "loss": 0.3657,
      "num_input_tokens_seen": 7143984,
      "step": 10900
    },
    {
      "epoch": 5.715408805031447,
      "grad_norm": 0.5414590239524841,
      "learning_rate": 4.492673449815257e-05,
      "loss": 0.4831,
      "num_input_tokens_seen": 7146608,
      "step": 10905
    },
    {
      "epoch": 5.718029350104822,
      "grad_norm": 2.3776700496673584,
      "learning_rate": 4.4919827389674496e-05,
      "loss": 0.4669,
      "num_input_tokens_seen": 7149296,
      "step": 10910
    },
    {
      "epoch": 5.720649895178197,
      "grad_norm": 0.9014016389846802,
      "learning_rate": 4.491291611419579e-05,
      "loss": 0.5403,
      "num_input_tokens_seen": 7155440,
      "step": 10915
    },
    {
      "epoch": 5.723270440251572,
      "grad_norm": 0.3517223000526428,
      "learning_rate": 4.4906000673162216e-05,
      "loss": 0.5207,
      "num_input_tokens_seen": 7158768,
      "step": 10920
    },
    {
      "epoch": 5.725890985324948,
      "grad_norm": 0.48759058117866516,
      "learning_rate": 4.48990810680204e-05,
      "loss": 0.4396,
      "num_input_tokens_seen": 7162160,
      "step": 10925
    },
    {
      "epoch": 5.728511530398323,
      "grad_norm": 0.41872143745422363,
      "learning_rate": 4.4892157300217844e-05,
      "loss": 0.2986,
      "num_input_tokens_seen": 7165296,
      "step": 10930
    },
    {
      "epoch": 5.731132075471698,
      "grad_norm": 0.6474673748016357,
      "learning_rate": 4.488522937120293e-05,
      "loss": 0.5149,
      "num_input_tokens_seen": 7168912,
      "step": 10935
    },
    {
      "epoch": 5.733752620545073,
      "grad_norm": 0.6844433546066284,
      "learning_rate": 4.4878297282424886e-05,
      "loss": 0.4596,
      "num_input_tokens_seen": 7171536,
      "step": 10940
    },
    {
      "epoch": 5.736373165618449,
      "grad_norm": 0.4924286901950836,
      "learning_rate": 4.4871361035333836e-05,
      "loss": 0.4647,
      "num_input_tokens_seen": 7174800,
      "step": 10945
    },
    {
      "epoch": 5.738993710691824,
      "grad_norm": 0.3782116174697876,
      "learning_rate": 4.486442063138076e-05,
      "loss": 0.4352,
      "num_input_tokens_seen": 7177520,
      "step": 10950
    },
    {
      "epoch": 5.7416142557651995,
      "grad_norm": 0.7579432725906372,
      "learning_rate": 4.4857476072017516e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 7179728,
      "step": 10955
    },
    {
      "epoch": 5.744234800838575,
      "grad_norm": 0.3642421364784241,
      "learning_rate": 4.485052735869682e-05,
      "loss": 0.4536,
      "num_input_tokens_seen": 7183024,
      "step": 10960
    },
    {
      "epoch": 5.74685534591195,
      "grad_norm": 0.37105563282966614,
      "learning_rate": 4.484357449287228e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 7185712,
      "step": 10965
    },
    {
      "epoch": 5.749475890985325,
      "grad_norm": 0.4218185544013977,
      "learning_rate": 4.483661747599833e-05,
      "loss": 0.4266,
      "num_input_tokens_seen": 7190160,
      "step": 10970
    },
    {
      "epoch": 5.7520964360587,
      "grad_norm": 0.3294750452041626,
      "learning_rate": 4.482965630953031e-05,
      "loss": 0.4653,
      "num_input_tokens_seen": 7193392,
      "step": 10975
    },
    {
      "epoch": 5.754716981132075,
      "grad_norm": 0.5248281359672546,
      "learning_rate": 4.4822690994924425e-05,
      "loss": 0.4174,
      "num_input_tokens_seen": 7196784,
      "step": 10980
    },
    {
      "epoch": 5.7573375262054505,
      "grad_norm": 0.5718879699707031,
      "learning_rate": 4.481572153363773e-05,
      "loss": 0.6506,
      "num_input_tokens_seen": 7200400,
      "step": 10985
    },
    {
      "epoch": 5.759958071278826,
      "grad_norm": 0.6410782337188721,
      "learning_rate": 4.480874792712817e-05,
      "loss": 0.368,
      "num_input_tokens_seen": 7204048,
      "step": 10990
    },
    {
      "epoch": 5.762578616352201,
      "grad_norm": 0.23786459863185883,
      "learning_rate": 4.480177017685452e-05,
      "loss": 0.4537,
      "num_input_tokens_seen": 7207184,
      "step": 10995
    },
    {
      "epoch": 5.765199161425577,
      "grad_norm": 0.7134338617324829,
      "learning_rate": 4.479478828427647e-05,
      "loss": 0.3534,
      "num_input_tokens_seen": 7210640,
      "step": 11000
    },
    {
      "epoch": 5.767819706498952,
      "grad_norm": 0.39185118675231934,
      "learning_rate": 4.478780225085454e-05,
      "loss": 0.5009,
      "num_input_tokens_seen": 7214672,
      "step": 11005
    },
    {
      "epoch": 5.770440251572327,
      "grad_norm": 0.46453776955604553,
      "learning_rate": 4.4780812078050125e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 7217264,
      "step": 11010
    },
    {
      "epoch": 5.773060796645702,
      "grad_norm": 0.5267873406410217,
      "learning_rate": 4.47738177673255e-05,
      "loss": 0.4888,
      "num_input_tokens_seen": 7220848,
      "step": 11015
    },
    {
      "epoch": 5.7756813417190775,
      "grad_norm": 0.4306488037109375,
      "learning_rate": 4.4766819320143785e-05,
      "loss": 0.3899,
      "num_input_tokens_seen": 7223408,
      "step": 11020
    },
    {
      "epoch": 5.778301886792453,
      "grad_norm": 0.5009642839431763,
      "learning_rate": 4.475981673796899e-05,
      "loss": 0.5992,
      "num_input_tokens_seen": 7225744,
      "step": 11025
    },
    {
      "epoch": 5.780922431865828,
      "grad_norm": 0.5997614860534668,
      "learning_rate": 4.475281002226596e-05,
      "loss": 0.4842,
      "num_input_tokens_seen": 7228528,
      "step": 11030
    },
    {
      "epoch": 5.783542976939203,
      "grad_norm": 0.4517015814781189,
      "learning_rate": 4.4745799174500414e-05,
      "loss": 0.6131,
      "num_input_tokens_seen": 7231920,
      "step": 11035
    },
    {
      "epoch": 5.786163522012579,
      "grad_norm": 0.2872943878173828,
      "learning_rate": 4.473878419613896e-05,
      "loss": 0.3822,
      "num_input_tokens_seen": 7234864,
      "step": 11040
    },
    {
      "epoch": 5.788784067085954,
      "grad_norm": 0.6688061952590942,
      "learning_rate": 4.473176508864905e-05,
      "loss": 0.6186,
      "num_input_tokens_seen": 7237648,
      "step": 11045
    },
    {
      "epoch": 5.7914046121593294,
      "grad_norm": 0.4327128529548645,
      "learning_rate": 4.4724741853498995e-05,
      "loss": 0.5276,
      "num_input_tokens_seen": 7241232,
      "step": 11050
    },
    {
      "epoch": 5.794025157232705,
      "grad_norm": 0.42820295691490173,
      "learning_rate": 4.471771449215797e-05,
      "loss": 0.4633,
      "num_input_tokens_seen": 7245008,
      "step": 11055
    },
    {
      "epoch": 5.79664570230608,
      "grad_norm": 0.5562082529067993,
      "learning_rate": 4.471068300609603e-05,
      "loss": 0.5464,
      "num_input_tokens_seen": 7248368,
      "step": 11060
    },
    {
      "epoch": 5.799266247379455,
      "grad_norm": 0.3699032962322235,
      "learning_rate": 4.4703647396784065e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 7251920,
      "step": 11065
    },
    {
      "epoch": 5.80188679245283,
      "grad_norm": 0.3044646978378296,
      "learning_rate": 4.469660766569387e-05,
      "loss": 0.3551,
      "num_input_tokens_seen": 7254800,
      "step": 11070
    },
    {
      "epoch": 5.804507337526205,
      "grad_norm": 0.6089727878570557,
      "learning_rate": 4.4689563814298057e-05,
      "loss": 0.5206,
      "num_input_tokens_seen": 7257552,
      "step": 11075
    },
    {
      "epoch": 5.8071278825995805,
      "grad_norm": 0.5242647528648376,
      "learning_rate": 4.468251584407012e-05,
      "loss": 0.4425,
      "num_input_tokens_seen": 7261328,
      "step": 11080
    },
    {
      "epoch": 5.809748427672956,
      "grad_norm": 0.4405508041381836,
      "learning_rate": 4.467546375648443e-05,
      "loss": 0.488,
      "num_input_tokens_seen": 7264432,
      "step": 11085
    },
    {
      "epoch": 5.812368972746331,
      "grad_norm": 0.6051580905914307,
      "learning_rate": 4.4668407553016176e-05,
      "loss": 0.5646,
      "num_input_tokens_seen": 7266864,
      "step": 11090
    },
    {
      "epoch": 5.814989517819707,
      "grad_norm": 0.23821042478084564,
      "learning_rate": 4.466134723514146e-05,
      "loss": 0.6703,
      "num_input_tokens_seen": 7269936,
      "step": 11095
    },
    {
      "epoch": 5.817610062893082,
      "grad_norm": 0.5073774456977844,
      "learning_rate": 4.465428280433721e-05,
      "loss": 0.3628,
      "num_input_tokens_seen": 7273552,
      "step": 11100
    },
    {
      "epoch": 5.820230607966457,
      "grad_norm": 0.41686326265335083,
      "learning_rate": 4.464721426208123e-05,
      "loss": 0.4677,
      "num_input_tokens_seen": 7276752,
      "step": 11105
    },
    {
      "epoch": 5.822851153039832,
      "grad_norm": 0.7483634948730469,
      "learning_rate": 4.464014160985217e-05,
      "loss": 0.3791,
      "num_input_tokens_seen": 7278704,
      "step": 11110
    },
    {
      "epoch": 5.8254716981132075,
      "grad_norm": 0.5165201425552368,
      "learning_rate": 4.463306484912955e-05,
      "loss": 0.4876,
      "num_input_tokens_seen": 7281488,
      "step": 11115
    },
    {
      "epoch": 5.828092243186583,
      "grad_norm": 0.3409101068973541,
      "learning_rate": 4.462598398139375e-05,
      "loss": 0.4078,
      "num_input_tokens_seen": 7284624,
      "step": 11120
    },
    {
      "epoch": 5.830712788259958,
      "grad_norm": 0.3861719071865082,
      "learning_rate": 4.4618899008126005e-05,
      "loss": 0.3562,
      "num_input_tokens_seen": 7287664,
      "step": 11125
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 0.5628684163093567,
      "learning_rate": 4.461180993080841e-05,
      "loss": 0.4493,
      "num_input_tokens_seen": 7290992,
      "step": 11130
    },
    {
      "epoch": 5.835953878406709,
      "grad_norm": 0.5067611336708069,
      "learning_rate": 4.460471675092393e-05,
      "loss": 0.3928,
      "num_input_tokens_seen": 7294256,
      "step": 11135
    },
    {
      "epoch": 5.838574423480084,
      "grad_norm": 0.3601117730140686,
      "learning_rate": 4.4597619469956356e-05,
      "loss": 0.4391,
      "num_input_tokens_seen": 7297424,
      "step": 11140
    },
    {
      "epoch": 5.841194968553459,
      "grad_norm": 0.5847686529159546,
      "learning_rate": 4.459051808939038e-05,
      "loss": 0.4916,
      "num_input_tokens_seen": 7300592,
      "step": 11145
    },
    {
      "epoch": 5.843815513626835,
      "grad_norm": 0.330532968044281,
      "learning_rate": 4.458341261071151e-05,
      "loss": 0.4579,
      "num_input_tokens_seen": 7303568,
      "step": 11150
    },
    {
      "epoch": 5.84643605870021,
      "grad_norm": 0.3568776547908783,
      "learning_rate": 4.457630303540615e-05,
      "loss": 0.4091,
      "num_input_tokens_seen": 7306736,
      "step": 11155
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 0.3892345130443573,
      "learning_rate": 4.456918936496154e-05,
      "loss": 0.4293,
      "num_input_tokens_seen": 7309488,
      "step": 11160
    },
    {
      "epoch": 5.85167714884696,
      "grad_norm": 0.30173423886299133,
      "learning_rate": 4.456207160086575e-05,
      "loss": 0.4542,
      "num_input_tokens_seen": 7312400,
      "step": 11165
    },
    {
      "epoch": 5.854297693920335,
      "grad_norm": 0.3127512037754059,
      "learning_rate": 4.4554949744607775e-05,
      "loss": 0.409,
      "num_input_tokens_seen": 7316112,
      "step": 11170
    },
    {
      "epoch": 5.8569182389937104,
      "grad_norm": 0.37113434076309204,
      "learning_rate": 4.4547823797677405e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 7319216,
      "step": 11175
    },
    {
      "epoch": 5.859538784067086,
      "grad_norm": 0.36973631381988525,
      "learning_rate": 4.45406937615653e-05,
      "loss": 0.3966,
      "num_input_tokens_seen": 7322192,
      "step": 11180
    },
    {
      "epoch": 5.862159329140461,
      "grad_norm": 0.4991072118282318,
      "learning_rate": 4.4533559637763006e-05,
      "loss": 0.3827,
      "num_input_tokens_seen": 7325072,
      "step": 11185
    },
    {
      "epoch": 5.864779874213837,
      "grad_norm": 0.49998220801353455,
      "learning_rate": 4.452642142776288e-05,
      "loss": 0.5302,
      "num_input_tokens_seen": 7327600,
      "step": 11190
    },
    {
      "epoch": 5.867400419287212,
      "grad_norm": 0.5986753106117249,
      "learning_rate": 4.451927913305816e-05,
      "loss": 0.4892,
      "num_input_tokens_seen": 7330576,
      "step": 11195
    },
    {
      "epoch": 5.870020964360587,
      "grad_norm": 0.4964129626750946,
      "learning_rate": 4.4512132755142935e-05,
      "loss": 0.4642,
      "num_input_tokens_seen": 7333360,
      "step": 11200
    },
    {
      "epoch": 5.872641509433962,
      "grad_norm": 0.49186986684799194,
      "learning_rate": 4.450498229551214e-05,
      "loss": 0.4937,
      "num_input_tokens_seen": 7336656,
      "step": 11205
    },
    {
      "epoch": 5.8752620545073375,
      "grad_norm": 0.4864296615123749,
      "learning_rate": 4.449782775566158e-05,
      "loss": 0.4238,
      "num_input_tokens_seen": 7341008,
      "step": 11210
    },
    {
      "epoch": 5.877882599580713,
      "grad_norm": 0.3689470887184143,
      "learning_rate": 4.449066913708789e-05,
      "loss": 0.4931,
      "num_input_tokens_seen": 7343856,
      "step": 11215
    },
    {
      "epoch": 5.880503144654088,
      "grad_norm": 0.326610803604126,
      "learning_rate": 4.448350644128858e-05,
      "loss": 0.6394,
      "num_input_tokens_seen": 7347056,
      "step": 11220
    },
    {
      "epoch": 5.883123689727463,
      "grad_norm": 0.3819367587566376,
      "learning_rate": 4.4476339669762e-05,
      "loss": 0.4306,
      "num_input_tokens_seen": 7350320,
      "step": 11225
    },
    {
      "epoch": 5.885744234800838,
      "grad_norm": 0.4454681873321533,
      "learning_rate": 4.446916882400737e-05,
      "loss": 0.4903,
      "num_input_tokens_seen": 7353392,
      "step": 11230
    },
    {
      "epoch": 5.888364779874214,
      "grad_norm": 0.3809988796710968,
      "learning_rate": 4.446199390552472e-05,
      "loss": 0.4182,
      "num_input_tokens_seen": 7356400,
      "step": 11235
    },
    {
      "epoch": 5.890985324947589,
      "grad_norm": 0.28227731585502625,
      "learning_rate": 4.445481491581499e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 7359216,
      "step": 11240
    },
    {
      "epoch": 5.893605870020965,
      "grad_norm": 0.39745181798934937,
      "learning_rate": 4.444763185637993e-05,
      "loss": 0.4122,
      "num_input_tokens_seen": 7361904,
      "step": 11245
    },
    {
      "epoch": 5.89622641509434,
      "grad_norm": 0.3826032280921936,
      "learning_rate": 4.4440444728722166e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 7365040,
      "step": 11250
    },
    {
      "epoch": 5.898846960167715,
      "grad_norm": 0.2842246890068054,
      "learning_rate": 4.4433253534345136e-05,
      "loss": 0.4107,
      "num_input_tokens_seen": 7368368,
      "step": 11255
    },
    {
      "epoch": 5.90146750524109,
      "grad_norm": 0.37999609112739563,
      "learning_rate": 4.442605827475318e-05,
      "loss": 0.4835,
      "num_input_tokens_seen": 7370960,
      "step": 11260
    },
    {
      "epoch": 5.904088050314465,
      "grad_norm": 0.3376215696334839,
      "learning_rate": 4.4418858951451457e-05,
      "loss": 0.3682,
      "num_input_tokens_seen": 7374192,
      "step": 11265
    },
    {
      "epoch": 5.90670859538784,
      "grad_norm": 0.49391746520996094,
      "learning_rate": 4.441165556594599e-05,
      "loss": 0.5478,
      "num_input_tokens_seen": 7377456,
      "step": 11270
    },
    {
      "epoch": 5.909329140461216,
      "grad_norm": 0.3868251144886017,
      "learning_rate": 4.440444811974363e-05,
      "loss": 0.5685,
      "num_input_tokens_seen": 7380688,
      "step": 11275
    },
    {
      "epoch": 5.911949685534591,
      "grad_norm": 0.560699999332428,
      "learning_rate": 4.43972366143521e-05,
      "loss": 0.5661,
      "num_input_tokens_seen": 7384112,
      "step": 11280
    },
    {
      "epoch": 5.914570230607967,
      "grad_norm": 0.3566477596759796,
      "learning_rate": 4.4390021051279975e-05,
      "loss": 0.5256,
      "num_input_tokens_seen": 7388240,
      "step": 11285
    },
    {
      "epoch": 5.917190775681342,
      "grad_norm": 0.425085186958313,
      "learning_rate": 4.438280143203665e-05,
      "loss": 0.4583,
      "num_input_tokens_seen": 7391600,
      "step": 11290
    },
    {
      "epoch": 5.919811320754717,
      "grad_norm": 0.5801260471343994,
      "learning_rate": 4.43755777581324e-05,
      "loss": 0.5004,
      "num_input_tokens_seen": 7395184,
      "step": 11295
    },
    {
      "epoch": 5.922431865828092,
      "grad_norm": 0.3365832269191742,
      "learning_rate": 4.4368350031078334e-05,
      "loss": 0.5876,
      "num_input_tokens_seen": 7398032,
      "step": 11300
    },
    {
      "epoch": 5.9250524109014675,
      "grad_norm": 0.7054300308227539,
      "learning_rate": 4.43611182523864e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 7401360,
      "step": 11305
    },
    {
      "epoch": 5.927672955974843,
      "grad_norm": 0.5825355648994446,
      "learning_rate": 4.4353882423569416e-05,
      "loss": 0.4878,
      "num_input_tokens_seen": 7404112,
      "step": 11310
    },
    {
      "epoch": 5.930293501048218,
      "grad_norm": 0.46643322706222534,
      "learning_rate": 4.434664254614103e-05,
      "loss": 0.4234,
      "num_input_tokens_seen": 7407056,
      "step": 11315
    },
    {
      "epoch": 5.932914046121593,
      "grad_norm": 0.8167864084243774,
      "learning_rate": 4.4339398621615734e-05,
      "loss": 0.5201,
      "num_input_tokens_seen": 7410288,
      "step": 11320
    },
    {
      "epoch": 5.935534591194968,
      "grad_norm": 0.37808147072792053,
      "learning_rate": 4.433215065150889e-05,
      "loss": 0.5151,
      "num_input_tokens_seen": 7413520,
      "step": 11325
    },
    {
      "epoch": 5.938155136268344,
      "grad_norm": 0.40990152955055237,
      "learning_rate": 4.4324898637336674e-05,
      "loss": 0.4389,
      "num_input_tokens_seen": 7415920,
      "step": 11330
    },
    {
      "epoch": 5.940775681341719,
      "grad_norm": 0.48730549216270447,
      "learning_rate": 4.431764258061615e-05,
      "loss": 0.3596,
      "num_input_tokens_seen": 7419056,
      "step": 11335
    },
    {
      "epoch": 5.943396226415095,
      "grad_norm": 0.4889754056930542,
      "learning_rate": 4.431038248286517e-05,
      "loss": 0.5602,
      "num_input_tokens_seen": 7423216,
      "step": 11340
    },
    {
      "epoch": 5.94601677148847,
      "grad_norm": 0.6836706399917603,
      "learning_rate": 4.430311834560248e-05,
      "loss": 0.4826,
      "num_input_tokens_seen": 7426352,
      "step": 11345
    },
    {
      "epoch": 5.948637316561845,
      "grad_norm": 0.42309170961380005,
      "learning_rate": 4.4295850170347664e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 7429712,
      "step": 11350
    },
    {
      "epoch": 5.95125786163522,
      "grad_norm": 0.27360835671424866,
      "learning_rate": 4.428857795862113e-05,
      "loss": 0.4289,
      "num_input_tokens_seen": 7432912,
      "step": 11355
    },
    {
      "epoch": 5.953878406708595,
      "grad_norm": 0.2674335539340973,
      "learning_rate": 4.428130171194414e-05,
      "loss": 0.3758,
      "num_input_tokens_seen": 7436592,
      "step": 11360
    },
    {
      "epoch": 5.95649895178197,
      "grad_norm": 0.3572365641593933,
      "learning_rate": 4.427402143183879e-05,
      "loss": 0.3355,
      "num_input_tokens_seen": 7439696,
      "step": 11365
    },
    {
      "epoch": 5.959119496855346,
      "grad_norm": 0.6052301526069641,
      "learning_rate": 4.426673711982806e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 7442704,
      "step": 11370
    },
    {
      "epoch": 5.961740041928721,
      "grad_norm": 0.35959699749946594,
      "learning_rate": 4.425944877743573e-05,
      "loss": 0.5342,
      "num_input_tokens_seen": 7446224,
      "step": 11375
    },
    {
      "epoch": 5.964360587002097,
      "grad_norm": 0.423391193151474,
      "learning_rate": 4.425215640618644e-05,
      "loss": 0.5474,
      "num_input_tokens_seen": 7449008,
      "step": 11380
    },
    {
      "epoch": 5.966981132075472,
      "grad_norm": 0.2913306951522827,
      "learning_rate": 4.424486000760567e-05,
      "loss": 0.325,
      "num_input_tokens_seen": 7453008,
      "step": 11385
    },
    {
      "epoch": 5.969601677148847,
      "grad_norm": 0.396420419216156,
      "learning_rate": 4.423755958321974e-05,
      "loss": 0.5046,
      "num_input_tokens_seen": 7456144,
      "step": 11390
    },
    {
      "epoch": 5.972222222222222,
      "grad_norm": 0.6367719173431396,
      "learning_rate": 4.423025513455582e-05,
      "loss": 0.4443,
      "num_input_tokens_seen": 7459504,
      "step": 11395
    },
    {
      "epoch": 5.9748427672955975,
      "grad_norm": 0.31932708621025085,
      "learning_rate": 4.422294666314192e-05,
      "loss": 0.5129,
      "num_input_tokens_seen": 7463760,
      "step": 11400
    },
    {
      "epoch": 5.977463312368973,
      "grad_norm": 0.3186148703098297,
      "learning_rate": 4.421563417050688e-05,
      "loss": 0.3611,
      "num_input_tokens_seen": 7467280,
      "step": 11405
    },
    {
      "epoch": 5.980083857442348,
      "grad_norm": 0.44478556513786316,
      "learning_rate": 4.42083176581804e-05,
      "loss": 0.3831,
      "num_input_tokens_seen": 7470096,
      "step": 11410
    },
    {
      "epoch": 5.982704402515723,
      "grad_norm": 0.48826277256011963,
      "learning_rate": 4.4200997127693e-05,
      "loss": 0.4632,
      "num_input_tokens_seen": 7472624,
      "step": 11415
    },
    {
      "epoch": 5.985324947589098,
      "grad_norm": 0.39160460233688354,
      "learning_rate": 4.419367258057606e-05,
      "loss": 0.5342,
      "num_input_tokens_seen": 7475760,
      "step": 11420
    },
    {
      "epoch": 5.987945492662474,
      "grad_norm": 0.6365559101104736,
      "learning_rate": 4.418634401836179e-05,
      "loss": 0.5533,
      "num_input_tokens_seen": 7478480,
      "step": 11425
    },
    {
      "epoch": 5.990566037735849,
      "grad_norm": 0.3613879382610321,
      "learning_rate": 4.417901144258324e-05,
      "loss": 0.4187,
      "num_input_tokens_seen": 7481616,
      "step": 11430
    },
    {
      "epoch": 5.993186582809225,
      "grad_norm": 0.33682337403297424,
      "learning_rate": 4.4171674854774305e-05,
      "loss": 0.3576,
      "num_input_tokens_seen": 7484560,
      "step": 11435
    },
    {
      "epoch": 5.9958071278826,
      "grad_norm": 0.32838815450668335,
      "learning_rate": 4.416433425646971e-05,
      "loss": 0.4673,
      "num_input_tokens_seen": 7488048,
      "step": 11440
    },
    {
      "epoch": 5.998427672955975,
      "grad_norm": 0.3863224983215332,
      "learning_rate": 4.415698964920503e-05,
      "loss": 0.5159,
      "num_input_tokens_seen": 7490672,
      "step": 11445
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.48420068621635437,
      "eval_runtime": 9.6325,
      "eval_samples_per_second": 88.035,
      "eval_steps_per_second": 22.009,
      "num_input_tokens_seen": 7491776,
      "step": 11448
    },
    {
      "epoch": 6.00104821802935,
      "grad_norm": 0.34973010420799255,
      "learning_rate": 4.414964103451667e-05,
      "loss": 0.3849,
      "num_input_tokens_seen": 7493024,
      "step": 11450
    },
    {
      "epoch": 6.003668763102725,
      "grad_norm": 0.3573627471923828,
      "learning_rate": 4.414228841394187e-05,
      "loss": 0.5322,
      "num_input_tokens_seen": 7496448,
      "step": 11455
    },
    {
      "epoch": 6.0062893081761,
      "grad_norm": 0.3968656063079834,
      "learning_rate": 4.413493178901872e-05,
      "loss": 0.4789,
      "num_input_tokens_seen": 7499584,
      "step": 11460
    },
    {
      "epoch": 6.008909853249476,
      "grad_norm": 0.8073579668998718,
      "learning_rate": 4.412757116128614e-05,
      "loss": 0.4315,
      "num_input_tokens_seen": 7502432,
      "step": 11465
    },
    {
      "epoch": 6.011530398322851,
      "grad_norm": 0.4667278230190277,
      "learning_rate": 4.41202065322839e-05,
      "loss": 0.5456,
      "num_input_tokens_seen": 7506176,
      "step": 11470
    },
    {
      "epoch": 6.014150943396227,
      "grad_norm": 0.23983383178710938,
      "learning_rate": 4.411283790355259e-05,
      "loss": 0.3351,
      "num_input_tokens_seen": 7509248,
      "step": 11475
    },
    {
      "epoch": 6.016771488469602,
      "grad_norm": 0.36556950211524963,
      "learning_rate": 4.410546527663364e-05,
      "loss": 0.5699,
      "num_input_tokens_seen": 7512320,
      "step": 11480
    },
    {
      "epoch": 6.019392033542977,
      "grad_norm": 0.41542401909828186,
      "learning_rate": 4.409808865306931e-05,
      "loss": 0.581,
      "num_input_tokens_seen": 7515232,
      "step": 11485
    },
    {
      "epoch": 6.022012578616352,
      "grad_norm": 0.507300615310669,
      "learning_rate": 4.409070803440273e-05,
      "loss": 0.3655,
      "num_input_tokens_seen": 7518752,
      "step": 11490
    },
    {
      "epoch": 6.0246331236897275,
      "grad_norm": 0.32913345098495483,
      "learning_rate": 4.4083323422177824e-05,
      "loss": 0.4502,
      "num_input_tokens_seen": 7522880,
      "step": 11495
    },
    {
      "epoch": 6.027253668763103,
      "grad_norm": 0.269456148147583,
      "learning_rate": 4.407593481793937e-05,
      "loss": 0.3112,
      "num_input_tokens_seen": 7525952,
      "step": 11500
    },
    {
      "epoch": 6.029874213836478,
      "grad_norm": 0.2832964062690735,
      "learning_rate": 4.406854222323298e-05,
      "loss": 0.5558,
      "num_input_tokens_seen": 7529088,
      "step": 11505
    },
    {
      "epoch": 6.032494758909853,
      "grad_norm": 0.5411089658737183,
      "learning_rate": 4.406114563960511e-05,
      "loss": 0.5555,
      "num_input_tokens_seen": 7535616,
      "step": 11510
    },
    {
      "epoch": 6.035115303983228,
      "grad_norm": 0.30980566143989563,
      "learning_rate": 4.4053745068603026e-05,
      "loss": 0.4583,
      "num_input_tokens_seen": 7538240,
      "step": 11515
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 0.29258427023887634,
      "learning_rate": 4.404634051177485e-05,
      "loss": 0.5442,
      "num_input_tokens_seen": 7541312,
      "step": 11520
    },
    {
      "epoch": 6.040356394129979,
      "grad_norm": 0.5369760394096375,
      "learning_rate": 4.403893197066953e-05,
      "loss": 0.4342,
      "num_input_tokens_seen": 7544160,
      "step": 11525
    },
    {
      "epoch": 6.0429769392033545,
      "grad_norm": 0.355415940284729,
      "learning_rate": 4.4031519446836855e-05,
      "loss": 0.5232,
      "num_input_tokens_seen": 7546976,
      "step": 11530
    },
    {
      "epoch": 6.04559748427673,
      "grad_norm": 0.5008522272109985,
      "learning_rate": 4.402410294182743e-05,
      "loss": 0.4107,
      "num_input_tokens_seen": 7549536,
      "step": 11535
    },
    {
      "epoch": 6.048218029350105,
      "grad_norm": 0.2010369598865509,
      "learning_rate": 4.4016682457192714e-05,
      "loss": 0.5362,
      "num_input_tokens_seen": 7552608,
      "step": 11540
    },
    {
      "epoch": 6.05083857442348,
      "grad_norm": 0.42406558990478516,
      "learning_rate": 4.400925799448497e-05,
      "loss": 0.4026,
      "num_input_tokens_seen": 7556352,
      "step": 11545
    },
    {
      "epoch": 6.053459119496855,
      "grad_norm": 0.41197195649147034,
      "learning_rate": 4.400182955525733e-05,
      "loss": 0.4476,
      "num_input_tokens_seen": 7559072,
      "step": 11550
    },
    {
      "epoch": 6.05607966457023,
      "grad_norm": 0.5126252174377441,
      "learning_rate": 4.3994397141063734e-05,
      "loss": 0.6607,
      "num_input_tokens_seen": 7562080,
      "step": 11555
    },
    {
      "epoch": 6.058700209643606,
      "grad_norm": 0.46078744530677795,
      "learning_rate": 4.398696075345896e-05,
      "loss": 0.4079,
      "num_input_tokens_seen": 7564704,
      "step": 11560
    },
    {
      "epoch": 6.061320754716981,
      "grad_norm": 0.5357528924942017,
      "learning_rate": 4.39795203939986e-05,
      "loss": 0.4762,
      "num_input_tokens_seen": 7567776,
      "step": 11565
    },
    {
      "epoch": 6.063941299790357,
      "grad_norm": 0.5725979804992676,
      "learning_rate": 4.397207606423911e-05,
      "loss": 0.4343,
      "num_input_tokens_seen": 7570976,
      "step": 11570
    },
    {
      "epoch": 6.066561844863732,
      "grad_norm": 0.2061680555343628,
      "learning_rate": 4.396462776573775e-05,
      "loss": 0.3471,
      "num_input_tokens_seen": 7575168,
      "step": 11575
    },
    {
      "epoch": 6.069182389937107,
      "grad_norm": 0.3492540419101715,
      "learning_rate": 4.395717550005262e-05,
      "loss": 0.4277,
      "num_input_tokens_seen": 7577952,
      "step": 11580
    },
    {
      "epoch": 6.071802935010482,
      "grad_norm": 0.3237641155719757,
      "learning_rate": 4.394971926874265e-05,
      "loss": 0.494,
      "num_input_tokens_seen": 7580512,
      "step": 11585
    },
    {
      "epoch": 6.0744234800838575,
      "grad_norm": 0.4894042909145355,
      "learning_rate": 4.3942259073367604e-05,
      "loss": 0.2938,
      "num_input_tokens_seen": 7584032,
      "step": 11590
    },
    {
      "epoch": 6.077044025157233,
      "grad_norm": 0.43998661637306213,
      "learning_rate": 4.3934794915488065e-05,
      "loss": 0.434,
      "num_input_tokens_seen": 7587680,
      "step": 11595
    },
    {
      "epoch": 6.079664570230608,
      "grad_norm": 0.2862246632575989,
      "learning_rate": 4.392732679666544e-05,
      "loss": 0.3962,
      "num_input_tokens_seen": 7590816,
      "step": 11600
    },
    {
      "epoch": 6.082285115303983,
      "grad_norm": 0.5922172665596008,
      "learning_rate": 4.3919854718461986e-05,
      "loss": 0.5475,
      "num_input_tokens_seen": 7594016,
      "step": 11605
    },
    {
      "epoch": 6.084905660377358,
      "grad_norm": 0.5000983476638794,
      "learning_rate": 4.391237868244078e-05,
      "loss": 0.542,
      "num_input_tokens_seen": 7597344,
      "step": 11610
    },
    {
      "epoch": 6.087526205450734,
      "grad_norm": 0.47284454107284546,
      "learning_rate": 4.390489869016571e-05,
      "loss": 0.521,
      "num_input_tokens_seen": 7600128,
      "step": 11615
    },
    {
      "epoch": 6.090146750524109,
      "grad_norm": 1.1563544273376465,
      "learning_rate": 4.389741474320152e-05,
      "loss": 0.4876,
      "num_input_tokens_seen": 7602816,
      "step": 11620
    },
    {
      "epoch": 6.0927672955974845,
      "grad_norm": 0.6254989504814148,
      "learning_rate": 4.388992684311376e-05,
      "loss": 0.3908,
      "num_input_tokens_seen": 7605696,
      "step": 11625
    },
    {
      "epoch": 6.09538784067086,
      "grad_norm": 0.743889331817627,
      "learning_rate": 4.388243499146881e-05,
      "loss": 0.5805,
      "num_input_tokens_seen": 7608672,
      "step": 11630
    },
    {
      "epoch": 6.098008385744235,
      "grad_norm": 0.8798679709434509,
      "learning_rate": 4.3874939189833876e-05,
      "loss": 0.5313,
      "num_input_tokens_seen": 7611936,
      "step": 11635
    },
    {
      "epoch": 6.10062893081761,
      "grad_norm": 0.37760061025619507,
      "learning_rate": 4.3867439439777005e-05,
      "loss": 0.6248,
      "num_input_tokens_seen": 7614496,
      "step": 11640
    },
    {
      "epoch": 6.103249475890985,
      "grad_norm": 0.4431547224521637,
      "learning_rate": 4.385993574286705e-05,
      "loss": 0.4377,
      "num_input_tokens_seen": 7617536,
      "step": 11645
    },
    {
      "epoch": 6.10587002096436,
      "grad_norm": 0.6588441729545593,
      "learning_rate": 4.3852428100673705e-05,
      "loss": 0.4589,
      "num_input_tokens_seen": 7620448,
      "step": 11650
    },
    {
      "epoch": 6.1084905660377355,
      "grad_norm": 0.5116970539093018,
      "learning_rate": 4.3844916514767484e-05,
      "loss": 0.5213,
      "num_input_tokens_seen": 7622592,
      "step": 11655
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 0.44492107629776,
      "learning_rate": 4.383740098671972e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 7626304,
      "step": 11660
    },
    {
      "epoch": 6.113731656184487,
      "grad_norm": 0.3272726535797119,
      "learning_rate": 4.3829881518102576e-05,
      "loss": 0.4466,
      "num_input_tokens_seen": 7630624,
      "step": 11665
    },
    {
      "epoch": 6.116352201257862,
      "grad_norm": 0.625187873840332,
      "learning_rate": 4.3822358110489045e-05,
      "loss": 0.4706,
      "num_input_tokens_seen": 7634848,
      "step": 11670
    },
    {
      "epoch": 6.118972746331237,
      "grad_norm": 0.300449013710022,
      "learning_rate": 4.381483076545294e-05,
      "loss": 0.3684,
      "num_input_tokens_seen": 7639712,
      "step": 11675
    },
    {
      "epoch": 6.121593291404612,
      "grad_norm": 0.6354237198829651,
      "learning_rate": 4.380729948456888e-05,
      "loss": 0.5081,
      "num_input_tokens_seen": 7642784,
      "step": 11680
    },
    {
      "epoch": 6.1242138364779874,
      "grad_norm": 1.0698033571243286,
      "learning_rate": 4.3799764269412336e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 7645568,
      "step": 11685
    },
    {
      "epoch": 6.126834381551363,
      "grad_norm": 0.8429985046386719,
      "learning_rate": 4.379222512155958e-05,
      "loss": 0.5454,
      "num_input_tokens_seen": 7648352,
      "step": 11690
    },
    {
      "epoch": 6.129454926624738,
      "grad_norm": 0.26073378324508667,
      "learning_rate": 4.378468204258773e-05,
      "loss": 0.4733,
      "num_input_tokens_seen": 7651840,
      "step": 11695
    },
    {
      "epoch": 6.132075471698113,
      "grad_norm": 0.5229993462562561,
      "learning_rate": 4.37771350340747e-05,
      "loss": 0.3929,
      "num_input_tokens_seen": 7655328,
      "step": 11700
    },
    {
      "epoch": 6.134696016771488,
      "grad_norm": 0.5764406323432922,
      "learning_rate": 4.376958409759925e-05,
      "loss": 0.5699,
      "num_input_tokens_seen": 7658464,
      "step": 11705
    },
    {
      "epoch": 6.137316561844864,
      "grad_norm": 0.35979756712913513,
      "learning_rate": 4.376202923474093e-05,
      "loss": 0.5727,
      "num_input_tokens_seen": 7661568,
      "step": 11710
    },
    {
      "epoch": 6.139937106918239,
      "grad_norm": 0.3499952256679535,
      "learning_rate": 4.3754470447080145e-05,
      "loss": 0.4317,
      "num_input_tokens_seen": 7665312,
      "step": 11715
    },
    {
      "epoch": 6.1425576519916145,
      "grad_norm": 0.5311763882637024,
      "learning_rate": 4.374690773619811e-05,
      "loss": 0.4641,
      "num_input_tokens_seen": 7668000,
      "step": 11720
    },
    {
      "epoch": 6.14517819706499,
      "grad_norm": 0.338386207818985,
      "learning_rate": 4.373934110367685e-05,
      "loss": 0.3339,
      "num_input_tokens_seen": 7671936,
      "step": 11725
    },
    {
      "epoch": 6.147798742138365,
      "grad_norm": 0.4940475821495056,
      "learning_rate": 4.373177055109922e-05,
      "loss": 0.3915,
      "num_input_tokens_seen": 7675520,
      "step": 11730
    },
    {
      "epoch": 6.15041928721174,
      "grad_norm": 0.5612630248069763,
      "learning_rate": 4.3724196080048896e-05,
      "loss": 0.3709,
      "num_input_tokens_seen": 7678720,
      "step": 11735
    },
    {
      "epoch": 6.153039832285115,
      "grad_norm": 0.5023675560951233,
      "learning_rate": 4.3716617692110364e-05,
      "loss": 0.4861,
      "num_input_tokens_seen": 7681856,
      "step": 11740
    },
    {
      "epoch": 6.15566037735849,
      "grad_norm": 0.6287562251091003,
      "learning_rate": 4.3709035388868943e-05,
      "loss": 0.3512,
      "num_input_tokens_seen": 7685824,
      "step": 11745
    },
    {
      "epoch": 6.1582809224318655,
      "grad_norm": 0.61770099401474,
      "learning_rate": 4.3701449171910767e-05,
      "loss": 0.3869,
      "num_input_tokens_seen": 7689056,
      "step": 11750
    },
    {
      "epoch": 6.160901467505241,
      "grad_norm": 0.2751685678958893,
      "learning_rate": 4.3693859042822774e-05,
      "loss": 0.4337,
      "num_input_tokens_seen": 7691744,
      "step": 11755
    },
    {
      "epoch": 6.163522012578617,
      "grad_norm": 0.34696561098098755,
      "learning_rate": 4.368626500319275e-05,
      "loss": 0.399,
      "num_input_tokens_seen": 7696416,
      "step": 11760
    },
    {
      "epoch": 6.166142557651992,
      "grad_norm": 0.3652190864086151,
      "learning_rate": 4.367866705460927e-05,
      "loss": 0.4972,
      "num_input_tokens_seen": 7699232,
      "step": 11765
    },
    {
      "epoch": 6.168763102725367,
      "grad_norm": 0.3041223883628845,
      "learning_rate": 4.367106519866173e-05,
      "loss": 0.5153,
      "num_input_tokens_seen": 7703168,
      "step": 11770
    },
    {
      "epoch": 6.171383647798742,
      "grad_norm": 0.4985767900943756,
      "learning_rate": 4.366345943694037e-05,
      "loss": 0.3639,
      "num_input_tokens_seen": 7706912,
      "step": 11775
    },
    {
      "epoch": 6.174004192872117,
      "grad_norm": 0.5348847508430481,
      "learning_rate": 4.365584977103622e-05,
      "loss": 0.5116,
      "num_input_tokens_seen": 7709504,
      "step": 11780
    },
    {
      "epoch": 6.176624737945493,
      "grad_norm": 0.6405344605445862,
      "learning_rate": 4.364823620254113e-05,
      "loss": 0.4351,
      "num_input_tokens_seen": 7711872,
      "step": 11785
    },
    {
      "epoch": 6.179245283018868,
      "grad_norm": 0.438869833946228,
      "learning_rate": 4.364061873304778e-05,
      "loss": 0.3517,
      "num_input_tokens_seen": 7714912,
      "step": 11790
    },
    {
      "epoch": 6.181865828092243,
      "grad_norm": 0.36842080950737,
      "learning_rate": 4.363299736414965e-05,
      "loss": 0.4875,
      "num_input_tokens_seen": 7718304,
      "step": 11795
    },
    {
      "epoch": 6.184486373165618,
      "grad_norm": 0.7612301707267761,
      "learning_rate": 4.3625372097441045e-05,
      "loss": 0.5093,
      "num_input_tokens_seen": 7721024,
      "step": 11800
    },
    {
      "epoch": 6.187106918238993,
      "grad_norm": 0.4064621329307556,
      "learning_rate": 4.3617742934517094e-05,
      "loss": 0.3461,
      "num_input_tokens_seen": 7723872,
      "step": 11805
    },
    {
      "epoch": 6.189727463312369,
      "grad_norm": 0.4439699053764343,
      "learning_rate": 4.361010987697373e-05,
      "loss": 0.4313,
      "num_input_tokens_seen": 7726688,
      "step": 11810
    },
    {
      "epoch": 6.1923480083857445,
      "grad_norm": 0.6924927234649658,
      "learning_rate": 4.360247292640768e-05,
      "loss": 0.4468,
      "num_input_tokens_seen": 7728992,
      "step": 11815
    },
    {
      "epoch": 6.19496855345912,
      "grad_norm": 0.3970593810081482,
      "learning_rate": 4.359483208441653e-05,
      "loss": 0.482,
      "num_input_tokens_seen": 7732096,
      "step": 11820
    },
    {
      "epoch": 6.197589098532495,
      "grad_norm": 0.5448378920555115,
      "learning_rate": 4.3587187352598646e-05,
      "loss": 0.587,
      "num_input_tokens_seen": 7734624,
      "step": 11825
    },
    {
      "epoch": 6.20020964360587,
      "grad_norm": 0.31036314368247986,
      "learning_rate": 4.357953873255323e-05,
      "loss": 0.3835,
      "num_input_tokens_seen": 7737792,
      "step": 11830
    },
    {
      "epoch": 6.202830188679245,
      "grad_norm": 0.7416166067123413,
      "learning_rate": 4.357188622588027e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 7740864,
      "step": 11835
    },
    {
      "epoch": 6.20545073375262,
      "grad_norm": 0.3940972089767456,
      "learning_rate": 4.356422983418059e-05,
      "loss": 0.3685,
      "num_input_tokens_seen": 7744288,
      "step": 11840
    },
    {
      "epoch": 6.2080712788259955,
      "grad_norm": 0.40744590759277344,
      "learning_rate": 4.3556569559055824e-05,
      "loss": 0.4742,
      "num_input_tokens_seen": 7747328,
      "step": 11845
    },
    {
      "epoch": 6.210691823899371,
      "grad_norm": 0.6830626726150513,
      "learning_rate": 4.3548905402108406e-05,
      "loss": 0.4316,
      "num_input_tokens_seen": 7750304,
      "step": 11850
    },
    {
      "epoch": 6.213312368972747,
      "grad_norm": 0.5794103145599365,
      "learning_rate": 4.35412373649416e-05,
      "loss": 0.3258,
      "num_input_tokens_seen": 7753088,
      "step": 11855
    },
    {
      "epoch": 6.215932914046122,
      "grad_norm": 0.32586270570755005,
      "learning_rate": 4.353356544915947e-05,
      "loss": 0.4134,
      "num_input_tokens_seen": 7756384,
      "step": 11860
    },
    {
      "epoch": 6.218553459119497,
      "grad_norm": 0.3310903012752533,
      "learning_rate": 4.352588965636689e-05,
      "loss": 0.3519,
      "num_input_tokens_seen": 7759200,
      "step": 11865
    },
    {
      "epoch": 6.221174004192872,
      "grad_norm": 0.6685644388198853,
      "learning_rate": 4.351820998816954e-05,
      "loss": 0.5496,
      "num_input_tokens_seen": 7761952,
      "step": 11870
    },
    {
      "epoch": 6.223794549266247,
      "grad_norm": 0.3033061921596527,
      "learning_rate": 4.3510526446173935e-05,
      "loss": 0.3192,
      "num_input_tokens_seen": 7765088,
      "step": 11875
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 0.5897775292396545,
      "learning_rate": 4.350283903198737e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 7767776,
      "step": 11880
    },
    {
      "epoch": 6.229035639412998,
      "grad_norm": 0.2978099584579468,
      "learning_rate": 4.349514774721797e-05,
      "loss": 0.4828,
      "num_input_tokens_seen": 7770240,
      "step": 11885
    },
    {
      "epoch": 6.231656184486373,
      "grad_norm": 0.8868696689605713,
      "learning_rate": 4.348745259347467e-05,
      "loss": 0.4601,
      "num_input_tokens_seen": 7773024,
      "step": 11890
    },
    {
      "epoch": 6.234276729559748,
      "grad_norm": 0.5800454616546631,
      "learning_rate": 4.347975357236721e-05,
      "loss": 0.3555,
      "num_input_tokens_seen": 7775648,
      "step": 11895
    },
    {
      "epoch": 6.236897274633123,
      "grad_norm": 0.7324624061584473,
      "learning_rate": 4.347205068550612e-05,
      "loss": 0.5038,
      "num_input_tokens_seen": 7779168,
      "step": 11900
    },
    {
      "epoch": 6.239517819706499,
      "grad_norm": 0.6125400066375732,
      "learning_rate": 4.3464343934502774e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 7783264,
      "step": 11905
    },
    {
      "epoch": 6.2421383647798745,
      "grad_norm": 0.4851604402065277,
      "learning_rate": 4.345663332096932e-05,
      "loss": 0.5659,
      "num_input_tokens_seen": 7785920,
      "step": 11910
    },
    {
      "epoch": 6.24475890985325,
      "grad_norm": 0.45074012875556946,
      "learning_rate": 4.344891884651875e-05,
      "loss": 0.5577,
      "num_input_tokens_seen": 7788640,
      "step": 11915
    },
    {
      "epoch": 6.247379454926625,
      "grad_norm": 0.5080593824386597,
      "learning_rate": 4.344120051276483e-05,
      "loss": 0.5438,
      "num_input_tokens_seen": 7792736,
      "step": 11920
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.7100957036018372,
      "learning_rate": 4.3433478321322155e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 7796128,
      "step": 11925
    },
    {
      "epoch": 6.252620545073375,
      "grad_norm": 1.2055319547653198,
      "learning_rate": 4.3425752273806106e-05,
      "loss": 0.7299,
      "num_input_tokens_seen": 7799488,
      "step": 11930
    },
    {
      "epoch": 6.25524109014675,
      "grad_norm": 0.6646183133125305,
      "learning_rate": 4.34180223718329e-05,
      "loss": 0.5408,
      "num_input_tokens_seen": 7803264,
      "step": 11935
    },
    {
      "epoch": 6.2578616352201255,
      "grad_norm": 0.656345009803772,
      "learning_rate": 4.341028861701953e-05,
      "loss": 0.5273,
      "num_input_tokens_seen": 7805920,
      "step": 11940
    },
    {
      "epoch": 6.260482180293501,
      "grad_norm": 0.41039952635765076,
      "learning_rate": 4.3402551010983825e-05,
      "loss": 0.3989,
      "num_input_tokens_seen": 7809728,
      "step": 11945
    },
    {
      "epoch": 6.263102725366876,
      "grad_norm": 0.53815758228302,
      "learning_rate": 4.33948095553444e-05,
      "loss": 0.5069,
      "num_input_tokens_seen": 7813056,
      "step": 11950
    },
    {
      "epoch": 6.265723270440252,
      "grad_norm": 0.312420517206192,
      "learning_rate": 4.338706425172067e-05,
      "loss": 0.5791,
      "num_input_tokens_seen": 7815904,
      "step": 11955
    },
    {
      "epoch": 6.268343815513627,
      "grad_norm": 0.6056864857673645,
      "learning_rate": 4.3379315101732866e-05,
      "loss": 0.5823,
      "num_input_tokens_seen": 7819744,
      "step": 11960
    },
    {
      "epoch": 6.270964360587002,
      "grad_norm": 0.25682833790779114,
      "learning_rate": 4.3371562107002026e-05,
      "loss": 0.4108,
      "num_input_tokens_seen": 7823104,
      "step": 11965
    },
    {
      "epoch": 6.273584905660377,
      "grad_norm": 0.8446240425109863,
      "learning_rate": 4.3363805269149984e-05,
      "loss": 0.4428,
      "num_input_tokens_seen": 7825728,
      "step": 11970
    },
    {
      "epoch": 6.276205450733753,
      "grad_norm": 0.5289098024368286,
      "learning_rate": 4.335604458979939e-05,
      "loss": 0.5076,
      "num_input_tokens_seen": 7828288,
      "step": 11975
    },
    {
      "epoch": 6.278825995807128,
      "grad_norm": 0.2692432999610901,
      "learning_rate": 4.3348280070573686e-05,
      "loss": 0.5353,
      "num_input_tokens_seen": 7830912,
      "step": 11980
    },
    {
      "epoch": 6.281446540880503,
      "grad_norm": 0.5740308165550232,
      "learning_rate": 4.334051171309712e-05,
      "loss": 0.358,
      "num_input_tokens_seen": 7833728,
      "step": 11985
    },
    {
      "epoch": 6.284067085953878,
      "grad_norm": 0.31257691979408264,
      "learning_rate": 4.333273951899474e-05,
      "loss": 0.4726,
      "num_input_tokens_seen": 7837632,
      "step": 11990
    },
    {
      "epoch": 6.286687631027253,
      "grad_norm": 0.6341008543968201,
      "learning_rate": 4.33249634898924e-05,
      "loss": 0.3638,
      "num_input_tokens_seen": 7840672,
      "step": 11995
    },
    {
      "epoch": 6.289308176100629,
      "grad_norm": 0.34350523352622986,
      "learning_rate": 4.331718362741677e-05,
      "loss": 0.5204,
      "num_input_tokens_seen": 7843488,
      "step": 12000
    },
    {
      "epoch": 6.2919287211740045,
      "grad_norm": 0.5110820531845093,
      "learning_rate": 4.330939993319529e-05,
      "loss": 0.6461,
      "num_input_tokens_seen": 7845952,
      "step": 12005
    },
    {
      "epoch": 6.29454926624738,
      "grad_norm": 1.1734557151794434,
      "learning_rate": 4.330161240885623e-05,
      "loss": 0.4668,
      "num_input_tokens_seen": 7849088,
      "step": 12010
    },
    {
      "epoch": 6.297169811320755,
      "grad_norm": 0.23176738619804382,
      "learning_rate": 4.3293821056028654e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 7853792,
      "step": 12015
    },
    {
      "epoch": 6.29979035639413,
      "grad_norm": 0.6459108591079712,
      "learning_rate": 4.3286025876342415e-05,
      "loss": 0.5723,
      "num_input_tokens_seen": 7856448,
      "step": 12020
    },
    {
      "epoch": 6.302410901467505,
      "grad_norm": 0.36337101459503174,
      "learning_rate": 4.327822687142819e-05,
      "loss": 0.4288,
      "num_input_tokens_seen": 7859200,
      "step": 12025
    },
    {
      "epoch": 6.30503144654088,
      "grad_norm": 0.29390832781791687,
      "learning_rate": 4.327042404291742e-05,
      "loss": 0.6019,
      "num_input_tokens_seen": 7864064,
      "step": 12030
    },
    {
      "epoch": 6.3076519916142555,
      "grad_norm": 0.4638126492500305,
      "learning_rate": 4.326261739244239e-05,
      "loss": 0.3797,
      "num_input_tokens_seen": 7867552,
      "step": 12035
    },
    {
      "epoch": 6.310272536687631,
      "grad_norm": 0.4431093633174896,
      "learning_rate": 4.325480692163615e-05,
      "loss": 0.4337,
      "num_input_tokens_seen": 7871360,
      "step": 12040
    },
    {
      "epoch": 6.312893081761006,
      "grad_norm": 0.39185792207717896,
      "learning_rate": 4.324699263213256e-05,
      "loss": 0.5068,
      "num_input_tokens_seen": 7874656,
      "step": 12045
    },
    {
      "epoch": 6.315513626834382,
      "grad_norm": 0.35950446128845215,
      "learning_rate": 4.3239174525566285e-05,
      "loss": 0.4744,
      "num_input_tokens_seen": 7877792,
      "step": 12050
    },
    {
      "epoch": 6.318134171907757,
      "grad_norm": 0.4661211371421814,
      "learning_rate": 4.3231352603572784e-05,
      "loss": 0.4156,
      "num_input_tokens_seen": 7881760,
      "step": 12055
    },
    {
      "epoch": 6.320754716981132,
      "grad_norm": 0.6017948389053345,
      "learning_rate": 4.322352686778832e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 7887968,
      "step": 12060
    },
    {
      "epoch": 6.323375262054507,
      "grad_norm": 0.475221186876297,
      "learning_rate": 4.3215697319849926e-05,
      "loss": 0.4368,
      "num_input_tokens_seen": 7892832,
      "step": 12065
    },
    {
      "epoch": 6.325995807127883,
      "grad_norm": 0.349998414516449,
      "learning_rate": 4.3207863961395476e-05,
      "loss": 0.5845,
      "num_input_tokens_seen": 7896512,
      "step": 12070
    },
    {
      "epoch": 6.328616352201258,
      "grad_norm": 0.534155547618866,
      "learning_rate": 4.320002679406361e-05,
      "loss": 0.4635,
      "num_input_tokens_seen": 7898720,
      "step": 12075
    },
    {
      "epoch": 6.331236897274633,
      "grad_norm": 0.5545229911804199,
      "learning_rate": 4.319218581949378e-05,
      "loss": 0.4099,
      "num_input_tokens_seen": 7902528,
      "step": 12080
    },
    {
      "epoch": 6.333857442348008,
      "grad_norm": 0.5004413723945618,
      "learning_rate": 4.318434103932622e-05,
      "loss": 0.5378,
      "num_input_tokens_seen": 7906336,
      "step": 12085
    },
    {
      "epoch": 6.336477987421383,
      "grad_norm": 0.4820014536380768,
      "learning_rate": 4.317649245520197e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 7910240,
      "step": 12090
    },
    {
      "epoch": 6.339098532494759,
      "grad_norm": 0.48425501585006714,
      "learning_rate": 4.3168640068762875e-05,
      "loss": 0.4105,
      "num_input_tokens_seen": 7912800,
      "step": 12095
    },
    {
      "epoch": 6.3417190775681345,
      "grad_norm": 0.3325752019882202,
      "learning_rate": 4.316078388165155e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 7916544,
      "step": 12100
    },
    {
      "epoch": 6.34433962264151,
      "grad_norm": 0.3787189722061157,
      "learning_rate": 4.315292389551143e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 7920192,
      "step": 12105
    },
    {
      "epoch": 6.346960167714885,
      "grad_norm": 0.43200910091400146,
      "learning_rate": 4.3145060111986726e-05,
      "loss": 0.4636,
      "num_input_tokens_seen": 7923264,
      "step": 12110
    },
    {
      "epoch": 6.34958071278826,
      "grad_norm": 0.464191734790802,
      "learning_rate": 4.313719253272246e-05,
      "loss": 0.3648,
      "num_input_tokens_seen": 7926080,
      "step": 12115
    },
    {
      "epoch": 6.352201257861635,
      "grad_norm": 0.48634129762649536,
      "learning_rate": 4.3129321159364436e-05,
      "loss": 0.38,
      "num_input_tokens_seen": 7929120,
      "step": 12120
    },
    {
      "epoch": 6.35482180293501,
      "grad_norm": 0.7746837139129639,
      "learning_rate": 4.312144599355925e-05,
      "loss": 0.4611,
      "num_input_tokens_seen": 7931968,
      "step": 12125
    },
    {
      "epoch": 6.3574423480083855,
      "grad_norm": 0.4088068902492523,
      "learning_rate": 4.31135670369543e-05,
      "loss": 0.4772,
      "num_input_tokens_seen": 7935232,
      "step": 12130
    },
    {
      "epoch": 6.360062893081761,
      "grad_norm": 0.38585150241851807,
      "learning_rate": 4.310568429119779e-05,
      "loss": 0.5221,
      "num_input_tokens_seen": 7939840,
      "step": 12135
    },
    {
      "epoch": 6.362683438155136,
      "grad_norm": 1.191784381866455,
      "learning_rate": 4.309779775793868e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 7942752,
      "step": 12140
    },
    {
      "epoch": 6.365303983228512,
      "grad_norm": 0.31143951416015625,
      "learning_rate": 4.308990743882675e-05,
      "loss": 0.5333,
      "num_input_tokens_seen": 7947552,
      "step": 12145
    },
    {
      "epoch": 6.367924528301887,
      "grad_norm": 0.7271791696548462,
      "learning_rate": 4.308201333551256e-05,
      "loss": 0.5656,
      "num_input_tokens_seen": 7951424,
      "step": 12150
    },
    {
      "epoch": 6.370545073375262,
      "grad_norm": 0.28351226449012756,
      "learning_rate": 4.3074115449647476e-05,
      "loss": 0.5633,
      "num_input_tokens_seen": 7954560,
      "step": 12155
    },
    {
      "epoch": 6.373165618448637,
      "grad_norm": 0.3472062051296234,
      "learning_rate": 4.306621378288364e-05,
      "loss": 0.4848,
      "num_input_tokens_seen": 7959616,
      "step": 12160
    },
    {
      "epoch": 6.3757861635220126,
      "grad_norm": 0.3891472816467285,
      "learning_rate": 4.305830833687399e-05,
      "loss": 0.4298,
      "num_input_tokens_seen": 7962912,
      "step": 12165
    },
    {
      "epoch": 6.378406708595388,
      "grad_norm": 0.4378274083137512,
      "learning_rate": 4.305039911327225e-05,
      "loss": 0.4496,
      "num_input_tokens_seen": 7966656,
      "step": 12170
    },
    {
      "epoch": 6.381027253668763,
      "grad_norm": 1.475053071975708,
      "learning_rate": 4.304248611373296e-05,
      "loss": 0.7387,
      "num_input_tokens_seen": 7969280,
      "step": 12175
    },
    {
      "epoch": 6.383647798742138,
      "grad_norm": 0.5630595684051514,
      "learning_rate": 4.30345693399114e-05,
      "loss": 0.5044,
      "num_input_tokens_seen": 7972352,
      "step": 12180
    },
    {
      "epoch": 6.386268343815513,
      "grad_norm": 0.6892962455749512,
      "learning_rate": 4.30266487934637e-05,
      "loss": 0.4273,
      "num_input_tokens_seen": 7975136,
      "step": 12185
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 0.7911378741264343,
      "learning_rate": 4.3018724476046714e-05,
      "loss": 0.5084,
      "num_input_tokens_seen": 7977632,
      "step": 12190
    },
    {
      "epoch": 6.3915094339622645,
      "grad_norm": 0.33009639382362366,
      "learning_rate": 4.301079638931814e-05,
      "loss": 0.4554,
      "num_input_tokens_seen": 7980960,
      "step": 12195
    },
    {
      "epoch": 6.39412997903564,
      "grad_norm": 0.6353901028633118,
      "learning_rate": 4.3002864534936444e-05,
      "loss": 0.4803,
      "num_input_tokens_seen": 7983776,
      "step": 12200
    },
    {
      "epoch": 6.396750524109015,
      "grad_norm": 0.7143780589103699,
      "learning_rate": 4.2994928914560874e-05,
      "loss": 0.4568,
      "num_input_tokens_seen": 7986752,
      "step": 12205
    },
    {
      "epoch": 6.39937106918239,
      "grad_norm": 0.5575690865516663,
      "learning_rate": 4.298698952985147e-05,
      "loss": 0.449,
      "num_input_tokens_seen": 7989728,
      "step": 12210
    },
    {
      "epoch": 6.401991614255765,
      "grad_norm": 0.47728490829467773,
      "learning_rate": 4.297904638246906e-05,
      "loss": 0.5694,
      "num_input_tokens_seen": 7992384,
      "step": 12215
    },
    {
      "epoch": 6.40461215932914,
      "grad_norm": 0.9240630269050598,
      "learning_rate": 4.297109947407527e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 7994784,
      "step": 12220
    },
    {
      "epoch": 6.4072327044025155,
      "grad_norm": 0.40133631229400635,
      "learning_rate": 4.2963148806332486e-05,
      "loss": 0.3937,
      "num_input_tokens_seen": 7998976,
      "step": 12225
    },
    {
      "epoch": 6.409853249475891,
      "grad_norm": 0.5471147298812866,
      "learning_rate": 4.2955194380903915e-05,
      "loss": 0.5231,
      "num_input_tokens_seen": 8001728,
      "step": 12230
    },
    {
      "epoch": 6.412473794549266,
      "grad_norm": 0.5098037719726562,
      "learning_rate": 4.2947236199453524e-05,
      "loss": 0.6965,
      "num_input_tokens_seen": 8004800,
      "step": 12235
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 0.29422739148139954,
      "learning_rate": 4.2939274263646065e-05,
      "loss": 0.3346,
      "num_input_tokens_seen": 8007872,
      "step": 12240
    },
    {
      "epoch": 6.417714884696017,
      "grad_norm": 0.3313255310058594,
      "learning_rate": 4.29313085751471e-05,
      "loss": 0.5111,
      "num_input_tokens_seen": 8010944,
      "step": 12245
    },
    {
      "epoch": 6.420335429769392,
      "grad_norm": 0.7525586485862732,
      "learning_rate": 4.292333913562294e-05,
      "loss": 0.4158,
      "num_input_tokens_seen": 8013312,
      "step": 12250
    },
    {
      "epoch": 6.422955974842767,
      "grad_norm": 0.44079577922821045,
      "learning_rate": 4.291536594674073e-05,
      "loss": 0.5859,
      "num_input_tokens_seen": 8015808,
      "step": 12255
    },
    {
      "epoch": 6.4255765199161425,
      "grad_norm": 0.4928062856197357,
      "learning_rate": 4.2907389010168344e-05,
      "loss": 0.5281,
      "num_input_tokens_seen": 8019552,
      "step": 12260
    },
    {
      "epoch": 6.428197064989518,
      "grad_norm": 0.3285909593105316,
      "learning_rate": 4.2899408327574475e-05,
      "loss": 0.381,
      "num_input_tokens_seen": 8023296,
      "step": 12265
    },
    {
      "epoch": 6.430817610062893,
      "grad_norm": 0.37604930996894836,
      "learning_rate": 4.289142390062859e-05,
      "loss": 0.5509,
      "num_input_tokens_seen": 8025952,
      "step": 12270
    },
    {
      "epoch": 6.433438155136268,
      "grad_norm": 0.5643285512924194,
      "learning_rate": 4.2883435731000946e-05,
      "loss": 0.4992,
      "num_input_tokens_seen": 8028928,
      "step": 12275
    },
    {
      "epoch": 6.436058700209643,
      "grad_norm": 0.4176575541496277,
      "learning_rate": 4.287544382036256e-05,
      "loss": 0.3916,
      "num_input_tokens_seen": 8031808,
      "step": 12280
    },
    {
      "epoch": 6.438679245283019,
      "grad_norm": 0.47940102219581604,
      "learning_rate": 4.286744817038527e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 8037728,
      "step": 12285
    },
    {
      "epoch": 6.441299790356394,
      "grad_norm": 0.6098403334617615,
      "learning_rate": 4.2859448782741666e-05,
      "loss": 0.3928,
      "num_input_tokens_seen": 8040544,
      "step": 12290
    },
    {
      "epoch": 6.44392033542977,
      "grad_norm": 0.3236217200756073,
      "learning_rate": 4.2851445659105126e-05,
      "loss": 0.4409,
      "num_input_tokens_seen": 8043616,
      "step": 12295
    },
    {
      "epoch": 6.446540880503145,
      "grad_norm": 0.6307655572891235,
      "learning_rate": 4.284343880114982e-05,
      "loss": 0.4611,
      "num_input_tokens_seen": 8046816,
      "step": 12300
    },
    {
      "epoch": 6.44916142557652,
      "grad_norm": 0.3584336042404175,
      "learning_rate": 4.2835428210550676e-05,
      "loss": 0.3933,
      "num_input_tokens_seen": 8050016,
      "step": 12305
    },
    {
      "epoch": 6.451781970649895,
      "grad_norm": 0.3349487781524658,
      "learning_rate": 4.282741388898344e-05,
      "loss": 0.4014,
      "num_input_tokens_seen": 8053888,
      "step": 12310
    },
    {
      "epoch": 6.45440251572327,
      "grad_norm": 0.33885741233825684,
      "learning_rate": 4.281939583812459e-05,
      "loss": 0.5162,
      "num_input_tokens_seen": 8057280,
      "step": 12315
    },
    {
      "epoch": 6.4570230607966455,
      "grad_norm": 0.47469136118888855,
      "learning_rate": 4.281137405965143e-05,
      "loss": 0.4002,
      "num_input_tokens_seen": 8060768,
      "step": 12320
    },
    {
      "epoch": 6.459643605870021,
      "grad_norm": 0.48809072375297546,
      "learning_rate": 4.280334855524203e-05,
      "loss": 0.4464,
      "num_input_tokens_seen": 8064576,
      "step": 12325
    },
    {
      "epoch": 6.462264150943396,
      "grad_norm": 0.6712958216667175,
      "learning_rate": 4.279531932657521e-05,
      "loss": 0.4335,
      "num_input_tokens_seen": 8067808,
      "step": 12330
    },
    {
      "epoch": 6.464884696016772,
      "grad_norm": 0.38184455037117004,
      "learning_rate": 4.278728637533062e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 8071680,
      "step": 12335
    },
    {
      "epoch": 6.467505241090147,
      "grad_norm": 0.5447492003440857,
      "learning_rate": 4.2779249703188626e-05,
      "loss": 0.4484,
      "num_input_tokens_seen": 8074880,
      "step": 12340
    },
    {
      "epoch": 6.470125786163522,
      "grad_norm": 0.5430605411529541,
      "learning_rate": 4.277120931183045e-05,
      "loss": 0.3824,
      "num_input_tokens_seen": 8080192,
      "step": 12345
    },
    {
      "epoch": 6.472746331236897,
      "grad_norm": 0.4355144500732422,
      "learning_rate": 4.276316520293803e-05,
      "loss": 0.5198,
      "num_input_tokens_seen": 8082784,
      "step": 12350
    },
    {
      "epoch": 6.4753668763102725,
      "grad_norm": 0.399528443813324,
      "learning_rate": 4.27551173781941e-05,
      "loss": 0.4923,
      "num_input_tokens_seen": 8085216,
      "step": 12355
    },
    {
      "epoch": 6.477987421383648,
      "grad_norm": 0.35409578680992126,
      "learning_rate": 4.274706583928216e-05,
      "loss": 0.3496,
      "num_input_tokens_seen": 8087968,
      "step": 12360
    },
    {
      "epoch": 6.480607966457023,
      "grad_norm": 0.9978720545768738,
      "learning_rate": 4.273901058788654e-05,
      "loss": 0.4025,
      "num_input_tokens_seen": 8091200,
      "step": 12365
    },
    {
      "epoch": 6.483228511530398,
      "grad_norm": 0.3424592912197113,
      "learning_rate": 4.2730951625692275e-05,
      "loss": 0.4626,
      "num_input_tokens_seen": 8094464,
      "step": 12370
    },
    {
      "epoch": 6.485849056603773,
      "grad_norm": 0.5741156339645386,
      "learning_rate": 4.27228889543852e-05,
      "loss": 0.491,
      "num_input_tokens_seen": 8097216,
      "step": 12375
    },
    {
      "epoch": 6.488469601677149,
      "grad_norm": 0.2924659550189972,
      "learning_rate": 4.2714822575651966e-05,
      "loss": 0.4259,
      "num_input_tokens_seen": 8100672,
      "step": 12380
    },
    {
      "epoch": 6.491090146750524,
      "grad_norm": 0.4201439619064331,
      "learning_rate": 4.270675249117994e-05,
      "loss": 0.512,
      "num_input_tokens_seen": 8103808,
      "step": 12385
    },
    {
      "epoch": 6.4937106918239,
      "grad_norm": 0.613565981388092,
      "learning_rate": 4.269867870265731e-05,
      "loss": 0.4988,
      "num_input_tokens_seen": 8106496,
      "step": 12390
    },
    {
      "epoch": 6.496331236897275,
      "grad_norm": 0.30647164583206177,
      "learning_rate": 4.269060121177301e-05,
      "loss": 0.4636,
      "num_input_tokens_seen": 8109888,
      "step": 12395
    },
    {
      "epoch": 6.49895178197065,
      "grad_norm": 0.33196720480918884,
      "learning_rate": 4.2682520020216764e-05,
      "loss": 0.3519,
      "num_input_tokens_seen": 8113056,
      "step": 12400
    },
    {
      "epoch": 6.501572327044025,
      "grad_norm": 0.3697899580001831,
      "learning_rate": 4.267443512967906e-05,
      "loss": 0.3862,
      "num_input_tokens_seen": 8116384,
      "step": 12405
    },
    {
      "epoch": 6.5041928721174,
      "grad_norm": 0.5134750008583069,
      "learning_rate": 4.2666346541851165e-05,
      "loss": 0.4531,
      "num_input_tokens_seen": 8119680,
      "step": 12410
    },
    {
      "epoch": 6.506813417190775,
      "grad_norm": 1.1285370588302612,
      "learning_rate": 4.265825425842512e-05,
      "loss": 0.4787,
      "num_input_tokens_seen": 8123520,
      "step": 12415
    },
    {
      "epoch": 6.509433962264151,
      "grad_norm": 0.4280617833137512,
      "learning_rate": 4.265015828109375e-05,
      "loss": 0.4275,
      "num_input_tokens_seen": 8126464,
      "step": 12420
    },
    {
      "epoch": 6.512054507337526,
      "grad_norm": 0.7504937052726746,
      "learning_rate": 4.264205861155063e-05,
      "loss": 0.5857,
      "num_input_tokens_seen": 8129728,
      "step": 12425
    },
    {
      "epoch": 6.514675052410902,
      "grad_norm": 0.6328748464584351,
      "learning_rate": 4.263395525149011e-05,
      "loss": 0.5917,
      "num_input_tokens_seen": 8132352,
      "step": 12430
    },
    {
      "epoch": 6.517295597484277,
      "grad_norm": 0.3421046733856201,
      "learning_rate": 4.262584820260734e-05,
      "loss": 0.4364,
      "num_input_tokens_seen": 8135328,
      "step": 12435
    },
    {
      "epoch": 6.519916142557652,
      "grad_norm": 0.4234776198863983,
      "learning_rate": 4.261773746659821e-05,
      "loss": 0.4641,
      "num_input_tokens_seen": 8137920,
      "step": 12440
    },
    {
      "epoch": 6.522536687631027,
      "grad_norm": 0.4500305652618408,
      "learning_rate": 4.260962304515939e-05,
      "loss": 0.5526,
      "num_input_tokens_seen": 8141152,
      "step": 12445
    },
    {
      "epoch": 6.5251572327044025,
      "grad_norm": 0.41974562406539917,
      "learning_rate": 4.2601504939988335e-05,
      "loss": 0.5199,
      "num_input_tokens_seen": 8143808,
      "step": 12450
    },
    {
      "epoch": 6.527777777777778,
      "grad_norm": 0.24208799004554749,
      "learning_rate": 4.259338315278325e-05,
      "loss": 0.4149,
      "num_input_tokens_seen": 8147392,
      "step": 12455
    },
    {
      "epoch": 6.530398322851153,
      "grad_norm": 0.5184647440910339,
      "learning_rate": 4.258525768524313e-05,
      "loss": 0.6888,
      "num_input_tokens_seen": 8150720,
      "step": 12460
    },
    {
      "epoch": 6.533018867924528,
      "grad_norm": 0.2495000660419464,
      "learning_rate": 4.2577128539067725e-05,
      "loss": 0.4846,
      "num_input_tokens_seen": 8153792,
      "step": 12465
    },
    {
      "epoch": 6.535639412997903,
      "grad_norm": 0.5596246123313904,
      "learning_rate": 4.256899571595756e-05,
      "loss": 0.3528,
      "num_input_tokens_seen": 8157184,
      "step": 12470
    },
    {
      "epoch": 6.538259958071279,
      "grad_norm": 0.43710318207740784,
      "learning_rate": 4.256085921761393e-05,
      "loss": 0.6151,
      "num_input_tokens_seen": 8160480,
      "step": 12475
    },
    {
      "epoch": 6.540880503144654,
      "grad_norm": 0.7053130269050598,
      "learning_rate": 4.25527190457389e-05,
      "loss": 0.5874,
      "num_input_tokens_seen": 8163776,
      "step": 12480
    },
    {
      "epoch": 6.54350104821803,
      "grad_norm": 0.4238725006580353,
      "learning_rate": 4.2544575202035285e-05,
      "loss": 0.5072,
      "num_input_tokens_seen": 8166912,
      "step": 12485
    },
    {
      "epoch": 6.546121593291405,
      "grad_norm": 0.4675327241420746,
      "learning_rate": 4.253642768820671e-05,
      "loss": 0.3904,
      "num_input_tokens_seen": 8169664,
      "step": 12490
    },
    {
      "epoch": 6.54874213836478,
      "grad_norm": 0.4880945086479187,
      "learning_rate": 4.252827650595752e-05,
      "loss": 0.4738,
      "num_input_tokens_seen": 8172768,
      "step": 12495
    },
    {
      "epoch": 6.551362683438155,
      "grad_norm": 0.30056875944137573,
      "learning_rate": 4.252012165699286e-05,
      "loss": 0.5183,
      "num_input_tokens_seen": 8176064,
      "step": 12500
    },
    {
      "epoch": 6.55398322851153,
      "grad_norm": 0.609900712966919,
      "learning_rate": 4.2511963143018634e-05,
      "loss": 0.4747,
      "num_input_tokens_seen": 8179840,
      "step": 12505
    },
    {
      "epoch": 6.556603773584905,
      "grad_norm": 0.6490330100059509,
      "learning_rate": 4.25038009657415e-05,
      "loss": 0.3378,
      "num_input_tokens_seen": 8183840,
      "step": 12510
    },
    {
      "epoch": 6.559224318658281,
      "grad_norm": 0.27596116065979004,
      "learning_rate": 4.24956351268689e-05,
      "loss": 0.4236,
      "num_input_tokens_seen": 8187456,
      "step": 12515
    },
    {
      "epoch": 6.561844863731656,
      "grad_norm": 0.5139244198799133,
      "learning_rate": 4.2487465628109036e-05,
      "loss": 0.3892,
      "num_input_tokens_seen": 8190528,
      "step": 12520
    },
    {
      "epoch": 6.564465408805032,
      "grad_norm": 0.6452215313911438,
      "learning_rate": 4.247929247117087e-05,
      "loss": 0.607,
      "num_input_tokens_seen": 8193344,
      "step": 12525
    },
    {
      "epoch": 6.567085953878407,
      "grad_norm": 0.9572401642799377,
      "learning_rate": 4.247111565776413e-05,
      "loss": 0.4374,
      "num_input_tokens_seen": 8195680,
      "step": 12530
    },
    {
      "epoch": 6.569706498951782,
      "grad_norm": 0.3041152358055115,
      "learning_rate": 4.246293518959932e-05,
      "loss": 0.5059,
      "num_input_tokens_seen": 8198656,
      "step": 12535
    },
    {
      "epoch": 6.572327044025157,
      "grad_norm": 0.6740264892578125,
      "learning_rate": 4.245475106838769e-05,
      "loss": 0.4542,
      "num_input_tokens_seen": 8201408,
      "step": 12540
    },
    {
      "epoch": 6.5749475890985325,
      "grad_norm": 0.49470701813697815,
      "learning_rate": 4.2446563295841266e-05,
      "loss": 0.562,
      "num_input_tokens_seen": 8204832,
      "step": 12545
    },
    {
      "epoch": 6.577568134171908,
      "grad_norm": 0.23889094591140747,
      "learning_rate": 4.243837187367285e-05,
      "loss": 0.5217,
      "num_input_tokens_seen": 8209088,
      "step": 12550
    },
    {
      "epoch": 6.580188679245283,
      "grad_norm": 0.33476194739341736,
      "learning_rate": 4.2430176803595987e-05,
      "loss": 0.3825,
      "num_input_tokens_seen": 8211744,
      "step": 12555
    },
    {
      "epoch": 6.582809224318658,
      "grad_norm": 0.4657536447048187,
      "learning_rate": 4.2421978087324985e-05,
      "loss": 0.3564,
      "num_input_tokens_seen": 8214176,
      "step": 12560
    },
    {
      "epoch": 6.585429769392033,
      "grad_norm": 0.673004150390625,
      "learning_rate": 4.241377572657493e-05,
      "loss": 0.3031,
      "num_input_tokens_seen": 8217056,
      "step": 12565
    },
    {
      "epoch": 6.588050314465409,
      "grad_norm": 0.5113753080368042,
      "learning_rate": 4.2405569723061654e-05,
      "loss": 0.433,
      "num_input_tokens_seen": 8219200,
      "step": 12570
    },
    {
      "epoch": 6.590670859538784,
      "grad_norm": 0.2512945532798767,
      "learning_rate": 4.239736007850178e-05,
      "loss": 0.49,
      "num_input_tokens_seen": 8222080,
      "step": 12575
    },
    {
      "epoch": 6.59329140461216,
      "grad_norm": 0.4061591625213623,
      "learning_rate": 4.238914679461264e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 8224960,
      "step": 12580
    },
    {
      "epoch": 6.595911949685535,
      "grad_norm": 0.25901108980178833,
      "learning_rate": 4.2380929873112386e-05,
      "loss": 0.569,
      "num_input_tokens_seen": 8227936,
      "step": 12585
    },
    {
      "epoch": 6.59853249475891,
      "grad_norm": 0.4192439317703247,
      "learning_rate": 4.2372709315719896e-05,
      "loss": 0.5095,
      "num_input_tokens_seen": 8231648,
      "step": 12590
    },
    {
      "epoch": 6.601153039832285,
      "grad_norm": 0.38626351952552795,
      "learning_rate": 4.2364485124154806e-05,
      "loss": 0.3903,
      "num_input_tokens_seen": 8234848,
      "step": 12595
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 0.5306388735771179,
      "learning_rate": 4.235625730013754e-05,
      "loss": 0.434,
      "num_input_tokens_seen": 8237536,
      "step": 12600
    },
    {
      "epoch": 6.606394129979035,
      "grad_norm": 1.0864008665084839,
      "learning_rate": 4.234802584538926e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 8240064,
      "step": 12605
    },
    {
      "epoch": 6.609014675052411,
      "grad_norm": 0.43036362528800964,
      "learning_rate": 4.2339790761631884e-05,
      "loss": 0.5811,
      "num_input_tokens_seen": 8243360,
      "step": 12610
    },
    {
      "epoch": 6.611635220125786,
      "grad_norm": 0.4911383092403412,
      "learning_rate": 4.233155205058811e-05,
      "loss": 0.5584,
      "num_input_tokens_seen": 8247648,
      "step": 12615
    },
    {
      "epoch": 6.614255765199162,
      "grad_norm": 0.42250221967697144,
      "learning_rate": 4.232330971398137e-05,
      "loss": 0.4188,
      "num_input_tokens_seen": 8251328,
      "step": 12620
    },
    {
      "epoch": 6.616876310272537,
      "grad_norm": 0.24020107090473175,
      "learning_rate": 4.2315063753535875e-05,
      "loss": 0.4562,
      "num_input_tokens_seen": 8254688,
      "step": 12625
    },
    {
      "epoch": 6.619496855345912,
      "grad_norm": 0.48896121978759766,
      "learning_rate": 4.230681417097658e-05,
      "loss": 0.5764,
      "num_input_tokens_seen": 8258144,
      "step": 12630
    },
    {
      "epoch": 6.622117400419287,
      "grad_norm": 0.3536459505558014,
      "learning_rate": 4.229856096802921e-05,
      "loss": 0.3776,
      "num_input_tokens_seen": 8261856,
      "step": 12635
    },
    {
      "epoch": 6.6247379454926625,
      "grad_norm": 0.4669049084186554,
      "learning_rate": 4.229030414642024e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 8265120,
      "step": 12640
    },
    {
      "epoch": 6.627358490566038,
      "grad_norm": 0.3162456750869751,
      "learning_rate": 4.2282043707876896e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 8267936,
      "step": 12645
    },
    {
      "epoch": 6.629979035639413,
      "grad_norm": 0.8017546534538269,
      "learning_rate": 4.2273779654127175e-05,
      "loss": 0.4741,
      "num_input_tokens_seen": 8271584,
      "step": 12650
    },
    {
      "epoch": 6.632599580712788,
      "grad_norm": 0.9864010810852051,
      "learning_rate": 4.226551198689982e-05,
      "loss": 0.444,
      "num_input_tokens_seen": 8275072,
      "step": 12655
    },
    {
      "epoch": 6.635220125786163,
      "grad_norm": 0.4688577950000763,
      "learning_rate": 4.225724070792434e-05,
      "loss": 0.3899,
      "num_input_tokens_seen": 8277856,
      "step": 12660
    },
    {
      "epoch": 6.637840670859539,
      "grad_norm": 0.20987765491008759,
      "learning_rate": 4.2248965818930975e-05,
      "loss": 0.5493,
      "num_input_tokens_seen": 8281888,
      "step": 12665
    },
    {
      "epoch": 6.640461215932914,
      "grad_norm": 0.3627394139766693,
      "learning_rate": 4.224068732165075e-05,
      "loss": 0.3623,
      "num_input_tokens_seen": 8288032,
      "step": 12670
    },
    {
      "epoch": 6.6430817610062896,
      "grad_norm": 0.5745804905891418,
      "learning_rate": 4.223240521781543e-05,
      "loss": 0.4673,
      "num_input_tokens_seen": 8291520,
      "step": 12675
    },
    {
      "epoch": 6.645702306079665,
      "grad_norm": 0.455117404460907,
      "learning_rate": 4.2224119509157535e-05,
      "loss": 0.496,
      "num_input_tokens_seen": 8295360,
      "step": 12680
    },
    {
      "epoch": 6.64832285115304,
      "grad_norm": 0.38492169976234436,
      "learning_rate": 4.221583019741033e-05,
      "loss": 0.6257,
      "num_input_tokens_seen": 8299776,
      "step": 12685
    },
    {
      "epoch": 6.650943396226415,
      "grad_norm": 0.8164406418800354,
      "learning_rate": 4.2207537284307876e-05,
      "loss": 0.4734,
      "num_input_tokens_seen": 8303072,
      "step": 12690
    },
    {
      "epoch": 6.65356394129979,
      "grad_norm": 0.44965940713882446,
      "learning_rate": 4.219924077158493e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 8305824,
      "step": 12695
    },
    {
      "epoch": 6.656184486373165,
      "grad_norm": 0.7785863280296326,
      "learning_rate": 4.219094066097703e-05,
      "loss": 0.4189,
      "num_input_tokens_seen": 8308704,
      "step": 12700
    },
    {
      "epoch": 6.658805031446541,
      "grad_norm": 0.5631717443466187,
      "learning_rate": 4.218263695422047e-05,
      "loss": 0.5951,
      "num_input_tokens_seen": 8311904,
      "step": 12705
    },
    {
      "epoch": 6.661425576519916,
      "grad_norm": 0.6198883652687073,
      "learning_rate": 4.217432965305229e-05,
      "loss": 0.5195,
      "num_input_tokens_seen": 8315296,
      "step": 12710
    },
    {
      "epoch": 6.664046121593291,
      "grad_norm": 0.3954162001609802,
      "learning_rate": 4.2166018759210286e-05,
      "loss": 0.4427,
      "num_input_tokens_seen": 8318112,
      "step": 12715
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.44871947169303894,
      "learning_rate": 4.215770427443299e-05,
      "loss": 0.3929,
      "num_input_tokens_seen": 8321600,
      "step": 12720
    },
    {
      "epoch": 6.669287211740042,
      "grad_norm": 0.2588209807872772,
      "learning_rate": 4.214938620045971e-05,
      "loss": 0.4741,
      "num_input_tokens_seen": 8324352,
      "step": 12725
    },
    {
      "epoch": 6.671907756813417,
      "grad_norm": 0.32495585083961487,
      "learning_rate": 4.214106453903048e-05,
      "loss": 0.4789,
      "num_input_tokens_seen": 8327648,
      "step": 12730
    },
    {
      "epoch": 6.6745283018867925,
      "grad_norm": 0.28340578079223633,
      "learning_rate": 4.2132739291886114e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 8330688,
      "step": 12735
    },
    {
      "epoch": 6.677148846960168,
      "grad_norm": 0.724153995513916,
      "learning_rate": 4.212441046076814e-05,
      "loss": 0.5421,
      "num_input_tokens_seen": 8333216,
      "step": 12740
    },
    {
      "epoch": 6.679769392033543,
      "grad_norm": 0.3777068257331848,
      "learning_rate": 4.211607804741887e-05,
      "loss": 0.4841,
      "num_input_tokens_seen": 8336928,
      "step": 12745
    },
    {
      "epoch": 6.682389937106918,
      "grad_norm": 0.47471746802330017,
      "learning_rate": 4.2107742053581344e-05,
      "loss": 0.59,
      "num_input_tokens_seen": 8339712,
      "step": 12750
    },
    {
      "epoch": 6.685010482180293,
      "grad_norm": 0.4880756735801697,
      "learning_rate": 4.209940248099935e-05,
      "loss": 0.6552,
      "num_input_tokens_seen": 8342208,
      "step": 12755
    },
    {
      "epoch": 6.687631027253669,
      "grad_norm": 0.7061354517936707,
      "learning_rate": 4.2091059331417445e-05,
      "loss": 0.4946,
      "num_input_tokens_seen": 8345280,
      "step": 12760
    },
    {
      "epoch": 6.690251572327044,
      "grad_norm": 0.529285192489624,
      "learning_rate": 4.208271260658091e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 8348640,
      "step": 12765
    },
    {
      "epoch": 6.6928721174004195,
      "grad_norm": 0.5936887264251709,
      "learning_rate": 4.2074362308235785e-05,
      "loss": 0.3955,
      "num_input_tokens_seen": 8351840,
      "step": 12770
    },
    {
      "epoch": 6.695492662473795,
      "grad_norm": 0.4608258605003357,
      "learning_rate": 4.206600843812886e-05,
      "loss": 0.3875,
      "num_input_tokens_seen": 8354912,
      "step": 12775
    },
    {
      "epoch": 6.69811320754717,
      "grad_norm": 0.5155589580535889,
      "learning_rate": 4.205765099800767e-05,
      "loss": 0.3903,
      "num_input_tokens_seen": 8357792,
      "step": 12780
    },
    {
      "epoch": 6.700733752620545,
      "grad_norm": 0.22144193947315216,
      "learning_rate": 4.204928998962049e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 8361408,
      "step": 12785
    },
    {
      "epoch": 6.70335429769392,
      "grad_norm": 1.658024549484253,
      "learning_rate": 4.204092541471636e-05,
      "loss": 0.5892,
      "num_input_tokens_seen": 8364128,
      "step": 12790
    },
    {
      "epoch": 6.705974842767295,
      "grad_norm": 0.2662132978439331,
      "learning_rate": 4.203255727504504e-05,
      "loss": 0.4916,
      "num_input_tokens_seen": 8367904,
      "step": 12795
    },
    {
      "epoch": 6.7085953878406706,
      "grad_norm": 0.49427324533462524,
      "learning_rate": 4.202418557235706e-05,
      "loss": 0.4414,
      "num_input_tokens_seen": 8371328,
      "step": 12800
    },
    {
      "epoch": 6.711215932914046,
      "grad_norm": 0.48095089197158813,
      "learning_rate": 4.201581030840368e-05,
      "loss": 0.3701,
      "num_input_tokens_seen": 8374464,
      "step": 12805
    },
    {
      "epoch": 6.713836477987421,
      "grad_norm": 0.4267335534095764,
      "learning_rate": 4.200743148493691e-05,
      "loss": 0.5732,
      "num_input_tokens_seen": 8377888,
      "step": 12810
    },
    {
      "epoch": 6.716457023060797,
      "grad_norm": 1.2161191701889038,
      "learning_rate": 4.199904910370951e-05,
      "loss": 0.5509,
      "num_input_tokens_seen": 8380864,
      "step": 12815
    },
    {
      "epoch": 6.719077568134172,
      "grad_norm": 0.522638201713562,
      "learning_rate": 4.1990663166474964e-05,
      "loss": 0.4523,
      "num_input_tokens_seen": 8383616,
      "step": 12820
    },
    {
      "epoch": 6.721698113207547,
      "grad_norm": 0.47979408502578735,
      "learning_rate": 4.1982273674987524e-05,
      "loss": 0.455,
      "num_input_tokens_seen": 8386784,
      "step": 12825
    },
    {
      "epoch": 6.7243186582809225,
      "grad_norm": 0.3772539794445038,
      "learning_rate": 4.197388063100218e-05,
      "loss": 0.3879,
      "num_input_tokens_seen": 8390400,
      "step": 12830
    },
    {
      "epoch": 6.726939203354298,
      "grad_norm": 0.5963979363441467,
      "learning_rate": 4.196548403627465e-05,
      "loss": 0.4056,
      "num_input_tokens_seen": 8393152,
      "step": 12835
    },
    {
      "epoch": 6.729559748427673,
      "grad_norm": 0.4868620038032532,
      "learning_rate": 4.195708389256142e-05,
      "loss": 0.4511,
      "num_input_tokens_seen": 8395776,
      "step": 12840
    },
    {
      "epoch": 6.732180293501048,
      "grad_norm": 0.4014447331428528,
      "learning_rate": 4.194868020161969e-05,
      "loss": 0.4814,
      "num_input_tokens_seen": 8399648,
      "step": 12845
    },
    {
      "epoch": 6.734800838574423,
      "grad_norm": 0.473865807056427,
      "learning_rate": 4.1940272965207425e-05,
      "loss": 0.3815,
      "num_input_tokens_seen": 8402624,
      "step": 12850
    },
    {
      "epoch": 6.737421383647799,
      "grad_norm": 0.529590368270874,
      "learning_rate": 4.1931862185083326e-05,
      "loss": 0.5146,
      "num_input_tokens_seen": 8406944,
      "step": 12855
    },
    {
      "epoch": 6.740041928721174,
      "grad_norm": 0.2999720275402069,
      "learning_rate": 4.192344786300682e-05,
      "loss": 0.5935,
      "num_input_tokens_seen": 8410144,
      "step": 12860
    },
    {
      "epoch": 6.7426624737945495,
      "grad_norm": 0.6251363158226013,
      "learning_rate": 4.19150300007381e-05,
      "loss": 0.5286,
      "num_input_tokens_seen": 8413280,
      "step": 12865
    },
    {
      "epoch": 6.745283018867925,
      "grad_norm": 0.3552848994731903,
      "learning_rate": 4.190660860003806e-05,
      "loss": 0.5414,
      "num_input_tokens_seen": 8416832,
      "step": 12870
    },
    {
      "epoch": 6.7479035639413,
      "grad_norm": 0.39468908309936523,
      "learning_rate": 4.189818366266841e-05,
      "loss": 0.5767,
      "num_input_tokens_seen": 8420480,
      "step": 12875
    },
    {
      "epoch": 6.750524109014675,
      "grad_norm": 0.4275830388069153,
      "learning_rate": 4.188975519039151e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 8424032,
      "step": 12880
    },
    {
      "epoch": 6.75314465408805,
      "grad_norm": 0.5688655376434326,
      "learning_rate": 4.188132318497051e-05,
      "loss": 0.4781,
      "num_input_tokens_seen": 8426880,
      "step": 12885
    },
    {
      "epoch": 6.755765199161425,
      "grad_norm": 0.7723092436790466,
      "learning_rate": 4.1872887648169304e-05,
      "loss": 0.4121,
      "num_input_tokens_seen": 8429632,
      "step": 12890
    },
    {
      "epoch": 6.7583857442348005,
      "grad_norm": 0.3923814296722412,
      "learning_rate": 4.186444858175248e-05,
      "loss": 0.489,
      "num_input_tokens_seen": 8432512,
      "step": 12895
    },
    {
      "epoch": 6.761006289308176,
      "grad_norm": 0.5324649214744568,
      "learning_rate": 4.1856005987485424e-05,
      "loss": 0.4325,
      "num_input_tokens_seen": 8436192,
      "step": 12900
    },
    {
      "epoch": 6.763626834381551,
      "grad_norm": 0.3130945861339569,
      "learning_rate": 4.1847559867134215e-05,
      "loss": 0.5195,
      "num_input_tokens_seen": 8439968,
      "step": 12905
    },
    {
      "epoch": 6.766247379454927,
      "grad_norm": 0.5374826788902283,
      "learning_rate": 4.1839110222465693e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 8442816,
      "step": 12910
    },
    {
      "epoch": 6.768867924528302,
      "grad_norm": 0.6664875149726868,
      "learning_rate": 4.183065705524742e-05,
      "loss": 0.4404,
      "num_input_tokens_seen": 8445216,
      "step": 12915
    },
    {
      "epoch": 6.771488469601677,
      "grad_norm": 0.3478686809539795,
      "learning_rate": 4.1822200367247696e-05,
      "loss": 0.4265,
      "num_input_tokens_seen": 8448768,
      "step": 12920
    },
    {
      "epoch": 6.774109014675052,
      "grad_norm": 0.3549349904060364,
      "learning_rate": 4.181374016023559e-05,
      "loss": 0.3728,
      "num_input_tokens_seen": 8451616,
      "step": 12925
    },
    {
      "epoch": 6.776729559748428,
      "grad_norm": 0.7964330911636353,
      "learning_rate": 4.1805276435980854e-05,
      "loss": 0.5788,
      "num_input_tokens_seen": 8453856,
      "step": 12930
    },
    {
      "epoch": 6.779350104821803,
      "grad_norm": 0.5280107855796814,
      "learning_rate": 4.1796809196254014e-05,
      "loss": 0.5808,
      "num_input_tokens_seen": 8457248,
      "step": 12935
    },
    {
      "epoch": 6.781970649895178,
      "grad_norm": 0.7679905295372009,
      "learning_rate": 4.178833844282632e-05,
      "loss": 0.5846,
      "num_input_tokens_seen": 8459808,
      "step": 12940
    },
    {
      "epoch": 6.784591194968553,
      "grad_norm": 0.36422720551490784,
      "learning_rate": 4.177986417746976e-05,
      "loss": 0.4559,
      "num_input_tokens_seen": 8463040,
      "step": 12945
    },
    {
      "epoch": 6.787211740041929,
      "grad_norm": 0.44786518812179565,
      "learning_rate": 4.177138640195704e-05,
      "loss": 0.3893,
      "num_input_tokens_seen": 8466016,
      "step": 12950
    },
    {
      "epoch": 6.789832285115304,
      "grad_norm": 0.3287036120891571,
      "learning_rate": 4.176290511806162e-05,
      "loss": 0.5094,
      "num_input_tokens_seen": 8468544,
      "step": 12955
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 0.7023809552192688,
      "learning_rate": 4.1754420327557706e-05,
      "loss": 0.4324,
      "num_input_tokens_seen": 8471968,
      "step": 12960
    },
    {
      "epoch": 6.795073375262055,
      "grad_norm": 0.4745625853538513,
      "learning_rate": 4.1745932032220195e-05,
      "loss": 0.5904,
      "num_input_tokens_seen": 8474272,
      "step": 12965
    },
    {
      "epoch": 6.79769392033543,
      "grad_norm": 0.6407644748687744,
      "learning_rate": 4.173744023382474e-05,
      "loss": 0.5628,
      "num_input_tokens_seen": 8478848,
      "step": 12970
    },
    {
      "epoch": 6.800314465408805,
      "grad_norm": 0.35142385959625244,
      "learning_rate": 4.172894493414776e-05,
      "loss": 0.3481,
      "num_input_tokens_seen": 8481568,
      "step": 12975
    },
    {
      "epoch": 6.80293501048218,
      "grad_norm": 0.5351846218109131,
      "learning_rate": 4.172044613496634e-05,
      "loss": 0.4,
      "num_input_tokens_seen": 8485280,
      "step": 12980
    },
    {
      "epoch": 6.805555555555555,
      "grad_norm": 0.42949721217155457,
      "learning_rate": 4.171194383805835e-05,
      "loss": 0.439,
      "num_input_tokens_seen": 8488160,
      "step": 12985
    },
    {
      "epoch": 6.8081761006289305,
      "grad_norm": 0.6443914771080017,
      "learning_rate": 4.170343804520237e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 8491264,
      "step": 12990
    },
    {
      "epoch": 6.810796645702306,
      "grad_norm": 0.7137472629547119,
      "learning_rate": 4.169492875817771e-05,
      "loss": 0.5632,
      "num_input_tokens_seen": 8494176,
      "step": 12995
    },
    {
      "epoch": 6.813417190775681,
      "grad_norm": 0.43544885516166687,
      "learning_rate": 4.1686415978764415e-05,
      "loss": 0.7338,
      "num_input_tokens_seen": 8497504,
      "step": 13000
    },
    {
      "epoch": 6.816037735849057,
      "grad_norm": 0.4362770915031433,
      "learning_rate": 4.167789970874327e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 8500992,
      "step": 13005
    },
    {
      "epoch": 6.818658280922432,
      "grad_norm": 0.3692219853401184,
      "learning_rate": 4.166937994989578e-05,
      "loss": 0.4979,
      "num_input_tokens_seen": 8504800,
      "step": 13010
    },
    {
      "epoch": 6.821278825995807,
      "grad_norm": 0.3003621995449066,
      "learning_rate": 4.166085670400418e-05,
      "loss": 0.3912,
      "num_input_tokens_seen": 8508864,
      "step": 13015
    },
    {
      "epoch": 6.823899371069182,
      "grad_norm": 0.36067622900009155,
      "learning_rate": 4.165232997285143e-05,
      "loss": 0.5012,
      "num_input_tokens_seen": 8511872,
      "step": 13020
    },
    {
      "epoch": 6.826519916142558,
      "grad_norm": 0.6666020750999451,
      "learning_rate": 4.164379975822124e-05,
      "loss": 0.3799,
      "num_input_tokens_seen": 8515200,
      "step": 13025
    },
    {
      "epoch": 6.829140461215933,
      "grad_norm": 0.2970692217350006,
      "learning_rate": 4.1635266061898015e-05,
      "loss": 0.5916,
      "num_input_tokens_seen": 8518240,
      "step": 13030
    },
    {
      "epoch": 6.831761006289308,
      "grad_norm": 0.44435998797416687,
      "learning_rate": 4.162672888566691e-05,
      "loss": 0.4796,
      "num_input_tokens_seen": 8521472,
      "step": 13035
    },
    {
      "epoch": 6.834381551362683,
      "grad_norm": 0.43544065952301025,
      "learning_rate": 4.1618188231313816e-05,
      "loss": 0.3638,
      "num_input_tokens_seen": 8524288,
      "step": 13040
    },
    {
      "epoch": 6.837002096436059,
      "grad_norm": 0.3513374328613281,
      "learning_rate": 4.1609644100625335e-05,
      "loss": 0.3987,
      "num_input_tokens_seen": 8526912,
      "step": 13045
    },
    {
      "epoch": 6.839622641509434,
      "grad_norm": 0.6304656267166138,
      "learning_rate": 4.16010964953888e-05,
      "loss": 0.4429,
      "num_input_tokens_seen": 8530048,
      "step": 13050
    },
    {
      "epoch": 6.8422431865828095,
      "grad_norm": 0.5813458561897278,
      "learning_rate": 4.159254541739227e-05,
      "loss": 0.5112,
      "num_input_tokens_seen": 8534144,
      "step": 13055
    },
    {
      "epoch": 6.844863731656185,
      "grad_norm": 0.6594177484512329,
      "learning_rate": 4.158399086842454e-05,
      "loss": 0.4152,
      "num_input_tokens_seen": 8537536,
      "step": 13060
    },
    {
      "epoch": 6.84748427672956,
      "grad_norm": 0.41301339864730835,
      "learning_rate": 4.157543285027512e-05,
      "loss": 0.5019,
      "num_input_tokens_seen": 8541056,
      "step": 13065
    },
    {
      "epoch": 6.850104821802935,
      "grad_norm": 0.326372891664505,
      "learning_rate": 4.156687136473425e-05,
      "loss": 0.4345,
      "num_input_tokens_seen": 8545248,
      "step": 13070
    },
    {
      "epoch": 6.85272536687631,
      "grad_norm": 0.6439953446388245,
      "learning_rate": 4.1558306413592894e-05,
      "loss": 0.434,
      "num_input_tokens_seen": 8548288,
      "step": 13075
    },
    {
      "epoch": 6.855345911949685,
      "grad_norm": 0.3509591817855835,
      "learning_rate": 4.1549737998642745e-05,
      "loss": 0.3616,
      "num_input_tokens_seen": 8551456,
      "step": 13080
    },
    {
      "epoch": 6.8579664570230605,
      "grad_norm": 0.6938080191612244,
      "learning_rate": 4.154116612167622e-05,
      "loss": 0.6048,
      "num_input_tokens_seen": 8554688,
      "step": 13085
    },
    {
      "epoch": 6.860587002096436,
      "grad_norm": 0.24184298515319824,
      "learning_rate": 4.1532590784486444e-05,
      "loss": 0.4694,
      "num_input_tokens_seen": 8558016,
      "step": 13090
    },
    {
      "epoch": 6.863207547169811,
      "grad_norm": 0.7411203980445862,
      "learning_rate": 4.1524011988867294e-05,
      "loss": 0.4611,
      "num_input_tokens_seen": 8561312,
      "step": 13095
    },
    {
      "epoch": 6.865828092243187,
      "grad_norm": 0.5292640328407288,
      "learning_rate": 4.151542973661336e-05,
      "loss": 0.3464,
      "num_input_tokens_seen": 8565152,
      "step": 13100
    },
    {
      "epoch": 6.868448637316562,
      "grad_norm": 0.3598749041557312,
      "learning_rate": 4.150684402951994e-05,
      "loss": 0.542,
      "num_input_tokens_seen": 8567936,
      "step": 13105
    },
    {
      "epoch": 6.871069182389937,
      "grad_norm": 0.4685792922973633,
      "learning_rate": 4.149825486938306e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 8571104,
      "step": 13110
    },
    {
      "epoch": 6.873689727463312,
      "grad_norm": 0.5026623010635376,
      "learning_rate": 4.148966225799949e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 8573728,
      "step": 13115
    },
    {
      "epoch": 6.876310272536688,
      "grad_norm": 0.5282192230224609,
      "learning_rate": 4.14810661971667e-05,
      "loss": 0.5141,
      "num_input_tokens_seen": 8576672,
      "step": 13120
    },
    {
      "epoch": 6.878930817610063,
      "grad_norm": 0.6811478137969971,
      "learning_rate": 4.147246668868289e-05,
      "loss": 0.5426,
      "num_input_tokens_seen": 8579232,
      "step": 13125
    },
    {
      "epoch": 6.881551362683438,
      "grad_norm": 0.517440915107727,
      "learning_rate": 4.1463863734346964e-05,
      "loss": 0.5005,
      "num_input_tokens_seen": 8581824,
      "step": 13130
    },
    {
      "epoch": 6.884171907756813,
      "grad_norm": 0.8560224771499634,
      "learning_rate": 4.145525733595859e-05,
      "loss": 0.4455,
      "num_input_tokens_seen": 8586080,
      "step": 13135
    },
    {
      "epoch": 6.886792452830189,
      "grad_norm": 0.40968427062034607,
      "learning_rate": 4.14466474953181e-05,
      "loss": 0.405,
      "num_input_tokens_seen": 8590400,
      "step": 13140
    },
    {
      "epoch": 6.889412997903564,
      "grad_norm": 0.494361937046051,
      "learning_rate": 4.14380342142266e-05,
      "loss": 0.5336,
      "num_input_tokens_seen": 8593728,
      "step": 13145
    },
    {
      "epoch": 6.8920335429769395,
      "grad_norm": 0.319976270198822,
      "learning_rate": 4.142941749448587e-05,
      "loss": 0.3814,
      "num_input_tokens_seen": 8596832,
      "step": 13150
    },
    {
      "epoch": 6.894654088050315,
      "grad_norm": 0.44115903973579407,
      "learning_rate": 4.142079733789843e-05,
      "loss": 0.3755,
      "num_input_tokens_seen": 8601024,
      "step": 13155
    },
    {
      "epoch": 6.89727463312369,
      "grad_norm": 0.6063717007637024,
      "learning_rate": 4.141217374626753e-05,
      "loss": 0.594,
      "num_input_tokens_seen": 8603936,
      "step": 13160
    },
    {
      "epoch": 6.899895178197065,
      "grad_norm": 0.45527172088623047,
      "learning_rate": 4.140354672139713e-05,
      "loss": 0.4045,
      "num_input_tokens_seen": 8606720,
      "step": 13165
    },
    {
      "epoch": 6.90251572327044,
      "grad_norm": 0.29358431696891785,
      "learning_rate": 4.1394916265091886e-05,
      "loss": 0.5944,
      "num_input_tokens_seen": 8612352,
      "step": 13170
    },
    {
      "epoch": 6.905136268343815,
      "grad_norm": 0.48813366889953613,
      "learning_rate": 4.1386282379157214e-05,
      "loss": 0.3551,
      "num_input_tokens_seen": 8615168,
      "step": 13175
    },
    {
      "epoch": 6.9077568134171905,
      "grad_norm": 0.38480907678604126,
      "learning_rate": 4.137764506539921e-05,
      "loss": 0.5617,
      "num_input_tokens_seen": 8618304,
      "step": 13180
    },
    {
      "epoch": 6.910377358490566,
      "grad_norm": 0.4456060826778412,
      "learning_rate": 4.136900432562471e-05,
      "loss": 0.5372,
      "num_input_tokens_seen": 8621216,
      "step": 13185
    },
    {
      "epoch": 6.912997903563941,
      "grad_norm": 0.5975179076194763,
      "learning_rate": 4.136036016164123e-05,
      "loss": 0.34,
      "num_input_tokens_seen": 8624256,
      "step": 13190
    },
    {
      "epoch": 6.915618448637317,
      "grad_norm": 0.6436276435852051,
      "learning_rate": 4.135171257525707e-05,
      "loss": 0.4894,
      "num_input_tokens_seen": 8627968,
      "step": 13195
    },
    {
      "epoch": 6.918238993710692,
      "grad_norm": 0.5778671503067017,
      "learning_rate": 4.1343061568281195e-05,
      "loss": 0.4353,
      "num_input_tokens_seen": 8630848,
      "step": 13200
    },
    {
      "epoch": 6.920859538784067,
      "grad_norm": 0.312673419713974,
      "learning_rate": 4.133440714252328e-05,
      "loss": 0.4596,
      "num_input_tokens_seen": 8634880,
      "step": 13205
    },
    {
      "epoch": 6.923480083857442,
      "grad_norm": 0.4026774764060974,
      "learning_rate": 4.132574929979375e-05,
      "loss": 0.5892,
      "num_input_tokens_seen": 8638432,
      "step": 13210
    },
    {
      "epoch": 6.926100628930818,
      "grad_norm": 0.6461926102638245,
      "learning_rate": 4.1317088041903714e-05,
      "loss": 0.5676,
      "num_input_tokens_seen": 8641888,
      "step": 13215
    },
    {
      "epoch": 6.928721174004193,
      "grad_norm": 0.5180462002754211,
      "learning_rate": 4.1308423370665015e-05,
      "loss": 0.5821,
      "num_input_tokens_seen": 8645760,
      "step": 13220
    },
    {
      "epoch": 6.931341719077568,
      "grad_norm": 0.39145416021347046,
      "learning_rate": 4.1299755287890206e-05,
      "loss": 0.4437,
      "num_input_tokens_seen": 8649120,
      "step": 13225
    },
    {
      "epoch": 6.933962264150943,
      "grad_norm": 0.6323971152305603,
      "learning_rate": 4.129108379539255e-05,
      "loss": 0.5114,
      "num_input_tokens_seen": 8652416,
      "step": 13230
    },
    {
      "epoch": 6.936582809224319,
      "grad_norm": 0.28212329745292664,
      "learning_rate": 4.128240889498602e-05,
      "loss": 0.3049,
      "num_input_tokens_seen": 8655040,
      "step": 13235
    },
    {
      "epoch": 6.939203354297694,
      "grad_norm": 0.5022902488708496,
      "learning_rate": 4.127373058848531e-05,
      "loss": 0.4069,
      "num_input_tokens_seen": 8658720,
      "step": 13240
    },
    {
      "epoch": 6.9418238993710695,
      "grad_norm": 0.42182445526123047,
      "learning_rate": 4.126504887770582e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 8662144,
      "step": 13245
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 0.49021369218826294,
      "learning_rate": 4.1256363764463666e-05,
      "loss": 0.3811,
      "num_input_tokens_seen": 8664960,
      "step": 13250
    },
    {
      "epoch": 6.94706498951782,
      "grad_norm": 0.3445865213871002,
      "learning_rate": 4.1247675250575676e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 8668352,
      "step": 13255
    },
    {
      "epoch": 6.949685534591195,
      "grad_norm": 0.46854761242866516,
      "learning_rate": 4.1238983337859396e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 8670784,
      "step": 13260
    },
    {
      "epoch": 6.95230607966457,
      "grad_norm": 0.29538899660110474,
      "learning_rate": 4.123028802813305e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 8674272,
      "step": 13265
    },
    {
      "epoch": 6.954926624737945,
      "grad_norm": 0.3658522069454193,
      "learning_rate": 4.1221589323215625e-05,
      "loss": 0.4219,
      "num_input_tokens_seen": 8677696,
      "step": 13270
    },
    {
      "epoch": 6.9575471698113205,
      "grad_norm": 0.4917001724243164,
      "learning_rate": 4.121288722492678e-05,
      "loss": 0.5679,
      "num_input_tokens_seen": 8680640,
      "step": 13275
    },
    {
      "epoch": 6.960167714884696,
      "grad_norm": 0.328633189201355,
      "learning_rate": 4.120418173508688e-05,
      "loss": 0.4271,
      "num_input_tokens_seen": 8683680,
      "step": 13280
    },
    {
      "epoch": 6.962788259958071,
      "grad_norm": 0.4159906804561615,
      "learning_rate": 4.1195472855517045e-05,
      "loss": 0.4001,
      "num_input_tokens_seen": 8687136,
      "step": 13285
    },
    {
      "epoch": 6.965408805031447,
      "grad_norm": 0.4826846420764923,
      "learning_rate": 4.1186760588039045e-05,
      "loss": 0.3498,
      "num_input_tokens_seen": 8690048,
      "step": 13290
    },
    {
      "epoch": 6.968029350104822,
      "grad_norm": 0.46470996737480164,
      "learning_rate": 4.11780449344754e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 8693248,
      "step": 13295
    },
    {
      "epoch": 6.970649895178197,
      "grad_norm": 0.4400692284107208,
      "learning_rate": 4.1169325896649326e-05,
      "loss": 0.4256,
      "num_input_tokens_seen": 8700800,
      "step": 13300
    },
    {
      "epoch": 6.973270440251572,
      "grad_norm": 0.4245144724845886,
      "learning_rate": 4.116060347638474e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 8705824,
      "step": 13305
    },
    {
      "epoch": 6.975890985324948,
      "grad_norm": 0.32766208052635193,
      "learning_rate": 4.115187767550628e-05,
      "loss": 0.4324,
      "num_input_tokens_seen": 8708768,
      "step": 13310
    },
    {
      "epoch": 6.978511530398323,
      "grad_norm": 0.44392403960227966,
      "learning_rate": 4.114314849583928e-05,
      "loss": 0.3685,
      "num_input_tokens_seen": 8711936,
      "step": 13315
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 0.5094349980354309,
      "learning_rate": 4.113441593920979e-05,
      "loss": 0.4589,
      "num_input_tokens_seen": 8714880,
      "step": 13320
    },
    {
      "epoch": 6.983752620545073,
      "grad_norm": 0.5466775298118591,
      "learning_rate": 4.112568000744454e-05,
      "loss": 0.3599,
      "num_input_tokens_seen": 8718272,
      "step": 13325
    },
    {
      "epoch": 6.986373165618449,
      "grad_norm": 0.4033624529838562,
      "learning_rate": 4.1116940702371025e-05,
      "loss": 0.5064,
      "num_input_tokens_seen": 8720928,
      "step": 13330
    },
    {
      "epoch": 6.988993710691824,
      "grad_norm": 0.3805805444717407,
      "learning_rate": 4.1108198025817376e-05,
      "loss": 0.4344,
      "num_input_tokens_seen": 8723712,
      "step": 13335
    },
    {
      "epoch": 6.9916142557651995,
      "grad_norm": 0.7203850746154785,
      "learning_rate": 4.1099451979612476e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 8726752,
      "step": 13340
    },
    {
      "epoch": 6.994234800838575,
      "grad_norm": 0.4210554361343384,
      "learning_rate": 4.109070256558589e-05,
      "loss": 0.5777,
      "num_input_tokens_seen": 8730336,
      "step": 13345
    },
    {
      "epoch": 6.99685534591195,
      "grad_norm": 0.4200596213340759,
      "learning_rate": 4.108194978556791e-05,
      "loss": 0.4939,
      "num_input_tokens_seen": 8732896,
      "step": 13350
    },
    {
      "epoch": 6.999475890985325,
      "grad_norm": 0.7664488554000854,
      "learning_rate": 4.1073193641389495e-05,
      "loss": 0.4684,
      "num_input_tokens_seen": 8735648,
      "step": 13355
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.4811994433403015,
      "eval_runtime": 9.6765,
      "eval_samples_per_second": 87.635,
      "eval_steps_per_second": 21.909,
      "num_input_tokens_seen": 8735728,
      "step": 13356
    },
    {
      "epoch": 7.0020964360587,
      "grad_norm": 0.5649755597114563,
      "learning_rate": 4.1064434134882345e-05,
      "loss": 0.3987,
      "num_input_tokens_seen": 8738288,
      "step": 13360
    },
    {
      "epoch": 7.004716981132075,
      "grad_norm": 0.32263728976249695,
      "learning_rate": 4.1055671267878847e-05,
      "loss": 0.3768,
      "num_input_tokens_seen": 8742352,
      "step": 13365
    },
    {
      "epoch": 7.0073375262054505,
      "grad_norm": 0.6792131662368774,
      "learning_rate": 4.1046905042212106e-05,
      "loss": 0.4505,
      "num_input_tokens_seen": 8744880,
      "step": 13370
    },
    {
      "epoch": 7.009958071278826,
      "grad_norm": 0.23968353867530823,
      "learning_rate": 4.103813545971589e-05,
      "loss": 0.3996,
      "num_input_tokens_seen": 8751280,
      "step": 13375
    },
    {
      "epoch": 7.012578616352202,
      "grad_norm": 0.4064466059207916,
      "learning_rate": 4.10293625222247e-05,
      "loss": 0.5035,
      "num_input_tokens_seen": 8754704,
      "step": 13380
    },
    {
      "epoch": 7.015199161425577,
      "grad_norm": 0.43084636330604553,
      "learning_rate": 4.102058623157376e-05,
      "loss": 0.6184,
      "num_input_tokens_seen": 8757840,
      "step": 13385
    },
    {
      "epoch": 7.017819706498952,
      "grad_norm": 0.5648661255836487,
      "learning_rate": 4.101180658959894e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 8761136,
      "step": 13390
    },
    {
      "epoch": 7.020440251572327,
      "grad_norm": 0.25082021951675415,
      "learning_rate": 4.1003023598136866e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 8764976,
      "step": 13395
    },
    {
      "epoch": 7.023060796645702,
      "grad_norm": 0.4958304166793823,
      "learning_rate": 4.099423725902482e-05,
      "loss": 0.5027,
      "num_input_tokens_seen": 8768624,
      "step": 13400
    },
    {
      "epoch": 7.0256813417190775,
      "grad_norm": 0.3372681438922882,
      "learning_rate": 4.098544757410081e-05,
      "loss": 0.6103,
      "num_input_tokens_seen": 8772880,
      "step": 13405
    },
    {
      "epoch": 7.028301886792453,
      "grad_norm": 0.527504026889801,
      "learning_rate": 4.097665454520354e-05,
      "loss": 0.4172,
      "num_input_tokens_seen": 8775824,
      "step": 13410
    },
    {
      "epoch": 7.030922431865828,
      "grad_norm": 0.819198489189148,
      "learning_rate": 4.096785817417241e-05,
      "loss": 0.5322,
      "num_input_tokens_seen": 8778480,
      "step": 13415
    },
    {
      "epoch": 7.033542976939203,
      "grad_norm": 0.4516012668609619,
      "learning_rate": 4.0959058462847524e-05,
      "loss": 0.4033,
      "num_input_tokens_seen": 8781072,
      "step": 13420
    },
    {
      "epoch": 7.036163522012578,
      "grad_norm": 0.520115315914154,
      "learning_rate": 4.095025541306968e-05,
      "loss": 0.3833,
      "num_input_tokens_seen": 8784592,
      "step": 13425
    },
    {
      "epoch": 7.038784067085954,
      "grad_norm": 0.2983372211456299,
      "learning_rate": 4.0941449026680374e-05,
      "loss": 0.3756,
      "num_input_tokens_seen": 8788656,
      "step": 13430
    },
    {
      "epoch": 7.0414046121593294,
      "grad_norm": 0.418790727853775,
      "learning_rate": 4.09326393055218e-05,
      "loss": 0.4302,
      "num_input_tokens_seen": 8792016,
      "step": 13435
    },
    {
      "epoch": 7.044025157232705,
      "grad_norm": 0.6822171807289124,
      "learning_rate": 4.0923826251436845e-05,
      "loss": 0.4039,
      "num_input_tokens_seen": 8794768,
      "step": 13440
    },
    {
      "epoch": 7.04664570230608,
      "grad_norm": 0.3779838979244232,
      "learning_rate": 4.0915009866269115e-05,
      "loss": 0.5795,
      "num_input_tokens_seen": 8797776,
      "step": 13445
    },
    {
      "epoch": 7.049266247379455,
      "grad_norm": 0.34670865535736084,
      "learning_rate": 4.090619015186288e-05,
      "loss": 0.5163,
      "num_input_tokens_seen": 8801488,
      "step": 13450
    },
    {
      "epoch": 7.05188679245283,
      "grad_norm": 0.5163935422897339,
      "learning_rate": 4.0897367110063145e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 8805360,
      "step": 13455
    },
    {
      "epoch": 7.054507337526205,
      "grad_norm": 0.4006754159927368,
      "learning_rate": 4.088854074271557e-05,
      "loss": 0.4723,
      "num_input_tokens_seen": 8808848,
      "step": 13460
    },
    {
      "epoch": 7.0571278825995805,
      "grad_norm": 0.9825135469436646,
      "learning_rate": 4.0879711051666534e-05,
      "loss": 0.5381,
      "num_input_tokens_seen": 8812336,
      "step": 13465
    },
    {
      "epoch": 7.059748427672956,
      "grad_norm": 0.9213544130325317,
      "learning_rate": 4.087087803876313e-05,
      "loss": 0.5245,
      "num_input_tokens_seen": 8814768,
      "step": 13470
    },
    {
      "epoch": 7.062368972746331,
      "grad_norm": 0.7079169154167175,
      "learning_rate": 4.086204170585308e-05,
      "loss": 0.3316,
      "num_input_tokens_seen": 8818512,
      "step": 13475
    },
    {
      "epoch": 7.064989517819707,
      "grad_norm": 0.5308299660682678,
      "learning_rate": 4.085320205478488e-05,
      "loss": 0.5392,
      "num_input_tokens_seen": 8821232,
      "step": 13480
    },
    {
      "epoch": 7.067610062893082,
      "grad_norm": 0.7392623424530029,
      "learning_rate": 4.084435908740768e-05,
      "loss": 0.4676,
      "num_input_tokens_seen": 8825616,
      "step": 13485
    },
    {
      "epoch": 7.070230607966457,
      "grad_norm": 0.4261646866798401,
      "learning_rate": 4.0835512805571306e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 8828144,
      "step": 13490
    },
    {
      "epoch": 7.072851153039832,
      "grad_norm": 0.6596032977104187,
      "learning_rate": 4.082666321112633e-05,
      "loss": 0.5874,
      "num_input_tokens_seen": 8830960,
      "step": 13495
    },
    {
      "epoch": 7.0754716981132075,
      "grad_norm": 0.621761679649353,
      "learning_rate": 4.081781030592396e-05,
      "loss": 0.4506,
      "num_input_tokens_seen": 8833680,
      "step": 13500
    },
    {
      "epoch": 7.078092243186583,
      "grad_norm": 0.316886305809021,
      "learning_rate": 4.080895409181613e-05,
      "loss": 0.5548,
      "num_input_tokens_seen": 8837008,
      "step": 13505
    },
    {
      "epoch": 7.080712788259958,
      "grad_norm": 0.5149868726730347,
      "learning_rate": 4.080009457065547e-05,
      "loss": 0.4149,
      "num_input_tokens_seen": 8840528,
      "step": 13510
    },
    {
      "epoch": 7.083333333333333,
      "grad_norm": 0.4097732901573181,
      "learning_rate": 4.0791231744295286e-05,
      "loss": 0.4491,
      "num_input_tokens_seen": 8844080,
      "step": 13515
    },
    {
      "epoch": 7.085953878406708,
      "grad_norm": 0.3289467394351959,
      "learning_rate": 4.078236561458957e-05,
      "loss": 0.338,
      "num_input_tokens_seen": 8847184,
      "step": 13520
    },
    {
      "epoch": 7.088574423480084,
      "grad_norm": 0.46484822034835815,
      "learning_rate": 4.077349618339302e-05,
      "loss": 0.4638,
      "num_input_tokens_seen": 8850032,
      "step": 13525
    },
    {
      "epoch": 7.091194968553459,
      "grad_norm": 0.21820271015167236,
      "learning_rate": 4.076462345256103e-05,
      "loss": 0.3975,
      "num_input_tokens_seen": 8853328,
      "step": 13530
    },
    {
      "epoch": 7.093815513626835,
      "grad_norm": 0.5088518261909485,
      "learning_rate": 4.0755747423949655e-05,
      "loss": 0.4102,
      "num_input_tokens_seen": 8856432,
      "step": 13535
    },
    {
      "epoch": 7.09643605870021,
      "grad_norm": 0.6417772173881531,
      "learning_rate": 4.074686809941568e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 8859696,
      "step": 13540
    },
    {
      "epoch": 7.099056603773585,
      "grad_norm": 0.1983002871274948,
      "learning_rate": 4.0737985480816546e-05,
      "loss": 0.3565,
      "num_input_tokens_seen": 8862736,
      "step": 13545
    },
    {
      "epoch": 7.10167714884696,
      "grad_norm": 0.31841781735420227,
      "learning_rate": 4.0729099570010395e-05,
      "loss": 0.4536,
      "num_input_tokens_seen": 8865424,
      "step": 13550
    },
    {
      "epoch": 7.104297693920335,
      "grad_norm": 0.7265649437904358,
      "learning_rate": 4.072021036885607e-05,
      "loss": 0.5049,
      "num_input_tokens_seen": 8868368,
      "step": 13555
    },
    {
      "epoch": 7.1069182389937104,
      "grad_norm": 0.35827934741973877,
      "learning_rate": 4.0711317879213075e-05,
      "loss": 0.4954,
      "num_input_tokens_seen": 8870960,
      "step": 13560
    },
    {
      "epoch": 7.109538784067086,
      "grad_norm": 0.5712363719940186,
      "learning_rate": 4.070242210294162e-05,
      "loss": 0.6014,
      "num_input_tokens_seen": 8874256,
      "step": 13565
    },
    {
      "epoch": 7.112159329140461,
      "grad_norm": 0.8993622660636902,
      "learning_rate": 4.069352304190261e-05,
      "loss": 0.5057,
      "num_input_tokens_seen": 8877456,
      "step": 13570
    },
    {
      "epoch": 7.114779874213837,
      "grad_norm": 0.37936073541641235,
      "learning_rate": 4.068462069795761e-05,
      "loss": 0.4625,
      "num_input_tokens_seen": 8880976,
      "step": 13575
    },
    {
      "epoch": 7.117400419287212,
      "grad_norm": 0.5254959464073181,
      "learning_rate": 4.067571507296891e-05,
      "loss": 0.3622,
      "num_input_tokens_seen": 8883568,
      "step": 13580
    },
    {
      "epoch": 7.120020964360587,
      "grad_norm": 0.7444589734077454,
      "learning_rate": 4.0666806168799455e-05,
      "loss": 0.5566,
      "num_input_tokens_seen": 8887088,
      "step": 13585
    },
    {
      "epoch": 7.122641509433962,
      "grad_norm": 0.3085569441318512,
      "learning_rate": 4.0657893987312876e-05,
      "loss": 0.5049,
      "num_input_tokens_seen": 8890704,
      "step": 13590
    },
    {
      "epoch": 7.1252620545073375,
      "grad_norm": 0.5405668020248413,
      "learning_rate": 4.064897853037352e-05,
      "loss": 0.4913,
      "num_input_tokens_seen": 8894160,
      "step": 13595
    },
    {
      "epoch": 7.127882599580713,
      "grad_norm": 0.34484148025512695,
      "learning_rate": 4.064005979984639e-05,
      "loss": 0.4354,
      "num_input_tokens_seen": 8897008,
      "step": 13600
    },
    {
      "epoch": 7.130503144654088,
      "grad_norm": 0.7144900560379028,
      "learning_rate": 4.0631137797597176e-05,
      "loss": 0.32,
      "num_input_tokens_seen": 8899696,
      "step": 13605
    },
    {
      "epoch": 7.133123689727463,
      "grad_norm": 0.39415910840034485,
      "learning_rate": 4.0622212525492266e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 8903568,
      "step": 13610
    },
    {
      "epoch": 7.135744234800838,
      "grad_norm": 0.49188071489334106,
      "learning_rate": 4.061328398539872e-05,
      "loss": 0.504,
      "num_input_tokens_seen": 8907216,
      "step": 13615
    },
    {
      "epoch": 7.138364779874214,
      "grad_norm": 1.1944576501846313,
      "learning_rate": 4.060435217918429e-05,
      "loss": 0.4505,
      "num_input_tokens_seen": 8910544,
      "step": 13620
    },
    {
      "epoch": 7.140985324947589,
      "grad_norm": 0.710622251033783,
      "learning_rate": 4.0595417108717405e-05,
      "loss": 0.4873,
      "num_input_tokens_seen": 8913872,
      "step": 13625
    },
    {
      "epoch": 7.143605870020965,
      "grad_norm": 0.4770604372024536,
      "learning_rate": 4.0586478775867185e-05,
      "loss": 0.3524,
      "num_input_tokens_seen": 8917456,
      "step": 13630
    },
    {
      "epoch": 7.14622641509434,
      "grad_norm": 0.39409855008125305,
      "learning_rate": 4.057753718250343e-05,
      "loss": 0.5258,
      "num_input_tokens_seen": 8920848,
      "step": 13635
    },
    {
      "epoch": 7.148846960167715,
      "grad_norm": 0.59394371509552,
      "learning_rate": 4.056859233049661e-05,
      "loss": 0.4613,
      "num_input_tokens_seen": 8923632,
      "step": 13640
    },
    {
      "epoch": 7.15146750524109,
      "grad_norm": 0.27942150831222534,
      "learning_rate": 4.05596442217179e-05,
      "loss": 0.442,
      "num_input_tokens_seen": 8927952,
      "step": 13645
    },
    {
      "epoch": 7.154088050314465,
      "grad_norm": 0.5955811738967896,
      "learning_rate": 4.055069285803912e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 8931792,
      "step": 13650
    },
    {
      "epoch": 7.15670859538784,
      "grad_norm": 0.3019745945930481,
      "learning_rate": 4.0541738241332815e-05,
      "loss": 0.3494,
      "num_input_tokens_seen": 8934672,
      "step": 13655
    },
    {
      "epoch": 7.159329140461216,
      "grad_norm": 0.46354442834854126,
      "learning_rate": 4.053278037347217e-05,
      "loss": 0.4148,
      "num_input_tokens_seen": 8938128,
      "step": 13660
    },
    {
      "epoch": 7.161949685534591,
      "grad_norm": 0.7362797260284424,
      "learning_rate": 4.052381925633109e-05,
      "loss": 0.4566,
      "num_input_tokens_seen": 8940784,
      "step": 13665
    },
    {
      "epoch": 7.164570230607967,
      "grad_norm": 0.3088594973087311,
      "learning_rate": 4.051485489178412e-05,
      "loss": 0.3893,
      "num_input_tokens_seen": 8944208,
      "step": 13670
    },
    {
      "epoch": 7.167190775681342,
      "grad_norm": 0.33156293630599976,
      "learning_rate": 4.05058872817065e-05,
      "loss": 0.63,
      "num_input_tokens_seen": 8947216,
      "step": 13675
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 0.38349196314811707,
      "learning_rate": 4.0496916427974177e-05,
      "loss": 0.4189,
      "num_input_tokens_seen": 8950480,
      "step": 13680
    },
    {
      "epoch": 7.172431865828092,
      "grad_norm": 0.37042275071144104,
      "learning_rate": 4.048794233246373e-05,
      "loss": 0.5286,
      "num_input_tokens_seen": 8953232,
      "step": 13685
    },
    {
      "epoch": 7.1750524109014675,
      "grad_norm": 0.3205021917819977,
      "learning_rate": 4.0478964997052435e-05,
      "loss": 0.3523,
      "num_input_tokens_seen": 8956304,
      "step": 13690
    },
    {
      "epoch": 7.177672955974843,
      "grad_norm": 0.3491024971008301,
      "learning_rate": 4.046998442361827e-05,
      "loss": 0.4418,
      "num_input_tokens_seen": 8960048,
      "step": 13695
    },
    {
      "epoch": 7.180293501048218,
      "grad_norm": 0.3994803726673126,
      "learning_rate": 4.046100061403984e-05,
      "loss": 0.397,
      "num_input_tokens_seen": 8963184,
      "step": 13700
    },
    {
      "epoch": 7.182914046121593,
      "grad_norm": 1.0015370845794678,
      "learning_rate": 4.045201357019648e-05,
      "loss": 0.4724,
      "num_input_tokens_seen": 8965776,
      "step": 13705
    },
    {
      "epoch": 7.185534591194968,
      "grad_norm": 0.4377724230289459,
      "learning_rate": 4.044302329396816e-05,
      "loss": 0.4951,
      "num_input_tokens_seen": 8968400,
      "step": 13710
    },
    {
      "epoch": 7.188155136268344,
      "grad_norm": 0.4486476182937622,
      "learning_rate": 4.0434029787235564e-05,
      "loss": 0.5837,
      "num_input_tokens_seen": 8971152,
      "step": 13715
    },
    {
      "epoch": 7.190775681341719,
      "grad_norm": 0.3910665810108185,
      "learning_rate": 4.042503305188e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 8974288,
      "step": 13720
    },
    {
      "epoch": 7.193396226415095,
      "grad_norm": 0.2686678469181061,
      "learning_rate": 4.041603308978351e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 8977840,
      "step": 13725
    },
    {
      "epoch": 7.19601677148847,
      "grad_norm": 0.43105942010879517,
      "learning_rate": 4.040702990282878e-05,
      "loss": 0.3927,
      "num_input_tokens_seen": 8981552,
      "step": 13730
    },
    {
      "epoch": 7.198637316561845,
      "grad_norm": 0.23400352895259857,
      "learning_rate": 4.039802349289914e-05,
      "loss": 0.3698,
      "num_input_tokens_seen": 8984400,
      "step": 13735
    },
    {
      "epoch": 7.20125786163522,
      "grad_norm": 0.630917489528656,
      "learning_rate": 4.038901386187868e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 8986992,
      "step": 13740
    },
    {
      "epoch": 7.203878406708595,
      "grad_norm": 0.7698590755462646,
      "learning_rate": 4.038000101165208e-05,
      "loss": 0.5019,
      "num_input_tokens_seen": 8989488,
      "step": 13745
    },
    {
      "epoch": 7.20649895178197,
      "grad_norm": 0.3396550118923187,
      "learning_rate": 4.0370984944104725e-05,
      "loss": 0.4362,
      "num_input_tokens_seen": 8992752,
      "step": 13750
    },
    {
      "epoch": 7.209119496855346,
      "grad_norm": 0.35216671228408813,
      "learning_rate": 4.0361965661122695e-05,
      "loss": 0.4832,
      "num_input_tokens_seen": 8995504,
      "step": 13755
    },
    {
      "epoch": 7.211740041928721,
      "grad_norm": 1.0388461351394653,
      "learning_rate": 4.035294316459271e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 8998256,
      "step": 13760
    },
    {
      "epoch": 7.214360587002097,
      "grad_norm": 0.4450562000274658,
      "learning_rate": 4.034391745640215e-05,
      "loss": 0.4979,
      "num_input_tokens_seen": 9001168,
      "step": 13765
    },
    {
      "epoch": 7.216981132075472,
      "grad_norm": 0.7012620568275452,
      "learning_rate": 4.033488853843913e-05,
      "loss": 0.3765,
      "num_input_tokens_seen": 9004176,
      "step": 13770
    },
    {
      "epoch": 7.219601677148847,
      "grad_norm": 0.5878544449806213,
      "learning_rate": 4.0325856412592363e-05,
      "loss": 0.4777,
      "num_input_tokens_seen": 9007088,
      "step": 13775
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 0.4980071783065796,
      "learning_rate": 4.031682108075129e-05,
      "loss": 0.4394,
      "num_input_tokens_seen": 9010192,
      "step": 13780
    },
    {
      "epoch": 7.2248427672955975,
      "grad_norm": 0.2950246036052704,
      "learning_rate": 4.030778254480599e-05,
      "loss": 0.417,
      "num_input_tokens_seen": 9013744,
      "step": 13785
    },
    {
      "epoch": 7.227463312368973,
      "grad_norm": 0.5042386651039124,
      "learning_rate": 4.029874080664722e-05,
      "loss": 0.5328,
      "num_input_tokens_seen": 9016720,
      "step": 13790
    },
    {
      "epoch": 7.230083857442348,
      "grad_norm": 0.4290347993373871,
      "learning_rate": 4.0289695868166414e-05,
      "loss": 0.4415,
      "num_input_tokens_seen": 9020336,
      "step": 13795
    },
    {
      "epoch": 7.232704402515723,
      "grad_norm": 0.3872722089290619,
      "learning_rate": 4.028064773125566e-05,
      "loss": 0.5427,
      "num_input_tokens_seen": 9022896,
      "step": 13800
    },
    {
      "epoch": 7.235324947589098,
      "grad_norm": 0.34535714983940125,
      "learning_rate": 4.0271596397807734e-05,
      "loss": 0.4074,
      "num_input_tokens_seen": 9025808,
      "step": 13805
    },
    {
      "epoch": 7.237945492662474,
      "grad_norm": 0.34681436419487,
      "learning_rate": 4.026254186971607e-05,
      "loss": 0.3487,
      "num_input_tokens_seen": 9028528,
      "step": 13810
    },
    {
      "epoch": 7.240566037735849,
      "grad_norm": 0.4888305068016052,
      "learning_rate": 4.025348414887477e-05,
      "loss": 0.4797,
      "num_input_tokens_seen": 9031984,
      "step": 13815
    },
    {
      "epoch": 7.243186582809225,
      "grad_norm": 0.6135165691375732,
      "learning_rate": 4.024442323717861e-05,
      "loss": 0.5588,
      "num_input_tokens_seen": 9034480,
      "step": 13820
    },
    {
      "epoch": 7.2458071278826,
      "grad_norm": 0.6430482864379883,
      "learning_rate": 4.023535913652302e-05,
      "loss": 0.4125,
      "num_input_tokens_seen": 9037904,
      "step": 13825
    },
    {
      "epoch": 7.248427672955975,
      "grad_norm": 0.6593138575553894,
      "learning_rate": 4.022629184880412e-05,
      "loss": 0.5246,
      "num_input_tokens_seen": 9040016,
      "step": 13830
    },
    {
      "epoch": 7.25104821802935,
      "grad_norm": 0.3703632056713104,
      "learning_rate": 4.021722137591867e-05,
      "loss": 0.516,
      "num_input_tokens_seen": 9044176,
      "step": 13835
    },
    {
      "epoch": 7.253668763102725,
      "grad_norm": 0.8565771579742432,
      "learning_rate": 4.020814771976412e-05,
      "loss": 0.4003,
      "num_input_tokens_seen": 9047920,
      "step": 13840
    },
    {
      "epoch": 7.2562893081761,
      "grad_norm": 0.4195462167263031,
      "learning_rate": 4.019907088223857e-05,
      "loss": 0.3556,
      "num_input_tokens_seen": 9051024,
      "step": 13845
    },
    {
      "epoch": 7.258909853249476,
      "grad_norm": 0.5185359716415405,
      "learning_rate": 4.018999086524079e-05,
      "loss": 0.6499,
      "num_input_tokens_seen": 9055280,
      "step": 13850
    },
    {
      "epoch": 7.261530398322851,
      "grad_norm": 0.5272156596183777,
      "learning_rate": 4.018090767067022e-05,
      "loss": 0.4417,
      "num_input_tokens_seen": 9058128,
      "step": 13855
    },
    {
      "epoch": 7.264150943396227,
      "grad_norm": 0.3744884729385376,
      "learning_rate": 4.017182130042696e-05,
      "loss": 0.5524,
      "num_input_tokens_seen": 9061008,
      "step": 13860
    },
    {
      "epoch": 7.266771488469602,
      "grad_norm": 0.328702837228775,
      "learning_rate": 4.0162731756411764e-05,
      "loss": 0.4682,
      "num_input_tokens_seen": 9063856,
      "step": 13865
    },
    {
      "epoch": 7.269392033542977,
      "grad_norm": 0.5246071219444275,
      "learning_rate": 4.015363904052608e-05,
      "loss": 0.3928,
      "num_input_tokens_seen": 9067664,
      "step": 13870
    },
    {
      "epoch": 7.272012578616352,
      "grad_norm": 0.4814213514328003,
      "learning_rate": 4.014454315467199e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 9070320,
      "step": 13875
    },
    {
      "epoch": 7.2746331236897275,
      "grad_norm": 0.38136589527130127,
      "learning_rate": 4.013544410075224e-05,
      "loss": 0.5202,
      "num_input_tokens_seen": 9073264,
      "step": 13880
    },
    {
      "epoch": 7.277253668763103,
      "grad_norm": 0.5800631642341614,
      "learning_rate": 4.0126341880670273e-05,
      "loss": 0.4896,
      "num_input_tokens_seen": 9076368,
      "step": 13885
    },
    {
      "epoch": 7.279874213836478,
      "grad_norm": 0.5328190326690674,
      "learning_rate": 4.011723649633014e-05,
      "loss": 0.4158,
      "num_input_tokens_seen": 9079568,
      "step": 13890
    },
    {
      "epoch": 7.282494758909853,
      "grad_norm": 0.591285765171051,
      "learning_rate": 4.0108127949636606e-05,
      "loss": 0.5198,
      "num_input_tokens_seen": 9083184,
      "step": 13895
    },
    {
      "epoch": 7.285115303983228,
      "grad_norm": 0.407329261302948,
      "learning_rate": 4.009901624249507e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 9086832,
      "step": 13900
    },
    {
      "epoch": 7.287735849056604,
      "grad_norm": 0.5454021692276001,
      "learning_rate": 4.008990137681159e-05,
      "loss": 0.4103,
      "num_input_tokens_seen": 9089488,
      "step": 13905
    },
    {
      "epoch": 7.290356394129979,
      "grad_norm": 0.7319362759590149,
      "learning_rate": 4.00807833544929e-05,
      "loss": 0.4823,
      "num_input_tokens_seen": 9092144,
      "step": 13910
    },
    {
      "epoch": 7.2929769392033545,
      "grad_norm": 0.7428013682365417,
      "learning_rate": 4.0071662177446376e-05,
      "loss": 0.465,
      "num_input_tokens_seen": 9095152,
      "step": 13915
    },
    {
      "epoch": 7.29559748427673,
      "grad_norm": 0.39289307594299316,
      "learning_rate": 4.006253784758008e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 9098448,
      "step": 13920
    },
    {
      "epoch": 7.298218029350105,
      "grad_norm": 0.44298258423805237,
      "learning_rate": 4.0053410366802704e-05,
      "loss": 0.3275,
      "num_input_tokens_seen": 9101072,
      "step": 13925
    },
    {
      "epoch": 7.30083857442348,
      "grad_norm": 0.6345914602279663,
      "learning_rate": 4.004427973702362e-05,
      "loss": 0.394,
      "num_input_tokens_seen": 9103600,
      "step": 13930
    },
    {
      "epoch": 7.303459119496855,
      "grad_norm": 0.3621063530445099,
      "learning_rate": 4.003514596015285e-05,
      "loss": 0.5176,
      "num_input_tokens_seen": 9107344,
      "step": 13935
    },
    {
      "epoch": 7.30607966457023,
      "grad_norm": 0.8132867217063904,
      "learning_rate": 4.002600903810108e-05,
      "loss": 0.5477,
      "num_input_tokens_seen": 9110672,
      "step": 13940
    },
    {
      "epoch": 7.308700209643606,
      "grad_norm": 0.25264954566955566,
      "learning_rate": 4.001686897277963e-05,
      "loss": 0.4428,
      "num_input_tokens_seen": 9113936,
      "step": 13945
    },
    {
      "epoch": 7.311320754716981,
      "grad_norm": 0.4531826376914978,
      "learning_rate": 4.000772576610052e-05,
      "loss": 0.2997,
      "num_input_tokens_seen": 9116688,
      "step": 13950
    },
    {
      "epoch": 7.313941299790357,
      "grad_norm": 0.286950021982193,
      "learning_rate": 3.99985794199764e-05,
      "loss": 0.3941,
      "num_input_tokens_seen": 9120304,
      "step": 13955
    },
    {
      "epoch": 7.316561844863732,
      "grad_norm": 0.7910771369934082,
      "learning_rate": 3.998942993632058e-05,
      "loss": 0.5179,
      "num_input_tokens_seen": 9123440,
      "step": 13960
    },
    {
      "epoch": 7.319182389937107,
      "grad_norm": 1.0630724430084229,
      "learning_rate": 3.998027731704702e-05,
      "loss": 0.5632,
      "num_input_tokens_seen": 9126096,
      "step": 13965
    },
    {
      "epoch": 7.321802935010482,
      "grad_norm": 0.49809518456459045,
      "learning_rate": 3.9971121564070354e-05,
      "loss": 0.3468,
      "num_input_tokens_seen": 9129168,
      "step": 13970
    },
    {
      "epoch": 7.3244234800838575,
      "grad_norm": 0.6841040253639221,
      "learning_rate": 3.996196267930585e-05,
      "loss": 0.3869,
      "num_input_tokens_seen": 9132688,
      "step": 13975
    },
    {
      "epoch": 7.327044025157233,
      "grad_norm": 0.5178110003471375,
      "learning_rate": 3.995280066466945e-05,
      "loss": 0.5257,
      "num_input_tokens_seen": 9135216,
      "step": 13980
    },
    {
      "epoch": 7.329664570230608,
      "grad_norm": 0.6897820830345154,
      "learning_rate": 3.9943635522077746e-05,
      "loss": 0.5033,
      "num_input_tokens_seen": 9138768,
      "step": 13985
    },
    {
      "epoch": 7.332285115303983,
      "grad_norm": 0.31648632884025574,
      "learning_rate": 3.993446725344797e-05,
      "loss": 0.4516,
      "num_input_tokens_seen": 9141904,
      "step": 13990
    },
    {
      "epoch": 7.334905660377358,
      "grad_norm": 0.253042995929718,
      "learning_rate": 3.992529586069803e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 9145744,
      "step": 13995
    },
    {
      "epoch": 7.337526205450734,
      "grad_norm": 0.47145238518714905,
      "learning_rate": 3.991612134574648e-05,
      "loss": 0.6705,
      "num_input_tokens_seen": 9151760,
      "step": 14000
    },
    {
      "epoch": 7.340146750524109,
      "grad_norm": 0.4949408769607544,
      "learning_rate": 3.99069437105125e-05,
      "loss": 0.5177,
      "num_input_tokens_seen": 9154704,
      "step": 14005
    },
    {
      "epoch": 7.3427672955974845,
      "grad_norm": 0.47753411531448364,
      "learning_rate": 3.9897762956915976e-05,
      "loss": 0.4246,
      "num_input_tokens_seen": 9157520,
      "step": 14010
    },
    {
      "epoch": 7.34538784067086,
      "grad_norm": 0.5896091461181641,
      "learning_rate": 3.9888579086877403e-05,
      "loss": 0.5609,
      "num_input_tokens_seen": 9162128,
      "step": 14015
    },
    {
      "epoch": 7.348008385744235,
      "grad_norm": 0.35364800691604614,
      "learning_rate": 3.987939210231794e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 9165616,
      "step": 14020
    },
    {
      "epoch": 7.35062893081761,
      "grad_norm": 0.2805870473384857,
      "learning_rate": 3.987020200515941e-05,
      "loss": 0.3843,
      "num_input_tokens_seen": 9168400,
      "step": 14025
    },
    {
      "epoch": 7.353249475890985,
      "grad_norm": 0.48686733841896057,
      "learning_rate": 3.9861008797324255e-05,
      "loss": 0.5468,
      "num_input_tokens_seen": 9172176,
      "step": 14030
    },
    {
      "epoch": 7.35587002096436,
      "grad_norm": 0.3167025148868561,
      "learning_rate": 3.9851812480735605e-05,
      "loss": 0.4266,
      "num_input_tokens_seen": 9175184,
      "step": 14035
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 0.47595906257629395,
      "learning_rate": 3.9842613057317235e-05,
      "loss": 0.4107,
      "num_input_tokens_seen": 9178320,
      "step": 14040
    },
    {
      "epoch": 7.361111111111111,
      "grad_norm": 0.43066492676734924,
      "learning_rate": 3.9833410528993534e-05,
      "loss": 0.388,
      "num_input_tokens_seen": 9181264,
      "step": 14045
    },
    {
      "epoch": 7.363731656184487,
      "grad_norm": 0.6622921824455261,
      "learning_rate": 3.9824204897689585e-05,
      "loss": 0.4299,
      "num_input_tokens_seen": 9184272,
      "step": 14050
    },
    {
      "epoch": 7.366352201257862,
      "grad_norm": 0.4691953659057617,
      "learning_rate": 3.98149961653311e-05,
      "loss": 0.4774,
      "num_input_tokens_seen": 9187024,
      "step": 14055
    },
    {
      "epoch": 7.368972746331237,
      "grad_norm": 0.43455368280410767,
      "learning_rate": 3.980578433384442e-05,
      "loss": 0.4919,
      "num_input_tokens_seen": 9189456,
      "step": 14060
    },
    {
      "epoch": 7.371593291404612,
      "grad_norm": 0.28758907318115234,
      "learning_rate": 3.979656940515659e-05,
      "loss": 0.4745,
      "num_input_tokens_seen": 9192368,
      "step": 14065
    },
    {
      "epoch": 7.3742138364779874,
      "grad_norm": 0.3963577151298523,
      "learning_rate": 3.9787351381195235e-05,
      "loss": 0.4491,
      "num_input_tokens_seen": 9195568,
      "step": 14070
    },
    {
      "epoch": 7.376834381551363,
      "grad_norm": 0.509597659111023,
      "learning_rate": 3.977813026388868e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 9198384,
      "step": 14075
    },
    {
      "epoch": 7.379454926624738,
      "grad_norm": 0.4193321764469147,
      "learning_rate": 3.976890605516587e-05,
      "loss": 0.4284,
      "num_input_tokens_seen": 9201232,
      "step": 14080
    },
    {
      "epoch": 7.382075471698113,
      "grad_norm": 0.5415794849395752,
      "learning_rate": 3.97596787569564e-05,
      "loss": 0.5451,
      "num_input_tokens_seen": 9208560,
      "step": 14085
    },
    {
      "epoch": 7.384696016771488,
      "grad_norm": 0.5972034931182861,
      "learning_rate": 3.975044837119053e-05,
      "loss": 0.4767,
      "num_input_tokens_seen": 9211856,
      "step": 14090
    },
    {
      "epoch": 7.387316561844864,
      "grad_norm": 0.2934004068374634,
      "learning_rate": 3.974121489979914e-05,
      "loss": 0.4401,
      "num_input_tokens_seen": 9215280,
      "step": 14095
    },
    {
      "epoch": 7.389937106918239,
      "grad_norm": 0.5509296655654907,
      "learning_rate": 3.973197834471377e-05,
      "loss": 0.6205,
      "num_input_tokens_seen": 9218352,
      "step": 14100
    },
    {
      "epoch": 7.3925576519916145,
      "grad_norm": 0.38451695442199707,
      "learning_rate": 3.9722738707866595e-05,
      "loss": 0.4172,
      "num_input_tokens_seen": 9221552,
      "step": 14105
    },
    {
      "epoch": 7.39517819706499,
      "grad_norm": 0.7309607267379761,
      "learning_rate": 3.971349599119046e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 9224496,
      "step": 14110
    },
    {
      "epoch": 7.397798742138365,
      "grad_norm": 0.41923147439956665,
      "learning_rate": 3.970425019661881e-05,
      "loss": 0.5217,
      "num_input_tokens_seen": 9227696,
      "step": 14115
    },
    {
      "epoch": 7.40041928721174,
      "grad_norm": 0.45206108689308167,
      "learning_rate": 3.9695001326085785e-05,
      "loss": 0.3596,
      "num_input_tokens_seen": 9231792,
      "step": 14120
    },
    {
      "epoch": 7.403039832285115,
      "grad_norm": 0.4664232134819031,
      "learning_rate": 3.968574938152613e-05,
      "loss": 0.5021,
      "num_input_tokens_seen": 9235280,
      "step": 14125
    },
    {
      "epoch": 7.40566037735849,
      "grad_norm": 0.3889026939868927,
      "learning_rate": 3.9676494364875247e-05,
      "loss": 0.4688,
      "num_input_tokens_seen": 9238672,
      "step": 14130
    },
    {
      "epoch": 7.4082809224318655,
      "grad_norm": 0.2999117076396942,
      "learning_rate": 3.966723627806918e-05,
      "loss": 0.6409,
      "num_input_tokens_seen": 9241968,
      "step": 14135
    },
    {
      "epoch": 7.410901467505241,
      "grad_norm": 0.5316811800003052,
      "learning_rate": 3.9657975123044624e-05,
      "loss": 0.5824,
      "num_input_tokens_seen": 9244976,
      "step": 14140
    },
    {
      "epoch": 7.413522012578617,
      "grad_norm": 0.5580628514289856,
      "learning_rate": 3.964871090173888e-05,
      "loss": 0.4265,
      "num_input_tokens_seen": 9247856,
      "step": 14145
    },
    {
      "epoch": 7.416142557651992,
      "grad_norm": 0.36754992604255676,
      "learning_rate": 3.963944361608995e-05,
      "loss": 0.3777,
      "num_input_tokens_seen": 9252656,
      "step": 14150
    },
    {
      "epoch": 7.418763102725367,
      "grad_norm": 0.3588995933532715,
      "learning_rate": 3.963017326803644e-05,
      "loss": 0.5552,
      "num_input_tokens_seen": 9254960,
      "step": 14155
    },
    {
      "epoch": 7.421383647798742,
      "grad_norm": 0.40160632133483887,
      "learning_rate": 3.962089985951758e-05,
      "loss": 0.3888,
      "num_input_tokens_seen": 9258288,
      "step": 14160
    },
    {
      "epoch": 7.424004192872117,
      "grad_norm": 0.42552387714385986,
      "learning_rate": 3.961162339247327e-05,
      "loss": 0.4427,
      "num_input_tokens_seen": 9262032,
      "step": 14165
    },
    {
      "epoch": 7.426624737945493,
      "grad_norm": 0.38404762744903564,
      "learning_rate": 3.9602343868844055e-05,
      "loss": 0.5848,
      "num_input_tokens_seen": 9265520,
      "step": 14170
    },
    {
      "epoch": 7.429245283018868,
      "grad_norm": 1.2170124053955078,
      "learning_rate": 3.959306129057109e-05,
      "loss": 0.5589,
      "num_input_tokens_seen": 9268208,
      "step": 14175
    },
    {
      "epoch": 7.431865828092243,
      "grad_norm": 0.49346598982810974,
      "learning_rate": 3.958377565959619e-05,
      "loss": 0.5009,
      "num_input_tokens_seen": 9270768,
      "step": 14180
    },
    {
      "epoch": 7.434486373165618,
      "grad_norm": 0.5086880326271057,
      "learning_rate": 3.95744869778618e-05,
      "loss": 0.4292,
      "num_input_tokens_seen": 9274064,
      "step": 14185
    },
    {
      "epoch": 7.437106918238994,
      "grad_norm": 0.5437180399894714,
      "learning_rate": 3.956519524731101e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 9277456,
      "step": 14190
    },
    {
      "epoch": 7.439727463312369,
      "grad_norm": 0.40155139565467834,
      "learning_rate": 3.9555900469887555e-05,
      "loss": 0.5482,
      "num_input_tokens_seen": 9280208,
      "step": 14195
    },
    {
      "epoch": 7.4423480083857445,
      "grad_norm": 0.328036904335022,
      "learning_rate": 3.954660264753577e-05,
      "loss": 0.4977,
      "num_input_tokens_seen": 9283664,
      "step": 14200
    },
    {
      "epoch": 7.44496855345912,
      "grad_norm": 0.31851881742477417,
      "learning_rate": 3.953730178220067e-05,
      "loss": 0.383,
      "num_input_tokens_seen": 9288944,
      "step": 14205
    },
    {
      "epoch": 7.447589098532495,
      "grad_norm": 0.443013459444046,
      "learning_rate": 3.952799787582789e-05,
      "loss": 0.4412,
      "num_input_tokens_seen": 9292976,
      "step": 14210
    },
    {
      "epoch": 7.45020964360587,
      "grad_norm": 0.48060911893844604,
      "learning_rate": 3.9518690930363694e-05,
      "loss": 0.4611,
      "num_input_tokens_seen": 9295600,
      "step": 14215
    },
    {
      "epoch": 7.452830188679245,
      "grad_norm": 0.3766896426677704,
      "learning_rate": 3.9509380947754994e-05,
      "loss": 0.4292,
      "num_input_tokens_seen": 9298512,
      "step": 14220
    },
    {
      "epoch": 7.45545073375262,
      "grad_norm": 0.8860974311828613,
      "learning_rate": 3.9500067929949336e-05,
      "loss": 0.5805,
      "num_input_tokens_seen": 9301616,
      "step": 14225
    },
    {
      "epoch": 7.4580712788259955,
      "grad_norm": 0.619283139705658,
      "learning_rate": 3.9490751878894895e-05,
      "loss": 0.3971,
      "num_input_tokens_seen": 9304112,
      "step": 14230
    },
    {
      "epoch": 7.460691823899371,
      "grad_norm": 0.5165513753890991,
      "learning_rate": 3.948143279654047e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 9307312,
      "step": 14235
    },
    {
      "epoch": 7.463312368972747,
      "grad_norm": 0.5164913535118103,
      "learning_rate": 3.947211068483553e-05,
      "loss": 0.4502,
      "num_input_tokens_seen": 9310960,
      "step": 14240
    },
    {
      "epoch": 7.465932914046122,
      "grad_norm": 0.6430625319480896,
      "learning_rate": 3.946278554573013e-05,
      "loss": 0.5342,
      "num_input_tokens_seen": 9313168,
      "step": 14245
    },
    {
      "epoch": 7.468553459119497,
      "grad_norm": 0.22326213121414185,
      "learning_rate": 3.9453457381175e-05,
      "loss": 0.409,
      "num_input_tokens_seen": 9316624,
      "step": 14250
    },
    {
      "epoch": 7.471174004192872,
      "grad_norm": 0.7899209260940552,
      "learning_rate": 3.9444126193121474e-05,
      "loss": 0.4902,
      "num_input_tokens_seen": 9319920,
      "step": 14255
    },
    {
      "epoch": 7.473794549266247,
      "grad_norm": 0.5190792083740234,
      "learning_rate": 3.943479198352153e-05,
      "loss": 0.483,
      "num_input_tokens_seen": 9322640,
      "step": 14260
    },
    {
      "epoch": 7.476415094339623,
      "grad_norm": 0.39981991052627563,
      "learning_rate": 3.94254547543278e-05,
      "loss": 0.5288,
      "num_input_tokens_seen": 9325424,
      "step": 14265
    },
    {
      "epoch": 7.479035639412998,
      "grad_norm": 0.6232850551605225,
      "learning_rate": 3.9416114507493504e-05,
      "loss": 0.3765,
      "num_input_tokens_seen": 9328816,
      "step": 14270
    },
    {
      "epoch": 7.481656184486373,
      "grad_norm": 0.5832687020301819,
      "learning_rate": 3.940677124497252e-05,
      "loss": 0.3953,
      "num_input_tokens_seen": 9331472,
      "step": 14275
    },
    {
      "epoch": 7.484276729559748,
      "grad_norm": 0.47689366340637207,
      "learning_rate": 3.939742496871935e-05,
      "loss": 0.4975,
      "num_input_tokens_seen": 9334544,
      "step": 14280
    },
    {
      "epoch": 7.486897274633124,
      "grad_norm": 0.41272708773612976,
      "learning_rate": 3.9388075680689136e-05,
      "loss": 0.4854,
      "num_input_tokens_seen": 9337488,
      "step": 14285
    },
    {
      "epoch": 7.489517819706499,
      "grad_norm": 0.3855089545249939,
      "learning_rate": 3.9378723382837634e-05,
      "loss": 0.669,
      "num_input_tokens_seen": 9341168,
      "step": 14290
    },
    {
      "epoch": 7.4921383647798745,
      "grad_norm": 0.2674019932746887,
      "learning_rate": 3.936936807712124e-05,
      "loss": 0.4629,
      "num_input_tokens_seen": 9344112,
      "step": 14295
    },
    {
      "epoch": 7.49475890985325,
      "grad_norm": 0.6641794443130493,
      "learning_rate": 3.936000976549698e-05,
      "loss": 0.4805,
      "num_input_tokens_seen": 9347920,
      "step": 14300
    },
    {
      "epoch": 7.497379454926625,
      "grad_norm": 0.44747620820999146,
      "learning_rate": 3.935064844992251e-05,
      "loss": 0.4826,
      "num_input_tokens_seen": 9350800,
      "step": 14305
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.3461919128894806,
      "learning_rate": 3.93412841323561e-05,
      "loss": 0.4778,
      "num_input_tokens_seen": 9353712,
      "step": 14310
    },
    {
      "epoch": 7.502620545073375,
      "grad_norm": 0.3323427736759186,
      "learning_rate": 3.933191681475666e-05,
      "loss": 0.4505,
      "num_input_tokens_seen": 9357456,
      "step": 14315
    },
    {
      "epoch": 7.50524109014675,
      "grad_norm": 0.5344644784927368,
      "learning_rate": 3.932254649908373e-05,
      "loss": 0.5871,
      "num_input_tokens_seen": 9360752,
      "step": 14320
    },
    {
      "epoch": 7.5078616352201255,
      "grad_norm": 0.43298014998435974,
      "learning_rate": 3.931317318729748e-05,
      "loss": 0.4801,
      "num_input_tokens_seen": 9364528,
      "step": 14325
    },
    {
      "epoch": 7.510482180293501,
      "grad_norm": 0.46262186765670776,
      "learning_rate": 3.930379688135869e-05,
      "loss": 0.4355,
      "num_input_tokens_seen": 9368336,
      "step": 14330
    },
    {
      "epoch": 7.513102725366876,
      "grad_norm": 0.5202972292900085,
      "learning_rate": 3.929441758322878e-05,
