{
  "best_global_step": 17172,
  "best_metric": 0.45358505845069885,
  "best_model_checkpoint": "saves_multiple/p-tuning/llama-3-8b-instruct/train_codealpacapy_42_1760664611/checkpoint-17172",
  "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": 36.37511444091797,
      "learning_rate": 1.0482180293501048e-06,
      "loss": 4.3366,
      "num_input_tokens_seen": 3264,
      "step": 5
    },
    {
      "epoch": 0.005241090146750524,
      "grad_norm": 47.15357208251953,
      "learning_rate": 2.358490566037736e-06,
      "loss": 4.3961,
      "num_input_tokens_seen": 5888,
      "step": 10
    },
    {
      "epoch": 0.007861635220125786,
      "grad_norm": 54.218666076660156,
      "learning_rate": 3.668763102725367e-06,
      "loss": 4.3377,
      "num_input_tokens_seen": 8480,
      "step": 15
    },
    {
      "epoch": 0.010482180293501049,
      "grad_norm": 44.877288818359375,
      "learning_rate": 4.979035639412998e-06,
      "loss": 3.2407,
      "num_input_tokens_seen": 11488,
      "step": 20
    },
    {
      "epoch": 0.01310272536687631,
      "grad_norm": 32.66029739379883,
      "learning_rate": 6.289308176100629e-06,
      "loss": 2.08,
      "num_input_tokens_seen": 14880,
      "step": 25
    },
    {
      "epoch": 0.015723270440251572,
      "grad_norm": 10.666071891784668,
      "learning_rate": 7.59958071278826e-06,
      "loss": 1.0257,
      "num_input_tokens_seen": 18592,
      "step": 30
    },
    {
      "epoch": 0.018343815513626835,
      "grad_norm": 15.194086074829102,
      "learning_rate": 8.90985324947589e-06,
      "loss": 0.8789,
      "num_input_tokens_seen": 21568,
      "step": 35
    },
    {
      "epoch": 0.020964360587002098,
      "grad_norm": 5.819694995880127,
      "learning_rate": 1.0220125786163522e-05,
      "loss": 0.7281,
      "num_input_tokens_seen": 24160,
      "step": 40
    },
    {
      "epoch": 0.02358490566037736,
      "grad_norm": 7.504000663757324,
      "learning_rate": 1.1530398322851153e-05,
      "loss": 0.7915,
      "num_input_tokens_seen": 27168,
      "step": 45
    },
    {
      "epoch": 0.02620545073375262,
      "grad_norm": 3.151390314102173,
      "learning_rate": 1.2840670859538784e-05,
      "loss": 0.6498,
      "num_input_tokens_seen": 30080,
      "step": 50
    },
    {
      "epoch": 0.028825995807127882,
      "grad_norm": 4.705784320831299,
      "learning_rate": 1.4150943396226415e-05,
      "loss": 0.7485,
      "num_input_tokens_seen": 32608,
      "step": 55
    },
    {
      "epoch": 0.031446540880503145,
      "grad_norm": 1.960841417312622,
      "learning_rate": 1.5461215932914046e-05,
      "loss": 0.7152,
      "num_input_tokens_seen": 36448,
      "step": 60
    },
    {
      "epoch": 0.034067085953878404,
      "grad_norm": 6.21964168548584,
      "learning_rate": 1.6771488469601677e-05,
      "loss": 0.4345,
      "num_input_tokens_seen": 38944,
      "step": 65
    },
    {
      "epoch": 0.03668763102725367,
      "grad_norm": 9.175660133361816,
      "learning_rate": 1.8081761006289308e-05,
      "loss": 0.5775,
      "num_input_tokens_seen": 41824,
      "step": 70
    },
    {
      "epoch": 0.03930817610062893,
      "grad_norm": 1.8137201070785522,
      "learning_rate": 1.9392033542976942e-05,
      "loss": 0.5619,
      "num_input_tokens_seen": 44896,
      "step": 75
    },
    {
      "epoch": 0.041928721174004195,
      "grad_norm": 2.437303304672241,
      "learning_rate": 2.070230607966457e-05,
      "loss": 0.5432,
      "num_input_tokens_seen": 47616,
      "step": 80
    },
    {
      "epoch": 0.044549266247379454,
      "grad_norm": 3.2654027938842773,
      "learning_rate": 2.20125786163522e-05,
      "loss": 0.6556,
      "num_input_tokens_seen": 50368,
      "step": 85
    },
    {
      "epoch": 0.04716981132075472,
      "grad_norm": 2.930898904800415,
      "learning_rate": 2.3322851153039832e-05,
      "loss": 0.6779,
      "num_input_tokens_seen": 54016,
      "step": 90
    },
    {
      "epoch": 0.04979035639412998,
      "grad_norm": 1.0492274761199951,
      "learning_rate": 2.4633123689727463e-05,
      "loss": 0.4229,
      "num_input_tokens_seen": 57888,
      "step": 95
    },
    {
      "epoch": 0.05241090146750524,
      "grad_norm": 1.727480173110962,
      "learning_rate": 2.5943396226415097e-05,
      "loss": 0.5745,
      "num_input_tokens_seen": 61024,
      "step": 100
    },
    {
      "epoch": 0.055031446540880505,
      "grad_norm": 1.9369148015975952,
      "learning_rate": 2.7253668763102725e-05,
      "loss": 0.5809,
      "num_input_tokens_seen": 64672,
      "step": 105
    },
    {
      "epoch": 0.057651991614255764,
      "grad_norm": 2.284978151321411,
      "learning_rate": 2.8563941299790356e-05,
      "loss": 0.5085,
      "num_input_tokens_seen": 67616,
      "step": 110
    },
    {
      "epoch": 0.06027253668763103,
      "grad_norm": 1.127700686454773,
      "learning_rate": 2.987421383647799e-05,
      "loss": 0.521,
      "num_input_tokens_seen": 71232,
      "step": 115
    },
    {
      "epoch": 0.06289308176100629,
      "grad_norm": 4.249385833740234,
      "learning_rate": 3.118448637316562e-05,
      "loss": 0.4112,
      "num_input_tokens_seen": 74880,
      "step": 120
    },
    {
      "epoch": 0.06551362683438156,
      "grad_norm": 6.756195068359375,
      "learning_rate": 3.2494758909853245e-05,
      "loss": 0.5647,
      "num_input_tokens_seen": 78432,
      "step": 125
    },
    {
      "epoch": 0.06813417190775681,
      "grad_norm": 1.5354418754577637,
      "learning_rate": 3.380503144654088e-05,
      "loss": 0.4921,
      "num_input_tokens_seen": 81088,
      "step": 130
    },
    {
      "epoch": 0.07075471698113207,
      "grad_norm": 0.7773059606552124,
      "learning_rate": 3.5115303983228514e-05,
      "loss": 0.5339,
      "num_input_tokens_seen": 84896,
      "step": 135
    },
    {
      "epoch": 0.07337526205450734,
      "grad_norm": 1.359603762626648,
      "learning_rate": 3.642557651991614e-05,
      "loss": 0.5942,
      "num_input_tokens_seen": 88736,
      "step": 140
    },
    {
      "epoch": 0.0759958071278826,
      "grad_norm": 3.1003196239471436,
      "learning_rate": 3.7735849056603776e-05,
      "loss": 0.8808,
      "num_input_tokens_seen": 92224,
      "step": 145
    },
    {
      "epoch": 0.07861635220125786,
      "grad_norm": 2.669731855392456,
      "learning_rate": 3.90461215932914e-05,
      "loss": 0.5362,
      "num_input_tokens_seen": 95168,
      "step": 150
    },
    {
      "epoch": 0.08123689727463312,
      "grad_norm": 1.6232050657272339,
      "learning_rate": 4.035639412997904e-05,
      "loss": 0.5562,
      "num_input_tokens_seen": 98976,
      "step": 155
    },
    {
      "epoch": 0.08385744234800839,
      "grad_norm": 1.8857656717300415,
      "learning_rate": 4.1666666666666665e-05,
      "loss": 0.469,
      "num_input_tokens_seen": 102176,
      "step": 160
    },
    {
      "epoch": 0.08647798742138364,
      "grad_norm": 1.2677682638168335,
      "learning_rate": 4.29769392033543e-05,
      "loss": 0.464,
      "num_input_tokens_seen": 105824,
      "step": 165
    },
    {
      "epoch": 0.08909853249475891,
      "grad_norm": 1.5558921098709106,
      "learning_rate": 4.4287211740041934e-05,
      "loss": 0.5105,
      "num_input_tokens_seen": 108512,
      "step": 170
    },
    {
      "epoch": 0.09171907756813417,
      "grad_norm": 1.2208294868469238,
      "learning_rate": 4.5597484276729555e-05,
      "loss": 0.523,
      "num_input_tokens_seen": 111872,
      "step": 175
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 1.3901151418685913,
      "learning_rate": 4.690775681341719e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 115552,
      "step": 180
    },
    {
      "epoch": 0.09696016771488469,
      "grad_norm": 1.0770810842514038,
      "learning_rate": 4.8218029350104823e-05,
      "loss": 0.6313,
      "num_input_tokens_seen": 118784,
      "step": 185
    },
    {
      "epoch": 0.09958071278825996,
      "grad_norm": 2.5382447242736816,
      "learning_rate": 4.952830188679246e-05,
      "loss": 0.4303,
      "num_input_tokens_seen": 122528,
      "step": 190
    },
    {
      "epoch": 0.10220125786163523,
      "grad_norm": 1.802834153175354,
      "learning_rate": 5.0838574423480085e-05,
      "loss": 0.5259,
      "num_input_tokens_seen": 125024,
      "step": 195
    },
    {
      "epoch": 0.10482180293501048,
      "grad_norm": 1.0965396165847778,
      "learning_rate": 5.214884696016771e-05,
      "loss": 0.5996,
      "num_input_tokens_seen": 128096,
      "step": 200
    },
    {
      "epoch": 0.10744234800838574,
      "grad_norm": 2.107889413833618,
      "learning_rate": 5.345911949685535e-05,
      "loss": 0.6745,
      "num_input_tokens_seen": 130496,
      "step": 205
    },
    {
      "epoch": 0.11006289308176101,
      "grad_norm": 1.0323224067687988,
      "learning_rate": 5.4769392033542975e-05,
      "loss": 0.5892,
      "num_input_tokens_seen": 133152,
      "step": 210
    },
    {
      "epoch": 0.11268343815513626,
      "grad_norm": 4.541623592376709,
      "learning_rate": 5.607966457023061e-05,
      "loss": 0.6939,
      "num_input_tokens_seen": 136384,
      "step": 215
    },
    {
      "epoch": 0.11530398322851153,
      "grad_norm": 1.3010039329528809,
      "learning_rate": 5.7389937106918244e-05,
      "loss": 0.6202,
      "num_input_tokens_seen": 139456,
      "step": 220
    },
    {
      "epoch": 0.1179245283018868,
      "grad_norm": 4.006557464599609,
      "learning_rate": 5.870020964360587e-05,
      "loss": 0.626,
      "num_input_tokens_seen": 142816,
      "step": 225
    },
    {
      "epoch": 0.12054507337526206,
      "grad_norm": 1.498618245124817,
      "learning_rate": 6.00104821802935e-05,
      "loss": 0.5322,
      "num_input_tokens_seen": 145696,
      "step": 230
    },
    {
      "epoch": 0.12316561844863731,
      "grad_norm": 0.944551408290863,
      "learning_rate": 6.132075471698113e-05,
      "loss": 0.6419,
      "num_input_tokens_seen": 149024,
      "step": 235
    },
    {
      "epoch": 0.12578616352201258,
      "grad_norm": 1.9880057573318481,
      "learning_rate": 6.263102725366875e-05,
      "loss": 0.5641,
      "num_input_tokens_seen": 151840,
      "step": 240
    },
    {
      "epoch": 0.12840670859538783,
      "grad_norm": 1.0167557001113892,
      "learning_rate": 6.39412997903564e-05,
      "loss": 0.6886,
      "num_input_tokens_seen": 155456,
      "step": 245
    },
    {
      "epoch": 0.1310272536687631,
      "grad_norm": 1.7375400066375732,
      "learning_rate": 6.525157232704402e-05,
      "loss": 0.5696,
      "num_input_tokens_seen": 158528,
      "step": 250
    },
    {
      "epoch": 0.13364779874213836,
      "grad_norm": 2.5441417694091797,
      "learning_rate": 6.656184486373166e-05,
      "loss": 0.5468,
      "num_input_tokens_seen": 166496,
      "step": 255
    },
    {
      "epoch": 0.13626834381551362,
      "grad_norm": 0.8839701414108276,
      "learning_rate": 6.787211740041929e-05,
      "loss": 0.5205,
      "num_input_tokens_seen": 169568,
      "step": 260
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 1.2689623832702637,
      "learning_rate": 6.918238993710691e-05,
      "loss": 0.6078,
      "num_input_tokens_seen": 172704,
      "step": 265
    },
    {
      "epoch": 0.14150943396226415,
      "grad_norm": 0.7321489453315735,
      "learning_rate": 7.049266247379455e-05,
      "loss": 0.7384,
      "num_input_tokens_seen": 176480,
      "step": 270
    },
    {
      "epoch": 0.1441299790356394,
      "grad_norm": 0.4416513442993164,
      "learning_rate": 7.180293501048218e-05,
      "loss": 0.5469,
      "num_input_tokens_seen": 181408,
      "step": 275
    },
    {
      "epoch": 0.14675052410901468,
      "grad_norm": 0.8268011212348938,
      "learning_rate": 7.311320754716982e-05,
      "loss": 0.5923,
      "num_input_tokens_seen": 184736,
      "step": 280
    },
    {
      "epoch": 0.14937106918238993,
      "grad_norm": 0.8864881992340088,
      "learning_rate": 7.442348008385745e-05,
      "loss": 0.5122,
      "num_input_tokens_seen": 188352,
      "step": 285
    },
    {
      "epoch": 0.1519916142557652,
      "grad_norm": 5.784894943237305,
      "learning_rate": 7.573375262054507e-05,
      "loss": 0.4922,
      "num_input_tokens_seen": 193408,
      "step": 290
    },
    {
      "epoch": 0.15461215932914046,
      "grad_norm": 0.9055310487747192,
      "learning_rate": 7.70440251572327e-05,
      "loss": 0.4207,
      "num_input_tokens_seen": 196384,
      "step": 295
    },
    {
      "epoch": 0.15723270440251572,
      "grad_norm": 1.1681617498397827,
      "learning_rate": 7.835429769392034e-05,
      "loss": 0.6115,
      "num_input_tokens_seen": 199360,
      "step": 300
    },
    {
      "epoch": 0.159853249475891,
      "grad_norm": 0.4760703444480896,
      "learning_rate": 7.966457023060797e-05,
      "loss": 0.5228,
      "num_input_tokens_seen": 202848,
      "step": 305
    },
    {
      "epoch": 0.16247379454926625,
      "grad_norm": 0.7115180492401123,
      "learning_rate": 8.09748427672956e-05,
      "loss": 0.4791,
      "num_input_tokens_seen": 207104,
      "step": 310
    },
    {
      "epoch": 0.1650943396226415,
      "grad_norm": 0.8621534109115601,
      "learning_rate": 8.228511530398323e-05,
      "loss": 0.637,
      "num_input_tokens_seen": 209600,
      "step": 315
    },
    {
      "epoch": 0.16771488469601678,
      "grad_norm": 0.7642444968223572,
      "learning_rate": 8.359538784067086e-05,
      "loss": 0.611,
      "num_input_tokens_seen": 213152,
      "step": 320
    },
    {
      "epoch": 0.17033542976939203,
      "grad_norm": 1.320351481437683,
      "learning_rate": 8.490566037735848e-05,
      "loss": 0.7027,
      "num_input_tokens_seen": 215616,
      "step": 325
    },
    {
      "epoch": 0.17295597484276728,
      "grad_norm": 0.5799564719200134,
      "learning_rate": 8.621593291404613e-05,
      "loss": 0.6089,
      "num_input_tokens_seen": 218528,
      "step": 330
    },
    {
      "epoch": 0.17557651991614256,
      "grad_norm": 2.252868175506592,
      "learning_rate": 8.752620545073375e-05,
      "loss": 0.4405,
      "num_input_tokens_seen": 221824,
      "step": 335
    },
    {
      "epoch": 0.17819706498951782,
      "grad_norm": 0.41030770540237427,
      "learning_rate": 8.883647798742137e-05,
      "loss": 0.5055,
      "num_input_tokens_seen": 225280,
      "step": 340
    },
    {
      "epoch": 0.18081761006289307,
      "grad_norm": 0.9709662795066833,
      "learning_rate": 9.014675052410902e-05,
      "loss": 0.4572,
      "num_input_tokens_seen": 230912,
      "step": 345
    },
    {
      "epoch": 0.18343815513626835,
      "grad_norm": 0.7299672365188599,
      "learning_rate": 9.145702306079664e-05,
      "loss": 0.5646,
      "num_input_tokens_seen": 234592,
      "step": 350
    },
    {
      "epoch": 0.1860587002096436,
      "grad_norm": 0.5802033543586731,
      "learning_rate": 9.276729559748428e-05,
      "loss": 0.493,
      "num_input_tokens_seen": 237600,
      "step": 355
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 0.5321568250656128,
      "learning_rate": 9.407756813417191e-05,
      "loss": 0.6188,
      "num_input_tokens_seen": 241088,
      "step": 360
    },
    {
      "epoch": 0.19129979035639413,
      "grad_norm": 1.0294125080108643,
      "learning_rate": 9.538784067085953e-05,
      "loss": 0.5569,
      "num_input_tokens_seen": 243520,
      "step": 365
    },
    {
      "epoch": 0.19392033542976939,
      "grad_norm": 2.6575582027435303,
      "learning_rate": 9.669811320754718e-05,
      "loss": 0.742,
      "num_input_tokens_seen": 247456,
      "step": 370
    },
    {
      "epoch": 0.19654088050314467,
      "grad_norm": 0.44868069887161255,
      "learning_rate": 9.80083857442348e-05,
      "loss": 0.4974,
      "num_input_tokens_seen": 250784,
      "step": 375
    },
    {
      "epoch": 0.19916142557651992,
      "grad_norm": 0.7755410671234131,
      "learning_rate": 9.931865828092243e-05,
      "loss": 0.542,
      "num_input_tokens_seen": 255040,
      "step": 380
    },
    {
      "epoch": 0.20178197064989517,
      "grad_norm": 1.2239811420440674,
      "learning_rate": 0.00010062893081761007,
      "loss": 0.6188,
      "num_input_tokens_seen": 258240,
      "step": 385
    },
    {
      "epoch": 0.20440251572327045,
      "grad_norm": 0.3172089457511902,
      "learning_rate": 0.0001019392033542977,
      "loss": 0.4524,
      "num_input_tokens_seen": 261152,
      "step": 390
    },
    {
      "epoch": 0.2070230607966457,
      "grad_norm": 0.5608502626419067,
      "learning_rate": 0.00010324947589098532,
      "loss": 0.5425,
      "num_input_tokens_seen": 264544,
      "step": 395
    },
    {
      "epoch": 0.20964360587002095,
      "grad_norm": 0.8350383639335632,
      "learning_rate": 0.00010455974842767296,
      "loss": 0.6098,