      "loss": 0.5335,
      "num_input_tokens_seen": 9372336,
      "step": 14335
    },
    {
      "epoch": 7.515723270440252,
      "grad_norm": 0.9434369206428528,
      "learning_rate": 3.928503529486979e-05,
      "loss": 0.4868,
      "num_input_tokens_seen": 9375280,
      "step": 14340
    },
    {
      "epoch": 7.518343815513627,
      "grad_norm": 0.513911247253418,
      "learning_rate": 3.927565001824438e-05,
      "loss": 0.3849,
      "num_input_tokens_seen": 9377936,
      "step": 14345
    },
    {
      "epoch": 7.520964360587002,
      "grad_norm": 0.441182941198349,
      "learning_rate": 3.926626175531586e-05,
      "loss": 0.5901,
      "num_input_tokens_seen": 9380816,
      "step": 14350
    },
    {
      "epoch": 7.523584905660377,
      "grad_norm": 0.28181982040405273,
      "learning_rate": 3.925687050804814e-05,
      "loss": 0.4057,
      "num_input_tokens_seen": 9383920,
      "step": 14355
    },
    {
      "epoch": 7.526205450733753,
      "grad_norm": 0.4836117625236511,
      "learning_rate": 3.924747627840575e-05,
      "loss": 0.5764,
      "num_input_tokens_seen": 9386320,
      "step": 14360
    },
    {
      "epoch": 7.528825995807128,
      "grad_norm": 0.38942715525627136,
      "learning_rate": 3.923807906835388e-05,
      "loss": 0.5468,
      "num_input_tokens_seen": 9389392,
      "step": 14365
    },
    {
      "epoch": 7.531446540880503,
      "grad_norm": 0.2958337068557739,
      "learning_rate": 3.92286788798583e-05,
      "loss": 0.3985,
      "num_input_tokens_seen": 9393424,
      "step": 14370
    },
    {
      "epoch": 7.534067085953878,
      "grad_norm": 0.33738967776298523,
      "learning_rate": 3.921927571488543e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 9396208,
      "step": 14375
    },
    {
      "epoch": 7.536687631027254,
      "grad_norm": 0.554251492023468,
      "learning_rate": 3.9209869575402296e-05,
      "loss": 0.4407,
      "num_input_tokens_seen": 9399504,
      "step": 14380
    },
    {
      "epoch": 7.539308176100629,
      "grad_norm": 0.3840298354625702,
      "learning_rate": 3.9200460463376566e-05,
      "loss": 0.4832,
      "num_input_tokens_seen": 9403728,
      "step": 14385
    },
    {
      "epoch": 7.5419287211740045,
      "grad_norm": 0.3849649727344513,
      "learning_rate": 3.9191048380776515e-05,
      "loss": 0.4041,
      "num_input_tokens_seen": 9406832,
      "step": 14390
    },
    {
      "epoch": 7.54454926624738,
      "grad_norm": 0.37944185733795166,
      "learning_rate": 3.918163332957104e-05,
      "loss": 0.4046,
      "num_input_tokens_seen": 9410544,
      "step": 14395
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 0.33544933795928955,
      "learning_rate": 3.9172215311729665e-05,
      "loss": 0.5325,
      "num_input_tokens_seen": 9414160,
      "step": 14400
    },
    {
      "epoch": 7.54979035639413,
      "grad_norm": 1.2109041213989258,
      "learning_rate": 3.916279432922254e-05,
      "loss": 0.4028,
      "num_input_tokens_seen": 9417680,
      "step": 14405
    },
    {
      "epoch": 7.552410901467505,
      "grad_norm": 0.41618025302886963,
      "learning_rate": 3.915337038402041e-05,
      "loss": 0.344,
      "num_input_tokens_seen": 9420272,
      "step": 14410
    },
    {
      "epoch": 7.55503144654088,
      "grad_norm": 0.36242982745170593,
      "learning_rate": 3.914394347809467e-05,
      "loss": 0.4276,
      "num_input_tokens_seen": 9424560,
      "step": 14415
    },
    {
      "epoch": 7.5576519916142555,
      "grad_norm": 0.40292948484420776,
      "learning_rate": 3.9134513613417315e-05,
      "loss": 0.3814,
      "num_input_tokens_seen": 9427184,
      "step": 14420
    },
    {
      "epoch": 7.560272536687631,
      "grad_norm": 0.5915669202804565,
      "learning_rate": 3.912508079196097e-05,
      "loss": 0.4275,
      "num_input_tokens_seen": 9429648,
      "step": 14425
    },
    {
      "epoch": 7.562893081761006,
      "grad_norm": 0.3922523856163025,
      "learning_rate": 3.9115645015698866e-05,
      "loss": 0.5561,
      "num_input_tokens_seen": 9432688,
      "step": 14430
    },
    {
      "epoch": 7.565513626834382,
      "grad_norm": 0.36957618594169617,
      "learning_rate": 3.9106206286604865e-05,
      "loss": 0.4316,
      "num_input_tokens_seen": 9435600,
      "step": 14435
    },
    {
      "epoch": 7.568134171907757,
      "grad_norm": 0.36787885427474976,
      "learning_rate": 3.909676460665345e-05,
      "loss": 0.306,
      "num_input_tokens_seen": 9438448,
      "step": 14440
    },
    {
      "epoch": 7.570754716981132,
      "grad_norm": 0.6669439077377319,
      "learning_rate": 3.908731997781969e-05,
      "loss": 0.515,
      "num_input_tokens_seen": 9441040,
      "step": 14445
    },
    {
      "epoch": 7.573375262054507,
      "grad_norm": 0.49106764793395996,
      "learning_rate": 3.907787240207932e-05,
      "loss": 0.3232,
      "num_input_tokens_seen": 9444560,
      "step": 14450
    },
    {
      "epoch": 7.575995807127883,
      "grad_norm": 1.8840856552124023,
      "learning_rate": 3.9068421881408645e-05,
      "loss": 0.502,
      "num_input_tokens_seen": 9447760,
      "step": 14455
    },
    {
      "epoch": 7.578616352201258,
      "grad_norm": 0.37233346700668335,
      "learning_rate": 3.9058968417784614e-05,
      "loss": 0.4954,
      "num_input_tokens_seen": 9450608,
      "step": 14460
    },
    {
      "epoch": 7.581236897274633,
      "grad_norm": 0.2726753056049347,
      "learning_rate": 3.9049512013184785e-05,
      "loss": 0.3355,
      "num_input_tokens_seen": 9454928,
      "step": 14465
    },
    {
      "epoch": 7.583857442348008,
      "grad_norm": 0.5403198003768921,
      "learning_rate": 3.9040052669587325e-05,
      "loss": 0.5175,
      "num_input_tokens_seen": 9458832,
      "step": 14470
    },
    {
      "epoch": 7.586477987421384,
      "grad_norm": 0.47273463010787964,
      "learning_rate": 3.9030590388971026e-05,
      "loss": 0.3706,
      "num_input_tokens_seen": 9461488,
      "step": 14475
    },
    {
      "epoch": 7.589098532494759,
      "grad_norm": 0.5477659106254578,
      "learning_rate": 3.902112517331529e-05,
      "loss": 0.3826,
      "num_input_tokens_seen": 9464336,
      "step": 14480
    },
    {
      "epoch": 7.5917190775681345,
      "grad_norm": 0.2371819168329239,
      "learning_rate": 3.901165702460012e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 9467600,
      "step": 14485
    },
    {
      "epoch": 7.59433962264151,
      "grad_norm": 0.5041053295135498,
      "learning_rate": 3.900218594480616e-05,
      "loss": 0.5017,
      "num_input_tokens_seen": 9470384,
      "step": 14490
    },
    {
      "epoch": 7.596960167714885,
      "grad_norm": 0.269273579120636,
      "learning_rate": 3.899271193591465e-05,
      "loss": 0.3163,
      "num_input_tokens_seen": 9474256,
      "step": 14495
    },
    {
      "epoch": 7.59958071278826,
      "grad_norm": 0.32386717200279236,
      "learning_rate": 3.8983234999907435e-05,
      "loss": 0.3275,
      "num_input_tokens_seen": 9477584,
      "step": 14500
    },
    {
      "epoch": 7.602201257861635,
      "grad_norm": 0.3099769055843353,
      "learning_rate": 3.897375513876699e-05,
      "loss": 0.4381,
      "num_input_tokens_seen": 9480912,
      "step": 14505
    },
    {
      "epoch": 7.60482180293501,
      "grad_norm": 0.3477505147457123,
      "learning_rate": 3.8964272354476396e-05,
      "loss": 0.5022,
      "num_input_tokens_seen": 9483696,
      "step": 14510
    },
    {
      "epoch": 7.6074423480083855,
      "grad_norm": 0.322188138961792,
      "learning_rate": 3.895478664901933e-05,
      "loss": 0.407,
      "num_input_tokens_seen": 9486768,
      "step": 14515
    },
    {
      "epoch": 7.610062893081761,
      "grad_norm": 0.3436605930328369,
      "learning_rate": 3.89452980243801e-05,
      "loss": 0.446,
      "num_input_tokens_seen": 9491088,
      "step": 14520
    },
    {
      "epoch": 7.612683438155136,
      "grad_norm": 0.40278035402297974,
      "learning_rate": 3.893580648254362e-05,
      "loss": 0.4918,
      "num_input_tokens_seen": 9493840,
      "step": 14525
    },
    {
      "epoch": 7.615303983228512,
      "grad_norm": 0.4263468086719513,
      "learning_rate": 3.8926312025495414e-05,
      "loss": 0.5537,
      "num_input_tokens_seen": 9496752,
      "step": 14530
    },
    {
      "epoch": 7.617924528301887,
      "grad_norm": 0.4020291268825531,
      "learning_rate": 3.891681465522161e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 9500144,
      "step": 14535
    },
    {
      "epoch": 7.620545073375262,
      "grad_norm": 0.43010395765304565,
      "learning_rate": 3.890731437370896e-05,
      "loss": 0.4296,
      "num_input_tokens_seen": 9503568,
      "step": 14540
    },
    {
      "epoch": 7.623165618448637,
      "grad_norm": 0.3828120827674866,
      "learning_rate": 3.88978111829448e-05,
      "loss": 0.5273,
      "num_input_tokens_seen": 9507472,
      "step": 14545
    },
    {
      "epoch": 7.6257861635220126,
      "grad_norm": 0.5778111815452576,
      "learning_rate": 3.88883050849171e-05,
      "loss": 0.4551,
      "num_input_tokens_seen": 9510224,
      "step": 14550
    },
    {
      "epoch": 7.628406708595388,
      "grad_norm": 0.3446017801761627,
      "learning_rate": 3.8878796081614417e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 9512976,
      "step": 14555
    },
    {
      "epoch": 7.631027253668763,
      "grad_norm": 0.29448869824409485,
      "learning_rate": 3.8869284175025935e-05,
      "loss": 0.4826,
      "num_input_tokens_seen": 9516016,
      "step": 14560
    },
    {
      "epoch": 7.633647798742138,
      "grad_norm": 0.3614143431186676,
      "learning_rate": 3.885976936714143e-05,
      "loss": 0.4542,
      "num_input_tokens_seen": 9518800,
      "step": 14565
    },
    {
      "epoch": 7.636268343815514,
      "grad_norm": 0.4402613043785095,
      "learning_rate": 3.885025165995129e-05,
      "loss": 0.4566,
      "num_input_tokens_seen": 9524688,
      "step": 14570
    },
    {
      "epoch": 7.638888888888889,
      "grad_norm": 0.6622611880302429,
      "learning_rate": 3.884073105544652e-05,
      "loss": 0.3703,
      "num_input_tokens_seen": 9527344,
      "step": 14575
    },
    {
      "epoch": 7.6415094339622645,
      "grad_norm": 0.5886029005050659,
      "learning_rate": 3.8831207555618723e-05,
      "loss": 0.3769,
      "num_input_tokens_seen": 9530224,
      "step": 14580
    },
    {
      "epoch": 7.64412997903564,
      "grad_norm": 0.4159293472766876,
      "learning_rate": 3.882168116246009e-05,
      "loss": 0.6768,
      "num_input_tokens_seen": 9533200,
      "step": 14585
    },
    {
      "epoch": 7.646750524109015,
      "grad_norm": 0.5590254068374634,
      "learning_rate": 3.881215187796344e-05,
      "loss": 0.4432,
      "num_input_tokens_seen": 9535888,
      "step": 14590
    },
    {
      "epoch": 7.64937106918239,
      "grad_norm": 0.46283572912216187,
      "learning_rate": 3.8802619704122204e-05,
      "loss": 0.4675,
      "num_input_tokens_seen": 9538576,
      "step": 14595
    },
    {
      "epoch": 7.651991614255765,
      "grad_norm": 0.4083844721317291,
      "learning_rate": 3.8793084642930387e-05,
      "loss": 0.4774,
      "num_input_tokens_seen": 9542128,
      "step": 14600
    },
    {
      "epoch": 7.65461215932914,
      "grad_norm": 0.5010654330253601,
      "learning_rate": 3.878354669638263e-05,
      "loss": 0.4599,
      "num_input_tokens_seen": 9545232,
      "step": 14605
    },
    {
      "epoch": 7.6572327044025155,
      "grad_norm": 0.4697131812572479,
      "learning_rate": 3.8774005866474155e-05,
      "loss": 0.5462,
      "num_input_tokens_seen": 9548080,
      "step": 14610
    },
    {
      "epoch": 7.659853249475891,
      "grad_norm": 0.6551068425178528,
      "learning_rate": 3.876446215520079e-05,
      "loss": 0.2624,
      "num_input_tokens_seen": 9550640,
      "step": 14615
    },
    {
      "epoch": 7.662473794549266,
      "grad_norm": 0.46253159642219543,
      "learning_rate": 3.875491556455898e-05,
      "loss": 0.4402,
      "num_input_tokens_seen": 9553136,
      "step": 14620
    },
    {
      "epoch": 7.665094339622642,
      "grad_norm": 0.30348914861679077,
      "learning_rate": 3.8745366096545754e-05,
      "loss": 0.3898,
      "num_input_tokens_seen": 9556208,
      "step": 14625
    },
    {
      "epoch": 7.667714884696017,
      "grad_norm": 0.4733288288116455,
      "learning_rate": 3.873581375315876e-05,
      "loss": 0.3476,
      "num_input_tokens_seen": 9560080,
      "step": 14630
    },
    {
      "epoch": 7.670335429769392,
      "grad_norm": 0.48641934990882874,
      "learning_rate": 3.872625853639623e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 9562704,
      "step": 14635
    },
    {
      "epoch": 7.672955974842767,
      "grad_norm": 0.622916579246521,
      "learning_rate": 3.871670044825701e-05,
      "loss": 0.5286,
      "num_input_tokens_seen": 9565680,
      "step": 14640
    },
    {
      "epoch": 7.6755765199161425,
      "grad_norm": 0.4507465064525604,
      "learning_rate": 3.870713949074056e-05,
      "loss": 0.439,
      "num_input_tokens_seen": 9569136,
      "step": 14645
    },
    {
      "epoch": 7.678197064989518,
      "grad_norm": 0.3817097842693329,
      "learning_rate": 3.86975756658469e-05,
      "loss": 0.5425,
      "num_input_tokens_seen": 9572144,
      "step": 14650
    },
    {
      "epoch": 7.680817610062893,
      "grad_norm": 0.5575520396232605,
      "learning_rate": 3.868800897557667e-05,
      "loss": 0.5932,
      "num_input_tokens_seen": 9575184,
      "step": 14655
    },
    {
      "epoch": 7.683438155136268,
      "grad_norm": 0.44536224007606506,
      "learning_rate": 3.867843942193113e-05,
      "loss": 0.4265,
      "num_input_tokens_seen": 9579632,
      "step": 14660
    },
    {
      "epoch": 7.686058700209644,
      "grad_norm": 0.47440579533576965,
      "learning_rate": 3.866886700691212e-05,
      "loss": 0.4094,
      "num_input_tokens_seen": 9582576,
      "step": 14665
    },
    {
      "epoch": 7.688679245283019,
      "grad_norm": 0.6753228306770325,
      "learning_rate": 3.8659291732522075e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 9586224,
      "step": 14670
    },
    {
      "epoch": 7.691299790356394,
      "grad_norm": 0.6267331838607788,
      "learning_rate": 3.8649713600764026e-05,
      "loss": 0.3835,
      "num_input_tokens_seen": 9593136,
      "step": 14675
    },
    {
      "epoch": 7.69392033542977,
      "grad_norm": 0.7334138751029968,
      "learning_rate": 3.864013261364162e-05,
      "loss": 0.4299,
      "num_input_tokens_seen": 9596336,
      "step": 14680
    },
    {
      "epoch": 7.696540880503145,
      "grad_norm": 0.438753217458725,
      "learning_rate": 3.8630548773159084e-05,
      "loss": 0.3687,
      "num_input_tokens_seen": 9599760,
      "step": 14685
    },
    {
      "epoch": 7.69916142557652,
      "grad_norm": 0.35880154371261597,
      "learning_rate": 3.862096208132125e-05,
      "loss": 0.592,
      "num_input_tokens_seen": 9603248,
      "step": 14690
    },
    {
      "epoch": 7.701781970649895,
      "grad_norm": 0.43378946185112,
      "learning_rate": 3.861137254013355e-05,
      "loss": 0.3905,
      "num_input_tokens_seen": 9606384,
      "step": 14695
    },
    {
      "epoch": 7.70440251572327,
      "grad_norm": 0.4319189786911011,
      "learning_rate": 3.8601780151601996e-05,
      "loss": 0.4629,
      "num_input_tokens_seen": 9609424,
      "step": 14700
    },
    {
      "epoch": 7.7070230607966455,
      "grad_norm": 0.3557435870170593,
      "learning_rate": 3.859218491773321e-05,
      "loss": 0.4237,
      "num_input_tokens_seen": 9611984,
      "step": 14705
    },
    {
      "epoch": 7.709643605870021,
      "grad_norm": 0.7413803935050964,
      "learning_rate": 3.858258684053441e-05,
      "loss": 0.4913,
      "num_input_tokens_seen": 9614768,
      "step": 14710
    },
    {
      "epoch": 7.712264150943396,
      "grad_norm": 0.537896990776062,
      "learning_rate": 3.8572985922013405e-05,
      "loss": 0.3178,
      "num_input_tokens_seen": 9617392,
      "step": 14715
    },
    {
      "epoch": 7.714884696016772,
      "grad_norm": 0.542536199092865,
      "learning_rate": 3.856338216417859e-05,
      "loss": 0.3917,
      "num_input_tokens_seen": 9620368,
      "step": 14720
    },
    {
      "epoch": 7.717505241090147,
      "grad_norm": 0.33308300375938416,
      "learning_rate": 3.855377556903897e-05,
      "loss": 0.3851,
      "num_input_tokens_seen": 9623600,
      "step": 14725
    },
    {
      "epoch": 7.720125786163522,
      "grad_norm": 0.7960119247436523,
      "learning_rate": 3.8544166138604124e-05,
      "loss": 0.5621,
      "num_input_tokens_seen": 9626160,
      "step": 14730
    },
    {
      "epoch": 7.722746331236897,
      "grad_norm": 0.5158860087394714,
      "learning_rate": 3.8534553874884244e-05,
      "loss": 0.3455,
      "num_input_tokens_seen": 9629136,
      "step": 14735
    },
    {
      "epoch": 7.7253668763102725,
      "grad_norm": 0.4254954755306244,
      "learning_rate": 3.852493877989011e-05,
      "loss": 0.43,
      "num_input_tokens_seen": 9632080,
      "step": 14740
    },
    {
      "epoch": 7.727987421383648,
      "grad_norm": 0.31969359517097473,
      "learning_rate": 3.851532085563308e-05,
      "loss": 0.5726,
      "num_input_tokens_seen": 9635024,
      "step": 14745
    },
    {
      "epoch": 7.730607966457023,
      "grad_norm": 0.4530249238014221,
      "learning_rate": 3.850570010412511e-05,
      "loss": 0.5041,
      "num_input_tokens_seen": 9638832,
      "step": 14750
    },
    {
      "epoch": 7.733228511530398,
      "grad_norm": 0.4636765718460083,
      "learning_rate": 3.8496076527378766e-05,
      "loss": 0.4078,
      "num_input_tokens_seen": 9641936,
      "step": 14755
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 0.4175536334514618,
      "learning_rate": 3.848645012740718e-05,
      "loss": 0.5174,
      "num_input_tokens_seen": 9645872,
      "step": 14760
    },
    {
      "epoch": 7.738469601677149,
      "grad_norm": 0.27767348289489746,
      "learning_rate": 3.8476820906224096e-05,
      "loss": 0.4472,
      "num_input_tokens_seen": 9649872,
      "step": 14765
    },
    {
      "epoch": 7.741090146750524,
      "grad_norm": 0.668461263179779,
      "learning_rate": 3.846718886584382e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 9653712,
      "step": 14770
    },
    {
      "epoch": 7.7437106918239,
      "grad_norm": 0.6591158509254456,
      "learning_rate": 3.845755400828128e-05,
      "loss": 0.5375,
      "num_input_tokens_seen": 9656560,
      "step": 14775
    },
    {
      "epoch": 7.746331236897275,
      "grad_norm": 0.5460144281387329,
      "learning_rate": 3.844791633555198e-05,
      "loss": 0.5037,
      "num_input_tokens_seen": 9660592,
      "step": 14780
    },
    {
      "epoch": 7.74895178197065,
      "grad_norm": 0.23923639953136444,
      "learning_rate": 3.8438275849671996e-05,
      "loss": 0.5428,
      "num_input_tokens_seen": 9664656,
      "step": 14785
    },
    {
      "epoch": 7.751572327044025,
      "grad_norm": 0.30152615904808044,
      "learning_rate": 3.842863255265802e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 9667888,
      "step": 14790
    },
    {
      "epoch": 7.7541928721174,
      "grad_norm": 0.4210125505924225,
      "learning_rate": 3.841898644652732e-05,
      "loss": 0.4566,
      "num_input_tokens_seen": 9671248,
      "step": 14795
    },
    {
      "epoch": 7.756813417190775,
      "grad_norm": 0.2954987585544586,
      "learning_rate": 3.840933753329774e-05,
      "loss": 0.3726,
      "num_input_tokens_seen": 9674416,
      "step": 14800
    },
    {
      "epoch": 7.759433962264151,
      "grad_norm": 0.6448689103126526,
      "learning_rate": 3.839968581498773e-05,
      "loss": 0.4829,
      "num_input_tokens_seen": 9677808,
      "step": 14805
    },
    {
      "epoch": 7.762054507337526,
      "grad_norm": 0.3972366154193878,
      "learning_rate": 3.839003129361632e-05,
      "loss": 0.4675,
      "num_input_tokens_seen": 9680720,
      "step": 14810
    },
    {
      "epoch": 7.764675052410902,
      "grad_norm": 0.49196815490722656,
      "learning_rate": 3.8380373971203134e-05,
      "loss": 0.563,
      "num_input_tokens_seen": 9684656,
      "step": 14815
    },
    {
      "epoch": 7.767295597484277,
      "grad_norm": 0.49747294187545776,
      "learning_rate": 3.8370713849768346e-05,
      "loss": 0.398,
      "num_input_tokens_seen": 9687760,
      "step": 14820
    },
    {
      "epoch": 7.769916142557652,
      "grad_norm": 0.4867576062679291,
      "learning_rate": 3.8361050931332774e-05,
      "loss": 0.5401,
      "num_input_tokens_seen": 9692016,
      "step": 14825
    },
    {
      "epoch": 7.772536687631027,
      "grad_norm": 0.3117658793926239,
      "learning_rate": 3.835138521791778e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 9695856,
      "step": 14830
    },
    {
      "epoch": 7.7751572327044025,
      "grad_norm": 0.6429283618927002,
      "learning_rate": 3.8341716711545314e-05,
      "loss": 0.6299,
      "num_input_tokens_seen": 9699824,
      "step": 14835
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 0.5408644080162048,
      "learning_rate": 3.833204541423793e-05,
      "loss": 0.4092,
      "num_input_tokens_seen": 9702224,
      "step": 14840
    },
    {
      "epoch": 7.780398322851153,
      "grad_norm": 0.32903504371643066,
      "learning_rate": 3.832237132801874e-05,
      "loss": 0.5622,
      "num_input_tokens_seen": 9704656,
      "step": 14845
    },
    {
      "epoch": 7.783018867924528,
      "grad_norm": 0.36463725566864014,
      "learning_rate": 3.831269445491146e-05,
      "loss": 0.4643,
      "num_input_tokens_seen": 9707472,
      "step": 14850
    },
    {
      "epoch": 7.785639412997903,
      "grad_norm": 0.3846313953399658,
      "learning_rate": 3.8303014796940375e-05,
      "loss": 0.4772,
      "num_input_tokens_seen": 9710896,
      "step": 14855
    },
    {
      "epoch": 7.788259958071279,
      "grad_norm": 0.2848917841911316,
      "learning_rate": 3.829333235613036e-05,
      "loss": 0.5143,
      "num_input_tokens_seen": 9713616,
      "step": 14860
    },
    {
      "epoch": 7.790880503144654,
      "grad_norm": 0.38250818848609924,
      "learning_rate": 3.828364713450688e-05,
      "loss": 0.4502,
      "num_input_tokens_seen": 9716336,
      "step": 14865
    },
    {
      "epoch": 7.79350104821803,
      "grad_norm": 0.571740448474884,
      "learning_rate": 3.827395913409597e-05,
      "loss": 0.4071,
      "num_input_tokens_seen": 9719088,
      "step": 14870
    },
    {
      "epoch": 7.796121593291405,
      "grad_norm": 0.4491315484046936,
      "learning_rate": 3.826426835692424e-05,
      "loss": 0.3655,
      "num_input_tokens_seen": 9722256,
      "step": 14875
    },
    {
      "epoch": 7.79874213836478,
      "grad_norm": 0.4287451505661011,
      "learning_rate": 3.825457480501891e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 9725648,
      "step": 14880
    },
    {
      "epoch": 7.801362683438155,
      "grad_norm": 0.39960601925849915,
      "learning_rate": 3.824487848040773e-05,
      "loss": 0.4072,
      "num_input_tokens_seen": 9728976,
      "step": 14885
    },
    {
      "epoch": 7.80398322851153,
      "grad_norm": 0.44239434599876404,
      "learning_rate": 3.823517938511908e-05,
      "loss": 0.3415,
      "num_input_tokens_seen": 9735472,
      "step": 14890
    },
    {
      "epoch": 7.806603773584905,
      "grad_norm": 0.2655397951602936,
      "learning_rate": 3.8225477521181906e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 9738512,
      "step": 14895
    },
    {
      "epoch": 7.809224318658281,
      "grad_norm": 0.31256169080734253,
      "learning_rate": 3.821577289062571e-05,
      "loss": 0.6942,
      "num_input_tokens_seen": 9742416,
      "step": 14900
    },
    {
      "epoch": 7.811844863731656,
      "grad_norm": 0.6859028935432434,
      "learning_rate": 3.82060654954806e-05,
      "loss": 0.4576,
      "num_input_tokens_seen": 9746800,
      "step": 14905
    },
    {
      "epoch": 7.814465408805032,
      "grad_norm": 0.37982413172721863,
      "learning_rate": 3.819635533777726e-05,
      "loss": 0.37,
      "num_input_tokens_seen": 9750768,
      "step": 14910
    },
    {
      "epoch": 7.817085953878407,
      "grad_norm": 0.3057609796524048,
      "learning_rate": 3.8186642419546924e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 9753968,
      "step": 14915
    },
    {
      "epoch": 7.819706498951782,
      "grad_norm": 5.342855453491211,
      "learning_rate": 3.817692674282145e-05,
      "loss": 0.486,
      "num_input_tokens_seen": 9757744,
      "step": 14920
    },
    {
      "epoch": 7.822327044025157,
      "grad_norm": 0.3625357449054718,
      "learning_rate": 3.816720830963322e-05,
      "loss": 0.5299,
      "num_input_tokens_seen": 9760016,
      "step": 14925
    },
    {
      "epoch": 7.8249475890985325,
      "grad_norm": 0.3065020442008972,
      "learning_rate": 3.815748712201523e-05,
      "loss": 0.4539,
      "num_input_tokens_seen": 9762960,
      "step": 14930
    },
    {
      "epoch": 7.827568134171908,
      "grad_norm": 0.5636818408966064,
      "learning_rate": 3.814776318200105e-05,
      "loss": 0.5413,
      "num_input_tokens_seen": 9767376,
      "step": 14935
    },
    {
      "epoch": 7.830188679245283,
      "grad_norm": 0.47578781843185425,
      "learning_rate": 3.8138036491624805e-05,
      "loss": 0.3607,
      "num_input_tokens_seen": 9770800,
      "step": 14940
    },
    {
      "epoch": 7.832809224318658,
      "grad_norm": 0.45757824182510376,
      "learning_rate": 3.812830705292122e-05,
      "loss": 0.4849,
      "num_input_tokens_seen": 9774288,
      "step": 14945
    },
    {
      "epoch": 7.835429769392033,
      "grad_norm": 0.3315238654613495,
      "learning_rate": 3.8118574867925564e-05,
      "loss": 0.4962,
      "num_input_tokens_seen": 9777296,
      "step": 14950
    },
    {
      "epoch": 7.838050314465409,
      "grad_norm": 0.33540984988212585,
      "learning_rate": 3.810883993867371e-05,
      "loss": 0.3967,
      "num_input_tokens_seen": 9779888,
      "step": 14955
    },
    {
      "epoch": 7.840670859538784,
      "grad_norm": 0.4118601977825165,
      "learning_rate": 3.80991022672021e-05,
      "loss": 0.3297,
      "num_input_tokens_seen": 9783120,
      "step": 14960
    },
    {
      "epoch": 7.84329140461216,
      "grad_norm": 0.7201200127601624,
      "learning_rate": 3.8089361855547744e-05,
      "loss": 0.4814,
      "num_input_tokens_seen": 9786576,
      "step": 14965
    },
    {
      "epoch": 7.845911949685535,
      "grad_norm": 0.4885975420475006,
      "learning_rate": 3.8079618705748214e-05,
      "loss": 0.4028,
      "num_input_tokens_seen": 9790384,
      "step": 14970
    },
    {
      "epoch": 7.84853249475891,
      "grad_norm": 0.5614059567451477,
      "learning_rate": 3.806987281984167e-05,
      "loss": 0.457,
      "num_input_tokens_seen": 9793392,
      "step": 14975
    },
    {
      "epoch": 7.851153039832285,
      "grad_norm": 0.8340751528739929,
      "learning_rate": 3.806012419986684e-05,
      "loss": 0.5937,
      "num_input_tokens_seen": 9796656,
      "step": 14980
    },
    {
      "epoch": 7.85377358490566,
      "grad_norm": 0.5696045160293579,
      "learning_rate": 3.805037284786303e-05,
      "loss": 0.4829,
      "num_input_tokens_seen": 9800720,
      "step": 14985
    },
    {
      "epoch": 7.856394129979035,
      "grad_norm": 0.7735720276832581,
      "learning_rate": 3.804061876587009e-05,
      "loss": 0.3523,
      "num_input_tokens_seen": 9805040,
      "step": 14990
    },
    {
      "epoch": 7.859014675052411,
      "grad_norm": 0.2895691394805908,
      "learning_rate": 3.80308619559285e-05,
      "loss": 0.3383,
      "num_input_tokens_seen": 9808304,
      "step": 14995
    },
    {
      "epoch": 7.861635220125786,
      "grad_norm": 0.5036974549293518,
      "learning_rate": 3.802110242007925e-05,
      "loss": 0.4052,
      "num_input_tokens_seen": 9811120,
      "step": 15000
    },
    {
      "epoch": 7.864255765199162,
      "grad_norm": 0.3375188708305359,
      "learning_rate": 3.801134016036392e-05,
      "loss": 0.512,
      "num_input_tokens_seen": 9813776,
      "step": 15005
    },
    {
      "epoch": 7.866876310272537,
      "grad_norm": 0.43774205446243286,
      "learning_rate": 3.8001575178824665e-05,
      "loss": 0.3629,
      "num_input_tokens_seen": 9816976,
      "step": 15010
    },
    {
      "epoch": 7.869496855345912,
      "grad_norm": 0.3897300958633423,
      "learning_rate": 3.799180747750422e-05,
      "loss": 0.552,
      "num_input_tokens_seen": 9819728,
      "step": 15015
    },
    {
      "epoch": 7.872117400419287,
      "grad_norm": 0.4369770884513855,
      "learning_rate": 3.798203705844586e-05,
      "loss": 0.4999,
      "num_input_tokens_seen": 9823344,
      "step": 15020
    },
    {
      "epoch": 7.8747379454926625,
      "grad_norm": 0.7386045455932617,
      "learning_rate": 3.7972263923693476e-05,
      "loss": 0.4395,
      "num_input_tokens_seen": 9826480,
      "step": 15025
    },
    {
      "epoch": 7.877358490566038,
      "grad_norm": 0.8426761627197266,
      "learning_rate": 3.796248807529145e-05,
      "loss": 0.6231,
      "num_input_tokens_seen": 9829744,
      "step": 15030
    },
    {
      "epoch": 7.879979035639413,
      "grad_norm": 0.3134734034538269,
      "learning_rate": 3.795270951528482e-05,
      "loss": 0.4709,
      "num_input_tokens_seen": 9832624,
      "step": 15035
    },
    {
      "epoch": 7.882599580712788,
      "grad_norm": 0.36391744017601013,
      "learning_rate": 3.7942928245719124e-05,
      "loss": 0.4833,
      "num_input_tokens_seen": 9835600,
      "step": 15040
    },
    {
      "epoch": 7.885220125786163,
      "grad_norm": 0.40328913927078247,
      "learning_rate": 3.7933144268640495e-05,
      "loss": 0.3977,
      "num_input_tokens_seen": 9838448,
      "step": 15045
    },
    {
      "epoch": 7.887840670859539,
      "grad_norm": 0.6195444464683533,
      "learning_rate": 3.7923357586095645e-05,
      "loss": 0.3937,
      "num_input_tokens_seen": 9840528,
      "step": 15050
    },
    {
      "epoch": 7.890461215932914,
      "grad_norm": 0.2965794503688812,
      "learning_rate": 3.7913568200131823e-05,
      "loss": 0.502,
      "num_input_tokens_seen": 9844816,
      "step": 15055
    },
    {
      "epoch": 7.8930817610062896,
      "grad_norm": 0.47261935472488403,
      "learning_rate": 3.790377611279685e-05,
      "loss": 0.4688,
      "num_input_tokens_seen": 9847312,
      "step": 15060
    },
    {
      "epoch": 7.895702306079665,
      "grad_norm": 0.390064001083374,
      "learning_rate": 3.789398132613914e-05,
      "loss": 0.4427,
      "num_input_tokens_seen": 9850960,
      "step": 15065
    },
    {
      "epoch": 7.89832285115304,
      "grad_norm": 0.590141773223877,
      "learning_rate": 3.788418384220763e-05,
      "loss": 0.4715,
      "num_input_tokens_seen": 9854192,
      "step": 15070
    },
    {
      "epoch": 7.900943396226415,
      "grad_norm": 0.4031340181827545,
      "learning_rate": 3.7874383663051846e-05,
      "loss": 0.4862,
      "num_input_tokens_seen": 9857136,
      "step": 15075
    },
    {
      "epoch": 7.90356394129979,
      "grad_norm": 0.321481853723526,
      "learning_rate": 3.7864580790721894e-05,
      "loss": 0.4477,
      "num_input_tokens_seen": 9860304,
      "step": 15080
    },
    {
      "epoch": 7.906184486373165,
      "grad_norm": 0.5555768609046936,
      "learning_rate": 3.785477522726839e-05,
      "loss": 0.4736,
      "num_input_tokens_seen": 9863312,
      "step": 15085
    },
    {
      "epoch": 7.908805031446541,
      "grad_norm": 0.6171147227287292,
      "learning_rate": 3.784496697474258e-05,
      "loss": 0.5019,
      "num_input_tokens_seen": 9866320,
      "step": 15090
    },
    {
      "epoch": 7.911425576519916,
      "grad_norm": 0.6250103712081909,
      "learning_rate": 3.783515603519622e-05,
      "loss": 0.4636,
      "num_input_tokens_seen": 9869840,
      "step": 15095
    },
    {
      "epoch": 7.914046121593291,
      "grad_norm": 0.9850631356239319,
      "learning_rate": 3.782534241068164e-05,
      "loss": 0.3812,
      "num_input_tokens_seen": 9873040,
      "step": 15100
    },
    {
      "epoch": 7.916666666666667,
      "grad_norm": 0.27668729424476624,
      "learning_rate": 3.781552610325175e-05,
      "loss": 0.32,
      "num_input_tokens_seen": 9876240,
      "step": 15105
    },
    {
      "epoch": 7.919287211740042,
      "grad_norm": 0.7263554334640503,
      "learning_rate": 3.780570711496001e-05,
      "loss": 0.4314,
      "num_input_tokens_seen": 9879536,
      "step": 15110
    },
    {
      "epoch": 7.921907756813417,
      "grad_norm": 1.5776242017745972,
      "learning_rate": 3.779588544786044e-05,
      "loss": 0.755,
      "num_input_tokens_seen": 9882256,
      "step": 15115
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 0.6235623359680176,
      "learning_rate": 3.778606110400761e-05,
      "loss": 0.5587,
      "num_input_tokens_seen": 9885936,
      "step": 15120
    },
    {
      "epoch": 7.927148846960168,
      "grad_norm": 0.3908642828464508,
      "learning_rate": 3.7776234085456665e-05,
      "loss": 0.2798,
      "num_input_tokens_seen": 9889072,
      "step": 15125
    },
    {
      "epoch": 7.929769392033543,
      "grad_norm": 0.2717592418193817,
      "learning_rate": 3.776640439426332e-05,
      "loss": 0.4478,
      "num_input_tokens_seen": 9892240,
      "step": 15130
    },
    {
      "epoch": 7.932389937106918,
      "grad_norm": 0.5718669295310974,
      "learning_rate": 3.775657203248381e-05,
      "loss": 0.5896,
      "num_input_tokens_seen": 9895312,
      "step": 15135
    },
    {
      "epoch": 7.935010482180293,
      "grad_norm": 0.2047015279531479,
      "learning_rate": 3.774673700217497e-05,
      "loss": 0.3358,
      "num_input_tokens_seen": 9898800,
      "step": 15140
    },
    {
      "epoch": 7.937631027253669,
      "grad_norm": 0.5791856050491333,
      "learning_rate": 3.773689930539418e-05,
      "loss": 0.4399,
      "num_input_tokens_seen": 9901680,
      "step": 15145
    },
    {
      "epoch": 7.940251572327044,
      "grad_norm": 0.24236047267913818,
      "learning_rate": 3.7727058944199354e-05,
      "loss": 0.307,
      "num_input_tokens_seen": 9905456,
      "step": 15150
    },
    {
      "epoch": 7.9428721174004195,
      "grad_norm": 0.4710235893726349,
      "learning_rate": 3.7717215920648994e-05,
      "loss": 0.3205,
      "num_input_tokens_seen": 9908304,
      "step": 15155
    },
    {
      "epoch": 7.945492662473795,
      "grad_norm": 0.30087974667549133,
      "learning_rate": 3.770737023680215e-05,
      "loss": 0.5064,
      "num_input_tokens_seen": 9911760,
      "step": 15160
    },
    {
      "epoch": 7.94811320754717,
      "grad_norm": 0.38549408316612244,
      "learning_rate": 3.7697521894718436e-05,
      "loss": 0.3553,
      "num_input_tokens_seen": 9914928,
      "step": 15165
    },
    {
      "epoch": 7.950733752620545,
      "grad_norm": 0.6858106851577759,
      "learning_rate": 3.7687670896458e-05,
      "loss": 0.713,
      "num_input_tokens_seen": 9918992,
      "step": 15170
    },
    {
      "epoch": 7.95335429769392,
      "grad_norm": 0.5099686980247498,
      "learning_rate": 3.7677817244081556e-05,
      "loss": 0.4889,
      "num_input_tokens_seen": 9922960,
      "step": 15175
    },
    {
      "epoch": 7.955974842767295,
      "grad_norm": 0.5294631123542786,
      "learning_rate": 3.766796093965038e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 9926320,
      "step": 15180
    },
    {
      "epoch": 7.9585953878406706,
      "grad_norm": 0.41011548042297363,
      "learning_rate": 3.76581019852263e-05,
      "loss": 0.3501,
      "num_input_tokens_seen": 9930000,
      "step": 15185
    },
    {
      "epoch": 7.961215932914046,
      "grad_norm": 0.502672553062439,
      "learning_rate": 3.7648240382871705e-05,
      "loss": 0.4769,
      "num_input_tokens_seen": 9934160,
      "step": 15190
    },
    {
      "epoch": 7.963836477987421,
      "grad_norm": 0.10832806676626205,
      "learning_rate": 3.763837613464951e-05,
      "loss": 0.362,
      "num_input_tokens_seen": 9939760,
      "step": 15195
    },
    {
      "epoch": 7.966457023060797,
      "grad_norm": 0.8162806034088135,
      "learning_rate": 3.7628509242623215e-05,
      "loss": 0.539,
      "num_input_tokens_seen": 9942768,
      "step": 15200
    },
    {
      "epoch": 7.969077568134172,
      "grad_norm": 0.34726032614707947,
      "learning_rate": 3.761863970885686e-05,
      "loss": 0.5482,
      "num_input_tokens_seen": 9946864,
      "step": 15205
    },
    {
      "epoch": 7.971698113207547,
      "grad_norm": 0.5599056482315063,
      "learning_rate": 3.760876753541504e-05,
      "loss": 0.4237,
      "num_input_tokens_seen": 9951664,
      "step": 15210
    },
    {
      "epoch": 7.9743186582809225,
      "grad_norm": 0.824755072593689,
      "learning_rate": 3.7598892724362907e-05,
      "loss": 0.4275,
      "num_input_tokens_seen": 9954096,
      "step": 15215
    },
    {
      "epoch": 7.976939203354298,
      "grad_norm": 0.3998405635356903,
      "learning_rate": 3.758901527776615e-05,
      "loss": 0.5265,
      "num_input_tokens_seen": 9957456,
      "step": 15220
    },
    {
      "epoch": 7.979559748427673,
      "grad_norm": 0.4089649021625519,
      "learning_rate": 3.7579135197691015e-05,
      "loss": 0.5009,
      "num_input_tokens_seen": 9960240,
      "step": 15225
    },
    {
      "epoch": 7.982180293501048,
      "grad_norm": 0.5856416821479797,
      "learning_rate": 3.7569252486204306e-05,
      "loss": 0.579,
      "num_input_tokens_seen": 9962736,
      "step": 15230
    },
    {
      "epoch": 7.984800838574423,
      "grad_norm": 0.39959874749183655,
      "learning_rate": 3.755936714537338e-05,
      "loss": 0.3742,
      "num_input_tokens_seen": 9965424,
      "step": 15235
    },
    {
      "epoch": 7.987421383647799,
      "grad_norm": 0.39151865243911743,
      "learning_rate": 3.754947917726612e-05,
      "loss": 0.36,
      "num_input_tokens_seen": 9968464,
      "step": 15240
    },
    {
      "epoch": 7.990041928721174,
      "grad_norm": 0.4857766032218933,
      "learning_rate": 3.7539588583951e-05,
      "loss": 0.5161,
      "num_input_tokens_seen": 9971152,
      "step": 15245
    },
    {
      "epoch": 7.9926624737945495,
      "grad_norm": 0.44787880778312683,
      "learning_rate": 3.752969536749701e-05,
      "loss": 0.3885,
      "num_input_tokens_seen": 9973712,
      "step": 15250
    },
    {
      "epoch": 7.995283018867925,
      "grad_norm": 0.5555245280265808,
      "learning_rate": 3.7519799529973685e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 9976560,
      "step": 15255
    },
    {
      "epoch": 7.9979035639413,
      "grad_norm": 0.6710485816001892,
      "learning_rate": 3.7509901073451134e-05,
      "loss": 0.4356,
      "num_input_tokens_seen": 9979344,
      "step": 15260
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.4800127148628235,
      "eval_runtime": 9.6398,
      "eval_samples_per_second": 87.969,
      "eval_steps_per_second": 21.992,
      "num_input_tokens_seen": 9981168,
      "step": 15264
    },
    {
      "epoch": 8.000524109014675,
      "grad_norm": 0.31321626901626587,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.3274,
      "num_input_tokens_seen": 9981648,
      "step": 15265
    },
    {
      "epoch": 8.00314465408805,
      "grad_norm": 0.5643694400787354,
      "learning_rate": 3.749009631169147e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 9984048,
      "step": 15270
    },
    {
      "epoch": 8.005765199161425,
      "grad_norm": 0.39389973878860474,
      "learning_rate": 3.74801900105973e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 9987056,
      "step": 15275
    },
    {
      "epoch": 8.0083857442348,
      "grad_norm": 0.45553159713745117,
      "learning_rate": 3.747028109878975e-05,
      "loss": 0.7002,
      "num_input_tokens_seen": 9990512,
      "step": 15280
    },
    {
      "epoch": 8.011006289308176,
      "grad_norm": 0.4021334648132324,
      "learning_rate": 3.746036957834167e-05,
      "loss": 0.396,
      "num_input_tokens_seen": 9993104,
      "step": 15285
    },
    {
      "epoch": 8.01362683438155,
      "grad_norm": 0.6345365047454834,
      "learning_rate": 3.7450455451326415e-05,
      "loss": 0.4337,
      "num_input_tokens_seen": 9996016,
      "step": 15290
    },
    {
      "epoch": 8.016247379454926,
      "grad_norm": 0.4931938648223877,
      "learning_rate": 3.7440538719817935e-05,
      "loss": 0.5309,
      "num_input_tokens_seen": 9998864,
      "step": 15295
    },
    {
      "epoch": 8.018867924528301,
      "grad_norm": 0.4133201539516449,
      "learning_rate": 3.743061938589068e-05,
      "loss": 0.6371,
      "num_input_tokens_seen": 10001424,
      "step": 15300
    },
    {
      "epoch": 8.021488469601676,
      "grad_norm": 0.38599276542663574,
      "learning_rate": 3.742069745161966e-05,
      "loss": 0.4081,
      "num_input_tokens_seen": 10004848,
      "step": 15305
    },
    {
      "epoch": 8.024109014675052,
      "grad_norm": 0.45482051372528076,
      "learning_rate": 3.741077291908045e-05,
      "loss": 0.4224,
      "num_input_tokens_seen": 10008560,
      "step": 15310
    },
    {
      "epoch": 8.026729559748428,
      "grad_norm": 0.2557140588760376,
      "learning_rate": 3.740084579034913e-05,
      "loss": 0.5138,
      "num_input_tokens_seen": 10011664,
      "step": 15315
    },
    {
      "epoch": 8.029350104821804,
      "grad_norm": 0.44990307092666626,
      "learning_rate": 3.739091606750234e-05,
      "loss": 0.4276,
      "num_input_tokens_seen": 10015120,
      "step": 15320
    },
    {
      "epoch": 8.031970649895179,
      "grad_norm": 0.37908026576042175,
      "learning_rate": 3.7380983752617275e-05,
      "loss": 0.5697,
      "num_input_tokens_seen": 10018672,
      "step": 15325
    },
    {
      "epoch": 8.034591194968554,
      "grad_norm": 0.5158067345619202,
      "learning_rate": 3.737104884777166e-05,
      "loss": 0.6364,
      "num_input_tokens_seen": 10022928,
      "step": 15330
    },
    {
      "epoch": 8.03721174004193,
      "grad_norm": 0.4782876670360565,
      "learning_rate": 3.736111135504376e-05,
      "loss": 0.4423,
      "num_input_tokens_seen": 10027216,
      "step": 15335
    },
    {
      "epoch": 8.039832285115304,
      "grad_norm": 0.42506107687950134,
      "learning_rate": 3.735117127651239e-05,
      "loss": 0.4516,
      "num_input_tokens_seen": 10029872,
      "step": 15340
    },
    {
      "epoch": 8.04245283018868,
      "grad_norm": 0.3674429953098297,
      "learning_rate": 3.734122861425689e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 10034352,
      "step": 15345
    },
    {
      "epoch": 8.045073375262055,
      "grad_norm": 0.7156550288200378,
      "learning_rate": 3.7331283370357173e-05,
      "loss": 0.4343,
      "num_input_tokens_seen": 10037520,
      "step": 15350
    },
    {
      "epoch": 8.04769392033543,
      "grad_norm": 0.35040515661239624,
      "learning_rate": 3.732133554689365e-05,
      "loss": 0.4565,
      "num_input_tokens_seen": 10040720,
      "step": 15355
    },
    {
      "epoch": 8.050314465408805,
      "grad_norm": 0.262976735830307,
      "learning_rate": 3.73113851459473e-05,
      "loss": 0.5329,
      "num_input_tokens_seen": 10043760,
      "step": 15360
    },
    {
      "epoch": 8.05293501048218,
      "grad_norm": 0.5211671590805054,
      "learning_rate": 3.730143216959964e-05,
      "loss": 0.4135,
      "num_input_tokens_seen": 10047504,
      "step": 15365
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 0.24408991634845734,
      "learning_rate": 3.729147661993271e-05,
      "loss": 0.3552,
      "num_input_tokens_seen": 10050480,
      "step": 15370
    },
    {
      "epoch": 8.05817610062893,
      "grad_norm": 0.5724484920501709,
      "learning_rate": 3.728151849902909e-05,
      "loss": 0.4263,
      "num_input_tokens_seen": 10053488,
      "step": 15375
    },
    {
      "epoch": 8.060796645702306,
      "grad_norm": 0.33957016468048096,
      "learning_rate": 3.727155780897193e-05,
      "loss": 0.5124,
      "num_input_tokens_seen": 10057744,
      "step": 15380
    },
    {
      "epoch": 8.06341719077568,
      "grad_norm": 0.82330721616745,
      "learning_rate": 3.7261594551844865e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 10060688,
      "step": 15385
    },
    {
      "epoch": 8.066037735849056,
      "grad_norm": 0.4636780321598053,
      "learning_rate": 3.7251628729732116e-05,
      "loss": 0.485,
      "num_input_tokens_seen": 10063408,
      "step": 15390
    },
    {
      "epoch": 8.068658280922431,
      "grad_norm": 0.5386529564857483,
      "learning_rate": 3.724166034471841e-05,
      "loss": 0.4474,
      "num_input_tokens_seen": 10067120,
      "step": 15395
    },
    {
      "epoch": 8.071278825995806,
      "grad_norm": 0.3449230194091797,
      "learning_rate": 3.723168939888901e-05,
      "loss": 0.461,
      "num_input_tokens_seen": 10070768,
      "step": 15400
    },
    {
      "epoch": 8.073899371069182,
      "grad_norm": 0.37806418538093567,
      "learning_rate": 3.7221715894329747e-05,
      "loss": 0.6861,
      "num_input_tokens_seen": 10073776,
      "step": 15405
    },
    {
      "epoch": 8.076519916142558,
      "grad_norm": 0.4060131311416626,
      "learning_rate": 3.7211739833126945e-05,
      "loss": 0.5297,
      "num_input_tokens_seen": 10077488,
      "step": 15410
    },
    {
      "epoch": 8.079140461215934,
      "grad_norm": 0.4523075222969055,
      "learning_rate": 3.720176121736748e-05,
      "loss": 0.5643,
      "num_input_tokens_seen": 10080208,
      "step": 15415
    },
    {
      "epoch": 8.081761006289309,
      "grad_norm": 0.4443657398223877,
      "learning_rate": 3.719178004913879e-05,
      "loss": 0.4289,
      "num_input_tokens_seen": 10083664,
      "step": 15420
    },
    {
      "epoch": 8.084381551362684,
      "grad_norm": 0.5776609182357788,
      "learning_rate": 3.7181796330528796e-05,
      "loss": 0.4374,
      "num_input_tokens_seen": 10086672,
      "step": 15425
    },
    {
      "epoch": 8.08700209643606,
      "grad_norm": 0.46912121772766113,
      "learning_rate": 3.717181006362599e-05,
      "loss": 0.419,
      "num_input_tokens_seen": 10090000,
      "step": 15430
    },
    {
      "epoch": 8.089622641509434,
      "grad_norm": 0.4332001805305481,
      "learning_rate": 3.7161821250519384e-05,
      "loss": 0.4954,
      "num_input_tokens_seen": 10093136,
      "step": 15435
    },
    {
      "epoch": 8.09224318658281,
      "grad_norm": 0.5097130537033081,
      "learning_rate": 3.7151829893298515e-05,
      "loss": 0.4889,
      "num_input_tokens_seen": 10096176,
      "step": 15440
    },
    {
      "epoch": 8.094863731656185,
      "grad_norm": 0.37322160601615906,
      "learning_rate": 3.714183599405347e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 10099728,
      "step": 15445
    },
    {
      "epoch": 8.09748427672956,
      "grad_norm": 0.48447370529174805,
      "learning_rate": 3.713183955487487e-05,
      "loss": 0.4705,
      "num_input_tokens_seen": 10102896,
      "step": 15450
    },
    {
      "epoch": 8.100104821802935,
      "grad_norm": 0.25711220502853394,
      "learning_rate": 3.7121840577853815e-05,
      "loss": 0.3756,
      "num_input_tokens_seen": 10106416,
      "step": 15455
    },
    {
      "epoch": 8.10272536687631,
      "grad_norm": 0.6110478043556213,
      "learning_rate": 3.7111839065082027e-05,
      "loss": 0.518,
      "num_input_tokens_seen": 10108848,
      "step": 15460
    },
    {
      "epoch": 8.105345911949685,
      "grad_norm": 0.41483068466186523,
      "learning_rate": 3.710183501865168e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 10112496,
      "step": 15465
    },
    {
      "epoch": 8.10796645702306,
      "grad_norm": 0.5117811560630798,
      "learning_rate": 3.7091828440655514e-05,
      "loss": 0.3788,
      "num_input_tokens_seen": 10115472,
      "step": 15470
    },
    {
      "epoch": 8.110587002096436,
      "grad_norm": 0.5070021152496338,
      "learning_rate": 3.708181933318678e-05,
      "loss": 0.4274,
      "num_input_tokens_seen": 10117872,
      "step": 15475
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 0.39281409978866577,
      "learning_rate": 3.707180769833929e-05,
      "loss": 0.4665,
      "num_input_tokens_seen": 10120720,
      "step": 15480
    },
    {
      "epoch": 8.115828092243186,
      "grad_norm": 0.7561473250389099,
      "learning_rate": 3.706179353820734e-05,
      "loss": 0.4329,
      "num_input_tokens_seen": 10123152,
      "step": 15485
    },
    {
      "epoch": 8.118448637316561,
      "grad_norm": 0.3831307888031006,
      "learning_rate": 3.7051776854885794e-05,
      "loss": 0.649,
      "num_input_tokens_seen": 10126288,
      "step": 15490
    },
    {
      "epoch": 8.121069182389936,
      "grad_norm": 0.46085307002067566,
      "learning_rate": 3.7041757650470025e-05,
      "loss": 0.3683,
      "num_input_tokens_seen": 10129232,
      "step": 15495
    },
    {
      "epoch": 8.123689727463312,
      "grad_norm": 0.5832905173301697,
      "learning_rate": 3.703173592705594e-05,
      "loss": 0.437,
      "num_input_tokens_seen": 10131664,
      "step": 15500
    },
    {
      "epoch": 8.126310272536688,
      "grad_norm": 0.38318002223968506,
      "learning_rate": 3.702171168673996e-05,
      "loss": 0.2918,
      "num_input_tokens_seen": 10134736,
      "step": 15505
    },
    {
      "epoch": 8.128930817610064,
      "grad_norm": 0.6138721704483032,
      "learning_rate": 3.701168493161905e-05,
      "loss": 0.4277,
      "num_input_tokens_seen": 10138192,
      "step": 15510
    },
    {
      "epoch": 8.131551362683439,
      "grad_norm": 0.345540314912796,
      "learning_rate": 3.7001655663790676e-05,
      "loss": 0.3521,
      "num_input_tokens_seen": 10141232,
      "step": 15515
    },
    {
      "epoch": 8.134171907756814,
      "grad_norm": 0.45676130056381226,
      "learning_rate": 3.699162388535288e-05,
      "loss": 0.5251,
      "num_input_tokens_seen": 10144464,
      "step": 15520
    },
    {
      "epoch": 8.13679245283019,
      "grad_norm": 0.3927009701728821,
      "learning_rate": 3.698158959840416e-05,
      "loss": 0.5394,
      "num_input_tokens_seen": 10148048,
      "step": 15525
    },
    {
      "epoch": 8.139412997903564,
      "grad_norm": 0.4460362493991852,
      "learning_rate": 3.6971552805043585e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 10151600,
      "step": 15530
    },
    {
      "epoch": 8.14203354297694,
      "grad_norm": 0.41594377160072327,
      "learning_rate": 3.6961513507370754e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 10154032,
      "step": 15535
    },
    {
      "epoch": 8.144654088050315,
      "grad_norm": 0.5089478492736816,
      "learning_rate": 3.695147170748576e-05,
      "loss": 0.4893,
      "num_input_tokens_seen": 10156368,
      "step": 15540
    },
    {
      "epoch": 8.14727463312369,
      "grad_norm": 0.2847643196582794,
      "learning_rate": 3.694142740748923e-05,
      "loss": 0.3308,
      "num_input_tokens_seen": 10159504,
      "step": 15545
    },
    {
      "epoch": 8.149895178197065,
      "grad_norm": 0.40541860461235046,
      "learning_rate": 3.693138060948232e-05,
      "loss": 0.3252,
      "num_input_tokens_seen": 10162864,
      "step": 15550
    },
    {
      "epoch": 8.15251572327044,
      "grad_norm": 0.4953452944755554,
      "learning_rate": 3.692133131556671e-05,
      "loss": 0.4607,
      "num_input_tokens_seen": 10165744,
      "step": 15555
    },
    {
      "epoch": 8.155136268343815,
      "grad_norm": 0.5393906831741333,
      "learning_rate": 3.6911279527844596e-05,
      "loss": 0.4345,
      "num_input_tokens_seen": 10169456,
      "step": 15560
    },
    {
      "epoch": 8.15775681341719,
      "grad_norm": 0.3413529396057129,
      "learning_rate": 3.69012252484187e-05,
      "loss": 0.4335,
      "num_input_tokens_seen": 10172048,
      "step": 15565
    },
    {
      "epoch": 8.160377358490566,
      "grad_norm": 0.30958813428878784,
      "learning_rate": 3.6891168479392245e-05,
      "loss": 0.5157,
      "num_input_tokens_seen": 10175088,
      "step": 15570
    },
    {
      "epoch": 8.16299790356394,
      "grad_norm": 0.3844253718852997,
      "learning_rate": 3.688110922286901e-05,
      "loss": 0.4906,
      "num_input_tokens_seen": 10177744,
      "step": 15575
    },
    {
      "epoch": 8.165618448637316,
      "grad_norm": 0.5431311726570129,
      "learning_rate": 3.687104748095328e-05,
      "loss": 0.534,
      "num_input_tokens_seen": 10180528,
      "step": 15580
    },
    {
      "epoch": 8.168238993710691,
      "grad_norm": 0.6544564366340637,
      "learning_rate": 3.686098325574984e-05,
      "loss": 0.5562,
      "num_input_tokens_seen": 10182896,
      "step": 15585
    },
    {
      "epoch": 8.170859538784066,
      "grad_norm": 0.33799806237220764,
      "learning_rate": 3.685091654936403e-05,
      "loss": 0.3975,
      "num_input_tokens_seen": 10186224,
      "step": 15590
    },
    {
      "epoch": 8.173480083857442,
      "grad_norm": 0.35818082094192505,
      "learning_rate": 3.6840847363901675e-05,
      "loss": 0.3893,
      "num_input_tokens_seen": 10190064,
      "step": 15595
    },
    {
      "epoch": 8.176100628930818,
      "grad_norm": 0.6708997488021851,
      "learning_rate": 3.683077570146914e-05,
      "loss": 0.4796,
      "num_input_tokens_seen": 10192720,
      "step": 15600
    },
    {
      "epoch": 8.178721174004194,
      "grad_norm": 0.4221639931201935,
      "learning_rate": 3.68207015641733e-05,
      "loss": 0.4056,
      "num_input_tokens_seen": 10196848,
      "step": 15605
    },
    {
      "epoch": 8.181341719077569,
      "grad_norm": 0.36012011766433716,
      "learning_rate": 3.681062495412155e-05,
      "loss": 0.456,
      "num_input_tokens_seen": 10200400,
      "step": 15610
    },
    {
      "epoch": 8.183962264150944,
      "grad_norm": 0.5642812848091125,
      "learning_rate": 3.68005458734218e-05,
      "loss": 0.43,
      "num_input_tokens_seen": 10204368,
      "step": 15615
    },
    {
      "epoch": 8.18658280922432,
      "grad_norm": 0.5294137597084045,
      "learning_rate": 3.679046432418248e-05,
      "loss": 0.3931,
      "num_input_tokens_seen": 10208592,
      "step": 15620
    },
    {
      "epoch": 8.189203354297694,
      "grad_norm": 0.3527871072292328,
      "learning_rate": 3.6780380308512543e-05,
      "loss": 0.4334,
      "num_input_tokens_seen": 10212304,
      "step": 15625
    },
    {
      "epoch": 8.19182389937107,
      "grad_norm": 0.552361249923706,
      "learning_rate": 3.677029382852144e-05,
      "loss": 0.3721,
      "num_input_tokens_seen": 10215504,
      "step": 15630
    },
    {
      "epoch": 8.194444444444445,
      "grad_norm": 0.6101449131965637,
      "learning_rate": 3.676020488631915e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 10218768,
      "step": 15635
    },
    {
      "epoch": 8.19706498951782,
      "grad_norm": 0.5612964630126953,
      "learning_rate": 3.675011348401617e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 10222640,
      "step": 15640
    },
    {
      "epoch": 8.199685534591195,
      "grad_norm": 0.800303041934967,
      "learning_rate": 3.6740019623723494e-05,
      "loss": 0.5055,
      "num_input_tokens_seen": 10225008,
      "step": 15645
    },
    {
      "epoch": 8.20230607966457,
      "grad_norm": 0.4356004595756531,
      "learning_rate": 3.672992330755265e-05,
      "loss": 0.4277,
      "num_input_tokens_seen": 10228208,
      "step": 15650
    },
    {
      "epoch": 8.204926624737945,
      "grad_norm": 0.514667272567749,
      "learning_rate": 3.671982453761568e-05,
      "loss": 0.4653,
      "num_input_tokens_seen": 10232080,
      "step": 15655
    },
    {
      "epoch": 8.20754716981132,
      "grad_norm": 0.4543689489364624,
      "learning_rate": 3.6709723316025114e-05,
      "loss": 0.4454,
      "num_input_tokens_seen": 10234704,
      "step": 15660
    },
    {
      "epoch": 8.210167714884696,
      "grad_norm": 0.7312958240509033,
      "learning_rate": 3.669961964489403e-05,
      "loss": 0.3885,
      "num_input_tokens_seen": 10237872,
      "step": 15665
    },
    {
      "epoch": 8.21278825995807,
      "grad_norm": 0.43127191066741943,
      "learning_rate": 3.668951352633599e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 10240784,
      "step": 15670
    },
    {
      "epoch": 8.215408805031446,
      "grad_norm": 0.7656546831130981,
      "learning_rate": 3.6679404962465084e-05,
      "loss": 0.383,
      "num_input_tokens_seen": 10243440,
      "step": 15675
    },
    {
      "epoch": 8.218029350104821,
      "grad_norm": 0.2608190178871155,
      "learning_rate": 3.66692939553959e-05,
      "loss": 0.4605,
      "num_input_tokens_seen": 10246864,
      "step": 15680
    },
    {
      "epoch": 8.220649895178196,
      "grad_norm": 0.39194947481155396,
      "learning_rate": 3.6659180507243554e-05,
      "loss": 0.5128,
      "num_input_tokens_seen": 10249360,
      "step": 15685
    },
    {
      "epoch": 8.223270440251572,
      "grad_norm": 0.3914473354816437,
      "learning_rate": 3.664906462012366e-05,
      "loss": 0.5091,
      "num_input_tokens_seen": 10252848,
      "step": 15690
    },
    {
      "epoch": 8.225890985324948,
      "grad_norm": 0.29648372530937195,
      "learning_rate": 3.663894629615234e-05,
      "loss": 0.5295,
      "num_input_tokens_seen": 10257680,
      "step": 15695
    },
    {
      "epoch": 8.228511530398324,
      "grad_norm": 0.31302306056022644,
      "learning_rate": 3.6628825537446246e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 10261520,
      "step": 15700
    },
    {
      "epoch": 8.231132075471699,
      "grad_norm": 0.2619929611682892,
      "learning_rate": 3.6618702346122525e-05,
      "loss": 0.3668,
      "num_input_tokens_seen": 10264624,
      "step": 15705
    },
    {
      "epoch": 8.233752620545074,
      "grad_norm": 0.3919757306575775,
      "learning_rate": 3.6608576724298817e-05,
      "loss": 0.4554,
      "num_input_tokens_seen": 10267664,
      "step": 15710
    },
    {
      "epoch": 8.23637316561845,
      "grad_norm": 0.7625667452812195,
      "learning_rate": 3.65984486740933e-05,
      "loss": 0.5229,
      "num_input_tokens_seen": 10270000,
      "step": 15715
    },
    {
      "epoch": 8.238993710691824,
      "grad_norm": 0.4042195677757263,
      "learning_rate": 3.6588318197624644e-05,
      "loss": 0.4468,
      "num_input_tokens_seen": 10273424,
      "step": 15720
    },
    {
      "epoch": 8.2416142557652,
      "grad_norm": 0.47934627532958984,
      "learning_rate": 3.6578185297012015e-05,
      "loss": 0.4738,
      "num_input_tokens_seen": 10277264,
      "step": 15725
    },
    {
      "epoch": 8.244234800838575,
      "grad_norm": 0.38866791129112244,
      "learning_rate": 3.656804997437512e-05,
      "loss": 0.4689,
      "num_input_tokens_seen": 10280080,
      "step": 15730
    },
    {
      "epoch": 8.24685534591195,
      "grad_norm": 0.3334875702857971,
      "learning_rate": 3.6557912231834155e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 10283216,
      "step": 15735
    },
    {
      "epoch": 8.249475890985325,
      "grad_norm": 0.5467429757118225,
      "learning_rate": 3.65477720715098e-05,
      "loss": 0.4987,
      "num_input_tokens_seen": 10285968,
      "step": 15740
    },
    {
      "epoch": 8.2520964360587,
      "grad_norm": 0.38454607129096985,
      "learning_rate": 3.653762949552328e-05,
      "loss": 0.3617,
      "num_input_tokens_seen": 10288784,
      "step": 15745
    },
    {
      "epoch": 8.254716981132075,
      "grad_norm": 0.42080381512641907,
      "learning_rate": 3.6527484505996295e-05,
      "loss": 0.394,
      "num_input_tokens_seen": 10293904,
      "step": 15750
    },
    {
      "epoch": 8.25733752620545,
      "grad_norm": 0.4762851893901825,
      "learning_rate": 3.651733710505106e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 10297488,
      "step": 15755
    },
    {
      "epoch": 8.259958071278826,
      "grad_norm": 0.5992440581321716,
      "learning_rate": 3.6507187294810305e-05,
      "loss": 0.5649,
      "num_input_tokens_seen": 10301392,
      "step": 15760
    },
    {
      "epoch": 8.2625786163522,
      "grad_norm": 0.34598490595817566,
      "learning_rate": 3.649703507739724e-05,
      "loss": 0.3491,
      "num_input_tokens_seen": 10303888,
      "step": 15765
    },
    {
      "epoch": 8.265199161425576,
      "grad_norm": 0.3624088168144226,
      "learning_rate": 3.6486880454935606e-05,
      "loss": 0.3101,
      "num_input_tokens_seen": 10307888,
      "step": 15770
    },
    {
      "epoch": 8.267819706498951,
      "grad_norm": 1.0680395364761353,
      "learning_rate": 3.647672342954963e-05,
      "loss": 0.5775,
      "num_input_tokens_seen": 10310768,
      "step": 15775
    },
    {
      "epoch": 8.270440251572326,
      "grad_norm": 0.9285266399383545,
      "learning_rate": 3.646656400336405e-05,
      "loss": 0.4654,
      "num_input_tokens_seen": 10313936,
      "step": 15780
    },
    {
      "epoch": 8.273060796645701,
      "grad_norm": 0.28277698159217834,
      "learning_rate": 3.6456402178504094e-05,
      "loss": 0.4208,
      "num_input_tokens_seen": 10319632,
      "step": 15785
    },
    {
      "epoch": 8.275681341719078,
      "grad_norm": 0.38375985622406006,
      "learning_rate": 3.6446237957095505e-05,
      "loss": 0.5365,
      "num_input_tokens_seen": 10322800,
      "step": 15790
    },
    {
      "epoch": 8.278301886792454,
      "grad_norm": 0.5261811017990112,
      "learning_rate": 3.643607134126452e-05,
      "loss": 0.4708,
      "num_input_tokens_seen": 10326224,
      "step": 15795
    },
    {
      "epoch": 8.280922431865829,
      "grad_norm": 0.5025452375411987,
      "learning_rate": 3.642590233313789e-05,
      "loss": 0.6123,
      "num_input_tokens_seen": 10328912,
      "step": 15800
    },
    {
      "epoch": 8.283542976939204,
      "grad_norm": 0.3298271596431732,
      "learning_rate": 3.6415730934842827e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 10331728,
      "step": 15805
    },
    {
      "epoch": 8.286163522012579,
      "grad_norm": 0.3071727454662323,
      "learning_rate": 3.64055571485071e-05,
      "loss": 0.4215,
      "num_input_tokens_seen": 10335056,
      "step": 15810
    },
    {
      "epoch": 8.288784067085954,
      "grad_norm": 0.32967376708984375,
      "learning_rate": 3.639538097625894e-05,
      "loss": 0.4815,
      "num_input_tokens_seen": 10338384,
      "step": 15815
    },
    {
      "epoch": 8.29140461215933,
      "grad_norm": 0.3176080286502838,
      "learning_rate": 3.638520242022708e-05,
      "loss": 0.3901,
      "num_input_tokens_seen": 10341008,
      "step": 15820
    },
    {
      "epoch": 8.294025157232705,
      "grad_norm": 0.2982490360736847,
      "learning_rate": 3.637502148254077e-05,
      "loss": 0.5158,
      "num_input_tokens_seen": 10344336,
      "step": 15825
    },
    {
      "epoch": 8.29664570230608,
      "grad_norm": 0.2275988757610321,
      "learning_rate": 3.636483816532974e-05,
      "loss": 0.3418,
      "num_input_tokens_seen": 10348432,
      "step": 15830
    },
    {
      "epoch": 8.299266247379455,
      "grad_norm": 0.5947254300117493,
      "learning_rate": 3.6354652470724217e-05,
      "loss": 0.4905,
      "num_input_tokens_seen": 10351056,
      "step": 15835
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 0.30016452074050903,
      "learning_rate": 3.634446440085494e-05,
      "loss": 0.3148,
      "num_input_tokens_seen": 10353680,
      "step": 15840
    },
    {
      "epoch": 8.304507337526205,
      "grad_norm": 0.4593086242675781,
      "learning_rate": 3.633427395785313e-05,
      "loss": 0.4122,
      "num_input_tokens_seen": 10358160,
      "step": 15845
    },
    {
      "epoch": 8.30712788259958,
      "grad_norm": 0.4241732060909271,
      "learning_rate": 3.6324081143850516e-05,
      "loss": 0.4236,
      "num_input_tokens_seen": 10360976,
      "step": 15850
    },
    {
      "epoch": 8.309748427672956,
      "grad_norm": 0.6788349747657776,
      "learning_rate": 3.6313885960979326e-05,
      "loss": 0.3839,
      "num_input_tokens_seen": 10363600,
      "step": 15855
    },
    {
      "epoch": 8.31236897274633,
      "grad_norm": 0.383948415517807,
      "learning_rate": 3.630368841137226e-05,
      "loss": 0.4917,
      "num_input_tokens_seen": 10365744,
      "step": 15860
    },
    {
      "epoch": 8.314989517819706,
      "grad_norm": 0.3040434420108795,
      "learning_rate": 3.6293488497162554e-05,
      "loss": 0.5367,
      "num_input_tokens_seen": 10369104,
      "step": 15865
    },
    {
      "epoch": 8.317610062893081,
      "grad_norm": 0.3206924796104431,
      "learning_rate": 3.628328622048388e-05,
      "loss": 0.4779,
      "num_input_tokens_seen": 10371696,
      "step": 15870
    },
    {
      "epoch": 8.320230607966456,
      "grad_norm": 0.52820885181427,
      "learning_rate": 3.627308158347046e-05,
      "loss": 0.5305,
      "num_input_tokens_seen": 10375184,
      "step": 15875
    },
    {
      "epoch": 8.322851153039831,
      "grad_norm": 0.5829315185546875,
      "learning_rate": 3.626287458825699e-05,
      "loss": 0.4171,
      "num_input_tokens_seen": 10378288,
      "step": 15880
    },
    {
      "epoch": 8.325471698113208,
      "grad_norm": 0.6169010996818542,
      "learning_rate": 3.625266523697864e-05,
      "loss": 0.3827,
      "num_input_tokens_seen": 10380848,
      "step": 15885
    },
    {
      "epoch": 8.328092243186584,
      "grad_norm": 0.473436176776886,
      "learning_rate": 3.62424535317711e-05,
      "loss": 0.5057,
      "num_input_tokens_seen": 10383888,
      "step": 15890
    },
    {
      "epoch": 8.330712788259959,
      "grad_norm": 0.6509637832641602,
      "learning_rate": 3.623223947477055e-05,
      "loss": 0.3294,
      "num_input_tokens_seen": 10388368,
      "step": 15895
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.38287708163261414,
      "learning_rate": 3.6222023068113635e-05,
      "loss": 0.4614,
      "num_input_tokens_seen": 10391280,
      "step": 15900
    },
    {
      "epoch": 8.335953878406709,
      "grad_norm": 0.3009696900844574,
      "learning_rate": 3.621180431393754e-05,
      "loss": 0.3975,
      "num_input_tokens_seen": 10394384,
      "step": 15905
    },
    {
      "epoch": 8.338574423480084,
      "grad_norm": 0.5011253356933594,
      "learning_rate": 3.6201583214379874e-05,
      "loss": 0.3125,
      "num_input_tokens_seen": 10397744,
      "step": 15910
    },
    {
      "epoch": 8.34119496855346,
      "grad_norm": 0.5074350237846375,
      "learning_rate": 3.61913597715788e-05,
      "loss": 0.4067,
      "num_input_tokens_seen": 10401712,
      "step": 15915
    },
    {
      "epoch": 8.343815513626835,
      "grad_norm": 0.35309770703315735,
      "learning_rate": 3.6181133987672944e-05,
      "loss": 0.3551,
      "num_input_tokens_seen": 10404592,
      "step": 15920
    },
    {
      "epoch": 8.34643605870021,
      "grad_norm": 0.5151453614234924,
      "learning_rate": 3.617090586480142e-05,
      "loss": 0.5519,
      "num_input_tokens_seen": 10411120,
      "step": 15925
    },
    {
      "epoch": 8.349056603773585,
      "grad_norm": 0.39565804600715637,
      "learning_rate": 3.6160675405103826e-05,
      "loss": 0.43,
      "num_input_tokens_seen": 10414416,
      "step": 15930
    },
    {
      "epoch": 8.35167714884696,
      "grad_norm": 0.48478564620018005,
      "learning_rate": 3.615044261072027e-05,
      "loss": 0.4885,
      "num_input_tokens_seen": 10416944,
      "step": 15935
    },
    {
      "epoch": 8.354297693920335,
      "grad_norm": 0.3159739077091217,
      "learning_rate": 3.614020748379133e-05,
      "loss": 0.4933,
      "num_input_tokens_seen": 10419472,
      "step": 15940
    },
    {
      "epoch": 8.35691823899371,
      "grad_norm": 0.3481249213218689,
      "learning_rate": 3.6129970026458076e-05,
      "loss": 0.4455,
      "num_input_tokens_seen": 10422384,
      "step": 15945
    },
    {
      "epoch": 8.359538784067086,
      "grad_norm": 0.6329312920570374,
      "learning_rate": 3.611973024086207e-05,
      "loss": 0.5922,
      "num_input_tokens_seen": 10426832,
      "step": 15950
    },
    {
      "epoch": 8.36215932914046,
      "grad_norm": 0.4566398859024048,
      "learning_rate": 3.610948812914536e-05,
      "loss": 0.5798,
      "num_input_tokens_seen": 10429680,
      "step": 15955
    },
    {
      "epoch": 8.364779874213836,
      "grad_norm": 0.39137521386146545,
      "learning_rate": 3.609924369345048e-05,
      "loss": 0.4567,
      "num_input_tokens_seen": 10432240,
      "step": 15960
    },
    {
      "epoch": 8.367400419287211,
      "grad_norm": 1.178003191947937,
      "learning_rate": 3.6088996935920444e-05,
      "loss": 0.5664,
      "num_input_tokens_seen": 10434608,
      "step": 15965
    },
    {
      "epoch": 8.370020964360586,
      "grad_norm": 0.31517934799194336,
      "learning_rate": 3.607874785869876e-05,
      "loss": 0.534,
      "num_input_tokens_seen": 10437648,
      "step": 15970
    },
    {
      "epoch": 8.372641509433961,
      "grad_norm": 0.4342285096645355,
      "learning_rate": 3.6068496463929416e-05,
      "loss": 0.3367,
      "num_input_tokens_seen": 10440688,
      "step": 15975
    },
    {
      "epoch": 8.375262054507338,
      "grad_norm": 0.42431655526161194,
      "learning_rate": 3.605824275375689e-05,
      "loss": 0.4309,
      "num_input_tokens_seen": 10442992,
      "step": 15980
    },
    {
      "epoch": 8.377882599580714,
      "grad_norm": 0.6232920289039612,
      "learning_rate": 3.604798673032613e-05,
      "loss": 0.4202,
      "num_input_tokens_seen": 10445776,
      "step": 15985
    },
    {
      "epoch": 8.380503144654089,
      "grad_norm": 0.37245461344718933,
      "learning_rate": 3.60377283957826e-05,
      "loss": 0.4345,
      "num_input_tokens_seen": 10449040,
      "step": 15990
    },
    {
      "epoch": 8.383123689727464,
      "grad_norm": 0.43510183691978455,
      "learning_rate": 3.6027467752272203e-05,
      "loss": 0.4113,
      "num_input_tokens_seen": 10453104,
      "step": 15995
    },
    {
      "epoch": 8.385744234800839,
      "grad_norm": 0.46460890769958496,
      "learning_rate": 3.601720480194137e-05,
      "loss": 0.3872,
      "num_input_tokens_seen": 10456912,
      "step": 16000
    },
    {
      "epoch": 8.388364779874214,
      "grad_norm": 0.6088123321533203,
      "learning_rate": 3.600693954693698e-05,
      "loss": 0.4797,
      "num_input_tokens_seen": 10459920,
      "step": 16005
    },
    {
      "epoch": 8.39098532494759,
      "grad_norm": 0.6110098361968994,
      "learning_rate": 3.5996671989406414e-05,
      "loss": 0.3762,
      "num_input_tokens_seen": 10462992,
      "step": 16010
    },
    {
      "epoch": 8.393605870020965,
      "grad_norm": 0.6034775972366333,
      "learning_rate": 3.598640213149752e-05,
      "loss": 0.5251,
      "num_input_tokens_seen": 10466000,
      "step": 16015
    },
    {
      "epoch": 8.39622641509434,
      "grad_norm": 0.293232798576355,
      "learning_rate": 3.597612997535864e-05,
      "loss": 0.4279,
      "num_input_tokens_seen": 10470288,
      "step": 16020
    },
    {
      "epoch": 8.398846960167715,
      "grad_norm": 0.4493834972381592,
      "learning_rate": 3.596585552313859e-05,
      "loss": 0.405,
      "num_input_tokens_seen": 10473776,
      "step": 16025
    },
    {
      "epoch": 8.40146750524109,
      "grad_norm": 0.4066992402076721,
      "learning_rate": 3.595557877698666e-05,
      "loss": 0.5622,
      "num_input_tokens_seen": 10477200,
      "step": 16030
    },
    {
      "epoch": 8.404088050314465,
      "grad_norm": 0.6944375038146973,
      "learning_rate": 3.594529973905266e-05,
      "loss": 0.4327,
      "num_input_tokens_seen": 10480208,
      "step": 16035
    },
    {
      "epoch": 8.40670859538784,
      "grad_norm": 0.7119035124778748,
      "learning_rate": 3.593501841148681e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 10483536,
      "step": 16040
    },
    {
      "epoch": 8.409329140461216,
      "grad_norm": 0.3818866014480591,
      "learning_rate": 3.5924734796439864e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 10486704,
      "step": 16045
    },
    {
      "epoch": 8.41194968553459,
      "grad_norm": 0.6067362427711487,
      "learning_rate": 3.5914448896063044e-05,
      "loss": 0.4719,
      "num_input_tokens_seen": 10489648,
      "step": 16050
    },
    {
      "epoch": 8.414570230607966,
      "grad_norm": 0.6031538844108582,
      "learning_rate": 3.590416071250802e-05,
      "loss": 0.5162,
      "num_input_tokens_seen": 10493200,
      "step": 16055
    },
    {
      "epoch": 8.417190775681341,
      "grad_norm": 0.5556190013885498,
      "learning_rate": 3.589387024792699e-05,
      "loss": 0.4646,
      "num_input_tokens_seen": 10495248,
      "step": 16060
    },
    {
      "epoch": 8.419811320754716,
      "grad_norm": 0.9166865348815918,
      "learning_rate": 3.5883577504472576e-05,
      "loss": 0.4838,
      "num_input_tokens_seen": 10499088,
      "step": 16065
    },
    {
      "epoch": 8.422431865828091,
      "grad_norm": 0.2751018702983856,
      "learning_rate": 3.587328248429792e-05,
      "loss": 0.461,
      "num_input_tokens_seen": 10502224,
      "step": 16070
    },
    {
      "epoch": 8.425052410901468,
      "grad_norm": 0.3655848503112793,
      "learning_rate": 3.586298518955661e-05,
      "loss": 0.4124,
      "num_input_tokens_seen": 10505200,
      "step": 16075
    },
    {
      "epoch": 8.427672955974844,
      "grad_norm": 0.5745839476585388,
      "learning_rate": 3.5852685622402746e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 10509168,
      "step": 16080
    },
    {
      "epoch": 8.430293501048219,
      "grad_norm": 0.5334933996200562,
      "learning_rate": 3.584238378499085e-05,
      "loss": 0.5129,
      "num_input_tokens_seen": 10512432,
      "step": 16085
    },
    {
      "epoch": 8.432914046121594,
      "grad_norm": 0.4332554042339325,
      "learning_rate": 3.5832079679475974e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 10514832,
      "step": 16090
    },
    {
      "epoch": 8.435534591194969,
      "grad_norm": 0.2584438621997833,
      "learning_rate": 3.582177330801359e-05,
      "loss": 0.4344,
      "num_input_tokens_seen": 10517840,
      "step": 16095
    },
    {
      "epoch": 8.438155136268344,
      "grad_norm": 0.38120320439338684,
      "learning_rate": 3.58114646727597e-05,
      "loss": 0.3172,
      "num_input_tokens_seen": 10520912,
      "step": 16100
    },
    {
      "epoch": 8.44077568134172,
      "grad_norm": 0.31566348671913147,
      "learning_rate": 3.580115377587075e-05,
      "loss": 0.6178,
      "num_input_tokens_seen": 10524048,
      "step": 16105
    },
    {
      "epoch": 8.443396226415095,
      "grad_norm": 0.3748965263366699,
      "learning_rate": 3.5790840619503654e-05,
      "loss": 0.4658,
      "num_input_tokens_seen": 10527120,
      "step": 16110
    },
    {
      "epoch": 8.44601677148847,
      "grad_norm": 0.3335381746292114,
      "learning_rate": 3.57805252058158e-05,
      "loss": 0.4893,
      "num_input_tokens_seen": 10530704,
      "step": 16115
    },
    {
      "epoch": 8.448637316561845,
      "grad_norm": 0.5328220129013062,
      "learning_rate": 3.577020753696507e-05,
      "loss": 0.4719,
      "num_input_tokens_seen": 10536560,
      "step": 16120
    },
    {
      "epoch": 8.45125786163522,
      "grad_norm": 0.3607673645019531,
      "learning_rate": 3.57598876151098e-05,
      "loss": 0.3697,
      "num_input_tokens_seen": 10539824,
      "step": 16125
    },
    {
      "epoch": 8.453878406708595,
      "grad_norm": 0.5468893647193909,
      "learning_rate": 3.5749565442408796e-05,
      "loss": 0.5307,
      "num_input_tokens_seen": 10542352,
      "step": 16130
    },
    {
      "epoch": 8.45649895178197,
      "grad_norm": 0.5942029356956482,
      "learning_rate": 3.573924102102133e-05,
      "loss": 0.5229,
      "num_input_tokens_seen": 10545872,
      "step": 16135
    },
    {
      "epoch": 8.459119496855346,
      "grad_norm": 0.344964861869812,
      "learning_rate": 3.572891435310717e-05,
      "loss": 0.4852,
      "num_input_tokens_seen": 10550960,
      "step": 16140
    },
    {
      "epoch": 8.46174004192872,
      "grad_norm": 1.3087986707687378,
      "learning_rate": 3.5718585440826527e-05,
      "loss": 0.6374,
      "num_input_tokens_seen": 10553712,
      "step": 16145
    },
    {
      "epoch": 8.464360587002096,
      "grad_norm": 0.550052285194397,
      "learning_rate": 3.570825428634009e-05,
      "loss": 0.3803,
      "num_input_tokens_seen": 10556560,
      "step": 16150
    },
    {
      "epoch": 8.466981132075471,
      "grad_norm": 0.3503340482711792,
      "learning_rate": 3.569792089180903e-05,
      "loss": 0.5515,
      "num_input_tokens_seen": 10560112,
      "step": 16155
    },
    {
      "epoch": 8.469601677148846,
      "grad_norm": 0.4480438530445099,
      "learning_rate": 3.568758525939497e-05,
      "loss": 0.4866,
      "num_input_tokens_seen": 10563504,
      "step": 16160
    },
    {
      "epoch": 8.472222222222221,
      "grad_norm": 0.46228456497192383,
      "learning_rate": 3.5677247391259995e-05,
      "loss": 0.5003,
      "num_input_tokens_seen": 10567568,
      "step": 16165
    },
    {
      "epoch": 8.474842767295598,
      "grad_norm": 0.4738307297229767,
      "learning_rate": 3.566690728956668e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 10570448,
      "step": 16170
    },
    {
      "epoch": 8.477463312368974,
      "grad_norm": 0.6751120090484619,
      "learning_rate": 3.5656564956478056e-05,
      "loss": 0.4719,
      "num_input_tokens_seen": 10573264,
      "step": 16175
    },
    {
      "epoch": 8.480083857442349,
      "grad_norm": 0.7927449941635132,
      "learning_rate": 3.5646220394157626e-05,
      "loss": 0.4976,
      "num_input_tokens_seen": 10577264,
      "step": 16180
    },
    {
      "epoch": 8.482704402515724,
      "grad_norm": 0.5119681358337402,
      "learning_rate": 3.563587360476933e-05,
      "loss": 0.4507,
      "num_input_tokens_seen": 10580080,
      "step": 16185
    },
    {
      "epoch": 8.485324947589099,
      "grad_norm": 0.3829830288887024,
      "learning_rate": 3.562552459047763e-05,
      "loss": 0.4555,
      "num_input_tokens_seen": 10582736,
      "step": 16190
    },
    {
      "epoch": 8.487945492662474,
      "grad_norm": 0.4166693687438965,
      "learning_rate": 3.561517335344741e-05,
      "loss": 0.4072,
      "num_input_tokens_seen": 10586064,
      "step": 16195
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 0.4817931652069092,
      "learning_rate": 3.560481989584402e-05,
      "loss": 0.4112,
      "num_input_tokens_seen": 10589136,
      "step": 16200
    },
    {
      "epoch": 8.493186582809225,
      "grad_norm": 0.39279359579086304,
      "learning_rate": 3.55944642198333e-05,
      "loss": 0.4424,
      "num_input_tokens_seen": 10592784,
      "step": 16205
    },
    {
      "epoch": 8.4958071278826,
      "grad_norm": 0.3914693593978882,
      "learning_rate": 3.5584106327581536e-05,
      "loss": 0.4866,
      "num_input_tokens_seen": 10595888,
      "step": 16210
    },
    {
      "epoch": 8.498427672955975,
      "grad_norm": 0.314578652381897,
      "learning_rate": 3.557374622125547e-05,
      "loss": 0.4308,
      "num_input_tokens_seen": 10598960,
      "step": 16215
    },
    {
      "epoch": 8.50104821802935,
      "grad_norm": 0.3955861032009125,
      "learning_rate": 3.556338390302234e-05,
      "loss": 0.563,
      "num_input_tokens_seen": 10602736,
      "step": 16220
    },
    {
      "epoch": 8.503668763102725,
      "grad_norm": 0.4345249533653259,
      "learning_rate": 3.55530193750498e-05,
      "loss": 0.4572,
      "num_input_tokens_seen": 10605712,
      "step": 16225
    },
    {
      "epoch": 8.5062893081761,
      "grad_norm": 0.4399009048938751,
      "learning_rate": 3.5542652639505996e-05,
      "loss": 0.423,
      "num_input_tokens_seen": 10608848,
      "step": 16230
    },
    {
      "epoch": 8.508909853249476,
      "grad_norm": 0.36381933093070984,
      "learning_rate": 3.5532283698559556e-05,
      "loss": 0.4241,
      "num_input_tokens_seen": 10612944,
      "step": 16235
    },
    {
      "epoch": 8.51153039832285,
      "grad_norm": 0.7773443460464478,
      "learning_rate": 3.55219125543795e-05,
      "loss": 0.3975,
      "num_input_tokens_seen": 10616464,
      "step": 16240
    },
    {
      "epoch": 8.514150943396226,
      "grad_norm": 0.41175299882888794,
      "learning_rate": 3.5511539209135396e-05,
      "loss": 0.5363,
      "num_input_tokens_seen": 10619280,
      "step": 16245
    },
    {
      "epoch": 8.516771488469601,
      "grad_norm": 0.3493255376815796,
      "learning_rate": 3.550116366499721e-05,
      "loss": 0.4387,
      "num_input_tokens_seen": 10621872,
      "step": 16250
    },
    {
      "epoch": 8.519392033542976,
      "grad_norm": 0.39251306653022766,
      "learning_rate": 3.549078592413538e-05,
      "loss": 0.4469,
      "num_input_tokens_seen": 10625360,
      "step": 16255
    },
    {
      "epoch": 8.522012578616351,
      "grad_norm": 0.6385846138000488,
      "learning_rate": 3.548040598872083e-05,
      "loss": 0.3424,
      "num_input_tokens_seen": 10628752,
      "step": 16260
    },
    {
      "epoch": 8.524633123689728,
      "grad_norm": 0.3031903803348541,
      "learning_rate": 3.5470023860924905e-05,
      "loss": 0.399,
      "num_input_tokens_seen": 10632176,
      "step": 16265
    },
    {
      "epoch": 8.527253668763104,
      "grad_norm": 0.35896235704421997,
      "learning_rate": 3.5459639542919444e-05,
      "loss": 0.3846,
      "num_input_tokens_seen": 10636880,
      "step": 16270
    },
    {
      "epoch": 8.529874213836479,
      "grad_norm": 0.5432657599449158,
      "learning_rate": 3.5449253036876715e-05,
      "loss": 0.4619,
      "num_input_tokens_seen": 10640112,
      "step": 16275
    },
    {
      "epoch": 8.532494758909854,
      "grad_norm": 0.7093208432197571,
      "learning_rate": 3.543886434496947e-05,
      "loss": 0.5094,
      "num_input_tokens_seen": 10643856,
      "step": 16280
    },
    {
      "epoch": 8.535115303983229,
      "grad_norm": 0.7191697359085083,
      "learning_rate": 3.542847346937089e-05,
      "loss": 0.4667,
      "num_input_tokens_seen": 10646960,
      "step": 16285
    },
    {
      "epoch": 8.537735849056604,
      "grad_norm": 0.8213671445846558,
      "learning_rate": 3.541808041225464e-05,
      "loss": 0.43,
      "num_input_tokens_seen": 10653456,
      "step": 16290
    },
    {
      "epoch": 8.54035639412998,
      "grad_norm": 0.6397095322608948,
      "learning_rate": 3.540768517579481e-05,
      "loss": 0.4529,
      "num_input_tokens_seen": 10656656,
      "step": 16295
    },
    {
      "epoch": 8.542976939203355,
      "grad_norm": 0.8707200288772583,
      "learning_rate": 3.539728776216599e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 10660304,
      "step": 16300
    },
    {
      "epoch": 8.54559748427673,
      "grad_norm": 0.5923447012901306,
      "learning_rate": 3.5386888173543185e-05,
      "loss": 0.346,
      "num_input_tokens_seen": 10663696,
      "step": 16305
    },
    {
      "epoch": 8.548218029350105,
      "grad_norm": 0.38973110914230347,
      "learning_rate": 3.5376486412101875e-05,
      "loss": 0.4187,
      "num_input_tokens_seen": 10667344,
      "step": 16310
    },
    {
      "epoch": 8.55083857442348,
      "grad_norm": 0.374416321516037,
      "learning_rate": 3.5366082480017994e-05,
      "loss": 0.4791,
      "num_input_tokens_seen": 10670480,
      "step": 16315
    },
    {
      "epoch": 8.553459119496855,
      "grad_norm": 0.34461167454719543,
      "learning_rate": 3.535567637946792e-05,
      "loss": 0.5578,
      "num_input_tokens_seen": 10674064,
      "step": 16320
    },
    {
      "epoch": 8.55607966457023,
      "grad_norm": 0.35213926434516907,
      "learning_rate": 3.534526811262848e-05,
      "loss": 0.5354,
      "num_input_tokens_seen": 10676848,
      "step": 16325
    },
    {
      "epoch": 8.558700209643606,
      "grad_norm": 0.39964261651039124,
      "learning_rate": 3.533485768167698e-05,
      "loss": 0.5136,
      "num_input_tokens_seen": 10679440,
      "step": 16330
    },
    {
      "epoch": 8.56132075471698,
      "grad_norm": 0.4734300971031189,
      "learning_rate": 3.532444508879117e-05,
      "loss": 0.3967,
      "num_input_tokens_seen": 10682672,
      "step": 16335
    },
    {
      "epoch": 8.563941299790356,
      "grad_norm": 0.27724936604499817,
      "learning_rate": 3.531403033614922e-05,
      "loss": 0.3969,
      "num_input_tokens_seen": 10686224,
      "step": 16340
    },
    {
      "epoch": 8.566561844863731,
      "grad_norm": 0.2974637746810913,
      "learning_rate": 3.530361342592981e-05,
      "loss": 0.4365,
      "num_input_tokens_seen": 10690128,
      "step": 16345
    },
    {
      "epoch": 8.569182389937106,
      "grad_norm": 0.6294756531715393,
      "learning_rate": 3.5293194360312006e-05,
      "loss": 0.3921,
      "num_input_tokens_seen": 10693616,
      "step": 16350
    },
    {
      "epoch": 8.571802935010481,
      "grad_norm": 0.6925328969955444,
      "learning_rate": 3.528277314147538e-05,
      "loss": 0.5302,
      "num_input_tokens_seen": 10696720,
      "step": 16355
    },
    {
      "epoch": 8.574423480083858,
      "grad_norm": 0.6889406442642212,
      "learning_rate": 3.5272349771599925e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 10699920,
      "step": 16360
    },
    {
      "epoch": 8.577044025157234,
      "grad_norm": 0.5284407734870911,
      "learning_rate": 3.526192425286609e-05,
      "loss": 0.3568,
      "num_input_tokens_seen": 10702864,
      "step": 16365
    },
    {
      "epoch": 8.579664570230609,
      "grad_norm": 0.5512686967849731,
      "learning_rate": 3.525149658745476e-05,
      "loss": 0.5173,
      "num_input_tokens_seen": 10706416,
      "step": 16370
    },
    {
      "epoch": 8.582285115303984,
      "grad_norm": 0.4260249733924866,
      "learning_rate": 3.5241066777547315e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 10710128,
      "step": 16375
    },
    {
      "epoch": 8.584905660377359,
      "grad_norm": 0.6723129153251648,
      "learning_rate": 3.523063482532554e-05,
      "loss": 0.568,
      "num_input_tokens_seen": 10712432,
      "step": 16380
    },
    {
      "epoch": 8.587526205450734,
      "grad_norm": 0.7620450854301453,
      "learning_rate": 3.5220200732971665e-05,
      "loss": 0.3987,
      "num_input_tokens_seen": 10716720,
      "step": 16385
    },
    {
      "epoch": 8.59014675052411,
      "grad_norm": 0.40059027075767517,
      "learning_rate": 3.52097645026684e-05,
      "loss": 0.5396,
      "num_input_tokens_seen": 10720656,
      "step": 16390
    },
    {
      "epoch": 8.592767295597485,
      "grad_norm": 0.4251652657985687,
      "learning_rate": 3.519932613659889e-05,
      "loss": 0.4797,
      "num_input_tokens_seen": 10723984,
      "step": 16395
    },
    {
      "epoch": 8.59538784067086,
      "grad_norm": 0.47944149374961853,
      "learning_rate": 3.5188885636946696e-05,
      "loss": 0.4256,
      "num_input_tokens_seen": 10727024,
      "step": 16400
    },
    {
      "epoch": 8.598008385744235,
      "grad_norm": 0.43120190501213074,
      "learning_rate": 3.5178443005895884e-05,
      "loss": 0.5452,
      "num_input_tokens_seen": 10730160,
      "step": 16405
    },
    {
      "epoch": 8.60062893081761,
      "grad_norm": 0.8975232243537903,
      "learning_rate": 3.51679982456309e-05,
      "loss": 0.4337,
      "num_input_tokens_seen": 10732560,
      "step": 16410
    },
    {
      "epoch": 8.603249475890985,
      "grad_norm": 0.6073652505874634,
      "learning_rate": 3.51575513583367e-05,
      "loss": 0.5861,
      "num_input_tokens_seen": 10735568,
      "step": 16415
    },
    {
      "epoch": 8.60587002096436,
      "grad_norm": 0.28808343410491943,
      "learning_rate": 3.5147102346198635e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 10739344,
      "step": 16420
    },
    {
      "epoch": 8.608490566037736,
      "grad_norm": 0.5181177258491516,
      "learning_rate": 3.5136651211402515e-05,
      "loss": 0.5261,
      "num_input_tokens_seen": 10742064,
      "step": 16425
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 0.508968710899353,
      "learning_rate": 3.5126197956134614e-05,
      "loss": 0.4487,
      "num_input_tokens_seen": 10745680,
      "step": 16430
    },
    {
      "epoch": 8.613731656184486,
      "grad_norm": 0.636643648147583,
      "learning_rate": 3.5115742582581626e-05,
      "loss": 0.4271,
      "num_input_tokens_seen": 10748784,
      "step": 16435
    },
    {
      "epoch": 8.616352201257861,
      "grad_norm": 0.4559679627418518,
      "learning_rate": 3.51052850929307e-05,
      "loss": 0.4224,
      "num_input_tokens_seen": 10752144,
      "step": 16440
    },
    {
      "epoch": 8.618972746331236,
      "grad_norm": 0.41355791687965393,
      "learning_rate": 3.509482548936941e-05,
      "loss": 0.4693,
      "num_input_tokens_seen": 10755664,
      "step": 16445
    },
    {
      "epoch": 8.621593291404611,
      "grad_norm": 0.44261130690574646,
      "learning_rate": 3.5084363774085804e-05,
      "loss": 0.3782,
      "num_input_tokens_seen": 10759728,
      "step": 16450
    },
    {
      "epoch": 8.624213836477988,
      "grad_norm": 0.7566593289375305,
      "learning_rate": 3.507389994926834e-05,
      "loss": 0.6083,
      "num_input_tokens_seen": 10763184,
      "step": 16455
    },
    {
      "epoch": 8.626834381551364,
      "grad_norm": 0.5042046308517456,
      "learning_rate": 3.5063434017105946e-05,
      "loss": 0.5444,
      "num_input_tokens_seen": 10766320,
      "step": 16460
    },
    {
      "epoch": 8.629454926624739,
      "grad_norm": 0.7034949064254761,
      "learning_rate": 3.505296597978795e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 10769104,
      "step": 16465
    },
    {
      "epoch": 8.632075471698114,
      "grad_norm": 0.4239383041858673,
      "learning_rate": 3.504249583950417e-05,
      "loss": 0.4004,
      "num_input_tokens_seen": 10772016,
      "step": 16470
    },
    {
      "epoch": 8.634696016771489,
      "grad_norm": 0.6676696538925171,
      "learning_rate": 3.503202359844483e-05,
      "loss": 0.3296,
      "num_input_tokens_seen": 10774960,
      "step": 16475
    },
    {
      "epoch": 8.637316561844864,
      "grad_norm": 0.4076266884803772,
      "learning_rate": 3.50215492588006e-05,
      "loss": 0.3724,
      "num_input_tokens_seen": 10778288,
      "step": 16480
    },
    {
      "epoch": 8.63993710691824,
      "grad_norm": 0.44230717420578003,
      "learning_rate": 3.5011072822762605e-05,
      "loss": 0.4627,
      "num_input_tokens_seen": 10781744,
      "step": 16485
    },
    {
      "epoch": 8.642557651991615,
      "grad_norm": 0.33477863669395447,
      "learning_rate": 3.500059429252238e-05,
      "loss": 0.4867,
      "num_input_tokens_seen": 10784624,
      "step": 16490
    },
    {
      "epoch": 8.64517819706499,
      "grad_norm": 0.331482470035553,
      "learning_rate": 3.499011367027193e-05,
      "loss": 0.4,
      "num_input_tokens_seen": 10787536,
      "step": 16495
    },
    {
      "epoch": 8.647798742138365,
      "grad_norm": 0.3931543231010437,
      "learning_rate": 3.4979630958203665e-05,
      "loss": 0.4202,
      "num_input_tokens_seen": 10790000,
      "step": 16500
    },
    {
      "epoch": 8.65041928721174,
      "grad_norm": 0.4295133650302887,
      "learning_rate": 3.496914615851046e-05,
      "loss": 0.4094,
      "num_input_tokens_seen": 10793872,
      "step": 16505
    },
    {
      "epoch": 8.653039832285115,
      "grad_norm": 0.7153326869010925,
      "learning_rate": 3.495865927338562e-05,
      "loss": 0.5082,
      "num_input_tokens_seen": 10796720,
      "step": 16510
    },
    {
      "epoch": 8.65566037735849,
      "grad_norm": 0.42006921768188477,
      "learning_rate": 3.494817030502286e-05,
      "loss": 0.4455,
      "num_input_tokens_seen": 10799856,
      "step": 16515
    },
    {
      "epoch": 8.658280922431866,
      "grad_norm": 0.5322287678718567,
      "learning_rate": 3.493767925561638e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 10804272,
      "step": 16520
    },
    {
      "epoch": 8.66090146750524,
      "grad_norm": 0.47317636013031006,
      "learning_rate": 3.492718612736077e-05,
      "loss": 0.5134,
      "num_input_tokens_seen": 10806896,
      "step": 16525
    },
    {
      "epoch": 8.663522012578616,
      "grad_norm": 0.4229969382286072,
      "learning_rate": 3.491669092245108e-05,
      "loss": 0.4334,
      "num_input_tokens_seen": 10810128,
      "step": 16530
    },
    {
      "epoch": 8.666142557651991,
      "grad_norm": 0.894757091999054,
      "learning_rate": 3.4906193643082784e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 10812816,
      "step": 16535
    },
    {
      "epoch": 8.668763102725366,
      "grad_norm": 0.5662396550178528,
      "learning_rate": 3.4895694291451795e-05,
      "loss": 0.6276,
      "num_input_tokens_seen": 10815408,
      "step": 16540
    },
    {
      "epoch": 8.671383647798741,
      "grad_norm": 0.543458878993988,
      "learning_rate": 3.4885192869754444e-05,
      "loss": 0.4944,
      "num_input_tokens_seen": 10818928,
      "step": 16545
    },
    {
      "epoch": 8.674004192872118,
      "grad_norm": 0.36065003275871277,
      "learning_rate": 3.4874689380187537e-05,
      "loss": 0.4053,
      "num_input_tokens_seen": 10822288,
      "step": 16550
    },
    {
      "epoch": 8.676624737945493,
      "grad_norm": 0.4140338599681854,
      "learning_rate": 3.486418382494826e-05,
      "loss": 0.3514,
      "num_input_tokens_seen": 10826512,
      "step": 16555
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 0.4679838716983795,
      "learning_rate": 3.485367620623426e-05,
      "loss": 0.4012,
      "num_input_tokens_seen": 10829072,
      "step": 16560
    },
    {
      "epoch": 8.681865828092244,
      "grad_norm": 0.5246437788009644,
      "learning_rate": 3.484316652624362e-05,
      "loss": 0.5305,
      "num_input_tokens_seen": 10832112,
      "step": 16565
    },
    {
      "epoch": 8.684486373165619,
      "grad_norm": 0.7312932014465332,
      "learning_rate": 3.483265478717484e-05,
      "loss": 0.589,
      "num_input_tokens_seen": 10835152,
      "step": 16570
    },
    {
      "epoch": 8.687106918238994,
      "grad_norm": 0.5533487796783447,
      "learning_rate": 3.482214099122686e-05,
      "loss": 0.457,
      "num_input_tokens_seen": 10838416,
      "step": 16575
    },
    {
      "epoch": 8.68972746331237,
      "grad_norm": 0.45252886414527893,
      "learning_rate": 3.481162514059903e-05,
      "loss": 0.3799,
      "num_input_tokens_seen": 10841136,
      "step": 16580
    },
    {
      "epoch": 8.692348008385745,
      "grad_norm": 0.7865087985992432,
      "learning_rate": 3.480110723749116e-05,
      "loss": 0.5255,
      "num_input_tokens_seen": 10844560,
      "step": 16585
    },
    {
      "epoch": 8.69496855345912,
      "grad_norm": 0.25241202116012573,
      "learning_rate": 3.479058728410348e-05,
      "loss": 0.4404,
      "num_input_tokens_seen": 10849040,
      "step": 16590
    },
    {
      "epoch": 8.697589098532495,
      "grad_norm": 0.2305998057126999,
      "learning_rate": 3.478006528263663e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 10852656,
      "step": 16595
    },
    {
      "epoch": 8.70020964360587,
      "grad_norm": 0.6979044675827026,
      "learning_rate": 3.4769541235291704e-05,
      "loss": 0.3967,
      "num_input_tokens_seen": 10855248,
      "step": 16600
    },
    {
      "epoch": 8.702830188679245,
      "grad_norm": 0.3921349048614502,
      "learning_rate": 3.475901514427021e-05,
      "loss": 0.4116,
      "num_input_tokens_seen": 10858352,
      "step": 16605
    },
    {
      "epoch": 8.70545073375262,
      "grad_norm": 0.7844857573509216,
      "learning_rate": 3.4748487011774075e-05,
      "loss": 0.4821,
      "num_input_tokens_seen": 10860944,
      "step": 16610
    },
    {
      "epoch": 8.708071278825996,
      "grad_norm": 0.5811801552772522,
      "learning_rate": 3.473795684000569e-05,
      "loss": 0.477,
      "num_input_tokens_seen": 10864560,
      "step": 16615
    },
    {
      "epoch": 8.71069182389937,
      "grad_norm": 0.26976466178894043,
      "learning_rate": 3.4727424631167815e-05,
      "loss": 0.4207,
      "num_input_tokens_seen": 10868144,
      "step": 16620
    },
    {
      "epoch": 8.713312368972746,
      "grad_norm": 0.4344055950641632,
      "learning_rate": 3.471689038746369e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 10871792,
      "step": 16625
    },
    {
      "epoch": 8.715932914046121,
      "grad_norm": 0.28787553310394287,
      "learning_rate": 3.470635411109695e-05,
      "loss": 0.4311,
      "num_input_tokens_seen": 10874800,
      "step": 16630
    },
    {
      "epoch": 8.718553459119496,
      "grad_norm": 0.31410229206085205,
      "learning_rate": 3.469581580427167e-05,
      "loss": 0.7102,
      "num_input_tokens_seen": 10877680,
      "step": 16635
    },
    {
      "epoch": 8.721174004192871,
      "grad_norm": 0.45030587911605835,
      "learning_rate": 3.4685275469192335e-05,
      "loss": 0.5264,
      "num_input_tokens_seen": 10880464,
      "step": 16640
    },
    {
      "epoch": 8.723794549266248,
      "grad_norm": 0.5720878839492798,
      "learning_rate": 3.467473310806387e-05,
      "loss": 0.5505,
      "num_input_tokens_seen": 10883856,
      "step": 16645
    },
    {
      "epoch": 8.726415094339622,
      "grad_norm": 0.5888451933860779,
      "learning_rate": 3.466418872309161e-05,
      "loss": 0.6173,
      "num_input_tokens_seen": 10888016,
      "step": 16650
    },
    {
      "epoch": 8.729035639412999,
      "grad_norm": 0.4202443063259125,
      "learning_rate": 3.465364231648133e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 10891088,
      "step": 16655
    },
    {
      "epoch": 8.731656184486374,
      "grad_norm": 0.580311119556427,
      "learning_rate": 3.464309389043921e-05,
      "loss": 0.5506,
      "num_input_tokens_seen": 10893808,
      "step": 16660
    },
    {
      "epoch": 8.734276729559749,
      "grad_norm": 0.6326735615730286,
      "learning_rate": 3.4632543447171856e-05,
      "loss": 0.5055,
      "num_input_tokens_seen": 10896176,
      "step": 16665
    },
    {
      "epoch": 8.736897274633124,
      "grad_norm": 0.33111661672592163,
      "learning_rate": 3.4621990988886315e-05,
      "loss": 0.3391,
      "num_input_tokens_seen": 10899376,
      "step": 16670
    },
    {
      "epoch": 8.7395178197065,
      "grad_norm": 0.46831798553466797,
      "learning_rate": 3.461143651779003e-05,
      "loss": 0.5084,
      "num_input_tokens_seen": 10902800,
      "step": 16675
    },
    {
      "epoch": 8.742138364779874,
      "grad_norm": 0.6868222951889038,
      "learning_rate": 3.4600880036090876e-05,
      "loss": 0.5351,
      "num_input_tokens_seen": 10906352,
      "step": 16680
    },
    {
      "epoch": 8.74475890985325,
      "grad_norm": 0.2756568193435669,
      "learning_rate": 3.459032154599715e-05,
      "loss": 0.5019,
      "num_input_tokens_seen": 10910256,
      "step": 16685
    },
    {
      "epoch": 8.747379454926625,
      "grad_norm": 0.432502806186676,
      "learning_rate": 3.457976104971757e-05,
      "loss": 0.6858,
      "num_input_tokens_seen": 10913168,
      "step": 16690
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.6034643054008484,
      "learning_rate": 3.4569198549461276e-05,
      "loss": 0.507,
      "num_input_tokens_seen": 10915984,
      "step": 16695
    },
    {
      "epoch": 8.752620545073375,
      "grad_norm": 0.6865081191062927,
      "learning_rate": 3.455863404743781e-05,
      "loss": 0.417,
      "num_input_tokens_seen": 10919024,
      "step": 16700
    },
    {
      "epoch": 8.75524109014675,
      "grad_norm": 0.29861724376678467,
      "learning_rate": 3.454806754585716e-05,
      "loss": 0.4358,
      "num_input_tokens_seen": 10922960,
      "step": 16705
    },
    {
      "epoch": 8.757861635220126,
      "grad_norm": 0.538636326789856,
      "learning_rate": 3.453749904692971e-05,
      "loss": 0.4633,
      "num_input_tokens_seen": 10926064,
      "step": 16710
    },
    {
      "epoch": 8.7604821802935,
      "grad_norm": 0.49364396929740906,
      "learning_rate": 3.452692855286626e-05,
      "loss": 0.3945,
      "num_input_tokens_seen": 10929168,
      "step": 16715
    },
    {
      "epoch": 8.763102725366876,
      "grad_norm": 0.3148796856403351,
      "learning_rate": 3.4516356065878055e-05,
      "loss": 0.3773,
      "num_input_tokens_seen": 10932176,
      "step": 16720
    },
    {
      "epoch": 8.765723270440251,
      "grad_norm": 0.37201234698295593,
      "learning_rate": 3.450578158817671e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 10935312,
      "step": 16725
    },
    {
      "epoch": 8.768343815513626,
      "grad_norm": 0.6082959771156311,
      "learning_rate": 3.4495205121974325e-05,
      "loss": 0.5293,
      "num_input_tokens_seen": 10937904,
      "step": 16730
    },
    {
      "epoch": 8.770964360587001,
      "grad_norm": 0.6135914325714111,
      "learning_rate": 3.448462666948335e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 10942160,
      "step": 16735
    },
    {
      "epoch": 8.773584905660378,
      "grad_norm": 0.47575461864471436,
      "learning_rate": 3.4474046232916674e-05,
      "loss": 0.3842,
      "num_input_tokens_seen": 10946256,
      "step": 16740
    },
    {
      "epoch": 8.776205450733752,
      "grad_norm": 0.3985670208930969,
      "learning_rate": 3.4463463814487616e-05,
      "loss": 0.6315,
      "num_input_tokens_seen": 10949776,
      "step": 16745
    },
    {
      "epoch": 8.778825995807129,
      "grad_norm": 0.4113439619541168,
      "learning_rate": 3.44528794164099e-05,
      "loss": 0.3662,
      "num_input_tokens_seen": 10952912,
      "step": 16750
    },
    {
      "epoch": 8.781446540880504,
      "grad_norm": 0.6017640233039856,
      "learning_rate": 3.4442293040897646e-05,
      "loss": 0.5129,
      "num_input_tokens_seen": 10956656,
      "step": 16755
    },
    {
      "epoch": 8.784067085953879,
      "grad_norm": 0.47635844349861145,
      "learning_rate": 3.443170469016542e-05,
      "loss": 0.4523,
      "num_input_tokens_seen": 10960240,
      "step": 16760
    },
    {
      "epoch": 8.786687631027254,
      "grad_norm": 0.47231051325798035,
      "learning_rate": 3.442111436642817e-05,
      "loss": 0.3998,
      "num_input_tokens_seen": 10963504,
      "step": 16765
    },
    {
      "epoch": 8.78930817610063,
      "grad_norm": 0.255913645029068,
      "learning_rate": 3.441052207190127e-05,
      "loss": 0.5031,
      "num_input_tokens_seen": 10968432,
      "step": 16770
    },
    {
      "epoch": 8.791928721174004,
      "grad_norm": 0.45000579953193665,
      "learning_rate": 3.4399927808800526e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 10971472,
      "step": 16775
    },
    {
      "epoch": 8.79454926624738,
      "grad_norm": 0.3898962736129761,
      "learning_rate": 3.438933157934211e-05,
      "loss": 0.5628,
      "num_input_tokens_seen": 10974384,
      "step": 16780
    },
    {
      "epoch": 8.797169811320755,
      "grad_norm": 0.6876570582389832,
      "learning_rate": 3.4378733385742653e-05,
      "loss": 0.6315,
      "num_input_tokens_seen": 10977296,
      "step": 16785
    },
    {
      "epoch": 8.79979035639413,
      "grad_norm": 0.5371579527854919,
      "learning_rate": 3.436813323021918e-05,
      "loss": 0.5209,
      "num_input_tokens_seen": 10980528,
      "step": 16790
    },
    {
      "epoch": 8.802410901467505,
      "grad_norm": 0.6379220485687256,
      "learning_rate": 3.43575311149891e-05,
      "loss": 0.3855,
      "num_input_tokens_seen": 10983408,
      "step": 16795
    },
    {
      "epoch": 8.80503144654088,
      "grad_norm": 0.6121867895126343,
      "learning_rate": 3.434692704227028e-05,
      "loss": 0.4036,
      "num_input_tokens_seen": 10986000,
      "step": 16800
    },
    {
      "epoch": 8.807651991614255,
      "grad_norm": 0.4931182563304901,
      "learning_rate": 3.4336321014280944e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 10989168,
      "step": 16805
    },
    {
      "epoch": 8.81027253668763,
      "grad_norm": 0.2740795612335205,
      "learning_rate": 3.432571303323977e-05,
      "loss": 0.4974,
      "num_input_tokens_seen": 10992528,
      "step": 16810
    },
    {
      "epoch": 8.812893081761006,
      "grad_norm": 0.5211848020553589,
      "learning_rate": 3.431510310136583e-05,
      "loss": 0.4586,
      "num_input_tokens_seen": 10995120,
      "step": 16815
    },
    {
      "epoch": 8.815513626834381,
      "grad_norm": 0.242223858833313,
      "learning_rate": 3.4304491220878585e-05,
      "loss": 0.5023,
      "num_input_tokens_seen": 10998416,
      "step": 16820
    },
    {
      "epoch": 8.818134171907756,
      "grad_norm": 0.47104185819625854,
      "learning_rate": 3.429387739399793e-05,
      "loss": 0.4478,
      "num_input_tokens_seen": 11001712,
      "step": 16825
    },
    {
      "epoch": 8.820754716981131,
      "grad_norm": 0.2755703926086426,
      "learning_rate": 3.428326162294416e-05,
      "loss": 0.5325,
      "num_input_tokens_seen": 11005200,
      "step": 16830
    },
    {
      "epoch": 8.823375262054507,
      "grad_norm": 0.5113876461982727,
      "learning_rate": 3.427264390993795e-05,
      "loss": 0.3657,
      "num_input_tokens_seen": 11008496,
      "step": 16835
    },
    {
      "epoch": 8.825995807127882,
      "grad_norm": 0.31519970297813416,
      "learning_rate": 3.426202425720044e-05,
      "loss": 0.4636,
      "num_input_tokens_seen": 11011792,
      "step": 16840
    },
    {
      "epoch": 8.828616352201259,
      "grad_norm": 0.2354181557893753,
      "learning_rate": 3.425140266695312e-05,
      "loss": 0.375,
      "num_input_tokens_seen": 11015888,
      "step": 16845
    },
    {
      "epoch": 8.831236897274634,
      "grad_norm": 0.34157487750053406,
      "learning_rate": 3.4240779141417905e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 11019184,
      "step": 16850
    },
    {
      "epoch": 8.833857442348009,
      "grad_norm": 0.3457704782485962,
      "learning_rate": 3.423015368281711e-05,
      "loss": 0.4195,
      "num_input_tokens_seen": 11022320,
      "step": 16855
    },
    {
      "epoch": 8.836477987421384,
      "grad_norm": 0.610571026802063,
      "learning_rate": 3.421952629337348e-05,
      "loss": 0.4169,
      "num_input_tokens_seen": 11024784,
      "step": 16860
    },
    {
      "epoch": 8.83909853249476,
      "grad_norm": 0.7166480422019958,
      "learning_rate": 3.420889697531013e-05,
      "loss": 0.4122,
      "num_input_tokens_seen": 11027728,
      "step": 16865
    },
    {
      "epoch": 8.841719077568134,
      "grad_norm": 0.9420936107635498,
      "learning_rate": 3.419826573085058e-05,
      "loss": 0.4485,
      "num_input_tokens_seen": 11031888,
      "step": 16870
    },
    {
      "epoch": 8.84433962264151,
      "grad_norm": 0.4513722062110901,
      "learning_rate": 3.418763256221879e-05,
      "loss": 0.4104,
      "num_input_tokens_seen": 11034256,
      "step": 16875
    },
    {
      "epoch": 8.846960167714885,
      "grad_norm": 0.9277600646018982,
      "learning_rate": 3.4176997471639085e-05,
      "loss": 0.4364,
      "num_input_tokens_seen": 11036976,
      "step": 16880
    },
    {
      "epoch": 8.84958071278826,
      "grad_norm": 0.495679646730423,
      "learning_rate": 3.4166360461336206e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 11039856,
      "step": 16885
    },
    {
      "epoch": 8.852201257861635,
      "grad_norm": 0.6205777525901794,
      "learning_rate": 3.41557215335353e-05,
      "loss": 0.4307,
      "num_input_tokens_seen": 11042480,
      "step": 16890
    },
    {
      "epoch": 8.85482180293501,
      "grad_norm": 0.3909720182418823,
      "learning_rate": 3.4145080690461887e-05,
      "loss": 0.5194,
      "num_input_tokens_seen": 11045360,
      "step": 16895
    },
    {
      "epoch": 8.857442348008385,
      "grad_norm": 0.3322901427745819,
      "learning_rate": 3.4134437934341936e-05,
      "loss": 0.432,
      "num_input_tokens_seen": 11049680,
      "step": 16900
    },
    {
      "epoch": 8.86006289308176,
      "grad_norm": 0.21885855495929718,
      "learning_rate": 3.4123793267401784e-05,
      "loss": 0.3732,
      "num_input_tokens_seen": 11052816,
      "step": 16905
    },
    {
      "epoch": 8.862683438155136,
      "grad_norm": 0.5615055561065674,
      "learning_rate": 3.411314669186816e-05,
      "loss": 0.5452,
      "num_input_tokens_seen": 11055408,
      "step": 16910
    },
    {
      "epoch": 8.865303983228511,
      "grad_norm": 0.4121466875076294,
      "learning_rate": 3.410249820996823e-05,
      "loss": 0.4406,
      "num_input_tokens_seen": 11058704,
      "step": 16915
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 0.6076706647872925,
      "learning_rate": 3.4091847823929513e-05,
      "loss": 0.4134,
      "num_input_tokens_seen": 11061616,
      "step": 16920
    },
    {
      "epoch": 8.870545073375261,
      "grad_norm": 0.49081841111183167,
      "learning_rate": 3.4081195535979946e-05,
      "loss": 0.3847,
      "num_input_tokens_seen": 11064304,
      "step": 16925
    },
    {
      "epoch": 8.873165618448636,
      "grad_norm": 0.535990834236145,
      "learning_rate": 3.40705413483479e-05,
      "loss": 0.3908,
      "num_input_tokens_seen": 11067120,
      "step": 16930
    },
    {
      "epoch": 8.875786163522012,
      "grad_norm": 0.3294816017150879,
      "learning_rate": 3.405988526326207e-05,
      "loss": 0.3721,
      "num_input_tokens_seen": 11071088,
      "step": 16935
    },
    {
      "epoch": 8.878406708595389,
      "grad_norm": 0.398980051279068,
      "learning_rate": 3.40492272829516e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 11073808,
      "step": 16940
    },
    {
      "epoch": 8.881027253668764,
      "grad_norm": 0.3904721438884735,
      "learning_rate": 3.403856740964604e-05,
      "loss": 0.4297,
      "num_input_tokens_seen": 11077744,
      "step": 16945
    },
    {
      "epoch": 8.883647798742139,
      "grad_norm": 0.38300153613090515,
      "learning_rate": 3.4027905645575275e-05,
      "loss": 0.4278,
      "num_input_tokens_seen": 11080688,
      "step": 16950
    },
    {
      "epoch": 8.886268343815514,
      "grad_norm": 0.3719659745693207,
      "learning_rate": 3.401724199296966e-05,
      "loss": 0.456,
      "num_input_tokens_seen": 11083792,
      "step": 16955
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 0.30608657002449036,
      "learning_rate": 3.400657645405989e-05,
      "loss": 0.4209,
      "num_input_tokens_seen": 11086640,
      "step": 16960
    },
    {
      "epoch": 8.891509433962264,
      "grad_norm": 0.8583053946495056,
      "learning_rate": 3.3995909031077065e-05,
      "loss": 0.4759,
      "num_input_tokens_seen": 11089936,
      "step": 16965
    },
    {
      "epoch": 8.89412997903564,
      "grad_norm": 0.4632149338722229,
      "learning_rate": 3.398523972625272e-05,
      "loss": 0.4835,
      "num_input_tokens_seen": 11093520,
      "step": 16970
    },
    {
      "epoch": 8.896750524109015,
      "grad_norm": 0.49223896861076355,
      "learning_rate": 3.3974568541818723e-05,
      "loss": 0.4509,
      "num_input_tokens_seen": 11096016,
      "step": 16975
    },
    {
      "epoch": 8.89937106918239,
      "grad_norm": 0.7444644570350647,
      "learning_rate": 3.396389548000738e-05,
      "loss": 0.4706,
      "num_input_tokens_seen": 11099280,
      "step": 16980
    },
    {
      "epoch": 8.901991614255765,
      "grad_norm": 0.6178092360496521,
      "learning_rate": 3.395322054305137e-05,
      "loss": 0.4615,
      "num_input_tokens_seen": 11101808,
      "step": 16985
    },
    {
      "epoch": 8.90461215932914,
      "grad_norm": 0.5307134985923767,
      "learning_rate": 3.394254373318376e-05,
      "loss": 0.4916,
      "num_input_tokens_seen": 11104624,
      "step": 16990
    },
    {
      "epoch": 8.907232704402515,
      "grad_norm": 0.4169444739818573,
      "learning_rate": 3.3931865052638026e-05,
      "loss": 0.4652,
      "num_input_tokens_seen": 11108592,
      "step": 16995
    },
    {
      "epoch": 8.90985324947589,
      "grad_norm": 2.2852084636688232,
      "learning_rate": 3.392118450364802e-05,
      "loss": 0.5003,
      "num_input_tokens_seen": 11111536,
      "step": 17000
    },
    {
      "epoch": 8.912473794549266,
      "grad_norm": 0.45041126012802124,
      "learning_rate": 3.391050208844799e-05,
      "loss": 0.332,
      "num_input_tokens_seen": 11114832,
      "step": 17005
    },
    {
      "epoch": 8.915094339622641,
      "grad_norm": 0.35194891691207886,
      "learning_rate": 3.389981780927259e-05,
      "loss": 0.472,
      "num_input_tokens_seen": 11118128,
      "step": 17010
    },
    {
      "epoch": 8.917714884696016,
      "grad_norm": 0.7820870280265808,
      "learning_rate": 3.3889131668356834e-05,
      "loss": 0.3181,
      "num_input_tokens_seen": 11123792,
      "step": 17015
    },
    {
      "epoch": 8.920335429769391,
      "grad_norm": 0.3912254571914673,
      "learning_rate": 3.387844366793614e-05,
      "loss": 0.4737,
      "num_input_tokens_seen": 11127280,
      "step": 17020
    },
    {
      "epoch": 8.922955974842766,
      "grad_norm": 0.36136436462402344,
      "learning_rate": 3.3867753810246326e-05,
      "loss": 0.4763,
      "num_input_tokens_seen": 11130896,
      "step": 17025
    },
    {
      "epoch": 8.925576519916142,
      "grad_norm": 0.5116523504257202,
      "learning_rate": 3.3857062097523586e-05,
      "loss": 0.4065,
      "num_input_tokens_seen": 11133936,
      "step": 17030
    },
    {
      "epoch": 8.928197064989519,
      "grad_norm": 0.6422320604324341,
      "learning_rate": 3.38463685320045e-05,
      "loss": 0.4668,
      "num_input_tokens_seen": 11137104,
      "step": 17035
    },
    {
      "epoch": 8.930817610062894,
      "grad_norm": 0.4703216552734375,
      "learning_rate": 3.3835673115926045e-05,
      "loss": 0.4422,
      "num_input_tokens_seen": 11139984,
      "step": 17040
    },
    {
      "epoch": 8.933438155136269,
      "grad_norm": 0.595187246799469,
      "learning_rate": 3.382497585152557e-05,
      "loss": 0.4264,
      "num_input_tokens_seen": 11142672,
      "step": 17045
    },
    {
      "epoch": 8.936058700209644,
      "grad_norm": 0.44621777534484863,
      "learning_rate": 3.3814276741040835e-05,
      "loss": 0.5595,
      "num_input_tokens_seen": 11146096,
      "step": 17050
    },
    {
      "epoch": 8.93867924528302,
      "grad_norm": 0.5544081330299377,
      "learning_rate": 3.380357578670996e-05,
      "loss": 0.3428,
      "num_input_tokens_seen": 11148656,
      "step": 17055
    },
    {
      "epoch": 8.941299790356394,
      "grad_norm": 0.35608193278312683,
      "learning_rate": 3.379287299077146e-05,
      "loss": 0.3383,
      "num_input_tokens_seen": 11153008,
      "step": 17060
    },
    {
      "epoch": 8.94392033542977,
      "grad_norm": 0.37101641297340393,
      "learning_rate": 3.3782168355464263e-05,
      "loss": 0.4334,
      "num_input_tokens_seen": 11155984,
      "step": 17065
    },
    {
      "epoch": 8.946540880503145,
      "grad_norm": 0.29827946424484253,
      "learning_rate": 3.377146188302762e-05,
      "loss": 0.3118,
      "num_input_tokens_seen": 11158960,
      "step": 17070
    },
    {
      "epoch": 8.94916142557652,
      "grad_norm": 0.3520875573158264,
      "learning_rate": 3.376075357570124e-05,
      "loss": 0.5392,
      "num_input_tokens_seen": 11162576,
      "step": 17075
    },
    {
      "epoch": 8.951781970649895,
      "grad_norm": 0.7519758939743042,
      "learning_rate": 3.3750043435725144e-05,
      "loss": 0.4198,
      "num_input_tokens_seen": 11166160,
      "step": 17080
    },
    {
      "epoch": 8.95440251572327,
      "grad_norm": 0.3063581883907318,
      "learning_rate": 3.373933146533979e-05,
      "loss": 0.4001,
      "num_input_tokens_seen": 11173296,
      "step": 17085
    },
    {
      "epoch": 8.957023060796645,
      "grad_norm": 0.43224307894706726,
      "learning_rate": 3.3728617666786e-05,
      "loss": 0.4266,
      "num_input_tokens_seen": 11176848,
      "step": 17090
    },
    {
      "epoch": 8.95964360587002,
      "grad_norm": 0.4129657745361328,
      "learning_rate": 3.371790204230497e-05,
      "loss": 0.4719,
      "num_input_tokens_seen": 11179312,
      "step": 17095
    },
    {
      "epoch": 8.962264150943396,
      "grad_norm": 0.4704054296016693,
      "learning_rate": 3.3707184594138295e-05,
      "loss": 0.5079,
      "num_input_tokens_seen": 11182384,
      "step": 17100
    },
    {
      "epoch": 8.964884696016771,
      "grad_norm": 0.639825165271759,
      "learning_rate": 3.369646532452795e-05,
      "loss": 0.5995,
      "num_input_tokens_seen": 11185744,
      "step": 17105
    },
    {
      "epoch": 8.967505241090146,
      "grad_norm": 0.49571603536605835,
      "learning_rate": 3.3685744235716254e-05,
      "loss": 0.5487,
      "num_input_tokens_seen": 11188880,
      "step": 17110
    },
    {
      "epoch": 8.970125786163521,
      "grad_norm": 0.36885878443717957,
      "learning_rate": 3.367502132994598e-05,
      "loss": 0.5361,
      "num_input_tokens_seen": 11192464,
      "step": 17115
    },
    {
      "epoch": 8.972746331236896,
      "grad_norm": 0.4717079699039459,
      "learning_rate": 3.366429660946019e-05,
      "loss": 0.4335,
      "num_input_tokens_seen": 11195888,
      "step": 17120
    },
    {
      "epoch": 8.975366876310272,
      "grad_norm": 0.2860824763774872,
      "learning_rate": 3.365357007650241e-05,
      "loss": 0.4982,
      "num_input_tokens_seen": 11199056,
      "step": 17125
    },
    {
      "epoch": 8.977987421383649,
      "grad_norm": 0.4378119111061096,
      "learning_rate": 3.364284173331649e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 11201552,
      "step": 17130
    },
    {
      "epoch": 8.980607966457024,
      "grad_norm": 0.900518000125885,
      "learning_rate": 3.3632111582146686e-05,
      "loss": 0.5034,
      "num_input_tokens_seen": 11204432,
      "step": 17135
    },
    {
      "epoch": 8.983228511530399,
      "grad_norm": 0.4357946515083313,
      "learning_rate": 3.362137962523761e-05,
      "loss": 0.3394,
      "num_input_tokens_seen": 11207504,
      "step": 17140
    },
    {
      "epoch": 8.985849056603774,
      "grad_norm": 0.9189314842224121,
      "learning_rate": 3.361064586483429e-05,
      "loss": 0.4567,
      "num_input_tokens_seen": 11210416,
      "step": 17145
    },
    {
      "epoch": 8.98846960167715,
      "grad_norm": 0.3841745853424072,
      "learning_rate": 3.359991030318208e-05,
      "loss": 0.4023,
      "num_input_tokens_seen": 11212688,
      "step": 17150
    },
    {
      "epoch": 8.991090146750524,
      "grad_norm": 0.6016098260879517,
      "learning_rate": 3.358917294252675e-05,
      "loss": 0.4541,
      "num_input_tokens_seen": 11215600,
      "step": 17155
    },
    {
      "epoch": 8.9937106918239,
      "grad_norm": 1.0663260221481323,
      "learning_rate": 3.357843378511443e-05,
      "loss": 0.5016,
      "num_input_tokens_seen": 11219088,
      "step": 17160
    },
    {
      "epoch": 8.996331236897275,
      "grad_norm": 0.37183433771133423,
      "learning_rate": 3.3567692833191635e-05,
      "loss": 0.4975,
      "num_input_tokens_seen": 11222800,
      "step": 17165
    },
    {
      "epoch": 8.99895178197065,
      "grad_norm": 0.6985437273979187,
      "learning_rate": 3.355695008900525e-05,
      "loss": 0.5734,
      "num_input_tokens_seen": 11226640,
      "step": 17170
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.4789770841598511,
      "eval_runtime": 9.6454,
      "eval_samples_per_second": 87.917,
      "eval_steps_per_second": 21.979,
      "num_input_tokens_seen": 11227560,
      "step": 17172
    },
    {
      "epoch": 9.001572327044025,
      "grad_norm": 0.4816417098045349,
      "learning_rate": 3.354620555480251e-05,
      "loss": 0.3296,
      "num_input_tokens_seen": 11229416,
      "step": 17175
    },
    {
      "epoch": 9.0041928721174,
      "grad_norm": 0.5632688999176025,
      "learning_rate": 3.353545923283108e-05,
      "loss": 0.4254,
      "num_input_tokens_seen": 11232744,
      "step": 17180
    },
    {
      "epoch": 9.006813417190775,
      "grad_norm": 0.2693389356136322,
      "learning_rate": 3.352471112533895e-05,
      "loss": 0.4729,
      "num_input_tokens_seen": 11235880,
      "step": 17185
    },
    {
      "epoch": 9.00943396226415,
      "grad_norm": 0.2976149618625641,
      "learning_rate": 3.351396123457451e-05,
      "loss": 0.4044,
      "num_input_tokens_seen": 11240008,
      "step": 17190
    },
    {
      "epoch": 9.012054507337526,
      "grad_norm": 0.770907998085022,
      "learning_rate": 3.350320956278652e-05,
      "loss": 0.6015,
      "num_input_tokens_seen": 11242568,
      "step": 17195
    },
    {
      "epoch": 9.014675052410901,
      "grad_norm": 0.2625845670700073,
      "learning_rate": 3.349245611222409e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 11245928,
      "step": 17200
    },
    {
      "epoch": 9.017295597484276,
      "grad_norm": 0.42796748876571655,
      "learning_rate": 3.348170088513673e-05,
      "loss": 0.4199,
      "num_input_tokens_seen": 11249768,
      "step": 17205
    },
    {
      "epoch": 9.019916142557651,
      "grad_norm": 0.6350656747817993,
      "learning_rate": 3.347094388377431e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 11253384,
      "step": 17210
    },
    {
      "epoch": 9.022536687631026,
      "grad_norm": 0.40199190378189087,
      "learning_rate": 3.3460185110387076e-05,
      "loss": 0.4611,
      "num_input_tokens_seen": 11256904,
      "step": 17215
    },
    {
      "epoch": 9.025157232704403,
      "grad_norm": 0.44526207447052,
      "learning_rate": 3.344942456722563e-05,
      "loss": 0.3703,
      "num_input_tokens_seen": 11259496,
      "step": 17220
    },
    {
      "epoch": 9.027777777777779,
      "grad_norm": 0.6295145750045776,
      "learning_rate": 3.343866225654096e-05,
      "loss": 0.5394,
      "num_input_tokens_seen": 11262728,
      "step": 17225
    },
    {
      "epoch": 9.030398322851154,
      "grad_norm": 0.3350708782672882,
      "learning_rate": 3.342789818058443e-05,
      "loss": 0.3345,
      "num_input_tokens_seen": 11265448,
      "step": 17230
    },
    {
      "epoch": 9.033018867924529,
      "grad_norm": 0.42154765129089355,
      "learning_rate": 3.341713234160774e-05,
      "loss": 0.4525,
      "num_input_tokens_seen": 11269032,
      "step": 17235
    },
    {
      "epoch": 9.035639412997904,
      "grad_norm": 0.4195624589920044,
      "learning_rate": 3.3406364741862986e-05,
      "loss": 0.6019,
      "num_input_tokens_seen": 11271912,
      "step": 17240
    },
    {
      "epoch": 9.03825995807128,
      "grad_norm": 0.4431452751159668,
      "learning_rate": 3.3395595383602644e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 11274728,
      "step": 17245
    },
    {
      "epoch": 9.040880503144654,
      "grad_norm": 0.386161744594574,
      "learning_rate": 3.338482426907953e-05,
      "loss": 0.4314,
      "num_input_tokens_seen": 11278280,
      "step": 17250
    },
    {
      "epoch": 9.04350104821803,
      "grad_norm": 0.7585834860801697,
      "learning_rate": 3.337405140054683e-05,
      "loss": 0.4306,
      "num_input_tokens_seen": 11281960,
      "step": 17255
    },
    {
      "epoch": 9.046121593291405,
      "grad_norm": 0.32990574836730957,
      "learning_rate": 3.336327678025813e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 11285096,
      "step": 17260
    },
    {
      "epoch": 9.04874213836478,
      "grad_norm": 0.42435842752456665,
      "learning_rate": 3.3352500410467324e-05,
      "loss": 0.3704,
      "num_input_tokens_seen": 11288104,
      "step": 17265
    },
    {
      "epoch": 9.051362683438155,
      "grad_norm": 0.3187551200389862,
      "learning_rate": 3.3341722293428726e-05,
      "loss": 0.6965,
      "num_input_tokens_seen": 11290760,
      "step": 17270
    },
    {
      "epoch": 9.05398322851153,
      "grad_norm": 0.4046933948993683,
      "learning_rate": 3.3330942431397e-05,
      "loss": 0.5116,
      "num_input_tokens_seen": 11293800,
      "step": 17275
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 0.4659285843372345,
      "learning_rate": 3.332016082662715e-05,
      "loss": 0.5047,
      "num_input_tokens_seen": 11296520,
      "step": 17280
    },
    {
      "epoch": 9.05922431865828,
      "grad_norm": 0.26538774371147156,
      "learning_rate": 3.3309377481374584e-05,
      "loss": 0.4904,
      "num_input_tokens_seen": 11299656,
      "step": 17285
    },
    {
      "epoch": 9.061844863731656,
      "grad_norm": 0.6621416807174683,
      "learning_rate": 3.3298592397895046e-05,
      "loss": 0.4076,
      "num_input_tokens_seen": 11303080,
      "step": 17290
    },
    {
      "epoch": 9.064465408805031,
      "grad_norm": 0.5017789602279663,
      "learning_rate": 3.3287805578444645e-05,
      "loss": 0.5195,
      "num_input_tokens_seen": 11305928,
      "step": 17295
    },
    {
      "epoch": 9.067085953878406,
      "grad_norm": 0.5175681114196777,
      "learning_rate": 3.327701702527989e-05,
      "loss": 0.4981,
      "num_input_tokens_seen": 11309160,
      "step": 17300
    },
    {
      "epoch": 9.069706498951781,
      "grad_norm": 0.6335952877998352,
      "learning_rate": 3.3266226740657584e-05,
      "loss": 0.4569,
      "num_input_tokens_seen": 11311976,
      "step": 17305
    },
    {
      "epoch": 9.072327044025156,
      "grad_norm": 0.7184887528419495,
      "learning_rate": 3.3255434726834954e-05,
      "loss": 0.6251,
      "num_input_tokens_seen": 11314792,
      "step": 17310
    },
    {
      "epoch": 9.074947589098532,
      "grad_norm": 0.6431415677070618,
      "learning_rate": 3.324464098606957e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 11317736,
      "step": 17315
    },
    {
      "epoch": 9.077568134171909,
      "grad_norm": 0.28259921073913574,
      "learning_rate": 3.323384552061934e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 11320616,
      "step": 17320
    },
    {
      "epoch": 9.080188679245284,
      "grad_norm": 0.5148313641548157,
      "learning_rate": 3.322304833274257e-05,
      "loss": 0.4738,
      "num_input_tokens_seen": 11324712,
      "step": 17325
    },
    {
      "epoch": 9.082809224318659,
      "grad_norm": 0.37570834159851074,
      "learning_rate": 3.32122494246979e-05,
      "loss": 0.5255,
      "num_input_tokens_seen": 11327496,
      "step": 17330
    },
    {
      "epoch": 9.085429769392034,
      "grad_norm": 0.27795496582984924,
      "learning_rate": 3.320144879874434e-05,
      "loss": 0.4796,
      "num_input_tokens_seen": 11330408,
      "step": 17335
    },
    {
      "epoch": 9.08805031446541,
      "grad_norm": 0.40928182005882263,
      "learning_rate": 3.319064645714126e-05,
      "loss": 0.4325,
      "num_input_tokens_seen": 11333640,
      "step": 17340
    },
    {
      "epoch": 9.090670859538784,
      "grad_norm": 0.4002944827079773,
      "learning_rate": 3.3179842402148386e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 11336360,
      "step": 17345
    },
    {
      "epoch": 9.09329140461216,
      "grad_norm": 0.33413317799568176,
      "learning_rate": 3.316903663602581e-05,
      "loss": 0.3336,
      "num_input_tokens_seen": 11338696,
      "step": 17350
    },
    {
      "epoch": 9.095911949685535,
      "grad_norm": 0.46264341473579407,
      "learning_rate": 3.315822916103395e-05,
      "loss": 0.4812,
      "num_input_tokens_seen": 11343048,
      "step": 17355
    },
    {
      "epoch": 9.09853249475891,
      "grad_norm": 0.5189201831817627,
      "learning_rate": 3.3147419979433635e-05,
      "loss": 0.4762,
      "num_input_tokens_seen": 11345384,
      "step": 17360
    },
    {
      "epoch": 9.101153039832285,
      "grad_norm": 0.40802982449531555,
      "learning_rate": 3.313660909348601e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 11348232,
      "step": 17365
    },
    {
      "epoch": 9.10377358490566,
      "grad_norm": 0.4750986695289612,
      "learning_rate": 3.312579650545259e-05,
      "loss": 0.5175,
      "num_input_tokens_seen": 11351368,
      "step": 17370
    },
    {
      "epoch": 9.106394129979035,
      "grad_norm": 0.2979245185852051,
      "learning_rate": 3.311498221759524e-05,
      "loss": 0.5512,
      "num_input_tokens_seen": 11354984,
      "step": 17375
    },
    {
      "epoch": 9.10901467505241,
      "grad_norm": 0.5136995911598206,
      "learning_rate": 3.310416623217619e-05,
      "loss": 0.4676,
      "num_input_tokens_seen": 11358152,
      "step": 17380
    },
    {
      "epoch": 9.111635220125786,
      "grad_norm": 0.3654588758945465,
      "learning_rate": 3.309334855145803e-05,
      "loss": 0.3502,
      "num_input_tokens_seen": 11361128,
      "step": 17385
    },
    {
      "epoch": 9.114255765199161,
      "grad_norm": 0.3717794120311737,
      "learning_rate": 3.3082529177703684e-05,
      "loss": 0.4636,
      "num_input_tokens_seen": 11364648,
      "step": 17390
    },
    {
      "epoch": 9.116876310272536,
      "grad_norm": 0.48403775691986084,
      "learning_rate": 3.3071708113176435e-05,
      "loss": 0.4112,
      "num_input_tokens_seen": 11371624,
      "step": 17395
    },
    {
      "epoch": 9.119496855345911,
      "grad_norm": 0.5162243247032166,
      "learning_rate": 3.306088536013995e-05,
      "loss": 0.4613,
      "num_input_tokens_seen": 11374824,
      "step": 17400
    },
    {
      "epoch": 9.122117400419286,
      "grad_norm": 0.32628169655799866,
      "learning_rate": 3.3050060920858206e-05,
      "loss": 0.4923,
      "num_input_tokens_seen": 11378408,
      "step": 17405
    },
    {
      "epoch": 9.124737945492662,
      "grad_norm": 0.3937745988368988,
      "learning_rate": 3.3039234797595544e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 11382664,
      "step": 17410
    },
    {
      "epoch": 9.127358490566039,
      "grad_norm": 0.4780619442462921,
      "learning_rate": 3.3028406992616695e-05,
      "loss": 0.5009,
      "num_input_tokens_seen": 11386536,
      "step": 17415
    },
    {
      "epoch": 9.129979035639414,
      "grad_norm": 0.4407377541065216,
      "learning_rate": 3.301757750818669e-05,
      "loss": 0.491,
      "num_input_tokens_seen": 11389672,
      "step": 17420
    },
    {
      "epoch": 9.132599580712789,
      "grad_norm": 0.3992946147918701,
      "learning_rate": 3.300674634657094e-05,
      "loss": 0.3413,
      "num_input_tokens_seen": 11392872,
      "step": 17425
    },
    {
      "epoch": 9.135220125786164,
      "grad_norm": 0.46826934814453125,
      "learning_rate": 3.2995913510035206e-05,
      "loss": 0.3946,
      "num_input_tokens_seen": 11396104,
      "step": 17430
    },
    {
      "epoch": 9.13784067085954,
      "grad_norm": 0.3153223991394043,
      "learning_rate": 3.298507900084558e-05,
      "loss": 0.3849,
      "num_input_tokens_seen": 11398696,
      "step": 17435
    },
    {
      "epoch": 9.140461215932914,
      "grad_norm": 0.7157824039459229,
      "learning_rate": 3.2974242821268524e-05,
      "loss": 0.5586,
      "num_input_tokens_seen": 11401288,
      "step": 17440
    },
    {
      "epoch": 9.14308176100629,
      "grad_norm": 0.3613167405128479,
      "learning_rate": 3.2963404973570855e-05,
      "loss": 0.4158,
      "num_input_tokens_seen": 11403560,
      "step": 17445
    },
    {
      "epoch": 9.145702306079665,
      "grad_norm": 0.38073569536209106,
      "learning_rate": 3.2952565460019715e-05,
      "loss": 0.5034,
      "num_input_tokens_seen": 11405800,
      "step": 17450
    },
    {
      "epoch": 9.14832285115304,
      "grad_norm": 0.7734383344650269,
      "learning_rate": 3.29417242828826e-05,
      "loss": 0.3447,
      "num_input_tokens_seen": 11408936,
      "step": 17455
    },
    {
      "epoch": 9.150943396226415,
      "grad_norm": 0.5639548897743225,
      "learning_rate": 3.2930881444427374e-05,
      "loss": 0.4546,
      "num_input_tokens_seen": 11412296,
      "step": 17460
    },
    {
      "epoch": 9.15356394129979,
      "grad_norm": 0.3844287693500519,
      "learning_rate": 3.292003694692223e-05,
      "loss": 0.4173,
      "num_input_tokens_seen": 11415496,
      "step": 17465
    },
    {
      "epoch": 9.156184486373165,
      "grad_norm": 0.3516579270362854,
      "learning_rate": 3.290919079263572e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 11418216,
      "step": 17470
    },
    {
      "epoch": 9.15880503144654,
      "grad_norm": 0.34166234731674194,
      "learning_rate": 3.2898342983836715e-05,
      "loss": 0.4698,
      "num_input_tokens_seen": 11421608,
      "step": 17475
    },
    {
      "epoch": 9.161425576519916,
      "grad_norm": 0.3357316851615906,
      "learning_rate": 3.288749352279447e-05,
      "loss": 0.5028,
      "num_input_tokens_seen": 11424200,
      "step": 17480
    },
    {
      "epoch": 9.164046121593291,
      "grad_norm": 0.5208860635757446,
      "learning_rate": 3.2876642411778573e-05,
      "loss": 0.5639,
      "num_input_tokens_seen": 11427272,
      "step": 17485
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 0.34520232677459717,
      "learning_rate": 3.286578965305894e-05,
      "loss": 0.4193,
      "num_input_tokens_seen": 11429576,
      "step": 17490
    },
    {
      "epoch": 9.169287211740041,
      "grad_norm": 0.31382250785827637,
      "learning_rate": 3.285493524890585e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 11433288,
      "step": 17495
    },
    {
      "epoch": 9.171907756813416,
      "grad_norm": 0.6768757700920105,
      "learning_rate": 3.284407920158992e-05,
      "loss": 0.3649,
      "num_input_tokens_seen": 11436392,
      "step": 17500
    },
    {
      "epoch": 9.174528301886792,
      "grad_norm": 1.2298529148101807,
      "learning_rate": 3.283322151338211e-05,
      "loss": 0.4039,
      "num_input_tokens_seen": 11439848,
      "step": 17505
    },
    {
      "epoch": 9.177148846960169,
      "grad_norm": 0.3611311614513397,
      "learning_rate": 3.282236218655373e-05,
      "loss": 0.3669,
      "num_input_tokens_seen": 11442824,
      "step": 17510
    },
    {
      "epoch": 9.179769392033544,
      "grad_norm": 0.5455644130706787,
      "learning_rate": 3.281150122337642e-05,
      "loss": 0.4921,
      "num_input_tokens_seen": 11445640,
      "step": 17515
    },
    {
      "epoch": 9.182389937106919,
      "grad_norm": 0.37915128469467163,
      "learning_rate": 3.280063862612216e-05,
      "loss": 0.4347,
      "num_input_tokens_seen": 11449928,
      "step": 17520
    },
    {
      "epoch": 9.185010482180294,
      "grad_norm": 1.2900731563568115,
      "learning_rate": 3.2789774397063314e-05,
      "loss": 0.6301,
      "num_input_tokens_seen": 11452328,
      "step": 17525
    },
    {
      "epoch": 9.18763102725367,
      "grad_norm": 0.2712387442588806,
      "learning_rate": 3.277890853847253e-05,
      "loss": 0.3163,
      "num_input_tokens_seen": 11456424,
      "step": 17530
    },
    {
      "epoch": 9.190251572327044,
      "grad_norm": 0.37574276328086853,
      "learning_rate": 3.2768041052622824e-05,
      "loss": 0.5189,
      "num_input_tokens_seen": 11460168,
      "step": 17535
    },
    {
      "epoch": 9.19287211740042,
      "grad_norm": 0.44903433322906494,
      "learning_rate": 3.275717194178755e-05,
      "loss": 0.4873,
      "num_input_tokens_seen": 11462600,
      "step": 17540
    },
    {
      "epoch": 9.195492662473795,
      "grad_norm": 0.3868100941181183,
      "learning_rate": 3.27463012082404e-05,
      "loss": 0.6001,
      "num_input_tokens_seen": 11465160,
      "step": 17545
    },
    {
      "epoch": 9.19811320754717,
      "grad_norm": 0.31685173511505127,
      "learning_rate": 3.2735428854255416e-05,
      "loss": 0.4329,
      "num_input_tokens_seen": 11468680,
      "step": 17550
    },
    {
      "epoch": 9.200733752620545,
      "grad_norm": 0.3853674829006195,
      "learning_rate": 3.272455488210697e-05,
      "loss": 0.4139,
      "num_input_tokens_seen": 11472776,
      "step": 17555
    },
    {
      "epoch": 9.20335429769392,
      "grad_norm": 0.36442694067955017,
      "learning_rate": 3.271367929406977e-05,
      "loss": 0.4067,
      "num_input_tokens_seen": 11477672,
      "step": 17560
    },
    {
      "epoch": 9.205974842767295,
      "grad_norm": 0.5255393981933594,
      "learning_rate": 3.270280209241886e-05,
      "loss": 0.3692,
      "num_input_tokens_seen": 11481544,
      "step": 17565
    },
    {
      "epoch": 9.20859538784067,
      "grad_norm": 0.3964151442050934,
      "learning_rate": 3.2691923279429624e-05,
      "loss": 0.4121,
      "num_input_tokens_seen": 11484648,
      "step": 17570
    },
    {
      "epoch": 9.211215932914046,
      "grad_norm": 0.4969254732131958,
      "learning_rate": 3.2681042857377804e-05,
      "loss": 0.4288,
      "num_input_tokens_seen": 11487496,
      "step": 17575
    },
    {
      "epoch": 9.213836477987421,
      "grad_norm": 0.3140166103839874,
      "learning_rate": 3.2670160828539435e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 11490088,
      "step": 17580
    },
    {
      "epoch": 9.216457023060796,
      "grad_norm": 0.5594607591629028,
      "learning_rate": 3.2659277195190934e-05,
      "loss": 0.3987,
      "num_input_tokens_seen": 11493192,
      "step": 17585
    },
    {
      "epoch": 9.219077568134171,
      "grad_norm": 0.6023274064064026,
      "learning_rate": 3.2648391959609026e-05,
      "loss": 0.4883,
      "num_input_tokens_seen": 11496584,
      "step": 17590
    },
    {
      "epoch": 9.221698113207546,
      "grad_norm": 0.5593783855438232,
      "learning_rate": 3.263750512407078e-05,
      "loss": 0.3713,
      "num_input_tokens_seen": 11499528,
      "step": 17595
    },
    {
      "epoch": 9.224318658280922,
      "grad_norm": 0.5220418572425842,
      "learning_rate": 3.262661669085359e-05,
      "loss": 0.5188,
      "num_input_tokens_seen": 11502344,
      "step": 17600
    },
    {
      "epoch": 9.226939203354299,
      "grad_norm": 0.3692466914653778,
      "learning_rate": 3.2615726662235205e-05,
      "loss": 0.3833,
      "num_input_tokens_seen": 11505128,
      "step": 17605
    },
    {
      "epoch": 9.229559748427674,
      "grad_norm": 0.4711281657218933,
      "learning_rate": 3.2604835040493676e-05,
      "loss": 0.4029,
      "num_input_tokens_seen": 11508232,
      "step": 17610
    },
    {
      "epoch": 9.232180293501049,
      "grad_norm": 0.35326874256134033,
      "learning_rate": 3.259394182790742e-05,
      "loss": 0.5418,
      "num_input_tokens_seen": 11512136,
      "step": 17615
    },
    {
      "epoch": 9.234800838574424,
      "grad_norm": 0.546234667301178,
      "learning_rate": 3.2583047026755174e-05,
      "loss": 0.4886,
      "num_input_tokens_seen": 11514920,
      "step": 17620
    },
    {
      "epoch": 9.2374213836478,
      "grad_norm": 0.2693645656108856,
      "learning_rate": 3.2572150639316e-05,
      "loss": 0.3569,
      "num_input_tokens_seen": 11518280,
      "step": 17625
    },
    {
      "epoch": 9.240041928721174,
      "grad_norm": 0.3103209435939789,
      "learning_rate": 3.256125266786931e-05,
      "loss": 0.3063,
      "num_input_tokens_seen": 11520904,
      "step": 17630
    },
    {
      "epoch": 9.24266247379455,
      "grad_norm": 0.4757659137248993,
      "learning_rate": 3.255035311469481e-05,
      "loss": 0.4941,
      "num_input_tokens_seen": 11524904,
      "step": 17635
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 0.5201685428619385,
      "learning_rate": 3.2539451982072586e-05,
      "loss": 0.5082,
      "num_input_tokens_seen": 11528680,
      "step": 17640
    },
    {
      "epoch": 9.2479035639413,
      "grad_norm": 1.142702341079712,
      "learning_rate": 3.2528549272283016e-05,
      "loss": 0.3883,
      "num_input_tokens_seen": 11531176,
      "step": 17645
    },
    {
      "epoch": 9.250524109014675,
      "grad_norm": 0.41764047741889954,
      "learning_rate": 3.251764498760683e-05,
      "loss": 0.4705,
      "num_input_tokens_seen": 11534536,
      "step": 17650
    },
    {
      "epoch": 9.25314465408805,
      "grad_norm": 0.7535741925239563,
      "learning_rate": 3.2506739130325084e-05,
      "loss": 0.3818,
      "num_input_tokens_seen": 11537224,
      "step": 17655
    },
    {
      "epoch": 9.255765199161425,
      "grad_norm": 0.5549479126930237,
      "learning_rate": 3.2495831702719146e-05,
      "loss": 0.2921,
      "num_input_tokens_seen": 11540456,
      "step": 17660
    },
    {
      "epoch": 9.2583857442348,
      "grad_norm": 0.5670521855354309,
      "learning_rate": 3.248492270707074e-05,
      "loss": 0.3304,
      "num_input_tokens_seen": 11542824,
      "step": 17665
    },
    {
      "epoch": 9.261006289308176,
      "grad_norm": 0.3851185739040375,
      "learning_rate": 3.247401214566189e-05,
      "loss": 0.7564,
      "num_input_tokens_seen": 11546440,
      "step": 17670
    },
    {
      "epoch": 9.26362683438155,
      "grad_norm": 0.3299659192562103,
      "learning_rate": 3.2463100020774965e-05,
      "loss": 0.4043,
      "num_input_tokens_seen": 11549416,
      "step": 17675
    },
    {
      "epoch": 9.266247379454926,
      "grad_norm": 0.4885091185569763,
      "learning_rate": 3.245218633469267e-05,
      "loss": 0.4878,
      "num_input_tokens_seen": 11553000,
      "step": 17680
    },
    {
      "epoch": 9.268867924528301,
      "grad_norm": 0.38146334886550903,
      "learning_rate": 3.244127108969801e-05,
      "loss": 0.4049,
      "num_input_tokens_seen": 11555848,
      "step": 17685
    },
    {
      "epoch": 9.271488469601676,
      "grad_norm": 0.8177158236503601,
      "learning_rate": 3.2430354288074314e-05,
      "loss": 0.5413,
      "num_input_tokens_seen": 11559176,
      "step": 17690
    },
    {
      "epoch": 9.274109014675052,
      "grad_norm": 0.639996349811554,
      "learning_rate": 3.241943593210529e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 11562376,
      "step": 17695
    },
    {
      "epoch": 9.276729559748428,
      "grad_norm": 0.5557040572166443,
      "learning_rate": 3.2408516024074916e-05,
      "loss": 0.4866,
      "num_input_tokens_seen": 11565192,
      "step": 17700
    },
    {
      "epoch": 9.279350104821804,
      "grad_norm": 0.347477525472641,
      "learning_rate": 3.2397594566267505e-05,
      "loss": 0.324,
      "num_input_tokens_seen": 11568392,
      "step": 17705
    },
    {
      "epoch": 9.281970649895179,
      "grad_norm": 0.301212877035141,
      "learning_rate": 3.23866715609677e-05,
      "loss": 0.4237,
      "num_input_tokens_seen": 11572168,
      "step": 17710
    },
    {
      "epoch": 9.284591194968554,
      "grad_norm": 0.4855177402496338,
      "learning_rate": 3.237574701046049e-05,
      "loss": 0.4134,
      "num_input_tokens_seen": 11575016,
      "step": 17715
    },
    {
      "epoch": 9.28721174004193,
      "grad_norm": 0.5932342410087585,
      "learning_rate": 3.236482091703115e-05,
      "loss": 0.4347,
      "num_input_tokens_seen": 11578056,
      "step": 17720
    },
    {
      "epoch": 9.289832285115304,
      "grad_norm": 0.33949559926986694,
      "learning_rate": 3.235389328296528e-05,
      "loss": 0.5193,
      "num_input_tokens_seen": 11580520,
      "step": 17725
    },
    {
      "epoch": 9.29245283018868,
      "grad_norm": 0.3717341125011444,
      "learning_rate": 3.2342964110548845e-05,
      "loss": 0.3392,
      "num_input_tokens_seen": 11584040,
      "step": 17730
    },
    {
      "epoch": 9.295073375262055,
      "grad_norm": 0.27621230483055115,
      "learning_rate": 3.2332033402068093e-05,
      "loss": 0.4298,
      "num_input_tokens_seen": 11589000,
      "step": 17735
    },
    {
      "epoch": 9.29769392033543,
      "grad_norm": 0.403839647769928,
      "learning_rate": 3.2321101159809605e-05,
      "loss": 0.7306,
      "num_input_tokens_seen": 11592328,
      "step": 17740
    },
    {
      "epoch": 9.300314465408805,
      "grad_norm": 0.5997262001037598,
      "learning_rate": 3.231016738606027e-05,
      "loss": 0.4069,
      "num_input_tokens_seen": 11595080,
      "step": 17745
    },
    {
      "epoch": 9.30293501048218,
      "grad_norm": 0.5207831263542175,
      "learning_rate": 3.2299232083107313e-05,
      "loss": 0.5073,
      "num_input_tokens_seen": 11598088,
      "step": 17750
    },
    {
      "epoch": 9.305555555555555,
      "grad_norm": 0.48694196343421936,
      "learning_rate": 3.228829525323829e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 11601672,
      "step": 17755
    },
    {
      "epoch": 9.30817610062893,
      "grad_norm": 0.5683486461639404,
      "learning_rate": 3.227735689874104e-05,
      "loss": 0.4581,
      "num_input_tokens_seen": 11604648,
      "step": 17760
    },
    {
      "epoch": 9.310796645702306,
      "grad_norm": 0.37636637687683105,
      "learning_rate": 3.226641702190375e-05,
      "loss": 0.497,
      "num_input_tokens_seen": 11607688,
      "step": 17765
    },
    {
      "epoch": 9.31341719077568,
      "grad_norm": 0.35994696617126465,
      "learning_rate": 3.225547562501493e-05,
      "loss": 0.5514,
      "num_input_tokens_seen": 11610920,
      "step": 17770
    },
    {
      "epoch": 9.316037735849056,
      "grad_norm": 0.6109917759895325,
      "learning_rate": 3.2244532710363375e-05,
      "loss": 0.4422,
      "num_input_tokens_seen": 11613992,
      "step": 17775
    },
    {
      "epoch": 9.318658280922431,
      "grad_norm": 0.5495603084564209,
      "learning_rate": 3.2233588280238224e-05,
      "loss": 0.4288,
      "num_input_tokens_seen": 11616776,
      "step": 17780
    },
    {
      "epoch": 9.321278825995806,
      "grad_norm": 0.4013109803199768,
      "learning_rate": 3.2222642336928944e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 11619592,
      "step": 17785
    },
    {
      "epoch": 9.323899371069182,
      "grad_norm": 0.6790323257446289,
      "learning_rate": 3.2211694882725275e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 11622952,
      "step": 17790
    },
    {
      "epoch": 9.326519916142558,
      "grad_norm": 0.3521025478839874,
      "learning_rate": 3.220074591991731e-05,
      "loss": 0.382,
      "num_input_tokens_seen": 11626344,
      "step": 17795
    },
    {
      "epoch": 9.329140461215934,
      "grad_norm": 0.43676114082336426,
      "learning_rate": 3.218979545079546e-05,
      "loss": 0.4719,
      "num_input_tokens_seen": 11629832,
      "step": 17800
    },
    {
      "epoch": 9.331761006289309,
      "grad_norm": 0.7112559080123901,
      "learning_rate": 3.217884347765042e-05,
      "loss": 0.4077,
      "num_input_tokens_seen": 11632584,
      "step": 17805
    },
    {
      "epoch": 9.334381551362684,
      "grad_norm": 0.5124146342277527,
      "learning_rate": 3.216789000277323e-05,
      "loss": 0.4646,
      "num_input_tokens_seen": 11635560,
      "step": 17810
    },
    {
      "epoch": 9.33700209643606,
      "grad_norm": 0.37022170424461365,
      "learning_rate": 3.2156935028455224e-05,
      "loss": 0.4066,
      "num_input_tokens_seen": 11638856,
      "step": 17815
    },
    {
      "epoch": 9.339622641509434,
      "grad_norm": 0.6431238055229187,
      "learning_rate": 3.2145978556988056e-05,
      "loss": 0.5853,
      "num_input_tokens_seen": 11641448,
      "step": 17820
    },
    {
      "epoch": 9.34224318658281,
      "grad_norm": 0.3519207537174225,
      "learning_rate": 3.213502059066371e-05,
      "loss": 0.5184,
      "num_input_tokens_seen": 11645864,
      "step": 17825
    },
    {
      "epoch": 9.344863731656185,
      "grad_norm": 0.39056122303009033,
      "learning_rate": 3.212406113177445e-05,
      "loss": 0.4417,
      "num_input_tokens_seen": 11648872,
      "step": 17830
    },
    {
      "epoch": 9.34748427672956,
      "grad_norm": 0.5216460824012756,
      "learning_rate": 3.211310018261287e-05,
      "loss": 0.4071,
      "num_input_tokens_seen": 11651976,
      "step": 17835
    },
    {
      "epoch": 9.350104821802935,
      "grad_norm": 0.49556005001068115,
      "learning_rate": 3.21021377454719e-05,
      "loss": 0.5125,
      "num_input_tokens_seen": 11655432,
      "step": 17840
    },
    {
      "epoch": 9.35272536687631,
      "grad_norm": 0.4643397033214569,
      "learning_rate": 3.209117382264473e-05,
      "loss": 0.6407,
      "num_input_tokens_seen": 11658376,
      "step": 17845
    },
    {
      "epoch": 9.355345911949685,
      "grad_norm": 0.4323780834674835,
      "learning_rate": 3.208020841642489e-05,
      "loss": 0.3365,
      "num_input_tokens_seen": 11662120,
      "step": 17850
    },
    {
      "epoch": 9.35796645702306,
      "grad_norm": 0.735758900642395,
      "learning_rate": 3.206924152910623e-05,
      "loss": 0.4452,
      "num_input_tokens_seen": 11665512,
      "step": 17855
    },
    {
      "epoch": 9.360587002096436,
      "grad_norm": 0.3116666078567505,
      "learning_rate": 3.205827316298289e-05,
      "loss": 0.4397,
      "num_input_tokens_seen": 11668264,
      "step": 17860
    },
    {
      "epoch": 9.36320754716981,
      "grad_norm": 0.6887289881706238,
      "learning_rate": 3.204730332034934e-05,
      "loss": 0.4591,
      "num_input_tokens_seen": 11671592,
      "step": 17865
    },
    {
      "epoch": 9.365828092243186,
      "grad_norm": 0.437082976102829,
      "learning_rate": 3.2036332003500324e-05,
      "loss": 0.5818,
      "num_input_tokens_seen": 11675368,
      "step": 17870
    },
    {
      "epoch": 9.368448637316561,
      "grad_norm": 0.4668407738208771,
      "learning_rate": 3.2025359214730924e-05,
      "loss": 0.4138,
      "num_input_tokens_seen": 11678600,
      "step": 17875
    },
    {
      "epoch": 9.371069182389936,
      "grad_norm": 0.428043007850647,
      "learning_rate": 3.201438495633654e-05,
      "loss": 0.5544,
      "num_input_tokens_seen": 11681672,
      "step": 17880
    },
    {
      "epoch": 9.373689727463312,
      "grad_norm": 0.6919742226600647,
      "learning_rate": 3.200340923061284e-05,
      "loss": 0.5177,
      "num_input_tokens_seen": 11684648,
      "step": 17885
    },
    {
      "epoch": 9.376310272536688,
      "grad_norm": 0.5031125545501709,
      "learning_rate": 3.199243203985583e-05,
      "loss": 0.4014,
      "num_input_tokens_seen": 11687400,
      "step": 17890
    },
    {
      "epoch": 9.378930817610064,
      "grad_norm": 0.8136667013168335,
      "learning_rate": 3.1981453386361804e-05,
      "loss": 0.3756,
      "num_input_tokens_seen": 11689768,
      "step": 17895
    },
    {
      "epoch": 9.381551362683439,
      "grad_norm": 0.49837785959243774,
      "learning_rate": 3.1970473272427386e-05,
      "loss": 0.3806,
      "num_input_tokens_seen": 11692456,
      "step": 17900
    },
    {
      "epoch": 9.384171907756814,
      "grad_norm": 0.29741188883781433,
      "learning_rate": 3.195949170034948e-05,
      "loss": 0.37,
      "num_input_tokens_seen": 11696360,
      "step": 17905
    },
    {
      "epoch": 9.38679245283019,
      "grad_norm": 0.6488888263702393,
      "learning_rate": 3.19485086724253e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 11699816,
      "step": 17910
    },
    {
      "epoch": 9.389412997903564,
      "grad_norm": 0.5605972409248352,
      "learning_rate": 3.193752419095239e-05,
      "loss": 0.5185,
      "num_input_tokens_seen": 11702664,
      "step": 17915
    },
    {
      "epoch": 9.39203354297694,
      "grad_norm": 0.32964327931404114,
      "learning_rate": 3.192653825822857e-05,
      "loss": 0.4676,
      "num_input_tokens_seen": 11706696,
      "step": 17920
    },
    {
      "epoch": 9.394654088050315,
      "grad_norm": 0.46062394976615906,
      "learning_rate": 3.1915550876551964e-05,
      "loss": 0.5651,
      "num_input_tokens_seen": 11709928,
      "step": 17925
    },
    {
      "epoch": 9.39727463312369,
      "grad_norm": 0.4502320885658264,
      "learning_rate": 3.190456204822101e-05,
      "loss": 0.3374,
      "num_input_tokens_seen": 11713256,
      "step": 17930
    },
    {
      "epoch": 9.399895178197065,
      "grad_norm": 0.6928263902664185,
      "learning_rate": 3.1893571775534445e-05,
      "loss": 0.5923,
      "num_input_tokens_seen": 11717256,
      "step": 17935
    },
    {
      "epoch": 9.40251572327044,
      "grad_norm": 0.5154154896736145,
      "learning_rate": 3.1882580060791315e-05,
      "loss": 0.5104,
      "num_input_tokens_seen": 11720168,
      "step": 17940
    },
    {
      "epoch": 9.405136268343815,
      "grad_norm": 0.7934046387672424,
      "learning_rate": 3.187158690629095e-05,
      "loss": 0.5079,
      "num_input_tokens_seen": 11722952,
      "step": 17945
    },
    {
      "epoch": 9.40775681341719,
      "grad_norm": 0.5648139715194702,
      "learning_rate": 3.1860592314333e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 11725128,
      "step": 17950
    },
    {
      "epoch": 9.410377358490566,
      "grad_norm": 0.5426859855651855,
      "learning_rate": 3.184959628721742e-05,
      "loss": 0.4257,
      "num_input_tokens_seen": 11727720,
      "step": 17955
    },
    {
      "epoch": 9.41299790356394,
      "grad_norm": 0.38214391469955444,
      "learning_rate": 3.183859882724443e-05,
      "loss": 0.3924,
      "num_input_tokens_seen": 11731016,
      "step": 17960
    },
    {
      "epoch": 9.415618448637316,
      "grad_norm": 0.27949345111846924,
      "learning_rate": 3.1827599936714584e-05,
      "loss": 0.3004,
      "num_input_tokens_seen": 11733992,
      "step": 17965
    },
    {
      "epoch": 9.418238993710691,
      "grad_norm": 0.4728333353996277,
      "learning_rate": 3.1816599617928736e-05,
      "loss": 0.4809,
      "num_input_tokens_seen": 11737448,
      "step": 17970
    },
    {
      "epoch": 9.420859538784066,
      "grad_norm": 0.40692323446273804,
      "learning_rate": 3.1805597873188e-05,
      "loss": 0.3954,
      "num_input_tokens_seen": 11741256,
      "step": 17975
    },
    {
      "epoch": 9.423480083857442,
      "grad_norm": 0.4244961738586426,
      "learning_rate": 3.179459470479384e-05,
      "loss": 0.3398,
      "num_input_tokens_seen": 11743624,
      "step": 17980
    },
    {
      "epoch": 9.426100628930818,
      "grad_norm": 0.6500338912010193,
      "learning_rate": 3.1783590115047983e-05,
      "loss": 0.5581,
      "num_input_tokens_seen": 11746184,
      "step": 17985
    },
    {
      "epoch": 9.428721174004194,
      "grad_norm": 0.43151625990867615,
      "learning_rate": 3.177258410625246e-05,
      "loss": 0.3726,
      "num_input_tokens_seen": 11749544,
      "step": 17990
    },
    {
      "epoch": 9.431341719077569,
      "grad_norm": 0.4550926387310028,
      "learning_rate": 3.176157668070962e-05,
      "loss": 0.5121,
      "num_input_tokens_seen": 11752456,
      "step": 17995
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 0.5907080769538879,
      "learning_rate": 3.175056784072207e-05,
      "loss": 0.4771,
      "num_input_tokens_seen": 11755720,
      "step": 18000
    },
    {
      "epoch": 9.43658280922432,
      "grad_norm": 0.5143696069717407,
      "learning_rate": 3.173955758859274e-05,
      "loss": 0.3652,
      "num_input_tokens_seen": 11758664,
      "step": 18005
    },
    {
      "epoch": 9.439203354297694,
      "grad_norm": 0.7788112759590149,
      "learning_rate": 3.172854592662486e-05,
      "loss": 0.6635,
      "num_input_tokens_seen": 11761576,
      "step": 18010
    },
    {
      "epoch": 9.44182389937107,
      "grad_norm": 0.27723953127861023,
      "learning_rate": 3.1717532857121926e-05,
      "loss": 0.4011,
      "num_input_tokens_seen": 11765832,
      "step": 18015
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 0.39338961243629456,
      "learning_rate": 3.170651838238776e-05,
      "loss": 0.3982,
      "num_input_tokens_seen": 11773160,
      "step": 18020
    },
    {
      "epoch": 9.44706498951782,
      "grad_norm": 0.39529159665107727,
      "learning_rate": 3.169550250472646e-05,
      "loss": 0.4455,
      "num_input_tokens_seen": 11776648,
      "step": 18025
    },
    {
      "epoch": 9.449685534591195,
      "grad_norm": 0.39279311895370483,
      "learning_rate": 3.168448522644243e-05,
      "loss": 0.3668,
      "num_input_tokens_seen": 11779976,
      "step": 18030
    },
    {
      "epoch": 9.45230607966457,
      "grad_norm": 0.6892783641815186,
      "learning_rate": 3.167346654984034e-05,
      "loss": 0.3308,
      "num_input_tokens_seen": 11783784,
      "step": 18035
    },
    {
      "epoch": 9.454926624737945,
      "grad_norm": 0.48027071356773376,
      "learning_rate": 3.166244647722519e-05,
      "loss": 0.5784,
      "num_input_tokens_seen": 11787176,
      "step": 18040
    },
    {
      "epoch": 9.45754716981132,
      "grad_norm": 0.5480319261550903,
      "learning_rate": 3.1651425010902244e-05,
      "loss": 0.4979,
      "num_input_tokens_seen": 11789864,
      "step": 18045
    },
    {
      "epoch": 9.460167714884696,
      "grad_norm": 0.33225885033607483,
      "learning_rate": 3.164040215317706e-05,
      "loss": 0.3535,
      "num_input_tokens_seen": 11795432,
      "step": 18050
    },
    {
      "epoch": 9.46278825995807,
      "grad_norm": 0.7677196860313416,
      "learning_rate": 3.162937790635552e-05,
      "loss": 0.3771,
      "num_input_tokens_seen": 11798472,
      "step": 18055
    },
    {
      "epoch": 9.465408805031446,
      "grad_norm": 1.1224662065505981,
      "learning_rate": 3.1618352272743746e-05,
      "loss": 0.5543,
      "num_input_tokens_seen": 11800872,
      "step": 18060
    },
    {
      "epoch": 9.468029350104821,
      "grad_norm": 0.31548604369163513,
      "learning_rate": 3.160732525464818e-05,
      "loss": 0.3599,
      "num_input_tokens_seen": 11804392,
      "step": 18065
    },
    {
      "epoch": 9.470649895178196,
      "grad_norm": 0.4838969111442566,
      "learning_rate": 3.1596296854375554e-05,
      "loss": 0.5716,
      "num_input_tokens_seen": 11808520,
      "step": 18070
    },
    {
      "epoch": 9.473270440251572,
      "grad_norm": 0.3941914141178131,
      "learning_rate": 3.1585267074232884e-05,
      "loss": 0.3433,
      "num_input_tokens_seen": 11812008,
      "step": 18075
    },
    {
      "epoch": 9.475890985324948,
      "grad_norm": 0.5796436667442322,
      "learning_rate": 3.157423591652746e-05,
      "loss": 0.4505,
      "num_input_tokens_seen": 11815304,
      "step": 18080
    },
    {
      "epoch": 9.478511530398324,
      "grad_norm": 0.7859403491020203,
      "learning_rate": 3.1563203383566894e-05,
      "loss": 0.5625,
      "num_input_tokens_seen": 11818376,
      "step": 18085
    },
    {
      "epoch": 9.481132075471699,
      "grad_norm": 0.3451259434223175,
      "learning_rate": 3.155216947765906e-05,
      "loss": 0.4268,
      "num_input_tokens_seen": 11821000,
      "step": 18090
    },
    {
      "epoch": 9.483752620545074,
      "grad_norm": 0.5177667737007141,
      "learning_rate": 3.154113420111211e-05,
      "loss": 0.4763,
      "num_input_tokens_seen": 11823656,
      "step": 18095
    },
    {
      "epoch": 9.48637316561845,
      "grad_norm": 0.4886195957660675,
      "learning_rate": 3.1530097556234526e-05,
      "loss": 0.479,
      "num_input_tokens_seen": 11826824,
      "step": 18100
    },
    {
      "epoch": 9.488993710691824,
      "grad_norm": 0.5847464203834534,
      "learning_rate": 3.151905954533502e-05,
      "loss": 0.3722,
      "num_input_tokens_seen": 11830280,
      "step": 18105
    },
    {
      "epoch": 9.4916142557652,
      "grad_norm": 0.331276535987854,
      "learning_rate": 3.150802017072263e-05,
      "loss": 0.4918,
      "num_input_tokens_seen": 11833768,
      "step": 18110
    },
    {
      "epoch": 9.494234800838575,
      "grad_norm": 0.5481605529785156,
      "learning_rate": 3.149697943470668e-05,
      "loss": 0.463,
      "num_input_tokens_seen": 11836328,
      "step": 18115
    },
    {
      "epoch": 9.49685534591195,
      "grad_norm": 0.2935040295124054,
      "learning_rate": 3.1485937339596736e-05,
      "loss": 0.3416,
      "num_input_tokens_seen": 11839112,
      "step": 18120
    },
    {
      "epoch": 9.499475890985325,
      "grad_norm": 0.5715702772140503,
      "learning_rate": 3.147489388770271e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 11843592,
      "step": 18125
    },
    {
      "epoch": 9.5020964360587,
      "grad_norm": 0.8803171515464783,
      "learning_rate": 3.146384908133475e-05,
      "loss": 0.4669,
      "num_input_tokens_seen": 11846088,
      "step": 18130
    },
    {
      "epoch": 9.504716981132075,
      "grad_norm": 0.41213560104370117,
      "learning_rate": 3.14528029228033e-05,
      "loss": 0.5706,
      "num_input_tokens_seen": 11848872,
      "step": 18135
    },
    {
      "epoch": 9.50733752620545,
      "grad_norm": 0.9600390195846558,
      "learning_rate": 3.144175541441911e-05,
      "loss": 0.4433,
      "num_input_tokens_seen": 11852136,
      "step": 18140
    },
    {
      "epoch": 9.509958071278826,
      "grad_norm": 0.07529425621032715,
      "learning_rate": 3.1430706558493166e-05,
      "loss": 0.3663,
      "num_input_tokens_seen": 11858600,
      "step": 18145
    },
    {
      "epoch": 9.5125786163522,
      "grad_norm": 0.3532361686229706,
      "learning_rate": 3.141965635733677e-05,
      "loss": 0.5811,
      "num_input_tokens_seen": 11862344,
      "step": 18150
    },
    {
      "epoch": 9.515199161425576,
      "grad_norm": 0.3625589609146118,
      "learning_rate": 3.140860481326152e-05,
      "loss": 0.4039,
      "num_input_tokens_seen": 11864808,
      "step": 18155
    },
    {
      "epoch": 9.517819706498951,
      "grad_norm": 0.44902604818344116,
      "learning_rate": 3.139755192857925e-05,
      "loss": 0.51,
      "num_input_tokens_seen": 11867880,
      "step": 18160
    },
    {
      "epoch": 9.520440251572326,
      "grad_norm": 0.3927319347858429,
      "learning_rate": 3.1386497705602094e-05,
      "loss": 0.3473,
      "num_input_tokens_seen": 11871304,
      "step": 18165
    },
    {
      "epoch": 9.523060796645701,
      "grad_norm": 0.3628002405166626,
      "learning_rate": 3.137544214664249e-05,
      "loss": 0.4326,
      "num_input_tokens_seen": 11874312,
      "step": 18170
    },
    {
      "epoch": 9.525681341719078,
      "grad_norm": 0.44180798530578613,
      "learning_rate": 3.136438525401312e-05,
      "loss": 0.4257,
      "num_input_tokens_seen": 11878408,
      "step": 18175
    },
    {
      "epoch": 9.528301886792454,
      "grad_norm": 1.0617661476135254,
      "learning_rate": 3.135332703002696e-05,
      "loss": 0.5142,
      "num_input_tokens_seen": 11882408,
      "step": 18180
    },
    {
      "epoch": 9.530922431865829,
      "grad_norm": 0.4175862669944763,
      "learning_rate": 3.1342267476997276e-05,
      "loss": 0.386,
      "num_input_tokens_seen": 11885640,
      "step": 18185
    },
    {
      "epoch": 9.533542976939204,
      "grad_norm": 0.5757549405097961,
      "learning_rate": 3.1331206597237584e-05,
      "loss": 0.444,
      "num_input_tokens_seen": 11889064,
      "step": 18190
    },
    {
      "epoch": 9.536163522012579,
      "grad_norm": 0.48629897832870483,
      "learning_rate": 3.1320144393061715e-05,
      "loss": 0.5947,
      "num_input_tokens_seen": 11892264,
      "step": 18195
    },
    {
      "epoch": 9.538784067085954,
      "grad_norm": 0.4681134521961212,
      "learning_rate": 3.130908086678373e-05,
      "loss": 0.3143,
      "num_input_tokens_seen": 11896200,
      "step": 18200
    },
    {
      "epoch": 9.54140461215933,
      "grad_norm": 0.4785352349281311,
      "learning_rate": 3.1298016020718014e-05,
      "loss": 0.4359,
      "num_input_tokens_seen": 11899048,
      "step": 18205
    },
    {
      "epoch": 9.544025157232705,
      "grad_norm": 1.4471404552459717,
      "learning_rate": 3.128694985717919e-05,
      "loss": 0.3571,
      "num_input_tokens_seen": 11901416,
      "step": 18210
    },
    {
      "epoch": 9.54664570230608,
      "grad_norm": 0.5406640768051147,
      "learning_rate": 3.1275882378482185e-05,
      "loss": 0.328,
      "num_input_tokens_seen": 11905224,
      "step": 18215
    },
    {
      "epoch": 9.549266247379455,
      "grad_norm": 0.5661073327064514,
      "learning_rate": 3.1264813586942185e-05,
      "loss": 0.4832,
      "num_input_tokens_seen": 11909224,
      "step": 18220
    },
    {
      "epoch": 9.55188679245283,
      "grad_norm": 0.4803616404533386,
      "learning_rate": 3.125374348487466e-05,
      "loss": 0.489,
      "num_input_tokens_seen": 11912520,
      "step": 18225
    },
    {
      "epoch": 9.554507337526205,
      "grad_norm": 0.5391477942466736,
      "learning_rate": 3.1242672074595334e-05,
      "loss": 0.4395,
      "num_input_tokens_seen": 11916392,
      "step": 18230
    },
    {
      "epoch": 9.55712788259958,
      "grad_norm": 0.3770046830177307,
      "learning_rate": 3.123159935842024e-05,
      "loss": 0.4005,
      "num_input_tokens_seen": 11918824,
      "step": 18235
    },
    {
      "epoch": 9.559748427672956,
      "grad_norm": 0.2997472286224365,
      "learning_rate": 3.122052533866565e-05,
      "loss": 0.3651,
      "num_input_tokens_seen": 11921864,
      "step": 18240
    },
    {
      "epoch": 9.56236897274633,
      "grad_norm": 0.4332044720649719,
      "learning_rate": 3.120945001764813e-05,
      "loss": 0.522,
      "num_input_tokens_seen": 11924904,
      "step": 18245
    },
    {
      "epoch": 9.564989517819706,
      "grad_norm": 0.6228184700012207,
      "learning_rate": 3.11983733976845e-05,
      "loss": 0.3803,
      "num_input_tokens_seen": 11930152,
      "step": 18250
    },
    {
      "epoch": 9.567610062893081,
      "grad_norm": 0.5516433715820312,
      "learning_rate": 3.1187295481091876e-05,
      "loss": 0.4697,
      "num_input_tokens_seen": 11933288,
      "step": 18255
    },
    {
      "epoch": 9.570230607966456,
      "grad_norm": 0.8913248777389526,
      "learning_rate": 3.1176216270187625e-05,
      "loss": 0.3974,
      "num_input_tokens_seen": 11936744,
      "step": 18260
    },
    {
      "epoch": 9.572851153039831,
      "grad_norm": 0.8455912470817566,
      "learning_rate": 3.116513576728939e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 11939912,
      "step": 18265
    },
    {
      "epoch": 9.575471698113208,
      "grad_norm": 0.4964645504951477,
      "learning_rate": 3.1154053974715085e-05,
      "loss": 0.4405,
      "num_input_tokens_seen": 11942568,
      "step": 18270
    },
    {
      "epoch": 9.578092243186584,
      "grad_norm": 0.6934157013893127,
      "learning_rate": 3.1142970894782906e-05,
      "loss": 0.3656,
      "num_input_tokens_seen": 11945768,
      "step": 18275
    },
    {
      "epoch": 9.580712788259959,
      "grad_norm": 0.44396671652793884,
      "learning_rate": 3.1131886529811296e-05,
      "loss": 0.4011,
      "num_input_tokens_seen": 11948904,
      "step": 18280
    },
    {
      "epoch": 9.583333333333334,
      "grad_norm": 0.4452945291996002,
      "learning_rate": 3.1120800882118985e-05,
      "loss": 0.3672,
      "num_input_tokens_seen": 11953256,
      "step": 18285
    },
    {
      "epoch": 9.585953878406709,
      "grad_norm": 0.5315406322479248,
      "learning_rate": 3.1109713954024945e-05,
      "loss": 0.395,
      "num_input_tokens_seen": 11956328,
      "step": 18290
    },
    {
      "epoch": 9.588574423480084,
      "grad_norm": 0.4760340452194214,
      "learning_rate": 3.109862574784846e-05,
      "loss": 0.2694,
      "num_input_tokens_seen": 11958728,
      "step": 18295
    },
    {
      "epoch": 9.59119496855346,
      "grad_norm": 0.52742600440979,
      "learning_rate": 3.108753626590905e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 11961736,
      "step": 18300
    },
    {
      "epoch": 9.593815513626835,
      "grad_norm": 0.6285080909729004,
      "learning_rate": 3.1076445510526486e-05,
      "loss": 0.7517,
      "num_input_tokens_seen": 11964456,
      "step": 18305
    },
    {
      "epoch": 9.59643605870021,
      "grad_norm": 0.4988253116607666,
      "learning_rate": 3.106535348402086e-05,
      "loss": 0.477,
      "num_input_tokens_seen": 11967272,
      "step": 18310
    },
    {
      "epoch": 9.599056603773585,
      "grad_norm": 0.3986961543560028,
      "learning_rate": 3.105426018871247e-05,
      "loss": 0.5226,
      "num_input_tokens_seen": 11970760,
      "step": 18315
    },
    {
      "epoch": 9.60167714884696,
      "grad_norm": 0.456479012966156,
      "learning_rate": 3.104316562692193e-05,
      "loss": 0.4413,
      "num_input_tokens_seen": 11974632,
      "step": 18320
    },
    {
      "epoch": 9.604297693920335,
      "grad_norm": 0.46635252237319946,
      "learning_rate": 3.103206980097009e-05,
      "loss": 0.388,
      "num_input_tokens_seen": 11978568,
      "step": 18325
    },
    {
      "epoch": 9.60691823899371,
      "grad_norm": 0.34744998812675476,
      "learning_rate": 3.102097271317805e-05,
      "loss": 0.3536,
      "num_input_tokens_seen": 11981992,
      "step": 18330
    },
    {
      "epoch": 9.609538784067086,
      "grad_norm": 0.3934241235256195,
      "learning_rate": 3.100987436586721e-05,
      "loss": 0.3589,
      "num_input_tokens_seen": 11984616,
      "step": 18335
    },
    {
      "epoch": 9.61215932914046,
      "grad_norm": 0.5057207942008972,
      "learning_rate": 3.099877476135923e-05,
      "loss": 0.4035,
      "num_input_tokens_seen": 11990280,
      "step": 18340
    },
    {
      "epoch": 9.614779874213836,
      "grad_norm": 0.22839325666427612,
      "learning_rate": 3.0987673901976e-05,
      "loss": 0.3806,
      "num_input_tokens_seen": 11993608,
      "step": 18345
    },
    {
      "epoch": 9.617400419287211,
      "grad_norm": 0.3623104691505432,
      "learning_rate": 3.0976571790039697e-05,
      "loss": 0.4264,
      "num_input_tokens_seen": 11996840,
      "step": 18350
    },
    {
      "epoch": 9.620020964360586,
      "grad_norm": 0.43441203236579895,
      "learning_rate": 3.096546842787277e-05,
      "loss": 0.5125,
      "num_input_tokens_seen": 12000392,
      "step": 18355
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 0.36122986674308777,
      "learning_rate": 3.095436381779789e-05,
      "loss": 0.4398,
      "num_input_tokens_seen": 12002888,
      "step": 18360
    },
    {
      "epoch": 9.625262054507338,
      "grad_norm": 0.8246618509292603,
      "learning_rate": 3.0943257962138045e-05,
      "loss": 0.6206,
      "num_input_tokens_seen": 12005160,
      "step": 18365
    },
    {
      "epoch": 9.627882599580714,
      "grad_norm": 0.4820043742656708,
      "learning_rate": 3.093215086321644e-05,
      "loss": 0.432,
      "num_input_tokens_seen": 12007240,
      "step": 18370
    },
    {
      "epoch": 9.630503144654089,
      "grad_norm": 0.2495594173669815,
      "learning_rate": 3.092104252335654e-05,
      "loss": 0.3596,
      "num_input_tokens_seen": 12011720,
      "step": 18375
    },
    {
      "epoch": 9.633123689727464,
      "grad_norm": 0.6707715392112732,
      "learning_rate": 3.09099329448821e-05,
      "loss": 0.5852,
      "num_input_tokens_seen": 12014216,
      "step": 18380
    },
    {
      "epoch": 9.635744234800839,
      "grad_norm": 0.6785979270935059,
      "learning_rate": 3.089882213011711e-05,
      "loss": 0.3878,
      "num_input_tokens_seen": 12018504,
      "step": 18385
    },
    {
      "epoch": 9.638364779874214,
      "grad_norm": 0.7003193497657776,
      "learning_rate": 3.088771008138583e-05,
      "loss": 0.4765,
      "num_input_tokens_seen": 12021128,
      "step": 18390
    },
    {
      "epoch": 9.64098532494759,
      "grad_norm": 0.6352038979530334,
      "learning_rate": 3.087659680101278e-05,
      "loss": 0.4923,
      "num_input_tokens_seen": 12023432,
      "step": 18395
    },
    {
      "epoch": 9.643605870020965,
      "grad_norm": 0.46410319209098816,
      "learning_rate": 3.086548229132271e-05,
      "loss": 0.5042,
      "num_input_tokens_seen": 12025768,
      "step": 18400
    },
    {
      "epoch": 9.64622641509434,
      "grad_norm": 0.40189918875694275,
      "learning_rate": 3.085436655464067e-05,
      "loss": 0.4894,
      "num_input_tokens_seen": 12029064,
      "step": 18405
    },
    {
      "epoch": 9.648846960167715,
      "grad_norm": 0.24609656631946564,
      "learning_rate": 3.084324959329193e-05,
      "loss": 0.3511,
      "num_input_tokens_seen": 12031944,
      "step": 18410
    },
    {
      "epoch": 9.65146750524109,
      "grad_norm": 0.40959280729293823,
      "learning_rate": 3.083213140960204e-05,
      "loss": 0.4705,
      "num_input_tokens_seen": 12034792,
      "step": 18415
    },
    {
      "epoch": 9.654088050314465,
      "grad_norm": 0.2936334013938904,
      "learning_rate": 3.08210120058968e-05,
      "loss": 0.3149,
      "num_input_tokens_seen": 12038632,
      "step": 18420
    },
    {
      "epoch": 9.65670859538784,
      "grad_norm": 0.9622254967689514,
      "learning_rate": 3.080989138450225e-05,
      "loss": 0.6185,
      "num_input_tokens_seen": 12041672,
      "step": 18425
    },
    {
      "epoch": 9.659329140461216,
      "grad_norm": 0.5273007154464722,
      "learning_rate": 3.0798769547744713e-05,
      "loss": 0.4266,
      "num_input_tokens_seen": 12044392,
      "step": 18430
    },
    {
      "epoch": 9.66194968553459,
      "grad_norm": 0.6523393988609314,
      "learning_rate": 3.0787646497950736e-05,
      "loss": 0.434,
      "num_input_tokens_seen": 12048520,
      "step": 18435
    },
    {
      "epoch": 9.664570230607966,
      "grad_norm": 0.3905702233314514,
      "learning_rate": 3.0776522237447146e-05,
      "loss": 0.4484,
      "num_input_tokens_seen": 12051496,
      "step": 18440
    },
    {
      "epoch": 9.667190775681341,
      "grad_norm": 0.6464378237724304,
      "learning_rate": 3.076539676856101e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 12054472,
      "step": 18445
    },
    {
      "epoch": 9.669811320754716,
      "grad_norm": 0.5331742167472839,
      "learning_rate": 3.0754270093619634e-05,
      "loss": 0.5113,
      "num_input_tokens_seen": 12057512,
      "step": 18450
    },
    {
      "epoch": 9.672431865828091,
      "grad_norm": 0.5609127283096313,
      "learning_rate": 3.074314221495061e-05,
      "loss": 0.4143,
      "num_input_tokens_seen": 12062856,
      "step": 18455
    },
    {
      "epoch": 9.675052410901468,
      "grad_norm": 0.5868579745292664,
      "learning_rate": 3.0732013134881756e-05,
      "loss": 0.4985,
      "num_input_tokens_seen": 12065672,
      "step": 18460
    },
    {
      "epoch": 9.677672955974844,
      "grad_norm": 0.7065322399139404,
      "learning_rate": 3.072088285574114e-05,
      "loss": 0.5028,
      "num_input_tokens_seen": 12068328,
      "step": 18465
    },
    {
      "epoch": 9.680293501048219,
      "grad_norm": 0.3514057397842407,
      "learning_rate": 3.070975137985711e-05,
      "loss": 0.3941,
      "num_input_tokens_seen": 12070984,
      "step": 18470
    },
    {
      "epoch": 9.682914046121594,
      "grad_norm": 0.46572616696357727,
      "learning_rate": 3.0698618709558226e-05,
      "loss": 0.4348,
      "num_input_tokens_seen": 12073896,
      "step": 18475
    },
    {
      "epoch": 9.685534591194969,
      "grad_norm": 0.4833914637565613,
      "learning_rate": 3.068748484717332e-05,
      "loss": 0.4824,
      "num_input_tokens_seen": 12077704,
      "step": 18480
    },
    {
      "epoch": 9.688155136268344,
      "grad_norm": 0.8063849806785583,
      "learning_rate": 3.067634979503147e-05,
      "loss": 0.7242,
      "num_input_tokens_seen": 12080584,
      "step": 18485
    },
    {
      "epoch": 9.69077568134172,
      "grad_norm": 0.43081825971603394,
      "learning_rate": 3.0665213555462e-05,
      "loss": 0.3973,
      "num_input_tokens_seen": 12083688,
      "step": 18490
    },
    {
      "epoch": 9.693396226415095,
      "grad_norm": 0.4893452823162079,
      "learning_rate": 3.0654076130794494e-05,
      "loss": 0.5025,
      "num_input_tokens_seen": 12088552,
      "step": 18495
    },
    {
      "epoch": 9.69601677148847,
      "grad_norm": 0.3805047869682312,
      "learning_rate": 3.064293752335876e-05,
      "loss": 0.3728,
      "num_input_tokens_seen": 12091848,
      "step": 18500
    },
    {
      "epoch": 9.698637316561845,
      "grad_norm": 0.5499202013015747,
      "learning_rate": 3.063179773548487e-05,
      "loss": 0.3745,
      "num_input_tokens_seen": 12094568,
      "step": 18505
    },
    {
      "epoch": 9.70125786163522,
      "grad_norm": 0.7122259736061096,
      "learning_rate": 3.0620656769503164e-05,
      "loss": 0.4835,
      "num_input_tokens_seen": 12098312,
      "step": 18510
    },
    {
      "epoch": 9.703878406708595,
      "grad_norm": 0.7917917966842651,
      "learning_rate": 3.060951462774417e-05,
      "loss": 0.4606,
      "num_input_tokens_seen": 12101512,
      "step": 18515
    },
    {
      "epoch": 9.70649895178197,
      "grad_norm": 0.33854079246520996,
      "learning_rate": 3.0598371312538714e-05,
      "loss": 0.6053,
      "num_input_tokens_seen": 12104200,
      "step": 18520
    },
    {
      "epoch": 9.709119496855346,
      "grad_norm": 0.30760273337364197,
      "learning_rate": 3.058722682621785e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 12107112,
      "step": 18525
    },
    {
      "epoch": 9.71174004192872,
      "grad_norm": 0.40984755754470825,
      "learning_rate": 3.0576081171112875e-05,
      "loss": 0.5052,
      "num_input_tokens_seen": 12110376,
      "step": 18530
    },
    {
      "epoch": 9.714360587002096,
      "grad_norm": 0.6587627530097961,
      "learning_rate": 3.056493434955534e-05,
      "loss": 0.5613,
      "num_input_tokens_seen": 12113448,
      "step": 18535
    },
    {
      "epoch": 9.716981132075471,
      "grad_norm": 0.39321181178092957,
      "learning_rate": 3.0553786363877027e-05,
      "loss": 0.5444,
      "num_input_tokens_seen": 12117576,
      "step": 18540
    },
    {
      "epoch": 9.719601677148846,
      "grad_norm": 0.3444618582725525,
      "learning_rate": 3.054263721640996e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 12121416,
      "step": 18545
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 0.376402884721756,
      "learning_rate": 3.053148690948642e-05,
      "loss": 0.4415,
      "num_input_tokens_seen": 12124072,
      "step": 18550
    },
    {
      "epoch": 9.724842767295598,
      "grad_norm": 0.4678923785686493,
      "learning_rate": 3.0520335445438935e-05,
      "loss": 0.4201,
      "num_input_tokens_seen": 12127272,
      "step": 18555
    },
    {
      "epoch": 9.727463312368974,
      "grad_norm": 0.3874358534812927,
      "learning_rate": 3.0509182826600246e-05,
      "loss": 0.3856,
      "num_input_tokens_seen": 12131240,
      "step": 18560
    },
    {
      "epoch": 9.730083857442349,
      "grad_norm": 0.31471481919288635,
      "learning_rate": 3.049802905530337e-05,
      "loss": 0.355,
      "num_input_tokens_seen": 12134664,
      "step": 18565
    },
    {
      "epoch": 9.732704402515724,
      "grad_norm": 0.6361053586006165,
      "learning_rate": 3.0486874133881533e-05,
      "loss": 0.4646,
      "num_input_tokens_seen": 12137352,
      "step": 18570
    },
    {
      "epoch": 9.735324947589099,
      "grad_norm": 0.49771571159362793,
      "learning_rate": 3.047571806466823e-05,
      "loss": 0.5544,
      "num_input_tokens_seen": 12140776,
      "step": 18575
    },
    {
      "epoch": 9.737945492662474,
      "grad_norm": 0.6616657376289368,
      "learning_rate": 3.0464560849997177e-05,
      "loss": 0.588,
      "num_input_tokens_seen": 12143112,
      "step": 18580
    },
    {
      "epoch": 9.74056603773585,
      "grad_norm": 1.104009747505188,
      "learning_rate": 3.0453402492202325e-05,
      "loss": 0.538,
      "num_input_tokens_seen": 12145992,
      "step": 18585
    },
    {
      "epoch": 9.743186582809225,
      "grad_norm": 0.365575909614563,
      "learning_rate": 3.0442242993617904e-05,
      "loss": 0.5286,
      "num_input_tokens_seen": 12149672,
      "step": 18590
    },
    {
      "epoch": 9.7458071278826,
      "grad_norm": 0.46504777669906616,
      "learning_rate": 3.0431082356578334e-05,
      "loss": 0.4178,
      "num_input_tokens_seen": 12152872,
      "step": 18595
    },
    {
      "epoch": 9.748427672955975,
      "grad_norm": 0.6082614660263062,
      "learning_rate": 3.0419920583418294e-05,
      "loss": 0.3511,
      "num_input_tokens_seen": 12155656,
      "step": 18600
    },
    {
      "epoch": 9.75104821802935,
      "grad_norm": 0.32616373896598816,
      "learning_rate": 3.0408757676472694e-05,
      "loss": 0.6854,
      "num_input_tokens_seen": 12159528,
      "step": 18605
    },
    {
      "epoch": 9.753668763102725,
      "grad_norm": 0.7126193642616272,
      "learning_rate": 3.0397593638076705e-05,
      "loss": 0.5107,
      "num_input_tokens_seen": 12163208,
      "step": 18610
    },
    {
      "epoch": 9.7562893081761,
      "grad_norm": 0.46937671303749084,
      "learning_rate": 3.0386428470565702e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 12166504,
      "step": 18615
    },
    {
      "epoch": 9.758909853249476,
      "grad_norm": 0.4369599521160126,
      "learning_rate": 3.0375262176275308e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 12169928,
      "step": 18620
    },
    {
      "epoch": 9.76153039832285,
      "grad_norm": 0.33724337816238403,
      "learning_rate": 3.03640947575414e-05,
      "loss": 0.4443,
      "num_input_tokens_seen": 12176232,
      "step": 18625
    },
    {
      "epoch": 9.764150943396226,
      "grad_norm": 0.5407045483589172,
      "learning_rate": 3.0352926216700067e-05,
      "loss": 0.5397,
      "num_input_tokens_seen": 12179592,
      "step": 18630
    },
    {
      "epoch": 9.766771488469601,
      "grad_norm": 0.47815948724746704,
      "learning_rate": 3.034175655608763e-05,
      "loss": 0.4836,
      "num_input_tokens_seen": 12183240,
      "step": 18635
    },
    {
      "epoch": 9.769392033542976,
      "grad_norm": 0.3858913481235504,
      "learning_rate": 3.0330585778040677e-05,
      "loss": 0.4166,
      "num_input_tokens_seen": 12186280,
      "step": 18640
    },
    {
      "epoch": 9.772012578616351,
      "grad_norm": 0.37062153220176697,
      "learning_rate": 3.031941388489598e-05,
      "loss": 0.4952,
      "num_input_tokens_seen": 12189160,
      "step": 18645
    },
    {
      "epoch": 9.774633123689728,
      "grad_norm": 0.3418489992618561,
      "learning_rate": 3.0308240878990584e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 12191784,
      "step": 18650
    },
    {
      "epoch": 9.777253668763104,
      "grad_norm": 0.3089541792869568,
      "learning_rate": 3.0297066762661773e-05,
      "loss": 0.5385,
      "num_input_tokens_seen": 12195080,
      "step": 18655
    },
    {
      "epoch": 9.779874213836479,
      "grad_norm": 0.8825511932373047,
      "learning_rate": 3.0285891538247015e-05,
      "loss": 0.4439,
      "num_input_tokens_seen": 12197736,
      "step": 18660
    },
    {
      "epoch": 9.782494758909854,
      "grad_norm": 0.7357693314552307,
      "learning_rate": 3.0274715208084053e-05,
      "loss": 0.5237,
      "num_input_tokens_seen": 12200392,
      "step": 18665
    },
    {
      "epoch": 9.785115303983229,
      "grad_norm": 0.23574233055114746,
      "learning_rate": 3.0263537774510847e-05,
      "loss": 0.3319,
      "num_input_tokens_seen": 12203816,
      "step": 18670
    },
    {
      "epoch": 9.787735849056604,
      "grad_norm": 0.37034329771995544,
      "learning_rate": 3.0252359239865585e-05,
      "loss": 0.4243,
      "num_input_tokens_seen": 12208264,
      "step": 18675
    },
    {
      "epoch": 9.79035639412998,
      "grad_norm": 0.47159403562545776,
      "learning_rate": 3.0241179606486704e-05,
      "loss": 0.4206,
      "num_input_tokens_seen": 12211048,
      "step": 18680
    },
    {
      "epoch": 9.792976939203355,
      "grad_norm": 0.270731121301651,
      "learning_rate": 3.0229998876712827e-05,
      "loss": 0.5638,
      "num_input_tokens_seen": 12213928,
      "step": 18685
    },
    {
      "epoch": 9.79559748427673,
      "grad_norm": 0.4890733063220978,
      "learning_rate": 3.0218817052882858e-05,
      "loss": 0.5547,
      "num_input_tokens_seen": 12217032,
      "step": 18690
    },
    {
      "epoch": 9.798218029350105,
      "grad_norm": 0.9872231483459473,
      "learning_rate": 3.020763413733591e-05,
      "loss": 0.4664,
      "num_input_tokens_seen": 12220456,
      "step": 18695
    },
    {
      "epoch": 9.80083857442348,
      "grad_norm": 0.3476005494594574,
      "learning_rate": 3.0196450132411302e-05,
      "loss": 0.5526,
      "num_input_tokens_seen": 12223336,
      "step": 18700
    },
    {
      "epoch": 9.803459119496855,
      "grad_norm": 0.5852622389793396,
      "learning_rate": 3.0185265040448616e-05,
      "loss": 0.5298,
      "num_input_tokens_seen": 12226632,
      "step": 18705
    },
    {
      "epoch": 9.80607966457023,
      "grad_norm": 0.48544180393218994,
      "learning_rate": 3.017407886378764e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 12229512,
      "step": 18710
    },
    {
      "epoch": 9.808700209643606,
      "grad_norm": 0.5260866284370422,
      "learning_rate": 3.016289160476839e-05,
      "loss": 0.3522,
      "num_input_tokens_seen": 12232808,
      "step": 18715
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 0.8328986167907715,
      "learning_rate": 3.0151703265731125e-05,
      "loss": 0.3536,
      "num_input_tokens_seen": 12237288,
      "step": 18720
    },
    {
      "epoch": 9.813941299790356,
      "grad_norm": 0.4779738485813141,
      "learning_rate": 3.0140513849016305e-05,
      "loss": 0.321,
      "num_input_tokens_seen": 12241096,
      "step": 18725
    },
    {
      "epoch": 9.816561844863731,
      "grad_norm": 0.7860069274902344,
      "learning_rate": 3.0129323356964627e-05,
      "loss": 0.5006,
      "num_input_tokens_seen": 12244552,
      "step": 18730
    },
    {
      "epoch": 9.819182389937106,
      "grad_norm": 0.7116873264312744,
      "learning_rate": 3.0118131791917032e-05,
      "loss": 0.5745,
      "num_input_tokens_seen": 12247016,
      "step": 18735
    },
    {
      "epoch": 9.821802935010481,
      "grad_norm": 0.29764458537101746,
      "learning_rate": 3.010693915621466e-05,
      "loss": 0.4183,
      "num_input_tokens_seen": 12250184,
      "step": 18740
    },
    {
      "epoch": 9.824423480083858,
      "grad_norm": 0.3484274744987488,
      "learning_rate": 3.009574545219887e-05,
      "loss": 0.4353,
      "num_input_tokens_seen": 12253064,
      "step": 18745
    },
    {
      "epoch": 9.827044025157234,
      "grad_norm": 0.32364749908447266,
      "learning_rate": 3.0084550682211272e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 12256328,
      "step": 18750
    },
    {
      "epoch": 9.829664570230609,
      "grad_norm": 0.6897169947624207,
      "learning_rate": 3.0073354848593677e-05,
      "loss": 0.4584,
      "num_input_tokens_seen": 12258888,
      "step": 18755
    },
    {
      "epoch": 9.832285115303984,
      "grad_norm": 0.3072650730609894,
      "learning_rate": 3.0062157953688125e-05,
      "loss": 0.4104,
      "num_input_tokens_seen": 12262536,
      "step": 18760
    },
    {
      "epoch": 9.834905660377359,
      "grad_norm": 0.33233192563056946,
      "learning_rate": 3.0050959999836887e-05,
      "loss": 0.5087,
      "num_input_tokens_seen": 12266760,
      "step": 18765
    },
    {
      "epoch": 9.837526205450734,
      "grad_norm": 0.5852479338645935,
      "learning_rate": 3.003976098938244e-05,
      "loss": 0.4595,
      "num_input_tokens_seen": 12269576,
      "step": 18770
    },
    {
      "epoch": 9.84014675052411,
      "grad_norm": 0.45692935585975647,
      "learning_rate": 3.0028560924667487e-05,
      "loss": 0.433,
      "num_input_tokens_seen": 12273544,
      "step": 18775
    },
    {
      "epoch": 9.842767295597485,
      "grad_norm": 0.3246449828147888,
      "learning_rate": 3.001735980803496e-05,
      "loss": 0.6278,
      "num_input_tokens_seen": 12276456,
      "step": 18780
    },
    {
      "epoch": 9.84538784067086,
      "grad_norm": 0.7366659641265869,
      "learning_rate": 3.0006157641828004e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 12279368,
      "step": 18785
    },
    {
      "epoch": 9.848008385744235,
      "grad_norm": 0.40402883291244507,
      "learning_rate": 2.9994954428389977e-05,
      "loss": 0.5417,
      "num_input_tokens_seen": 12282280,
      "step": 18790
    },
    {
      "epoch": 9.85062893081761,
      "grad_norm": 0.7807847261428833,
      "learning_rate": 2.9983750170064472e-05,
      "loss": 0.5781,
      "num_input_tokens_seen": 12286600,
      "step": 18795
    },
    {
      "epoch": 9.853249475890985,
      "grad_norm": 0.29550901055336,
      "learning_rate": 2.9972544869195292e-05,
      "loss": 0.2935,
      "num_input_tokens_seen": 12289992,
      "step": 18800
    },
    {
      "epoch": 9.85587002096436,
      "grad_norm": 0.30600878596305847,
      "learning_rate": 2.9961338528126447e-05,
      "loss": 0.5386,
      "num_input_tokens_seen": 12294248,
      "step": 18805
    },
    {
      "epoch": 9.858490566037736,
      "grad_norm": 0.38341790437698364,
      "learning_rate": 2.9950131149202192e-05,
      "loss": 0.3091,
      "num_input_tokens_seen": 12299560,
      "step": 18810
    },
    {
      "epoch": 9.86111111111111,
      "grad_norm": 0.7851564288139343,
      "learning_rate": 2.9938922734766973e-05,
      "loss": 0.4227,
      "num_input_tokens_seen": 12302632,
      "step": 18815
    },
    {
      "epoch": 9.863731656184486,
      "grad_norm": 0.5464969277381897,
      "learning_rate": 2.992771328716546e-05,
      "loss": 0.3798,
      "num_input_tokens_seen": 12305576,
      "step": 18820
    },
    {
      "epoch": 9.866352201257861,
      "grad_norm": 0.7184147834777832,
      "learning_rate": 2.991650280874255e-05,
      "loss": 0.4835,
      "num_input_tokens_seen": 12308456,
      "step": 18825
    },
    {
      "epoch": 9.868972746331236,
      "grad_norm": 0.3784220516681671,
      "learning_rate": 2.990529130184334e-05,
      "loss": 0.3808,
      "num_input_tokens_seen": 12312040,
      "step": 18830
    },
    {
      "epoch": 9.871593291404611,
      "grad_norm": 0.6076700687408447,
      "learning_rate": 2.989407876881316e-05,
      "loss": 0.6431,
      "num_input_tokens_seen": 12315528,
      "step": 18835
    },
    {
      "epoch": 9.874213836477988,
      "grad_norm": 0.3291487395763397,
      "learning_rate": 2.988286521199753e-05,
      "loss": 0.438,
      "num_input_tokens_seen": 12318632,
      "step": 18840
    },
    {
      "epoch": 9.876834381551364,
      "grad_norm": 0.3318800926208496,
      "learning_rate": 2.9871650633742198e-05,
      "loss": 0.6707,
      "num_input_tokens_seen": 12321480,
      "step": 18845
    },
    {
      "epoch": 9.879454926624739,
      "grad_norm": 0.46452975273132324,
      "learning_rate": 2.9860435036393142e-05,
      "loss": 0.4331,
      "num_input_tokens_seen": 12325288,
      "step": 18850
    },
    {
      "epoch": 9.882075471698114,
      "grad_norm": 0.4780445098876953,
      "learning_rate": 2.9849218422296527e-05,
      "loss": 0.5318,
      "num_input_tokens_seen": 12328040,
      "step": 18855
    },
    {
      "epoch": 9.884696016771489,
      "grad_norm": 0.33492061495780945,
      "learning_rate": 2.983800079379873e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 12331848,
      "step": 18860
    },
    {
      "epoch": 9.887316561844864,
      "grad_norm": 0.3795499801635742,
      "learning_rate": 2.982678215324638e-05,
      "loss": 0.5046,
      "num_input_tokens_seen": 12334664,
      "step": 18865
    },
    {
      "epoch": 9.88993710691824,
      "grad_norm": 0.3966444730758667,
      "learning_rate": 2.9815562502986256e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 12338152,
      "step": 18870
    },
    {
      "epoch": 9.892557651991615,
      "grad_norm": 0.41405507922172546,
      "learning_rate": 2.9804341845365398e-05,
      "loss": 0.3956,
      "num_input_tokens_seen": 12342568,
      "step": 18875
    },
    {
      "epoch": 9.89517819706499,
      "grad_norm": 0.558620274066925,
      "learning_rate": 2.9793120182731037e-05,
      "loss": 0.5515,
      "num_input_tokens_seen": 12346088,
      "step": 18880
    },
    {
      "epoch": 9.897798742138365,
      "grad_norm": 0.3353066146373749,
      "learning_rate": 2.9781897517430617e-05,
      "loss": 0.4346,
      "num_input_tokens_seen": 12349384,
      "step": 18885
    },
    {
      "epoch": 9.90041928721174,
      "grad_norm": 0.38730689883232117,
      "learning_rate": 2.977067385181179e-05,
      "loss": 0.4643,
      "num_input_tokens_seen": 12351912,
      "step": 18890
    },
    {
      "epoch": 9.903039832285115,
      "grad_norm": 0.6210649609565735,
      "learning_rate": 2.9759449188222426e-05,
      "loss": 0.3722,
      "num_input_tokens_seen": 12355592,
      "step": 18895
    },
    {
      "epoch": 9.90566037735849,
      "grad_norm": 0.5155217051506042,
      "learning_rate": 2.9748223529010582e-05,
      "loss": 0.4883,
      "num_input_tokens_seen": 12358856,
      "step": 18900
    },
    {
      "epoch": 9.908280922431866,
      "grad_norm": 0.7275372743606567,
      "learning_rate": 2.9736996876524546e-05,
      "loss": 0.483,
      "num_input_tokens_seen": 12361800,
      "step": 18905
    },
    {
      "epoch": 9.91090146750524,
      "grad_norm": 0.3740609884262085,
      "learning_rate": 2.9725769233112816e-05,
      "loss": 0.4625,
      "num_input_tokens_seen": 12364936,
      "step": 18910
    },
    {
      "epoch": 9.913522012578616,
      "grad_norm": 0.4639403223991394,
      "learning_rate": 2.9714540601124074e-05,
      "loss": 0.4076,
      "num_input_tokens_seen": 12367592,
      "step": 18915
    },
    {
      "epoch": 9.916142557651991,
      "grad_norm": 0.40917810797691345,
      "learning_rate": 2.9703310982907218e-05,
      "loss": 0.3963,
      "num_input_tokens_seen": 12371176,
      "step": 18920
    },
    {
      "epoch": 9.918763102725366,
      "grad_norm": 0.25379565358161926,
      "learning_rate": 2.9692080380811365e-05,
      "loss": 0.4121,
      "num_input_tokens_seen": 12375464,
      "step": 18925
    },
    {
      "epoch": 9.921383647798741,
      "grad_norm": 0.5929741263389587,
      "learning_rate": 2.9680848797185828e-05,
      "loss": 0.4647,
      "num_input_tokens_seen": 12378248,
      "step": 18930
    },
    {
      "epoch": 9.924004192872118,
      "grad_norm": 0.5115299224853516,
      "learning_rate": 2.966961623438012e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 12381288,
      "step": 18935
    },
    {
      "epoch": 9.926624737945493,
      "grad_norm": 0.33694177865982056,
      "learning_rate": 2.9658382694743973e-05,
      "loss": 0.3991,
      "num_input_tokens_seen": 12385192,
      "step": 18940
    },
    {
      "epoch": 9.929245283018869,
      "grad_norm": 0.34820234775543213,
      "learning_rate": 2.9647148180627317e-05,
      "loss": 0.317,
      "num_input_tokens_seen": 12388392,
      "step": 18945
    },
    {
      "epoch": 9.931865828092244,
      "grad_norm": 0.35187166929244995,
      "learning_rate": 2.9635912694380275e-05,
      "loss": 0.3296,
      "num_input_tokens_seen": 12391112,
      "step": 18950
    },
    {
      "epoch": 9.934486373165619,
      "grad_norm": 0.5645958185195923,
      "learning_rate": 2.962467623835319e-05,
      "loss": 0.3967,
      "num_input_tokens_seen": 12394088,
      "step": 18955
    },
    {
      "epoch": 9.937106918238994,
      "grad_norm": 0.47940051555633545,
      "learning_rate": 2.961343881489659e-05,
      "loss": 0.7081,
      "num_input_tokens_seen": 12396840,
      "step": 18960
    },
    {
      "epoch": 9.93972746331237,
      "grad_norm": 0.631404459476471,
      "learning_rate": 2.9602200426361226e-05,
      "loss": 0.4266,
      "num_input_tokens_seen": 12400776,
      "step": 18965
    },
    {
      "epoch": 9.942348008385745,
      "grad_norm": 0.32388800382614136,
      "learning_rate": 2.9590961075098038e-05,
      "loss": 0.4099,
      "num_input_tokens_seen": 12403848,
      "step": 18970
    },
    {
      "epoch": 9.94496855345912,
      "grad_norm": 0.6202534437179565,
      "learning_rate": 2.957972076345817e-05,
      "loss": 0.4098,
      "num_input_tokens_seen": 12407016,
      "step": 18975
    },
    {
      "epoch": 9.947589098532495,
      "grad_norm": 0.3129224181175232,
      "learning_rate": 2.9568479493792967e-05,
      "loss": 0.4153,
      "num_input_tokens_seen": 12410312,
      "step": 18980
    },
    {
      "epoch": 9.95020964360587,
      "grad_norm": 0.34671634435653687,
      "learning_rate": 2.955723726845397e-05,
      "loss": 0.4371,
      "num_input_tokens_seen": 12413672,
      "step": 18985
    },
    {
      "epoch": 9.952830188679245,
      "grad_norm": 0.4527193605899811,
      "learning_rate": 2.9545994089792928e-05,
      "loss": 0.3866,
      "num_input_tokens_seen": 12416936,
      "step": 18990
    },
    {
      "epoch": 9.95545073375262,
      "grad_norm": 0.3280191123485565,
      "learning_rate": 2.9534749960161795e-05,
      "loss": 0.4646,
      "num_input_tokens_seen": 12419368,
      "step": 18995
    },
    {
      "epoch": 9.958071278825996,
      "grad_norm": 0.4071759581565857,
      "learning_rate": 2.9523504881912694e-05,
      "loss": 0.3479,
      "num_input_tokens_seen": 12422152,
      "step": 19000
    },
    {
      "epoch": 9.96069182389937,
      "grad_norm": 1.2565422058105469,
      "learning_rate": 2.9512258857397978e-05,
      "loss": 0.3277,
      "num_input_tokens_seen": 12424840,
      "step": 19005
    },
    {
      "epoch": 9.963312368972746,
      "grad_norm": 0.22118492424488068,
      "learning_rate": 2.9501011888970203e-05,
      "loss": 0.4142,
      "num_input_tokens_seen": 12429576,
      "step": 19010
    },
    {
      "epoch": 9.965932914046121,
      "grad_norm": 0.5265933871269226,
      "learning_rate": 2.9489763978982075e-05,
      "loss": 0.5718,
      "num_input_tokens_seen": 12433096,
      "step": 19015
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.38246652483940125,
      "learning_rate": 2.9478515129786556e-05,
      "loss": 0.3398,
      "num_input_tokens_seen": 12436744,
      "step": 19020
    },
    {
      "epoch": 9.971174004192871,
      "grad_norm": 0.9781017303466797,
      "learning_rate": 2.946726534373676e-05,
      "loss": 0.504,
      "num_input_tokens_seen": 12439464,
      "step": 19025
    },
    {
      "epoch": 9.973794549266248,
      "grad_norm": 0.3328745663166046,
      "learning_rate": 2.945601462318601e-05,
      "loss": 0.456,
      "num_input_tokens_seen": 12442920,
      "step": 19030
    },
    {
      "epoch": 9.976415094339622,
      "grad_norm": 0.9270390868186951,
      "learning_rate": 2.944476297048786e-05,
      "loss": 0.4538,
      "num_input_tokens_seen": 12446920,
      "step": 19035
    },
    {
      "epoch": 9.979035639412999,
      "grad_norm": 0.6628504991531372,
      "learning_rate": 2.9433510387995987e-05,
      "loss": 0.5232,
      "num_input_tokens_seen": 12449768,
      "step": 19040
    },
    {
      "epoch": 9.981656184486374,
      "grad_norm": 0.8887121081352234,
      "learning_rate": 2.9422256878064325e-05,
      "loss": 0.4552,
      "num_input_tokens_seen": 12452712,
      "step": 19045
    },
    {
      "epoch": 9.984276729559749,
      "grad_norm": 0.43586987257003784,
      "learning_rate": 2.9411002443046982e-05,
      "loss": 0.5105,
      "num_input_tokens_seen": 12456008,
      "step": 19050
    },
    {
      "epoch": 9.986897274633124,
      "grad_norm": 0.8631776571273804,
      "learning_rate": 2.9399747085298245e-05,
      "loss": 0.564,
      "num_input_tokens_seen": 12458792,
      "step": 19055
    },
    {
      "epoch": 9.9895178197065,
      "grad_norm": 0.3512836694717407,
      "learning_rate": 2.9388490807172614e-05,
      "loss": 0.3225,
      "num_input_tokens_seen": 12461896,
      "step": 19060
    },
    {
      "epoch": 9.992138364779874,
      "grad_norm": 0.33337166905403137,
      "learning_rate": 2.9377233611024775e-05,
      "loss": 0.452,
      "num_input_tokens_seen": 12464616,
      "step": 19065
    },
    {
      "epoch": 9.99475890985325,
      "grad_norm": 0.4279584288597107,
      "learning_rate": 2.9365975499209596e-05,
      "loss": 0.5596,
      "num_input_tokens_seen": 12467880,
      "step": 19070
    },
    {
      "epoch": 9.997379454926625,
      "grad_norm": 0.522316038608551,
      "learning_rate": 2.9354716474082154e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 12470760,
      "step": 19075
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.41945382952690125,
      "learning_rate": 2.9343456537997714e-05,
      "loss": 0.444,
      "num_input_tokens_seen": 12474576,
      "step": 19080
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.4780687987804413,
      "eval_runtime": 9.7049,
      "eval_samples_per_second": 87.379,
      "eval_steps_per_second": 21.845,
      "num_input_tokens_seen": 12474576,
      "step": 19080
    },
    {
      "epoch": 10.002620545073375,
      "grad_norm": 0.8173631429672241,
      "learning_rate": 2.9332195693311714e-05,
      "loss": 0.302,
      "num_input_tokens_seen": 12477296,
      "step": 19085
    },
    {
      "epoch": 10.00524109014675,
      "grad_norm": 0.34604212641716003,
      "learning_rate": 2.9320933942379803e-05,
      "loss": 0.5151,
      "num_input_tokens_seen": 12482608,
      "step": 19090
    },
    {
      "epoch": 10.007861635220126,
      "grad_norm": 0.30886727571487427,
      "learning_rate": 2.9309671287557805e-05,
      "loss": 0.4029,
      "num_input_tokens_seen": 12485328,
      "step": 19095
    },
    {
      "epoch": 10.0104821802935,
      "grad_norm": 0.4039437174797058,
      "learning_rate": 2.9298407731201737e-05,
      "loss": 0.4538,
      "num_input_tokens_seen": 12488944,
      "step": 19100
    },
    {
      "epoch": 10.013102725366876,
      "grad_norm": 0.3863140642642975,
      "learning_rate": 2.928714327566781e-05,
      "loss": 0.4849,
      "num_input_tokens_seen": 12491952,
      "step": 19105
    },
    {
      "epoch": 10.015723270440251,
      "grad_norm": 0.4848407506942749,
      "learning_rate": 2.9275877923312427e-05,
      "loss": 0.3917,
      "num_input_tokens_seen": 12495600,
      "step": 19110
    },
    {
      "epoch": 10.018343815513626,
      "grad_norm": 0.7015286684036255,
      "learning_rate": 2.9264611676492165e-05,
      "loss": 0.3308,
      "num_input_tokens_seen": 12499152,
      "step": 19115
    },
    {
      "epoch": 10.020964360587001,
      "grad_norm": 0.6793297529220581,
      "learning_rate": 2.9253344537563782e-05,
      "loss": 0.5031,
      "num_input_tokens_seen": 12502032,
      "step": 19120
    },
    {
      "epoch": 10.023584905660377,
      "grad_norm": 0.569173276424408,
      "learning_rate": 2.924207650888426e-05,
      "loss": 0.516,
      "num_input_tokens_seen": 12505264,
      "step": 19125
    },
    {
      "epoch": 10.026205450733753,
      "grad_norm": 0.33076363801956177,
      "learning_rate": 2.9230807592810723e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 12510384,
      "step": 19130
    },
    {
      "epoch": 10.028825995807129,
      "grad_norm": 0.4169284701347351,
      "learning_rate": 2.92195377917005e-05,
      "loss": 0.7011,
      "num_input_tokens_seen": 12513680,
      "step": 19135
    },
    {
      "epoch": 10.031446540880504,
      "grad_norm": 0.5522189736366272,
      "learning_rate": 2.920826710791112e-05,
      "loss": 0.5773,
      "num_input_tokens_seen": 12517328,
      "step": 19140
    },
    {
      "epoch": 10.034067085953879,
      "grad_norm": 0.48940423130989075,
      "learning_rate": 2.919699554380026e-05,
      "loss": 0.4291,
      "num_input_tokens_seen": 12521872,
      "step": 19145
    },
    {
      "epoch": 10.036687631027254,
      "grad_norm": 0.6093086004257202,
      "learning_rate": 2.9185723101725825e-05,
      "loss": 0.453,
      "num_input_tokens_seen": 12524752,
      "step": 19150
    },
    {
      "epoch": 10.03930817610063,
      "grad_norm": 0.8527776598930359,
      "learning_rate": 2.9174449784045867e-05,
      "loss": 0.5202,
      "num_input_tokens_seen": 12527216,
      "step": 19155
    },
    {
      "epoch": 10.041928721174004,
      "grad_norm": 0.6053761839866638,
      "learning_rate": 2.916317559311863e-05,
      "loss": 0.4572,
      "num_input_tokens_seen": 12530416,
      "step": 19160
    },
    {
      "epoch": 10.04454926624738,
      "grad_norm": 0.7458449006080627,
      "learning_rate": 2.9151900531302566e-05,
      "loss": 0.4469,
      "num_input_tokens_seen": 12534064,
      "step": 19165
    },
    {
      "epoch": 10.047169811320755,
      "grad_norm": 0.3904799818992615,
      "learning_rate": 2.9140624600956272e-05,
      "loss": 0.3804,
      "num_input_tokens_seen": 12536816,
      "step": 19170
    },
    {
      "epoch": 10.04979035639413,
      "grad_norm": 0.45865318179130554,
      "learning_rate": 2.912934780443855e-05,
      "loss": 0.3379,
      "num_input_tokens_seen": 12540048,
      "step": 19175
    },
    {
      "epoch": 10.052410901467505,
      "grad_norm": 0.32859015464782715,
      "learning_rate": 2.9118070144108388e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 12542992,
      "step": 19180
    },
    {
      "epoch": 10.05503144654088,
      "grad_norm": 0.3541281521320343,
      "learning_rate": 2.9106791622324924e-05,
      "loss": 0.5559,
      "num_input_tokens_seen": 12547504,
      "step": 19185
    },
    {
      "epoch": 10.057651991614255,
      "grad_norm": 0.5729717016220093,
      "learning_rate": 2.909551224144751e-05,
      "loss": 0.386,
      "num_input_tokens_seen": 12549904,
      "step": 19190
    },
    {
      "epoch": 10.06027253668763,
      "grad_norm": 0.4189419448375702,
      "learning_rate": 2.9084232003835653e-05,
      "loss": 0.3547,
      "num_input_tokens_seen": 12552912,
      "step": 19195
    },
    {
      "epoch": 10.062893081761006,
      "grad_norm": 0.5575315952301025,
      "learning_rate": 2.9072950911849058e-05,
      "loss": 0.3572,
      "num_input_tokens_seen": 12556560,
      "step": 19200
    },
    {
      "epoch": 10.065513626834381,
      "grad_norm": 0.23741665482521057,
      "learning_rate": 2.90616689678476e-05,
      "loss": 0.4957,
      "num_input_tokens_seen": 12559600,
      "step": 19205
    },
    {
      "epoch": 10.068134171907756,
      "grad_norm": 0.22497984766960144,
      "learning_rate": 2.9050386174191337e-05,
      "loss": 0.3774,
      "num_input_tokens_seen": 12563056,
      "step": 19210
    },
    {
      "epoch": 10.070754716981131,
      "grad_norm": 0.5491334795951843,
      "learning_rate": 2.903910253324048e-05,
      "loss": 0.3853,
      "num_input_tokens_seen": 12567088,
      "step": 19215
    },
    {
      "epoch": 10.073375262054507,
      "grad_norm": 0.39129361510276794,
      "learning_rate": 2.9027818047355477e-05,
      "loss": 0.3163,
      "num_input_tokens_seen": 12571024,
      "step": 19220
    },
    {
      "epoch": 10.075995807127883,
      "grad_norm": 0.46219828724861145,
      "learning_rate": 2.9016532718896873e-05,
      "loss": 0.4332,
      "num_input_tokens_seen": 12573392,
      "step": 19225
    },
    {
      "epoch": 10.078616352201259,
      "grad_norm": 0.4912542402744293,
      "learning_rate": 2.900524655022545e-05,
      "loss": 0.4245,
      "num_input_tokens_seen": 12576752,
      "step": 19230
    },
    {
      "epoch": 10.081236897274634,
      "grad_norm": 0.2561753988265991,
      "learning_rate": 2.899395954370215e-05,
      "loss": 0.4,
      "num_input_tokens_seen": 12581648,
      "step": 19235
    },
    {
      "epoch": 10.083857442348009,
      "grad_norm": 0.27532240748405457,
      "learning_rate": 2.898267170168807e-05,
      "loss": 0.5013,
      "num_input_tokens_seen": 12585744,
      "step": 19240
    },
    {
      "epoch": 10.086477987421384,
      "grad_norm": 0.4101257622241974,
      "learning_rate": 2.8971383026544507e-05,
      "loss": 0.3288,
      "num_input_tokens_seen": 12591216,
      "step": 19245
    },
    {
      "epoch": 10.08909853249476,
      "grad_norm": 0.5164194107055664,
      "learning_rate": 2.896009352063293e-05,
      "loss": 0.6112,
      "num_input_tokens_seen": 12595152,
      "step": 19250
    },
    {
      "epoch": 10.091719077568134,
      "grad_norm": 0.6256662011146545,
      "learning_rate": 2.894880318631496e-05,
      "loss": 0.3786,
      "num_input_tokens_seen": 12598096,
      "step": 19255
    },
    {
      "epoch": 10.09433962264151,
      "grad_norm": 0.4921885132789612,
      "learning_rate": 2.8937512025952417e-05,
      "loss": 0.4971,
      "num_input_tokens_seen": 12601424,
      "step": 19260
    },
    {
      "epoch": 10.096960167714885,
      "grad_norm": 0.35346436500549316,
      "learning_rate": 2.8926220041907283e-05,
      "loss": 0.3677,
      "num_input_tokens_seen": 12605040,
      "step": 19265
    },
    {
      "epoch": 10.09958071278826,
      "grad_norm": 0.5053911805152893,
      "learning_rate": 2.8914927236541706e-05,
      "loss": 0.4482,
      "num_input_tokens_seen": 12607984,
      "step": 19270
    },
    {
      "epoch": 10.102201257861635,
      "grad_norm": 0.3948405683040619,
      "learning_rate": 2.8903633612218012e-05,
      "loss": 0.393,
      "num_input_tokens_seen": 12611728,
      "step": 19275
    },
    {
      "epoch": 10.10482180293501,
      "grad_norm": 0.47959861159324646,
      "learning_rate": 2.8892339171298703e-05,
      "loss": 0.4858,
      "num_input_tokens_seen": 12614960,
      "step": 19280
    },
    {
      "epoch": 10.107442348008385,
      "grad_norm": 1.4349747896194458,
      "learning_rate": 2.888104391614645e-05,
      "loss": 0.5828,
      "num_input_tokens_seen": 12617872,
      "step": 19285
    },
    {
      "epoch": 10.11006289308176,
      "grad_norm": 0.7464232444763184,
      "learning_rate": 2.886974784912408e-05,
      "loss": 0.6044,
      "num_input_tokens_seen": 12620528,
      "step": 19290
    },
    {
      "epoch": 10.112683438155136,
      "grad_norm": 0.28704413771629333,
      "learning_rate": 2.8858450972594612e-05,
      "loss": 0.3635,
      "num_input_tokens_seen": 12623152,
      "step": 19295
    },
    {
      "epoch": 10.115303983228511,
      "grad_norm": 0.27474990487098694,
      "learning_rate": 2.8847153288921226e-05,
      "loss": 0.3544,
      "num_input_tokens_seen": 12626832,
      "step": 19300
    },
    {
      "epoch": 10.117924528301886,
      "grad_norm": 0.6494747996330261,
      "learning_rate": 2.883585480046725e-05,
      "loss": 0.6135,
      "num_input_tokens_seen": 12629456,
      "step": 19305
    },
    {
      "epoch": 10.120545073375261,
      "grad_norm": 0.4967104196548462,
      "learning_rate": 2.8824555509596223e-05,
      "loss": 0.4363,
      "num_input_tokens_seen": 12632912,
      "step": 19310
    },
    {
      "epoch": 10.123165618448636,
      "grad_norm": 0.49750253558158875,
      "learning_rate": 2.8813255418671815e-05,
      "loss": 0.5997,
      "num_input_tokens_seen": 12636272,
      "step": 19315
    },
    {
      "epoch": 10.125786163522013,
      "grad_norm": 0.3495972454547882,
      "learning_rate": 2.8801954530057874e-05,
      "loss": 0.5364,
      "num_input_tokens_seen": 12639152,
      "step": 19320
    },
    {
      "epoch": 10.128406708595389,
      "grad_norm": 0.7477861046791077,
      "learning_rate": 2.879065284611842e-05,
      "loss": 0.5536,
      "num_input_tokens_seen": 12641552,
      "step": 19325
    },
    {
      "epoch": 10.131027253668764,
      "grad_norm": 0.9970418810844421,
      "learning_rate": 2.8779350369217634e-05,
      "loss": 0.475,
      "num_input_tokens_seen": 12644976,
      "step": 19330
    },
    {
      "epoch": 10.133647798742139,
      "grad_norm": 0.4189894199371338,
      "learning_rate": 2.876804710171987e-05,
      "loss": 0.4569,
      "num_input_tokens_seen": 12649072,
      "step": 19335
    },
    {
      "epoch": 10.136268343815514,
      "grad_norm": 0.2554369866847992,
      "learning_rate": 2.8756743045989643e-05,
      "loss": 0.4025,
      "num_input_tokens_seen": 12651760,
      "step": 19340
    },
    {
      "epoch": 10.13888888888889,
      "grad_norm": 0.30213505029678345,
      "learning_rate": 2.8745438204391623e-05,
      "loss": 0.5431,
      "num_input_tokens_seen": 12654576,
      "step": 19345
    },
    {
      "epoch": 10.141509433962264,
      "grad_norm": 1.1951258182525635,
      "learning_rate": 2.8734132579290675e-05,
      "loss": 0.4962,
      "num_input_tokens_seen": 12657520,
      "step": 19350
    },
    {
      "epoch": 10.14412997903564,
      "grad_norm": 0.3711563050746918,
      "learning_rate": 2.8722826173051782e-05,
      "loss": 0.4246,
      "num_input_tokens_seen": 12660560,
      "step": 19355
    },
    {
      "epoch": 10.146750524109015,
      "grad_norm": 0.4385834336280823,
      "learning_rate": 2.871151898804012e-05,
      "loss": 0.466,
      "num_input_tokens_seen": 12664080,
      "step": 19360
    },
    {
      "epoch": 10.14937106918239,
      "grad_norm": 0.48099786043167114,
      "learning_rate": 2.8700211026621043e-05,
      "loss": 0.6114,
      "num_input_tokens_seen": 12667792,
      "step": 19365
    },
    {
      "epoch": 10.151991614255765,
      "grad_norm": 0.5337663888931274,
      "learning_rate": 2.8688902291160025e-05,
      "loss": 0.3918,
      "num_input_tokens_seen": 12670704,
      "step": 19370
    },
    {
      "epoch": 10.15461215932914,
      "grad_norm": 1.1803677082061768,
      "learning_rate": 2.867759278402273e-05,
      "loss": 0.5169,
      "num_input_tokens_seen": 12673136,
      "step": 19375
    },
    {
      "epoch": 10.157232704402515,
      "grad_norm": 0.5690073370933533,
      "learning_rate": 2.8666282507574994e-05,
      "loss": 0.5622,
      "num_input_tokens_seen": 12677040,
      "step": 19380
    },
    {
      "epoch": 10.15985324947589,
      "grad_norm": 2.483741044998169,
      "learning_rate": 2.8654971464182773e-05,
      "loss": 0.4641,
      "num_input_tokens_seen": 12679984,
      "step": 19385
    },
    {
      "epoch": 10.162473794549266,
      "grad_norm": 0.3579897880554199,
      "learning_rate": 2.8643659656212222e-05,
      "loss": 0.4328,
      "num_input_tokens_seen": 12682832,
      "step": 19390
    },
    {
      "epoch": 10.165094339622641,
      "grad_norm": 0.4795098602771759,
      "learning_rate": 2.8632347086029642e-05,
      "loss": 0.4582,
      "num_input_tokens_seen": 12685968,
      "step": 19395
    },
    {
      "epoch": 10.167714884696016,
      "grad_norm": 0.6921142339706421,
      "learning_rate": 2.8621033756001486e-05,
      "loss": 0.4439,
      "num_input_tokens_seen": 12688624,
      "step": 19400
    },
    {
      "epoch": 10.170335429769391,
      "grad_norm": 0.7906513810157776,
      "learning_rate": 2.8609719668494394e-05,
      "loss": 0.5503,
      "num_input_tokens_seen": 12691728,
      "step": 19405
    },
    {
      "epoch": 10.172955974842766,
      "grad_norm": 0.4055326581001282,
      "learning_rate": 2.8598404825875113e-05,
      "loss": 0.3892,
      "num_input_tokens_seen": 12694416,
      "step": 19410
    },
    {
      "epoch": 10.175576519916143,
      "grad_norm": 0.27365654706954956,
      "learning_rate": 2.8587089230510604e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 12697360,
      "step": 19415
    },
    {
      "epoch": 10.178197064989519,
      "grad_norm": 0.4173250198364258,
      "learning_rate": 2.857577288476795e-05,
      "loss": 0.3173,
      "num_input_tokens_seen": 12700560,
      "step": 19420
    },
    {
      "epoch": 10.180817610062894,
      "grad_norm": 0.44159480929374695,
      "learning_rate": 2.85644557910144e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 12703024,
      "step": 19425
    },
    {
      "epoch": 10.183438155136269,
      "grad_norm": 0.6883893609046936,
      "learning_rate": 2.8553137951617366e-05,
      "loss": 0.4694,
      "num_input_tokens_seen": 12706864,
      "step": 19430
    },
    {
      "epoch": 10.186058700209644,
      "grad_norm": 0.39758142828941345,
      "learning_rate": 2.854181936894441e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 12709776,
      "step": 19435
    },
    {
      "epoch": 10.18867924528302,
      "grad_norm": 0.5941957831382751,
      "learning_rate": 2.8530500045363247e-05,
      "loss": 0.4449,
      "num_input_tokens_seen": 12713392,
      "step": 19440
    },
    {
      "epoch": 10.191299790356394,
      "grad_norm": 0.39500677585601807,
      "learning_rate": 2.851917998324175e-05,
      "loss": 0.4464,
      "num_input_tokens_seen": 12716112,
      "step": 19445
    },
    {
      "epoch": 10.19392033542977,
      "grad_norm": 0.7974594235420227,
      "learning_rate": 2.8507859184947957e-05,
      "loss": 0.4353,
      "num_input_tokens_seen": 12721392,
      "step": 19450
    },
    {
      "epoch": 10.196540880503145,
      "grad_norm": 0.2919536232948303,
      "learning_rate": 2.8496537652850043e-05,
      "loss": 0.5158,
      "num_input_tokens_seen": 12724592,
      "step": 19455
    },
    {
      "epoch": 10.19916142557652,
      "grad_norm": 0.2881413698196411,
      "learning_rate": 2.8485215389316333e-05,
      "loss": 0.3831,
      "num_input_tokens_seen": 12728176,
      "step": 19460
    },
    {
      "epoch": 10.201781970649895,
      "grad_norm": 0.34668031334877014,
      "learning_rate": 2.8473892396715328e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 12732176,
      "step": 19465
    },
    {
      "epoch": 10.20440251572327,
      "grad_norm": 0.3381885588169098,
      "learning_rate": 2.846256867741567e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 12734864,
      "step": 19470
    },
    {
      "epoch": 10.207023060796645,
      "grad_norm": 0.5059210062026978,
      "learning_rate": 2.845124423378614e-05,
      "loss": 0.3505,
      "num_input_tokens_seen": 12737968,
      "step": 19475
    },
    {
      "epoch": 10.20964360587002,
      "grad_norm": 0.3462825119495392,
      "learning_rate": 2.8439919068195697e-05,
      "loss": 0.49,
      "num_input_tokens_seen": 12741968,
      "step": 19480
    },
    {
      "epoch": 10.212264150943396,
      "grad_norm": 0.7306724190711975,
      "learning_rate": 2.842859318301343e-05,
      "loss": 0.5017,
      "num_input_tokens_seen": 12744976,
      "step": 19485
    },
    {
      "epoch": 10.214884696016771,
      "grad_norm": 0.43953925371170044,
      "learning_rate": 2.841726658060858e-05,
      "loss": 0.5006,
      "num_input_tokens_seen": 12748176,
      "step": 19490
    },
    {
      "epoch": 10.217505241090146,
      "grad_norm": 0.6743193864822388,
      "learning_rate": 2.8405939263350555e-05,
      "loss": 0.3094,
      "num_input_tokens_seen": 12750640,
      "step": 19495
    },
    {
      "epoch": 10.220125786163521,
      "grad_norm": 0.4221574366092682,
      "learning_rate": 2.8394611233608886e-05,
      "loss": 0.3443,
      "num_input_tokens_seen": 12753552,
      "step": 19500
    },
    {
      "epoch": 10.222746331236896,
      "grad_norm": 0.6275908946990967,
      "learning_rate": 2.8383282493753283e-05,
      "loss": 0.5332,
      "num_input_tokens_seen": 12756400,
      "step": 19505
    },
    {
      "epoch": 10.225366876310272,
      "grad_norm": 0.42889246344566345,
      "learning_rate": 2.8371953046153575e-05,
      "loss": 0.4101,
      "num_input_tokens_seen": 12759632,
      "step": 19510
    },
    {
      "epoch": 10.227987421383649,
      "grad_norm": 0.46324193477630615,
      "learning_rate": 2.8360622893179757e-05,
      "loss": 0.3874,
      "num_input_tokens_seen": 12762672,
      "step": 19515
    },
    {
      "epoch": 10.230607966457024,
      "grad_norm": 0.41048571467399597,
      "learning_rate": 2.834929203720197e-05,
      "loss": 0.352,
      "num_input_tokens_seen": 12766192,
      "step": 19520
    },
    {
      "epoch": 10.233228511530399,
      "grad_norm": 0.2848590910434723,
      "learning_rate": 2.83379604805905e-05,
      "loss": 0.4006,
      "num_input_tokens_seen": 12770064,
      "step": 19525
    },
    {
      "epoch": 10.235849056603774,
      "grad_norm": 0.4906681478023529,
      "learning_rate": 2.832662822571578e-05,
      "loss": 0.5345,
      "num_input_tokens_seen": 12773488,
      "step": 19530
    },
    {
      "epoch": 10.23846960167715,
      "grad_norm": 0.35782188177108765,
      "learning_rate": 2.8315295274948395e-05,
      "loss": 0.4936,
      "num_input_tokens_seen": 12776080,
      "step": 19535
    },
    {
      "epoch": 10.241090146750524,
      "grad_norm": 0.39627161622047424,
      "learning_rate": 2.8303961630659047e-05,
      "loss": 0.3082,
      "num_input_tokens_seen": 12779280,
      "step": 19540
    },
    {
      "epoch": 10.2437106918239,
      "grad_norm": 0.33253565430641174,
      "learning_rate": 2.8292627295218627e-05,
      "loss": 0.5257,
      "num_input_tokens_seen": 12783120,
      "step": 19545
    },
    {
      "epoch": 10.246331236897275,
      "grad_norm": 0.32044023275375366,
      "learning_rate": 2.828129227099815e-05,
      "loss": 0.5877,
      "num_input_tokens_seen": 12785968,
      "step": 19550
    },
    {
      "epoch": 10.24895178197065,
      "grad_norm": 0.7145405411720276,
      "learning_rate": 2.826995656036875e-05,
      "loss": 0.505,
      "num_input_tokens_seen": 12789168,
      "step": 19555
    },
    {
      "epoch": 10.251572327044025,
      "grad_norm": 0.34082621335983276,
      "learning_rate": 2.8258620165701747e-05,
      "loss": 0.3806,
      "num_input_tokens_seen": 12793328,
      "step": 19560
    },
    {
      "epoch": 10.2541928721174,
      "grad_norm": 0.4641399383544922,
      "learning_rate": 2.8247283089368586e-05,
      "loss": 0.4693,
      "num_input_tokens_seen": 12796048,
      "step": 19565
    },
    {
      "epoch": 10.256813417190775,
      "grad_norm": 0.4295702278614044,
      "learning_rate": 2.8235945333740842e-05,
      "loss": 0.4269,
      "num_input_tokens_seen": 12800336,
      "step": 19570
    },
    {
      "epoch": 10.25943396226415,
      "grad_norm": 0.41701218485832214,
      "learning_rate": 2.822460690119026e-05,
      "loss": 0.473,
      "num_input_tokens_seen": 12803088,
      "step": 19575
    },
    {
      "epoch": 10.262054507337526,
      "grad_norm": 0.5690174102783203,
      "learning_rate": 2.82132677940887e-05,
      "loss": 0.5787,
      "num_input_tokens_seen": 12806320,
      "step": 19580
    },
    {
      "epoch": 10.264675052410901,
      "grad_norm": 0.5857667922973633,
      "learning_rate": 2.820192801480817e-05,
      "loss": 0.4941,
      "num_input_tokens_seen": 12809488,
      "step": 19585
    },
    {
      "epoch": 10.267295597484276,
      "grad_norm": 0.8255484104156494,
      "learning_rate": 2.8190587565720842e-05,
      "loss": 0.4571,
      "num_input_tokens_seen": 12813232,
      "step": 19590
    },
    {
      "epoch": 10.269916142557651,
      "grad_norm": 0.6502721905708313,
      "learning_rate": 2.8179246449198986e-05,
      "loss": 0.552,
      "num_input_tokens_seen": 12816048,
      "step": 19595
    },
    {
      "epoch": 10.272536687631026,
      "grad_norm": 0.6362029314041138,
      "learning_rate": 2.8167904667615048e-05,
      "loss": 0.4436,
      "num_input_tokens_seen": 12818960,
      "step": 19600
    },
    {
      "epoch": 10.275157232704402,
      "grad_norm": 0.6203923225402832,
      "learning_rate": 2.8156562223341592e-05,
      "loss": 0.4601,
      "num_input_tokens_seen": 12822384,
      "step": 19605
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 0.5413923263549805,
      "learning_rate": 2.814521911875133e-05,
      "loss": 0.4472,
      "num_input_tokens_seen": 12825136,
      "step": 19610
    },
    {
      "epoch": 10.280398322851154,
      "grad_norm": 0.3008306324481964,
      "learning_rate": 2.8133875356217114e-05,
      "loss": 0.35,
      "num_input_tokens_seen": 12829488,
      "step": 19615
    },
    {
      "epoch": 10.283018867924529,
      "grad_norm": 0.4437774419784546,
      "learning_rate": 2.8122530938111923e-05,
      "loss": 0.4626,
      "num_input_tokens_seen": 12832432,
      "step": 19620
    },
    {
      "epoch": 10.285639412997904,
      "grad_norm": 0.376735121011734,
      "learning_rate": 2.8111185866808888e-05,
      "loss": 0.4738,
      "num_input_tokens_seen": 12835824,
      "step": 19625
    },
    {
      "epoch": 10.28825995807128,
      "grad_norm": 0.47078636288642883,
      "learning_rate": 2.809984014468126e-05,
      "loss": 0.5927,
      "num_input_tokens_seen": 12839504,
      "step": 19630
    },
    {
      "epoch": 10.290880503144654,
      "grad_norm": 0.33361098170280457,
      "learning_rate": 2.8088493774102443e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 12842448,
      "step": 19635
    },
    {
      "epoch": 10.29350104821803,
      "grad_norm": 0.3025000989437103,
      "learning_rate": 2.807714675744596e-05,
      "loss": 0.4291,
      "num_input_tokens_seen": 12845392,
      "step": 19640
    },
    {
      "epoch": 10.296121593291405,
      "grad_norm": 0.44289088249206543,
      "learning_rate": 2.8065799097085475e-05,
      "loss": 0.5197,
      "num_input_tokens_seen": 12849424,
      "step": 19645
    },
    {
      "epoch": 10.29874213836478,
      "grad_norm": 0.6201605796813965,
      "learning_rate": 2.8054450795394804e-05,
      "loss": 0.3465,
      "num_input_tokens_seen": 12852368,
      "step": 19650
    },
    {
      "epoch": 10.301362683438155,
      "grad_norm": 0.4808075428009033,
      "learning_rate": 2.8043101854747878e-05,
      "loss": 0.5409,
      "num_input_tokens_seen": 12855920,
      "step": 19655
    },
    {
      "epoch": 10.30398322851153,
      "grad_norm": 0.6097260117530823,
      "learning_rate": 2.8031752277518748e-05,
      "loss": 0.4071,
      "num_input_tokens_seen": 12859056,
      "step": 19660
    },
    {
      "epoch": 10.306603773584905,
      "grad_norm": 0.472372442483902,
      "learning_rate": 2.802040206608164e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 12861712,
      "step": 19665
    },
    {
      "epoch": 10.30922431865828,
      "grad_norm": 0.4138677716255188,
      "learning_rate": 2.8009051222810877e-05,
      "loss": 0.4239,
      "num_input_tokens_seen": 12864880,
      "step": 19670
    },
    {
      "epoch": 10.311844863731656,
      "grad_norm": 0.4503040909767151,
      "learning_rate": 2.7997699750080918e-05,
      "loss": 0.3891,
      "num_input_tokens_seen": 12867664,
      "step": 19675
    },
    {
      "epoch": 10.314465408805031,
      "grad_norm": 0.349638432264328,
      "learning_rate": 2.7986347650266386e-05,
      "loss": 0.436,
      "num_input_tokens_seen": 12872112,
      "step": 19680
    },
    {
      "epoch": 10.317085953878406,
      "grad_norm": 0.31206366419792175,
      "learning_rate": 2.797499492574199e-05,
      "loss": 0.4706,
      "num_input_tokens_seen": 12874928,
      "step": 19685
    },
    {
      "epoch": 10.319706498951781,
      "grad_norm": 0.4117976725101471,
      "learning_rate": 2.7963641578882594e-05,
      "loss": 0.6407,
      "num_input_tokens_seen": 12878224,
      "step": 19690
    },
    {
      "epoch": 10.322327044025156,
      "grad_norm": 0.5625401735305786,
      "learning_rate": 2.7952287612063193e-05,
      "loss": 0.4793,
      "num_input_tokens_seen": 12882128,
      "step": 19695
    },
    {
      "epoch": 10.324947589098532,
      "grad_norm": 0.3927326500415802,
      "learning_rate": 2.79409330276589e-05,
      "loss": 0.4472,
      "num_input_tokens_seen": 12885360,
      "step": 19700
    },
    {
      "epoch": 10.327568134171909,
      "grad_norm": 0.3943522274494171,
      "learning_rate": 2.7929577828044974e-05,
      "loss": 0.3485,
      "num_input_tokens_seen": 12888656,
      "step": 19705
    },
    {
      "epoch": 10.330188679245284,
      "grad_norm": 0.2533826529979706,
      "learning_rate": 2.7918222015596784e-05,
      "loss": 0.4537,
      "num_input_tokens_seen": 12892400,
      "step": 19710
    },
    {
      "epoch": 10.332809224318659,
      "grad_norm": 0.32340583205223083,
      "learning_rate": 2.7906865592689847e-05,
      "loss": 0.527,
      "num_input_tokens_seen": 12895088,
      "step": 19715
    },
    {
      "epoch": 10.335429769392034,
      "grad_norm": 0.5095556378364563,
      "learning_rate": 2.789550856169979e-05,
      "loss": 0.4599,
      "num_input_tokens_seen": 12897712,
      "step": 19720
    },
    {
      "epoch": 10.33805031446541,
      "grad_norm": 0.4139721691608429,
      "learning_rate": 2.7884150925002374e-05,
      "loss": 0.4528,
      "num_input_tokens_seen": 12901520,
      "step": 19725
    },
    {
      "epoch": 10.340670859538784,
      "grad_norm": 0.5334341526031494,
      "learning_rate": 2.787279268497349e-05,
      "loss": 0.5017,
      "num_input_tokens_seen": 12906064,
      "step": 19730
    },
    {
      "epoch": 10.34329140461216,
      "grad_norm": 1.1347379684448242,
      "learning_rate": 2.786143384398915e-05,
      "loss": 0.5953,
      "num_input_tokens_seen": 12908400,
      "step": 19735
    },
    {
      "epoch": 10.345911949685535,
      "grad_norm": 0.5336925387382507,
      "learning_rate": 2.785007440442549e-05,
      "loss": 0.3523,
      "num_input_tokens_seen": 12911088,
      "step": 19740
    },
    {
      "epoch": 10.34853249475891,
      "grad_norm": 0.40065625309944153,
      "learning_rate": 2.7838714368658785e-05,
      "loss": 0.5575,
      "num_input_tokens_seen": 12913424,
      "step": 19745
    },
    {
      "epoch": 10.351153039832285,
      "grad_norm": 0.4040432870388031,
      "learning_rate": 2.7827353739065427e-05,
      "loss": 0.5301,
      "num_input_tokens_seen": 12917104,
      "step": 19750
    },
    {
      "epoch": 10.35377358490566,
      "grad_norm": 0.5024396777153015,
      "learning_rate": 2.781599251802191e-05,
      "loss": 0.2885,
      "num_input_tokens_seen": 12919984,
      "step": 19755
    },
    {
      "epoch": 10.356394129979035,
      "grad_norm": 0.34248101711273193,
      "learning_rate": 2.780463070790489e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 12923792,
      "step": 19760
    },
    {
      "epoch": 10.35901467505241,
      "grad_norm": 0.4098190367221832,
      "learning_rate": 2.7793268311091125e-05,
      "loss": 0.5228,
      "num_input_tokens_seen": 12926320,
      "step": 19765
    },
    {
      "epoch": 10.361635220125786,
      "grad_norm": 0.3661225140094757,
      "learning_rate": 2.7781905329957496e-05,
      "loss": 0.4516,
      "num_input_tokens_seen": 12929104,
      "step": 19770
    },
    {
      "epoch": 10.364255765199161,
      "grad_norm": 0.3566299378871918,
      "learning_rate": 2.7770541766881008e-05,
      "loss": 0.3634,
      "num_input_tokens_seen": 12932496,
      "step": 19775
    },
    {
      "epoch": 10.366876310272536,
      "grad_norm": 0.47028878331184387,
      "learning_rate": 2.7759177624238786e-05,
      "loss": 0.4076,
      "num_input_tokens_seen": 12935760,
      "step": 19780
    },
    {
      "epoch": 10.369496855345911,
      "grad_norm": 0.4656876027584076,
      "learning_rate": 2.7747812904408094e-05,
      "loss": 0.3785,
      "num_input_tokens_seen": 12938736,
      "step": 19785
    },
    {
      "epoch": 10.372117400419286,
      "grad_norm": 0.48295971751213074,
      "learning_rate": 2.7736447609766286e-05,
      "loss": 0.4132,
      "num_input_tokens_seen": 12941616,
      "step": 19790
    },
    {
      "epoch": 10.374737945492662,
      "grad_norm": 0.5412299036979675,
      "learning_rate": 2.772508174269085e-05,
      "loss": 0.3746,
      "num_input_tokens_seen": 12944368,
      "step": 19795
    },
    {
      "epoch": 10.377358490566039,
      "grad_norm": 0.8560463786125183,
      "learning_rate": 2.7713715305559414e-05,
      "loss": 0.519,
      "num_input_tokens_seen": 12950224,
      "step": 19800
    },
    {
      "epoch": 10.379979035639414,
      "grad_norm": 1.0403521060943604,
      "learning_rate": 2.7702348300749692e-05,
      "loss": 0.4522,
      "num_input_tokens_seen": 12952656,
      "step": 19805
    },
    {
      "epoch": 10.382599580712789,
      "grad_norm": 0.523546040058136,
      "learning_rate": 2.7690980730639533e-05,
      "loss": 0.29,
      "num_input_tokens_seen": 12957520,
      "step": 19810
    },
    {
      "epoch": 10.385220125786164,
      "grad_norm": 0.28288477659225464,
      "learning_rate": 2.7679612597606902e-05,
      "loss": 0.4711,
      "num_input_tokens_seen": 12961392,
      "step": 19815
    },
    {
      "epoch": 10.38784067085954,
      "grad_norm": 0.7041094303131104,
      "learning_rate": 2.7668243904029894e-05,
      "loss": 0.5118,
      "num_input_tokens_seen": 12968784,
      "step": 19820
    },
    {
      "epoch": 10.390461215932914,
      "grad_norm": 0.2770914137363434,
      "learning_rate": 2.76568746522867e-05,
      "loss": 0.3886,
      "num_input_tokens_seen": 12972176,
      "step": 19825
    },
    {
      "epoch": 10.39308176100629,
      "grad_norm": 0.4556155502796173,
      "learning_rate": 2.7645504844755637e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 12974672,
      "step": 19830
    },
    {
      "epoch": 10.395702306079665,
      "grad_norm": 0.339233934879303,
      "learning_rate": 2.7634134483815143e-05,
      "loss": 0.4253,
      "num_input_tokens_seen": 12978224,
      "step": 19835
    },
    {
      "epoch": 10.39832285115304,
      "grad_norm": 0.5150198936462402,
      "learning_rate": 2.762276357184377e-05,
      "loss": 0.3791,
      "num_input_tokens_seen": 12981840,
      "step": 19840
    },
    {
      "epoch": 10.400943396226415,
      "grad_norm": 0.4082224369049072,
      "learning_rate": 2.7611392111220176e-05,
      "loss": 0.5505,
      "num_input_tokens_seen": 12984496,
      "step": 19845
    },
    {
      "epoch": 10.40356394129979,
      "grad_norm": 0.32325297594070435,
      "learning_rate": 2.7600020104323154e-05,
      "loss": 0.4943,
      "num_input_tokens_seen": 12986992,
      "step": 19850
    },
    {
      "epoch": 10.406184486373165,
      "grad_norm": 0.4374060332775116,
      "learning_rate": 2.7588647553531576e-05,
      "loss": 0.5327,
      "num_input_tokens_seen": 12989968,
      "step": 19855
    },
    {
      "epoch": 10.40880503144654,
      "grad_norm": 0.32975834608078003,
      "learning_rate": 2.7577274461224463e-05,
      "loss": 0.4583,
      "num_input_tokens_seen": 12993200,
      "step": 19860
    },
    {
      "epoch": 10.411425576519916,
      "grad_norm": 0.43019941449165344,
      "learning_rate": 2.756590082978095e-05,
      "loss": 0.3501,
      "num_input_tokens_seen": 12996432,
      "step": 19865
    },
    {
      "epoch": 10.414046121593291,
      "grad_norm": 0.4202328026294708,
      "learning_rate": 2.7554526661580243e-05,
      "loss": 0.5652,
      "num_input_tokens_seen": 12999600,
      "step": 19870
    },
    {
      "epoch": 10.416666666666666,
      "grad_norm": 0.5833422541618347,
      "learning_rate": 2.7543151959001708e-05,
      "loss": 0.4455,
      "num_input_tokens_seen": 13002800,
      "step": 19875
    },
    {
      "epoch": 10.419287211740041,
      "grad_norm": 0.4053517282009125,
      "learning_rate": 2.7531776724424797e-05,
      "loss": 0.3781,
      "num_input_tokens_seen": 13005680,
      "step": 19880
    },
    {
      "epoch": 10.421907756813416,
      "grad_norm": 0.3770594000816345,
      "learning_rate": 2.752040096022907e-05,
      "loss": 0.5432,
      "num_input_tokens_seen": 13008784,
      "step": 19885
    },
    {
      "epoch": 10.424528301886792,
      "grad_norm": 1.6307454109191895,
      "learning_rate": 2.750902466879423e-05,
      "loss": 0.4597,
      "num_input_tokens_seen": 13011824,
      "step": 19890
    },
    {
      "epoch": 10.427148846960169,
      "grad_norm": 0.32702431082725525,
      "learning_rate": 2.7497647852500048e-05,
      "loss": 0.3804,
      "num_input_tokens_seen": 13015568,
      "step": 19895
    },
    {
      "epoch": 10.429769392033544,
      "grad_norm": 0.5976666808128357,
      "learning_rate": 2.7486270513726424e-05,
      "loss": 0.4201,
      "num_input_tokens_seen": 13018320,
      "step": 19900
    },
    {
      "epoch": 10.432389937106919,
      "grad_norm": 0.6404386758804321,
      "learning_rate": 2.7474892654853384e-05,
      "loss": 0.3011,
      "num_input_tokens_seen": 13020816,
      "step": 19905
    },
    {
      "epoch": 10.435010482180294,
      "grad_norm": 0.37449896335601807,
      "learning_rate": 2.7463514278261034e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 13023920,
      "step": 19910
    },
    {
      "epoch": 10.43763102725367,
      "grad_norm": 0.4835715889930725,
      "learning_rate": 2.7452135386329596e-05,
      "loss": 0.3801,
      "num_input_tokens_seen": 13027312,
      "step": 19915
    },
    {
      "epoch": 10.440251572327044,
      "grad_norm": 0.29327523708343506,
      "learning_rate": 2.7440755981439414e-05,
      "loss": 0.4045,
      "num_input_tokens_seen": 13032560,
      "step": 19920
    },
    {
      "epoch": 10.44287211740042,
      "grad_norm": 0.8787369132041931,
      "learning_rate": 2.7429376065970924e-05,
      "loss": 0.5049,
      "num_input_tokens_seen": 13034544,
      "step": 19925
    },
    {
      "epoch": 10.445492662473795,
      "grad_norm": 0.44249415397644043,
      "learning_rate": 2.7417995642304678e-05,
      "loss": 0.4806,
      "num_input_tokens_seen": 13038096,
      "step": 19930
    },
    {
      "epoch": 10.44811320754717,
      "grad_norm": 0.40024295449256897,
      "learning_rate": 2.7406614712821333e-05,
      "loss": 0.5407,
      "num_input_tokens_seen": 13041168,
      "step": 19935
    },
    {
      "epoch": 10.450733752620545,
      "grad_norm": 0.3582644462585449,
      "learning_rate": 2.739523327990164e-05,
      "loss": 0.4885,
      "num_input_tokens_seen": 13044240,
      "step": 19940
    },
    {
      "epoch": 10.45335429769392,
      "grad_norm": 0.3481682538986206,
      "learning_rate": 2.7383851345926477e-05,
      "loss": 0.4403,
      "num_input_tokens_seen": 13047888,
      "step": 19945
    },
    {
      "epoch": 10.455974842767295,
      "grad_norm": 0.62037193775177,
      "learning_rate": 2.7372468913276804e-05,
      "loss": 0.5288,
      "num_input_tokens_seen": 13051120,
      "step": 19950
    },
    {
      "epoch": 10.45859538784067,
      "grad_norm": 0.4072948694229126,
      "learning_rate": 2.7361085984333705e-05,
      "loss": 0.356,
      "num_input_tokens_seen": 13054320,
      "step": 19955
    },
    {
      "epoch": 10.461215932914046,
      "grad_norm": 0.27233853936195374,
      "learning_rate": 2.7349702561478356e-05,
      "loss": 0.412,
      "num_input_tokens_seen": 13058736,
      "step": 19960
    },
    {
      "epoch": 10.463836477987421,
      "grad_norm": 0.4873991310596466,
      "learning_rate": 2.733831864709203e-05,
      "loss": 0.4046,
      "num_input_tokens_seen": 13061904,
      "step": 19965
    },
    {
      "epoch": 10.466457023060796,
      "grad_norm": 0.5670714974403381,
      "learning_rate": 2.732693424355613e-05,
      "loss": 0.3797,
      "num_input_tokens_seen": 13064976,
      "step": 19970
    },
    {
      "epoch": 10.469077568134171,
      "grad_norm": 0.5502851605415344,
      "learning_rate": 2.731554935325213e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 13067472,
      "step": 19975
    },
    {
      "epoch": 10.471698113207546,
      "grad_norm": 0.4463476836681366,
      "learning_rate": 2.7304163978561613e-05,
      "loss": 0.556,
      "num_input_tokens_seen": 13070224,
      "step": 19980
    },
    {
      "epoch": 10.474318658280922,
      "grad_norm": 0.4392985999584198,
      "learning_rate": 2.729277812186629e-05,
      "loss": 0.3075,
      "num_input_tokens_seen": 13073008,
      "step": 19985
    },
    {
      "epoch": 10.476939203354299,
      "grad_norm": 0.47206541895866394,
      "learning_rate": 2.7281391785547945e-05,
      "loss": 0.4181,
      "num_input_tokens_seen": 13076752,
      "step": 19990
    },
    {
      "epoch": 10.479559748427674,
      "grad_norm": 0.40531593561172485,
      "learning_rate": 2.7270004971988456e-05,
      "loss": 0.3342,
      "num_input_tokens_seen": 13081040,
      "step": 19995
    },
    {
      "epoch": 10.482180293501049,
      "grad_norm": 0.49446019530296326,
      "learning_rate": 2.7258617683569825e-05,
      "loss": 0.3829,
      "num_input_tokens_seen": 13083632,
      "step": 20000
    },
    {
      "epoch": 10.484800838574424,
      "grad_norm": 0.48039206862449646,
      "learning_rate": 2.7247229922674143e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 13086928,
      "step": 20005
    },
    {
      "epoch": 10.4874213836478,
      "grad_norm": 0.4300840198993683,
      "learning_rate": 2.7235841691683605e-05,
      "loss": 0.3827,
      "num_input_tokens_seen": 13089424,
      "step": 20010
    },
    {
      "epoch": 10.490041928721174,
      "grad_norm": 0.6086318492889404,
      "learning_rate": 2.7224452992980487e-05,
      "loss": 0.4487,
      "num_input_tokens_seen": 13092688,
      "step": 20015
    },
    {
      "epoch": 10.49266247379455,
      "grad_norm": 0.2374255508184433,
      "learning_rate": 2.7213063828947182e-05,
      "loss": 0.3476,
      "num_input_tokens_seen": 13096048,
      "step": 20020
    },
    {
      "epoch": 10.495283018867925,
      "grad_norm": 0.4395695924758911,
      "learning_rate": 2.7201674201966176e-05,
      "loss": 0.5703,
      "num_input_tokens_seen": 13099760,
      "step": 20025
    },
    {
      "epoch": 10.4979035639413,
      "grad_norm": 0.5994921326637268,
      "learning_rate": 2.7190284114420035e-05,
      "loss": 0.3633,
      "num_input_tokens_seen": 13103856,
      "step": 20030
    },
    {
      "epoch": 10.500524109014675,
      "grad_norm": 0.6948838233947754,
      "learning_rate": 2.717889356869146e-05,
      "loss": 0.4452,
      "num_input_tokens_seen": 13106288,
      "step": 20035
    },
    {
      "epoch": 10.50314465408805,
      "grad_norm": 0.6787749528884888,
      "learning_rate": 2.7167502567163194e-05,
      "loss": 0.4842,
      "num_input_tokens_seen": 13108752,
      "step": 20040
    },
    {
      "epoch": 10.505765199161425,
      "grad_norm": 0.3583753705024719,
      "learning_rate": 2.715611111221813e-05,
      "loss": 0.4746,
      "num_input_tokens_seen": 13111600,
      "step": 20045
    },
    {
      "epoch": 10.5083857442348,
      "grad_norm": 0.34844061732292175,
      "learning_rate": 2.7144719206239217e-05,
      "loss": 0.3976,
      "num_input_tokens_seen": 13114352,
      "step": 20050
    },
    {
      "epoch": 10.511006289308176,
      "grad_norm": 0.4827231466770172,
      "learning_rate": 2.7133326851609508e-05,
      "loss": 0.4341,
      "num_input_tokens_seen": 13117552,
      "step": 20055
    },
    {
      "epoch": 10.51362683438155,
      "grad_norm": 0.3583219349384308,
      "learning_rate": 2.7121934050712166e-05,
      "loss": 0.3177,
      "num_input_tokens_seen": 13120368,
      "step": 20060
    },
    {
      "epoch": 10.516247379454926,
      "grad_norm": 0.28969529271125793,
      "learning_rate": 2.711054080593043e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 13123728,
      "step": 20065
    },
    {
      "epoch": 10.518867924528301,
      "grad_norm": 0.5311833024024963,
      "learning_rate": 2.7099147119647627e-05,
      "loss": 0.4547,
      "num_input_tokens_seen": 13127440,
      "step": 20070
    },
    {
      "epoch": 10.521488469601676,
      "grad_norm": 0.44677311182022095,
      "learning_rate": 2.7087752994247206e-05,
      "loss": 0.4699,
      "num_input_tokens_seen": 13130320,
      "step": 20075
    },
    {
      "epoch": 10.524109014675052,
      "grad_norm": 0.4872855246067047,
      "learning_rate": 2.707635843211267e-05,
      "loss": 0.4229,
      "num_input_tokens_seen": 13133424,
      "step": 20080
    },
    {
      "epoch": 10.526729559748428,
      "grad_norm": 0.5275387167930603,
      "learning_rate": 2.7064963435627637e-05,
      "loss": 0.5459,
      "num_input_tokens_seen": 13137136,
      "step": 20085
    },
    {
      "epoch": 10.529350104821804,
      "grad_norm": 0.3256334662437439,
      "learning_rate": 2.7053568007175823e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 13141264,
      "step": 20090
    },
    {
      "epoch": 10.531970649895179,
      "grad_norm": 0.31231996417045593,
      "learning_rate": 2.7042172149140994e-05,
      "loss": 0.5366,
      "num_input_tokens_seen": 13144624,
      "step": 20095
    },
    {
      "epoch": 10.534591194968554,
      "grad_norm": 0.5866744518280029,
      "learning_rate": 2.703077586390706e-05,
      "loss": 0.3909,
      "num_input_tokens_seen": 13147216,
      "step": 20100
    },
    {
      "epoch": 10.53721174004193,
      "grad_norm": 0.5652526617050171,
      "learning_rate": 2.7019379153857982e-05,
      "loss": 0.5038,
      "num_input_tokens_seen": 13149296,
      "step": 20105
    },
    {
      "epoch": 10.539832285115304,
      "grad_norm": 0.5183078050613403,
      "learning_rate": 2.700798202137782e-05,
      "loss": 0.4131,
      "num_input_tokens_seen": 13153904,
      "step": 20110
    },
    {
      "epoch": 10.54245283018868,
      "grad_norm": 0.4114084243774414,
      "learning_rate": 2.6996584468850732e-05,
      "loss": 0.4174,
      "num_input_tokens_seen": 13157104,
      "step": 20115
    },
    {
      "epoch": 10.545073375262055,
      "grad_norm": 0.8305720090866089,
      "learning_rate": 2.6985186498660946e-05,
      "loss": 0.4696,
      "num_input_tokens_seen": 13159376,
      "step": 20120
    },
    {
      "epoch": 10.54769392033543,
      "grad_norm": 0.8679777979850769,
      "learning_rate": 2.69737881131928e-05,
      "loss": 0.5493,
      "num_input_tokens_seen": 13162384,
      "step": 20125
    },
    {
      "epoch": 10.550314465408805,
      "grad_norm": 0.5058850049972534,
      "learning_rate": 2.6962389314830692e-05,
      "loss": 0.4296,
      "num_input_tokens_seen": 13165744,
      "step": 20130
    },
    {
      "epoch": 10.55293501048218,
      "grad_norm": 0.22387149930000305,
      "learning_rate": 2.6950990105959135e-05,
      "loss": 0.5218,
      "num_input_tokens_seen": 13169808,
      "step": 20135
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 0.8538710474967957,
      "learning_rate": 2.6939590488962703e-05,
      "loss": 0.5467,
      "num_input_tokens_seen": 13172624,
      "step": 20140
    },
    {
      "epoch": 10.55817610062893,
      "grad_norm": 0.41009950637817383,
      "learning_rate": 2.6928190466226067e-05,
      "loss": 0.479,
      "num_input_tokens_seen": 13176208,
      "step": 20145
    },
    {
      "epoch": 10.560796645702306,
      "grad_norm": 0.8882958889007568,
      "learning_rate": 2.691679004013398e-05,
      "loss": 0.4936,
      "num_input_tokens_seen": 13179216,
      "step": 20150
    },
    {
      "epoch": 10.56341719077568,
      "grad_norm": 0.4786587953567505,
      "learning_rate": 2.6905389213071285e-05,
      "loss": 0.5072,
      "num_input_tokens_seen": 13182960,
      "step": 20155
    },
    {
      "epoch": 10.566037735849056,
      "grad_norm": 0.4552858769893646,
      "learning_rate": 2.6893987987422907e-05,
      "loss": 0.2856,
      "num_input_tokens_seen": 13186384,
      "step": 20160
    },
    {
      "epoch": 10.568658280922431,
      "grad_norm": 0.33799049258232117,
      "learning_rate": 2.6882586365573854e-05,
      "loss": 0.2796,
      "num_input_tokens_seen": 13189712,
      "step": 20165
    },
    {
      "epoch": 10.571278825995806,
      "grad_norm": 0.3523010015487671,
      "learning_rate": 2.6871184349909206e-05,
      "loss": 0.4914,
      "num_input_tokens_seen": 13193424,
      "step": 20170
    },
    {
      "epoch": 10.573899371069182,
      "grad_norm": 0.3835509419441223,
      "learning_rate": 2.685978194281414e-05,
      "loss": 0.3814,
      "num_input_tokens_seen": 13196400,
      "step": 20175
    },
    {
      "epoch": 10.576519916142558,
      "grad_norm": 0.48262450098991394,
      "learning_rate": 2.6848379146673914e-05,
      "loss": 0.491,
      "num_input_tokens_seen": 13200048,
      "step": 20180
    },
    {
      "epoch": 10.579140461215934,
      "grad_norm": 0.47537344694137573,
      "learning_rate": 2.6836975963873843e-05,
      "loss": 0.49,
      "num_input_tokens_seen": 13202896,
      "step": 20185
    },
    {
      "epoch": 10.581761006289309,
      "grad_norm": 0.40686702728271484,
      "learning_rate": 2.6825572396799377e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 13205520,
      "step": 20190
    },
    {
      "epoch": 10.584381551362684,
      "grad_norm": 0.9385026097297668,
      "learning_rate": 2.6814168447835986e-05,
      "loss": 0.4034,
      "num_input_tokens_seen": 13208784,
      "step": 20195
    },
    {
      "epoch": 10.58700209643606,
      "grad_norm": 0.40212079882621765,
      "learning_rate": 2.6802764119369245e-05,
      "loss": 0.3876,
      "num_input_tokens_seen": 13211952,
      "step": 20200
    },
    {
      "epoch": 10.589622641509434,
      "grad_norm": 0.4324948191642761,
      "learning_rate": 2.6791359413784828e-05,
      "loss": 0.3999,
      "num_input_tokens_seen": 13214960,
      "step": 20205
    },
    {
      "epoch": 10.59224318658281,
      "grad_norm": 0.5804754495620728,
      "learning_rate": 2.677995433346845e-05,
      "loss": 0.3891,
      "num_input_tokens_seen": 13219312,
      "step": 20210
    },
    {
      "epoch": 10.594863731656185,
      "grad_norm": 0.5523722767829895,
      "learning_rate": 2.6768548880805934e-05,
      "loss": 0.5733,
      "num_input_tokens_seen": 13222192,
      "step": 20215
    },
    {
      "epoch": 10.59748427672956,
      "grad_norm": 0.6540147066116333,
      "learning_rate": 2.6757143058183177e-05,
      "loss": 0.493,
      "num_input_tokens_seen": 13225360,
      "step": 20220
    },
    {
      "epoch": 10.600104821802935,
      "grad_norm": 0.32901719212532043,
      "learning_rate": 2.6745736867986132e-05,
      "loss": 0.3604,
      "num_input_tokens_seen": 13227792,
      "step": 20225
    },
    {
      "epoch": 10.60272536687631,
      "grad_norm": 0.29094117879867554,
      "learning_rate": 2.673433031260085e-05,
      "loss": 0.3891,
      "num_input_tokens_seen": 13231376,
      "step": 20230
    },
    {
      "epoch": 10.605345911949685,
      "grad_norm": 0.4954665005207062,
      "learning_rate": 2.6722923394413456e-05,
      "loss": 0.3999,
      "num_input_tokens_seen": 13234224,
      "step": 20235
    },
    {
      "epoch": 10.60796645702306,
      "grad_norm": 0.9944257736206055,
      "learning_rate": 2.6711516115810136e-05,
      "loss": 0.3566,
      "num_input_tokens_seen": 13239056,
      "step": 20240
    },
    {
      "epoch": 10.610587002096436,
      "grad_norm": 0.4041811227798462,
      "learning_rate": 2.6700108479177188e-05,
      "loss": 0.4904,
      "num_input_tokens_seen": 13242672,
      "step": 20245
    },
    {
      "epoch": 10.61320754716981,
      "grad_norm": 0.7399139404296875,
      "learning_rate": 2.6688700486900926e-05,
      "loss": 0.3573,
      "num_input_tokens_seen": 13246736,
      "step": 20250
    },
    {
      "epoch": 10.615828092243186,
      "grad_norm": 0.5997503399848938,
      "learning_rate": 2.6677292141367787e-05,
      "loss": 0.536,
      "num_input_tokens_seen": 13249456,
      "step": 20255
    },
    {
      "epoch": 10.618448637316561,
      "grad_norm": 0.37906160950660706,
      "learning_rate": 2.6665883444964278e-05,
      "loss": 0.5114,
      "num_input_tokens_seen": 13252400,
      "step": 20260
    },
    {
      "epoch": 10.621069182389936,
      "grad_norm": 0.4448961317539215,
      "learning_rate": 2.665447440007695e-05,
      "loss": 0.3437,
      "num_input_tokens_seen": 13255696,
      "step": 20265
    },
    {
      "epoch": 10.623689727463312,
      "grad_norm": 0.6816656589508057,
      "learning_rate": 2.664306500909246e-05,
      "loss": 0.4872,
      "num_input_tokens_seen": 13259888,
      "step": 20270
    },
    {
      "epoch": 10.626310272536688,
      "grad_norm": 0.4167565107345581,
      "learning_rate": 2.663165527439751e-05,
      "loss": 0.4607,
      "num_input_tokens_seen": 13263472,
      "step": 20275
    },
    {
      "epoch": 10.628930817610064,
      "grad_norm": 0.40458807349205017,
      "learning_rate": 2.662024519837889e-05,
      "loss": 0.5281,
      "num_input_tokens_seen": 13266224,
      "step": 20280
    },
    {
      "epoch": 10.631551362683439,
      "grad_norm": 0.7208436131477356,
      "learning_rate": 2.6608834783423465e-05,
      "loss": 0.5908,
      "num_input_tokens_seen": 13268880,
      "step": 20285
    },
    {
      "epoch": 10.634171907756814,
      "grad_norm": 0.35876646637916565,
      "learning_rate": 2.6597424031918156e-05,
      "loss": 0.4159,
      "num_input_tokens_seen": 13271760,
      "step": 20290
    },
    {
      "epoch": 10.63679245283019,
      "grad_norm": 0.3578992784023285,
      "learning_rate": 2.6586012946249954e-05,
      "loss": 0.3551,
      "num_input_tokens_seen": 13275152,
      "step": 20295
    },
    {
      "epoch": 10.639412997903564,
      "grad_norm": 0.7020267844200134,
      "learning_rate": 2.6574601528805953e-05,
      "loss": 0.4026,
      "num_input_tokens_seen": 13278480,
      "step": 20300
    },
    {
      "epoch": 10.64203354297694,
      "grad_norm": 0.34672048687934875,
      "learning_rate": 2.6563189781973268e-05,
      "loss": 0.385,
      "num_input_tokens_seen": 13281552,
      "step": 20305
    },
    {
      "epoch": 10.644654088050315,
      "grad_norm": 0.2974678874015808,
      "learning_rate": 2.655177770813912e-05,
      "loss": 0.549,
      "num_input_tokens_seen": 13284816,
      "step": 20310
    },
    {
      "epoch": 10.64727463312369,
      "grad_norm": 0.8791460394859314,
      "learning_rate": 2.654036530969077e-05,
      "loss": 0.3425,
      "num_input_tokens_seen": 13287600,
      "step": 20315
    },
    {
      "epoch": 10.649895178197065,
      "grad_norm": 1.082455039024353,
      "learning_rate": 2.6528952589015576e-05,
      "loss": 0.3763,
      "num_input_tokens_seen": 13290512,
      "step": 20320
    },
    {
      "epoch": 10.65251572327044,
      "grad_norm": 0.39690425992012024,
      "learning_rate": 2.6517539548500947e-05,
      "loss": 0.4143,
      "num_input_tokens_seen": 13293392,
      "step": 20325
    },
    {
      "epoch": 10.655136268343815,
      "grad_norm": 0.4848822057247162,
      "learning_rate": 2.6506126190534354e-05,
      "loss": 0.5843,
      "num_input_tokens_seen": 13296816,
      "step": 20330
    },
    {
      "epoch": 10.65775681341719,
      "grad_norm": 0.554365873336792,
      "learning_rate": 2.649471251750334e-05,
      "loss": 0.3791,
      "num_input_tokens_seen": 13300336,
      "step": 20335
    },
    {
      "epoch": 10.660377358490566,
      "grad_norm": 0.6673331260681152,
      "learning_rate": 2.6483298531795527e-05,
      "loss": 0.6805,
      "num_input_tokens_seen": 13303312,
      "step": 20340
    },
    {
      "epoch": 10.66299790356394,
      "grad_norm": 0.23706378042697906,
      "learning_rate": 2.6471884235798583e-05,
      "loss": 0.2877,
      "num_input_tokens_seen": 13305968,
      "step": 20345
    },
    {
      "epoch": 10.665618448637316,
      "grad_norm": 0.47130391001701355,
      "learning_rate": 2.6460469631900247e-05,
      "loss": 0.387,
      "num_input_tokens_seen": 13309392,
      "step": 20350
    },
    {
      "epoch": 10.668238993710691,
      "grad_norm": 0.6630795001983643,
      "learning_rate": 2.644905472248832e-05,
      "loss": 0.4029,
      "num_input_tokens_seen": 13312336,
      "step": 20355
    },
    {
      "epoch": 10.670859538784066,
      "grad_norm": 0.7403320670127869,
      "learning_rate": 2.6437639509950686e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 13315312,
      "step": 20360
    },
    {
      "epoch": 10.673480083857442,
      "grad_norm": 0.3772774636745453,
      "learning_rate": 2.6426223996675264e-05,
      "loss": 0.4613,
      "num_input_tokens_seen": 13318384,
      "step": 20365
    },
    {
      "epoch": 10.676100628930818,
      "grad_norm": 0.5092613697052002,
      "learning_rate": 2.641480818505005e-05,
      "loss": 0.4858,
      "num_input_tokens_seen": 13322224,
      "step": 20370
    },
    {
      "epoch": 10.678721174004194,
      "grad_norm": 1.0972580909729004,
      "learning_rate": 2.6403392077463108e-05,
      "loss": 0.7244,
      "num_input_tokens_seen": 13324368,
      "step": 20375
    },
    {
      "epoch": 10.681341719077569,
      "grad_norm": 0.45771223306655884,
      "learning_rate": 2.639197567630256e-05,
      "loss": 0.3942,
      "num_input_tokens_seen": 13327632,
      "step": 20380
    },
    {
      "epoch": 10.683962264150944,
      "grad_norm": 0.6583049893379211,
      "learning_rate": 2.638055898395657e-05,
      "loss": 0.434,
      "num_input_tokens_seen": 13330064,
      "step": 20385
    },
    {
      "epoch": 10.68658280922432,
      "grad_norm": 0.2825111746788025,
      "learning_rate": 2.6369142002813403e-05,
      "loss": 0.5407,
      "num_input_tokens_seen": 13336944,
      "step": 20390
    },
    {
      "epoch": 10.689203354297694,
      "grad_norm": 0.4828643500804901,
      "learning_rate": 2.6357724735261335e-05,
      "loss": 0.4848,
      "num_input_tokens_seen": 13340048,
      "step": 20395
    },
    {
      "epoch": 10.69182389937107,
      "grad_norm": 1.0226900577545166,
      "learning_rate": 2.6346307183688752e-05,
      "loss": 0.445,
      "num_input_tokens_seen": 13343024,
      "step": 20400
    },
    {
      "epoch": 10.694444444444445,
      "grad_norm": 0.5512526631355286,
      "learning_rate": 2.6334889350484064e-05,
      "loss": 0.4274,
      "num_input_tokens_seen": 13345968,
      "step": 20405
    },
    {
      "epoch": 10.69706498951782,
      "grad_norm": 0.2979440987110138,
      "learning_rate": 2.632347123803575e-05,
      "loss": 0.4742,
      "num_input_tokens_seen": 13349776,
      "step": 20410
    },
    {
      "epoch": 10.699685534591195,
      "grad_norm": 0.6648659110069275,
      "learning_rate": 2.6312052848732353e-05,
      "loss": 0.384,
      "num_input_tokens_seen": 13352080,
      "step": 20415
    },
    {
      "epoch": 10.70230607966457,
      "grad_norm": 0.48233333230018616,
      "learning_rate": 2.6300634184962475e-05,
      "loss": 0.4437,
      "num_input_tokens_seen": 13354672,
      "step": 20420
    },
    {
      "epoch": 10.704926624737945,
      "grad_norm": 0.378936231136322,
      "learning_rate": 2.628921524911475e-05,
      "loss": 0.3445,
      "num_input_tokens_seen": 13357776,
      "step": 20425
    },
    {
      "epoch": 10.70754716981132,
      "grad_norm": 0.4193604290485382,
      "learning_rate": 2.6277796043577924e-05,
      "loss": 0.423,
      "num_input_tokens_seen": 13360240,
      "step": 20430
    },
    {
      "epoch": 10.710167714884696,
      "grad_norm": 0.45419973134994507,
      "learning_rate": 2.6266376570740725e-05,
      "loss": 0.4635,
      "num_input_tokens_seen": 13363280,
      "step": 20435
    },
    {
      "epoch": 10.71278825995807,
      "grad_norm": 0.40667372941970825,
      "learning_rate": 2.6254956832992005e-05,
      "loss": 0.507,
      "num_input_tokens_seen": 13366256,
      "step": 20440
    },
    {
      "epoch": 10.715408805031446,
      "grad_norm": 0.30623090267181396,
      "learning_rate": 2.6243536832720627e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 13369968,
      "step": 20445
    },
    {
      "epoch": 10.718029350104821,
      "grad_norm": 0.4027574062347412,
      "learning_rate": 2.623211657231553e-05,
      "loss": 0.4285,
      "num_input_tokens_seen": 13372784,
      "step": 20450
    },
    {
      "epoch": 10.720649895178196,
      "grad_norm": 0.4383350610733032,
      "learning_rate": 2.6220696054165704e-05,
      "loss": 0.5032,
      "num_input_tokens_seen": 13375984,
      "step": 20455
    },
    {
      "epoch": 10.723270440251572,
      "grad_norm": 0.3839549124240875,
      "learning_rate": 2.6209275280660194e-05,
      "loss": 0.4049,
      "num_input_tokens_seen": 13379120,
      "step": 20460
    },
    {
      "epoch": 10.725890985324948,
      "grad_norm": 0.6878730654716492,
      "learning_rate": 2.6197854254188085e-05,
      "loss": 0.4622,
      "num_input_tokens_seen": 13383664,
      "step": 20465
    },
    {
      "epoch": 10.728511530398324,
      "grad_norm": 0.6020787954330444,
      "learning_rate": 2.618643297713854e-05,
      "loss": 0.4567,
      "num_input_tokens_seen": 13386736,
      "step": 20470
    },
    {
      "epoch": 10.731132075471699,
      "grad_norm": 0.4216744005680084,
      "learning_rate": 2.617501145190075e-05,
      "loss": 0.421,
      "num_input_tokens_seen": 13390480,
      "step": 20475
    },
    {
      "epoch": 10.733752620545074,
      "grad_norm": 0.7597498297691345,
      "learning_rate": 2.616358968086397e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 13393680,
      "step": 20480
    },
    {
      "epoch": 10.73637316561845,
      "grad_norm": 0.5387913584709167,
      "learning_rate": 2.6152167666417493e-05,
      "loss": 0.3401,
      "num_input_tokens_seen": 13399760,
      "step": 20485
    },
    {
      "epoch": 10.738993710691824,
      "grad_norm": 0.43434661626815796,
      "learning_rate": 2.6140745410950694e-05,
      "loss": 0.5104,
      "num_input_tokens_seen": 13402064,
      "step": 20490
    },
    {
      "epoch": 10.7416142557652,
      "grad_norm": 0.513732373714447,
      "learning_rate": 2.6129322916852973e-05,
      "loss": 0.3401,
      "num_input_tokens_seen": 13405456,
      "step": 20495
    },
    {
      "epoch": 10.744234800838575,
      "grad_norm": 0.8315087556838989,
      "learning_rate": 2.611790018651377e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 13408368,
      "step": 20500
    },
    {
      "epoch": 10.74685534591195,
      "grad_norm": 0.47082993388175964,
      "learning_rate": 2.6106477222322613e-05,
      "loss": 0.4646,
      "num_input_tokens_seen": 13411504,
      "step": 20505
    },
    {
      "epoch": 10.749475890985325,
      "grad_norm": 0.4817207455635071,
      "learning_rate": 2.6095054026669037e-05,
      "loss": 0.5288,
      "num_input_tokens_seen": 13414672,
      "step": 20510
    },
    {
      "epoch": 10.7520964360587,
      "grad_norm": 0.7718202471733093,
      "learning_rate": 2.608363060194265e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 13417232,
      "step": 20515
    },
    {
      "epoch": 10.754716981132075,
      "grad_norm": 0.7229754328727722,
      "learning_rate": 2.6072206950533117e-05,
      "loss": 0.4472,
      "num_input_tokens_seen": 13420688,
      "step": 20520
    },
    {
      "epoch": 10.75733752620545,
      "grad_norm": 0.816056489944458,
      "learning_rate": 2.606078307483011e-05,
      "loss": 0.3219,
      "num_input_tokens_seen": 13423920,
      "step": 20525
    },
    {
      "epoch": 10.759958071278826,
      "grad_norm": 0.6184305548667908,
      "learning_rate": 2.6049358977223392e-05,
      "loss": 0.4395,
      "num_input_tokens_seen": 13426672,
      "step": 20530
    },
    {
      "epoch": 10.7625786163522,
      "grad_norm": 0.6306840777397156,
      "learning_rate": 2.6037934660102747e-05,
      "loss": 0.4485,
      "num_input_tokens_seen": 13429904,
      "step": 20535
    },
    {
      "epoch": 10.765199161425576,
      "grad_norm": 0.33044999837875366,
      "learning_rate": 2.6026510125858006e-05,
      "loss": 0.4682,
      "num_input_tokens_seen": 13433168,
      "step": 20540
    },
    {
      "epoch": 10.767819706498951,
      "grad_norm": 0.34245139360427856,
      "learning_rate": 2.6015085376879063e-05,
      "loss": 0.298,
      "num_input_tokens_seen": 13436624,
      "step": 20545
    },
    {
      "epoch": 10.770440251572326,
      "grad_norm": 0.40475893020629883,
      "learning_rate": 2.600366041555584e-05,
      "loss": 0.4542,
      "num_input_tokens_seen": 13439984,
      "step": 20550
    },
    {
      "epoch": 10.773060796645701,
      "grad_norm": 0.3033801019191742,
      "learning_rate": 2.5992235244278302e-05,
      "loss": 0.4873,
      "num_input_tokens_seen": 13443760,
      "step": 20555
    },
    {
      "epoch": 10.775681341719078,
      "grad_norm": 0.6811304092407227,
      "learning_rate": 2.5980809865436485e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 13446288,
      "step": 20560
    },
    {
      "epoch": 10.778301886792454,
      "grad_norm": 0.32374307513237,
      "learning_rate": 2.5969384281420424e-05,
      "loss": 0.4071,
      "num_input_tokens_seen": 13449168,
      "step": 20565
    },
    {
      "epoch": 10.780922431865829,
      "grad_norm": 1.0268089771270752,
      "learning_rate": 2.5957958494620227e-05,
      "loss": 0.538,
      "num_input_tokens_seen": 13452432,
      "step": 20570
    },
    {
      "epoch": 10.783542976939204,
      "grad_norm": 0.28521403670310974,
      "learning_rate": 2.594653250742605e-05,
      "loss": 0.4513,
      "num_input_tokens_seen": 13455728,
      "step": 20575
    },
    {
      "epoch": 10.786163522012579,
      "grad_norm": 0.5409173369407654,
      "learning_rate": 2.5935106322228065e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 13458832,
      "step": 20580
    },
    {
      "epoch": 10.788784067085954,
      "grad_norm": 0.41508030891418457,
      "learning_rate": 2.5923679941416514e-05,
      "loss": 0.6141,
      "num_input_tokens_seen": 13462928,
      "step": 20585
    },
    {
      "epoch": 10.79140461215933,
      "grad_norm": 0.2886795401573181,
      "learning_rate": 2.591225336738165e-05,
      "loss": 0.4882,
      "num_input_tokens_seen": 13465968,
      "step": 20590
    },
    {
      "epoch": 10.794025157232705,
      "grad_norm": 0.7637646794319153,
      "learning_rate": 2.5900826602513782e-05,
      "loss": 0.5154,
      "num_input_tokens_seen": 13469072,
      "step": 20595
    },
    {
      "epoch": 10.79664570230608,
      "grad_norm": 0.3488087058067322,
      "learning_rate": 2.5889399649203277e-05,
      "loss": 0.5204,
      "num_input_tokens_seen": 13472368,
      "step": 20600
    },
    {
      "epoch": 10.799266247379455,
      "grad_norm": 0.49020063877105713,
      "learning_rate": 2.5877972509840514e-05,
      "loss": 0.4039,
      "num_input_tokens_seen": 13475440,
      "step": 20605
    },
    {
      "epoch": 10.80188679245283,
      "grad_norm": 0.37052878737449646,
      "learning_rate": 2.5866545186815912e-05,
      "loss": 0.4948,
      "num_input_tokens_seen": 13478864,
      "step": 20610
    },
    {
      "epoch": 10.804507337526205,
      "grad_norm": 0.6787365078926086,
      "learning_rate": 2.5855117682519954e-05,
      "loss": 0.6594,
      "num_input_tokens_seen": 13482448,
      "step": 20615
    },
    {
      "epoch": 10.80712788259958,
      "grad_norm": 0.37823817133903503,
      "learning_rate": 2.584368999934312e-05,
      "loss": 0.3738,
      "num_input_tokens_seen": 13485776,
      "step": 20620
    },
    {
      "epoch": 10.809748427672956,
      "grad_norm": 0.3849793076515198,
      "learning_rate": 2.5832262139675966e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 13488848,
      "step": 20625
    },
    {
      "epoch": 10.81236897274633,
      "grad_norm": 0.28727349638938904,
      "learning_rate": 2.582083410590907e-05,
      "loss": 0.4034,
      "num_input_tokens_seen": 13492432,
      "step": 20630
    },
    {
      "epoch": 10.814989517819706,
      "grad_norm": 0.37992632389068604,
      "learning_rate": 2.5809405900433036e-05,
      "loss": 0.3969,
      "num_input_tokens_seen": 13496432,
      "step": 20635
    },
    {
      "epoch": 10.817610062893081,
      "grad_norm": 0.44167983531951904,
      "learning_rate": 2.579797752563853e-05,
      "loss": 0.4158,
      "num_input_tokens_seen": 13499408,
      "step": 20640
    },
    {
      "epoch": 10.820230607966456,
      "grad_norm": 0.4268491566181183,
      "learning_rate": 2.5786548983916232e-05,
      "loss": 0.5111,
      "num_input_tokens_seen": 13503312,
      "step": 20645
    },
    {
      "epoch": 10.822851153039831,
      "grad_norm": 0.32769840955734253,
      "learning_rate": 2.577512027765685e-05,
      "loss": 0.3323,
      "num_input_tokens_seen": 13506096,
      "step": 20650
    },
    {
      "epoch": 10.825471698113208,
      "grad_norm": 0.588828444480896,
      "learning_rate": 2.5763691409251155e-05,
      "loss": 0.3659,
      "num_input_tokens_seen": 13508688,
      "step": 20655
    },
    {
      "epoch": 10.828092243186584,
      "grad_norm": 0.48982110619544983,
      "learning_rate": 2.5752262381089932e-05,
      "loss": 0.3407,
      "num_input_tokens_seen": 13512112,
      "step": 20660
    },
    {
      "epoch": 10.830712788259959,
      "grad_norm": 0.7238737940788269,
      "learning_rate": 2.5740833195563996e-05,
      "loss": 0.5901,
      "num_input_tokens_seen": 13516304,
      "step": 20665
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 0.6292600035667419,
      "learning_rate": 2.572940385506421e-05,
      "loss": 0.4841,
      "num_input_tokens_seen": 13520368,
      "step": 20670
    },
    {
      "epoch": 10.835953878406709,
      "grad_norm": 0.31609877943992615,
      "learning_rate": 2.5717974361981457e-05,
      "loss": 0.4028,
      "num_input_tokens_seen": 13523600,
      "step": 20675
    },
    {
      "epoch": 10.838574423480084,
      "grad_norm": 0.4621676802635193,
      "learning_rate": 2.5706544718706665e-05,
      "loss": 0.4666,
      "num_input_tokens_seen": 13527088,
      "step": 20680
    },
    {
      "epoch": 10.84119496855346,
      "grad_norm": 0.37502631545066833,
      "learning_rate": 2.569511492763077e-05,
      "loss": 0.4418,
      "num_input_tokens_seen": 13530384,
      "step": 20685
    },
    {
      "epoch": 10.843815513626835,
      "grad_norm": 0.42345380783081055,
      "learning_rate": 2.5683684991144764e-05,
      "loss": 0.4054,
      "num_input_tokens_seen": 13533520,
      "step": 20690
    },
    {
      "epoch": 10.84643605870021,
      "grad_norm": 0.5175102949142456,
      "learning_rate": 2.5672254911639666e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 13536560,
      "step": 20695
    },
    {
      "epoch": 10.849056603773585,
      "grad_norm": 0.2819845676422119,
      "learning_rate": 2.5660824691506496e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 13539504,
      "step": 20700
    },
    {
      "epoch": 10.85167714884696,
      "grad_norm": 0.5019439458847046,
      "learning_rate": 2.5649394333136356e-05,
      "loss": 0.3749,
      "num_input_tokens_seen": 13542864,
      "step": 20705
    },
    {
      "epoch": 10.854297693920335,
      "grad_norm": 0.5211907625198364,
      "learning_rate": 2.5637963838920316e-05,
      "loss": 0.4724,
      "num_input_tokens_seen": 13545328,
      "step": 20710
    },
    {
      "epoch": 10.85691823899371,
      "grad_norm": 0.27591371536254883,
      "learning_rate": 2.5626533211249525e-05,
      "loss": 0.501,
      "num_input_tokens_seen": 13549072,
      "step": 20715
    },
    {
      "epoch": 10.859538784067086,
      "grad_norm": 0.5976735353469849,
      "learning_rate": 2.5615102452515138e-05,
      "loss": 0.3776,
      "num_input_tokens_seen": 13551760,
      "step": 20720
    },
    {
      "epoch": 10.86215932914046,
      "grad_norm": 0.7514457106590271,
      "learning_rate": 2.5603671565108328e-05,
      "loss": 0.4993,
      "num_input_tokens_seen": 13554896,
      "step": 20725
    },
    {
      "epoch": 10.864779874213836,
      "grad_norm": 0.6590315699577332,
      "learning_rate": 2.5592240551420317e-05,
      "loss": 0.4166,
      "num_input_tokens_seen": 13557712,
      "step": 20730
    },
    {
      "epoch": 10.867400419287211,
      "grad_norm": 0.46250128746032715,
      "learning_rate": 2.5580809413842343e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 13561296,
      "step": 20735
    },
    {
      "epoch": 10.870020964360586,
      "grad_norm": 0.5512210726737976,
      "learning_rate": 2.5569378154765655e-05,
      "loss": 0.5167,
      "num_input_tokens_seen": 13564368,
      "step": 20740
    },
    {
      "epoch": 10.872641509433961,
      "grad_norm": 0.621526837348938,
      "learning_rate": 2.5557946776581572e-05,
      "loss": 0.3932,
      "num_input_tokens_seen": 13566832,
      "step": 20745
    },
    {
      "epoch": 10.875262054507338,
      "grad_norm": 0.2942240238189697,
      "learning_rate": 2.554651528168137e-05,
      "loss": 0.477,
      "num_input_tokens_seen": 13570544,
      "step": 20750
    },
    {
      "epoch": 10.877882599580714,
      "grad_norm": 0.4897836446762085,
      "learning_rate": 2.5535083672456407e-05,
      "loss": 0.5686,
      "num_input_tokens_seen": 13575376,
      "step": 20755
    },
    {
      "epoch": 10.880503144654089,
      "grad_norm": 0.34628957509994507,
      "learning_rate": 2.5523651951298055e-05,
      "loss": 0.4149,
      "num_input_tokens_seen": 13578320,
      "step": 20760
    },
    {
      "epoch": 10.883123689727464,
      "grad_norm": 0.8457726836204529,
      "learning_rate": 2.5512220120597686e-05,
      "loss": 0.4175,
      "num_input_tokens_seen": 13581520,
      "step": 20765
    },
    {
      "epoch": 10.885744234800839,
      "grad_norm": 0.6841984987258911,
      "learning_rate": 2.550078818274671e-05,
      "loss": 0.5384,
      "num_input_tokens_seen": 13584304,
      "step": 20770
    },
    {
      "epoch": 10.888364779874214,
      "grad_norm": 0.42335933446884155,
      "learning_rate": 2.548935614013656e-05,
      "loss": 0.387,
      "num_input_tokens_seen": 13587376,
      "step": 20775
    },
    {
      "epoch": 10.89098532494759,
      "grad_norm": 0.3548223674297333,
      "learning_rate": 2.5477923995158683e-05,
      "loss": 0.5007,
      "num_input_tokens_seen": 13590160,
      "step": 20780
    },
    {
      "epoch": 10.893605870020965,
      "grad_norm": 0.7981169819831848,
      "learning_rate": 2.5466491750204562e-05,
      "loss": 0.4777,
      "num_input_tokens_seen": 13593328,
      "step": 20785
    },
    {
      "epoch": 10.89622641509434,
      "grad_norm": 0.5025973916053772,
      "learning_rate": 2.5455059407665693e-05,
      "loss": 0.4298,
      "num_input_tokens_seen": 13596496,
      "step": 20790
    },
    {
      "epoch": 10.898846960167715,
      "grad_norm": 0.40858256816864014,
      "learning_rate": 2.544362696993358e-05,
      "loss": 0.4591,
      "num_input_tokens_seen": 13599728,
      "step": 20795
    },
    {
      "epoch": 10.90146750524109,
      "grad_norm": 0.6425467729568481,
      "learning_rate": 2.5432194439399775e-05,
      "loss": 0.5035,
      "num_input_tokens_seen": 13602544,
      "step": 20800
    },
    {
      "epoch": 10.904088050314465,
      "grad_norm": 0.848822295665741,
      "learning_rate": 2.542076181845582e-05,
      "loss": 0.5726,
      "num_input_tokens_seen": 13605360,
      "step": 20805
    },
    {
      "epoch": 10.90670859538784,
      "grad_norm": 0.48063549399375916,
      "learning_rate": 2.540932910949329e-05,
      "loss": 0.4117,
      "num_input_tokens_seen": 13608272,
      "step": 20810
    },
    {
      "epoch": 10.909329140461216,
      "grad_norm": 0.5149195194244385,
      "learning_rate": 2.5397896314903785e-05,
      "loss": 0.3268,
      "num_input_tokens_seen": 13611504,
      "step": 20815
    },
    {
      "epoch": 10.91194968553459,
      "grad_norm": 0.33193495869636536,
      "learning_rate": 2.5386463437078905e-05,
      "loss": 0.3662,
      "num_input_tokens_seen": 13615024,
      "step": 20820
    },
    {
      "epoch": 10.914570230607966,
      "grad_norm": 0.5415549874305725,
      "learning_rate": 2.5375030478410282e-05,
      "loss": 0.405,
      "num_input_tokens_seen": 13618320,
      "step": 20825
    },
    {
      "epoch": 10.917190775681341,
      "grad_norm": 0.4079395830631256,
      "learning_rate": 2.536359744128957e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 13621488,
      "step": 20830
    },
    {
      "epoch": 10.919811320754716,
      "grad_norm": 0.5950683951377869,
      "learning_rate": 2.5352164328108413e-05,
      "loss": 0.491,
      "num_input_tokens_seen": 13624624,
      "step": 20835
    },
    {
      "epoch": 10.922431865828091,
      "grad_norm": 0.4509585201740265,
      "learning_rate": 2.53407311412585e-05,
      "loss": 0.4472,
      "num_input_tokens_seen": 13627408,
      "step": 20840
    },
    {
      "epoch": 10.925052410901468,
      "grad_norm": 0.855510413646698,
      "learning_rate": 2.5329297883131526e-05,
      "loss": 0.4674,
      "num_input_tokens_seen": 13630160,
      "step": 20845
    },
    {
      "epoch": 10.927672955974844,
      "grad_norm": 0.3713502585887909,
      "learning_rate": 2.5317864556119182e-05,
      "loss": 0.7703,
      "num_input_tokens_seen": 13634320,
      "step": 20850
    },
    {
      "epoch": 10.930293501048219,
      "grad_norm": 0.3407402038574219,
      "learning_rate": 2.53064311626132e-05,
      "loss": 0.4344,
      "num_input_tokens_seen": 13638608,
      "step": 20855
    },
    {
      "epoch": 10.932914046121594,
      "grad_norm": 0.4083385765552521,
      "learning_rate": 2.5294997705005318e-05,
      "loss": 0.3527,
      "num_input_tokens_seen": 13641616,
      "step": 20860
    },
    {
      "epoch": 10.935534591194969,
      "grad_norm": 0.40833020210266113,
      "learning_rate": 2.5283564185687282e-05,
      "loss": 0.5009,
      "num_input_tokens_seen": 13644496,
      "step": 20865
    },
    {
      "epoch": 10.938155136268344,
      "grad_norm": 0.27665114402770996,
      "learning_rate": 2.5272130607050848e-05,
      "loss": 0.555,
      "num_input_tokens_seen": 13647696,
      "step": 20870
    },
    {
      "epoch": 10.94077568134172,
      "grad_norm": 0.5295302867889404,
      "learning_rate": 2.5260696971487802e-05,
      "loss": 0.4743,
      "num_input_tokens_seen": 13650256,
      "step": 20875
    },
    {
      "epoch": 10.943396226415095,
      "grad_norm": 0.44904375076293945,
      "learning_rate": 2.5249263281389923e-05,
      "loss": 0.5384,
      "num_input_tokens_seen": 13652816,
      "step": 20880
    },
    {
      "epoch": 10.94601677148847,
      "grad_norm": 0.37440225481987,
      "learning_rate": 2.5237829539149004e-05,
      "loss": 0.3883,
      "num_input_tokens_seen": 13655216,
      "step": 20885
    },
    {
      "epoch": 10.948637316561845,
      "grad_norm": 0.27121302485466003,
      "learning_rate": 2.5226395747156866e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 13658704,
      "step": 20890
    },
    {
      "epoch": 10.95125786163522,
      "grad_norm": 0.520858645439148,
      "learning_rate": 2.5214961907805317e-05,
      "loss": 0.5803,
      "num_input_tokens_seen": 13661936,
      "step": 20895
    },
    {
      "epoch": 10.953878406708595,
      "grad_norm": 1.077667474746704,
      "learning_rate": 2.520352802348619e-05,
      "loss": 0.4393,
      "num_input_tokens_seen": 13665136,
      "step": 20900
    },
    {
      "epoch": 10.95649895178197,
      "grad_norm": 1.414362907409668,
      "learning_rate": 2.5192094096591323e-05,
      "loss": 0.3514,
      "num_input_tokens_seen": 13668112,
      "step": 20905
    },
    {
      "epoch": 10.959119496855346,
      "grad_norm": 0.4729622006416321,
      "learning_rate": 2.5180660129512556e-05,
      "loss": 0.4111,
      "num_input_tokens_seen": 13671344,
      "step": 20910
    },
    {
      "epoch": 10.96174004192872,
      "grad_norm": 0.4035741686820984,
      "learning_rate": 2.516922612464176e-05,
      "loss": 0.5981,
      "num_input_tokens_seen": 13674608,
      "step": 20915
    },
    {
      "epoch": 10.964360587002096,
      "grad_norm": 0.518516480922699,
      "learning_rate": 2.5157792084370783e-05,
      "loss": 0.4078,
      "num_input_tokens_seen": 13678000,
      "step": 20920
    },
    {
      "epoch": 10.966981132075471,
      "grad_norm": 0.8360491394996643,
      "learning_rate": 2.5146358011091502e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 13681008,
      "step": 20925
    },
    {
      "epoch": 10.969601677148846,
      "grad_norm": 0.46915069222450256,
      "learning_rate": 2.5134923907195805e-05,
      "loss": 0.5034,
      "num_input_tokens_seen": 13684656,
      "step": 20930
    },
    {
      "epoch": 10.972222222222221,
      "grad_norm": 0.6053932905197144,
      "learning_rate": 2.5123489775075558e-05,
      "loss": 0.5438,
      "num_input_tokens_seen": 13688016,
      "step": 20935
    },
    {
      "epoch": 10.974842767295598,
      "grad_norm": 0.4236968159675598,
      "learning_rate": 2.511205561712266e-05,
      "loss": 0.4668,
      "num_input_tokens_seen": 13691344,
      "step": 20940
    },
    {
      "epoch": 10.977463312368974,
      "grad_norm": 0.4937721788883209,
      "learning_rate": 2.5100621435729004e-05,
      "loss": 0.6901,
      "num_input_tokens_seen": 13694768,
      "step": 20945
    },
    {
      "epoch": 10.980083857442349,
      "grad_norm": 0.392834335565567,
      "learning_rate": 2.5089187233286492e-05,
      "loss": 0.4141,
      "num_input_tokens_seen": 13698224,
      "step": 20950
    },
    {
      "epoch": 10.982704402515724,
      "grad_norm": 0.4953322112560272,
      "learning_rate": 2.5077753012187026e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 13701008,
      "step": 20955
    },
    {
      "epoch": 10.985324947589099,
      "grad_norm": 0.5953294634819031,
      "learning_rate": 2.506631877482252e-05,
      "loss": 0.6378,
      "num_input_tokens_seen": 13703920,
      "step": 20960
    },
    {
      "epoch": 10.987945492662474,
      "grad_norm": 0.9713189601898193,
      "learning_rate": 2.505488452358488e-05,
      "loss": 0.4903,
      "num_input_tokens_seen": 13706448,
      "step": 20965
    },
    {
      "epoch": 10.99056603773585,
      "grad_norm": 0.3616548776626587,
      "learning_rate": 2.504345026086603e-05,
      "loss": 0.5121,
      "num_input_tokens_seen": 13710128,
      "step": 20970
    },
    {
      "epoch": 10.993186582809225,
      "grad_norm": 0.4913422465324402,
      "learning_rate": 2.5032015989057877e-05,
      "loss": 0.3546,
      "num_input_tokens_seen": 13712880,
      "step": 20975
    },
    {
      "epoch": 10.9958071278826,
      "grad_norm": 0.9039394855499268,
      "learning_rate": 2.5020581710552348e-05,
      "loss": 0.4694,
      "num_input_tokens_seen": 13716336,
      "step": 20980
    },
    {
      "epoch": 10.998427672955975,
      "grad_norm": 0.6038282513618469,
      "learning_rate": 2.5009147427741358e-05,
      "loss": 0.3451,
      "num_input_tokens_seen": 13719088,
      "step": 20985
    },
    {
      "epoch": 11.0,
      "eval_loss": 0.4773971736431122,
      "eval_runtime": 9.6135,
      "eval_samples_per_second": 88.209,
      "eval_steps_per_second": 22.052,
      "num_input_tokens_seen": 13719896,
      "step": 20988
    },
    {
      "epoch": 11.00104821802935,
      "grad_norm": 0.33636021614074707,
      "learning_rate": 2.4997713143016822e-05,
      "loss": 0.5495,
      "num_input_tokens_seen": 13722872,
      "step": 20990
    },
    {
      "epoch": 11.003668763102725,
      "grad_norm": 0.26902368664741516,
      "learning_rate": 2.498627885877068e-05,
      "loss": 0.4291,
      "num_input_tokens_seen": 13729624,
      "step": 20995
    },
    {
      "epoch": 11.0062893081761,
      "grad_norm": 0.5192081928253174,
      "learning_rate": 2.4974844577394846e-05,
      "loss": 0.4131,
      "num_input_tokens_seen": 13731896,
      "step": 21000
    },
    {
      "epoch": 11.008909853249476,
      "grad_norm": 0.4393714368343353,
      "learning_rate": 2.496341030128123e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 13735160,
      "step": 21005
    },
    {
      "epoch": 11.01153039832285,
      "grad_norm": 0.48288702964782715,
      "learning_rate": 2.495197603282177e-05,
      "loss": 0.3898,
      "num_input_tokens_seen": 13738488,
      "step": 21010
    },
    {
      "epoch": 11.014150943396226,
      "grad_norm": 0.42887917160987854,
      "learning_rate": 2.4940541774408368e-05,
      "loss": 0.4451,
      "num_input_tokens_seen": 13741752,
      "step": 21015
    },
    {
      "epoch": 11.016771488469601,
      "grad_norm": 0.29952049255371094,
      "learning_rate": 2.4929107528432953e-05,
      "loss": 0.3113,
      "num_input_tokens_seen": 13744248,
      "step": 21020
    },
    {
      "epoch": 11.019392033542976,
      "grad_norm": 0.7082865834236145,
      "learning_rate": 2.4917673297287446e-05,
      "loss": 0.4634,
      "num_input_tokens_seen": 13748056,
      "step": 21025
    },
    {
      "epoch": 11.022012578616351,
      "grad_norm": 0.3520447015762329,
      "learning_rate": 2.4906239083363728e-05,
      "loss": 0.4734,
      "num_input_tokens_seen": 13751512,
      "step": 21030
    },
    {
      "epoch": 11.024633123689727,
      "grad_norm": 0.7883553504943848,
      "learning_rate": 2.4894804889053732e-05,
      "loss": 0.5525,
      "num_input_tokens_seen": 13755384,
      "step": 21035
    },
    {
      "epoch": 11.027253668763104,
      "grad_norm": 0.6889050006866455,
      "learning_rate": 2.4883370716749347e-05,
      "loss": 0.5037,
      "num_input_tokens_seen": 13758904,
      "step": 21040
    },
    {
      "epoch": 11.029874213836479,
      "grad_norm": 0.3885016441345215,
      "learning_rate": 2.4871936568842486e-05,
      "loss": 0.49,
      "num_input_tokens_seen": 13762712,
      "step": 21045
    },
    {
      "epoch": 11.032494758909854,
      "grad_norm": 0.5130318403244019,
      "learning_rate": 2.4860502447725033e-05,
      "loss": 0.4744,
      "num_input_tokens_seen": 13765784,
      "step": 21050
    },
    {
      "epoch": 11.035115303983229,
      "grad_norm": 0.32735925912857056,
      "learning_rate": 2.4849068355788873e-05,
      "loss": 0.4527,
      "num_input_tokens_seen": 13769304,
      "step": 21055
    },
    {
      "epoch": 11.037735849056604,
      "grad_norm": 0.5422608852386475,
      "learning_rate": 2.4837634295425892e-05,
      "loss": 0.5303,
      "num_input_tokens_seen": 13772920,
      "step": 21060
    },
    {
      "epoch": 11.04035639412998,
      "grad_norm": 0.4611113369464874,
      "learning_rate": 2.4826200269027957e-05,
      "loss": 0.4449,
      "num_input_tokens_seen": 13776088,
      "step": 21065
    },
    {
      "epoch": 11.042976939203355,
      "grad_norm": 0.2808421552181244,
      "learning_rate": 2.481476627898695e-05,
      "loss": 0.3748,
      "num_input_tokens_seen": 13780824,
      "step": 21070
    },
    {
      "epoch": 11.04559748427673,
      "grad_norm": 0.26677295565605164,
      "learning_rate": 2.4803332327694724e-05,
      "loss": 0.4362,
      "num_input_tokens_seen": 13784696,
      "step": 21075
    },
    {
      "epoch": 11.048218029350105,
      "grad_norm": 0.21073096990585327,
      "learning_rate": 2.4791898417543136e-05,
      "loss": 0.3201,
      "num_input_tokens_seen": 13789432,
      "step": 21080
    },
    {
      "epoch": 11.05083857442348,
      "grad_norm": 0.5325319170951843,
      "learning_rate": 2.4780464550924014e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 13792632,
      "step": 21085
    },
    {
      "epoch": 11.053459119496855,
      "grad_norm": 0.6491798162460327,
      "learning_rate": 2.4769030730229213e-05,
      "loss": 0.467,
      "num_input_tokens_seen": 13795640,
      "step": 21090
    },
    {
      "epoch": 11.05607966457023,
      "grad_norm": 0.4738463759422302,
      "learning_rate": 2.475759695785054e-05,
      "loss": 0.3469,
      "num_input_tokens_seen": 13798680,
      "step": 21095
    },
    {
      "epoch": 11.058700209643606,
      "grad_norm": 0.8213015794754028,
      "learning_rate": 2.4746163236179822e-05,
      "loss": 0.4493,
      "num_input_tokens_seen": 13801272,
      "step": 21100
    },
    {
      "epoch": 11.06132075471698,
      "grad_norm": 0.3420228958129883,
      "learning_rate": 2.4734729567608858e-05,
      "loss": 0.3798,
      "num_input_tokens_seen": 13804376,
      "step": 21105
    },
    {
      "epoch": 11.063941299790356,
      "grad_norm": 0.4033488631248474,
      "learning_rate": 2.472329595452945e-05,
      "loss": 0.3921,
      "num_input_tokens_seen": 13809272,
      "step": 21110
    },
    {
      "epoch": 11.066561844863731,
      "grad_norm": 0.8658339381217957,
      "learning_rate": 2.4711862399333367e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 13812312,
      "step": 21115
    },
    {
      "epoch": 11.069182389937106,
      "grad_norm": 0.45885512232780457,
      "learning_rate": 2.4700428904412377e-05,
      "loss": 0.4726,
      "num_input_tokens_seen": 13815096,
      "step": 21120
    },
    {
      "epoch": 11.071802935010481,
      "grad_norm": 0.4879235327243805,
      "learning_rate": 2.4688995472158245e-05,
      "loss": 0.5333,
      "num_input_tokens_seen": 13818424,
      "step": 21125
    },
    {
      "epoch": 11.074423480083857,
      "grad_norm": 0.37153732776641846,
      "learning_rate": 2.4677562104962712e-05,
      "loss": 0.422,
      "num_input_tokens_seen": 13821656,
      "step": 21130
    },
    {
      "epoch": 11.077044025157234,
      "grad_norm": 0.5228050947189331,
      "learning_rate": 2.4666128805217513e-05,
      "loss": 0.3798,
      "num_input_tokens_seen": 13824376,
      "step": 21135
    },
    {
      "epoch": 11.079664570230609,
      "grad_norm": 0.731954038143158,
      "learning_rate": 2.465469557531436e-05,
      "loss": 0.5735,
      "num_input_tokens_seen": 13827320,
      "step": 21140
    },
    {
      "epoch": 11.082285115303984,
      "grad_norm": 0.5326417684555054,
      "learning_rate": 2.4643262417644944e-05,
      "loss": 0.3998,
      "num_input_tokens_seen": 13829912,
      "step": 21145
    },
    {
      "epoch": 11.084905660377359,
      "grad_norm": 0.4616711437702179,
      "learning_rate": 2.463182933460096e-05,
      "loss": 0.3698,
      "num_input_tokens_seen": 13832536,
      "step": 21150
    },
    {
      "epoch": 11.087526205450734,
      "grad_norm": 0.3937532603740692,
      "learning_rate": 2.4620396328574087e-05,
      "loss": 0.3504,
      "num_input_tokens_seen": 13836088,
      "step": 21155
    },
    {
      "epoch": 11.09014675052411,
      "grad_norm": 0.783228874206543,
      "learning_rate": 2.4608963401955965e-05,
      "loss": 0.5389,
      "num_input_tokens_seen": 13839160,
      "step": 21160
    },
    {
      "epoch": 11.092767295597485,
      "grad_norm": 0.49876269698143005,