      "num_input_tokens_seen": 267136,
      "step": 400
    },
    {
      "epoch": 0.21226415094339623,
      "grad_norm": 0.9740188121795654,
      "learning_rate": 0.00010587002096436059,
      "loss": 0.4776,
      "num_input_tokens_seen": 270464,
      "step": 405
    },
    {
      "epoch": 0.2148846960167715,
      "grad_norm": 0.8400376439094543,
      "learning_rate": 0.00010718029350104821,
      "loss": 0.5049,
      "num_input_tokens_seen": 274528,
      "step": 410
    },
    {
      "epoch": 0.21750524109014674,
      "grad_norm": 0.6592617034912109,
      "learning_rate": 0.00010849056603773586,
      "loss": 0.4909,
      "num_input_tokens_seen": 277376,
      "step": 415
    },
    {
      "epoch": 0.22012578616352202,
      "grad_norm": 0.889051079750061,
      "learning_rate": 0.00010980083857442348,
      "loss": 0.395,
      "num_input_tokens_seen": 279616,
      "step": 420
    },
    {
      "epoch": 0.22274633123689727,
      "grad_norm": 1.306659460067749,
      "learning_rate": 0.0001111111111111111,
      "loss": 0.4555,
      "num_input_tokens_seen": 282656,
      "step": 425
    },
    {
      "epoch": 0.22536687631027252,
      "grad_norm": 0.8371270298957825,
      "learning_rate": 0.00011242138364779875,
      "loss": 0.5687,
      "num_input_tokens_seen": 285600,
      "step": 430
    },
    {
      "epoch": 0.2279874213836478,
      "grad_norm": 0.3181367516517639,
      "learning_rate": 0.00011373165618448637,
      "loss": 0.395,
      "num_input_tokens_seen": 289568,
      "step": 435
    },
    {
      "epoch": 0.23060796645702306,
      "grad_norm": 0.7335044145584106,
      "learning_rate": 0.00011504192872117402,
      "loss": 0.592,
      "num_input_tokens_seen": 292768,
      "step": 440
    },
    {
      "epoch": 0.23322851153039834,
      "grad_norm": 1.0750073194503784,
      "learning_rate": 0.00011635220125786164,
      "loss": 0.6336,
      "num_input_tokens_seen": 295456,
      "step": 445
    },
    {
      "epoch": 0.2358490566037736,
      "grad_norm": 0.7022571563720703,
      "learning_rate": 0.00011766247379454926,
      "loss": 0.5672,
      "num_input_tokens_seen": 299488,
      "step": 450
    },
    {
      "epoch": 0.23846960167714884,
      "grad_norm": 1.6382282972335815,
      "learning_rate": 0.00011897274633123691,
      "loss": 0.4676,
      "num_input_tokens_seen": 302976,
      "step": 455
    },
    {
      "epoch": 0.24109014675052412,
      "grad_norm": 0.5302943587303162,
      "learning_rate": 0.00012028301886792453,
      "loss": 0.5342,
      "num_input_tokens_seen": 306560,
      "step": 460
    },
    {
      "epoch": 0.24371069182389937,
      "grad_norm": 0.5208105444908142,
      "learning_rate": 0.00012159329140461216,
      "loss": 0.4462,
      "num_input_tokens_seen": 309536,
      "step": 465
    },
    {
      "epoch": 0.24633123689727462,
      "grad_norm": 0.4882085621356964,
      "learning_rate": 0.0001229035639412998,
      "loss": 0.5008,
      "num_input_tokens_seen": 312576,
      "step": 470
    },
    {
      "epoch": 0.2489517819706499,
      "grad_norm": 0.33355027437210083,
      "learning_rate": 0.00012421383647798743,
      "loss": 0.4987,
      "num_input_tokens_seen": 315616,
      "step": 475
    },
    {
      "epoch": 0.25157232704402516,
      "grad_norm": 0.41390642523765564,
      "learning_rate": 0.00012552410901467507,
      "loss": 0.5167,
      "num_input_tokens_seen": 319136,
      "step": 480
    },
    {
      "epoch": 0.25419287211740044,
      "grad_norm": 0.7818740606307983,
      "learning_rate": 0.00012683438155136267,
      "loss": 0.6978,
      "num_input_tokens_seen": 321376,
      "step": 485
    },
    {
      "epoch": 0.25681341719077566,
      "grad_norm": 1.1563949584960938,
      "learning_rate": 0.0001281446540880503,
      "loss": 0.5501,
      "num_input_tokens_seen": 324928,
      "step": 490
    },
    {
      "epoch": 0.25943396226415094,
      "grad_norm": 0.6280922889709473,
      "learning_rate": 0.00012945492662473794,
      "loss": 0.6535,
      "num_input_tokens_seen": 329248,
      "step": 495
    },
    {
      "epoch": 0.2620545073375262,
      "grad_norm": 0.5391634702682495,
      "learning_rate": 0.00013076519916142558,
      "loss": 0.6412,
      "num_input_tokens_seen": 332288,
      "step": 500
    },
    {
      "epoch": 0.26467505241090145,
      "grad_norm": 0.46183428168296814,
      "learning_rate": 0.0001320754716981132,
      "loss": 0.5217,
      "num_input_tokens_seen": 334752,
      "step": 505
    },
    {
      "epoch": 0.2672955974842767,
      "grad_norm": 0.3989781141281128,
      "learning_rate": 0.00013338574423480085,
      "loss": 0.5572,
      "num_input_tokens_seen": 337856,
      "step": 510
    },
    {
      "epoch": 0.269916142557652,
      "grad_norm": 0.32025647163391113,
      "learning_rate": 0.00013469601677148845,
      "loss": 0.6406,
      "num_input_tokens_seen": 341792,
      "step": 515
    },
    {
      "epoch": 0.27253668763102723,
      "grad_norm": 0.7233796119689941,
      "learning_rate": 0.0001360062893081761,
      "loss": 0.6394,
      "num_input_tokens_seen": 344768,
      "step": 520
    },
    {
      "epoch": 0.2751572327044025,
      "grad_norm": 0.6780195236206055,
      "learning_rate": 0.00013731656184486375,
      "loss": 0.4626,
      "num_input_tokens_seen": 347360,
      "step": 525
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 0.4932557940483093,
      "learning_rate": 0.00013862683438155136,
      "loss": 0.4874,
      "num_input_tokens_seen": 350592,
      "step": 530
    },
    {
      "epoch": 0.280398322851153,
      "grad_norm": 0.5400628447532654,
      "learning_rate": 0.000139937106918239,
      "loss": 0.491,
      "num_input_tokens_seen": 353568,
      "step": 535
    },
    {
      "epoch": 0.2830188679245283,
      "grad_norm": 0.49257761240005493,
      "learning_rate": 0.00014124737945492663,
      "loss": 0.6103,
      "num_input_tokens_seen": 356320,
      "step": 540
    },
    {
      "epoch": 0.2856394129979036,
      "grad_norm": 0.9981381893157959,
      "learning_rate": 0.00014255765199161423,
      "loss": 0.5264,
      "num_input_tokens_seen": 359968,
      "step": 545
    },
    {
      "epoch": 0.2882599580712788,
      "grad_norm": 0.6025562882423401,
      "learning_rate": 0.0001438679245283019,
      "loss": 0.4495,
      "num_input_tokens_seen": 362752,
      "step": 550
    },
    {
      "epoch": 0.2908805031446541,
      "grad_norm": 0.6628862023353577,
      "learning_rate": 0.00014517819706498953,
      "loss": 0.5891,
      "num_input_tokens_seen": 365312,
      "step": 555
    },
    {
      "epoch": 0.29350104821802936,
      "grad_norm": 0.32889747619628906,
      "learning_rate": 0.00014648846960167716,
      "loss": 0.4835,
      "num_input_tokens_seen": 368896,
      "step": 560
    },
    {
      "epoch": 0.29612159329140464,
      "grad_norm": 0.39700159430503845,
      "learning_rate": 0.00014779874213836477,
      "loss": 0.5361,
      "num_input_tokens_seen": 371648,
      "step": 565
    },
    {
      "epoch": 0.29874213836477986,
      "grad_norm": 0.39810144901275635,
      "learning_rate": 0.00014910901467505243,
      "loss": 0.5799,
      "num_input_tokens_seen": 374848,
      "step": 570
    },
    {
      "epoch": 0.30136268343815514,
      "grad_norm": 0.5036880373954773,
      "learning_rate": 0.00015041928721174007,
      "loss": 0.5398,
      "num_input_tokens_seen": 377184,
      "step": 575
    },
    {
      "epoch": 0.3039832285115304,
      "grad_norm": 0.34127914905548096,
      "learning_rate": 0.00015172955974842767,
      "loss": 0.6253,
      "num_input_tokens_seen": 381504,
      "step": 580
    },
    {
      "epoch": 0.30660377358490565,
      "grad_norm": 0.46368762850761414,
      "learning_rate": 0.0001530398322851153,
      "loss": 0.5353,
      "num_input_tokens_seen": 384672,
      "step": 585
    },
    {
      "epoch": 0.30922431865828093,
      "grad_norm": 0.48236772418022156,
      "learning_rate": 0.00015435010482180294,
      "loss": 0.5229,
      "num_input_tokens_seen": 389376,
      "step": 590
    },
    {
      "epoch": 0.3118448637316562,
      "grad_norm": 0.40100327134132385,
      "learning_rate": 0.00015566037735849058,
      "loss": 0.589,
      "num_input_tokens_seen": 392512,
      "step": 595
    },
    {
      "epoch": 0.31446540880503143,
      "grad_norm": 0.4037320017814636,
      "learning_rate": 0.0001569706498951782,
      "loss": 0.5162,
      "num_input_tokens_seen": 396864,
      "step": 600
    },
    {
      "epoch": 0.3170859538784067,
      "grad_norm": 0.6549465656280518,
      "learning_rate": 0.00015828092243186584,
      "loss": 0.4766,
      "num_input_tokens_seen": 399520,
      "step": 605
    },
    {
      "epoch": 0.319706498951782,
      "grad_norm": 1.6604889631271362,
      "learning_rate": 0.00015959119496855345,
      "loss": 0.4859,
      "num_input_tokens_seen": 402560,
      "step": 610
    },
    {
      "epoch": 0.3223270440251572,
      "grad_norm": 0.5272001624107361,
      "learning_rate": 0.00016090146750524109,
      "loss": 0.6413,
      "num_input_tokens_seen": 406240,
      "step": 615
    },
    {
      "epoch": 0.3249475890985325,
      "grad_norm": 0.4569492042064667,
      "learning_rate": 0.00016221174004192875,
      "loss": 0.5483,
      "num_input_tokens_seen": 408960,
      "step": 620
    },
    {
      "epoch": 0.3275681341719078,
      "grad_norm": 1.445087194442749,
      "learning_rate": 0.00016352201257861635,
      "loss": 0.6812,
      "num_input_tokens_seen": 412224,
      "step": 625
    },
    {
      "epoch": 0.330188679245283,
      "grad_norm": 0.30364173650741577,
      "learning_rate": 0.000164832285115304,
      "loss": 0.4919,
      "num_input_tokens_seen": 415296,
      "step": 630
    },
    {
      "epoch": 0.3328092243186583,
      "grad_norm": 0.5615999698638916,
      "learning_rate": 0.00016614255765199162,
      "loss": 0.4547,
      "num_input_tokens_seen": 417504,
      "step": 635
    },
    {
      "epoch": 0.33542976939203356,
      "grad_norm": 0.5971509218215942,
      "learning_rate": 0.00016745283018867923,
      "loss": 0.5495,
      "num_input_tokens_seen": 420288,
      "step": 640
    },
    {
      "epoch": 0.3380503144654088,
      "grad_norm": 0.43733957409858704,
      "learning_rate": 0.0001687631027253669,
      "loss": 0.6852,
      "num_input_tokens_seen": 422560,
      "step": 645
    },
    {
      "epoch": 0.34067085953878407,
      "grad_norm": 0.6991249322891235,
      "learning_rate": 0.00017007337526205453,
      "loss": 0.6101,
      "num_input_tokens_seen": 425472,
      "step": 650
    },
    {
      "epoch": 0.34329140461215935,
      "grad_norm": 0.4327264428138733,
      "learning_rate": 0.00017138364779874213,
      "loss": 0.5264,
      "num_input_tokens_seen": 429056,
      "step": 655
    },
    {
      "epoch": 0.34591194968553457,
      "grad_norm": 0.3444565534591675,
      "learning_rate": 0.00017269392033542977,
      "loss": 0.492,
      "num_input_tokens_seen": 433024,
      "step": 660
    },
    {
      "epoch": 0.34853249475890985,
      "grad_norm": 0.4016338586807251,
      "learning_rate": 0.0001740041928721174,
      "loss": 0.5435,
      "num_input_tokens_seen": 435776,
      "step": 665
    },
    {
      "epoch": 0.35115303983228513,
      "grad_norm": 0.38772329688072205,
      "learning_rate": 0.00017531446540880504,
      "loss": 0.5188,
      "num_input_tokens_seen": 438432,
      "step": 670
    },
    {
      "epoch": 0.35377358490566035,
      "grad_norm": 0.40205273032188416,
      "learning_rate": 0.00017662473794549267,
      "loss": 0.5188,
      "num_input_tokens_seen": 442400,
      "step": 675
    },
    {
      "epoch": 0.35639412997903563,
      "grad_norm": 0.5548549294471741,
      "learning_rate": 0.0001779350104821803,
      "loss": 0.6068,
      "num_input_tokens_seen": 445888,
      "step": 680
    },
    {
      "epoch": 0.3590146750524109,
      "grad_norm": 0.13598626852035522,
      "learning_rate": 0.0001792452830188679,
      "loss": 0.4774,
      "num_input_tokens_seen": 448864,
      "step": 685
    },
    {
      "epoch": 0.36163522012578614,
      "grad_norm": 0.30231255292892456,
      "learning_rate": 0.00018055555555555555,
      "loss": 0.6968,
      "num_input_tokens_seen": 452640,
      "step": 690
    },
    {
      "epoch": 0.3642557651991614,
      "grad_norm": 0.3976934254169464,
      "learning_rate": 0.0001818658280922432,
      "loss": 0.6116,
      "num_input_tokens_seen": 455296,
      "step": 695
    },
    {
      "epoch": 0.3668763102725367,
      "grad_norm": 1.4828462600708008,
      "learning_rate": 0.00018317610062893082,
      "loss": 0.5421,
      "num_input_tokens_seen": 457888,
      "step": 700
    },
    {
      "epoch": 0.3694968553459119,
      "grad_norm": 2.0073349475860596,
      "learning_rate": 0.00018448637316561845,
      "loss": 0.6139,
      "num_input_tokens_seen": 461664,
      "step": 705
    },
    {
      "epoch": 0.3721174004192872,
      "grad_norm": 0.3797239661216736,
      "learning_rate": 0.00018579664570230608,
      "loss": 0.6136,
      "num_input_tokens_seen": 464352,
      "step": 710
    },
    {
      "epoch": 0.3747379454926625,
      "grad_norm": 0.42687880992889404,
      "learning_rate": 0.0001871069182389937,
      "loss": 0.6077,
      "num_input_tokens_seen": 468256,
      "step": 715
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 1.0455635786056519,
      "learning_rate": 0.00018841719077568135,
      "loss": 0.5133,
      "num_input_tokens_seen": 471104,
      "step": 720
    },
    {
      "epoch": 0.379979035639413,
      "grad_norm": 0.9295072555541992,
      "learning_rate": 0.000189727463312369,
      "loss": 0.6115,
      "num_input_tokens_seen": 473696,
      "step": 725
    },
    {
      "epoch": 0.38259958071278827,
      "grad_norm": 0.4942203760147095,
      "learning_rate": 0.00019103773584905662,
      "loss": 0.5122,
      "num_input_tokens_seen": 476512,
      "step": 730
    },
    {
      "epoch": 0.38522012578616355,
      "grad_norm": 0.5037869811058044,
      "learning_rate": 0.00019234800838574423,
      "loss": 0.5904,
      "num_input_tokens_seen": 480128,
      "step": 735
    },
    {
      "epoch": 0.38784067085953877,
      "grad_norm": 0.6260188221931458,
      "learning_rate": 0.00019365828092243186,
      "loss": 0.6301,
      "num_input_tokens_seen": 483808,
      "step": 740
    },
    {
      "epoch": 0.39046121593291405,
      "grad_norm": 0.5473939180374146,
      "learning_rate": 0.00019496855345911953,
      "loss": 0.5939,
      "num_input_tokens_seen": 486240,
      "step": 745
    },
    {
      "epoch": 0.39308176100628933,
      "grad_norm": 0.4274798631668091,
      "learning_rate": 0.00019627882599580713,
      "loss": 0.4792,
      "num_input_tokens_seen": 488448,
      "step": 750
    },
    {
      "epoch": 0.39570230607966456,
      "grad_norm": 0.6462203860282898,
      "learning_rate": 0.00019758909853249477,
      "loss": 0.5778,
      "num_input_tokens_seen": 491392,
      "step": 755
    },
    {
      "epoch": 0.39832285115303984,
      "grad_norm": 0.4451272189617157,
      "learning_rate": 0.0001988993710691824,
      "loss": 0.5299,
      "num_input_tokens_seen": 495232,
      "step": 760
    },
    {
      "epoch": 0.4009433962264151,
      "grad_norm": 0.4326593279838562,
      "learning_rate": 0.00020020964360587,
      "loss": 0.4787,
      "num_input_tokens_seen": 498016,
      "step": 765
    },
    {
      "epoch": 0.40356394129979034,
      "grad_norm": 0.2858069837093353,
      "learning_rate": 0.00020151991614255767,
      "loss": 0.5109,
      "num_input_tokens_seen": 501152,
      "step": 770
    },
    {
      "epoch": 0.4061844863731656,
      "grad_norm": 0.41115859150886536,
      "learning_rate": 0.0002028301886792453,
      "loss": 0.496,
      "num_input_tokens_seen": 503872,
      "step": 775
    },
    {
      "epoch": 0.4088050314465409,
      "grad_norm": 0.37958356738090515,
      "learning_rate": 0.0002041404612159329,
      "loss": 0.5207,
      "num_input_tokens_seen": 507328,
      "step": 780
    },
    {
      "epoch": 0.4114255765199161,
      "grad_norm": 0.32550397515296936,
      "learning_rate": 0.00020545073375262055,
      "loss": 0.4149,
      "num_input_tokens_seen": 510912,
      "step": 785
    },
    {
      "epoch": 0.4140461215932914,
      "grad_norm": 0.32484132051467896,
      "learning_rate": 0.00020676100628930818,
      "loss": 0.5954,
      "num_input_tokens_seen": 514304,
      "step": 790
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.493306040763855,
      "learning_rate": 0.00020807127882599581,
      "loss": 0.4442,
      "num_input_tokens_seen": 517440,
      "step": 795
    },
    {
      "epoch": 0.4192872117400419,
      "grad_norm": 0.39065760374069214,
      "learning_rate": 0.00020938155136268345,
      "loss": 0.6763,