      "learning_rate": 2.4597530557138247e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 13841560,
      "step": 21165
    },
    {
      "epoch": 11.09538784067086,
      "grad_norm": 0.7118752002716064,
      "learning_rate": 2.4586097796512538e-05,
      "loss": 0.4401,
      "num_input_tokens_seen": 13844248,
      "step": 21170
    },
    {
      "epoch": 11.098008385744235,
      "grad_norm": 0.3455096483230591,
      "learning_rate": 2.457466512247045e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 13847960,
      "step": 21175
    },
    {
      "epoch": 11.10062893081761,
      "grad_norm": 0.3848511576652527,
      "learning_rate": 2.4563232537403563e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 13851000,
      "step": 21180
    },
    {
      "epoch": 11.103249475890985,
      "grad_norm": 0.8917334675788879,
      "learning_rate": 2.4551800043703442e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 13853304,
      "step": 21185
    },
    {
      "epoch": 11.10587002096436,
      "grad_norm": 0.31653574109077454,
      "learning_rate": 2.454036764376165e-05,
      "loss": 0.3947,
      "num_input_tokens_seen": 13856344,
      "step": 21190
    },
    {
      "epoch": 11.108490566037736,
      "grad_norm": 0.6251389980316162,
      "learning_rate": 2.4528935339969705e-05,
      "loss": 0.5018,
      "num_input_tokens_seen": 13859672,
      "step": 21195
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 0.4737272262573242,
      "learning_rate": 2.45175031347191e-05,
      "loss": 0.6635,
      "num_input_tokens_seen": 13862328,
      "step": 21200
    },
    {
      "epoch": 11.113731656184486,
      "grad_norm": 0.8406775593757629,
      "learning_rate": 2.4506071030401344e-05,
      "loss": 0.3922,
      "num_input_tokens_seen": 13865592,
      "step": 21205
    },
    {
      "epoch": 11.116352201257861,
      "grad_norm": 0.39908382296562195,
      "learning_rate": 2.4494639029407884e-05,
      "loss": 0.5919,
      "num_input_tokens_seen": 13868920,
      "step": 21210
    },
    {
      "epoch": 11.118972746331236,
      "grad_norm": 0.693001925945282,
      "learning_rate": 2.448320713413018e-05,
      "loss": 0.5016,
      "num_input_tokens_seen": 13872024,
      "step": 21215
    },
    {
      "epoch": 11.121593291404611,
      "grad_norm": 0.5275025963783264,
      "learning_rate": 2.4471775346959654e-05,
      "loss": 0.3658,
      "num_input_tokens_seen": 13875352,
      "step": 21220
    },
    {
      "epoch": 11.124213836477987,
      "grad_norm": 0.5101861357688904,
      "learning_rate": 2.446034367028769e-05,
      "loss": 0.3524,
      "num_input_tokens_seen": 13879672,
      "step": 21225
    },
    {
      "epoch": 11.126834381551364,
      "grad_norm": 0.5973419547080994,
      "learning_rate": 2.444891210650567e-05,
      "loss": 0.3524,
      "num_input_tokens_seen": 13882552,
      "step": 21230
    },
    {
      "epoch": 11.129454926624739,
      "grad_norm": 0.3464837074279785,
      "learning_rate": 2.4437480658004958e-05,
      "loss": 0.51,
      "num_input_tokens_seen": 13887928,
      "step": 21235
    },
    {
      "epoch": 11.132075471698114,
      "grad_norm": 0.7078356146812439,
      "learning_rate": 2.4426049327176868e-05,
      "loss": 0.4473,
      "num_input_tokens_seen": 13890808,
      "step": 21240
    },
    {
      "epoch": 11.134696016771489,
      "grad_norm": 0.30889561772346497,
      "learning_rate": 2.4414618116412716e-05,
      "loss": 0.5265,
      "num_input_tokens_seen": 13894520,
      "step": 21245
    },
    {
      "epoch": 11.137316561844864,
      "grad_norm": 0.3040284812450409,
      "learning_rate": 2.440318702810378e-05,
      "loss": 0.3569,
      "num_input_tokens_seen": 13897976,
      "step": 21250
    },
    {
      "epoch": 11.13993710691824,
      "grad_norm": 0.4059452712535858,
      "learning_rate": 2.43917560646413e-05,
      "loss": 0.5005,
      "num_input_tokens_seen": 13901528,
      "step": 21255
    },
    {
      "epoch": 11.142557651991615,
      "grad_norm": 0.4263049364089966,
      "learning_rate": 2.4380325228416518e-05,
      "loss": 0.4676,
      "num_input_tokens_seen": 13905272,
      "step": 21260
    },
    {
      "epoch": 11.14517819706499,
      "grad_norm": 0.5073153376579285,
      "learning_rate": 2.4368894521820624e-05,
      "loss": 0.3972,
      "num_input_tokens_seen": 13908760,
      "step": 21265
    },
    {
      "epoch": 11.147798742138365,
      "grad_norm": 0.48368433117866516,
      "learning_rate": 2.43574639472448e-05,
      "loss": 0.393,
      "num_input_tokens_seen": 13912312,
      "step": 21270
    },
    {
      "epoch": 11.15041928721174,
      "grad_norm": 0.3421502709388733,
      "learning_rate": 2.4346033507080178e-05,
      "loss": 0.499,
      "num_input_tokens_seen": 13914712,
      "step": 21275
    },
    {
      "epoch": 11.153039832285115,
      "grad_norm": 0.1996377557516098,
      "learning_rate": 2.4334603203717903e-05,
      "loss": 0.4667,
      "num_input_tokens_seen": 13917784,
      "step": 21280
    },
    {
      "epoch": 11.15566037735849,
      "grad_norm": 0.6820748448371887,
      "learning_rate": 2.432317303954904e-05,
      "loss": 0.51,
      "num_input_tokens_seen": 13920760,
      "step": 21285
    },
    {
      "epoch": 11.158280922431866,
      "grad_norm": 0.4862842261791229,
      "learning_rate": 2.4311743016964648e-05,
      "loss": 0.4209,
      "num_input_tokens_seen": 13923864,
      "step": 21290
    },
    {
      "epoch": 11.16090146750524,
      "grad_norm": 0.5450080633163452,
      "learning_rate": 2.4300313138355772e-05,
      "loss": 0.4334,
      "num_input_tokens_seen": 13926616,
      "step": 21295
    },
    {
      "epoch": 11.163522012578616,
      "grad_norm": 0.3369053900241852,
      "learning_rate": 2.4288883406113396e-05,
      "loss": 0.578,
      "num_input_tokens_seen": 13931128,
      "step": 21300
    },
    {
      "epoch": 11.166142557651991,
      "grad_norm": 0.4835279583930969,
      "learning_rate": 2.4277453822628512e-05,
      "loss": 0.3323,
      "num_input_tokens_seen": 13934616,
      "step": 21305
    },
    {
      "epoch": 11.168763102725366,
      "grad_norm": 0.3487063944339752,
      "learning_rate": 2.4266024390292037e-05,
      "loss": 0.381,
      "num_input_tokens_seen": 13936952,
      "step": 21310
    },
    {
      "epoch": 11.171383647798741,
      "grad_norm": 0.4965004324913025,
      "learning_rate": 2.425459511149489e-05,
      "loss": 0.6751,
      "num_input_tokens_seen": 13940376,
      "step": 21315
    },
    {
      "epoch": 11.174004192872117,
      "grad_norm": 0.5764280557632446,
      "learning_rate": 2.424316598862793e-05,
      "loss": 0.5576,
      "num_input_tokens_seen": 13943000,
      "step": 21320
    },
    {
      "epoch": 11.176624737945493,
      "grad_norm": 0.7789319753646851,
      "learning_rate": 2.4231737024082023e-05,
      "loss": 0.3854,
      "num_input_tokens_seen": 13946776,
      "step": 21325
    },
    {
      "epoch": 11.179245283018869,
      "grad_norm": 0.43586912751197815,
      "learning_rate": 2.4220308220247954e-05,
      "loss": 0.5029,
      "num_input_tokens_seen": 13949112,
      "step": 21330
    },
    {
      "epoch": 11.181865828092244,
      "grad_norm": 0.46005192399024963,
      "learning_rate": 2.4208879579516528e-05,
      "loss": 0.4802,
      "num_input_tokens_seen": 13951896,
      "step": 21335
    },
    {
      "epoch": 11.184486373165619,
      "grad_norm": 0.4210183322429657,
      "learning_rate": 2.419745110427846e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 13955672,
      "step": 21340
    },
    {
      "epoch": 11.187106918238994,
      "grad_norm": 0.3773685097694397,
      "learning_rate": 2.418602279692446e-05,
      "loss": 0.4007,
      "num_input_tokens_seen": 13959512,
      "step": 21345
    },
    {
      "epoch": 11.18972746331237,
      "grad_norm": 0.5740013122558594,
      "learning_rate": 2.417459465984521e-05,
      "loss": 0.5706,
      "num_input_tokens_seen": 13963768,
      "step": 21350
    },
    {
      "epoch": 11.192348008385745,
      "grad_norm": 0.44336560368537903,
      "learning_rate": 2.4163166695431336e-05,
      "loss": 0.3833,
      "num_input_tokens_seen": 13967032,
      "step": 21355
    },
    {
      "epoch": 11.19496855345912,
      "grad_norm": 0.44068795442581177,
      "learning_rate": 2.4151738906073446e-05,
      "loss": 0.3608,
      "num_input_tokens_seen": 13970648,
      "step": 21360
    },
    {
      "epoch": 11.197589098532495,
      "grad_norm": 0.3706226050853729,
      "learning_rate": 2.4140311294162105e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 13973080,
      "step": 21365
    },
    {
      "epoch": 11.20020964360587,
      "grad_norm": 0.6388086676597595,
      "learning_rate": 2.412888386208782e-05,
      "loss": 0.477,
      "num_input_tokens_seen": 13976696,
      "step": 21370
    },
    {
      "epoch": 11.202830188679245,
      "grad_norm": 0.4801539480686188,
      "learning_rate": 2.4117456612241103e-05,
      "loss": 0.3692,
      "num_input_tokens_seen": 13980248,
      "step": 21375
    },
    {
      "epoch": 11.20545073375262,
      "grad_norm": 0.4872561991214752,
      "learning_rate": 2.410602954701239e-05,
      "loss": 0.5392,
      "num_input_tokens_seen": 13983320,
      "step": 21380
    },
    {
      "epoch": 11.208071278825996,
      "grad_norm": 0.27688515186309814,
      "learning_rate": 2.409460266879209e-05,
      "loss": 0.4409,
      "num_input_tokens_seen": 13986680,
      "step": 21385
    },
    {
      "epoch": 11.21069182389937,
      "grad_norm": 0.27723070979118347,
      "learning_rate": 2.4083175979970592e-05,
      "loss": 0.4231,
      "num_input_tokens_seen": 13992504,
      "step": 21390
    },
    {
      "epoch": 11.213312368972746,
      "grad_norm": 0.47152233123779297,
      "learning_rate": 2.4071749482938226e-05,
      "loss": 0.4289,
      "num_input_tokens_seen": 13995416,
      "step": 21395
    },
    {
      "epoch": 11.215932914046121,
      "grad_norm": 0.30534103512763977,
      "learning_rate": 2.4060323180085266e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 13999320,
      "step": 21400
    },
    {
      "epoch": 11.218553459119496,
      "grad_norm": 0.8953856229782104,
      "learning_rate": 2.404889707380198e-05,
      "loss": 0.5275,
      "num_input_tokens_seen": 14002552,
      "step": 21405
    },
    {
      "epoch": 11.221174004192871,
      "grad_norm": 0.6850642561912537,
      "learning_rate": 2.403747116647857e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 14006040,
      "step": 21410
    },
    {
      "epoch": 11.223794549266247,
      "grad_norm": 0.5140559077262878,
      "learning_rate": 2.4026045460505216e-05,
      "loss": 0.4506,
      "num_input_tokens_seen": 14009400,
      "step": 21415
    },
    {
      "epoch": 11.226415094339623,
      "grad_norm": 0.5254111886024475,
      "learning_rate": 2.4014619958272047e-05,
      "loss": 0.4423,
      "num_input_tokens_seen": 14012824,
      "step": 21420
    },
    {
      "epoch": 11.229035639412999,
      "grad_norm": 0.5175868272781372,
      "learning_rate": 2.4003194662169132e-05,
      "loss": 0.4579,
      "num_input_tokens_seen": 14016792,
      "step": 21425
    },
    {
      "epoch": 11.231656184486374,
      "grad_norm": 0.3048422038555145,
      "learning_rate": 2.3991769574586522e-05,
      "loss": 0.476,
      "num_input_tokens_seen": 14019864,
      "step": 21430
    },
    {
      "epoch": 11.234276729559749,
      "grad_norm": 0.5877376198768616,
      "learning_rate": 2.3980344697914215e-05,
      "loss": 0.4358,
      "num_input_tokens_seen": 14022552,
      "step": 21435
    },
    {
      "epoch": 11.236897274633124,
      "grad_norm": 0.905091404914856,
      "learning_rate": 2.3968920034542168e-05,
      "loss": 0.5045,
      "num_input_tokens_seen": 14025592,
      "step": 21440
    },
    {
      "epoch": 11.2395178197065,
      "grad_norm": 0.4249229431152344,
      "learning_rate": 2.3957495586860278e-05,
      "loss": 0.4682,
      "num_input_tokens_seen": 14029720,
      "step": 21445
    },
    {
      "epoch": 11.242138364779874,
      "grad_norm": 0.35834744572639465,
      "learning_rate": 2.394607135725843e-05,
      "loss": 0.4141,
      "num_input_tokens_seen": 14032824,
      "step": 21450
    },
    {
      "epoch": 11.24475890985325,
      "grad_norm": 0.26528263092041016,
      "learning_rate": 2.3934647348126426e-05,
      "loss": 0.3147,
      "num_input_tokens_seen": 14035352,
      "step": 21455
    },
    {
      "epoch": 11.247379454926625,
      "grad_norm": 0.5056616067886353,
      "learning_rate": 2.3923223561854035e-05,
      "loss": 0.3653,
      "num_input_tokens_seen": 14039928,
      "step": 21460
    },
    {
      "epoch": 11.25,
      "grad_norm": 0.9099351763725281,
      "learning_rate": 2.391180000083099e-05,
      "loss": 0.4488,
      "num_input_tokens_seen": 14042296,
      "step": 21465
    },
    {
      "epoch": 11.252620545073375,
      "grad_norm": 0.559449315071106,
      "learning_rate": 2.390037666744698e-05,
      "loss": 0.5036,
      "num_input_tokens_seen": 14045624,
      "step": 21470
    },
    {
      "epoch": 11.25524109014675,
      "grad_norm": 0.5450436472892761,
      "learning_rate": 2.388895356409162e-05,
      "loss": 0.3048,
      "num_input_tokens_seen": 14048248,
      "step": 21475
    },
    {
      "epoch": 11.257861635220126,
      "grad_norm": 0.3570950925350189,
      "learning_rate": 2.3877530693154506e-05,
      "loss": 0.4051,
      "num_input_tokens_seen": 14050904,
      "step": 21480
    },
    {
      "epoch": 11.2604821802935,
      "grad_norm": 0.9092350602149963,
      "learning_rate": 2.3866108057025162e-05,
      "loss": 0.3807,
      "num_input_tokens_seen": 14054424,
      "step": 21485
    },
    {
      "epoch": 11.263102725366876,
      "grad_norm": 0.3172011971473694,
      "learning_rate": 2.385468565809307e-05,
      "loss": 0.6356,
      "num_input_tokens_seen": 14057592,
      "step": 21490
    },
    {
      "epoch": 11.265723270440251,
      "grad_norm": 0.11552801728248596,
      "learning_rate": 2.3843263498747675e-05,
      "loss": 0.4326,
      "num_input_tokens_seen": 14063160,
      "step": 21495
    },
    {
      "epoch": 11.268343815513626,
      "grad_norm": 0.4663749635219574,
      "learning_rate": 2.383184158137836e-05,
      "loss": 0.5212,
      "num_input_tokens_seen": 14065880,
      "step": 21500
    },
    {
      "epoch": 11.270964360587001,
      "grad_norm": 0.6473397612571716,
      "learning_rate": 2.3820419908374465e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 14069688,
      "step": 21505
    },
    {
      "epoch": 11.273584905660377,
      "grad_norm": 0.7155306339263916,
      "learning_rate": 2.3808998482125265e-05,
      "loss": 0.3897,
      "num_input_tokens_seen": 14072248,
      "step": 21510
    },
    {
      "epoch": 11.276205450733753,
      "grad_norm": 0.4520663022994995,
      "learning_rate": 2.3797577305019987e-05,
      "loss": 0.4238,
      "num_input_tokens_seen": 14075672,
      "step": 21515
    },
    {
      "epoch": 11.278825995807129,
      "grad_norm": 0.46814724802970886,
      "learning_rate": 2.378615637944783e-05,
      "loss": 0.4882,
      "num_input_tokens_seen": 14079128,
      "step": 21520
    },
    {
      "epoch": 11.281446540880504,
      "grad_norm": 0.3298031687736511,
      "learning_rate": 2.3774735707797895e-05,
      "loss": 0.4884,
      "num_input_tokens_seen": 14082584,
      "step": 21525
    },
    {
      "epoch": 11.284067085953879,
      "grad_norm": 0.5634803175926208,
      "learning_rate": 2.376331529245928e-05,
      "loss": 0.4551,
      "num_input_tokens_seen": 14085720,
      "step": 21530
    },
    {
      "epoch": 11.286687631027254,
      "grad_norm": 0.36617615818977356,
      "learning_rate": 2.3751895135821006e-05,
      "loss": 0.4389,
      "num_input_tokens_seen": 14089432,
      "step": 21535
    },
    {
      "epoch": 11.28930817610063,
      "grad_norm": 0.5586175918579102,
      "learning_rate": 2.374047524027202e-05,
      "loss": 0.473,
      "num_input_tokens_seen": 14092440,
      "step": 21540
    },
    {
      "epoch": 11.291928721174004,
      "grad_norm": 0.6199929714202881,
      "learning_rate": 2.3729055608201243e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 14095320,
      "step": 21545
    },
    {
      "epoch": 11.29454926624738,
      "grad_norm": 0.5679246783256531,
      "learning_rate": 2.3717636241997537e-05,
      "loss": 0.4192,
      "num_input_tokens_seen": 14098232,
      "step": 21550
    },
    {
      "epoch": 11.297169811320755,
      "grad_norm": 0.5066136717796326,
      "learning_rate": 2.3706217144049696e-05,
      "loss": 0.4264,
      "num_input_tokens_seen": 14101528,
      "step": 21555
    },
    {
      "epoch": 11.29979035639413,
      "grad_norm": 0.7485718131065369,
      "learning_rate": 2.369479831674647e-05,
      "loss": 0.5026,
      "num_input_tokens_seen": 14104376,
      "step": 21560
    },
    {
      "epoch": 11.302410901467505,
      "grad_norm": 0.40684428811073303,
      "learning_rate": 2.3683379762476555e-05,
      "loss": 0.3604,
      "num_input_tokens_seen": 14107864,
      "step": 21565
    },
    {
      "epoch": 11.30503144654088,
      "grad_norm": 0.595955491065979,
      "learning_rate": 2.3671961483628562e-05,
      "loss": 0.5173,
      "num_input_tokens_seen": 14110360,
      "step": 21570
    },
    {
      "epoch": 11.307651991614255,
      "grad_norm": 0.4421381652355194,
      "learning_rate": 2.3660543482591078e-05,
      "loss": 0.2959,
      "num_input_tokens_seen": 14114200,
      "step": 21575
    },
    {
      "epoch": 11.31027253668763,
      "grad_norm": 0.5226427912712097,
      "learning_rate": 2.364912576175261e-05,
      "loss": 0.5918,
      "num_input_tokens_seen": 14118616,
      "step": 21580
    },
    {
      "epoch": 11.312893081761006,
      "grad_norm": 0.5557103157043457,
      "learning_rate": 2.363770832350163e-05,
      "loss": 0.3959,
      "num_input_tokens_seen": 14121496,
      "step": 21585
    },
    {
      "epoch": 11.315513626834381,
      "grad_norm": 0.4347175359725952,
      "learning_rate": 2.362629117022653e-05,
      "loss": 0.4579,
      "num_input_tokens_seen": 14124280,
      "step": 21590
    },
    {
      "epoch": 11.318134171907756,
      "grad_norm": 0.4728849530220032,
      "learning_rate": 2.3614874304315633e-05,
      "loss": 0.4842,
      "num_input_tokens_seen": 14128216,
      "step": 21595
    },
    {
      "epoch": 11.320754716981131,
      "grad_norm": 0.632885754108429,
      "learning_rate": 2.3603457728157237e-05,
      "loss": 0.5234,
      "num_input_tokens_seen": 14130456,
      "step": 21600
    },
    {
      "epoch": 11.323375262054507,
      "grad_norm": 0.43966686725616455,
      "learning_rate": 2.3592041444139547e-05,
      "loss": 0.4392,
      "num_input_tokens_seen": 14133720,
      "step": 21605
    },
    {
      "epoch": 11.325995807127883,
      "grad_norm": 0.48773568868637085,
      "learning_rate": 2.358062545465073e-05,
      "loss": 0.4378,
      "num_input_tokens_seen": 14137176,
      "step": 21610
    },
    {
      "epoch": 11.328616352201259,
      "grad_norm": 0.6274536848068237,
      "learning_rate": 2.3569209762078866e-05,
      "loss": 0.6703,
      "num_input_tokens_seen": 14141240,
      "step": 21615
    },
    {
      "epoch": 11.331236897274634,
      "grad_norm": 0.5973449945449829,
      "learning_rate": 2.355779436881201e-05,
      "loss": 0.4388,
      "num_input_tokens_seen": 14144728,
      "step": 21620
    },
    {
      "epoch": 11.333857442348009,
      "grad_norm": 0.563537061214447,
      "learning_rate": 2.3546379277238107e-05,
      "loss": 0.3865,
      "num_input_tokens_seen": 14148760,
      "step": 21625
    },
    {
      "epoch": 11.336477987421384,
      "grad_norm": 0.4215533137321472,
      "learning_rate": 2.353496448974508e-05,
      "loss": 0.4197,
      "num_input_tokens_seen": 14151704,
      "step": 21630
    },
    {
      "epoch": 11.33909853249476,
      "grad_norm": 0.5621517896652222,
      "learning_rate": 2.3523550008720763e-05,
      "loss": 0.3967,
      "num_input_tokens_seen": 14154456,
      "step": 21635
    },
    {
      "epoch": 11.341719077568134,
      "grad_norm": 0.47921136021614075,
      "learning_rate": 2.3512135836552945e-05,
      "loss": 0.5691,
      "num_input_tokens_seen": 14157016,
      "step": 21640
    },
    {
      "epoch": 11.34433962264151,
      "grad_norm": 0.47970789670944214,
      "learning_rate": 2.350072197562933e-05,
      "loss": 0.3511,
      "num_input_tokens_seen": 14159384,
      "step": 21645
    },
    {
      "epoch": 11.346960167714885,
      "grad_norm": 0.571708619594574,
      "learning_rate": 2.3489308428337586e-05,
      "loss": 0.6466,
      "num_input_tokens_seen": 14162360,
      "step": 21650
    },
    {
      "epoch": 11.34958071278826,
      "grad_norm": 0.5484784245491028,
      "learning_rate": 2.3477895197065275e-05,
      "loss": 0.6172,
      "num_input_tokens_seen": 14165592,
      "step": 21655
    },
    {
      "epoch": 11.352201257861635,
      "grad_norm": 0.30204513669013977,
      "learning_rate": 2.346648228419992e-05,
      "loss": 0.456,
      "num_input_tokens_seen": 14168088,
      "step": 21660
    },
    {
      "epoch": 11.35482180293501,
      "grad_norm": 0.43376874923706055,
      "learning_rate": 2.3455069692128982e-05,
      "loss": 0.4843,
      "num_input_tokens_seen": 14171032,
      "step": 21665
    },
    {
      "epoch": 11.357442348008385,
      "grad_norm": 0.3911760151386261,
      "learning_rate": 2.3443657423239832e-05,
      "loss": 0.5367,
      "num_input_tokens_seen": 14173720,
      "step": 21670
    },
    {
      "epoch": 11.36006289308176,
      "grad_norm": 0.5531674027442932,
      "learning_rate": 2.34322454799198e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 14176664,
      "step": 21675
    },
    {
      "epoch": 11.362683438155136,
      "grad_norm": 0.4319046437740326,
      "learning_rate": 2.3420833864556133e-05,
      "loss": 0.4953,
      "num_input_tokens_seen": 14180024,
      "step": 21680
    },
    {
      "epoch": 11.365303983228511,
      "grad_norm": 0.4900084435939789,
      "learning_rate": 2.3409422579535995e-05,
      "loss": 0.386,
      "num_input_tokens_seen": 14182328,
      "step": 21685
    },
    {
      "epoch": 11.367924528301886,
      "grad_norm": 0.4244673550128937,
      "learning_rate": 2.339801162724651e-05,
      "loss": 0.516,
      "num_input_tokens_seen": 14185784,
      "step": 21690
    },
    {
      "epoch": 11.370545073375261,
      "grad_norm": 0.478694349527359,
      "learning_rate": 2.338660101007471e-05,
      "loss": 0.3936,
      "num_input_tokens_seen": 14188536,
      "step": 21695
    },
    {
      "epoch": 11.373165618448636,
      "grad_norm": 0.31851768493652344,
      "learning_rate": 2.3375190730407578e-05,
      "loss": 0.4258,
      "num_input_tokens_seen": 14192504,
      "step": 21700
    },
    {
      "epoch": 11.375786163522013,
      "grad_norm": 0.4307136833667755,
      "learning_rate": 2.336378079063202e-05,
      "loss": 0.4744,
      "num_input_tokens_seen": 14195256,
      "step": 21705
    },
    {
      "epoch": 11.378406708595389,
      "grad_norm": 0.3785759508609772,
      "learning_rate": 2.335237119313484e-05,
      "loss": 0.3794,
      "num_input_tokens_seen": 14198296,
      "step": 21710
    },
    {
      "epoch": 11.381027253668764,
      "grad_norm": 0.43667665123939514,
      "learning_rate": 2.3340961940302807e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 14201880,
      "step": 21715
    },
    {
      "epoch": 11.383647798742139,
      "grad_norm": 0.5013404488563538,
      "learning_rate": 2.3329553034522616e-05,
      "loss": 0.5604,
      "num_input_tokens_seen": 14204792,
      "step": 21720
    },
    {
      "epoch": 11.386268343815514,
      "grad_norm": 0.6312384605407715,
      "learning_rate": 2.3318144478180866e-05,
      "loss": 0.3418,
      "num_input_tokens_seen": 14212216,
      "step": 21725
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 0.3815370798110962,
      "learning_rate": 2.330673627366411e-05,
      "loss": 0.4785,
      "num_input_tokens_seen": 14215384,
      "step": 21730
    },
    {
      "epoch": 11.391509433962264,
      "grad_norm": 0.44091248512268066,
      "learning_rate": 2.329532842335881e-05,
      "loss": 0.3614,
      "num_input_tokens_seen": 14219736,
      "step": 21735
    },
    {
      "epoch": 11.39412997903564,
      "grad_norm": 0.35222944617271423,
      "learning_rate": 2.328392092965135e-05,
      "loss": 0.3817,
      "num_input_tokens_seen": 14222232,
      "step": 21740
    },
    {
      "epoch": 11.396750524109015,
      "grad_norm": 0.44331884384155273,
      "learning_rate": 2.3272513794928054e-05,
      "loss": 0.3999,
      "num_input_tokens_seen": 14225016,
      "step": 21745
    },
    {
      "epoch": 11.39937106918239,
      "grad_norm": 0.6330385804176331,
      "learning_rate": 2.326110702157516e-05,
      "loss": 0.5052,
      "num_input_tokens_seen": 14228568,
      "step": 21750
    },
    {
      "epoch": 11.401991614255765,
      "grad_norm": 0.9117884039878845,
      "learning_rate": 2.3249700611978847e-05,
      "loss": 0.4112,
      "num_input_tokens_seen": 14231960,
      "step": 21755
    },
    {
      "epoch": 11.40461215932914,
      "grad_norm": 0.5197013020515442,
      "learning_rate": 2.323829456852519e-05,
      "loss": 0.626,
      "num_input_tokens_seen": 14236344,
      "step": 21760
    },
    {
      "epoch": 11.407232704402515,
      "grad_norm": 0.6512666344642639,
      "learning_rate": 2.3226888893600223e-05,
      "loss": 0.3969,
      "num_input_tokens_seen": 14239640,
      "step": 21765
    },
    {
      "epoch": 11.40985324947589,
      "grad_norm": 0.5717538595199585,
      "learning_rate": 2.3215483589589865e-05,
      "loss": 0.5288,
      "num_input_tokens_seen": 14242488,
      "step": 21770
    },
    {
      "epoch": 11.412473794549266,
      "grad_norm": 1.002035140991211,
      "learning_rate": 2.3204078658879974e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 14245784,
      "step": 21775
    },
    {
      "epoch": 11.415094339622641,
      "grad_norm": 0.47516342997550964,
      "learning_rate": 2.3192674103856337e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 14248792,
      "step": 21780
    },
    {
      "epoch": 11.417714884696016,
      "grad_norm": 0.5457086563110352,
      "learning_rate": 2.318126992690467e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 14252216,
      "step": 21785
    },
    {
      "epoch": 11.420335429769391,
      "grad_norm": 0.41260120272636414,
      "learning_rate": 2.3169866130410585e-05,
      "loss": 0.3772,
      "num_input_tokens_seen": 14255384,
      "step": 21790
    },
    {
      "epoch": 11.422955974842766,
      "grad_norm": 0.3885829448699951,
      "learning_rate": 2.315846271675962e-05,
      "loss": 0.3224,
      "num_input_tokens_seen": 14257656,
      "step": 21795
    },
    {
      "epoch": 11.425576519916142,
      "grad_norm": 0.43685221672058105,
      "learning_rate": 2.3147059688337244e-05,
      "loss": 0.4624,
      "num_input_tokens_seen": 14260664,
      "step": 21800
    },
    {
      "epoch": 11.428197064989519,
      "grad_norm": 0.6187393665313721,
      "learning_rate": 2.313565704752884e-05,
      "loss": 0.4504,
      "num_input_tokens_seen": 14263032,
      "step": 21805
    },
    {
      "epoch": 11.430817610062894,
      "grad_norm": 0.4563755393028259,
      "learning_rate": 2.3124254796719718e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 14265784,
      "step": 21810
    },
    {
      "epoch": 11.433438155136269,
      "grad_norm": 0.507357120513916,
      "learning_rate": 2.3112852938295088e-05,
      "loss": 0.4186,
      "num_input_tokens_seen": 14269080,
      "step": 21815
    },
    {
      "epoch": 11.436058700209644,
      "grad_norm": 0.8780151605606079,
      "learning_rate": 2.3101451474640104e-05,
      "loss": 0.3764,
      "num_input_tokens_seen": 14272120,
      "step": 21820
    },
    {
      "epoch": 11.43867924528302,
      "grad_norm": 0.38269859552383423,
      "learning_rate": 2.3090050408139808e-05,
      "loss": 0.3662,
      "num_input_tokens_seen": 14275256,
      "step": 21825
    },
    {
      "epoch": 11.441299790356394,
      "grad_norm": 0.8296228051185608,
      "learning_rate": 2.307864974117917e-05,
      "loss": 0.4386,
      "num_input_tokens_seen": 14279096,
      "step": 21830
    },
    {
      "epoch": 11.44392033542977,
      "grad_norm": 0.3581344783306122,
      "learning_rate": 2.3067249476143087e-05,
      "loss": 0.361,
      "num_input_tokens_seen": 14282072,
      "step": 21835
    },
    {
      "epoch": 11.446540880503145,
      "grad_norm": 0.4758983254432678,
      "learning_rate": 2.3055849615416363e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 14285080,
      "step": 21840
    },
    {
      "epoch": 11.44916142557652,
      "grad_norm": 0.32382526993751526,
      "learning_rate": 2.304445016138372e-05,
      "loss": 0.4126,
      "num_input_tokens_seen": 14288600,
      "step": 21845
    },
    {
      "epoch": 11.451781970649895,
      "grad_norm": 0.3679221570491791,
      "learning_rate": 2.303305111642981e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 14291928,
      "step": 21850
    },
    {
      "epoch": 11.45440251572327,
      "grad_norm": 0.6617632508277893,
      "learning_rate": 2.3021652482939144e-05,
      "loss": 0.3494,
      "num_input_tokens_seen": 14294840,
      "step": 21855
    },
    {
      "epoch": 11.457023060796645,
      "grad_norm": 0.35248610377311707,
      "learning_rate": 2.3010254263296214e-05,
      "loss": 0.3648,
      "num_input_tokens_seen": 14297944,
      "step": 21860
    },
    {
      "epoch": 11.45964360587002,
      "grad_norm": 0.6702277660369873,
      "learning_rate": 2.2998856459885396e-05,
      "loss": 0.5288,
      "num_input_tokens_seen": 14300952,
      "step": 21865
    },
    {
      "epoch": 11.462264150943396,
      "grad_norm": 0.6767899394035339,
      "learning_rate": 2.298745907509097e-05,
      "loss": 0.4919,
      "num_input_tokens_seen": 14303704,
      "step": 21870
    },
    {
      "epoch": 11.464884696016771,
      "grad_norm": 0.2915022373199463,
      "learning_rate": 2.2976062111297156e-05,
      "loss": 0.3363,
      "num_input_tokens_seen": 14307320,
      "step": 21875
    },
    {
      "epoch": 11.467505241090146,
      "grad_norm": 0.3827628195285797,
      "learning_rate": 2.296466557088805e-05,
      "loss": 0.4459,
      "num_input_tokens_seen": 14310136,
      "step": 21880
    },
    {
      "epoch": 11.470125786163521,
      "grad_norm": 0.5114974975585938,
      "learning_rate": 2.2953269456247688e-05,
      "loss": 0.4277,
      "num_input_tokens_seen": 14313016,
      "step": 21885
    },
    {
      "epoch": 11.472746331236896,
      "grad_norm": 0.4421977996826172,
      "learning_rate": 2.2941873769760006e-05,
      "loss": 0.4655,
      "num_input_tokens_seen": 14316152,
      "step": 21890
    },
    {
      "epoch": 11.475366876310272,
      "grad_norm": 0.5168741345405579,
      "learning_rate": 2.2930478513808844e-05,
      "loss": 0.5316,
      "num_input_tokens_seen": 14319672,
      "step": 21895
    },
    {
      "epoch": 11.477987421383649,
      "grad_norm": 0.3234981596469879,
      "learning_rate": 2.2919083690777976e-05,
      "loss": 0.3547,
      "num_input_tokens_seen": 14323192,
      "step": 21900
    },
    {
      "epoch": 11.480607966457024,
      "grad_norm": 0.6914440989494324,
      "learning_rate": 2.2907689303051062e-05,
      "loss": 0.4211,
      "num_input_tokens_seen": 14326392,
      "step": 21905
    },
    {
      "epoch": 11.483228511530399,
      "grad_norm": 0.3494112193584442,
      "learning_rate": 2.289629535301167e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 14329688,
      "step": 21910
    },
    {
      "epoch": 11.485849056603774,
      "grad_norm": 0.530754804611206,
      "learning_rate": 2.288490184304329e-05,
      "loss": 0.4915,
      "num_input_tokens_seen": 14332504,
      "step": 21915
    },
    {
      "epoch": 11.48846960167715,
      "grad_norm": 0.5216919779777527,
      "learning_rate": 2.2873508775529308e-05,
      "loss": 0.4045,
      "num_input_tokens_seen": 14335736,
      "step": 21920
    },
    {
      "epoch": 11.491090146750524,
      "grad_norm": 0.27174636721611023,
      "learning_rate": 2.2862116152853034e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 14338392,
      "step": 21925
    },
    {
      "epoch": 11.4937106918239,
      "grad_norm": 0.5837008953094482,
      "learning_rate": 2.285072397739767e-05,
      "loss": 0.403,
      "num_input_tokens_seen": 14341656,
      "step": 21930
    },
    {
      "epoch": 11.496331236897275,
      "grad_norm": 0.31887388229370117,
      "learning_rate": 2.283933225154634e-05,
      "loss": 0.4677,
      "num_input_tokens_seen": 14345304,
      "step": 21935
    },
    {
      "epoch": 11.49895178197065,
      "grad_norm": 0.35885992646217346,
      "learning_rate": 2.2827940977682045e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 14348568,
      "step": 21940
    },
    {
      "epoch": 11.501572327044025,
      "grad_norm": 0.4255984127521515,
      "learning_rate": 2.2816550158187717e-05,
      "loss": 0.541,
      "num_input_tokens_seen": 14351960,
      "step": 21945
    },
    {
      "epoch": 11.5041928721174,
      "grad_norm": 0.7377384901046753,
      "learning_rate": 2.2805159795446182e-05,
      "loss": 0.4186,
      "num_input_tokens_seen": 14355192,
      "step": 21950
    },
    {
      "epoch": 11.506813417190775,
      "grad_norm": 0.42058730125427246,
      "learning_rate": 2.2793769891840184e-05,
      "loss": 0.4163,
      "num_input_tokens_seen": 14357816,
      "step": 21955
    },
    {
      "epoch": 11.50943396226415,
      "grad_norm": 0.4336099922657013,
      "learning_rate": 2.278238044975236e-05,
      "loss": 0.4268,
      "num_input_tokens_seen": 14360536,
      "step": 21960
    },
    {
      "epoch": 11.512054507337526,
      "grad_norm": 0.5306544899940491,
      "learning_rate": 2.2770991471565236e-05,
      "loss": 0.3997,
      "num_input_tokens_seen": 14364248,
      "step": 21965
    },
    {
      "epoch": 11.514675052410901,
      "grad_norm": 0.31319665908813477,
      "learning_rate": 2.2759602959661265e-05,
      "loss": 0.3759,
      "num_input_tokens_seen": 14367736,
      "step": 21970
    },
    {
      "epoch": 11.517295597484276,
      "grad_norm": 0.4000089168548584,
      "learning_rate": 2.2748214916422796e-05,
      "loss": 0.4273,
      "num_input_tokens_seen": 14370808,
      "step": 21975
    },
    {
      "epoch": 11.519916142557651,
      "grad_norm": 0.7742359638214111,
      "learning_rate": 2.2736827344232082e-05,
      "loss": 0.462,
      "num_input_tokens_seen": 14373752,
      "step": 21980
    },
    {
      "epoch": 11.522536687631026,
      "grad_norm": 0.36341166496276855,
      "learning_rate": 2.272544024547126e-05,
      "loss": 0.4717,
      "num_input_tokens_seen": 14377336,
      "step": 21985
    },
    {
      "epoch": 11.525157232704402,
      "grad_norm": 1.633536696434021,
      "learning_rate": 2.27140536225224e-05,
      "loss": 0.2794,
      "num_input_tokens_seen": 14379768,
      "step": 21990
    },
    {
      "epoch": 11.527777777777779,
      "grad_norm": 0.728465735912323,
      "learning_rate": 2.270266747776744e-05,
      "loss": 0.4614,
      "num_input_tokens_seen": 14381944,
      "step": 21995
    },
    {
      "epoch": 11.530398322851154,
      "grad_norm": 0.4845198690891266,
      "learning_rate": 2.2691281813588226e-05,
      "loss": 0.6008,
      "num_input_tokens_seen": 14385048,
      "step": 22000
    },
    {
      "epoch": 11.533018867924529,
      "grad_norm": 0.30072876811027527,
      "learning_rate": 2.267989663236652e-05,
      "loss": 0.3932,
      "num_input_tokens_seen": 14388120,
      "step": 22005
    },
    {
      "epoch": 11.535639412997904,
      "grad_norm": 0.38774144649505615,
      "learning_rate": 2.2668511936483975e-05,
      "loss": 0.3984,
      "num_input_tokens_seen": 14391032,
      "step": 22010
    },
    {
      "epoch": 11.53825995807128,
      "grad_norm": 0.47986656427383423,
      "learning_rate": 2.2657127728322124e-05,
      "loss": 0.4104,
      "num_input_tokens_seen": 14394392,
      "step": 22015
    },
    {
      "epoch": 11.540880503144654,
      "grad_norm": 0.46666425466537476,
      "learning_rate": 2.264574401026244e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 14397208,
      "step": 22020
    },
    {
      "epoch": 11.54350104821803,
      "grad_norm": 0.3684278726577759,
      "learning_rate": 2.2634360784686242e-05,
      "loss": 0.4224,
      "num_input_tokens_seen": 14400632,
      "step": 22025
    },
    {
      "epoch": 11.546121593291405,
      "grad_norm": 0.5411916971206665,
      "learning_rate": 2.2622978053974776e-05,
      "loss": 0.4646,
      "num_input_tokens_seen": 14403192,
      "step": 22030
    },
    {
      "epoch": 11.54874213836478,
      "grad_norm": 0.39253929257392883,
      "learning_rate": 2.2611595820509186e-05,
      "loss": 0.4196,
      "num_input_tokens_seen": 14407224,
      "step": 22035
    },
    {
      "epoch": 11.551362683438155,
      "grad_norm": 0.3245483934879303,
      "learning_rate": 2.2600214086670494e-05,
      "loss": 0.444,
      "num_input_tokens_seen": 14410040,
      "step": 22040
    },
    {
      "epoch": 11.55398322851153,
      "grad_norm": 0.4115021228790283,
      "learning_rate": 2.2588832854839644e-05,
      "loss": 0.445,
      "num_input_tokens_seen": 14413816,
      "step": 22045
    },
    {
      "epoch": 11.556603773584905,
      "grad_norm": 1.0170389413833618,
      "learning_rate": 2.2577452127397463e-05,
      "loss": 0.5474,
      "num_input_tokens_seen": 14416536,
      "step": 22050
    },
    {
      "epoch": 11.55922431865828,
      "grad_norm": 0.4597635865211487,
      "learning_rate": 2.256607190672464e-05,
      "loss": 0.4608,
      "num_input_tokens_seen": 14419672,
      "step": 22055
    },
    {
      "epoch": 11.561844863731656,
      "grad_norm": 0.6102385520935059,
      "learning_rate": 2.2554692195201817e-05,
      "loss": 0.5005,
      "num_input_tokens_seen": 14424344,
      "step": 22060
    },
    {
      "epoch": 11.564465408805031,
      "grad_norm": 0.7387434840202332,
      "learning_rate": 2.254331299520948e-05,
      "loss": 0.6135,
      "num_input_tokens_seen": 14426968,
      "step": 22065
    },
    {
      "epoch": 11.567085953878406,
      "grad_norm": 0.44353511929512024,
      "learning_rate": 2.253193430912805e-05,
      "loss": 0.4589,
      "num_input_tokens_seen": 14429752,
      "step": 22070
    },
    {
      "epoch": 11.569706498951781,
      "grad_norm": 0.32218706607818604,
      "learning_rate": 2.2520556139337803e-05,
      "loss": 0.3993,
      "num_input_tokens_seen": 14434488,
      "step": 22075
    },
    {
      "epoch": 11.572327044025156,
      "grad_norm": 0.7325944900512695,
      "learning_rate": 2.250917848821892e-05,
      "loss": 0.5459,
      "num_input_tokens_seen": 14437592,
      "step": 22080
    },
    {
      "epoch": 11.574947589098532,
      "grad_norm": 0.5528491735458374,
      "learning_rate": 2.2497801358151472e-05,
      "loss": 0.4323,
      "num_input_tokens_seen": 14441144,
      "step": 22085
    },
    {
      "epoch": 11.577568134171909,
      "grad_norm": 0.40086808800697327,
      "learning_rate": 2.248642475151545e-05,
      "loss": 0.3908,
      "num_input_tokens_seen": 14444472,
      "step": 22090
    },
    {
      "epoch": 11.580188679245284,
      "grad_norm": 0.32833945751190186,
      "learning_rate": 2.2475048670690678e-05,
      "loss": 0.4917,
      "num_input_tokens_seen": 14448120,
      "step": 22095
    },
    {
      "epoch": 11.582809224318659,
      "grad_norm": 0.3290785253047943,
      "learning_rate": 2.2463673118056927e-05,
      "loss": 0.4143,
      "num_input_tokens_seen": 14451672,
      "step": 22100
    },
    {
      "epoch": 11.585429769392034,
      "grad_norm": 0.3941251039505005,
      "learning_rate": 2.245229809599383e-05,
      "loss": 0.3463,
      "num_input_tokens_seen": 14454552,
      "step": 22105
    },
    {
      "epoch": 11.58805031446541,
      "grad_norm": 0.27626824378967285,
      "learning_rate": 2.2440923606880896e-05,
      "loss": 0.4407,
      "num_input_tokens_seen": 14458296,
      "step": 22110
    },
    {
      "epoch": 11.590670859538784,
      "grad_norm": 0.4435197412967682,
      "learning_rate": 2.242954965309755e-05,
      "loss": 0.4914,
      "num_input_tokens_seen": 14462072,
      "step": 22115
    },
    {
      "epoch": 11.59329140461216,
      "grad_norm": 0.24345706403255463,
      "learning_rate": 2.2418176237023088e-05,
      "loss": 0.492,
      "num_input_tokens_seen": 14465688,
      "step": 22120
    },
    {
      "epoch": 11.595911949685535,
      "grad_norm": 0.4042970836162567,
      "learning_rate": 2.2406803361036706e-05,
      "loss": 0.4373,
      "num_input_tokens_seen": 14471576,
      "step": 22125
    },
    {
      "epoch": 11.59853249475891,
      "grad_norm": 0.35653021931648254,
      "learning_rate": 2.2395431027517475e-05,
      "loss": 0.5256,
      "num_input_tokens_seen": 14475288,
      "step": 22130
    },
    {
      "epoch": 11.601153039832285,
      "grad_norm": 0.4969547688961029,
      "learning_rate": 2.2384059238844364e-05,
      "loss": 0.3971,
      "num_input_tokens_seen": 14477816,
      "step": 22135
    },
    {
      "epoch": 11.60377358490566,
      "grad_norm": 0.5212722420692444,
      "learning_rate": 2.2372687997396215e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 14481112,
      "step": 22140
    },
    {
      "epoch": 11.606394129979035,
      "grad_norm": 0.2833975851535797,
      "learning_rate": 2.2361317305551756e-05,
      "loss": 0.4544,
      "num_input_tokens_seen": 14484408,
      "step": 22145
    },
    {
      "epoch": 11.60901467505241,
      "grad_norm": 0.4329333007335663,
      "learning_rate": 2.2349947165689615e-05,
      "loss": 0.4818,
      "num_input_tokens_seen": 14487448,
      "step": 22150
    },
    {
      "epoch": 11.611635220125786,
      "grad_norm": 0.3891667425632477,
      "learning_rate": 2.2338577580188296e-05,
      "loss": 0.4682,
      "num_input_tokens_seen": 14490680,
      "step": 22155
    },
    {
      "epoch": 11.614255765199161,
      "grad_norm": 0.408088356256485,
      "learning_rate": 2.2327208551426188e-05,
      "loss": 0.5084,
      "num_input_tokens_seen": 14493496,
      "step": 22160
    },
    {
      "epoch": 11.616876310272536,
      "grad_norm": 0.7071548104286194,
      "learning_rate": 2.2315840081781546e-05,
      "loss": 0.4344,
      "num_input_tokens_seen": 14496472,
      "step": 22165
    },
    {
      "epoch": 11.619496855345911,
      "grad_norm": 0.5866991281509399,
      "learning_rate": 2.230447217363255e-05,
      "loss": 0.5784,
      "num_input_tokens_seen": 14500120,
      "step": 22170
    },
    {
      "epoch": 11.622117400419286,
      "grad_norm": 0.44192326068878174,
      "learning_rate": 2.2293104829357212e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 14502968,
      "step": 22175
    },
    {
      "epoch": 11.624737945492662,
      "grad_norm": 0.2865114212036133,
      "learning_rate": 2.2281738051333467e-05,
      "loss": 0.3272,
      "num_input_tokens_seen": 14505720,
      "step": 22180
    },
    {
      "epoch": 11.627358490566039,
      "grad_norm": 0.511290431022644,
      "learning_rate": 2.2270371841939105e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 14508728,
      "step": 22185
    },
    {
      "epoch": 11.629979035639414,
      "grad_norm": 0.4752768278121948,
      "learning_rate": 2.2259006203551824e-05,
      "loss": 0.4531,
      "num_input_tokens_seen": 14512344,
      "step": 22190
    },
    {
      "epoch": 11.632599580712789,
      "grad_norm": 0.6978552341461182,
      "learning_rate": 2.2247641138549168e-05,
      "loss": 0.4582,
      "num_input_tokens_seen": 14515032,
      "step": 22195
    },
    {
      "epoch": 11.635220125786164,
      "grad_norm": 0.46327951550483704,
      "learning_rate": 2.2236276649308583e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 14517560,
      "step": 22200
    },
    {
      "epoch": 11.63784067085954,
      "grad_norm": 0.4953775703907013,
      "learning_rate": 2.2224912738207395e-05,
      "loss": 0.6326,
      "num_input_tokens_seen": 14521720,
      "step": 22205
    },
    {
      "epoch": 11.640461215932914,
      "grad_norm": 0.5604206919670105,
      "learning_rate": 2.2213549407622796e-05,
      "loss": 0.2871,
      "num_input_tokens_seen": 14524216,
      "step": 22210
    },
    {
      "epoch": 11.64308176100629,
      "grad_norm": 0.455171138048172,
      "learning_rate": 2.220218665993188e-05,
      "loss": 0.4881,
      "num_input_tokens_seen": 14526744,
      "step": 22215
    },
    {
      "epoch": 11.645702306079665,
      "grad_norm": 0.41208553314208984,
      "learning_rate": 2.21908244975116e-05,
      "loss": 0.5261,
      "num_input_tokens_seen": 14530840,
      "step": 22220
    },
    {
      "epoch": 11.64832285115304,
      "grad_norm": 0.8445853590965271,
      "learning_rate": 2.2179462922738776e-05,
      "loss": 0.4342,
      "num_input_tokens_seen": 14533784,
      "step": 22225
    },
    {
      "epoch": 11.650943396226415,
      "grad_norm": 0.4509298801422119,
      "learning_rate": 2.2168101937990135e-05,
      "loss": 0.4552,
      "num_input_tokens_seen": 14536856,
      "step": 22230
    },
    {
      "epoch": 11.65356394129979,
      "grad_norm": 0.6433769464492798,
      "learning_rate": 2.2156741545642253e-05,
      "loss": 0.6101,
      "num_input_tokens_seen": 14539992,
      "step": 22235
    },
    {
      "epoch": 11.656184486373165,
      "grad_norm": 0.41284406185150146,
      "learning_rate": 2.2145381748071606e-05,
      "loss": 0.4206,
      "num_input_tokens_seen": 14543672,
      "step": 22240
    },
    {
      "epoch": 11.65880503144654,
      "grad_norm": 0.34407421946525574,
      "learning_rate": 2.2134022547654538e-05,
      "loss": 0.3157,
      "num_input_tokens_seen": 14546744,
      "step": 22245
    },
    {
      "epoch": 11.661425576519916,
      "grad_norm": 0.4198891222476959,
      "learning_rate": 2.212266394676725e-05,
      "loss": 0.4293,
      "num_input_tokens_seen": 14549688,
      "step": 22250
    },
    {
      "epoch": 11.664046121593291,
      "grad_norm": 0.4048694372177124,
      "learning_rate": 2.2111305947785836e-05,
      "loss": 0.4206,
      "num_input_tokens_seen": 14552440,
      "step": 22255
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 0.7306393384933472,
      "learning_rate": 2.2099948553086263e-05,
      "loss": 0.3591,
      "num_input_tokens_seen": 14555160,
      "step": 22260
    },
    {
      "epoch": 11.669287211740041,
      "grad_norm": 0.3049521744251251,
      "learning_rate": 2.2088591765044365e-05,
      "loss": 0.3613,
      "num_input_tokens_seen": 14558360,
      "step": 22265
    },
    {
      "epoch": 11.671907756813416,
      "grad_norm": 0.4401460587978363,
      "learning_rate": 2.207723558603586e-05,
      "loss": 0.3808,
      "num_input_tokens_seen": 14561208,
      "step": 22270
    },
    {
      "epoch": 11.674528301886792,
      "grad_norm": 0.27131006121635437,
      "learning_rate": 2.206588001843633e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 14565208,
      "step": 22275
    },
    {
      "epoch": 11.677148846960169,
      "grad_norm": 0.4638371765613556,
      "learning_rate": 2.2054525064621216e-05,
      "loss": 0.3867,
      "num_input_tokens_seen": 14568088,
      "step": 22280
    },
    {
      "epoch": 11.679769392033544,
      "grad_norm": 0.49351009726524353,
      "learning_rate": 2.2043170726965858e-05,
      "loss": 0.5463,
      "num_input_tokens_seen": 14571480,
      "step": 22285
    },
    {
      "epoch": 11.682389937106919,
      "grad_norm": 0.6398478746414185,
      "learning_rate": 2.2031817007845447e-05,
      "loss": 0.2853,
      "num_input_tokens_seen": 14574040,
      "step": 22290
    },
    {
      "epoch": 11.685010482180294,
      "grad_norm": 0.4777049124240875,
      "learning_rate": 2.202046390963506e-05,
      "loss": 0.5697,
      "num_input_tokens_seen": 14576312,
      "step": 22295
    },
    {
      "epoch": 11.68763102725367,
      "grad_norm": 0.47999513149261475,
      "learning_rate": 2.2009111434709627e-05,
      "loss": 0.4271,
      "num_input_tokens_seen": 14579608,
      "step": 22300
    },
    {
      "epoch": 11.690251572327044,
      "grad_norm": 0.4815494418144226,
      "learning_rate": 2.1997759585443968e-05,
      "loss": 0.4937,
      "num_input_tokens_seen": 14582392,
      "step": 22305
    },
    {
      "epoch": 11.69287211740042,
      "grad_norm": 0.706051230430603,
      "learning_rate": 2.198640836421275e-05,
      "loss": 0.626,
      "num_input_tokens_seen": 14585016,
      "step": 22310
    },
    {
      "epoch": 11.695492662473795,
      "grad_norm": 0.9625207781791687,
      "learning_rate": 2.1975057773390515e-05,
      "loss": 0.396,
      "num_input_tokens_seen": 14588344,
      "step": 22315
    },
    {
      "epoch": 11.69811320754717,
      "grad_norm": 0.46284160017967224,
      "learning_rate": 2.1963707815351683e-05,
      "loss": 0.4091,
      "num_input_tokens_seen": 14590744,
      "step": 22320
    },
    {
      "epoch": 11.700733752620545,
      "grad_norm": 0.74403977394104,
      "learning_rate": 2.1952358492470544e-05,
      "loss": 0.4354,
      "num_input_tokens_seen": 14594136,
      "step": 22325
    },
    {
      "epoch": 11.70335429769392,
      "grad_norm": 0.43870311975479126,
      "learning_rate": 2.194100980712124e-05,
      "loss": 0.5282,
      "num_input_tokens_seen": 14597208,
      "step": 22330
    },
    {
      "epoch": 11.705974842767295,
      "grad_norm": 0.4620625674724579,
      "learning_rate": 2.1929661761677774e-05,
      "loss": 0.4444,
      "num_input_tokens_seen": 14600888,
      "step": 22335
    },
    {
      "epoch": 11.70859538784067,
      "grad_norm": 0.7399479746818542,
      "learning_rate": 2.191831435851405e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 14604376,
      "step": 22340
    },
    {
      "epoch": 11.711215932914046,
      "grad_norm": 0.41332751512527466,
      "learning_rate": 2.1906967600003793e-05,
      "loss": 0.3361,
      "num_input_tokens_seen": 14607448,
      "step": 22345
    },
    {
      "epoch": 11.713836477987421,
      "grad_norm": 0.7356878519058228,
      "learning_rate": 2.1895621488520636e-05,
      "loss": 0.5371,
      "num_input_tokens_seen": 14610520,
      "step": 22350
    },
    {
      "epoch": 11.716457023060796,
      "grad_norm": 0.329958975315094,
      "learning_rate": 2.1884276026438044e-05,
      "loss": 0.382,
      "num_input_tokens_seen": 14613752,
      "step": 22355
    },
    {
      "epoch": 11.719077568134171,
      "grad_norm": 0.41194191575050354,
      "learning_rate": 2.187293121612937e-05,
      "loss": 0.4746,
      "num_input_tokens_seen": 14618136,
      "step": 22360
    },
    {
      "epoch": 11.721698113207546,
      "grad_norm": 0.29992979764938354,
      "learning_rate": 2.1861587059967804e-05,
      "loss": 0.5514,
      "num_input_tokens_seen": 14621880,
      "step": 22365
    },
    {
      "epoch": 11.724318658280922,
      "grad_norm": 0.4337192177772522,
      "learning_rate": 2.1850243560326414e-05,
      "loss": 0.3967,
      "num_input_tokens_seen": 14624696,
      "step": 22370
    },
    {
      "epoch": 11.726939203354299,
      "grad_norm": 0.6382522583007812,
      "learning_rate": 2.1838900719578143e-05,
      "loss": 0.3405,
      "num_input_tokens_seen": 14628440,
      "step": 22375
    },
    {
      "epoch": 11.729559748427674,
      "grad_norm": 0.5245485901832581,
      "learning_rate": 2.1827558540095772e-05,
      "loss": 0.4485,
      "num_input_tokens_seen": 14631544,
      "step": 22380
    },
    {
      "epoch": 11.732180293501049,
      "grad_norm": 0.5901239514350891,
      "learning_rate": 2.1816217024251966e-05,
      "loss": 0.563,
      "num_input_tokens_seen": 14634744,
      "step": 22385
    },
    {
      "epoch": 11.734800838574424,
      "grad_norm": 0.3798380494117737,
      "learning_rate": 2.1804876174419242e-05,
      "loss": 0.3697,
      "num_input_tokens_seen": 14637272,
      "step": 22390
    },
    {
      "epoch": 11.7374213836478,
      "grad_norm": 1.0701028108596802,
      "learning_rate": 2.179353599296996e-05,
      "loss": 0.4874,
      "num_input_tokens_seen": 14639544,
      "step": 22395
    },
    {
      "epoch": 11.740041928721174,
      "grad_norm": 0.3391619920730591,
      "learning_rate": 2.1782196482276368e-05,
      "loss": 0.5681,
      "num_input_tokens_seen": 14642648,
      "step": 22400
    },
    {
      "epoch": 11.74266247379455,
      "grad_norm": 0.732315719127655,
      "learning_rate": 2.1770857644710567e-05,
      "loss": 0.615,
      "num_input_tokens_seen": 14646168,
      "step": 22405
    },
    {
      "epoch": 11.745283018867925,
      "grad_norm": 0.7204757332801819,
      "learning_rate": 2.1759519482644495e-05,
      "loss": 0.6034,
      "num_input_tokens_seen": 14648760,
      "step": 22410
    },
    {
      "epoch": 11.7479035639413,
      "grad_norm": 0.5488326549530029,
      "learning_rate": 2.174818199844999e-05,
      "loss": 0.4459,
      "num_input_tokens_seen": 14651768,
      "step": 22415
    },
    {
      "epoch": 11.750524109014675,
      "grad_norm": 0.8843138217926025,
      "learning_rate": 2.173684519449872e-05,
      "loss": 0.4849,
      "num_input_tokens_seen": 14654168,
      "step": 22420
    },
    {
      "epoch": 11.75314465408805,
      "grad_norm": 0.5471056699752808,
      "learning_rate": 2.1725509073162188e-05,
      "loss": 0.2942,
      "num_input_tokens_seen": 14657112,
      "step": 22425
    },
    {
      "epoch": 11.755765199161425,
      "grad_norm": 0.716580331325531,
      "learning_rate": 2.171417363681181e-05,
      "loss": 0.4222,
      "num_input_tokens_seen": 14662008,
      "step": 22430
    },
    {
      "epoch": 11.7583857442348,
      "grad_norm": 0.48234453797340393,
      "learning_rate": 2.1702838887818816e-05,
      "loss": 0.6308,
      "num_input_tokens_seen": 14665304,
      "step": 22435
    },
    {
      "epoch": 11.761006289308176,
      "grad_norm": 0.43977195024490356,
      "learning_rate": 2.1691504828554315e-05,
      "loss": 0.5106,
      "num_input_tokens_seen": 14668920,
      "step": 22440
    },
    {
      "epoch": 11.76362683438155,
      "grad_norm": 0.5020380020141602,
      "learning_rate": 2.1680171461389265e-05,
      "loss": 0.6401,
      "num_input_tokens_seen": 14671608,
      "step": 22445
    },
    {
      "epoch": 11.766247379454926,
      "grad_norm": 0.5777364373207092,
      "learning_rate": 2.166883878869446e-05,
      "loss": 0.4299,
      "num_input_tokens_seen": 14674680,
      "step": 22450
    },
    {
      "epoch": 11.768867924528301,
      "grad_norm": 0.35975033044815063,
      "learning_rate": 2.1657506812840582e-05,
      "loss": 0.4025,
      "num_input_tokens_seen": 14677720,
      "step": 22455
    },
    {
      "epoch": 11.771488469601676,
      "grad_norm": 0.5868021845817566,
      "learning_rate": 2.164617553619814e-05,
      "loss": 0.3768,
      "num_input_tokens_seen": 14680888,
      "step": 22460
    },
    {
      "epoch": 11.774109014675052,
      "grad_norm": 0.614270806312561,
      "learning_rate": 2.1634844961137517e-05,
      "loss": 0.3847,
      "num_input_tokens_seen": 14685144,
      "step": 22465
    },
    {
      "epoch": 11.776729559748428,
      "grad_norm": 0.46799176931381226,
      "learning_rate": 2.162351509002893e-05,
      "loss": 0.5512,
      "num_input_tokens_seen": 14689336,
      "step": 22470
    },
    {
      "epoch": 11.779350104821804,
      "grad_norm": 0.5138627886772156,
      "learning_rate": 2.161218592524248e-05,
      "loss": 0.339,
      "num_input_tokens_seen": 14692344,
      "step": 22475
    },
    {
      "epoch": 11.781970649895179,
      "grad_norm": 0.4132396876811981,
      "learning_rate": 2.160085746914807e-05,
      "loss": 0.364,
      "num_input_tokens_seen": 14696152,
      "step": 22480
    },
    {
      "epoch": 11.784591194968554,
      "grad_norm": 0.3030405342578888,
      "learning_rate": 2.1589529724115503e-05,
      "loss": 0.5291,
      "num_input_tokens_seen": 14699480,
      "step": 22485
    },
    {
      "epoch": 11.78721174004193,
      "grad_norm": 1.1998951435089111,
      "learning_rate": 2.1578202692514407e-05,
      "loss": 0.3957,
      "num_input_tokens_seen": 14702264,
      "step": 22490
    },
    {
      "epoch": 11.789832285115304,
      "grad_norm": 0.4524605870246887,
      "learning_rate": 2.1566876376714274e-05,
      "loss": 0.4528,
      "num_input_tokens_seen": 14705720,
      "step": 22495
    },
    {
      "epoch": 11.79245283018868,
      "grad_norm": 0.36234140396118164,
      "learning_rate": 2.155555077908443e-05,
      "loss": 0.4744,
      "num_input_tokens_seen": 14708600,
      "step": 22500
    },
    {
      "epoch": 11.795073375262055,
      "grad_norm": 0.2534167170524597,
      "learning_rate": 2.1544225901994077e-05,
      "loss": 0.5444,
      "num_input_tokens_seen": 14711992,
      "step": 22505
    },
    {
      "epoch": 11.79769392033543,
      "grad_norm": 0.5788962841033936,
      "learning_rate": 2.1532901747812234e-05,
      "loss": 0.4351,
      "num_input_tokens_seen": 14715736,
      "step": 22510
    },
    {
      "epoch": 11.800314465408805,
      "grad_norm": 0.37553685903549194,
      "learning_rate": 2.1521578318907787e-05,
      "loss": 0.3875,
      "num_input_tokens_seen": 14718840,
      "step": 22515
    },
    {
      "epoch": 11.80293501048218,
      "grad_norm": 0.4448736011981964,
      "learning_rate": 2.1510255617649476e-05,
      "loss": 0.3546,
      "num_input_tokens_seen": 14721496,
      "step": 22520
    },
    {
      "epoch": 11.805555555555555,
      "grad_norm": 0.38165226578712463,
      "learning_rate": 2.149893364640587e-05,
      "loss": 0.3913,
      "num_input_tokens_seen": 14728824,
      "step": 22525
    },
    {
      "epoch": 11.80817610062893,
      "grad_norm": 0.325970858335495,
      "learning_rate": 2.1487612407545417e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 14731608,
      "step": 22530
    },
    {
      "epoch": 11.810796645702306,
      "grad_norm": 0.48199814558029175,
      "learning_rate": 2.147629190343637e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 14735192,
      "step": 22535
    },
    {
      "epoch": 11.81341719077568,
      "grad_norm": 0.5435806512832642,
      "learning_rate": 2.1464972136446852e-05,
      "loss": 0.481,
      "num_input_tokens_seen": 14738744,
      "step": 22540
    },
    {
      "epoch": 11.816037735849056,
      "grad_norm": 0.5602591037750244,
      "learning_rate": 2.1453653108944838e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 14742360,
      "step": 22545
    },
    {
      "epoch": 11.818658280922431,
      "grad_norm": 0.5968480110168457,
      "learning_rate": 2.1442334823298128e-05,
      "loss": 0.3752,
      "num_input_tokens_seen": 14745560,
      "step": 22550
    },
    {
      "epoch": 11.821278825995806,
      "grad_norm": 0.47247326374053955,
      "learning_rate": 2.1431017281874386e-05,
      "loss": 0.4806,
      "num_input_tokens_seen": 14750168,
      "step": 22555
    },
    {
      "epoch": 11.823899371069182,
      "grad_norm": 0.3059004843235016,
      "learning_rate": 2.1419700487041123e-05,
      "loss": 0.4158,
      "num_input_tokens_seen": 14753880,
      "step": 22560
    },
    {
      "epoch": 11.826519916142558,
      "grad_norm": 0.353230744600296,
      "learning_rate": 2.140838444116566e-05,
      "loss": 0.4668,
      "num_input_tokens_seen": 14757400,
      "step": 22565
    },
    {
      "epoch": 11.829140461215934,
      "grad_norm": 0.33481326699256897,
      "learning_rate": 2.13970691466152e-05,
      "loss": 0.595,
      "num_input_tokens_seen": 14760664,
      "step": 22570
    },
    {
      "epoch": 11.831761006289309,
      "grad_norm": 1.1149938106536865,
      "learning_rate": 2.1385754605756773e-05,
      "loss": 0.37,
      "num_input_tokens_seen": 14763672,
      "step": 22575
    },
    {
      "epoch": 11.834381551362684,
      "grad_norm": 0.36575838923454285,
      "learning_rate": 2.137444082095724e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 14767384,
      "step": 22580
    },
    {
      "epoch": 11.83700209643606,
      "grad_norm": 0.5203818082809448,
      "learning_rate": 2.1363127794583335e-05,
      "loss": 0.6535,
      "num_input_tokens_seen": 14770552,
      "step": 22585
    },
    {
      "epoch": 11.839622641509434,
      "grad_norm": 1.1129783391952515,
      "learning_rate": 2.1351815529001605e-05,
      "loss": 0.4413,
      "num_input_tokens_seen": 14773688,
      "step": 22590
    },
    {
      "epoch": 11.84224318658281,
      "grad_norm": 0.2763827443122864,
      "learning_rate": 2.1340504026578438e-05,
      "loss": 0.3823,
      "num_input_tokens_seen": 14776568,
      "step": 22595
    },
    {
      "epoch": 11.844863731656185,
      "grad_norm": 0.5781798958778381,
      "learning_rate": 2.1329193289680084e-05,
      "loss": 0.4918,
      "num_input_tokens_seen": 14779256,
      "step": 22600
    },
    {
      "epoch": 11.84748427672956,
      "grad_norm": 0.9738225340843201,
      "learning_rate": 2.1317883320672608e-05,
      "loss": 0.5516,
      "num_input_tokens_seen": 14782200,
      "step": 22605
    },
    {
      "epoch": 11.850104821802935,
      "grad_norm": 0.3002690374851227,
      "learning_rate": 2.1306574121921942e-05,
      "loss": 0.3042,
      "num_input_tokens_seen": 14784600,
      "step": 22610
    },
    {
      "epoch": 11.85272536687631,
      "grad_norm": 0.40331703424453735,
      "learning_rate": 2.1295265695793836e-05,
      "loss": 0.4671,
      "num_input_tokens_seen": 14788440,
      "step": 22615
    },
    {
      "epoch": 11.855345911949685,
      "grad_norm": 0.4665159285068512,
      "learning_rate": 2.128395804465387e-05,
      "loss": 0.5881,
      "num_input_tokens_seen": 14791512,
      "step": 22620
    },
    {
      "epoch": 11.85796645702306,
      "grad_norm": 0.45599910616874695,
      "learning_rate": 2.1272651170867487e-05,
      "loss": 0.4555,
      "num_input_tokens_seen": 14794712,
      "step": 22625
    },
    {
      "epoch": 11.860587002096436,
      "grad_norm": 0.3769814968109131,
      "learning_rate": 2.1261345076799955e-05,
      "loss": 0.5635,
      "num_input_tokens_seen": 14798520,
      "step": 22630
    },
    {
      "epoch": 11.86320754716981,
      "grad_norm": 0.7755874395370483,
      "learning_rate": 2.125003976481638e-05,
      "loss": 0.4551,
      "num_input_tokens_seen": 14802456,
      "step": 22635
    },
    {
      "epoch": 11.865828092243186,
      "grad_norm": 0.3338426351547241,
      "learning_rate": 2.123873523728171e-05,
      "loss": 0.3415,
      "num_input_tokens_seen": 14808184,
      "step": 22640
    },
    {
      "epoch": 11.868448637316561,
      "grad_norm": 0.5970845818519592,
      "learning_rate": 2.122743149656072e-05,
      "loss": 0.4365,
      "num_input_tokens_seen": 14810552,
      "step": 22645
    },
    {
      "epoch": 11.871069182389936,
      "grad_norm": 0.3338460922241211,
      "learning_rate": 2.1216128545018012e-05,
      "loss": 0.4092,
      "num_input_tokens_seen": 14814456,
      "step": 22650
    },
    {
      "epoch": 11.873689727463312,
      "grad_norm": 0.5539346933364868,
      "learning_rate": 2.1204826385018052e-05,
      "loss": 0.4185,
      "num_input_tokens_seen": 14818072,
      "step": 22655
    },
    {
      "epoch": 11.876310272536688,
      "grad_norm": 0.514743447303772,
      "learning_rate": 2.1193525018925106e-05,
      "loss": 0.3886,
      "num_input_tokens_seen": 14821016,
      "step": 22660
    },
    {
      "epoch": 11.878930817610064,
      "grad_norm": 0.3509393334388733,
      "learning_rate": 2.1182224449103304e-05,
      "loss": 0.4403,
      "num_input_tokens_seen": 14824792,
      "step": 22665
    },
    {
      "epoch": 11.881551362683439,
      "grad_norm": 0.7168620824813843,
      "learning_rate": 2.117092467791659e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 14828376,
      "step": 22670
    },
    {
      "epoch": 11.884171907756814,
      "grad_norm": 0.4056895077228546,
      "learning_rate": 2.115962570772876e-05,
      "loss": 0.4081,
      "num_input_tokens_seen": 14831608,
      "step": 22675
    },
    {
      "epoch": 11.88679245283019,
      "grad_norm": 0.9738031029701233,
      "learning_rate": 2.1148327540903408e-05,
      "loss": 0.4984,
      "num_input_tokens_seen": 14834200,
      "step": 22680
    },
    {
      "epoch": 11.889412997903564,
      "grad_norm": 0.5558691620826721,
      "learning_rate": 2.113703017980399e-05,
      "loss": 0.38,
      "num_input_tokens_seen": 14837624,
      "step": 22685
    },
    {
      "epoch": 11.89203354297694,
      "grad_norm": 0.3865150511264801,
      "learning_rate": 2.1125733626793793e-05,
      "loss": 0.4157,
      "num_input_tokens_seen": 14840728,
      "step": 22690
    },
    {
      "epoch": 11.894654088050315,
      "grad_norm": 0.6414215564727783,
      "learning_rate": 2.1114437884235915e-05,
      "loss": 0.5267,
      "num_input_tokens_seen": 14843160,
      "step": 22695
    },
    {
      "epoch": 11.89727463312369,
      "grad_norm": 0.390889436006546,
      "learning_rate": 2.1103142954493313e-05,
      "loss": 0.5031,
      "num_input_tokens_seen": 14845976,
      "step": 22700
    },
    {
      "epoch": 11.899895178197065,
      "grad_norm": 0.4477856159210205,
      "learning_rate": 2.1091848839928745e-05,
      "loss": 0.4193,
      "num_input_tokens_seen": 14849048,
      "step": 22705
    },
    {
      "epoch": 11.90251572327044,
      "grad_norm": 0.5346909165382385,
      "learning_rate": 2.1080555542904805e-05,
      "loss": 0.4433,
      "num_input_tokens_seen": 14852536,
      "step": 22710
    },
    {
      "epoch": 11.905136268343815,
      "grad_norm": 0.5686900019645691,
      "learning_rate": 2.106926306578393e-05,
      "loss": 0.4163,
      "num_input_tokens_seen": 14855128,
      "step": 22715
    },
    {
      "epoch": 11.90775681341719,
      "grad_norm": 0.3114893138408661,
      "learning_rate": 2.1057971410928383e-05,
      "loss": 0.4907,
      "num_input_tokens_seen": 14859224,
      "step": 22720
    },
    {
      "epoch": 11.910377358490566,
      "grad_norm": 0.3850562870502472,
      "learning_rate": 2.104668058070024e-05,
      "loss": 0.4563,
      "num_input_tokens_seen": 14863288,
      "step": 22725
    },
    {
      "epoch": 11.91299790356394,
      "grad_norm": 0.5628864765167236,
      "learning_rate": 2.1035390577461427e-05,
      "loss": 0.4767,
      "num_input_tokens_seen": 14867032,
      "step": 22730
    },
    {
      "epoch": 11.915618448637316,
      "grad_norm": 0.7314074039459229,
      "learning_rate": 2.102410140357367e-05,
      "loss": 0.3403,
      "num_input_tokens_seen": 14869752,
      "step": 22735
    },
    {
      "epoch": 11.918238993710691,
      "grad_norm": 0.47219789028167725,
      "learning_rate": 2.1012813061398534e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 14872856,
      "step": 22740
    },
    {
      "epoch": 11.920859538784066,
      "grad_norm": 0.2964116334915161,
      "learning_rate": 2.1001525553297422e-05,
      "loss": 0.5109,
      "num_input_tokens_seen": 14878648,
      "step": 22745
    },
    {
      "epoch": 11.923480083857442,
      "grad_norm": 0.2609241306781769,
      "learning_rate": 2.099023888163154e-05,
      "loss": 0.4228,
      "num_input_tokens_seen": 14882072,
      "step": 22750
    },
    {
      "epoch": 11.926100628930818,
      "grad_norm": 0.5975109934806824,
      "learning_rate": 2.0978953048761947e-05,
      "loss": 0.5236,
      "num_input_tokens_seen": 14884344,
      "step": 22755
    },
    {
      "epoch": 11.928721174004194,
      "grad_norm": 0.518669068813324,
      "learning_rate": 2.0967668057049507e-05,
      "loss": 0.4684,
      "num_input_tokens_seen": 14886584,
      "step": 22760
    },
    {
      "epoch": 11.931341719077569,
      "grad_norm": 0.6549491882324219,
      "learning_rate": 2.0956383908854895e-05,
      "loss": 0.5208,
      "num_input_tokens_seen": 14889784,
      "step": 22765
    },
    {
      "epoch": 11.933962264150944,
      "grad_norm": 0.4094485938549042,
      "learning_rate": 2.0945100606538644e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 14892664,
      "step": 22770
    },
    {
      "epoch": 11.93658280922432,
      "grad_norm": 0.3312686085700989,
      "learning_rate": 2.0933818152461077e-05,
      "loss": 0.3959,
      "num_input_tokens_seen": 14895576,
      "step": 22775
    },
    {
      "epoch": 11.939203354297694,
      "grad_norm": 0.5569171905517578,
      "learning_rate": 2.092253654898237e-05,
      "loss": 0.4372,
      "num_input_tokens_seen": 14898136,
      "step": 22780
    },
    {
      "epoch": 11.94182389937107,
      "grad_norm": 0.23857665061950684,
      "learning_rate": 2.0911255798462494e-05,
      "loss": 0.4311,
      "num_input_tokens_seen": 14901112,
      "step": 22785
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 0.9032405018806458,
      "learning_rate": 2.0899975903261266e-05,
      "loss": 0.3673,
      "num_input_tokens_seen": 14904056,
      "step": 22790
    },
    {
      "epoch": 11.94706498951782,
      "grad_norm": 0.4755770266056061,
      "learning_rate": 2.0888696865738295e-05,
      "loss": 0.3416,
      "num_input_tokens_seen": 14907352,
      "step": 22795
    },
    {
      "epoch": 11.949685534591195,
      "grad_norm": 0.6895006895065308,
      "learning_rate": 2.087741868825304e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 14910712,
      "step": 22800
    },
    {
      "epoch": 11.95230607966457,
      "grad_norm": 0.6732204556465149,
      "learning_rate": 2.0866141373164758e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 14913080,
      "step": 22805
    },
    {
      "epoch": 11.954926624737945,
      "grad_norm": 0.5509355068206787,
      "learning_rate": 2.0854864922832544e-05,
      "loss": 0.436,
      "num_input_tokens_seen": 14916152,
      "step": 22810
    },
    {
      "epoch": 11.95754716981132,
      "grad_norm": 0.5624083280563354,
      "learning_rate": 2.0843589339615304e-05,
      "loss": 0.5841,
      "num_input_tokens_seen": 14918584,
      "step": 22815
    },
    {
      "epoch": 11.960167714884696,
      "grad_norm": 0.28472593426704407,
      "learning_rate": 2.0832314625871742e-05,
      "loss": 0.5067,
      "num_input_tokens_seen": 14921784,
      "step": 22820
    },
    {
      "epoch": 11.96278825995807,
      "grad_norm": 0.6878684759140015,
      "learning_rate": 2.0821040783960423e-05,
      "loss": 0.3673,
      "num_input_tokens_seen": 14924664,
      "step": 22825
    },
    {
      "epoch": 11.965408805031446,
      "grad_norm": 0.588533878326416,
      "learning_rate": 2.0809767816239686e-05,
      "loss": 0.4464,
      "num_input_tokens_seen": 14927384,
      "step": 22830
    },
    {
      "epoch": 11.968029350104821,
      "grad_norm": 0.3431227207183838,
      "learning_rate": 2.0798495725067725e-05,
      "loss": 0.4976,
      "num_input_tokens_seen": 14930232,
      "step": 22835
    },
    {
      "epoch": 11.970649895178196,
      "grad_norm": 0.40653905272483826,
      "learning_rate": 2.078722451280252e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 14933272,
      "step": 22840
    },
    {
      "epoch": 11.973270440251572,
      "grad_norm": 0.3908213675022125,
      "learning_rate": 2.0775954181801893e-05,
      "loss": 0.4941,
      "num_input_tokens_seen": 14937208,
      "step": 22845
    },
    {
      "epoch": 11.975890985324948,
      "grad_norm": 0.41299670934677124,
      "learning_rate": 2.0764684734423455e-05,
      "loss": 0.3473,
      "num_input_tokens_seen": 14940120,
      "step": 22850
    },
    {
      "epoch": 11.978511530398324,
      "grad_norm": 1.2686493396759033,
      "learning_rate": 2.075341617302465e-05,
      "loss": 0.4729,
      "num_input_tokens_seen": 14942968,
      "step": 22855
    },
    {
      "epoch": 11.981132075471699,
      "grad_norm": 0.46771350502967834,
      "learning_rate": 2.074214849996274e-05,
      "loss": 0.4087,
      "num_input_tokens_seen": 14945784,
      "step": 22860
    },
    {
      "epoch": 11.983752620545074,
      "grad_norm": 0.5990303754806519,
      "learning_rate": 2.0730881717594776e-05,
      "loss": 0.3966,
      "num_input_tokens_seen": 14949432,
      "step": 22865
    },
    {
      "epoch": 11.98637316561845,
      "grad_norm": 0.3478381633758545,
      "learning_rate": 2.0719615828277657e-05,
      "loss": 0.4441,
      "num_input_tokens_seen": 14952344,
      "step": 22870
    },
    {
      "epoch": 11.988993710691824,
      "grad_norm": 0.5800449848175049,
      "learning_rate": 2.0708350834368083e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 14955288,
      "step": 22875
    },
    {
      "epoch": 11.9916142557652,
      "grad_norm": 0.7490730881690979,
      "learning_rate": 2.069708673822255e-05,
      "loss": 0.4808,
      "num_input_tokens_seen": 14958776,
      "step": 22880
    },
    {
      "epoch": 11.994234800838575,
      "grad_norm": 0.7913994193077087,
      "learning_rate": 2.068582354219737e-05,
      "loss": 0.3638,
      "num_input_tokens_seen": 14961624,
      "step": 22885
    },
    {
      "epoch": 11.99685534591195,
      "grad_norm": 0.33158326148986816,
      "learning_rate": 2.0674561248648697e-05,
      "loss": 0.4606,
      "num_input_tokens_seen": 14966392,
      "step": 22890
    },
    {
      "epoch": 11.999475890985325,
      "grad_norm": 0.4468001425266266,
      "learning_rate": 2.0663299859932456e-05,
      "loss": 0.5939,
      "num_input_tokens_seen": 14969976,
      "step": 22895
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.4773597717285156,
      "eval_runtime": 9.732,
      "eval_samples_per_second": 87.135,
      "eval_steps_per_second": 21.784,
      "num_input_tokens_seen": 14970024,
      "step": 22896
    },
    {
      "epoch": 12.0020964360587,
      "grad_norm": 1.456311821937561,
      "learning_rate": 2.065203937840442e-05,
      "loss": 0.2253,
      "num_input_tokens_seen": 14972840,
      "step": 22900
    },
    {
      "epoch": 12.004716981132075,
      "grad_norm": 0.5113627910614014,
      "learning_rate": 2.0640779806420136e-05,
      "loss": 0.4322,
      "num_input_tokens_seen": 14975720,
      "step": 22905
    },
    {
      "epoch": 12.00733752620545,
      "grad_norm": 0.3475560247898102,
      "learning_rate": 2.0629521146334978e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 14979016,
      "step": 22910
    },
    {
      "epoch": 12.009958071278826,
      "grad_norm": 0.4125777781009674,
      "learning_rate": 2.061826340050414e-05,
      "loss": 0.5535,
      "num_input_tokens_seen": 14982472,
      "step": 22915
    },
    {
      "epoch": 12.0125786163522,
      "grad_norm": 0.4532913565635681,
      "learning_rate": 2.0607006571282605e-05,
      "loss": 0.4063,
      "num_input_tokens_seen": 14986184,
      "step": 22920
    },
    {
      "epoch": 12.015199161425576,
      "grad_norm": 0.33002954721450806,
      "learning_rate": 2.0595750661025183e-05,
      "loss": 0.4278,
      "num_input_tokens_seen": 14988904,
      "step": 22925
    },
    {
      "epoch": 12.017819706498951,
      "grad_norm": 0.46413952112197876,
      "learning_rate": 2.0584495672086483e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 14991432,
      "step": 22930
    },
    {
      "epoch": 12.020440251572326,
      "grad_norm": 0.3962741494178772,
      "learning_rate": 2.05732416068209e-05,
      "loss": 0.328,
      "num_input_tokens_seen": 14994184,
      "step": 22935
    },
    {
      "epoch": 12.023060796645701,
      "grad_norm": 0.4117549657821655,
      "learning_rate": 2.0561988467582676e-05,
      "loss": 0.468,
      "num_input_tokens_seen": 14998280,
      "step": 22940
    },
    {
      "epoch": 12.025681341719078,
      "grad_norm": 0.5677744746208191,
      "learning_rate": 2.0550736256725828e-05,
      "loss": 0.6053,
      "num_input_tokens_seen": 15001224,
      "step": 22945
    },
    {
      "epoch": 12.028301886792454,
      "grad_norm": 0.3353107273578644,
      "learning_rate": 2.0539484976604196e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 15004392,
      "step": 22950
    },
    {
      "epoch": 12.030922431865829,
      "grad_norm": 0.7215291261672974,
      "learning_rate": 2.052823462957142e-05,
      "loss": 0.4472,
      "num_input_tokens_seen": 15007272,
      "step": 22955
    },
    {
      "epoch": 12.033542976939204,
      "grad_norm": 0.457084983587265,
      "learning_rate": 2.0516985217980953e-05,
      "loss": 0.3686,
      "num_input_tokens_seen": 15011048,
      "step": 22960
    },
    {
      "epoch": 12.036163522012579,
      "grad_norm": 0.37249287962913513,
      "learning_rate": 2.050573674418602e-05,
      "loss": 0.5507,
      "num_input_tokens_seen": 15014376,
      "step": 22965
    },
    {
      "epoch": 12.038784067085954,
      "grad_norm": 0.4288389980792999,
      "learning_rate": 2.0494489210539693e-05,
      "loss": 0.5039,
      "num_input_tokens_seen": 15018408,
      "step": 22970
    },
    {
      "epoch": 12.04140461215933,
      "grad_norm": 0.45568618178367615,
      "learning_rate": 2.0483242619394817e-05,
      "loss": 0.3969,
      "num_input_tokens_seen": 15021512,
      "step": 22975
    },
    {
      "epoch": 12.044025157232705,
      "grad_norm": 0.539598286151886,
      "learning_rate": 2.0471996973104063e-05,
      "loss": 0.4139,
      "num_input_tokens_seen": 15023688,
      "step": 22980
    },
    {
      "epoch": 12.04664570230608,
      "grad_norm": 0.49297675490379333,
      "learning_rate": 2.0460752274019886e-05,
      "loss": 0.3838,
      "num_input_tokens_seen": 15026760,
      "step": 22985
    },
    {
      "epoch": 12.049266247379455,
      "grad_norm": 0.3963184058666229,
      "learning_rate": 2.0449508524494537e-05,
      "loss": 0.4169,
      "num_input_tokens_seen": 15030248,
      "step": 22990
    },
    {
      "epoch": 12.05188679245283,
      "grad_norm": 0.6600430607795715,
      "learning_rate": 2.0438265726880092e-05,
      "loss": 0.4174,
      "num_input_tokens_seen": 15032872,
      "step": 22995
    },
    {
      "epoch": 12.054507337526205,
      "grad_norm": 0.5892731547355652,
      "learning_rate": 2.042702388352841e-05,
      "loss": 0.5018,
      "num_input_tokens_seen": 15035560,
      "step": 23000
    },
    {
      "epoch": 12.05712788259958,
      "grad_norm": 0.3706341087818146,
      "learning_rate": 2.041578299679117e-05,
      "loss": 0.4675,
      "num_input_tokens_seen": 15038504,
      "step": 23005
    },
    {
      "epoch": 12.059748427672956,
      "grad_norm": 0.3708091080188751,
      "learning_rate": 2.0404543069019818e-05,
      "loss": 0.5392,
      "num_input_tokens_seen": 15041352,
      "step": 23010
    },
    {
      "epoch": 12.06236897274633,
      "grad_norm": 0.5260536670684814,
      "learning_rate": 2.0393304102565642e-05,
      "loss": 0.4569,
      "num_input_tokens_seen": 15044584,
      "step": 23015
    },
    {
      "epoch": 12.064989517819706,
      "grad_norm": 0.5368385314941406,
      "learning_rate": 2.0382066099779685e-05,
      "loss": 0.3298,
      "num_input_tokens_seen": 15047496,
      "step": 23020
    },
    {
      "epoch": 12.067610062893081,
      "grad_norm": 0.41870155930519104,
      "learning_rate": 2.0370829063012808e-05,
      "loss": 0.4839,
      "num_input_tokens_seen": 15050408,
      "step": 23025
    },
    {
      "epoch": 12.070230607966456,
      "grad_norm": 0.4468826353549957,
      "learning_rate": 2.035959299461568e-05,
      "loss": 0.4333,
      "num_input_tokens_seen": 15053128,
      "step": 23030
    },
    {
      "epoch": 12.072851153039831,
      "grad_norm": 0.5820249915122986,
      "learning_rate": 2.0348357896938762e-05,
      "loss": 0.4037,
      "num_input_tokens_seen": 15056040,
      "step": 23035
    },
    {
      "epoch": 12.075471698113208,
      "grad_norm": 0.4205307960510254,
      "learning_rate": 2.03371237723323e-05,
      "loss": 0.3998,
      "num_input_tokens_seen": 15059208,
      "step": 23040
    },
    {
      "epoch": 12.078092243186584,
      "grad_norm": 0.7360280156135559,
      "learning_rate": 2.032589062314636e-05,
      "loss": 0.445,
      "num_input_tokens_seen": 15062472,
      "step": 23045
    },
    {
      "epoch": 12.080712788259959,
      "grad_norm": 0.3660444915294647,
      "learning_rate": 2.0314658451730767e-05,
      "loss": 0.4627,
      "num_input_tokens_seen": 15066120,
      "step": 23050
    },
    {
      "epoch": 12.083333333333334,
      "grad_norm": 0.5025942921638489,
      "learning_rate": 2.030342726043517e-05,
      "loss": 0.4119,
      "num_input_tokens_seen": 15069544,
      "step": 23055
    },
    {
      "epoch": 12.085953878406709,
      "grad_norm": 0.48793506622314453,
      "learning_rate": 2.0292197051609013e-05,
      "loss": 0.5426,
      "num_input_tokens_seen": 15073096,
      "step": 23060
    },
    {
      "epoch": 12.088574423480084,
      "grad_norm": 0.5442632436752319,
      "learning_rate": 2.0280967827601518e-05,
      "loss": 0.5139,
      "num_input_tokens_seen": 15076296,
      "step": 23065
    },
    {
      "epoch": 12.09119496855346,
      "grad_norm": 0.6234867572784424,
      "learning_rate": 2.0269739590761723e-05,
      "loss": 0.4351,
      "num_input_tokens_seen": 15079176,
      "step": 23070
    },
    {
      "epoch": 12.093815513626835,
      "grad_norm": 0.49360015988349915,
      "learning_rate": 2.0258512343438442e-05,
      "loss": 0.4919,
      "num_input_tokens_seen": 15082312,
      "step": 23075
    },
    {
      "epoch": 12.09643605870021,
      "grad_norm": 0.4530845880508423,
      "learning_rate": 2.0247286087980276e-05,
      "loss": 0.5954,
      "num_input_tokens_seen": 15084872,
      "step": 23080
    },
    {
      "epoch": 12.099056603773585,
      "grad_norm": 0.3844693601131439,
      "learning_rate": 2.0236060826735644e-05,
      "loss": 0.3637,
      "num_input_tokens_seen": 15089896,
      "step": 23085
    },
    {
      "epoch": 12.10167714884696,
      "grad_norm": 0.4995766580104828,
      "learning_rate": 2.0224836562052732e-05,
      "loss": 0.5719,
      "num_input_tokens_seen": 15092424,
      "step": 23090
    },
    {
      "epoch": 12.104297693920335,
      "grad_norm": 0.8943659067153931,
      "learning_rate": 2.0213613296279533e-05,
      "loss": 0.431,
      "num_input_tokens_seen": 15096392,
      "step": 23095
    },
    {
      "epoch": 12.10691823899371,
      "grad_norm": 0.3669390082359314,
      "learning_rate": 2.0202391031763844e-05,
      "loss": 0.468,
      "num_input_tokens_seen": 15099464,
      "step": 23100
    },
    {
      "epoch": 12.109538784067086,
      "grad_norm": 0.5368056297302246,
      "learning_rate": 2.0191169770853195e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 15102088,
      "step": 23105
    },
    {
      "epoch": 12.11215932914046,
      "grad_norm": 0.5686761140823364,
      "learning_rate": 2.0179949515894975e-05,
      "loss": 0.4905,
      "num_input_tokens_seen": 15105288,
      "step": 23110
    },
    {
      "epoch": 12.114779874213836,
      "grad_norm": 0.5560217499732971,
      "learning_rate": 2.016873026923633e-05,
      "loss": 0.3566,
      "num_input_tokens_seen": 15107592,
      "step": 23115
    },
    {
      "epoch": 12.117400419287211,
      "grad_norm": 0.42598217725753784,
      "learning_rate": 2.0157512033224187e-05,
      "loss": 0.3968,
      "num_input_tokens_seen": 15109960,
      "step": 23120
    },
    {
      "epoch": 12.120020964360586,
      "grad_norm": 0.41954049468040466,
      "learning_rate": 2.0146294810205287e-05,
      "loss": 0.3103,
      "num_input_tokens_seen": 15113576,
      "step": 23125
    },
    {
      "epoch": 12.122641509433961,
      "grad_norm": 0.6300427913665771,
      "learning_rate": 2.0135078602526146e-05,
      "loss": 0.5583,
      "num_input_tokens_seen": 15116392,
      "step": 23130
    },
    {
      "epoch": 12.125262054507338,
      "grad_norm": 0.38476336002349854,
      "learning_rate": 2.0123863412533046e-05,
      "loss": 0.3426,
      "num_input_tokens_seen": 15118920,
      "step": 23135
    },
    {
      "epoch": 12.127882599580714,
      "grad_norm": 0.5104121565818787,
      "learning_rate": 2.0112649242572095e-05,
      "loss": 0.4031,
      "num_input_tokens_seen": 15123752,
      "step": 23140
    },
    {
      "epoch": 12.130503144654089,
      "grad_norm": 0.5137814283370972,
      "learning_rate": 2.0101436094989164e-05,
      "loss": 0.355,
      "num_input_tokens_seen": 15127784,
      "step": 23145
    },
    {
      "epoch": 12.133123689727464,
      "grad_norm": 0.44460228085517883,
      "learning_rate": 2.0090223972129918e-05,
      "loss": 0.4478,
      "num_input_tokens_seen": 15131080,
      "step": 23150
    },
    {
      "epoch": 12.135744234800839,
      "grad_norm": 0.865588366985321,
      "learning_rate": 2.0079012876339797e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 15134792,
      "step": 23155
    },
    {
      "epoch": 12.138364779874214,
      "grad_norm": 0.8371056914329529,
      "learning_rate": 2.0067802809964055e-05,
      "loss": 0.4987,
      "num_input_tokens_seen": 15138216,
      "step": 23160
    },
    {
      "epoch": 12.14098532494759,
      "grad_norm": 0.4791332483291626,
      "learning_rate": 2.0056593775347686e-05,
      "loss": 0.3599,
      "num_input_tokens_seen": 15142408,
      "step": 23165
    },
    {
      "epoch": 12.143605870020965,
      "grad_norm": 0.7185300588607788,
      "learning_rate": 2.00453857748355e-05,
      "loss": 0.3837,
      "num_input_tokens_seen": 15145768,
      "step": 23170
    },
    {
      "epoch": 12.14622641509434,
      "grad_norm": 0.5069531202316284,
      "learning_rate": 2.0034178810772082e-05,
      "loss": 0.3388,
      "num_input_tokens_seen": 15148872,
      "step": 23175
    },
    {
      "epoch": 12.148846960167715,
      "grad_norm": 0.30284371972084045,
      "learning_rate": 2.002297288550181e-05,
      "loss": 0.408,
      "num_input_tokens_seen": 15151560,
      "step": 23180
    },
    {
      "epoch": 12.15146750524109,
      "grad_norm": 0.49889513850212097,
      "learning_rate": 2.0011768001368835e-05,
      "loss": 0.5053,
      "num_input_tokens_seen": 15153992,
      "step": 23185
    },
    {
      "epoch": 12.154088050314465,
      "grad_norm": 0.5386846661567688,
      "learning_rate": 2.0000564160717076e-05,
      "loss": 0.5437,
      "num_input_tokens_seen": 15158248,
      "step": 23190
    },
    {
      "epoch": 12.15670859538784,
      "grad_norm": 0.5278210043907166,
      "learning_rate": 1.9989361365890263e-05,
      "loss": 0.407,
      "num_input_tokens_seen": 15160776,
      "step": 23195
    },
    {
      "epoch": 12.159329140461216,
      "grad_norm": 0.3585485517978668,
      "learning_rate": 1.997815961923188e-05,
      "loss": 0.4131,
      "num_input_tokens_seen": 15163816,
      "step": 23200
    },
    {
      "epoch": 12.16194968553459,
      "grad_norm": 0.4806639552116394,
      "learning_rate": 1.996695892308522e-05,
      "loss": 0.4156,
      "num_input_tokens_seen": 15168424,
      "step": 23205
    },
    {
      "epoch": 12.164570230607966,
      "grad_norm": 0.5236414670944214,
      "learning_rate": 1.9955759279793327e-05,
      "loss": 0.4622,
      "num_input_tokens_seen": 15171560,
      "step": 23210
    },
    {
      "epoch": 12.167190775681341,
      "grad_norm": 0.37272003293037415,
      "learning_rate": 1.9944560691699057e-05,
      "loss": 0.5302,
      "num_input_tokens_seen": 15174536,
      "step": 23215
    },
    {
      "epoch": 12.169811320754716,
      "grad_norm": 0.4156387448310852,
      "learning_rate": 1.9933363161145006e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 15177800,
      "step": 23220
    },
    {
      "epoch": 12.172431865828091,
      "grad_norm": 0.41263630986213684,
      "learning_rate": 1.9922166690473575e-05,
      "loss": 0.4075,
      "num_input_tokens_seen": 15180872,
      "step": 23225
    },
    {
      "epoch": 12.175052410901468,
      "grad_norm": 0.4596535563468933,
      "learning_rate": 1.991097128202695e-05,
      "loss": 0.5048,
      "num_input_tokens_seen": 15183912,
      "step": 23230
    },
    {
      "epoch": 12.177672955974844,
      "grad_norm": 0.7615252137184143,
      "learning_rate": 1.9899776938147065e-05,
      "loss": 0.4202,
      "num_input_tokens_seen": 15186664,
      "step": 23235
    },
    {
      "epoch": 12.180293501048219,
      "grad_norm": 0.7455654740333557,
      "learning_rate": 1.9888583661175665e-05,
      "loss": 0.4767,
      "num_input_tokens_seen": 15190248,
      "step": 23240
    },
    {
      "epoch": 12.182914046121594,
      "grad_norm": 0.25278547406196594,
      "learning_rate": 1.9877391453454253e-05,
      "loss": 0.3775,
      "num_input_tokens_seen": 15193736,
      "step": 23245
    },
    {
      "epoch": 12.185534591194969,
      "grad_norm": 0.41654253005981445,
      "learning_rate": 1.98662003173241e-05,
      "loss": 0.4763,
      "num_input_tokens_seen": 15196296,
      "step": 23250
    },
    {
      "epoch": 12.188155136268344,
      "grad_norm": 0.39895182847976685,
      "learning_rate": 1.9855010255126266e-05,
      "loss": 0.444,
      "num_input_tokens_seen": 15199272,
      "step": 23255
    },
    {
      "epoch": 12.19077568134172,
      "grad_norm": 0.35322973132133484,
      "learning_rate": 1.9843821269201598e-05,
      "loss": 0.4673,
      "num_input_tokens_seen": 15202152,
      "step": 23260
    },
    {
      "epoch": 12.193396226415095,
      "grad_norm": 0.6878146529197693,
      "learning_rate": 1.983263336189069e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 15205320,
      "step": 23265
    },
    {
      "epoch": 12.19601677148847,
      "grad_norm": 0.5353782176971436,
      "learning_rate": 1.9821446535533937e-05,
      "loss": 0.4173,
      "num_input_tokens_seen": 15208936,
      "step": 23270
    },
    {
      "epoch": 12.198637316561845,
      "grad_norm": 0.38025379180908203,
      "learning_rate": 1.9810260792471492e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 15212392,
      "step": 23275
    },
    {
      "epoch": 12.20125786163522,
      "grad_norm": 0.18254454433918,
      "learning_rate": 1.979907613504327e-05,
      "loss": 0.3405,
      "num_input_tokens_seen": 15216328,
      "step": 23280
    },
    {
      "epoch": 12.203878406708595,
      "grad_norm": 0.2977007329463959,
      "learning_rate": 1.978789256558899e-05,
      "loss": 0.478,
      "num_input_tokens_seen": 15220136,
      "step": 23285
    },
    {
      "epoch": 12.20649895178197,
      "grad_norm": 0.4180670380592346,
      "learning_rate": 1.977671008644812e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 15222792,
      "step": 23290
    },
    {
      "epoch": 12.209119496855346,
      "grad_norm": 0.3155987858772278,
      "learning_rate": 1.9765528699959912e-05,
      "loss": 0.5386,
      "num_input_tokens_seen": 15225512,
      "step": 23295
    },
    {
      "epoch": 12.21174004192872,
      "grad_norm": 0.3228461742401123,
      "learning_rate": 1.975434840846339e-05,
      "loss": 0.4146,
      "num_input_tokens_seen": 15228648,
      "step": 23300
    },
    {
      "epoch": 12.214360587002096,
      "grad_norm": 0.38257721066474915,
      "learning_rate": 1.9743169214297326e-05,
      "loss": 0.5363,
      "num_input_tokens_seen": 15231784,
      "step": 23305
    },
    {
      "epoch": 12.216981132075471,
      "grad_norm": 0.5991618633270264,
      "learning_rate": 1.973199111980029e-05,
      "loss": 0.3775,
      "num_input_tokens_seen": 15234760,
      "step": 23310
    },
    {
      "epoch": 12.219601677148846,
      "grad_norm": 0.4556563198566437,
      "learning_rate": 1.9720814127310608e-05,
      "loss": 0.3018,
      "num_input_tokens_seen": 15237800,
      "step": 23315
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 0.5393933653831482,
      "learning_rate": 1.9709638239166386e-05,
      "loss": 0.4033,
      "num_input_tokens_seen": 15241544,
      "step": 23320
    },
    {
      "epoch": 12.224842767295598,
      "grad_norm": 0.32923561334609985,
      "learning_rate": 1.9698463457705485e-05,
      "loss": 0.5697,
      "num_input_tokens_seen": 15246760,
      "step": 23325
    },
    {
      "epoch": 12.227463312368974,
      "grad_norm": 0.30805811285972595,
      "learning_rate": 1.968728978526556e-05,
      "loss": 0.3994,
      "num_input_tokens_seen": 15250184,
      "step": 23330
    },
    {
      "epoch": 12.230083857442349,
      "grad_norm": 0.7501058578491211,
      "learning_rate": 1.9676117224183986e-05,
      "loss": 0.4751,
      "num_input_tokens_seen": 15253576,
      "step": 23335
    },
    {
      "epoch": 12.232704402515724,
      "grad_norm": 0.9180174469947815,
      "learning_rate": 1.9664945776797956e-05,
      "loss": 0.4382,
      "num_input_tokens_seen": 15256808,
      "step": 23340
    },
    {
      "epoch": 12.235324947589099,
      "grad_norm": 0.3492831289768219,
      "learning_rate": 1.96537754454444e-05,
      "loss": 0.4402,
      "num_input_tokens_seen": 15260040,
      "step": 23345
    },
    {
      "epoch": 12.237945492662474,
      "grad_norm": 0.6072824597358704,
      "learning_rate": 1.964260623246003e-05,
      "loss": 0.4182,
      "num_input_tokens_seen": 15262664,
      "step": 23350
    },
    {
      "epoch": 12.24056603773585,
      "grad_norm": 0.6217004656791687,
      "learning_rate": 1.9631438140181324e-05,
      "loss": 0.4674,
      "num_input_tokens_seen": 15265768,
      "step": 23355
    },
    {
      "epoch": 12.243186582809225,
      "grad_norm": 0.32219555974006653,
      "learning_rate": 1.9620271170944503e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 15269352,
      "step": 23360
    },
    {
      "epoch": 12.2458071278826,
      "grad_norm": 0.9563642144203186,
      "learning_rate": 1.960910532708558e-05,
      "loss": 0.6968,
      "num_input_tokens_seen": 15272392,
      "step": 23365
    },
    {
      "epoch": 12.248427672955975,
      "grad_norm": 0.6861942410469055,
      "learning_rate": 1.9597940610940315e-05,
      "loss": 0.4559,
      "num_input_tokens_seen": 15275176,
      "step": 23370
    },
    {
      "epoch": 12.25104821802935,
      "grad_norm": 0.5008842349052429,
      "learning_rate": 1.958677702484425e-05,
      "loss": 0.7434,
      "num_input_tokens_seen": 15278536,
      "step": 23375
    },
    {
      "epoch": 12.253668763102725,
      "grad_norm": 0.4818980395793915,
      "learning_rate": 1.957561457113267e-05,
      "loss": 0.349,
      "num_input_tokens_seen": 15282152,
      "step": 23380
    },
    {
      "epoch": 12.2562893081761,
      "grad_norm": 0.6252796649932861,
      "learning_rate": 1.9564453252140648e-05,
      "loss": 0.379,
      "num_input_tokens_seen": 15285832,
      "step": 23385
    },
    {
      "epoch": 12.258909853249476,
      "grad_norm": 0.5238023400306702,
      "learning_rate": 1.9553293070202984e-05,
      "loss": 0.5453,
      "num_input_tokens_seen": 15289064,
      "step": 23390
    },
    {
      "epoch": 12.26153039832285,
      "grad_norm": 0.5685635805130005,
      "learning_rate": 1.9542134027654267e-05,
      "loss": 0.486,
      "num_input_tokens_seen": 15292104,
      "step": 23395
    },
    {
      "epoch": 12.264150943396226,
      "grad_norm": 0.5752062201499939,
      "learning_rate": 1.953097612682885e-05,
      "loss": 0.4524,
      "num_input_tokens_seen": 15295496,
      "step": 23400
    },
    {
      "epoch": 12.266771488469601,
      "grad_norm": 0.48903149366378784,
      "learning_rate": 1.9519819370060832e-05,
      "loss": 0.3136,
      "num_input_tokens_seen": 15298472,
      "step": 23405
    },
    {
      "epoch": 12.269392033542976,
      "grad_norm": 0.365896999835968,
      "learning_rate": 1.9508663759684074e-05,
      "loss": 0.4006,
      "num_input_tokens_seen": 15302408,
      "step": 23410
    },
    {
      "epoch": 12.272012578616351,
      "grad_norm": 0.3240440785884857,
      "learning_rate": 1.9497509298032224e-05,
      "loss": 0.4387,
      "num_input_tokens_seen": 15306440,
      "step": 23415
    },
    {
      "epoch": 12.274633123689728,
      "grad_norm": 0.634823739528656,
      "learning_rate": 1.9486355987438647e-05,
      "loss": 0.6263,
      "num_input_tokens_seen": 15310920,
      "step": 23420
    },
    {
      "epoch": 12.277253668763104,
      "grad_norm": 0.36591437458992004,
      "learning_rate": 1.947520383023649e-05,
      "loss": 0.3947,
      "num_input_tokens_seen": 15313384,
      "step": 23425
    },
    {
      "epoch": 12.279874213836479,
      "grad_norm": 0.9171949625015259,
      "learning_rate": 1.9464052828758666e-05,
      "loss": 0.5007,
      "num_input_tokens_seen": 15316520,
      "step": 23430
    },
    {
      "epoch": 12.282494758909854,
      "grad_norm": 0.3643542230129242,
      "learning_rate": 1.945290298533783e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 15319240,
      "step": 23435
    },
    {
      "epoch": 12.285115303983229,
      "grad_norm": 0.3878520131111145,
      "learning_rate": 1.9441754302306406e-05,
      "loss": 0.4405,
      "num_input_tokens_seen": 15322120,
      "step": 23440
    },
    {
      "epoch": 12.287735849056604,
      "grad_norm": 0.51227205991745,
      "learning_rate": 1.9430606781996578e-05,
      "loss": 0.4274,
      "num_input_tokens_seen": 15326376,
      "step": 23445
    },
    {
      "epoch": 12.29035639412998,
      "grad_norm": 0.34789639711380005,
      "learning_rate": 1.941946042674026e-05,
      "loss": 0.4006,
      "num_input_tokens_seen": 15330280,
      "step": 23450
    },
    {
      "epoch": 12.292976939203355,
      "grad_norm": 0.5103552341461182,
      "learning_rate": 1.940831523886916e-05,
      "loss": 0.3607,
      "num_input_tokens_seen": 15333128,
      "step": 23455
    },
    {
      "epoch": 12.29559748427673,
      "grad_norm": 0.26525524258613586,
      "learning_rate": 1.9397171220714714e-05,
      "loss": 0.3499,
      "num_input_tokens_seen": 15335944,
      "step": 23460
    },
    {
      "epoch": 12.298218029350105,
      "grad_norm": 0.36126232147216797,
      "learning_rate": 1.9386028374608134e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 15338568,
      "step": 23465
    },
    {
      "epoch": 12.30083857442348,
      "grad_norm": 0.39790984988212585,
      "learning_rate": 1.9374886702880372e-05,
      "loss": 0.4291,
      "num_input_tokens_seen": 15343400,
      "step": 23470
    },
    {
      "epoch": 12.303459119496855,
      "grad_norm": 0.448495090007782,
      "learning_rate": 1.936374620786213e-05,
      "loss": 0.3483,
      "num_input_tokens_seen": 15346472,
      "step": 23475
    },
    {
      "epoch": 12.30607966457023,
      "grad_norm": 0.5148624181747437,
      "learning_rate": 1.9352606891883885e-05,
      "loss": 0.4761,
      "num_input_tokens_seen": 15350024,
      "step": 23480
    },
    {
      "epoch": 12.308700209643606,
      "grad_norm": 0.4571794271469116,
      "learning_rate": 1.9341468757275844e-05,
      "loss": 0.527,
      "num_input_tokens_seen": 15352488,
      "step": 23485
    },
    {
      "epoch": 12.31132075471698,
      "grad_norm": 0.42115217447280884,
      "learning_rate": 1.933033180636798e-05,
      "loss": 0.3627,
      "num_input_tokens_seen": 15355560,
      "step": 23490
    },
    {
      "epoch": 12.313941299790356,
      "grad_norm": 0.9087288975715637,
      "learning_rate": 1.9319196041490027e-05,
      "loss": 0.3464,
      "num_input_tokens_seen": 15358216,
      "step": 23495
    },
    {
      "epoch": 12.316561844863731,
      "grad_norm": 0.41402068734169006,
      "learning_rate": 1.930806146497146e-05,
      "loss": 0.3902,
      "num_input_tokens_seen": 15361864,
      "step": 23500
    },
    {
      "epoch": 12.319182389937106,
      "grad_norm": 0.41602930426597595,
      "learning_rate": 1.9296928079141485e-05,
      "loss": 0.3517,
      "num_input_tokens_seen": 15367592,
      "step": 23505
    },
    {
      "epoch": 12.321802935010481,
      "grad_norm": 0.8953512907028198,
      "learning_rate": 1.9285795886329093e-05,
      "loss": 0.5584,
      "num_input_tokens_seen": 15370184,
      "step": 23510
    },
    {
      "epoch": 12.324423480083858,
      "grad_norm": 0.5521827340126038,
      "learning_rate": 1.927466488886301e-05,
      "loss": 0.4981,
      "num_input_tokens_seen": 15372840,
      "step": 23515
    },
    {
      "epoch": 12.327044025157234,
      "grad_norm": 0.7215569019317627,
      "learning_rate": 1.926353508907172e-05,
      "loss": 0.4014,
      "num_input_tokens_seen": 15375624,
      "step": 23520
    },
    {
      "epoch": 12.329664570230609,
      "grad_norm": 0.7931508421897888,
      "learning_rate": 1.9252406489283437e-05,
      "loss": 0.3077,
      "num_input_tokens_seen": 15379144,
      "step": 23525
    },
    {
      "epoch": 12.332285115303984,
      "grad_norm": 0.712357223033905,
      "learning_rate": 1.9241279091826155e-05,
      "loss": 0.466,
      "num_input_tokens_seen": 15381672,
      "step": 23530
    },
    {
      "epoch": 12.334905660377359,
      "grad_norm": 0.5978911519050598,
      "learning_rate": 1.923015289902758e-05,
      "loss": 0.3764,
      "num_input_tokens_seen": 15385064,
      "step": 23535
    },
    {
      "epoch": 12.337526205450734,
      "grad_norm": 0.46544554829597473,
      "learning_rate": 1.9219027913215193e-05,
      "loss": 0.3547,
      "num_input_tokens_seen": 15387592,
      "step": 23540
    },
    {
      "epoch": 12.34014675052411,
      "grad_norm": 1.1953240633010864,
      "learning_rate": 1.9207904136716214e-05,
      "loss": 0.4027,
      "num_input_tokens_seen": 15391528,
      "step": 23545
    },
    {
      "epoch": 12.342767295597485,
      "grad_norm": 0.5843059420585632,
      "learning_rate": 1.9196781571857608e-05,
      "loss": 0.4903,
      "num_input_tokens_seen": 15394760,
      "step": 23550
    },
    {
      "epoch": 12.34538784067086,
      "grad_norm": 0.5105969905853271,
      "learning_rate": 1.9185660220966097e-05,
      "loss": 0.3348,
      "num_input_tokens_seen": 15397992,
      "step": 23555
    },
    {
      "epoch": 12.348008385744235,
      "grad_norm": 0.47523465752601624,
      "learning_rate": 1.9174540086368135e-05,
      "loss": 0.7644,
      "num_input_tokens_seen": 15400776,
      "step": 23560
    },
    {
      "epoch": 12.35062893081761,
      "grad_norm": 0.5016756057739258,
      "learning_rate": 1.9163421170389918e-05,
      "loss": 0.406,
      "num_input_tokens_seen": 15404136,
      "step": 23565
    },
    {
      "epoch": 12.353249475890985,
      "grad_norm": 0.3083198368549347,
      "learning_rate": 1.9152303475357402e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 15407688,
      "step": 23570
    },
    {
      "epoch": 12.35587002096436,
      "grad_norm": 0.4586935341358185,
      "learning_rate": 1.9141187003596295e-05,
      "loss": 0.4396,
      "num_input_tokens_seen": 15412360,
      "step": 23575
    },
    {
      "epoch": 12.358490566037736,
      "grad_norm": 0.5247047543525696,
      "learning_rate": 1.913007175743202e-05,
      "loss": 0.4268,
      "num_input_tokens_seen": 15416264,
      "step": 23580
    },
    {
      "epoch": 12.36111111111111,
      "grad_norm": 0.4154978096485138,
      "learning_rate": 1.9118957739189773e-05,
      "loss": 0.3811,
      "num_input_tokens_seen": 15419528,
      "step": 23585
    },
    {
      "epoch": 12.363731656184486,
      "grad_norm": 0.49139270186424255,
      "learning_rate": 1.910784495119447e-05,
      "loss": 0.3429,
      "num_input_tokens_seen": 15422792,
      "step": 23590
    },
    {
      "epoch": 12.366352201257861,
      "grad_norm": 0.875369131565094,
      "learning_rate": 1.9096733395770778e-05,
      "loss": 0.4322,
      "num_input_tokens_seen": 15425416,
      "step": 23595
    },
    {
      "epoch": 12.368972746331236,
      "grad_norm": 0.45874667167663574,
      "learning_rate": 1.9085623075243114e-05,
      "loss": 0.6353,
      "num_input_tokens_seen": 15430504,
      "step": 23600
    },
    {
      "epoch": 12.371593291404611,
      "grad_norm": 0.45489832758903503,
      "learning_rate": 1.9074513991935623e-05,
      "loss": 0.5031,
      "num_input_tokens_seen": 15433512,
      "step": 23605
    },
    {
      "epoch": 12.374213836477987,
      "grad_norm": 0.8628419041633606,
      "learning_rate": 1.906340614817221e-05,
      "loss": 0.3532,
      "num_input_tokens_seen": 15436168,
      "step": 23610
    },
    {
      "epoch": 12.376834381551364,
      "grad_norm": 1.174426794052124,
      "learning_rate": 1.9052299546276506e-05,
      "loss": 0.3912,
      "num_input_tokens_seen": 15439720,
      "step": 23615
    },
    {
      "epoch": 12.379454926624739,
      "grad_norm": 0.4787018597126007,
      "learning_rate": 1.904119418857187e-05,
      "loss": 0.3508,
      "num_input_tokens_seen": 15442184,
      "step": 23620
    },
    {
      "epoch": 12.382075471698114,
      "grad_norm": 0.5772161483764648,
      "learning_rate": 1.903009007738144e-05,
      "loss": 0.2775,
      "num_input_tokens_seen": 15444936,
      "step": 23625
    },
    {
      "epoch": 12.384696016771489,
      "grad_norm": 0.3295195996761322,
      "learning_rate": 1.9018987215028045e-05,
      "loss": 0.3306,
      "num_input_tokens_seen": 15447656,
      "step": 23630
    },
    {
      "epoch": 12.387316561844864,
      "grad_norm": 0.5649035573005676,
      "learning_rate": 1.9007885603834298e-05,
      "loss": 0.5558,
      "num_input_tokens_seen": 15451112,
      "step": 23635
    },
    {
      "epoch": 12.38993710691824,
      "grad_norm": 0.6065310835838318,
      "learning_rate": 1.8996785246122526e-05,
      "loss": 0.6701,
      "num_input_tokens_seen": 15454568,
      "step": 23640
    },
    {
      "epoch": 12.392557651991615,
      "grad_norm": 0.3690823018550873,
      "learning_rate": 1.898568614421478e-05,
      "loss": 0.4219,
      "num_input_tokens_seen": 15457480,
      "step": 23645
    },
    {
      "epoch": 12.39517819706499,
      "grad_norm": 0.591634213924408,
      "learning_rate": 1.897458830043288e-05,
      "loss": 0.5387,
      "num_input_tokens_seen": 15460296,
      "step": 23650
    },
    {
      "epoch": 12.397798742138365,
      "grad_norm": 0.4748404324054718,
      "learning_rate": 1.8963491717098372e-05,
      "loss": 0.4365,
      "num_input_tokens_seen": 15464104,
      "step": 23655
    },
    {
      "epoch": 12.40041928721174,
      "grad_norm": 0.3765857219696045,
      "learning_rate": 1.8952396396532522e-05,
      "loss": 0.5456,
      "num_input_tokens_seen": 15468040,
      "step": 23660
    },
    {
      "epoch": 12.403039832285115,
      "grad_norm": 0.3671342134475708,
      "learning_rate": 1.8941302341056356e-05,
      "loss": 0.3972,
      "num_input_tokens_seen": 15470984,
      "step": 23665
    },
    {
      "epoch": 12.40566037735849,
      "grad_norm": 0.7821651697158813,
      "learning_rate": 1.8930209552990625e-05,
      "loss": 0.4591,
      "num_input_tokens_seen": 15473640,
      "step": 23670
    },
    {
      "epoch": 12.408280922431866,
      "grad_norm": 0.47815924882888794,
      "learning_rate": 1.8919118034655796e-05,
      "loss": 0.4861,
      "num_input_tokens_seen": 15476456,
      "step": 23675
    },
    {
      "epoch": 12.41090146750524,
      "grad_norm": 0.565713107585907,
      "learning_rate": 1.8908027788372107e-05,
      "loss": 0.548,
      "num_input_tokens_seen": 15479016,
      "step": 23680
    },
    {
      "epoch": 12.413522012578616,
      "grad_norm": 0.5645127296447754,
      "learning_rate": 1.88969388164595e-05,
      "loss": 0.4701,
      "num_input_tokens_seen": 15483016,
      "step": 23685
    },
    {
      "epoch": 12.416142557651991,
      "grad_norm": 0.29252415895462036,
      "learning_rate": 1.8885851121237666e-05,
      "loss": 0.4577,
      "num_input_tokens_seen": 15486312,
      "step": 23690
    },
    {
      "epoch": 12.418763102725366,
      "grad_norm": 0.8832845091819763,
      "learning_rate": 1.8874764705026023e-05,
      "loss": 0.5865,
      "num_input_tokens_seen": 15490408,
      "step": 23695
    },
    {
      "epoch": 12.421383647798741,
      "grad_norm": 0.48487141728401184,
      "learning_rate": 1.8863679570143728e-05,
      "loss": 0.3647,
      "num_input_tokens_seen": 15493128,
      "step": 23700
    },
    {
      "epoch": 12.424004192872117,
      "grad_norm": 0.40486565232276917,
      "learning_rate": 1.885259571890966e-05,
      "loss": 0.4423,
      "num_input_tokens_seen": 15496680,
      "step": 23705
    },
    {
      "epoch": 12.426624737945493,
      "grad_norm": 0.2977479100227356,
      "learning_rate": 1.8841513153642427e-05,
      "loss": 0.3948,
      "num_input_tokens_seen": 15500648,
      "step": 23710
    },
    {
      "epoch": 12.429245283018869,
      "grad_norm": 0.6168103814125061,
      "learning_rate": 1.8830431876660388e-05,
      "loss": 0.4796,
      "num_input_tokens_seen": 15503400,
      "step": 23715
    },
    {
      "epoch": 12.431865828092244,
      "grad_norm": 0.43843433260917664,
      "learning_rate": 1.8819351890281607e-05,
      "loss": 0.5465,
      "num_input_tokens_seen": 15506312,
      "step": 23720
    },
    {
      "epoch": 12.434486373165619,
      "grad_norm": 0.5180979371070862,
      "learning_rate": 1.880827319682391e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 15509224,
      "step": 23725
    },
    {
      "epoch": 12.437106918238994,
      "grad_norm": 0.6914905309677124,
      "learning_rate": 1.879719579860481e-05,
      "loss": 0.4877,
      "num_input_tokens_seen": 15511976,
      "step": 23730
    },
    {
      "epoch": 12.43972746331237,
      "grad_norm": 0.5029873251914978,
      "learning_rate": 1.878611969794159e-05,
      "loss": 0.4595,
      "num_input_tokens_seen": 15514824,
      "step": 23735
    },
    {
      "epoch": 12.442348008385745,
      "grad_norm": 0.959099292755127,
      "learning_rate": 1.8775044897151232e-05,
      "loss": 0.5696,
      "num_input_tokens_seen": 15517512,
      "step": 23740
    },
    {
      "epoch": 12.44496855345912,
      "grad_norm": 0.3686795234680176,
      "learning_rate": 1.876397139855047e-05,
      "loss": 0.5405,
      "num_input_tokens_seen": 15521640,
      "step": 23745
    },
    {
      "epoch": 12.447589098532495,
      "grad_norm": 0.940668523311615,
      "learning_rate": 1.875289920445574e-05,
      "loss": 0.4195,
      "num_input_tokens_seen": 15524264,
      "step": 23750
    },
    {
      "epoch": 12.45020964360587,
      "grad_norm": 0.5788503289222717,
      "learning_rate": 1.8741828317183234e-05,
      "loss": 0.424,
      "num_input_tokens_seen": 15526824,
      "step": 23755
    },
    {
      "epoch": 12.452830188679245,
      "grad_norm": 0.52239990234375,
      "learning_rate": 1.873075873904885e-05,
      "loss": 0.5153,
      "num_input_tokens_seen": 15529864,
      "step": 23760
    },
    {
      "epoch": 12.45545073375262,
      "grad_norm": 0.4661538898944855,
      "learning_rate": 1.87196904723682e-05,
      "loss": 0.4749,
      "num_input_tokens_seen": 15534056,
      "step": 23765
    },
    {
      "epoch": 12.458071278825996,
      "grad_norm": 0.7602227926254272,
      "learning_rate": 1.8708623519456665e-05,
      "loss": 0.4486,
      "num_input_tokens_seen": 15536840,
      "step": 23770
    },
    {
      "epoch": 12.46069182389937,
      "grad_norm": 0.4336165487766266,
      "learning_rate": 1.8697557882629308e-05,
      "loss": 0.4937,
      "num_input_tokens_seen": 15540040,
      "step": 23775
    },
    {
      "epoch": 12.463312368972746,
      "grad_norm": 0.5207415819168091,
      "learning_rate": 1.868649356420094e-05,
      "loss": 0.3319,
      "num_input_tokens_seen": 15542728,
      "step": 23780
    },
    {
      "epoch": 12.465932914046121,
      "grad_norm": 0.5317745804786682,
      "learning_rate": 1.86754305664861e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 15545352,
      "step": 23785
    },
    {
      "epoch": 12.468553459119496,
      "grad_norm": 0.8571552038192749,
      "learning_rate": 1.866436889179902e-05,
      "loss": 0.5522,
      "num_input_tokens_seen": 15549288,
      "step": 23790
    },
    {
      "epoch": 12.471174004192871,
      "grad_norm": 0.645460307598114,
      "learning_rate": 1.865330854245369e-05,
      "loss": 0.5609,
      "num_input_tokens_seen": 15552040,
      "step": 23795
    },
    {
      "epoch": 12.473794549266247,
      "grad_norm": 0.49406951665878296,
      "learning_rate": 1.8642249520763804e-05,
      "loss": 0.511,
      "num_input_tokens_seen": 15555880,
      "step": 23800
    },
    {
      "epoch": 12.476415094339623,
      "grad_norm": 0.4184897541999817,
      "learning_rate": 1.863119182904278e-05,
      "loss": 0.478,
      "num_input_tokens_seen": 15558664,
      "step": 23805
    },
    {
      "epoch": 12.479035639412999,
      "grad_norm": 0.4841580092906952,
      "learning_rate": 1.8620135469603777e-05,
      "loss": 0.3632,
      "num_input_tokens_seen": 15562120,
      "step": 23810
    },
    {
      "epoch": 12.481656184486374,
      "grad_norm": 0.4632049798965454,
      "learning_rate": 1.8609080444759653e-05,
      "loss": 0.435,
      "num_input_tokens_seen": 15565128,
      "step": 23815
    },
    {
      "epoch": 12.484276729559749,
      "grad_norm": 0.3162877559661865,
      "learning_rate": 1.8598026756822978e-05,
      "loss": 0.4289,
      "num_input_tokens_seen": 15568840,
      "step": 23820
    },
    {
      "epoch": 12.486897274633124,
      "grad_norm": 0.7374488711357117,
      "learning_rate": 1.8586974408106074e-05,
      "loss": 0.6723,
      "num_input_tokens_seen": 15571304,
      "step": 23825
    },
    {
      "epoch": 12.4895178197065,
      "grad_norm": 0.47418081760406494,
      "learning_rate": 1.8575923400920954e-05,
      "loss": 0.4885,
      "num_input_tokens_seen": 15574280,
      "step": 23830
    },
    {
      "epoch": 12.492138364779874,
      "grad_norm": 0.7470788359642029,
      "learning_rate": 1.856487373757937e-05,
      "loss": 0.5508,
      "num_input_tokens_seen": 15577096,
      "step": 23835
    },
    {
      "epoch": 12.49475890985325,
      "grad_norm": 0.5931780934333801,
      "learning_rate": 1.8553825420392792e-05,
      "loss": 0.5635,
      "num_input_tokens_seen": 15580680,
      "step": 23840
    },
    {
      "epoch": 12.497379454926625,
      "grad_norm": 0.4213126301765442,
      "learning_rate": 1.8542778451672385e-05,
      "loss": 0.4629,
      "num_input_tokens_seen": 15584200,
      "step": 23845
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.3917595148086548,
      "learning_rate": 1.8531732833729063e-05,
      "loss": 0.4195,
      "num_input_tokens_seen": 15587784,
      "step": 23850
    },
    {
      "epoch": 12.502620545073375,
      "grad_norm": 1.0543298721313477,
      "learning_rate": 1.852068856887343e-05,
      "loss": 0.6316,
      "num_input_tokens_seen": 15591336,
      "step": 23855
    },
    {
      "epoch": 12.50524109014675,
      "grad_norm": 0.33122965693473816,
      "learning_rate": 1.8509645659415837e-05,
      "loss": 0.3884,
      "num_input_tokens_seen": 15594568,
      "step": 23860
    },
    {
      "epoch": 12.507861635220126,
      "grad_norm": 0.3931643068790436,
      "learning_rate": 1.849860410766632e-05,
      "loss": 0.4333,
      "num_input_tokens_seen": 15597704,
      "step": 23865
    },
    {
      "epoch": 12.5104821802935,
      "grad_norm": 0.35042959451675415,
      "learning_rate": 1.8487563915934664e-05,
      "loss": 0.4147,
      "num_input_tokens_seen": 15601928,
      "step": 23870
    },
    {
      "epoch": 12.513102725366876,
      "grad_norm": 0.8565240502357483,
      "learning_rate": 1.847652508653033e-05,
      "loss": 0.4981,
      "num_input_tokens_seen": 15604872,
      "step": 23875
    },
    {
      "epoch": 12.515723270440251,
      "grad_norm": 0.45012935996055603,
      "learning_rate": 1.846548762176252e-05,
      "loss": 0.3744,
      "num_input_tokens_seen": 15607560,
      "step": 23880
    },
    {
      "epoch": 12.518343815513626,
      "grad_norm": 0.2634742259979248,
      "learning_rate": 1.845445152394015e-05,
      "loss": 0.3864,
      "num_input_tokens_seen": 15611336,
      "step": 23885
    },
    {
      "epoch": 12.520964360587001,
      "grad_norm": 0.4807347357273102,
      "learning_rate": 1.8443416795371853e-05,
      "loss": 0.4388,
      "num_input_tokens_seen": 15614888,
      "step": 23890
    },
    {
      "epoch": 12.523584905660378,
      "grad_norm": 1.2607834339141846,
      "learning_rate": 1.8432383438365957e-05,
      "loss": 0.4336,
      "num_input_tokens_seen": 15618216,
      "step": 23895
    },
    {
      "epoch": 12.526205450733752,
      "grad_norm": 0.714236319065094,
      "learning_rate": 1.8421351455230528e-05,
      "loss": 0.5035,
      "num_input_tokens_seen": 15622120,
      "step": 23900
    },
    {
      "epoch": 12.528825995807129,
      "grad_norm": 0.6926404237747192,
      "learning_rate": 1.8410320848273315e-05,
      "loss": 0.3919,
      "num_input_tokens_seen": 15625000,
      "step": 23905
    },
    {
      "epoch": 12.531446540880504,
      "grad_norm": 0.4697229266166687,
      "learning_rate": 1.8399291619801794e-05,
      "loss": 0.5307,
      "num_input_tokens_seen": 15627496,
      "step": 23910
    },
    {
      "epoch": 12.534067085953879,
      "grad_norm": 0.46742087602615356,
      "learning_rate": 1.838826377212317e-05,
      "loss": 0.4126,
      "num_input_tokens_seen": 15630888,
      "step": 23915
    },
    {
      "epoch": 12.536687631027254,
      "grad_norm": 0.3626342713832855,
      "learning_rate": 1.8377237307544326e-05,
      "loss": 0.418,
      "num_input_tokens_seen": 15634888,
      "step": 23920
    },
    {
      "epoch": 12.53930817610063,
      "grad_norm": 0.4239584505558014,
      "learning_rate": 1.8366212228371894e-05,
      "loss": 0.3158,
      "num_input_tokens_seen": 15637672,
      "step": 23925
    },
    {
      "epoch": 12.541928721174004,
      "grad_norm": 0.41412124037742615,
      "learning_rate": 1.8355188536912173e-05,
      "loss": 0.5029,
      "num_input_tokens_seen": 15641224,
      "step": 23930
    },
    {
      "epoch": 12.54454926624738,
      "grad_norm": 0.3611448407173157,
      "learning_rate": 1.83441662354712e-05,
      "loss": 0.4808,
      "num_input_tokens_seen": 15644136,
      "step": 23935
    },
    {
      "epoch": 12.547169811320755,
      "grad_norm": 0.391242116689682,
      "learning_rate": 1.8333145326354722e-05,
      "loss": 0.5224,
      "num_input_tokens_seen": 15648072,
      "step": 23940
    },
    {
      "epoch": 12.54979035639413,
      "grad_norm": 0.3804553151130676,
      "learning_rate": 1.8322125811868175e-05,
      "loss": 0.4635,
      "num_input_tokens_seen": 15650952,
      "step": 23945
    },
    {
      "epoch": 12.552410901467505,
      "grad_norm": 0.742909848690033,
      "learning_rate": 1.831110769431673e-05,
      "loss": 0.3911,
      "num_input_tokens_seen": 15654216,
      "step": 23950
    },
    {
      "epoch": 12.55503144654088,
      "grad_norm": 0.31971636414527893,
      "learning_rate": 1.8300090976005247e-05,
      "loss": 0.3747,
      "num_input_tokens_seen": 15658184,
      "step": 23955
    },
    {
      "epoch": 12.557651991614255,
      "grad_norm": 0.2925550937652588,
      "learning_rate": 1.8289075659238285e-05,
      "loss": 0.4257,
      "num_input_tokens_seen": 15661672,
      "step": 23960
    },
    {
      "epoch": 12.56027253668763,
      "grad_norm": 0.5222625136375427,
      "learning_rate": 1.8278061746320134e-05,
      "loss": 0.4005,
      "num_input_tokens_seen": 15664104,
      "step": 23965
    },
    {
      "epoch": 12.562893081761006,
      "grad_norm": 0.39065006375312805,
      "learning_rate": 1.8267049239554786e-05,
      "loss": 0.4664,
      "num_input_tokens_seen": 15667560,
      "step": 23970
    },
    {
      "epoch": 12.565513626834381,
      "grad_norm": 0.6331239342689514,
      "learning_rate": 1.8256038141245912e-05,
      "loss": 0.3517,
      "num_input_tokens_seen": 15670952,
      "step": 23975
    },
    {
      "epoch": 12.568134171907756,
      "grad_norm": 0.45660027861595154,
      "learning_rate": 1.824502845369693e-05,
      "loss": 0.3129,
      "num_input_tokens_seen": 15673640,
      "step": 23980
    },
    {
      "epoch": 12.570754716981131,
      "grad_norm": 0.5752968192100525,
      "learning_rate": 1.8234020179210938e-05,
      "loss": 0.5564,
      "num_input_tokens_seen": 15676456,
      "step": 23985
    },
    {
      "epoch": 12.573375262054507,
      "grad_norm": 0.4509489834308624,
      "learning_rate": 1.8223013320090715e-05,
      "loss": 0.466,
      "num_input_tokens_seen": 15680008,
      "step": 23990
    },
    {
      "epoch": 12.575995807127882,
      "grad_norm": 0.44611480832099915,
      "learning_rate": 1.82120078786388e-05,
      "loss": 0.3852,
      "num_input_tokens_seen": 15683400,
      "step": 23995
    },
    {
      "epoch": 12.578616352201259,
      "grad_norm": 0.20142118632793427,
      "learning_rate": 1.820100385715739e-05,
      "loss": 0.4339,
      "num_input_tokens_seen": 15686632,
      "step": 24000
    },
    {
      "epoch": 12.581236897274634,
      "grad_norm": 0.47770604491233826,
      "learning_rate": 1.8190001257948414e-05,
      "loss": 0.4062,
      "num_input_tokens_seen": 15689832,
      "step": 24005
    },
    {
      "epoch": 12.583857442348009,
      "grad_norm": 0.4930919110774994,
      "learning_rate": 1.8179000083313483e-05,
      "loss": 0.6976,
      "num_input_tokens_seen": 15693256,
      "step": 24010
    },
    {
      "epoch": 12.586477987421384,
      "grad_norm": 0.5810826420783997,
      "learning_rate": 1.8168000335553907e-05,
      "loss": 0.6149,
      "num_input_tokens_seen": 15696776,
      "step": 24015
    },
    {
      "epoch": 12.58909853249476,
      "grad_norm": 0.38897284865379333,
      "learning_rate": 1.8157002016970723e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 15699624,
      "step": 24020
    },
    {
      "epoch": 12.591719077568134,
      "grad_norm": 0.6128069758415222,
      "learning_rate": 1.8146005129864644e-05,
      "loss": 0.3885,
      "num_input_tokens_seen": 15703176,
      "step": 24025
    },
    {
      "epoch": 12.59433962264151,
      "grad_norm": 0.7180180549621582,
      "learning_rate": 1.81350096765361e-05,
      "loss": 0.3713,
      "num_input_tokens_seen": 15706312,
      "step": 24030
    },
    {
      "epoch": 12.596960167714885,
      "grad_norm": 0.5986908674240112,
      "learning_rate": 1.812401565928521e-05,
      "loss": 0.4067,
      "num_input_tokens_seen": 15710056,
      "step": 24035
    },
    {
      "epoch": 12.59958071278826,
      "grad_norm": 0.3885052502155304,
      "learning_rate": 1.8113023080411806e-05,
      "loss": 0.3975,
      "num_input_tokens_seen": 15713864,
      "step": 24040
    },
    {
      "epoch": 12.602201257861635,
      "grad_norm": 0.4569917321205139,
      "learning_rate": 1.810203194221539e-05,
      "loss": 0.4209,
      "num_input_tokens_seen": 15717672,
      "step": 24045
    },
    {
      "epoch": 12.60482180293501,
      "grad_norm": 0.5117300152778625,
      "learning_rate": 1.8091042246995206e-05,
      "loss": 0.482,
      "num_input_tokens_seen": 15720264,
      "step": 24050
    },
    {
      "epoch": 12.607442348008385,
      "grad_norm": 0.5785900354385376,
      "learning_rate": 1.8080053997050154e-05,
      "loss": 0.4709,
      "num_input_tokens_seen": 15722952,
      "step": 24055
    },
    {
      "epoch": 12.61006289308176,
      "grad_norm": 0.3126155436038971,
      "learning_rate": 1.8069067194678867e-05,
      "loss": 0.4621,
      "num_input_tokens_seen": 15725832,
      "step": 24060
    },
    {
      "epoch": 12.612683438155136,
      "grad_norm": 0.44264867901802063,
      "learning_rate": 1.8058081842179642e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 15729704,
      "step": 24065
    },
    {
      "epoch": 12.615303983228511,
      "grad_norm": 0.5732283592224121,
      "learning_rate": 1.804709794185051e-05,
      "loss": 0.4835,
      "num_input_tokens_seen": 15732424,
      "step": 24070
    },
    {
      "epoch": 12.617924528301886,
      "grad_norm": 0.6587660312652588,
      "learning_rate": 1.8036115495989164e-05,
      "loss": 0.5554,
      "num_input_tokens_seen": 15735144,
      "step": 24075
    },
    {
      "epoch": 12.620545073375261,
      "grad_norm": 0.6404339075088501,
      "learning_rate": 1.8025134506892997e-05,
      "loss": 0.4888,
      "num_input_tokens_seen": 15737704,
      "step": 24080
    },
    {
      "epoch": 12.623165618448636,
      "grad_norm": 0.5817682147026062,
      "learning_rate": 1.8014154976859126e-05,
      "loss": 0.3935,
      "num_input_tokens_seen": 15740904,
      "step": 24085
    },
    {
      "epoch": 12.625786163522012,
      "grad_norm": 0.6402411460876465,
      "learning_rate": 1.800317690818433e-05,
      "loss": 0.4802,
      "num_input_tokens_seen": 15744072,
      "step": 24090
    },
    {
      "epoch": 12.628406708595389,
      "grad_norm": 0.4942255914211273,
      "learning_rate": 1.799220030316511e-05,
      "loss": 0.4766,
      "num_input_tokens_seen": 15751176,
      "step": 24095
    },
    {
      "epoch": 12.631027253668764,
      "grad_norm": 0.9596999883651733,
      "learning_rate": 1.7981225164097633e-05,
      "loss": 0.644,
      "num_input_tokens_seen": 15754216,
      "step": 24100
    },
    {
      "epoch": 12.633647798742139,
      "grad_norm": 0.6245629191398621,
      "learning_rate": 1.7970251493277774e-05,
      "loss": 0.4126,
      "num_input_tokens_seen": 15757224,
      "step": 24105
    },
    {
      "epoch": 12.636268343815514,
      "grad_norm": 0.525829017162323,
      "learning_rate": 1.7959279293001102e-05,
      "loss": 0.431,
      "num_input_tokens_seen": 15760680,
      "step": 24110
    },
    {
      "epoch": 12.63888888888889,
      "grad_norm": 0.4247928559780121,
      "learning_rate": 1.7948308565562876e-05,
      "loss": 0.4823,
      "num_input_tokens_seen": 15763464,
      "step": 24115
    },
    {
      "epoch": 12.641509433962264,
      "grad_norm": 0.30232182145118713,
      "learning_rate": 1.793733931325805e-05,
      "loss": 0.4202,
      "num_input_tokens_seen": 15766504,
      "step": 24120
    },
    {
      "epoch": 12.64412997903564,
      "grad_norm": 0.3141838312149048,
      "learning_rate": 1.792637153838127e-05,
      "loss": 0.425,
      "num_input_tokens_seen": 15769416,
      "step": 24125
    },
    {
      "epoch": 12.646750524109015,
      "grad_norm": 0.5008410215377808,
      "learning_rate": 1.7915405243226864e-05,
      "loss": 0.5519,
      "num_input_tokens_seen": 15772776,
      "step": 24130
    },
    {
      "epoch": 12.64937106918239,
      "grad_norm": 0.49944227933883667,
      "learning_rate": 1.7904440430088847e-05,
      "loss": 0.405,
      "num_input_tokens_seen": 15775720,
      "step": 24135
    },
    {
      "epoch": 12.651991614255765,
      "grad_norm": 0.43412649631500244,
      "learning_rate": 1.7893477101260947e-05,
      "loss": 0.3893,
      "num_input_tokens_seen": 15778632,
      "step": 24140
    },
    {
      "epoch": 12.65461215932914,
      "grad_norm": 0.8834192156791687,
      "learning_rate": 1.7882515259036558e-05,
      "loss": 0.6104,
      "num_input_tokens_seen": 15781832,
      "step": 24145
    },
    {
      "epoch": 12.657232704402515,
      "grad_norm": 0.5268839001655579,
      "learning_rate": 1.787155490570878e-05,
      "loss": 0.3328,
      "num_input_tokens_seen": 15787848,
      "step": 24150
    },
    {
      "epoch": 12.65985324947589,
      "grad_norm": 0.44373801350593567,
      "learning_rate": 1.7860596043570398e-05,
      "loss": 0.4185,
      "num_input_tokens_seen": 15790568,
      "step": 24155
    },
    {
      "epoch": 12.662473794549266,
      "grad_norm": 0.45525702834129333,
      "learning_rate": 1.7849638674913863e-05,
      "loss": 0.565,
      "num_input_tokens_seen": 15793800,
      "step": 24160
    },
    {
      "epoch": 12.665094339622641,
      "grad_norm": 0.3107070326805115,
      "learning_rate": 1.7838682802031344e-05,
      "loss": 0.5065,
      "num_input_tokens_seen": 15797032,
      "step": 24165
    },
    {
      "epoch": 12.667714884696016,
      "grad_norm": 0.4855060875415802,
      "learning_rate": 1.782772842721468e-05,
      "loss": 0.4146,
      "num_input_tokens_seen": 15799720,
      "step": 24170
    },
    {
      "epoch": 12.670335429769391,
      "grad_norm": 0.5460391044616699,
      "learning_rate": 1.7816775552755406e-05,
      "loss": 0.4903,
      "num_input_tokens_seen": 15803912,
      "step": 24175
    },
    {
      "epoch": 12.672955974842766,
      "grad_norm": 0.3239974081516266,
      "learning_rate": 1.780582418094473e-05,
      "loss": 0.4885,
      "num_input_tokens_seen": 15806760,
      "step": 24180
    },
    {
      "epoch": 12.675576519916142,
      "grad_norm": 0.471181184053421,
      "learning_rate": 1.7794874314073573e-05,
      "loss": 0.3669,
      "num_input_tokens_seen": 15811176,
      "step": 24185
    },
    {
      "epoch": 12.678197064989519,
      "grad_norm": 0.8158292174339294,
      "learning_rate": 1.7783925954432503e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 15814440,
      "step": 24190
    },
    {
      "epoch": 12.680817610062894,
      "grad_norm": 0.30419883131980896,
      "learning_rate": 1.777297910431179e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 15818568,
      "step": 24195
    },
    {
      "epoch": 12.683438155136269,
      "grad_norm": 0.3436792492866516,
      "learning_rate": 1.7762033766001402e-05,
      "loss": 0.3344,
      "num_input_tokens_seen": 15822952,
      "step": 24200
    },
    {
      "epoch": 12.686058700209644,
      "grad_norm": 0.4116522967815399,
      "learning_rate": 1.7751089941790977e-05,
      "loss": 0.3739,
      "num_input_tokens_seen": 15826440,
      "step": 24205
    },
    {
      "epoch": 12.68867924528302,
      "grad_norm": 0.5842615962028503,
      "learning_rate": 1.7740147633969844e-05,
      "loss": 0.402,
      "num_input_tokens_seen": 15830568,
      "step": 24210
    },
    {
      "epoch": 12.691299790356394,
      "grad_norm": 0.42630326747894287,
      "learning_rate": 1.772920684482699e-05,
      "loss": 0.4563,
      "num_input_tokens_seen": 15834728,
      "step": 24215
    },
    {
      "epoch": 12.69392033542977,
      "grad_norm": 0.2731771469116211,
      "learning_rate": 1.771826757665112e-05,
      "loss": 0.437,
      "num_input_tokens_seen": 15837896,
      "step": 24220
    },
    {
      "epoch": 12.696540880503145,
      "grad_norm": 0.5901691317558289,
      "learning_rate": 1.770732983173059e-05,
      "loss": 0.4498,
      "num_input_tokens_seen": 15841096,
      "step": 24225
    },
    {
      "epoch": 12.69916142557652,
      "grad_norm": 0.5283163189888,
      "learning_rate": 1.7696393612353467e-05,
      "loss": 0.4765,
      "num_input_tokens_seen": 15844584,
      "step": 24230
    },
    {
      "epoch": 12.701781970649895,
      "grad_norm": 0.2772309184074402,
      "learning_rate": 1.7685458920807474e-05,
      "loss": 0.4055,
      "num_input_tokens_seen": 15848200,
      "step": 24235
    },
    {
      "epoch": 12.70440251572327,
      "grad_norm": 0.4254292845726013,
      "learning_rate": 1.7674525759380033e-05,
      "loss": 0.3991,
      "num_input_tokens_seen": 15850728,
      "step": 24240
    },
    {
      "epoch": 12.707023060796645,
      "grad_norm": 0.8800473809242249,
      "learning_rate": 1.766359413035823e-05,
      "loss": 0.4892,
      "num_input_tokens_seen": 15854152,
      "step": 24245
    },
    {
      "epoch": 12.70964360587002,
      "grad_norm": 0.3068011999130249,
      "learning_rate": 1.7652664036028827e-05,
      "loss": 0.4162,
      "num_input_tokens_seen": 15856744,
      "step": 24250
    },
    {
      "epoch": 12.712264150943396,
      "grad_norm": 0.6823535561561584,
      "learning_rate": 1.7641735478678288e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 15859752,
      "step": 24255
    },
    {
      "epoch": 12.714884696016771,
      "grad_norm": 0.5190622806549072,
      "learning_rate": 1.763080846059274e-05,
      "loss": 0.5492,
      "num_input_tokens_seen": 15862280,
      "step": 24260
    },
    {
      "epoch": 12.717505241090146,
      "grad_norm": 0.6318172812461853,
      "learning_rate": 1.7619882984057987e-05,
      "loss": 0.4522,
      "num_input_tokens_seen": 15865608,
      "step": 24265
    },
    {
      "epoch": 12.720125786163521,
      "grad_norm": 0.2873168885707855,
      "learning_rate": 1.760895905135953e-05,
      "loss": 0.3588,
      "num_input_tokens_seen": 15869832,
      "step": 24270
    },
    {
      "epoch": 12.722746331236896,
      "grad_norm": 0.5724205374717712,
      "learning_rate": 1.7598036664782508e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 15873160,
      "step": 24275
    },
    {
      "epoch": 12.725366876310272,
      "grad_norm": 0.41526660323143005,
      "learning_rate": 1.7587115826611766e-05,
      "loss": 0.4498,
      "num_input_tokens_seen": 15876072,
      "step": 24280
    },
    {
      "epoch": 12.727987421383649,
      "grad_norm": 0.46120747923851013,
      "learning_rate": 1.7576196539131828e-05,
      "loss": 0.4783,
      "num_input_tokens_seen": 15878920,
      "step": 24285
    },
    {
      "epoch": 12.730607966457024,
      "grad_norm": 0.6714768409729004,
      "learning_rate": 1.7565278804626872e-05,
      "loss": 0.4275,
      "num_input_tokens_seen": 15882728,
      "step": 24290
    },
    {
      "epoch": 12.733228511530399,
      "grad_norm": 0.4236920177936554,
      "learning_rate": 1.755436262538078e-05,
      "loss": 0.4195,
      "num_input_tokens_seen": 15886888,
      "step": 24295
    },
    {
      "epoch": 12.735849056603774,
      "grad_norm": 1.1877548694610596,
      "learning_rate": 1.7543448003677082e-05,
      "loss": 0.4105,
      "num_input_tokens_seen": 15890216,
      "step": 24300
    },
    {
      "epoch": 12.73846960167715,
      "grad_norm": 0.5632490515708923,
      "learning_rate": 1.7532534941798985e-05,
      "loss": 0.4029,
      "num_input_tokens_seen": 15892392,
      "step": 24305
    },
    {
      "epoch": 12.741090146750524,
      "grad_norm": 0.42061924934387207,
      "learning_rate": 1.752162344202939e-05,
      "loss": 0.4358,
      "num_input_tokens_seen": 15895176,
      "step": 24310
    },
    {
      "epoch": 12.7437106918239,
      "grad_norm": 0.7049232721328735,
      "learning_rate": 1.751071350665085e-05,
      "loss": 0.4404,
      "num_input_tokens_seen": 15898696,
      "step": 24315
    },
    {
      "epoch": 12.746331236897275,
      "grad_norm": 0.38389846682548523,
      "learning_rate": 1.7499805137945602e-05,
      "loss": 0.5228,
      "num_input_tokens_seen": 15902792,
      "step": 24320
    },
    {
      "epoch": 12.74895178197065,
      "grad_norm": 0.42486006021499634,
      "learning_rate": 1.7488898338195566e-05,
      "loss": 0.3064,
      "num_input_tokens_seen": 15905192,
      "step": 24325
    },
    {
      "epoch": 12.751572327044025,
      "grad_norm": 0.9861467480659485,
      "learning_rate": 1.7477993109682288e-05,
      "loss": 0.6589,
      "num_input_tokens_seen": 15912008,
      "step": 24330
    },
    {
      "epoch": 12.7541928721174,
      "grad_norm": 0.5122083425521851,
      "learning_rate": 1.7467089454687043e-05,
      "loss": 0.6023,
      "num_input_tokens_seen": 15914920,
      "step": 24335
    },
    {
      "epoch": 12.756813417190775,
      "grad_norm": 0.3881281018257141,
      "learning_rate": 1.745618737549074e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 15918408,
      "step": 24340
    },
    {
      "epoch": 12.75943396226415,
      "grad_norm": 0.8946919441223145,
      "learning_rate": 1.7445286874373974e-05,
      "loss": 0.4273,
      "num_input_tokens_seen": 15921032,
      "step": 24345
    },
    {
      "epoch": 12.762054507337526,
      "grad_norm": 0.5855112671852112,
      "learning_rate": 1.743438795361701e-05,
      "loss": 0.5169,
      "num_input_tokens_seen": 15923592,
      "step": 24350
    },
    {
      "epoch": 12.764675052410901,
      "grad_norm": 0.6463603973388672,
      "learning_rate": 1.742349061549978e-05,
      "loss": 0.5265,
      "num_input_tokens_seen": 15926152,
      "step": 24355
    },
    {
      "epoch": 12.767295597484276,
      "grad_norm": 0.3706563413143158,
      "learning_rate": 1.741259486230186e-05,
      "loss": 0.513,
      "num_input_tokens_seen": 15929352,
      "step": 24360
    },
    {
      "epoch": 12.769916142557651,
      "grad_norm": 0.37197887897491455,
      "learning_rate": 1.7401700696302546e-05,
      "loss": 0.3925,
      "num_input_tokens_seen": 15932168,
      "step": 24365
    },
    {
      "epoch": 12.772536687631026,
      "grad_norm": 0.4808916449546814,
      "learning_rate": 1.739080811978075e-05,
      "loss": 0.5274,
      "num_input_tokens_seen": 15935656,
      "step": 24370
    },
    {
      "epoch": 12.775157232704402,
      "grad_norm": 0.53402179479599,
      "learning_rate": 1.7379917135015093e-05,
      "loss": 0.4274,
      "num_input_tokens_seen": 15938504,
      "step": 24375
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 0.46653327345848083,
      "learning_rate": 1.736902774428384e-05,
      "loss": 0.4273,
      "num_input_tokens_seen": 15945032,
      "step": 24380
    },
    {
      "epoch": 12.780398322851154,
      "grad_norm": 0.38628697395324707,
      "learning_rate": 1.735813994986491e-05,
      "loss": 0.3092,
      "num_input_tokens_seen": 15948584,
      "step": 24385
    },
    {
      "epoch": 12.783018867924529,
      "grad_norm": 0.7247396111488342,
      "learning_rate": 1.7347253754035933e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 15951816,
      "step": 24390
    },
    {
      "epoch": 12.785639412997904,
      "grad_norm": 0.528146505355835,
      "learning_rate": 1.7336369159074154e-05,
      "loss": 0.6378,
      "num_input_tokens_seen": 15954632,
      "step": 24395
    },
    {
      "epoch": 12.78825995807128,
      "grad_norm": 0.39858031272888184,
      "learning_rate": 1.7325486167256522e-05,
      "loss": 0.5893,
      "num_input_tokens_seen": 15958152,
      "step": 24400
    },
    {
      "epoch": 12.790880503144654,
      "grad_norm": 0.40274786949157715,
      "learning_rate": 1.7314604780859623e-05,
      "loss": 0.4242,
      "num_input_tokens_seen": 15961064,
      "step": 24405
    },
    {
      "epoch": 12.79350104821803,
      "grad_norm": 0.35778340697288513,
      "learning_rate": 1.730372500215973e-05,
      "loss": 0.5226,
      "num_input_tokens_seen": 15964200,
      "step": 24410
    },
    {
      "epoch": 12.796121593291405,
      "grad_norm": 0.4930480420589447,
      "learning_rate": 1.7292846833432763e-05,
      "loss": 0.3665,
      "num_input_tokens_seen": 15967400,
      "step": 24415
    },
    {
      "epoch": 12.79874213836478,
      "grad_norm": 0.5292941927909851,
      "learning_rate": 1.7281970276954305e-05,
      "loss": 0.5919,
      "num_input_tokens_seen": 15971048,
      "step": 24420
    },
    {
      "epoch": 12.801362683438155,
      "grad_norm": 0.7964038848876953,
      "learning_rate": 1.727109533499961e-05,
      "loss": 0.4744,
      "num_input_tokens_seen": 15973928,
      "step": 24425
    },
    {
      "epoch": 12.80398322851153,
      "grad_norm": 0.5474677085876465,
      "learning_rate": 1.72602220098436e-05,
      "loss": 0.499,
      "num_input_tokens_seen": 15977704,
      "step": 24430
    },
    {
      "epoch": 12.806603773584905,
      "grad_norm": 0.6596693396568298,
      "learning_rate": 1.7249350303760846e-05,
      "loss": 0.4583,
      "num_input_tokens_seen": 15980616,
      "step": 24435
    },
    {
      "epoch": 12.80922431865828,
      "grad_norm": 0.2551756501197815,
      "learning_rate": 1.7238480219025593e-05,
      "loss": 0.4177,
      "num_input_tokens_seen": 15985064,
      "step": 24440
    },
    {
      "epoch": 12.811844863731656,
      "grad_norm": 0.8498877286911011,
      "learning_rate": 1.7227611757911723e-05,
      "loss": 0.4108,
      "num_input_tokens_seen": 15987816,
      "step": 24445
    },
    {
      "epoch": 12.814465408805031,
      "grad_norm": 0.3809891641139984,
      "learning_rate": 1.7216744922692794e-05,
      "loss": 0.374,
      "num_input_tokens_seen": 15992072,
      "step": 24450
    },
    {
      "epoch": 12.817085953878406,
      "grad_norm": 0.5256991982460022,
      "learning_rate": 1.7205879715642037e-05,
      "loss": 0.4311,
      "num_input_tokens_seen": 15995048,
      "step": 24455
    },
    {
      "epoch": 12.819706498951781,
      "grad_norm": 0.3521091341972351,
      "learning_rate": 1.719501613903232e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 15998664,
      "step": 24460
    },
    {
      "epoch": 12.822327044025156,
      "grad_norm": 0.7580031752586365,
      "learning_rate": 1.7184154195136192e-05,
      "loss": 0.3669,
      "num_input_tokens_seen": 16001512,
      "step": 24465
    },
    {
      "epoch": 12.824947589098532,
      "grad_norm": 0.57419753074646,
      "learning_rate": 1.7173293886225837e-05,
      "loss": 0.5125,
      "num_input_tokens_seen": 16005096,
      "step": 24470
    },
    {
      "epoch": 12.827568134171909,
      "grad_norm": 0.29961782693862915,
      "learning_rate": 1.71624352145731e-05,
      "loss": 0.4033,
      "num_input_tokens_seen": 16008616,
      "step": 24475
    },
    {
      "epoch": 12.830188679245284,
      "grad_norm": 0.666193425655365,
      "learning_rate": 1.7151578182449508e-05,
      "loss": 0.5245,
      "num_input_tokens_seen": 16012200,
      "step": 24480
    },
    {
      "epoch": 12.832809224318659,
      "grad_norm": 0.5474724173545837,
      "learning_rate": 1.7140722792126212e-05,
      "loss": 0.4433,
      "num_input_tokens_seen": 16014824,
      "step": 24485
    },
    {
      "epoch": 12.835429769392034,
      "grad_norm": 0.4493643343448639,
      "learning_rate": 1.7129869045874053e-05,
      "loss": 0.4171,
      "num_input_tokens_seen": 16017736,
      "step": 24490
    },
    {
      "epoch": 12.83805031446541,
      "grad_norm": 0.47559329867362976,
      "learning_rate": 1.7119016945963506e-05,
      "loss": 0.3292,
      "num_input_tokens_seen": 16020552,
      "step": 24495
    },
    {
      "epoch": 12.840670859538784,
      "grad_norm": 0.4151448607444763,
      "learning_rate": 1.7108166494664694e-05,
      "loss": 0.4159,
      "num_input_tokens_seen": 16023528,
      "step": 24500
    },
    {
      "epoch": 12.84329140461216,
      "grad_norm": 0.40706121921539307,
      "learning_rate": 1.7097317694247417e-05,
      "loss": 0.3754,
      "num_input_tokens_seen": 16026472,
      "step": 24505
    },
    {
      "epoch": 12.845911949685535,
      "grad_norm": 0.6440954804420471,
      "learning_rate": 1.7086470546981122e-05,
      "loss": 0.6813,
      "num_input_tokens_seen": 16028776,
      "step": 24510
    },
    {
      "epoch": 12.84853249475891,
      "grad_norm": 0.36885538697242737,
      "learning_rate": 1.7075625055134904e-05,
      "loss": 0.453,
      "num_input_tokens_seen": 16032488,
      "step": 24515
    },
    {
      "epoch": 12.851153039832285,
      "grad_norm": 0.7841477394104004,
      "learning_rate": 1.706478122097752e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 16035816,
      "step": 24520
    },
    {
      "epoch": 12.85377358490566,
      "grad_norm": 0.6052539348602295,
      "learning_rate": 1.7053939046777384e-05,
      "loss": 0.3898,
      "num_input_tokens_seen": 16039144,
      "step": 24525
    },
    {
      "epoch": 12.856394129979035,
      "grad_norm": 0.48314961791038513,
      "learning_rate": 1.7043098534802533e-05,
      "loss": 0.4072,
      "num_input_tokens_seen": 16041960,
      "step": 24530
    },
    {
      "epoch": 12.85901467505241,
      "grad_norm": 0.8073805570602417,
      "learning_rate": 1.7032259687320695e-05,
      "loss": 0.5243,
      "num_input_tokens_seen": 16045064,
      "step": 24535
    },
    {
      "epoch": 12.861635220125786,
      "grad_norm": 0.31984636187553406,
      "learning_rate": 1.7021422506599228e-05,
      "loss": 0.4092,
      "num_input_tokens_seen": 16048520,
      "step": 24540
    },
    {
      "epoch": 12.864255765199161,
      "grad_norm": 0.5057579278945923,
      "learning_rate": 1.7010586994905155e-05,
      "loss": 0.3679,
      "num_input_tokens_seen": 16052584,
      "step": 24545
    },
    {
      "epoch": 12.866876310272536,
      "grad_norm": 0.3445151150226593,
      "learning_rate": 1.6999753154505123e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 16055080,
      "step": 24550
    },
    {
      "epoch": 12.869496855345911,
      "grad_norm": 0.4406908452510834,
      "learning_rate": 1.6988920987665476e-05,
      "loss": 0.2845,
      "num_input_tokens_seen": 16057736,
      "step": 24555
    },
    {
      "epoch": 12.872117400419286,
      "grad_norm": 0.4516426920890808,
      "learning_rate": 1.6978090496652156e-05,
      "loss": 0.3703,
      "num_input_tokens_seen": 16060552,
      "step": 24560
    },
    {
      "epoch": 12.874737945492662,
      "grad_norm": 0.4118500351905823,
      "learning_rate": 1.6967261683730778e-05,
      "loss": 0.463,
      "num_input_tokens_seen": 16063464,
      "step": 24565
    },
    {
      "epoch": 12.877358490566039,
      "grad_norm": 0.583449125289917,