      "num_input_tokens_seen": 520416,
      "step": 800
    },
    {
      "epoch": 0.4219077568134172,
      "grad_norm": 0.38310790061950684,
      "learning_rate": 0.00021069182389937108,
      "loss": 0.6537,
      "num_input_tokens_seen": 523456,
      "step": 805
    },
    {
      "epoch": 0.42452830188679247,
      "grad_norm": 0.3452560603618622,
      "learning_rate": 0.0002120020964360587,
      "loss": 0.5032,
      "num_input_tokens_seen": 526720,
      "step": 810
    },
    {
      "epoch": 0.4271488469601677,
      "grad_norm": 0.6603071689605713,
      "learning_rate": 0.00021331236897274632,
      "loss": 0.4356,
      "num_input_tokens_seen": 530272,
      "step": 815
    },
    {
      "epoch": 0.429769392033543,
      "grad_norm": 0.6812957525253296,
      "learning_rate": 0.00021462264150943399,
      "loss": 0.6282,
      "num_input_tokens_seen": 533408,
      "step": 820
    },
    {
      "epoch": 0.43238993710691825,
      "grad_norm": 1.0582618713378906,
      "learning_rate": 0.0002159329140461216,
      "loss": 0.6366,
      "num_input_tokens_seen": 536768,
      "step": 825
    },
    {
      "epoch": 0.4350104821802935,
      "grad_norm": 0.2061627060174942,
      "learning_rate": 0.00021724318658280923,
      "loss": 0.6108,
      "num_input_tokens_seen": 541184,
      "step": 830
    },
    {
      "epoch": 0.43763102725366876,
      "grad_norm": 0.6013889312744141,
      "learning_rate": 0.00021855345911949686,
      "loss": 0.4637,
      "num_input_tokens_seen": 544576,
      "step": 835
    },
    {
      "epoch": 0.44025157232704404,
      "grad_norm": 0.35411372780799866,
      "learning_rate": 0.00021986373165618447,
      "loss": 0.5865,
      "num_input_tokens_seen": 549408,
      "step": 840
    },
    {
      "epoch": 0.44287211740041926,
      "grad_norm": 0.4917649030685425,
      "learning_rate": 0.00022117400419287213,
      "loss": 0.468,
      "num_input_tokens_seen": 552032,
      "step": 845
    },
    {
      "epoch": 0.44549266247379454,
      "grad_norm": 0.3555688261985779,
      "learning_rate": 0.00022248427672955977,
      "loss": 0.5104,
      "num_input_tokens_seen": 555872,
      "step": 850
    },
    {
      "epoch": 0.4481132075471698,
      "grad_norm": 0.2595771849155426,
      "learning_rate": 0.00022379454926624737,
      "loss": 0.6376,
      "num_input_tokens_seen": 559136,
      "step": 855
    },
    {
      "epoch": 0.45073375262054505,
      "grad_norm": 0.2751867175102234,
      "learning_rate": 0.000225104821802935,
      "loss": 0.5296,
      "num_input_tokens_seen": 561984,
      "step": 860
    },
    {
      "epoch": 0.4533542976939203,
      "grad_norm": 0.3163353502750397,
      "learning_rate": 0.00022641509433962264,
      "loss": 0.4832,
      "num_input_tokens_seen": 564448,
      "step": 865
    },
    {
      "epoch": 0.4559748427672956,
      "grad_norm": 0.4697161018848419,
      "learning_rate": 0.00022772536687631028,
      "loss": 0.5687,
      "num_input_tokens_seen": 566752,
      "step": 870
    },
    {
      "epoch": 0.4585953878406709,
      "grad_norm": 0.35889360308647156,
      "learning_rate": 0.0002290356394129979,
      "loss": 0.5288,
      "num_input_tokens_seen": 569856,
      "step": 875
    },
    {
      "epoch": 0.4612159329140461,
      "grad_norm": 0.3321026563644409,
      "learning_rate": 0.00023034591194968554,
      "loss": 0.5104,
      "num_input_tokens_seen": 573056,
      "step": 880
    },
    {
      "epoch": 0.4638364779874214,
      "grad_norm": 0.41044166684150696,
      "learning_rate": 0.00023165618448637318,
      "loss": 0.5814,
      "num_input_tokens_seen": 575712,
      "step": 885
    },
    {
      "epoch": 0.46645702306079667,
      "grad_norm": 0.338021457195282,
      "learning_rate": 0.00023296645702306079,
      "loss": 0.4417,
      "num_input_tokens_seen": 578880,
      "step": 890
    },
    {
      "epoch": 0.4690775681341719,
      "grad_norm": 0.24231953918933868,
      "learning_rate": 0.00023427672955974845,
      "loss": 0.4844,
      "num_input_tokens_seen": 581920,
      "step": 895
    },
    {
      "epoch": 0.4716981132075472,
      "grad_norm": 0.25651973485946655,
      "learning_rate": 0.00023558700209643608,
      "loss": 0.5446,
      "num_input_tokens_seen": 584672,
      "step": 900
    },
    {
      "epoch": 0.47431865828092246,
      "grad_norm": 0.5090682506561279,
      "learning_rate": 0.0002368972746331237,
      "loss": 0.4763,
      "num_input_tokens_seen": 588704,
      "step": 905
    },
    {
      "epoch": 0.4769392033542977,
      "grad_norm": 0.4438055455684662,
      "learning_rate": 0.00023820754716981132,
      "loss": 0.5383,
      "num_input_tokens_seen": 591712,
      "step": 910
    },
    {
      "epoch": 0.47955974842767296,
      "grad_norm": 0.56185382604599,
      "learning_rate": 0.00023951781970649896,
      "loss": 0.5598,
      "num_input_tokens_seen": 595104,
      "step": 915
    },
    {
      "epoch": 0.48218029350104824,
      "grad_norm": 0.26994797587394714,
      "learning_rate": 0.0002408280922431866,
      "loss": 0.494,
      "num_input_tokens_seen": 599808,
      "step": 920
    },
    {
      "epoch": 0.48480083857442346,
      "grad_norm": 0.7096908092498779,
      "learning_rate": 0.00024213836477987423,
      "loss": 0.4524,
      "num_input_tokens_seen": 602720,
      "step": 925
    },
    {
      "epoch": 0.48742138364779874,
      "grad_norm": 0.2845062017440796,
      "learning_rate": 0.00024344863731656186,
      "loss": 0.4922,
      "num_input_tokens_seen": 606176,
      "step": 930
    },
    {
      "epoch": 0.490041928721174,
      "grad_norm": 0.19293244183063507,
      "learning_rate": 0.00024475890985324947,
      "loss": 0.4674,
      "num_input_tokens_seen": 608864,
      "step": 935
    },
    {
      "epoch": 0.49266247379454925,
      "grad_norm": 0.2707439064979553,
      "learning_rate": 0.0002460691823899371,
      "loss": 0.5117,
      "num_input_tokens_seen": 611968,
      "step": 940
    },
    {
      "epoch": 0.49528301886792453,
      "grad_norm": 0.4684884250164032,
      "learning_rate": 0.00024737945492662474,
      "loss": 0.6172,
      "num_input_tokens_seen": 614848,
      "step": 945
    },
    {
      "epoch": 0.4979035639412998,
      "grad_norm": 0.2459111213684082,
      "learning_rate": 0.00024868972746331237,
      "loss": 0.665,
      "num_input_tokens_seen": 618080,
      "step": 950
    },
    {
      "epoch": 0.500524109014675,
      "grad_norm": 0.32144564390182495,
      "learning_rate": 0.00025,
      "loss": 0.3781,
      "num_input_tokens_seen": 620736,
      "step": 955
    },
    {
      "epoch": 0.5031446540880503,
      "grad_norm": 0.6321529150009155,
      "learning_rate": 0.00025131027253668764,
      "loss": 0.4699,
      "num_input_tokens_seen": 624480,
      "step": 960
    },
    {
      "epoch": 0.5057651991614256,
      "grad_norm": 0.3725711405277252,
      "learning_rate": 0.0002526205450733753,
      "loss": 0.4593,
      "num_input_tokens_seen": 628448,
      "step": 965
    },
    {
      "epoch": 0.5083857442348009,
      "grad_norm": 0.18621298670768738,
      "learning_rate": 0.0002539308176100629,
      "loss": 0.6066,
      "num_input_tokens_seen": 632000,
      "step": 970
    },
    {
      "epoch": 0.5110062893081762,
      "grad_norm": 0.22298184037208557,
      "learning_rate": 0.0002552410901467505,
      "loss": 0.6094,
      "num_input_tokens_seen": 635680,
      "step": 975
    },
    {
      "epoch": 0.5136268343815513,
      "grad_norm": 0.5000988841056824,
      "learning_rate": 0.0002565513626834381,
      "loss": 0.7542,
      "num_input_tokens_seen": 638496,
      "step": 980
    },
    {
      "epoch": 0.5162473794549266,
      "grad_norm": 0.3263683021068573,
      "learning_rate": 0.0002578616352201258,
      "loss": 0.4697,
      "num_input_tokens_seen": 642272,
      "step": 985
    },
    {
      "epoch": 0.5188679245283019,
      "grad_norm": 0.29571521282196045,
      "learning_rate": 0.00025917190775681345,
      "loss": 0.5283,
      "num_input_tokens_seen": 645888,
      "step": 990
    },
    {
      "epoch": 0.5214884696016772,
      "grad_norm": 0.39655929803848267,
      "learning_rate": 0.0002604821802935011,
      "loss": 0.3726,
      "num_input_tokens_seen": 651808,
      "step": 995
    },
    {
      "epoch": 0.5241090146750524,
      "grad_norm": 0.30409377813339233,
      "learning_rate": 0.0002617924528301887,
      "loss": 0.5356,
      "num_input_tokens_seen": 655488,
      "step": 1000
    },
    {
      "epoch": 0.5267295597484277,
      "grad_norm": 0.15399502217769623,
      "learning_rate": 0.0002631027253668763,
      "loss": 0.4621,
      "num_input_tokens_seen": 658976,
      "step": 1005
    },
    {
      "epoch": 0.5293501048218029,
      "grad_norm": 0.29966670274734497,
      "learning_rate": 0.00026441299790356393,
      "loss": 0.4489,
      "num_input_tokens_seen": 662048,
      "step": 1010
    },
    {
      "epoch": 0.5319706498951782,
      "grad_norm": 0.2110607773065567,
      "learning_rate": 0.00026572327044025156,
      "loss": 0.5845,
      "num_input_tokens_seen": 664640,
      "step": 1015
    },
    {
      "epoch": 0.5345911949685535,
      "grad_norm": 0.2930071949958801,
      "learning_rate": 0.0002670335429769392,
      "loss": 0.489,
      "num_input_tokens_seen": 669056,
      "step": 1020
    },
    {
      "epoch": 0.5372117400419287,
      "grad_norm": 0.3236556053161621,
      "learning_rate": 0.00026834381551362683,
      "loss": 0.5195,
      "num_input_tokens_seen": 672128,
      "step": 1025
    },
    {
      "epoch": 0.539832285115304,
      "grad_norm": 0.7644017934799194,
      "learning_rate": 0.0002696540880503145,
      "loss": 0.4863,
      "num_input_tokens_seen": 675136,
      "step": 1030
    },
    {
      "epoch": 0.5424528301886793,
      "grad_norm": 0.3087236285209656,
      "learning_rate": 0.0002709643605870021,
      "loss": 0.6039,
      "num_input_tokens_seen": 678976,
      "step": 1035
    },
    {
      "epoch": 0.5450733752620545,
      "grad_norm": 0.6482839584350586,
      "learning_rate": 0.00027227463312368973,
      "loss": 0.6647,
      "num_input_tokens_seen": 682176,
      "step": 1040
    },
    {
      "epoch": 0.5476939203354297,
      "grad_norm": 0.2930818498134613,
      "learning_rate": 0.00027358490566037737,
      "loss": 0.5717,
      "num_input_tokens_seen": 684768,
      "step": 1045
    },
    {
      "epoch": 0.550314465408805,
      "grad_norm": 0.2553316354751587,
      "learning_rate": 0.000274895178197065,
      "loss": 0.5619,
      "num_input_tokens_seen": 687392,
      "step": 1050
    },
    {
      "epoch": 0.5529350104821803,
      "grad_norm": 0.6335535645484924,
      "learning_rate": 0.00027620545073375264,
      "loss": 0.4905,
      "num_input_tokens_seen": 690496,
      "step": 1055
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.19580091536045074,
      "learning_rate": 0.00027751572327044027,
      "loss": 0.6701,
      "num_input_tokens_seen": 693376,
      "step": 1060
    },
    {
      "epoch": 0.5581761006289309,
      "grad_norm": 0.4174307584762573,
      "learning_rate": 0.00027882599580712785,
      "loss": 0.4841,
      "num_input_tokens_seen": 695808,
      "step": 1065
    },
    {
      "epoch": 0.560796645702306,
      "grad_norm": 0.21251614391803741,
      "learning_rate": 0.0002801362683438155,
      "loss": 0.491,
      "num_input_tokens_seen": 699424,
      "step": 1070
    },
    {
      "epoch": 0.5634171907756813,
      "grad_norm": 0.21296368539333344,
      "learning_rate": 0.0002814465408805031,
      "loss": 0.5349,
      "num_input_tokens_seen": 703584,
      "step": 1075
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 0.27232101559638977,
      "learning_rate": 0.0002827568134171908,
      "loss": 0.5218,
      "num_input_tokens_seen": 706144,
      "step": 1080
    },
    {
      "epoch": 0.5686582809224319,
      "grad_norm": 0.2693198025226593,
      "learning_rate": 0.00028406708595387844,
      "loss": 0.5281,
      "num_input_tokens_seen": 708864,
      "step": 1085
    },
    {
      "epoch": 0.5712788259958071,
      "grad_norm": 0.2605394423007965,
      "learning_rate": 0.0002853773584905661,
      "loss": 0.5503,
      "num_input_tokens_seen": 712352,
      "step": 1090
    },
    {
      "epoch": 0.5738993710691824,
      "grad_norm": 0.3070299029350281,
      "learning_rate": 0.00028668763102725366,
      "loss": 0.7352,
      "num_input_tokens_seen": 715296,
      "step": 1095
    },
    {
      "epoch": 0.5765199161425576,
      "grad_norm": 0.2347191870212555,
      "learning_rate": 0.0002879979035639413,
      "loss": 0.4537,
      "num_input_tokens_seen": 718528,
      "step": 1100
    },
    {
      "epoch": 0.5791404612159329,
      "grad_norm": 0.17859429121017456,
      "learning_rate": 0.00028930817610062893,
      "loss": 0.4627,
      "num_input_tokens_seen": 722752,
      "step": 1105
    },
    {
      "epoch": 0.5817610062893082,
      "grad_norm": 0.20949424803256989,
      "learning_rate": 0.00029061844863731656,
      "loss": 0.4673,
      "num_input_tokens_seen": 725824,
      "step": 1110
    },
    {
      "epoch": 0.5843815513626834,
      "grad_norm": 0.18730264902114868,
      "learning_rate": 0.0002919287211740042,
      "loss": 0.5829,
      "num_input_tokens_seen": 730080,
      "step": 1115
    },
    {
      "epoch": 0.5870020964360587,
      "grad_norm": 0.27852848172187805,
      "learning_rate": 0.00029323899371069183,
      "loss": 0.5032,
      "num_input_tokens_seen": 733216,
      "step": 1120
    },
    {
      "epoch": 0.589622641509434,
      "grad_norm": 0.36160746216773987,
      "learning_rate": 0.00029454926624737946,
      "loss": 0.6646,
      "num_input_tokens_seen": 736416,
      "step": 1125
    },
    {
      "epoch": 0.5922431865828093,
      "grad_norm": 0.1805410534143448,
      "learning_rate": 0.0002958595387840671,
      "loss": 0.4559,
      "num_input_tokens_seen": 739808,
      "step": 1130
    },
    {
      "epoch": 0.5948637316561844,
      "grad_norm": 0.3030797839164734,
      "learning_rate": 0.00029716981132075473,
      "loss": 0.5518,
      "num_input_tokens_seen": 742400,
      "step": 1135
    },
    {
      "epoch": 0.5974842767295597,
      "grad_norm": 0.4321233630180359,
      "learning_rate": 0.00029848008385744237,
      "loss": 0.4005,
      "num_input_tokens_seen": 745760,
      "step": 1140
    },
    {
      "epoch": 0.600104821802935,
      "grad_norm": 0.3557263910770416,
      "learning_rate": 0.00029979035639413,
      "loss": 0.5925,
      "num_input_tokens_seen": 748736,
      "step": 1145
    },
    {
      "epoch": 0.6027253668763103,
      "grad_norm": 0.21365858614444733,
      "learning_rate": 0.00030110062893081764,
      "loss": 0.5357,
      "num_input_tokens_seen": 752032,
      "step": 1150
    },
    {
      "epoch": 0.6053459119496856,
      "grad_norm": 0.15353770554065704,
      "learning_rate": 0.00030241090146750527,
      "loss": 0.491,
      "num_input_tokens_seen": 756736,
      "step": 1155
    },
    {
      "epoch": 0.6079664570230608,
      "grad_norm": 0.2292877733707428,
      "learning_rate": 0.00030372117400419285,
      "loss": 0.4877,
      "num_input_tokens_seen": 759680,
      "step": 1160
    },
    {
      "epoch": 0.610587002096436,
      "grad_norm": 0.376157283782959,
      "learning_rate": 0.0003050314465408805,
      "loss": 0.5316,
      "num_input_tokens_seen": 762560,
      "step": 1165
    },
    {
      "epoch": 0.6132075471698113,
      "grad_norm": 0.21052594482898712,
      "learning_rate": 0.0003063417190775681,
      "loss": 0.4494,
      "num_input_tokens_seen": 765600,
      "step": 1170
    },
    {
      "epoch": 0.6158280922431866,
      "grad_norm": 0.15436799824237823,
      "learning_rate": 0.00030765199161425575,
      "loss": 0.4,
      "num_input_tokens_seen": 768736,
      "step": 1175
    },
    {
      "epoch": 0.6184486373165619,
      "grad_norm": 0.25230199098587036,
      "learning_rate": 0.00030896226415094344,
      "loss": 0.6684,
      "num_input_tokens_seen": 771872,
      "step": 1180
    },
    {
      "epoch": 0.6210691823899371,
      "grad_norm": 0.2537326216697693,
      "learning_rate": 0.0003102725366876311,
      "loss": 0.619,
      "num_input_tokens_seen": 775008,
      "step": 1185
    },
    {
      "epoch": 0.6236897274633124,
      "grad_norm": 0.21003709733486176,
      "learning_rate": 0.00031158280922431866,
      "loss": 0.4269,
      "num_input_tokens_seen": 777920,
      "step": 1190
    },
    {
      "epoch": 0.6263102725366876,
      "grad_norm": 0.1605054885149002,
      "learning_rate": 0.0003128930817610063,
      "loss": 0.5171,
      "num_input_tokens_seen": 781056,
      "step": 1195
    },
    {
      "epoch": 0.6289308176100629,
      "grad_norm": 0.2101152390241623,
      "learning_rate": 0.0003142033542976939,
      "loss": 0.5798,
      "num_input_tokens_seen": 784128,
      "step": 1200
    },
    {
      "epoch": 0.6315513626834381,
      "grad_norm": 0.22440187633037567,
      "learning_rate": 0.00031551362683438156,
      "loss": 0.5388,
      "num_input_tokens_seen": 786624,
      "step": 1205
    },
    {
      "epoch": 0.6341719077568134,
      "grad_norm": 0.3241609036922455,
      "learning_rate": 0.0003168238993710692,
      "loss": 0.5321,
      "num_input_tokens_seen": 789664,
      "step": 1210
    },
    {
      "epoch": 0.6367924528301887,
      "grad_norm": 0.2078603059053421,
      "learning_rate": 0.00031813417190775683,
      "loss": 0.4326,
      "num_input_tokens_seen": 793824,
      "step": 1215
    },
    {
      "epoch": 0.639412997903564,
      "grad_norm": 0.4095917344093323,
      "learning_rate": 0.0003194444444444444,
      "loss": 0.6322,
      "num_input_tokens_seen": 796544,
      "step": 1220
    },
    {
      "epoch": 0.6420335429769392,
      "grad_norm": 0.17480215430259705,
      "learning_rate": 0.00032075471698113204,
      "loss": 0.4127,
      "num_input_tokens_seen": 799520,
      "step": 1225
    },
    {
      "epoch": 0.6446540880503144,
      "grad_norm": 0.3564589321613312,
      "learning_rate": 0.00032206498951781973,
      "loss": 0.3456,
      "num_input_tokens_seen": 802464,
      "step": 1230
    },
    {
      "epoch": 0.6472746331236897,
      "grad_norm": 0.3239567279815674,
      "learning_rate": 0.00032337526205450737,
      "loss": 0.5828,
      "num_input_tokens_seen": 805536,
      "step": 1235
    },
    {
      "epoch": 0.649895178197065,
      "grad_norm": 0.21953368186950684,
      "learning_rate": 0.000324685534591195,
      "loss": 0.3669,
      "num_input_tokens_seen": 808448,
      "step": 1240
    },
    {
      "epoch": 0.6525157232704403,
      "grad_norm": 0.3092052638530731,
      "learning_rate": 0.00032599580712788263,
      "loss": 0.4455,
      "num_input_tokens_seen": 812160,
      "step": 1245
    },
    {
      "epoch": 0.6551362683438156,
      "grad_norm": 0.5536033511161804,
      "learning_rate": 0.0003273060796645702,
      "loss": 0.637,
      "num_input_tokens_seen": 814752,
      "step": 1250
    },
    {
      "epoch": 0.6577568134171907,
      "grad_norm": 0.3831644654273987,
      "learning_rate": 0.00032861635220125785,
      "loss": 0.4277,
      "num_input_tokens_seen": 817568,
      "step": 1255
    },
    {
      "epoch": 0.660377358490566,
      "grad_norm": 0.23082125186920166,
      "learning_rate": 0.0003299266247379455,
      "loss": 0.5027,
      "num_input_tokens_seen": 820384,
      "step": 1260
    },
    {
      "epoch": 0.6629979035639413,
      "grad_norm": 0.1872601956129074,
      "learning_rate": 0.0003312368972746331,
      "loss": 0.3046,
      "num_input_tokens_seen": 823040,
      "step": 1265
    },
    {
      "epoch": 0.6656184486373166,
      "grad_norm": 0.25367698073387146,
      "learning_rate": 0.00033254716981132075,
      "loss": 0.5936,
      "num_input_tokens_seen": 825600,
      "step": 1270
    },
    {
      "epoch": 0.6682389937106918,
      "grad_norm": 0.2750582695007324,
      "learning_rate": 0.0003338574423480084,
      "loss": 0.5951,
      "num_input_tokens_seen": 828032,
      "step": 1275
    },
    {
      "epoch": 0.6708595387840671,
      "grad_norm": 0.2764155864715576,
      "learning_rate": 0.0003351677148846961,
      "loss": 0.4522,
      "num_input_tokens_seen": 831296,
      "step": 1280
    },
    {
      "epoch": 0.6734800838574424,
      "grad_norm": 0.1840190589427948,
      "learning_rate": 0.00033647798742138366,
      "loss": 0.6076,
      "num_input_tokens_seen": 834208,
      "step": 1285
    },
    {
      "epoch": 0.6761006289308176,
      "grad_norm": 0.32448646426200867,
      "learning_rate": 0.0003377882599580713,
      "loss": 0.4781,
      "num_input_tokens_seen": 838048,
      "step": 1290
    },
    {
      "epoch": 0.6787211740041929,
      "grad_norm": 0.4407355487346649,
      "learning_rate": 0.0003390985324947589,
      "loss": 0.5903,
      "num_input_tokens_seen": 841536,
      "step": 1295
    },
    {
      "epoch": 0.6813417190775681,
      "grad_norm": 0.16768911480903625,
      "learning_rate": 0.00034040880503144656,
      "loss": 0.4476,
      "num_input_tokens_seen": 845184,
      "step": 1300
    },
    {
      "epoch": 0.6839622641509434,
      "grad_norm": 0.16010695695877075,
      "learning_rate": 0.0003417190775681342,
      "loss": 0.442,
      "num_input_tokens_seen": 848000,
      "step": 1305
    },
    {
      "epoch": 0.6865828092243187,
      "grad_norm": 0.30624786019325256,
      "learning_rate": 0.00034302935010482183,
      "loss": 0.5777,
      "num_input_tokens_seen": 851040,
      "step": 1310
    },
    {
      "epoch": 0.689203354297694,
      "grad_norm": 0.3250081241130829,
      "learning_rate": 0.0003443396226415094,
      "loss": 0.5948,
      "num_input_tokens_seen": 853920,
      "step": 1315
    },
    {
      "epoch": 0.6918238993710691,
      "grad_norm": 0.27084842324256897,
      "learning_rate": 0.00034564989517819704,
      "loss": 0.513,
      "num_input_tokens_seen": 857312,
      "step": 1320
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 0.20950250327587128,
      "learning_rate": 0.0003469601677148847,
      "loss": 0.4137,
      "num_input_tokens_seen": 859904,
      "step": 1325
    },
    {
      "epoch": 0.6970649895178197,
      "grad_norm": 0.2360249161720276,
      "learning_rate": 0.00034827044025157236,
      "loss": 0.5575,
      "num_input_tokens_seen": 863680,
      "step": 1330
    },
    {
      "epoch": 0.699685534591195,
      "grad_norm": 0.262787401676178,
      "learning_rate": 0.00034958071278826,
      "loss": 0.5729,
      "num_input_tokens_seen": 866880,
      "step": 1335
    },
    {
      "epoch": 0.7023060796645703,
      "grad_norm": 0.47497716546058655,
      "learning_rate": 0.00035089098532494763,
      "loss": 0.4186,
      "num_input_tokens_seen": 869856,
      "step": 1340
    },
    {
      "epoch": 0.7049266247379455,
      "grad_norm": 0.30667394399642944,
      "learning_rate": 0.0003522012578616352,
      "loss": 0.5439,
      "num_input_tokens_seen": 873472,
      "step": 1345
    },
    {
      "epoch": 0.7075471698113207,
      "grad_norm": 0.19851931929588318,
      "learning_rate": 0.00035351153039832285,
      "loss": 0.4465,
      "num_input_tokens_seen": 876704,
      "step": 1350
    },
    {
      "epoch": 0.710167714884696,
      "grad_norm": 0.14134572446346283,
      "learning_rate": 0.0003548218029350105,
      "loss": 0.4769,
      "num_input_tokens_seen": 880576,
      "step": 1355
    },
    {
      "epoch": 0.7127882599580713,
      "grad_norm": 0.19098827242851257,
      "learning_rate": 0.0003561320754716981,
      "loss": 0.4906,
      "num_input_tokens_seen": 883680,
      "step": 1360
    },
    {
      "epoch": 0.7154088050314465,
      "grad_norm": 0.19019347429275513,
      "learning_rate": 0.00035744234800838575,
      "loss": 0.5653,
      "num_input_tokens_seen": 886464,
      "step": 1365
    },
    {
      "epoch": 0.7180293501048218,
      "grad_norm": 0.21790744364261627,
      "learning_rate": 0.0003587526205450734,
      "loss": 0.5247,
      "num_input_tokens_seen": 889536,
      "step": 1370
    },
    {
      "epoch": 0.7206498951781971,
      "grad_norm": 0.27587077021598816,
      "learning_rate": 0.00036006289308176097,
      "loss": 0.507,
      "num_input_tokens_seen": 894272,
      "step": 1375
    },
    {
      "epoch": 0.7232704402515723,
      "grad_norm": 0.14009474217891693,
      "learning_rate": 0.00036137316561844865,
      "loss": 0.4721,
      "num_input_tokens_seen": 897120,
      "step": 1380
    },
    {
      "epoch": 0.7258909853249476,
      "grad_norm": 0.25120308995246887,
      "learning_rate": 0.0003626834381551363,
      "loss": 0.4942,
      "num_input_tokens_seen": 899808,
      "step": 1385
    },
    {
      "epoch": 0.7285115303983228,
      "grad_norm": 0.33952027559280396,
      "learning_rate": 0.0003639937106918239,
      "loss": 0.4896,
      "num_input_tokens_seen": 903520,
      "step": 1390
    },
    {
      "epoch": 0.7311320754716981,
      "grad_norm": 0.19995561242103577,
      "learning_rate": 0.00036530398322851156,
      "loss": 0.5451,
      "num_input_tokens_seen": 907200,
      "step": 1395
    },
    {
      "epoch": 0.7337526205450734,
      "grad_norm": 0.1367632895708084,
      "learning_rate": 0.0003666142557651992,
      "loss": 0.4845,
      "num_input_tokens_seen": 911392,
      "step": 1400
    },
    {
      "epoch": 0.7363731656184487,
      "grad_norm": 0.27509599924087524,
      "learning_rate": 0.00036792452830188677,
      "loss": 0.5145,
      "num_input_tokens_seen": 913440,
      "step": 1405
    },
    {
      "epoch": 0.7389937106918238,
      "grad_norm": 0.22603419423103333,
      "learning_rate": 0.0003692348008385744,
      "loss": 0.4958,
      "num_input_tokens_seen": 917888,
      "step": 1410
    },
    {
      "epoch": 0.7416142557651991,
      "grad_norm": 0.11688148975372314,
      "learning_rate": 0.00037054507337526204,
      "loss": 0.5504,
      "num_input_tokens_seen": 921376,
      "step": 1415
    },
    {
      "epoch": 0.7442348008385744,
      "grad_norm": 0.19329074025154114,
      "learning_rate": 0.0003718553459119497,
      "loss": 0.5504,
      "num_input_tokens_seen": 924800,
      "step": 1420
    },
    {
      "epoch": 0.7468553459119497,
      "grad_norm": 0.795558750629425,
      "learning_rate": 0.0003731656184486373,
      "loss": 0.64,
      "num_input_tokens_seen": 927648,
      "step": 1425
    },
    {
      "epoch": 0.749475890985325,
      "grad_norm": 0.38749271631240845,
      "learning_rate": 0.000374475890985325,
      "loss": 0.4149,
      "num_input_tokens_seen": 930144,
      "step": 1430
    },
    {
      "epoch": 0.7520964360587002,
      "grad_norm": 0.17275112867355347,
      "learning_rate": 0.00037578616352201263,
      "loss": 0.413,
      "num_input_tokens_seen": 933184,
      "step": 1435
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 0.16183999180793762,
      "learning_rate": 0.0003770964360587002,
      "loss": 0.5526,
      "num_input_tokens_seen": 937376,
      "step": 1440
    },
    {
      "epoch": 0.7573375262054507,
      "grad_norm": 0.24270489811897278,
      "learning_rate": 0.00037840670859538785,
      "loss": 0.4662,
      "num_input_tokens_seen": 940160,
      "step": 1445
    },
    {
      "epoch": 0.759958071278826,
      "grad_norm": 0.27121391892433167,
      "learning_rate": 0.0003797169811320755,
      "loss": 0.5484,
      "num_input_tokens_seen": 942816,
      "step": 1450
    },
    {
      "epoch": 0.7625786163522013,
      "grad_norm": 0.3597170114517212,
      "learning_rate": 0.0003810272536687631,
      "loss": 0.5784,
      "num_input_tokens_seen": 945952,
      "step": 1455
    },
    {
      "epoch": 0.7651991614255765,
      "grad_norm": 0.34672558307647705,
      "learning_rate": 0.00038233752620545075,
      "loss": 0.3904,
      "num_input_tokens_seen": 949472,
      "step": 1460
    },
    {
      "epoch": 0.7678197064989518,
      "grad_norm": 0.4023905098438263,
      "learning_rate": 0.0003836477987421384,
      "loss": 0.5353,
      "num_input_tokens_seen": 952672,
      "step": 1465
    },
    {
      "epoch": 0.7704402515723271,
      "grad_norm": 0.15619751811027527,
      "learning_rate": 0.00038495807127882596,
      "loss": 0.5603,
      "num_input_tokens_seen": 955520,
      "step": 1470
    },
    {
      "epoch": 0.7730607966457023,
      "grad_norm": 0.21208497881889343,
      "learning_rate": 0.0003862683438155136,
      "loss": 0.5678,
      "num_input_tokens_seen": 958464,
      "step": 1475
    },
    {
      "epoch": 0.7756813417190775,
      "grad_norm": 0.38263896107673645,
      "learning_rate": 0.0003875786163522013,
      "loss": 0.4988,
      "num_input_tokens_seen": 960768,
      "step": 1480
    },
    {
      "epoch": 0.7783018867924528,
      "grad_norm": 0.220529705286026,
      "learning_rate": 0.0003888888888888889,
      "loss": 0.6904,
      "num_input_tokens_seen": 963616,
      "step": 1485
    },
    {
      "epoch": 0.7809224318658281,
      "grad_norm": 0.11573688685894012,
      "learning_rate": 0.00039019916142557656,
      "loss": 0.7019,
      "num_input_tokens_seen": 967936,
      "step": 1490
    },
    {
      "epoch": 0.7835429769392034,
      "grad_norm": 0.11527425050735474,
      "learning_rate": 0.0003915094339622642,
      "loss": 0.481,
      "num_input_tokens_seen": 971744,
      "step": 1495
    },
    {
      "epoch": 0.7861635220125787,
      "grad_norm": 0.30255427956581116,
      "learning_rate": 0.00039281970649895177,
      "loss": 0.5369,
      "num_input_tokens_seen": 974080,
      "step": 1500
    },
    {
      "epoch": 0.7887840670859538,
      "grad_norm": 0.15047305822372437,
      "learning_rate": 0.0003941299790356394,
      "loss": 0.4812,
      "num_input_tokens_seen": 976992,
      "step": 1505
    },
    {
      "epoch": 0.7914046121593291,
      "grad_norm": 0.24886628985404968,
      "learning_rate": 0.00039544025157232704,
      "loss": 0.5068,
      "num_input_tokens_seen": 979648,
      "step": 1510
    },
    {
      "epoch": 0.7940251572327044,
      "grad_norm": 0.2130061537027359,
      "learning_rate": 0.0003967505241090147,
      "loss": 0.5554,
      "num_input_tokens_seen": 983136,
      "step": 1515
    },
    {
      "epoch": 0.7966457023060797,
      "grad_norm": 0.13536806404590607,
      "learning_rate": 0.0003980607966457023,
      "loss": 0.4946,
      "num_input_tokens_seen": 986976,
      "step": 1520
    },
    {
      "epoch": 0.799266247379455,
      "grad_norm": 0.10803242772817612,
      "learning_rate": 0.00039937106918238994,
      "loss": 0.4699,
      "num_input_tokens_seen": 990272,
      "step": 1525
    },
    {
      "epoch": 0.8018867924528302,
      "grad_norm": 0.16849279403686523,
      "learning_rate": 0.0004006813417190776,
      "loss": 0.4808,
      "num_input_tokens_seen": 994368,
      "step": 1530
    },
    {
      "epoch": 0.8045073375262054,
      "grad_norm": 0.2097587287425995,
      "learning_rate": 0.0004019916142557652,
      "loss": 0.4407,
      "num_input_tokens_seen": 997504,
      "step": 1535
    },
    {
      "epoch": 0.8071278825995807,
      "grad_norm": 0.23893022537231445,
      "learning_rate": 0.00040330188679245284,
      "loss": 0.568,
      "num_input_tokens_seen": 1000352,
      "step": 1540
    },
    {
      "epoch": 0.809748427672956,
      "grad_norm": 0.1539527326822281,
      "learning_rate": 0.0004046121593291405,
      "loss": 0.58,
      "num_input_tokens_seen": 1003616,
      "step": 1545
    },
    {
      "epoch": 0.8123689727463312,
      "grad_norm": 0.6487398743629456,
      "learning_rate": 0.0004059224318658281,
      "loss": 0.5682,
      "num_input_tokens_seen": 1006240,
      "step": 1550
    },
    {
      "epoch": 0.8149895178197065,
      "grad_norm": 0.19436608254909515,
      "learning_rate": 0.00040723270440251575,
      "loss": 0.5371,
      "num_input_tokens_seen": 1009440,
      "step": 1555
    },
    {
      "epoch": 0.8176100628930818,
      "grad_norm": 0.20720188319683075,
      "learning_rate": 0.00040854297693920333,
      "loss": 0.5506,
      "num_input_tokens_seen": 1011936,
      "step": 1560
    },
    {
      "epoch": 0.820230607966457,
      "grad_norm": 0.18255719542503357,
      "learning_rate": 0.00040985324947589096,
      "loss": 0.4716,
      "num_input_tokens_seen": 1015040,
      "step": 1565
    },
    {
      "epoch": 0.8228511530398323,
      "grad_norm": 0.15628600120544434,
      "learning_rate": 0.0004111635220125786,
      "loss": 0.5998,
      "num_input_tokens_seen": 1018240,
      "step": 1570
    },
    {
      "epoch": 0.8254716981132075,
      "grad_norm": 0.08414579927921295,
      "learning_rate": 0.00041247379454926623,
      "loss": 0.503,
      "num_input_tokens_seen": 1022336,
      "step": 1575
    },
    {
      "epoch": 0.8280922431865828,
      "grad_norm": 0.23803626000881195,
      "learning_rate": 0.0004137840670859539,
      "loss": 0.5135,
      "num_input_tokens_seen": 1025184,
      "step": 1580
    },
    {
      "epoch": 0.8307127882599581,
      "grad_norm": 0.3339661955833435,
      "learning_rate": 0.00041509433962264155,
      "loss": 0.5793,
      "num_input_tokens_seen": 1028448,
      "step": 1585
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.26734867691993713,
      "learning_rate": 0.00041640461215932913,
      "loss": 0.5545,
      "num_input_tokens_seen": 1031072,
      "step": 1590
    },
    {
      "epoch": 0.8359538784067087,
      "grad_norm": 0.17525891959667206,
      "learning_rate": 0.00041771488469601677,
      "loss": 0.476,
      "num_input_tokens_seen": 1034208,
      "step": 1595
    },
    {
      "epoch": 0.8385744234800838,
      "grad_norm": 0.22810140252113342,
      "learning_rate": 0.0004190251572327044,
      "loss": 0.4526,