      "learning_rate": 1.695643455116662e-05,
      "loss": 0.3584,
      "num_input_tokens_seen": 16066504,
      "step": 24570
    },
    {
      "epoch": 12.879979035639414,
      "grad_norm": 0.3114434480667114,
      "learning_rate": 1.694560910122458e-05,
      "loss": 0.4041,
      "num_input_tokens_seen": 16070312,
      "step": 24575
    },
    {
      "epoch": 12.882599580712789,
      "grad_norm": 0.36192819476127625,
      "learning_rate": 1.693478533616924e-05,
      "loss": 0.3447,
      "num_input_tokens_seen": 16073384,
      "step": 24580
    },
    {
      "epoch": 12.885220125786164,
      "grad_norm": 0.5117918252944946,
      "learning_rate": 1.692396325826479e-05,
      "loss": 0.48,
      "num_input_tokens_seen": 16076296,
      "step": 24585
    },
    {
      "epoch": 12.88784067085954,
      "grad_norm": 0.7401407957077026,
      "learning_rate": 1.6913142869775088e-05,
      "loss": 0.3614,
      "num_input_tokens_seen": 16080616,
      "step": 24590
    },
    {
      "epoch": 12.890461215932914,
      "grad_norm": 0.4607275724411011,
      "learning_rate": 1.6902324172963635e-05,
      "loss": 0.5253,
      "num_input_tokens_seen": 16084552,
      "step": 24595
    },
    {
      "epoch": 12.89308176100629,
      "grad_norm": 0.3639852702617645,
      "learning_rate": 1.6891507170093586e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 16087464,
      "step": 24600
    },
    {
      "epoch": 12.895702306079665,
      "grad_norm": 0.42191076278686523,
      "learning_rate": 1.6880691863427725e-05,
      "loss": 0.3259,
      "num_input_tokens_seen": 16090664,
      "step": 24605
    },
    {
      "epoch": 12.89832285115304,
      "grad_norm": 0.44347837567329407,
      "learning_rate": 1.686987825522851e-05,
      "loss": 0.3862,
      "num_input_tokens_seen": 16094312,
      "step": 24610
    },
    {
      "epoch": 12.900943396226415,
      "grad_norm": 0.4977334141731262,
      "learning_rate": 1.6859066347758002e-05,
      "loss": 0.5259,
      "num_input_tokens_seen": 16097000,
      "step": 24615
    },
    {
      "epoch": 12.90356394129979,
      "grad_norm": 0.3274233639240265,
      "learning_rate": 1.6848256143277933e-05,
      "loss": 0.5139,
      "num_input_tokens_seen": 16100008,
      "step": 24620
    },
    {
      "epoch": 12.906184486373165,
      "grad_norm": 0.34049081802368164,
      "learning_rate": 1.683744764404968e-05,
      "loss": 0.3994,
      "num_input_tokens_seen": 16102952,
      "step": 24625
    },
    {
      "epoch": 12.90880503144654,
      "grad_norm": 0.38105782866477966,
      "learning_rate": 1.6826640852334257e-05,
      "loss": 0.3755,
      "num_input_tokens_seen": 16105768,
      "step": 24630
    },
    {
      "epoch": 12.911425576519916,
      "grad_norm": 0.5905722975730896,
      "learning_rate": 1.6815835770392324e-05,
      "loss": 0.5322,
      "num_input_tokens_seen": 16108296,
      "step": 24635
    },
    {
      "epoch": 12.914046121593291,
      "grad_norm": 0.4668818414211273,
      "learning_rate": 1.6805032400484184e-05,
      "loss": 0.4019,
      "num_input_tokens_seen": 16110984,
      "step": 24640
    },
    {
      "epoch": 12.916666666666666,
      "grad_norm": 0.5169419050216675,
      "learning_rate": 1.6794230744869767e-05,
      "loss": 0.3284,
      "num_input_tokens_seen": 16114536,
      "step": 24645
    },
    {
      "epoch": 12.919287211740041,
      "grad_norm": 0.4056136906147003,
      "learning_rate": 1.6783430805808658e-05,
      "loss": 0.3114,
      "num_input_tokens_seen": 16117832,
      "step": 24650
    },
    {
      "epoch": 12.921907756813416,
      "grad_norm": 0.5352941751480103,
      "learning_rate": 1.6772632585560084e-05,
      "loss": 0.5363,
      "num_input_tokens_seen": 16121064,
      "step": 24655
    },
    {
      "epoch": 12.924528301886792,
      "grad_norm": 0.28108465671539307,
      "learning_rate": 1.6761836086382908e-05,
      "loss": 0.3459,
      "num_input_tokens_seen": 16124168,
      "step": 24660
    },
    {
      "epoch": 12.927148846960169,
      "grad_norm": 0.5234488248825073,
      "learning_rate": 1.6751041310535653e-05,
      "loss": 0.4999,
      "num_input_tokens_seen": 16127080,
      "step": 24665
    },
    {
      "epoch": 12.929769392033544,
      "grad_norm": 0.42116624116897583,
      "learning_rate": 1.6740248260276437e-05,
      "loss": 0.3988,
      "num_input_tokens_seen": 16131048,
      "step": 24670
    },
    {
      "epoch": 12.932389937106919,
      "grad_norm": 0.32161951065063477,
      "learning_rate": 1.6729456937863047e-05,
      "loss": 0.3562,
      "num_input_tokens_seen": 16134728,
      "step": 24675
    },
    {
      "epoch": 12.935010482180294,
      "grad_norm": 0.36101987957954407,
      "learning_rate": 1.6718667345552918e-05,
      "loss": 0.5531,
      "num_input_tokens_seen": 16140424,
      "step": 24680
    },
    {
      "epoch": 12.93763102725367,
      "grad_norm": 0.4375644624233246,
      "learning_rate": 1.6707879485603097e-05,
      "loss": 0.3631,
      "num_input_tokens_seen": 16143816,
      "step": 24685
    },
    {
      "epoch": 12.940251572327044,
      "grad_norm": 0.4224044978618622,
      "learning_rate": 1.669709336027029e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 16147144,
      "step": 24690
    },
    {
      "epoch": 12.94287211740042,
      "grad_norm": 0.4170876145362854,
      "learning_rate": 1.6686308971810833e-05,
      "loss": 0.4311,
      "num_input_tokens_seen": 16151400,
      "step": 24695
    },
    {
      "epoch": 12.945492662473795,
      "grad_norm": 0.5702067613601685,
      "learning_rate": 1.6675526322480677e-05,
      "loss": 0.5822,
      "num_input_tokens_seen": 16154152,
      "step": 24700
    },
    {
      "epoch": 12.94811320754717,
      "grad_norm": 0.6378082633018494,
      "learning_rate": 1.6664745414535454e-05,
      "loss": 0.3332,
      "num_input_tokens_seen": 16157000,
      "step": 24705
    },
    {
      "epoch": 12.950733752620545,
      "grad_norm": 0.5400651097297668,
      "learning_rate": 1.665396625023039e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 16159592,
      "step": 24710
    },
    {
      "epoch": 12.95335429769392,
      "grad_norm": 0.4028383195400238,
      "learning_rate": 1.6643188831820375e-05,
      "loss": 0.2729,
      "num_input_tokens_seen": 16164488,
      "step": 24715
    },
    {
      "epoch": 12.955974842767295,
      "grad_norm": 0.43063315749168396,
      "learning_rate": 1.6632413161559912e-05,
      "loss": 0.3881,
      "num_input_tokens_seen": 16167176,
      "step": 24720
    },
    {
      "epoch": 12.95859538784067,
      "grad_norm": 0.4700472354888916,
      "learning_rate": 1.6621639241703162e-05,
      "loss": 0.2948,
      "num_input_tokens_seen": 16173032,
      "step": 24725
    },
    {
      "epoch": 12.961215932914046,
      "grad_norm": 0.33493879437446594,
      "learning_rate": 1.6610867074503893e-05,
      "loss": 0.5221,
      "num_input_tokens_seen": 16175816,
      "step": 24730
    },
    {
      "epoch": 12.963836477987421,
      "grad_norm": 0.5514540076255798,
      "learning_rate": 1.660009666221552e-05,
      "loss": 0.4364,
      "num_input_tokens_seen": 16178952,
      "step": 24735
    },
    {
      "epoch": 12.966457023060796,
      "grad_norm": 0.5916787981987,
      "learning_rate": 1.6589328007091093e-05,
      "loss": 0.4487,
      "num_input_tokens_seen": 16182760,
      "step": 24740
    },
    {
      "epoch": 12.969077568134171,
      "grad_norm": 0.39040789008140564,
      "learning_rate": 1.65785611113833e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 16185768,
      "step": 24745
    },
    {
      "epoch": 12.971698113207546,
      "grad_norm": 0.5376111268997192,
      "learning_rate": 1.6567795977344447e-05,
      "loss": 0.4719,
      "num_input_tokens_seen": 16188904,
      "step": 24750
    },
    {
      "epoch": 12.974318658280922,
      "grad_norm": 0.6079456210136414,
      "learning_rate": 1.6557032607226466e-05,
      "loss": 0.5082,
      "num_input_tokens_seen": 16191816,
      "step": 24755
    },
    {
      "epoch": 12.976939203354299,
      "grad_norm": 0.5687761902809143,
      "learning_rate": 1.654627100328095e-05,
      "loss": 0.4819,
      "num_input_tokens_seen": 16194792,
      "step": 24760
    },
    {
      "epoch": 12.979559748427674,
      "grad_norm": 0.311908483505249,
      "learning_rate": 1.6535511167759085e-05,
      "loss": 0.4781,
      "num_input_tokens_seen": 16197608,
      "step": 24765
    },
    {
      "epoch": 12.982180293501049,
      "grad_norm": 0.7588208317756653,
      "learning_rate": 1.652475310291172e-05,
      "loss": 0.5086,
      "num_input_tokens_seen": 16200904,
      "step": 24770
    },
    {
      "epoch": 12.984800838574424,
      "grad_norm": 0.5741252303123474,
      "learning_rate": 1.6513996810989317e-05,
      "loss": 0.4423,
      "num_input_tokens_seen": 16204136,
      "step": 24775
    },
    {
      "epoch": 12.9874213836478,
      "grad_norm": 0.4663604199886322,
      "learning_rate": 1.6503242294241973e-05,
      "loss": 0.4704,
      "num_input_tokens_seen": 16207624,
      "step": 24780
    },
    {
      "epoch": 12.990041928721174,
      "grad_norm": 0.5119190812110901,
      "learning_rate": 1.6492489554919404e-05,
      "loss": 0.4746,
      "num_input_tokens_seen": 16210856,
      "step": 24785
    },
    {
      "epoch": 12.99266247379455,
      "grad_norm": 0.3360772132873535,
      "learning_rate": 1.6481738595270953e-05,
      "loss": 0.5392,
      "num_input_tokens_seen": 16214280,
      "step": 24790
    },
    {
      "epoch": 12.995283018867925,
      "grad_norm": 0.4592287838459015,
      "learning_rate": 1.6470989417545608e-05,
      "loss": 0.4848,
      "num_input_tokens_seen": 16217128,
      "step": 24795
    },
    {
      "epoch": 12.9979035639413,
      "grad_norm": 0.44940948486328125,
      "learning_rate": 1.646024202399197e-05,
      "loss": 0.3983,
      "num_input_tokens_seen": 16220840,
      "step": 24800
    },
    {
      "epoch": 13.0,
      "eval_loss": 0.4770670235157013,
      "eval_runtime": 9.7312,
      "eval_samples_per_second": 87.142,
      "eval_steps_per_second": 21.786,
      "num_input_tokens_seen": 16222408,
      "step": 24804
    },
    {
      "epoch": 13.000524109014675,
      "grad_norm": 0.4081278443336487,
      "learning_rate": 1.6449496416858284e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 16222792,
      "step": 24805
    },
    {
      "epoch": 13.00314465408805,
      "grad_norm": 0.3971358835697174,
      "learning_rate": 1.64387525983924e-05,
      "loss": 0.4429,
      "num_input_tokens_seen": 16226664,
      "step": 24810
    },
    {
      "epoch": 13.005765199161425,
      "grad_norm": 0.5868306159973145,
      "learning_rate": 1.6428010570841788e-05,
      "loss": 0.3785,
      "num_input_tokens_seen": 16230536,
      "step": 24815
    },
    {
      "epoch": 13.0083857442348,
      "grad_norm": 0.8847299814224243,
      "learning_rate": 1.6417270336453573e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 16233160,
      "step": 24820
    },
    {
      "epoch": 13.011006289308176,
      "grad_norm": 0.4194187521934509,
      "learning_rate": 1.6406531897474492e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 16237160,
      "step": 24825
    },
    {
      "epoch": 13.01362683438155,
      "grad_norm": 0.38886016607284546,
      "learning_rate": 1.639579525615089e-05,
      "loss": 0.3982,
      "num_input_tokens_seen": 16239720,
      "step": 24830
    },
    {
      "epoch": 13.016247379454926,
      "grad_norm": 0.45731237530708313,
      "learning_rate": 1.638506041472877e-05,
      "loss": 0.4436,
      "num_input_tokens_seen": 16242440,
      "step": 24835
    },
    {
      "epoch": 13.018867924528301,
      "grad_norm": 0.40627387166023254,
      "learning_rate": 1.6374327375453724e-05,
      "loss": 0.4065,
      "num_input_tokens_seen": 16245416,
      "step": 24840
    },
    {
      "epoch": 13.021488469601676,
      "grad_norm": 0.6291938424110413,
      "learning_rate": 1.6363596140570982e-05,
      "loss": 0.4017,
      "num_input_tokens_seen": 16248936,
      "step": 24845
    },
    {
      "epoch": 13.024109014675052,
      "grad_norm": 0.5047767162322998,
      "learning_rate": 1.6352866712325395e-05,
      "loss": 0.537,
      "num_input_tokens_seen": 16252072,
      "step": 24850
    },
    {
      "epoch": 13.026729559748428,
      "grad_norm": 0.47656452655792236,
      "learning_rate": 1.634213909296144e-05,
      "loss": 0.4493,
      "num_input_tokens_seen": 16255784,
      "step": 24855
    },
    {
      "epoch": 13.029350104821804,
      "grad_norm": 0.6533024907112122,
      "learning_rate": 1.6331413284723217e-05,
      "loss": 0.405,
      "num_input_tokens_seen": 16259528,
      "step": 24860
    },
    {
      "epoch": 13.031970649895179,
      "grad_norm": 0.4700944423675537,
      "learning_rate": 1.6320689289854442e-05,
      "loss": 0.4865,
      "num_input_tokens_seen": 16262920,
      "step": 24865
    },
    {
      "epoch": 13.034591194968554,
      "grad_norm": 0.5853138566017151,
      "learning_rate": 1.6309967110598438e-05,
      "loss": 0.3915,
      "num_input_tokens_seen": 16266152,
      "step": 24870
    },
    {
      "epoch": 13.03721174004193,
      "grad_norm": 1.090850591659546,
      "learning_rate": 1.6299246749198183e-05,
      "loss": 0.4594,
      "num_input_tokens_seen": 16269608,
      "step": 24875
    },
    {
      "epoch": 13.039832285115304,
      "grad_norm": 0.2939610183238983,
      "learning_rate": 1.6288528207896236e-05,
      "loss": 0.3557,
      "num_input_tokens_seen": 16274056,
      "step": 24880
    },
    {
      "epoch": 13.04245283018868,
      "grad_norm": 0.6217824816703796,
      "learning_rate": 1.627781148893481e-05,
      "loss": 0.3921,
      "num_input_tokens_seen": 16276680,
      "step": 24885
    },
    {
      "epoch": 13.045073375262055,
      "grad_norm": 0.7271012663841248,
      "learning_rate": 1.6267096594555703e-05,
      "loss": 0.4001,
      "num_input_tokens_seen": 16278696,
      "step": 24890
    },
    {
      "epoch": 13.04769392033543,
      "grad_norm": 0.5621318817138672,
      "learning_rate": 1.6256383527000372e-05,
      "loss": 0.5545,
      "num_input_tokens_seen": 16281800,
      "step": 24895
    },
    {
      "epoch": 13.050314465408805,
      "grad_norm": 0.501848042011261,
      "learning_rate": 1.6245672288509843e-05,
      "loss": 0.3437,
      "num_input_tokens_seen": 16284968,
      "step": 24900
    },
    {
      "epoch": 13.05293501048218,
      "grad_norm": 0.8452006578445435,
      "learning_rate": 1.62349628813248e-05,
      "loss": 0.408,
      "num_input_tokens_seen": 16287592,
      "step": 24905
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 0.33356887102127075,
      "learning_rate": 1.6224255307685525e-05,
      "loss": 0.4847,
      "num_input_tokens_seen": 16291080,
      "step": 24910
    },
    {
      "epoch": 13.05817610062893,
      "grad_norm": 2.2786033153533936,
      "learning_rate": 1.6213549569831922e-05,
      "loss": 0.4965,
      "num_input_tokens_seen": 16293768,
      "step": 24915
    },
    {
      "epoch": 13.060796645702306,
      "grad_norm": 0.3208876848220825,
      "learning_rate": 1.62028456700035e-05,
      "loss": 0.4666,
      "num_input_tokens_seen": 16298440,
      "step": 24920
    },
    {
      "epoch": 13.06341719077568,
      "grad_norm": 0.4787388741970062,
      "learning_rate": 1.619214361043942e-05,
      "loss": 0.4473,
      "num_input_tokens_seen": 16301768,
      "step": 24925
    },
    {
      "epoch": 13.066037735849056,
      "grad_norm": 0.49423861503601074,
      "learning_rate": 1.61814433933784e-05,
      "loss": 0.41,
      "num_input_tokens_seen": 16304872,
      "step": 24930
    },
    {
      "epoch": 13.068658280922431,
      "grad_norm": 0.5558385252952576,
      "learning_rate": 1.617074502105881e-05,
      "loss": 0.5354,
      "num_input_tokens_seen": 16308456,
      "step": 24935
    },
    {
      "epoch": 13.071278825995806,
      "grad_norm": 0.37036073207855225,
      "learning_rate": 1.616004849571864e-05,
      "loss": 0.3947,
      "num_input_tokens_seen": 16311816,
      "step": 24940
    },
    {
      "epoch": 13.073899371069182,
      "grad_norm": 0.5660409331321716,
      "learning_rate": 1.6149353819595463e-05,
      "loss": 0.52,
      "num_input_tokens_seen": 16315048,
      "step": 24945
    },
    {
      "epoch": 13.076519916142558,
      "grad_norm": 0.639015793800354,
      "learning_rate": 1.6138660994926508e-05,
      "loss": 0.5358,
      "num_input_tokens_seen": 16318120,
      "step": 24950
    },
    {
      "epoch": 13.079140461215934,
      "grad_norm": 0.2639061212539673,
      "learning_rate": 1.612797002394857e-05,
      "loss": 0.3201,
      "num_input_tokens_seen": 16321256,
      "step": 24955
    },
    {
      "epoch": 13.081761006289309,
      "grad_norm": 0.5028044581413269,
      "learning_rate": 1.611728090889808e-05,
      "loss": 0.4273,
      "num_input_tokens_seen": 16325096,
      "step": 24960
    },
    {
      "epoch": 13.084381551362684,
      "grad_norm": 0.3243633210659027,
      "learning_rate": 1.610659365201109e-05,
      "loss": 0.3582,
      "num_input_tokens_seen": 16328424,
      "step": 24965
    },
    {
      "epoch": 13.08700209643606,
      "grad_norm": 0.41838380694389343,
      "learning_rate": 1.609590825552324e-05,
      "loss": 0.3682,
      "num_input_tokens_seen": 16332200,
      "step": 24970
    },
    {
      "epoch": 13.089622641509434,
      "grad_norm": 0.49350705742836,
      "learning_rate": 1.6085224721669807e-05,
      "loss": 0.5216,
      "num_input_tokens_seen": 16334760,
      "step": 24975
    },
    {
      "epoch": 13.09224318658281,
      "grad_norm": 0.4749518930912018,
      "learning_rate": 1.607454305268566e-05,
      "loss": 0.4736,
      "num_input_tokens_seen": 16337384,
      "step": 24980
    },
    {
      "epoch": 13.094863731656185,
      "grad_norm": 0.633493185043335,
      "learning_rate": 1.606386325080528e-05,
      "loss": 0.4276,
      "num_input_tokens_seen": 16340488,
      "step": 24985
    },
    {
      "epoch": 13.09748427672956,
      "grad_norm": 0.5832304358482361,
      "learning_rate": 1.6053185318262746e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 16343336,
      "step": 24990
    },
    {
      "epoch": 13.100104821802935,
      "grad_norm": 0.3728271424770355,
      "learning_rate": 1.6042509257291783e-05,
      "loss": 0.4072,
      "num_input_tokens_seen": 16347176,
      "step": 24995
    },
    {
      "epoch": 13.10272536687631,
      "grad_norm": 0.4040848910808563,
      "learning_rate": 1.6031835070125685e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 16349864,
      "step": 25000
    },
    {
      "epoch": 13.105345911949685,
      "grad_norm": 0.6153953671455383,
      "learning_rate": 1.6021162758997382e-05,
      "loss": 0.5289,
      "num_input_tokens_seen": 16352904,
      "step": 25005
    },
    {
      "epoch": 13.10796645702306,
      "grad_norm": 0.4176477789878845,
      "learning_rate": 1.6010492326139398e-05,
      "loss": 0.5302,
      "num_input_tokens_seen": 16356072,
      "step": 25010
    },
    {
      "epoch": 13.110587002096436,
      "grad_norm": 0.3589712977409363,
      "learning_rate": 1.599982377378385e-05,
      "loss": 0.3704,
      "num_input_tokens_seen": 16360360,
      "step": 25015
    },
    {
      "epoch": 13.11320754716981,
      "grad_norm": 1.493160605430603,
      "learning_rate": 1.5989157104162496e-05,
      "loss": 0.586,
      "num_input_tokens_seen": 16363240,
      "step": 25020
    },
    {
      "epoch": 13.115828092243186,
      "grad_norm": 0.47053495049476624,
      "learning_rate": 1.5978492319506673e-05,
      "loss": 0.3674,
      "num_input_tokens_seen": 16365992,
      "step": 25025
    },
    {
      "epoch": 13.118448637316561,
      "grad_norm": 0.5552544593811035,
      "learning_rate": 1.596782942204733e-05,
      "loss": 0.4727,
      "num_input_tokens_seen": 16369064,
      "step": 25030
    },
    {
      "epoch": 13.121069182389936,
      "grad_norm": 0.7286280393600464,
      "learning_rate": 1.595716841401504e-05,
      "loss": 0.4016,
      "num_input_tokens_seen": 16372456,
      "step": 25035
    },
    {
      "epoch": 13.123689727463312,
      "grad_norm": 0.3721131980419159,
      "learning_rate": 1.594650929763994e-05,
      "loss": 0.4487,
      "num_input_tokens_seen": 16376328,
      "step": 25040
    },
    {
      "epoch": 13.126310272536688,
      "grad_norm": 0.36537835001945496,
      "learning_rate": 1.5935852075151812e-05,
      "loss": 0.6072,
      "num_input_tokens_seen": 16378952,
      "step": 25045
    },
    {
      "epoch": 13.128930817610064,
      "grad_norm": 0.5001424551010132,
      "learning_rate": 1.5925196748780016e-05,
      "loss": 0.5982,
      "num_input_tokens_seen": 16381640,
      "step": 25050
    },
    {
      "epoch": 13.131551362683439,
      "grad_norm": 0.3200109004974365,
      "learning_rate": 1.5914543320753528e-05,
      "loss": 0.3572,
      "num_input_tokens_seen": 16385032,
      "step": 25055
    },
    {
      "epoch": 13.134171907756814,
      "grad_norm": 0.6351261138916016,
      "learning_rate": 1.5903891793300932e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 16388520,
      "step": 25060
    },
    {
      "epoch": 13.13679245283019,
      "grad_norm": 0.6265797019004822,
      "learning_rate": 1.5893242168650403e-05,
      "loss": 0.5485,
      "num_input_tokens_seen": 16391400,
      "step": 25065
    },
    {
      "epoch": 13.139412997903564,
      "grad_norm": 0.28260087966918945,
      "learning_rate": 1.5882594449029705e-05,
      "loss": 0.4788,
      "num_input_tokens_seen": 16394376,
      "step": 25070
    },
    {
      "epoch": 13.14203354297694,
      "grad_norm": 0.4210667610168457,
      "learning_rate": 1.587194863666624e-05,
      "loss": 0.3637,
      "num_input_tokens_seen": 16398024,
      "step": 25075
    },
    {
      "epoch": 13.144654088050315,
      "grad_norm": 1.230601191520691,
      "learning_rate": 1.5861304733786974e-05,
      "loss": 0.4234,
      "num_input_tokens_seen": 16400648,
      "step": 25080
    },
    {
      "epoch": 13.14727463312369,
      "grad_norm": 0.400634765625,
      "learning_rate": 1.5850662742618505e-05,
      "loss": 0.3778,
      "num_input_tokens_seen": 16404552,
      "step": 25085
    },
    {
      "epoch": 13.149895178197065,
      "grad_norm": 0.45523086190223694,
      "learning_rate": 1.5840022665387004e-05,
      "loss": 0.4301,
      "num_input_tokens_seen": 16407240,
      "step": 25090
    },
    {
      "epoch": 13.15251572327044,
      "grad_norm": 0.45749977231025696,
      "learning_rate": 1.5829384504318268e-05,
      "loss": 0.583,
      "num_input_tokens_seen": 16410344,
      "step": 25095
    },
    {
      "epoch": 13.155136268343815,
      "grad_norm": 0.6652722954750061,
      "learning_rate": 1.5818748261637662e-05,
      "loss": 0.4636,
      "num_input_tokens_seen": 16413128,
      "step": 25100
    },
    {
      "epoch": 13.15775681341719,
      "grad_norm": 0.43221092224121094,
      "learning_rate": 1.5808113939570173e-05,
      "loss": 0.3643,
      "num_input_tokens_seen": 16416040,
      "step": 25105
    },
    {
      "epoch": 13.160377358490566,
      "grad_norm": 0.3384433090686798,
      "learning_rate": 1.5797481540340386e-05,
      "loss": 0.4656,
      "num_input_tokens_seen": 16420776,
      "step": 25110
    },
    {
      "epoch": 13.16299790356394,
      "grad_norm": 0.32022151350975037,
      "learning_rate": 1.5786851066172466e-05,
      "loss": 0.4594,
      "num_input_tokens_seen": 16423496,
      "step": 25115
    },
    {
      "epoch": 13.165618448637316,
      "grad_norm": 0.3861553966999054,
      "learning_rate": 1.5776222519290208e-05,
      "loss": 0.4299,
      "num_input_tokens_seen": 16425864,
      "step": 25120
    },
    {
      "epoch": 13.168238993710691,
      "grad_norm": 0.5185946822166443,
      "learning_rate": 1.576559590191696e-05,
      "loss": 0.4691,
      "num_input_tokens_seen": 16428872,
      "step": 25125
    },
    {
      "epoch": 13.170859538784066,
      "grad_norm": 0.4581664502620697,
      "learning_rate": 1.5754971216275692e-05,
      "loss": 0.3128,
      "num_input_tokens_seen": 16432808,
      "step": 25130
    },
    {
      "epoch": 13.173480083857442,
      "grad_norm": 0.23175321519374847,
      "learning_rate": 1.5744348464588975e-05,
      "loss": 0.5622,
      "num_input_tokens_seen": 16436136,
      "step": 25135
    },
    {
      "epoch": 13.176100628930818,
      "grad_norm": 0.48330119252204895,
      "learning_rate": 1.5733727649078968e-05,
      "loss": 0.484,
      "num_input_tokens_seen": 16440008,
      "step": 25140
    },
    {
      "epoch": 13.178721174004194,
      "grad_norm": 0.9740814566612244,
      "learning_rate": 1.5723108771967414e-05,
      "loss": 0.6493,
      "num_input_tokens_seen": 16443240,
      "step": 25145
    },
    {
      "epoch": 13.181341719077569,
      "grad_norm": 0.37513497471809387,
      "learning_rate": 1.5712491835475685e-05,
      "loss": 0.5087,
      "num_input_tokens_seen": 16446600,
      "step": 25150
    },
    {
      "epoch": 13.183962264150944,
      "grad_norm": 0.6027389764785767,
      "learning_rate": 1.5701876841824698e-05,
      "loss": 0.4654,
      "num_input_tokens_seen": 16449480,
      "step": 25155
    },
    {
      "epoch": 13.18658280922432,
      "grad_norm": 0.4472504258155823,
      "learning_rate": 1.5691263793234993e-05,
      "loss": 0.3701,
      "num_input_tokens_seen": 16452360,
      "step": 25160
    },
    {
      "epoch": 13.189203354297694,
      "grad_norm": 0.47564756870269775,
      "learning_rate": 1.5680652691926706e-05,
      "loss": 0.4887,
      "num_input_tokens_seen": 16455048,
      "step": 25165
    },
    {
      "epoch": 13.19182389937107,
      "grad_norm": 0.5877525806427002,
      "learning_rate": 1.5670043540119553e-05,
      "loss": 0.366,
      "num_input_tokens_seen": 16458248,
      "step": 25170
    },
    {
      "epoch": 13.194444444444445,
      "grad_norm": 0.5160966515541077,
      "learning_rate": 1.5659436340032855e-05,
      "loss": 0.449,
      "num_input_tokens_seen": 16461896,
      "step": 25175
    },
    {
      "epoch": 13.19706498951782,
      "grad_norm": 0.6382184624671936,
      "learning_rate": 1.5648831093885515e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 16464712,
      "step": 25180
    },
    {
      "epoch": 13.199685534591195,
      "grad_norm": 0.6356901526451111,
      "learning_rate": 1.5638227803896015e-05,
      "loss": 0.376,
      "num_input_tokens_seen": 16467784,
      "step": 25185
    },
    {
      "epoch": 13.20230607966457,
      "grad_norm": 0.2584974765777588,
      "learning_rate": 1.562762647228246e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 16471400,
      "step": 25190
    },
    {
      "epoch": 13.204926624737945,
      "grad_norm": 0.4464767277240753,
      "learning_rate": 1.5617027101262515e-05,
      "loss": 0.3688,
      "num_input_tokens_seen": 16475016,
      "step": 25195
    },
    {
      "epoch": 13.20754716981132,
      "grad_norm": 0.7142561674118042,
      "learning_rate": 1.5606429693053463e-05,
      "loss": 0.4958,
      "num_input_tokens_seen": 16479016,
      "step": 25200
    },
    {
      "epoch": 13.210167714884696,
      "grad_norm": 0.6117770671844482,
      "learning_rate": 1.5595834249872145e-05,
      "loss": 0.369,
      "num_input_tokens_seen": 16482696,
      "step": 25205
    },
    {
      "epoch": 13.21278825995807,
      "grad_norm": 0.7831028699874878,
      "learning_rate": 1.558524077393502e-05,
      "loss": 0.5501,
      "num_input_tokens_seen": 16486408,
      "step": 25210
    },
    {
      "epoch": 13.215408805031446,
      "grad_norm": 0.7472196817398071,
      "learning_rate": 1.557464926745811e-05,
      "loss": 0.4267,
      "num_input_tokens_seen": 16489992,
      "step": 25215
    },
    {
      "epoch": 13.218029350104821,
      "grad_norm": 0.44998618960380554,
      "learning_rate": 1.5564059732657045e-05,
      "loss": 0.5457,
      "num_input_tokens_seen": 16493512,
      "step": 25220
    },
    {
      "epoch": 13.220649895178196,
      "grad_norm": 0.5834546089172363,
      "learning_rate": 1.555347217174703e-05,
      "loss": 0.418,
      "num_input_tokens_seen": 16499240,
      "step": 25225
    },
    {
      "epoch": 13.223270440251572,
      "grad_norm": 0.48055216670036316,
      "learning_rate": 1.554288658694287e-05,
      "loss": 0.3806,
      "num_input_tokens_seen": 16502120,
      "step": 25230
    },
    {
      "epoch": 13.225890985324948,
      "grad_norm": 0.51677006483078,
      "learning_rate": 1.553230298045895e-05,
      "loss": 0.7261,
      "num_input_tokens_seen": 16505448,
      "step": 25235
    },
    {
      "epoch": 13.228511530398324,
      "grad_norm": 0.5162584185600281,
      "learning_rate": 1.5521721354509223e-05,
      "loss": 0.4438,
      "num_input_tokens_seen": 16507592,
      "step": 25240
    },
    {
      "epoch": 13.231132075471699,
      "grad_norm": 0.9146361947059631,
      "learning_rate": 1.5511141711307258e-05,
      "loss": 0.5771,
      "num_input_tokens_seen": 16510600,
      "step": 25245
    },
    {
      "epoch": 13.233752620545074,
      "grad_norm": 0.8619827032089233,
      "learning_rate": 1.5500564053066184e-05,
      "loss": 0.5878,
      "num_input_tokens_seen": 16513288,
      "step": 25250
    },
    {
      "epoch": 13.23637316561845,
      "grad_norm": 0.5849536657333374,
      "learning_rate": 1.5489988381998743e-05,
      "loss": 0.4436,
      "num_input_tokens_seen": 16515752,
      "step": 25255
    },
    {
      "epoch": 13.238993710691824,
      "grad_norm": 0.3046723008155823,
      "learning_rate": 1.5479414700317237e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 16519496,
      "step": 25260
    },
    {
      "epoch": 13.2416142557652,
      "grad_norm": 0.44326353073120117,
      "learning_rate": 1.546884301023356e-05,
      "loss": 0.4212,
      "num_input_tokens_seen": 16526120,
      "step": 25265
    },
    {
      "epoch": 13.244234800838575,
      "grad_norm": 0.5247043967247009,
      "learning_rate": 1.5458273313959186e-05,
      "loss": 0.5507,
      "num_input_tokens_seen": 16529544,
      "step": 25270
    },
    {
      "epoch": 13.24685534591195,
      "grad_norm": 0.40003785490989685,
      "learning_rate": 1.5447705613705177e-05,
      "loss": 0.4696,
      "num_input_tokens_seen": 16533288,
      "step": 25275
    },
    {
      "epoch": 13.249475890985325,
      "grad_norm": 0.38701149821281433,
      "learning_rate": 1.5437139911682176e-05,
      "loss": 0.4054,
      "num_input_tokens_seen": 16537256,
      "step": 25280
    },
    {
      "epoch": 13.2520964360587,
      "grad_norm": 0.713243305683136,
      "learning_rate": 1.54265762101004e-05,
      "loss": 0.3822,
      "num_input_tokens_seen": 16540744,
      "step": 25285
    },
    {
      "epoch": 13.254716981132075,
      "grad_norm": 0.3114863634109497,
      "learning_rate": 1.5416014511169663e-05,
      "loss": 0.4815,
      "num_input_tokens_seen": 16543784,
      "step": 25290
    },
    {
      "epoch": 13.25733752620545,
      "grad_norm": 0.4281182289123535,
      "learning_rate": 1.5405454817099363e-05,
      "loss": 0.4745,
      "num_input_tokens_seen": 16547368,
      "step": 25295
    },
    {
      "epoch": 13.259958071278826,
      "grad_norm": 0.3838338553905487,
      "learning_rate": 1.5394897130098446e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 16550952,
      "step": 25300
    },
    {
      "epoch": 13.2625786163522,
      "grad_norm": 0.6012200713157654,
      "learning_rate": 1.5384341452375465e-05,
      "loss": 0.4485,
      "num_input_tokens_seen": 16553960,
      "step": 25305
    },
    {
      "epoch": 13.265199161425576,
      "grad_norm": 0.5995935201644897,
      "learning_rate": 1.5373787786138556e-05,
      "loss": 0.3954,
      "num_input_tokens_seen": 16556680,
      "step": 25310
    },
    {
      "epoch": 13.267819706498951,
      "grad_norm": 0.3467071056365967,
      "learning_rate": 1.5363236133595416e-05,
      "loss": 0.4526,
      "num_input_tokens_seen": 16560680,
      "step": 25315
    },
    {
      "epoch": 13.270440251572326,
      "grad_norm": 0.48135265707969666,
      "learning_rate": 1.5352686496953344e-05,
      "loss": 0.3958,
      "num_input_tokens_seen": 16564936,
      "step": 25320
    },
    {
      "epoch": 13.273060796645701,
      "grad_norm": 0.5761799812316895,
      "learning_rate": 1.534213887841919e-05,
      "loss": 0.4313,
      "num_input_tokens_seen": 16567880,
      "step": 25325
    },
    {
      "epoch": 13.275681341719078,
      "grad_norm": 0.39615678787231445,
      "learning_rate": 1.5331593280199395e-05,
      "loss": 0.441,
      "num_input_tokens_seen": 16570984,
      "step": 25330
    },
    {
      "epoch": 13.278301886792454,
      "grad_norm": 0.3252353370189667,
      "learning_rate": 1.532104970449999e-05,
      "loss": 0.3343,
      "num_input_tokens_seen": 16574632,
      "step": 25335
    },
    {
      "epoch": 13.280922431865829,
      "grad_norm": 0.7076736688613892,
      "learning_rate": 1.5310508153526555e-05,
      "loss": 0.4198,
      "num_input_tokens_seen": 16577320,
      "step": 25340
    },
    {
      "epoch": 13.283542976939204,
      "grad_norm": 0.44885778427124023,
      "learning_rate": 1.5299968629484276e-05,
      "loss": 0.4769,
      "num_input_tokens_seen": 16580136,
      "step": 25345
    },
    {
      "epoch": 13.286163522012579,
      "grad_norm": 0.3962424099445343,
      "learning_rate": 1.5289431134577904e-05,
      "loss": 0.432,
      "num_input_tokens_seen": 16583080,
      "step": 25350
    },
    {
      "epoch": 13.288784067085954,
      "grad_norm": 0.6057462692260742,
      "learning_rate": 1.527889567101174e-05,
      "loss": 0.7658,
      "num_input_tokens_seen": 16585512,
      "step": 25355
    },
    {
      "epoch": 13.29140461215933,
      "grad_norm": 0.659416139125824,
      "learning_rate": 1.5268362240989707e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 16588456,
      "step": 25360
    },
    {
      "epoch": 13.294025157232705,
      "grad_norm": 0.33165857195854187,
      "learning_rate": 1.525783084671526e-05,
      "loss": 0.4971,
      "num_input_tokens_seen": 16592040,
      "step": 25365
    },
    {
      "epoch": 13.29664570230608,
      "grad_norm": 0.3778662383556366,
      "learning_rate": 1.5247301490391458e-05,
      "loss": 0.3856,
      "num_input_tokens_seen": 16594728,
      "step": 25370
    },
    {
      "epoch": 13.299266247379455,
      "grad_norm": 0.4515862464904785,
      "learning_rate": 1.5236774174220922e-05,
      "loss": 0.437,
      "num_input_tokens_seen": 16597512,
      "step": 25375
    },
    {
      "epoch": 13.30188679245283,
      "grad_norm": 0.7184445261955261,
      "learning_rate": 1.5226248900405852e-05,
      "loss": 0.5968,
      "num_input_tokens_seen": 16601064,
      "step": 25380
    },
    {
      "epoch": 13.304507337526205,
      "grad_norm": 0.7421137094497681,
      "learning_rate": 1.5215725671147996e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 16603944,
      "step": 25385
    },
    {
      "epoch": 13.30712788259958,
      "grad_norm": 0.37277650833129883,
      "learning_rate": 1.5205204488648709e-05,
      "loss": 0.5187,
      "num_input_tokens_seen": 16608712,
      "step": 25390
    },
    {
      "epoch": 13.309748427672956,
      "grad_norm": 0.5160108804702759,
      "learning_rate": 1.5194685355108893e-05,
      "loss": 0.4933,
      "num_input_tokens_seen": 16611752,
      "step": 25395
    },
    {
      "epoch": 13.31236897274633,
      "grad_norm": 0.9976527690887451,
      "learning_rate": 1.518416827272904e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 16615144,
      "step": 25400
    },
    {
      "epoch": 13.314989517819706,
      "grad_norm": 0.5025090575218201,
      "learning_rate": 1.5173653243709207e-05,
      "loss": 0.452,
      "num_input_tokens_seen": 16619496,
      "step": 25405
    },
    {
      "epoch": 13.317610062893081,
      "grad_norm": 0.5253115296363831,
      "learning_rate": 1.5163140270249e-05,
      "loss": 0.3759,
      "num_input_tokens_seen": 16622088,
      "step": 25410
    },
    {
      "epoch": 13.320230607966456,
      "grad_norm": 0.6915678977966309,
      "learning_rate": 1.5152629354547626e-05,
      "loss": 0.3247,
      "num_input_tokens_seen": 16626120,
      "step": 25415
    },
    {
      "epoch": 13.322851153039831,
      "grad_norm": 0.4523760676383972,
      "learning_rate": 1.5142120498803842e-05,
      "loss": 0.6312,
      "num_input_tokens_seen": 16629096,
      "step": 25420
    },
    {
      "epoch": 13.325471698113208,
      "grad_norm": 0.33715641498565674,
      "learning_rate": 1.5131613705215991e-05,
      "loss": 0.5356,
      "num_input_tokens_seen": 16632296,
      "step": 25425
    },
    {
      "epoch": 13.328092243186584,
      "grad_norm": 0.5063401460647583,
      "learning_rate": 1.5121108975981965e-05,
      "loss": 0.5128,
      "num_input_tokens_seen": 16635368,
      "step": 25430
    },
    {
      "epoch": 13.330712788259959,
      "grad_norm": 0.7181898951530457,
      "learning_rate": 1.5110606313299248e-05,
      "loss": 0.4259,
      "num_input_tokens_seen": 16639080,
      "step": 25435
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 0.38652679324150085,
      "learning_rate": 1.5100105719364855e-05,
      "loss": 0.3884,
      "num_input_tokens_seen": 16642440,
      "step": 25440
    },
    {
      "epoch": 13.335953878406709,
      "grad_norm": 0.571295440196991,
      "learning_rate": 1.50896071963754e-05,
      "loss": 0.3612,
      "num_input_tokens_seen": 16645928,
      "step": 25445
    },
    {
      "epoch": 13.338574423480084,
      "grad_norm": 0.3366576135158539,
      "learning_rate": 1.5079110746527055e-05,
      "loss": 0.3191,
      "num_input_tokens_seen": 16650376,
      "step": 25450
    },
    {
      "epoch": 13.34119496855346,
      "grad_norm": 0.5752995014190674,
      "learning_rate": 1.5068616372015565e-05,
      "loss": 0.3815,
      "num_input_tokens_seen": 16653096,
      "step": 25455
    },
    {
      "epoch": 13.343815513626835,
      "grad_norm": 0.4575340747833252,
      "learning_rate": 1.5058124075036223e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 16656360,
      "step": 25460
    },
    {
      "epoch": 13.34643605870021,
      "grad_norm": 0.37630435824394226,
      "learning_rate": 1.5047633857783908e-05,
      "loss": 0.3764,
      "num_input_tokens_seen": 16659784,
      "step": 25465
    },
    {
      "epoch": 13.349056603773585,
      "grad_norm": 0.44058969616889954,
      "learning_rate": 1.5037145722453044e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 16663400,
      "step": 25470
    },
    {
      "epoch": 13.35167714884696,
      "grad_norm": 0.25374341011047363,
      "learning_rate": 1.5026659671237624e-05,
      "loss": 0.3983,
      "num_input_tokens_seen": 16667592,
      "step": 25475
    },
    {
      "epoch": 13.354297693920335,
      "grad_norm": 0.6176823377609253,
      "learning_rate": 1.501617570633123e-05,
      "loss": 0.5165,
      "num_input_tokens_seen": 16671080,
      "step": 25480
    },
    {
      "epoch": 13.35691823899371,
      "grad_norm": 0.5290617942810059,
      "learning_rate": 1.500569382992697e-05,
      "loss": 0.4195,
      "num_input_tokens_seen": 16674664,
      "step": 25485
    },
    {
      "epoch": 13.359538784067086,
      "grad_norm": 0.6489535570144653,
      "learning_rate": 1.4995214044217547e-05,
      "loss": 0.4273,
      "num_input_tokens_seen": 16677864,
      "step": 25490
    },
    {
      "epoch": 13.36215932914046,
      "grad_norm": 0.7316802740097046,
      "learning_rate": 1.49847363513952e-05,
      "loss": 0.3135,
      "num_input_tokens_seen": 16681544,
      "step": 25495
    },
    {
      "epoch": 13.364779874213836,
      "grad_norm": 0.29398733377456665,
      "learning_rate": 1.4974260753651748e-05,
      "loss": 0.3187,
      "num_input_tokens_seen": 16684104,
      "step": 25500
    },
    {
      "epoch": 13.367400419287211,
      "grad_norm": 0.512448787689209,
      "learning_rate": 1.4963787253178568e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 16688136,
      "step": 25505
    },
    {
      "epoch": 13.370020964360586,
      "grad_norm": 0.8582059741020203,
      "learning_rate": 1.4953315852166594e-05,
      "loss": 0.4183,
      "num_input_tokens_seen": 16691592,
      "step": 25510
    },
    {
      "epoch": 13.372641509433961,
      "grad_norm": 0.46289968490600586,
      "learning_rate": 1.4942846552806322e-05,
      "loss": 0.3449,
      "num_input_tokens_seen": 16695112,
      "step": 25515
    },
    {
      "epoch": 13.375262054507338,
      "grad_norm": 0.5604698061943054,
      "learning_rate": 1.4932379357287829e-05,
      "loss": 0.4413,
      "num_input_tokens_seen": 16697480,
      "step": 25520
    },
    {
      "epoch": 13.377882599580714,
      "grad_norm": 0.28053635358810425,
      "learning_rate": 1.49219142678007e-05,
      "loss": 0.5482,
      "num_input_tokens_seen": 16700168,
      "step": 25525
    },
    {
      "epoch": 13.380503144654089,
      "grad_norm": 0.37530943751335144,
      "learning_rate": 1.4911451286534128e-05,
      "loss": 0.5432,
      "num_input_tokens_seen": 16703400,
      "step": 25530
    },
    {
      "epoch": 13.383123689727464,
      "grad_norm": 0.5109317302703857,
      "learning_rate": 1.490099041567686e-05,
      "loss": 0.5225,
      "num_input_tokens_seen": 16706152,
      "step": 25535
    },
    {
      "epoch": 13.385744234800839,
      "grad_norm": 0.3812336027622223,
      "learning_rate": 1.4890531657417176e-05,
      "loss": 0.4596,
      "num_input_tokens_seen": 16709288,
      "step": 25540
    },
    {
      "epoch": 13.388364779874214,
      "grad_norm": 0.2654878497123718,
      "learning_rate": 1.4880075013942938e-05,
      "loss": 0.4172,
      "num_input_tokens_seen": 16712424,
      "step": 25545
    },
    {
      "epoch": 13.39098532494759,
      "grad_norm": 0.36996421217918396,
      "learning_rate": 1.4869620487441558e-05,
      "loss": 0.4891,
      "num_input_tokens_seen": 16716168,
      "step": 25550
    },
    {
      "epoch": 13.393605870020965,
      "grad_norm": 0.9289466142654419,
      "learning_rate": 1.4859168080099988e-05,
      "loss": 0.539,
      "num_input_tokens_seen": 16719464,
      "step": 25555
    },
    {
      "epoch": 13.39622641509434,
      "grad_norm": 0.47668954730033875,
      "learning_rate": 1.484871779410477e-05,
      "loss": 0.3501,
      "num_input_tokens_seen": 16723368,
      "step": 25560
    },
    {
      "epoch": 13.398846960167715,
      "grad_norm": 0.4282396733760834,
      "learning_rate": 1.4838269631641971e-05,
      "loss": 0.4456,
      "num_input_tokens_seen": 16727080,
      "step": 25565
    },
    {
      "epoch": 13.40146750524109,
      "grad_norm": 0.3960167169570923,
      "learning_rate": 1.482782359489724e-05,
      "loss": 0.4135,
      "num_input_tokens_seen": 16730600,
      "step": 25570
    },
    {
      "epoch": 13.404088050314465,
      "grad_norm": 0.6101580262184143,
      "learning_rate": 1.4817379686055754e-05,
      "loss": 0.4214,
      "num_input_tokens_seen": 16733352,
      "step": 25575
    },
    {
      "epoch": 13.40670859538784,
      "grad_norm": 0.25427326560020447,
      "learning_rate": 1.4806937907302282e-05,
      "loss": 0.3583,
      "num_input_tokens_seen": 16736232,
      "step": 25580
    },
    {
      "epoch": 13.409329140461216,
      "grad_norm": 0.4299165904521942,
      "learning_rate": 1.4796498260821103e-05,
      "loss": 0.3252,
      "num_input_tokens_seen": 16741288,
      "step": 25585
    },
    {
      "epoch": 13.41194968553459,
      "grad_norm": 0.2722390592098236,
      "learning_rate": 1.4786060748796076e-05,
      "loss": 0.5356,
      "num_input_tokens_seen": 16744424,
      "step": 25590
    },
    {
      "epoch": 13.414570230607966,
      "grad_norm": 0.46718481183052063,
      "learning_rate": 1.4775625373410617e-05,
      "loss": 0.5801,
      "num_input_tokens_seen": 16747880,
      "step": 25595
    },
    {
      "epoch": 13.417190775681341,
      "grad_norm": 1.0878428220748901,
      "learning_rate": 1.4765192136847685e-05,
      "loss": 0.6032,
      "num_input_tokens_seen": 16750984,
      "step": 25600
    },
    {
      "epoch": 13.419811320754716,
      "grad_norm": 0.6356806755065918,
      "learning_rate": 1.47547610412898e-05,
      "loss": 0.4246,
      "num_input_tokens_seen": 16753672,
      "step": 25605
    },
    {
      "epoch": 13.422431865828091,
      "grad_norm": 0.3647036850452423,
      "learning_rate": 1.4744332088919011e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 16756968,
      "step": 25610
    },
    {
      "epoch": 13.425052410901468,
      "grad_norm": 0.47632139921188354,
      "learning_rate": 1.4733905281916954e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 16760680,
      "step": 25615
    },
    {
      "epoch": 13.427672955974844,
      "grad_norm": 0.7457202672958374,
      "learning_rate": 1.472348062246478e-05,
      "loss": 0.4894,
      "num_input_tokens_seen": 16763624,
      "step": 25620
    },
    {
      "epoch": 13.430293501048219,
      "grad_norm": 0.29949116706848145,
      "learning_rate": 1.4713058112743228e-05,
      "loss": 0.4858,
      "num_input_tokens_seen": 16766728,
      "step": 25625
    },
    {
      "epoch": 13.432914046121594,
      "grad_norm": 0.3684508502483368,
      "learning_rate": 1.4702637754932553e-05,
      "loss": 0.5053,
      "num_input_tokens_seen": 16769768,
      "step": 25630
    },
    {
      "epoch": 13.435534591194969,
      "grad_norm": 0.5894736051559448,
      "learning_rate": 1.4692219551212587e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 16773384,
      "step": 25635
    },
    {
      "epoch": 13.438155136268344,
      "grad_norm": 0.37713614106178284,
      "learning_rate": 1.4681803503762695e-05,
      "loss": 0.4228,
      "num_input_tokens_seen": 16776520,
      "step": 25640
    },
    {
      "epoch": 13.44077568134172,
      "grad_norm": 0.7911726236343384,
      "learning_rate": 1.4671389614761788e-05,
      "loss": 0.3876,
      "num_input_tokens_seen": 16779336,
      "step": 25645
    },
    {
      "epoch": 13.443396226415095,
      "grad_norm": 0.5307169556617737,
      "learning_rate": 1.4660977886388339e-05,
      "loss": 0.5256,
      "num_input_tokens_seen": 16781576,
      "step": 25650
    },
    {
      "epoch": 13.44601677148847,
      "grad_norm": 0.42072027921676636,
      "learning_rate": 1.4650568320820362e-05,
      "loss": 0.5035,
      "num_input_tokens_seen": 16785064,
      "step": 25655
    },
    {
      "epoch": 13.448637316561845,
      "grad_norm": 0.6267650127410889,
      "learning_rate": 1.4640160920235423e-05,
      "loss": 0.4531,
      "num_input_tokens_seen": 16789128,
      "step": 25660
    },
    {
      "epoch": 13.45125786163522,
      "grad_norm": 0.5433320999145508,
      "learning_rate": 1.4629755686810631e-05,
      "loss": 0.3127,
      "num_input_tokens_seen": 16793064,
      "step": 25665
    },
    {
      "epoch": 13.453878406708595,
      "grad_norm": 0.4076659083366394,
      "learning_rate": 1.4619352622722632e-05,
      "loss": 0.5633,
      "num_input_tokens_seen": 16796200,
      "step": 25670
    },
    {
      "epoch": 13.45649895178197,
      "grad_norm": 0.4522848427295685,
      "learning_rate": 1.4608951730147635e-05,
      "loss": 0.4173,
      "num_input_tokens_seen": 16799016,
      "step": 25675
    },
    {
      "epoch": 13.459119496855346,
      "grad_norm": 0.4526651203632355,
      "learning_rate": 1.4598553011261395e-05,
      "loss": 0.5323,
      "num_input_tokens_seen": 16805608,
      "step": 25680
    },
    {
      "epoch": 13.46174004192872,
      "grad_norm": 0.3576548397541046,
      "learning_rate": 1.4588156468239192e-05,
      "loss": 0.4178,
      "num_input_tokens_seen": 16809448,
      "step": 25685
    },
    {
      "epoch": 13.464360587002096,
      "grad_norm": 0.33313634991645813,
      "learning_rate": 1.4577762103255882e-05,
      "loss": 0.445,
      "num_input_tokens_seen": 16813160,
      "step": 25690
    },
    {
      "epoch": 13.466981132075471,
      "grad_norm": 0.5622760653495789,
      "learning_rate": 1.4567369918485828e-05,
      "loss": 0.5285,
      "num_input_tokens_seen": 16817384,
      "step": 25695
    },
    {
      "epoch": 13.469601677148846,
      "grad_norm": 0.37684720754623413,
      "learning_rate": 1.4556979916102976e-05,
      "loss": 0.4118,
      "num_input_tokens_seen": 16820136,
      "step": 25700
    },
    {
      "epoch": 13.472222222222221,
      "grad_norm": 0.5320557951927185,
      "learning_rate": 1.454659209828077e-05,
      "loss": 0.5337,
      "num_input_tokens_seen": 16823240,
      "step": 25705
    },
    {
      "epoch": 13.474842767295598,
      "grad_norm": 0.579196035861969,
      "learning_rate": 1.4536206467192243e-05,
      "loss": 0.5157,
      "num_input_tokens_seen": 16825928,
      "step": 25710
    },
    {
      "epoch": 13.477463312368974,
      "grad_norm": 0.49148455262184143,
      "learning_rate": 1.4525823025009944e-05,
      "loss": 0.428,
      "num_input_tokens_seen": 16828808,
      "step": 25715
    },
    {
      "epoch": 13.480083857442349,
      "grad_norm": 0.7956348061561584,
      "learning_rate": 1.4515441773905983e-05,
      "loss": 0.3242,
      "num_input_tokens_seen": 16831336,
      "step": 25720
    },
    {
      "epoch": 13.482704402515724,
      "grad_norm": 0.4224705994129181,
      "learning_rate": 1.4505062716051988e-05,
      "loss": 0.4623,
      "num_input_tokens_seen": 16834184,
      "step": 25725
    },
    {
      "epoch": 13.485324947589099,
      "grad_norm": 0.4133986830711365,
      "learning_rate": 1.4494685853619133e-05,
      "loss": 0.4108,
      "num_input_tokens_seen": 16837576,
      "step": 25730
    },
    {
      "epoch": 13.487945492662474,
      "grad_norm": 0.5051171183586121,
      "learning_rate": 1.4484311188778143e-05,
      "loss": 0.5094,
      "num_input_tokens_seen": 16840616,
      "step": 25735
    },
    {
      "epoch": 13.49056603773585,
      "grad_norm": 0.5273492932319641,
      "learning_rate": 1.4473938723699285e-05,
      "loss": 0.4545,
      "num_input_tokens_seen": 16843592,
      "step": 25740
    },
    {
      "epoch": 13.493186582809225,
      "grad_norm": 0.470116525888443,
      "learning_rate": 1.4463568460552368e-05,
      "loss": 0.3623,
      "num_input_tokens_seen": 16846344,
      "step": 25745
    },
    {
      "epoch": 13.4958071278826,
      "grad_norm": 0.32328560948371887,
      "learning_rate": 1.4453200401506723e-05,
      "loss": 0.4927,
      "num_input_tokens_seen": 16849000,
      "step": 25750
    },
    {
      "epoch": 13.498427672955975,
      "grad_norm": 0.6217056512832642,
      "learning_rate": 1.4442834548731218e-05,
      "loss": 0.4495,
      "num_input_tokens_seen": 16852232,
      "step": 25755
    },
    {
      "epoch": 13.50104821802935,
      "grad_norm": 0.5534654855728149,
      "learning_rate": 1.4432470904394285e-05,
      "loss": 0.2822,
      "num_input_tokens_seen": 16855944,
      "step": 25760
    },
    {
      "epoch": 13.503668763102725,
      "grad_norm": 0.5213167667388916,
      "learning_rate": 1.4422109470663878e-05,
      "loss": 0.4115,
      "num_input_tokens_seen": 16859976,
      "step": 25765
    },
    {
      "epoch": 13.5062893081761,
      "grad_norm": 0.8915036916732788,
      "learning_rate": 1.441175024970749e-05,
      "loss": 0.3842,
      "num_input_tokens_seen": 16862440,
      "step": 25770
    },
    {
      "epoch": 13.508909853249476,
      "grad_norm": 0.3610125780105591,
      "learning_rate": 1.4401393243692163e-05,
      "loss": 0.396,
      "num_input_tokens_seen": 16865992,
      "step": 25775
    },
    {
      "epoch": 13.51153039832285,
      "grad_norm": 0.29482293128967285,
      "learning_rate": 1.4391038454784455e-05,
      "loss": 0.4248,
      "num_input_tokens_seen": 16869064,
      "step": 25780
    },
    {
      "epoch": 13.514150943396226,
      "grad_norm": 0.5200616717338562,
      "learning_rate": 1.4380685885150458e-05,
      "loss": 0.4447,
      "num_input_tokens_seen": 16872296,
      "step": 25785
    },
    {
      "epoch": 13.516771488469601,
      "grad_norm": 0.3441637456417084,
      "learning_rate": 1.437033553695582e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 16876136,
      "step": 25790
    },
    {
      "epoch": 13.519392033542976,
      "grad_norm": 1.0416487455368042,
      "learning_rate": 1.4359987412365721e-05,
      "loss": 0.4712,
      "num_input_tokens_seen": 16879592,
      "step": 25795
    },
    {
      "epoch": 13.522012578616351,
      "grad_norm": 0.43746718764305115,
      "learning_rate": 1.4349641513544876e-05,
      "loss": 0.5731,
      "num_input_tokens_seen": 16882248,
      "step": 25800
    },
    {
      "epoch": 13.524633123689728,
      "grad_norm": 1.320241093635559,
      "learning_rate": 1.4339297842657523e-05,
      "loss": 0.3823,
      "num_input_tokens_seen": 16884712,
      "step": 25805
    },
    {
      "epoch": 13.527253668763104,
      "grad_norm": 0.6839555501937866,
      "learning_rate": 1.4328956401867427e-05,
      "loss": 0.454,
      "num_input_tokens_seen": 16887528,
      "step": 25810
    },
    {
      "epoch": 13.529874213836479,
      "grad_norm": 0.3668331801891327,
      "learning_rate": 1.4318617193337908e-05,
      "loss": 0.4236,
      "num_input_tokens_seen": 16890856,
      "step": 25815
    },
    {
      "epoch": 13.532494758909854,
      "grad_norm": 0.4731201231479645,
      "learning_rate": 1.4308280219231812e-05,
      "loss": 0.3574,
      "num_input_tokens_seen": 16894600,
      "step": 25820
    },
    {
      "epoch": 13.535115303983229,
      "grad_norm": 0.4916442632675171,
      "learning_rate": 1.4297945481711527e-05,
      "loss": 0.4672,
      "num_input_tokens_seen": 16897416,
      "step": 25825
    },
    {
      "epoch": 13.537735849056604,
      "grad_norm": 0.48115238547325134,
      "learning_rate": 1.4287612982938946e-05,
      "loss": 0.3555,
      "num_input_tokens_seen": 16902664,
      "step": 25830
    },
    {
      "epoch": 13.54035639412998,
      "grad_norm": 0.32660728693008423,
      "learning_rate": 1.4277282725075523e-05,
      "loss": 0.3351,
      "num_input_tokens_seen": 16905608,
      "step": 25835
    },
    {
      "epoch": 13.542976939203355,
      "grad_norm": 0.5805600881576538,
      "learning_rate": 1.426695471028221e-05,
      "loss": 0.58,
      "num_input_tokens_seen": 16908552,
      "step": 25840
    },
    {
      "epoch": 13.54559748427673,
      "grad_norm": 1.0379410982131958,
      "learning_rate": 1.4256628940719524e-05,
      "loss": 0.3594,
      "num_input_tokens_seen": 16912296,
      "step": 25845
    },
    {
      "epoch": 13.548218029350105,
      "grad_norm": 0.5324119329452515,
      "learning_rate": 1.42463054185475e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 16915368,
      "step": 25850
    },
    {
      "epoch": 13.55083857442348,
      "grad_norm": 0.7787272930145264,
      "learning_rate": 1.4235984145925705e-05,
      "loss": 0.409,
      "num_input_tokens_seen": 16919080,
      "step": 25855
    },
    {
      "epoch": 13.553459119496855,
      "grad_norm": 0.7248687148094177,
      "learning_rate": 1.4225665125013224e-05,
      "loss": 0.4154,
      "num_input_tokens_seen": 16922024,
      "step": 25860
    },
    {
      "epoch": 13.55607966457023,
      "grad_norm": 0.8247166872024536,
      "learning_rate": 1.4215348357968669e-05,
      "loss": 0.3839,
      "num_input_tokens_seen": 16925416,
      "step": 25865
    },
    {
      "epoch": 13.558700209643606,
      "grad_norm": 0.49526309967041016,
      "learning_rate": 1.42050338469502e-05,
      "loss": 0.3515,
      "num_input_tokens_seen": 16928712,
      "step": 25870
    },
    {
      "epoch": 13.56132075471698,
      "grad_norm": 0.28354743123054504,
      "learning_rate": 1.4194721594115495e-05,
      "loss": 0.5686,
      "num_input_tokens_seen": 16931912,
      "step": 25875
    },
    {
      "epoch": 13.563941299790356,
      "grad_norm": 0.4768980145454407,
      "learning_rate": 1.4184411601621769e-05,
      "loss": 0.4775,
      "num_input_tokens_seen": 16935496,
      "step": 25880
    },
    {
      "epoch": 13.566561844863731,
      "grad_norm": 0.8558230996131897,
      "learning_rate": 1.4174103871625738e-05,
      "loss": 0.464,
      "num_input_tokens_seen": 16938216,
      "step": 25885
    },
    {
      "epoch": 13.569182389937106,
      "grad_norm": 0.3309226632118225,
      "learning_rate": 1.4163798406283674e-05,
      "loss": 0.4336,
      "num_input_tokens_seen": 16941736,
      "step": 25890
    },
    {
      "epoch": 13.571802935010481,
      "grad_norm": 0.5567119717597961,
      "learning_rate": 1.415349520775135e-05,
      "loss": 0.3935,
      "num_input_tokens_seen": 16944776,
      "step": 25895
    },
    {
      "epoch": 13.574423480083858,
      "grad_norm": 0.3089925944805145,
      "learning_rate": 1.4143194278184086e-05,
      "loss": 0.4763,
      "num_input_tokens_seen": 16948456,
      "step": 25900
    },
    {
      "epoch": 13.577044025157234,
      "grad_norm": 0.4698188006877899,
      "learning_rate": 1.4132895619736724e-05,
      "loss": 0.364,
      "num_input_tokens_seen": 16951208,
      "step": 25905
    },
    {
      "epoch": 13.579664570230609,
      "grad_norm": 0.3786347508430481,
      "learning_rate": 1.4122599234563613e-05,
      "loss": 0.4303,
      "num_input_tokens_seen": 16955208,
      "step": 25910
    },
    {
      "epoch": 13.582285115303984,
      "grad_norm": 0.7540761232376099,
      "learning_rate": 1.4112305124818647e-05,
      "loss": 0.4752,
      "num_input_tokens_seen": 16957992,
      "step": 25915
    },
    {
      "epoch": 13.584905660377359,
      "grad_norm": 0.2936040759086609,
      "learning_rate": 1.4102013292655242e-05,
      "loss": 0.3398,
      "num_input_tokens_seen": 16960872,
      "step": 25920
    },
    {
      "epoch": 13.587526205450734,
      "grad_norm": 0.6444059014320374,
      "learning_rate": 1.409172374022632e-05,
      "loss": 0.4371,
      "num_input_tokens_seen": 16963816,
      "step": 25925
    },
    {
      "epoch": 13.59014675052411,
      "grad_norm": 0.3497620224952698,
      "learning_rate": 1.4081436469684339e-05,
      "loss": 0.5002,
      "num_input_tokens_seen": 16966920,
      "step": 25930
    },
    {
      "epoch": 13.592767295597485,
      "grad_norm": 0.6072225570678711,
      "learning_rate": 1.4071151483181294e-05,
      "loss": 0.4256,
      "num_input_tokens_seen": 16969768,
      "step": 25935
    },
    {
      "epoch": 13.59538784067086,
      "grad_norm": 0.2440144121646881,
      "learning_rate": 1.4060868782868669e-05,
      "loss": 0.5818,
      "num_input_tokens_seen": 16975176,
      "step": 25940
    },
    {
      "epoch": 13.598008385744235,
      "grad_norm": 1.0669854879379272,
      "learning_rate": 1.405058837089749e-05,
      "loss": 0.4711,
      "num_input_tokens_seen": 16977960,
      "step": 25945
    },
    {
      "epoch": 13.60062893081761,
      "grad_norm": 0.9034468531608582,
      "learning_rate": 1.404031024941832e-05,
      "loss": 0.4832,
      "num_input_tokens_seen": 16981512,
      "step": 25950
    },
    {
      "epoch": 13.603249475890985,
      "grad_norm": 0.8060078620910645,
      "learning_rate": 1.40300344205812e-05,
      "loss": 0.4994,
      "num_input_tokens_seen": 16984104,
      "step": 25955
    },
    {
      "epoch": 13.60587002096436,
      "grad_norm": 0.46548891067504883,
      "learning_rate": 1.4019760886535738e-05,
      "loss": 0.3944,
      "num_input_tokens_seen": 16987464,
      "step": 25960
    },
    {
      "epoch": 13.608490566037736,
      "grad_norm": 0.5676167607307434,
      "learning_rate": 1.4009489649431018e-05,
      "loss": 0.4417,
      "num_input_tokens_seen": 16989768,
      "step": 25965
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 0.45034295320510864,
      "learning_rate": 1.3999220711415675e-05,
      "loss": 0.5498,
      "num_input_tokens_seen": 16992520,
      "step": 25970
    },
    {
      "epoch": 13.613731656184486,
      "grad_norm": 0.4273066222667694,
      "learning_rate": 1.3988954074637867e-05,
      "loss": 0.4449,
      "num_input_tokens_seen": 16995336,
      "step": 25975
    },
    {
      "epoch": 13.616352201257861,
      "grad_norm": 0.5862645506858826,
      "learning_rate": 1.3978689741245238e-05,
      "loss": 0.4075,
      "num_input_tokens_seen": 16998088,
      "step": 25980
    },
    {
      "epoch": 13.618972746331236,
      "grad_norm": 1.18955659866333,
      "learning_rate": 1.3968427713384986e-05,
      "loss": 0.5644,
      "num_input_tokens_seen": 17001128,
      "step": 25985
    },
    {
      "epoch": 13.621593291404611,
      "grad_norm": 0.3897000551223755,
      "learning_rate": 1.3958167993203791e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 17004808,
      "step": 25990
    },
    {
      "epoch": 13.624213836477988,
      "grad_norm": 0.44566020369529724,
      "learning_rate": 1.3947910582847879e-05,
      "loss": 0.465,
      "num_input_tokens_seen": 17008488,
      "step": 25995
    },
    {
      "epoch": 13.626834381551364,
      "grad_norm": 0.4722404479980469,
      "learning_rate": 1.3937655484462985e-05,
      "loss": 0.4555,
      "num_input_tokens_seen": 17011720,
      "step": 26000
    },
    {
      "epoch": 13.629454926624739,
      "grad_norm": 0.6645393371582031,
      "learning_rate": 1.392740270019437e-05,
      "loss": 0.3047,
      "num_input_tokens_seen": 17015976,
      "step": 26005
    },
    {
      "epoch": 13.632075471698114,
      "grad_norm": 0.5650595426559448,
      "learning_rate": 1.3917152232186776e-05,
      "loss": 0.3505,
      "num_input_tokens_seen": 17018568,
      "step": 26010
    },
    {
      "epoch": 13.634696016771489,
      "grad_norm": 0.7807236313819885,
      "learning_rate": 1.3906904082584504e-05,
      "loss": 0.4382,
      "num_input_tokens_seen": 17021896,
      "step": 26015
    },
    {
      "epoch": 13.637316561844864,
      "grad_norm": 0.49855872988700867,
      "learning_rate": 1.3896658253531334e-05,
      "loss": 0.6537,
      "num_input_tokens_seen": 17025256,
      "step": 26020
    },
    {
      "epoch": 13.63993710691824,
      "grad_norm": 0.5207127928733826,
      "learning_rate": 1.3886414747170581e-05,
      "loss": 0.5154,
      "num_input_tokens_seen": 17028136,
      "step": 26025
    },
    {
      "epoch": 13.642557651991615,
      "grad_norm": 0.48718273639678955,
      "learning_rate": 1.3876173565645072e-05,
      "loss": 0.434,
      "num_input_tokens_seen": 17030632,
      "step": 26030
    },
    {
      "epoch": 13.64517819706499,
      "grad_norm": 0.5847026705741882,
      "learning_rate": 1.3865934711097156e-05,
      "loss": 0.4801,
      "num_input_tokens_seen": 17033544,
      "step": 26035
    },
    {
      "epoch": 13.647798742138365,
      "grad_norm": 0.38354942202568054,
      "learning_rate": 1.3855698185668675e-05,
      "loss": 0.3788,
      "num_input_tokens_seen": 17036712,
      "step": 26040
    },
    {
      "epoch": 13.65041928721174,
      "grad_norm": 0.350432425737381,
      "learning_rate": 1.3845463991500985e-05,
      "loss": 0.4469,
      "num_input_tokens_seen": 17039240,
      "step": 26045
    },
    {
      "epoch": 13.653039832285115,
      "grad_norm": 0.7488237619400024,
      "learning_rate": 1.3835232130734968e-05,
      "loss": 0.4684,
      "num_input_tokens_seen": 17041864,
      "step": 26050
    },
    {
      "epoch": 13.65566037735849,
      "grad_norm": 0.5373026728630066,
      "learning_rate": 1.3825002605511017e-05,
      "loss": 0.3489,
      "num_input_tokens_seen": 17044424,
      "step": 26055
    },
    {
      "epoch": 13.658280922431866,
      "grad_norm": 0.5667520761489868,
      "learning_rate": 1.3814775417969037e-05,
      "loss": 0.5487,
      "num_input_tokens_seen": 17047368,
      "step": 26060
    },
    {
      "epoch": 13.66090146750524,
      "grad_norm": 0.34547924995422363,
      "learning_rate": 1.3804550570248431e-05,
      "loss": 0.3786,
      "num_input_tokens_seen": 17050696,
      "step": 26065
    },
    {
      "epoch": 13.663522012578616,
      "grad_norm": 0.5888931751251221,
      "learning_rate": 1.3794328064488116e-05,
      "loss": 0.4016,
      "num_input_tokens_seen": 17053832,
      "step": 26070
    },
    {
      "epoch": 13.666142557651991,
      "grad_norm": 0.5063986778259277,
      "learning_rate": 1.3784107902826523e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 17056712,
      "step": 26075
    },
    {
      "epoch": 13.668763102725366,
      "grad_norm": 0.33259767293930054,
      "learning_rate": 1.3773890087401603e-05,
      "loss": 0.3771,
      "num_input_tokens_seen": 17059464,
      "step": 26080
    },
    {
      "epoch": 13.671383647798741,
      "grad_norm": 0.6412482261657715,
      "learning_rate": 1.3763674620350802e-05,
      "loss": 0.4237,
      "num_input_tokens_seen": 17062216,
      "step": 26085
    },
    {
      "epoch": 13.674004192872118,
      "grad_norm": 0.6717536449432373,
      "learning_rate": 1.3753461503811089e-05,
      "loss": 0.3643,
      "num_input_tokens_seen": 17064456,
      "step": 26090
    },
    {
      "epoch": 13.676624737945493,
      "grad_norm": 0.7338694334030151,
      "learning_rate": 1.3743250739918922e-05,
      "loss": 0.4132,
      "num_input_tokens_seen": 17068296,
      "step": 26095
    },
    {
      "epoch": 13.679245283018869,
      "grad_norm": 0.5841419100761414,
      "learning_rate": 1.3733042330810265e-05,
      "loss": 0.574,
      "num_input_tokens_seen": 17071464,
      "step": 26100
    },
    {
      "epoch": 13.681865828092244,
      "grad_norm": 0.2991854250431061,
      "learning_rate": 1.3722836278620616e-05,
      "loss": 0.5057,
      "num_input_tokens_seen": 17075048,
      "step": 26105
    },
    {
      "epoch": 13.684486373165619,
      "grad_norm": 0.6630850434303284,
      "learning_rate": 1.3712632585484957e-05,
      "loss": 0.429,
      "num_input_tokens_seen": 17077768,
      "step": 26110
    },
    {
      "epoch": 13.687106918238994,
      "grad_norm": 0.9235053658485413,
      "learning_rate": 1.37024312535378e-05,
      "loss": 0.2827,
      "num_input_tokens_seen": 17082344,
      "step": 26115
    },
    {
      "epoch": 13.68972746331237,
      "grad_norm": 1.049217700958252,
      "learning_rate": 1.3692232284913133e-05,
      "loss": 0.4054,
      "num_input_tokens_seen": 17085352,
      "step": 26120
    },
    {
      "epoch": 13.692348008385745,
      "grad_norm": 0.5328983664512634,
      "learning_rate": 1.3682035681744459e-05,
      "loss": 0.3844,
      "num_input_tokens_seen": 17088360,
      "step": 26125
    },
    {
      "epoch": 13.69496855345912,
      "grad_norm": 0.5727859139442444,
      "learning_rate": 1.3671841446164795e-05,
      "loss": 0.4957,
      "num_input_tokens_seen": 17092008,
      "step": 26130
    },
    {
      "epoch": 13.697589098532495,
      "grad_norm": 0.305767685174942,
      "learning_rate": 1.3661649580306669e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 17095752,
      "step": 26135
    },
    {
      "epoch": 13.70020964360587,
      "grad_norm": 0.6534205675125122,
      "learning_rate": 1.3651460086302087e-05,
      "loss": 0.6348,
      "num_input_tokens_seen": 17098664,
      "step": 26140
    },
    {
      "epoch": 13.702830188679245,
      "grad_norm": 0.5343591570854187,
      "learning_rate": 1.3641272966282609e-05,
      "loss": 0.4704,
      "num_input_tokens_seen": 17101736,
      "step": 26145
    },
    {
      "epoch": 13.70545073375262,
      "grad_norm": 0.718477189540863,
      "learning_rate": 1.3631088222379218e-05,
      "loss": 0.4099,
      "num_input_tokens_seen": 17105544,
      "step": 26150
    },
    {
      "epoch": 13.708071278825996,
      "grad_norm": 0.2619595229625702,
      "learning_rate": 1.3620905856722468e-05,
      "loss": 0.3243,
      "num_input_tokens_seen": 17109704,
      "step": 26155
    },
    {
      "epoch": 13.71069182389937,
      "grad_norm": 0.5055640339851379,
      "learning_rate": 1.3610725871442392e-05,
      "loss": 0.5087,
      "num_input_tokens_seen": 17112040,
      "step": 26160
    },
    {
      "epoch": 13.713312368972746,
      "grad_norm": 0.28753942251205444,
      "learning_rate": 1.3600548268668532e-05,
      "loss": 0.427,
      "num_input_tokens_seen": 17116392,
      "step": 26165
    },
    {
      "epoch": 13.715932914046121,
      "grad_norm": 0.7462248206138611,
      "learning_rate": 1.3590373050529927e-05,
      "loss": 0.4056,
      "num_input_tokens_seen": 17120168,
      "step": 26170
    },
    {
      "epoch": 13.718553459119496,
      "grad_norm": 0.6151148080825806,
      "learning_rate": 1.3580200219155116e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 17124264,
      "step": 26175
    },
    {
      "epoch": 13.721174004192871,
      "grad_norm": 0.4404536187648773,
      "learning_rate": 1.3570029776672122e-05,
      "loss": 0.4482,
      "num_input_tokens_seen": 17126888,
      "step": 26180
    },
    {
      "epoch": 13.723794549266248,
      "grad_norm": 0.32582762837409973,
      "learning_rate": 1.35598617252085e-05,
      "loss": 0.4377,
      "num_input_tokens_seen": 17129224,
      "step": 26185
    },
    {
      "epoch": 13.726415094339622,
      "grad_norm": 0.49307435750961304,
      "learning_rate": 1.354969606689129e-05,
      "loss": 0.42,
      "num_input_tokens_seen": 17132168,
      "step": 26190
    },
    {
      "epoch": 13.729035639412999,
      "grad_norm": 0.44884446263313293,
      "learning_rate": 1.353953280384704e-05,
      "loss": 0.3347,
      "num_input_tokens_seen": 17134952,
      "step": 26195
    },
    {
      "epoch": 13.731656184486374,
      "grad_norm": 1.222486138343811,
      "learning_rate": 1.352937193820177e-05,
      "loss": 0.4402,
      "num_input_tokens_seen": 17137736,
      "step": 26200
    },
    {
      "epoch": 13.734276729559749,
      "grad_norm": 0.4670118987560272,
      "learning_rate": 1.3519213472081038e-05,
      "loss": 0.3191,
      "num_input_tokens_seen": 17140520,
      "step": 26205
    },
    {
      "epoch": 13.736897274633124,
      "grad_norm": 0.6326296925544739,
      "learning_rate": 1.3509057407609854e-05,
      "loss": 0.693,
      "num_input_tokens_seen": 17143304,
      "step": 26210
    },
    {
      "epoch": 13.7395178197065,
      "grad_norm": 0.38055500388145447,
      "learning_rate": 1.3498903746912769e-05,
      "loss": 0.4586,
      "num_input_tokens_seen": 17146440,
      "step": 26215
    },
    {
      "epoch": 13.742138364779874,
      "grad_norm": 0.5336074233055115,
      "learning_rate": 1.3488752492113807e-05,
      "loss": 0.5231,
      "num_input_tokens_seen": 17149512,
      "step": 26220
    },
    {
      "epoch": 13.74475890985325,
      "grad_norm": 0.8268663883209229,
      "learning_rate": 1.3478603645336508e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 17152712,
      "step": 26225
    },
    {
      "epoch": 13.747379454926625,
      "grad_norm": 0.6190255284309387,
      "learning_rate": 1.3468457208703877e-05,
      "loss": 0.3815,
      "num_input_tokens_seen": 17155528,
      "step": 26230
    },
    {
      "epoch": 13.75,
      "grad_norm": 0.26255035400390625,
      "learning_rate": 1.3458313184338437e-05,
      "loss": 0.3755,
      "num_input_tokens_seen": 17159240,
      "step": 26235
    },
    {
      "epoch": 13.752620545073375,
      "grad_norm": 0.5220652222633362,
      "learning_rate": 1.3448171574362201e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 17162280,
      "step": 26240
    },
    {
      "epoch": 13.75524109014675,
      "grad_norm": 0.39819425344467163,
      "learning_rate": 1.3438032380896681e-05,
      "loss": 0.4885,
      "num_input_tokens_seen": 17165608,
      "step": 26245
    },
    {
      "epoch": 13.757861635220126,
      "grad_norm": 0.8913461565971375,
      "learning_rate": 1.342789560606289e-05,
      "loss": 0.5639,
      "num_input_tokens_seen": 17168392,
      "step": 26250
    },
    {
      "epoch": 13.7604821802935,
      "grad_norm": 0.3689485490322113,
      "learning_rate": 1.341776125198131e-05,
      "loss": 0.3786,
      "num_input_tokens_seen": 17171432,
      "step": 26255
    },
    {
      "epoch": 13.763102725366876,
      "grad_norm": 0.5068019032478333,
      "learning_rate": 1.3407629320771948e-05,
      "loss": 0.5982,
      "num_input_tokens_seen": 17175496,
      "step": 26260
    },
    {
      "epoch": 13.765723270440251,
      "grad_norm": 0.5860087871551514,
      "learning_rate": 1.3397499814554266e-05,
      "loss": 0.5609,
      "num_input_tokens_seen": 17177960,
      "step": 26265
    },
    {
      "epoch": 13.768343815513626,
      "grad_norm": 0.43081027269363403,
      "learning_rate": 1.3387372735447256e-05,
      "loss": 0.4222,
      "num_input_tokens_seen": 17181160,
      "step": 26270
    },
    {
      "epoch": 13.770964360587001,
      "grad_norm": 0.4001021683216095,
      "learning_rate": 1.3377248085569399e-05,
      "loss": 0.5196,
      "num_input_tokens_seen": 17184136,
      "step": 26275
    },
    {
      "epoch": 13.773584905660378,
      "grad_norm": 0.4938690662384033,
      "learning_rate": 1.336712586703863e-05,
      "loss": 0.3522,
      "num_input_tokens_seen": 17187592,
      "step": 26280
    },
    {
      "epoch": 13.776205450733752,
      "grad_norm": 0.9384855628013611,
      "learning_rate": 1.3357006081972414e-05,
      "loss": 0.4096,
      "num_input_tokens_seen": 17189768,
      "step": 26285
    },
    {
      "epoch": 13.778825995807129,
      "grad_norm": 0.43754130601882935,
      "learning_rate": 1.3346888732487706e-05,
      "loss": 0.5238,
      "num_input_tokens_seen": 17192552,
      "step": 26290
    },
    {
      "epoch": 13.781446540880504,
      "grad_norm": 0.9057106375694275,
      "learning_rate": 1.3336773820700921e-05,
      "loss": 0.4516,
      "num_input_tokens_seen": 17195336,
      "step": 26295
    },
    {
      "epoch": 13.784067085953879,
      "grad_norm": 0.6859248280525208,
      "learning_rate": 1.3326661348727992e-05,
      "loss": 0.4424,
      "num_input_tokens_seen": 17198696,
      "step": 26300
    },
    {
      "epoch": 13.786687631027254,
      "grad_norm": 0.5027488470077515,
      "learning_rate": 1.3316551318684342e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 17201736,
      "step": 26305
    },
    {
      "epoch": 13.78930817610063,
      "grad_norm": 0.4564143419265747,
      "learning_rate": 1.3306443732684854e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 17204968,
      "step": 26310
    },
    {
      "epoch": 13.791928721174004,
      "grad_norm": 0.5579565167427063,
      "learning_rate": 1.3296338592843929e-05,
      "loss": 0.3799,
      "num_input_tokens_seen": 17208008,
      "step": 26315
    },
    {
      "epoch": 13.79454926624738,
      "grad_norm": 0.6015067100524902,
      "learning_rate": 1.3286235901275457e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 17210504,
      "step": 26320
    },
    {
      "epoch": 13.797169811320755,
      "grad_norm": 0.40783441066741943,
      "learning_rate": 1.327613566009279e-05,
      "loss": 0.5028,
      "num_input_tokens_seen": 17214088,
      "step": 26325
    },
    {
      "epoch": 13.79979035639413,
      "grad_norm": 0.36983418464660645,
      "learning_rate": 1.3266037871408798e-05,
      "loss": 0.3994,
      "num_input_tokens_seen": 17216680,
      "step": 26330
    },
    {
      "epoch": 13.802410901467505,
      "grad_norm": 0.34976309537887573,
      "learning_rate": 1.3255942537335805e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 17220456,
      "step": 26335
    },
    {
      "epoch": 13.80503144654088,
      "grad_norm": 0.6234809756278992,
      "learning_rate": 1.3245849659985652e-05,
      "loss": 0.4104,
      "num_input_tokens_seen": 17222728,
      "step": 26340
    },
    {
      "epoch": 13.807651991614255,
      "grad_norm": 0.4716395139694214,
      "learning_rate": 1.3235759241469658e-05,
      "loss": 0.3997,
      "num_input_tokens_seen": 17225672,
      "step": 26345
    },
    {
      "epoch": 13.81027253668763,
      "grad_norm": 0.345723420381546,
      "learning_rate": 1.322567128389861e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 17229736,
      "step": 26350
    },
    {
      "epoch": 13.812893081761006,
      "grad_norm": 0.40175095200538635,
      "learning_rate": 1.3215585789382812e-05,
      "loss": 0.5632,
      "num_input_tokens_seen": 17233576,
      "step": 26355
    },
    {
      "epoch": 13.815513626834381,
      "grad_norm": 0.6759024262428284,
      "learning_rate": 1.320550276003201e-05,
      "loss": 0.4402,
      "num_input_tokens_seen": 17236328,
      "step": 26360
    },
    {
      "epoch": 13.818134171907756,
      "grad_norm": 0.5382693409919739,
      "learning_rate": 1.3195422197955473e-05,
      "loss": 0.3339,
      "num_input_tokens_seen": 17239208,
      "step": 26365
    },
    {
      "epoch": 13.820754716981131,
      "grad_norm": 0.32263556122779846,
      "learning_rate": 1.3185344105261937e-05,
      "loss": 0.4857,
      "num_input_tokens_seen": 17243624,
      "step": 26370
    },
    {
      "epoch": 13.823375262054507,
      "grad_norm": 0.5429776310920715,
      "learning_rate": 1.3175268484059639e-05,
      "loss": 0.3799,
      "num_input_tokens_seen": 17247144,
      "step": 26375
    },
    {
      "epoch": 13.825995807127882,
      "grad_norm": 0.49476346373558044,
      "learning_rate": 1.316519533645626e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 17249672,
      "step": 26380
    },
    {
      "epoch": 13.828616352201259,
      "grad_norm": 0.3928261399269104,
      "learning_rate": 1.3155124664559004e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 17254056,
      "step": 26385
    },
    {
      "epoch": 13.831236897274634,
      "grad_norm": 0.8607785701751709,
      "learning_rate": 1.3145056470474525e-05,
      "loss": 0.3201,
      "num_input_tokens_seen": 17257672,
      "step": 26390
    },
    {
      "epoch": 13.833857442348009,
      "grad_norm": 0.40430667996406555,
      "learning_rate": 1.313499075630899e-05,
      "loss": 0.3255,
      "num_input_tokens_seen": 17260456,
      "step": 26395
    },
    {
      "epoch": 13.836477987421384,
      "grad_norm": 0.8843283653259277,
      "learning_rate": 1.312492752416802e-05,
      "loss": 0.3566,
      "num_input_tokens_seen": 17262920,
      "step": 26400
    },
    {
      "epoch": 13.83909853249476,
      "grad_norm": 0.6287440061569214,
      "learning_rate": 1.3114866776156748e-05,
      "loss": 0.4383,
      "num_input_tokens_seen": 17265320,
      "step": 26405
    },
    {
      "epoch": 13.841719077568134,
      "grad_norm": 0.5340094566345215,
      "learning_rate": 1.3104808514379748e-05,
      "loss": 0.4587,
      "num_input_tokens_seen": 17268104,
      "step": 26410
    },
    {
      "epoch": 13.84433962264151,
      "grad_norm": 0.3646594285964966,
      "learning_rate": 1.3094752740941094e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 17273128,
      "step": 26415
    },
    {
      "epoch": 13.846960167714885,
      "grad_norm": 0.544403076171875,
      "learning_rate": 1.3084699457944344e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 17275624,
      "step": 26420
    },
    {
      "epoch": 13.84958071278826,
      "grad_norm": 0.38301053643226624,
      "learning_rate": 1.3074648667492528e-05,
      "loss": 0.4893,
      "num_input_tokens_seen": 17278696,
      "step": 26425
    },
    {
      "epoch": 13.852201257861635,
      "grad_norm": 0.31169593334198,
      "learning_rate": 1.3064600371688171e-05,
      "loss": 0.4491,
      "num_input_tokens_seen": 17283112,
      "step": 26430
    },
    {
      "epoch": 13.85482180293501,
      "grad_norm": 0.3995395004749298,
      "learning_rate": 1.3054554572633249e-05,
      "loss": 0.4195,
      "num_input_tokens_seen": 17287272,
      "step": 26435
    },
    {
      "epoch": 13.857442348008385,
      "grad_norm": 1.7043625116348267,
      "learning_rate": 1.3044511272429222e-05,
      "loss": 0.3696,
      "num_input_tokens_seen": 17289608,
      "step": 26440
    },
    {
      "epoch": 13.86006289308176,
      "grad_norm": 0.4737403690814972,
      "learning_rate": 1.3034470473177038e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 17292936,
      "step": 26445
    },
    {
      "epoch": 13.862683438155136,
      "grad_norm": 0.4860375225543976,
      "learning_rate": 1.3024432176977125e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 17296392,
      "step": 26450
    },
    {
      "epoch": 13.865303983228511,
      "grad_norm": 0.3998044729232788,
      "learning_rate": 1.3014396385929372e-05,
      "loss": 0.3373,
      "num_input_tokens_seen": 17300104,
      "step": 26455
    },
    {
      "epoch": 13.867924528301886,
      "grad_norm": 0.4735070765018463,
      "learning_rate": 1.3004363102133165e-05,
      "loss": 0.388,
      "num_input_tokens_seen": 17303080,
      "step": 26460
    },
    {
      "epoch": 13.870545073375261,
      "grad_norm": 0.4175903797149658,
      "learning_rate": 1.2994332327687348e-05,
      "loss": 0.3464,
      "num_input_tokens_seen": 17305288,
      "step": 26465
    },
    {
      "epoch": 13.873165618448636,
      "grad_norm": 0.6303237080574036,
      "learning_rate": 1.2984304064690228e-05,
      "loss": 0.4223,
      "num_input_tokens_seen": 17308200,
      "step": 26470
    },
    {
      "epoch": 13.875786163522012,
      "grad_norm": 0.24524417519569397,
      "learning_rate": 1.2974278315239619e-05,
      "loss": 0.6682,
      "num_input_tokens_seen": 17310888,
      "step": 26475
    },
    {
      "epoch": 13.878406708595389,
      "grad_norm": 0.8584855198860168,
      "learning_rate": 1.2964255081432786e-05,
      "loss": 0.4563,
      "num_input_tokens_seen": 17313384,
      "step": 26480
    },
    {
      "epoch": 13.881027253668764,
      "grad_norm": 0.5420374870300293,
      "learning_rate": 1.2954234365366488e-05,
      "loss": 0.3716,
      "num_input_tokens_seen": 17316104,
      "step": 26485
    },
    {
      "epoch": 13.883647798742139,
      "grad_norm": 0.8068959712982178,
      "learning_rate": 1.2944216169136936e-05,
      "loss": 0.4509,
      "num_input_tokens_seen": 17319784,
      "step": 26490
    },
    {
      "epoch": 13.886268343815514,
      "grad_norm": 0.5540191531181335,
      "learning_rate": 1.2934200494839815e-05,
      "loss": 0.4698,
      "num_input_tokens_seen": 17323304,
      "step": 26495
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 0.8718456029891968,
      "learning_rate": 1.2924187344570293e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 17326408,
      "step": 26500
    },
    {
      "epoch": 13.891509433962264,
      "grad_norm": 0.8082026839256287,
      "learning_rate": 1.2914176720423011e-05,
      "loss": 0.5216,
      "num_input_tokens_seen": 17329800,
      "step": 26505
    },
    {
      "epoch": 13.89412997903564,
      "grad_norm": 0.33485230803489685,
      "learning_rate": 1.2904168624492088e-05,
      "loss": 0.4289,
      "num_input_tokens_seen": 17332648,
      "step": 26510
    },
    {
      "epoch": 13.896750524109015,
      "grad_norm": 0.27087998390197754,
      "learning_rate": 1.2894163058871092e-05,
      "loss": 0.3909,
      "num_input_tokens_seen": 17336648,
      "step": 26515
    },
    {
      "epoch": 13.89937106918239,
      "grad_norm": 1.0478798151016235,
      "learning_rate": 1.2884160025653064e-05,
      "loss": 0.4789,
      "num_input_tokens_seen": 17339720,
      "step": 26520
    },
    {
      "epoch": 13.901991614255765,
      "grad_norm": 0.5409146547317505,
      "learning_rate": 1.2874159526930532e-05,
      "loss": 0.4677,
      "num_input_tokens_seen": 17343080,
      "step": 26525
    },
    {
      "epoch": 13.90461215932914,
      "grad_norm": 0.31493279337882996,
      "learning_rate": 1.286416156479549e-05,
      "loss": 0.4095,
      "num_input_tokens_seen": 17346248,
      "step": 26530
    },
    {
      "epoch": 13.907232704402515,
      "grad_norm": 0.43225589394569397,
      "learning_rate": 1.2854166141339394e-05,
      "loss": 0.4009,
      "num_input_tokens_seen": 17349992,
      "step": 26535
    },
    {
      "epoch": 13.90985324947589,
      "grad_norm": 0.4958292543888092,
      "learning_rate": 1.2844173258653183e-05,
      "loss": 0.5345,
      "num_input_tokens_seen": 17352904,
      "step": 26540
    },
    {
      "epoch": 13.912473794549266,
      "grad_norm": 0.47709575295448303,
      "learning_rate": 1.2834182918827247e-05,
      "loss": 0.4284,
      "num_input_tokens_seen": 17356392,
      "step": 26545
    },
    {
      "epoch": 13.915094339622641,
      "grad_norm": 0.6937934160232544,
      "learning_rate": 1.282419512395144e-05,
      "loss": 0.396,
      "num_input_tokens_seen": 17359048,
      "step": 26550
    },
    {
      "epoch": 13.917714884696016,
      "grad_norm": 0.5706237554550171,
      "learning_rate": 1.2814209876115102e-05,
      "loss": 0.3538,
      "num_input_tokens_seen": 17362664,
      "step": 26555
    },
    {
      "epoch": 13.920335429769391,
      "grad_norm": 0.4090804159641266,
      "learning_rate": 1.2804227177407038e-05,
      "loss": 0.6709,
      "num_input_tokens_seen": 17365352,
      "step": 26560
    },
    {
      "epoch": 13.922955974842766,
      "grad_norm": 0.29227033257484436,
      "learning_rate": 1.279424702991552e-05,
      "loss": 0.3107,
      "num_input_tokens_seen": 17368392,
      "step": 26565
    },
    {
      "epoch": 13.925576519916142,
      "grad_norm": 0.3799603581428528,
      "learning_rate": 1.2784269435728264e-05,
      "loss": 0.4029,
      "num_input_tokens_seen": 17371304,
      "step": 26570
    },
    {
      "epoch": 13.928197064989519,
      "grad_norm": 0.9831981062889099,
      "learning_rate": 1.2774294396932488e-05,
      "loss": 0.3624,
      "num_input_tokens_seen": 17374600,
      "step": 26575
    },
    {
      "epoch": 13.930817610062894,
      "grad_norm": 0.5049298405647278,
      "learning_rate": 1.2764321915614835e-05,
      "loss": 0.509,
      "num_input_tokens_seen": 17377448,
      "step": 26580
    },
    {
      "epoch": 13.933438155136269,
      "grad_norm": 0.4012776017189026,
      "learning_rate": 1.2754351993861444e-05,
      "loss": 0.4085,
      "num_input_tokens_seen": 17381704,
      "step": 26585
    },
    {
      "epoch": 13.936058700209644,
      "grad_norm": 0.5406469702720642,
      "learning_rate": 1.2744384633757922e-05,
      "loss": 0.4472,
      "num_input_tokens_seen": 17384424,
      "step": 26590
    },
    {
      "epoch": 13.93867924528302,
      "grad_norm": 0.6634474396705627,
      "learning_rate": 1.2734419837389306e-05,
      "loss": 0.426,
      "num_input_tokens_seen": 17387976,
      "step": 26595
    },
    {
      "epoch": 13.941299790356394,
      "grad_norm": 0.41536352038383484,
      "learning_rate": 1.2724457606840123e-05,
      "loss": 0.347,
      "num_input_tokens_seen": 17394472,
      "step": 26600
    },
    {
      "epoch": 13.94392033542977,
      "grad_norm": 0.5451582670211792,
      "learning_rate": 1.2714497944194376e-05,
      "loss": 0.5423,
      "num_input_tokens_seen": 17400712,
      "step": 26605
    },
    {
      "epoch": 13.946540880503145,
      "grad_norm": 0.8290956616401672,
      "learning_rate": 1.2704540851535484e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 17403496,
      "step": 26610
    },
    {
      "epoch": 13.94916142557652,
      "grad_norm": 0.5527154207229614,
      "learning_rate": 1.2694586330946376e-05,
      "loss": 0.4942,
      "num_input_tokens_seen": 17406632,
      "step": 26615
    },
    {
      "epoch": 13.951781970649895,
      "grad_norm": 0.34060436487197876,
      "learning_rate": 1.2684634384509427e-05,
      "loss": 0.4263,
      "num_input_tokens_seen": 17409832,
      "step": 26620
    },
    {
      "epoch": 13.95440251572327,
      "grad_norm": 0.6236517429351807,
      "learning_rate": 1.2674685014306456e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 17412776,
      "step": 26625
    },
    {
      "epoch": 13.957023060796645,
      "grad_norm": 0.4094306528568268,
      "learning_rate": 1.2664738222418776e-05,
      "loss": 0.4298,
      "num_input_tokens_seen": 17416392,
      "step": 26630
    },
    {
      "epoch": 13.95964360587002,
      "grad_norm": 0.7594686150550842,
      "learning_rate": 1.2654794010927118e-05,
      "loss": 0.444,
      "num_input_tokens_seen": 17419080,
      "step": 26635
    },
    {
      "epoch": 13.962264150943396,
      "grad_norm": 0.5036583542823792,
      "learning_rate": 1.2644852381911715e-05,
      "loss": 0.4757,
      "num_input_tokens_seen": 17423048,
      "step": 26640
    },
    {
      "epoch": 13.964884696016771,
      "grad_norm": 0.7998973727226257,
      "learning_rate": 1.2634913337452248e-05,
      "loss": 0.3861,
      "num_input_tokens_seen": 17426824,
      "step": 26645
    },
    {
      "epoch": 13.967505241090146,
      "grad_norm": 0.6654404401779175,
      "learning_rate": 1.2624976879627832e-05,
      "loss": 0.357,
      "num_input_tokens_seen": 17429384,
      "step": 26650
    },
    {
      "epoch": 13.970125786163521,
      "grad_norm": 0.7343633770942688,
      "learning_rate": 1.2615043010517069e-05,
      "loss": 0.3117,
      "num_input_tokens_seen": 17432808,
      "step": 26655
    },
    {
      "epoch": 13.972746331236896,
      "grad_norm": 0.5725370049476624,
      "learning_rate": 1.2605111732198027e-05,
      "loss": 0.5345,
      "num_input_tokens_seen": 17435816,
      "step": 26660
    },
    {
      "epoch": 13.975366876310272,
      "grad_norm": 0.49805116653442383,
      "learning_rate": 1.2595183046748188e-05,
      "loss": 0.585,
      "num_input_tokens_seen": 17439624,
      "step": 26665
    },
    {
      "epoch": 13.977987421383649,
      "grad_norm": 0.4084390103816986,
      "learning_rate": 1.2585256956244545e-05,
      "loss": 0.4296,
      "num_input_tokens_seen": 17442664,
      "step": 26670
    },
    {
      "epoch": 13.980607966457024,
      "grad_norm": 0.3233877122402191,
      "learning_rate": 1.2575333462763506e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 17447816,
      "step": 26675
    },
    {
      "epoch": 13.983228511530399,
      "grad_norm": 0.7010835409164429,
      "learning_rate": 1.2565412568380957e-05,
      "loss": 0.5223,
      "num_input_tokens_seen": 17451912,
      "step": 26680
    },
    {
      "epoch": 13.985849056603774,
      "grad_norm": 0.508074939250946,
      "learning_rate": 1.2555494275172237e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 17454856,
      "step": 26685
    },
    {
      "epoch": 13.98846960167715,
      "grad_norm": 0.43967074155807495,
      "learning_rate": 1.2545578585212148e-05,
      "loss": 0.3797,
      "num_input_tokens_seen": 17459944,
      "step": 26690
    },
    {
      "epoch": 13.991090146750524,
      "grad_norm": 0.5720836520195007,
      "learning_rate": 1.2535665500574922e-05,
      "loss": 0.5812,
      "num_input_tokens_seen": 17463176,
      "step": 26695
    },
    {
      "epoch": 13.9937106918239,
      "grad_norm": 0.4348117709159851,
      "learning_rate": 1.2525755023334285e-05,
      "loss": 0.3956,
      "num_input_tokens_seen": 17467624,
      "step": 26700
    },
    {
      "epoch": 13.996331236897275,
      "grad_norm": 0.6211995482444763,
      "learning_rate": 1.251584715556337e-05,
      "loss": 0.4428,
      "num_input_tokens_seen": 17469992,
      "step": 26705
    },
    {
      "epoch": 13.99895178197065,
      "grad_norm": 0.659787654876709,
      "learning_rate": 1.2505941899334805e-05,
      "loss": 0.5125,
      "num_input_tokens_seen": 17474088,
      "step": 26710
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.4770318269729614,
      "eval_runtime": 9.6975,
      "eval_samples_per_second": 87.445,
      "eval_steps_per_second": 21.861,
      "num_input_tokens_seen": 17474664,
      "step": 26712
    },
    {
      "epoch": 14.001572327044025,
      "grad_norm": 0.5369515419006348,
      "learning_rate": 1.2496039256720665e-05,
      "loss": 0.421,
      "num_input_tokens_seen": 17476712,
      "step": 26715
    },
    {
      "epoch": 14.0041928721174,
      "grad_norm": 0.42388710379600525,
      "learning_rate": 1.2486139229792452e-05,
      "loss": 0.509,
      "num_input_tokens_seen": 17479528,
      "step": 26720
    },
    {
      "epoch": 14.006813417190775,
      "grad_norm": 0.3954678475856781,
      "learning_rate": 1.2476241820621152e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 17483560,
      "step": 26725
    },
    {
      "epoch": 14.00943396226415,
      "grad_norm": 0.7351123094558716,
      "learning_rate": 1.2466347031277176e-05,
      "loss": 0.5533,
      "num_input_tokens_seen": 17485832,
      "step": 26730
    },
    {
      "epoch": 14.012054507337526,
      "grad_norm": 0.4196907579898834,
      "learning_rate": 1.2456454863830411e-05,
      "loss": 0.3031,
      "num_input_tokens_seen": 17489192,
      "step": 26735
    },
    {
      "epoch": 14.014675052410901,
      "grad_norm": 0.7982561588287354,
      "learning_rate": 1.2446565320350185e-05,
      "loss": 0.6495,
      "num_input_tokens_seen": 17492072,
      "step": 26740
    },
    {
      "epoch": 14.017295597484276,
      "grad_norm": 0.49584367871284485,
      "learning_rate": 1.2436678402905283e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 17495656,
      "step": 26745
    },
    {
      "epoch": 14.019916142557651,
      "grad_norm": 0.5078381299972534,
      "learning_rate": 1.2426794113563933e-05,
      "loss": 0.3875,
      "num_input_tokens_seen": 17497960,
      "step": 26750
    },
    {
      "epoch": 14.022536687631026,
      "grad_norm": 0.4977566599845886,
      "learning_rate": 1.24169124543938e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 17501480,
      "step": 26755
    },
    {
      "epoch": 14.025157232704403,
      "grad_norm": 0.5263508558273315,
      "learning_rate": 1.2407033427462026e-05,
      "loss": 0.5589,
      "num_input_tokens_seen": 17504360,
      "step": 26760
    },
    {
      "epoch": 14.027777777777779,
      "grad_norm": 0.34987205266952515,
      "learning_rate": 1.2397157034835192e-05,
      "loss": 0.3429,
      "num_input_tokens_seen": 17507016,
      "step": 26765
    },
    {
      "epoch": 14.030398322851154,
      "grad_norm": 0.3592981994152069,
      "learning_rate": 1.2387283278579325e-05,
      "loss": 0.5366,
      "num_input_tokens_seen": 17511144,
      "step": 26770
    },
    {
      "epoch": 14.033018867924529,
      "grad_norm": 0.39509567618370056,
      "learning_rate": 1.237741216075991e-05,
      "loss": 0.4007,
      "num_input_tokens_seen": 17514440,
      "step": 26775
    },
    {
      "epoch": 14.035639412997904,
      "grad_norm": 0.6056687235832214,
      "learning_rate": 1.2367543683441868e-05,
      "loss": 0.5498,
      "num_input_tokens_seen": 17517768,
      "step": 26780
    },
    {
      "epoch": 14.03825995807128,
      "grad_norm": 0.4395424723625183,
      "learning_rate": 1.2357677848689558e-05,
      "loss": 0.5366,
      "num_input_tokens_seen": 17520968,
      "step": 26785
    },
    {
      "epoch": 14.040880503144654,
      "grad_norm": 0.357697457075119,
      "learning_rate": 1.2347814658566811e-05,
      "loss": 0.3333,
      "num_input_tokens_seen": 17524136,
      "step": 26790
    },
    {
      "epoch": 14.04350104821803,
      "grad_norm": 0.5626691579818726,
      "learning_rate": 1.2337954115136894e-05,
      "loss": 0.3474,
      "num_input_tokens_seen": 17528136,
      "step": 26795
    },
    {
      "epoch": 14.046121593291405,
      "grad_norm": 0.758618950843811,
      "learning_rate": 1.2328096220462529e-05,
      "loss": 0.4241,
      "num_input_tokens_seen": 17530856,
      "step": 26800
    },
    {
      "epoch": 14.04874213836478,
      "grad_norm": 0.38098740577697754,
      "learning_rate": 1.2318240976605865e-05,
      "loss": 0.405,
      "num_input_tokens_seen": 17534248,
      "step": 26805
    },
    {
      "epoch": 14.051362683438155,
      "grad_norm": 1.0388174057006836,
      "learning_rate": 1.23083883856285e-05,
      "loss": 0.3222,
      "num_input_tokens_seen": 17536648,
      "step": 26810
    },
    {
      "epoch": 14.05398322851153,
      "grad_norm": 0.4687235653400421,
      "learning_rate": 1.229853844959149e-05,
      "loss": 0.4358,
      "num_input_tokens_seen": 17539592,
      "step": 26815
    },
    {
      "epoch": 14.056603773584905,
      "grad_norm": 0.37501657009124756,
      "learning_rate": 1.228869117055533e-05,
      "loss": 0.555,
      "num_input_tokens_seen": 17543240,
      "step": 26820
    },
    {
      "epoch": 14.05922431865828,
      "grad_norm": 0.49770617485046387,
      "learning_rate": 1.227884655057997e-05,
      "loss": 0.4651,
      "num_input_tokens_seen": 17545768,
      "step": 26825
    },
    {
      "epoch": 14.061844863731656,
      "grad_norm": 0.38338062167167664,
      "learning_rate": 1.2269004591724777e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 17548968,
      "step": 26830
    },
    {
      "epoch": 14.064465408805031,
      "grad_norm": 0.3772045969963074,
      "learning_rate": 1.2259165296048575e-05,
      "loss": 0.5271,
      "num_input_tokens_seen": 17553480,
      "step": 26835
    },
    {
      "epoch": 14.067085953878406,
      "grad_norm": 0.36499086022377014,
      "learning_rate": 1.2249328665609638e-05,
      "loss": 0.6268,
      "num_input_tokens_seen": 17557960,
      "step": 26840
    },
    {
      "epoch": 14.069706498951781,
      "grad_norm": 0.509089469909668,
      "learning_rate": 1.2239494702465676e-05,
      "loss": 0.3435,
      "num_input_tokens_seen": 17561352,
      "step": 26845
    },
    {
      "epoch": 14.072327044025156,
      "grad_norm": 0.24740208685398102,
      "learning_rate": 1.2229663408673842e-05,
      "loss": 0.435,
      "num_input_tokens_seen": 17565736,
      "step": 26850
    },
    {
      "epoch": 14.074947589098532,
      "grad_norm": 0.5671676397323608,
      "learning_rate": 1.221983478629074e-05,
      "loss": 0.4748,
      "num_input_tokens_seen": 17568040,
      "step": 26855
    },
    {
      "epoch": 14.077568134171909,
      "grad_norm": 0.4260542690753937,
      "learning_rate": 1.22100088373724e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 17570888,
      "step": 26860
    },
    {
      "epoch": 14.080188679245284,
      "grad_norm": 0.40870165824890137,
      "learning_rate": 1.2200185563974283e-05,
      "loss": 0.5478,
      "num_input_tokens_seen": 17574248,
      "step": 26865
    },
    {
      "epoch": 14.082809224318659,
      "grad_norm": 0.7467341423034668,
      "learning_rate": 1.219036496815132e-05,
      "loss": 0.4112,
      "num_input_tokens_seen": 17577992,
      "step": 26870
    },
    {
      "epoch": 14.085429769392034,
      "grad_norm": 0.31359124183654785,
      "learning_rate": 1.2180547051957863e-05,
      "loss": 0.4437,
      "num_input_tokens_seen": 17581352,
      "step": 26875
    },
    {
      "epoch": 14.08805031446541,
      "grad_norm": 0.7249298095703125,
      "learning_rate": 1.2170731817447723e-05,
      "loss": 0.6097,
      "num_input_tokens_seen": 17584264,
      "step": 26880
    },
    {
      "epoch": 14.090670859538784,
      "grad_norm": 0.8067558407783508,
      "learning_rate": 1.216091926667412e-05,
      "loss": 0.5171,
      "num_input_tokens_seen": 17587496,
      "step": 26885
    },
    {
      "epoch": 14.09329140461216,
      "grad_norm": 0.49385035037994385,
      "learning_rate": 1.2151109401689723e-05,
      "loss": 0.4747,
      "num_input_tokens_seen": 17591080,
      "step": 26890
    },
    {
      "epoch": 14.095911949685535,
      "grad_norm": 0.5135359764099121,
      "learning_rate": 1.214130222454665e-05,
      "loss": 0.39,
      "num_input_tokens_seen": 17593992,
      "step": 26895
    },
    {
      "epoch": 14.09853249475891,
      "grad_norm": 0.5565561056137085,
      "learning_rate": 1.2131497737296452e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 17596936,
      "step": 26900
    },
    {
      "epoch": 14.101153039832285,
      "grad_norm": 1.0152077674865723,
      "learning_rate": 1.2121695941990124e-05,
      "loss": 0.3836,
      "num_input_tokens_seen": 17599336,
      "step": 26905
    },
    {
      "epoch": 14.10377358490566,
      "grad_norm": 0.42028215527534485,
      "learning_rate": 1.2111896840678075e-05,
      "loss": 0.5495,
      "num_input_tokens_seen": 17603336,
      "step": 26910
    },
    {
      "epoch": 14.106394129979035,
      "grad_norm": 0.5524863600730896,
      "learning_rate": 1.210210043541018e-05,
      "loss": 0.3431,
      "num_input_tokens_seen": 17606568,
      "step": 26915
    },
    {
      "epoch": 14.10901467505241,
      "grad_norm": 0.5395916104316711,
      "learning_rate": 1.2092306728235717e-05,
      "loss": 0.4335,
      "num_input_tokens_seen": 17609064,
      "step": 26920
    },
    {
      "epoch": 14.111635220125786,
      "grad_norm": 0.5237599015235901,
      "learning_rate": 1.2082515721203427e-05,
      "loss": 0.4244,
      "num_input_tokens_seen": 17611752,
      "step": 26925
    },
    {
      "epoch": 14.114255765199161,
      "grad_norm": 0.5779109001159668,
      "learning_rate": 1.2072727416361479e-05,
      "loss": 0.4138,
      "num_input_tokens_seen": 17614888,
      "step": 26930
    },
    {
      "epoch": 14.116876310272536,
      "grad_norm": 0.491593599319458,
      "learning_rate": 1.2062941815757484e-05,
      "loss": 0.386,
      "num_input_tokens_seen": 17617448,
      "step": 26935
    },
    {
      "epoch": 14.119496855345911,
      "grad_norm": 0.4234355688095093,
      "learning_rate": 1.2053158921438454e-05,
      "loss": 0.331,
      "num_input_tokens_seen": 17621480,
      "step": 26940
    },
    {
      "epoch": 14.122117400419286,
      "grad_norm": 0.7022008299827576,
      "learning_rate": 1.2043378735450885e-05,
      "loss": 0.3949,
      "num_input_tokens_seen": 17624488,
      "step": 26945
    },
    {
      "epoch": 14.124737945492662,
      "grad_norm": 0.350739061832428,
      "learning_rate": 1.2033601259840657e-05,
      "loss": 0.5187,
      "num_input_tokens_seen": 17628008,
      "step": 26950
    },
    {
      "epoch": 14.127358490566039,
      "grad_norm": 0.5732079744338989,
      "learning_rate": 1.2023826496653118e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 17630696,
      "step": 26955
    },
    {
      "epoch": 14.129979035639414,
      "grad_norm": 0.4232800006866455,
      "learning_rate": 1.2014054447933044e-05,
      "loss": 0.4904,
      "num_input_tokens_seen": 17634216,
      "step": 26960
    },
    {
      "epoch": 14.132599580712789,
      "grad_norm": 0.42757612466812134,
      "learning_rate": 1.200428511572462e-05,
      "loss": 0.5407,
      "num_input_tokens_seen": 17637832,
      "step": 26965
    },
    {
      "epoch": 14.135220125786164,
      "grad_norm": 0.3200725317001343,
      "learning_rate": 1.1994518502071484e-05,
      "loss": 0.4127,
      "num_input_tokens_seen": 17641640,
      "step": 26970
    },
    {
      "epoch": 14.13784067085954,
      "grad_norm": 0.43677443265914917,
      "learning_rate": 1.1984754609016716e-05,
      "loss": 0.4302,
      "num_input_tokens_seen": 17645256,
      "step": 26975
    },
    {
      "epoch": 14.140461215932914,
      "grad_norm": 0.40950074791908264,
      "learning_rate": 1.1974993438602785e-05,
      "loss": 0.4822,
      "num_input_tokens_seen": 17647944,
      "step": 26980
    },
    {
      "epoch": 14.14308176100629,
      "grad_norm": 0.38100236654281616,
      "learning_rate": 1.1965234992871641e-05,
      "loss": 0.5175,
      "num_input_tokens_seen": 17650568,
      "step": 26985
    },
    {
      "epoch": 14.145702306079665,
      "grad_norm": 0.43377089500427246,
      "learning_rate": 1.195547927386462e-05,
      "loss": 0.4069,
      "num_input_tokens_seen": 17653640,
      "step": 26990
    },
    {
      "epoch": 14.14832285115304,
      "grad_norm": 0.3925807774066925,
      "learning_rate": 1.1945726283622514e-05,
      "loss": 0.5202,
      "num_input_tokens_seen": 17656616,
      "step": 26995
    },
    {
      "epoch": 14.150943396226415,
      "grad_norm": 0.5622520446777344,
      "learning_rate": 1.1935976024185547e-05,
      "loss": 0.449,
      "num_input_tokens_seen": 17659976,
      "step": 27000
    },
    {
      "epoch": 14.15356394129979,
      "grad_norm": 0.2955296039581299,
      "learning_rate": 1.1926228497593341e-05,
      "loss": 0.5171,
      "num_input_tokens_seen": 17663496,
      "step": 27005
    },
    {
      "epoch": 14.156184486373165,
      "grad_norm": 0.3674708306789398,
      "learning_rate": 1.1916483705884985e-05,
      "loss": 0.324,
      "num_input_tokens_seen": 17666920,
      "step": 27010
    },
    {
      "epoch": 14.15880503144654,
      "grad_norm": 0.753079354763031,
      "learning_rate": 1.1906741651098983e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 17669096,
      "step": 27015
    },
    {
      "epoch": 14.161425576519916,
      "grad_norm": 0.507016122341156,
      "learning_rate": 1.1897002335273239e-05,
      "loss": 0.4368,
      "num_input_tokens_seen": 17672424,
      "step": 27020
    },
    {
      "epoch": 14.164046121593291,
      "grad_norm": 0.408216655254364,
      "learning_rate": 1.1887265760445121e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 17675112,
      "step": 27025
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 0.48461055755615234,
      "learning_rate": 1.1877531928651417e-05,
      "loss": 0.4292,
      "num_input_tokens_seen": 17678376,
      "step": 27030
    },
    {
      "epoch": 14.169287211740041,
      "grad_norm": 0.3610103726387024,
      "learning_rate": 1.1867800841928313e-05,
      "loss": 0.437,
      "num_input_tokens_seen": 17682280,
      "step": 27035
    },
    {
      "epoch": 14.171907756813416,
      "grad_norm": 0.455623060464859,
      "learning_rate": 1.1858072502311465e-05,
      "loss": 0.395,
      "num_input_tokens_seen": 17685000,
      "step": 27040
    },
    {
      "epoch": 14.174528301886792,
      "grad_norm": 0.27311086654663086,
      "learning_rate": 1.1848346911835908e-05,
      "loss": 0.4463,
      "num_input_tokens_seen": 17687944,
      "step": 27045
    },
    {
      "epoch": 14.177148846960169,
      "grad_norm": 0.4288448989391327,
      "learning_rate": 1.1838624072536137e-05,
      "loss": 0.4275,
      "num_input_tokens_seen": 17690792,
      "step": 27050
    },
    {
      "epoch": 14.179769392033544,
      "grad_norm": 0.3177943527698517,
      "learning_rate": 1.1828903986446055e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 17694984,
      "step": 27055
    },
    {
      "epoch": 14.182389937106919,
      "grad_norm": 0.8076359629631042,
      "learning_rate": 1.1819186655599007e-05,
      "loss": 0.5155,
      "num_input_tokens_seen": 17697352,
      "step": 27060
    },
    {
      "epoch": 14.185010482180294,
      "grad_norm": 0.9530956149101257,
      "learning_rate": 1.1809472082027734e-05,
      "loss": 0.5131,
      "num_input_tokens_seen": 17701928,
      "step": 27065
    },
    {
      "epoch": 14.18763102725367,
      "grad_norm": 0.5542440414428711,
      "learning_rate": 1.1799760267764407e-05,
      "loss": 0.4863,
      "num_input_tokens_seen": 17704968,
      "step": 27070
    },
    {
      "epoch": 14.190251572327044,
      "grad_norm": 0.512622594833374,
      "learning_rate": 1.1790051214840634e-05,
      "loss": 0.4551,
      "num_input_tokens_seen": 17708360,
      "step": 27075
    },
    {
      "epoch": 14.19287211740042,
      "grad_norm": 0.7097141742706299,
      "learning_rate": 1.1780344925287443e-05,
      "loss": 0.4872,
      "num_input_tokens_seen": 17711464,
      "step": 27080
    },
    {
      "epoch": 14.195492662473795,
      "grad_norm": 0.7066447734832764,
      "learning_rate": 1.1770641401135282e-05,
      "loss": 0.4928,
      "num_input_tokens_seen": 17715336,
      "step": 27085
    },
    {
      "epoch": 14.19811320754717,
      "grad_norm": 0.764487624168396,
      "learning_rate": 1.1760940644414003e-05,
      "loss": 0.5182,
      "num_input_tokens_seen": 17717512,
      "step": 27090
    },
    {
      "epoch": 14.200733752620545,
      "grad_norm": 1.106804609298706,
      "learning_rate": 1.1751242657152914e-05,
      "loss": 0.4396,
      "num_input_tokens_seen": 17720072,
      "step": 27095
    },
    {
      "epoch": 14.20335429769392,
      "grad_norm": 0.35922303795814514,
      "learning_rate": 1.1741547441380698e-05,
      "loss": 0.3901,
      "num_input_tokens_seen": 17723528,
      "step": 27100
    },
    {
      "epoch": 14.205974842767295,
      "grad_norm": 0.375413179397583,
      "learning_rate": 1.17318549991255e-05,
      "loss": 0.3558,
      "num_input_tokens_seen": 17725960,
      "step": 27105
    },
    {
      "epoch": 14.20859538784067,
      "grad_norm": 0.43997639417648315,
      "learning_rate": 1.1722165332414863e-05,
      "loss": 0.4704,
      "num_input_tokens_seen": 17729928,
      "step": 27110
    },
    {
      "epoch": 14.211215932914046,
      "grad_norm": 0.6676433086395264,
      "learning_rate": 1.1712478443275768e-05,
      "loss": 0.5095,
      "num_input_tokens_seen": 17732712,
      "step": 27115
    },
    {
      "epoch": 14.213836477987421,
      "grad_norm": 0.5651285648345947,
      "learning_rate": 1.170279433373459e-05,
      "loss": 0.4525,
      "num_input_tokens_seen": 17736072,
      "step": 27120
    },
    {
      "epoch": 14.216457023060796,
      "grad_norm": 0.5733287334442139,
      "learning_rate": 1.1693113005817125e-05,
      "loss": 0.4717,
      "num_input_tokens_seen": 17739080,
      "step": 27125
    },
    {
      "epoch": 14.219077568134171,
      "grad_norm": 0.41031414270401,
      "learning_rate": 1.1683434461548606e-05,
      "loss": 0.3395,
      "num_input_tokens_seen": 17742440,
      "step": 27130
    },
    {
      "epoch": 14.221698113207546,
      "grad_norm": 0.7238239049911499,
      "learning_rate": 1.1673758702953675e-05,
      "loss": 0.5675,
      "num_input_tokens_seen": 17745736,
      "step": 27135
    },
    {
      "epoch": 14.224318658280922,
      "grad_norm": 0.2792488932609558,
      "learning_rate": 1.1664085732056398e-05,
      "loss": 0.3622,
      "num_input_tokens_seen": 17748552,
      "step": 27140
    },
    {
      "epoch": 14.226939203354299,
      "grad_norm": 0.3805942237377167,
      "learning_rate": 1.1654415550880243e-05,
      "loss": 0.4952,
      "num_input_tokens_seen": 17751592,
      "step": 27145
    },
    {
      "epoch": 14.229559748427674,
      "grad_norm": 0.47359350323677063,
      "learning_rate": 1.1644748161448088e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 17755080,
      "step": 27150
    },
    {
      "epoch": 14.232180293501049,
      "grad_norm": 0.3275109827518463,
      "learning_rate": 1.1635083565782257e-05,
      "loss": 0.4057,
      "num_input_tokens_seen": 17758504,
      "step": 27155
    },
    {
      "epoch": 14.234800838574424,
      "grad_norm": 0.4818483293056488,
      "learning_rate": 1.1625421765904465e-05,
      "loss": 0.4383,
      "num_input_tokens_seen": 17762888,
      "step": 27160
    },
    {
      "epoch": 14.2374213836478,
      "grad_norm": 0.42828938364982605,
      "learning_rate": 1.1615762763835858e-05,
      "loss": 0.2729,
      "num_input_tokens_seen": 17769384,
      "step": 27165
    },
    {
      "epoch": 14.240041928721174,
      "grad_norm": 0.36993521451950073,
      "learning_rate": 1.1606106561596993e-05,
      "loss": 0.6321,
      "num_input_tokens_seen": 17772904,
      "step": 27170
    },
    {
      "epoch": 14.24266247379455,
      "grad_norm": 0.8152502775192261,
      "learning_rate": 1.1596453161207829e-05,
      "loss": 0.5783,
      "num_input_tokens_seen": 17775752,
      "step": 27175
    },
    {
      "epoch": 14.245283018867925,
      "grad_norm": 0.5688480138778687,
      "learning_rate": 1.1586802564687741e-05,
      "loss": 0.4718,
      "num_input_tokens_seen": 17778184,
      "step": 27180
    },
    {
      "epoch": 14.2479035639413,
      "grad_norm": 0.7410316467285156,
      "learning_rate": 1.1577154774055528e-05,
      "loss": 0.3998,
      "num_input_tokens_seen": 17781384,
      "step": 27185
    },
    {
      "epoch": 14.250524109014675,
      "grad_norm": 0.5700952410697937,
      "learning_rate": 1.1567509791329401e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 17784136,
      "step": 27190
    },
    {
      "epoch": 14.25314465408805,
      "grad_norm": 0.4130907952785492,
      "learning_rate": 1.155786761852699e-05,
      "loss": 0.4162,
      "num_input_tokens_seen": 17790376,
      "step": 27195
    },
    {
      "epoch": 14.255765199161425,
      "grad_norm": 0.5626381635665894,
      "learning_rate": 1.1548228257665317e-05,
      "loss": 0.4654,
      "num_input_tokens_seen": 17793928,
      "step": 27200
    },
    {
      "epoch": 14.2583857442348,
      "grad_norm": 0.5361654162406921,
      "learning_rate": 1.1538591710760815e-05,
      "loss": 0.5068,
      "num_input_tokens_seen": 17796712,
      "step": 27205
    },
    {
      "epoch": 14.261006289308176,
      "grad_norm": 0.5240490436553955,
      "learning_rate": 1.1528957979829355e-05,
      "loss": 0.4566,
      "num_input_tokens_seen": 17799432,
      "step": 27210
    },
    {
      "epoch": 14.26362683438155,
      "grad_norm": 0.3828558623790741,
      "learning_rate": 1.1519327066886198e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 17802888,
      "step": 27215
    },
    {
      "epoch": 14.266247379454926,
      "grad_norm": 0.5813097357749939,
      "learning_rate": 1.1509698973946032e-05,
      "loss": 0.4201,
      "num_input_tokens_seen": 17806408,
      "step": 27220
    },
    {
      "epoch": 14.268867924528301,
      "grad_norm": 0.34301096200942993,
      "learning_rate": 1.1500073703022926e-05,
      "loss": 0.3767,
      "num_input_tokens_seen": 17810632,
      "step": 27225
    },
    {
      "epoch": 14.271488469601676,
      "grad_norm": 0.4979609251022339,
      "learning_rate": 1.1490451256130394e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 17813480,
      "step": 27230
    },
    {
      "epoch": 14.274109014675052,
      "grad_norm": 0.5882885456085205,
      "learning_rate": 1.1480831635281328e-05,
      "loss": 0.4921,
      "num_input_tokens_seen": 17816968,
      "step": 27235
    },
    {
      "epoch": 14.276729559748428,
      "grad_norm": 0.48797476291656494,
      "learning_rate": 1.1471214842488046e-05,
      "loss": 0.6003,
      "num_input_tokens_seen": 17819944,
      "step": 27240
    },
    {
      "epoch": 14.279350104821804,
      "grad_norm": 0.47348347306251526,
      "learning_rate": 1.1461600879762277e-05,
      "loss": 0.4496,
      "num_input_tokens_seen": 17822984,
      "step": 27245
    },
    {
      "epoch": 14.281970649895179,
      "grad_norm": 0.5108006596565247,
      "learning_rate": 1.1451989749115157e-05,
      "loss": 0.4814,
      "num_input_tokens_seen": 17825832,
      "step": 27250
    },
    {
      "epoch": 14.284591194968554,
      "grad_norm": 0.4530637264251709,
      "learning_rate": 1.1442381452557219e-05,
      "loss": 0.4078,
      "num_input_tokens_seen": 17828936,
      "step": 27255
    },
    {
      "epoch": 14.28721174004193,
      "grad_norm": 0.40686699748039246,
      "learning_rate": 1.1432775992098397e-05,
      "loss": 0.3538,
      "num_input_tokens_seen": 17831848,
      "step": 27260
    },
    {
      "epoch": 14.289832285115304,
      "grad_norm": 0.4629060626029968,
      "learning_rate": 1.1423173369748056e-05,
      "loss": 0.3464,
      "num_input_tokens_seen": 17834856,
      "step": 27265
    },
    {
      "epoch": 14.29245283018868,
      "grad_norm": 0.49759790301322937,
      "learning_rate": 1.1413573587514952e-05,
      "loss": 0.5333,
      "num_input_tokens_seen": 17838856,
      "step": 27270
    },
    {
      "epoch": 14.295073375262055,
      "grad_norm": 0.6082776784896851,
      "learning_rate": 1.140397664740726e-05,
      "loss": 0.3416,
      "num_input_tokens_seen": 17841960,
      "step": 27275
    },
    {
      "epoch": 14.29769392033543,
      "grad_norm": 0.41462215781211853,
      "learning_rate": 1.1394382551432534e-05,
      "loss": 0.4271,
      "num_input_tokens_seen": 17844584,
      "step": 27280
    },
    {
      "epoch": 14.300314465408805,
      "grad_norm": 0.7119799256324768,
      "learning_rate": 1.1384791301597767e-05,
      "loss": 0.5369,
      "num_input_tokens_seen": 17848328,
      "step": 27285
    },
    {
      "epoch": 14.30293501048218,
      "grad_norm": 0.30175113677978516,
      "learning_rate": 1.1375202899909317e-05,
      "loss": 0.3257,
      "num_input_tokens_seen": 17852552,
      "step": 27290
    },
    {
      "epoch": 14.305555555555555,
      "grad_norm": 0.30516982078552246,
      "learning_rate": 1.1365617348372984e-05,
      "loss": 0.3664,
      "num_input_tokens_seen": 17855912,
      "step": 27295
    },
    {
      "epoch": 14.30817610062893,
      "grad_norm": 0.3384946286678314,
      "learning_rate": 1.135603464899396e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 17859240,
      "step": 27300
    },
    {
      "epoch": 14.310796645702306,
      "grad_norm": 0.7086400985717773,
      "learning_rate": 1.1346454803776821e-05,
      "loss": 0.4173,
      "num_input_tokens_seen": 17862568,
      "step": 27305
    },
    {
      "epoch": 14.31341719077568,
      "grad_norm": 0.302864134311676,
      "learning_rate": 1.1336877814725566e-05,
      "loss": 0.4418,
      "num_input_tokens_seen": 17866152,
      "step": 27310
    },
    {
      "epoch": 14.316037735849056,
      "grad_norm": 0.5093854665756226,
      "learning_rate": 1.1327303683843607e-05,
      "loss": 0.4648,
      "num_input_tokens_seen": 17869128,
      "step": 27315
    },
    {
      "epoch": 14.318658280922431,
      "grad_norm": 0.4452230930328369,
      "learning_rate": 1.1317732413133723e-05,
      "loss": 0.3382,
      "num_input_tokens_seen": 17872072,
      "step": 27320
    },
    {
      "epoch": 14.321278825995806,
      "grad_norm": 0.39352962374687195,
      "learning_rate": 1.1308164004598118e-05,
      "loss": 0.4049,
      "num_input_tokens_seen": 17875560,
      "step": 27325
    },
    {
      "epoch": 14.323899371069182,
      "grad_norm": 0.5166761875152588,
      "learning_rate": 1.129859846023841e-05,
      "loss": 0.4748,
      "num_input_tokens_seen": 17878632,
      "step": 27330
    },
    {
      "epoch": 14.326519916142558,
      "grad_norm": 0.5295729041099548,
      "learning_rate": 1.1289035782055582e-05,
      "loss": 0.5054,
      "num_input_tokens_seen": 17882248,
      "step": 27335
    },
    {
      "epoch": 14.329140461215934,
      "grad_norm": 0.4195975363254547,
      "learning_rate": 1.1279475972050044e-05,
      "loss": 0.3806,
      "num_input_tokens_seen": 17885736,
      "step": 27340
    },
    {
      "epoch": 14.331761006289309,
      "grad_norm": 1.1735649108886719,
      "learning_rate": 1.1269919032221612e-05,
      "loss": 0.3097,
      "num_input_tokens_seen": 17888424,
      "step": 27345
    },
    {
      "epoch": 14.334381551362684,
      "grad_norm": 0.5458644032478333,
      "learning_rate": 1.1260364964569466e-05,
      "loss": 0.3107,
      "num_input_tokens_seen": 17891144,
      "step": 27350
    },
    {
      "epoch": 14.33700209643606,
      "grad_norm": 0.4685528576374054,
      "learning_rate": 1.125081377109223e-05,
      "loss": 0.4397,
      "num_input_tokens_seen": 17894024,
      "step": 27355
    },
    {
      "epoch": 14.339622641509434,
      "grad_norm": 0.6624406576156616,
      "learning_rate": 1.1241265453787882e-05,
      "loss": 0.3824,
      "num_input_tokens_seen": 17896776,
      "step": 27360
    },
    {
      "epoch": 14.34224318658281,
      "grad_norm": 0.4612761437892914,
      "learning_rate": 1.1231720014653837e-05,
      "loss": 0.4354,
      "num_input_tokens_seen": 17899688,
      "step": 27365
    },
    {
      "epoch": 14.344863731656185,
      "grad_norm": 0.5182008743286133,
      "learning_rate": 1.1222177455686896e-05,
      "loss": 0.4097,
      "num_input_tokens_seen": 17903240,
      "step": 27370
    },
    {
      "epoch": 14.34748427672956,
      "grad_norm": 1.4336920976638794,
      "learning_rate": 1.1212637778883237e-05,
      "loss": 0.4138,
      "num_input_tokens_seen": 17906312,
      "step": 27375
    },
    {
      "epoch": 14.350104821802935,
      "grad_norm": 0.3964850604534149,
      "learning_rate": 1.1203100986238468e-05,
      "loss": 0.6095,
      "num_input_tokens_seen": 17909640,
      "step": 27380
    },
    {
      "epoch": 14.35272536687631,
      "grad_norm": 0.6281522512435913,
      "learning_rate": 1.1193567079747561e-05,
      "loss": 0.3881,
      "num_input_tokens_seen": 17912232,
      "step": 27385
    },
    {
      "epoch": 14.355345911949685,
      "grad_norm": 0.5738101005554199,
      "learning_rate": 1.1184036061404907e-05,
      "loss": 0.3773,
      "num_input_tokens_seen": 17915080,
      "step": 27390
    },
    {
      "epoch": 14.35796645702306,
      "grad_norm": 0.3447226285934448,
      "learning_rate": 1.117450793320429e-05,
      "loss": 0.3532,
      "num_input_tokens_seen": 17918600,
      "step": 27395
    },
    {
      "epoch": 14.360587002096436,
      "grad_norm": 0.5391371846199036,
      "learning_rate": 1.1164982697138893e-05,
      "loss": 0.4262,
      "num_input_tokens_seen": 17922536,
      "step": 27400
    },
    {
      "epoch": 14.36320754716981,
      "grad_norm": 0.45260530710220337,
      "learning_rate": 1.1155460355201267e-05,
      "loss": 0.3504,
      "num_input_tokens_seen": 17927208,
      "step": 27405
    },
    {
      "epoch": 14.365828092243186,
      "grad_norm": 0.5115600228309631,
      "learning_rate": 1.1145940909383399e-05,
      "loss": 0.3263,
      "num_input_tokens_seen": 17930312,
      "step": 27410
    },
    {
      "epoch": 14.368448637316561,
      "grad_norm": 1.1796507835388184,
      "learning_rate": 1.1136424361676626e-05,
      "loss": 0.6347,
      "num_input_tokens_seen": 17933256,
      "step": 27415
    },
    {
      "epoch": 14.371069182389936,
      "grad_norm": 0.5089554190635681,
      "learning_rate": 1.1126910714071711e-05,
      "loss": 0.4334,
      "num_input_tokens_seen": 17936264,
      "step": 27420
    },
    {
      "epoch": 14.373689727463312,
      "grad_norm": 0.4130765497684479,
      "learning_rate": 1.1117399968558806e-05,
      "loss": 0.609,
      "num_input_tokens_seen": 17939432,
      "step": 27425
    },
    {
      "epoch": 14.376310272536688,
      "grad_norm": 0.2961166501045227,
      "learning_rate": 1.1107892127127454e-05,
      "loss": 0.4185,
      "num_input_tokens_seen": 17943848,
      "step": 27430
    },
    {
      "epoch": 14.378930817610064,
      "grad_norm": 0.40945741534233093,
      "learning_rate": 1.109838719176658e-05,
      "loss": 0.4602,
      "num_input_tokens_seen": 17947368,
      "step": 27435
    },
    {
      "epoch": 14.381551362683439,
      "grad_norm": 0.43653377890586853,
      "learning_rate": 1.1088885164464497e-05,
      "loss": 0.3001,
      "num_input_tokens_seen": 17950344,
      "step": 27440
    },
    {
      "epoch": 14.384171907756814,
      "grad_norm": 0.6908408403396606,
      "learning_rate": 1.1079386047208932e-05,
      "loss": 0.2894,
      "num_input_tokens_seen": 17952648,
      "step": 27445
    },
    {
      "epoch": 14.38679245283019,
      "grad_norm": 0.44133105874061584,
      "learning_rate": 1.1069889841986991e-05,
      "loss": 0.4975,
      "num_input_tokens_seen": 17956392,
      "step": 27450
    },
    {
      "epoch": 14.389412997903564,
      "grad_norm": 0.460492342710495,
      "learning_rate": 1.1060396550785182e-05,
      "loss": 0.4196,
      "num_input_tokens_seen": 17959816,
      "step": 27455
    },
    {
      "epoch": 14.39203354297694,
      "grad_norm": 0.5170839428901672,
      "learning_rate": 1.1050906175589382e-05,
      "loss": 0.3569,
      "num_input_tokens_seen": 17963144,
      "step": 27460
    },
    {
      "epoch": 14.394654088050315,
      "grad_norm": 0.5263219475746155,
      "learning_rate": 1.1041418718384863e-05,
      "loss": 0.4641,
      "num_input_tokens_seen": 17965640,
      "step": 27465
    },
    {
      "epoch": 14.39727463312369,
      "grad_norm": 0.4261118471622467,
      "learning_rate": 1.1031934181156298e-05,
      "loss": 0.3535,
      "num_input_tokens_seen": 17968168,
      "step": 27470
    },
    {
      "epoch": 14.399895178197065,
      "grad_norm": 0.25866732001304626,
      "learning_rate": 1.1022452565887747e-05,
      "loss": 0.2422,
      "num_input_tokens_seen": 17971816,
      "step": 27475
    },
    {
      "epoch": 14.40251572327044,
      "grad_norm": 0.3332207500934601,
      "learning_rate": 1.1012973874562657e-05,
      "loss": 0.4234,
      "num_input_tokens_seen": 17974376,
      "step": 27480
    },
    {
      "epoch": 14.405136268343815,
      "grad_norm": 0.5679412484169006,
      "learning_rate": 1.1003498109163865e-05,
      "loss": 0.4113,
      "num_input_tokens_seen": 17977608,
      "step": 27485
    },
    {
      "epoch": 14.40775681341719,
      "grad_norm": 0.41965755820274353,
      "learning_rate": 1.0994025271673592e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 17980808,
      "step": 27490
    },
    {
      "epoch": 14.410377358490566,
      "grad_norm": 0.26045283675193787,
      "learning_rate": 1.0984555364073431e-05,
      "loss": 0.3966,
      "num_input_tokens_seen": 17984392,
      "step": 27495
    },
    {
      "epoch": 14.41299790356394,
      "grad_norm": 0.7406543493270874,
      "learning_rate": 1.0975088388344393e-05,
      "loss": 0.379,
      "num_input_tokens_seen": 17987432,
      "step": 27500
    },
    {
      "epoch": 14.415618448637316,
      "grad_norm": 0.5175172090530396,
      "learning_rate": 1.096562434646686e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 17991400,
      "step": 27505
    },
    {
      "epoch": 14.418238993710691,
      "grad_norm": 0.41581302881240845,
      "learning_rate": 1.0956163240420611e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 17994536,
      "step": 27510
    },
    {
      "epoch": 14.420859538784066,
      "grad_norm": 0.34565240144729614,
      "learning_rate": 1.094670507218479e-05,
      "loss": 0.3696,
      "num_input_tokens_seen": 17997352,
      "step": 27515
    },
    {
      "epoch": 14.423480083857442,
      "grad_norm": 0.378900945186615,
      "learning_rate": 1.093724984373793e-05,
      "loss": 0.3435,
      "num_input_tokens_seen": 18001320,
      "step": 27520
    },
    {
      "epoch": 14.426100628930818,
      "grad_norm": 0.5395210385322571,
      "learning_rate": 1.0927797557057968e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 18005032,
      "step": 27525
    },
    {
      "epoch": 14.428721174004194,
      "grad_norm": 0.744270384311676,
      "learning_rate": 1.0918348214122209e-05,
      "loss": 0.5051,
      "num_input_tokens_seen": 18008456,
      "step": 27530
    },
    {
      "epoch": 14.431341719077569,
      "grad_norm": 0.49457091093063354,
      "learning_rate": 1.0908901816907366e-05,
      "loss": 0.4306,
      "num_input_tokens_seen": 18012552,
      "step": 27535
    },
    {
      "epoch": 14.433962264150944,
      "grad_norm": 0.484605073928833,
      "learning_rate": 1.08994583673895e-05,
      "loss": 0.3765,
      "num_input_tokens_seen": 18017704,
      "step": 27540
    },
    {
      "epoch": 14.43658280922432,
      "grad_norm": 0.6298871636390686,
      "learning_rate": 1.0890017867544068e-05,
      "loss": 0.3518,
      "num_input_tokens_seen": 18021032,
      "step": 27545
    },
    {
      "epoch": 14.439203354297694,
      "grad_norm": 0.5286217331886292,
      "learning_rate": 1.088058031934592e-05,
      "loss": 0.5036,
      "num_input_tokens_seen": 18023944,
      "step": 27550
    },
    {
      "epoch": 14.44182389937107,
      "grad_norm": 0.3984656035900116,
      "learning_rate": 1.087114572476929e-05,
      "loss": 0.3903,
      "num_input_tokens_seen": 18027816,
      "step": 27555
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 0.4105580747127533,
      "learning_rate": 1.0861714085787783e-05,
      "loss": 0.3658,
      "num_input_tokens_seen": 18031080,
      "step": 27560
    },
    {
      "epoch": 14.44706498951782,
      "grad_norm": 0.4648188650608063,
      "learning_rate": 1.0852285404374402e-05,
      "loss": 0.2981,
      "num_input_tokens_seen": 18035496,
      "step": 27565
    },
    {
      "epoch": 14.449685534591195,
      "grad_norm": 0.6940786838531494,
      "learning_rate": 1.0842859682501505e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 18039720,
      "step": 27570
    },
    {
      "epoch": 14.45230607966457,
      "grad_norm": 0.5417872071266174,
      "learning_rate": 1.0833436922140841e-05,
      "loss": 0.5009,
      "num_input_tokens_seen": 18042696,
      "step": 27575
    },
    {
      "epoch": 14.454926624737945,
      "grad_norm": 0.4429219961166382,
      "learning_rate": 1.0824017125263555e-05,
      "loss": 0.4528,
      "num_input_tokens_seen": 18045864,
      "step": 27580
    },
    {
      "epoch": 14.45754716981132,
      "grad_norm": 0.3742965757846832,
      "learning_rate": 1.0814600293840154e-05,
      "loss": 0.3951,
      "num_input_tokens_seen": 18049544,
      "step": 27585
    },
    {
      "epoch": 14.460167714884696,
      "grad_norm": 0.38428542017936707,
      "learning_rate": 1.080518642984055e-05,
      "loss": 0.3651,
      "num_input_tokens_seen": 18052712,
      "step": 27590
    },
    {
      "epoch": 14.46278825995807,
      "grad_norm": 0.7867276668548584,
      "learning_rate": 1.0795775535233988e-05,
      "loss": 0.4713,
      "num_input_tokens_seen": 18055976,
      "step": 27595
    },
    {
      "epoch": 14.465408805031446,
      "grad_norm": 0.720099687576294,
      "learning_rate": 1.0786367611989145e-05,
      "loss": 0.4151,
      "num_input_tokens_seen": 18058632,
      "step": 27600
    },
    {
      "epoch": 14.468029350104821,
      "grad_norm": 0.5090008974075317,
      "learning_rate": 1.0776962662074031e-05,
      "loss": 0.4098,
      "num_input_tokens_seen": 18061160,
      "step": 27605
    },
    {
      "epoch": 14.470649895178196,
      "grad_norm": 0.27753958106040955,
      "learning_rate": 1.0767560687456062e-05,
      "loss": 0.3632,
      "num_input_tokens_seen": 18064328,
      "step": 27610
    },
    {
      "epoch": 14.473270440251572,
      "grad_norm": 0.4323285222053528,
      "learning_rate": 1.0758161690102036e-05,
      "loss": 0.4365,
      "num_input_tokens_seen": 18067464,
      "step": 27615
    },
    {
      "epoch": 14.475890985324948,
      "grad_norm": 0.5167339444160461,
      "learning_rate": 1.0748765671978095e-05,
      "loss": 0.5092,
      "num_input_tokens_seen": 18071912,
      "step": 27620
    },
    {
      "epoch": 14.478511530398324,
      "grad_norm": 0.8744324445724487,
      "learning_rate": 1.0739372635049797e-05,
      "loss": 0.3853,
      "num_input_tokens_seen": 18074728,
      "step": 27625
    },
    {
      "epoch": 14.481132075471699,
      "grad_norm": 0.35874131321907043,
      "learning_rate": 1.0729982581282042e-05,
      "loss": 0.4317,
      "num_input_tokens_seen": 18078568,
      "step": 27630
    },
    {
      "epoch": 14.483752620545074,
      "grad_norm": 0.4059484302997589,
      "learning_rate": 1.0720595512639129e-05,
      "loss": 0.3863,
      "num_input_tokens_seen": 18081128,
      "step": 27635
    },
    {
      "epoch": 14.48637316561845,
      "grad_norm": 0.5785218477249146,
      "learning_rate": 1.0711211431084728e-05,
      "loss": 0.4183,
      "num_input_tokens_seen": 18083976,
      "step": 27640
    },
    {
      "epoch": 14.488993710691824,
      "grad_norm": 0.286555677652359,
      "learning_rate": 1.0701830338581887e-05,
      "loss": 0.436,
      "num_input_tokens_seen": 18087720,
      "step": 27645
    },
    {
      "epoch": 14.4916142557652,
      "grad_norm": 0.5888288617134094,
      "learning_rate": 1.069245223709301e-05,
      "loss": 0.5596,
      "num_input_tokens_seen": 18090536,
      "step": 27650
    },
    {
      "epoch": 14.494234800838575,
      "grad_norm": 0.461858868598938,
      "learning_rate": 1.0683077128579904e-05,
      "loss": 0.3929,
      "num_input_tokens_seen": 18093960,
      "step": 27655
    },
    {
      "epoch": 14.49685534591195,
      "grad_norm": 0.38024160265922546,
      "learning_rate": 1.0673705015003718e-05,
      "loss": 0.5372,
      "num_input_tokens_seen": 18097128,
      "step": 27660
    },
    {
      "epoch": 14.499475890985325,
      "grad_norm": 0.46294480562210083,
      "learning_rate": 1.0664335898325002e-05,
      "loss": 0.465,
      "num_input_tokens_seen": 18100008,
      "step": 27665
    },
    {
      "epoch": 14.5020964360587,
      "grad_norm": 0.5904019474983215,
      "learning_rate": 1.065496978050367e-05,
      "loss": 0.6009,
      "num_input_tokens_seen": 18103080,
      "step": 27670
    },
    {
      "epoch": 14.504716981132075,
      "grad_norm": 0.36700162291526794,
      "learning_rate": 1.0645606663499e-05,
      "loss": 0.3262,
      "num_input_tokens_seen": 18106152,
      "step": 27675
    },
    {
      "epoch": 14.50733752620545,
      "grad_norm": 0.37760981917381287,
      "learning_rate": 1.0636246549269649e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 18109224,
      "step": 27680
    },
    {
      "epoch": 14.509958071278826,
      "grad_norm": 1.033045768737793,
      "learning_rate": 1.062688943977366e-05,
      "loss": 0.5198,
      "num_input_tokens_seen": 18112072,
      "step": 27685
    },
    {
      "epoch": 14.5125786163522,
      "grad_norm": 0.6874695420265198,
      "learning_rate": 1.0617535336968413e-05,
      "loss": 0.3967,
      "num_input_tokens_seen": 18114760,
      "step": 27690
    },
    {
      "epoch": 14.515199161425576,
      "grad_norm": 0.2667478919029236,
      "learning_rate": 1.0608184242810698e-05,
      "loss": 0.4153,
      "num_input_tokens_seen": 18121640,
      "step": 27695
    },
    {
      "epoch": 14.517819706498951,
      "grad_norm": 0.3873177766799927,
      "learning_rate": 1.0598836159256642e-05,
      "loss": 0.3903,
      "num_input_tokens_seen": 18124904,
      "step": 27700
    },
    {
      "epoch": 14.520440251572326,
      "grad_norm": 0.3963620364665985,
      "learning_rate": 1.0589491088261763e-05,
      "loss": 0.5197,
      "num_input_tokens_seen": 18127656,
      "step": 27705
    },
    {
      "epoch": 14.523060796645701,
      "grad_norm": 0.5507503151893616,
      "learning_rate": 1.0580149031780945e-05,
      "loss": 0.4363,
      "num_input_tokens_seen": 18130120,
      "step": 27710
    },
    {
      "epoch": 14.525681341719078,
      "grad_norm": 0.7211785316467285,
      "learning_rate": 1.0570809991768447e-05,
      "loss": 0.6111,
      "num_input_tokens_seen": 18132584,
      "step": 27715
    },
    {
      "epoch": 14.528301886792454,
      "grad_norm": 0.5817405581474304,
      "learning_rate": 1.0561473970177873e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 18138248,
      "step": 27720
    },
    {
      "epoch": 14.530922431865829,
      "grad_norm": 0.4401341676712036,
      "learning_rate": 1.0552140968962232e-05,
      "loss": 0.3302,
      "num_input_tokens_seen": 18140904,
      "step": 27725
    },
    {
      "epoch": 14.533542976939204,
      "grad_norm": 0.5791760683059692,
      "learning_rate": 1.054281099007386e-05,
      "loss": 0.4588,
      "num_input_tokens_seen": 18144360,
      "step": 27730
    },
    {
      "epoch": 14.536163522012579,
      "grad_norm": 0.45998963713645935,
      "learning_rate": 1.0533484035464492e-05,
      "loss": 0.4308,
      "num_input_tokens_seen": 18148712,
      "step": 27735
    },
    {
      "epoch": 14.538784067085954,
      "grad_norm": 0.3305354118347168,
      "learning_rate": 1.0524160107085231e-05,
      "loss": 0.334,
      "num_input_tokens_seen": 18151240,
      "step": 27740
    },
    {
      "epoch": 14.54140461215933,
      "grad_norm": 0.46527424454689026,
      "learning_rate": 1.0514839206886518e-05,
      "loss": 0.4917,
      "num_input_tokens_seen": 18154056,
      "step": 27745
    },
    {
      "epoch": 14.544025157232705,
      "grad_norm": 0.7904834151268005,
      "learning_rate": 1.0505521336818195e-05,
      "loss": 0.5231,
      "num_input_tokens_seen": 18158088,
      "step": 27750
    },
    {
      "epoch": 14.54664570230608,
      "grad_norm": 0.6681926250457764,
      "learning_rate": 1.0496206498829438e-05,
      "loss": 0.4658,
      "num_input_tokens_seen": 18161096,
      "step": 27755
    },
    {
      "epoch": 14.549266247379455,
      "grad_norm": 0.29736730456352234,
      "learning_rate": 1.0486894694868814e-05,
      "loss": 0.5076,
      "num_input_tokens_seen": 18164200,
      "step": 27760
    },
    {
      "epoch": 14.55188679245283,
      "grad_norm": 0.47693341970443726,
      "learning_rate": 1.0477585926884246e-05,
      "loss": 0.4498,
      "num_input_tokens_seen": 18168104,
      "step": 27765
    },
    {
      "epoch": 14.554507337526205,
      "grad_norm": 0.31387171149253845,
      "learning_rate": 1.046828019682303e-05,
      "loss": 0.4991,
      "num_input_tokens_seen": 18171528,
      "step": 27770
    },
    {
      "epoch": 14.55712788259958,
      "grad_norm": 0.6913873553276062,
      "learning_rate": 1.0458977506631808e-05,
      "loss": 0.4301,
      "num_input_tokens_seen": 18174568,
      "step": 27775
    },
    {
      "epoch": 14.559748427672956,
      "grad_norm": 0.40092596411705017,
      "learning_rate": 1.0449677858256593e-05,
      "loss": 0.6307,
      "num_input_tokens_seen": 18177480,
      "step": 27780
    },
    {
      "epoch": 14.56236897274633,
      "grad_norm": 0.4343433976173401,
      "learning_rate": 1.0440381253642773e-05,
      "loss": 0.3118,
      "num_input_tokens_seen": 18180168,
      "step": 27785
    },
    {
      "epoch": 14.564989517819706,
      "grad_norm": 0.4430420398712158,
      "learning_rate": 1.0431087694735087e-05,
      "loss": 0.348,
      "num_input_tokens_seen": 18183624,
      "step": 27790
    },
    {
      "epoch": 14.567610062893081,
      "grad_norm": 0.5823338627815247,
      "learning_rate": 1.0421797183477644e-05,
      "loss": 0.4206,
      "num_input_tokens_seen": 18186632,
      "step": 27795
    },
    {
      "epoch": 14.570230607966456,
      "grad_norm": 0.4149233400821686,
      "learning_rate": 1.041250972181392e-05,
      "loss": 0.3744,
      "num_input_tokens_seen": 18189320,
      "step": 27800
    },
    {
      "epoch": 14.572851153039831,
      "grad_norm": 0.3735992908477783,
      "learning_rate": 1.0403225311686743e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 18192392,
      "step": 27805
    },
    {
      "epoch": 14.575471698113208,
      "grad_norm": 0.5670915842056274,
      "learning_rate": 1.0393943955038293e-05,
      "loss": 0.4874,
      "num_input_tokens_seen": 18195112,
      "step": 27810
    },
    {
      "epoch": 14.578092243186584,
      "grad_norm": 0.6068394184112549,
      "learning_rate": 1.038466565381013e-05,
      "loss": 0.5346,
      "num_input_tokens_seen": 18198664,
      "step": 27815
    },
    {
      "epoch": 14.580712788259959,
      "grad_norm": 0.786653459072113,
      "learning_rate": 1.0375390409943175e-05,
      "loss": 0.4459,
      "num_input_tokens_seen": 18201800,
      "step": 27820
    },
    {
      "epoch": 14.583333333333334,
      "grad_norm": 0.4938866198062897,
      "learning_rate": 1.0366118225377708e-05,
      "loss": 0.3568,
      "num_input_tokens_seen": 18205096,
      "step": 27825
    },
    {
      "epoch": 14.585953878406709,
      "grad_norm": 0.4106466472148895,
      "learning_rate": 1.0356849102053353e-05,
      "loss": 0.3291,
      "num_input_tokens_seen": 18208008,
      "step": 27830
    },
    {
      "epoch": 14.588574423480084,
      "grad_norm": 0.2870727479457855,
      "learning_rate": 1.0347583041909104e-05,
      "loss": 0.463,
      "num_input_tokens_seen": 18211528,
      "step": 27835
    },
    {
      "epoch": 14.59119496855346,
      "grad_norm": 0.8165878653526306,
      "learning_rate": 1.0338320046883315e-05,
      "loss": 0.3941,
      "num_input_tokens_seen": 18214472,
      "step": 27840
    },
    {
      "epoch": 14.593815513626835,
      "grad_norm": 0.6464263796806335,
      "learning_rate": 1.0329060118913705e-05,
      "loss": 0.4769,
      "num_input_tokens_seen": 18217064,
      "step": 27845
    },
    {
      "epoch": 14.59643605870021,
      "grad_norm": 0.40407371520996094,
      "learning_rate": 1.0319803259937353e-05,
      "loss": 0.4097,
      "num_input_tokens_seen": 18220168,
      "step": 27850
    },
    {
      "epoch": 14.599056603773585,
      "grad_norm": 0.6586127877235413,
      "learning_rate": 1.0310549471890676e-05,
      "loss": 0.3597,
      "num_input_tokens_seen": 18222888,
      "step": 27855
    },
    {
      "epoch": 14.60167714884696,
      "grad_norm": 0.6431657075881958,
      "learning_rate": 1.0301298756709454e-05,
      "loss": 0.3774,
      "num_input_tokens_seen": 18225128,
      "step": 27860
    },
    {
      "epoch": 14.604297693920335,
      "grad_norm": 0.3093867301940918,
      "learning_rate": 1.0292051116328844e-05,
      "loss": 0.4028,
      "num_input_tokens_seen": 18228168,
      "step": 27865
    },
    {
      "epoch": 14.60691823899371,
      "grad_norm": 0.7299073338508606,
      "learning_rate": 1.028280655268334e-05,
      "loss": 0.4593,
      "num_input_tokens_seen": 18231816,
      "step": 27870
    },
    {
      "epoch": 14.609538784067086,
      "grad_norm": 0.6041938662528992,
      "learning_rate": 1.0273565067706806e-05,
      "loss": 0.4657,
      "num_input_tokens_seen": 18234120,
      "step": 27875
    },
    {
      "epoch": 14.61215932914046,
      "grad_norm": 0.6187328100204468,
      "learning_rate": 1.0264326663332457e-05,
      "loss": 0.4052,
      "num_input_tokens_seen": 18237320,
      "step": 27880
    },
    {
      "epoch": 14.614779874213836,
      "grad_norm": 0.4564744830131531,
      "learning_rate": 1.0255091341492859e-05,
      "loss": 0.5529,
      "num_input_tokens_seen": 18240008,
      "step": 27885
    },
    {
      "epoch": 14.617400419287211,
      "grad_norm": 0.33015599846839905,
      "learning_rate": 1.0245859104119923e-05,
      "loss": 0.4381,
      "num_input_tokens_seen": 18243720,
      "step": 27890
    },
    {
      "epoch": 14.620020964360586,
      "grad_norm": 0.5441310405731201,
      "learning_rate": 1.0236629953144938e-05,
      "loss": 0.3819,
      "num_input_tokens_seen": 18246088,
      "step": 27895
    },
    {
      "epoch": 14.622641509433961,
      "grad_norm": 0.35883012413978577,
      "learning_rate": 1.0227403890498537e-05,
      "loss": 0.4676,
      "num_input_tokens_seen": 18248904,
      "step": 27900
    },
    {
      "epoch": 14.625262054507338,
      "grad_norm": 0.3458663523197174,
      "learning_rate": 1.021818091811072e-05,
      "loss": 0.4375,
      "num_input_tokens_seen": 18252648,
      "step": 27905
    },
    {
      "epoch": 14.627882599580714,
      "grad_norm": 0.44128796458244324,
      "learning_rate": 1.0208961037910813e-05,
      "loss": 0.5082,
      "num_input_tokens_seen": 18255464,
      "step": 27910
    },
    {
      "epoch": 14.630503144654089,
      "grad_norm": 0.508922815322876,
      "learning_rate": 1.0199744251827503e-05,
      "loss": 0.5293,
      "num_input_tokens_seen": 18262408,
      "step": 27915
    },
    {
      "epoch": 14.633123689727464,
      "grad_norm": 0.6047188639640808,
      "learning_rate": 1.0190530561788845e-05,
      "loss": 0.3649,
      "num_input_tokens_seen": 18264968,
      "step": 27920
    },
    {
      "epoch": 14.635744234800839,
      "grad_norm": 0.5141869187355042,
      "learning_rate": 1.0181319969722236e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 18268104,
      "step": 27925
    },
    {
      "epoch": 14.638364779874214,
      "grad_norm": 0.43003180623054504,
      "learning_rate": 1.0172112477554441e-05,
      "loss": 0.577,
      "num_input_tokens_seen": 18271048,
      "step": 27930
    },
    {
      "epoch": 14.64098532494759,
      "grad_norm": 0.40556347370147705,
      "learning_rate": 1.016290808721154e-05,
      "loss": 0.3764,
      "num_input_tokens_seen": 18274440,
      "step": 27935
    },
    {
      "epoch": 14.643605870020965,
      "grad_norm": 0.5315119028091431,
      "learning_rate": 1.0153706800619007e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 18277672,
      "step": 27940
    },
    {
      "epoch": 14.64622641509434,
      "grad_norm": 0.3601570427417755,
      "learning_rate": 1.0144508619701623e-05,
      "loss": 0.3653,
      "num_input_tokens_seen": 18281640,
      "step": 27945
    },
    {
      "epoch": 14.648846960167715,
      "grad_norm": 0.6066627502441406,
      "learning_rate": 1.013531354638356e-05,
      "loss": 0.4424,
      "num_input_tokens_seen": 18284648,
      "step": 27950
    },
    {
      "epoch": 14.65146750524109,
      "grad_norm": 1.139675259590149,
      "learning_rate": 1.0126121582588316e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 18287624,
      "step": 27955
    },
    {
      "epoch": 14.654088050314465,
      "grad_norm": 0.5126003623008728,
      "learning_rate": 1.0116932730238754e-05,
      "loss": 0.3586,
      "num_input_tokens_seen": 18291752,
      "step": 27960
    },
    {
      "epoch": 14.65670859538784,
      "grad_norm": 0.4323464632034302,
      "learning_rate": 1.0107746991257063e-05,
      "loss": 0.3561,
      "num_input_tokens_seen": 18295400,