      "num_input_tokens_seen": 1038624,
      "step": 1600
    },
    {
      "epoch": 0.8411949685534591,
      "grad_norm": 0.1563606709241867,
      "learning_rate": 0.00042033542976939204,
      "loss": 0.5629,
      "num_input_tokens_seen": 1041152,
      "step": 1605
    },
    {
      "epoch": 0.8438155136268344,
      "grad_norm": 0.15499696135520935,
      "learning_rate": 0.00042164570230607967,
      "loss": 0.4165,
      "num_input_tokens_seen": 1043904,
      "step": 1610
    },
    {
      "epoch": 0.8464360587002097,
      "grad_norm": 0.12436691671609879,
      "learning_rate": 0.0004229559748427673,
      "loss": 0.5203,
      "num_input_tokens_seen": 1047456,
      "step": 1615
    },
    {
      "epoch": 0.8490566037735849,
      "grad_norm": 0.16743552684783936,
      "learning_rate": 0.00042426624737945494,
      "loss": 0.5663,
      "num_input_tokens_seen": 1050304,
      "step": 1620
    },
    {
      "epoch": 0.8516771488469602,
      "grad_norm": 0.15623731911182404,
      "learning_rate": 0.0004255765199161425,
      "loss": 0.5146,
      "num_input_tokens_seen": 1053600,
      "step": 1625
    },
    {
      "epoch": 0.8542976939203354,
      "grad_norm": 0.19614475965499878,
      "learning_rate": 0.0004268867924528302,
      "loss": 0.3679,
      "num_input_tokens_seen": 1056672,
      "step": 1630
    },
    {
      "epoch": 0.8569182389937107,
      "grad_norm": 0.22827044129371643,
      "learning_rate": 0.00042819706498951784,
      "loss": 0.3992,
      "num_input_tokens_seen": 1059904,
      "step": 1635
    },
    {
      "epoch": 0.859538784067086,
      "grad_norm": 0.1400059461593628,
      "learning_rate": 0.0004295073375262055,
      "loss": 0.5048,
      "num_input_tokens_seen": 1063360,
      "step": 1640
    },
    {
      "epoch": 0.8621593291404612,
      "grad_norm": 0.22172528505325317,
      "learning_rate": 0.0004308176100628931,
      "loss": 0.508,
      "num_input_tokens_seen": 1065728,
      "step": 1645
    },
    {
      "epoch": 0.8647798742138365,
      "grad_norm": 0.17494088411331177,
      "learning_rate": 0.00043212788259958075,
      "loss": 0.481,
      "num_input_tokens_seen": 1068448,
      "step": 1650
    },
    {
      "epoch": 0.8674004192872118,
      "grad_norm": 0.19010473787784576,
      "learning_rate": 0.0004334381551362683,
      "loss": 0.5102,
      "num_input_tokens_seen": 1071104,
      "step": 1655
    },
    {
      "epoch": 0.870020964360587,
      "grad_norm": 0.25421711802482605,
      "learning_rate": 0.00043474842767295596,
      "loss": 0.4178,
      "num_input_tokens_seen": 1075072,
      "step": 1660
    },
    {
      "epoch": 0.8726415094339622,
      "grad_norm": 0.15948368608951569,
      "learning_rate": 0.0004360587002096436,
      "loss": 0.5128,
      "num_input_tokens_seen": 1077824,
      "step": 1665
    },
    {
      "epoch": 0.8752620545073375,
      "grad_norm": 0.18929976224899292,
      "learning_rate": 0.00043736897274633123,
      "loss": 0.4771,
      "num_input_tokens_seen": 1080832,
      "step": 1670
    },
    {
      "epoch": 0.8778825995807128,
      "grad_norm": 0.20844584703445435,
      "learning_rate": 0.00043867924528301886,
      "loss": 0.543,
      "num_input_tokens_seen": 1084096,
      "step": 1675
    },
    {
      "epoch": 0.8805031446540881,
      "grad_norm": 0.17855149507522583,
      "learning_rate": 0.00043998951781970655,
      "loss": 0.5159,
      "num_input_tokens_seen": 1087104,
      "step": 1680
    },
    {
      "epoch": 0.8831236897274634,
      "grad_norm": 0.15844635665416718,
      "learning_rate": 0.00044129979035639413,
      "loss": 0.5715,
      "num_input_tokens_seen": 1090304,
      "step": 1685
    },
    {
      "epoch": 0.8857442348008385,
      "grad_norm": 0.23449528217315674,
      "learning_rate": 0.00044261006289308177,
      "loss": 0.7883,
      "num_input_tokens_seen": 1094112,
      "step": 1690
    },
    {
      "epoch": 0.8883647798742138,
      "grad_norm": 0.147230863571167,
      "learning_rate": 0.0004439203354297694,
      "loss": 0.5333,
      "num_input_tokens_seen": 1098496,
      "step": 1695
    },
    {
      "epoch": 0.8909853249475891,
      "grad_norm": 0.1603125035762787,
      "learning_rate": 0.00044523060796645704,
      "loss": 0.4503,
      "num_input_tokens_seen": 1100992,
      "step": 1700
    },
    {
      "epoch": 0.8936058700209644,
      "grad_norm": 0.2007765769958496,
      "learning_rate": 0.00044654088050314467,
      "loss": 0.5451,
      "num_input_tokens_seen": 1103776,
      "step": 1705
    },
    {
      "epoch": 0.8962264150943396,
      "grad_norm": 0.2120979279279709,
      "learning_rate": 0.0004478511530398323,
      "loss": 0.452,
      "num_input_tokens_seen": 1106368,
      "step": 1710
    },
    {
      "epoch": 0.8988469601677149,
      "grad_norm": 0.18852809071540833,
      "learning_rate": 0.0004491614255765199,
      "loss": 0.4424,
      "num_input_tokens_seen": 1109376,
      "step": 1715
    },
    {
      "epoch": 0.9014675052410901,
      "grad_norm": 0.1265508532524109,
      "learning_rate": 0.0004504716981132075,
      "loss": 0.4653,
      "num_input_tokens_seen": 1112448,
      "step": 1720
    },
    {
      "epoch": 0.9040880503144654,
      "grad_norm": 0.099153071641922,
      "learning_rate": 0.00045178197064989515,
      "loss": 0.4825,
      "num_input_tokens_seen": 1115520,
      "step": 1725
    },
    {
      "epoch": 0.9067085953878407,
      "grad_norm": 0.23374107480049133,
      "learning_rate": 0.00045309224318658284,
      "loss": 0.6451,
      "num_input_tokens_seen": 1118784,
      "step": 1730
    },
    {
      "epoch": 0.9093291404612159,
      "grad_norm": 0.5880999565124512,
      "learning_rate": 0.0004544025157232705,
      "loss": 0.7135,
      "num_input_tokens_seen": 1122112,
      "step": 1735
    },
    {
      "epoch": 0.9119496855345912,
      "grad_norm": 0.1722518503665924,
      "learning_rate": 0.0004557127882599581,
      "loss": 0.4702,
      "num_input_tokens_seen": 1125952,
      "step": 1740
    },
    {
      "epoch": 0.9145702306079665,
      "grad_norm": 0.3001502454280853,
      "learning_rate": 0.0004570230607966457,
      "loss": 0.657,
      "num_input_tokens_seen": 1129024,
      "step": 1745
    },
    {
      "epoch": 0.9171907756813418,
      "grad_norm": 0.20568612217903137,
      "learning_rate": 0.0004583333333333333,
      "loss": 0.4666,
      "num_input_tokens_seen": 1131488,
      "step": 1750
    },
    {
      "epoch": 0.9198113207547169,
      "grad_norm": 0.2138822078704834,
      "learning_rate": 0.00045964360587002096,
      "loss": 0.4721,
      "num_input_tokens_seen": 1134176,
      "step": 1755
    },
    {
      "epoch": 0.9224318658280922,
      "grad_norm": 0.0828217938542366,
      "learning_rate": 0.0004609538784067086,
      "loss": 0.5269,
      "num_input_tokens_seen": 1137888,
      "step": 1760
    },
    {
      "epoch": 0.9250524109014675,
      "grad_norm": 0.43347835540771484,
      "learning_rate": 0.00046226415094339623,
      "loss": 0.615,
      "num_input_tokens_seen": 1141184,
      "step": 1765
    },
    {
      "epoch": 0.9276729559748428,
      "grad_norm": 0.3117639124393463,
      "learning_rate": 0.00046357442348008386,
      "loss": 0.5396,
      "num_input_tokens_seen": 1144064,
      "step": 1770
    },
    {
      "epoch": 0.9302935010482181,
      "grad_norm": 0.32724037766456604,
      "learning_rate": 0.00046488469601677155,
      "loss": 0.5017,
      "num_input_tokens_seen": 1147584,
      "step": 1775
    },
    {
      "epoch": 0.9329140461215933,
      "grad_norm": 0.08151181787252426,
      "learning_rate": 0.00046619496855345913,
      "loss": 0.4822,
      "num_input_tokens_seen": 1154784,
      "step": 1780
    },
    {
      "epoch": 0.9355345911949685,
      "grad_norm": 0.40098288655281067,
      "learning_rate": 0.00046750524109014677,
      "loss": 0.4725,
      "num_input_tokens_seen": 1157440,
      "step": 1785
    },
    {
      "epoch": 0.9381551362683438,
      "grad_norm": 0.0957418754696846,
      "learning_rate": 0.0004688155136268344,
      "loss": 0.6405,
      "num_input_tokens_seen": 1160224,
      "step": 1790
    },
    {
      "epoch": 0.9407756813417191,
      "grad_norm": 0.25346237421035767,
      "learning_rate": 0.00047012578616352203,
      "loss": 0.5172,
      "num_input_tokens_seen": 1164480,
      "step": 1795
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 0.2394961714744568,
      "learning_rate": 0.00047143605870020967,
      "loss": 0.3977,
      "num_input_tokens_seen": 1168704,
      "step": 1800
    },
    {
      "epoch": 0.9460167714884696,
      "grad_norm": 0.16837278008460999,
      "learning_rate": 0.0004727463312368973,
      "loss": 0.4354,
      "num_input_tokens_seen": 1172384,
      "step": 1805
    },
    {
      "epoch": 0.9486373165618449,
      "grad_norm": 0.18241256475448608,
      "learning_rate": 0.0004740566037735849,
      "loss": 0.4943,
      "num_input_tokens_seen": 1175392,
      "step": 1810
    },
    {
      "epoch": 0.9512578616352201,
      "grad_norm": 0.15058289468288422,
      "learning_rate": 0.0004753668763102725,
      "loss": 0.4631,
      "num_input_tokens_seen": 1178368,
      "step": 1815
    },
    {
      "epoch": 0.9538784067085954,
      "grad_norm": 0.20277461409568787,
      "learning_rate": 0.00047667714884696015,
      "loss": 0.4071,
      "num_input_tokens_seen": 1182240,
      "step": 1820
    },
    {
      "epoch": 0.9564989517819706,
      "grad_norm": 0.12284812331199646,
      "learning_rate": 0.00047798742138364784,
      "loss": 0.4638,
      "num_input_tokens_seen": 1185504,
      "step": 1825
    },
    {
      "epoch": 0.9591194968553459,
      "grad_norm": 0.26939207315444946,
      "learning_rate": 0.0004792976939203355,
      "loss": 0.4824,
      "num_input_tokens_seen": 1188288,
      "step": 1830
    },
    {
      "epoch": 0.9617400419287212,
      "grad_norm": 0.17868545651435852,
      "learning_rate": 0.0004806079664570231,
      "loss": 0.457,
      "num_input_tokens_seen": 1192064,
      "step": 1835
    },
    {
      "epoch": 0.9643605870020965,
      "grad_norm": 0.11155866831541061,
      "learning_rate": 0.0004819182389937107,
      "loss": 0.461,
      "num_input_tokens_seen": 1195104,
      "step": 1840
    },
    {
      "epoch": 0.9669811320754716,
      "grad_norm": 0.17326685786247253,
      "learning_rate": 0.0004832285115303983,
      "loss": 0.6006,
      "num_input_tokens_seen": 1197920,
      "step": 1845
    },
    {
      "epoch": 0.9696016771488469,
      "grad_norm": 0.15923936665058136,
      "learning_rate": 0.00048453878406708596,
      "loss": 0.4591,
      "num_input_tokens_seen": 1200896,
      "step": 1850
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 0.1824704110622406,
      "learning_rate": 0.0004858490566037736,
      "loss": 0.4771,
      "num_input_tokens_seen": 1204032,
      "step": 1855
    },
    {
      "epoch": 0.9748427672955975,
      "grad_norm": 0.13187338411808014,
      "learning_rate": 0.0004871593291404612,
      "loss": 0.4047,
      "num_input_tokens_seen": 1207264,
      "step": 1860
    },
    {
      "epoch": 0.9774633123689728,
      "grad_norm": 0.23110811412334442,
      "learning_rate": 0.0004884696016771489,
      "loss": 0.5478,
      "num_input_tokens_seen": 1209984,
      "step": 1865
    },
    {
      "epoch": 0.980083857442348,
      "grad_norm": 0.2048482596874237,
      "learning_rate": 0.0004897798742138365,
      "loss": 0.4362,
      "num_input_tokens_seen": 1213856,
      "step": 1870
    },
    {
      "epoch": 0.9827044025157232,
      "grad_norm": 0.11667975783348083,
      "learning_rate": 0.0004910901467505241,
      "loss": 0.4379,
      "num_input_tokens_seen": 1217408,
      "step": 1875
    },
    {
      "epoch": 0.9853249475890985,
      "grad_norm": 0.12775109708309174,
      "learning_rate": 0.0004924004192872118,
      "loss": 0.5103,
      "num_input_tokens_seen": 1221024,
      "step": 1880
    },
    {
      "epoch": 0.9879454926624738,
      "grad_norm": 0.17716601490974426,
      "learning_rate": 0.0004937106918238993,
      "loss": 0.4078,
      "num_input_tokens_seen": 1224320,
      "step": 1885
    },
    {
      "epoch": 0.9905660377358491,
      "grad_norm": 0.1361325979232788,
      "learning_rate": 0.000495020964360587,
      "loss": 0.3956,
      "num_input_tokens_seen": 1228000,
      "step": 1890
    },
    {
      "epoch": 0.9931865828092243,
      "grad_norm": 0.20175008475780487,
      "learning_rate": 0.0004963312368972746,
      "loss": 0.4922,
      "num_input_tokens_seen": 1232320,
      "step": 1895
    },
    {
      "epoch": 0.9958071278825996,
      "grad_norm": 0.17874129116535187,
      "learning_rate": 0.0004976415094339623,
      "loss": 0.5838,
      "num_input_tokens_seen": 1238272,
      "step": 1900
    },
    {
      "epoch": 0.9984276729559748,
      "grad_norm": 0.32082101702690125,
      "learning_rate": 0.0004989517819706499,
      "loss": 0.4991,
      "num_input_tokens_seen": 1241344,
      "step": 1905
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.47990021109580994,
      "eval_runtime": 13.5377,
      "eval_samples_per_second": 62.64,
      "eval_steps_per_second": 15.66,
      "num_input_tokens_seen": 1243048,
      "step": 1908
    },
    {
      "epoch": 1.00104821802935,
      "grad_norm": 0.20935001969337463,
      "learning_rate": 0.0005002620545073376,
      "loss": 0.4387,
      "num_input_tokens_seen": 1244520,
      "step": 1910
    },
    {
      "epoch": 1.0036687631027255,
      "grad_norm": 0.18767164647579193,
      "learning_rate": 0.0005015723270440253,
      "loss": 0.4998,
      "num_input_tokens_seen": 1247528,
      "step": 1915
    },
    {
      "epoch": 1.0062893081761006,
      "grad_norm": 0.12031282484531403,
      "learning_rate": 0.0005028825995807128,
      "loss": 0.5969,
      "num_input_tokens_seen": 1250824,
      "step": 1920
    },
    {
      "epoch": 1.0089098532494758,
      "grad_norm": 0.19817157089710236,
      "learning_rate": 0.0005041928721174004,
      "loss": 0.5149,
      "num_input_tokens_seen": 1254024,
      "step": 1925
    },
    {
      "epoch": 1.0115303983228512,
      "grad_norm": 0.19522832334041595,
      "learning_rate": 0.0005055031446540881,
      "loss": 0.4696,
      "num_input_tokens_seen": 1260328,
      "step": 1930
    },
    {
      "epoch": 1.0141509433962264,
      "grad_norm": 0.22197531163692474,
      "learning_rate": 0.0005068134171907757,
      "loss": 0.6895,
      "num_input_tokens_seen": 1263656,
      "step": 1935
    },
    {
      "epoch": 1.0167714884696017,
      "grad_norm": 0.16695758700370789,
      "learning_rate": 0.0005081236897274634,
      "loss": 0.5358,
      "num_input_tokens_seen": 1267528,
      "step": 1940
    },
    {
      "epoch": 1.019392033542977,
      "grad_norm": 0.29449930787086487,
      "learning_rate": 0.000509433962264151,
      "loss": 0.6001,
      "num_input_tokens_seen": 1270536,
      "step": 1945
    },
    {
      "epoch": 1.0220125786163523,
      "grad_norm": 0.1664009541273117,
      "learning_rate": 0.0005107442348008385,
      "loss": 0.399,
      "num_input_tokens_seen": 1273416,
      "step": 1950
    },
    {
      "epoch": 1.0246331236897275,
      "grad_norm": 0.21839790046215057,
      "learning_rate": 0.0005120545073375262,
      "loss": 0.375,
      "num_input_tokens_seen": 1276232,
      "step": 1955
    },
    {
      "epoch": 1.0272536687631026,
      "grad_norm": 0.10716330260038376,
      "learning_rate": 0.0005133647798742138,
      "loss": 0.402,
      "num_input_tokens_seen": 1279400,
      "step": 1960
    },
    {
      "epoch": 1.029874213836478,
      "grad_norm": 0.1333862990140915,
      "learning_rate": 0.0005146750524109015,
      "loss": 0.4461,
      "num_input_tokens_seen": 1283752,
      "step": 1965
    },
    {
      "epoch": 1.0324947589098532,
      "grad_norm": 0.15896683931350708,
      "learning_rate": 0.0005159853249475891,
      "loss": 0.4364,
      "num_input_tokens_seen": 1287048,
      "step": 1970
    },
    {
      "epoch": 1.0351153039832286,
      "grad_norm": 0.12141911685466766,
      "learning_rate": 0.0005172955974842768,
      "loss": 0.5042,
      "num_input_tokens_seen": 1289544,
      "step": 1975
    },
    {
      "epoch": 1.0377358490566038,
      "grad_norm": 0.17657220363616943,
      "learning_rate": 0.0005186058700209643,
      "loss": 0.4788,
      "num_input_tokens_seen": 1292392,
      "step": 1980
    },
    {
      "epoch": 1.040356394129979,
      "grad_norm": 0.19084644317626953,
      "learning_rate": 0.0005199161425576519,
      "loss": 0.5652,
      "num_input_tokens_seen": 1295400,
      "step": 1985
    },
    {
      "epoch": 1.0429769392033543,
      "grad_norm": 0.12087299674749374,
      "learning_rate": 0.0005212264150943396,
      "loss": 0.5139,
      "num_input_tokens_seen": 1298152,
      "step": 1990
    },
    {
      "epoch": 1.0455974842767295,
      "grad_norm": 0.20267285406589508,
      "learning_rate": 0.0005225366876310272,
      "loss": 0.5943,
      "num_input_tokens_seen": 1301384,
      "step": 1995
    },
    {
      "epoch": 1.0482180293501049,
      "grad_norm": 0.2292216569185257,
      "learning_rate": 0.000523846960167715,
      "loss": 0.5438,
      "num_input_tokens_seen": 1304296,
      "step": 2000
    },
    {
      "epoch": 1.05083857442348,
      "grad_norm": 0.15236414968967438,
      "learning_rate": 0.0005251572327044026,
      "loss": 0.5838,
      "num_input_tokens_seen": 1307048,
      "step": 2005
    },
    {
      "epoch": 1.0534591194968554,
      "grad_norm": 0.12364363670349121,
      "learning_rate": 0.0005264675052410901,
      "loss": 0.4636,
      "num_input_tokens_seen": 1310792,
      "step": 2010
    },
    {
      "epoch": 1.0560796645702306,
      "grad_norm": 0.1711197793483734,
      "learning_rate": 0.0005277777777777778,
      "loss": 0.6042,
      "num_input_tokens_seen": 1313448,
      "step": 2015
    },
    {
      "epoch": 1.0587002096436058,
      "grad_norm": 0.24530576169490814,
      "learning_rate": 0.0005290880503144654,
      "loss": 0.4096,
      "num_input_tokens_seen": 1316200,
      "step": 2020
    },
    {
      "epoch": 1.0613207547169812,
      "grad_norm": 0.12846577167510986,
      "learning_rate": 0.0005303983228511531,
      "loss": 0.7931,
      "num_input_tokens_seen": 1319720,
      "step": 2025
    },
    {
      "epoch": 1.0639412997903563,
      "grad_norm": 0.28097450733184814,
      "learning_rate": 0.0005317085953878407,
      "loss": 0.5444,
      "num_input_tokens_seen": 1323432,
      "step": 2030
    },
    {
      "epoch": 1.0665618448637317,
      "grad_norm": 0.13446180522441864,
      "learning_rate": 0.0005330188679245284,
      "loss": 0.4949,
      "num_input_tokens_seen": 1326760,
      "step": 2035
    },
    {
      "epoch": 1.069182389937107,
      "grad_norm": 0.15264663100242615,
      "learning_rate": 0.000534329140461216,
      "loss": 0.4619,
      "num_input_tokens_seen": 1330120,
      "step": 2040
    },
    {
      "epoch": 1.0718029350104823,
      "grad_norm": 0.15428535640239716,
      "learning_rate": 0.0005356394129979035,
      "loss": 0.5418,
      "num_input_tokens_seen": 1333032,
      "step": 2045
    },
    {
      "epoch": 1.0744234800838575,
      "grad_norm": 0.27598488330841064,
      "learning_rate": 0.0005369496855345912,
      "loss": 0.5808,
      "num_input_tokens_seen": 1336296,
      "step": 2050
    },
    {
      "epoch": 1.0770440251572326,
      "grad_norm": 0.08560232818126678,
      "learning_rate": 0.0005382599580712788,
      "loss": 0.3735,
      "num_input_tokens_seen": 1339592,
      "step": 2055
    },
    {
      "epoch": 1.079664570230608,
      "grad_norm": 0.12769083678722382,
      "learning_rate": 0.0005395702306079665,
      "loss": 0.4939,
      "num_input_tokens_seen": 1342664,
      "step": 2060
    },
    {
      "epoch": 1.0822851153039832,
      "grad_norm": 0.13877876102924347,
      "learning_rate": 0.0005408805031446541,
      "loss": 0.5871,
      "num_input_tokens_seen": 1345288,
      "step": 2065
    },
    {
      "epoch": 1.0849056603773586,
      "grad_norm": 0.12640418112277985,
      "learning_rate": 0.0005421907756813418,
      "loss": 0.4675,
      "num_input_tokens_seen": 1348488,
      "step": 2070
    },
    {
      "epoch": 1.0875262054507338,
      "grad_norm": 0.1636268049478531,
      "learning_rate": 0.0005435010482180293,
      "loss": 0.6155,
      "num_input_tokens_seen": 1352712,
      "step": 2075
    },
    {
      "epoch": 1.090146750524109,
      "grad_norm": 0.10503066331148148,
      "learning_rate": 0.0005448113207547169,
      "loss": 0.4465,
      "num_input_tokens_seen": 1355496,
      "step": 2080
    },
    {
      "epoch": 1.0927672955974843,
      "grad_norm": 0.08051334321498871,
      "learning_rate": 0.0005461215932914046,
      "loss": 0.52,
      "num_input_tokens_seen": 1360552,
      "step": 2085
    },
    {
      "epoch": 1.0953878406708595,
      "grad_norm": 0.12718378007411957,
      "learning_rate": 0.0005474318658280922,
      "loss": 0.428,
      "num_input_tokens_seen": 1363432,
      "step": 2090
    },
    {
      "epoch": 1.0980083857442349,
      "grad_norm": 0.1769132763147354,
      "learning_rate": 0.0005487421383647799,
      "loss": 0.3868,
      "num_input_tokens_seen": 1366824,
      "step": 2095
    },
    {
      "epoch": 1.10062893081761,
      "grad_norm": 0.2622573673725128,
      "learning_rate": 0.0005500524109014676,
      "loss": 0.5607,
      "num_input_tokens_seen": 1370312,
      "step": 2100
    },
    {
      "epoch": 1.1032494758909852,
      "grad_norm": 0.2028270810842514,
      "learning_rate": 0.0005513626834381551,
      "loss": 0.7055,
      "num_input_tokens_seen": 1373160,
      "step": 2105
    },
    {
      "epoch": 1.1058700209643606,
      "grad_norm": 0.25572049617767334,
      "learning_rate": 0.0005526729559748428,
      "loss": 0.5473,
      "num_input_tokens_seen": 1376008,
      "step": 2110
    },
    {
      "epoch": 1.1084905660377358,
      "grad_norm": 0.16920681297779083,
      "learning_rate": 0.0005539832285115304,
      "loss": 0.5465,
      "num_input_tokens_seen": 1378440,
      "step": 2115
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.10549373924732208,
      "learning_rate": 0.0005552935010482181,
      "loss": 0.5314,
      "num_input_tokens_seen": 1382152,
      "step": 2120
    },
    {
      "epoch": 1.1137316561844863,
      "grad_norm": 0.14079216122627258,
      "learning_rate": 0.0005566037735849057,
      "loss": 0.6587,
      "num_input_tokens_seen": 1385416,
      "step": 2125
    },
    {
      "epoch": 1.1163522012578617,
      "grad_norm": 0.30529147386550903,
      "learning_rate": 0.0005579140461215934,
      "loss": 0.4586,
      "num_input_tokens_seen": 1389288,
      "step": 2130
    },
    {
      "epoch": 1.118972746331237,
      "grad_norm": 0.17221766710281372,
      "learning_rate": 0.000559224318658281,
      "loss": 0.5157,
      "num_input_tokens_seen": 1392136,
      "step": 2135
    },
    {
      "epoch": 1.121593291404612,
      "grad_norm": 0.13909631967544556,
      "learning_rate": 0.0005605345911949685,
      "loss": 0.4965,
      "num_input_tokens_seen": 1395208,
      "step": 2140
    },
    {
      "epoch": 1.1242138364779874,
      "grad_norm": 0.16987499594688416,
      "learning_rate": 0.0005618448637316562,
      "loss": 0.4232,
      "num_input_tokens_seen": 1398024,
      "step": 2145
    },
    {
      "epoch": 1.1268343815513626,
      "grad_norm": 0.11385642737150192,
      "learning_rate": 0.0005631551362683438,
      "loss": 0.4574,
      "num_input_tokens_seen": 1401256,
      "step": 2150
    },
    {
      "epoch": 1.129454926624738,
      "grad_norm": 0.18146328628063202,
      "learning_rate": 0.0005644654088050315,
      "loss": 0.3676,
      "num_input_tokens_seen": 1403848,
      "step": 2155
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 0.19240796566009521,
      "learning_rate": 0.0005657756813417191,
      "loss": 0.5891,
      "num_input_tokens_seen": 1408040,
      "step": 2160
    },
    {
      "epoch": 1.1346960167714886,
      "grad_norm": 0.2889261543750763,
      "learning_rate": 0.0005670859538784067,
      "loss": 0.6175,
      "num_input_tokens_seen": 1411688,
      "step": 2165
    },
    {
      "epoch": 1.1373165618448637,
      "grad_norm": 0.14031274616718292,
      "learning_rate": 0.0005683962264150943,
      "loss": 0.6036,
      "num_input_tokens_seen": 1415464,
      "step": 2170
    },
    {
      "epoch": 1.139937106918239,
      "grad_norm": 0.10512494295835495,
      "learning_rate": 0.0005697064989517819,
      "loss": 0.4133,
      "num_input_tokens_seen": 1418664,
      "step": 2175
    },
    {
      "epoch": 1.1425576519916143,
      "grad_norm": 0.14247387647628784,
      "learning_rate": 0.0005710167714884696,
      "loss": 0.4336,
      "num_input_tokens_seen": 1422088,
      "step": 2180
    },
    {
      "epoch": 1.1451781970649895,
      "grad_norm": 0.1935369223356247,
      "learning_rate": 0.0005723270440251572,
      "loss": 0.4393,
      "num_input_tokens_seen": 1424872,
      "step": 2185
    },
    {
      "epoch": 1.1477987421383649,
      "grad_norm": 0.10872030258178711,
      "learning_rate": 0.0005736373165618449,
      "loss": 0.6248,
      "num_input_tokens_seen": 1429320,
      "step": 2190
    },
    {
      "epoch": 1.15041928721174,
      "grad_norm": 0.09420914202928543,
      "learning_rate": 0.0005749475890985325,
      "loss": 0.4801,
      "num_input_tokens_seen": 1433704,
      "step": 2195
    },
    {
      "epoch": 1.1530398322851152,
      "grad_norm": 0.12686710059642792,
      "learning_rate": 0.0005762578616352201,
      "loss": 0.5103,
      "num_input_tokens_seen": 1436872,
      "step": 2200
    },
    {
      "epoch": 1.1556603773584906,
      "grad_norm": 0.29707685112953186,
      "learning_rate": 0.0005775681341719078,
      "loss": 0.4957,
      "num_input_tokens_seen": 1440104,
      "step": 2205
    },
    {
      "epoch": 1.1582809224318658,
      "grad_norm": 0.12571631371974945,
      "learning_rate": 0.0005788784067085954,
      "loss": 0.6599,
      "num_input_tokens_seen": 1443176,
      "step": 2210
    },
    {
      "epoch": 1.1609014675052411,
      "grad_norm": 0.10530340671539307,
      "learning_rate": 0.0005801886792452831,
      "loss": 0.4622,
      "num_input_tokens_seen": 1446440,
      "step": 2215
    },
    {
      "epoch": 1.1635220125786163,
      "grad_norm": 0.273885577917099,
      "learning_rate": 0.0005814989517819707,
      "loss": 0.4721,
      "num_input_tokens_seen": 1450120,
      "step": 2220
    },
    {
      "epoch": 1.1661425576519917,
      "grad_norm": 0.14458487927913666,
      "learning_rate": 0.0005828092243186583,
      "loss": 0.4728,
      "num_input_tokens_seen": 1453608,
      "step": 2225
    },
    {
      "epoch": 1.1687631027253669,
      "grad_norm": 0.1297215223312378,
      "learning_rate": 0.000584119496855346,
      "loss": 0.6287,
      "num_input_tokens_seen": 1457832,
      "step": 2230
    },
    {
      "epoch": 1.171383647798742,
      "grad_norm": 0.12850576639175415,
      "learning_rate": 0.0005854297693920335,
      "loss": 0.543,
      "num_input_tokens_seen": 1462280,
      "step": 2235
    },
    {
      "epoch": 1.1740041928721174,
      "grad_norm": 0.13934198021888733,
      "learning_rate": 0.0005867400419287212,
      "loss": 0.4937,
      "num_input_tokens_seen": 1466024,
      "step": 2240
    },
    {
      "epoch": 1.1766247379454926,
      "grad_norm": 0.12378193438053131,
      "learning_rate": 0.0005880503144654088,
      "loss": 0.4739,
      "num_input_tokens_seen": 1468872,
      "step": 2245
    },
    {
      "epoch": 1.179245283018868,
      "grad_norm": 0.18240033090114594,
      "learning_rate": 0.0005893605870020965,
      "loss": 0.4959,
      "num_input_tokens_seen": 1471336,
      "step": 2250
    },
    {
      "epoch": 1.1818658280922432,
      "grad_norm": 0.20081904530525208,
      "learning_rate": 0.0005906708595387841,
      "loss": 0.6223,
      "num_input_tokens_seen": 1473928,
      "step": 2255
    },
    {
      "epoch": 1.1844863731656186,
      "grad_norm": 0.1638210415840149,
      "learning_rate": 0.0005919811320754716,
      "loss": 0.428,
      "num_input_tokens_seen": 1478728,
      "step": 2260
    },
    {
      "epoch": 1.1871069182389937,
      "grad_norm": 0.14548270404338837,
      "learning_rate": 0.0005932914046121593,
      "loss": 0.4867,
      "num_input_tokens_seen": 1481576,
      "step": 2265
    },
    {
      "epoch": 1.189727463312369,
      "grad_norm": 0.18083396553993225,
      "learning_rate": 0.0005946016771488469,
      "loss": 0.5201,
      "num_input_tokens_seen": 1484104,
      "step": 2270
    },
    {
      "epoch": 1.1923480083857443,
      "grad_norm": 0.1858469843864441,
      "learning_rate": 0.0005959119496855346,
      "loss": 0.5939,
      "num_input_tokens_seen": 1487048,
      "step": 2275
    },
    {
      "epoch": 1.1949685534591195,
      "grad_norm": 0.11496371030807495,
      "learning_rate": 0.0005972222222222222,
      "loss": 0.4562,
      "num_input_tokens_seen": 1490152,
      "step": 2280
    },
    {
      "epoch": 1.1975890985324948,
      "grad_norm": 0.13072414696216583,
      "learning_rate": 0.0005985324947589099,
      "loss": 0.4346,
      "num_input_tokens_seen": 1492648,
      "step": 2285
    },
    {
      "epoch": 1.20020964360587,
      "grad_norm": 0.1829564869403839,
      "learning_rate": 0.0005998427672955975,
      "loss": 0.5848,
      "num_input_tokens_seen": 1495272,
      "step": 2290
    },
    {
      "epoch": 1.2028301886792452,
      "grad_norm": 0.12358658015727997,
      "learning_rate": 0.000601153039832285,
      "loss": 0.5072,
      "num_input_tokens_seen": 1497832,
      "step": 2295
    },
    {
      "epoch": 1.2054507337526206,
      "grad_norm": 0.15049874782562256,
      "learning_rate": 0.0006024633123689728,
      "loss": 0.4003,
      "num_input_tokens_seen": 1501896,
      "step": 2300
    },
    {
      "epoch": 1.2080712788259957,
      "grad_norm": 0.20930978655815125,
      "learning_rate": 0.0006037735849056604,
      "loss": 0.4761,
      "num_input_tokens_seen": 1504328,
      "step": 2305
    },
    {
      "epoch": 1.2106918238993711,
      "grad_norm": 0.14787644147872925,
      "learning_rate": 0.0006050838574423481,
      "loss": 0.5255,
      "num_input_tokens_seen": 1508264,
      "step": 2310
    },
    {
      "epoch": 1.2133123689727463,
      "grad_norm": 0.1683129519224167,
      "learning_rate": 0.0006063941299790357,
      "loss": 0.5938,
      "num_input_tokens_seen": 1511048,
      "step": 2315
    },
    {
      "epoch": 1.2159329140461215,
      "grad_norm": 0.21464566886425018,
      "learning_rate": 0.0006077044025157233,
      "loss": 0.679,
      "num_input_tokens_seen": 1513928,
      "step": 2320
    },
    {
      "epoch": 1.2185534591194969,
      "grad_norm": 0.11492037773132324,
      "learning_rate": 0.000609014675052411,
      "loss": 0.4412,
      "num_input_tokens_seen": 1517512,
      "step": 2325
    },
    {
      "epoch": 1.221174004192872,
      "grad_norm": 0.20094537734985352,
      "learning_rate": 0.0006103249475890985,
      "loss": 0.391,
      "num_input_tokens_seen": 1521256,
      "step": 2330
    },
    {
      "epoch": 1.2237945492662474,
      "grad_norm": 0.2168809473514557,
      "learning_rate": 0.0006116352201257862,
      "loss": 0.6336,
      "num_input_tokens_seen": 1523144,
      "step": 2335
    },
    {
      "epoch": 1.2264150943396226,
      "grad_norm": 0.17274968326091766,
      "learning_rate": 0.0006129454926624738,
      "loss": 0.5005,
      "num_input_tokens_seen": 1527624,
      "step": 2340
    },
    {
      "epoch": 1.229035639412998,
      "grad_norm": 0.13957679271697998,
      "learning_rate": 0.0006142557651991615,
      "loss": 0.4744,
      "num_input_tokens_seen": 1531560,
      "step": 2345
    },
    {
      "epoch": 1.2316561844863732,
      "grad_norm": 0.18725748360157013,
      "learning_rate": 0.0006155660377358491,
      "loss": 0.4903,
      "num_input_tokens_seen": 1533832,
      "step": 2350
    },
    {
      "epoch": 1.2342767295597485,
      "grad_norm": 0.09959086775779724,
      "learning_rate": 0.0006168763102725366,
      "loss": 0.5364,
      "num_input_tokens_seen": 1536584,
      "step": 2355
    },
    {
      "epoch": 1.2368972746331237,
      "grad_norm": 0.19480255246162415,
      "learning_rate": 0.0006181865828092243,
      "loss": 0.4857,
      "num_input_tokens_seen": 1539720,
      "step": 2360
    },
    {
      "epoch": 1.2395178197064989,
      "grad_norm": 0.11603730171918869,
      "learning_rate": 0.0006194968553459119,
      "loss": 0.3794,
      "num_input_tokens_seen": 1543112,
      "step": 2365
    },
    {
      "epoch": 1.2421383647798743,
      "grad_norm": 0.16428713500499725,
      "learning_rate": 0.0006208071278825996,
      "loss": 0.519,
      "num_input_tokens_seen": 1546568,
      "step": 2370
    },
    {
      "epoch": 1.2447589098532494,
      "grad_norm": 0.2394178956747055,
      "learning_rate": 0.0006221174004192872,
      "loss": 0.5325,
      "num_input_tokens_seen": 1549864,
      "step": 2375
    },
    {
      "epoch": 1.2473794549266248,
      "grad_norm": 0.11512283980846405,
      "learning_rate": 0.0006234276729559748,
      "loss": 0.4411,
      "num_input_tokens_seen": 1552456,
      "step": 2380
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.3895910382270813,
      "learning_rate": 0.0006247379454926625,
      "loss": 0.5765,
      "num_input_tokens_seen": 1555240,
      "step": 2385
    },
    {
      "epoch": 1.2526205450733752,
      "grad_norm": 0.13309869170188904,
      "learning_rate": 0.00062604821802935,
      "loss": 0.4217,