      "step": 27965
    },
    {
      "epoch": 14.659329140461216,
      "grad_norm": 0.7409646511077881,
      "learning_rate": 1.0098564367564811e-05,
      "loss": 0.4049,
      "num_input_tokens_seen": 18298440,
      "step": 27970
    },
    {
      "epoch": 14.66194968553459,
      "grad_norm": 0.4887104332447052,
      "learning_rate": 1.0089384861082876e-05,
      "loss": 0.4614,
      "num_input_tokens_seen": 18301480,
      "step": 27975
    },
    {
      "epoch": 14.664570230607966,
      "grad_norm": 0.6440389156341553,
      "learning_rate": 1.0080208473731523e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 18304744,
      "step": 27980
    },
    {
      "epoch": 14.667190775681341,
      "grad_norm": 0.42650169134140015,
      "learning_rate": 1.0071035207430352e-05,
      "loss": 0.4559,
      "num_input_tokens_seen": 18308328,
      "step": 27985
    },
    {
      "epoch": 14.669811320754716,
      "grad_norm": 0.458039790391922,
      "learning_rate": 1.0061865064098285e-05,
      "loss": 0.4359,
      "num_input_tokens_seen": 18311240,
      "step": 27990
    },
    {
      "epoch": 14.672431865828091,
      "grad_norm": 0.5388805270195007,
      "learning_rate": 1.0052698045653625e-05,
      "loss": 0.4407,
      "num_input_tokens_seen": 18314344,
      "step": 27995
    },
    {
      "epoch": 14.675052410901468,
      "grad_norm": 0.20503997802734375,
      "learning_rate": 1.0043534154014015e-05,
      "loss": 0.4039,
      "num_input_tokens_seen": 18317544,
      "step": 28000
    },
    {
      "epoch": 14.677672955974844,
      "grad_norm": 0.36150428652763367,
      "learning_rate": 1.003437339109642e-05,
      "loss": 0.5289,
      "num_input_tokens_seen": 18320776,
      "step": 28005
    },
    {
      "epoch": 14.680293501048219,
      "grad_norm": 0.589545726776123,
      "learning_rate": 1.0025215758817183e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 18325096,
      "step": 28010
    },
    {
      "epoch": 14.682914046121594,
      "grad_norm": 0.48121803998947144,
      "learning_rate": 1.0016061259091958e-05,
      "loss": 0.4849,
      "num_input_tokens_seen": 18328392,
      "step": 28015
    },
    {
      "epoch": 14.685534591194969,
      "grad_norm": 0.6784305572509766,
      "learning_rate": 1.000690989383577e-05,
      "loss": 0.4957,
      "num_input_tokens_seen": 18331720,
      "step": 28020
    },
    {
      "epoch": 14.688155136268344,
      "grad_norm": 0.4795266091823578,
      "learning_rate": 9.997761664962993e-06,
      "loss": 0.3729,
      "num_input_tokens_seen": 18334536,
      "step": 28025
    },
    {
      "epoch": 14.69077568134172,
      "grad_norm": 0.331089049577713,
      "learning_rate": 9.988616574387314e-06,
      "loss": 0.4479,
      "num_input_tokens_seen": 18339304,
      "step": 28030
    },
    {
      "epoch": 14.693396226415095,
      "grad_norm": 0.46280327439308167,
      "learning_rate": 9.979474624021787e-06,
      "loss": 0.3721,
      "num_input_tokens_seen": 18341896,
      "step": 28035
    },
    {
      "epoch": 14.69601677148847,
      "grad_norm": 0.6818332076072693,
      "learning_rate": 9.970335815778815e-06,
      "loss": 0.3514,
      "num_input_tokens_seen": 18345736,
      "step": 28040
    },
    {
      "epoch": 14.698637316561845,
      "grad_norm": 0.6176140308380127,
      "learning_rate": 9.96120015157011e-06,
      "loss": 0.481,
      "num_input_tokens_seen": 18348552,
      "step": 28045
    },
    {
      "epoch": 14.70125786163522,
      "grad_norm": 0.544853150844574,
      "learning_rate": 9.952067633306767e-06,
      "loss": 0.5315,
      "num_input_tokens_seen": 18351624,
      "step": 28050
    },
    {
      "epoch": 14.703878406708595,
      "grad_norm": 0.3322210907936096,
      "learning_rate": 9.942938262899204e-06,
      "loss": 0.4432,
      "num_input_tokens_seen": 18354632,
      "step": 28055
    },
    {
      "epoch": 14.70649895178197,
      "grad_norm": 0.5560405254364014,
      "learning_rate": 9.93381204225717e-06,
      "loss": 0.3222,
      "num_input_tokens_seen": 18357736,
      "step": 28060
    },
    {
      "epoch": 14.709119496855346,
      "grad_norm": 0.337740421295166,
      "learning_rate": 9.924688973289782e-06,
      "loss": 0.4169,
      "num_input_tokens_seen": 18362120,
      "step": 28065
    },
    {
      "epoch": 14.71174004192872,
      "grad_norm": 0.4098382890224457,
      "learning_rate": 9.915569057905463e-06,
      "loss": 0.4502,
      "num_input_tokens_seen": 18366312,
      "step": 28070
    },
    {
      "epoch": 14.714360587002096,
      "grad_norm": 0.5080645680427551,
      "learning_rate": 9.906452298012003e-06,
      "loss": 0.4095,
      "num_input_tokens_seen": 18369704,
      "step": 28075
    },
    {
      "epoch": 14.716981132075471,
      "grad_norm": 0.3493153154850006,
      "learning_rate": 9.897338695516529e-06,
      "loss": 0.4744,
      "num_input_tokens_seen": 18372840,
      "step": 28080
    },
    {
      "epoch": 14.719601677148846,
      "grad_norm": 0.4010658264160156,
      "learning_rate": 9.888228252325507e-06,
      "loss": 0.4027,
      "num_input_tokens_seen": 18376392,
      "step": 28085
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 0.5155462622642517,
      "learning_rate": 9.87912097034473e-06,
      "loss": 0.6528,
      "num_input_tokens_seen": 18379752,
      "step": 28090
    },
    {
      "epoch": 14.724842767295598,
      "grad_norm": 0.2889532148838043,
      "learning_rate": 9.870016851479327e-06,
      "loss": 0.321,
      "num_input_tokens_seen": 18383048,
      "step": 28095
    },
    {
      "epoch": 14.727463312368974,
      "grad_norm": 0.6213849186897278,
      "learning_rate": 9.860915897633791e-06,
      "loss": 0.5813,
      "num_input_tokens_seen": 18386312,
      "step": 28100
    },
    {
      "epoch": 14.730083857442349,
      "grad_norm": 0.4549620449542999,
      "learning_rate": 9.85181811071193e-06,
      "loss": 0.6329,
      "num_input_tokens_seen": 18388936,
      "step": 28105
    },
    {
      "epoch": 14.732704402515724,
      "grad_norm": 0.4897574186325073,
      "learning_rate": 9.842723492616913e-06,
      "loss": 0.3772,
      "num_input_tokens_seen": 18393448,
      "step": 28110
    },
    {
      "epoch": 14.735324947589099,
      "grad_norm": 0.5839179158210754,
      "learning_rate": 9.833632045251207e-06,
      "loss": 0.5259,
      "num_input_tokens_seen": 18396456,
      "step": 28115
    },
    {
      "epoch": 14.737945492662474,
      "grad_norm": 0.4586019217967987,
      "learning_rate": 9.824543770516656e-06,
      "loss": 0.3896,
      "num_input_tokens_seen": 18399688,
      "step": 28120
    },
    {
      "epoch": 14.74056603773585,
      "grad_norm": 0.6615382432937622,
      "learning_rate": 9.815458670314412e-06,
      "loss": 0.3687,
      "num_input_tokens_seen": 18402664,
      "step": 28125
    },
    {
      "epoch": 14.743186582809225,
      "grad_norm": 0.48906999826431274,
      "learning_rate": 9.80637674654498e-06,
      "loss": 0.368,
      "num_input_tokens_seen": 18405640,
      "step": 28130
    },
    {
      "epoch": 14.7458071278826,
      "grad_norm": 0.5378610491752625,
      "learning_rate": 9.797298001108193e-06,
      "loss": 0.4911,
      "num_input_tokens_seen": 18408424,
      "step": 28135
    },
    {
      "epoch": 14.748427672955975,
      "grad_norm": 0.5743562579154968,
      "learning_rate": 9.788222435903232e-06,
      "loss": 0.5375,
      "num_input_tokens_seen": 18412488,
      "step": 28140
    },
    {
      "epoch": 14.75104821802935,
      "grad_norm": 0.4265981614589691,
      "learning_rate": 9.779150052828596e-06,
      "loss": 0.4088,
      "num_input_tokens_seen": 18415944,
      "step": 28145
    },
    {
      "epoch": 14.753668763102725,
      "grad_norm": 0.28525540232658386,
      "learning_rate": 9.770080853782108e-06,
      "loss": 0.3832,
      "num_input_tokens_seen": 18420136,
      "step": 28150
    },
    {
      "epoch": 14.7562893081761,
      "grad_norm": 0.5737215876579285,
      "learning_rate": 9.76101484066095e-06,
      "loss": 0.4893,
      "num_input_tokens_seen": 18424552,
      "step": 28155
    },
    {
      "epoch": 14.758909853249476,
      "grad_norm": 1.5014476776123047,
      "learning_rate": 9.751952015361637e-06,
      "loss": 0.4699,
      "num_input_tokens_seen": 18427112,
      "step": 28160
    },
    {
      "epoch": 14.76153039832285,
      "grad_norm": 0.842544674873352,
      "learning_rate": 9.742892379780003e-06,
      "loss": 0.474,
      "num_input_tokens_seen": 18430216,
      "step": 28165
    },
    {
      "epoch": 14.764150943396226,
      "grad_norm": 0.48650220036506653,
      "learning_rate": 9.733835935811237e-06,
      "loss": 0.5529,
      "num_input_tokens_seen": 18436232,
      "step": 28170
    },
    {
      "epoch": 14.766771488469601,
      "grad_norm": 1.0471851825714111,
      "learning_rate": 9.724782685349809e-06,
      "loss": 0.503,
      "num_input_tokens_seen": 18439528,
      "step": 28175
    },
    {
      "epoch": 14.769392033542976,
      "grad_norm": 0.44379469752311707,
      "learning_rate": 9.715732630289576e-06,
      "loss": 0.347,
      "num_input_tokens_seen": 18442728,
      "step": 28180
    },
    {
      "epoch": 14.772012578616351,
      "grad_norm": 0.5240036845207214,
      "learning_rate": 9.706685772523701e-06,
      "loss": 0.602,
      "num_input_tokens_seen": 18445640,
      "step": 28185
    },
    {
      "epoch": 14.774633123689728,
      "grad_norm": 0.3107626736164093,
      "learning_rate": 9.697642113944692e-06,
      "loss": 0.365,
      "num_input_tokens_seen": 18448584,
      "step": 28190
    },
    {
      "epoch": 14.777253668763104,
      "grad_norm": 1.0294404029846191,
      "learning_rate": 9.688601656444381e-06,
      "loss": 0.4868,
      "num_input_tokens_seen": 18451112,
      "step": 28195
    },
    {
      "epoch": 14.779874213836479,
      "grad_norm": 0.6577557921409607,
      "learning_rate": 9.679564401913921e-06,
      "loss": 0.415,
      "num_input_tokens_seen": 18453512,
      "step": 28200
    },
    {
      "epoch": 14.782494758909854,
      "grad_norm": 0.3286266326904297,
      "learning_rate": 9.670530352243798e-06,
      "loss": 0.5681,
      "num_input_tokens_seen": 18457320,
      "step": 28205
    },
    {
      "epoch": 14.785115303983229,
      "grad_norm": 0.46650251746177673,
      "learning_rate": 9.661499509323835e-06,
      "loss": 0.4102,
      "num_input_tokens_seen": 18460712,
      "step": 28210
    },
    {
      "epoch": 14.787735849056604,
      "grad_norm": 0.5061944127082825,
      "learning_rate": 9.652471875043182e-06,
      "loss": 0.3202,
      "num_input_tokens_seen": 18463720,
      "step": 28215
    },
    {
      "epoch": 14.79035639412998,
      "grad_norm": 0.8628416061401367,
      "learning_rate": 9.64344745129033e-06,
      "loss": 0.5935,
      "num_input_tokens_seen": 18466344,
      "step": 28220
    },
    {
      "epoch": 14.792976939203355,
      "grad_norm": 0.5003984570503235,
      "learning_rate": 9.634426239953073e-06,
      "loss": 0.4569,
      "num_input_tokens_seen": 18469320,
      "step": 28225
    },
    {
      "epoch": 14.79559748427673,
      "grad_norm": 0.46715518832206726,
      "learning_rate": 9.62540824291854e-06,
      "loss": 0.4648,
      "num_input_tokens_seen": 18472360,
      "step": 28230
    },
    {
      "epoch": 14.798218029350105,
      "grad_norm": 0.4487735331058502,
      "learning_rate": 9.616393462073197e-06,
      "loss": 0.3473,
      "num_input_tokens_seen": 18475080,
      "step": 28235
    },
    {
      "epoch": 14.80083857442348,
      "grad_norm": 0.9979318976402283,
      "learning_rate": 9.607381899302839e-06,
      "loss": 0.383,
      "num_input_tokens_seen": 18477672,
      "step": 28240
    },
    {
      "epoch": 14.803459119496855,
      "grad_norm": 0.8860336542129517,
      "learning_rate": 9.59837355649258e-06,
      "loss": 0.412,
      "num_input_tokens_seen": 18480776,
      "step": 28245
    },
    {
      "epoch": 14.80607966457023,
      "grad_norm": 0.8329493403434753,
      "learning_rate": 9.589368435526868e-06,
      "loss": 0.4505,
      "num_input_tokens_seen": 18484104,
      "step": 28250
    },
    {
      "epoch": 14.808700209643606,
      "grad_norm": 0.49662837386131287,
      "learning_rate": 9.580366538289468e-06,
      "loss": 0.5104,
      "num_input_tokens_seen": 18487560,
      "step": 28255
    },
    {
      "epoch": 14.81132075471698,
      "grad_norm": 0.2521352171897888,
      "learning_rate": 9.57136786666346e-06,
      "loss": 0.4921,
      "num_input_tokens_seen": 18491112,
      "step": 28260
    },
    {
      "epoch": 14.813941299790356,
      "grad_norm": 0.609491765499115,
      "learning_rate": 9.562372422531276e-06,
      "loss": 0.4215,
      "num_input_tokens_seen": 18494184,
      "step": 28265
    },
    {
      "epoch": 14.816561844863731,
      "grad_norm": 0.9120352864265442,
      "learning_rate": 9.553380207774662e-06,
      "loss": 0.4751,
      "num_input_tokens_seen": 18497864,
      "step": 28270
    },
    {
      "epoch": 14.819182389937106,
      "grad_norm": 0.546209990978241,
      "learning_rate": 9.54439122427469e-06,
      "loss": 0.4038,
      "num_input_tokens_seen": 18500264,
      "step": 28275
    },
    {
      "epoch": 14.821802935010481,
      "grad_norm": 0.9352949857711792,
      "learning_rate": 9.535405473911748e-06,
      "loss": 0.4928,
      "num_input_tokens_seen": 18503240,
      "step": 28280
    },
    {
      "epoch": 14.824423480083858,
      "grad_norm": 0.7319332361221313,
      "learning_rate": 9.526422958565546e-06,
      "loss": 0.414,
      "num_input_tokens_seen": 18506920,
      "step": 28285
    },
    {
      "epoch": 14.827044025157234,
      "grad_norm": 0.5562358498573303,
      "learning_rate": 9.517443680115124e-06,
      "loss": 0.4609,
      "num_input_tokens_seen": 18509896,
      "step": 28290
    },
    {
      "epoch": 14.829664570230609,
      "grad_norm": 1.1236438751220703,
      "learning_rate": 9.50846764043885e-06,
      "loss": 0.4547,
      "num_input_tokens_seen": 18514056,
      "step": 28295
    },
    {
      "epoch": 14.832285115303984,
      "grad_norm": 0.6305713057518005,
      "learning_rate": 9.499494841414422e-06,
      "loss": 0.6174,
      "num_input_tokens_seen": 18516616,
      "step": 28300
    },
    {
      "epoch": 14.834905660377359,
      "grad_norm": 0.6686530709266663,
      "learning_rate": 9.490525284918822e-06,
      "loss": 0.3801,
      "num_input_tokens_seen": 18519496,
      "step": 28305
    },
    {
      "epoch": 14.837526205450734,
      "grad_norm": 0.42187878489494324,
      "learning_rate": 9.4815589728284e-06,
      "loss": 0.5666,
      "num_input_tokens_seen": 18523272,
      "step": 28310
    },
    {
      "epoch": 14.84014675052411,
      "grad_norm": 0.35998114943504333,
      "learning_rate": 9.472595907018788e-06,
      "loss": 0.5777,
      "num_input_tokens_seen": 18528424,
      "step": 28315
    },
    {
      "epoch": 14.842767295597485,
      "grad_norm": 0.5595922470092773,
      "learning_rate": 9.463636089364968e-06,
      "loss": 0.3825,
      "num_input_tokens_seen": 18531112,
      "step": 28320
    },
    {
      "epoch": 14.84538784067086,
      "grad_norm": 0.6150668859481812,
      "learning_rate": 9.454679521741239e-06,
      "loss": 0.4244,
      "num_input_tokens_seen": 18534024,
      "step": 28325
    },
    {
      "epoch": 14.848008385744235,
      "grad_norm": 0.5511452555656433,
      "learning_rate": 9.445726206021194e-06,
      "loss": 0.4168,
      "num_input_tokens_seen": 18537320,
      "step": 28330
    },
    {
      "epoch": 14.85062893081761,
      "grad_norm": 0.4012407064437866,
      "learning_rate": 9.436776144077772e-06,
      "loss": 0.4634,
      "num_input_tokens_seen": 18539848,
      "step": 28335
    },
    {
      "epoch": 14.853249475890985,
      "grad_norm": 0.3862428665161133,
      "learning_rate": 9.427829337783237e-06,
      "loss": 0.3555,
      "num_input_tokens_seen": 18542536,
      "step": 28340
    },
    {
      "epoch": 14.85587002096436,
      "grad_norm": 0.33518701791763306,
      "learning_rate": 9.418885789009138e-06,
      "loss": 0.4561,
      "num_input_tokens_seen": 18545768,
      "step": 28345
    },
    {
      "epoch": 14.858490566037736,
      "grad_norm": 0.5263813138008118,
      "learning_rate": 9.409945499626374e-06,
      "loss": 0.6666,
      "num_input_tokens_seen": 18549512,
      "step": 28350
    },
    {
      "epoch": 14.86111111111111,
      "grad_norm": 0.338716059923172,
      "learning_rate": 9.40100847150516e-06,
      "loss": 0.4175,
      "num_input_tokens_seen": 18553320,
      "step": 28355
    },
    {
      "epoch": 14.863731656184486,
      "grad_norm": 0.5605983138084412,
      "learning_rate": 9.392074706515003e-06,
      "loss": 0.5132,
      "num_input_tokens_seen": 18557064,
      "step": 28360
    },
    {
      "epoch": 14.866352201257861,
      "grad_norm": 0.6571401953697205,
      "learning_rate": 9.383144206524753e-06,
      "loss": 0.4993,
      "num_input_tokens_seen": 18560968,
      "step": 28365
    },
    {
      "epoch": 14.868972746331236,
      "grad_norm": 0.3087805211544037,
      "learning_rate": 9.374216973402575e-06,
      "loss": 0.4048,
      "num_input_tokens_seen": 18564136,
      "step": 28370
    },
    {
      "epoch": 14.871593291404611,
      "grad_norm": 0.32030314207077026,
      "learning_rate": 9.365293009015926e-06,
      "loss": 0.511,
      "num_input_tokens_seen": 18567144,
      "step": 28375
    },
    {
      "epoch": 14.874213836477988,
      "grad_norm": 0.385304719209671,
      "learning_rate": 9.356372315231623e-06,
      "loss": 0.4886,
      "num_input_tokens_seen": 18570856,
      "step": 28380
    },
    {
      "epoch": 14.876834381551364,
      "grad_norm": 0.5229487419128418,
      "learning_rate": 9.347454893915749e-06,
      "loss": 0.4948,
      "num_input_tokens_seen": 18573704,
      "step": 28385
    },
    {
      "epoch": 14.879454926624739,
      "grad_norm": 1.095288634300232,
      "learning_rate": 9.338540746933741e-06,
      "loss": 0.4764,
      "num_input_tokens_seen": 18576392,
      "step": 28390
    },
    {
      "epoch": 14.882075471698114,
      "grad_norm": 0.5744603276252747,
      "learning_rate": 9.329629876150339e-06,
      "loss": 0.479,
      "num_input_tokens_seen": 18579688,
      "step": 28395
    },
    {
      "epoch": 14.884696016771489,
      "grad_norm": 0.6018508076667786,
      "learning_rate": 9.320722283429581e-06,
      "loss": 0.3608,
      "num_input_tokens_seen": 18582376,
      "step": 28400
    },
    {
      "epoch": 14.887316561844864,
      "grad_norm": 0.3639920651912689,
      "learning_rate": 9.311817970634854e-06,
      "loss": 0.4169,
      "num_input_tokens_seen": 18584904,
      "step": 28405
    },
    {
      "epoch": 14.88993710691824,
      "grad_norm": 0.859661877155304,
      "learning_rate": 9.302916939628814e-06,
      "loss": 0.4732,
      "num_input_tokens_seen": 18588328,
      "step": 28410
    },
    {
      "epoch": 14.892557651991615,
      "grad_norm": 0.663942813873291,
      "learning_rate": 9.29401919227347e-06,
      "loss": 0.4536,
      "num_input_tokens_seen": 18591176,
      "step": 28415
    },
    {
      "epoch": 14.89517819706499,
      "grad_norm": 0.26396483182907104,
      "learning_rate": 9.28512473043013e-06,
      "loss": 0.3653,
      "num_input_tokens_seen": 18594536,
      "step": 28420
    },
    {
      "epoch": 14.897798742138365,
      "grad_norm": 0.3990524709224701,
      "learning_rate": 9.276233555959419e-06,
      "loss": 0.4583,
      "num_input_tokens_seen": 18597800,
      "step": 28425
    },
    {
      "epoch": 14.90041928721174,
      "grad_norm": 0.4712826907634735,
      "learning_rate": 9.267345670721254e-06,
      "loss": 0.4078,
      "num_input_tokens_seen": 18601416,
      "step": 28430
    },
    {
      "epoch": 14.903039832285115,
      "grad_norm": 0.32880058884620667,
      "learning_rate": 9.258461076574895e-06,
      "loss": 0.418,
      "num_input_tokens_seen": 18604136,
      "step": 28435
    },
    {
      "epoch": 14.90566037735849,
      "grad_norm": 0.4062502086162567,
      "learning_rate": 9.24957977537888e-06,
      "loss": 0.4281,
      "num_input_tokens_seen": 18607016,
      "step": 28440
    },
    {
      "epoch": 14.908280922431866,
      "grad_norm": 0.7171783447265625,
      "learning_rate": 9.240701768991086e-06,
      "loss": 0.3578,
      "num_input_tokens_seen": 18610728,
      "step": 28445
    },
    {
      "epoch": 14.91090146750524,
      "grad_norm": 0.449497252702713,
      "learning_rate": 9.231827059268692e-06,
      "loss": 0.4449,
      "num_input_tokens_seen": 18613960,
      "step": 28450
    },
    {
      "epoch": 14.913522012578616,
      "grad_norm": 0.6187649369239807,
      "learning_rate": 9.222955648068193e-06,
      "loss": 0.385,
      "num_input_tokens_seen": 18617032,
      "step": 28455
    },
    {
      "epoch": 14.916142557651991,
      "grad_norm": 0.8746821880340576,
      "learning_rate": 9.214087537245378e-06,
      "loss": 0.41,
      "num_input_tokens_seen": 18619816,
      "step": 28460
    },
    {
      "epoch": 14.918763102725366,
      "grad_norm": 0.3743959069252014,
      "learning_rate": 9.205222728655344e-06,
      "loss": 0.3916,
      "num_input_tokens_seen": 18623560,
      "step": 28465
    },
    {
      "epoch": 14.921383647798741,
      "grad_norm": 0.6332958936691284,
      "learning_rate": 9.196361224152522e-06,
      "loss": 0.6342,
      "num_input_tokens_seen": 18626120,
      "step": 28470
    },
    {
      "epoch": 14.924004192872118,
      "grad_norm": 0.6029223203659058,
      "learning_rate": 9.187503025590632e-06,
      "loss": 0.5538,
      "num_input_tokens_seen": 18629000,
      "step": 28475
    },
    {
      "epoch": 14.926624737945493,
      "grad_norm": 0.4215571880340576,
      "learning_rate": 9.17864813482272e-06,
      "loss": 0.4551,
      "num_input_tokens_seen": 18631464,
      "step": 28480
    },
    {
      "epoch": 14.929245283018869,
      "grad_norm": 0.4688447117805481,
      "learning_rate": 9.169796553701122e-06,
      "loss": 0.4104,
      "num_input_tokens_seen": 18635208,
      "step": 28485
    },
    {
      "epoch": 14.931865828092244,
      "grad_norm": 0.3833421468734741,
      "learning_rate": 9.160948284077473e-06,
      "loss": 0.4247,
      "num_input_tokens_seen": 18638632,
      "step": 28490
    },
    {
      "epoch": 14.934486373165619,
      "grad_norm": 0.40980562567710876,
      "learning_rate": 9.152103327802738e-06,
      "loss": 0.3895,
      "num_input_tokens_seen": 18641864,
      "step": 28495
    },
    {
      "epoch": 14.937106918238994,
      "grad_norm": 0.5619413256645203,
      "learning_rate": 9.143261686727189e-06,
      "loss": 0.4041,
      "num_input_tokens_seen": 18644328,
      "step": 28500
    },
    {
      "epoch": 14.93972746331237,
      "grad_norm": 0.5351954102516174,
      "learning_rate": 9.134423362700392e-06,
      "loss": 0.5286,
      "num_input_tokens_seen": 18647976,
      "step": 28505
    },
    {
      "epoch": 14.942348008385745,
      "grad_norm": 0.5814465880393982,
      "learning_rate": 9.12558835757123e-06,
      "loss": 0.5314,
      "num_input_tokens_seen": 18650920,
      "step": 28510
    },
    {
      "epoch": 14.94496855345912,
      "grad_norm": 0.5369204878807068,
      "learning_rate": 9.116756673187878e-06,
      "loss": 0.3344,
      "num_input_tokens_seen": 18653224,
      "step": 28515
    },
    {
      "epoch": 14.947589098532495,
      "grad_norm": 0.2529407739639282,
      "learning_rate": 9.107928311397817e-06,
      "loss": 0.3765,
      "num_input_tokens_seen": 18656136,
      "step": 28520
    },
    {
      "epoch": 14.95020964360587,
      "grad_norm": 0.36563459038734436,
      "learning_rate": 9.09910327404785e-06,
      "loss": 0.4195,
      "num_input_tokens_seen": 18660072,
      "step": 28525
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.4929691255092621,
      "learning_rate": 9.090281562984068e-06,
      "loss": 0.3567,
      "num_input_tokens_seen": 18663368,
      "step": 28530
    },
    {
      "epoch": 14.95545073375262,
      "grad_norm": 0.5672488212585449,
      "learning_rate": 9.081463180051883e-06,
      "loss": 0.5302,
      "num_input_tokens_seen": 18666952,
      "step": 28535
    },
    {
      "epoch": 14.958071278825996,
      "grad_norm": 0.3115173280239105,
      "learning_rate": 9.072648127095996e-06,
      "loss": 0.4071,
      "num_input_tokens_seen": 18670344,
      "step": 28540
    },
    {
      "epoch": 14.96069182389937,
      "grad_norm": 0.27374154329299927,
      "learning_rate": 9.0638364059604e-06,
      "loss": 0.4187,
      "num_input_tokens_seen": 18673768,
      "step": 28545
    },
    {
      "epoch": 14.963312368972746,
      "grad_norm": 0.48041239380836487,
      "learning_rate": 9.055028018488417e-06,
      "loss": 0.5684,
      "num_input_tokens_seen": 18677544,
      "step": 28550
    },
    {
      "epoch": 14.965932914046121,
      "grad_norm": 0.44431257247924805,
      "learning_rate": 9.046222966522665e-06,
      "loss": 0.4589,
      "num_input_tokens_seen": 18681064,
      "step": 28555
    },
    {
      "epoch": 14.968553459119496,
      "grad_norm": 0.7793449759483337,
      "learning_rate": 9.037421251905059e-06,
      "loss": 0.5967,
      "num_input_tokens_seen": 18683880,
      "step": 28560
    },
    {
      "epoch": 14.971174004192871,
      "grad_norm": 0.5407832860946655,
      "learning_rate": 9.028622876476833e-06,
      "loss": 0.4145,
      "num_input_tokens_seen": 18687240,
      "step": 28565
    },
    {
      "epoch": 14.973794549266248,
      "grad_norm": 0.6071179509162903,
      "learning_rate": 9.019827842078474e-06,
      "loss": 0.5118,
      "num_input_tokens_seen": 18690280,
      "step": 28570
    },
    {
      "epoch": 14.976415094339622,
      "grad_norm": 0.46709468960762024,
      "learning_rate": 9.011036150549817e-06,
      "loss": 0.3819,
      "num_input_tokens_seen": 18693928,
      "step": 28575
    },
    {
      "epoch": 14.979035639412999,
      "grad_norm": 0.752576470375061,
      "learning_rate": 9.002247803729987e-06,
      "loss": 0.385,
      "num_input_tokens_seen": 18696680,
      "step": 28580
    },
    {
      "epoch": 14.981656184486374,
      "grad_norm": 0.43833693861961365,
      "learning_rate": 8.993462803457404e-06,
      "loss": 0.55,
      "num_input_tokens_seen": 18700488,
      "step": 28585
    },
    {
      "epoch": 14.984276729559749,
      "grad_norm": 0.6189985275268555,
      "learning_rate": 8.984681151569799e-06,
      "loss": 0.4282,
      "num_input_tokens_seen": 18703240,
      "step": 28590
    },
    {
      "epoch": 14.986897274633124,
      "grad_norm": 0.5929753184318542,
      "learning_rate": 8.975902849904183e-06,
      "loss": 0.3845,
      "num_input_tokens_seen": 18706184,
      "step": 28595
    },
    {
      "epoch": 14.9895178197065,
      "grad_norm": 0.4336085915565491,
      "learning_rate": 8.967127900296874e-06,
      "loss": 0.3171,
      "num_input_tokens_seen": 18710024,
      "step": 28600
    },
    {
      "epoch": 14.992138364779874,
      "grad_norm": 0.4932886064052582,
      "learning_rate": 8.95835630458349e-06,
      "loss": 0.5256,
      "num_input_tokens_seen": 18713448,
      "step": 28605
    },
    {
      "epoch": 14.99475890985325,
      "grad_norm": 0.6025798320770264,
      "learning_rate": 8.949588064598958e-06,
      "loss": 0.3729,
      "num_input_tokens_seen": 18716136,
      "step": 28610
    },
    {
      "epoch": 14.997379454926625,
      "grad_norm": 0.4439403712749481,
      "learning_rate": 8.940823182177499e-06,
      "loss": 0.352,
      "num_input_tokens_seen": 18719464,
      "step": 28615
    },
    {
      "epoch": 15.0,
      "grad_norm": 2.2418830394744873,
      "learning_rate": 8.932061659152607e-06,
      "loss": 0.4654,
      "num_input_tokens_seen": 18722440,
      "step": 28620
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.47713857889175415,
      "eval_runtime": 9.7189,
      "eval_samples_per_second": 87.253,
      "eval_steps_per_second": 21.813,
      "num_input_tokens_seen": 18722440,
      "step": 28620
    },
    {
      "epoch": 15.002620545073375,
      "grad_norm": 0.2328372448682785,
      "learning_rate": 8.92330349735711e-06,
      "loss": 0.3742,
      "num_input_tokens_seen": 18725288,
      "step": 28625
    },
    {
      "epoch": 15.00524109014675,
      "grad_norm": 0.43431270122528076,
      "learning_rate": 8.914548698623104e-06,
      "loss": 0.5395,
      "num_input_tokens_seen": 18728072,
      "step": 28630
    },
    {
      "epoch": 15.007861635220126,
      "grad_norm": 0.404285192489624,
      "learning_rate": 8.905797264781995e-06,
      "loss": 0.5893,
      "num_input_tokens_seen": 18734728,
      "step": 28635
    },
    {
      "epoch": 15.0104821802935,
      "grad_norm": 0.4577624797821045,
      "learning_rate": 8.89704919766449e-06,
      "loss": 0.4208,
      "num_input_tokens_seen": 18737352,
      "step": 28640
    },
    {
      "epoch": 15.013102725366876,
      "grad_norm": 0.4402674436569214,
      "learning_rate": 8.888304499100586e-06,
      "loss": 0.4351,
      "num_input_tokens_seen": 18741544,
      "step": 28645
    },
    {
      "epoch": 15.015723270440251,
      "grad_norm": 0.44619521498680115,
      "learning_rate": 8.879563170919566e-06,
      "loss": 0.3806,
      "num_input_tokens_seen": 18745032,
      "step": 28650
    },
    {
      "epoch": 15.018343815513626,
      "grad_norm": 0.4006041884422302,
      "learning_rate": 8.870825214950016e-06,
      "loss": 0.4082,
      "num_input_tokens_seen": 18750088,
      "step": 28655
    },
    {
      "epoch": 15.020964360587001,
      "grad_norm": 0.503549337387085,
      "learning_rate": 8.862090633019818e-06,
      "loss": 0.51,
      "num_input_tokens_seen": 18753864,
      "step": 28660
    },
    {
      "epoch": 15.023584905660377,
      "grad_norm": 0.6863192915916443,
      "learning_rate": 8.853359426956146e-06,
      "loss": 0.5127,
      "num_input_tokens_seen": 18757256,
      "step": 28665
    },
    {
      "epoch": 15.026205450733753,
      "grad_norm": 1.2085373401641846,
      "learning_rate": 8.844631598585482e-06,
      "loss": 0.4639,
      "num_input_tokens_seen": 18759528,
      "step": 28670
    },
    {
      "epoch": 15.028825995807129,
      "grad_norm": 0.4636845588684082,
      "learning_rate": 8.835907149733569e-06,
      "loss": 0.555,
      "num_input_tokens_seen": 18763016,
      "step": 28675
    },
    {
      "epoch": 15.031446540880504,
      "grad_norm": 0.3282536268234253,
      "learning_rate": 8.827186082225477e-06,
      "loss": 0.3962,
      "num_input_tokens_seen": 18766408,
      "step": 28680
    },
    {
      "epoch": 15.034067085953879,
      "grad_norm": 0.4274196922779083,
      "learning_rate": 8.81846839788554e-06,
      "loss": 0.4155,
      "num_input_tokens_seen": 18769064,
      "step": 28685
    },
    {
      "epoch": 15.036687631027254,
      "grad_norm": 0.3618665933609009,
      "learning_rate": 8.809754098537407e-06,
      "loss": 0.3667,
      "num_input_tokens_seen": 18771720,
      "step": 28690
    },
    {
      "epoch": 15.03930817610063,
      "grad_norm": 0.4647749960422516,
      "learning_rate": 8.801043186004016e-06,
      "loss": 0.4078,
      "num_input_tokens_seen": 18776232,
      "step": 28695
    },
    {
      "epoch": 15.041928721174004,
      "grad_norm": 0.3721223473548889,
      "learning_rate": 8.792335662107575e-06,
      "loss": 0.412,
      "num_input_tokens_seen": 18779272,
      "step": 28700
    },
    {
      "epoch": 15.04454926624738,
      "grad_norm": 0.47835829854011536,
      "learning_rate": 8.783631528669605e-06,
      "loss": 0.4123,
      "num_input_tokens_seen": 18781896,
      "step": 28705
    },
    {
      "epoch": 15.047169811320755,
      "grad_norm": 0.42208433151245117,
      "learning_rate": 8.774930787510924e-06,
      "loss": 0.3511,
      "num_input_tokens_seen": 18785384,
      "step": 28710
    },
    {
      "epoch": 15.04979035639413,
      "grad_norm": 0.8313445448875427,
      "learning_rate": 8.766233440451608e-06,
      "loss": 0.3359,
      "num_input_tokens_seen": 18787816,
      "step": 28715
    },
    {
      "epoch": 15.052410901467505,
      "grad_norm": 0.8196111917495728,
      "learning_rate": 8.75753948931105e-06,
      "loss": 0.4558,
      "num_input_tokens_seen": 18791400,
      "step": 28720
    },
    {
      "epoch": 15.05503144654088,
      "grad_norm": 0.8316233158111572,
      "learning_rate": 8.748848935907936e-06,
      "loss": 0.4857,
      "num_input_tokens_seen": 18794728,
      "step": 28725
    },
    {
      "epoch": 15.057651991614255,
      "grad_norm": 0.660615861415863,
      "learning_rate": 8.740161782060214e-06,
      "loss": 0.5508,
      "num_input_tokens_seen": 18798248,
      "step": 28730
    },
    {
      "epoch": 15.06027253668763,
      "grad_norm": 0.36567384004592896,
      "learning_rate": 8.731478029585144e-06,
      "loss": 0.3911,
      "num_input_tokens_seen": 18801256,
      "step": 28735
    },
    {
      "epoch": 15.062893081761006,
      "grad_norm": 0.36139634251594543,
      "learning_rate": 8.722797680299278e-06,
      "loss": 0.3915,
      "num_input_tokens_seen": 18805960,
      "step": 28740
    },
    {
      "epoch": 15.065513626834381,
      "grad_norm": 0.32720229029655457,
      "learning_rate": 8.71412073601843e-06,
      "loss": 0.4802,
      "num_input_tokens_seen": 18809160,
      "step": 28745
    },
    {
      "epoch": 15.068134171907756,
      "grad_norm": 0.9005676507949829,
      "learning_rate": 8.705447198557731e-06,
      "loss": 0.4669,
      "num_input_tokens_seen": 18811944,
      "step": 28750
    },
    {
      "epoch": 15.070754716981131,
      "grad_norm": 0.6830553412437439,
      "learning_rate": 8.696777069731574e-06,
      "loss": 0.3864,
      "num_input_tokens_seen": 18815304,
      "step": 28755
    },
    {
      "epoch": 15.073375262054507,
      "grad_norm": 0.3063199520111084,
      "learning_rate": 8.688110351353654e-06,
      "loss": 0.3837,
      "num_input_tokens_seen": 18818344,
      "step": 28760
    },
    {
      "epoch": 15.075995807127883,
      "grad_norm": 0.5338236689567566,
      "learning_rate": 8.679447045236962e-06,
      "loss": 0.414,
      "num_input_tokens_seen": 18820616,
      "step": 28765
    },
    {
      "epoch": 15.078616352201259,
      "grad_norm": 0.44694581627845764,
      "learning_rate": 8.670787153193746e-06,
      "loss": 0.427,
      "num_input_tokens_seen": 18823624,
      "step": 28770
    },
    {
      "epoch": 15.081236897274634,
      "grad_norm": 0.48577672243118286,
      "learning_rate": 8.662130677035574e-06,
      "loss": 0.4719,
      "num_input_tokens_seen": 18826376,
      "step": 28775
    },
    {
      "epoch": 15.083857442348009,
      "grad_norm": 0.6919603943824768,
      "learning_rate": 8.65347761857326e-06,
      "loss": 0.3985,
      "num_input_tokens_seen": 18829000,
      "step": 28780
    },
    {
      "epoch": 15.086477987421384,
      "grad_norm": 0.37203505635261536,
      "learning_rate": 8.64482797961694e-06,
      "loss": 0.4107,
      "num_input_tokens_seen": 18831848,
      "step": 28785
    },
    {
      "epoch": 15.08909853249476,
      "grad_norm": 0.2823689877986908,
      "learning_rate": 8.636181761976016e-06,
      "loss": 0.5063,
      "num_input_tokens_seen": 18835688,
      "step": 28790
    },
    {
      "epoch": 15.091719077568134,
      "grad_norm": 0.48249056935310364,
      "learning_rate": 8.62753896745919e-06,
      "loss": 0.4605,
      "num_input_tokens_seen": 18838344,
      "step": 28795
    },
    {
      "epoch": 15.09433962264151,
      "grad_norm": 0.28911128640174866,
      "learning_rate": 8.618899597874413e-06,
      "loss": 0.3611,
      "num_input_tokens_seen": 18841512,
      "step": 28800
    },
    {
      "epoch": 15.096960167714885,
      "grad_norm": 0.4706289768218994,
      "learning_rate": 8.610263655028964e-06,
      "loss": 0.5624,
      "num_input_tokens_seen": 18845160,
      "step": 28805
    },
    {
      "epoch": 15.09958071278826,
      "grad_norm": 0.8172373175621033,
      "learning_rate": 8.601631140729366e-06,
      "loss": 0.3672,
      "num_input_tokens_seen": 18847848,
      "step": 28810
    },
    {
      "epoch": 15.102201257861635,
      "grad_norm": 0.5651489496231079,
      "learning_rate": 8.593002056781451e-06,
      "loss": 0.4361,
      "num_input_tokens_seen": 18851592,
      "step": 28815
    },
    {
      "epoch": 15.10482180293501,
      "grad_norm": 0.36283013224601746,
      "learning_rate": 8.584376404990326e-06,
      "loss": 0.3901,
      "num_input_tokens_seen": 18854568,
      "step": 28820
    },
    {
      "epoch": 15.107442348008385,
      "grad_norm": 0.5847052335739136,
      "learning_rate": 8.575754187160384e-06,
      "loss": 0.5114,
      "num_input_tokens_seen": 18858216,
      "step": 28825
    },
    {
      "epoch": 15.11006289308176,
      "grad_norm": 0.5791652202606201,
      "learning_rate": 8.567135405095291e-06,
      "loss": 0.5783,
      "num_input_tokens_seen": 18861192,
      "step": 28830
    },
    {
      "epoch": 15.112683438155136,
      "grad_norm": 0.6452985405921936,
      "learning_rate": 8.558520060597985e-06,
      "loss": 0.3774,
      "num_input_tokens_seen": 18865256,
      "step": 28835
    },
    {
      "epoch": 15.115303983228511,
      "grad_norm": 0.6408486366271973,
      "learning_rate": 8.54990815547071e-06,
      "loss": 0.4648,
      "num_input_tokens_seen": 18867912,
      "step": 28840
    },
    {
      "epoch": 15.117924528301886,
      "grad_norm": 0.9074154496192932,
      "learning_rate": 8.541299691514974e-06,
      "loss": 0.388,
      "num_input_tokens_seen": 18870792,
      "step": 28845
    },
    {
      "epoch": 15.120545073375261,
      "grad_norm": 0.5249502658843994,
      "learning_rate": 8.53269467053158e-06,
      "loss": 0.4978,
      "num_input_tokens_seen": 18873480,
      "step": 28850
    },
    {
      "epoch": 15.123165618448636,
      "grad_norm": 0.6273297667503357,
      "learning_rate": 8.524093094320593e-06,
      "loss": 0.5147,
      "num_input_tokens_seen": 18876808,
      "step": 28855
    },
    {
      "epoch": 15.125786163522013,
      "grad_norm": 0.3866637945175171,
      "learning_rate": 8.515494964681357e-06,
      "loss": 0.5449,
      "num_input_tokens_seen": 18879432,
      "step": 28860
    },
    {
      "epoch": 15.128406708595389,
      "grad_norm": 0.5201178193092346,
      "learning_rate": 8.506900283412506e-06,
      "loss": 0.4934,
      "num_input_tokens_seen": 18882344,
      "step": 28865
    },
    {
      "epoch": 15.131027253668764,
      "grad_norm": 0.35896775126457214,
      "learning_rate": 8.498309052311953e-06,
      "loss": 0.3093,
      "num_input_tokens_seen": 18885992,
      "step": 28870
    },
    {
      "epoch": 15.133647798742139,
      "grad_norm": 0.6191550493240356,
      "learning_rate": 8.489721273176887e-06,
      "loss": 0.3144,
      "num_input_tokens_seen": 18889256,
      "step": 28875
    },
    {
      "epoch": 15.136268343815514,
      "grad_norm": 0.4829705059528351,
      "learning_rate": 8.481136947803777e-06,
      "loss": 0.3949,
      "num_input_tokens_seen": 18892552,
      "step": 28880
    },
    {
      "epoch": 15.13888888888889,
      "grad_norm": 0.5582820773124695,
      "learning_rate": 8.472556077988362e-06,
      "loss": 0.3834,
      "num_input_tokens_seen": 18896168,
      "step": 28885
    },
    {
      "epoch": 15.141509433962264,
      "grad_norm": 0.41305187344551086,
      "learning_rate": 8.46397866552565e-06,
      "loss": 0.3142,
      "num_input_tokens_seen": 18899560,
      "step": 28890
    },
    {
      "epoch": 15.14412997903564,
      "grad_norm": 0.6146633625030518,
      "learning_rate": 8.455404712209952e-06,
      "loss": 0.4252,
      "num_input_tokens_seen": 18903336,
      "step": 28895
    },
    {
      "epoch": 15.146750524109015,
      "grad_norm": 0.33356958627700806,
      "learning_rate": 8.446834219834836e-06,
      "loss": 0.4544,
      "num_input_tokens_seen": 18906696,
      "step": 28900
    },
    {
      "epoch": 15.14937106918239,
      "grad_norm": 0.40584126114845276,
      "learning_rate": 8.43826719019316e-06,
      "loss": 0.4796,
      "num_input_tokens_seen": 18909992,
      "step": 28905
    },
    {
      "epoch": 15.151991614255765,
      "grad_norm": 0.4064614176750183,
      "learning_rate": 8.429703625077043e-06,
      "loss": 0.4359,
      "num_input_tokens_seen": 18913352,
      "step": 28910
    },
    {
      "epoch": 15.15461215932914,
      "grad_norm": 0.7365368008613586,
      "learning_rate": 8.421143526277875e-06,
      "loss": 0.4429,
      "num_input_tokens_seen": 18917032,
      "step": 28915
    },
    {
      "epoch": 15.157232704402515,
      "grad_norm": 0.6038554310798645,
      "learning_rate": 8.41258689558634e-06,
      "loss": 0.4354,
      "num_input_tokens_seen": 18919912,
      "step": 28920
    },
    {
      "epoch": 15.15985324947589,
      "grad_norm": 0.6527988314628601,
      "learning_rate": 8.404033734792386e-06,
      "loss": 0.4937,
      "num_input_tokens_seen": 18923656,
      "step": 28925
    },
    {
      "epoch": 15.162473794549266,
      "grad_norm": 0.7076516151428223,
      "learning_rate": 8.395484045685248e-06,
      "loss": 0.2674,
      "num_input_tokens_seen": 18926344,
      "step": 28930
    },
    {
      "epoch": 15.165094339622641,
      "grad_norm": 0.4861942231655121,
      "learning_rate": 8.386937830053412e-06,
      "loss": 0.4244,
      "num_input_tokens_seen": 18929640,
      "step": 28935
    },
    {
      "epoch": 15.167714884696016,
      "grad_norm": 0.6888955235481262,
      "learning_rate": 8.37839508968464e-06,
      "loss": 0.4678,
      "num_input_tokens_seen": 18932648,
      "step": 28940
    },
    {
      "epoch": 15.170335429769391,
      "grad_norm": 1.2582857608795166,
      "learning_rate": 8.369855826365988e-06,
      "loss": 0.6669,
      "num_input_tokens_seen": 18936456,
      "step": 28945
    },
    {
      "epoch": 15.172955974842766,
      "grad_norm": 0.45682188868522644,
      "learning_rate": 8.361320041883772e-06,
      "loss": 0.3332,
      "num_input_tokens_seen": 18938984,
      "step": 28950
    },
    {
      "epoch": 15.175576519916143,
      "grad_norm": 0.4979459047317505,
      "learning_rate": 8.352787738023576e-06,
      "loss": 0.3497,
      "num_input_tokens_seen": 18941896,
      "step": 28955
    },
    {
      "epoch": 15.178197064989519,
      "grad_norm": 0.3403055965900421,
      "learning_rate": 8.344258916570274e-06,
      "loss": 0.4574,
      "num_input_tokens_seen": 18945704,
      "step": 28960
    },
    {
      "epoch": 15.180817610062894,
      "grad_norm": 0.6156893372535706,
      "learning_rate": 8.335733579307988e-06,
      "loss": 0.2811,
      "num_input_tokens_seen": 18948392,
      "step": 28965
    },
    {
      "epoch": 15.183438155136269,
      "grad_norm": 0.2863057255744934,
      "learning_rate": 8.327211728020113e-06,
      "loss": 0.4021,
      "num_input_tokens_seen": 18952264,
      "step": 28970
    },
    {
      "epoch": 15.186058700209644,
      "grad_norm": 0.5481739640235901,
      "learning_rate": 8.31869336448933e-06,
      "loss": 0.4915,
      "num_input_tokens_seen": 18955144,
      "step": 28975
    },
    {
      "epoch": 15.18867924528302,
      "grad_norm": 0.6486265659332275,
      "learning_rate": 8.310178490497586e-06,
      "loss": 0.4528,
      "num_input_tokens_seen": 18957992,
      "step": 28980
    },
    {
      "epoch": 15.191299790356394,
      "grad_norm": 0.5300329327583313,
      "learning_rate": 8.301667107826103e-06,
      "loss": 0.4086,
      "num_input_tokens_seen": 18960552,
      "step": 28985
    },
    {
      "epoch": 15.19392033542977,
      "grad_norm": 0.5572439432144165,
      "learning_rate": 8.293159218255345e-06,
      "loss": 0.4093,
      "num_input_tokens_seen": 18963816,
      "step": 28990
    },
    {
      "epoch": 15.196540880503145,
      "grad_norm": 0.9347187280654907,
      "learning_rate": 8.284654823565088e-06,
      "loss": 0.5184,
      "num_input_tokens_seen": 18966088,
      "step": 28995
    },
    {
      "epoch": 15.19916142557652,
      "grad_norm": 0.7017068862915039,
      "learning_rate": 8.276153925534333e-06,
      "loss": 0.3796,
      "num_input_tokens_seen": 18969544,
      "step": 29000
    },
    {
      "epoch": 15.201781970649895,
      "grad_norm": 0.39359110593795776,
      "learning_rate": 8.267656525941383e-06,
      "loss": 0.4975,
      "num_input_tokens_seen": 18972840,
      "step": 29005
    },
    {
      "epoch": 15.20440251572327,
      "grad_norm": 0.48914700746536255,
      "learning_rate": 8.259162626563801e-06,
      "loss": 0.5632,
      "num_input_tokens_seen": 18976712,
      "step": 29010
    },
    {
      "epoch": 15.207023060796645,
      "grad_norm": 1.2477376461029053,
      "learning_rate": 8.250672229178402e-06,
      "loss": 0.4431,
      "num_input_tokens_seen": 18979336,
      "step": 29015
    },
    {
      "epoch": 15.20964360587002,
      "grad_norm": 0.5238872170448303,
      "learning_rate": 8.242185335561295e-06,
      "loss": 0.4484,
      "num_input_tokens_seen": 18982664,
      "step": 29020
    },
    {
      "epoch": 15.212264150943396,
      "grad_norm": 0.4850461483001709,
      "learning_rate": 8.23370194748782e-06,
      "loss": 0.4549,
      "num_input_tokens_seen": 18985832,
      "step": 29025
    },
    {
      "epoch": 15.214884696016771,
      "grad_norm": 0.39078837633132935,
      "learning_rate": 8.22522206673262e-06,
      "loss": 0.5607,
      "num_input_tokens_seen": 18988680,
      "step": 29030
    },
    {
      "epoch": 15.217505241090146,
      "grad_norm": 0.4083174467086792,
      "learning_rate": 8.216745695069589e-06,
      "loss": 0.4345,
      "num_input_tokens_seen": 18991592,
      "step": 29035
    },
    {
      "epoch": 15.220125786163521,
      "grad_norm": 0.6231839060783386,
      "learning_rate": 8.208272834271894e-06,
      "loss": 0.431,
      "num_input_tokens_seen": 18994536,
      "step": 29040
    },
    {
      "epoch": 15.222746331236896,
      "grad_norm": 0.42693275213241577,
      "learning_rate": 8.19980348611194e-06,
      "loss": 0.3569,
      "num_input_tokens_seen": 18998312,
      "step": 29045
    },
    {
      "epoch": 15.225366876310272,
      "grad_norm": 0.4343807101249695,
      "learning_rate": 8.191337652361439e-06,
      "loss": 0.4777,
      "num_input_tokens_seen": 19001416,
      "step": 29050
    },
    {
      "epoch": 15.227987421383649,
      "grad_norm": 0.3435872793197632,
      "learning_rate": 8.182875334791332e-06,
      "loss": 0.4815,
      "num_input_tokens_seen": 19004648,
      "step": 29055
    },
    {
      "epoch": 15.230607966457024,
      "grad_norm": 0.5239810943603516,
      "learning_rate": 8.174416535171841e-06,
      "loss": 0.4225,
      "num_input_tokens_seen": 19008360,
      "step": 29060
    },
    {
      "epoch": 15.233228511530399,
      "grad_norm": 0.46506673097610474,
      "learning_rate": 8.165961255272467e-06,
      "loss": 0.4823,
      "num_input_tokens_seen": 19011368,
      "step": 29065
    },
    {
      "epoch": 15.235849056603774,
      "grad_norm": 0.5418808460235596,
      "learning_rate": 8.157509496861935e-06,
      "loss": 0.4111,
      "num_input_tokens_seen": 19015368,
      "step": 29070
    },
    {
      "epoch": 15.23846960167715,
      "grad_norm": 0.357223778963089,
      "learning_rate": 8.149061261708266e-06,
      "loss": 0.4728,
      "num_input_tokens_seen": 19019240,
      "step": 29075
    },
    {
      "epoch": 15.241090146750524,
      "grad_norm": 0.5580798983573914,
      "learning_rate": 8.140616551578745e-06,
      "loss": 0.6065,
      "num_input_tokens_seen": 19022152,
      "step": 29080
    },
    {
      "epoch": 15.2437106918239,
      "grad_norm": 0.6536417603492737,
      "learning_rate": 8.13217536823989e-06,
      "loss": 0.4257,
      "num_input_tokens_seen": 19025448,
      "step": 29085
    },
    {
      "epoch": 15.246331236897275,
      "grad_norm": 0.5674160718917847,
      "learning_rate": 8.12373771345752e-06,
      "loss": 0.3818,
      "num_input_tokens_seen": 19029288,
      "step": 29090
    },
    {
      "epoch": 15.24895178197065,
      "grad_norm": 0.5828611850738525,
      "learning_rate": 8.115303588996676e-06,
      "loss": 0.4038,
      "num_input_tokens_seen": 19032136,
      "step": 29095
    },
    {
      "epoch": 15.251572327044025,
      "grad_norm": 0.2918580174446106,
      "learning_rate": 8.10687299662169e-06,
      "loss": 0.4748,
      "num_input_tokens_seen": 19035592,
      "step": 29100
    },
    {
      "epoch": 15.2541928721174,
      "grad_norm": 0.43119654059410095,
      "learning_rate": 8.098445938096147e-06,
      "loss": 0.4171,
      "num_input_tokens_seen": 19039784,
      "step": 29105
    },
    {
      "epoch": 15.256813417190775,
      "grad_norm": 0.4073396623134613,
      "learning_rate": 8.090022415182898e-06,
      "loss": 0.4095,
      "num_input_tokens_seen": 19042856,
      "step": 29110
    },
    {
      "epoch": 15.25943396226415,
      "grad_norm": 0.534350574016571,
      "learning_rate": 8.081602429644033e-06,
      "loss": 0.3735,
      "num_input_tokens_seen": 19046152,
      "step": 29115
    },
    {
      "epoch": 15.262054507337526,
      "grad_norm": 0.45647335052490234,
      "learning_rate": 8.073185983240932e-06,
      "loss": 0.3984,
      "num_input_tokens_seen": 19049416,
      "step": 29120
    },
    {
      "epoch": 15.264675052410901,
      "grad_norm": 0.40708127617836,
      "learning_rate": 8.064773077734206e-06,
      "loss": 0.5439,
      "num_input_tokens_seen": 19052680,
      "step": 29125
    },
    {
      "epoch": 15.267295597484276,
      "grad_norm": 0.71822589635849,
      "learning_rate": 8.05636371488374e-06,
      "loss": 0.4067,
      "num_input_tokens_seen": 19055560,
      "step": 29130
    },
    {
      "epoch": 15.269916142557651,
      "grad_norm": 0.4290073812007904,
      "learning_rate": 8.047957896448696e-06,
      "loss": 0.4718,
      "num_input_tokens_seen": 19058696,
      "step": 29135
    },
    {
      "epoch": 15.272536687631026,
      "grad_norm": 0.5183464884757996,
      "learning_rate": 8.039555624187451e-06,
      "loss": 0.5439,
      "num_input_tokens_seen": 19062248,
      "step": 29140
    },
    {
      "epoch": 15.275157232704402,
      "grad_norm": 0.4597265124320984,
      "learning_rate": 8.031156899857681e-06,
      "loss": 0.4886,
      "num_input_tokens_seen": 19065096,
      "step": 29145
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 0.5573505163192749,
      "learning_rate": 8.022761725216288e-06,
      "loss": 0.3584,
      "num_input_tokens_seen": 19068392,
      "step": 29150
    },
    {
      "epoch": 15.280398322851154,
      "grad_norm": 0.6471738815307617,
      "learning_rate": 8.014370102019456e-06,
      "loss": 0.4716,
      "num_input_tokens_seen": 19070888,
      "step": 29155
    },
    {
      "epoch": 15.283018867924529,
      "grad_norm": 0.3201558291912079,
      "learning_rate": 8.005982032022616e-06,
      "loss": 0.4507,
      "num_input_tokens_seen": 19074248,
      "step": 29160
    },
    {
      "epoch": 15.285639412997904,
      "grad_norm": 0.936698853969574,
      "learning_rate": 7.997597516980467e-06,
      "loss": 0.4935,
      "num_input_tokens_seen": 19077736,
      "step": 29165
    },
    {
      "epoch": 15.28825995807128,
      "grad_norm": 0.5306326150894165,
      "learning_rate": 7.989216558646942e-06,
      "loss": 0.4789,
      "num_input_tokens_seen": 19081640,
      "step": 29170
    },
    {
      "epoch": 15.290880503144654,
      "grad_norm": 0.40970224142074585,
      "learning_rate": 7.980839158775235e-06,
      "loss": 0.5256,
      "num_input_tokens_seen": 19085384,
      "step": 29175
    },
    {
      "epoch": 15.29350104821803,
      "grad_norm": 0.3504211902618408,
      "learning_rate": 7.972465319117814e-06,
      "loss": 0.3673,
      "num_input_tokens_seen": 19088488,
      "step": 29180
    },
    {
      "epoch": 15.296121593291405,
      "grad_norm": 0.7418977618217468,
      "learning_rate": 7.964095041426386e-06,
      "loss": 0.5511,
      "num_input_tokens_seen": 19092776,
      "step": 29185
    },
    {
      "epoch": 15.29874213836478,
      "grad_norm": 0.4038311243057251,
      "learning_rate": 7.95572832745192e-06,
      "loss": 0.3503,
      "num_input_tokens_seen": 19095560,
      "step": 29190
    },
    {
      "epoch": 15.301362683438155,
      "grad_norm": 0.5095683336257935,
      "learning_rate": 7.947365178944643e-06,
      "loss": 0.4979,
      "num_input_tokens_seen": 19098824,
      "step": 29195
    },
    {
      "epoch": 15.30398322851153,
      "grad_norm": 0.31809425354003906,
      "learning_rate": 7.939005597654025e-06,
      "loss": 0.4504,
      "num_input_tokens_seen": 19103208,
      "step": 29200
    },
    {
      "epoch": 15.306603773584905,
      "grad_norm": 0.5946217179298401,
      "learning_rate": 7.930649585328787e-06,
      "loss": 0.5198,
      "num_input_tokens_seen": 19106088,
      "step": 29205
    },
    {
      "epoch": 15.30922431865828,
      "grad_norm": 0.7314202785491943,
      "learning_rate": 7.922297143716918e-06,
      "loss": 0.5181,
      "num_input_tokens_seen": 19109256,
      "step": 29210
    },
    {
      "epoch": 15.311844863731656,
      "grad_norm": 0.8513268232345581,
      "learning_rate": 7.913948274565652e-06,
      "loss": 0.3232,
      "num_input_tokens_seen": 19112296,
      "step": 29215
    },
    {
      "epoch": 15.314465408805031,
      "grad_norm": 0.5117412209510803,
      "learning_rate": 7.905602979621491e-06,
      "loss": 0.4731,
      "num_input_tokens_seen": 19115816,
      "step": 29220
    },
    {
      "epoch": 15.317085953878406,
      "grad_norm": 0.21713417768478394,
      "learning_rate": 7.897261260630159e-06,
      "loss": 0.4912,
      "num_input_tokens_seen": 19118632,
      "step": 29225
    },
    {
      "epoch": 15.319706498951781,
      "grad_norm": 0.5677416324615479,
      "learning_rate": 7.888923119336647e-06,
      "loss": 0.3811,
      "num_input_tokens_seen": 19121608,
      "step": 29230
    },
    {
      "epoch": 15.322327044025156,
      "grad_norm": 0.4298165440559387,
      "learning_rate": 7.880588557485203e-06,
      "loss": 0.5107,
      "num_input_tokens_seen": 19124904,
      "step": 29235
    },
    {
      "epoch": 15.324947589098532,
      "grad_norm": 0.36987388134002686,
      "learning_rate": 7.872257576819325e-06,
      "loss": 0.4318,
      "num_input_tokens_seen": 19127944,
      "step": 29240
    },
    {
      "epoch": 15.327568134171909,
      "grad_norm": 0.7467027902603149,
      "learning_rate": 7.863930179081769e-06,
      "loss": 0.3505,
      "num_input_tokens_seen": 19134344,
      "step": 29245
    },
    {
      "epoch": 15.330188679245284,
      "grad_norm": 0.6750571727752686,
      "learning_rate": 7.855606366014517e-06,
      "loss": 0.3544,
      "num_input_tokens_seen": 19137192,
      "step": 29250
    },
    {
      "epoch": 15.332809224318659,
      "grad_norm": 1.1811891794204712,
      "learning_rate": 7.847286139358814e-06,
      "loss": 0.4361,
      "num_input_tokens_seen": 19139816,
      "step": 29255
    },
    {
      "epoch": 15.335429769392034,
      "grad_norm": 0.4664722681045532,
      "learning_rate": 7.838969500855162e-06,
      "loss": 0.4183,
      "num_input_tokens_seen": 19142504,
      "step": 29260
    },
    {
      "epoch": 15.33805031446541,
      "grad_norm": 0.26201802492141724,
      "learning_rate": 7.830656452243307e-06,
      "loss": 0.4171,
      "num_input_tokens_seen": 19146920,
      "step": 29265
    },
    {
      "epoch": 15.340670859538784,
      "grad_norm": 0.44496282935142517,
      "learning_rate": 7.822346995262241e-06,
      "loss": 0.371,
      "num_input_tokens_seen": 19149448,
      "step": 29270
    },
    {
      "epoch": 15.34329140461216,
      "grad_norm": 0.4924173355102539,
      "learning_rate": 7.814041131650223e-06,
      "loss": 0.4051,
      "num_input_tokens_seen": 19151848,
      "step": 29275
    },
    {
      "epoch": 15.345911949685535,
      "grad_norm": 0.3790093958377838,
      "learning_rate": 7.805738863144731e-06,
      "loss": 0.5023,
      "num_input_tokens_seen": 19154600,
      "step": 29280
    },
    {
      "epoch": 15.34853249475891,
      "grad_norm": 0.538650631904602,
      "learning_rate": 7.797440191482496e-06,
      "loss": 0.346,
      "num_input_tokens_seen": 19157128,
      "step": 29285
    },
    {
      "epoch": 15.351153039832285,
      "grad_norm": 0.44217967987060547,
      "learning_rate": 7.789145118399518e-06,
      "loss": 0.4232,
      "num_input_tokens_seen": 19160072,
      "step": 29290
    },
    {
      "epoch": 15.35377358490566,
      "grad_norm": 0.70816570520401,
      "learning_rate": 7.780853645631032e-06,
      "loss": 0.4005,
      "num_input_tokens_seen": 19163624,
      "step": 29295
    },
    {
      "epoch": 15.356394129979035,
      "grad_norm": 0.3106728196144104,
      "learning_rate": 7.772565774911522e-06,
      "loss": 0.5248,
      "num_input_tokens_seen": 19167432,
      "step": 29300
    },
    {
      "epoch": 15.35901467505241,
      "grad_norm": 0.7591602206230164,
      "learning_rate": 7.76428150797471e-06,
      "loss": 0.4504,
      "num_input_tokens_seen": 19170408,
      "step": 29305
    },
    {
      "epoch": 15.361635220125786,
      "grad_norm": 0.3867955505847931,
      "learning_rate": 7.756000846553562e-06,
      "loss": 0.4967,
      "num_input_tokens_seen": 19173384,
      "step": 29310
    },
    {
      "epoch": 15.364255765199161,
      "grad_norm": 0.4968152642250061,
      "learning_rate": 7.74772379238031e-06,
      "loss": 0.6522,
      "num_input_tokens_seen": 19176264,
      "step": 29315
    },
    {
      "epoch": 15.366876310272536,
      "grad_norm": 0.4756484925746918,
      "learning_rate": 7.739450347186417e-06,
      "loss": 0.5562,
      "num_input_tokens_seen": 19179304,
      "step": 29320
    },
    {
      "epoch": 15.369496855345911,
      "grad_norm": 0.3755505084991455,
      "learning_rate": 7.731180512702599e-06,
      "loss": 0.3834,
      "num_input_tokens_seen": 19182376,
      "step": 29325
    },
    {
      "epoch": 15.372117400419286,
      "grad_norm": 0.4986802935600281,
      "learning_rate": 7.722914290658795e-06,
      "loss": 0.4081,
      "num_input_tokens_seen": 19185224,
      "step": 29330
    },
    {
      "epoch": 15.374737945492662,
      "grad_norm": 0.5584988594055176,
      "learning_rate": 7.714651682784224e-06,
      "loss": 0.6041,
      "num_input_tokens_seen": 19188488,
      "step": 29335
    },
    {
      "epoch": 15.377358490566039,
      "grad_norm": 0.3657341003417969,
      "learning_rate": 7.706392690807313e-06,
      "loss": 0.5385,
      "num_input_tokens_seen": 19191912,
      "step": 29340
    },
    {
      "epoch": 15.379979035639414,
      "grad_norm": 0.33307531476020813,
      "learning_rate": 7.69813731645575e-06,
      "loss": 0.4574,
      "num_input_tokens_seen": 19195336,
      "step": 29345
    },
    {
      "epoch": 15.382599580712789,
      "grad_norm": 0.46804165840148926,
      "learning_rate": 7.689885561456475e-06,
      "loss": 0.3561,
      "num_input_tokens_seen": 19198184,
      "step": 29350
    },
    {
      "epoch": 15.385220125786164,
      "grad_norm": 0.5546596646308899,
      "learning_rate": 7.681637427535663e-06,
      "loss": 0.3733,
      "num_input_tokens_seen": 19201000,
      "step": 29355
    },
    {
      "epoch": 15.38784067085954,
      "grad_norm": 0.47311288118362427,
      "learning_rate": 7.673392916418715e-06,
      "loss": 0.3931,
      "num_input_tokens_seen": 19205032,
      "step": 29360
    },
    {
      "epoch": 15.390461215932914,
      "grad_norm": 0.4856659173965454,
      "learning_rate": 7.665152029830303e-06,
      "loss": 0.4248,
      "num_input_tokens_seen": 19207624,
      "step": 29365
    },
    {
      "epoch": 15.39308176100629,
      "grad_norm": 0.5711731910705566,
      "learning_rate": 7.656914769494314e-06,
      "loss": 0.5436,
      "num_input_tokens_seen": 19210984,
      "step": 29370
    },
    {
      "epoch": 15.395702306079665,
      "grad_norm": 0.4871707558631897,
      "learning_rate": 7.648681137133892e-06,
      "loss": 0.4469,
      "num_input_tokens_seen": 19213480,
      "step": 29375
    },
    {
      "epoch": 15.39832285115304,
      "grad_norm": 0.33873629570007324,
      "learning_rate": 7.640451134471432e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 19217640,
      "step": 29380
    },
    {
      "epoch": 15.400943396226415,
      "grad_norm": 0.5159749984741211,
      "learning_rate": 7.632224763228538e-06,
      "loss": 0.4421,
      "num_input_tokens_seen": 19220328,
      "step": 29385
    },
    {
      "epoch": 15.40356394129979,
      "grad_norm": 0.40622273087501526,
      "learning_rate": 7.6240020251260896e-06,
      "loss": 0.3735,
      "num_input_tokens_seen": 19224232,
      "step": 29390
    },
    {
      "epoch": 15.406184486373165,
      "grad_norm": 0.9487040638923645,
      "learning_rate": 7.615782921884174e-06,
      "loss": 0.4782,
      "num_input_tokens_seen": 19226696,
      "step": 29395
    },
    {
      "epoch": 15.40880503144654,
      "grad_norm": 0.3831096887588501,
      "learning_rate": 7.607567455222142e-06,
      "loss": 0.568,
      "num_input_tokens_seen": 19229608,
      "step": 29400
    },
    {
      "epoch": 15.411425576519916,
      "grad_norm": 0.47641631960868835,
      "learning_rate": 7.599355626858582e-06,
      "loss": 0.4279,
      "num_input_tokens_seen": 19232488,
      "step": 29405
    },
    {
      "epoch": 15.414046121593291,
      "grad_norm": 0.41295579075813293,
      "learning_rate": 7.591147438511298e-06,
      "loss": 0.3942,
      "num_input_tokens_seen": 19236424,
      "step": 29410
    },
    {
      "epoch": 15.416666666666666,
      "grad_norm": 0.4863453209400177,
      "learning_rate": 7.5829428918973625e-06,
      "loss": 0.539,
      "num_input_tokens_seen": 19239592,
      "step": 29415
    },
    {
      "epoch": 15.419287211740041,
      "grad_norm": 0.38971173763275146,
      "learning_rate": 7.574741988733075e-06,
      "loss": 0.4266,
      "num_input_tokens_seen": 19242312,
      "step": 29420
    },
    {
      "epoch": 15.421907756813416,
      "grad_norm": 0.5018089413642883,
      "learning_rate": 7.56654473073396e-06,
      "loss": 0.4182,
      "num_input_tokens_seen": 19245256,
      "step": 29425
    },
    {
      "epoch": 15.424528301886792,
      "grad_norm": 0.3274651765823364,
      "learning_rate": 7.558351119614796e-06,
      "loss": 0.3801,
      "num_input_tokens_seen": 19248296,
      "step": 29430
    },
    {
      "epoch": 15.427148846960169,
      "grad_norm": 0.5707536935806274,
      "learning_rate": 7.5501611570896e-06,
      "loss": 0.4905,
      "num_input_tokens_seen": 19252264,
      "step": 29435
    },
    {
      "epoch": 15.429769392033544,
      "grad_norm": 1.1516400575637817,
      "learning_rate": 7.541974844871602e-06,
      "loss": 0.4906,
      "num_input_tokens_seen": 19255016,
      "step": 29440
    },
    {
      "epoch": 15.432389937106919,
      "grad_norm": 0.3343196511268616,
      "learning_rate": 7.5337921846732965e-06,
      "loss": 0.5001,
      "num_input_tokens_seen": 19259624,
      "step": 29445
    },
    {
      "epoch": 15.435010482180294,
      "grad_norm": 0.3513895571231842,
      "learning_rate": 7.525613178206409e-06,
      "loss": 0.4953,
      "num_input_tokens_seen": 19263336,
      "step": 29450
    },
    {
      "epoch": 15.43763102725367,
      "grad_norm": 0.268731951713562,
      "learning_rate": 7.517437827181878e-06,
      "loss": 0.3962,
      "num_input_tokens_seen": 19267688,
      "step": 29455
    },
    {
      "epoch": 15.440251572327044,
      "grad_norm": 0.43688175082206726,
      "learning_rate": 7.509266133309908e-06,
      "loss": 0.385,
      "num_input_tokens_seen": 19270408,
      "step": 29460
    },
    {
      "epoch": 15.44287211740042,
      "grad_norm": 0.7171322107315063,
      "learning_rate": 7.501098098299914e-06,
      "loss": 0.3521,
      "num_input_tokens_seen": 19273672,
      "step": 29465
    },
    {
      "epoch": 15.445492662473795,
      "grad_norm": 0.5599861145019531,
      "learning_rate": 7.492933723860557e-06,
      "loss": 0.3989,
      "num_input_tokens_seen": 19276744,
      "step": 29470
    },
    {
      "epoch": 15.44811320754717,
      "grad_norm": 0.45092085003852844,
      "learning_rate": 7.4847730116997334e-06,
      "loss": 0.4477,
      "num_input_tokens_seen": 19279880,
      "step": 29475
    },
    {
      "epoch": 15.450733752620545,
      "grad_norm": 0.7838711738586426,
      "learning_rate": 7.476615963524583e-06,
      "loss": 0.3761,
      "num_input_tokens_seen": 19282376,
      "step": 29480
    },
    {
      "epoch": 15.45335429769392,
      "grad_norm": 0.6544823050498962,
      "learning_rate": 7.468462581041452e-06,
      "loss": 0.5497,
      "num_input_tokens_seen": 19285000,
      "step": 29485
    },
    {
      "epoch": 15.455974842767295,
      "grad_norm": 0.511649489402771,
      "learning_rate": 7.460312865955932e-06,
      "loss": 0.5595,
      "num_input_tokens_seen": 19288264,
      "step": 29490
    },
    {
      "epoch": 15.45859538784067,
      "grad_norm": 0.28606459498405457,
      "learning_rate": 7.4521668199728584e-06,
      "loss": 0.4189,
      "num_input_tokens_seen": 19291176,
      "step": 29495
    },
    {
      "epoch": 15.461215932914046,
      "grad_norm": 0.49752187728881836,
      "learning_rate": 7.4440244447962884e-06,
      "loss": 0.4522,
      "num_input_tokens_seen": 19294408,
      "step": 29500
    },
    {
      "epoch": 15.463836477987421,
      "grad_norm": 0.5317540764808655,
      "learning_rate": 7.435885742129523e-06,
      "loss": 0.4,
      "num_input_tokens_seen": 19297512,
      "step": 29505
    },
    {
      "epoch": 15.466457023060796,
      "grad_norm": 0.4169999659061432,
      "learning_rate": 7.427750713675071e-06,
      "loss": 0.5103,
      "num_input_tokens_seen": 19300456,
      "step": 29510
    },
    {
      "epoch": 15.469077568134171,
      "grad_norm": 0.6642100214958191,
      "learning_rate": 7.419619361134702e-06,
      "loss": 0.509,
      "num_input_tokens_seen": 19303144,
      "step": 29515
    },
    {
      "epoch": 15.471698113207546,
      "grad_norm": 0.4884651303291321,
      "learning_rate": 7.411491686209387e-06,
      "loss": 0.3376,
      "num_input_tokens_seen": 19306344,
      "step": 29520
    },
    {
      "epoch": 15.474318658280922,
      "grad_norm": 0.32126909494400024,
      "learning_rate": 7.4033676905993525e-06,
      "loss": 0.4982,
      "num_input_tokens_seen": 19309736,
      "step": 29525
    },
    {
      "epoch": 15.476939203354299,
      "grad_norm": 1.1360828876495361,
      "learning_rate": 7.3952473760040444e-06,
      "loss": 0.4726,
      "num_input_tokens_seen": 19312264,
      "step": 29530
    },
    {
      "epoch": 15.479559748427674,
      "grad_norm": 0.7804445028305054,
      "learning_rate": 7.387130744122148e-06,
      "loss": 0.4812,
      "num_input_tokens_seen": 19315720,
      "step": 29535
    },
    {
      "epoch": 15.482180293501049,
      "grad_norm": 0.4686402380466461,
      "learning_rate": 7.379017796651558e-06,
      "loss": 0.4203,
      "num_input_tokens_seen": 19318984,
      "step": 29540
    },
    {
      "epoch": 15.484800838574424,
      "grad_norm": 0.49463340640068054,
      "learning_rate": 7.3709085352894085e-06,
      "loss": 0.3599,
      "num_input_tokens_seen": 19323624,
      "step": 29545
    },
    {
      "epoch": 15.4874213836478,
      "grad_norm": 0.671160876750946,
      "learning_rate": 7.362802961732071e-06,
      "loss": 0.593,
      "num_input_tokens_seen": 19326056,
      "step": 29550
    },
    {
      "epoch": 15.490041928721174,
      "grad_norm": 0.4114731550216675,
      "learning_rate": 7.354701077675136e-06,
      "loss": 0.5151,
      "num_input_tokens_seen": 19329352,
      "step": 29555
    },
    {
      "epoch": 15.49266247379455,
      "grad_norm": 0.703932523727417,
      "learning_rate": 7.346602884813439e-06,
      "loss": 0.474,
      "num_input_tokens_seen": 19332104,
      "step": 29560
    },
    {
      "epoch": 15.495283018867925,
      "grad_norm": 0.749728798866272,
      "learning_rate": 7.338508384841014e-06,
      "loss": 0.5246,
      "num_input_tokens_seen": 19335752,
      "step": 29565
    },
    {
      "epoch": 15.4979035639413,
      "grad_norm": 0.9017061591148376,
      "learning_rate": 7.3304175794511345e-06,
      "loss": 0.5058,
      "num_input_tokens_seen": 19338952,
      "step": 29570
    },
    {
      "epoch": 15.500524109014675,
      "grad_norm": 0.343398779630661,
      "learning_rate": 7.3223304703363135e-06,
      "loss": 0.3921,
      "num_input_tokens_seen": 19344008,
      "step": 29575
    },
    {
      "epoch": 15.50314465408805,
      "grad_norm": 0.5290107131004333,
      "learning_rate": 7.314247059188281e-06,
      "loss": 0.4716,
      "num_input_tokens_seen": 19347368,
      "step": 29580
    },
    {
      "epoch": 15.505765199161425,
      "grad_norm": 0.692886471748352,
      "learning_rate": 7.306167347697992e-06,
      "loss": 0.4232,
      "num_input_tokens_seen": 19351656,
      "step": 29585
    },
    {
      "epoch": 15.5083857442348,
      "grad_norm": 0.5158631205558777,
      "learning_rate": 7.29809133755564e-06,
      "loss": 0.4434,
      "num_input_tokens_seen": 19354664,
      "step": 29590
    },
    {
      "epoch": 15.511006289308176,
      "grad_norm": 0.3227289617061615,
      "learning_rate": 7.290019030450629e-06,
      "loss": 0.4912,
      "num_input_tokens_seen": 19358376,
      "step": 29595
    },
    {
      "epoch": 15.51362683438155,
      "grad_norm": 0.4156606197357178,
      "learning_rate": 7.2819504280715794e-06,
      "loss": 0.3242,
      "num_input_tokens_seen": 19361736,
      "step": 29600
    },
    {
      "epoch": 15.516247379454926,
      "grad_norm": 0.37544432282447815,
      "learning_rate": 7.273885532106364e-06,
      "loss": 0.6456,
      "num_input_tokens_seen": 19364168,
      "step": 29605
    },
    {
      "epoch": 15.518867924528301,
      "grad_norm": 0.4304060637950897,
      "learning_rate": 7.265824344242064e-06,
      "loss": 0.5329,
      "num_input_tokens_seen": 19367624,
      "step": 29610
    },
    {
      "epoch": 15.521488469601676,
      "grad_norm": 0.3854959309101105,
      "learning_rate": 7.257766866165e-06,
      "loss": 0.4351,
      "num_input_tokens_seen": 19371080,
      "step": 29615
    },
    {
      "epoch": 15.524109014675052,
      "grad_norm": 0.5711207389831543,
      "learning_rate": 7.24971309956069e-06,
      "loss": 0.5626,
      "num_input_tokens_seen": 19373672,
      "step": 29620
    },
    {
      "epoch": 15.526729559748428,
      "grad_norm": 0.4069248139858246,
      "learning_rate": 7.241663046113887e-06,
      "loss": 0.3652,
      "num_input_tokens_seen": 19376808,
      "step": 29625
    },
    {
      "epoch": 15.529350104821804,
      "grad_norm": 0.40129995346069336,
      "learning_rate": 7.233616707508576e-06,
      "loss": 0.4708,
      "num_input_tokens_seen": 19379656,
      "step": 29630
    },
    {
      "epoch": 15.531970649895179,
      "grad_norm": 0.44094452261924744,
      "learning_rate": 7.225574085427961e-06,
      "loss": 0.3128,
      "num_input_tokens_seen": 19382120,
      "step": 29635
    },
    {
      "epoch": 15.534591194968554,
      "grad_norm": 0.3125288188457489,
      "learning_rate": 7.217535181554474e-06,
      "loss": 0.3095,
      "num_input_tokens_seen": 19388232,
      "step": 29640
    },
    {
      "epoch": 15.53721174004193,
      "grad_norm": 0.48194044828414917,
      "learning_rate": 7.209499997569747e-06,
      "loss": 0.4184,
      "num_input_tokens_seen": 19392648,
      "step": 29645
    },
    {
      "epoch": 15.539832285115304,
      "grad_norm": 0.28783008456230164,
      "learning_rate": 7.201468535154663e-06,
      "loss": 0.4441,
      "num_input_tokens_seen": 19395880,
      "step": 29650
    },
    {
      "epoch": 15.54245283018868,
      "grad_norm": 0.3620312213897705,
      "learning_rate": 7.193440795989295e-06,
      "loss": 0.5004,
      "num_input_tokens_seen": 19398824,
      "step": 29655
    },
    {
      "epoch": 15.545073375262055,
      "grad_norm": 0.7221571207046509,
      "learning_rate": 7.185416781752965e-06,
      "loss": 0.3931,
      "num_input_tokens_seen": 19402344,
      "step": 29660
    },
    {
      "epoch": 15.54769392033543,
      "grad_norm": 0.4233047366142273,
      "learning_rate": 7.177396494124206e-06,
      "loss": 0.4965,
      "num_input_tokens_seen": 19405800,
      "step": 29665
    },
    {
      "epoch": 15.550314465408805,
      "grad_norm": 0.806205153465271,
      "learning_rate": 7.169379934780779e-06,
      "loss": 0.4179,
      "num_input_tokens_seen": 19410472,
      "step": 29670
    },
    {
      "epoch": 15.55293501048218,
      "grad_norm": 0.38926202058792114,
      "learning_rate": 7.161367105399644e-06,
      "loss": 0.3477,
      "num_input_tokens_seen": 19413448,
      "step": 29675
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 0.8515443205833435,
      "learning_rate": 7.153358007656991e-06,
      "loss": 0.4321,
      "num_input_tokens_seen": 19416104,
      "step": 29680
    },
    {
      "epoch": 15.55817610062893,
      "grad_norm": 0.3806880712509155,
      "learning_rate": 7.145352643228237e-06,
      "loss": 0.3157,
      "num_input_tokens_seen": 19418536,
      "step": 29685
    },
    {
      "epoch": 15.560796645702306,
      "grad_norm": 0.4125250279903412,
      "learning_rate": 7.137351013788013e-06,
      "loss": 0.3118,
      "num_input_tokens_seen": 19421288,
      "step": 29690
    },
    {
      "epoch": 15.56341719077568,
      "grad_norm": 0.6012617945671082,
      "learning_rate": 7.129353121010177e-06,
      "loss": 0.5134,
      "num_input_tokens_seen": 19424040,
      "step": 29695
    },
    {
      "epoch": 15.566037735849056,
      "grad_norm": 0.5885464549064636,
      "learning_rate": 7.121358966567779e-06,
      "loss": 0.4473,
      "num_input_tokens_seen": 19431336,
      "step": 29700
    },
    {
      "epoch": 15.568658280922431,
      "grad_norm": 0.9304638504981995,
      "learning_rate": 7.113368552133126e-06,
      "loss": 0.5145,
      "num_input_tokens_seen": 19434184,
      "step": 29705
    },
    {
      "epoch": 15.571278825995806,
      "grad_norm": 0.30530619621276855,
      "learning_rate": 7.105381879377701e-06,
      "loss": 0.3054,
      "num_input_tokens_seen": 19437960,
      "step": 29710
    },
    {
      "epoch": 15.573899371069182,
      "grad_norm": 0.4531291723251343,
      "learning_rate": 7.097398949972234e-06,
      "loss": 0.5167,
      "num_input_tokens_seen": 19441416,
      "step": 29715
    },
    {
      "epoch": 15.576519916142558,
      "grad_norm": 0.31192493438720703,
      "learning_rate": 7.089419765586672e-06,
      "loss": 0.4082,
      "num_input_tokens_seen": 19444712,
      "step": 29720
    },
    {
      "epoch": 15.579140461215934,
      "grad_norm": 0.473472535610199,
      "learning_rate": 7.081444327890152e-06,
      "loss": 0.476,
      "num_input_tokens_seen": 19448584,
      "step": 29725
    },
    {
      "epoch": 15.581761006289309,
      "grad_norm": 0.5112980604171753,
      "learning_rate": 7.073472638551054e-06,
      "loss": 0.3423,
      "num_input_tokens_seen": 19451048,
      "step": 29730
    },
    {
      "epoch": 15.584381551362684,
      "grad_norm": 0.33629006147384644,
      "learning_rate": 7.065504699236969e-06,
      "loss": 0.4165,
      "num_input_tokens_seen": 19454088,
      "step": 29735
    },
    {
      "epoch": 15.58700209643606,
      "grad_norm": 0.2965138852596283,
      "learning_rate": 7.057540511614688e-06,
      "loss": 0.4239,
      "num_input_tokens_seen": 19457832,
      "step": 29740
    },
    {
      "epoch": 15.589622641509434,
      "grad_norm": 0.335895836353302,
      "learning_rate": 7.049580077350232e-06,
      "loss": 0.3874,
      "num_input_tokens_seen": 19460968,
      "step": 29745
    },
    {
      "epoch": 15.59224318658281,
      "grad_norm": 0.42231589555740356,
      "learning_rate": 7.041623398108843e-06,
      "loss": 0.5747,
      "num_input_tokens_seen": 19463880,
      "step": 29750
    },
    {
      "epoch": 15.594863731656185,
      "grad_norm": 0.9376871585845947,
      "learning_rate": 7.033670475554949e-06,
      "loss": 0.3962,
      "num_input_tokens_seen": 19467624,
      "step": 29755
    },
    {
      "epoch": 15.59748427672956,
      "grad_norm": 0.6195992827415466,
      "learning_rate": 7.025721311352224e-06,
      "loss": 0.3411,
      "num_input_tokens_seen": 19472328,
      "step": 29760
    },
    {
      "epoch": 15.600104821802935,
      "grad_norm": 0.46718186140060425,
      "learning_rate": 7.017775907163546e-06,
      "loss": 0.4798,
      "num_input_tokens_seen": 19475336,
      "step": 29765
    },
    {
      "epoch": 15.60272536687631,
      "grad_norm": 0.5871198177337646,
      "learning_rate": 7.009834264650989e-06,
      "loss": 0.4404,
      "num_input_tokens_seen": 19477800,
      "step": 29770
    },
    {
      "epoch": 15.605345911949685,
      "grad_norm": 0.3492794334888458,
      "learning_rate": 7.001896385475867e-06,
      "loss": 0.2916,
      "num_input_tokens_seen": 19480968,
      "step": 29775
    },
    {
      "epoch": 15.60796645702306,
      "grad_norm": 0.4594140946865082,
      "learning_rate": 6.993962271298682e-06,
      "loss": 0.4999,
      "num_input_tokens_seen": 19484488,
      "step": 29780
    },
    {
      "epoch": 15.610587002096436,
      "grad_norm": 0.4931572377681732,
      "learning_rate": 6.986031923779166e-06,
      "loss": 0.7387,
      "num_input_tokens_seen": 19486952,
      "step": 29785
    },
    {
      "epoch": 15.61320754716981,
      "grad_norm": 0.7385359406471252,
      "learning_rate": 6.978105344576264e-06,
      "loss": 0.4174,
      "num_input_tokens_seen": 19489768,
      "step": 29790
    },
    {
      "epoch": 15.615828092243186,
      "grad_norm": 0.40216436982154846,
      "learning_rate": 6.970182535348111e-06,
      "loss": 0.3209,
      "num_input_tokens_seen": 19494088,
      "step": 29795
    },
    {
      "epoch": 15.618448637316561,
      "grad_norm": 0.5275047421455383,
      "learning_rate": 6.962263497752086e-06,
      "loss": 0.4691,
      "num_input_tokens_seen": 19497128,
      "step": 29800
    },
    {
      "epoch": 15.621069182389936,
      "grad_norm": 0.5579429268836975,
      "learning_rate": 6.9543482334447436e-06,
      "loss": 0.4153,
      "num_input_tokens_seen": 19500776,
      "step": 29805
    },
    {
      "epoch": 15.623689727463312,
      "grad_norm": 0.5577704906463623,
      "learning_rate": 6.946436744081875e-06,
      "loss": 0.5344,
      "num_input_tokens_seen": 19504104,
      "step": 29810
    },
    {
      "epoch": 15.626310272536688,
      "grad_norm": 0.36636224389076233,
      "learning_rate": 6.938529031318472e-06,
      "loss": 0.3869,
      "num_input_tokens_seen": 19506664,
      "step": 29815
    },
    {
      "epoch": 15.628930817610064,
      "grad_norm": 0.44481441378593445,
      "learning_rate": 6.9306250968087485e-06,
      "loss": 0.4971,
      "num_input_tokens_seen": 19509800,
      "step": 29820
    },
    {
      "epoch": 15.631551362683439,
      "grad_norm": 0.4159216284751892,
      "learning_rate": 6.922724942206102e-06,
      "loss": 0.3766,
      "num_input_tokens_seen": 19513064,
      "step": 29825
    },
    {
      "epoch": 15.634171907756814,
      "grad_norm": 0.5749322772026062,
      "learning_rate": 6.914828569163167e-06,
      "loss": 0.6667,
      "num_input_tokens_seen": 19515784,
      "step": 29830
    },
    {
      "epoch": 15.63679245283019,
      "grad_norm": 0.2645788788795471,
      "learning_rate": 6.906935979331763e-06,
      "loss": 0.3852,
      "num_input_tokens_seen": 19519880,
      "step": 29835
    },
    {
      "epoch": 15.639412997903564,
      "grad_norm": 0.6419863700866699,
      "learning_rate": 6.8990471743629356e-06,
      "loss": 0.5216,
      "num_input_tokens_seen": 19522376,
      "step": 29840
    },
    {
      "epoch": 15.64203354297694,
      "grad_norm": 0.7322667241096497,
      "learning_rate": 6.89116215590693e-06,
      "loss": 0.3989,
      "num_input_tokens_seen": 19525288,
      "step": 29845
    },
    {
      "epoch": 15.644654088050315,
      "grad_norm": 0.4670164883136749,
      "learning_rate": 6.8832809256132146e-06,
      "loss": 0.4821,
      "num_input_tokens_seen": 19528264,
      "step": 29850
    },
    {
      "epoch": 15.64727463312369,
      "grad_norm": 0.6896824240684509,
      "learning_rate": 6.875403485130444e-06,
      "loss": 0.4264,
      "num_input_tokens_seen": 19531496,
      "step": 29855
    },
    {
      "epoch": 15.649895178197065,
      "grad_norm": 0.506862461566925,
      "learning_rate": 6.867529836106479e-06,
      "loss": 0.3303,
      "num_input_tokens_seen": 19535048,
      "step": 29860
    },
    {
      "epoch": 15.65251572327044,
      "grad_norm": 0.518598735332489,
      "learning_rate": 6.8596599801884045e-06,
      "loss": 0.4451,
      "num_input_tokens_seen": 19537768,
      "step": 29865
    },
    {
      "epoch": 15.655136268343815,
      "grad_norm": 0.5422894954681396,
      "learning_rate": 6.851793919022509e-06,
      "loss": 0.3455,
      "num_input_tokens_seen": 19541032,
      "step": 29870
    },
    {
      "epoch": 15.65775681341719,
      "grad_norm": 0.6475769281387329,
      "learning_rate": 6.843931654254285e-06,
      "loss": 0.4227,
      "num_input_tokens_seen": 19544712,
      "step": 29875
    },
    {
      "epoch": 15.660377358490566,
      "grad_norm": 0.400588721036911,
      "learning_rate": 6.836073187528425e-06,
      "loss": 0.3976,
      "num_input_tokens_seen": 19548840,
      "step": 29880
    },
    {
      "epoch": 15.66299790356394,
      "grad_norm": 0.41065075993537903,
      "learning_rate": 6.828218520488821e-06,
      "loss": 0.4995,
      "num_input_tokens_seen": 19552168,
      "step": 29885
    },
    {
      "epoch": 15.665618448637316,
      "grad_norm": 0.5731022953987122,
      "learning_rate": 6.820367654778589e-06,
      "loss": 0.3525,
      "num_input_tokens_seen": 19555912,
      "step": 29890
    },
    {
      "epoch": 15.668238993710691,
      "grad_norm": 0.48445847630500793,
      "learning_rate": 6.812520592040039e-06,
      "loss": 0.5538,
      "num_input_tokens_seen": 19558984,
      "step": 29895
    },
    {
      "epoch": 15.670859538784066,
      "grad_norm": 0.5175164341926575,
      "learning_rate": 6.804677333914689e-06,
      "loss": 0.3523,
      "num_input_tokens_seen": 19561672,
      "step": 29900
    },
    {
      "epoch": 15.673480083857442,
      "grad_norm": 0.4045335054397583,
      "learning_rate": 6.796837882043261e-06,
      "loss": 0.4127,
      "num_input_tokens_seen": 19564520,
      "step": 29905
    },
    {
      "epoch": 15.676100628930818,
      "grad_norm": 0.414840966463089,
      "learning_rate": 6.7890022380656785e-06,
      "loss": 0.3307,
      "num_input_tokens_seen": 19567240,
      "step": 29910
    },
    {
      "epoch": 15.678721174004194,
      "grad_norm": 0.4923013746738434,
      "learning_rate": 6.781170403621056e-06,
      "loss": 0.3285,
      "num_input_tokens_seen": 19570792,
      "step": 29915
    },
    {
      "epoch": 15.681341719077569,
      "grad_norm": 0.26766741275787354,
      "learning_rate": 6.773342380347736e-06,
      "loss": 0.4497,
      "num_input_tokens_seen": 19574376,
      "step": 29920
    },
    {
      "epoch": 15.683962264150944,
      "grad_norm": 0.45953303575515747,
      "learning_rate": 6.765518169883248e-06,
      "loss": 0.3226,
      "num_input_tokens_seen": 19578088,
      "step": 29925
    },
    {
      "epoch": 15.68658280922432,
      "grad_norm": 0.671897828578949,
      "learning_rate": 6.757697773864338e-06,
      "loss": 0.5756,
      "num_input_tokens_seen": 19581224,
      "step": 29930
    },
    {
      "epoch": 15.689203354297694,
      "grad_norm": 1.1311194896697998,
      "learning_rate": 6.749881193926932e-06,
      "loss": 0.4272,
      "num_input_tokens_seen": 19583784,
      "step": 29935
    },
    {
      "epoch": 15.69182389937107,
      "grad_norm": 0.6375607848167419,
      "learning_rate": 6.742068431706167e-06,
      "loss": 0.362,
      "num_input_tokens_seen": 19586184,
      "step": 29940
    },
    {
      "epoch": 15.694444444444445,
      "grad_norm": 1.0119273662567139,
      "learning_rate": 6.734259488836386e-06,
      "loss": 0.5248,
      "num_input_tokens_seen": 19588200,
      "step": 29945
    },
    {
      "epoch": 15.69706498951782,
      "grad_norm": 0.6292349696159363,
      "learning_rate": 6.726454366951135e-06,
      "loss": 0.4109,
      "num_input_tokens_seen": 19591240,
      "step": 29950
    },
    {
      "epoch": 15.699685534591195,
      "grad_norm": 0.35519036650657654,
      "learning_rate": 6.7186530676831616e-06,
      "loss": 0.4621,
      "num_input_tokens_seen": 19596168,
      "step": 29955
    },
    {
      "epoch": 15.70230607966457,
      "grad_norm": 0.34077996015548706,
      "learning_rate": 6.710855592664403e-06,
      "loss": 0.3896,
      "num_input_tokens_seen": 19599880,
      "step": 29960
    },
    {
      "epoch": 15.704926624737945,
      "grad_norm": 0.31882402300834656,
      "learning_rate": 6.703061943525993e-06,
      "loss": 0.4172,
      "num_input_tokens_seen": 19603176,
      "step": 29965
    },
    {
      "epoch": 15.70754716981132,
      "grad_norm": 0.6201484799385071,
      "learning_rate": 6.69527212189828e-06,
      "loss": 0.4563,
      "num_input_tokens_seen": 19606344,
      "step": 29970
    },
    {
      "epoch": 15.710167714884696,
      "grad_norm": 0.5061672329902649,
      "learning_rate": 6.687486129410811e-06,
      "loss": 0.5298,
      "num_input_tokens_seen": 19609192,
      "step": 29975
    },
    {
      "epoch": 15.71278825995807,
      "grad_norm": 0.4057583212852478,
      "learning_rate": 6.679703967692322e-06,
      "loss": 0.4332,
      "num_input_tokens_seen": 19612744,
      "step": 29980
    },
    {
      "epoch": 15.715408805031446,
      "grad_norm": 0.45907706022262573,
      "learning_rate": 6.67192563837076e-06,
      "loss": 0.4129,
      "num_input_tokens_seen": 19615784,
      "step": 29985
    },
    {
      "epoch": 15.718029350104821,
      "grad_norm": 0.42286133766174316,
      "learning_rate": 6.664151143073258e-06,
      "loss": 0.4282,
      "num_input_tokens_seen": 19619176,
      "step": 29990
    },
    {
      "epoch": 15.720649895178196,
      "grad_norm": 0.871772050857544,
      "learning_rate": 6.656380483426141e-06,
      "loss": 0.5927,
      "num_input_tokens_seen": 19622632,
      "step": 29995
    },
    {
      "epoch": 15.723270440251572,
      "grad_norm": 0.5880774259567261,
      "learning_rate": 6.648613661054956e-06,
      "loss": 0.5686,
      "num_input_tokens_seen": 19625320,
      "step": 30000
    },
    {
      "epoch": 15.725890985324948,
      "grad_norm": 0.5586536526679993,
      "learning_rate": 6.6408506775844256e-06,
      "loss": 0.4589,
      "num_input_tokens_seen": 19629480,
      "step": 30005
    },
    {
      "epoch": 15.728511530398324,
      "grad_norm": 0.4262983202934265,
      "learning_rate": 6.633091534638492e-06,
      "loss": 0.4087,
      "num_input_tokens_seen": 19632296,
      "step": 30010
    },
    {
      "epoch": 15.731132075471699,
      "grad_norm": 0.5772345066070557,
      "learning_rate": 6.625336233840257e-06,
      "loss": 0.5078,
      "num_input_tokens_seen": 19635432,
      "step": 30015
    },
    {
      "epoch": 15.733752620545074,
      "grad_norm": 0.6542426943778992,
      "learning_rate": 6.617584776812064e-06,
      "loss": 0.396,
      "num_input_tokens_seen": 19639176,
      "step": 30020
    },
    {
      "epoch": 15.73637316561845,
      "grad_norm": 0.5069159269332886,
      "learning_rate": 6.6098371651754085e-06,
      "loss": 0.3858,
      "num_input_tokens_seen": 19642376,
      "step": 30025
    },
    {
      "epoch": 15.738993710691824,
      "grad_norm": 0.49738991260528564,
      "learning_rate": 6.602093400551012e-06,
      "loss": 0.6497,
      "num_input_tokens_seen": 19645224,
      "step": 30030
    },
    {
      "epoch": 15.7416142557652,
      "grad_norm": 0.43206095695495605,
      "learning_rate": 6.59435348455879e-06,
      "loss": 0.6322,
      "num_input_tokens_seen": 19648104,
      "step": 30035
    },
    {
      "epoch": 15.744234800838575,
      "grad_norm": 0.9630619287490845,
      "learning_rate": 6.586617418817828e-06,
      "loss": 0.3941,
      "num_input_tokens_seen": 19651912,
      "step": 30040
    },
    {
      "epoch": 15.74685534591195,
      "grad_norm": 0.9748889803886414,
      "learning_rate": 6.578885204946439e-06,
      "loss": 0.4983,
      "num_input_tokens_seen": 19654856,
      "step": 30045
    },
    {
      "epoch": 15.749475890985325,
      "grad_norm": 1.2717128992080688,
      "learning_rate": 6.571156844562098e-06,
      "loss": 0.7008,
      "num_input_tokens_seen": 19657640,
      "step": 30050
    },
    {
      "epoch": 15.7520964360587,
      "grad_norm": 0.4124138057231903,
      "learning_rate": 6.563432339281497e-06,
      "loss": 0.3402,
      "num_input_tokens_seen": 19661704,
      "step": 30055
    },
    {
      "epoch": 15.754716981132075,
      "grad_norm": 0.41835182905197144,
      "learning_rate": 6.555711690720517e-06,
      "loss": 0.354,
      "num_input_tokens_seen": 19664808,
      "step": 30060
    },
    {
      "epoch": 15.75733752620545,
      "grad_norm": 0.7028214931488037,
      "learning_rate": 6.547994900494234e-06,
      "loss": 0.36,
      "num_input_tokens_seen": 19668264,
      "step": 30065
    },
    {
      "epoch": 15.759958071278826,
      "grad_norm": 0.4350566565990448,
      "learning_rate": 6.540281970216899e-06,
      "loss": 0.4149,
      "num_input_tokens_seen": 19671176,
      "step": 30070
    },
    {
      "epoch": 15.7625786163522,
      "grad_norm": 0.9173377752304077,
      "learning_rate": 6.5325729015019845e-06,
      "loss": 0.5259,
      "num_input_tokens_seen": 19674504,
      "step": 30075
    },
    {
      "epoch": 15.765199161425576,
      "grad_norm": 0.569719135761261,
      "learning_rate": 6.524867695962122e-06,
      "loss": 0.3414,
      "num_input_tokens_seen": 19677864,
      "step": 30080
    },
    {
      "epoch": 15.767819706498951,
      "grad_norm": 1.1689225435256958,
      "learning_rate": 6.517166355209165e-06,
      "loss": 0.5266,
      "num_input_tokens_seen": 19680680,
      "step": 30085
    },
    {
      "epoch": 15.770440251572326,
      "grad_norm": 0.3057970404624939,
      "learning_rate": 6.50946888085415e-06,
      "loss": 0.4276,
      "num_input_tokens_seen": 19685096,
      "step": 30090
    },
    {
      "epoch": 15.773060796645701,
      "grad_norm": 0.6271118521690369,
      "learning_rate": 6.5017752745072855e-06,
      "loss": 0.364,
      "num_input_tokens_seen": 19688968,
      "step": 30095
    },
    {
      "epoch": 15.775681341719078,
      "grad_norm": 0.4431447386741638,
      "learning_rate": 6.494085537777994e-06,
      "loss": 0.4058,
      "num_input_tokens_seen": 19691528,
      "step": 30100
    },
    {
      "epoch": 15.778301886792454,
      "grad_norm": 0.3616977632045746,
      "learning_rate": 6.48639967227489e-06,
      "loss": 0.4277,
      "num_input_tokens_seen": 19694344,
      "step": 30105
    },
    {
      "epoch": 15.780922431865829,
      "grad_norm": 0.4677499234676361,
      "learning_rate": 6.47871767960575e-06,
      "loss": 0.5364,
      "num_input_tokens_seen": 19698664,
      "step": 30110
    },
    {
      "epoch": 15.783542976939204,
      "grad_norm": 0.5740287899971008,
      "learning_rate": 6.471039561377581e-06,
      "loss": 0.4157,
      "num_input_tokens_seen": 19701704,
      "step": 30115
    },
    {
      "epoch": 15.786163522012579,
      "grad_norm": 0.6807761192321777,
      "learning_rate": 6.463365319196538e-06,
      "loss": 0.3765,
      "num_input_tokens_seen": 19704904,
      "step": 30120
    },
    {
      "epoch": 15.788784067085954,
      "grad_norm": 1.0091612339019775,
      "learning_rate": 6.4556949546679905e-06,
      "loss": 0.46,
      "num_input_tokens_seen": 19707688,
      "step": 30125
    },
    {
      "epoch": 15.79140461215933,
      "grad_norm": 0.3967168629169464,
      "learning_rate": 6.448028469396497e-06,
      "loss": 0.4158,
      "num_input_tokens_seen": 19712264,
      "step": 30130
    },
    {
      "epoch": 15.794025157232705,
      "grad_norm": 0.6464819312095642,
      "learning_rate": 6.440365864985801e-06,
      "loss": 0.5058,
      "num_input_tokens_seen": 19714728,
      "step": 30135
    },
    {
      "epoch": 15.79664570230608,
      "grad_norm": 0.5589241981506348,
      "learning_rate": 6.432707143038818e-06,
      "loss": 0.4149,
      "num_input_tokens_seen": 19718312,
      "step": 30140
    },
    {
      "epoch": 15.799266247379455,
      "grad_norm": 0.7639501690864563,
      "learning_rate": 6.4250523051576834e-06,
      "loss": 0.4133,
      "num_input_tokens_seen": 19721672,
      "step": 30145
    },
    {
      "epoch": 15.80188679245283,
      "grad_norm": 0.5642570853233337,
      "learning_rate": 6.417401352943686e-06,
      "loss": 0.4051,
      "num_input_tokens_seen": 19724296,
      "step": 30150
    },
    {
      "epoch": 15.804507337526205,
      "grad_norm": 0.44574323296546936,
      "learning_rate": 6.409754287997322e-06,
      "loss": 0.5361,
      "num_input_tokens_seen": 19726952,
      "step": 30155
    },
    {
      "epoch": 15.80712788259958,
      "grad_norm": 0.3090016543865204,
      "learning_rate": 6.40211111191828e-06,
      "loss": 0.5145,
      "num_input_tokens_seen": 19730568,
      "step": 30160
    },
    {
      "epoch": 15.809748427672956,
      "grad_norm": 0.42281395196914673,
      "learning_rate": 6.394471826305409e-06,
      "loss": 0.2894,
      "num_input_tokens_seen": 19733288,
      "step": 30165
    },
    {
      "epoch": 15.81236897274633,
      "grad_norm": 0.42173245549201965,
      "learning_rate": 6.3868364327567795e-06,
      "loss": 0.3984,
      "num_input_tokens_seen": 19736008,
      "step": 30170
    },
    {
      "epoch": 15.814989517819706,
      "grad_norm": 0.40896567702293396,
      "learning_rate": 6.379204932869606e-06,
      "loss": 0.4782,
      "num_input_tokens_seen": 19739208,
      "step": 30175
    },
    {
      "epoch": 15.817610062893081,
      "grad_norm": 0.2765936553478241,
      "learning_rate": 6.371577328240327e-06,
      "loss": 0.4421,
      "num_input_tokens_seen": 19743080,
      "step": 30180
    },
    {
      "epoch": 15.820230607966456,
      "grad_norm": 0.6738669276237488,
      "learning_rate": 6.363953620464547e-06,
      "loss": 0.4643,
      "num_input_tokens_seen": 19745640,
      "step": 30185
    },
    {
      "epoch": 15.822851153039831,
      "grad_norm": 0.2936536371707916,
      "learning_rate": 6.356333811137064e-06,
      "loss": 0.4189,
      "num_input_tokens_seen": 19749032,
      "step": 30190
    },
    {
      "epoch": 15.825471698113208,
      "grad_norm": 0.5274679064750671,
      "learning_rate": 6.34871790185185e-06,
      "loss": 0.4457,
      "num_input_tokens_seen": 19752104,
      "step": 30195
    },
    {
      "epoch": 15.828092243186584,
      "grad_norm": 0.7075225710868835,
      "learning_rate": 6.341105894202057e-06,
      "loss": 0.457,
      "num_input_tokens_seen": 19755208,
      "step": 30200
    },
    {
      "epoch": 15.830712788259959,
      "grad_norm": 0.37815409898757935,
      "learning_rate": 6.333497789780041e-06,
      "loss": 0.3853,
      "num_input_tokens_seen": 19759944,
      "step": 30205
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 0.35152021050453186,
      "learning_rate": 6.325893590177329e-06,
      "loss": 0.5842,
      "num_input_tokens_seen": 19764872,
      "step": 30210
    },
    {
      "epoch": 15.835953878406709,
      "grad_norm": 0.6830692291259766,
      "learning_rate": 6.318293296984631e-06,
      "loss": 0.4884,
      "num_input_tokens_seen": 19767944,
      "step": 30215
    },
    {
      "epoch": 15.838574423480084,
      "grad_norm": 0.7939002513885498,
      "learning_rate": 6.3106969117918495e-06,
      "loss": 0.3984,
      "num_input_tokens_seen": 19771496,
      "step": 30220
    },
    {
      "epoch": 15.84119496855346,
      "grad_norm": 0.4200535714626312,
      "learning_rate": 6.303104436188057e-06,
      "loss": 0.5093,
      "num_input_tokens_seen": 19775080,
      "step": 30225
    },
    {
      "epoch": 15.843815513626835,
      "grad_norm": 0.585730791091919,
      "learning_rate": 6.2955158717615036e-06,
      "loss": 0.3438,
      "num_input_tokens_seen": 19777384,
      "step": 30230
    },
    {
      "epoch": 15.84643605870021,
      "grad_norm": 0.40731099247932434,
      "learning_rate": 6.287931220099638e-06,
      "loss": 0.63,
      "num_input_tokens_seen": 19780648,
      "step": 30235
    },
    {
      "epoch": 15.849056603773585,
      "grad_norm": 0.8182506561279297,
      "learning_rate": 6.280350482789082e-06,
      "loss": 0.5077,
      "num_input_tokens_seen": 19784072,
      "step": 30240
    },
    {
      "epoch": 15.85167714884696,
      "grad_norm": 0.3666708171367645,
      "learning_rate": 6.272773661415645e-06,
      "loss": 0.4541,
      "num_input_tokens_seen": 19787400,
      "step": 30245
    },
    {
      "epoch": 15.854297693920335,
      "grad_norm": 0.6808696985244751,
      "learning_rate": 6.26520075756431e-06,
      "loss": 0.3405,
      "num_input_tokens_seen": 19790248,
      "step": 30250
    },
    {
      "epoch": 15.85691823899371,
      "grad_norm": 0.6005519032478333,
      "learning_rate": 6.2576317728192304e-06,
      "loss": 0.3623,
      "num_input_tokens_seen": 19793224,
      "step": 30255
    },
    {
      "epoch": 15.859538784067086,
      "grad_norm": 0.6021909117698669,
      "learning_rate": 6.250066708763761e-06,
      "loss": 0.4856,
      "num_input_tokens_seen": 19796328,
      "step": 30260
    },
    {
      "epoch": 15.86215932914046,
      "grad_norm": 0.3956831693649292,
      "learning_rate": 6.242505566980422e-06,
      "loss": 0.3592,
      "num_input_tokens_seen": 19799176,
      "step": 30265
    },
    {
      "epoch": 15.864779874213836,
      "grad_norm": 0.35369622707366943,
      "learning_rate": 6.234948349050931e-06,
      "loss": 0.4707,
      "num_input_tokens_seen": 19802408,
      "step": 30270
    },
    {
      "epoch": 15.867400419287211,
      "grad_norm": 0.3460583984851837,
      "learning_rate": 6.227395056556162e-06,
      "loss": 0.3625,
      "num_input_tokens_seen": 19805320,
      "step": 30275
    },
    {
      "epoch": 15.870020964360586,
      "grad_norm": 0.47238120436668396,
      "learning_rate": 6.219845691076173e-06,
      "loss": 0.6859,
      "num_input_tokens_seen": 19809800,
      "step": 30280
    },
    {
      "epoch": 15.872641509433961,
      "grad_norm": 0.8703253865242004,
      "learning_rate": 6.212300254190206e-06,
      "loss": 0.4719,
      "num_input_tokens_seen": 19812072,
      "step": 30285
    },
    {
      "epoch": 15.875262054507338,
      "grad_norm": 0.3464801013469696,
      "learning_rate": 6.204758747476688e-06,
      "loss": 0.4122,
      "num_input_tokens_seen": 19815496,
      "step": 30290
    },
    {
      "epoch": 15.877882599580714,
      "grad_norm": 0.3127577304840088,
      "learning_rate": 6.1972211725132095e-06,
      "loss": 0.4881,
      "num_input_tokens_seen": 19819912,
      "step": 30295
    },
    {
      "epoch": 15.880503144654089,
      "grad_norm": 0.35652753710746765,
      "learning_rate": 6.189687530876559e-06,
      "loss": 0.432,
      "num_input_tokens_seen": 19823272,
      "step": 30300
    },
    {
      "epoch": 15.883123689727464,
      "grad_norm": 0.654811680316925,
      "learning_rate": 6.182157824142676e-06,
      "loss": 0.4577,
      "num_input_tokens_seen": 19827560,
      "step": 30305
    },
    {
      "epoch": 15.885744234800839,
      "grad_norm": 0.5038920640945435,
      "learning_rate": 6.174632053886681e-06,
      "loss": 0.5257,
      "num_input_tokens_seen": 19830184,
      "step": 30310
    },
    {
      "epoch": 15.888364779874214,
      "grad_norm": 0.632308840751648,
      "learning_rate": 6.167110221682893e-06,
      "loss": 0.4309,
      "num_input_tokens_seen": 19833640,
      "step": 30315
    },
    {
      "epoch": 15.89098532494759,
      "grad_norm": 0.6068304181098938,
      "learning_rate": 6.159592329104788e-06,
      "loss": 0.4517,
      "num_input_tokens_seen": 19836200,
      "step": 30320
    },
    {
      "epoch": 15.893605870020965,
      "grad_norm": 0.5728830695152283,
      "learning_rate": 6.152078377725032e-06,
      "loss": 0.5478,
      "num_input_tokens_seen": 19838888,
      "step": 30325
    },
    {
      "epoch": 15.89622641509434,
      "grad_norm": 0.38923272490501404,
      "learning_rate": 6.144568369115453e-06,
      "loss": 0.6639,
      "num_input_tokens_seen": 19841608,
      "step": 30330
    },
    {
      "epoch": 15.898846960167715,
      "grad_norm": 0.9739260077476501,
      "learning_rate": 6.137062304847046e-06,
      "loss": 0.4611,
      "num_input_tokens_seen": 19844648,
      "step": 30335
    },
    {
      "epoch": 15.90146750524109,
      "grad_norm": 0.48492616415023804,
      "learning_rate": 6.129560186490008e-06,
      "loss": 0.426,
      "num_input_tokens_seen": 19847432,
      "step": 30340
    },
    {
      "epoch": 15.904088050314465,
      "grad_norm": 0.49104583263397217,
      "learning_rate": 6.122062015613694e-06,
      "loss": 0.3461,
      "num_input_tokens_seen": 19850120,
      "step": 30345
    },
    {
      "epoch": 15.90670859538784,
      "grad_norm": 0.5918129682540894,
      "learning_rate": 6.114567793786641e-06,
      "loss": 0.4078,
      "num_input_tokens_seen": 19852776,
      "step": 30350
    },
    {
      "epoch": 15.909329140461216,
      "grad_norm": 0.5822403430938721,
      "learning_rate": 6.107077522576543e-06,
      "loss": 0.3816,
      "num_input_tokens_seen": 19856104,
      "step": 30355
    },
    {
      "epoch": 15.91194968553459,
      "grad_norm": 0.4223763942718506,
      "learning_rate": 6.099591203550292e-06,
      "loss": 0.402,
      "num_input_tokens_seen": 19859464,
      "step": 30360
    },
    {
      "epoch": 15.914570230607966,
      "grad_norm": 0.4081578552722931,
      "learning_rate": 6.092108838273927e-06,
      "loss": 0.5059,
      "num_input_tokens_seen": 19863208,
      "step": 30365
    },
    {
      "epoch": 15.917190775681341,
      "grad_norm": 0.6865708231925964,
      "learning_rate": 6.08463042831268e-06,
      "loss": 0.5464,
      "num_input_tokens_seen": 19865672,
      "step": 30370
    },
    {
      "epoch": 15.919811320754716,
      "grad_norm": 0.6991078853607178,
      "learning_rate": 6.0771559752309496e-06,
      "loss": 0.4751,
      "num_input_tokens_seen": 19868072,
      "step": 30375
    },
    {
      "epoch": 15.922431865828091,
      "grad_norm": 0.34953951835632324,
      "learning_rate": 6.069685480592313e-06,
      "loss": 0.4288,
      "num_input_tokens_seen": 19870952,
      "step": 30380
    },
    {
      "epoch": 15.925052410901468,
      "grad_norm": 0.4537391662597656,
      "learning_rate": 6.062218945959497e-06,
      "loss": 0.281,
      "num_input_tokens_seen": 19873832,
      "step": 30385
    },
    {
      "epoch": 15.927672955974844,
      "grad_norm": 0.5190725922584534,
      "learning_rate": 6.054756372894435e-06,
      "loss": 0.4773,
      "num_input_tokens_seen": 19876712,
      "step": 30390
    },
    {
      "epoch": 15.930293501048219,
      "grad_norm": 0.3447323739528656,
      "learning_rate": 6.04729776295819e-06,
      "loss": 0.3458,
      "num_input_tokens_seen": 19880008,
      "step": 30395
    },
    {
      "epoch": 15.932914046121594,
      "grad_norm": 0.38957780599594116,
      "learning_rate": 6.0398431177110306e-06,
      "loss": 0.4755,
      "num_input_tokens_seen": 19883240,
      "step": 30400
    },
    {
      "epoch": 15.935534591194969,
      "grad_norm": 0.4896058142185211,
      "learning_rate": 6.032392438712389e-06,
      "loss": 0.4565,
      "num_input_tokens_seen": 19886472,
      "step": 30405
    },
    {
      "epoch": 15.938155136268344,
      "grad_norm": 0.4417877197265625,
      "learning_rate": 6.024945727520847e-06,
      "loss": 0.3481,
      "num_input_tokens_seen": 19888904,
      "step": 30410
    },
    {
      "epoch": 15.94077568134172,
      "grad_norm": 0.2791753113269806,
      "learning_rate": 6.01750298569419e-06,
      "loss": 0.4785,
      "num_input_tokens_seen": 19892744,
      "step": 30415
    },
    {
      "epoch": 15.943396226415095,
      "grad_norm": 0.4339469075202942,
      "learning_rate": 6.010064214789335e-06,
      "loss": 0.5083,
      "num_input_tokens_seen": 19896488,
      "step": 30420
    },
    {
      "epoch": 15.94601677148847,
      "grad_norm": 0.9246313571929932,
      "learning_rate": 6.002629416362399e-06,
      "loss": 0.5352,
      "num_input_tokens_seen": 19899912,
      "step": 30425
    },
    {
      "epoch": 15.948637316561845,
      "grad_norm": 0.5024076700210571,
      "learning_rate": 5.995198591968662e-06,
      "loss": 0.4167,
      "num_input_tokens_seen": 19903432,
      "step": 30430
    },
    {
      "epoch": 15.95125786163522,
      "grad_norm": 0.4153388738632202,
      "learning_rate": 5.987771743162554e-06,
      "loss": 0.4223,
      "num_input_tokens_seen": 19906184,
      "step": 30435
    },
    {
      "epoch": 15.953878406708595,
      "grad_norm": 0.5691100358963013,
      "learning_rate": 5.9803488714976955e-06,
      "loss": 0.5564,
      "num_input_tokens_seen": 19909064,
      "step": 30440
    },
    {
      "epoch": 15.95649895178197,
      "grad_norm": 0.410470575094223,
      "learning_rate": 5.972929978526872e-06,
      "loss": 0.524,
      "num_input_tokens_seen": 19915944,
      "step": 30445
    },
    {
      "epoch": 15.959119496855346,
      "grad_norm": 0.520711362361908,
      "learning_rate": 5.965515065802019e-06,
      "loss": 0.3373,
      "num_input_tokens_seen": 19919784,
      "step": 30450
    },
    {
      "epoch": 15.96174004192872,
      "grad_norm": 0.3070763647556305,
      "learning_rate": 5.958104134874254e-06,
      "loss": 0.3758,
      "num_input_tokens_seen": 19922600,
      "step": 30455
    },
    {
      "epoch": 15.964360587002096,
      "grad_norm": 0.6720112562179565,
      "learning_rate": 5.950697187293872e-06,
      "loss": 0.4106,
      "num_input_tokens_seen": 19925704,
      "step": 30460
    },
    {
      "epoch": 15.966981132075471,
      "grad_norm": 0.4418190121650696,
      "learning_rate": 5.943294224610305e-06,
      "loss": 0.571,
      "num_input_tokens_seen": 19928296,
      "step": 30465
    },
    {
      "epoch": 15.969601677148846,
      "grad_norm": 0.3209335505962372,
      "learning_rate": 5.935895248372175e-06,
      "loss": 0.3717,
      "num_input_tokens_seen": 19931432,
      "step": 30470
    },
    {
      "epoch": 15.972222222222221,
      "grad_norm": 0.677839994430542,
      "learning_rate": 5.928500260127273e-06,
      "loss": 0.3761,
      "num_input_tokens_seen": 19934728,
      "step": 30475
    },
    {
      "epoch": 15.974842767295598,
      "grad_norm": 0.5034793615341187,
      "learning_rate": 5.921109261422531e-06,
      "loss": 0.4666,
      "num_input_tokens_seen": 19937800,
      "step": 30480
    },
    {
      "epoch": 15.977463312368974,
      "grad_norm": 0.3691592216491699,
      "learning_rate": 5.913722253804071e-06,
      "loss": 0.4528,
      "num_input_tokens_seen": 19941608,
      "step": 30485
    },
    {
      "epoch": 15.980083857442349,
      "grad_norm": 0.48423054814338684,
      "learning_rate": 5.906339238817163e-06,
      "loss": 0.4065,
      "num_input_tokens_seen": 19944008,
      "step": 30490
    },
    {
      "epoch": 15.982704402515724,
      "grad_norm": 0.34832292795181274,
      "learning_rate": 5.8989602180062554e-06,
      "loss": 0.4175,
      "num_input_tokens_seen": 19947016,
      "step": 30495
    },
    {
      "epoch": 15.985324947589099,
      "grad_norm": 0.4206872284412384,
      "learning_rate": 5.891585192914953e-06,
      "loss": 0.3058,
      "num_input_tokens_seen": 19951784,
      "step": 30500
    },
    {
      "epoch": 15.987945492662474,
      "grad_norm": 0.37103739380836487,
      "learning_rate": 5.884214165086036e-06,
      "loss": 0.3994,
      "num_input_tokens_seen": 19956136,
      "step": 30505
    },
    {
      "epoch": 15.99056603773585,
      "grad_norm": 0.4280370771884918,
      "learning_rate": 5.876847136061428e-06,
      "loss": 0.6769,
      "num_input_tokens_seen": 19959592,
      "step": 30510
    },
    {
      "epoch": 15.993186582809225,
      "grad_norm": 0.7447920441627502,
      "learning_rate": 5.869484107382228e-06,
      "loss": 0.4448,
      "num_input_tokens_seen": 19962920,
      "step": 30515
    },
    {
      "epoch": 15.9958071278826,
      "grad_norm": 0.6320412755012512,
      "learning_rate": 5.862125080588696e-06,
      "loss": 0.4392,
      "num_input_tokens_seen": 19965992,
      "step": 30520
    },
    {
      "epoch": 15.998427672955975,
      "grad_norm": 0.8228473663330078,
      "learning_rate": 5.854770057220263e-06,
      "loss": 0.4001,
      "num_input_tokens_seen": 19968904,
      "step": 30525
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.4767840802669525,
      "eval_runtime": 9.6444,
      "eval_samples_per_second": 87.927,
      "eval_steps_per_second": 21.982,
      "num_input_tokens_seen": 19970248,
      "step": 30528
    },
    {
      "epoch": 16.00104821802935,
      "grad_norm": 0.43003004789352417,
      "learning_rate": 5.8474190388155216e-06,
      "loss": 0.3382,
      "num_input_tokens_seen": 19971464,
      "step": 30530
    },
    {
      "epoch": 16.003668763102727,
      "grad_norm": 0.3709757626056671,
      "learning_rate": 5.840072026912205e-06,
      "loss": 0.3814,
      "num_input_tokens_seen": 19974760,
      "step": 30535
    },
    {
      "epoch": 16.0062893081761,
      "grad_norm": 0.5069681406021118,
      "learning_rate": 5.83272902304724e-06,
      "loss": 0.429,
      "num_input_tokens_seen": 19977992,
      "step": 30540
    },
    {
      "epoch": 16.008909853249477,
      "grad_norm": 0.5178096294403076,
      "learning_rate": 5.825390028756683e-06,
      "loss": 0.3496,
      "num_input_tokens_seen": 19981672,
      "step": 30545
    },
    {
      "epoch": 16.01153039832285,
      "grad_norm": 0.3950120806694031,
      "learning_rate": 5.818055045575776e-06,
      "loss": 0.3848,
      "num_input_tokens_seen": 19984520,
      "step": 30550
    },
    {
      "epoch": 16.014150943396228,
      "grad_norm": 0.2628556787967682,
      "learning_rate": 5.810724075038917e-06,
      "loss": 0.3304,
      "num_input_tokens_seen": 19987720,
      "step": 30555
    },
    {
      "epoch": 16.0167714884696,
      "grad_norm": 0.7048048377037048,
      "learning_rate": 5.803397118679665e-06,
      "loss": 0.5172,
      "num_input_tokens_seen": 19990344,
      "step": 30560
    },
    {
      "epoch": 16.019392033542978,
      "grad_norm": 0.5453453660011292,
      "learning_rate": 5.796074178030727e-06,
      "loss": 0.4502,
      "num_input_tokens_seen": 19993224,
      "step": 30565
    },
    {
      "epoch": 16.02201257861635,
      "grad_norm": 0.9457423090934753,
      "learning_rate": 5.788755254623976e-06,
      "loss": 0.4843,
      "num_input_tokens_seen": 19995688,
      "step": 30570
    },
    {
      "epoch": 16.02463312368973,
      "grad_norm": 0.43329259753227234,
      "learning_rate": 5.7814403499904525e-06,
      "loss": 0.3879,
      "num_input_tokens_seen": 19998536,
      "step": 30575
    },
    {
      "epoch": 16.0272536687631,
      "grad_norm": 0.5991328358650208,
      "learning_rate": 5.774129465660352e-06,
      "loss": 0.4525,
      "num_input_tokens_seen": 20001320,
      "step": 30580
    },
    {
      "epoch": 16.02987421383648,
      "grad_norm": 0.5406339764595032,
      "learning_rate": 5.7668226031630246e-06,
      "loss": 0.3995,
      "num_input_tokens_seen": 20004104,
      "step": 30585
    },
    {
      "epoch": 16.032494758909852,
      "grad_norm": 0.47615018486976624,
      "learning_rate": 5.759519764026999e-06,
      "loss": 0.4268,
      "num_input_tokens_seen": 20007880,
      "step": 30590
    },
    {
      "epoch": 16.03511530398323,
      "grad_norm": 0.5205857157707214,
      "learning_rate": 5.7522209497799185e-06,
      "loss": 0.4468,
      "num_input_tokens_seen": 20011816,
      "step": 30595
    },
    {
      "epoch": 16.037735849056602,
      "grad_norm": 0.2974132001399994,
      "learning_rate": 5.7449261619486246e-06,
      "loss": 0.3736,
      "num_input_tokens_seen": 20016008,
      "step": 30600
    },
    {
      "epoch": 16.04035639412998,
      "grad_norm": 0.45683515071868896,
      "learning_rate": 5.737635402059105e-06,
      "loss": 0.347,
      "num_input_tokens_seen": 20019112,
      "step": 30605
    },
    {
      "epoch": 16.042976939203353,
      "grad_norm": 1.0357714891433716,
      "learning_rate": 5.730348671636501e-06,
      "loss": 0.3882,
      "num_input_tokens_seen": 20021544,
      "step": 30610
    },
    {
      "epoch": 16.04559748427673,
      "grad_norm": 0.5551290512084961,
      "learning_rate": 5.723065972205124e-06,
      "loss": 0.3526,
      "num_input_tokens_seen": 20023976,
      "step": 30615
    },
    {
      "epoch": 16.048218029350103,
      "grad_norm": 0.37518152594566345,
      "learning_rate": 5.715787305288423e-06,
      "loss": 0.514,
      "num_input_tokens_seen": 20027560,
      "step": 30620
    },
    {
      "epoch": 16.05083857442348,
      "grad_norm": 0.43366768956184387,
      "learning_rate": 5.7085126724090024e-06,
      "loss": 0.453,
      "num_input_tokens_seen": 20030696,
      "step": 30625
    },
    {
      "epoch": 16.053459119496857,
      "grad_norm": 0.3450336456298828,
      "learning_rate": 5.7012420750886445e-06,
      "loss": 0.4205,
      "num_input_tokens_seen": 20033800,
      "step": 30630
    },
    {
      "epoch": 16.05607966457023,
      "grad_norm": 0.5813555717468262,
      "learning_rate": 5.693975514848271e-06,
      "loss": 0.3381,
      "num_input_tokens_seen": 20036136,
      "step": 30635
    },
    {
      "epoch": 16.058700209643607,
      "grad_norm": 0.3918152153491974,
      "learning_rate": 5.686712993207971e-06,
      "loss": 0.5391,
      "num_input_tokens_seen": 20039368,
      "step": 30640
    },
    {
      "epoch": 16.06132075471698,
      "grad_norm": 1.2758474349975586,
      "learning_rate": 5.679454511686977e-06,
      "loss": 0.3438,
      "num_input_tokens_seen": 20042376,
      "step": 30645
    },
    {
      "epoch": 16.063941299790358,
      "grad_norm": 0.537850022315979,
      "learning_rate": 5.672200071803671e-06,
      "loss": 0.3885,
      "num_input_tokens_seen": 20045608,
      "step": 30650
    },
    {
      "epoch": 16.06656184486373,
      "grad_norm": 0.3487362563610077,
      "learning_rate": 5.664949675075607e-06,
      "loss": 0.4271,
      "num_input_tokens_seen": 20048456,
      "step": 30655
    },
    {
      "epoch": 16.069182389937108,
      "grad_norm": 1.298641562461853,
      "learning_rate": 5.657703323019486e-06,
      "loss": 0.4321,
      "num_input_tokens_seen": 20051272,
      "step": 30660
    },
    {
      "epoch": 16.07180293501048,
      "grad_norm": 0.3932504653930664,
      "learning_rate": 5.65046101715116e-06,
      "loss": 0.4192,
      "num_input_tokens_seen": 20054440,
      "step": 30665
    },
    {
      "epoch": 16.07442348008386,
      "grad_norm": 0.6821528077125549,
      "learning_rate": 5.643222758985644e-06,
      "loss": 0.4317,
      "num_input_tokens_seen": 20056904,
      "step": 30670
    },
    {
      "epoch": 16.07704402515723,
      "grad_norm": 0.3579932451248169,
      "learning_rate": 5.635988550037094e-06,
      "loss": 0.4884,
      "num_input_tokens_seen": 20061352,
      "step": 30675
    },
    {
      "epoch": 16.07966457023061,
      "grad_norm": 0.4587358832359314,
      "learning_rate": 5.628758391818817e-06,
      "loss": 0.3848,
      "num_input_tokens_seen": 20064392,
      "step": 30680
    },
    {
      "epoch": 16.082285115303982,
      "grad_norm": 0.6843833923339844,
      "learning_rate": 5.621532285843287e-06,
      "loss": 0.4941,
      "num_input_tokens_seen": 20067528,
      "step": 30685
    },
    {
      "epoch": 16.08490566037736,
      "grad_norm": 0.6632573008537292,
      "learning_rate": 5.614310233622119e-06,
      "loss": 0.4151,
      "num_input_tokens_seen": 20070376,
      "step": 30690
    },
    {
      "epoch": 16.087526205450732,
      "grad_norm": 0.5704191327095032,
      "learning_rate": 5.607092236666093e-06,
      "loss": 0.3712,
      "num_input_tokens_seen": 20074440,
      "step": 30695
    },
    {
      "epoch": 16.09014675052411,
      "grad_norm": 0.5198731422424316,
      "learning_rate": 5.599878296485125e-06,
      "loss": 0.5182,
      "num_input_tokens_seen": 20080392,
      "step": 30700
    },
    {
      "epoch": 16.092767295597483,
      "grad_norm": 0.9907459616661072,
      "learning_rate": 5.592668414588284e-06,
      "loss": 0.4401,
      "num_input_tokens_seen": 20083336,
      "step": 30705
    },
    {
      "epoch": 16.09538784067086,
      "grad_norm": 0.7009323835372925,
      "learning_rate": 5.585462592483801e-06,
      "loss": 0.4714,
      "num_input_tokens_seen": 20086056,
      "step": 30710
    },
    {
      "epoch": 16.098008385744233,
      "grad_norm": 0.6873860955238342,
      "learning_rate": 5.578260831679047e-06,
      "loss": 0.406,
      "num_input_tokens_seen": 20093480,
      "step": 30715
    },
    {
      "epoch": 16.10062893081761,
      "grad_norm": 0.6668714880943298,
      "learning_rate": 5.571063133680562e-06,
      "loss": 0.3717,
      "num_input_tokens_seen": 20096648,
      "step": 30720
    },
    {
      "epoch": 16.103249475890987,
      "grad_norm": 0.9047843813896179,
      "learning_rate": 5.563869499994004e-06,
      "loss": 0.369,
      "num_input_tokens_seen": 20099080,
      "step": 30725
    },
    {
      "epoch": 16.10587002096436,
      "grad_norm": 0.46931177377700806,
      "learning_rate": 5.556679932124215e-06,
      "loss": 0.5599,
      "num_input_tokens_seen": 20101768,
      "step": 30730
    },
    {
      "epoch": 16.108490566037737,
      "grad_norm": 0.42650243639945984,
      "learning_rate": 5.549494431575159e-06,
      "loss": 0.3555,
      "num_input_tokens_seen": 20105192,
      "step": 30735
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 0.48413947224617004,
      "learning_rate": 5.542312999849961e-06,
      "loss": 0.3878,
      "num_input_tokens_seen": 20108072,
      "step": 30740
    },
    {
      "epoch": 16.113731656184488,
      "grad_norm": 0.30739280581474304,
      "learning_rate": 5.5351356384509e-06,
      "loss": 0.4246,
      "num_input_tokens_seen": 20112392,
      "step": 30745
    },
    {
      "epoch": 16.11635220125786,
      "grad_norm": 0.5856643319129944,
      "learning_rate": 5.527962348879401e-06,
      "loss": 0.3779,
      "num_input_tokens_seen": 20114792,
      "step": 30750
    },
    {
      "epoch": 16.118972746331238,
      "grad_norm": 0.5202236175537109,
      "learning_rate": 5.520793132636027e-06,
      "loss": 0.4034,
      "num_input_tokens_seen": 20117672,
      "step": 30755
    },
    {
      "epoch": 16.12159329140461,
      "grad_norm": 0.43077048659324646,
      "learning_rate": 5.513627991220502e-06,
      "loss": 0.4506,
      "num_input_tokens_seen": 20121096,
      "step": 30760
    },
    {
      "epoch": 16.12421383647799,
      "grad_norm": 0.4332990050315857,
      "learning_rate": 5.506466926131684e-06,
      "loss": 0.4877,
      "num_input_tokens_seen": 20123624,
      "step": 30765
    },
    {
      "epoch": 16.12683438155136,
      "grad_norm": 0.5487675070762634,
      "learning_rate": 5.4993099388675916e-06,
      "loss": 0.5356,
      "num_input_tokens_seen": 20126920,
      "step": 30770
    },
    {
      "epoch": 16.12945492662474,
      "grad_norm": 0.3481157124042511,
      "learning_rate": 5.49215703092539e-06,
      "loss": 0.4357,
      "num_input_tokens_seen": 20130312,
      "step": 30775
    },
    {
      "epoch": 16.132075471698112,
      "grad_norm": 0.397477388381958,
      "learning_rate": 5.4850082038013755e-06,
      "loss": 0.3614,
      "num_input_tokens_seen": 20133416,
      "step": 30780
    },
    {
      "epoch": 16.13469601677149,
      "grad_norm": 0.365492582321167,
      "learning_rate": 5.4778634589910115e-06,
      "loss": 0.4303,
      "num_input_tokens_seen": 20136936,
      "step": 30785
    },
    {
      "epoch": 16.137316561844862,
      "grad_norm": 0.42225977778434753,
      "learning_rate": 5.4707227979888834e-06,
      "loss": 0.4714,
      "num_input_tokens_seen": 20139752,
      "step": 30790
    },
    {
      "epoch": 16.13993710691824,
      "grad_norm": 1.1972705125808716,
      "learning_rate": 5.463586222288747e-06,
      "loss": 0.4973,
      "num_input_tokens_seen": 20143112,
      "step": 30795
    },
    {
      "epoch": 16.142557651991613,
      "grad_norm": 0.30641987919807434,
      "learning_rate": 5.456453733383493e-06,
      "loss": 0.547,
      "num_input_tokens_seen": 20146280,
      "step": 30800
    },
    {
      "epoch": 16.14517819706499,
      "grad_norm": 0.6172212958335876,
      "learning_rate": 5.449325332765151e-06,
      "loss": 0.6228,
      "num_input_tokens_seen": 20148840,
      "step": 30805
    },
    {
      "epoch": 16.147798742138363,
      "grad_norm": 0.3733736574649811,
      "learning_rate": 5.4422010219249e-06,
      "loss": 0.4968,
      "num_input_tokens_seen": 20152328,
      "step": 30810
    },
    {
      "epoch": 16.15041928721174,
      "grad_norm": 0.6281514167785645,
      "learning_rate": 5.435080802353077e-06,
      "loss": 0.4151,
      "num_input_tokens_seen": 20155592,
      "step": 30815
    },
    {
      "epoch": 16.153039832285117,
      "grad_norm": 0.5489354133605957,
      "learning_rate": 5.4279646755391325e-06,
      "loss": 0.4846,
      "num_input_tokens_seen": 20158216,
      "step": 30820
    },
    {
      "epoch": 16.15566037735849,
      "grad_norm": 0.5093144774436951,
      "learning_rate": 5.420852642971691e-06,
      "loss": 0.4766,
      "num_input_tokens_seen": 20162376,
      "step": 30825
    },
    {
      "epoch": 16.158280922431867,
      "grad_norm": 0.592062771320343,
      "learning_rate": 5.41374470613851e-06,
      "loss": 0.5581,
      "num_input_tokens_seen": 20165736,
      "step": 30830
    },
    {
      "epoch": 16.16090146750524,
      "grad_norm": 0.8750829696655273,
      "learning_rate": 5.40664086652648e-06,
      "loss": 0.3642,
      "num_input_tokens_seen": 20169544,
      "step": 30835
    },
    {
      "epoch": 16.163522012578618,
      "grad_norm": 0.5355885028839111,
      "learning_rate": 5.399541125621646e-06,
      "loss": 0.4358,
      "num_input_tokens_seen": 20173128,
      "step": 30840
    },
    {
      "epoch": 16.16614255765199,
      "grad_norm": 0.40136751532554626,
      "learning_rate": 5.392445484909203e-06,
      "loss": 0.3713,
      "num_input_tokens_seen": 20176776,
      "step": 30845
    },
    {
      "epoch": 16.168763102725368,
      "grad_norm": 0.4167482256889343,
      "learning_rate": 5.385353945873464e-06,
      "loss": 0.4321,
      "num_input_tokens_seen": 20180904,
      "step": 30850
    },
    {
      "epoch": 16.17138364779874,
      "grad_norm": 0.3276969790458679,
      "learning_rate": 5.378266509997909e-06,
      "loss": 0.3998,
      "num_input_tokens_seen": 20185416,
      "step": 30855
    },
    {
      "epoch": 16.17400419287212,
      "grad_norm": 0.48826828598976135,
      "learning_rate": 5.371183178765135e-06,
      "loss": 0.5057,
      "num_input_tokens_seen": 20188456,
      "step": 30860
    },
    {
      "epoch": 16.17662473794549,
      "grad_norm": 0.576265275478363,
      "learning_rate": 5.364103953656904e-06,
      "loss": 0.4369,
      "num_input_tokens_seen": 20191368,
      "step": 30865
    },
    {
      "epoch": 16.17924528301887,
      "grad_norm": 0.3508623242378235,
      "learning_rate": 5.35702883615411e-06,
      "loss": 0.5048,
      "num_input_tokens_seen": 20194120,
      "step": 30870
    },
    {
      "epoch": 16.181865828092242,
      "grad_norm": 0.5425946712493896,
      "learning_rate": 5.34995782773679e-06,
      "loss": 0.4304,
      "num_input_tokens_seen": 20197320,
      "step": 30875
    },
    {
      "epoch": 16.18448637316562,
      "grad_norm": 0.32493114471435547,
      "learning_rate": 5.3428909298841115e-06,
      "loss": 0.4697,
      "num_input_tokens_seen": 20200168,
      "step": 30880
    },
    {
      "epoch": 16.187106918238992,
      "grad_norm": 0.5826460123062134,
      "learning_rate": 5.335828144074384e-06,
      "loss": 0.5252,
      "num_input_tokens_seen": 20203304,
      "step": 30885
    },
    {
      "epoch": 16.18972746331237,
      "grad_norm": 0.5154299736022949,
      "learning_rate": 5.3287694717850695e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 20206184,
      "step": 30890
    },
    {
      "epoch": 16.192348008385743,
      "grad_norm": 0.591738224029541,
      "learning_rate": 5.321714914492759e-06,
      "loss": 0.4193,
      "num_input_tokens_seen": 20209320,
      "step": 30895
    },
    {
      "epoch": 16.19496855345912,
      "grad_norm": 0.37791991233825684,
      "learning_rate": 5.314664473673195e-06,
      "loss": 0.484,
      "num_input_tokens_seen": 20212616,
      "step": 30900
    },
    {
      "epoch": 16.197589098532493,
      "grad_norm": 0.500738263130188,
      "learning_rate": 5.307618150801239e-06,
      "loss": 0.3803,
      "num_input_tokens_seen": 20215272,
      "step": 30905
    },
    {
      "epoch": 16.20020964360587,
      "grad_norm": 0.33231571316719055,
      "learning_rate": 5.300575947350897e-06,
      "loss": 0.4162,
      "num_input_tokens_seen": 20218248,
      "step": 30910
    },
    {
      "epoch": 16.202830188679247,
      "grad_norm": 1.068648338317871,
      "learning_rate": 5.293537864795325e-06,
      "loss": 0.4191,
      "num_input_tokens_seen": 20220584,
      "step": 30915
    },
    {
      "epoch": 16.20545073375262,
      "grad_norm": 0.6031101942062378,
      "learning_rate": 5.2865039046068075e-06,
      "loss": 0.4789,
      "num_input_tokens_seen": 20223912,
      "step": 30920
    },
    {
      "epoch": 16.208071278825997,
      "grad_norm": 0.39748725295066833,
      "learning_rate": 5.279474068256767e-06,
      "loss": 0.3398,
      "num_input_tokens_seen": 20227592,
      "step": 30925
    },
    {
      "epoch": 16.21069182389937,
      "grad_norm": 0.5647238492965698,
      "learning_rate": 5.272448357215779e-06,
      "loss": 0.3813,
      "num_input_tokens_seen": 20230696,
      "step": 30930
    },
    {
      "epoch": 16.213312368972748,
      "grad_norm": 0.5934594869613647,
      "learning_rate": 5.26542677295353e-06,
      "loss": 0.5114,
      "num_input_tokens_seen": 20234248,
      "step": 30935
    },
    {
      "epoch": 16.21593291404612,
      "grad_norm": 0.5576099753379822,
      "learning_rate": 5.258409316938848e-06,
      "loss": 0.4585,
      "num_input_tokens_seen": 20237416,
      "step": 30940
    },
    {
      "epoch": 16.218553459119498,
      "grad_norm": 0.43279725313186646,
      "learning_rate": 5.251395990639713e-06,
      "loss": 0.3796,
      "num_input_tokens_seen": 20240392,
      "step": 30945
    },
    {
      "epoch": 16.22117400419287,
      "grad_norm": 0.4608955383300781,
      "learning_rate": 5.244386795523235e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 20242856,
      "step": 30950
    },
    {
      "epoch": 16.22379454926625,
      "grad_norm": 0.5364538431167603,
      "learning_rate": 5.23738173305566e-06,
      "loss": 0.3713,
      "num_input_tokens_seen": 20245608,
      "step": 30955
    },
    {
      "epoch": 16.22641509433962,
      "grad_norm": 0.5750195980072021,
      "learning_rate": 5.230380804702362e-06,
      "loss": 0.3582,
      "num_input_tokens_seen": 20248136,
      "step": 30960
    },
    {
      "epoch": 16.229035639413,
      "grad_norm": 0.588326632976532,
      "learning_rate": 5.223384011927851e-06,
      "loss": 0.3824,
      "num_input_tokens_seen": 20251688,
      "step": 30965
    },
    {
      "epoch": 16.231656184486372,
      "grad_norm": 0.3465929925441742,
      "learning_rate": 5.216391356195779e-06,
      "loss": 0.5123,
      "num_input_tokens_seen": 20254664,
      "step": 30970
    },
    {
      "epoch": 16.23427672955975,
      "grad_norm": 0.2966039180755615,
      "learning_rate": 5.209402838968935e-06,
      "loss": 0.4147,
      "num_input_tokens_seen": 20258440,
      "step": 30975
    },
    {
      "epoch": 16.236897274633122,
      "grad_norm": 0.5688019394874573,
      "learning_rate": 5.202418461709233e-06,
      "loss": 0.3125,
      "num_input_tokens_seen": 20261480,
      "step": 30980
    },
    {
      "epoch": 16.2395178197065,
      "grad_norm": 0.4459270238876343,
      "learning_rate": 5.195438225877738e-06,
      "loss": 0.3646,
      "num_input_tokens_seen": 20264520,
      "step": 30985
    },
    {
      "epoch": 16.242138364779873,
      "grad_norm": 0.9653254151344299,
      "learning_rate": 5.18846213293461e-06,
      "loss": 0.4546,
      "num_input_tokens_seen": 20267528,
      "step": 30990
    },
    {
      "epoch": 16.24475890985325,
      "grad_norm": 0.4759824573993683,
      "learning_rate": 5.181490184339185e-06,
      "loss": 0.4489,
      "num_input_tokens_seen": 20270568,
      "step": 30995
    },
    {
      "epoch": 16.247379454926623,
      "grad_norm": 1.0905530452728271,
      "learning_rate": 5.17452238154991e-06,
      "loss": 0.4877,
      "num_input_tokens_seen": 20272840,
      "step": 31000
    },
    {
      "epoch": 16.25,
      "grad_norm": 0.5597645044326782,
      "learning_rate": 5.1675587260243715e-06,
      "loss": 0.3828,
      "num_input_tokens_seen": 20275656,
      "step": 31005
    },
    {
      "epoch": 16.252620545073377,
      "grad_norm": 0.5191055536270142,
      "learning_rate": 5.160599219219295e-06,
      "loss": 0.4994,
      "num_input_tokens_seen": 20278824,
      "step": 31010
    },
    {
      "epoch": 16.25524109014675,
      "grad_norm": 0.5767310261726379,
      "learning_rate": 5.1536438625905185e-06,
      "loss": 0.4318,
      "num_input_tokens_seen": 20282408,
      "step": 31015
    },
    {
      "epoch": 16.257861635220127,
      "grad_norm": 0.28271347284317017,
      "learning_rate": 5.146692657593025e-06,
      "loss": 0.3402,
      "num_input_tokens_seen": 20285640,
      "step": 31020
    },
    {
      "epoch": 16.2604821802935,
      "grad_norm": 0.6186947226524353,
      "learning_rate": 5.139745605680926e-06,
      "loss": 0.4006,
      "num_input_tokens_seen": 20288872,
      "step": 31025
    },
    {
      "epoch": 16.263102725366878,
      "grad_norm": 0.383083701133728,
      "learning_rate": 5.132802708307469e-06,
      "loss": 0.4099,
      "num_input_tokens_seen": 20292264,
      "step": 31030
    },
    {
      "epoch": 16.26572327044025,
      "grad_norm": 0.5083844661712646,
      "learning_rate": 5.125863966925035e-06,
      "loss": 0.3325,
      "num_input_tokens_seen": 20296232,
      "step": 31035
    },
    {
      "epoch": 16.268343815513628,
      "grad_norm": 0.8523531556129456,
      "learning_rate": 5.118929382985116e-06,
      "loss": 0.484,
      "num_input_tokens_seen": 20299048,
      "step": 31040
    },
    {
      "epoch": 16.270964360587,
      "grad_norm": 0.349443644285202,
      "learning_rate": 5.111998957938363e-06,
      "loss": 0.4425,
      "num_input_tokens_seen": 20301864,
      "step": 31045
    },
    {
      "epoch": 16.27358490566038,
      "grad_norm": 0.9788272976875305,
      "learning_rate": 5.105072693234525e-06,
      "loss": 0.3607,
      "num_input_tokens_seen": 20305256,
      "step": 31050
    },
    {
      "epoch": 16.27620545073375,
      "grad_norm": 0.9331353902816772,
      "learning_rate": 5.098150590322506e-06,
      "loss": 0.3946,
      "num_input_tokens_seen": 20308616,
      "step": 31055
    },
    {
      "epoch": 16.27882599580713,
      "grad_norm": 0.6377644538879395,
      "learning_rate": 5.091232650650335e-06,
      "loss": 0.5283,
      "num_input_tokens_seen": 20312200,
      "step": 31060
    },
    {
      "epoch": 16.281446540880502,
      "grad_norm": 0.5933554768562317,
      "learning_rate": 5.0843188756651645e-06,
      "loss": 0.4163,
      "num_input_tokens_seen": 20316200,
      "step": 31065
    },
    {
      "epoch": 16.28406708595388,
      "grad_norm": 0.43344253301620483,
      "learning_rate": 5.077409266813276e-06,
      "loss": 0.4654,
      "num_input_tokens_seen": 20319944,
      "step": 31070
    },
    {
      "epoch": 16.286687631027252,
      "grad_norm": 0.5498464107513428,
      "learning_rate": 5.0705038255400765e-06,
      "loss": 0.4455,
      "num_input_tokens_seen": 20323240,
      "step": 31075
    },
    {
      "epoch": 16.28930817610063,
      "grad_norm": 0.3105725944042206,
      "learning_rate": 5.063602553290106e-06,
      "loss": 0.4847,
      "num_input_tokens_seen": 20327080,
      "step": 31080
    },
    {
      "epoch": 16.291928721174003,
      "grad_norm": 0.3300004303455353,
      "learning_rate": 5.056705451507035e-06,
      "loss": 0.3965,
      "num_input_tokens_seen": 20329800,
      "step": 31085
    },
    {
      "epoch": 16.29454926624738,
      "grad_norm": 0.21881785988807678,
      "learning_rate": 5.049812521633668e-06,
      "loss": 0.3614,
      "num_input_tokens_seen": 20333480,
      "step": 31090
    },
    {
      "epoch": 16.297169811320753,
      "grad_norm": 0.5912506580352783,
      "learning_rate": 5.042923765111907e-06,
      "loss": 0.3912,
      "num_input_tokens_seen": 20336616,
      "step": 31095
    },
    {
      "epoch": 16.29979035639413,
      "grad_norm": 0.5674607753753662,
      "learning_rate": 5.036039183382821e-06,
      "loss": 0.3259,
      "num_input_tokens_seen": 20340776,
      "step": 31100
    },
    {
      "epoch": 16.302410901467507,
      "grad_norm": 0.37658077478408813,
      "learning_rate": 5.02915877788657e-06,
      "loss": 0.4417,
      "num_input_tokens_seen": 20343752,
      "step": 31105
    },
    {
      "epoch": 16.30503144654088,
      "grad_norm": 0.5187650322914124,
      "learning_rate": 5.0222825500624625e-06,
      "loss": 0.4994,
      "num_input_tokens_seen": 20347496,
      "step": 31110
    },
    {
      "epoch": 16.307651991614257,
      "grad_norm": 0.3477337658405304,
      "learning_rate": 5.015410501348935e-06,
      "loss": 0.4139,
      "num_input_tokens_seen": 20351400,
      "step": 31115
    },
    {
      "epoch": 16.31027253668763,
      "grad_norm": 0.9632290005683899,
      "learning_rate": 5.008542633183527e-06,
      "loss": 0.4405,
      "num_input_tokens_seen": 20355848,
      "step": 31120
    },
    {
      "epoch": 16.312893081761008,
      "grad_norm": 0.46841302514076233,
      "learning_rate": 5.001678947002927e-06,
      "loss": 0.4889,
      "num_input_tokens_seen": 20359208,
      "step": 31125
    },
    {
      "epoch": 16.31551362683438,
      "grad_norm": 0.3750988245010376,
      "learning_rate": 4.994819444242943e-06,
      "loss": 0.3861,
      "num_input_tokens_seen": 20362664,
      "step": 31130
    },
    {
      "epoch": 16.318134171907758,
      "grad_norm": 0.4439634084701538,
      "learning_rate": 4.9879641263384956e-06,
      "loss": 0.4243,
      "num_input_tokens_seen": 20365736,
      "step": 31135
    },
    {
      "epoch": 16.32075471698113,
      "grad_norm": 0.2797672152519226,
      "learning_rate": 4.981112994723644e-06,
      "loss": 0.5011,
      "num_input_tokens_seen": 20369608,
      "step": 31140
    },
    {
      "epoch": 16.32337526205451,
      "grad_norm": 0.5625189542770386,
      "learning_rate": 4.974266050831572e-06,
      "loss": 0.4027,
      "num_input_tokens_seen": 20372168,
      "step": 31145
    },
    {
      "epoch": 16.32599580712788,
      "grad_norm": 0.3969637453556061,
      "learning_rate": 4.967423296094567e-06,
      "loss": 0.3217,
      "num_input_tokens_seen": 20374856,
      "step": 31150
    },
    {
      "epoch": 16.32861635220126,
      "grad_norm": 0.5236446857452393,
      "learning_rate": 4.960584731944076e-06,
      "loss": 0.5344,
      "num_input_tokens_seen": 20378152,
      "step": 31155
    },
    {
      "epoch": 16.331236897274632,
      "grad_norm": 0.6257985234260559,
      "learning_rate": 4.953750359810627e-06,
      "loss": 0.4405,
      "num_input_tokens_seen": 20380936,
      "step": 31160
    },
    {
      "epoch": 16.33385744234801,
      "grad_norm": 0.3145245313644409,
      "learning_rate": 4.946920181123904e-06,
      "loss": 0.4569,
      "num_input_tokens_seen": 20384584,
      "step": 31165
    },
    {
      "epoch": 16.336477987421382,
      "grad_norm": 0.4341670870780945,
      "learning_rate": 4.940094197312709e-06,
      "loss": 0.4387,
      "num_input_tokens_seen": 20387848,
      "step": 31170
    },
    {
      "epoch": 16.33909853249476,
      "grad_norm": 0.7224346995353699,
      "learning_rate": 4.933272409804946e-06,
      "loss": 0.6525,
      "num_input_tokens_seen": 20390632,
      "step": 31175
    },
    {
      "epoch": 16.341719077568133,
      "grad_norm": 0.644141674041748,
      "learning_rate": 4.926454820027663e-06,
      "loss": 0.3207,
      "num_input_tokens_seen": 20393544,
      "step": 31180
    },
    {
      "epoch": 16.34433962264151,
      "grad_norm": 0.391855925321579,
      "learning_rate": 4.919641429407024e-06,
      "loss": 0.3818,
      "num_input_tokens_seen": 20397256,
      "step": 31185
    },
    {
      "epoch": 16.346960167714883,
      "grad_norm": 0.5679445266723633,
      "learning_rate": 4.912832239368304e-06,
      "loss": 0.5697,
      "num_input_tokens_seen": 20401032,
      "step": 31190
    },
    {
      "epoch": 16.34958071278826,
      "grad_norm": 0.39295244216918945,
      "learning_rate": 4.906027251335918e-06,
      "loss": 0.5993,
      "num_input_tokens_seen": 20404456,
      "step": 31195
    },
    {
      "epoch": 16.352201257861637,
      "grad_norm": 0.4030938744544983,
      "learning_rate": 4.899226466733381e-06,
      "loss": 0.3254,
      "num_input_tokens_seen": 20407624,
      "step": 31200
    },
    {
      "epoch": 16.35482180293501,
      "grad_norm": 0.5594366192817688,
      "learning_rate": 4.892429886983346e-06,
      "loss": 0.5072,
      "num_input_tokens_seen": 20411112,
      "step": 31205
    },
    {
      "epoch": 16.357442348008387,
      "grad_norm": 0.53615802526474,
      "learning_rate": 4.88563751350758e-06,
      "loss": 0.42,
      "num_input_tokens_seen": 20413640,
      "step": 31210
    },
    {
      "epoch": 16.36006289308176,
      "grad_norm": 0.3644702136516571,
      "learning_rate": 4.878849347726977e-06,
      "loss": 0.5122,
      "num_input_tokens_seen": 20416424,
      "step": 31215
    },
    {
      "epoch": 16.362683438155138,
      "grad_norm": 0.43454840779304504,
      "learning_rate": 4.872065391061526e-06,
      "loss": 0.4758,
      "num_input_tokens_seen": 20419816,
      "step": 31220
    },
    {
      "epoch": 16.36530398322851,
      "grad_norm": 0.35831815004348755,
      "learning_rate": 4.865285644930373e-06,
      "loss": 0.4635,
      "num_input_tokens_seen": 20423432,
      "step": 31225
    },
    {
      "epoch": 16.367924528301888,
      "grad_norm": 0.37493735551834106,
      "learning_rate": 4.858510110751746e-06,
      "loss": 0.3977,
      "num_input_tokens_seen": 20426792,
      "step": 31230
    },
    {
      "epoch": 16.37054507337526,
      "grad_norm": 0.4642384350299835,
      "learning_rate": 4.851738789943019e-06,
      "loss": 0.4331,
      "num_input_tokens_seen": 20430536,
      "step": 31235
    },
    {
      "epoch": 16.37316561844864,
      "grad_norm": 0.8138437867164612,
      "learning_rate": 4.844971683920674e-06,
      "loss": 0.4145,
      "num_input_tokens_seen": 20433256,
      "step": 31240
    },
    {
      "epoch": 16.37578616352201,
      "grad_norm": 0.3722066879272461,
      "learning_rate": 4.838208794100316e-06,
      "loss": 0.4898,
      "num_input_tokens_seen": 20436744,
      "step": 31245
    },
    {
      "epoch": 16.37840670859539,
      "grad_norm": 0.4234342873096466,
      "learning_rate": 4.831450121896661e-06,
      "loss": 0.3447,
      "num_input_tokens_seen": 20441672,
      "step": 31250
    },
    {
      "epoch": 16.381027253668762,
      "grad_norm": 0.9916822910308838,
      "learning_rate": 4.824695668723542e-06,
      "loss": 0.4582,
      "num_input_tokens_seen": 20444776,
      "step": 31255
    },
    {
      "epoch": 16.38364779874214,
      "grad_norm": 1.2120074033737183,
      "learning_rate": 4.817945435993915e-06,
      "loss": 0.4629,
      "num_input_tokens_seen": 20447400,
      "step": 31260
    },
    {
      "epoch": 16.386268343815512,
      "grad_norm": 0.419006884098053,
      "learning_rate": 4.811199425119853e-06,
      "loss": 0.4185,
      "num_input_tokens_seen": 20450280,
      "step": 31265
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 0.6887242794036865,
      "learning_rate": 4.804457637512552e-06,
      "loss": 0.3619,
      "num_input_tokens_seen": 20453320,
      "step": 31270
    },
    {
      "epoch": 16.391509433962263,
      "grad_norm": 0.40912964940071106,
      "learning_rate": 4.79772007458231e-06,
      "loss": 0.4384,
      "num_input_tokens_seen": 20456008,
      "step": 31275
    },
    {
      "epoch": 16.39412997903564,
      "grad_norm": 1.1559489965438843,
      "learning_rate": 4.7909867377385424e-06,
      "loss": 0.3147,
      "num_input_tokens_seen": 20458120,
      "step": 31280
    },
    {
      "epoch": 16.396750524109013,
      "grad_norm": 0.48637017607688904,
      "learning_rate": 4.784257628389794e-06,
      "loss": 0.5018,
      "num_input_tokens_seen": 20461448,
      "step": 31285
    },
    {
      "epoch": 16.39937106918239,
      "grad_norm": 0.5547441840171814,
      "learning_rate": 4.7775327479437166e-06,
      "loss": 0.4227,
      "num_input_tokens_seen": 20465096,
      "step": 31290
    },
    {
      "epoch": 16.401991614255767,
      "grad_norm": 0.448246568441391,
      "learning_rate": 4.770812097807076e-06,
      "loss": 0.456,
      "num_input_tokens_seen": 20467912,
      "step": 31295
    },
    {
      "epoch": 16.40461215932914,
      "grad_norm": 0.46172258257865906,
      "learning_rate": 4.764095679385766e-06,
      "loss": 0.4311,
      "num_input_tokens_seen": 20470728,
      "step": 31300
    },
    {
      "epoch": 16.407232704402517,
      "grad_norm": 0.5437982678413391,
      "learning_rate": 4.7573834940847745e-06,
      "loss": 0.3895,
      "num_input_tokens_seen": 20473064,
      "step": 31305
    },
    {
      "epoch": 16.40985324947589,
      "grad_norm": 0.598105251789093,
      "learning_rate": 4.750675543308212e-06,
      "loss": 0.4676,
      "num_input_tokens_seen": 20475720,
      "step": 31310
    },
    {
      "epoch": 16.412473794549268,
      "grad_norm": 0.5154401063919067,
      "learning_rate": 4.743971828459307e-06,
      "loss": 0.4698,
      "num_input_tokens_seen": 20478568,
      "step": 31315
    },
    {
      "epoch": 16.41509433962264,
      "grad_norm": 0.2599973976612091,
      "learning_rate": 4.7372723509404025e-06,
      "loss": 0.3862,
      "num_input_tokens_seen": 20481640,
      "step": 31320
    },
    {
      "epoch": 16.417714884696018,
      "grad_norm": 0.8351848721504211,
      "learning_rate": 4.7305771121529615e-06,
      "loss": 0.4416,
      "num_input_tokens_seen": 20484680,
      "step": 31325
    },
    {
      "epoch": 16.42033542976939,
      "grad_norm": 0.28955766558647156,
      "learning_rate": 4.723886113497542e-06,
      "loss": 0.5071,
      "num_input_tokens_seen": 20491688,
      "step": 31330
    },
    {
      "epoch": 16.42295597484277,
      "grad_norm": 0.4238063097000122,
      "learning_rate": 4.7171993563738195e-06,
      "loss": 0.4343,
      "num_input_tokens_seen": 20494408,
      "step": 31335
    },
    {
      "epoch": 16.42557651991614,
      "grad_norm": 0.348652184009552,
      "learning_rate": 4.7105168421805924e-06,
      "loss": 0.5881,
      "num_input_tokens_seen": 20497672,
      "step": 31340
    },
    {
      "epoch": 16.42819706498952,
      "grad_norm": 0.28034117817878723,
      "learning_rate": 4.703838572315767e-06,
      "loss": 0.4002,
      "num_input_tokens_seen": 20501192,
      "step": 31345
    },
    {
      "epoch": 16.430817610062892,
      "grad_norm": 0.6673721671104431,
      "learning_rate": 4.697164548176367e-06,
      "loss": 0.6095,
      "num_input_tokens_seen": 20504168,
      "step": 31350
    },
    {
      "epoch": 16.43343815513627,
      "grad_norm": 0.4798484146595001,
      "learning_rate": 4.690494771158519e-06,
      "loss": 0.4428,
      "num_input_tokens_seen": 20507176,
      "step": 31355
    },
    {
      "epoch": 16.436058700209642,
      "grad_norm": 0.37991970777511597,
      "learning_rate": 4.683829242657453e-06,
      "loss": 0.4016,
      "num_input_tokens_seen": 20510344,
      "step": 31360
    },
    {
      "epoch": 16.43867924528302,
      "grad_norm": 0.5891596674919128,
      "learning_rate": 4.677167964067533e-06,
      "loss": 0.5314,
      "num_input_tokens_seen": 20512904,
      "step": 31365
    },
    {
      "epoch": 16.441299790356393,
      "grad_norm": 0.5025710463523865,
      "learning_rate": 4.670510936782216e-06,
      "loss": 0.4195,
      "num_input_tokens_seen": 20515720,
      "step": 31370