      "num_input_tokens_seen": 1558568,
      "step": 2390
    },
    {
      "epoch": 1.2552410901467506,
      "grad_norm": 0.16216053068637848,
      "learning_rate": 0.0006273584905660377,
      "loss": 0.546,
      "num_input_tokens_seen": 1561288,
      "step": 2395
    },
    {
      "epoch": 1.2578616352201257,
      "grad_norm": 0.22484998404979706,
      "learning_rate": 0.0006286687631027254,
      "loss": 0.5648,
      "num_input_tokens_seen": 1564200,
      "step": 2400
    },
    {
      "epoch": 1.2604821802935011,
      "grad_norm": 0.11463344842195511,
      "learning_rate": 0.0006299790356394131,
      "loss": 0.5169,
      "num_input_tokens_seen": 1567208,
      "step": 2405
    },
    {
      "epoch": 1.2631027253668763,
      "grad_norm": 0.10659648478031158,
      "learning_rate": 0.0006312893081761007,
      "loss": 0.4424,
      "num_input_tokens_seen": 1571240,
      "step": 2410
    },
    {
      "epoch": 1.2657232704402515,
      "grad_norm": 0.270296573638916,
      "learning_rate": 0.0006325995807127883,
      "loss": 0.504,
      "num_input_tokens_seen": 1574376,
      "step": 2415
    },
    {
      "epoch": 1.2683438155136268,
      "grad_norm": 0.14634987711906433,
      "learning_rate": 0.000633909853249476,
      "loss": 0.5658,
      "num_input_tokens_seen": 1577672,
      "step": 2420
    },
    {
      "epoch": 1.270964360587002,
      "grad_norm": 0.11856269091367722,
      "learning_rate": 0.0006352201257861635,
      "loss": 0.4414,
      "num_input_tokens_seen": 1581192,
      "step": 2425
    },
    {
      "epoch": 1.2735849056603774,
      "grad_norm": 0.08228956162929535,
      "learning_rate": 0.0006365303983228512,
      "loss": 0.4289,
      "num_input_tokens_seen": 1584424,
      "step": 2430
    },
    {
      "epoch": 1.2762054507337526,
      "grad_norm": 0.16405542194843292,
      "learning_rate": 0.0006378406708595388,
      "loss": 0.5018,
      "num_input_tokens_seen": 1587912,
      "step": 2435
    },
    {
      "epoch": 1.2788259958071277,
      "grad_norm": 0.15245285630226135,
      "learning_rate": 0.0006391509433962265,
      "loss": 0.3716,
      "num_input_tokens_seen": 1591624,
      "step": 2440
    },
    {
      "epoch": 1.2814465408805031,
      "grad_norm": 0.09896109253168106,
      "learning_rate": 0.0006404612159329141,
      "loss": 0.5233,
      "num_input_tokens_seen": 1594376,
      "step": 2445
    },
    {
      "epoch": 1.2840670859538785,
      "grad_norm": 0.11186987161636353,
      "learning_rate": 0.0006417714884696016,
      "loss": 0.4619,
      "num_input_tokens_seen": 1597608,
      "step": 2450
    },
    {
      "epoch": 1.2866876310272537,
      "grad_norm": 0.15616416931152344,
      "learning_rate": 0.0006430817610062893,
      "loss": 0.4558,
      "num_input_tokens_seen": 1600136,
      "step": 2455
    },
    {
      "epoch": 1.2893081761006289,
      "grad_norm": 0.1307230144739151,
      "learning_rate": 0.0006443920335429769,
      "loss": 0.495,
      "num_input_tokens_seen": 1603592,
      "step": 2460
    },
    {
      "epoch": 1.2919287211740043,
      "grad_norm": 0.1795375496149063,
      "learning_rate": 0.0006457023060796646,
      "loss": 0.6001,
      "num_input_tokens_seen": 1606504,
      "step": 2465
    },
    {
      "epoch": 1.2945492662473794,
      "grad_norm": 0.1496136337518692,
      "learning_rate": 0.0006470125786163522,
      "loss": 0.4403,
      "num_input_tokens_seen": 1613064,
      "step": 2470
    },
    {
      "epoch": 1.2971698113207548,
      "grad_norm": 0.08532848209142685,
      "learning_rate": 0.0006483228511530398,
      "loss": 0.4773,
      "num_input_tokens_seen": 1616136,
      "step": 2475
    },
    {
      "epoch": 1.29979035639413,
      "grad_norm": 0.12766116857528687,
      "learning_rate": 0.0006496331236897275,
      "loss": 0.4508,
      "num_input_tokens_seen": 1618792,
      "step": 2480
    },
    {
      "epoch": 1.3024109014675052,
      "grad_norm": 0.2380463182926178,
      "learning_rate": 0.000650943396226415,
      "loss": 0.4108,
      "num_input_tokens_seen": 1621096,
      "step": 2485
    },
    {
      "epoch": 1.3050314465408805,
      "grad_norm": 0.18558098375797272,
      "learning_rate": 0.0006522536687631027,
      "loss": 0.4194,
      "num_input_tokens_seen": 1623720,
      "step": 2490
    },
    {
      "epoch": 1.3076519916142557,
      "grad_norm": 0.11616940051317215,
      "learning_rate": 0.0006535639412997903,
      "loss": 0.51,
      "num_input_tokens_seen": 1626888,
      "step": 2495
    },
    {
      "epoch": 1.310272536687631,
      "grad_norm": 0.1474447101354599,
      "learning_rate": 0.0006548742138364781,
      "loss": 0.6266,
      "num_input_tokens_seen": 1630472,
      "step": 2500
    },
    {
      "epoch": 1.3128930817610063,
      "grad_norm": 0.12293053418397903,
      "learning_rate": 0.0006561844863731657,
      "loss": 0.4654,
      "num_input_tokens_seen": 1633192,
      "step": 2505
    },
    {
      "epoch": 1.3155136268343814,
      "grad_norm": 0.06373423337936401,
      "learning_rate": 0.0006574947589098533,
      "loss": 0.5465,
      "num_input_tokens_seen": 1637128,
      "step": 2510
    },
    {
      "epoch": 1.3181341719077568,
      "grad_norm": 0.09693603962659836,
      "learning_rate": 0.000658805031446541,
      "loss": 0.5135,
      "num_input_tokens_seen": 1640456,
      "step": 2515
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 0.14934556186199188,
      "learning_rate": 0.0006601153039832285,
      "loss": 0.6134,
      "num_input_tokens_seen": 1644008,
      "step": 2520
    },
    {
      "epoch": 1.3233752620545074,
      "grad_norm": 0.1754673570394516,
      "learning_rate": 0.0006614255765199162,
      "loss": 0.4634,
      "num_input_tokens_seen": 1647720,
      "step": 2525
    },
    {
      "epoch": 1.3259958071278826,
      "grad_norm": 0.20162436366081238,
      "learning_rate": 0.0006627358490566038,
      "loss": 0.6304,
      "num_input_tokens_seen": 1650440,
      "step": 2530
    },
    {
      "epoch": 1.3286163522012577,
      "grad_norm": 0.13487380743026733,
      "learning_rate": 0.0006640461215932914,
      "loss": 0.5914,
      "num_input_tokens_seen": 1653352,
      "step": 2535
    },
    {
      "epoch": 1.3312368972746331,
      "grad_norm": 0.1155925765633583,
      "learning_rate": 0.0006653563941299791,
      "loss": 0.4623,
      "num_input_tokens_seen": 1656360,
      "step": 2540
    },
    {
      "epoch": 1.3338574423480085,
      "grad_norm": 0.09433197975158691,
      "learning_rate": 0.0006666666666666666,
      "loss": 0.6201,
      "num_input_tokens_seen": 1659336,
      "step": 2545
    },
    {
      "epoch": 1.3364779874213837,
      "grad_norm": 0.4473826289176941,
      "learning_rate": 0.0006679769392033543,
      "loss": 0.7107,
      "num_input_tokens_seen": 1661896,
      "step": 2550
    },
    {
      "epoch": 1.3390985324947589,
      "grad_norm": 0.06731697916984558,
      "learning_rate": 0.0006692872117400419,
      "loss": 0.3833,
      "num_input_tokens_seen": 1665832,
      "step": 2555
    },
    {
      "epoch": 1.3417190775681342,
      "grad_norm": 0.16684333980083466,
      "learning_rate": 0.0006705974842767296,
      "loss": 0.5444,
      "num_input_tokens_seen": 1668520,
      "step": 2560
    },
    {
      "epoch": 1.3443396226415094,
      "grad_norm": 0.09926697611808777,
      "learning_rate": 0.0006719077568134172,
      "loss": 0.4662,
      "num_input_tokens_seen": 1671912,
      "step": 2565
    },
    {
      "epoch": 1.3469601677148848,
      "grad_norm": 0.1231752559542656,
      "learning_rate": 0.0006732180293501048,
      "loss": 0.4916,
      "num_input_tokens_seen": 1675080,
      "step": 2570
    },
    {
      "epoch": 1.34958071278826,
      "grad_norm": 0.13456685841083527,
      "learning_rate": 0.0006745283018867925,
      "loss": 0.3431,
      "num_input_tokens_seen": 1681512,
      "step": 2575
    },
    {
      "epoch": 1.3522012578616351,
      "grad_norm": 0.09951241314411163,
      "learning_rate": 0.00067583857442348,
      "loss": 0.4818,
      "num_input_tokens_seen": 1685000,
      "step": 2580
    },
    {
      "epoch": 1.3548218029350105,
      "grad_norm": 0.14158523082733154,
      "learning_rate": 0.0006771488469601677,
      "loss": 0.4459,
      "num_input_tokens_seen": 1687528,
      "step": 2585
    },
    {
      "epoch": 1.3574423480083857,
      "grad_norm": 0.1115216612815857,
      "learning_rate": 0.0006784591194968553,
      "loss": 0.6047,
      "num_input_tokens_seen": 1690280,
      "step": 2590
    },
    {
      "epoch": 1.360062893081761,
      "grad_norm": 0.1147770956158638,
      "learning_rate": 0.0006797693920335431,
      "loss": 0.4013,
      "num_input_tokens_seen": 1693352,
      "step": 2595
    },
    {
      "epoch": 1.3626834381551363,
      "grad_norm": 0.16186216473579407,
      "learning_rate": 0.0006810796645702307,
      "loss": 0.4774,
      "num_input_tokens_seen": 1696456,
      "step": 2600
    },
    {
      "epoch": 1.3653039832285114,
      "grad_norm": 0.09263705462217331,
      "learning_rate": 0.0006823899371069183,
      "loss": 0.4359,
      "num_input_tokens_seen": 1700488,
      "step": 2605
    },
    {
      "epoch": 1.3679245283018868,
      "grad_norm": 0.1966683566570282,
      "learning_rate": 0.0006837002096436059,
      "loss": 0.4523,
      "num_input_tokens_seen": 1703304,
      "step": 2610
    },
    {
      "epoch": 1.370545073375262,
      "grad_norm": 0.19328561425209045,
      "learning_rate": 0.0006850104821802935,
      "loss": 0.5371,
      "num_input_tokens_seen": 1705864,
      "step": 2615
    },
    {
      "epoch": 1.3731656184486374,
      "grad_norm": 0.10260830819606781,
      "learning_rate": 0.0006863207547169812,
      "loss": 0.6124,
      "num_input_tokens_seen": 1708616,
      "step": 2620
    },
    {
      "epoch": 1.3757861635220126,
      "grad_norm": 0.058688897639513016,
      "learning_rate": 0.0006876310272536688,
      "loss": 0.4809,
      "num_input_tokens_seen": 1712232,
      "step": 2625
    },
    {
      "epoch": 1.3784067085953877,
      "grad_norm": 0.15809966623783112,
      "learning_rate": 0.0006889412997903564,
      "loss": 0.3278,
      "num_input_tokens_seen": 1715144,
      "step": 2630
    },
    {
      "epoch": 1.381027253668763,
      "grad_norm": 0.19460247457027435,
      "learning_rate": 0.0006902515723270441,
      "loss": 0.4597,
      "num_input_tokens_seen": 1717800,
      "step": 2635
    },
    {
      "epoch": 1.3836477987421385,
      "grad_norm": 0.1441037803888321,
      "learning_rate": 0.0006915618448637316,
      "loss": 0.4859,
      "num_input_tokens_seen": 1721128,
      "step": 2640
    },
    {
      "epoch": 1.3862683438155137,
      "grad_norm": 0.07195868343114853,
      "learning_rate": 0.0006928721174004193,
      "loss": 0.4524,
      "num_input_tokens_seen": 1724328,
      "step": 2645
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 0.14030320942401886,
      "learning_rate": 0.0006941823899371069,
      "loss": 0.5577,
      "num_input_tokens_seen": 1728328,
      "step": 2650
    },
    {
      "epoch": 1.3915094339622642,
      "grad_norm": 0.10735902935266495,
      "learning_rate": 0.0006954926624737946,
      "loss": 0.473,
      "num_input_tokens_seen": 1731080,
      "step": 2655
    },
    {
      "epoch": 1.3941299790356394,
      "grad_norm": 0.12594665586948395,
      "learning_rate": 0.0006968029350104822,
      "loss": 0.5375,
      "num_input_tokens_seen": 1736008,
      "step": 2660
    },
    {
      "epoch": 1.3967505241090148,
      "grad_norm": 0.17212900519371033,
      "learning_rate": 0.0006981132075471698,
      "loss": 0.4075,
      "num_input_tokens_seen": 1739368,
      "step": 2665
    },
    {
      "epoch": 1.39937106918239,
      "grad_norm": 0.14201220870018005,
      "learning_rate": 0.0006994234800838574,
      "loss": 0.6998,
      "num_input_tokens_seen": 1742408,
      "step": 2670
    },
    {
      "epoch": 1.4019916142557651,
      "grad_norm": 0.19367967545986176,
      "learning_rate": 0.000700733752620545,
      "loss": 0.5335,
      "num_input_tokens_seen": 1746120,
      "step": 2675
    },
    {
      "epoch": 1.4046121593291405,
      "grad_norm": 0.12169313430786133,
      "learning_rate": 0.0007020440251572327,
      "loss": 0.5233,
      "num_input_tokens_seen": 1749096,
      "step": 2680
    },
    {
      "epoch": 1.4072327044025157,
      "grad_norm": 0.2912430763244629,
      "learning_rate": 0.0007033542976939203,
      "loss": 0.5474,
      "num_input_tokens_seen": 1751624,
      "step": 2685
    },
    {
      "epoch": 1.409853249475891,
      "grad_norm": 0.17176155745983124,
      "learning_rate": 0.0007046645702306079,
      "loss": 0.4435,
      "num_input_tokens_seen": 1754216,
      "step": 2690
    },
    {
      "epoch": 1.4124737945492662,
      "grad_norm": 0.10080456733703613,
      "learning_rate": 0.0007059748427672957,
      "loss": 0.5723,
      "num_input_tokens_seen": 1758280,
      "step": 2695
    },
    {
      "epoch": 1.4150943396226414,
      "grad_norm": 0.17383089661598206,
      "learning_rate": 0.0007072851153039833,
      "loss": 0.4881,
      "num_input_tokens_seen": 1761352,
      "step": 2700
    },
    {
      "epoch": 1.4177148846960168,
      "grad_norm": 0.11134834587574005,
      "learning_rate": 0.0007085953878406709,
      "loss": 0.4421,
      "num_input_tokens_seen": 1764648,
      "step": 2705
    },
    {
      "epoch": 1.420335429769392,
      "grad_norm": 0.12780265510082245,
      "learning_rate": 0.0007099056603773585,
      "loss": 0.6195,
      "num_input_tokens_seen": 1767592,
      "step": 2710
    },
    {
      "epoch": 1.4229559748427674,
      "grad_norm": 0.08496706187725067,
      "learning_rate": 0.0007112159329140462,
      "loss": 0.4565,
      "num_input_tokens_seen": 1774600,
      "step": 2715
    },
    {
      "epoch": 1.4255765199161425,
      "grad_norm": 0.16602623462677002,
      "learning_rate": 0.0007125262054507338,
      "loss": 0.5814,
      "num_input_tokens_seen": 1778248,
      "step": 2720
    },
    {
      "epoch": 1.4281970649895177,
      "grad_norm": 0.10184605419635773,
      "learning_rate": 0.0007138364779874214,
      "loss": 0.4708,
      "num_input_tokens_seen": 1781608,
      "step": 2725
    },
    {
      "epoch": 1.430817610062893,
      "grad_norm": 0.1427193284034729,
      "learning_rate": 0.0007151467505241091,
      "loss": 0.6384,
      "num_input_tokens_seen": 1784776,
      "step": 2730
    },
    {
      "epoch": 1.4334381551362683,
      "grad_norm": 0.12168832123279572,
      "learning_rate": 0.0007164570230607966,
      "loss": 0.5616,
      "num_input_tokens_seen": 1787688,
      "step": 2735
    },
    {
      "epoch": 1.4360587002096437,
      "grad_norm": 0.09070499986410141,
      "learning_rate": 0.0007177672955974843,
      "loss": 0.5882,
      "num_input_tokens_seen": 1791016,
      "step": 2740
    },
    {
      "epoch": 1.4386792452830188,
      "grad_norm": 0.41015687584877014,
      "learning_rate": 0.0007190775681341719,
      "loss": 0.5074,
      "num_input_tokens_seen": 1794280,
      "step": 2745
    },
    {
      "epoch": 1.441299790356394,
      "grad_norm": 0.15429195761680603,
      "learning_rate": 0.0007203878406708596,
      "loss": 0.3968,
      "num_input_tokens_seen": 1797480,
      "step": 2750
    },
    {
      "epoch": 1.4439203354297694,
      "grad_norm": 0.20932793617248535,
      "learning_rate": 0.0007216981132075472,
      "loss": 0.6271,
      "num_input_tokens_seen": 1801160,
      "step": 2755
    },
    {
      "epoch": 1.4465408805031448,
      "grad_norm": 0.14346763491630554,
      "learning_rate": 0.0007230083857442348,
      "loss": 0.6007,
      "num_input_tokens_seen": 1803464,
      "step": 2760
    },
    {
      "epoch": 1.44916142557652,
      "grad_norm": 0.3667054772377014,
      "learning_rate": 0.0007243186582809224,
      "loss": 0.4611,
      "num_input_tokens_seen": 1806728,
      "step": 2765
    },
    {
      "epoch": 1.4517819706498951,
      "grad_norm": 0.10567866265773773,
      "learning_rate": 0.00072562893081761,
      "loss": 0.4343,
      "num_input_tokens_seen": 1810088,
      "step": 2770
    },
    {
      "epoch": 1.4544025157232705,
      "grad_norm": 0.06356578320264816,
      "learning_rate": 0.0007269392033542977,
      "loss": 0.4314,
      "num_input_tokens_seen": 1815144,
      "step": 2775
    },
    {
      "epoch": 1.4570230607966457,
      "grad_norm": 0.07758204638957977,
      "learning_rate": 0.0007282494758909853,
      "loss": 0.4187,
      "num_input_tokens_seen": 1818792,
      "step": 2780
    },
    {
      "epoch": 1.459643605870021,
      "grad_norm": 0.11461982876062393,
      "learning_rate": 0.0007295597484276729,
      "loss": 0.6008,
      "num_input_tokens_seen": 1821768,
      "step": 2785
    },
    {
      "epoch": 1.4622641509433962,