    },
    {
      "epoch": 16.44392033542977,
      "grad_norm": 0.7024523615837097,
      "learning_rate": 4.6638581621940815e-06,
      "loss": 0.5319,
      "num_input_tokens_seen": 20519080,
      "step": 31375
    },
    {
      "epoch": 16.446540880503143,
      "grad_norm": 0.6308234930038452,
      "learning_rate": 4.657209641694821e-06,
      "loss": 0.3877,
      "num_input_tokens_seen": 20522536,
      "step": 31380
    },
    {
      "epoch": 16.44916142557652,
      "grad_norm": 0.512829601764679,
      "learning_rate": 4.65056537667522e-06,
      "loss": 0.4107,
      "num_input_tokens_seen": 20525800,
      "step": 31385
    },
    {
      "epoch": 16.451781970649897,
      "grad_norm": 0.6245517134666443,
      "learning_rate": 4.643925368525176e-06,
      "loss": 0.3871,
      "num_input_tokens_seen": 20528168,
      "step": 31390
    },
    {
      "epoch": 16.45440251572327,
      "grad_norm": 0.41934752464294434,
      "learning_rate": 4.637289618633711e-06,
      "loss": 0.3554,
      "num_input_tokens_seen": 20530984,
      "step": 31395
    },
    {
      "epoch": 16.457023060796647,
      "grad_norm": 0.760557234287262,
      "learning_rate": 4.630658128388948e-06,
      "loss": 0.4178,
      "num_input_tokens_seen": 20533928,
      "step": 31400
    },
    {
      "epoch": 16.45964360587002,
      "grad_norm": 0.2775118350982666,
      "learning_rate": 4.624030899178125e-06,
      "loss": 0.393,
      "num_input_tokens_seen": 20536744,
      "step": 31405
    },
    {
      "epoch": 16.462264150943398,
      "grad_norm": 0.765544593334198,
      "learning_rate": 4.617407932387569e-06,
      "loss": 0.4284,
      "num_input_tokens_seen": 20539304,
      "step": 31410
    },
    {
      "epoch": 16.46488469601677,
      "grad_norm": 0.3440967798233032,
      "learning_rate": 4.610789229402743e-06,
      "loss": 0.4134,
      "num_input_tokens_seen": 20541896,
      "step": 31415
    },
    {
      "epoch": 16.467505241090148,
      "grad_norm": 1.2383049726486206,
      "learning_rate": 4.604174791608193e-06,
      "loss": 0.4798,
      "num_input_tokens_seen": 20545000,
      "step": 31420
    },
    {
      "epoch": 16.47012578616352,
      "grad_norm": 0.5755031108856201,
      "learning_rate": 4.597564620387587e-06,
      "loss": 0.4254,
      "num_input_tokens_seen": 20547432,
      "step": 31425
    },
    {
      "epoch": 16.4727463312369,
      "grad_norm": 1.166327714920044,
      "learning_rate": 4.590958717123708e-06,
      "loss": 0.4782,
      "num_input_tokens_seen": 20550376,
      "step": 31430
    },
    {
      "epoch": 16.47536687631027,
      "grad_norm": 0.22381821274757385,
      "learning_rate": 4.5843570831984174e-06,
      "loss": 0.4272,
      "num_input_tokens_seen": 20555496,
      "step": 31435
    },
    {
      "epoch": 16.47798742138365,
      "grad_norm": 0.4525332450866699,
      "learning_rate": 4.577759719992719e-06,
      "loss": 0.3759,
      "num_input_tokens_seen": 20559752,
      "step": 31440
    },
    {
      "epoch": 16.480607966457022,
      "grad_norm": 0.3850131630897522,
      "learning_rate": 4.571166628886692e-06,
      "loss": 0.3675,
      "num_input_tokens_seen": 20563464,
      "step": 31445
    },
    {
      "epoch": 16.4832285115304,
      "grad_norm": 0.6576008200645447,
      "learning_rate": 4.564577811259541e-06,
      "loss": 0.5467,
      "num_input_tokens_seen": 20566728,
      "step": 31450
    },
    {
      "epoch": 16.485849056603772,
      "grad_norm": 0.6382924318313599,
      "learning_rate": 4.557993268489574e-06,
      "loss": 0.4045,
      "num_input_tokens_seen": 20569416,
      "step": 31455
    },
    {
      "epoch": 16.48846960167715,
      "grad_norm": 0.31383243203163147,
      "learning_rate": 4.551413001954205e-06,
      "loss": 0.3086,
      "num_input_tokens_seen": 20572456,
      "step": 31460
    },
    {
      "epoch": 16.491090146750523,
      "grad_norm": 0.35696282982826233,
      "learning_rate": 4.54483701302994e-06,
      "loss": 0.4527,
      "num_input_tokens_seen": 20576168,
      "step": 31465
    },
    {
      "epoch": 16.4937106918239,
      "grad_norm": 0.4502219259738922,
      "learning_rate": 4.538265303092415e-06,
      "loss": 0.4188,
      "num_input_tokens_seen": 20579112,
      "step": 31470
    },
    {
      "epoch": 16.496331236897273,
      "grad_norm": 0.5306179523468018,
      "learning_rate": 4.531697873516338e-06,
      "loss": 0.53,
      "num_input_tokens_seen": 20582824,
      "step": 31475
    },
    {
      "epoch": 16.49895178197065,
      "grad_norm": 0.8440129160881042,
      "learning_rate": 4.525134725675556e-06,
      "loss": 0.4226,
      "num_input_tokens_seen": 20586088,
      "step": 31480
    },
    {
      "epoch": 16.501572327044027,
      "grad_norm": 0.8149371147155762,
      "learning_rate": 4.518575860943003e-06,
      "loss": 0.5668,
      "num_input_tokens_seen": 20591560,
      "step": 31485
    },
    {
      "epoch": 16.5041928721174,
      "grad_norm": 0.5382925271987915,
      "learning_rate": 4.512021280690712e-06,
      "loss": 0.4066,
      "num_input_tokens_seen": 20594952,
      "step": 31490
    },
    {
      "epoch": 16.506813417190777,
      "grad_norm": 0.42383235692977905,
      "learning_rate": 4.505470986289828e-06,
      "loss": 0.3348,
      "num_input_tokens_seen": 20597992,
      "step": 31495
    },
    {
      "epoch": 16.50943396226415,
      "grad_norm": 0.5178347826004028,
      "learning_rate": 4.498924979110605e-06,
      "loss": 0.4896,
      "num_input_tokens_seen": 20600616,
      "step": 31500
    },
    {
      "epoch": 16.512054507337528,
      "grad_norm": 0.5529770255088806,
      "learning_rate": 4.492383260522384e-06,
      "loss": 0.3588,
      "num_input_tokens_seen": 20603432,
      "step": 31505
    },
    {
      "epoch": 16.5146750524109,
      "grad_norm": 1.1861015558242798,
      "learning_rate": 4.485845831893626e-06,
      "loss": 0.6191,
      "num_input_tokens_seen": 20606280,
      "step": 31510
    },
    {
      "epoch": 16.517295597484278,
      "grad_norm": 0.345634788274765,
      "learning_rate": 4.479312694591878e-06,
      "loss": 0.3846,
      "num_input_tokens_seen": 20609832,
      "step": 31515
    },
    {
      "epoch": 16.51991614255765,
      "grad_norm": 0.5352827310562134,
      "learning_rate": 4.4727838499838e-06,
      "loss": 0.4637,
      "num_input_tokens_seen": 20613416,
      "step": 31520
    },
    {
      "epoch": 16.52253668763103,
      "grad_norm": 0.5315558314323425,
      "learning_rate": 4.4662592994351545e-06,
      "loss": 0.4274,
      "num_input_tokens_seen": 20616712,
      "step": 31525
    },
    {
      "epoch": 16.5251572327044,
      "grad_norm": 0.43833500146865845,
      "learning_rate": 4.4597390443108065e-06,
      "loss": 0.492,
      "num_input_tokens_seen": 20619880,
      "step": 31530
    },
    {
      "epoch": 16.52777777777778,
      "grad_norm": 0.42437389492988586,
      "learning_rate": 4.453223085974712e-06,
      "loss": 0.3531,
      "num_input_tokens_seen": 20622728,
      "step": 31535
    },
    {
      "epoch": 16.530398322851152,
      "grad_norm": 0.7352066040039062,
      "learning_rate": 4.4467114257899425e-06,
      "loss": 0.4801,
      "num_input_tokens_seen": 20625544,
      "step": 31540
    },
    {
      "epoch": 16.53301886792453,
      "grad_norm": 0.30337777733802795,
      "learning_rate": 4.44020406511865e-06,
      "loss": 0.4451,
      "num_input_tokens_seen": 20629928,
      "step": 31545
    },
    {
      "epoch": 16.535639412997902,
      "grad_norm": 1.0083436965942383,
      "learning_rate": 4.43370100532211e-06,
      "loss": 0.5161,
      "num_input_tokens_seen": 20633608,
      "step": 31550
    },
    {
      "epoch": 16.53825995807128,
      "grad_norm": 0.41194063425064087,
      "learning_rate": 4.427202247760695e-06,
      "loss": 0.4442,
      "num_input_tokens_seen": 20637192,
      "step": 31555
    },
    {
      "epoch": 16.540880503144653,
      "grad_norm": 0.6827096343040466,
      "learning_rate": 4.420707793793854e-06,
      "loss": 0.4295,
      "num_input_tokens_seen": 20640136,
      "step": 31560
    },
    {
      "epoch": 16.54350104821803,
      "grad_norm": 0.8204889893531799,
      "learning_rate": 4.414217644780172e-06,
      "loss": 0.4777,
      "num_input_tokens_seen": 20644424,
      "step": 31565
    },
    {
      "epoch": 16.546121593291403,
      "grad_norm": 0.4798848330974579,
      "learning_rate": 4.407731802077294e-06,
      "loss": 0.4381,
      "num_input_tokens_seen": 20647528,
      "step": 31570
    },
    {
      "epoch": 16.54874213836478,
      "grad_norm": 0.4854891896247864,
      "learning_rate": 4.401250267041996e-06,
      "loss": 0.4192,
      "num_input_tokens_seen": 20651080,
      "step": 31575
    },
    {
      "epoch": 16.551362683438157,
      "grad_norm": 0.4747120141983032,
      "learning_rate": 4.394773041030137e-06,
      "loss": 0.372,
      "num_input_tokens_seen": 20654344,
      "step": 31580
    },
    {
      "epoch": 16.55398322851153,
      "grad_norm": 0.3568144142627716,
      "learning_rate": 4.388300125396685e-06,
      "loss": 0.3752,
      "num_input_tokens_seen": 20658056,
      "step": 31585
    },
    {
      "epoch": 16.556603773584907,
      "grad_norm": 0.6248685121536255,
      "learning_rate": 4.381831521495699e-06,
      "loss": 0.3844,
      "num_input_tokens_seen": 20661544,
      "step": 31590
    },
    {
      "epoch": 16.55922431865828,
      "grad_norm": 0.2681707739830017,
      "learning_rate": 4.37536723068033e-06,
      "loss": 0.4481,
      "num_input_tokens_seen": 20665768,
      "step": 31595
    },
    {
      "epoch": 16.561844863731658,
      "grad_norm": 0.3353504240512848,
      "learning_rate": 4.368907254302837e-06,
      "loss": 0.3735,
      "num_input_tokens_seen": 20668840,
      "step": 31600
    },
    {
      "epoch": 16.56446540880503,
      "grad_norm": 1.1235297918319702,
      "learning_rate": 4.3624515937145725e-06,
      "loss": 0.7289,
      "num_input_tokens_seen": 20671720,
      "step": 31605
    },
    {
      "epoch": 16.567085953878408,
      "grad_norm": 0.4026556611061096,
      "learning_rate": 4.3560002502659905e-06,
      "loss": 0.4218,
      "num_input_tokens_seen": 20675336,
      "step": 31610
    },
    {
      "epoch": 16.56970649895178,
      "grad_norm": 0.4813430905342102,
      "learning_rate": 4.349553225306644e-06,
      "loss": 0.5017,
      "num_input_tokens_seen": 20678056,
      "step": 31615
    },
    {
      "epoch": 16.572327044025158,
      "grad_norm": 0.6157224774360657,
      "learning_rate": 4.343110520185167e-06,
      "loss": 0.5125,
      "num_input_tokens_seen": 20681256,
      "step": 31620
    },
    {
      "epoch": 16.57494758909853,
      "grad_norm": 0.6620267033576965,
      "learning_rate": 4.336672136249298e-06,
      "loss": 0.3951,
      "num_input_tokens_seen": 20687112,
      "step": 31625
    },
    {
      "epoch": 16.57756813417191,
      "grad_norm": 0.5974211096763611,
      "learning_rate": 4.33023807484588e-06,
      "loss": 0.5167,
      "num_input_tokens_seen": 20689960,
      "step": 31630
    },
    {
      "epoch": 16.580188679245282,
      "grad_norm": 0.6191466450691223,
      "learning_rate": 4.323808337320842e-06,
      "loss": 0.3875,
      "num_input_tokens_seen": 20694984,
      "step": 31635
    },
    {
      "epoch": 16.58280922431866,
      "grad_norm": 0.2813461720943451,
      "learning_rate": 4.31738292501922e-06,
      "loss": 0.4292,
      "num_input_tokens_seen": 20699272,
      "step": 31640
    },
    {
      "epoch": 16.585429769392032,
      "grad_norm": 0.426501989364624,
      "learning_rate": 4.31096183928513e-06,
      "loss": 0.4353,
      "num_input_tokens_seen": 20702632,
      "step": 31645
    },
    {
      "epoch": 16.58805031446541,
      "grad_norm": 0.6799039840698242,
      "learning_rate": 4.304545081461783e-06,
      "loss": 0.3908,
      "num_input_tokens_seen": 20706152,
      "step": 31650
    },
    {
      "epoch": 16.590670859538783,
      "grad_norm": 0.3756597340106964,
      "learning_rate": 4.298132652891501e-06,
      "loss": 0.4244,
      "num_input_tokens_seen": 20709416,
      "step": 31655
    },
    {
      "epoch": 16.59329140461216,
      "grad_norm": 0.4487130045890808,
      "learning_rate": 4.291724554915688e-06,
      "loss": 0.3872,
      "num_input_tokens_seen": 20713480,
      "step": 31660
    },
    {
      "epoch": 16.595911949685533,
      "grad_norm": 0.33140724897384644,
      "learning_rate": 4.285320788874853e-06,
      "loss": 0.4272,
      "num_input_tokens_seen": 20717832,
      "step": 31665
    },
    {
      "epoch": 16.59853249475891,
      "grad_norm": 0.5494775772094727,
      "learning_rate": 4.278921356108581e-06,
      "loss": 0.379,
      "num_input_tokens_seen": 20720872,
      "step": 31670
    },
    {
      "epoch": 16.601153039832283,
      "grad_norm": 0.36084291338920593,
      "learning_rate": 4.272526257955556e-06,
      "loss": 0.3385,
      "num_input_tokens_seen": 20724424,
      "step": 31675
    },
    {
      "epoch": 16.60377358490566,
      "grad_norm": 0.6831348538398743,
      "learning_rate": 4.266135495753567e-06,
      "loss": 0.385,
      "num_input_tokens_seen": 20727688,
      "step": 31680
    },
    {
      "epoch": 16.606394129979037,
      "grad_norm": 0.4095783531665802,
      "learning_rate": 4.259749070839491e-06,
      "loss": 0.6224,
      "num_input_tokens_seen": 20731016,
      "step": 31685
    },
    {
      "epoch": 16.60901467505241,
      "grad_norm": 0.464602530002594,
      "learning_rate": 4.253366984549289e-06,
      "loss": 0.5043,
      "num_input_tokens_seen": 20733448,
      "step": 31690
    },
    {
      "epoch": 16.611635220125788,
      "grad_norm": 0.6833150386810303,
      "learning_rate": 4.24698923821803e-06,
      "loss": 0.5125,
      "num_input_tokens_seen": 20736680,
      "step": 31695
    },
    {
      "epoch": 16.61425576519916,
      "grad_norm": 0.32606378197669983,
      "learning_rate": 4.240615833179861e-06,
      "loss": 0.3122,
      "num_input_tokens_seen": 20739688,
      "step": 31700
    },
    {
      "epoch": 16.616876310272538,
      "grad_norm": 0.5492217540740967,
      "learning_rate": 4.234246770768016e-06,
      "loss": 0.5634,
      "num_input_tokens_seen": 20743560,
      "step": 31705
    },
    {
      "epoch": 16.61949685534591,
      "grad_norm": 0.4852062165737152,
      "learning_rate": 4.227882052314841e-06,
      "loss": 0.4362,
      "num_input_tokens_seen": 20746920,
      "step": 31710
    },
    {
      "epoch": 16.622117400419288,
      "grad_norm": 0.43486303091049194,
      "learning_rate": 4.2215216791517575e-06,
      "loss": 0.3878,
      "num_input_tokens_seen": 20750312,
      "step": 31715
    },
    {
      "epoch": 16.62473794549266,
      "grad_norm": 0.43514636158943176,
      "learning_rate": 4.215165652609293e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 20753960,
      "step": 31720
    },
    {
      "epoch": 16.62735849056604,
      "grad_norm": 0.3037307560443878,
      "learning_rate": 4.208813974017048e-06,
      "loss": 0.4669,
      "num_input_tokens_seen": 20757192,
      "step": 31725
    },
    {
      "epoch": 16.629979035639412,
      "grad_norm": 0.5045249462127686,
      "learning_rate": 4.202466644703712e-06,
      "loss": 0.3863,
      "num_input_tokens_seen": 20760360,
      "step": 31730
    },
    {
      "epoch": 16.63259958071279,
      "grad_norm": 0.694100558757782,
      "learning_rate": 4.196123665997087e-06,
      "loss": 0.4709,
      "num_input_tokens_seen": 20763592,
      "step": 31735
    },
    {
      "epoch": 16.635220125786162,
      "grad_norm": 0.48249301314353943,
      "learning_rate": 4.189785039224045e-06,
      "loss": 0.4114,
      "num_input_tokens_seen": 20766344,
      "step": 31740
    },
    {
      "epoch": 16.63784067085954,
      "grad_norm": 0.36254599690437317,
      "learning_rate": 4.183450765710567e-06,
      "loss": 0.3759,
      "num_input_tokens_seen": 20770248,
      "step": 31745
    },
    {
      "epoch": 16.640461215932913,
      "grad_norm": 0.4307093918323517,
      "learning_rate": 4.177120846781693e-06,
      "loss": 0.4997,
      "num_input_tokens_seen": 20773832,
      "step": 31750
    },
    {
      "epoch": 16.64308176100629,
      "grad_norm": 0.3792460560798645,
      "learning_rate": 4.170795283761583e-06,
      "loss": 0.4492,
      "num_input_tokens_seen": 20776136,
      "step": 31755
    },
    {
      "epoch": 16.645702306079663,
      "grad_norm": 1.1148791313171387,
      "learning_rate": 4.164474077973462e-06,
      "loss": 0.4297,
      "num_input_tokens_seen": 20779272,
      "step": 31760
    },
    {
      "epoch": 16.64832285115304,
      "grad_norm": 0.33566850423812866,
      "learning_rate": 4.158157230739659e-06,
      "loss": 0.3199,
      "num_input_tokens_seen": 20782984,
      "step": 31765
    },
    {
      "epoch": 16.650943396226417,
      "grad_norm": 0.36788544058799744,
      "learning_rate": 4.151844743381589e-06,
      "loss": 0.4389,
      "num_input_tokens_seen": 20787208,
      "step": 31770
    },
    {
      "epoch": 16.65356394129979,
      "grad_norm": 0.5706894993782043,
      "learning_rate": 4.145536617219756e-06,
      "loss": 0.3685,
      "num_input_tokens_seen": 20790216,
      "step": 31775
    },
    {
      "epoch": 16.656184486373167,
      "grad_norm": 0.8672971129417419,
      "learning_rate": 4.139232853573735e-06,
      "loss": 0.5151,
      "num_input_tokens_seen": 20793288,
      "step": 31780
    },
    {
      "epoch": 16.65880503144654,
      "grad_norm": 0.4529690444469452,
      "learning_rate": 4.132933453762214e-06,
      "loss": 0.5141,
      "num_input_tokens_seen": 20795688,
      "step": 31785
    },
    {
      "epoch": 16.661425576519918,
      "grad_norm": 0.28449559211730957,
      "learning_rate": 4.126638419102946e-06,
      "loss": 0.32,
      "num_input_tokens_seen": 20799496,
      "step": 31790
    },
    {
      "epoch": 16.66404612159329,
      "grad_norm": 0.4753890633583069,
      "learning_rate": 4.120347750912784e-06,
      "loss": 0.4649,
      "num_input_tokens_seen": 20802280,
      "step": 31795
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 0.4857238829135895,
      "learning_rate": 4.1140614505076724e-06,
      "loss": 0.3758,
      "num_input_tokens_seen": 20805352,
      "step": 31800
    },
    {
      "epoch": 16.66928721174004,
      "grad_norm": 0.5326230525970459,
      "learning_rate": 4.10777951920262e-06,
      "loss": 0.5609,
      "num_input_tokens_seen": 20808840,
      "step": 31805
    },
    {
      "epoch": 16.671907756813418,
      "grad_norm": 0.7195467352867126,
      "learning_rate": 4.1015019583117455e-06,
      "loss": 0.4838,
      "num_input_tokens_seen": 20811240,
      "step": 31810
    },
    {
      "epoch": 16.67452830188679,
      "grad_norm": 0.5076587200164795,
      "learning_rate": 4.095228769148232e-06,
      "loss": 0.4709,
      "num_input_tokens_seen": 20814632,
      "step": 31815
    },
    {
      "epoch": 16.67714884696017,
      "grad_norm": 0.3878437280654907,
      "learning_rate": 4.088959953024368e-06,
      "loss": 0.4873,
      "num_input_tokens_seen": 20818120,
      "step": 31820
    },
    {
      "epoch": 16.679769392033542,
      "grad_norm": 0.4803708493709564,
      "learning_rate": 4.082695511251522e-06,
      "loss": 0.4742,
      "num_input_tokens_seen": 20820776,
      "step": 31825
    },
    {
      "epoch": 16.68238993710692,
      "grad_norm": 0.7650853991508484,
      "learning_rate": 4.076435445140131e-06,
      "loss": 0.4313,
      "num_input_tokens_seen": 20823944,
      "step": 31830
    },
    {
      "epoch": 16.685010482180292,
      "grad_norm": 0.6104885935783386,
      "learning_rate": 4.070179755999737e-06,
      "loss": 0.3976,
      "num_input_tokens_seen": 20827144,
      "step": 31835
    },
    {
      "epoch": 16.68763102725367,
      "grad_norm": 0.9874148368835449,
      "learning_rate": 4.063928445138962e-06,
      "loss": 0.484,
      "num_input_tokens_seen": 20830376,
      "step": 31840
    },
    {
      "epoch": 16.690251572327043,
      "grad_norm": 0.637245774269104,
      "learning_rate": 4.057681513865499e-06,
      "loss": 0.3547,
      "num_input_tokens_seen": 20834888,
      "step": 31845
    },
    {
      "epoch": 16.69287211740042,
      "grad_norm": 0.4415205717086792,
      "learning_rate": 4.0514389634861426e-06,
      "loss": 0.396,
      "num_input_tokens_seen": 20837992,
      "step": 31850
    },
    {
      "epoch": 16.695492662473793,
      "grad_norm": 0.4260045886039734,
      "learning_rate": 4.045200795306767e-06,
      "loss": 0.4053,
      "num_input_tokens_seen": 20841448,
      "step": 31855
    },
    {
      "epoch": 16.69811320754717,
      "grad_norm": 0.6097859740257263,
      "learning_rate": 4.038967010632314e-06,
      "loss": 0.5404,
      "num_input_tokens_seen": 20844296,
      "step": 31860
    },
    {
      "epoch": 16.700733752620543,
      "grad_norm": 0.6384140253067017,
      "learning_rate": 4.032737610766826e-06,
      "loss": 0.4078,
      "num_input_tokens_seen": 20847080,
      "step": 31865
    },
    {
      "epoch": 16.70335429769392,
      "grad_norm": 0.33979707956314087,
      "learning_rate": 4.026512597013429e-06,
      "loss": 0.4875,
      "num_input_tokens_seen": 20850984,
      "step": 31870
    },
    {
      "epoch": 16.705974842767297,
      "grad_norm": 0.6019536852836609,
      "learning_rate": 4.020291970674312e-06,
      "loss": 0.4805,
      "num_input_tokens_seen": 20854344,
      "step": 31875
    },
    {
      "epoch": 16.70859538784067,
      "grad_norm": 0.5369818806648254,
      "learning_rate": 4.0140757330507705e-06,
      "loss": 0.4154,
      "num_input_tokens_seen": 20856840,
      "step": 31880
    },
    {
      "epoch": 16.711215932914047,
      "grad_norm": 0.3987909257411957,
      "learning_rate": 4.007863885443161e-06,
      "loss": 0.3382,
      "num_input_tokens_seen": 20860296,
      "step": 31885
    },
    {
      "epoch": 16.71383647798742,
      "grad_norm": 0.3039112985134125,
      "learning_rate": 4.001656429150935e-06,
      "loss": 0.3395,
      "num_input_tokens_seen": 20863656,
      "step": 31890
    },
    {
      "epoch": 16.716457023060798,
      "grad_norm": 0.40252184867858887,
      "learning_rate": 3.9954533654726235e-06,
      "loss": 0.5453,
      "num_input_tokens_seen": 20867016,
      "step": 31895
    },
    {
      "epoch": 16.71907756813417,
      "grad_norm": 0.4408138394355774,
      "learning_rate": 3.989254695705841e-06,
      "loss": 0.5156,
      "num_input_tokens_seen": 20870952,
      "step": 31900
    },
    {
      "epoch": 16.721698113207548,
      "grad_norm": 0.5313143730163574,
      "learning_rate": 3.983060421147272e-06,
      "loss": 0.5101,
      "num_input_tokens_seen": 20874184,
      "step": 31905
    },
    {
      "epoch": 16.72431865828092,
      "grad_norm": 0.49881380796432495,
      "learning_rate": 3.976870543092684e-06,
      "loss": 0.4489,
      "num_input_tokens_seen": 20877544,
      "step": 31910
    },
    {
      "epoch": 16.7269392033543,
      "grad_norm": 0.6334901452064514,
      "learning_rate": 3.970685062836932e-06,
      "loss": 0.5894,
      "num_input_tokens_seen": 20880904,
      "step": 31915
    },
    {
      "epoch": 16.729559748427672,
      "grad_norm": 0.4889291822910309,
      "learning_rate": 3.96450398167395e-06,
      "loss": 0.4397,
      "num_input_tokens_seen": 20884200,
      "step": 31920
    },
    {
      "epoch": 16.73218029350105,
      "grad_norm": 0.33281904458999634,
      "learning_rate": 3.958327300896758e-06,
      "loss": 0.3129,
      "num_input_tokens_seen": 20888328,
      "step": 31925
    },
    {
      "epoch": 16.734800838574422,
      "grad_norm": 0.5373958945274353,
      "learning_rate": 3.952155021797429e-06,
      "loss": 0.4392,
      "num_input_tokens_seen": 20892648,
      "step": 31930
    },
    {
      "epoch": 16.7374213836478,
      "grad_norm": 0.7863662242889404,
      "learning_rate": 3.945987145667151e-06,
      "loss": 0.5307,
      "num_input_tokens_seen": 20896296,
      "step": 31935
    },
    {
      "epoch": 16.740041928721173,
      "grad_norm": 0.4071097671985626,
      "learning_rate": 3.939823673796162e-06,
      "loss": 0.5345,
      "num_input_tokens_seen": 20899592,
      "step": 31940
    },
    {
      "epoch": 16.74266247379455,
      "grad_norm": 0.5414491295814514,
      "learning_rate": 3.933664607473789e-06,
      "loss": 0.6021,
      "num_input_tokens_seen": 20902440,
      "step": 31945
    },
    {
      "epoch": 16.745283018867923,
      "grad_norm": 0.8076689839363098,
      "learning_rate": 3.927509947988445e-06,
      "loss": 0.4758,
      "num_input_tokens_seen": 20905896,
      "step": 31950
    },
    {
      "epoch": 16.7479035639413,
      "grad_norm": 1.0037966966629028,
      "learning_rate": 3.9213596966276164e-06,
      "loss": 0.467,
      "num_input_tokens_seen": 20908808,
      "step": 31955
    },
    {
      "epoch": 16.750524109014677,
      "grad_norm": 0.4297393560409546,
      "learning_rate": 3.9152138546778625e-06,
      "loss": 0.3592,
      "num_input_tokens_seen": 20911592,
      "step": 31960
    },
    {
      "epoch": 16.75314465408805,
      "grad_norm": 0.6301850080490112,
      "learning_rate": 3.9090724234248175e-06,
      "loss": 0.5082,
      "num_input_tokens_seen": 20914664,
      "step": 31965
    },
    {
      "epoch": 16.755765199161427,
      "grad_norm": 0.5587096810340881,
      "learning_rate": 3.902935404153202e-06,
      "loss": 0.4228,
      "num_input_tokens_seen": 20917800,
      "step": 31970
    },
    {
      "epoch": 16.7583857442348,
      "grad_norm": 0.4625309109687805,
      "learning_rate": 3.896802798146809e-06,
      "loss": 0.4818,
      "num_input_tokens_seen": 20920680,
      "step": 31975
    },
    {
      "epoch": 16.761006289308177,
      "grad_norm": 0.45363426208496094,
      "learning_rate": 3.890674606688524e-06,
      "loss": 0.4569,
      "num_input_tokens_seen": 20923400,
      "step": 31980
    },
    {
      "epoch": 16.76362683438155,
      "grad_norm": 0.6637516617774963,
      "learning_rate": 3.88455083106028e-06,
      "loss": 0.3198,
      "num_input_tokens_seen": 20925320,
      "step": 31985
    },
    {
      "epoch": 16.766247379454928,
      "grad_norm": 0.6071697473526001,
      "learning_rate": 3.878431472543098e-06,
      "loss": 0.4451,
      "num_input_tokens_seen": 20928328,
      "step": 31990
    },
    {
      "epoch": 16.7688679245283,
      "grad_norm": 0.40639400482177734,
      "learning_rate": 3.872316532417086e-06,
      "loss": 0.3731,
      "num_input_tokens_seen": 20931304,
      "step": 31995
    },
    {
      "epoch": 16.771488469601678,
      "grad_norm": 0.5383502840995789,
      "learning_rate": 3.866206011961418e-06,
      "loss": 0.3707,
      "num_input_tokens_seen": 20934376,
      "step": 32000
    },
    {
      "epoch": 16.77410901467505,
      "grad_norm": 0.3893052637577057,
      "learning_rate": 3.860099912454346e-06,
      "loss": 0.6108,
      "num_input_tokens_seen": 20937224,
      "step": 32005
    },
    {
      "epoch": 16.77672955974843,
      "grad_norm": 0.44704669713974,
      "learning_rate": 3.853998235173201e-06,
      "loss": 0.429,
      "num_input_tokens_seen": 20940616,
      "step": 32010
    },
    {
      "epoch": 16.779350104821802,
      "grad_norm": 0.3936467170715332,
      "learning_rate": 3.847900981394379e-06,
      "loss": 0.3461,
      "num_input_tokens_seen": 20944488,
      "step": 32015
    },
    {
      "epoch": 16.78197064989518,
      "grad_norm": 0.5709092617034912,
      "learning_rate": 3.841808152393348e-06,
      "loss": 0.4084,
      "num_input_tokens_seen": 20947336,
      "step": 32020
    },
    {
      "epoch": 16.784591194968552,
      "grad_norm": 0.6528704166412354,
      "learning_rate": 3.835719749444666e-06,
      "loss": 0.4691,
      "num_input_tokens_seen": 20950280,
      "step": 32025
    },
    {
      "epoch": 16.78721174004193,
      "grad_norm": 0.47422289848327637,
      "learning_rate": 3.8296357738219606e-06,
      "loss": 0.3827,
      "num_input_tokens_seen": 20953160,
      "step": 32030
    },
    {
      "epoch": 16.789832285115303,
      "grad_norm": 0.925137996673584,
      "learning_rate": 3.823556226797931e-06,
      "loss": 0.3535,
      "num_input_tokens_seen": 20957064,
      "step": 32035
    },
    {
      "epoch": 16.79245283018868,
      "grad_norm": 0.5141469240188599,
      "learning_rate": 3.817481109644344e-06,
      "loss": 0.4377,
      "num_input_tokens_seen": 20959848,
      "step": 32040
    },
    {
      "epoch": 16.795073375262053,
      "grad_norm": 0.5028659105300903,
      "learning_rate": 3.811410423632042e-06,
      "loss": 0.5384,
      "num_input_tokens_seen": 20962536,
      "step": 32045
    },
    {
      "epoch": 16.79769392033543,
      "grad_norm": 0.37828925251960754,
      "learning_rate": 3.8053441700309473e-06,
      "loss": 0.4885,
      "num_input_tokens_seen": 20966120,
      "step": 32050
    },
    {
      "epoch": 16.800314465408803,
      "grad_norm": 0.7963162064552307,
      "learning_rate": 3.7992823501100507e-06,
      "loss": 0.4514,
      "num_input_tokens_seen": 20969064,
      "step": 32055
    },
    {
      "epoch": 16.80293501048218,
      "grad_norm": 0.4160585403442383,
      "learning_rate": 3.7932249651374207e-06,
      "loss": 0.3757,
      "num_input_tokens_seen": 20971656,
      "step": 32060
    },
    {
      "epoch": 16.805555555555557,
      "grad_norm": 0.38243332505226135,
      "learning_rate": 3.787172016380183e-06,
      "loss": 0.3734,
      "num_input_tokens_seen": 20974568,
      "step": 32065
    },
    {
      "epoch": 16.80817610062893,
      "grad_norm": 0.07804720848798752,
      "learning_rate": 3.7811235051045558e-06,
      "loss": 0.384,
      "num_input_tokens_seen": 20981576,
      "step": 32070
    },
    {
      "epoch": 16.810796645702307,
      "grad_norm": 0.7903638482093811,
      "learning_rate": 3.775079432575809e-06,
      "loss": 0.4905,
      "num_input_tokens_seen": 20985032,
      "step": 32075
    },
    {
      "epoch": 16.81341719077568,
      "grad_norm": 0.2952955961227417,
      "learning_rate": 3.7690398000582964e-06,
      "loss": 0.5889,
      "num_input_tokens_seen": 20989416,
      "step": 32080
    },
    {
      "epoch": 16.816037735849058,
      "grad_norm": 0.4025990068912506,
      "learning_rate": 3.763004608815443e-06,
      "loss": 0.4895,
      "num_input_tokens_seen": 20993320,
      "step": 32085
    },
    {
      "epoch": 16.81865828092243,
      "grad_norm": 0.6293625235557556,
      "learning_rate": 3.756973860109747e-06,
      "loss": 0.5586,
      "num_input_tokens_seen": 20996584,
      "step": 32090
    },
    {
      "epoch": 16.821278825995808,
      "grad_norm": 0.295783132314682,
      "learning_rate": 3.7509475552027663e-06,
      "loss": 0.3671,
      "num_input_tokens_seen": 20999432,
      "step": 32095
    },
    {
      "epoch": 16.82389937106918,
      "grad_norm": 0.5988446474075317,
      "learning_rate": 3.744925695355131e-06,
      "loss": 0.3708,
      "num_input_tokens_seen": 21002120,
      "step": 32100
    },
    {
      "epoch": 16.82651991614256,
      "grad_norm": 0.39556214213371277,
      "learning_rate": 3.7389082818265476e-06,
      "loss": 0.4086,
      "num_input_tokens_seen": 21005544,
      "step": 32105
    },
    {
      "epoch": 16.829140461215932,
      "grad_norm": 0.574505090713501,
      "learning_rate": 3.732895315875792e-06,
      "loss": 0.4862,
      "num_input_tokens_seen": 21008552,
      "step": 32110
    },
    {
      "epoch": 16.83176100628931,
      "grad_norm": 0.39914464950561523,
      "learning_rate": 3.726886798760715e-06,
      "loss": 0.423,
      "num_input_tokens_seen": 21013160,
      "step": 32115
    },
    {
      "epoch": 16.834381551362682,
      "grad_norm": 0.8765600919723511,
      "learning_rate": 3.720882731738215e-06,
      "loss": 0.3362,
      "num_input_tokens_seen": 21016648,
      "step": 32120
    },
    {
      "epoch": 16.83700209643606,
      "grad_norm": 0.5180799961090088,
      "learning_rate": 3.714883116064291e-06,
      "loss": 0.4912,
      "num_input_tokens_seen": 21019080,
      "step": 32125
    },
    {
      "epoch": 16.839622641509433,
      "grad_norm": 0.3658171594142914,
      "learning_rate": 3.708887952993978e-06,
      "loss": 0.3365,
      "num_input_tokens_seen": 21021512,
      "step": 32130
    },
    {
      "epoch": 16.84224318658281,
      "grad_norm": 0.9059314727783203,
      "learning_rate": 3.7028972437814057e-06,
      "loss": 0.4025,
      "num_input_tokens_seen": 21024008,
      "step": 32135
    },
    {
      "epoch": 16.844863731656183,
      "grad_norm": 1.1464909315109253,
      "learning_rate": 3.696910989679764e-06,
      "loss": 0.4823,
      "num_input_tokens_seen": 21027784,
      "step": 32140
    },
    {
      "epoch": 16.84748427672956,
      "grad_norm": 0.624821126461029,
      "learning_rate": 3.6909291919413003e-06,
      "loss": 0.3649,
      "num_input_tokens_seen": 21031144,
      "step": 32145
    },
    {
      "epoch": 16.850104821802937,
      "grad_norm": 0.3638294041156769,
      "learning_rate": 3.684951851817342e-06,
      "loss": 0.5167,
      "num_input_tokens_seen": 21034504,
      "step": 32150
    },
    {
      "epoch": 16.85272536687631,
      "grad_norm": 0.4407685399055481,
      "learning_rate": 3.678978970558286e-06,
      "loss": 0.4753,
      "num_input_tokens_seen": 21037160,
      "step": 32155
    },
    {
      "epoch": 16.855345911949687,
      "grad_norm": 0.3359622061252594,
      "learning_rate": 3.673010549413583e-06,
      "loss": 0.3837,
      "num_input_tokens_seen": 21041000,
      "step": 32160
    },
    {
      "epoch": 16.85796645702306,
      "grad_norm": 0.42111822962760925,
      "learning_rate": 3.6670465896317592e-06,
      "loss": 0.3559,
      "num_input_tokens_seen": 21043688,
      "step": 32165
    },
    {
      "epoch": 16.860587002096437,
      "grad_norm": 0.5413587093353271,
      "learning_rate": 3.661087092460419e-06,
      "loss": 0.4179,
      "num_input_tokens_seen": 21047144,
      "step": 32170
    },
    {
      "epoch": 16.86320754716981,
      "grad_norm": 0.3810743987560272,
      "learning_rate": 3.655132059146202e-06,
      "loss": 0.5273,
      "num_input_tokens_seen": 21050056,
      "step": 32175
    },
    {
      "epoch": 16.865828092243188,
      "grad_norm": 0.6696029901504517,
      "learning_rate": 3.6491814909348526e-06,
      "loss": 0.5774,
      "num_input_tokens_seen": 21054472,
      "step": 32180
    },
    {
      "epoch": 16.86844863731656,
      "grad_norm": 0.5683014392852783,
      "learning_rate": 3.6432353890711424e-06,
      "loss": 0.4623,
      "num_input_tokens_seen": 21057256,
      "step": 32185
    },
    {
      "epoch": 16.871069182389938,
      "grad_norm": 0.545005738735199,
      "learning_rate": 3.6372937547989425e-06,
      "loss": 0.5297,
      "num_input_tokens_seen": 21060744,
      "step": 32190
    },
    {
      "epoch": 16.87368972746331,
      "grad_norm": 0.3831162750720978,
      "learning_rate": 3.631356589361173e-06,
      "loss": 0.5755,
      "num_input_tokens_seen": 21063720,
      "step": 32195
    },
    {
      "epoch": 16.87631027253669,
      "grad_norm": 0.40682151913642883,
      "learning_rate": 3.625423893999813e-06,
      "loss": 0.3819,
      "num_input_tokens_seen": 21067112,
      "step": 32200
    },
    {
      "epoch": 16.878930817610062,
      "grad_norm": 0.4071725904941559,
      "learning_rate": 3.6194956699559217e-06,
      "loss": 0.6127,
      "num_input_tokens_seen": 21071208,
      "step": 32205
    },
    {
      "epoch": 16.88155136268344,
      "grad_norm": 0.7206317186355591,
      "learning_rate": 3.6135719184696188e-06,
      "loss": 0.5283,
      "num_input_tokens_seen": 21073960,
      "step": 32210
    },
    {
      "epoch": 16.884171907756812,
      "grad_norm": 0.47164252400398254,
      "learning_rate": 3.6076526407800766e-06,
      "loss": 0.4186,
      "num_input_tokens_seen": 21077320,
      "step": 32215
    },
    {
      "epoch": 16.88679245283019,
      "grad_norm": 0.7941756248474121,
      "learning_rate": 3.601737838125552e-06,
      "loss": 0.4091,
      "num_input_tokens_seen": 21081128,
      "step": 32220
    },
    {
      "epoch": 16.889412997903563,
      "grad_norm": 0.37939417362213135,
      "learning_rate": 3.595827511743341e-06,
      "loss": 0.439,
      "num_input_tokens_seen": 21083848,
      "step": 32225
    },
    {
      "epoch": 16.89203354297694,
      "grad_norm": 0.4832371175289154,
      "learning_rate": 3.5899216628698217e-06,
      "loss": 0.6324,
      "num_input_tokens_seen": 21086248,
      "step": 32230
    },
    {
      "epoch": 16.894654088050313,
      "grad_norm": 0.5897116661071777,
      "learning_rate": 3.584020292740431e-06,
      "loss": 0.3999,
      "num_input_tokens_seen": 21088808,
      "step": 32235
    },
    {
      "epoch": 16.89727463312369,
      "grad_norm": 0.4707588255405426,
      "learning_rate": 3.578123402589675e-06,
      "loss": 0.407,
      "num_input_tokens_seen": 21091752,
      "step": 32240
    },
    {
      "epoch": 16.899895178197063,
      "grad_norm": 0.39076510071754456,
      "learning_rate": 3.5722309936511044e-06,
      "loss": 0.4422,
      "num_input_tokens_seen": 21095176,
      "step": 32245
    },
    {
      "epoch": 16.90251572327044,
      "grad_norm": 0.4966999888420105,
      "learning_rate": 3.5663430671573572e-06,
      "loss": 0.364,
      "num_input_tokens_seen": 21097896,
      "step": 32250
    },
    {
      "epoch": 16.905136268343817,
      "grad_norm": 0.5497102737426758,
      "learning_rate": 3.5604596243401038e-06,
      "loss": 0.4359,
      "num_input_tokens_seen": 21100872,
      "step": 32255
    },
    {
      "epoch": 16.90775681341719,
      "grad_norm": 0.6664517521858215,
      "learning_rate": 3.554580666430102e-06,
      "loss": 0.4147,
      "num_input_tokens_seen": 21103432,
      "step": 32260
    },
    {
      "epoch": 16.910377358490567,
      "grad_norm": 0.6383840441703796,
      "learning_rate": 3.5487061946571614e-06,
      "loss": 0.3747,
      "num_input_tokens_seen": 21106760,
      "step": 32265
    },
    {
      "epoch": 16.91299790356394,
      "grad_norm": 0.34237566590309143,
      "learning_rate": 3.542836210250161e-06,
      "loss": 0.4704,
      "num_input_tokens_seen": 21110312,
      "step": 32270
    },
    {
      "epoch": 16.915618448637318,
      "grad_norm": 0.4788872003555298,
      "learning_rate": 3.536970714437032e-06,
      "loss": 0.4689,
      "num_input_tokens_seen": 21113160,
      "step": 32275
    },
    {
      "epoch": 16.91823899371069,
      "grad_norm": 0.6784825921058655,
      "learning_rate": 3.5311097084447575e-06,
      "loss": 0.4285,
      "num_input_tokens_seen": 21117096,
      "step": 32280
    },
    {
      "epoch": 16.920859538784068,
      "grad_norm": 0.3365013301372528,
      "learning_rate": 3.5252531934994033e-06,
      "loss": 0.4406,
      "num_input_tokens_seen": 21120552,
      "step": 32285
    },
    {
      "epoch": 16.92348008385744,
      "grad_norm": 0.7690746784210205,
      "learning_rate": 3.5194011708260866e-06,
      "loss": 0.4484,
      "num_input_tokens_seen": 21123816,
      "step": 32290
    },
    {
      "epoch": 16.92610062893082,
      "grad_norm": 0.5574589371681213,
      "learning_rate": 3.513553641648984e-06,
      "loss": 0.5131,
      "num_input_tokens_seen": 21126568,
      "step": 32295
    },
    {
      "epoch": 16.928721174004192,
      "grad_norm": 0.3912583589553833,
      "learning_rate": 3.5077106071913307e-06,
      "loss": 0.4845,
      "num_input_tokens_seen": 21129416,
      "step": 32300
    },
    {
      "epoch": 16.93134171907757,
      "grad_norm": 0.3378911316394806,
      "learning_rate": 3.5018720686754133e-06,
      "loss": 0.3607,
      "num_input_tokens_seen": 21132104,
      "step": 32305
    },
    {
      "epoch": 16.933962264150942,
      "grad_norm": 0.46800053119659424,
      "learning_rate": 3.4960380273225985e-06,
      "loss": 0.4556,
      "num_input_tokens_seen": 21135464,
      "step": 32310
    },
    {
      "epoch": 16.93658280922432,
      "grad_norm": 0.5876580476760864,
      "learning_rate": 3.490208484353297e-06,
      "loss": 0.4458,
      "num_input_tokens_seen": 21139144,
      "step": 32315
    },
    {
      "epoch": 16.939203354297693,
      "grad_norm": 0.5343331098556519,
      "learning_rate": 3.4843834409869824e-06,
      "loss": 0.5672,
      "num_input_tokens_seen": 21143016,
      "step": 32320
    },
    {
      "epoch": 16.94182389937107,
      "grad_norm": 0.7781245708465576,
      "learning_rate": 3.4785628984421977e-06,
      "loss": 0.444,
      "num_input_tokens_seen": 21145864,
      "step": 32325
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 0.5395739674568176,
      "learning_rate": 3.4727468579365235e-06,
      "loss": 0.5434,
      "num_input_tokens_seen": 21148904,
      "step": 32330
    },
    {
      "epoch": 16.94706498951782,
      "grad_norm": 0.5015100240707397,
      "learning_rate": 3.466935320686604e-06,
      "loss": 0.3551,
      "num_input_tokens_seen": 21151496,
      "step": 32335
    },
    {
      "epoch": 16.949685534591197,
      "grad_norm": 0.3947104215621948,
      "learning_rate": 3.4611282879081553e-06,
      "loss": 0.4244,
      "num_input_tokens_seen": 21154600,
      "step": 32340
    },
    {
      "epoch": 16.95230607966457,
      "grad_norm": 0.330318808555603,
      "learning_rate": 3.455325760815939e-06,
      "loss": 0.4821,
      "num_input_tokens_seen": 21157576,
      "step": 32345
    },
    {
      "epoch": 16.954926624737947,
      "grad_norm": 1.6356123685836792,
      "learning_rate": 3.4495277406237843e-06,
      "loss": 0.5069,
      "num_input_tokens_seen": 21160072,
      "step": 32350
    },
    {
      "epoch": 16.95754716981132,
      "grad_norm": 0.6129156351089478,
      "learning_rate": 3.443734228544565e-06,
      "loss": 0.4913,
      "num_input_tokens_seen": 21162856,
      "step": 32355
    },
    {
      "epoch": 16.960167714884697,
      "grad_norm": 0.5114707350730896,
      "learning_rate": 3.437945225790215e-06,
      "loss": 0.5191,
      "num_input_tokens_seen": 21165448,
      "step": 32360
    },
    {
      "epoch": 16.96278825995807,
      "grad_norm": 0.5074948072433472,
      "learning_rate": 3.432160733571729e-06,
      "loss": 0.4504,
      "num_input_tokens_seen": 21168840,
      "step": 32365
    },
    {
      "epoch": 16.965408805031448,
      "grad_norm": 0.5760220885276794,
      "learning_rate": 3.4263807530991595e-06,
      "loss": 0.4826,
      "num_input_tokens_seen": 21171592,
      "step": 32370
    },
    {
      "epoch": 16.96802935010482,
      "grad_norm": 0.34214943647384644,
      "learning_rate": 3.4206052855816155e-06,
      "loss": 0.4465,
      "num_input_tokens_seen": 21176168,
      "step": 32375
    },
    {
      "epoch": 16.970649895178198,
      "grad_norm": 0.5374078154563904,
      "learning_rate": 3.4148343322272567e-06,
      "loss": 0.4299,
      "num_input_tokens_seen": 21178632,
      "step": 32380
    },
    {
      "epoch": 16.97327044025157,
      "grad_norm": 0.4409995675086975,
      "learning_rate": 3.4090678942432923e-06,
      "loss": 0.4328,
      "num_input_tokens_seen": 21182184,
      "step": 32385
    },
    {
      "epoch": 16.97589098532495,
      "grad_norm": 0.32619422674179077,
      "learning_rate": 3.4033059728360018e-06,
      "loss": 0.344,
      "num_input_tokens_seen": 21187656,
      "step": 32390
    },
    {
      "epoch": 16.978511530398322,
      "grad_norm": 0.30898573994636536,
      "learning_rate": 3.397548569210715e-06,
      "loss": 0.4592,
      "num_input_tokens_seen": 21190952,
      "step": 32395
    },
    {
      "epoch": 16.9811320754717,
      "grad_norm": 0.6356412768363953,
      "learning_rate": 3.3917956845718136e-06,
      "loss": 0.4593,
      "num_input_tokens_seen": 21193640,
      "step": 32400
    },
    {
      "epoch": 16.983752620545072,
      "grad_norm": 0.3882860839366913,
      "learning_rate": 3.38604732012274e-06,
      "loss": 0.4409,
      "num_input_tokens_seen": 21197160,
      "step": 32405
    },
    {
      "epoch": 16.98637316561845,
      "grad_norm": 0.5778310894966125,
      "learning_rate": 3.3803034770659825e-06,
      "loss": 0.393,
      "num_input_tokens_seen": 21200296,
      "step": 32410
    },
    {
      "epoch": 16.988993710691823,
      "grad_norm": 0.2109857052564621,
      "learning_rate": 3.3745641566030826e-06,
      "loss": 0.4719,
      "num_input_tokens_seen": 21203528,
      "step": 32415
    },
    {
      "epoch": 16.9916142557652,
      "grad_norm": 0.5029614567756653,
      "learning_rate": 3.3688293599346442e-06,
      "loss": 0.4256,
      "num_input_tokens_seen": 21206600,
      "step": 32420
    },
    {
      "epoch": 16.994234800838573,
      "grad_norm": 0.6869811415672302,
      "learning_rate": 3.3630990882603213e-06,
      "loss": 0.5183,
      "num_input_tokens_seen": 21209928,
      "step": 32425
    },
    {
      "epoch": 16.99685534591195,
      "grad_norm": 0.44933754205703735,
      "learning_rate": 3.35737334277883e-06,
      "loss": 0.5572,
      "num_input_tokens_seen": 21213896,
      "step": 32430
    },
    {
      "epoch": 16.999475890985323,
      "grad_norm": 0.6326422691345215,
      "learning_rate": 3.351652124687918e-06,
      "loss": 0.493,
      "num_input_tokens_seen": 21217000,
      "step": 32435
    },
    {
      "epoch": 17.0,
      "eval_loss": 0.47668150067329407,
      "eval_runtime": 9.7383,
      "eval_samples_per_second": 87.079,
      "eval_steps_per_second": 21.77,
      "num_input_tokens_seen": 21217120,
      "step": 32436
    },
    {
      "epoch": 17.0020964360587,
      "grad_norm": 0.45749732851982117,
      "learning_rate": 3.345935435184408e-06,
      "loss": 0.6168,
      "num_input_tokens_seen": 21219808,
      "step": 32440
    },
    {
      "epoch": 17.004716981132077,
      "grad_norm": 0.6450797319412231,
      "learning_rate": 3.3402232754641576e-06,
      "loss": 0.468,
      "num_input_tokens_seen": 21222432,
      "step": 32445
    },
    {
      "epoch": 17.00733752620545,
      "grad_norm": 0.5168589949607849,
      "learning_rate": 3.334515646722089e-06,
      "loss": 0.5685,
      "num_input_tokens_seen": 21225856,
      "step": 32450
    },
    {
      "epoch": 17.009958071278827,
      "grad_norm": 0.6203747987747192,
      "learning_rate": 3.32881255015218e-06,
      "loss": 0.5064,
      "num_input_tokens_seen": 21228672,
      "step": 32455
    },
    {
      "epoch": 17.0125786163522,
      "grad_norm": 0.4356999099254608,
      "learning_rate": 3.3231139869474433e-06,
      "loss": 0.3367,
      "num_input_tokens_seen": 21231360,
      "step": 32460
    },
    {
      "epoch": 17.015199161425578,
      "grad_norm": 0.581614077091217,
      "learning_rate": 3.317419958299961e-06,
      "loss": 0.3732,
      "num_input_tokens_seen": 21233920,
      "step": 32465
    },
    {
      "epoch": 17.01781970649895,
      "grad_norm": 0.5776656270027161,
      "learning_rate": 3.31173046540085e-06,
      "loss": 0.4342,
      "num_input_tokens_seen": 21236576,
      "step": 32470
    },
    {
      "epoch": 17.020440251572328,
      "grad_norm": 0.5271552801132202,
      "learning_rate": 3.306045509440295e-06,
      "loss": 0.3733,
      "num_input_tokens_seen": 21239488,
      "step": 32475
    },
    {
      "epoch": 17.0230607966457,
      "grad_norm": 0.3463214933872223,
      "learning_rate": 3.300365091607521e-06,
      "loss": 0.3914,
      "num_input_tokens_seen": 21242880,
      "step": 32480
    },
    {
      "epoch": 17.02568134171908,
      "grad_norm": 0.3720220923423767,
      "learning_rate": 3.2946892130908127e-06,
      "loss": 0.4538,
      "num_input_tokens_seen": 21246240,
      "step": 32485
    },
    {
      "epoch": 17.028301886792452,
      "grad_norm": 0.6331580877304077,
      "learning_rate": 3.289017875077488e-06,
      "loss": 0.5194,
      "num_input_tokens_seen": 21249280,
      "step": 32490
    },
    {
      "epoch": 17.03092243186583,
      "grad_norm": 0.37895575165748596,
      "learning_rate": 3.283351078753938e-06,
      "loss": 0.4538,
      "num_input_tokens_seen": 21252800,
      "step": 32495
    },
    {
      "epoch": 17.033542976939202,
      "grad_norm": 0.4728778898715973,
      "learning_rate": 3.2776888253055777e-06,
      "loss": 0.3893,
      "num_input_tokens_seen": 21256160,
      "step": 32500
    },
    {
      "epoch": 17.03616352201258,
      "grad_norm": 0.5798147916793823,
      "learning_rate": 3.272031115916896e-06,
      "loss": 0.4492,
      "num_input_tokens_seen": 21259136,
      "step": 32505
    },
    {
      "epoch": 17.038784067085953,
      "grad_norm": 0.7979142665863037,
      "learning_rate": 3.2663779517714238e-06,
      "loss": 0.5023,
      "num_input_tokens_seen": 21261792,
      "step": 32510
    },
    {
      "epoch": 17.04140461215933,
      "grad_norm": 0.288604199886322,
      "learning_rate": 3.260729334051729e-06,
      "loss": 0.4556,
      "num_input_tokens_seen": 21265376,
      "step": 32515
    },
    {
      "epoch": 17.044025157232703,
      "grad_norm": 0.45081615447998047,
      "learning_rate": 3.2550852639394446e-06,
      "loss": 0.387,
      "num_input_tokens_seen": 21269056,
      "step": 32520
    },
    {
      "epoch": 17.04664570230608,
      "grad_norm": 0.35238975286483765,
      "learning_rate": 3.2494457426152452e-06,
      "loss": 0.3781,
      "num_input_tokens_seen": 21272416,
      "step": 32525
    },
    {
      "epoch": 17.049266247379453,
      "grad_norm": 0.48442158102989197,
      "learning_rate": 3.243810771258851e-06,
      "loss": 0.422,
      "num_input_tokens_seen": 21275904,
      "step": 32530
    },
    {
      "epoch": 17.05188679245283,
      "grad_norm": 0.46948015689849854,
      "learning_rate": 3.2381803510490387e-06,
      "loss": 0.4155,
      "num_input_tokens_seen": 21282112,
      "step": 32535
    },
    {
      "epoch": 17.054507337526207,
      "grad_norm": 0.6011528372764587,
      "learning_rate": 3.2325544831636217e-06,
      "loss": 0.3626,
      "num_input_tokens_seen": 21285056,
      "step": 32540
    },
    {
      "epoch": 17.05712788259958,
      "grad_norm": 0.6515294313430786,
      "learning_rate": 3.22693316877947e-06,
      "loss": 0.2762,
      "num_input_tokens_seen": 21288224,
      "step": 32545
    },
    {
      "epoch": 17.059748427672957,
      "grad_norm": 0.42934364080429077,
      "learning_rate": 3.2213164090725057e-06,
      "loss": 0.4272,
      "num_input_tokens_seen": 21291616,
      "step": 32550
    },
    {
      "epoch": 17.06236897274633,
      "grad_norm": 0.4300048053264618,
      "learning_rate": 3.2157042052176763e-06,
      "loss": 0.273,
      "num_input_tokens_seen": 21294976,
      "step": 32555
    },
    {
      "epoch": 17.064989517819708,
      "grad_norm": 0.6649267077445984,
      "learning_rate": 3.2100965583890003e-06,
      "loss": 0.3377,
      "num_input_tokens_seen": 21297824,
      "step": 32560
    },
    {
      "epoch": 17.06761006289308,
      "grad_norm": 0.6807359457015991,
      "learning_rate": 3.204493469759537e-06,
      "loss": 0.6185,
      "num_input_tokens_seen": 21300832,
      "step": 32565
    },
    {
      "epoch": 17.070230607966458,
      "grad_norm": 0.45845863223075867,
      "learning_rate": 3.1988949405013764e-06,
      "loss": 0.539,
      "num_input_tokens_seen": 21304960,
      "step": 32570
    },
    {
      "epoch": 17.07285115303983,
      "grad_norm": 0.43052801489830017,
      "learning_rate": 3.193300971785676e-06,
      "loss": 0.3956,
      "num_input_tokens_seen": 21308128,
      "step": 32575
    },
    {
      "epoch": 17.07547169811321,
      "grad_norm": 0.4223751127719879,
      "learning_rate": 3.1877115647826335e-06,
      "loss": 0.4581,
      "num_input_tokens_seen": 21311392,
      "step": 32580
    },
    {
      "epoch": 17.078092243186582,
      "grad_norm": 0.813983678817749,
      "learning_rate": 3.1821267206614797e-06,
      "loss": 0.3357,
      "num_input_tokens_seen": 21314016,
      "step": 32585
    },
    {
      "epoch": 17.08071278825996,
      "grad_norm": 0.7479026317596436,
      "learning_rate": 3.176546440590508e-06,
      "loss": 0.4935,
      "num_input_tokens_seen": 21317120,
      "step": 32590
    },
    {
      "epoch": 17.083333333333332,
      "grad_norm": 0.46018660068511963,
      "learning_rate": 3.170970725737041e-06,
      "loss": 0.4043,
      "num_input_tokens_seen": 21320384,
      "step": 32595
    },
    {
      "epoch": 17.08595387840671,
      "grad_norm": 0.26338034868240356,
      "learning_rate": 3.165399577267461e-06,
      "loss": 0.3784,
      "num_input_tokens_seen": 21323040,
      "step": 32600
    },
    {
      "epoch": 17.088574423480082,
      "grad_norm": 0.6445765495300293,
      "learning_rate": 3.159832996347184e-06,
      "loss": 0.4563,
      "num_input_tokens_seen": 21326176,
      "step": 32605
    },
    {
      "epoch": 17.09119496855346,
      "grad_norm": 0.4083012342453003,
      "learning_rate": 3.1542709841406865e-06,
      "loss": 0.4082,
      "num_input_tokens_seen": 21329376,
      "step": 32610
    },
    {
      "epoch": 17.093815513626833,
      "grad_norm": 0.5161946415901184,
      "learning_rate": 3.148713541811468e-06,
      "loss": 0.5302,
      "num_input_tokens_seen": 21332480,
      "step": 32615
    },
    {
      "epoch": 17.09643605870021,
      "grad_norm": 0.9101777672767639,
      "learning_rate": 3.1431606705220803e-06,
      "loss": 0.4589,
      "num_input_tokens_seen": 21336672,
      "step": 32620
    },
    {
      "epoch": 17.099056603773583,
      "grad_norm": 0.5885904431343079,
      "learning_rate": 3.1376123714341234e-06,
      "loss": 0.5859,
      "num_input_tokens_seen": 21339840,
      "step": 32625
    },
    {
      "epoch": 17.10167714884696,
      "grad_norm": 0.8919711112976074,
      "learning_rate": 3.1320686457082387e-06,
      "loss": 0.3771,
      "num_input_tokens_seen": 21342432,
      "step": 32630
    },
    {
      "epoch": 17.104297693920337,
      "grad_norm": 0.3080146312713623,
      "learning_rate": 3.126529494504113e-06,
      "loss": 0.4043,
      "num_input_tokens_seen": 21347456,
      "step": 32635
    },
    {
      "epoch": 17.10691823899371,
      "grad_norm": 0.8554834127426147,
      "learning_rate": 3.1209949189804744e-06,
      "loss": 0.3793,
      "num_input_tokens_seen": 21350816,
      "step": 32640
    },
    {
      "epoch": 17.109538784067087,
      "grad_norm": 0.7616277933120728,
      "learning_rate": 3.115464920295089e-06,
      "loss": 0.4547,
      "num_input_tokens_seen": 21354944,
      "step": 32645
    },
    {
      "epoch": 17.11215932914046,
      "grad_norm": 0.6282438635826111,
      "learning_rate": 3.1099394996047666e-06,
      "loss": 0.4331,
      "num_input_tokens_seen": 21358080,
      "step": 32650
    },
    {
      "epoch": 17.114779874213838,
      "grad_norm": 0.49628564715385437,
      "learning_rate": 3.104418658065367e-06,
      "loss": 0.4893,
      "num_input_tokens_seen": 21361152,
      "step": 32655
    },
    {
      "epoch": 17.11740041928721,
      "grad_norm": 0.8009805083274841,
      "learning_rate": 3.098902396831782e-06,
      "loss": 0.3377,
      "num_input_tokens_seen": 21363904,
      "step": 32660
    },
    {
      "epoch": 17.120020964360588,
      "grad_norm": 0.515153169631958,
      "learning_rate": 3.0933907170579646e-06,
      "loss": 0.5422,
      "num_input_tokens_seen": 21367392,
      "step": 32665
    },
    {
      "epoch": 17.12264150943396,
      "grad_norm": 0.45803913474082947,
      "learning_rate": 3.0878836198968835e-06,
      "loss": 0.4427,
      "num_input_tokens_seen": 21370272,
      "step": 32670
    },
    {
      "epoch": 17.12526205450734,
      "grad_norm": 0.402402400970459,
      "learning_rate": 3.0823811065005564e-06,
      "loss": 0.4102,
      "num_input_tokens_seen": 21373760,
      "step": 32675
    },
    {
      "epoch": 17.127882599580712,
      "grad_norm": 0.39524373412132263,
      "learning_rate": 3.0768831780200542e-06,
      "loss": 0.3615,
      "num_input_tokens_seen": 21378080,
      "step": 32680
    },
    {
      "epoch": 17.13050314465409,
      "grad_norm": 0.3939753472805023,
      "learning_rate": 3.071389835605476e-06,
      "loss": 0.5728,
      "num_input_tokens_seen": 21381440,
      "step": 32685
    },
    {
      "epoch": 17.133123689727462,
      "grad_norm": 0.6480967402458191,
      "learning_rate": 3.065901080405978e-06,
      "loss": 0.394,
      "num_input_tokens_seen": 21384128,
      "step": 32690
    },
    {
      "epoch": 17.13574423480084,
      "grad_norm": 0.6745597124099731,
      "learning_rate": 3.060416913569733e-06,
      "loss": 0.6318,
      "num_input_tokens_seen": 21386688,
      "step": 32695
    },
    {
      "epoch": 17.138364779874212,
      "grad_norm": 0.3830746114253998,
      "learning_rate": 3.054937336243968e-06,
      "loss": 0.3514,
      "num_input_tokens_seen": 21390208,
      "step": 32700
    },
    {
      "epoch": 17.14098532494759,
      "grad_norm": 0.1584429293870926,
      "learning_rate": 3.049462349574947e-06,
      "loss": 0.4167,
      "num_input_tokens_seen": 21394944,
      "step": 32705
    },
    {
      "epoch": 17.143605870020963,
      "grad_norm": 0.3902469575405121,
      "learning_rate": 3.043991954707981e-06,
      "loss": 0.4789,
      "num_input_tokens_seen": 21397760,
      "step": 32710
    },
    {
      "epoch": 17.14622641509434,
      "grad_norm": 0.7039769291877747,
      "learning_rate": 3.0385261527874075e-06,
      "loss": 0.3686,
      "num_input_tokens_seen": 21401024,
      "step": 32715
    },
    {
      "epoch": 17.148846960167713,
      "grad_norm": 0.27782994508743286,
      "learning_rate": 3.0330649449566205e-06,
      "loss": 0.3482,
      "num_input_tokens_seen": 21403712,
      "step": 32720
    },
    {
      "epoch": 17.15146750524109,
      "grad_norm": 0.44709712266921997,
      "learning_rate": 3.027608332358034e-06,
      "loss": 0.5711,
      "num_input_tokens_seen": 21407232,
      "step": 32725
    },
    {
      "epoch": 17.154088050314467,
      "grad_norm": 0.6305006146430969,
      "learning_rate": 3.022156316133104e-06,
      "loss": 0.4088,
      "num_input_tokens_seen": 21410112,
      "step": 32730
    },
    {
      "epoch": 17.15670859538784,
      "grad_norm": 0.3803694248199463,
      "learning_rate": 3.016708897422335e-06,
      "loss": 0.395,
      "num_input_tokens_seen": 21413792,
      "step": 32735
    },
    {
      "epoch": 17.159329140461217,
      "grad_norm": 0.47170451283454895,
      "learning_rate": 3.0112660773652685e-06,
      "loss": 0.3824,
      "num_input_tokens_seen": 21417440,
      "step": 32740
    },
    {
      "epoch": 17.16194968553459,
      "grad_norm": 0.40401002764701843,
      "learning_rate": 3.005827857100482e-06,
      "loss": 0.4281,
      "num_input_tokens_seen": 21421920,
      "step": 32745
    },
    {
      "epoch": 17.164570230607968,
      "grad_norm": 0.6739687919616699,
      "learning_rate": 3.000394237765586e-06,
      "loss": 0.4951,
      "num_input_tokens_seen": 21425280,
      "step": 32750
    },
    {
      "epoch": 17.16719077568134,
      "grad_norm": 0.3620153069496155,
      "learning_rate": 2.9949652204972254e-06,
      "loss": 0.3626,
      "num_input_tokens_seen": 21430240,
      "step": 32755
    },
    {
      "epoch": 17.169811320754718,
      "grad_norm": 0.5923346281051636,
      "learning_rate": 2.989540806431093e-06,
      "loss": 0.4231,
      "num_input_tokens_seen": 21433952,
      "step": 32760
    },
    {
      "epoch": 17.17243186582809,
      "grad_norm": 0.3661864399909973,
      "learning_rate": 2.984120996701917e-06,
      "loss": 0.4294,
      "num_input_tokens_seen": 21437312,
      "step": 32765
    },
    {
      "epoch": 17.17505241090147,
      "grad_norm": 0.7895545363426208,
      "learning_rate": 2.9787057924434618e-06,
      "loss": 0.3956,
      "num_input_tokens_seen": 21439968,
      "step": 32770
    },
    {
      "epoch": 17.177672955974842,
      "grad_norm": 1.5363097190856934,
      "learning_rate": 2.9732951947885166e-06,
      "loss": 0.3636,
      "num_input_tokens_seen": 21442848,
      "step": 32775
    },
    {
      "epoch": 17.18029350104822,
      "grad_norm": 0.35134440660476685,
      "learning_rate": 2.967889204868929e-06,
      "loss": 0.5564,
      "num_input_tokens_seen": 21446496,
      "step": 32780
    },
    {
      "epoch": 17.182914046121592,
      "grad_norm": 0.5263739824295044,
      "learning_rate": 2.96248782381556e-06,
      "loss": 0.4807,
      "num_input_tokens_seen": 21449856,
      "step": 32785
    },
    {
      "epoch": 17.18553459119497,
      "grad_norm": 0.3794224262237549,
      "learning_rate": 2.957091052758321e-06,
      "loss": 0.4835,
      "num_input_tokens_seen": 21454144,
      "step": 32790
    },
    {
      "epoch": 17.188155136268342,
      "grad_norm": 0.4258192479610443,
      "learning_rate": 2.9516988928261517e-06,
      "loss": 0.4263,
      "num_input_tokens_seen": 21458016,
      "step": 32795
    },
    {
      "epoch": 17.19077568134172,
      "grad_norm": 0.46688610315322876,
      "learning_rate": 2.9463113451470403e-06,
      "loss": 0.5242,
      "num_input_tokens_seen": 21461024,
      "step": 32800
    },
    {
      "epoch": 17.193396226415093,
      "grad_norm": 0.4780445694923401,
      "learning_rate": 2.94092841084799e-06,
      "loss": 0.3387,
      "num_input_tokens_seen": 21464064,
      "step": 32805
    },
    {
      "epoch": 17.19601677148847,
      "grad_norm": 0.469430148601532,
      "learning_rate": 2.9355500910550586e-06,
      "loss": 0.469,
      "num_input_tokens_seen": 21467552,
      "step": 32810
    },
    {
      "epoch": 17.198637316561843,
      "grad_norm": 0.33067354559898376,
      "learning_rate": 2.9301763868933157e-06,
      "loss": 0.3642,
      "num_input_tokens_seen": 21470432,
      "step": 32815
    },
    {
      "epoch": 17.20125786163522,
      "grad_norm": 0.5964471101760864,
      "learning_rate": 2.9248072994868896e-06,
      "loss": 0.4018,
      "num_input_tokens_seen": 21473888,
      "step": 32820
    },
    {
      "epoch": 17.203878406708597,
      "grad_norm": 1.7244162559509277,
      "learning_rate": 2.919442829958932e-06,
      "loss": 0.6494,
      "num_input_tokens_seen": 21476608,
      "step": 32825
    },
    {
      "epoch": 17.20649895178197,
      "grad_norm": 0.6258733868598938,
      "learning_rate": 2.9140829794316234e-06,
      "loss": 0.3794,
      "num_input_tokens_seen": 21479488,
      "step": 32830
    },
    {
      "epoch": 17.209119496855347,
      "grad_norm": 0.3226805329322815,
      "learning_rate": 2.908727749026191e-06,
      "loss": 0.5309,
      "num_input_tokens_seen": 21482816,
      "step": 32835
    },
    {
      "epoch": 17.21174004192872,
      "grad_norm": 0.5012903809547424,
      "learning_rate": 2.9033771398628756e-06,
      "loss": 0.4216,
      "num_input_tokens_seen": 21486400,
      "step": 32840
    },
    {
      "epoch": 17.214360587002098,
      "grad_norm": 0.4709845781326294,
      "learning_rate": 2.898031153060973e-06,
      "loss": 0.4477,
      "num_input_tokens_seen": 21489312,
      "step": 32845
    },
    {
      "epoch": 17.21698113207547,
      "grad_norm": 0.39408373832702637,
      "learning_rate": 2.8926897897388033e-06,
      "loss": 0.357,
      "num_input_tokens_seen": 21492256,
      "step": 32850
    },
    {
      "epoch": 17.219601677148848,
      "grad_norm": 0.38814401626586914,
      "learning_rate": 2.8873530510137136e-06,
      "loss": 0.3693,
      "num_input_tokens_seen": 21495424,
      "step": 32855
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 0.7919265627861023,
      "learning_rate": 2.8820209380020892e-06,
      "loss": 0.4036,
      "num_input_tokens_seen": 21498272,
      "step": 32860
    },
    {
      "epoch": 17.2248427672956,
      "grad_norm": 0.5743204951286316,
      "learning_rate": 2.8766934518193566e-06,
      "loss": 0.4903,
      "num_input_tokens_seen": 21501440,
      "step": 32865
    },
    {
      "epoch": 17.22746331236897,
      "grad_norm": 0.3339426815509796,
      "learning_rate": 2.87137059357995e-06,
      "loss": 0.3811,
      "num_input_tokens_seen": 21504192,
      "step": 32870
    },
    {
      "epoch": 17.23008385744235,
      "grad_norm": 0.3653189539909363,
      "learning_rate": 2.8660523643973618e-06,
      "loss": 0.6045,
      "num_input_tokens_seen": 21507936,
      "step": 32875
    },
    {
      "epoch": 17.232704402515722,
      "grad_norm": 0.8927925229072571,
      "learning_rate": 2.860738765384105e-06,
      "loss": 0.3817,
      "num_input_tokens_seen": 21510336,
      "step": 32880
    },
    {
      "epoch": 17.2353249475891,
      "grad_norm": 0.5908169746398926,
      "learning_rate": 2.8554297976517154e-06,
      "loss": 0.3611,
      "num_input_tokens_seen": 21514048,
      "step": 32885
    },
    {
      "epoch": 17.237945492662472,
      "grad_norm": 0.5365005135536194,
      "learning_rate": 2.8501254623107707e-06,
      "loss": 0.5259,
      "num_input_tokens_seen": 21516224,
      "step": 32890
    },
    {
      "epoch": 17.24056603773585,
      "grad_norm": 0.7149670124053955,
      "learning_rate": 2.84482576047089e-06,
      "loss": 0.4055,
      "num_input_tokens_seen": 21519616,
      "step": 32895
    },
    {
      "epoch": 17.243186582809223,
      "grad_norm": 0.5097280740737915,
      "learning_rate": 2.839530693240694e-06,
      "loss": 0.6258,
      "num_input_tokens_seen": 21522048,
      "step": 32900
    },
    {
      "epoch": 17.2458071278826,
      "grad_norm": 0.5439879298210144,
      "learning_rate": 2.8342402617278606e-06,
      "loss": 0.4611,
      "num_input_tokens_seen": 21525312,
      "step": 32905
    },
    {
      "epoch": 17.248427672955973,
      "grad_norm": 0.5800824165344238,
      "learning_rate": 2.82895446703908e-06,
      "loss": 0.4901,
      "num_input_tokens_seen": 21528960,
      "step": 32910
    },
    {
      "epoch": 17.25104821802935,
      "grad_norm": 0.6089904308319092,
      "learning_rate": 2.823673310280084e-06,
      "loss": 0.367,
      "num_input_tokens_seen": 21531648,
      "step": 32915
    },
    {
      "epoch": 17.253668763102727,
      "grad_norm": 0.29769137501716614,
      "learning_rate": 2.8183967925556305e-06,
      "loss": 0.3416,
      "num_input_tokens_seen": 21534240,
      "step": 32920
    },
    {
      "epoch": 17.2562893081761,
      "grad_norm": 0.5928912162780762,
      "learning_rate": 2.8131249149695115e-06,
      "loss": 0.4379,
      "num_input_tokens_seen": 21539584,
      "step": 32925
    },
    {
      "epoch": 17.258909853249477,
      "grad_norm": 0.4242931604385376,
      "learning_rate": 2.8078576786245397e-06,
      "loss": 0.3788,
      "num_input_tokens_seen": 21543392,
      "step": 32930
    },
    {
      "epoch": 17.26153039832285,
      "grad_norm": 0.6587229371070862,
      "learning_rate": 2.8025950846225557e-06,
      "loss": 0.5096,
      "num_input_tokens_seen": 21546112,
      "step": 32935
    },
    {
      "epoch": 17.264150943396228,
      "grad_norm": 0.4396766126155853,
      "learning_rate": 2.7973371340644373e-06,
      "loss": 0.5082,
      "num_input_tokens_seen": 21548896,
      "step": 32940
    },
    {
      "epoch": 17.2667714884696,
      "grad_norm": 0.6721100211143494,
      "learning_rate": 2.7920838280500907e-06,
      "loss": 0.5172,
      "num_input_tokens_seen": 21552064,
      "step": 32945
    },
    {
      "epoch": 17.269392033542978,
      "grad_norm": 0.40351149439811707,
      "learning_rate": 2.7868351676784504e-06,
      "loss": 0.4287,
      "num_input_tokens_seen": 21555872,
      "step": 32950
    },
    {
      "epoch": 17.27201257861635,
      "grad_norm": 0.6119697690010071,
      "learning_rate": 2.7815911540474665e-06,
      "loss": 0.6491,
      "num_input_tokens_seen": 21558880,
      "step": 32955
    },
    {
      "epoch": 17.27463312368973,
      "grad_norm": 0.2706655263900757,
      "learning_rate": 2.7763517882541364e-06,
      "loss": 0.3521,
      "num_input_tokens_seen": 21562272,
      "step": 32960
    },
    {
      "epoch": 17.2772536687631,
      "grad_norm": 0.5080409049987793,
      "learning_rate": 2.771117071394466e-06,
      "loss": 0.4849,
      "num_input_tokens_seen": 21565088,
      "step": 32965
    },
    {
      "epoch": 17.27987421383648,
      "grad_norm": 0.4516911208629608,
      "learning_rate": 2.7658870045635055e-06,
      "loss": 0.4751,
      "num_input_tokens_seen": 21568128,
      "step": 32970
    },
    {
      "epoch": 17.282494758909852,
      "grad_norm": 0.2896983325481415,
      "learning_rate": 2.7606615888553234e-06,
      "loss": 0.4911,
      "num_input_tokens_seen": 21571936,
      "step": 32975
    },
    {
      "epoch": 17.28511530398323,
      "grad_norm": 0.4913565516471863,
      "learning_rate": 2.755440825363023e-06,
      "loss": 0.3687,
      "num_input_tokens_seen": 21575808,
      "step": 32980
    },
    {
      "epoch": 17.287735849056602,
      "grad_norm": 0.4800967872142792,
      "learning_rate": 2.75022471517872e-06,
      "loss": 0.3906,
      "num_input_tokens_seen": 21578976,
      "step": 32985
    },
    {
      "epoch": 17.29035639412998,
      "grad_norm": 0.4143470525741577,
      "learning_rate": 2.745013259393564e-06,
      "loss": 0.4167,
      "num_input_tokens_seen": 21583008,
      "step": 32990
    },
    {
      "epoch": 17.292976939203353,
      "grad_norm": 0.5248459577560425,
      "learning_rate": 2.739806459097735e-06,
      "loss": 0.3728,
      "num_input_tokens_seen": 21585824,
      "step": 32995
    },
    {
      "epoch": 17.29559748427673,
      "grad_norm": 0.5321710705757141,
      "learning_rate": 2.7346043153804403e-06,
      "loss": 0.496,
      "num_input_tokens_seen": 21589568,
      "step": 33000
    },
    {
      "epoch": 17.298218029350103,
      "grad_norm": 0.5867923498153687,
      "learning_rate": 2.729406829329903e-06,
      "loss": 0.5768,
      "num_input_tokens_seen": 21593088,
      "step": 33005
    },
    {
      "epoch": 17.30083857442348,
      "grad_norm": 0.8730840682983398,
      "learning_rate": 2.72421400203339e-06,
      "loss": 0.4572,
      "num_input_tokens_seen": 21595488,
      "step": 33010
    },
    {
      "epoch": 17.303459119496857,
      "grad_norm": 0.44645559787750244,
      "learning_rate": 2.719025834577163e-06,
      "loss": 0.4948,
      "num_input_tokens_seen": 21598304,
      "step": 33015
    },
    {
      "epoch": 17.30607966457023,
      "grad_norm": 0.3576304316520691,
      "learning_rate": 2.713842328046537e-06,
      "loss": 0.4052,
      "num_input_tokens_seen": 21602432,
      "step": 33020
    },
    {
      "epoch": 17.308700209643607,
      "grad_norm": 0.6475261449813843,
      "learning_rate": 2.7086634835258416e-06,
      "loss": 0.4273,
      "num_input_tokens_seen": 21604736,
      "step": 33025
    },
    {
      "epoch": 17.31132075471698,
      "grad_norm": 0.633804976940155,
      "learning_rate": 2.7034893020984333e-06,
      "loss": 0.6845,
      "num_input_tokens_seen": 21608352,
      "step": 33030
    },
    {
      "epoch": 17.313941299790358,
      "grad_norm": 0.5947573781013489,
      "learning_rate": 2.6983197848466952e-06,
      "loss": 0.4749,
      "num_input_tokens_seen": 21610784,
      "step": 33035
    },
    {
      "epoch": 17.31656184486373,
      "grad_norm": 1.0388692617416382,
      "learning_rate": 2.6931549328520265e-06,
      "loss": 0.5139,
      "num_input_tokens_seen": 21614336,
      "step": 33040
    },
    {
      "epoch": 17.319182389937108,
      "grad_norm": 0.5475926995277405,
      "learning_rate": 2.6879947471948523e-06,
      "loss": 0.4736,
      "num_input_tokens_seen": 21616960,
      "step": 33045
    },
    {
      "epoch": 17.32180293501048,
      "grad_norm": 0.6665981411933899,
      "learning_rate": 2.6828392289546285e-06,
      "loss": 0.5279,
      "num_input_tokens_seen": 21619616,
      "step": 33050
    },
    {
      "epoch": 17.32442348008386,
      "grad_norm": 0.7163122892379761,
      "learning_rate": 2.677688379209828e-06,
      "loss": 0.3762,
      "num_input_tokens_seen": 21622336,
      "step": 33055
    },
    {
      "epoch": 17.32704402515723,
      "grad_norm": 0.6741851568222046,
      "learning_rate": 2.6725421990379586e-06,
      "loss": 0.5339,
      "num_input_tokens_seen": 21625408,
      "step": 33060
    },
    {
      "epoch": 17.32966457023061,
      "grad_norm": 0.3789743185043335,
      "learning_rate": 2.667400689515537e-06,
      "loss": 0.447,
      "num_input_tokens_seen": 21628640,
      "step": 33065
    },
    {
      "epoch": 17.332285115303982,
      "grad_norm": 0.4572788178920746,
      "learning_rate": 2.6622638517181e-06,
      "loss": 0.4636,
      "num_input_tokens_seen": 21632448,
      "step": 33070
    },
    {
      "epoch": 17.33490566037736,
      "grad_norm": 0.47970226407051086,
      "learning_rate": 2.657131686720224e-06,
      "loss": 0.4345,
      "num_input_tokens_seen": 21635424,
      "step": 33075
    },
    {
      "epoch": 17.337526205450732,
      "grad_norm": 0.5429767966270447,
      "learning_rate": 2.652004195595498e-06,
      "loss": 0.5954,
      "num_input_tokens_seen": 21638656,
      "step": 33080
    },
    {
      "epoch": 17.34014675052411,
      "grad_norm": 0.5874156355857849,
      "learning_rate": 2.6468813794165356e-06,
      "loss": 0.3937,
      "num_input_tokens_seen": 21641376,
      "step": 33085
    },
    {
      "epoch": 17.342767295597483,
      "grad_norm": 0.4670505225658417,
      "learning_rate": 2.641763239254977e-06,
      "loss": 0.3768,
      "num_input_tokens_seen": 21644480,
      "step": 33090
    },
    {
      "epoch": 17.34538784067086,
      "grad_norm": 0.3207525312900543,
      "learning_rate": 2.6366497761814713e-06,
      "loss": 0.3707,
      "num_input_tokens_seen": 21647712,
      "step": 33095
    },
    {
      "epoch": 17.348008385744233,
      "grad_norm": 0.5946163535118103,
      "learning_rate": 2.631540991265691e-06,
      "loss": 0.6344,
      "num_input_tokens_seen": 21650880,
      "step": 33100
    },
    {
      "epoch": 17.35062893081761,
      "grad_norm": 0.4480970799922943,
      "learning_rate": 2.6264368855763443e-06,
      "loss": 0.471,
      "num_input_tokens_seen": 21654048,
      "step": 33105
    },
    {
      "epoch": 17.353249475890987,
      "grad_norm": 0.42381909489631653,
      "learning_rate": 2.6213374601811504e-06,
      "loss": 0.4191,
      "num_input_tokens_seen": 21656896,
      "step": 33110
    },
    {
      "epoch": 17.35587002096436,
      "grad_norm": 1.7628443241119385,
      "learning_rate": 2.616242716146855e-06,
      "loss": 0.5474,
      "num_input_tokens_seen": 21659680,
      "step": 33115
    },
    {
      "epoch": 17.358490566037737,
      "grad_norm": 0.6050702929496765,
      "learning_rate": 2.6111526545392145e-06,
      "loss": 0.4177,
      "num_input_tokens_seen": 21662048,
      "step": 33120
    },
    {
      "epoch": 17.36111111111111,
      "grad_norm": 0.607552170753479,
      "learning_rate": 2.6060672764230106e-06,
      "loss": 0.5312,
      "num_input_tokens_seen": 21665952,
      "step": 33125
    },
    {
      "epoch": 17.363731656184488,
      "grad_norm": 0.66056227684021,
      "learning_rate": 2.600986582862047e-06,
      "loss": 0.4169,
      "num_input_tokens_seen": 21669376,
      "step": 33130
    },
    {
      "epoch": 17.36635220125786,
      "grad_norm": 0.500158965587616,
      "learning_rate": 2.5959105749191498e-06,
      "loss": 0.486,
      "num_input_tokens_seen": 21673440,
      "step": 33135
    },
    {
      "epoch": 17.368972746331238,
      "grad_norm": 0.4733177125453949,
      "learning_rate": 2.590839253656166e-06,
      "loss": 0.5584,
      "num_input_tokens_seen": 21677248,
      "step": 33140
    },
    {
      "epoch": 17.37159329140461,
      "grad_norm": 0.6820141673088074,
      "learning_rate": 2.5857726201339473e-06,
      "loss": 0.5031,
      "num_input_tokens_seen": 21680320,
      "step": 33145
    },
    {
      "epoch": 17.37421383647799,
      "grad_norm": 0.45808711647987366,
      "learning_rate": 2.5807106754123873e-06,
      "loss": 0.3976,
      "num_input_tokens_seen": 21683296,
      "step": 33150
    },
    {
      "epoch": 17.37683438155136,
      "grad_norm": 0.3771827816963196,
      "learning_rate": 2.575653420550378e-06,
      "loss": 0.4245,
      "num_input_tokens_seen": 21685888,
      "step": 33155
    },
    {
      "epoch": 17.37945492662474,
      "grad_norm": 0.4778707027435303,
      "learning_rate": 2.57060085660584e-06,
      "loss": 0.4709,
      "num_input_tokens_seen": 21689088,
      "step": 33160
    },
    {
      "epoch": 17.382075471698112,
      "grad_norm": 0.4317012131214142,
      "learning_rate": 2.565552984635716e-06,
      "loss": 0.4126,
      "num_input_tokens_seen": 21692896,
      "step": 33165
    },
    {
      "epoch": 17.38469601677149,
      "grad_norm": 0.792277455329895,
      "learning_rate": 2.5605098056959652e-06,
      "loss": 0.5159,
      "num_input_tokens_seen": 21695520,
      "step": 33170
    },
    {
      "epoch": 17.387316561844862,
      "grad_norm": 0.8605978488922119,
      "learning_rate": 2.555471320841557e-06,
      "loss": 0.5062,
      "num_input_tokens_seen": 21697824,
      "step": 33175
    },
    {
      "epoch": 17.38993710691824,
      "grad_norm": 0.48226118087768555,
      "learning_rate": 2.55043753112649e-06,
      "loss": 0.557,
      "num_input_tokens_seen": 21702336,
      "step": 33180
    },
    {
      "epoch": 17.392557651991613,
      "grad_norm": 0.43256527185440063,
      "learning_rate": 2.5454084376037717e-06,
      "loss": 0.4544,
      "num_input_tokens_seen": 21705728,
      "step": 33185
    },
    {
      "epoch": 17.39517819706499,
      "grad_norm": 0.5290156602859497,
      "learning_rate": 2.5403840413254304e-06,
      "loss": 0.3393,
      "num_input_tokens_seen": 21709472,
      "step": 33190
    },
    {
      "epoch": 17.397798742138363,
      "grad_norm": 0.3864929676055908,
      "learning_rate": 2.5353643433425196e-06,
      "loss": 0.4994,
      "num_input_tokens_seen": 21712704,
      "step": 33195
    },
    {
      "epoch": 17.40041928721174,
      "grad_norm": 0.9399967789649963,
      "learning_rate": 2.5303493447050965e-06,
      "loss": 0.3174,
      "num_input_tokens_seen": 21715040,
      "step": 33200
    },
    {
      "epoch": 17.403039832285117,
      "grad_norm": 0.7370752692222595,
      "learning_rate": 2.525339046462244e-06,
      "loss": 0.6861,
      "num_input_tokens_seen": 21717952,
      "step": 33205
    },
    {
      "epoch": 17.40566037735849,
      "grad_norm": 0.45208966732025146,
      "learning_rate": 2.5203334496620544e-06,
      "loss": 0.4536,
      "num_input_tokens_seen": 21721856,
      "step": 33210
    },
    {
      "epoch": 17.408280922431867,
      "grad_norm": 0.579863965511322,
      "learning_rate": 2.515332555351646e-06,
      "loss": 0.3389,
      "num_input_tokens_seen": 21726880,
      "step": 33215
    },
    {
      "epoch": 17.41090146750524,
      "grad_norm": 0.5211570858955383,
      "learning_rate": 2.510336364577154e-06,
      "loss": 0.5733,
      "num_input_tokens_seen": 21730048,
      "step": 33220
    },
    {
      "epoch": 17.413522012578618,
      "grad_norm": 0.5258392095565796,
      "learning_rate": 2.5053448783837143e-06,
      "loss": 0.4276,
      "num_input_tokens_seen": 21732768,
      "step": 33225
    },
    {
      "epoch": 17.41614255765199,
      "grad_norm": 0.4709720313549042,
      "learning_rate": 2.5003580978154917e-06,
      "loss": 0.3745,
      "num_input_tokens_seen": 21736064,
      "step": 33230
    },
    {
      "epoch": 17.418763102725368,
      "grad_norm": 0.709987998008728,
      "learning_rate": 2.4953760239156716e-06,
      "loss": 0.4965,
      "num_input_tokens_seen": 21738464,
      "step": 33235
    },
    {
      "epoch": 17.42138364779874,
      "grad_norm": 0.5825677514076233,
      "learning_rate": 2.4903986577264366e-06,
      "loss": 0.377,
      "num_input_tokens_seen": 21740800,
      "step": 33240
    },
    {
      "epoch": 17.42400419287212,
      "grad_norm": 0.2902831435203552,
      "learning_rate": 2.4854260002889986e-06,
      "loss": 0.4197,
      "num_input_tokens_seen": 21744096,
      "step": 33245
    },
    {
      "epoch": 17.42662473794549,
      "grad_norm": 0.4693946838378906,
      "learning_rate": 2.480458052643586e-06,
      "loss": 0.4197,
      "num_input_tokens_seen": 21746944,
      "step": 33250
    },
    {
      "epoch": 17.42924528301887,
      "grad_norm": 0.35597437620162964,
      "learning_rate": 2.4754948158294293e-06,
      "loss": 0.4631,
      "num_input_tokens_seen": 21750144,
      "step": 33255
    },
    {
      "epoch": 17.431865828092242,
      "grad_norm": 0.4731789231300354,
      "learning_rate": 2.4705362908847814e-06,
      "loss": 0.4806,
      "num_input_tokens_seen": 21752992,
      "step": 33260
    },
    {
      "epoch": 17.43448637316562,
      "grad_norm": 0.5699461102485657,
      "learning_rate": 2.4655824788469172e-06,
      "loss": 0.4483,
      "num_input_tokens_seen": 21755744,
      "step": 33265
    },
    {
      "epoch": 17.437106918238992,
      "grad_norm": 0.38611117005348206,
      "learning_rate": 2.4606333807521087e-06,
      "loss": 0.451,
      "num_input_tokens_seen": 21759392,
      "step": 33270
    },
    {
      "epoch": 17.43972746331237,
      "grad_norm": 0.9702040553092957,
      "learning_rate": 2.455688997635658e-06,
      "loss": 0.4575,
      "num_input_tokens_seen": 21762816,
      "step": 33275
    },
    {
      "epoch": 17.442348008385743,
      "grad_norm": 0.5364574193954468,
      "learning_rate": 2.4507493305318652e-06,
      "loss": 0.4737,
      "num_input_tokens_seen": 21766720,
      "step": 33280
    },
    {
      "epoch": 17.44496855345912,
      "grad_norm": 0.3553297519683838,
      "learning_rate": 2.445814380474057e-06,
      "loss": 0.425,
      "num_input_tokens_seen": 21770592,
      "step": 33285
    },
    {
      "epoch": 17.447589098532493,
      "grad_norm": 0.48768511414527893,
      "learning_rate": 2.4408841484945684e-06,
      "loss": 0.4455,
      "num_input_tokens_seen": 21774944,
      "step": 33290
    },
    {
      "epoch": 17.45020964360587,
      "grad_norm": 0.38955438137054443,
      "learning_rate": 2.435958635624755e-06,
      "loss": 0.454,
      "num_input_tokens_seen": 21777440,
      "step": 33295
    },
    {
      "epoch": 17.452830188679247,
      "grad_norm": 0.4818378984928131,
      "learning_rate": 2.4310378428949678e-06,
      "loss": 0.4052,
      "num_input_tokens_seen": 21781280,
      "step": 33300
    },
    {
      "epoch": 17.45545073375262,
      "grad_norm": 0.31832948327064514,
      "learning_rate": 2.4261217713345804e-06,
      "loss": 0.5019,
      "num_input_tokens_seen": 21784320,
      "step": 33305
    },
    {
      "epoch": 17.458071278825997,
      "grad_norm": 0.2691464424133301,
      "learning_rate": 2.421210421971981e-06,
      "loss": 0.4228,
      "num_input_tokens_seen": 21788832,
      "step": 33310
    },
    {
      "epoch": 17.46069182389937,
      "grad_norm": 0.48400112986564636,
      "learning_rate": 2.416303795834571e-06,
      "loss": 0.528,
      "num_input_tokens_seen": 21792320,
      "step": 33315
    },
    {
      "epoch": 17.463312368972748,
      "grad_norm": 0.44203290343284607,
      "learning_rate": 2.4114018939487615e-06,
      "loss": 0.3997,
      "num_input_tokens_seen": 21795296,
      "step": 33320
    },
    {
      "epoch": 17.46593291404612,
      "grad_norm": 0.4568311274051666,
      "learning_rate": 2.4065047173399662e-06,
      "loss": 0.4456,
      "num_input_tokens_seen": 21798176,
      "step": 33325
    },
    {
      "epoch": 17.468553459119498,
      "grad_norm": 0.5314241051673889,
      "learning_rate": 2.4016122670326295e-06,
      "loss": 0.4482,
      "num_input_tokens_seen": 21801120,
      "step": 33330
    },
    {
      "epoch": 17.47117400419287,
      "grad_norm": 0.3814983069896698,
      "learning_rate": 2.396724544050186e-06,
      "loss": 0.4256,
      "num_input_tokens_seen": 21804096,
      "step": 33335
    },
    {
      "epoch": 17.47379454926625,
      "grad_norm": 0.5782012343406677,
      "learning_rate": 2.3918415494150944e-06,
      "loss": 0.4301,
      "num_input_tokens_seen": 21806976,
      "step": 33340
    },
    {
      "epoch": 17.47641509433962,
      "grad_norm": 0.42589282989501953,
      "learning_rate": 2.3869632841488245e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 21810176,
      "step": 33345
    },
    {
      "epoch": 17.479035639413,
      "grad_norm": 0.7252336740493774,
      "learning_rate": 2.382089749271857e-06,
      "loss": 0.3359,
      "num_input_tokens_seen": 21817280,
      "step": 33350
    },
    {
      "epoch": 17.481656184486372,
      "grad_norm": 0.47128936648368835,
      "learning_rate": 2.3772209458036737e-06,
      "loss": 0.2669,
      "num_input_tokens_seen": 21819712,
      "step": 33355
    },
    {
      "epoch": 17.48427672955975,
      "grad_norm": 0.3640895485877991,
      "learning_rate": 2.372356874762771e-06,
      "loss": 0.3473,
      "num_input_tokens_seen": 21822816,
      "step": 33360
    },
    {
      "epoch": 17.486897274633122,
      "grad_norm": 0.40197518467903137,
      "learning_rate": 2.36749753716666e-06,
      "loss": 0.4483,
      "num_input_tokens_seen": 21826048,
      "step": 33365
    },
    {
      "epoch": 17.4895178197065,
      "grad_norm": 0.5951241254806519,
      "learning_rate": 2.3626429340318613e-06,
      "loss": 0.3697,
      "num_input_tokens_seen": 21829312,
      "step": 33370
    },
    {
      "epoch": 17.492138364779873,
      "grad_norm": 0.547693133354187,
      "learning_rate": 2.357793066373901e-06,
      "loss": 0.424,
      "num_input_tokens_seen": 21832736,
      "step": 33375
    },
    {
      "epoch": 17.49475890985325,
      "grad_norm": 0.9360001683235168,
      "learning_rate": 2.352947935207317e-06,
      "loss": 0.4525,
      "num_input_tokens_seen": 21836832,
      "step": 33380
    },
    {
      "epoch": 17.497379454926623,
      "grad_norm": 0.5936570763587952,
      "learning_rate": 2.3481075415456495e-06,
      "loss": 0.483,
      "num_input_tokens_seen": 21839584,
      "step": 33385
    },
    {
      "epoch": 17.5,
      "grad_norm": 0.29521071910858154,
      "learning_rate": 2.3432718864014576e-06,
      "loss": 0.3819,
      "num_input_tokens_seen": 21843072,
      "step": 33390
    },
    {
      "epoch": 17.502620545073377,
      "grad_norm": 0.5273264646530151,
      "learning_rate": 2.338440970786304e-06,
      "loss": 0.3715,
      "num_input_tokens_seen": 21846688,
      "step": 33395
    },
    {
      "epoch": 17.50524109014675,
      "grad_norm": 0.3825317323207855,
      "learning_rate": 2.3336147957107644e-06,
      "loss": 0.4659,
      "num_input_tokens_seen": 21849856,
      "step": 33400
    },
    {
      "epoch": 17.507861635220127,
      "grad_norm": 0.34959667921066284,
      "learning_rate": 2.3287933621844227e-06,
      "loss": 0.3887,
      "num_input_tokens_seen": 21853856,
      "step": 33405
    },
    {
      "epoch": 17.5104821802935,
      "grad_norm": 0.45459890365600586,
      "learning_rate": 2.3239766712158633e-06,
      "loss": 0.4104,
      "num_input_tokens_seen": 21860576,
      "step": 33410
    },
    {
      "epoch": 17.513102725366878,
      "grad_norm": 0.5810343027114868,
      "learning_rate": 2.319164723812675e-06,
      "loss": 0.4146,
      "num_input_tokens_seen": 21864192,
      "step": 33415
    },
    {
      "epoch": 17.51572327044025,
      "grad_norm": 1.0937894582748413,
      "learning_rate": 2.314357520981472e-06,
      "loss": 0.5149,
      "num_input_tokens_seen": 21866208,
      "step": 33420
    },
    {
      "epoch": 17.518343815513628,
      "grad_norm": 0.5907233953475952,
      "learning_rate": 2.309555063727864e-06,
      "loss": 0.4949,
      "num_input_tokens_seen": 21869760,
      "step": 33425
    },
    {
      "epoch": 17.520964360587,
      "grad_norm": 0.3534350097179413,
      "learning_rate": 2.304757353056475e-06,
      "loss": 0.4492,
      "num_input_tokens_seen": 21874560,
      "step": 33430
    },
    {
      "epoch": 17.52358490566038,
      "grad_norm": 0.3579021394252777,
      "learning_rate": 2.2999643899709267e-06,
      "loss": 0.2986,
      "num_input_tokens_seen": 21878336,
      "step": 33435
    },
    {
      "epoch": 17.52620545073375,
      "grad_norm": 1.323434829711914,
      "learning_rate": 2.2951761754738456e-06,
      "loss": 0.4171,
      "num_input_tokens_seen": 21881120,
      "step": 33440
    },
    {
      "epoch": 17.52882599580713,
      "grad_norm": 0.504032552242279,
      "learning_rate": 2.29039271056688e-06,
      "loss": 0.5128,
      "num_input_tokens_seen": 21884032,
      "step": 33445
    },
    {
      "epoch": 17.531446540880502,
      "grad_norm": 0.2764568030834198,
      "learning_rate": 2.285613996250674e-06,
      "loss": 0.3767,
      "num_input_tokens_seen": 21887776,
      "step": 33450
    },
    {
      "epoch": 17.53406708595388,
      "grad_norm": 0.4961960017681122,
      "learning_rate": 2.2808400335248865e-06,
      "loss": 0.5031,
      "num_input_tokens_seen": 21891488,
      "step": 33455
    },
    {
      "epoch": 17.536687631027252,
      "grad_norm": 0.5275465250015259,
      "learning_rate": 2.2760708233881654e-06,
      "loss": 0.3557,
      "num_input_tokens_seen": 21894176,
      "step": 33460
    },
    {
      "epoch": 17.53930817610063,
      "grad_norm": 0.3710968792438507,
      "learning_rate": 2.271306366838183e-06,
      "loss": 0.4462,
      "num_input_tokens_seen": 21897984,
      "step": 33465
    },
    {
      "epoch": 17.541928721174003,
      "grad_norm": 0.3136024475097656,
      "learning_rate": 2.266546664871602e-06,
      "loss": 0.444,
      "num_input_tokens_seen": 21900928,
      "step": 33470
    },
    {
      "epoch": 17.54454926624738,
      "grad_norm": 0.49622151255607605,
      "learning_rate": 2.261791718484102e-06,
      "loss": 0.34,
      "num_input_tokens_seen": 21904992,
      "step": 33475
    },
    {
      "epoch": 17.547169811320753,
      "grad_norm": 0.6385157704353333,
      "learning_rate": 2.257041528670362e-06,
      "loss": 0.3705,
      "num_input_tokens_seen": 21907584,
      "step": 33480
    },
    {
      "epoch": 17.54979035639413,
      "grad_norm": 0.8556817770004272,
      "learning_rate": 2.2522960964240753e-06,
      "loss": 0.5073,
      "num_input_tokens_seen": 21910208,
      "step": 33485
    },
    {
      "epoch": 17.552410901467507,
      "grad_norm": 0.6371718049049377,
      "learning_rate": 2.247555422737924e-06,
      "loss": 0.4577,
      "num_input_tokens_seen": 21913024,
      "step": 33490
    },
    {
      "epoch": 17.55503144654088,
      "grad_norm": 0.5486925840377808,
      "learning_rate": 2.2428195086036024e-06,
      "loss": 0.5436,
      "num_input_tokens_seen": 21916000,
      "step": 33495
    },
    {
      "epoch": 17.557651991614257,
      "grad_norm": 0.43574875593185425,
      "learning_rate": 2.238088355011811e-06,
      "loss": 0.4077,
      "num_input_tokens_seen": 21918880,
      "step": 33500
    },
    {
      "epoch": 17.56027253668763,
      "grad_norm": 0.7645496726036072,
      "learning_rate": 2.2333619629522518e-06,
      "loss": 0.4827,
      "num_input_tokens_seen": 21921728,
      "step": 33505
    },
    {
      "epoch": 17.562893081761008,
      "grad_norm": 0.33534592390060425,
      "learning_rate": 2.2286403334136424e-06,
      "loss": 0.5077,
      "num_input_tokens_seen": 21925568,
      "step": 33510
    },
    {
      "epoch": 17.56551362683438,
      "grad_norm": 0.9368852972984314,
      "learning_rate": 2.223923467383679e-06,
      "loss": 0.4002,
      "num_input_tokens_seen": 21928416,
      "step": 33515
    },
    {
      "epoch": 17.568134171907758,
      "grad_norm": 0.5927827954292297,
      "learning_rate": 2.2192113658490886e-06,
      "loss": 0.4558,
      "num_input_tokens_seen": 21932032,
      "step": 33520
    },
    {
      "epoch": 17.57075471698113,
      "grad_norm": 1.0775035619735718,
      "learning_rate": 2.2145040297955776e-06,
      "loss": 0.4393,
      "num_input_tokens_seen": 21936096,
      "step": 33525
    },
    {
      "epoch": 17.57337526205451,
      "grad_norm": 0.7020312547683716,
      "learning_rate": 2.209801460207875e-06,
      "loss": 0.5274,
      "num_input_tokens_seen": 21938528,
      "step": 33530
    },
    {
      "epoch": 17.57599580712788,
      "grad_norm": 0.6423866152763367,
      "learning_rate": 2.2051036580697042e-06,
      "loss": 0.4212,
      "num_input_tokens_seen": 21941184,
      "step": 33535
    },
    {
      "epoch": 17.57861635220126,
      "grad_norm": 0.22289827466011047,
      "learning_rate": 2.2004106243637885e-06,
      "loss": 0.413,
      "num_input_tokens_seen": 21944160,
      "step": 33540
    },
    {
      "epoch": 17.581236897274632,
      "grad_norm": 0.7186585068702698,
      "learning_rate": 2.1957223600718587e-06,
      "loss": 0.3603,
      "num_input_tokens_seen": 21946848,
      "step": 33545
    },
    {
      "epoch": 17.58385744234801,
      "grad_norm": 0.6861995458602905,
      "learning_rate": 2.1910388661746493e-06,
      "loss": 0.5054,
      "num_input_tokens_seen": 21949312,
      "step": 33550
    },
    {
      "epoch": 17.586477987421382,
      "grad_norm": 0.42622047662734985,
      "learning_rate": 2.186360143651886e-06,
      "loss": 0.6108,
      "num_input_tokens_seen": 21953184,
      "step": 33555
    },
    {
      "epoch": 17.58909853249476,
      "grad_norm": 0.6046077013015747,
      "learning_rate": 2.1816861934823103e-06,
      "loss": 0.3644,
      "num_input_tokens_seen": 21957568,
      "step": 33560
    },
    {
      "epoch": 17.591719077568133,
      "grad_norm": 1.5417040586471558,
      "learning_rate": 2.1770170166436595e-06,
      "loss": 0.5934,
      "num_input_tokens_seen": 21960896,
      "step": 33565
    },
    {
      "epoch": 17.59433962264151,
      "grad_norm": 0.8292446732521057,
      "learning_rate": 2.172352614112669e-06,
      "loss": 0.3702,
      "num_input_tokens_seen": 21963360,
      "step": 33570
    },
    {
      "epoch": 17.596960167714883,
      "grad_norm": 0.3338536322116852,
      "learning_rate": 2.167692986865083e-06,
      "loss": 0.3734,
      "num_input_tokens_seen": 21966976,
      "step": 33575
    },
    {
      "epoch": 17.59958071278826,
      "grad_norm": 0.6219679713249207,
      "learning_rate": 2.1630381358756362e-06,
      "loss": 0.2824,
      "num_input_tokens_seen": 21969824,
      "step": 33580
    },
    {
      "epoch": 17.602201257861637,
      "grad_norm": 0.5214774012565613,
      "learning_rate": 2.158388062118072e-06,
      "loss": 0.4952,
      "num_input_tokens_seen": 21972320,
      "step": 33585
    },
    {
      "epoch": 17.60482180293501,
      "grad_norm": 0.5758970379829407,
      "learning_rate": 2.1537427665651393e-06,
      "loss": 0.404,
      "num_input_tokens_seen": 21974624,
      "step": 33590
    },
    {
      "epoch": 17.607442348008387,
      "grad_norm": 0.6125354170799255,
      "learning_rate": 2.1491022501885726e-06,
      "loss": 0.3973,
      "num_input_tokens_seen": 21978432,
      "step": 33595
    },
    {
      "epoch": 17.61006289308176,
      "grad_norm": 0.3955400586128235,
      "learning_rate": 2.1444665139591174e-06,
      "loss": 0.4893,
      "num_input_tokens_seen": 21982016,
      "step": 33600
    },
    {
      "epoch": 17.612683438155138,
      "grad_norm": 0.3367362916469574,
      "learning_rate": 2.139835558846523e-06,
      "loss": 0.4084,
      "num_input_tokens_seen": 21985088,
      "step": 33605
    },
    {
      "epoch": 17.61530398322851,
      "grad_norm": 0.4204290211200714,
      "learning_rate": 2.1352093858195225e-06,
      "loss": 0.4909,
      "num_input_tokens_seen": 21988704,
      "step": 33610
    },
    {
      "epoch": 17.617924528301888,
      "grad_norm": 0.5913007855415344,
      "learning_rate": 2.1305879958458676e-06,
      "loss": 0.5224,
      "num_input_tokens_seen": 21991264,
      "step": 33615
    },
    {
      "epoch": 17.62054507337526,
      "grad_norm": 0.3702651262283325,
      "learning_rate": 2.1259713898922894e-06,
      "loss": 0.3495,
      "num_input_tokens_seen": 21994208,
      "step": 33620
    },
    {
      "epoch": 17.62316561844864,
      "grad_norm": 0.32900330424308777,
      "learning_rate": 2.1213595689245386e-06,
      "loss": 0.3659,
      "num_input_tokens_seen": 21997600,
      "step": 33625
    },
    {
      "epoch": 17.62578616352201,
      "grad_norm": 0.33397120237350464,
      "learning_rate": 2.1167525339073513e-06,
      "loss": 0.366,
      "num_input_tokens_seen": 22000512,
      "step": 33630
    },
    {
      "epoch": 17.62840670859539,
      "grad_norm": 0.45537903904914856,
      "learning_rate": 2.112150285804473e-06,
      "loss": 0.3404,
      "num_input_tokens_seen": 22003872,
      "step": 33635
    },
    {
      "epoch": 17.631027253668762,
      "grad_norm": 0.591462254524231,
      "learning_rate": 2.107552825578632e-06,
      "loss": 0.4077,
      "num_input_tokens_seen": 22007232,
      "step": 33640
    },
    {
      "epoch": 17.63364779874214,
      "grad_norm": 0.8185169100761414,
      "learning_rate": 2.102960154191572e-06,
      "loss": 0.4253,
      "num_input_tokens_seen": 22010624,
      "step": 33645
    },
    {
      "epoch": 17.636268343815512,
      "grad_norm": 0.48336154222488403,
      "learning_rate": 2.098372272604021e-06,
      "loss": 0.4542,
      "num_input_tokens_seen": 22013664,
      "step": 33650
    },
    {
      "epoch": 17.63888888888889,
      "grad_norm": 1.2248568534851074,
      "learning_rate": 2.0937891817757155e-06,
      "loss": 0.4462,
      "num_input_tokens_seen": 22016896,
      "step": 33655
    },
    {
      "epoch": 17.641509433962263,
      "grad_norm": 0.5562716126441956,
      "learning_rate": 2.089210882665385e-06,
      "loss": 0.4655,
      "num_input_tokens_seen": 22019744,
      "step": 33660
    },
    {
      "epoch": 17.64412997903564,
      "grad_norm": 0.4972453713417053,
      "learning_rate": 2.0846373762307635e-06,
      "loss": 0.3885,
      "num_input_tokens_seen": 22023232,
      "step": 33665
    },
    {
      "epoch": 17.646750524109013,
      "grad_norm": 0.2800089716911316,
      "learning_rate": 2.080068663428569e-06,
      "loss": 0.4358,
      "num_input_tokens_seen": 22026080,
      "step": 33670
    },
    {
      "epoch": 17.64937106918239,
      "grad_norm": 0.24806532263755798,
      "learning_rate": 2.075504745214521e-06,
      "loss": 0.5887,
      "num_input_tokens_seen": 22029248,
      "step": 33675
    },
    {
      "epoch": 17.651991614255767,
      "grad_norm": 0.5226616859436035,
      "learning_rate": 2.0709456225433437e-06,
      "loss": 0.4562,
      "num_input_tokens_seen": 22032128,
      "step": 33680
    },
    {
      "epoch": 17.65461215932914,
      "grad_norm": 0.36497023701667786,
      "learning_rate": 2.0663912963687554e-06,
      "loss": 0.4251,
      "num_input_tokens_seen": 22035488,
      "step": 33685
    },
    {
      "epoch": 17.657232704402517,
      "grad_norm": 0.4163546860218048,
      "learning_rate": 2.0618417676434713e-06,
      "loss": 0.382,
      "num_input_tokens_seen": 22038144,
      "step": 33690
    },
    {
      "epoch": 17.65985324947589,
      "grad_norm": 0.6336692571640015,
      "learning_rate": 2.0572970373191975e-06,
      "loss": 0.3988,
      "num_input_tokens_seen": 22041888,
      "step": 33695
    },
    {
      "epoch": 17.662473794549268,
      "grad_norm": 0.3218402862548828,
      "learning_rate": 2.052757106346634e-06,
      "loss": 0.4044,
      "num_input_tokens_seen": 22046624,
      "step": 33700
    },
    {
      "epoch": 17.66509433962264,
      "grad_norm": 0.4754990041255951,
      "learning_rate": 2.048221975675488e-06,
      "loss": 0.4417,
      "num_input_tokens_seen": 22049184,
      "step": 33705
    },
    {
      "epoch": 17.667714884696018,
      "grad_norm": 0.49619510769844055,
      "learning_rate": 2.043691646254456e-06,
      "loss": 0.3428,
      "num_input_tokens_seen": 22051712,
      "step": 33710
    },
    {
      "epoch": 17.67033542976939,
      "grad_norm": 0.5658960938453674,
      "learning_rate": 2.039166119031233e-06,
      "loss": 0.6193,
      "num_input_tokens_seen": 22055680,
      "step": 33715
    },
    {
      "epoch": 17.67295597484277,
      "grad_norm": 0.3804206848144531,
      "learning_rate": 2.034645394952511e-06,
      "loss": 0.46,
      "num_input_tokens_seen": 22058944,
      "step": 33720
    },
    {
      "epoch": 17.67557651991614,
      "grad_norm": 0.33842572569847107,
      "learning_rate": 2.030129474963971e-06,
      "loss": 0.5229,
      "num_input_tokens_seen": 22063424,
      "step": 33725
    },
    {
      "epoch": 17.67819706498952,
      "grad_norm": 0.34758737683296204,
      "learning_rate": 2.025618360010284e-06,
      "loss": 0.3455,
      "num_input_tokens_seen": 22066752,
      "step": 33730
    },
    {
      "epoch": 17.680817610062892,
      "grad_norm": 0.7554365396499634,
      "learning_rate": 2.0211120510351317e-06,
      "loss": 0.529,
      "num_input_tokens_seen": 22070432,
      "step": 33735
    },
    {
      "epoch": 17.68343815513627,
      "grad_norm": 0.9449359774589539,
      "learning_rate": 2.0166105489811822e-06,
      "loss": 0.5068,
      "num_input_tokens_seen": 22073088,
      "step": 33740
    },
    {
      "epoch": 17.686058700209642,
      "grad_norm": 0.44572991132736206,
      "learning_rate": 2.0121138547901016e-06,
      "loss": 0.3747,
      "num_input_tokens_seen": 22076032,
      "step": 33745
    },
    {
      "epoch": 17.68867924528302,
      "grad_norm": 0.5500967502593994,
      "learning_rate": 2.0076219694025427e-06,
      "loss": 0.4393,
      "num_input_tokens_seen": 22080736,
      "step": 33750
    },
    {
      "epoch": 17.691299790356393,
      "grad_norm": 0.5021515488624573,
      "learning_rate": 2.003134893758152e-06,
      "loss": 0.6097,
      "num_input_tokens_seen": 22083392,
      "step": 33755
    },
    {
      "epoch": 17.69392033542977,
      "grad_norm": 0.41986531019210815,
      "learning_rate": 1.9986526287955805e-06,
      "loss": 0.4703,
      "num_input_tokens_seen": 22086208,
      "step": 33760
    },
    {
      "epoch": 17.696540880503143,
      "grad_norm": 0.4940182566642761,
      "learning_rate": 1.9941751754524644e-06,
      "loss": 0.338,
      "num_input_tokens_seen": 22089792,
      "step": 33765
    },
    {
      "epoch": 17.69916142557652,
      "grad_norm": 0.4122755527496338,
      "learning_rate": 1.9897025346654412e-06,
      "loss": 0.355,
      "num_input_tokens_seen": 22092320,
      "step": 33770
    },
    {
      "epoch": 17.701781970649897,
      "grad_norm": 0.36634111404418945,
      "learning_rate": 1.985234707370134e-06,
      "loss": 0.3722,
      "num_input_tokens_seen": 22095616,
      "step": 33775
    },
    {
      "epoch": 17.70440251572327,
      "grad_norm": 0.48018330335617065,
      "learning_rate": 1.9807716945011544e-06,
      "loss": 0.3198,
      "num_input_tokens_seen": 22098624,
      "step": 33780
    },
    {
      "epoch": 17.707023060796647,
      "grad_norm": 0.6494651436805725,
      "learning_rate": 1.9763134969921215e-06,
      "loss": 0.4324,
      "num_input_tokens_seen": 22101344,
      "step": 33785
    },
    {
      "epoch": 17.70964360587002,
      "grad_norm": 0.5046547651290894,
      "learning_rate": 1.971860115775634e-06,
      "loss": 0.3186,
      "num_input_tokens_seen": 22103680,
      "step": 33790
    },
    {
      "epoch": 17.712264150943398,
      "grad_norm": 0.2854596674442291,
      "learning_rate": 1.967411551783294e-06,
      "loss": 0.4884,
      "num_input_tokens_seen": 22107008,
      "step": 33795
    },
    {
      "epoch": 17.71488469601677,
      "grad_norm": 0.7200257778167725,
      "learning_rate": 1.9629678059456895e-06,
      "loss": 0.327,
      "num_input_tokens_seen": 22109600,
      "step": 33800
    },
    {
      "epoch": 17.717505241090148,
      "grad_norm": 0.3138902485370636,
      "learning_rate": 1.9585288791924004e-06,
      "loss": 0.5493,
      "num_input_tokens_seen": 22112672,
      "step": 33805
    },
    {
      "epoch": 17.72012578616352,
      "grad_norm": 0.3733631372451782,
      "learning_rate": 1.9540947724519947e-06,
      "loss": 0.3881,
      "num_input_tokens_seen": 22115936,
      "step": 33810
    },
    {
      "epoch": 17.7227463312369,
      "grad_norm": 0.4253998398780823,
      "learning_rate": 1.9496654866520414e-06,
      "loss": 0.3799,
      "num_input_tokens_seen": 22119360,
      "step": 33815
    },
    {
      "epoch": 17.72536687631027,
      "grad_norm": 0.7569559216499329,
      "learning_rate": 1.9452410227190937e-06,
      "loss": 0.6178,
      "num_input_tokens_seen": 22122784,
      "step": 33820
    },
    {
      "epoch": 17.72798742138365,
      "grad_norm": 0.30153098702430725,
      "learning_rate": 1.9408213815787087e-06,
      "loss": 0.4736,
      "num_input_tokens_seen": 22125888,
      "step": 33825
    },
    {
      "epoch": 17.730607966457022,
      "grad_norm": 0.3517330288887024,
      "learning_rate": 1.936406564155413e-06,
      "loss": 0.5297,
      "num_input_tokens_seen": 22128960,
      "step": 33830
    },
    {
      "epoch": 17.7332285115304,
      "grad_norm": 0.3704054653644562,
      "learning_rate": 1.9319965713727435e-06,
      "loss": 0.3404,
      "num_input_tokens_seen": 22132512,
      "step": 33835
    },
    {
      "epoch": 17.735849056603772,
      "grad_norm": 0.39265623688697815,
      "learning_rate": 1.9275914041532144e-06,
      "loss": 0.3719,
      "num_input_tokens_seen": 22137056,
      "step": 33840
    },
    {
      "epoch": 17.73846960167715,
      "grad_norm": 0.5636959075927734,
      "learning_rate": 1.9231910634183394e-06,
      "loss": 0.3853,
      "num_input_tokens_seen": 22139360,
      "step": 33845
    },
    {
      "epoch": 17.741090146750523,
      "grad_norm": 0.45778775215148926,
      "learning_rate": 1.918795550088623e-06,
      "loss": 0.4435,
      "num_input_tokens_seen": 22142368,
      "step": 33850
    },
    {
      "epoch": 17.7437106918239,
      "grad_norm": 0.6224979758262634,
      "learning_rate": 1.9144048650835496e-06,
      "loss": 0.3514,
      "num_input_tokens_seen": 22145312,
      "step": 33855
    },
    {
      "epoch": 17.746331236897273,
      "grad_norm": 0.44037964940071106,
      "learning_rate": 1.9100190093216093e-06,
      "loss": 0.4235,
      "num_input_tokens_seen": 22148480,
      "step": 33860
    },
    {
      "epoch": 17.74895178197065,
      "grad_norm": 0.36119329929351807,
      "learning_rate": 1.9056379837202627e-06,
      "loss": 0.3176,
      "num_input_tokens_seen": 22150944,
      "step": 33865
    },
    {
      "epoch": 17.751572327044027,
      "grad_norm": 0.3327023386955261,
      "learning_rate": 1.9012617891959766e-06,
      "loss": 0.3294,
      "num_input_tokens_seen": 22154496,
      "step": 33870
    },
    {
      "epoch": 17.7541928721174,
      "grad_norm": 0.5008952617645264,
      "learning_rate": 1.8968904266641996e-06,
      "loss": 0.3534,
      "num_input_tokens_seen": 22158048,
      "step": 33875
    },
    {
      "epoch": 17.756813417190777,
      "grad_norm": 0.5005996227264404,
      "learning_rate": 1.892523897039372e-06,
      "loss": 0.5208,
      "num_input_tokens_seen": 22161920,
      "step": 33880
    },
    {
      "epoch": 17.75943396226415,
      "grad_norm": 0.5408786535263062,
      "learning_rate": 1.888162201234922e-06,
      "loss": 0.6176,
      "num_input_tokens_seen": 22164640,
      "step": 33885
    },
    {
      "epoch": 17.762054507337528,
      "grad_norm": 0.3221195340156555,
      "learning_rate": 1.8838053401632666e-06,
      "loss": 0.3684,
      "num_input_tokens_seen": 22168832,
      "step": 33890
    },
    {
      "epoch": 17.7646750524109,
      "grad_norm": 0.47169172763824463,
      "learning_rate": 1.8794533147358074e-06,
      "loss": 0.5264,
      "num_input_tokens_seen": 22172192,
      "step": 33895
    },
    {
      "epoch": 17.767295597484278,
      "grad_norm": 0.4443964660167694,
      "learning_rate": 1.875106125862941e-06,
      "loss": 0.5245,
      "num_input_tokens_seen": 22176128,
      "step": 33900
    },
    {
      "epoch": 17.76991614255765,
      "grad_norm": 0.3439239263534546,
      "learning_rate": 1.870763774454057e-06,
      "loss": 0.5568,
      "num_input_tokens_seen": 22179168,
      "step": 33905
    },
    {
      "epoch": 17.77253668763103,
      "grad_norm": 0.6937708854675293,
      "learning_rate": 1.8664262614175115e-06,
      "loss": 0.5465,
      "num_input_tokens_seen": 22181728,
      "step": 33910
    },
    {
      "epoch": 17.7751572327044,
      "grad_norm": 0.545188844203949,
      "learning_rate": 1.862093587660671e-06,
      "loss": 0.4034,
      "num_input_tokens_seen": 22184480,
      "step": 33915
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 0.46322646737098694,
      "learning_rate": 1.8577657540898851e-06,
      "loss": 0.4729,
      "num_input_tokens_seen": 22187552,
      "step": 33920
    },
    {
      "epoch": 17.780398322851152,
      "grad_norm": 0.6714254021644592,
      "learning_rate": 1.8534427616104798e-06,
      "loss": 0.4495,
      "num_input_tokens_seen": 22190176,
      "step": 33925
    },
    {
      "epoch": 17.78301886792453,
      "grad_norm": 0.49651750922203064,
      "learning_rate": 1.8491246111267846e-06,
      "loss": 0.4205,
      "num_input_tokens_seen": 22192864,
      "step": 33930
    },
    {
      "epoch": 17.785639412997902,
      "grad_norm": 0.5380036234855652,
      "learning_rate": 1.8448113035420934e-06,
      "loss": 0.424,
      "num_input_tokens_seen": 22197056,
      "step": 33935
    },
    {
      "epoch": 17.78825995807128,
      "grad_norm": 0.5409848690032959,
      "learning_rate": 1.840502839758712e-06,
      "loss": 0.3826,
      "num_input_tokens_seen": 22199680,
      "step": 33940
    },
    {
      "epoch": 17.790880503144653,
      "grad_norm": 0.42997050285339355,
      "learning_rate": 1.8361992206779226e-06,
      "loss": 0.462,
      "num_input_tokens_seen": 22202592,
      "step": 33945
    },
    {
      "epoch": 17.79350104821803,
      "grad_norm": 0.6750734448432922,
      "learning_rate": 1.8319004471999856e-06,
      "loss": 0.4996,
      "num_input_tokens_seen": 22205024,
      "step": 33950
    },
    {
      "epoch": 17.796121593291403,
      "grad_norm": 0.38968488574028015,
      "learning_rate": 1.8276065202241592e-06,
      "loss": 0.3835,
      "num_input_tokens_seen": 22208224,
      "step": 33955
    },
    {
      "epoch": 17.79874213836478,
      "grad_norm": 0.38673925399780273,
      "learning_rate": 1.8233174406486863e-06,
      "loss": 0.4422,
      "num_input_tokens_seen": 22211456,
      "step": 33960
    },
    {
      "epoch": 17.801362683438157,
      "grad_norm": 0.7511290907859802,
      "learning_rate": 1.8190332093707907e-06,
      "loss": 0.7885,
      "num_input_tokens_seen": 22214528,
      "step": 33965
    },
    {
      "epoch": 17.80398322851153,
      "grad_norm": 0.760403037071228,
      "learning_rate": 1.8147538272866833e-06,
      "loss": 0.4692,
      "num_input_tokens_seen": 22217312,
      "step": 33970
    },
    {
      "epoch": 17.806603773584907,
      "grad_norm": 0.3654293119907379,
      "learning_rate": 1.8104792952915672e-06,
      "loss": 0.3893,
      "num_input_tokens_seen": 22219936,
      "step": 33975
    },
    {
      "epoch": 17.80922431865828,
      "grad_norm": 0.36714258790016174,
      "learning_rate": 1.8062096142796192e-06,
      "loss": 0.3735,
      "num_input_tokens_seen": 22223520,
      "step": 33980
    },
    {
      "epoch": 17.811844863731658,
      "grad_norm": 0.5843338370323181,
      "learning_rate": 1.8019447851440163e-06,
      "loss": 0.3769,
      "num_input_tokens_seen": 22226176,
      "step": 33985
    },
    {
      "epoch": 17.81446540880503,
      "grad_norm": 0.29166775941848755,
      "learning_rate": 1.7976848087769032e-06,
      "loss": 0.351,
      "num_input_tokens_seen": 22229056,
      "step": 33990
    },
    {
      "epoch": 17.817085953878408,
      "grad_norm": 0.3065871596336365,
      "learning_rate": 1.79342968606942e-06,
      "loss": 0.37,
      "num_input_tokens_seen": 22232096,
      "step": 33995
    },
    {
      "epoch": 17.81970649895178,
      "grad_norm": 0.7071395516395569,
      "learning_rate": 1.7891794179116933e-06,
      "loss": 0.4435,
      "num_input_tokens_seen": 22234624,
      "step": 34000
    },
    {
      "epoch": 17.822327044025158,
      "grad_norm": 0.5034889578819275,
      "learning_rate": 1.7849340051928316e-06,
      "loss": 0.3864,
      "num_input_tokens_seen": 22238176,
      "step": 34005
    },
    {
      "epoch": 17.82494758909853,
      "grad_norm": 1.497527003288269,
      "learning_rate": 1.780693448800927e-06,
      "loss": 0.3331,
      "num_input_tokens_seen": 22241280,
      "step": 34010
    },
    {
      "epoch": 17.82756813417191,
      "grad_norm": 0.6229028105735779,
      "learning_rate": 1.7764577496230478e-06,
      "loss": 0.4156,
      "num_input_tokens_seen": 22245088,
      "step": 34015
    },
    {
      "epoch": 17.830188679245282,
      "grad_norm": 0.6048948168754578,
      "learning_rate": 1.7722269085452602e-06,
      "loss": 0.5867,
      "num_input_tokens_seen": 22247584,
      "step": 34020
    },
    {
      "epoch": 17.83280922431866,
      "grad_norm": 0.27173322439193726,
      "learning_rate": 1.7680009264526038e-06,
      "loss": 0.4148,
      "num_input_tokens_seen": 22250752,
      "step": 34025
    },
    {
      "epoch": 17.835429769392032,
      "grad_norm": 0.4325370192527771,
      "learning_rate": 1.7637798042291126e-06,
      "loss": 0.398,
      "num_input_tokens_seen": 22254720,
      "step": 34030
    },
    {
      "epoch": 17.83805031446541,
      "grad_norm": 0.34299737215042114,
      "learning_rate": 1.7595635427577944e-06,
      "loss": 0.4405,
      "num_input_tokens_seen": 22257600,
      "step": 34035
    },
    {
      "epoch": 17.840670859538783,
      "grad_norm": 0.509465217590332,
      "learning_rate": 1.7553521429206465e-06,
      "loss": 0.3101,
      "num_input_tokens_seen": 22260896,
      "step": 34040
    },
    {
      "epoch": 17.84329140461216,
      "grad_norm": 0.512565553188324,
      "learning_rate": 1.7511456055986335e-06,
      "loss": 0.3359,
      "num_input_tokens_seen": 22264864,
      "step": 34045
    },
    {
      "epoch": 17.845911949685533,
      "grad_norm": 0.841330885887146,
      "learning_rate": 1.746943931671724e-06,
      "loss": 0.4753,
      "num_input_tokens_seen": 22268064,
      "step": 34050
    },
    {
      "epoch": 17.84853249475891,
      "grad_norm": 0.527561366558075,
      "learning_rate": 1.742747122018859e-06,
      "loss": 0.4735,
      "num_input_tokens_seen": 22271488,
      "step": 34055
    },
    {
      "epoch": 17.851153039832283,
      "grad_norm": 0.38641175627708435,
      "learning_rate": 1.738555177517967e-06,
      "loss": 0.4476,
      "num_input_tokens_seen": 22275232,
      "step": 34060
    },
    {
      "epoch": 17.85377358490566,
      "grad_norm": 0.2700237035751343,
      "learning_rate": 1.734368099045952e-06,
      "loss": 0.3742,
      "num_input_tokens_seen": 22278720,
      "step": 34065
    },
    {
      "epoch": 17.856394129979037,
      "grad_norm": 0.2905307114124298,
      "learning_rate": 1.730185887478697e-06,
      "loss": 0.3391,
      "num_input_tokens_seen": 22281920,
      "step": 34070
    },
    {
      "epoch": 17.85901467505241,
      "grad_norm": 0.46963560581207275,
      "learning_rate": 1.7260085436910794e-06,
      "loss": 0.3841,
      "num_input_tokens_seen": 22284768,
      "step": 34075
    },
    {
      "epoch": 17.861635220125788,
      "grad_norm": 0.6764570474624634,
      "learning_rate": 1.7218360685569506e-06,
      "loss": 0.5044,
      "num_input_tokens_seen": 22287424,
      "step": 34080
    },
    {
      "epoch": 17.86425576519916,
      "grad_norm": 0.4022023379802704,
      "learning_rate": 1.7176684629491513e-06,
      "loss": 0.2989,
      "num_input_tokens_seen": 22293024,
      "step": 34085
    },
    {
      "epoch": 17.866876310272538,
      "grad_norm": 0.40993428230285645,
      "learning_rate": 1.7135057277394868e-06,
      "loss": 0.392,
      "num_input_tokens_seen": 22296256,
      "step": 34090
    },
    {
      "epoch": 17.86949685534591,
      "grad_norm": 0.3130556643009186,
      "learning_rate": 1.7093478637987548e-06,
      "loss": 0.4892,
      "num_input_tokens_seen": 22298752,
      "step": 34095
    },
    {
      "epoch": 17.872117400419288,
      "grad_norm": 0.33327528834342957,
      "learning_rate": 1.7051948719967375e-06,
      "loss": 0.5442,
      "num_input_tokens_seen": 22302176,
      "step": 34100
    },
    {
      "epoch": 17.87473794549266,
      "grad_norm": 0.46008312702178955,
      "learning_rate": 1.7010467532021901e-06,
      "loss": 0.418,
      "num_input_tokens_seen": 22305696,
      "step": 34105
    },
    {
      "epoch": 17.87735849056604,
      "grad_norm": 0.4639376401901245,
      "learning_rate": 1.6969035082828544e-06,
      "loss": 0.4225,
      "num_input_tokens_seen": 22308928,
      "step": 34110
    },
    {
      "epoch": 17.879979035639412,
      "grad_norm": 0.44589293003082275,
      "learning_rate": 1.692765138105451e-06,
      "loss": 0.6103,
      "num_input_tokens_seen": 22311936,
      "step": 34115
    },
    {
      "epoch": 17.88259958071279,
      "grad_norm": 0.3909783959388733,
      "learning_rate": 1.6886316435356792e-06,
      "loss": 0.4214,
      "num_input_tokens_seen": 22314624,
      "step": 34120
    },
    {
      "epoch": 17.885220125786162,
      "grad_norm": 0.5138084292411804,
      "learning_rate": 1.6845030254382138e-06,
      "loss": 0.4202,
      "num_input_tokens_seen": 22317696,
      "step": 34125
    },
    {
      "epoch": 17.88784067085954,
      "grad_norm": 0.567518413066864,
      "learning_rate": 1.6803792846767197e-06,
      "loss": 0.4462,
      "num_input_tokens_seen": 22324832,
      "step": 34130
    },
    {
      "epoch": 17.890461215932913,
      "grad_norm": 0.7130174040794373,
      "learning_rate": 1.6762604221138316e-06,
      "loss": 0.4189,
      "num_input_tokens_seen": 22328064,
      "step": 34135
    },
    {
      "epoch": 17.89308176100629,
      "grad_norm": 0.49601149559020996,
      "learning_rate": 1.6721464386111774e-06,
      "loss": 0.3756,
      "num_input_tokens_seen": 22330656,
      "step": 34140
    },
    {
      "epoch": 17.895702306079663,
      "grad_norm": 0.481457382440567,
      "learning_rate": 1.6680373350293516e-06,
      "loss": 0.4636,
      "num_input_tokens_seen": 22333760,
      "step": 34145
    },
    {
      "epoch": 17.89832285115304,
      "grad_norm": 0.5700123906135559,
      "learning_rate": 1.6639331122279223e-06,
      "loss": 0.6112,
      "num_input_tokens_seen": 22336256,
      "step": 34150
    },
    {
      "epoch": 17.900943396226417,
      "grad_norm": 0.4685802459716797,
      "learning_rate": 1.6598337710654583e-06,
      "loss": 0.5446,
      "num_input_tokens_seen": 22339520,
      "step": 34155
    },
    {
      "epoch": 17.90356394129979,
      "grad_norm": 0.7120426297187805,
      "learning_rate": 1.6557393123994874e-06,
      "loss": 0.4356,
      "num_input_tokens_seen": 22343200,
      "step": 34160
    },
    {
      "epoch": 17.906184486373167,
      "grad_norm": 0.7639592885971069,
      "learning_rate": 1.651649737086533e-06,
      "loss": 0.5706,
      "num_input_tokens_seen": 22347712,
      "step": 34165
    },
    {
      "epoch": 17.90880503144654,
      "grad_norm": 0.8598709106445312,
      "learning_rate": 1.6475650459820774e-06,
      "loss": 0.4706,
      "num_input_tokens_seen": 22350624,
      "step": 34170
    },
    {
      "epoch": 17.911425576519918,
      "grad_norm": 0.4624975025653839,
      "learning_rate": 1.643485239940598e-06,
      "loss": 0.3738,
      "num_input_tokens_seen": 22354112,
      "step": 34175
    },
    {
      "epoch": 17.91404612159329,
      "grad_norm": 0.4433470368385315,
      "learning_rate": 1.6394103198155375e-06,
      "loss": 0.4201,
      "num_input_tokens_seen": 22357696,
      "step": 34180
    },
    {
      "epoch": 17.916666666666668,
      "grad_norm": 0.6330662965774536,
      "learning_rate": 1.6353402864593276e-06,
      "loss": 0.472,
      "num_input_tokens_seen": 22361024,
      "step": 34185
    },
    {
      "epoch": 17.91928721174004,
      "grad_norm": 0.43500903248786926,
      "learning_rate": 1.6312751407233706e-06,
      "loss": 0.3655,
      "num_input_tokens_seen": 22365056,
      "step": 34190
    },
    {
      "epoch": 17.921907756813418,
      "grad_norm": 0.2917952537536621,
      "learning_rate": 1.6272148834580559e-06,
      "loss": 0.4644,
      "num_input_tokens_seen": 22368384,
      "step": 34195
    },
    {
      "epoch": 17.92452830188679,
      "grad_norm": 0.40543872117996216,
      "learning_rate": 1.6231595155127317e-06,
      "loss": 0.4445,
      "num_input_tokens_seen": 22371040,
      "step": 34200
    },
    {
      "epoch": 17.92714884696017,
      "grad_norm": 0.392965704202652,
      "learning_rate": 1.6191090377357414e-06,
      "loss": 0.3126,
      "num_input_tokens_seen": 22373664,
      "step": 34205
    },
    {
      "epoch": 17.929769392033542,
      "grad_norm": 0.48304760456085205,
      "learning_rate": 1.6150634509743939e-06,
      "loss": 0.4506,
      "num_input_tokens_seen": 22376672,
      "step": 34210
    },
    {
      "epoch": 17.93238993710692,
      "grad_norm": 0.6821635961532593,
      "learning_rate": 1.611022756074984e-06,
      "loss": 0.3932,
      "num_input_tokens_seen": 22379200,
      "step": 34215
    },
    {
      "epoch": 17.935010482180292,
      "grad_norm": 0.5163688063621521,
      "learning_rate": 1.606986953882783e-06,
      "loss": 0.3322,
      "num_input_tokens_seen": 22382368,
      "step": 34220
    },
    {
      "epoch": 17.93763102725367,
      "grad_norm": 0.8295137882232666,
      "learning_rate": 1.6029560452420238e-06,
      "loss": 0.5696,
      "num_input_tokens_seen": 22385728,
      "step": 34225
    },
    {
      "epoch": 17.940251572327043,
      "grad_norm": 0.5335580110549927,
      "learning_rate": 1.5989300309959377e-06,
      "loss": 0.5492,
      "num_input_tokens_seen": 22388352,
      "step": 34230
    },
    {
      "epoch": 17.94287211740042,
      "grad_norm": 0.3711804449558258,
      "learning_rate": 1.5949089119867116e-06,
      "loss": 0.4872,
      "num_input_tokens_seen": 22391456,
      "step": 34235
    },
    {
      "epoch": 17.945492662473793,
      "grad_norm": 0.405874103307724,
      "learning_rate": 1.5908926890555204e-06,
      "loss": 0.5837,
      "num_input_tokens_seen": 22395392,
      "step": 34240
    },
    {
      "epoch": 17.94811320754717,
      "grad_norm": 0.45126017928123474,
      "learning_rate": 1.5868813630425167e-06,
      "loss": 0.4905,
      "num_input_tokens_seen": 22398880,
      "step": 34245
    },
    {
      "epoch": 17.950733752620543,
      "grad_norm": 0.7941692471504211,
      "learning_rate": 1.5828749347868183e-06,
      "loss": 0.4009,
      "num_input_tokens_seen": 22401216,
      "step": 34250
    },
    {
      "epoch": 17.95335429769392,
      "grad_norm": 0.9606340527534485,
      "learning_rate": 1.5788734051265268e-06,
      "loss": 0.3286,
      "num_input_tokens_seen": 22404320,
      "step": 34255
    },
    {
      "epoch": 17.955974842767297,
      "grad_norm": 0.6352871060371399,
      "learning_rate": 1.5748767748987231e-06,
      "loss": 0.4543,
      "num_input_tokens_seen": 22407296,
      "step": 34260
    },
    {
      "epoch": 17.95859538784067,
      "grad_norm": 0.3844006657600403,
      "learning_rate": 1.5708850449394436e-06,
      "loss": 0.4065,
      "num_input_tokens_seen": 22409728,
      "step": 34265
    },
    {
      "epoch": 17.961215932914047,
      "grad_norm": 1.029091715812683,
      "learning_rate": 1.5668982160837176e-06,
      "loss": 0.4041,
      "num_input_tokens_seen": 22412864,
      "step": 34270
    },
    {
      "epoch": 17.96383647798742,
      "grad_norm": 0.543002188205719,
      "learning_rate": 1.562916289165553e-06,
      "loss": 0.3765,
      "num_input_tokens_seen": 22415968,
      "step": 34275
    },
    {
      "epoch": 17.966457023060798,
      "grad_norm": 0.4049091637134552,
      "learning_rate": 1.5589392650179114e-06,
      "loss": 0.4252,
      "num_input_tokens_seen": 22418784,
      "step": 34280
    },
    {
      "epoch": 17.96907756813417,
      "grad_norm": 0.2624971568584442,
      "learning_rate": 1.5549671444727437e-06,
      "loss": 0.3834,
      "num_input_tokens_seen": 22422880,
      "step": 34285
    },
    {
      "epoch": 17.971698113207548,
      "grad_norm": 0.3078940212726593,
      "learning_rate": 1.55099992836098e-06,
      "loss": 0.3992,
      "num_input_tokens_seen": 22426528,
      "step": 34290
    },
    {
      "epoch": 17.97431865828092,
      "grad_norm": 0.7490604519844055,
      "learning_rate": 1.5470376175125061e-06,
      "loss": 0.4151,
      "num_input_tokens_seen": 22429536,
      "step": 34295
    },
    {
      "epoch": 17.9769392033543,
      "grad_norm": 0.3750911056995392,
      "learning_rate": 1.5430802127562038e-06,
      "loss": 0.3932,
      "num_input_tokens_seen": 22433952,
      "step": 34300
    },
    {
      "epoch": 17.979559748427672,
      "grad_norm": 0.49598491191864014,
      "learning_rate": 1.5391277149199023e-06,
      "loss": 0.4607,
      "num_input_tokens_seen": 22437856,
      "step": 34305
    },
    {
      "epoch": 17.98218029350105,
      "grad_norm": 0.7540782690048218,
      "learning_rate": 1.5351801248304292e-06,
      "loss": 0.4143,
      "num_input_tokens_seen": 22440896,
      "step": 34310
    },
    {
      "epoch": 17.984800838574422,
      "grad_norm": 0.4704655408859253,
      "learning_rate": 1.5312374433135767e-06,
      "loss": 0.5518,
      "num_input_tokens_seen": 22444480,
      "step": 34315
    },
    {
      "epoch": 17.9874213836478,
      "grad_norm": 0.5054172873497009,
      "learning_rate": 1.5272996711941018e-06,
      "loss": 0.3134,
      "num_input_tokens_seen": 22448448,
      "step": 34320
    },
    {
      "epoch": 17.990041928721173,
      "grad_norm": 1.6528068780899048,
      "learning_rate": 1.5233668092957482e-06,
      "loss": 0.6564,
      "num_input_tokens_seen": 22451200,
      "step": 34325
    },
    {
      "epoch": 17.99266247379455,
      "grad_norm": 0.5294209122657776,
      "learning_rate": 1.5194388584412217e-06,
      "loss": 0.4337,
      "num_input_tokens_seen": 22454976,
      "step": 34330
    },
    {
      "epoch": 17.995283018867923,
      "grad_norm": 0.3417864143848419,
      "learning_rate": 1.5155158194522013e-06,
      "loss": 0.4328,
      "num_input_tokens_seen": 22458240,
      "step": 34335
    },
    {
      "epoch": 17.9979035639413,
      "grad_norm": 0.3845472037792206,
      "learning_rate": 1.5115976931493498e-06,
      "loss": 0.4032,
      "num_input_tokens_seen": 22461056,
      "step": 34340
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.4767856299877167,
      "eval_runtime": 9.6606,
      "eval_samples_per_second": 87.779,
      "eval_steps_per_second": 21.945,
      "num_input_tokens_seen": 22462808,
      "step": 34344
    },
    {
      "epoch": 18.000524109014677,
      "grad_norm": 0.5503536462783813,
      "learning_rate": 1.5076844803522922e-06,
      "loss": 0.3041,
      "num_input_tokens_seen": 22463384,
      "step": 34345
    },
    {
      "epoch": 18.00314465408805,
      "grad_norm": 0.596266508102417,
      "learning_rate": 1.5037761818796236e-06,
      "loss": 0.3525,
      "num_input_tokens_seen": 22467384,
      "step": 34350
    },
    {
      "epoch": 18.005765199161427,
      "grad_norm": 0.40325284004211426,
      "learning_rate": 1.4998727985489235e-06,
      "loss": 0.4818,
      "num_input_tokens_seen": 22470840,
      "step": 34355
    },
    {
      "epoch": 18.0083857442348,
      "grad_norm": 0.35568028688430786,
      "learning_rate": 1.4959743311767249e-06,
      "loss": 0.4413,
      "num_input_tokens_seen": 22474456,
      "step": 34360
    },
    {
      "epoch": 18.011006289308177,
      "grad_norm": 0.6028571128845215,
      "learning_rate": 1.4920807805785502e-06,
      "loss": 0.4193,
      "num_input_tokens_seen": 22477400,
      "step": 34365
    },
    {
      "epoch": 18.01362683438155,
      "grad_norm": 0.8318329453468323,
      "learning_rate": 1.4881921475688815e-06,
      "loss": 0.4735,
      "num_input_tokens_seen": 22480728,
      "step": 34370
    },
    {
      "epoch": 18.016247379454928,
      "grad_norm": 0.5366427302360535,
      "learning_rate": 1.4843084329611844e-06,
      "loss": 0.5527,
      "num_input_tokens_seen": 22483768,
      "step": 34375
    },
    {
      "epoch": 18.0188679245283,
      "grad_norm": 0.5908421874046326,
      "learning_rate": 1.480429637567879e-06,
      "loss": 0.5081,
      "num_input_tokens_seen": 22486808,
      "step": 34380
    },
    {
      "epoch": 18.021488469601678,
      "grad_norm": 0.35996803641319275,
      "learning_rate": 1.4765557622003684e-06,
      "loss": 0.3919,
      "num_input_tokens_seen": 22490168,
      "step": 34385
    },
    {
      "epoch": 18.02410901467505,
      "grad_norm": 0.41701918840408325,
      "learning_rate": 1.4726868076690186e-06,
      "loss": 0.4199,
      "num_input_tokens_seen": 22493528,
      "step": 34390
    },
    {
      "epoch": 18.02672955974843,
      "grad_norm": 0.552900493144989,
      "learning_rate": 1.4688227747831767e-06,
      "loss": 0.4347,
      "num_input_tokens_seen": 22496344,
      "step": 34395
    },
    {
      "epoch": 18.029350104821802,
      "grad_norm": 0.7525166273117065,
      "learning_rate": 1.4649636643511573e-06,
      "loss": 0.4748,
      "num_input_tokens_seen": 22498776,
      "step": 34400
    },
    {
      "epoch": 18.03197064989518,
      "grad_norm": 0.6160087585449219,
      "learning_rate": 1.4611094771802364e-06,
      "loss": 0.435,
      "num_input_tokens_seen": 22501464,
      "step": 34405
    },
    {
      "epoch": 18.034591194968552,
      "grad_norm": 0.3238847553730011,
      "learning_rate": 1.457260214076664e-06,
      "loss": 0.3872,
      "num_input_tokens_seen": 22504696,
      "step": 34410
    },
    {
      "epoch": 18.03721174004193,
      "grad_norm": 0.4732095003128052,
      "learning_rate": 1.4534158758456623e-06,
      "loss": 0.4346,
      "num_input_tokens_seen": 22507288,
      "step": 34415
    },
    {
      "epoch": 18.039832285115303,
      "grad_norm": 0.6998249888420105,
      "learning_rate": 1.4495764632914271e-06,
      "loss": 0.4563,
      "num_input_tokens_seen": 22509912,
      "step": 34420
    },
    {
      "epoch": 18.04245283018868,
      "grad_norm": 0.44294747710227966,
      "learning_rate": 1.4457419772171215e-06,
      "loss": 0.4612,
      "num_input_tokens_seen": 22513432,
      "step": 34425
    },
    {
      "epoch": 18.045073375262053,
      "grad_norm": 0.4316001832485199,
      "learning_rate": 1.4419124184248733e-06,
      "loss": 0.4244,
      "num_input_tokens_seen": 22516632,
      "step": 34430
    },
    {
      "epoch": 18.04769392033543,
      "grad_norm": 0.5577505230903625,
      "learning_rate": 1.4380877877157834e-06,
      "loss": 0.3407,
      "num_input_tokens_seen": 22520376,
      "step": 34435
    },
    {
      "epoch": 18.050314465408807,
      "grad_norm": 0.3120166063308716,
      "learning_rate": 1.4342680858899143e-06,
      "loss": 0.404,
      "num_input_tokens_seen": 22524728,
      "step": 34440
    },
    {
      "epoch": 18.05293501048218,
      "grad_norm": 0.36570286750793457,
      "learning_rate": 1.4304533137463134e-06,
      "loss": 0.4287,
      "num_input_tokens_seen": 22528408,
      "step": 34445
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 0.4067554771900177,
      "learning_rate": 1.4266434720829808e-06,
      "loss": 0.4484,
      "num_input_tokens_seen": 22531032,
      "step": 34450
    },
    {
      "epoch": 18.05817610062893,
      "grad_norm": 0.26561403274536133,
      "learning_rate": 1.4228385616968987e-06,
      "loss": 0.3761,
      "num_input_tokens_seen": 22533944,
      "step": 34455
    },
    {
      "epoch": 18.060796645702307,
      "grad_norm": 0.29445138573646545,
      "learning_rate": 1.419038583384008e-06,
      "loss": 0.4718,
      "num_input_tokens_seen": 22536952,
      "step": 34460
    },
    {
      "epoch": 18.06341719077568,
      "grad_norm": 0.3606609106063843,
      "learning_rate": 1.4152435379392148e-06,
      "loss": 0.4426,
      "num_input_tokens_seen": 22540376,
      "step": 34465
    },
    {
      "epoch": 18.066037735849058,
      "grad_norm": 0.46780088543891907,
      "learning_rate": 1.411453426156406e-06,
      "loss": 0.3966,
      "num_input_tokens_seen": 22543352,
      "step": 34470
    },
    {
      "epoch": 18.06865828092243,
      "grad_norm": 0.408916175365448,
      "learning_rate": 1.4076682488284282e-06,
      "loss": 0.3374,
      "num_input_tokens_seen": 22546808,
      "step": 34475
    },
    {
      "epoch": 18.071278825995808,
      "grad_norm": 0.3151216208934784,
      "learning_rate": 1.4038880067470978e-06,
      "loss": 0.3223,
      "num_input_tokens_seen": 22550008,
      "step": 34480
    },
    {
      "epoch": 18.07389937106918,
      "grad_norm": 0.48467451333999634,
      "learning_rate": 1.400112700703196e-06,
      "loss": 0.3315,
      "num_input_tokens_seen": 22552824,
      "step": 34485
    },
    {
      "epoch": 18.07651991614256,
      "grad_norm": 0.5168315172195435,
      "learning_rate": 1.3963423314864805e-06,
      "loss": 0.4309,
      "num_input_tokens_seen": 22555608,
      "step": 34490
    },
    {
      "epoch": 18.079140461215932,
      "grad_norm": 0.47781217098236084,
      "learning_rate": 1.3925768998856587e-06,
      "loss": 0.5224,
      "num_input_tokens_seen": 22559640,
      "step": 34495
    },
    {
      "epoch": 18.08176100628931,
      "grad_norm": 0.5510442852973938,
      "learning_rate": 1.388816406688423e-06,
      "loss": 0.4828,
      "num_input_tokens_seen": 22562936,
      "step": 34500
    },
    {
      "epoch": 18.084381551362682,
      "grad_norm": 0.5162073969841003,
      "learning_rate": 1.3850608526814246e-06,
      "loss": 0.4534,
      "num_input_tokens_seen": 22566424,
      "step": 34505
    },
    {
      "epoch": 18.08700209643606,
      "grad_norm": 0.4215719699859619,
      "learning_rate": 1.3813102386502853e-06,
      "loss": 0.4694,
      "num_input_tokens_seen": 22569944,
      "step": 34510
    },
    {
      "epoch": 18.089622641509433,
      "grad_norm": 0.4997224807739258,
      "learning_rate": 1.3775645653795881e-06,
      "loss": 0.4019,
      "num_input_tokens_seen": 22574264,
      "step": 34515
    },
    {
      "epoch": 18.09224318658281,
      "grad_norm": 0.760897696018219,
      "learning_rate": 1.373823833652882e-06,
      "loss": 0.3704,
      "num_input_tokens_seen": 22577080,
      "step": 34520
    },
    {
      "epoch": 18.094863731656183,
      "grad_norm": 0.3031993806362152,
      "learning_rate": 1.3700880442526876e-06,
      "loss": 0.4874,
      "num_input_tokens_seen": 22581048,
      "step": 34525
    },
    {
      "epoch": 18.09748427672956,
      "grad_norm": 0.5652111768722534,
      "learning_rate": 1.3663571979604884e-06,
      "loss": 0.5304,
      "num_input_tokens_seen": 22584088,
      "step": 34530
    },
    {
      "epoch": 18.100104821802937,
      "grad_norm": 0.3739142119884491,
      "learning_rate": 1.3626312955567406e-06,
      "loss": 0.3093,
      "num_input_tokens_seen": 22587480,
      "step": 34535
    },
    {
      "epoch": 18.10272536687631,
      "grad_norm": 0.3816404640674591,
      "learning_rate": 1.358910337820854e-06,
      "loss": 0.5685,
      "num_input_tokens_seen": 22591032,
      "step": 34540
    },
    {
      "epoch": 18.105345911949687,
      "grad_norm": 0.5971994400024414,
      "learning_rate": 1.3551943255312172e-06,
      "loss": 0.3292,
      "num_input_tokens_seen": 22594360,
      "step": 34545
    },
    {
      "epoch": 18.10796645702306,
      "grad_norm": 0.4125392436981201,
      "learning_rate": 1.3514832594651694e-06,
      "loss": 0.4052,
      "num_input_tokens_seen": 22597400,
      "step": 34550
    },
    {
      "epoch": 18.110587002096437,
      "grad_norm": 0.4906703233718872,
      "learning_rate": 1.347777140399026e-06,
      "loss": 0.4356,
      "num_input_tokens_seen": 22600056,
      "step": 34555
    },
    {
      "epoch": 18.11320754716981,
      "grad_norm": 0.34532132744789124,
      "learning_rate": 1.344075969108069e-06,
      "loss": 0.3958,
      "num_input_tokens_seen": 22603128,
      "step": 34560
    },
    {
      "epoch": 18.115828092243188,
      "grad_norm": 0.5308987498283386,
      "learning_rate": 1.3403797463665352e-06,
      "loss": 0.5045,
      "num_input_tokens_seen": 22606040,
      "step": 34565
    },
    {
      "epoch": 18.11844863731656,
      "grad_norm": 0.9754218459129333,
      "learning_rate": 1.3366884729476365e-06,
      "loss": 0.449,
      "num_input_tokens_seen": 22608632,
      "step": 34570
    },
    {
      "epoch": 18.121069182389938,
      "grad_norm": 0.7139639854431152,
      "learning_rate": 1.3330021496235467e-06,
      "loss": 0.4851,
      "num_input_tokens_seen": 22611512,
      "step": 34575
    },
    {
      "epoch": 18.12368972746331,
      "grad_norm": 0.7395851016044617,
      "learning_rate": 1.3293207771653965e-06,
      "loss": 0.3438,
      "num_input_tokens_seen": 22615352,
      "step": 34580
    },
    {
      "epoch": 18.12631027253669,
      "grad_norm": 0.3965725600719452,
      "learning_rate": 1.3256443563432891e-06,
      "loss": 0.4322,
      "num_input_tokens_seen": 22618328,
      "step": 34585
    },
    {
      "epoch": 18.128930817610062,
      "grad_norm": 0.6348749995231628,
      "learning_rate": 1.3219728879262956e-06,
      "loss": 0.428,
      "num_input_tokens_seen": 22621816,
      "step": 34590
    },
    {
      "epoch": 18.13155136268344,
      "grad_norm": 0.41487735509872437,
      "learning_rate": 1.3183063726824375e-06,
      "loss": 0.4663,
      "num_input_tokens_seen": 22626328,
      "step": 34595
    },
    {
      "epoch": 18.134171907756812,
      "grad_norm": 0.7633062601089478,
      "learning_rate": 1.3146448113787151e-06,
      "loss": 0.6298,
      "num_input_tokens_seen": 22630104,
      "step": 34600
    },
    {
      "epoch": 18.13679245283019,
      "grad_norm": 0.608248770236969,
      "learning_rate": 1.3109882047810773e-06,
      "loss": 0.3356,
      "num_input_tokens_seen": 22634040,
      "step": 34605
    },
    {
      "epoch": 18.139412997903563,
      "grad_norm": 0.8364596962928772,
      "learning_rate": 1.307336553654448e-06,
      "loss": 0.383,
      "num_input_tokens_seen": 22636792,
      "step": 34610
    },
    {
      "epoch": 18.14203354297694,
      "grad_norm": 0.5120129585266113,
      "learning_rate": 1.303689858762716e-06,
      "loss": 0.4494,
      "num_input_tokens_seen": 22640152,
      "step": 34615
    },
    {
      "epoch": 18.144654088050313,
      "grad_norm": 1.6131091117858887,
      "learning_rate": 1.3000481208687188e-06,
      "loss": 0.5712,
      "num_input_tokens_seen": 22643128,
      "step": 34620
    },
    {
      "epoch": 18.14727463312369,
      "grad_norm": 0.6840927600860596,
      "learning_rate": 1.2964113407342743e-06,
      "loss": 0.4671,
      "num_input_tokens_seen": 22646168,
      "step": 34625
    },
    {
      "epoch": 18.149895178197063,
      "grad_norm": 0.6189802885055542,
      "learning_rate": 1.292779519120152e-06,
      "loss": 0.4163,
      "num_input_tokens_seen": 22649144,
      "step": 34630
    },
    {
      "epoch": 18.15251572327044,
      "grad_norm": 0.6446345448493958,
      "learning_rate": 1.2891526567860856e-06,
      "loss": 0.4681,
      "num_input_tokens_seen": 22651704,
      "step": 34635
    },
    {
      "epoch": 18.155136268343817,
      "grad_norm": 0.6003907322883606,
      "learning_rate": 1.2855307544907825e-06,
      "loss": 0.5135,
      "num_input_tokens_seen": 22654840,
      "step": 34640
    },
    {
      "epoch": 18.15775681341719,
      "grad_norm": 0.4041139781475067,
      "learning_rate": 1.281913812991889e-06,
      "loss": 0.379,
      "num_input_tokens_seen": 22658392,
      "step": 34645
    },
    {
      "epoch": 18.160377358490567,
      "grad_norm": 0.3556372821331024,
      "learning_rate": 1.2783018330460389e-06,
      "loss": 0.4562,
      "num_input_tokens_seen": 22661592,
      "step": 34650
    },
    {
      "epoch": 18.16299790356394,
      "grad_norm": 0.6676614880561829,
      "learning_rate": 1.2746948154088111e-06,
      "loss": 0.5374,
      "num_input_tokens_seen": 22664312,
      "step": 34655
    },
    {
      "epoch": 18.165618448637318,
      "grad_norm": 0.5456894040107727,
      "learning_rate": 1.2710927608347605e-06,
      "loss": 0.5669,
      "num_input_tokens_seen": 22667032,
      "step": 34660
    },
    {
      "epoch": 18.16823899371069,
      "grad_norm": 0.32006216049194336,
      "learning_rate": 1.2674956700773872e-06,
      "loss": 0.542,
      "num_input_tokens_seen": 22670008,
      "step": 34665
    },
    {
      "epoch": 18.170859538784068,
      "grad_norm": 0.445211797952652,
      "learning_rate": 1.2639035438891666e-06,
      "loss": 0.3882,
      "num_input_tokens_seen": 22673656,
      "step": 34670
    },
    {
      "epoch": 18.17348008385744,
      "grad_norm": 0.6148659586906433,
      "learning_rate": 1.260316383021526e-06,
      "loss": 0.344,
      "num_input_tokens_seen": 22676792,
      "step": 34675
    },
    {
      "epoch": 18.17610062893082,
      "grad_norm": 0.7308721542358398,
      "learning_rate": 1.2567341882248645e-06,
      "loss": 0.5741,
      "num_input_tokens_seen": 22680056,
      "step": 34680
    },
    {
      "epoch": 18.178721174004192,
      "grad_norm": 1.1103413105010986,
      "learning_rate": 1.2531569602485305e-06,
      "loss": 0.4333,
      "num_input_tokens_seen": 22683224,
      "step": 34685
    },
    {
      "epoch": 18.18134171907757,
      "grad_norm": 0.311842143535614,
      "learning_rate": 1.2495846998408445e-06,
      "loss": 0.423,
      "num_input_tokens_seen": 22686648,
      "step": 34690
    },
    {
      "epoch": 18.183962264150942,
      "grad_norm": 0.46430879831314087,
      "learning_rate": 1.246017407749084e-06,
      "loss": 0.3863,
      "num_input_tokens_seen": 22688728,
      "step": 34695
    },
    {
      "epoch": 18.18658280922432,
      "grad_norm": 0.4595966041088104,
      "learning_rate": 1.2424550847194739e-06,
      "loss": 0.3768,
      "num_input_tokens_seen": 22692184,
      "step": 34700
    },
    {
      "epoch": 18.189203354297693,
      "grad_norm": 0.3069104850292206,
      "learning_rate": 1.2388977314972238e-06,
      "loss": 0.4164,
      "num_input_tokens_seen": 22695672,
      "step": 34705
    },
    {
      "epoch": 18.19182389937107,
      "grad_norm": 0.4143788516521454,
      "learning_rate": 1.235345348826486e-06,
      "loss": 0.3955,
      "num_input_tokens_seen": 22698520,
      "step": 34710
    },
    {
      "epoch": 18.194444444444443,
      "grad_norm": 0.42943209409713745,
      "learning_rate": 1.2317979374503825e-06,
      "loss": 0.4149,
      "num_input_tokens_seen": 22701784,
      "step": 34715
    },
    {
      "epoch": 18.19706498951782,
      "grad_norm": 0.4109768569469452,
      "learning_rate": 1.228255498110989e-06,
      "loss": 0.3514,
      "num_input_tokens_seen": 22705080,
      "step": 34720
    },
    {
      "epoch": 18.199685534591197,
      "grad_norm": 0.3624746799468994,
      "learning_rate": 1.224718031549338e-06,
      "loss": 0.3898,
      "num_input_tokens_seen": 22708152,
      "step": 34725
    },
    {
      "epoch": 18.20230607966457,
      "grad_norm": 0.661586582660675,
      "learning_rate": 1.2211855385054344e-06,
      "loss": 0.4084,
      "num_input_tokens_seen": 22711640,
      "step": 34730
    },
    {
      "epoch": 18.204926624737947,
      "grad_norm": 0.3774089515209198,
      "learning_rate": 1.217658019718232e-06,
      "loss": 0.4307,
      "num_input_tokens_seen": 22714936,
      "step": 34735
    },
    {
      "epoch": 18.20754716981132,
      "grad_norm": 0.5022189021110535,
      "learning_rate": 1.2141354759256484e-06,
      "loss": 0.4526,
      "num_input_tokens_seen": 22718456,
      "step": 34740
    },
    {
      "epoch": 18.210167714884697,
      "grad_norm": 0.6130495667457581,
      "learning_rate": 1.2106179078645641e-06,
      "loss": 0.5535,
      "num_input_tokens_seen": 22721400,
      "step": 34745
    },
    {
      "epoch": 18.21278825995807,
      "grad_norm": 0.42995721101760864,
      "learning_rate": 1.2071053162708096e-06,
      "loss": 0.4463,
      "num_input_tokens_seen": 22724312,
      "step": 34750
    },
    {
      "epoch": 18.215408805031448,
      "grad_norm": 0.5733381509780884,
      "learning_rate": 1.203597701879175e-06,
      "loss": 0.476,
      "num_input_tokens_seen": 22727128,
      "step": 34755
    },
    {
      "epoch": 18.21802935010482,
      "grad_norm": 0.292756587266922,
      "learning_rate": 1.2000950654234177e-06,
      "loss": 0.4841,
      "num_input_tokens_seen": 22730712,
      "step": 34760
    },
    {
      "epoch": 18.220649895178198,
      "grad_norm": 0.560217559337616,
      "learning_rate": 1.1965974076362491e-06,
      "loss": 0.4971,
      "num_input_tokens_seen": 22733592,
      "step": 34765
    },
    {
      "epoch": 18.22327044025157,
      "grad_norm": 0.3969316780567169,
      "learning_rate": 1.193104729249342e-06,
      "loss": 0.5951,
      "num_input_tokens_seen": 22736568,
      "step": 34770
    },
    {
      "epoch": 18.22589098532495,
      "grad_norm": 0.4700200855731964,
      "learning_rate": 1.189617030993323e-06,
      "loss": 0.4582,
      "num_input_tokens_seen": 22740696,
      "step": 34775
    },
    {
      "epoch": 18.228511530398322,
      "grad_norm": 0.47632279992103577,
      "learning_rate": 1.1861343135977753e-06,
      "loss": 0.3931,
      "num_input_tokens_seen": 22744728,
      "step": 34780
    },
    {
      "epoch": 18.2311320754717,
      "grad_norm": 0.7332766056060791,
      "learning_rate": 1.182656577791244e-06,
      "loss": 0.5999,
      "num_input_tokens_seen": 22747384,
      "step": 34785
    },
    {
      "epoch": 18.233752620545072,
      "grad_norm": 0.541248083114624,
      "learning_rate": 1.1791838243012359e-06,
      "loss": 0.4816,
      "num_input_tokens_seen": 22751224,
      "step": 34790
    },
    {
      "epoch": 18.23637316561845,
      "grad_norm": 0.6695720553398132,
      "learning_rate": 1.1757160538542117e-06,
      "loss": 0.5538,
      "num_input_tokens_seen": 22754392,
      "step": 34795
    },
    {
      "epoch": 18.238993710691823,
      "grad_norm": 0.5179870128631592,
      "learning_rate": 1.1722532671755882e-06,
      "loss": 0.3926,
      "num_input_tokens_seen": 22758776,
      "step": 34800
    },
    {
      "epoch": 18.2416142557652,
      "grad_norm": 0.4775524437427521,
      "learning_rate": 1.168795464989736e-06,
      "loss": 0.4605,
      "num_input_tokens_seen": 22762072,
      "step": 34805
    },
    {
      "epoch": 18.244234800838573,
      "grad_norm": 0.5790492296218872,
      "learning_rate": 1.165342648019993e-06,
      "loss": 0.4744,
      "num_input_tokens_seen": 22764856,
      "step": 34810
    },
    {
      "epoch": 18.24685534591195,
      "grad_norm": 0.38424864411354065,
      "learning_rate": 1.1618948169886457e-06,
      "loss": 0.4452,
      "num_input_tokens_seen": 22767736,
      "step": 34815
    },
    {
      "epoch": 18.249475890985323,
      "grad_norm": 0.46828126907348633,
      "learning_rate": 1.1584519726169446e-06,
      "loss": 0.3125,
      "num_input_tokens_seen": 22770840,
      "step": 34820
    },
    {
      "epoch": 18.2520964360587,
      "grad_norm": 0.6517965197563171,
      "learning_rate": 1.1550141156250942e-06,
      "loss": 0.4963,
      "num_input_tokens_seen": 22777208,
      "step": 34825
    },
    {
      "epoch": 18.254716981132077,
      "grad_norm": 0.5375543236732483,
      "learning_rate": 1.1515812467322556e-06,
      "loss": 0.5345,
      "num_input_tokens_seen": 22780312,
      "step": 34830
    },
    {
      "epoch": 18.25733752620545,
      "grad_norm": 0.509495198726654,
      "learning_rate": 1.1481533666565375e-06,
      "loss": 0.4858,
      "num_input_tokens_seen": 22784152,
      "step": 34835
    },
    {
      "epoch": 18.259958071278827,
      "grad_norm": 0.4626387357711792,
      "learning_rate": 1.144730476115019e-06,
      "loss": 0.54,
      "num_input_tokens_seen": 22788056,
      "step": 34840
    },
    {
      "epoch": 18.2625786163522,
      "grad_norm": 0.5037108063697815,
      "learning_rate": 1.1413125758237304e-06,
      "loss": 0.4118,
      "num_input_tokens_seen": 22791320,
      "step": 34845
    },
    {
      "epoch": 18.265199161425578,
      "grad_norm": 0.4152982234954834,
      "learning_rate": 1.1378996664976577e-06,
      "loss": 0.4434,
      "num_input_tokens_seen": 22794616,
      "step": 34850
    },
    {
      "epoch": 18.26781970649895,
      "grad_norm": 0.5411117672920227,
      "learning_rate": 1.134491748850741e-06,
      "loss": 0.4183,
      "num_input_tokens_seen": 22799256,
      "step": 34855
    },
    {
      "epoch": 18.270440251572328,
      "grad_norm": 0.35645467042922974,
      "learning_rate": 1.13108882359588e-06,
      "loss": 0.4241,
      "num_input_tokens_seen": 22802584,
      "step": 34860
    },
    {
      "epoch": 18.2730607966457,
      "grad_norm": 0.42631837725639343,
      "learning_rate": 1.127690891444924e-06,
      "loss": 0.3668,
      "num_input_tokens_seen": 22805272,
      "step": 34865
    },
    {
      "epoch": 18.27568134171908,
      "grad_norm": 0.3387836217880249,
      "learning_rate": 1.1242979531086829e-06,
      "loss": 0.4024,
      "num_input_tokens_seen": 22809016,
      "step": 34870
    },
    {
      "epoch": 18.278301886792452,
      "grad_norm": 0.4820742905139923,
      "learning_rate": 1.1209100092969244e-06,
      "loss": 0.5004,
      "num_input_tokens_seen": 22812248,
      "step": 34875
    },
    {
      "epoch": 18.28092243186583,
      "grad_norm": 1.4596540927886963,
      "learning_rate": 1.1175270607183625e-06,
      "loss": 0.4403,
      "num_input_tokens_seen": 22814648,
      "step": 34880
    },
    {
      "epoch": 18.283542976939202,
      "grad_norm": 0.5095796585083008,
      "learning_rate": 1.114149108080678e-06,
      "loss": 0.4848,
      "num_input_tokens_seen": 22816888,
      "step": 34885
    },
    {
      "epoch": 18.28616352201258,
      "grad_norm": 0.3646184504032135,
      "learning_rate": 1.1107761520904925e-06,
      "loss": 0.4274,
      "num_input_tokens_seen": 22821304,
      "step": 34890
    },
    {
      "epoch": 18.288784067085953,
      "grad_norm": 0.5320006012916565,
      "learning_rate": 1.1074081934533908e-06,
      "loss": 0.3437,
      "num_input_tokens_seen": 22824504,
      "step": 34895
    },
    {
      "epoch": 18.29140461215933,
      "grad_norm": 0.6832776665687561,
      "learning_rate": 1.1040452328739153e-06,
      "loss": 0.3429,
      "num_input_tokens_seen": 22828600,
      "step": 34900
    },
    {
      "epoch": 18.294025157232703,
      "grad_norm": 0.45616188645362854,
      "learning_rate": 1.1006872710555615e-06,
      "loss": 0.4323,
      "num_input_tokens_seen": 22834232,
      "step": 34905
    },
    {
      "epoch": 18.29664570230608,
      "grad_norm": 0.2836366891860962,
      "learning_rate": 1.0973343087007703e-06,
      "loss": 0.3948,
      "num_input_tokens_seen": 22838968,
      "step": 34910
    },
    {
      "epoch": 18.299266247379457,
      "grad_norm": 0.4404185116291046,
      "learning_rate": 1.0939863465109473e-06,
      "loss": 0.4321,
      "num_input_tokens_seen": 22841976,
      "step": 34915
    },
    {
      "epoch": 18.30188679245283,
      "grad_norm": 0.5668031573295593,
      "learning_rate": 1.0906433851864433e-06,
      "loss": 0.4378,
      "num_input_tokens_seen": 22845560,
      "step": 34920
    },
    {
      "epoch": 18.304507337526207,
      "grad_norm": 0.5406256318092346,
      "learning_rate": 1.0873054254265713e-06,
      "loss": 0.466,
      "num_input_tokens_seen": 22848696,
      "step": 34925
    },
    {
      "epoch": 18.30712788259958,
      "grad_norm": 0.3585834801197052,
      "learning_rate": 1.0839724679295976e-06,
      "loss": 0.3744,
      "num_input_tokens_seen": 22851320,
      "step": 34930
    },
    {
      "epoch": 18.309748427672957,
      "grad_norm": 0.4006875455379486,
      "learning_rate": 1.0806445133927311e-06,
      "loss": 0.386,
      "num_input_tokens_seen": 22855064,
      "step": 34935
    },
    {
      "epoch": 18.31236897274633,
      "grad_norm": 0.6948385834693909,
      "learning_rate": 1.0773215625121457e-06,
      "loss": 0.553,
      "num_input_tokens_seen": 22858392,
      "step": 34940
    },
    {
      "epoch": 18.314989517819708,
      "grad_norm": 0.32532799243927,
      "learning_rate": 1.0740036159829715e-06,
      "loss": 0.3783,
      "num_input_tokens_seen": 22861816,
      "step": 34945
    },
    {
      "epoch": 18.31761006289308,
      "grad_norm": 0.3922230303287506,
      "learning_rate": 1.0706906744992724e-06,
      "loss": 0.5423,
      "num_input_tokens_seen": 22864632,
      "step": 34950
    },
    {
      "epoch": 18.320230607966458,
      "grad_norm": 0.5354169011116028,
      "learning_rate": 1.0673827387540857e-06,
      "loss": 0.7288,
      "num_input_tokens_seen": 22867192,
      "step": 34955
    },
    {
      "epoch": 18.32285115303983,
      "grad_norm": 0.35277777910232544,
      "learning_rate": 1.0640798094393916e-06,
      "loss": 0.3995,
      "num_input_tokens_seen": 22869592,
      "step": 34960
    },
    {
      "epoch": 18.32547169811321,
      "grad_norm": 0.4044337272644043,
      "learning_rate": 1.060781887246126e-06,
      "loss": 0.4064,
      "num_input_tokens_seen": 22873560,
      "step": 34965
    },
    {
      "epoch": 18.328092243186582,
      "grad_norm": 0.9948843121528625,
      "learning_rate": 1.057488972864179e-06,
      "loss": 0.3102,
      "num_input_tokens_seen": 22877144,
      "step": 34970
    },
    {
      "epoch": 18.33071278825996,
      "grad_norm": 0.4862286150455475,
      "learning_rate": 1.0542010669823855e-06,
      "loss": 0.5032,
      "num_input_tokens_seen": 22880888,
      "step": 34975
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 0.3756372630596161,
      "learning_rate": 1.0509181702885395e-06,
      "loss": 0.3698,
      "num_input_tokens_seen": 22883896,
      "step": 34980
    },
    {
      "epoch": 18.33595387840671,
      "grad_norm": 0.4361593425273895,
      "learning_rate": 1.047640283469392e-06,
      "loss": 0.382,
      "num_input_tokens_seen": 22887384,
      "step": 34985
    },
    {
      "epoch": 18.338574423480082,
      "grad_norm": 0.9595755934715271,
      "learning_rate": 1.0443674072106308e-06,
      "loss": 0.4094,
      "num_input_tokens_seen": 22889880,
      "step": 34990
    },
    {
      "epoch": 18.34119496855346,
      "grad_norm": 0.4602247178554535,
      "learning_rate": 1.041099542196905e-06,
      "loss": 0.4089,
      "num_input_tokens_seen": 22893240,
      "step": 34995
    },
    {
      "epoch": 18.343815513626833,
      "grad_norm": 0.4622620642185211,
      "learning_rate": 1.0378366891118234e-06,
      "loss": 0.4484,
      "num_input_tokens_seen": 22897176,
      "step": 35000
    },
    {
      "epoch": 18.34643605870021,
      "grad_norm": 0.5999317765235901,
      "learning_rate": 1.0345788486379265e-06,
      "loss": 0.3326,
      "num_input_tokens_seen": 22899704,
      "step": 35005
    },
    {
      "epoch": 18.349056603773583,
      "grad_norm": 0.7608360648155212,
      "learning_rate": 1.0313260214567272e-06,
      "loss": 0.4282,
      "num_input_tokens_seen": 22902840,
      "step": 35010
    },
    {
      "epoch": 18.35167714884696,
      "grad_norm": 0.672353982925415,
      "learning_rate": 1.0280782082486729e-06,
      "loss": 0.5612,
      "num_input_tokens_seen": 22905912,
      "step": 35015
    },
    {
      "epoch": 18.354297693920337,
      "grad_norm": 0.38381537795066833,
      "learning_rate": 1.02483540969317e-06,
      "loss": 0.4887,
      "num_input_tokens_seen": 22909240,
      "step": 35020
    },
    {
      "epoch": 18.35691823899371,
      "grad_norm": 0.559782862663269,
      "learning_rate": 1.0215976264685763e-06,
      "loss": 0.4553,
      "num_input_tokens_seen": 22912216,
      "step": 35025
    },
    {
      "epoch": 18.359538784067087,
      "grad_norm": 0.5635309815406799,
      "learning_rate": 1.0183648592522021e-06,
      "loss": 0.4643,
      "num_input_tokens_seen": 22915064,
      "step": 35030
    },
    {
      "epoch": 18.36215932914046,
      "grad_norm": 0.3678596615791321,
      "learning_rate": 1.0151371087203015e-06,
      "loss": 0.5251,
      "num_input_tokens_seen": 22918296,
      "step": 35035
    },
    {
      "epoch": 18.364779874213838,
      "grad_norm": 0.569308876991272,
      "learning_rate": 1.0119143755480815e-06,
      "loss": 0.4691,
      "num_input_tokens_seen": 22921272,
      "step": 35040
    },
    {
      "epoch": 18.36740041928721,
      "grad_norm": 0.40902021527290344,
      "learning_rate": 1.0086966604097026e-06,
      "loss": 0.3215,
      "num_input_tokens_seen": 22927160,
      "step": 35045
    },
    {
      "epoch": 18.370020964360588,
      "grad_norm": 0.44596153497695923,
      "learning_rate": 1.0054839639782766e-06,
      "loss": 0.3529,
      "num_input_tokens_seen": 22929976,
      "step": 35050
    },
    {
      "epoch": 18.37264150943396,
      "grad_norm": 0.5522878766059875,
      "learning_rate": 1.00227628692586e-06,
      "loss": 0.477,
      "num_input_tokens_seen": 22932664,
      "step": 35055
    },
    {
      "epoch": 18.37526205450734,
      "grad_norm": 0.8611674904823303,
      "learning_rate": 9.990736299234692e-07,
      "loss": 0.4119,
      "num_input_tokens_seen": 22935576,
      "step": 35060
    },
    {
      "epoch": 18.377882599580712,
      "grad_norm": 0.4878283739089966,
      "learning_rate": 9.958759936410573e-07,
      "loss": 0.4034,
      "num_input_tokens_seen": 22939128,
      "step": 35065
    },
    {
      "epoch": 18.38050314465409,
      "grad_norm": 0.6600291728973389,
      "learning_rate": 9.926833787475304e-07,
      "loss": 0.4537,
      "num_input_tokens_seen": 22942616,
      "step": 35070
    },
    {
      "epoch": 18.383123689727462,
      "grad_norm": 0.665543258190155,
      "learning_rate": 9.89495785910749e-07,
      "loss": 0.4085,
      "num_input_tokens_seen": 22950168,
      "step": 35075
    },
    {
      "epoch": 18.38574423480084,
      "grad_norm": 0.5120550394058228,
      "learning_rate": 9.86313215797524e-07,
      "loss": 0.4833,
      "num_input_tokens_seen": 22954040,
      "step": 35080
    },
    {
      "epoch": 18.388364779874212,
      "grad_norm": 0.6721054911613464,
      "learning_rate": 9.831356690736149e-07,
      "loss": 0.4354,
      "num_input_tokens_seen": 22959256,
      "step": 35085
    },
    {
      "epoch": 18.39098532494759,
      "grad_norm": 0.3907923102378845,
      "learning_rate": 9.799631464037256e-07,
      "loss": 0.4357,
      "num_input_tokens_seen": 22962328,
      "step": 35090
    },
    {
      "epoch": 18.393605870020963,
      "grad_norm": 0.6300608515739441,
      "learning_rate": 9.76795648451509e-07,
      "loss": 0.478,
      "num_input_tokens_seen": 22967800,
      "step": 35095
    },
    {
      "epoch": 18.39622641509434,
      "grad_norm": 0.47168684005737305,
      "learning_rate": 9.73633175879568e-07,
      "loss": 0.3658,
      "num_input_tokens_seen": 22970936,
      "step": 35100
    },
    {
      "epoch": 18.398846960167713,
      "grad_norm": 0.45773378014564514,
      "learning_rate": 9.704757293494627e-07,
      "loss": 0.4341,
      "num_input_tokens_seen": 22973848,
      "step": 35105
    },
    {
      "epoch": 18.40146750524109,
      "grad_norm": 0.4892120063304901,
      "learning_rate": 9.673233095216922e-07,
      "loss": 0.4058,
      "num_input_tokens_seen": 22976888,
      "step": 35110
    },
    {
      "epoch": 18.404088050314467,
      "grad_norm": 1.5977340936660767,
      "learning_rate": 9.641759170557068e-07,
      "loss": 0.6907,
      "num_input_tokens_seen": 22979384,
      "step": 35115
    },
    {
      "epoch": 18.40670859538784,
      "grad_norm": 1.0646568536758423,
      "learning_rate": 9.610335526099024e-07,
      "loss": 0.4518,
      "num_input_tokens_seen": 22982904,
      "step": 35120
    },
    {
      "epoch": 18.409329140461217,
      "grad_norm": 0.4269910156726837,
      "learning_rate": 9.578962168416278e-07,
      "loss": 0.5259,
      "num_input_tokens_seen": 22986136,
      "step": 35125
    },
    {
      "epoch": 18.41194968553459,
      "grad_norm": 0.5382744669914246,
      "learning_rate": 9.547639104071775e-07,
      "loss": 0.2965,
      "num_input_tokens_seen": 22989752,
      "step": 35130
    },
    {
      "epoch": 18.414570230607968,
      "grad_norm": 0.7512073516845703,
      "learning_rate": 9.516366339617944e-07,
      "loss": 0.3537,
      "num_input_tokens_seen": 22993112,
      "step": 35135
    },
    {
      "epoch": 18.41719077568134,
      "grad_norm": 0.4630390405654907,
      "learning_rate": 9.485143881596714e-07,
      "loss": 0.3673,
      "num_input_tokens_seen": 22995672,
      "step": 35140
    },
    {
      "epoch": 18.419811320754718,
      "grad_norm": 0.8204203844070435,
      "learning_rate": 9.453971736539447e-07,
      "loss": 0.4822,
      "num_input_tokens_seen": 22998072,
      "step": 35145
    },
    {
      "epoch": 18.42243186582809,
      "grad_norm": 0.2583271265029907,
      "learning_rate": 9.42284991096698e-07,
      "loss": 0.5235,
      "num_input_tokens_seen": 23001752,
      "step": 35150
    },
    {
      "epoch": 18.42505241090147,
      "grad_norm": 0.7162860631942749,
      "learning_rate": 9.391778411389634e-07,
      "loss": 0.4202,
      "num_input_tokens_seen": 23005080,
      "step": 35155
    },
    {
      "epoch": 18.427672955974842,
      "grad_norm": 0.43197062611579895,
      "learning_rate": 9.360757244307267e-07,
      "loss": 0.4205,
      "num_input_tokens_seen": 23007768,
      "step": 35160
    },
    {
      "epoch": 18.43029350104822,
      "grad_norm": 0.7311506867408752,
      "learning_rate": 9.329786416209157e-07,
      "loss": 0.4724,
      "num_input_tokens_seen": 23010456,
      "step": 35165
    },
    {
      "epoch": 18.432914046121592,
      "grad_norm": 0.3618931770324707,
      "learning_rate": 9.298865933574014e-07,
      "loss": 0.3772,
      "num_input_tokens_seen": 23013368,
      "step": 35170
    },
    {
      "epoch": 18.43553459119497,
      "grad_norm": 0.3672195076942444,
      "learning_rate": 9.267995802870022e-07,
      "loss": 0.4495,
      "num_input_tokens_seen": 23015992,
      "step": 35175
    },
    {
      "epoch": 18.438155136268342,
      "grad_norm": 0.38479840755462646,
      "learning_rate": 9.237176030554878e-07,
      "loss": 0.4735,
      "num_input_tokens_seen": 23019544,
      "step": 35180
    },
    {
      "epoch": 18.44077568134172,
      "grad_norm": 0.3990580141544342,
      "learning_rate": 9.206406623075758e-07,
      "loss": 0.4811,
      "num_input_tokens_seen": 23022552,
      "step": 35185
    },
    {
      "epoch": 18.443396226415093,
      "grad_norm": 0.37979304790496826,
      "learning_rate": 9.175687586869292e-07,
      "loss": 0.6525,
      "num_input_tokens_seen": 23026008,
      "step": 35190
    },
    {
      "epoch": 18.44601677148847,
      "grad_norm": 0.6458427309989929,
      "learning_rate": 9.145018928361476e-07,
      "loss": 0.5406,
      "num_input_tokens_seen": 23029016,
      "step": 35195
    },
    {
      "epoch": 18.448637316561843,
      "grad_norm": 0.4636576175689697,
      "learning_rate": 9.114400653967903e-07,
      "loss": 0.5586,
      "num_input_tokens_seen": 23036792,
      "step": 35200
    },
    {
      "epoch": 18.45125786163522,
      "grad_norm": 0.8855462074279785,
      "learning_rate": 9.083832770093531e-07,
      "loss": 0.4772,
      "num_input_tokens_seen": 23039768,
      "step": 35205
    },
    {
      "epoch": 18.453878406708597,
      "grad_norm": 0.474043607711792,
      "learning_rate": 9.053315283132802e-07,
      "loss": 0.4467,
      "num_input_tokens_seen": 23043576,
      "step": 35210
    },
    {
      "epoch": 18.45649895178197,
      "grad_norm": 0.5630412697792053,
      "learning_rate": 9.022848199469663e-07,
      "loss": 0.4235,
      "num_input_tokens_seen": 23046456,
      "step": 35215
    },
    {
      "epoch": 18.459119496855347,
      "grad_norm": 0.4004395306110382,
      "learning_rate": 8.992431525477518e-07,
      "loss": 0.474,
      "num_input_tokens_seen": 23050328,
      "step": 35220
    },
    {
      "epoch": 18.46174004192872,
      "grad_norm": 0.5719766616821289,
      "learning_rate": 8.962065267519082e-07,
      "loss": 0.4075,
      "num_input_tokens_seen": 23052856,
      "step": 35225
    },
    {
      "epoch": 18.464360587002098,
      "grad_norm": 0.38723132014274597,
      "learning_rate": 8.931749431946718e-07,
      "loss": 0.3082,
      "num_input_tokens_seen": 23056440,
      "step": 35230
    },
    {
      "epoch": 18.46698113207547,
      "grad_norm": 0.3423706889152527,
      "learning_rate": 8.901484025102103e-07,
      "loss": 0.3601,
      "num_input_tokens_seen": 23059928,
      "step": 35235
    },
    {
      "epoch": 18.469601677148848,
      "grad_norm": 0.48641514778137207,
      "learning_rate": 8.871269053316422e-07,
      "loss": 0.5637,
      "num_input_tokens_seen": 23063128,
      "step": 35240
    },
    {
      "epoch": 18.47222222222222,
      "grad_norm": 0.46875423192977905,
      "learning_rate": 8.841104522910343e-07,
      "loss": 0.3605,
      "num_input_tokens_seen": 23065656,
      "step": 35245
    },
    {
      "epoch": 18.4748427672956,
      "grad_norm": 0.8450462818145752,
      "learning_rate": 8.8109904401939e-07,
      "loss": 0.4606,
      "num_input_tokens_seen": 23067992,
      "step": 35250
    },
    {
      "epoch": 18.47746331236897,
      "grad_norm": 0.8701163530349731,
      "learning_rate": 8.780926811466639e-07,
      "loss": 0.4619,
      "num_input_tokens_seen": 23070424,
      "step": 35255
    },
    {
      "epoch": 18.48008385744235,
      "grad_norm": 0.6251958012580872,
      "learning_rate": 8.7509136430175e-07,
      "loss": 0.6096,
      "num_input_tokens_seen": 23074520,
      "step": 35260
    },
    {
      "epoch": 18.482704402515722,
      "grad_norm": 0.4316754937171936,
      "learning_rate": 8.720950941124877e-07,
      "loss": 0.3919,
      "num_input_tokens_seen": 23077368,
      "step": 35265
    },
    {
      "epoch": 18.4853249475891,
      "grad_norm": 0.6627383828163147,
      "learning_rate": 8.691038712056732e-07,
      "loss": 0.5242,
      "num_input_tokens_seen": 23080024,
      "step": 35270
    },
    {
      "epoch": 18.487945492662472,
      "grad_norm": 0.9153573513031006,
      "learning_rate": 8.66117696207025e-07,
      "loss": 0.4125,
      "num_input_tokens_seen": 23082648,
      "step": 35275
    },
    {
      "epoch": 18.49056603773585,
      "grad_norm": 0.8699945211410522,
      "learning_rate": 8.631365697412186e-07,
      "loss": 0.6401,
      "num_input_tokens_seen": 23085720,
      "step": 35280
    },
    {
      "epoch": 18.493186582809223,
      "grad_norm": 0.5351300239562988,
      "learning_rate": 8.601604924318802e-07,
      "loss": 0.3554,
      "num_input_tokens_seen": 23088568,
      "step": 35285
    },
    {
      "epoch": 18.4958071278826,
      "grad_norm": 0.4155595600605011,
      "learning_rate": 8.571894649015588e-07,
      "loss": 0.352,
      "num_input_tokens_seen": 23092728,
      "step": 35290
    },
    {
      "epoch": 18.498427672955973,
      "grad_norm": 0.4809563457965851,
      "learning_rate": 8.542234877717686e-07,
      "loss": 0.3959,
      "num_input_tokens_seen": 23095864,
      "step": 35295
    },
    {
      "epoch": 18.50104821802935,
      "grad_norm": 0.30643120408058167,
      "learning_rate": 8.512625616629549e-07,
      "loss": 0.3851,
      "num_input_tokens_seen": 23099096,
      "step": 35300
    },
    {
      "epoch": 18.503668763102727,
      "grad_norm": 0.37819793820381165,
      "learning_rate": 8.483066871945083e-07,
      "loss": 0.3305,
      "num_input_tokens_seen": 23101944,
      "step": 35305
    },
    {
      "epoch": 18.5062893081761,
      "grad_norm": 0.37241414189338684,
      "learning_rate": 8.453558649847676e-07,
      "loss": 0.4476,
      "num_input_tokens_seen": 23105368,
      "step": 35310
    },
    {
      "epoch": 18.508909853249477,
      "grad_norm": 0.35381919145584106,
      "learning_rate": 8.424100956510112e-07,
      "loss": 0.4891,
      "num_input_tokens_seen": 23108568,
      "step": 35315
    },
    {
      "epoch": 18.51153039832285,
      "grad_norm": 0.5434849262237549,
      "learning_rate": 8.394693798094544e-07,
      "loss": 0.4605,
      "num_input_tokens_seen": 23112056,
      "step": 35320
    },
    {
      "epoch": 18.514150943396228,
      "grad_norm": 1.7047197818756104,
      "learning_rate": 8.365337180752691e-07,
      "loss": 0.4661,
      "num_input_tokens_seen": 23115864,
      "step": 35325
    },
    {
      "epoch": 18.5167714884696,
      "grad_norm": 0.4739479720592499,
      "learning_rate": 8.336031110625558e-07,
      "loss": 0.3774,
      "num_input_tokens_seen": 23118936,
      "step": 35330
    },
    {
      "epoch": 18.519392033542978,
      "grad_norm": 0.40261366963386536,
      "learning_rate": 8.306775593843657e-07,
      "loss": 0.3778,
      "num_input_tokens_seen": 23122680,
      "step": 35335
    },
    {
      "epoch": 18.52201257861635,
      "grad_norm": 0.4420808255672455,
      "learning_rate": 8.277570636526955e-07,
      "loss": 0.598,
      "num_input_tokens_seen": 23125400,
      "step": 35340
    },
    {
      "epoch": 18.52463312368973,
      "grad_norm": 0.7782948613166809,
      "learning_rate": 8.24841624478473e-07,
      "loss": 0.3705,
      "num_input_tokens_seen": 23128664,
      "step": 35345
    },
    {
      "epoch": 18.5272536687631,
      "grad_norm": 0.5344255566596985,
      "learning_rate": 8.219312424715825e-07,
      "loss": 0.3697,
      "num_input_tokens_seen": 23131032,
      "step": 35350
    },
    {
      "epoch": 18.52987421383648,
      "grad_norm": 0.37932634353637695,
      "learning_rate": 8.190259182408345e-07,
      "loss": 0.4588,
      "num_input_tokens_seen": 23133816,
      "step": 35355
    },
    {
      "epoch": 18.532494758909852,
      "grad_norm": 0.38982149958610535,
      "learning_rate": 8.161256523939925e-07,
      "loss": 0.4999,
      "num_input_tokens_seen": 23136600,
      "step": 35360
    },
    {
      "epoch": 18.53511530398323,
      "grad_norm": 0.43016085028648376,
      "learning_rate": 8.132304455377576e-07,
      "loss": 0.4808,
      "num_input_tokens_seen": 23139704,
      "step": 35365
    },
    {
      "epoch": 18.537735849056602,
      "grad_norm": 0.38813936710357666,
      "learning_rate": 8.103402982777813e-07,
      "loss": 0.3813,
      "num_input_tokens_seen": 23142712,
      "step": 35370
    },
    {
      "epoch": 18.54035639412998,
      "grad_norm": 0.32495182752609253,
      "learning_rate": 8.07455211218644e-07,
      "loss": 0.2726,
      "num_input_tokens_seen": 23146840,
      "step": 35375
    },
    {
      "epoch": 18.542976939203353,
      "grad_norm": 0.5328133702278137,
      "learning_rate": 8.045751849638738e-07,
      "loss": 0.4695,
      "num_input_tokens_seen": 23150488,
      "step": 35380
    },
    {
      "epoch": 18.54559748427673,
      "grad_norm": 0.6338895559310913,
      "learning_rate": 8.017002201159363e-07,
      "loss": 0.4609,
      "num_input_tokens_seen": 23153208,
      "step": 35385
    },
    {
      "epoch": 18.548218029350103,
      "grad_norm": 0.814505934715271,
      "learning_rate": 7.988303172762418e-07,
      "loss": 0.4537,
      "num_input_tokens_seen": 23155928,
      "step": 35390
    },
    {
      "epoch": 18.55083857442348,
      "grad_norm": 0.38660499453544617,
      "learning_rate": 7.959654770451463e-07,
      "loss": 0.5251,
      "num_input_tokens_seen": 23159352,
      "step": 35395
    },
    {
      "epoch": 18.553459119496857,
      "grad_norm": 0.4410436749458313,
      "learning_rate": 7.9310570002194e-07,
      "loss": 0.3378,
      "num_input_tokens_seen": 23163544,
      "step": 35400
    },
    {
      "epoch": 18.55607966457023,
      "grad_norm": 0.745786190032959,
      "learning_rate": 7.902509868048552e-07,
      "loss": 0.4148,
      "num_input_tokens_seen": 23166808,
      "step": 35405
    },
    {
      "epoch": 18.558700209643607,
      "grad_norm": 0.55075603723526,
      "learning_rate": 7.874013379910616e-07,
      "loss": 0.3832,
      "num_input_tokens_seen": 23170072,
      "step": 35410
    },
    {
      "epoch": 18.56132075471698,
      "grad_norm": 0.7847914099693298,
      "learning_rate": 7.845567541766769e-07,
      "loss": 0.345,
      "num_input_tokens_seen": 23172856,
      "step": 35415
    },
    {
      "epoch": 18.563941299790358,
      "grad_norm": 0.34781894087791443,
      "learning_rate": 7.817172359567526e-07,
      "loss": 0.5185,
      "num_input_tokens_seen": 23176024,
      "step": 35420
    },
    {
      "epoch": 18.56656184486373,
      "grad_norm": 0.6962189078330994,
      "learning_rate": 7.788827839252888e-07,
      "loss": 0.5326,
      "num_input_tokens_seen": 23179448,
      "step": 35425
    },
    {
      "epoch": 18.569182389937108,
      "grad_norm": 0.4204227328300476,
      "learning_rate": 7.760533986752222e-07,
      "loss": 0.4189,
      "num_input_tokens_seen": 23182808,
      "step": 35430
    },
    {
      "epoch": 18.57180293501048,
      "grad_norm": 0.5765255093574524,
      "learning_rate": 7.732290807984155e-07,
      "loss": 0.3725,
      "num_input_tokens_seen": 23186264,
      "step": 35435
    },
    {
      "epoch": 18.57442348008386,
      "grad_norm": 0.7153803110122681,
      "learning_rate": 7.704098308856933e-07,
      "loss": 0.4969,
      "num_input_tokens_seen": 23189176,
      "step": 35440
    },
    {
      "epoch": 18.57704402515723,
      "grad_norm": 0.40146005153656006,
      "learning_rate": 7.675956495268088e-07,
      "loss": 0.4861,
      "num_input_tokens_seen": 23192152,
      "step": 35445
    },
    {
      "epoch": 18.57966457023061,
      "grad_norm": 0.9796633720397949,
      "learning_rate": 7.647865373104551e-07,
      "loss": 0.4113,
      "num_input_tokens_seen": 23195448,
      "step": 35450
    },
    {
      "epoch": 18.582285115303982,
      "grad_norm": 0.42985185980796814,
      "learning_rate": 7.619824948242732e-07,
      "loss": 0.523,
      "num_input_tokens_seen": 23199352,
      "step": 35455
    },
    {
      "epoch": 18.58490566037736,
      "grad_norm": 0.78404301404953,
      "learning_rate": 7.5918352265483e-07,
      "loss": 0.3129,
      "num_input_tokens_seen": 23202072,
      "step": 35460
    },
    {
      "epoch": 18.587526205450732,
      "grad_norm": 0.7585442662239075,
      "learning_rate": 7.563896213876348e-07,
      "loss": 0.4308,
      "num_input_tokens_seen": 23204600,
      "step": 35465
    },
    {
      "epoch": 18.59014675052411,
      "grad_norm": 0.5448089838027954,
      "learning_rate": 7.53600791607148e-07,
      "loss": 0.6057,
      "num_input_tokens_seen": 23207416,
      "step": 35470
    },
    {
      "epoch": 18.592767295597483,
      "grad_norm": 0.6026710271835327,
      "learning_rate": 7.508170338967585e-07,
      "loss": 0.517,
      "num_input_tokens_seen": 23210616,
      "step": 35475
    },
    {
      "epoch": 18.59538784067086,
      "grad_norm": 0.5292016267776489,
      "learning_rate": 7.480383488387976e-07,
      "loss": 0.3952,
      "num_input_tokens_seen": 23213336,
      "step": 35480
    },
    {
      "epoch": 18.598008385744233,
      "grad_norm": 0.5232692956924438,
      "learning_rate": 7.452647370145338e-07,
      "loss": 0.5134,
      "num_input_tokens_seen": 23216856,
      "step": 35485
    },
    {
      "epoch": 18.60062893081761,
      "grad_norm": 0.5855368971824646,
      "learning_rate": 7.424961990041696e-07,
      "loss": 0.4466,
      "num_input_tokens_seen": 23219384,
      "step": 35490
    },
    {
      "epoch": 18.603249475890987,
      "grad_norm": 0.3214509189128876,
      "learning_rate": 7.397327353868583e-07,
      "loss": 0.4754,
      "num_input_tokens_seen": 23223064,
      "step": 35495
    },
    {
      "epoch": 18.60587002096436,
      "grad_norm": 0.3639785945415497,
      "learning_rate": 7.369743467406848e-07,
      "loss": 0.3378,
      "num_input_tokens_seen": 23226808,
      "step": 35500
    },
    {
      "epoch": 18.608490566037737,
      "grad_norm": 0.5765165686607361,
      "learning_rate": 7.34221033642668e-07,
      "loss": 0.524,
      "num_input_tokens_seen": 23229464,
      "step": 35505
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 0.8253651857376099,
      "learning_rate": 7.314727966687751e-07,
      "loss": 0.4436,
      "num_input_tokens_seen": 23232984,
      "step": 35510
    },
    {
      "epoch": 18.613731656184488,
      "grad_norm": 0.3954057991504669,
      "learning_rate": 7.287296363939045e-07,
      "loss": 0.4967,
      "num_input_tokens_seen": 23237048,
      "step": 35515
    },
    {
      "epoch": 18.61635220125786,
      "grad_norm": 0.5792689323425293,
      "learning_rate": 7.259915533918888e-07,
      "loss": 0.4405,
      "num_input_tokens_seen": 23239544,
      "step": 35520
    },
    {
      "epoch": 18.618972746331238,
      "grad_norm": 0.5882396697998047,
      "learning_rate": 7.232585482355086e-07,
      "loss": 0.4265,
      "num_input_tokens_seen": 23242936,
      "step": 35525
    },
    {
      "epoch": 18.62159329140461,
      "grad_norm": 1.0716352462768555,
      "learning_rate": 7.205306214964736e-07,
      "loss": 0.4332,
      "num_input_tokens_seen": 23246424,
      "step": 35530
    },
    {
      "epoch": 18.62421383647799,
      "grad_norm": 0.6377253532409668,
      "learning_rate": 7.178077737454436e-07,
      "loss": 0.4085,
      "num_input_tokens_seen": 23248984,
      "step": 35535
    },
    {
      "epoch": 18.62683438155136,
      "grad_norm": 0.7060366868972778,
      "learning_rate": 7.150900055519993e-07,
      "loss": 0.3578,
      "num_input_tokens_seen": 23251832,
      "step": 35540
    },
    {
      "epoch": 18.62945492662474,
      "grad_norm": 0.5309756994247437,
      "learning_rate": 7.123773174846693e-07,
      "loss": 0.4326,
      "num_input_tokens_seen": 23255032,
      "step": 35545
    },
    {
      "epoch": 18.632075471698112,
      "grad_norm": 0.26235994696617126,
      "learning_rate": 7.096697101109134e-07,
      "loss": 0.4588,
      "num_input_tokens_seen": 23257912,
      "step": 35550
    },
    {
      "epoch": 18.63469601677149,
      "grad_norm": 0.9701430201530457,
      "learning_rate": 7.069671839971371e-07,
      "loss": 0.5053,
      "num_input_tokens_seen": 23260504,
      "step": 35555
    },
    {
      "epoch": 18.637316561844862,
      "grad_norm": 0.5299291014671326,
      "learning_rate": 7.042697397086795e-07,
      "loss": 0.3658,
      "num_input_tokens_seen": 23262904,
      "step": 35560
    },
    {
      "epoch": 18.63993710691824,
      "grad_norm": 0.3282735347747803,
      "learning_rate": 7.015773778098089e-07,
      "loss": 0.3411,
      "num_input_tokens_seen": 23266584,
      "step": 35565
    },
    {
      "epoch": 18.642557651991613,
      "grad_norm": 0.4633568823337555,
      "learning_rate": 6.988900988637415e-07,
      "loss": 0.4054,
      "num_input_tokens_seen": 23269368,
      "step": 35570
    },
    {
      "epoch": 18.64517819706499,
      "grad_norm": 0.4164210855960846,
      "learning_rate": 6.962079034326218e-07,
      "loss": 0.2623,
      "num_input_tokens_seen": 23273208,
      "step": 35575
    },
    {
      "epoch": 18.647798742138363,
      "grad_norm": 0.6408993601799011,
      "learning_rate": 6.935307920775346e-07,
      "loss": 0.4766,
      "num_input_tokens_seen": 23276120,
      "step": 35580
    },
    {
      "epoch": 18.65041928721174,
      "grad_norm": 0.5102081894874573,
      "learning_rate": 6.90858765358507e-07,
      "loss": 0.4981,
      "num_input_tokens_seen": 23279192,
      "step": 35585
    },
    {
      "epoch": 18.653039832285117,
      "grad_norm": 0.38807418942451477,
      "learning_rate": 6.881918238344915e-07,
      "loss": 0.3842,
      "num_input_tokens_seen": 23282456,
      "step": 35590
    },
    {
      "epoch": 18.65566037735849,
      "grad_norm": 0.5565956830978394,
      "learning_rate": 6.855299680633808e-07,
      "loss": 0.4527,
      "num_input_tokens_seen": 23285528,
      "step": 35595
    },
    {
      "epoch": 18.658280922431867,
      "grad_norm": 0.3112674951553345,
      "learning_rate": 6.828731986020104e-07,
      "loss": 0.3552,
      "num_input_tokens_seen": 23289656,
      "step": 35600
    },
    {
      "epoch": 18.66090146750524,
      "grad_norm": 0.37924060225486755,
      "learning_rate": 6.802215160061381e-07,
      "loss": 0.403,
      "num_input_tokens_seen": 23292824,
      "step": 35605
    },
    {
      "epoch": 18.663522012578618,
      "grad_norm": 0.4322029948234558,
      "learning_rate": 6.775749208304704e-07,
      "loss": 0.4997,
      "num_input_tokens_seen": 23295832,
      "step": 35610
    },
    {
      "epoch": 18.66614255765199,
      "grad_norm": 0.32429423928260803,
      "learning_rate": 6.749334136286478e-07,
      "loss": 0.3575,
      "num_input_tokens_seen": 23299448,
      "step": 35615
    },
    {
      "epoch": 18.668763102725368,
      "grad_norm": 0.28253862261772156,
      "learning_rate": 6.722969949532365e-07,
      "loss": 0.375,
      "num_input_tokens_seen": 23301912,
      "step": 35620
    },
    {
      "epoch": 18.67138364779874,
      "grad_norm": 0.3673689663410187,
      "learning_rate": 6.69665665355751e-07,
      "loss": 0.3987,
      "num_input_tokens_seen": 23304664,
      "step": 35625
    },
    {
      "epoch": 18.67400419287212,
      "grad_norm": 0.39022284746170044,
      "learning_rate": 6.670394253866314e-07,
      "loss": 0.437,
      "num_input_tokens_seen": 23307288,
      "step": 35630
    },
    {
      "epoch": 18.67662473794549,
      "grad_norm": 0.8385114073753357,
      "learning_rate": 6.644182755952605e-07,
      "loss": 0.3904,
      "num_input_tokens_seen": 23310712,
      "step": 35635
    },
    {
      "epoch": 18.67924528301887,
      "grad_norm": 0.29283881187438965,
      "learning_rate": 6.618022165299526e-07,
      "loss": 0.4756,
      "num_input_tokens_seen": 23315448,
      "step": 35640
    },
    {
      "epoch": 18.681865828092242,
      "grad_norm": 0.3322240114212036,
      "learning_rate": 6.591912487379531e-07,
      "loss": 0.4347,
      "num_input_tokens_seen": 23318744,
      "step": 35645
    },
    {
      "epoch": 18.68448637316562,
      "grad_norm": 0.31423112750053406,
      "learning_rate": 6.565853727654503e-07,
      "loss": 0.5152,
      "num_input_tokens_seen": 23321784,
      "step": 35650
    },
    {
      "epoch": 18.687106918238992,
      "grad_norm": 0.3277439773082733,
      "learning_rate": 6.539845891575636e-07,
      "loss": 0.3836,
      "num_input_tokens_seen": 23325688,
      "step": 35655
    },
    {
      "epoch": 18.68972746331237,
      "grad_norm": 0.7321736812591553,
      "learning_rate": 6.513888984583466e-07,
      "loss": 0.5758,
      "num_input_tokens_seen": 23329528,
      "step": 35660
    },
    {
      "epoch": 18.692348008385743,
      "grad_norm": 0.7193148136138916,
      "learning_rate": 6.487983012107873e-07,
      "loss": 0.5243,
      "num_input_tokens_seen": 23331960,
      "step": 35665
    },
    {
      "epoch": 18.69496855345912,
      "grad_norm": 0.44581541419029236,
      "learning_rate": 6.462127979568133e-07,
      "loss": 0.3789,
      "num_input_tokens_seen": 23335288,
      "step": 35670
    },
    {
      "epoch": 18.697589098532493,
      "grad_norm": 0.36385592818260193,
      "learning_rate": 6.436323892372781e-07,
      "loss": 0.3206,
      "num_input_tokens_seen": 23337976,
      "step": 35675
    },
    {
      "epoch": 18.70020964360587,
      "grad_norm": 0.5788217782974243,
      "learning_rate": 6.410570755919721e-07,
      "loss": 0.3969,
      "num_input_tokens_seen": 23340824,
      "step": 35680
    },
    {
      "epoch": 18.702830188679247,
      "grad_norm": 1.0985411405563354,
      "learning_rate": 6.384868575596281e-07,
      "loss": 0.5515,
      "num_input_tokens_seen": 23343576,
      "step": 35685
    },
    {
      "epoch": 18.70545073375262,
      "grad_norm": 0.613633394241333,
      "learning_rate": 6.359217356779024e-07,
      "loss": 0.488,
      "num_input_tokens_seen": 23347192,
      "step": 35690
    },
    {
      "epoch": 18.708071278825997,
      "grad_norm": 0.49967852234840393,
      "learning_rate": 6.333617104833878e-07,
      "loss": 0.4751,
      "num_input_tokens_seen": 23351352,
      "step": 35695
    },
    {
      "epoch": 18.71069182389937,
      "grad_norm": 0.4680325984954834,
      "learning_rate": 6.308067825116142e-07,
      "loss": 0.4913,
      "num_input_tokens_seen": 23355064,
      "step": 35700
    },
    {
      "epoch": 18.713312368972748,
      "grad_norm": 0.42454150319099426,
      "learning_rate": 6.282569522970433e-07,
      "loss": 0.3829,
      "num_input_tokens_seen": 23359288,
      "step": 35705
    },
    {
      "epoch": 18.71593291404612,
      "grad_norm": 0.5157639384269714,
      "learning_rate": 6.257122203730731e-07,
      "loss": 0.4077,
      "num_input_tokens_seen": 23362296,
      "step": 35710
    },
    {
      "epoch": 18.718553459119498,
      "grad_norm": 0.6638316512107849,
      "learning_rate": 6.231725872720251e-07,
      "loss": 0.4137,
      "num_input_tokens_seen": 23364920,
      "step": 35715
    },
    {
      "epoch": 18.72117400419287,
      "grad_norm": 0.6861249208450317,
      "learning_rate": 6.206380535251716e-07,
      "loss": 0.4122,
      "num_input_tokens_seen": 23367608,
      "step": 35720
    },
    {
      "epoch": 18.72379454926625,
      "grad_norm": 0.4942310154438019,
      "learning_rate": 6.181086196626995e-07,
      "loss": 0.5313,
      "num_input_tokens_seen": 23370808,
      "step": 35725
    },
    {
      "epoch": 18.72641509433962,
      "grad_norm": 0.4709765613079071,
      "learning_rate": 6.155842862137384e-07,
      "loss": 0.4498,
      "num_input_tokens_seen": 23374104,
      "step": 35730
    },
    {
      "epoch": 18.729035639413,
      "grad_norm": 1.0098048448562622,
      "learning_rate": 6.130650537063548e-07,
      "loss": 0.4248,
      "num_input_tokens_seen": 23376856,
      "step": 35735
    },
    {
      "epoch": 18.731656184486372,
      "grad_norm": 0.7990301847457886,
      "learning_rate": 6.105509226675438e-07,
      "loss": 0.3928,
      "num_input_tokens_seen": 23379992,
      "step": 35740
    },
    {
      "epoch": 18.73427672955975,
      "grad_norm": 0.44023314118385315,
      "learning_rate": 6.08041893623229e-07,
      "loss": 0.4051,
      "num_input_tokens_seen": 23383832,
      "step": 35745
    },
    {
      "epoch": 18.736897274633122,
      "grad_norm": 1.0056864023208618,
      "learning_rate": 6.055379670982713e-07,
      "loss": 0.4503,
      "num_input_tokens_seen": 23386200,
      "step": 35750
    },
    {
      "epoch": 18.7395178197065,
      "grad_norm": 0.5153663754463196,
      "learning_rate": 6.030391436164628e-07,
      "loss": 0.4558,
      "num_input_tokens_seen": 23390648,
      "step": 35755
    },
    {
      "epoch": 18.742138364779873,
      "grad_norm": 1.017560601234436,
      "learning_rate": 6.005454237005298e-07,
      "loss": 0.4856,
      "num_input_tokens_seen": 23393080,
      "step": 35760
    },
    {
      "epoch": 18.74475890985325,
      "grad_norm": 0.7869441509246826,
      "learning_rate": 5.980568078721327e-07,
      "loss": 0.343,
      "num_input_tokens_seen": 23395960,
      "step": 35765
    },
    {
      "epoch": 18.747379454926623,
      "grad_norm": 0.6147339940071106,
      "learning_rate": 5.955732966518606e-07,
      "loss": 0.3243,
      "num_input_tokens_seen": 23398136,
      "step": 35770
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.5827088952064514,
      "learning_rate": 5.930948905592343e-07,
      "loss": 0.3706,
      "num_input_tokens_seen": 23400888,
      "step": 35775
    },
    {
      "epoch": 18.752620545073377,
      "grad_norm": 0.3834516406059265,
      "learning_rate": 5.906215901127054e-07,
      "loss": 0.4588,
      "num_input_tokens_seen": 23405368,
      "step": 35780
    },
    {
      "epoch": 18.75524109014675,
      "grad_norm": 0.5004130005836487,
      "learning_rate": 5.881533958296631e-07,
      "loss": 0.4165,
      "num_input_tokens_seen": 23408152,
      "step": 35785
    },
    {
      "epoch": 18.757861635220127,
      "grad_norm": 0.3160931468009949,
      "learning_rate": 5.856903082264248e-07,
      "loss": 0.4053,
      "num_input_tokens_seen": 23411800,
      "step": 35790
    },
    {
      "epoch": 18.7604821802935,
      "grad_norm": 0.5916451215744019,
      "learning_rate": 5.832323278182422e-07,
      "loss": 0.4454,
      "num_input_tokens_seen": 23414552,
      "step": 35795
    },
    {
      "epoch": 18.763102725366878,
      "grad_norm": 0.4669297933578491,
      "learning_rate": 5.807794551192986e-07,
      "loss": 0.4448,
      "num_input_tokens_seen": 23418008,
      "step": 35800
    },
    {
      "epoch": 18.76572327044025,
      "grad_norm": 0.43963924050331116,
      "learning_rate": 5.783316906427e-07,
      "loss": 0.5176,
      "num_input_tokens_seen": 23421304,
      "step": 35805
    },
    {
      "epoch": 18.768343815513628,
      "grad_norm": 0.36701709032058716,
      "learning_rate": 5.758890349004925e-07,
      "loss": 0.3415,
      "num_input_tokens_seen": 23425272,
      "step": 35810
    },
    {
      "epoch": 18.770964360587,
      "grad_norm": 0.5490483641624451,
      "learning_rate": 5.734514884036563e-07,
      "loss": 0.3739,
      "num_input_tokens_seen": 23428216,
      "step": 35815
    },
    {
      "epoch": 18.77358490566038,
      "grad_norm": 0.4191989302635193,
      "learning_rate": 5.710190516620945e-07,
      "loss": 0.4411,
      "num_input_tokens_seen": 23431064,
      "step": 35820
    },
    {
      "epoch": 18.77620545073375,
      "grad_norm": 0.4015599489212036,
      "learning_rate": 5.685917251846501e-07,
      "loss": 0.5245,
      "num_input_tokens_seen": 23434264,
      "step": 35825
    },
    {
      "epoch": 18.77882599580713,
      "grad_norm": 0.3756556808948517,
      "learning_rate": 5.661695094790892e-07,
      "loss": 0.4551,
      "num_input_tokens_seen": 23436280,
      "step": 35830
    },
    {
      "epoch": 18.781446540880502,
      "grad_norm": 0.3396424651145935,
      "learning_rate": 5.637524050521064e-07,
      "loss": 0.3311,
      "num_input_tokens_seen": 23440824,
      "step": 35835
    },
    {
      "epoch": 18.78406708595388,
      "grad_norm": 0.506359338760376,
      "learning_rate": 5.613404124093391e-07,
      "loss": 0.4592,
      "num_input_tokens_seen": 23443960,
      "step": 35840
    },
    {
      "epoch": 18.786687631027252,
      "grad_norm": 0.4841214120388031,
      "learning_rate": 5.589335320553446e-07,
      "loss": 0.5414,
      "num_input_tokens_seen": 23446680,
      "step": 35845
    },
    {
      "epoch": 18.78930817610063,
      "grad_norm": 0.42921146750450134,
      "learning_rate": 5.565317644936202e-07,
      "loss": 0.4995,
      "num_input_tokens_seen": 23450104,
      "step": 35850
    },
    {
      "epoch": 18.791928721174003,
      "grad_norm": 0.5769476890563965,
      "learning_rate": 5.541351102265863e-07,
      "loss": 0.5508,
      "num_input_tokens_seen": 23454200,
      "step": 35855
    },
    {
      "epoch": 18.79454926624738,
      "grad_norm": 0.43564102053642273,
      "learning_rate": 5.51743569755589e-07,
      "loss": 0.4945,
      "num_input_tokens_seen": 23457400,
      "step": 35860
    },
    {
      "epoch": 18.797169811320753,
      "grad_norm": 0.5484647154808044,
      "learning_rate": 5.493571435809197e-07,
      "loss": 0.3636,
      "num_input_tokens_seen": 23463160,
      "step": 35865
    },
    {
      "epoch": 18.79979035639413,
      "grad_norm": 0.33857834339141846,
      "learning_rate": 5.469758322017849e-07,
      "loss": 0.4339,
      "num_input_tokens_seen": 23465688,
      "step": 35870
    },
    {
      "epoch": 18.802410901467507,
      "grad_norm": 0.8529127836227417,
      "learning_rate": 5.445996361163358e-07,
      "loss": 0.2729,
      "num_input_tokens_seen": 23471608,
      "step": 35875
    },
    {
      "epoch": 18.80503144654088,
      "grad_norm": 0.45083075761795044,
      "learning_rate": 5.422285558216389e-07,
      "loss": 0.4574,
      "num_input_tokens_seen": 23474488,
      "step": 35880
    },
    {
      "epoch": 18.807651991614257,
      "grad_norm": 0.4161630868911743,
      "learning_rate": 5.398625918137001e-07,
      "loss": 0.4445,
      "num_input_tokens_seen": 23477272,
      "step": 35885
    },
    {
      "epoch": 18.81027253668763,
      "grad_norm": 0.43964648246765137,
      "learning_rate": 5.375017445874486e-07,
      "loss": 0.776,
      "num_input_tokens_seen": 23481112,
      "step": 35890
    },
    {
      "epoch": 18.812893081761008,
      "grad_norm": 0.31853538751602173,
      "learning_rate": 5.351460146367504e-07,
      "loss": 0.464,
      "num_input_tokens_seen": 23484632,
      "step": 35895
    },
    {
      "epoch": 18.81551362683438,
      "grad_norm": 0.6526290774345398,
      "learning_rate": 5.327954024543946e-07,
      "loss": 0.7011,
      "num_input_tokens_seen": 23487480,
      "step": 35900
    },
    {
      "epoch": 18.818134171907758,
      "grad_norm": 0.43204373121261597,
      "learning_rate": 5.304499085321074e-07,
      "loss": 0.392,
      "num_input_tokens_seen": 23490328,
      "step": 35905
    },
    {
      "epoch": 18.82075471698113,
      "grad_norm": 0.59471595287323,
      "learning_rate": 5.281095333605352e-07,
      "loss": 0.4549,
      "num_input_tokens_seen": 23492856,
      "step": 35910
    },
    {
      "epoch": 18.82337526205451,
      "grad_norm": 0.6826467514038086,
      "learning_rate": 5.257742774292556e-07,
      "loss": 0.549,
      "num_input_tokens_seen": 23496568,
      "step": 35915
    },
    {
      "epoch": 18.82599580712788,
      "grad_norm": 0.7087773084640503,
      "learning_rate": 5.234441412267837e-07,
      "loss": 0.4777,
      "num_input_tokens_seen": 23499288,
      "step": 35920
    },
    {
      "epoch": 18.82861635220126,
      "grad_norm": 0.5022704005241394,
      "learning_rate": 5.211191252405517e-07,
      "loss": 0.4919,
      "num_input_tokens_seen": 23501784,
      "step": 35925
    },
    {
      "epoch": 18.831236897274632,
      "grad_norm": 0.5419532656669617,
      "learning_rate": 5.187992299569316e-07,
      "loss": 0.5123,
      "num_input_tokens_seen": 23505720,
      "step": 35930
    },
    {
      "epoch": 18.83385744234801,
      "grad_norm": 0.6797417998313904,
      "learning_rate": 5.164844558612131e-07,
      "loss": 0.6226,
      "num_input_tokens_seen": 23508696,
      "step": 35935
    },
    {
      "epoch": 18.836477987421382,
      "grad_norm": 0.30751171708106995,
      "learning_rate": 5.141748034376254e-07,
      "loss": 0.3008,
      "num_input_tokens_seen": 23513112,
      "step": 35940
    },
    {
      "epoch": 18.83909853249476,
      "grad_norm": 0.4173077642917633,
      "learning_rate": 5.118702731693209e-07,
      "loss": 0.4582,
      "num_input_tokens_seen": 23515864,
      "step": 35945
    },
    {
      "epoch": 18.841719077568133,
      "grad_norm": 0.5445047616958618,
      "learning_rate": 5.095708655383751e-07,
      "loss": 0.5766,
      "num_input_tokens_seen": 23518424,
      "step": 35950
    },
    {
      "epoch": 18.84433962264151,
      "grad_norm": 0.7334421873092651,
      "learning_rate": 5.07276581025809e-07,
      "loss": 0.4063,
      "num_input_tokens_seen": 23521496,
      "step": 35955
    },
    {
      "epoch": 18.846960167714883,
      "grad_norm": 0.4274800419807434,
      "learning_rate": 5.049874201115495e-07,
      "loss": 0.4381,
      "num_input_tokens_seen": 23524984,
      "step": 35960
    },
    {
      "epoch": 18.84958071278826,
      "grad_norm": 0.41143181920051575,
      "learning_rate": 5.027033832744693e-07,
      "loss": 0.4182,
      "num_input_tokens_seen": 23527640,
      "step": 35965
    },
    {
      "epoch": 18.852201257861637,
      "grad_norm": 0.3943946361541748,
      "learning_rate": 5.004244709923639e-07,
      "loss": 0.3766,
      "num_input_tokens_seen": 23530424,
      "step": 35970
    },
    {
      "epoch": 18.85482180293501,
      "grad_norm": 0.7769948840141296,
      "learning_rate": 4.981506837419519e-07,
      "loss": 0.3882,
      "num_input_tokens_seen": 23533688,
      "step": 35975
    },
    {
      "epoch": 18.857442348008387,
      "grad_norm": 0.5165196061134338,
      "learning_rate": 4.958820219988836e-07,
      "loss": 0.3597,
      "num_input_tokens_seen": 23536760,
      "step": 35980
    },
    {
      "epoch": 18.86006289308176,
      "grad_norm": 0.3865417242050171,
      "learning_rate": 4.936184862377402e-07,
      "loss": 0.4414,
      "num_input_tokens_seen": 23539960,
      "step": 35985
    },
    {
      "epoch": 18.862683438155138,
      "grad_norm": 0.344372421503067,
      "learning_rate": 4.913600769320237e-07,
      "loss": 0.3906,
      "num_input_tokens_seen": 23543288,
      "step": 35990
    },
    {
      "epoch": 18.86530398322851,
      "grad_norm": 0.5867920517921448,
      "learning_rate": 4.891067945541728e-07,
      "loss": 0.5301,
      "num_input_tokens_seen": 23546936,
      "step": 35995
    },
    {
      "epoch": 18.867924528301888,
      "grad_norm": 0.3473816215991974,
      "learning_rate": 4.868586395755437e-07,
      "loss": 0.4311,
      "num_input_tokens_seen": 23550072,
      "step": 36000
    },
    {
      "epoch": 18.87054507337526,
      "grad_norm": 0.3977581560611725,
      "learning_rate": 4.846156124664242e-07,
      "loss": 0.3688,
      "num_input_tokens_seen": 23553400,
      "step": 36005
    },
    {
      "epoch": 18.87316561844864,
      "grad_norm": 0.37580883502960205,
      "learning_rate": 4.823777136960362e-07,
      "loss": 0.389,
      "num_input_tokens_seen": 23556856,
      "step": 36010
    },
    {
      "epoch": 18.87578616352201,
      "grad_norm": 0.6213849186897278,
      "learning_rate": 4.801449437325162e-07,
      "loss": 0.5114,
      "num_input_tokens_seen": 23560120,
      "step": 36015
    },
    {
      "epoch": 18.87840670859539,
      "grad_norm": 0.5470777750015259,
      "learning_rate": 4.779173030429379e-07,
      "loss": 0.4186,
      "num_input_tokens_seen": 23563256,
      "step": 36020
    },
    {
      "epoch": 18.881027253668762,
      "grad_norm": 0.514712393283844,
      "learning_rate": 4.7569479209329804e-07,
      "loss": 0.4249,
      "num_input_tokens_seen": 23567448,
      "step": 36025
    },
    {
      "epoch": 18.88364779874214,
      "grad_norm": 0.3950708210468292,
      "learning_rate": 4.734774113485163e-07,
      "loss": 0.4589,
      "num_input_tokens_seen": 23570488,
      "step": 36030
    },
    {
      "epoch": 18.886268343815512,
      "grad_norm": 0.4104955792427063,
      "learning_rate": 4.7126516127244945e-07,
      "loss": 0.4898,
      "num_input_tokens_seen": 23573624,
      "step": 36035
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 0.4159388542175293,
      "learning_rate": 4.690580423278718e-07,
      "loss": 0.3007,
      "num_input_tokens_seen": 23576312,
      "step": 36040
    },
    {
      "epoch": 18.891509433962263,
      "grad_norm": 0.7879481911659241,
      "learning_rate": 4.668560549764861e-07,
      "loss": 0.4713,
      "num_input_tokens_seen": 23579320,
      "step": 36045
    },
    {
      "epoch": 18.89412997903564,
      "grad_norm": 0.3927670419216156,
      "learning_rate": 4.646591996789268e-07,
      "loss": 0.3769,
      "num_input_tokens_seen": 23582776,
      "step": 36050
    },
    {
      "epoch": 18.896750524109013,
      "grad_norm": 0.29978692531585693,
      "learning_rate": 4.6246747689474847e-07,
      "loss": 0.4132,
      "num_input_tokens_seen": 23585208,
      "step": 36055
    },
    {
      "epoch": 18.89937106918239,
      "grad_norm": 0.6378452181816101,
      "learning_rate": 4.6028088708243434e-07,
      "loss": 0.474,
      "num_input_tokens_seen": 23588472,
      "step": 36060
    },
    {
      "epoch": 18.901991614255767,
      "grad_norm": 0.5111680030822754,
      "learning_rate": 4.580994306993991e-07,
      "loss": 0.4144,
      "num_input_tokens_seen": 23591768,
      "step": 36065
    },
    {
      "epoch": 18.90461215932914,
      "grad_norm": 0.5051275491714478,
      "learning_rate": 4.5592310820196936e-07,
      "loss": 0.4026,
      "num_input_tokens_seen": 23594488,
      "step": 36070
    },
    {
      "epoch": 18.907232704402517,
      "grad_norm": 0.29368358850479126,
      "learning_rate": 4.537519200454116e-07,
      "loss": 0.3348,
      "num_input_tokens_seen": 23597720,
      "step": 36075
    },
    {
      "epoch": 18.90985324947589,
      "grad_norm": 0.3356064558029175,
      "learning_rate": 4.515858666839151e-07,
      "loss": 0.3041,
      "num_input_tokens_seen": 23600696,
      "step": 36080
    },
    {
      "epoch": 18.912473794549268,
      "grad_norm": 0.34178343415260315,
      "learning_rate": 4.494249485705926e-07,
      "loss": 0.4434,
      "num_input_tokens_seen": 23603768,
      "step": 36085
    },
    {
      "epoch": 18.91509433962264,
      "grad_norm": 0.46628519892692566,
      "learning_rate": 4.472691661574824e-07,
      "loss": 0.492,
      "num_input_tokens_seen": 23606872,
      "step": 36090
    },
    {
      "epoch": 18.917714884696018,
      "grad_norm": 0.36717066168785095,
      "learning_rate": 4.4511851989555156e-07,
      "loss": 0.5185,
      "num_input_tokens_seen": 23610712,
      "step": 36095
    },
    {
      "epoch": 18.92033542976939,
      "grad_norm": 0.6707721948623657,
      "learning_rate": 4.429730102346874e-07,
      "loss": 0.5344,
      "num_input_tokens_seen": 23613784,
      "step": 36100
    },
    {
      "epoch": 18.92295597484277,
      "grad_norm": 0.8155465126037598,
      "learning_rate": 4.4083263762370873e-07,
      "loss": 0.4353,
      "num_input_tokens_seen": 23616888,
      "step": 36105
    },
    {
      "epoch": 18.92557651991614,
      "grad_norm": 0.41865724325180054,
      "learning_rate": 4.386974025103574e-07,
      "loss": 0.4121,
      "num_input_tokens_seen": 23619864,
      "step": 36110
    },
    {
      "epoch": 18.92819706498952,
      "grad_norm": 0.6174930930137634,
      "learning_rate": 4.365673053413011e-07,
      "loss": 0.4038,
      "num_input_tokens_seen": 23623864,
      "step": 36115
    },
    {
      "epoch": 18.930817610062892,
      "grad_norm": 0.6002963185310364,
      "learning_rate": 4.34442346562125e-07,
      "loss": 0.5066,
      "num_input_tokens_seen": 23627128,
      "step": 36120
    },
    {
      "epoch": 18.93343815513627,
      "grad_norm": 0.3399653136730194,
      "learning_rate": 4.323225266173542e-07,
      "loss": 0.4204,
      "num_input_tokens_seen": 23630840,
      "step": 36125
    },
    {
      "epoch": 18.936058700209642,
      "grad_norm": 0.7642581462860107,
      "learning_rate": 4.302078459504283e-07,
      "loss": 0.4813,
      "num_input_tokens_seen": 23634264,
      "step": 36130
    },
    {
      "epoch": 18.93867924528302,
      "grad_norm": 0.4060189127922058,
      "learning_rate": 4.2809830500371294e-07,
      "loss": 0.4354,
      "num_input_tokens_seen": 23637176,
      "step": 36135
    },
    {
      "epoch": 18.941299790356393,
      "grad_norm": 0.29716673493385315,
      "learning_rate": 4.2599390421850236e-07,
      "loss": 0.4263,
      "num_input_tokens_seen": 23640088,
      "step": 36140
    },
    {
      "epoch": 18.94392033542977,
      "grad_norm": 0.3116222620010376,
      "learning_rate": 4.2389464403501113e-07,
      "loss": 0.4466,
      "num_input_tokens_seen": 23643576,
      "step": 36145
    },
    {
      "epoch": 18.946540880503143,
      "grad_norm": 0.4463028311729431,
      "learning_rate": 4.218005248923823e-07,
      "loss": 0.5104,
      "num_input_tokens_seen": 23646424,
      "step": 36150
    },
    {
      "epoch": 18.94916142557652,
      "grad_norm": 0.29500260949134827,
      "learning_rate": 4.1971154722867946e-07,
      "loss": 0.4142,
      "num_input_tokens_seen": 23649688,
      "step": 36155
    },
    {
      "epoch": 18.951781970649897,
      "grad_norm": 1.5370360612869263,
      "learning_rate": 4.176277114808946e-07,
      "loss": 0.4884,
      "num_input_tokens_seen": 23652664,
      "step": 36160
    },
    {
      "epoch": 18.95440251572327,
      "grad_norm": 0.7816213369369507,
      "learning_rate": 4.155490180849431e-07,
      "loss": 0.6292,
      "num_input_tokens_seen": 23656088,
      "step": 36165
    },
    {
      "epoch": 18.957023060796647,
      "grad_norm": 0.6281497478485107,
      "learning_rate": 4.1347546747566304e-07,
      "loss": 0.4456,
      "num_input_tokens_seen": 23659544,
      "step": 36170
    },
    {
      "epoch": 18.95964360587002,
      "grad_norm": 0.3515547215938568,
      "learning_rate": 4.114070600868131e-07,
      "loss": 0.4261,
      "num_input_tokens_seen": 23663352,
      "step": 36175
    },
    {
      "epoch": 18.962264150943398,
      "grad_norm": 0.448647141456604,
      "learning_rate": 4.093437963510888e-07,
      "loss": 0.321,
      "num_input_tokens_seen": 23666232,
      "step": 36180
    },
    {
      "epoch": 18.96488469601677,
      "grad_norm": 0.6169120073318481,
      "learning_rate": 4.072856767000949e-07,
      "loss": 0.5036,
      "num_input_tokens_seen": 23669688,
      "step": 36185
    },
    {
      "epoch": 18.967505241090148,
      "grad_norm": 0.37826287746429443,
      "learning_rate": 4.052327015643703e-07,
      "loss": 0.5337,
      "num_input_tokens_seen": 23672696,
      "step": 36190
    },
    {
      "epoch": 18.97012578616352,
      "grad_norm": 0.5569135546684265,
      "learning_rate": 4.031848713733744e-07,
      "loss": 0.4205,
      "num_input_tokens_seen": 23676920,
      "step": 36195
    },
    {
      "epoch": 18.9727463312369,
      "grad_norm": 0.41329094767570496,
      "learning_rate": 4.011421865554837e-07,
      "loss": 0.4558,
      "num_input_tokens_seen": 23681336,
      "step": 36200
    },
    {
      "epoch": 18.97536687631027,
      "grad_norm": 0.34305721521377563,
      "learning_rate": 3.9910464753801224e-07,
      "loss": 0.5513,
      "num_input_tokens_seen": 23685336,
      "step": 36205
    },
    {
      "epoch": 18.97798742138365,
      "grad_norm": 0.46329793334007263,
      "learning_rate": 3.9707225474718557e-07,
      "loss": 0.35,
      "num_input_tokens_seen": 23687960,
      "step": 36210
    },
    {
      "epoch": 18.980607966457022,
      "grad_norm": 0.49428677558898926,
      "learning_rate": 3.950450086081553e-07,
      "loss": 0.3828,
      "num_input_tokens_seen": 23692184,
      "step": 36215
    },
    {
      "epoch": 18.9832285115304,
      "grad_norm": 0.40793970227241516,
      "learning_rate": 3.930229095450072e-07,
      "loss": 0.4195,
      "num_input_tokens_seen": 23695800,
      "step": 36220
    },
    {
      "epoch": 18.985849056603772,
      "grad_norm": 0.408278226852417,
      "learning_rate": 3.9100595798073345e-07,
      "loss": 0.5024,
      "num_input_tokens_seen": 23698872,
      "step": 36225
    },
    {
      "epoch": 18.98846960167715,
      "grad_norm": 0.300799697637558,
      "learning_rate": 3.8899415433725493e-07,
      "loss": 0.3977,
      "num_input_tokens_seen": 23702456,
      "step": 36230
    },
    {
      "epoch": 18.991090146750523,
      "grad_norm": 0.5056145787239075,
      "learning_rate": 3.8698749903542666e-07,
      "loss": 0.4993,
      "num_input_tokens_seen": 23705720,
      "step": 36235
    },
    {
      "epoch": 18.9937106918239,
      "grad_norm": 0.4835154116153717,
      "learning_rate": 3.8498599249501e-07,
      "loss": 0.5042,
      "num_input_tokens_seen": 23708824,
      "step": 36240
    },
    {
      "epoch": 18.996331236897273,
      "grad_norm": 0.7058393359184265,
      "learning_rate": 3.8298963513470345e-07,
      "loss": 0.4172,
      "num_input_tokens_seen": 23711416,
      "step": 36245
    },
    {
      "epoch": 18.99895178197065,
      "grad_norm": 0.6183128952980042,
      "learning_rate": 3.8099842737211734e-07,
      "loss": 0.3747,
      "num_input_tokens_seen": 23714392,
      "step": 36250
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.4768696129322052,
      "eval_runtime": 9.7243,
      "eval_samples_per_second": 87.204,
      "eval_steps_per_second": 21.801,
      "num_input_tokens_seen": 23715264,
      "step": 36252
    },
    {
      "epoch": 19.001572327044027,
      "grad_norm": 0.3967573940753937,
      "learning_rate": 3.7901236962379627e-07,
      "loss": 0.4316,
      "num_input_tokens_seen": 23716768,
      "step": 36255
    },
    {
      "epoch": 19.0041928721174,
      "grad_norm": 0.28513073921203613,
      "learning_rate": 3.7703146230519126e-07,
      "loss": 0.4629,
      "num_input_tokens_seen": 23720384,
      "step": 36260
    },
    {
      "epoch": 19.006813417190777,
      "grad_norm": 1.086474061012268,
      "learning_rate": 3.7505570583069306e-07,
      "loss": 0.5606,
      "num_input_tokens_seen": 23725024,
      "step": 36265
    },
    {
      "epoch": 19.00943396226415,
      "grad_norm": 0.34602034091949463,
      "learning_rate": 3.7308510061360435e-07,
      "loss": 0.3078,
      "num_input_tokens_seen": 23728640,
      "step": 36270
    },
    {
      "epoch": 19.012054507337528,
      "grad_norm": 0.5601246356964111,
      "learning_rate": 3.7111964706615374e-07,
      "loss": 0.7057,
      "num_input_tokens_seen": 23731392,
      "step": 36275
    },
    {
      "epoch": 19.0146750524109,
      "grad_norm": 0.4574359357357025,
      "learning_rate": 3.691593455994929e-07,
      "loss": 0.5054,
      "num_input_tokens_seen": 23734912,
      "step": 36280
    },
    {
      "epoch": 19.017295597484278,
      "grad_norm": 0.48722219467163086,
      "learning_rate": 3.672041966236911e-07,
      "loss": 0.4447,
      "num_input_tokens_seen": 23737856,
      "step": 36285
    },
    {
      "epoch": 19.01991614255765,
      "grad_norm": 0.5798856019973755,
      "learning_rate": 3.6525420054774884e-07,
      "loss": 0.4631,
      "num_input_tokens_seen": 23741408,
      "step": 36290
    },
    {
      "epoch": 19.02253668763103,
      "grad_norm": 0.33270758390426636,
      "learning_rate": 3.6330935777957597e-07,
      "loss": 0.4966,
      "num_input_tokens_seen": 23744448,
      "step": 36295
    },
    {
      "epoch": 19.0251572327044,
      "grad_norm": 0.489613801240921,
      "learning_rate": 3.613696687260193e-07,
      "loss": 0.259,
      "num_input_tokens_seen": 23748128,
      "step": 36300
    },
    {
      "epoch": 19.02777777777778,
      "grad_norm": 0.39878225326538086,
      "learning_rate": 3.59435133792832e-07,
      "loss": 0.4289,
      "num_input_tokens_seen": 23750880,
      "step": 36305
    },
    {
      "epoch": 19.030398322851152,
      "grad_norm": 0.4975855350494385,
      "learning_rate": 3.575057533847043e-07,
      "loss": 0.3776,
      "num_input_tokens_seen": 23754112,
      "step": 36310
    },
    {
      "epoch": 19.03301886792453,
      "grad_norm": 0.6051071882247925,
      "learning_rate": 3.555815279052327e-07,
      "loss": 0.4467,
      "num_input_tokens_seen": 23757792,
      "step": 36315
    },
    {
      "epoch": 19.035639412997902,
      "grad_norm": 0.3563620448112488,
      "learning_rate": 3.5366245775694805e-07,
      "loss": 0.4926,
      "num_input_tokens_seen": 23762336,
      "step": 36320
    },
    {
      "epoch": 19.03825995807128,
      "grad_norm": 0.5762856602668762,
      "learning_rate": 3.517485433412987e-07,
      "loss": 0.3254,
      "num_input_tokens_seen": 23764800,
      "step": 36325
    },
    {
      "epoch": 19.040880503144653,
      "grad_norm": 0.42036768794059753,
      "learning_rate": 3.4983978505865044e-07,
      "loss": 0.4941,
      "num_input_tokens_seen": 23768288,
      "step": 36330
    },
    {
      "epoch": 19.04350104821803,
      "grad_norm": 0.3521643579006195,
      "learning_rate": 3.47936183308295e-07,
      "loss": 0.3915,
      "num_input_tokens_seen": 23771456,
      "step": 36335
    },
    {
      "epoch": 19.046121593291403,
      "grad_norm": 0.582855761051178,
      "learning_rate": 3.460377384884472e-07,
      "loss": 0.4568,
      "num_input_tokens_seen": 23774144,
      "step": 36340
    },
    {
      "epoch": 19.04874213836478,
      "grad_norm": 0.3319235146045685,
      "learning_rate": 3.4414445099623384e-07,
      "loss": 0.398,
      "num_input_tokens_seen": 23777280,
      "step": 36345
    },
    {
      "epoch": 19.051362683438157,
      "grad_norm": 0.2549586892127991,
      "learning_rate": 3.42256321227713e-07,
      "loss": 0.4234,
      "num_input_tokens_seen": 23780672,
      "step": 36350
    },
    {
      "epoch": 19.05398322851153,
      "grad_norm": 0.5317942500114441,
      "learning_rate": 3.403733495778605e-07,
      "loss": 0.5268,
      "num_input_tokens_seen": 23784224,
      "step": 36355
    },
    {
      "epoch": 19.056603773584907,
      "grad_norm": 0.27256301045417786,
      "learning_rate": 3.3849553644056955e-07,
      "loss": 0.3707,
      "num_input_tokens_seen": 23787904,
      "step": 36360
    },
    {
      "epoch": 19.05922431865828,
      "grad_norm": 0.4456504285335541,
      "learning_rate": 3.36622882208662e-07,
      "loss": 0.4894,
      "num_input_tokens_seen": 23790784,
      "step": 36365
    },
    {
      "epoch": 19.061844863731658,
      "grad_norm": 0.5517325401306152,
      "learning_rate": 3.347553872738718e-07,
      "loss": 0.7455,
      "num_input_tokens_seen": 23794048,
      "step": 36370
    },
    {
      "epoch": 19.06446540880503,
      "grad_norm": 0.6259260177612305,
      "learning_rate": 3.3289305202685583e-07,
      "loss": 0.475,
      "num_input_tokens_seen": 23797792,
      "step": 36375
    },
    {
      "epoch": 19.067085953878408,
      "grad_norm": 0.9078395962715149,
      "learning_rate": 3.310358768571997e-07,
      "loss": 0.4381,
      "num_input_tokens_seen": 23800768,
      "step": 36380
    },
    {
      "epoch": 19.06970649895178,
      "grad_norm": 0.4353553354740143,
      "learning_rate": 3.2918386215339827e-07,
      "loss": 0.6044,
      "num_input_tokens_seen": 23803968,
      "step": 36385
    },
    {
      "epoch": 19.072327044025158,
      "grad_norm": 0.4795193672180176,
      "learning_rate": 3.2733700830287205e-07,
      "loss": 0.4925,
      "num_input_tokens_seen": 23807136,
      "step": 36390
    },
    {
      "epoch": 19.07494758909853,
      "grad_norm": 0.37434542179107666,
      "learning_rate": 3.254953156919649e-07,
      "loss": 0.4779,
      "num_input_tokens_seen": 23811520,
      "step": 36395
    },
    {
      "epoch": 19.07756813417191,
      "grad_norm": 0.5543133616447449,
      "learning_rate": 3.2365878470593527e-07,
      "loss": 0.4026,
      "num_input_tokens_seen": 23814624,
      "step": 36400
    },
    {
      "epoch": 19.080188679245282,
      "grad_norm": 0.36523309350013733,
      "learning_rate": 3.2182741572896747e-07,
      "loss": 0.4567,
      "num_input_tokens_seen": 23817856,
      "step": 36405
    },
    {
      "epoch": 19.08280922431866,
      "grad_norm": 0.4273943603038788,
      "learning_rate": 3.2000120914415786e-07,
      "loss": 0.4509,
      "num_input_tokens_seen": 23821472,
      "step": 36410
    },
    {
      "epoch": 19.085429769392032,
      "grad_norm": 0.40965232253074646,
      "learning_rate": 3.181801653335315e-07,
      "loss": 0.4643,
      "num_input_tokens_seen": 23824512,
      "step": 36415
    },
    {
      "epoch": 19.08805031446541,
      "grad_norm": 1.1639158725738525,
      "learning_rate": 3.163642846780307e-07,
      "loss": 0.3752,
      "num_input_tokens_seen": 23827776,
      "step": 36420
    },
    {
      "epoch": 19.090670859538783,
      "grad_norm": 0.6615884304046631,
      "learning_rate": 3.1455356755751565e-07,
      "loss": 0.4701,
      "num_input_tokens_seen": 23830688,
      "step": 36425
    },
    {
      "epoch": 19.09329140461216,
      "grad_norm": 0.49363258481025696,
      "learning_rate": 3.127480143507666e-07,
      "loss": 0.3674,
      "num_input_tokens_seen": 23833536,
      "step": 36430
    },
    {
      "epoch": 19.095911949685533,
      "grad_norm": 0.49551254510879517,
      "learning_rate": 3.1094762543548704e-07,
      "loss": 0.3868,
      "num_input_tokens_seen": 23838272,
      "step": 36435
    },
    {
      "epoch": 19.09853249475891,
      "grad_norm": 0.6893376111984253,
      "learning_rate": 3.091524011882951e-07,
      "loss": 0.5286,
      "num_input_tokens_seen": 23841600,
      "step": 36440
    },
    {
      "epoch": 19.101153039832287,
      "grad_norm": 0.749782383441925,
      "learning_rate": 3.073623419847349e-07,
      "loss": 0.4666,
      "num_input_tokens_seen": 23845120,
      "step": 36445
    },
    {
      "epoch": 19.10377358490566,
      "grad_norm": 1.1556978225708008,
      "learning_rate": 3.0557744819926235e-07,
      "loss": 0.4753,
      "num_input_tokens_seen": 23848480,
      "step": 36450
    },
    {
      "epoch": 19.106394129979037,
      "grad_norm": 0.47258877754211426,
      "learning_rate": 3.037977202052594e-07,
      "loss": 0.3858,
      "num_input_tokens_seen": 23852000,
      "step": 36455
    },
    {
      "epoch": 19.10901467505241,
      "grad_norm": 0.35882773995399475,
      "learning_rate": 3.020231583750255e-07,
      "loss": 0.3711,
      "num_input_tokens_seen": 23856224,
      "step": 36460
    },
    {
      "epoch": 19.111635220125788,
      "grad_norm": 0.4032195210456848,
      "learning_rate": 3.002537630797747e-07,
      "loss": 0.3429,
      "num_input_tokens_seen": 23859680,
      "step": 36465
    },
    {
      "epoch": 19.11425576519916,
      "grad_norm": 0.29046958684921265,
      "learning_rate": 2.984895346896499e-07,
      "loss": 0.3925,
      "num_input_tokens_seen": 23863168,
      "step": 36470
    },
    {
      "epoch": 19.116876310272538,
      "grad_norm": 0.5329388380050659,
      "learning_rate": 2.967304735737031e-07,
      "loss": 0.5003,
      "num_input_tokens_seen": 23866208,
      "step": 36475
    },
    {
      "epoch": 19.11949685534591,
      "grad_norm": 0.44414299726486206,
      "learning_rate": 2.9497658009991214e-07,
      "loss": 0.363,
      "num_input_tokens_seen": 23869152,
      "step": 36480
    },
    {
      "epoch": 19.122117400419288,
      "grad_norm": 0.4903024733066559,
      "learning_rate": 2.932278546351724e-07,
      "loss": 0.3183,
      "num_input_tokens_seen": 23872768,
      "step": 36485
    },
    {
      "epoch": 19.12473794549266,
      "grad_norm": 0.5645039677619934,
      "learning_rate": 2.9148429754529137e-07,
      "loss": 0.3949,
      "num_input_tokens_seen": 23875744,
      "step": 36490
    },
    {
      "epoch": 19.12735849056604,
      "grad_norm": 0.49511536955833435,
      "learning_rate": 2.897459091950078e-07,
      "loss": 0.3055,
      "num_input_tokens_seen": 23877984,
      "step": 36495
    },
    {
      "epoch": 19.129979035639412,
      "grad_norm": 0.5691615343093872,
      "learning_rate": 2.880126899479724e-07,
      "loss": 0.6469,
      "num_input_tokens_seen": 23881504,
      "step": 36500
    },
    {
      "epoch": 19.13259958071279,
      "grad_norm": 0.5839917063713074,
      "learning_rate": 2.862846401667507e-07,
      "loss": 0.3519,
      "num_input_tokens_seen": 23884864,
      "step": 36505
    },
    {
      "epoch": 19.135220125786162,
      "grad_norm": 0.5735644698143005,
      "learning_rate": 2.8456176021283697e-07,
      "loss": 0.5442,
      "num_input_tokens_seen": 23888000,
      "step": 36510
    },
    {
      "epoch": 19.13784067085954,
      "grad_norm": 0.3881551921367645,
      "learning_rate": 2.8284405044662897e-07,
      "loss": 0.5383,
      "num_input_tokens_seen": 23890528,
      "step": 36515
    },
    {
      "epoch": 19.140461215932913,
      "grad_norm": 0.6286579370498657,
      "learning_rate": 2.8113151122745874e-07,
      "loss": 0.3584,
      "num_input_tokens_seen": 23893184,
      "step": 36520
    },
    {
      "epoch": 19.14308176100629,
      "grad_norm": 0.5009663105010986,
      "learning_rate": 2.7942414291357023e-07,
      "loss": 0.4723,
      "num_input_tokens_seen": 23895968,
      "step": 36525
    },
    {
      "epoch": 19.145702306079663,
      "grad_norm": 0.3894546329975128,
      "learning_rate": 2.777219458621222e-07,
      "loss": 0.4229,
      "num_input_tokens_seen": 23898880,
      "step": 36530
    },
    {
      "epoch": 19.14832285115304,
      "grad_norm": 0.47715356945991516,
      "learning_rate": 2.7602492042919923e-07,
      "loss": 0.4625,
      "num_input_tokens_seen": 23901504,
      "step": 36535
    },
    {
      "epoch": 19.150943396226417,
      "grad_norm": 0.4235820770263672,
      "learning_rate": 2.7433306696979523e-07,
      "loss": 0.4526,
      "num_input_tokens_seen": 23904768,
      "step": 36540
    },
    {
      "epoch": 19.15356394129979,
      "grad_norm": 0.605679452419281,
      "learning_rate": 2.7264638583782973e-07,
      "loss": 0.4994,
      "num_input_tokens_seen": 23907872,
      "step": 36545
    },
    {
      "epoch": 19.156184486373167,
      "grad_norm": 0.3501911461353302,
      "learning_rate": 2.7096487738613177e-07,
      "loss": 0.3813,
      "num_input_tokens_seen": 23910816,
      "step": 36550
    },
    {
      "epoch": 19.15880503144654,
      "grad_norm": 0.5824812054634094,
      "learning_rate": 2.6928854196646147e-07,
      "loss": 0.4918,
      "num_input_tokens_seen": 23914048,
      "step": 36555
    },
    {
      "epoch": 19.161425576519918,
      "grad_norm": 0.43497487902641296,
      "learning_rate": 2.6761737992948565e-07,
      "loss": 0.3547,
      "num_input_tokens_seen": 23918016,
      "step": 36560
    },
    {
      "epoch": 19.16404612159329,
      "grad_norm": 0.6558104753494263,
      "learning_rate": 2.6595139162479134e-07,
      "loss": 0.371,
      "num_input_tokens_seen": 23921056,
      "step": 36565
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 0.5765544176101685,
      "learning_rate": 2.6429057740088303e-07,
      "loss": 0.5106,
      "num_input_tokens_seen": 23923648,
      "step": 36570
    },
    {
      "epoch": 19.16928721174004,
      "grad_norm": 0.5246214866638184,
      "learning_rate": 2.6263493760518564e-07,
      "loss": 0.3711,
      "num_input_tokens_seen": 23926592,
      "step": 36575
    },
    {
      "epoch": 19.171907756813418,
      "grad_norm": 0.4130973219871521,
      "learning_rate": 2.6098447258404157e-07,
      "loss": 0.4643,
      "num_input_tokens_seen": 23929856,
      "step": 36580
    },
    {
      "epoch": 19.17452830188679,
      "grad_norm": 0.6025440692901611,
      "learning_rate": 2.593391826827107e-07,
      "loss": 0.3838,
      "num_input_tokens_seen": 23932608,
      "step": 36585
    },
    {
      "epoch": 19.17714884696017,
      "grad_norm": 0.32864293456077576,
      "learning_rate": 2.5769906824536225e-07,
      "loss": 0.5844,
      "num_input_tokens_seen": 23935488,
      "step": 36590
    },
    {
      "epoch": 19.179769392033542,
      "grad_norm": 0.5923155546188354,
      "learning_rate": 2.560641296150967e-07,
      "loss": 0.4753,
      "num_input_tokens_seen": 23938432,
      "step": 36595
    },
    {
      "epoch": 19.18238993710692,
      "grad_norm": 0.4451850354671478,
      "learning_rate": 2.544343671339211e-07,
      "loss": 0.4992,
      "num_input_tokens_seen": 23941120,
      "step": 36600
    },
    {
      "epoch": 19.185010482180292,
      "grad_norm": 0.3234461545944214,
      "learning_rate": 2.5280978114276264e-07,
      "loss": 0.4446,
      "num_input_tokens_seen": 23944320,
      "step": 36605
    },
    {
      "epoch": 19.18763102725367,
      "grad_norm": 0.49286311864852905,
      "learning_rate": 2.511903719814662e-07,
      "loss": 0.5439,
      "num_input_tokens_seen": 23946624,
      "step": 36610
    },
    {
      "epoch": 19.190251572327043,
      "grad_norm": 0.463424414396286,
      "learning_rate": 2.4957613998879695e-07,
      "loss": 0.3723,
      "num_input_tokens_seen": 23949440,
      "step": 36615
    },
    {
      "epoch": 19.19287211740042,
      "grad_norm": 0.49395856261253357,
      "learning_rate": 2.479670855024291e-07,
      "loss": 0.3203,
      "num_input_tokens_seen": 23952224,
      "step": 36620
    },
    {
      "epoch": 19.195492662473793,
      "grad_norm": 0.44881826639175415,
      "learning_rate": 2.463632088589629e-07,
      "loss": 0.3747,
      "num_input_tokens_seen": 23955072,
      "step": 36625
    },
    {
      "epoch": 19.19811320754717,
      "grad_norm": 1.0179930925369263,
      "learning_rate": 2.447645103939078e-07,
      "loss": 0.4266,
      "num_input_tokens_seen": 23958336,
      "step": 36630
    },
    {
      "epoch": 19.200733752620547,
      "grad_norm": 0.38306915760040283,
      "learning_rate": 2.431709904416962e-07,
      "loss": 0.3012,
      "num_input_tokens_seen": 23961504,
      "step": 36635
    },
    {
      "epoch": 19.20335429769392,
      "grad_norm": 0.6335029602050781,
      "learning_rate": 2.4158264933567264e-07,
      "loss": 0.6953,
      "num_input_tokens_seen": 23964288,
      "step": 36640
    },
    {
      "epoch": 19.205974842767297,
      "grad_norm": 0.5416715741157532,
      "learning_rate": 2.399994874080963e-07,
      "loss": 0.4647,
      "num_input_tokens_seen": 23967776,
      "step": 36645
    },
    {
      "epoch": 19.20859538784067,
      "grad_norm": 0.6528046131134033,
      "learning_rate": 2.3842150499015504e-07,
      "loss": 0.3073,
      "num_input_tokens_seen": 23970688,
      "step": 36650
    },
    {
      "epoch": 19.211215932914047,
      "grad_norm": 0.23783813416957855,
      "learning_rate": 2.368487024119348e-07,
      "loss": 0.4254,
      "num_input_tokens_seen": 23974208,
      "step": 36655
    },
    {
      "epoch": 19.21383647798742,
      "grad_norm": 0.40198931097984314,
      "learning_rate": 2.3528108000245573e-07,
      "loss": 0.4739,
      "num_input_tokens_seen": 23978080,
      "step": 36660
    },
    {
      "epoch": 19.216457023060798,
      "grad_norm": 0.4176322817802429,
      "learning_rate": 2.3371863808964445e-07,
      "loss": 0.4931,
      "num_input_tokens_seen": 23982976,
      "step": 36665
    },
    {
      "epoch": 19.21907756813417,
      "grad_norm": 0.25287193059921265,
      "learning_rate": 2.32161377000345e-07,
      "loss": 0.4427,
      "num_input_tokens_seen": 23986528,
      "step": 36670
    },
    {
      "epoch": 19.221698113207548,
      "grad_norm": 0.5175942182540894,
      "learning_rate": 2.306092970603163e-07,
      "loss": 0.3904,
      "num_input_tokens_seen": 23989376,
      "step": 36675
    },
    {
      "epoch": 19.22431865828092,
      "grad_norm": 0.8394734263420105,
      "learning_rate": 2.2906239859424306e-07,
      "loss": 0.4462,
      "num_input_tokens_seen": 23991648,
      "step": 36680
    },
    {
      "epoch": 19.2269392033543,
      "grad_norm": 0.38706573843955994,
      "learning_rate": 2.2752068192571084e-07,
      "loss": 0.4378,
      "num_input_tokens_seen": 23994336,
      "step": 36685
    },
    {
      "epoch": 19.229559748427672,
      "grad_norm": 0.6382182836532593,
      "learning_rate": 2.2598414737723396e-07,
      "loss": 0.4283,
      "num_input_tokens_seen": 23997632,
      "step": 36690
    },
    {
      "epoch": 19.23218029350105,
      "grad_norm": 0.5709592700004578,
      "learning_rate": 2.2445279527023578e-07,
      "loss": 0.3862,
      "num_input_tokens_seen": 24000384,
      "step": 36695
    },
    {
      "epoch": 19.234800838574422,
      "grad_norm": 0.48105284571647644,
      "learning_rate": 2.2292662592506007e-07,
      "loss": 0.4977,
      "num_input_tokens_seen": 24003616,
      "step": 36700
    },
    {
      "epoch": 19.2374213836478,
      "grad_norm": 0.5479916334152222,
      "learning_rate": 2.2140563966095985e-07,
      "loss": 0.4244,
      "num_input_tokens_seen": 24006336,
      "step": 36705
    },
    {
      "epoch": 19.240041928721173,
      "grad_norm": 0.3885289430618286,
      "learning_rate": 2.1988983679611386e-07,
      "loss": 0.3558,
      "num_input_tokens_seen": 24011168,
      "step": 36710
    },
    {
      "epoch": 19.24266247379455,
      "grad_norm": 0.6321597695350647,
      "learning_rate": 2.1837921764760462e-07,
      "loss": 0.3466,
      "num_input_tokens_seen": 24014240,
      "step": 36715
    },
    {
      "epoch": 19.245283018867923,
      "grad_norm": 0.48356112837791443,
      "learning_rate": 2.1687378253144043e-07,
      "loss": 0.4825,
      "num_input_tokens_seen": 24017568,
      "step": 36720
    },
    {
      "epoch": 19.2479035639413,
      "grad_norm": 0.38668468594551086,
      "learning_rate": 2.153735317625416e-07,
      "loss": 0.5081,
      "num_input_tokens_seen": 24021376,
      "step": 36725
    },
    {
      "epoch": 19.250524109014677,
      "grad_norm": 0.5305458307266235,
      "learning_rate": 2.1387846565474045e-07,
      "loss": 0.3901,
      "num_input_tokens_seen": 24024928,
      "step": 36730
    },
    {
      "epoch": 19.25314465408805,
      "grad_norm": 0.6321129202842712,
      "learning_rate": 2.123885845207896e-07,
      "loss": 0.4165,
      "num_input_tokens_seen": 24028128,
      "step": 36735
    },
    {
      "epoch": 19.255765199161427,
      "grad_norm": 0.30889564752578735,
      "learning_rate": 2.1090388867235356e-07,
      "loss": 0.3669,
      "num_input_tokens_seen": 24031584,
      "step": 36740
    },
    {
      "epoch": 19.2583857442348,
      "grad_norm": 0.48903435468673706,
      "learning_rate": 2.0942437842001728e-07,
      "loss": 0.4897,
      "num_input_tokens_seen": 24034336,
      "step": 36745
    },
    {
      "epoch": 19.261006289308177,
      "grad_norm": 0.5690572261810303,
      "learning_rate": 2.0795005407327483e-07,
      "loss": 0.5133,
      "num_input_tokens_seen": 24037408,
      "step": 36750
    },
    {
      "epoch": 19.26362683438155,
      "grad_norm": 0.49508383870124817,
      "learning_rate": 2.0648091594053508e-07,
      "loss": 0.4559,
      "num_input_tokens_seen": 24040128,
      "step": 36755
    },
    {
      "epoch": 19.266247379454928,
      "grad_norm": 0.5053675770759583,
      "learning_rate": 2.0501696432913275e-07,
      "loss": 0.4295,
      "num_input_tokens_seen": 24042944,
      "step": 36760
    },
    {
      "epoch": 19.2688679245283,
      "grad_norm": 0.5816635489463806,
      "learning_rate": 2.035581995453034e-07,
      "loss": 0.4759,
      "num_input_tokens_seen": 24045952,
      "step": 36765
    },
    {
      "epoch": 19.271488469601678,
      "grad_norm": 0.8861939907073975,
      "learning_rate": 2.0210462189420577e-07,
      "loss": 0.5413,
      "num_input_tokens_seen": 24048768,
      "step": 36770
    },
    {
      "epoch": 19.27410901467505,
      "grad_norm": 0.538878858089447,
      "learning_rate": 2.006562316799132e-07,
      "loss": 0.4359,
      "num_input_tokens_seen": 24052704,
      "step": 36775
    },
    {
      "epoch": 19.27672955974843,
      "grad_norm": 0.672548234462738,
      "learning_rate": 1.992130292054084e-07,
      "loss": 0.5778,
      "num_input_tokens_seen": 24056896,
      "step": 36780
    },
    {
      "epoch": 19.279350104821802,
      "grad_norm": 0.6928320527076721,
      "learning_rate": 1.9777501477259708e-07,
      "loss": 0.5186,
      "num_input_tokens_seen": 24059168,
      "step": 36785
    },
    {
      "epoch": 19.28197064989518,
      "grad_norm": 0.6510777473449707,
      "learning_rate": 1.9634218868229414e-07,
      "loss": 0.3696,
      "num_input_tokens_seen": 24062176,
      "step": 36790
    },
    {
      "epoch": 19.284591194968552,
      "grad_norm": 0.44702860713005066,
      "learning_rate": 1.9491455123423207e-07,
      "loss": 0.3988,
      "num_input_tokens_seen": 24065536,
      "step": 36795
    },
    {
      "epoch": 19.28721174004193,
      "grad_norm": 0.6705323457717896,
      "learning_rate": 1.9349210272705253e-07,
      "loss": 0.4452,
      "num_input_tokens_seen": 24068608,
      "step": 36800
    },
    {
      "epoch": 19.289832285115303,
      "grad_norm": 0.578920304775238,
      "learning_rate": 1.920748434583175e-07,
      "loss": 0.4093,
      "num_input_tokens_seen": 24072928,
      "step": 36805
    },
    {
      "epoch": 19.29245283018868,
      "grad_norm": 0.3506568968296051,
      "learning_rate": 1.906627737245009e-07,
      "loss": 0.3872,
      "num_input_tokens_seen": 24078176,
      "step": 36810
    },
    {
      "epoch": 19.295073375262053,
      "grad_norm": 0.8610926270484924,
      "learning_rate": 1.8925589382099152e-07,
      "loss": 0.4003,
      "num_input_tokens_seen": 24080768,
      "step": 36815
    },
    {
      "epoch": 19.29769392033543,
      "grad_norm": 0.8264482021331787,
      "learning_rate": 1.8785420404209553e-07,
      "loss": 0.476,
      "num_input_tokens_seen": 24083936,
      "step": 36820
    },
    {
      "epoch": 19.300314465408803,
      "grad_norm": 0.30694225430488586,
      "learning_rate": 1.864577046810284e-07,
      "loss": 0.3738,
      "num_input_tokens_seen": 24087200,
      "step": 36825
    },
    {
      "epoch": 19.30293501048218,
      "grad_norm": 0.4358346164226532,
      "learning_rate": 1.8506639602992038e-07,
      "loss": 0.4097,
      "num_input_tokens_seen": 24090432,
      "step": 36830
    },
    {
      "epoch": 19.305555555555557,
      "grad_norm": 0.524168074131012,
      "learning_rate": 1.836802783798164e-07,
      "loss": 0.4577,
      "num_input_tokens_seen": 24093504,
      "step": 36835
    },
    {
      "epoch": 19.30817610062893,
      "grad_norm": 0.5356569290161133,
      "learning_rate": 1.822993520206817e-07,
      "loss": 0.4908,
      "num_input_tokens_seen": 24097088,
      "step": 36840
    },
    {
      "epoch": 19.310796645702307,
      "grad_norm": 0.5813986659049988,
      "learning_rate": 1.8092361724138528e-07,
      "loss": 0.3628,
      "num_input_tokens_seen": 24099552,
      "step": 36845
    },
    {
      "epoch": 19.31341719077568,
      "grad_norm": 0.5432133674621582,
      "learning_rate": 1.7955307432971914e-07,
      "loss": 0.4481,
      "num_input_tokens_seen": 24102816,
      "step": 36850
    },
    {
      "epoch": 19.316037735849058,
      "grad_norm": 0.9491338133811951,
      "learning_rate": 1.7818772357238445e-07,
      "loss": 0.4627,
      "num_input_tokens_seen": 24105856,
      "step": 36855
    },
    {
      "epoch": 19.31865828092243,
      "grad_norm": 0.6933731436729431,
      "learning_rate": 1.7682756525499166e-07,
      "loss": 0.4023,
      "num_input_tokens_seen": 24110848,
      "step": 36860
    },
    {
      "epoch": 19.321278825995808,
      "grad_norm": 0.4943411946296692,
      "learning_rate": 1.7547259966207708e-07,
      "loss": 0.3612,
      "num_input_tokens_seen": 24114144,
      "step": 36865
    },
    {
      "epoch": 19.32389937106918,
      "grad_norm": 0.43738576769828796,
      "learning_rate": 1.7412282707708062e-07,
      "loss": 0.3335,
      "num_input_tokens_seen": 24117472,
      "step": 36870
    },
    {
      "epoch": 19.32651991614256,
      "grad_norm": 0.39802712202072144,
      "learning_rate": 1.7277824778235974e-07,
      "loss": 0.3743,
      "num_input_tokens_seen": 24120576,
      "step": 36875
    },
    {
      "epoch": 19.329140461215932,
      "grad_norm": 0.5578474402427673,
      "learning_rate": 1.714388620591867e-07,
      "loss": 0.3992,
      "num_input_tokens_seen": 24123072,
      "step": 36880
    },
    {
      "epoch": 19.33176100628931,
      "grad_norm": 1.1617497205734253,
      "learning_rate": 1.701046701877429e-07,
      "loss": 0.3997,
      "num_input_tokens_seen": 24125952,
      "step": 36885
    },
    {
      "epoch": 19.334381551362682,
      "grad_norm": 0.5974417924880981,
      "learning_rate": 1.6877567244713011e-07,
      "loss": 0.4973,
      "num_input_tokens_seen": 24128896,
      "step": 36890
    },
    {
      "epoch": 19.33700209643606,
      "grad_norm": 0.3411010205745697,
      "learning_rate": 1.6745186911535372e-07,
      "loss": 0.3958,
      "num_input_tokens_seen": 24131936,
      "step": 36895
    },
    {
      "epoch": 19.339622641509433,
      "grad_norm": 0.5284488201141357,
      "learning_rate": 1.6613326046934497e-07,
      "loss": 0.5364,
      "num_input_tokens_seen": 24134912,
      "step": 36900
    },
    {
      "epoch": 19.34224318658281,
      "grad_norm": 0.8199689388275146,
      "learning_rate": 1.64819846784936e-07,
      "loss": 0.4722,
      "num_input_tokens_seen": 24137312,
      "step": 36905
    },
    {
      "epoch": 19.344863731656183,
      "grad_norm": 0.37246739864349365,
      "learning_rate": 1.6351162833688204e-07,
      "loss": 0.3753,
      "num_input_tokens_seen": 24141024,
      "step": 36910
    },
    {
      "epoch": 19.34748427672956,
      "grad_norm": 0.507530927658081,
      "learning_rate": 1.6220860539884475e-07,
      "loss": 0.3245,
      "num_input_tokens_seen": 24144512,
      "step": 36915
    },
    {
      "epoch": 19.350104821802937,
      "grad_norm": 0.6232202649116516,
      "learning_rate": 1.6091077824340327e-07,
      "loss": 0.3536,
      "num_input_tokens_seen": 24147392,
      "step": 36920
    },
    {
      "epoch": 19.35272536687631,
      "grad_norm": 0.8779141306877136,
      "learning_rate": 1.5961814714204605e-07,
      "loss": 0.464,
      "num_input_tokens_seen": 24149824,
      "step": 36925
    },
    {
      "epoch": 19.355345911949687,
      "grad_norm": 0.3221270442008972,
      "learning_rate": 1.5833071236518172e-07,
      "loss": 0.4991,
      "num_input_tokens_seen": 24152768,
      "step": 36930
    },
    {
      "epoch": 19.35796645702306,
      "grad_norm": 0.3757447302341461,
      "learning_rate": 1.5704847418212266e-07,
      "loss": 0.4341,
      "num_input_tokens_seen": 24156480,
      "step": 36935
    },
    {
      "epoch": 19.360587002096437,
      "grad_norm": 0.5516123175621033,
      "learning_rate": 1.5577143286109875e-07,
      "loss": 0.4639,
      "num_input_tokens_seen": 24159136,
      "step": 36940
    },
    {
      "epoch": 19.36320754716981,
      "grad_norm": 0.286069393157959,
      "learning_rate": 1.5449958866925185e-07,
      "loss": 0.3521,
      "num_input_tokens_seen": 24163232,
      "step": 36945
    },
    {
      "epoch": 19.365828092243188,
      "grad_norm": 0.6526319980621338,
      "learning_rate": 1.532329418726386e-07,
      "loss": 0.405,
      "num_input_tokens_seen": 24166624,
      "step": 36950
    },
    {
      "epoch": 19.36844863731656,
      "grad_norm": 0.5069071650505066,
      "learning_rate": 1.5197149273623036e-07,
      "loss": 0.4153,
      "num_input_tokens_seen": 24169440,
      "step": 36955
    },
    {
      "epoch": 19.371069182389938,
      "grad_norm": 0.3609291613101959,
      "learning_rate": 1.5071524152389936e-07,
      "loss": 0.3208,
      "num_input_tokens_seen": 24172960,
      "step": 36960
    },
    {
      "epoch": 19.37368972746331,
      "grad_norm": 0.563906729221344,
      "learning_rate": 1.4946418849844934e-07,
      "loss": 0.3425,
      "num_input_tokens_seen": 24177152,
      "step": 36965
    },
    {
      "epoch": 19.37631027253669,
      "grad_norm": 0.5142313241958618,
      "learning_rate": 1.4821833392157923e-07,
      "loss": 0.5224,
      "num_input_tokens_seen": 24180672,
      "step": 36970
    },
    {
      "epoch": 19.378930817610062,
      "grad_norm": 0.45312562584877014,
      "learning_rate": 1.4697767805390838e-07,
      "loss": 0.4383,
      "num_input_tokens_seen": 24184256,
      "step": 36975
    },
    {
      "epoch": 19.38155136268344,
      "grad_norm": 0.2167205810546875,
      "learning_rate": 1.4574222115497084e-07,
      "loss": 0.3218,
      "num_input_tokens_seen": 24189568,
      "step": 36980
    },
    {
      "epoch": 19.384171907756812,
      "grad_norm": 0.6348816156387329,
      "learning_rate": 1.4451196348320706e-07,
      "loss": 0.443,
      "num_input_tokens_seen": 24192288,
      "step": 36985
    },
    {
      "epoch": 19.38679245283019,
      "grad_norm": 0.4267576336860657,
      "learning_rate": 1.4328690529597232e-07,
      "loss": 0.4322,
      "num_input_tokens_seen": 24195584,
      "step": 36990
    },
    {
      "epoch": 19.389412997903563,
      "grad_norm": 0.5625284910202026,
      "learning_rate": 1.4206704684953943e-07,
      "loss": 0.4421,
      "num_input_tokens_seen": 24198048,
      "step": 36995
    },
    {
      "epoch": 19.39203354297694,
      "grad_norm": 1.0583579540252686,
      "learning_rate": 1.4085238839908476e-07,
      "loss": 0.3641,
      "num_input_tokens_seen": 24200896,
      "step": 37000
    },
    {
      "epoch": 19.394654088050313,
      "grad_norm": 0.5402756929397583,
      "learning_rate": 1.3964293019870235e-07,
      "loss": 0.4058,
      "num_input_tokens_seen": 24203968,
      "step": 37005
    },
    {
      "epoch": 19.39727463312369,
      "grad_norm": 0.38574013113975525,
      "learning_rate": 1.3843867250139807e-07,
      "loss": 0.3643,
      "num_input_tokens_seen": 24207712,
      "step": 37010
    },
    {
      "epoch": 19.399895178197063,
      "grad_norm": 0.3826046884059906,
      "learning_rate": 1.3723961555908715e-07,
      "loss": 0.4022,
      "num_input_tokens_seen": 24211040,
      "step": 37015
    },
    {
      "epoch": 19.40251572327044,
      "grad_norm": 0.39022400975227356,
      "learning_rate": 1.3604575962259947e-07,
      "loss": 0.442,
      "num_input_tokens_seen": 24214592,
      "step": 37020
    },
    {
      "epoch": 19.405136268343817,
      "grad_norm": 0.4134133458137512,
      "learning_rate": 1.3485710494167692e-07,
      "loss": 0.4486,
      "num_input_tokens_seen": 24217952,
      "step": 37025
    },
    {
      "epoch": 19.40775681341719,
      "grad_norm": 0.45597416162490845,
      "learning_rate": 1.3367365176496783e-07,
      "loss": 0.4513,
      "num_input_tokens_seen": 24221344,
      "step": 37030
    },
    {
      "epoch": 19.410377358490567,
      "grad_norm": 0.2911139726638794,
      "learning_rate": 1.324954003400436e-07,
      "loss": 0.6703,
      "num_input_tokens_seen": 24224640,
      "step": 37035
    },
    {
      "epoch": 19.41299790356394,
      "grad_norm": 0.4660566747188568,
      "learning_rate": 1.313223509133793e-07,
      "loss": 0.4939,
      "num_input_tokens_seen": 24227424,
      "step": 37040
    },
    {
      "epoch": 19.415618448637318,
      "grad_norm": 0.3249620795249939,
      "learning_rate": 1.301545037303592e-07,
      "loss": 0.4171,
      "num_input_tokens_seen": 24231424,
      "step": 37045
    },
    {
      "epoch": 19.41823899371069,
      "grad_norm": 0.3742001950740814,
      "learning_rate": 1.2899185903529066e-07,
      "loss": 0.5492,
      "num_input_tokens_seen": 24234176,
      "step": 37050
    },
    {
      "epoch": 19.420859538784068,
      "grad_norm": 0.5471071600914001,
      "learning_rate": 1.278344170713791e-07,
      "loss": 0.4597,
      "num_input_tokens_seen": 24236992,
      "step": 37055
    },
    {
      "epoch": 19.42348008385744,
      "grad_norm": 0.7046257257461548,
      "learning_rate": 1.2668217808075033e-07,
      "loss": 0.3967,
      "num_input_tokens_seen": 24239936,
      "step": 37060
    },
    {
      "epoch": 19.42610062893082,
      "grad_norm": 0.587823212146759,
      "learning_rate": 1.255351423044393e-07,
      "loss": 0.4004,
      "num_input_tokens_seen": 24242976,
      "step": 37065
    },
    {
      "epoch": 19.428721174004192,
      "grad_norm": 0.5511455535888672,
      "learning_rate": 1.2439330998239572e-07,
      "loss": 0.3966,
      "num_input_tokens_seen": 24245504,
      "step": 37070
    },
    {
      "epoch": 19.43134171907757,
      "grad_norm": 0.5666557550430298,
      "learning_rate": 1.2325668135347578e-07,
      "loss": 0.5393,
      "num_input_tokens_seen": 24248704,
      "step": 37075
    },
    {
      "epoch": 19.433962264150942,
      "grad_norm": 0.6187092661857605,
      "learning_rate": 1.221252566554476e-07,
      "loss": 0.4592,
      "num_input_tokens_seen": 24252288,
      "step": 37080
    },
    {
      "epoch": 19.43658280922432,
      "grad_norm": 0.47284936904907227,
      "learning_rate": 1.2099903612499685e-07,
      "loss": 0.4229,
      "num_input_tokens_seen": 24255776,
      "step": 37085
    },
    {
      "epoch": 19.439203354297693,
      "grad_norm": 0.3093314468860626,
      "learning_rate": 1.198780199977101e-07,
      "loss": 0.4564,
      "num_input_tokens_seen": 24258624,
      "step": 37090
    },
    {
      "epoch": 19.44182389937107,
      "grad_norm": 0.6779235601425171,
      "learning_rate": 1.1876220850809416e-07,
      "loss": 0.4707,
      "num_input_tokens_seen": 24261120,
      "step": 37095
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 0.5357060432434082,
      "learning_rate": 1.1765160188956514e-07,
      "loss": 0.3749,
      "num_input_tokens_seen": 24264032,
      "step": 37100
    },
    {
      "epoch": 19.44706498951782,
      "grad_norm": 0.4608322083950043,
      "learning_rate": 1.1654620037445107e-07,
      "loss": 0.2814,
      "num_input_tokens_seen": 24267424,
      "step": 37105
    },
    {
      "epoch": 19.449685534591197,
      "grad_norm": 0.4486667513847351,
      "learning_rate": 1.1544600419398088e-07,
      "loss": 0.5264,
      "num_input_tokens_seen": 24270976,
      "step": 37110
    },
    {
      "epoch": 19.45230607966457,
      "grad_norm": 0.4571413993835449,
      "learning_rate": 1.1435101357831213e-07,
      "loss": 0.4682,
      "num_input_tokens_seen": 24273856,
      "step": 37115
    },
    {
      "epoch": 19.454926624737947,
      "grad_norm": 0.32445791363716125,
      "learning_rate": 1.1326122875650047e-07,
      "loss": 0.4744,
      "num_input_tokens_seen": 24277536,
      "step": 37120
    },
    {
      "epoch": 19.45754716981132,
      "grad_norm": 0.4601949453353882,
      "learning_rate": 1.1217664995651633e-07,
      "loss": 0.4871,
      "num_input_tokens_seen": 24283520,
      "step": 37125
    },
    {
      "epoch": 19.460167714884697,
      "grad_norm": 0.5119982957839966,
      "learning_rate": 1.1109727740523935e-07,
      "loss": 0.4595,
      "num_input_tokens_seen": 24286592,
      "step": 37130
    },
    {
      "epoch": 19.46278825995807,
      "grad_norm": 0.512017011642456,
      "learning_rate": 1.1002311132846944e-07,
      "loss": 0.4411,
      "num_input_tokens_seen": 24290624,
      "step": 37135
    },
    {
      "epoch": 19.465408805031448,
      "grad_norm": 0.535412609577179,
      "learning_rate": 1.0895415195090186e-07,
      "loss": 0.4525,
      "num_input_tokens_seen": 24293568,
      "step": 37140
    },
    {
      "epoch": 19.46802935010482,
      "grad_norm": 0.5416945219039917,
      "learning_rate": 1.0789039949615221e-07,
      "loss": 0.5063,
      "num_input_tokens_seen": 24296288,
      "step": 37145
    },
    {
      "epoch": 19.470649895178198,
      "grad_norm": 0.8914316296577454,
      "learning_rate": 1.0683185418674801e-07,
      "loss": 0.4339,
      "num_input_tokens_seen": 24298592,
      "step": 37150
    },
    {
      "epoch": 19.47327044025157,
      "grad_norm": 0.4395000636577606,
      "learning_rate": 1.0577851624412604e-07,
      "loss": 0.421,
      "num_input_tokens_seen": 24301856,
      "step": 37155
    },
    {
      "epoch": 19.47589098532495,
      "grad_norm": 0.3466298580169678,
      "learning_rate": 1.0473038588862672e-07,
      "loss": 0.3413,
      "num_input_tokens_seen": 24304832,
      "step": 37160
    },
    {
      "epoch": 19.478511530398322,
      "grad_norm": 0.40961742401123047,
      "learning_rate": 1.0368746333951351e-07,
      "loss": 0.4129,
      "num_input_tokens_seen": 24308480,
      "step": 37165
    },
    {
      "epoch": 19.4811320754717,
      "grad_norm": 0.709498405456543,
      "learning_rate": 1.0264974881494805e-07,
      "loss": 0.4514,
      "num_input_tokens_seen": 24311552,
      "step": 37170
    },
    {
      "epoch": 19.483752620545072,
      "grad_norm": 0.4621434211730957,
      "learning_rate": 1.0161724253201222e-07,
      "loss": 0.5717,
      "num_input_tokens_seen": 24314336,
      "step": 37175
    },
    {
      "epoch": 19.48637316561845,
      "grad_norm": 0.4683607518672943,
      "learning_rate": 1.0058994470669436e-07,
      "loss": 0.5203,
      "num_input_tokens_seen": 24317184,
      "step": 37180
    },
    {
      "epoch": 19.488993710691823,
      "grad_norm": 0.4854583740234375,
      "learning_rate": 9.956785555388926e-08,
      "loss": 0.4698,
      "num_input_tokens_seen": 24320544,
      "step": 37185
    },
    {
      "epoch": 19.4916142557652,
      "grad_norm": 0.5948166847229004,
      "learning_rate": 9.855097528741475e-08,
      "loss": 0.4878,
      "num_input_tokens_seen": 24324192,
      "step": 37190
    },
    {
      "epoch": 19.494234800838573,
      "grad_norm": 0.3849221169948578,
      "learning_rate": 9.753930411998402e-08,
      "loss": 0.423,
      "num_input_tokens_seen": 24328000,
      "step": 37195
    },
    {
      "epoch": 19.49685534591195,
      "grad_norm": 0.5476910471916199,
      "learning_rate": 9.653284226322501e-08,
      "loss": 0.5097,
      "num_input_tokens_seen": 24331136,
      "step": 37200
    },
    {
      "epoch": 19.499475890985323,
      "grad_norm": 0.4419437050819397,
      "learning_rate": 9.553158992768318e-08,
      "loss": 0.4127,
      "num_input_tokens_seen": 24334368,
      "step": 37205
    },
    {
      "epoch": 19.5020964360587,
      "grad_norm": 0.4847414791584015,
      "learning_rate": 9.453554732281044e-08,
      "loss": 0.4585,
      "num_input_tokens_seen": 24337248,
      "step": 37210
    },
    {
      "epoch": 19.504716981132077,
      "grad_norm": 0.7123607993125916,
      "learning_rate": 9.354471465696236e-08,
      "loss": 0.4618,
      "num_input_tokens_seen": 24340352,
      "step": 37215
    },
    {
      "epoch": 19.50733752620545,
      "grad_norm": 0.4339683949947357,
      "learning_rate": 9.255909213741199e-08,
      "loss": 0.6399,
      "num_input_tokens_seen": 24343584,
      "step": 37220
    },
    {
      "epoch": 19.509958071278827,
      "grad_norm": 0.5090374946594238,
      "learning_rate": 9.157867997034164e-08,
      "loss": 0.484,
      "num_input_tokens_seen": 24346880,
      "step": 37225
    },
    {
      "epoch": 19.5125786163522,
      "grad_norm": 0.43359217047691345,
      "learning_rate": 9.060347836084004e-08,
      "loss": 0.2973,
      "num_input_tokens_seen": 24351552,
      "step": 37230
    },
    {
      "epoch": 19.515199161425578,
      "grad_norm": 0.46689969301223755,
      "learning_rate": 8.96334875129079e-08,
      "loss": 0.4056,
      "num_input_tokens_seen": 24355360,
      "step": 37235
    },
    {
      "epoch": 19.51781970649895,
      "grad_norm": 0.4279880225658417,
      "learning_rate": 8.866870762946067e-08,
      "loss": 0.3697,
      "num_input_tokens_seen": 24358336,
      "step": 37240
    },
    {
      "epoch": 19.520440251572328,
      "grad_norm": 0.575638473033905,
      "learning_rate": 8.77091389123147e-08,
      "loss": 0.4712,
      "num_input_tokens_seen": 24360992,
      "step": 37245
    },
    {
      "epoch": 19.5230607966457,
      "grad_norm": 0.6309239864349365,
      "learning_rate": 8.675478156220385e-08,
      "loss": 0.5191,
      "num_input_tokens_seen": 24363296,
      "step": 37250
    },
    {
      "epoch": 19.52568134171908,
      "grad_norm": 0.47414490580558777,
      "learning_rate": 8.580563577876566e-08,
      "loss": 0.3216,
      "num_input_tokens_seen": 24365568,
      "step": 37255
    },
    {
      "epoch": 19.528301886792452,
      "grad_norm": 0.4539867341518402,
      "learning_rate": 8.486170176055242e-08,
      "loss": 0.3845,
      "num_input_tokens_seen": 24368672,
      "step": 37260
    },
    {
      "epoch": 19.53092243186583,
      "grad_norm": 0.5566781163215637,
      "learning_rate": 8.392297970502282e-08,
      "loss": 0.3525,
      "num_input_tokens_seen": 24371456,
      "step": 37265
    },
    {
      "epoch": 19.533542976939202,
      "grad_norm": 0.272544801235199,
      "learning_rate": 8.298946980855315e-08,
      "loss": 0.4334,
      "num_input_tokens_seen": 24375200,
      "step": 37270
    },
    {
      "epoch": 19.53616352201258,
      "grad_norm": 0.5457293391227722,
      "learning_rate": 8.206117226641497e-08,
      "loss": 0.4631,
      "num_input_tokens_seen": 24377824,
      "step": 37275
    },
    {
      "epoch": 19.538784067085953,
      "grad_norm": 0.3873611092567444,
      "learning_rate": 8.113808727280015e-08,
      "loss": 0.3837,
      "num_input_tokens_seen": 24381184,
      "step": 37280
    },
    {
      "epoch": 19.54140461215933,
      "grad_norm": 0.3792857527732849,
      "learning_rate": 8.022021502080979e-08,
      "loss": 0.3452,
      "num_input_tokens_seen": 24384448,
      "step": 37285
    },
    {
      "epoch": 19.544025157232703,
      "grad_norm": 0.33682799339294434,
      "learning_rate": 7.930755570244863e-08,
      "loss": 0.3095,
      "num_input_tokens_seen": 24387104,
      "step": 37290
    },
    {
      "epoch": 19.54664570230608,
      "grad_norm": 0.44895321130752563,
      "learning_rate": 7.840010950863619e-08,
      "loss": 0.4715,
      "num_input_tokens_seen": 24389792,
      "step": 37295
    },
    {
      "epoch": 19.549266247379457,
      "grad_norm": 0.7286931276321411,
      "learning_rate": 7.749787662920116e-08,
      "loss": 0.4978,
      "num_input_tokens_seen": 24393632,
      "step": 37300
    },
    {
      "epoch": 19.55188679245283,
      "grad_norm": 0.5465759634971619,
      "learning_rate": 7.660085725288146e-08,
      "loss": 0.3695,
      "num_input_tokens_seen": 24397152,
      "step": 37305
    },
    {
      "epoch": 19.554507337526207,
      "grad_norm": 0.4530813992023468,
      "learning_rate": 7.570905156731867e-08,
      "loss": 0.4756,
      "num_input_tokens_seen": 24401152,
      "step": 37310
    },
    {
      "epoch": 19.55712788259958,
      "grad_norm": 0.30960512161254883,
      "learning_rate": 7.482245975907188e-08,
      "loss": 0.453,
      "num_input_tokens_seen": 24404896,
      "step": 37315
    },
    {
      "epoch": 19.559748427672957,
      "grad_norm": 0.4686301052570343,
      "learning_rate": 7.394108201360383e-08,
      "loss": 0.4291,
      "num_input_tokens_seen": 24407712,
      "step": 37320
    },
    {
      "epoch": 19.56236897274633,
      "grad_norm": 0.6090126633644104,
      "learning_rate": 7.306491851529206e-08,
      "loss": 0.4296,
      "num_input_tokens_seen": 24411232,
      "step": 37325
    },
    {
      "epoch": 19.564989517819708,
      "grad_norm": 0.8640155792236328,
      "learning_rate": 7.219396944741774e-08,
      "loss": 0.3686,
      "num_input_tokens_seen": 24414432,
      "step": 37330
    },
    {
      "epoch": 19.56761006289308,
      "grad_norm": 0.456109881401062,
      "learning_rate": 7.132823499217123e-08,
      "loss": 0.4256,
      "num_input_tokens_seen": 24417600,
      "step": 37335
    },
    {
      "epoch": 19.570230607966458,
      "grad_norm": 0.75421541929245,
      "learning_rate": 7.046771533065766e-08,
      "loss": 0.3906,
      "num_input_tokens_seen": 24420416,
      "step": 37340
    },
    {
      "epoch": 19.57285115303983,
      "grad_norm": 0.5313379764556885,
      "learning_rate": 6.961241064288859e-08,
      "loss": 0.4479,
      "num_input_tokens_seen": 24424000,
      "step": 37345
    },
    {
      "epoch": 19.57547169811321,
      "grad_norm": 0.6702365875244141,
      "learning_rate": 6.876232110778202e-08,
      "loss": 0.3888,
      "num_input_tokens_seen": 24427552,
      "step": 37350
    },
    {
      "epoch": 19.578092243186582,
      "grad_norm": 1.3289391994476318,
      "learning_rate": 6.791744690316515e-08,
      "loss": 0.4454,
      "num_input_tokens_seen": 24430176,
      "step": 37355
    },
    {
      "epoch": 19.58071278825996,
      "grad_norm": 0.5961002707481384,
      "learning_rate": 6.707778820577715e-08,
      "loss": 0.4247,
      "num_input_tokens_seen": 24433312,
      "step": 37360
    },
    {
      "epoch": 19.583333333333332,
      "grad_norm": 0.5218426585197449,
      "learning_rate": 6.624334519126919e-08,
      "loss": 0.3052,
      "num_input_tokens_seen": 24436608,
      "step": 37365
    },
    {
      "epoch": 19.58595387840671,
      "grad_norm": 0.4358714520931244,
      "learning_rate": 6.541411803419329e-08,
      "loss": 0.5175,
      "num_input_tokens_seen": 24439136,
      "step": 37370
    },
    {
      "epoch": 19.588574423480082,
      "grad_norm": 0.3155589997768402,
      "learning_rate": 6.459010690801626e-08,
      "loss": 0.3319,
      "num_input_tokens_seen": 24442880,
      "step": 37375
    },
    {
      "epoch": 19.59119496855346,
      "grad_norm": 0.6623464226722717,
      "learning_rate": 6.377131198510855e-08,
      "loss": 0.3913,
      "num_input_tokens_seen": 24446240,
      "step": 37380
    },
    {
      "epoch": 19.593815513626833,
      "grad_norm": 0.36218297481536865,
      "learning_rate": 6.295773343675537e-08,
      "loss": 0.5192,
      "num_input_tokens_seen": 24449376,
      "step": 37385
    },
    {
      "epoch": 19.59643605870021,
      "grad_norm": 0.31819936633110046,
      "learning_rate": 6.214937143314837e-08,
      "loss": 0.5417,
      "num_input_tokens_seen": 24452608,
      "step": 37390
    },
    {
      "epoch": 19.599056603773583,
      "grad_norm": 0.4976936876773834,
      "learning_rate": 6.134622614338559e-08,
      "loss": 0.5083,
      "num_input_tokens_seen": 24455008,
      "step": 37395
    },
    {
      "epoch": 19.60167714884696,
      "grad_norm": 0.4165554642677307,
      "learning_rate": 6.05482977354771e-08,
      "loss": 0.5119,
      "num_input_tokens_seen": 24459488,
      "step": 37400
    },
    {
      "epoch": 19.604297693920337,
      "grad_norm": 0.4602503776550293,
      "learning_rate": 5.975558637634215e-08,
      "loss": 0.5157,
      "num_input_tokens_seen": 24463136,
      "step": 37405
    },
    {
      "epoch": 19.60691823899371,
      "grad_norm": 0.8883538246154785,
      "learning_rate": 5.896809223180366e-08,
      "loss": 0.4454,
      "num_input_tokens_seen": 24466240,
      "step": 37410
    },
    {
      "epoch": 19.609538784067087,
      "grad_norm": 0.39050018787384033,
      "learning_rate": 5.8185815466596516e-08,
      "loss": 0.425,
      "num_input_tokens_seen": 24469344,
      "step": 37415
    },
    {
      "epoch": 19.61215932914046,
      "grad_norm": 0.7377591133117676,
      "learning_rate": 5.7408756244370365e-08,
      "loss": 0.4923,
      "num_input_tokens_seen": 24471872,
      "step": 37420
    },
    {
      "epoch": 19.614779874213838,
      "grad_norm": 0.6550792455673218,
      "learning_rate": 5.663691472766741e-08,
      "loss": 0.4626,
      "num_input_tokens_seen": 24474752,
      "step": 37425
    },
    {
      "epoch": 19.61740041928721,
      "grad_norm": 0.5100377798080444,
      "learning_rate": 5.587029107795572e-08,
      "loss": 0.5325,
      "num_input_tokens_seen": 24480000,
      "step": 37430
    },
    {
      "epoch": 19.620020964360588,
      "grad_norm": 0.5252776741981506,
      "learning_rate": 5.510888545559867e-08,
      "loss": 0.4392,
      "num_input_tokens_seen": 24482656,
      "step": 37435
    },
    {
      "epoch": 19.62264150943396,
      "grad_norm": 0.3436260223388672,
      "learning_rate": 5.435269801987997e-08,
      "loss": 0.4486,
      "num_input_tokens_seen": 24485856,
      "step": 37440
    },
    {
      "epoch": 19.62526205450734,
      "grad_norm": 0.5852154493331909,
      "learning_rate": 5.360172892897863e-08,
      "loss": 0.4782,
      "num_input_tokens_seen": 24489600,
      "step": 37445
    },
    {
      "epoch": 19.627882599580712,
      "grad_norm": 0.5774149298667908,
      "learning_rate": 5.285597833999401e-08,
      "loss": 0.357,
      "num_input_tokens_seen": 24492160,
      "step": 37450
    },
    {
      "epoch": 19.63050314465409,
      "grad_norm": 0.5804457068443298,
      "learning_rate": 5.2115446408926295e-08,
      "loss": 0.435,
      "num_input_tokens_seen": 24495200,
      "step": 37455
    },
    {
      "epoch": 19.633123689727462,
      "grad_norm": 0.5360037088394165,
      "learning_rate": 5.138013329068769e-08,
      "loss": 0.3938,
      "num_input_tokens_seen": 24498624,
      "step": 37460
    },
    {
      "epoch": 19.63574423480084,
      "grad_norm": 0.2974311411380768,
      "learning_rate": 5.065003913909405e-08,
      "loss": 0.4888,
      "num_input_tokens_seen": 24501760,
      "step": 37465
    },
    {
      "epoch": 19.638364779874212,
      "grad_norm": 0.9357287883758545,
      "learning_rate": 4.9925164106875975e-08,
      "loss": 0.4499,
      "num_input_tokens_seen": 24506560,
      "step": 37470
    },
    {
      "epoch": 19.64098532494759,
      "grad_norm": 0.6801042556762695,
      "learning_rate": 4.9205508345670506e-08,
      "loss": 0.4016,
      "num_input_tokens_seen": 24509472,
      "step": 37475
    },
    {
      "epoch": 19.643605870020963,
      "grad_norm": 0.43261322379112244,
      "learning_rate": 4.8491072006015546e-08,
      "loss": 0.3965,
      "num_input_tokens_seen": 24512608,
      "step": 37480
    },
    {
      "epoch": 19.64622641509434,
      "grad_norm": 0.32740074396133423,
      "learning_rate": 4.778185523736933e-08,
      "loss": 0.4034,
      "num_input_tokens_seen": 24516064,
      "step": 37485
    },
    {
      "epoch": 19.648846960167717,
      "grad_norm": 0.3511826694011688,
      "learning_rate": 4.7077858188085414e-08,
      "loss": 0.3198,
      "num_input_tokens_seen": 24519744,
      "step": 37490
    },
    {
      "epoch": 19.65146750524109,
      "grad_norm": 0.7550801634788513,
      "learning_rate": 4.6379081005437644e-08,
      "loss": 0.4499,
      "num_input_tokens_seen": 24523072,
      "step": 37495
    },
    {
      "epoch": 19.654088050314467,
      "grad_norm": 0.3591403067111969,
      "learning_rate": 4.568552383560076e-08,
      "loss": 0.3766,
      "num_input_tokens_seen": 24526752,
      "step": 37500
    },
    {
      "epoch": 19.65670859538784,
      "grad_norm": 0.6638772487640381,
      "learning_rate": 4.4997186823655945e-08,
      "loss": 0.4869,
      "num_input_tokens_seen": 24529184,
      "step": 37505
    },
    {
      "epoch": 19.659329140461217,
      "grad_norm": 0.619357168674469,
      "learning_rate": 4.431407011360189e-08,
      "loss": 0.4611,
      "num_input_tokens_seen": 24532576,
      "step": 37510
    },
    {
      "epoch": 19.66194968553459,
      "grad_norm": 0.6900088787078857,
      "learning_rate": 4.3636173848332627e-08,
      "loss": 0.349,
      "num_input_tokens_seen": 24535648,
      "step": 37515
    },
    {
      "epoch": 19.664570230607968,
      "grad_norm": 0.5283235907554626,
      "learning_rate": 4.2963498169656946e-08,
      "loss": 0.3963,
      "num_input_tokens_seen": 24539136,
      "step": 37520
    },
    {
      "epoch": 19.66719077568134,
      "grad_norm": 0.4547334611415863,
      "learning_rate": 4.229604321829561e-08,
      "loss": 0.4328,
      "num_input_tokens_seen": 24543008,
      "step": 37525
    },
    {
      "epoch": 19.669811320754718,
      "grad_norm": 0.35270828008651733,
      "learning_rate": 4.16338091338675e-08,
      "loss": 0.4128,
      "num_input_tokens_seen": 24549408,
      "step": 37530
    },
    {
      "epoch": 19.67243186582809,
      "grad_norm": 0.46713218092918396,
      "learning_rate": 4.097679605490623e-08,
      "loss": 0.3934,
      "num_input_tokens_seen": 24552960,
      "step": 37535
    },
    {
      "epoch": 19.67505241090147,
      "grad_norm": 0.5429545640945435,
      "learning_rate": 4.032500411885465e-08,
      "loss": 0.4795,
      "num_input_tokens_seen": 24555648,
      "step": 37540
    },
    {
      "epoch": 19.677672955974842,
      "grad_norm": 0.6036936640739441,
      "learning_rate": 3.967843346205369e-08,
      "loss": 0.4223,
      "num_input_tokens_seen": 24558976,
      "step": 37545
    },
    {
      "epoch": 19.68029350104822,
      "grad_norm": 0.5002838969230652,
      "learning_rate": 3.903708421976182e-08,
      "loss": 0.4861,
      "num_input_tokens_seen": 24562560,
      "step": 37550
    },
    {
      "epoch": 19.682914046121592,
      "grad_norm": 0.5205735564231873,
      "learning_rate": 3.840095652614395e-08,
      "loss": 0.5609,
      "num_input_tokens_seen": 24565664,
      "step": 37555
    },
    {
      "epoch": 19.68553459119497,
      "grad_norm": 0.4606568515300751,
      "learning_rate": 3.777005051426863e-08,
      "loss": 0.3807,
      "num_input_tokens_seen": 24569120,
      "step": 37560
    },
    {
      "epoch": 19.688155136268342,
      "grad_norm": 0.27116385102272034,
      "learning_rate": 3.71443663161164e-08,
      "loss": 0.3511,
      "num_input_tokens_seen": 24572320,
      "step": 37565
    },
    {
      "epoch": 19.69077568134172,
      "grad_norm": 0.8854178786277771,
      "learning_rate": 3.652390406256867e-08,
      "loss": 0.503,
      "num_input_tokens_seen": 24575808,
      "step": 37570
    },
    {
      "epoch": 19.693396226415093,
      "grad_norm": 0.5898087620735168,
      "learning_rate": 3.5908663883421624e-08,
      "loss": 0.3868,
      "num_input_tokens_seen": 24579072,
      "step": 37575
    },
    {
      "epoch": 19.69601677148847,
      "grad_norm": 0.8579962849617004,
      "learning_rate": 3.529864590737786e-08,
      "loss": 0.3664,
      "num_input_tokens_seen": 24581536,
      "step": 37580
    },
    {
      "epoch": 19.698637316561843,
      "grad_norm": 0.3923129737377167,
      "learning_rate": 3.4693850262046415e-08,
      "loss": 0.347,
      "num_input_tokens_seen": 24585472,
      "step": 37585
    },
    {
      "epoch": 19.70125786163522,
      "grad_norm": 0.4971328675746918,
      "learning_rate": 3.409427707394275e-08,
      "loss": 0.6014,
      "num_input_tokens_seen": 24588320,
      "step": 37590
    },
    {
      "epoch": 19.703878406708597,
      "grad_norm": 0.5185381174087524,
      "learning_rate": 3.349992646848599e-08,
      "loss": 0.3378,
      "num_input_tokens_seen": 24591328,
      "step": 37595
    },
    {
      "epoch": 19.70649895178197,
      "grad_norm": 0.37594911456108093,
      "learning_rate": 3.291079857001556e-08,
      "loss": 0.4491,
      "num_input_tokens_seen": 24594976,
      "step": 37600
    },
    {
      "epoch": 19.709119496855347,
      "grad_norm": 1.1850882768630981,
      "learning_rate": 3.2326893501763436e-08,
      "loss": 0.3862,
      "num_input_tokens_seen": 24597504,
      "step": 37605
    },
    {
      "epoch": 19.71174004192872,
      "grad_norm": 0.48439842462539673,
      "learning_rate": 3.174821138588191e-08,
      "loss": 0.4758,
      "num_input_tokens_seen": 24600128,
      "step": 37610
    },
    {
      "epoch": 19.714360587002098,
      "grad_norm": 0.3691757619380951,
      "learning_rate": 3.11747523434186e-08,
      "loss": 0.3378,
      "num_input_tokens_seen": 24603648,
      "step": 37615
    },
    {
      "epoch": 19.71698113207547,
      "grad_norm": 0.4066995084285736,
      "learning_rate": 3.060651649434143e-08,
      "loss": 0.5153,
      "num_input_tokens_seen": 24606688,
      "step": 37620
    },
    {
      "epoch": 19.719601677148848,
      "grad_norm": 0.4289129376411438,
      "learning_rate": 3.004350395751088e-08,
      "loss": 0.4182,
      "num_input_tokens_seen": 24610688,
      "step": 37625
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 0.40782856941223145,
      "learning_rate": 2.94857148507105e-08,
      "loss": 0.3324,
      "num_input_tokens_seen": 24615456,
      "step": 37630
    },
    {
      "epoch": 19.7248427672956,
      "grad_norm": 0.3539443910121918,
      "learning_rate": 2.893314929061641e-08,
      "loss": 0.3822,
      "num_input_tokens_seen": 24618720,
      "step": 37635
    },
    {
      "epoch": 19.72746331236897,
      "grad_norm": 0.324797660112381,
      "learning_rate": 2.8385807392825037e-08,
      "loss": 0.4018,
      "num_input_tokens_seen": 24621184,
      "step": 37640
    },
    {
      "epoch": 19.73008385744235,
      "grad_norm": 0.31121641397476196,
      "learning_rate": 2.7843689271830897e-08,
      "loss": 0.6109,
      "num_input_tokens_seen": 24624512,
      "step": 37645
    },
    {
      "epoch": 19.732704402515722,
      "grad_norm": 0.4283892512321472,
      "learning_rate": 2.7306795041040502e-08,
      "loss": 0.5068,
      "num_input_tokens_seen": 24626880,
      "step": 37650
    },
    {
      "epoch": 19.7353249475891,
      "grad_norm": 0.3309531807899475,
      "learning_rate": 2.677512481276401e-08,
      "loss": 0.5578,
      "num_input_tokens_seen": 24629888,
      "step": 37655
    },
    {
      "epoch": 19.737945492662472,
      "grad_norm": 0.3903804123401642,
      "learning_rate": 2.6248678698220785e-08,
      "loss": 0.361,
      "num_input_tokens_seen": 24633408,
      "step": 37660
    },
    {
      "epoch": 19.74056603773585,
      "grad_norm": 0.3390161991119385,
      "learning_rate": 2.5727456807542183e-08,
      "loss": 0.5313,
      "num_input_tokens_seen": 24636096,
      "step": 37665
    },
    {
      "epoch": 19.743186582809223,
      "grad_norm": 0.43527644872665405,
      "learning_rate": 2.5211459249754875e-08,
      "loss": 0.4393,
      "num_input_tokens_seen": 24639872,
      "step": 37670
    },
    {
      "epoch": 19.7458071278826,
      "grad_norm": 0.7521952390670776,
      "learning_rate": 2.4700686132803076e-08,
      "loss": 0.3972,
      "num_input_tokens_seen": 24642816,
      "step": 37675
    },
    {
      "epoch": 19.748427672955973,
      "grad_norm": 0.7442576289176941,
      "learning_rate": 2.4195137563534643e-08,
      "loss": 0.356,
      "num_input_tokens_seen": 24645312,
      "step": 37680
    },
    {
      "epoch": 19.75104821802935,
      "grad_norm": 0.6534457206726074,
      "learning_rate": 2.369481364770387e-08,
      "loss": 0.5479,
      "num_input_tokens_seen": 24648416,
      "step": 37685
    },
    {
      "epoch": 19.753668763102727,
      "grad_norm": 0.737277626991272,
      "learning_rate": 2.3199714489974266e-08,
      "loss": 0.4444,
      "num_input_tokens_seen": 24651808,
      "step": 37690
    },
    {
      "epoch": 19.7562893081761,
      "grad_norm": 0.6728066802024841,
      "learning_rate": 2.2709840193912978e-08,
      "loss": 0.3974,
      "num_input_tokens_seen": 24655232,
      "step": 37695
    },
    {
      "epoch": 19.758909853249477,
      "grad_norm": 0.43136274814605713,
      "learning_rate": 2.2225190861999146e-08,
      "loss": 0.8525,
      "num_input_tokens_seen": 24658208,
      "step": 37700
    },
    {
      "epoch": 19.76153039832285,
      "grad_norm": 0.41020360589027405,
      "learning_rate": 2.1745766595612783e-08,
      "loss": 0.4997,
      "num_input_tokens_seen": 24661536,
      "step": 37705
    },
    {
      "epoch": 19.764150943396228,
      "grad_norm": 0.3393065631389618,
      "learning_rate": 2.1271567495045886e-08,
      "loss": 0.4736,
      "num_input_tokens_seen": 24664832,
      "step": 37710
    },
    {
      "epoch": 19.7667714884696,
      "grad_norm": 0.39298754930496216,
      "learning_rate": 2.0802593659491333e-08,
      "loss": 0.4546,
      "num_input_tokens_seen": 24667872,
      "step": 37715
    },
    {
      "epoch": 19.769392033542978,
      "grad_norm": 0.33149099349975586,
      "learning_rate": 2.0338845187059528e-08,
      "loss": 0.3681,
      "num_input_tokens_seen": 24670976,
      "step": 37720
    },
    {
      "epoch": 19.77201257861635,
      "grad_norm": 0.3836200535297394,
      "learning_rate": 1.988032217475899e-08,
      "loss": 0.2281,
      "num_input_tokens_seen": 24677568,
      "step": 37725
    },
    {
      "epoch": 19.77463312368973,
      "grad_norm": 0.43680307269096375,
      "learning_rate": 1.942702471850466e-08,
      "loss": 0.4212,
      "num_input_tokens_seen": 24680768,
      "step": 37730
    },
    {
      "epoch": 19.7772536687631,
      "grad_norm": 0.43534910678863525,
      "learning_rate": 1.897895291312346e-08,
      "loss": 0.4854,
      "num_input_tokens_seen": 24684032,
      "step": 37735
    },
    {
      "epoch": 19.77987421383648,
      "grad_norm": 0.5338147878646851,
      "learning_rate": 1.8536106852348746e-08,
      "loss": 0.549,
      "num_input_tokens_seen": 24686976,
      "step": 37740
    },
    {
      "epoch": 19.782494758909852,
      "grad_norm": 0.42730727791786194,
      "learning_rate": 1.8098486628814748e-08,
      "loss": 0.3716,
      "num_input_tokens_seen": 24690560,
      "step": 37745
    },
    {
      "epoch": 19.78511530398323,
      "grad_norm": 0.6404996514320374,
      "learning_rate": 1.7666092334067684e-08,
      "loss": 0.4198,
      "num_input_tokens_seen": 24693344,
      "step": 37750
    },
    {
      "epoch": 19.787735849056602,
      "grad_norm": 0.6151076555252075,
      "learning_rate": 1.7238924058562977e-08,
      "loss": 0.4256,
      "num_input_tokens_seen": 24696160,
      "step": 37755
    },
    {
      "epoch": 19.79035639412998,
      "grad_norm": 0.48875802755355835,
      "learning_rate": 1.6816981891656924e-08,
      "loss": 0.4344,
      "num_input_tokens_seen": 24699232,
      "step": 37760
    },
    {
      "epoch": 19.792976939203353,
      "grad_norm": 0.3807647228240967,
      "learning_rate": 1.6400265921615032e-08,
      "loss": 0.3842,
      "num_input_tokens_seen": 24703104,
      "step": 37765
    },
    {
      "epoch": 19.79559748427673,
      "grad_norm": 0.4004932641983032,
      "learning_rate": 1.598877623560924e-08,
      "loss": 0.3831,
      "num_input_tokens_seen": 24706816,
      "step": 37770
    },
    {
      "epoch": 19.798218029350103,
      "grad_norm": 0.5355532169342041,
      "learning_rate": 1.5582512919717908e-08,
      "loss": 0.4829,
      "num_input_tokens_seen": 24709696,
      "step": 37775
    },
    {
      "epoch": 19.80083857442348,
      "grad_norm": 0.6291120648384094,
      "learning_rate": 1.5181476058928613e-08,
      "loss": 0.5272,
      "num_input_tokens_seen": 24713696,
      "step": 37780
    },
    {
      "epoch": 19.803459119496857,
      "grad_norm": 0.30253103375434875,
      "learning_rate": 1.4785665737129805e-08,
      "loss": 0.4084,
      "num_input_tokens_seen": 24716768,
      "step": 37785
    },
    {
      "epoch": 19.80607966457023,
      "grad_norm": 0.4499581456184387,
      "learning_rate": 1.4395082037127471e-08,
      "loss": 0.406,
      "num_input_tokens_seen": 24720448,
      "step": 37790
    },
    {
      "epoch": 19.808700209643607,
      "grad_norm": 0.368939608335495,
      "learning_rate": 1.4009725040620148e-08,
      "loss": 0.407,
      "num_input_tokens_seen": 24723680,
      "step": 37795
    },
    {
      "epoch": 19.81132075471698,
      "grad_norm": 0.42664262652397156,
      "learning_rate": 1.3629594828223902e-08,
      "loss": 0.4353,
      "num_input_tokens_seen": 24726848,
      "step": 37800
    },
    {
      "epoch": 19.813941299790358,
      "grad_norm": 0.3655843138694763,
      "learning_rate": 1.3254691479458458e-08,
      "loss": 0.4587,
      "num_input_tokens_seen": 24730016,
      "step": 37805
    },
    {
      "epoch": 19.81656184486373,
      "grad_norm": 0.7151139378547668,
      "learning_rate": 1.2885015072747197e-08,
      "loss": 0.5497,
      "num_input_tokens_seen": 24732480,
      "step": 37810
    },
    {
      "epoch": 19.819182389937108,
      "grad_norm": 0.5463820099830627,
      "learning_rate": 1.2520565685419927e-08,
      "loss": 0.5174,
      "num_input_tokens_seen": 24735712,
      "step": 37815
    },
    {
      "epoch": 19.82180293501048,
      "grad_norm": 0.5108484625816345,
      "learning_rate": 1.2161343393721214e-08,
      "loss": 0.4136,
      "num_input_tokens_seen": 24739008,
      "step": 37820
    },
    {
      "epoch": 19.82442348008386,
      "grad_norm": 0.4692295491695404,
      "learning_rate": 1.1807348272790952e-08,
      "loss": 0.5309,
      "num_input_tokens_seen": 24742432,
      "step": 37825
    },
    {
      "epoch": 19.82704402515723,
      "grad_norm": 0.6199067234992981,
      "learning_rate": 1.1458580396683794e-08,
      "loss": 0.448,
      "num_input_tokens_seen": 24745632,
      "step": 37830
    },
    {
      "epoch": 19.82966457023061,
      "grad_norm": 0.3107965886592865,
      "learning_rate": 1.1115039838358044e-08,
      "loss": 0.3141,
      "num_input_tokens_seen": 24748960,
      "step": 37835
    },
    {
      "epoch": 19.832285115303982,
      "grad_norm": 0.43121734261512756,
      "learning_rate": 1.0776726669678439e-08,
      "loss": 0.6287,
      "num_input_tokens_seen": 24751872,
      "step": 37840
    },
    {
      "epoch": 19.83490566037736,
      "grad_norm": 0.436054527759552,
      "learning_rate": 1.044364096141337e-08,
      "loss": 0.4791,
      "num_input_tokens_seen": 24759232,
      "step": 37845
    },
    {
      "epoch": 19.837526205450732,
      "grad_norm": 0.44911983609199524,
      "learning_rate": 1.0115782783243211e-08,
      "loss": 0.5423,
      "num_input_tokens_seen": 24762528,
      "step": 37850
    },
    {
      "epoch": 19.84014675052411,
      "grad_norm": 0.694342315196991,
      "learning_rate": 9.793152203751988e-09,
      "loss": 0.4671,
      "num_input_tokens_seen": 24765344,
      "step": 37855
    },
    {
      "epoch": 19.842767295597483,
      "grad_norm": 0.7395790219306946,
      "learning_rate": 9.475749290430158e-09,
      "loss": 0.5153,
      "num_input_tokens_seen": 24768512,
      "step": 37860
    },
    {
      "epoch": 19.84538784067086,
      "grad_norm": 0.4463697075843811,
      "learning_rate": 9.16357410967461e-09,
      "loss": 0.3324,
      "num_input_tokens_seen": 24771776,
      "step": 37865
    },
    {
      "epoch": 19.848008385744233,
      "grad_norm": 0.306222140789032,
      "learning_rate": 8.856626726788664e-09,
      "loss": 0.3825,
      "num_input_tokens_seen": 24774656,
      "step": 37870
    },
    {
      "epoch": 19.85062893081761,
      "grad_norm": 0.5278943777084351,
      "learning_rate": 8.554907205982066e-09,
      "loss": 0.4231,
      "num_input_tokens_seen": 24777568,
      "step": 37875
    },
    {
      "epoch": 19.853249475890987,
      "grad_norm": 0.35801225900650024,
      "learning_rate": 8.258415610373772e-09,
      "loss": 0.3107,
      "num_input_tokens_seen": 24780864,
      "step": 37880
    },
    {
      "epoch": 19.85587002096436,
      "grad_norm": 0.43264833092689514,
      "learning_rate": 7.967152001980838e-09,
      "loss": 0.3606,
      "num_input_tokens_seen": 24784256,
      "step": 37885
    },
    {
      "epoch": 19.858490566037737,
      "grad_norm": 0.8847585916519165,
      "learning_rate": 7.681116441737856e-09,
      "loss": 0.4603,
      "num_input_tokens_seen": 24787296,
      "step": 37890
    },
    {
      "epoch": 19.86111111111111,
      "grad_norm": 0.7084069848060608,
      "learning_rate": 7.4003089894747464e-09,
      "loss": 0.4697,
      "num_input_tokens_seen": 24792320,
      "step": 37895
    },
    {
      "epoch": 19.863731656184488,
      "grad_norm": 0.4851946532726288,
      "learning_rate": 7.12472970393896e-09,
      "loss": 0.3557,
      "num_input_tokens_seen": 24795072,
      "step": 37900
    },
    {
      "epoch": 19.86635220125786,
      "grad_norm": 0.2706320285797119,
      "learning_rate": 6.854378642776049e-09,
      "loss": 0.5307,
      "num_input_tokens_seen": 24798016,
      "step": 37905
    },
    {
      "epoch": 19.868972746331238,
      "grad_norm": 0.9654580950737,
      "learning_rate": 6.589255862538002e-09,
      "loss": 0.4849,
      "num_input_tokens_seen": 24800544,
      "step": 37910
    },
    {
      "epoch": 19.87159329140461,
      "grad_norm": 0.47930386662483215,
      "learning_rate": 6.32936141869156e-09,
      "loss": 0.4809,
      "num_input_tokens_seen": 24803872,
      "step": 37915
    },
    {
      "epoch": 19.87421383647799,
      "grad_norm": 0.6151717305183411,
      "learning_rate": 6.074695365596017e-09,
      "loss": 0.4666,
      "num_input_tokens_seen": 24806464,
      "step": 37920
    },
    {
      "epoch": 19.87683438155136,
      "grad_norm": 0.3478917181491852,
      "learning_rate": 5.8252577565309775e-09,
      "loss": 0.4322,
      "num_input_tokens_seen": 24814080,
      "step": 37925
    },
    {
      "epoch": 19.87945492662474,
      "grad_norm": 0.5828825831413269,
      "learning_rate": 5.581048643671371e-09,
      "loss": 0.4693,
      "num_input_tokens_seen": 24817248,
      "step": 37930
    },
    {
      "epoch": 19.882075471698112,
      "grad_norm": 0.5662754774093628,
      "learning_rate": 5.342068078106888e-09,
      "loss": 0.3075,
      "num_input_tokens_seen": 24820480,
      "step": 37935
    },
    {
      "epoch": 19.88469601677149,
      "grad_norm": 0.556885838508606,
      "learning_rate": 5.1083161098253175e-09,
      "loss": 0.4287,
      "num_input_tokens_seen": 24823200,
      "step": 37940
    },
    {
      "epoch": 19.887316561844862,
      "grad_norm": 0.6139037013053894,
      "learning_rate": 4.87979278772921e-09,
      "loss": 0.3767,
      "num_input_tokens_seen": 24826112,
      "step": 37945
    },
    {
      "epoch": 19.88993710691824,
      "grad_norm": 0.6184678077697754,
      "learning_rate": 4.656498159619216e-09,
      "loss": 0.4473,
      "num_input_tokens_seen": 24828576,
      "step": 37950
    },
    {
      "epoch": 19.892557651991613,
      "grad_norm": 0.3788638412952423,
      "learning_rate": 4.438432272210746e-09,
      "loss": 0.3315,
      "num_input_tokens_seen": 24832672,
      "step": 37955
    },
    {
      "epoch": 19.89517819706499,
      "grad_norm": 0.39048704504966736,
      "learning_rate": 4.225595171114538e-09,
      "loss": 0.4523,
      "num_input_tokens_seen": 24835616,
      "step": 37960
    },
    {
      "epoch": 19.897798742138363,
      "grad_norm": 0.3400722146034241,
      "learning_rate": 4.017986900858861e-09,
      "loss": 0.3921,
      "num_input_tokens_seen": 24839328,
      "step": 37965
    },
    {
      "epoch": 19.90041928721174,
      "grad_norm": 0.3751511871814728,
      "learning_rate": 3.815607504870089e-09,
      "loss": 0.3565,
      "num_input_tokens_seen": 24842176,
      "step": 37970
    },
    {
      "epoch": 19.903039832285117,
      "grad_norm": 0.5089302062988281,
      "learning_rate": 3.618457025486577e-09,
      "loss": 0.6305,
      "num_input_tokens_seen": 24845088,
      "step": 37975
    },
    {
      "epoch": 19.90566037735849,
      "grad_norm": 0.41269853711128235,
      "learning_rate": 3.426535503947559e-09,
      "loss": 0.389,
      "num_input_tokens_seen": 24848032,
      "step": 37980
    },
    {
      "epoch": 19.908280922431867,
      "grad_norm": 0.35669025778770447,
      "learning_rate": 3.2398429803986997e-09,
      "loss": 0.481,
      "num_input_tokens_seen": 24851584,
      "step": 37985
    },
    {
      "epoch": 19.91090146750524,
      "grad_norm": 0.489702433347702,
      "learning_rate": 3.058379493900421e-09,
      "loss": 0.4883,
      "num_input_tokens_seen": 24854464,
      "step": 37990
    },
    {
      "epoch": 19.913522012578618,
      "grad_norm": 0.8441889882087708,
      "learning_rate": 2.882145082405696e-09,
      "loss": 0.4539,
      "num_input_tokens_seen": 24857152,
      "step": 37995
    },
    {
      "epoch": 19.91614255765199,
      "grad_norm": 0.4467777609825134,
      "learning_rate": 2.711139782785033e-09,
      "loss": 0.4841,
      "num_input_tokens_seen": 24860480,
      "step": 38000
    },
    {
      "epoch": 19.918763102725368,
      "grad_norm": 0.3375874161720276,
      "learning_rate": 2.5453636308098163e-09,
      "loss": 0.4375,
      "num_input_tokens_seen": 24863680,
      "step": 38005
    },
    {
      "epoch": 19.92138364779874,
      "grad_norm": 0.4278984069824219,
      "learning_rate": 2.3848166611606383e-09,
      "loss": 0.4825,
      "num_input_tokens_seen": 24866976,
      "step": 38010
    },
    {
      "epoch": 19.92400419287212,
      "grad_norm": 0.6491400599479675,
      "learning_rate": 2.2294989074161943e-09,
      "loss": 0.4309,
      "num_input_tokens_seen": 24870240,
      "step": 38015
    },
    {
      "epoch": 19.92662473794549,
      "grad_norm": 0.5893769264221191,
      "learning_rate": 2.0794104020727125e-09,
      "loss": 0.6158,
      "num_input_tokens_seen": 24874400,
      "step": 38020
    },
    {
      "epoch": 19.92924528301887,
      "grad_norm": 1.118671178817749,
      "learning_rate": 1.9345511765272993e-09,
      "loss": 0.4865,
      "num_input_tokens_seen": 24878368,
      "step": 38025
    },
    {
      "epoch": 19.931865828092242,
      "grad_norm": 0.9429481625556946,
      "learning_rate": 1.794921261077942e-09,
      "loss": 0.357,
      "num_input_tokens_seen": 24881184,
      "step": 38030
    },
    {
      "epoch": 19.93448637316562,
      "grad_norm": 0.3559812605381012,
      "learning_rate": 1.6605206849373833e-09,
      "loss": 0.3929,
      "num_input_tokens_seen": 24884704,
      "step": 38035
    },
    {
      "epoch": 19.937106918238992,
      "grad_norm": 0.9073328375816345,
      "learning_rate": 1.531349476219246e-09,
      "loss": 0.6,
      "num_input_tokens_seen": 24887456,
      "step": 38040
    },
    {
      "epoch": 19.93972746331237,
      "grad_norm": 0.32540029287338257,
      "learning_rate": 1.4074076619463584e-09,
      "loss": 0.4109,
      "num_input_tokens_seen": 24890592,
      "step": 38045
    },
    {
      "epoch": 19.942348008385743,
      "grad_norm": 0.43301182985305786,
      "learning_rate": 1.2886952680452036e-09,
      "loss": 0.4513,
      "num_input_tokens_seen": 24893856,
      "step": 38050
    },
    {
      "epoch": 19.94496855345912,
      "grad_norm": 0.49628084897994995,
      "learning_rate": 1.1752123193459197e-09,
      "loss": 0.2896,
      "num_input_tokens_seen": 24896544,
      "step": 38055
    },
    {
      "epoch": 19.947589098532493,
      "grad_norm": 0.48270854353904724,
      "learning_rate": 1.0669588395934017e-09,
      "loss": 0.3627,
      "num_input_tokens_seen": 24898848,
      "step": 38060
    },
    {
      "epoch": 19.95020964360587,
      "grad_norm": 0.39884674549102783,
      "learning_rate": 9.639348514278722e-10,
      "loss": 0.3229,
      "num_input_tokens_seen": 24901472,
      "step": 38065
    },
    {
      "epoch": 19.952830188679247,
      "grad_norm": 0.32514774799346924,
      "learning_rate": 8.661403764043119e-10,
      "loss": 0.3999,
      "num_input_tokens_seen": 24904768,
      "step": 38070
    },
    {
      "epoch": 19.95545073375262,
      "grad_norm": 0.4655981957912445,
      "learning_rate": 7.735754349785795e-10,
      "loss": 0.4709,
      "num_input_tokens_seen": 24908512,
      "step": 38075
    },
    {
      "epoch": 19.958071278825997,
      "grad_norm": 0.511079728603363,
      "learning_rate": 6.862400465157403e-10,
      "loss": 0.5359,
      "num_input_tokens_seen": 24911488,
      "step": 38080
    },
    {
      "epoch": 19.96069182389937,
      "grad_norm": 0.36486631631851196,
      "learning_rate": 6.041342292817387e-10,
      "loss": 0.3978,
      "num_input_tokens_seen": 24914624,
      "step": 38085
    },
    {
      "epoch": 19.963312368972748,
      "grad_norm": 0.44164371490478516,
      "learning_rate": 5.272580004572758e-10,
      "loss": 0.4801,
      "num_input_tokens_seen": 24917632,
      "step": 38090
    },
    {
      "epoch": 19.96593291404612,
      "grad_norm": 0.5170221328735352,
      "learning_rate": 4.556113761211567e-10,
      "loss": 0.459,
      "num_input_tokens_seen": 24920768,
      "step": 38095
    },
    {
      "epoch": 19.968553459119498,
      "grad_norm": 0.47847339510917664,
      "learning_rate": 3.8919437125861657e-10,
      "loss": 0.6119,
      "num_input_tokens_seen": 24923936,
      "step": 38100
    },
    {
      "epoch": 19.97117400419287,
      "grad_norm": 0.5376165509223938,
      "learning_rate": 3.280069997696478e-10,
      "loss": 0.354,
      "num_input_tokens_seen": 24927424,
      "step": 38105
    },
    {
      "epoch": 19.97379454926625,
      "grad_norm": 0.6171088814735413,
      "learning_rate": 2.720492744467951e-10,
      "loss": 0.5323,
      "num_input_tokens_seen": 24930080,
      "step": 38110
    },
    {
      "epoch": 19.97641509433962,
      "grad_norm": 0.6289569735527039,
      "learning_rate": 2.2132120700291136e-10,
      "loss": 0.4419,
      "num_input_tokens_seen": 24933664,
      "step": 38115
    },
    {
      "epoch": 19.979035639413,
      "grad_norm": 0.5105796456336975,
      "learning_rate": 1.7582280804340213e-10,
      "loss": 0.4462,
      "num_input_tokens_seen": 24936800,
      "step": 38120
    },
    {
      "epoch": 19.981656184486372,
      "grad_norm": 1.6682298183441162,
      "learning_rate": 1.3555408708842977e-10,
      "loss": 0.5421,
      "num_input_tokens_seen": 24940352,
      "step": 38125
    },
    {
      "epoch": 19.98427672955975,
      "grad_norm": 0.43555259704589844,
      "learning_rate": 1.0051505256458704e-10,
      "loss": 0.4668,
      "num_input_tokens_seen": 24943808,
      "step": 38130
    },
    {
      "epoch": 19.986897274633122,
      "grad_norm": 0.4441698491573334,
      "learning_rate": 7.070571179657037e-11,
      "loss": 0.4371,
      "num_input_tokens_seen": 24946304,
      "step": 38135
    },
    {
      "epoch": 19.9895178197065,
      "grad_norm": 0.4403577446937561,
      "learning_rate": 4.612607102383315e-11,
      "loss": 0.4312,
      "num_input_tokens_seen": 24949440,
      "step": 38140
    },
    {
      "epoch": 19.992138364779873,
      "grad_norm": 0.5543020963668823,
      "learning_rate": 2.677613538670798e-11,
      "loss": 0.3517,
      "num_input_tokens_seen": 24953376,
      "step": 38145
    },
    {
      "epoch": 19.99475890985325,
      "grad_norm": 0.2862715423107147,
      "learning_rate": 1.2655908931957782e-11,
      "loss": 0.4133,
      "num_input_tokens_seen": 24957984,
      "step": 38150
    },
    {
      "epoch": 19.997379454926623,
      "grad_norm": 0.5771775841712952,
      "learning_rate": 3.765394615551366e-12,
      "loss": 0.4096,
      "num_input_tokens_seen": 24960832,
      "step": 38155
    },
    {
      "epoch": 20.0,
      "grad_norm": 1.5662710666656494,
      "learning_rate": 1.0459429711229262e-13,
      "loss": 0.6148,
      "num_input_tokens_seen": 24964664,
      "step": 38160
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.47683560848236084,
      "eval_runtime": 9.6169,
      "eval_samples_per_second": 88.178,
      "eval_steps_per_second": 22.045,
      "num_input_tokens_seen": 24964664,
      "step": 38160
    },
    {
      "epoch": 20.0,
      "num_input_tokens_seen": 24964664,
      "step": 38160,
      "total_flos": 1.1241874162436014e+18,
      "train_loss": 0.4822524391473964,
      "train_runtime": 6475.715,
      "train_samples_per_second": 23.562,
      "train_steps_per_second": 5.893
    }
  ],
  "logging_steps": 5,
  "max_steps": 38160,
  "num_input_tokens_seen": 24964664,
  "num_train_epochs": 20,
  "save_steps": 1908,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1241874162436014e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}