      "grad_norm": 0.14117862284183502,
      "learning_rate": 0.0007308700209643606,
      "loss": 0.6171,
      "num_input_tokens_seen": 1825512,
      "step": 2790
    },
    {
      "epoch": 1.4648846960167714,
      "grad_norm": 0.11547422409057617,
      "learning_rate": 0.0007321802935010483,
      "loss": 0.3789,
      "num_input_tokens_seen": 1828968,
      "step": 2795
    },
    {
      "epoch": 1.4675052410901468,
      "grad_norm": 0.12443093210458755,
      "learning_rate": 0.0007334905660377359,
      "loss": 0.5442,
      "num_input_tokens_seen": 1831336,
      "step": 2800
    },
    {
      "epoch": 1.470125786163522,
      "grad_norm": 0.07954557985067368,
      "learning_rate": 0.0007348008385744235,
      "loss": 0.5314,
      "num_input_tokens_seen": 1834536,
      "step": 2805
    },
    {
      "epoch": 1.4727463312368974,
      "grad_norm": 0.07832390069961548,
      "learning_rate": 0.0007361111111111112,
      "loss": 0.4684,
      "num_input_tokens_seen": 1837480,
      "step": 2810
    },
    {
      "epoch": 1.4753668763102725,
      "grad_norm": 0.23312847316265106,
      "learning_rate": 0.0007374213836477988,
      "loss": 0.5823,
      "num_input_tokens_seen": 1840808,
      "step": 2815
    },
    {
      "epoch": 1.4779874213836477,
      "grad_norm": 0.09403182566165924,
      "learning_rate": 0.0007387316561844864,
      "loss": 0.6257,
      "num_input_tokens_seen": 1844392,
      "step": 2820
    },
    {
      "epoch": 1.480607966457023,
      "grad_norm": 0.0957610160112381,
      "learning_rate": 0.0007400419287211741,
      "loss": 0.3899,
      "num_input_tokens_seen": 1848008,
      "step": 2825
    },
    {
      "epoch": 1.4832285115303983,
      "grad_norm": 0.12789064645767212,
      "learning_rate": 0.0007413522012578616,
      "loss": 0.4283,
      "num_input_tokens_seen": 1850888,
      "step": 2830
    },
    {
      "epoch": 1.4858490566037736,
      "grad_norm": 0.3502878248691559,
      "learning_rate": 0.0007426624737945493,
      "loss": 0.697,
      "num_input_tokens_seen": 1854312,
      "step": 2835
    },
    {
      "epoch": 1.4884696016771488,
      "grad_norm": 0.08732571452856064,
      "learning_rate": 0.0007439727463312369,
      "loss": 0.4895,
      "num_input_tokens_seen": 1858792,
      "step": 2840
    },
    {
      "epoch": 1.491090146750524,
      "grad_norm": 0.12488330155611038,
      "learning_rate": 0.0007452830188679245,
      "loss": 0.4691,
      "num_input_tokens_seen": 1861672,
      "step": 2845
    },
    {
      "epoch": 1.4937106918238994,
      "grad_norm": 0.14638788998126984,
      "learning_rate": 0.0007465932914046122,
      "loss": 0.4469,
      "num_input_tokens_seen": 1864360,
      "step": 2850
    },
    {
      "epoch": 1.4963312368972748,
      "grad_norm": 0.19452902674674988,
      "learning_rate": 0.0007479035639412998,
      "loss": 0.4787,
      "num_input_tokens_seen": 1867048,
      "step": 2855
    },
    {
      "epoch": 1.49895178197065,
      "grad_norm": 0.09995990246534348,
      "learning_rate": 0.0007492138364779874,
      "loss": 0.4734,
      "num_input_tokens_seen": 1870280,
      "step": 2860
    },
    {
      "epoch": 1.501572327044025,
      "grad_norm": 0.1393478363752365,
      "learning_rate": 0.000750524109014675,
      "loss": 0.6569,
      "num_input_tokens_seen": 1873896,
      "step": 2865
    },
    {
      "epoch": 1.5041928721174003,
      "grad_norm": 0.09196323901414871,
      "learning_rate": 0.0007518343815513627,
      "loss": 0.5101,
      "num_input_tokens_seen": 1876808,
      "step": 2870
    },
    {
      "epoch": 1.5068134171907757,
      "grad_norm": 0.10797059535980225,
      "learning_rate": 0.0007531446540880503,
      "loss": 0.5233,
      "num_input_tokens_seen": 1880168,
      "step": 2875
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 0.169708713889122,
      "learning_rate": 0.0007544549266247379,
      "loss": 0.4208,
      "num_input_tokens_seen": 1882824,
      "step": 2880
    },
    {
      "epoch": 1.5120545073375262,
      "grad_norm": 0.08964937925338745,
      "learning_rate": 0.0007557651991614256,
      "loss": 0.4526,
      "num_input_tokens_seen": 1886184,
      "step": 2885
    },
    {
      "epoch": 1.5146750524109014,
      "grad_norm": 0.11808939278125763,
      "learning_rate": 0.0007570754716981131,
      "loss": 0.526,
      "num_input_tokens_seen": 1888968,
      "step": 2890
    },
    {
      "epoch": 1.5172955974842768,
      "grad_norm": 0.1362273246049881,
      "learning_rate": 0.0007583857442348009,
      "loss": 0.5918,
      "num_input_tokens_seen": 1892232,
      "step": 2895
    },
    {
      "epoch": 1.519916142557652,
      "grad_norm": 0.10454388707876205,
      "learning_rate": 0.0007596960167714885,
      "loss": 0.4495,
      "num_input_tokens_seen": 1895240,
      "step": 2900
    },
    {
      "epoch": 1.5225366876310273,
      "grad_norm": 0.04979229345917702,
      "learning_rate": 0.0007610062893081762,
      "loss": 0.3535,
      "num_input_tokens_seen": 1899144,
      "step": 2905
    },
    {
      "epoch": 1.5251572327044025,
      "grad_norm": 0.11372412741184235,
      "learning_rate": 0.0007623165618448638,
      "loss": 0.4514,
      "num_input_tokens_seen": 1901896,
      "step": 2910
    },
    {
      "epoch": 1.5277777777777777,
      "grad_norm": 0.08344082534313202,
      "learning_rate": 0.0007636268343815514,
      "loss": 0.4711,
      "num_input_tokens_seen": 1904936,
      "step": 2915
    },
    {
      "epoch": 1.530398322851153,
      "grad_norm": 0.13003316521644592,
      "learning_rate": 0.0007649371069182391,
      "loss": 0.5089,
      "num_input_tokens_seen": 1908328,
      "step": 2920
    },
    {
      "epoch": 1.5330188679245285,
      "grad_norm": 0.10059721767902374,
      "learning_rate": 0.0007662473794549266,
      "loss": 0.5612,
      "num_input_tokens_seen": 1911496,
      "step": 2925
    },
    {
      "epoch": 1.5356394129979036,
      "grad_norm": 0.18789812922477722,
      "learning_rate": 0.0007675576519916143,
      "loss": 0.4345,
      "num_input_tokens_seen": 1913992,
      "step": 2930
    },
    {
      "epoch": 1.5382599580712788,
      "grad_norm": 0.1871994286775589,
      "learning_rate": 0.0007688679245283019,
      "loss": 0.5532,
      "num_input_tokens_seen": 1917640,
      "step": 2935
    },
    {
      "epoch": 1.540880503144654,
      "grad_norm": 0.10711561143398285,
      "learning_rate": 0.0007701781970649895,
      "loss": 0.3933,
      "num_input_tokens_seen": 1920616,
      "step": 2940
    },
    {
      "epoch": 1.5435010482180294,
      "grad_norm": 0.13010132312774658,
      "learning_rate": 0.0007714884696016772,
      "loss": 0.4593,
      "num_input_tokens_seen": 1923208,
      "step": 2945
    },
    {
      "epoch": 1.5461215932914047,
      "grad_norm": 0.0697295069694519,
      "learning_rate": 0.0007727987421383648,
      "loss": 0.4172,
      "num_input_tokens_seen": 1926952,
      "step": 2950
    },
    {
      "epoch": 1.54874213836478,
      "grad_norm": 0.16297698020935059,
      "learning_rate": 0.0007741090146750524,
      "loss": 0.4724,
      "num_input_tokens_seen": 1930696,
      "step": 2955
    },
    {
      "epoch": 1.551362683438155,
      "grad_norm": 0.08239731937646866,
      "learning_rate": 0.00077541928721174,
      "loss": 0.4842,
      "num_input_tokens_seen": 1933864,
      "step": 2960
    },
    {
      "epoch": 1.5539832285115303,
      "grad_norm": 0.12613733112812042,
      "learning_rate": 0.0007767295597484277,
      "loss": 0.5093,
      "num_input_tokens_seen": 1938440,
      "step": 2965
    },
    {
      "epoch": 1.5566037735849056,
      "grad_norm": 0.06465013325214386,
      "learning_rate": 0.0007780398322851153,
      "loss": 0.5786,
      "num_input_tokens_seen": 1942184,
      "step": 2970
    },
    {
      "epoch": 1.559224318658281,
      "grad_norm": 0.09704403579235077,
      "learning_rate": 0.0007793501048218029,
      "loss": 0.4438,
      "num_input_tokens_seen": 1945928,
      "step": 2975
    },
    {
      "epoch": 1.5618448637316562,
      "grad_norm": 0.16120180487632751,
      "learning_rate": 0.0007806603773584906,
      "loss": 0.5561,
      "num_input_tokens_seen": 1949416,
      "step": 2980
    },
    {
      "epoch": 1.5644654088050314,
      "grad_norm": 0.07854507863521576,
      "learning_rate": 0.0007819706498951781,
      "loss": 0.4447,
      "num_input_tokens_seen": 1952008,
      "step": 2985
    },
    {
      "epoch": 1.5670859538784065,
      "grad_norm": 0.1467886120080948,
      "learning_rate": 0.0007832809224318658,
      "loss": 0.4526,
      "num_input_tokens_seen": 1955656,
      "step": 2990
    },
    {
      "epoch": 1.569706498951782,
      "grad_norm": 0.1676732450723648,
      "learning_rate": 0.0007845911949685535,
      "loss": 0.6624,
      "num_input_tokens_seen": 1958632,
      "step": 2995
    },
    {
      "epoch": 1.5723270440251573,
      "grad_norm": 0.06834208220243454,
      "learning_rate": 0.0007859014675052411,
      "loss": 0.4898,
      "num_input_tokens_seen": 1961256,
      "step": 3000
    },
    {
      "epoch": 1.5749475890985325,
      "grad_norm": 0.0701621025800705,
      "learning_rate": 0.0007872117400419288,
      "loss": 0.4196,
      "num_input_tokens_seen": 1965480,
      "step": 3005
    },
    {
      "epoch": 1.5775681341719077,
      "grad_norm": 0.14063501358032227,
      "learning_rate": 0.0007885220125786164,
      "loss": 0.4771,
      "num_input_tokens_seen": 1968552,
      "step": 3010
    },
    {
      "epoch": 1.580188679245283,
      "grad_norm": 0.06775826960802078,
      "learning_rate": 0.0007898322851153041,
      "loss": 0.4646,
      "num_input_tokens_seen": 1971400,
      "step": 3015
    },
    {
      "epoch": 1.5828092243186582,
      "grad_norm": 0.07474400103092194,
      "learning_rate": 0.0007911425576519916,
      "loss": 0.398,
      "num_input_tokens_seen": 1974312,
      "step": 3020
    },
    {
      "epoch": 1.5854297693920336,
      "grad_norm": 0.06963302195072174,
      "learning_rate": 0.0007924528301886793,
      "loss": 0.3763,
      "num_input_tokens_seen": 1976904,
      "step": 3025
    },
    {
      "epoch": 1.5880503144654088,
      "grad_norm": 0.16966208815574646,
      "learning_rate": 0.0007937631027253669,
      "loss": 0.5424,
      "num_input_tokens_seen": 1979432,
      "step": 3030
    },
    {
      "epoch": 1.590670859538784,
      "grad_norm": 0.11685218662023544,
      "learning_rate": 0.0007950733752620545,
      "loss": 0.4991,
      "num_input_tokens_seen": 1982952,
      "step": 3035
    },
    {
      "epoch": 1.5932914046121593,
      "grad_norm": 0.18328255414962769,
      "learning_rate": 0.0007963836477987422,
      "loss": 0.4756,
      "num_input_tokens_seen": 1986408,
      "step": 3040
    },
    {
      "epoch": 1.5959119496855347,
      "grad_norm": 0.13670895993709564,
      "learning_rate": 0.0007976939203354298,
      "loss": 0.4414,
      "num_input_tokens_seen": 1988936,
      "step": 3045
    },
    {
      "epoch": 1.59853249475891,
      "grad_norm": 0.10173245519399643,
      "learning_rate": 0.0007990041928721174,
      "loss": 0.6301,
      "num_input_tokens_seen": 1992264,
      "step": 3050
    },
    {
      "epoch": 1.601153039832285,
      "grad_norm": 0.15248653292655945,
      "learning_rate": 0.000800314465408805,
      "loss": 0.591,
      "num_input_tokens_seen": 1995240,
      "step": 3055
    },
    {
      "epoch": 1.6037735849056602,
      "grad_norm": 0.10489282757043839,
      "learning_rate": 0.0008016247379454927,
      "loss": 0.5291,
      "num_input_tokens_seen": 1998408,
      "step": 3060
    },
    {
      "epoch": 1.6063941299790356,
      "grad_norm": 0.1256984919309616,
      "learning_rate": 0.0008029350104821803,
      "loss": 0.485,
      "num_input_tokens_seen": 2001416,
      "step": 3065
    },
    {
      "epoch": 1.609014675052411,
      "grad_norm": 0.0834575966000557,
      "learning_rate": 0.0008042452830188679,
      "loss": 0.4601,
      "num_input_tokens_seen": 2003752,
      "step": 3070
    },
    {
      "epoch": 1.6116352201257862,
      "grad_norm": 0.04799262806773186,
      "learning_rate": 0.0008055555555555556,
      "loss": 0.467,
      "num_input_tokens_seen": 2006728,
      "step": 3075
    },
    {
      "epoch": 1.6142557651991614,
      "grad_norm": 0.1056966632604599,
      "learning_rate": 0.0008068658280922431,
      "loss": 0.4623,
      "num_input_tokens_seen": 2009928,
      "step": 3080
    },
    {
      "epoch": 1.6168763102725365,
      "grad_norm": 0.055823665112257004,
      "learning_rate": 0.0008081761006289308,
      "loss": 0.3521,
      "num_input_tokens_seen": 2012648,
      "step": 3085
    },
    {
      "epoch": 1.619496855345912,
      "grad_norm": 0.06463918089866638,
      "learning_rate": 0.0008094863731656184,
      "loss": 0.4057,
      "num_input_tokens_seen": 2016936,
      "step": 3090
    },
    {
      "epoch": 1.6221174004192873,
      "grad_norm": 0.13881109654903412,
      "learning_rate": 0.0008107966457023061,
      "loss": 0.3439,
      "num_input_tokens_seen": 2019720,
      "step": 3095
    },
    {
      "epoch": 1.6247379454926625,
      "grad_norm": 0.13252778351306915,
      "learning_rate": 0.0008121069182389938,
      "loss": 0.4394,
      "num_input_tokens_seen": 2022760,
      "step": 3100
    },
    {
      "epoch": 1.6273584905660377,
      "grad_norm": 0.0836377739906311,
      "learning_rate": 0.0008134171907756814,
      "loss": 0.4408,
      "num_input_tokens_seen": 2025544,
      "step": 3105
    },
    {
      "epoch": 1.629979035639413,
      "grad_norm": 0.11498484015464783,
      "learning_rate": 0.0008147274633123691,
      "loss": 0.4214,
      "num_input_tokens_seen": 2028872,
      "step": 3110
    },
    {
      "epoch": 1.6325995807127882,
      "grad_norm": 0.1125246211886406,
      "learning_rate": 0.0008160377358490566,
      "loss": 0.4243,
      "num_input_tokens_seen": 2031816,
      "step": 3115
    },
    {
      "epoch": 1.6352201257861636,
      "grad_norm": 0.1171121895313263,
      "learning_rate": 0.0008173480083857443,
      "loss": 0.5943,
      "num_input_tokens_seen": 2034280,
      "step": 3120
    },
    {
      "epoch": 1.6378406708595388,
      "grad_norm": 0.08498336374759674,
      "learning_rate": 0.0008186582809224319,
      "loss": 0.4913,
      "num_input_tokens_seen": 2038056,
      "step": 3125
    },
    {
      "epoch": 1.640461215932914,
      "grad_norm": 0.12578152120113373,
      "learning_rate": 0.0008199685534591195,
      "loss": 0.5348,
      "num_input_tokens_seen": 2041192,
      "step": 3130
    },
    {
      "epoch": 1.6430817610062893,
      "grad_norm": 0.08654726296663284,
      "learning_rate": 0.0008212788259958072,
      "loss": 0.4764,
      "num_input_tokens_seen": 2043720,
      "step": 3135
    },
    {
      "epoch": 1.6457023060796647,
      "grad_norm": 0.09017306566238403,
      "learning_rate": 0.0008225890985324948,
      "loss": 0.5506,
      "num_input_tokens_seen": 2046792,
      "step": 3140
    },
    {
      "epoch": 1.64832285115304,
      "grad_norm": 0.0804893746972084,
      "learning_rate": 0.0008238993710691824,
      "loss": 0.633,
      "num_input_tokens_seen": 2049768,
      "step": 3145
    },
    {
      "epoch": 1.650943396226415,
      "grad_norm": 0.11381354182958603,
      "learning_rate": 0.00082520964360587,
      "loss": 0.5126,
      "num_input_tokens_seen": 2052840,
      "step": 3150
    },
    {
      "epoch": 1.6535639412997902,
      "grad_norm": 0.09524408727884293,
      "learning_rate": 0.0008265199161425576,
      "loss": 0.5228,
      "num_input_tokens_seen": 2055432,
      "step": 3155
    },
    {
      "epoch": 1.6561844863731656,
      "grad_norm": 0.11912812292575836,
      "learning_rate": 0.0008278301886792453,
      "loss": 0.6153,
      "num_input_tokens_seen": 2058792,
      "step": 3160
    },
    {
      "epoch": 1.658805031446541,
      "grad_norm": 0.08036439120769501,
      "learning_rate": 0.0008291404612159329,
      "loss": 0.4796,
      "num_input_tokens_seen": 2061960,
      "step": 3165
    },
    {
      "epoch": 1.6614255765199162,
      "grad_norm": 0.09492501616477966,
      "learning_rate": 0.0008304507337526206,
      "loss": 0.63,
      "num_input_tokens_seen": 2064712,
      "step": 3170
    },
    {
      "epoch": 1.6640461215932913,
      "grad_norm": 0.12854263186454773,
      "learning_rate": 0.0008317610062893081,
      "loss": 0.4393,
      "num_input_tokens_seen": 2068360,
      "step": 3175
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.10767249763011932,
      "learning_rate": 0.0008330712788259958,
      "loss": 0.4699,
      "num_input_tokens_seen": 2071304,
      "step": 3180
    },
    {
      "epoch": 1.669287211740042,
      "grad_norm": 0.0972372442483902,
      "learning_rate": 0.0008343815513626834,
      "loss": 0.6975,
      "num_input_tokens_seen": 2075016,
      "step": 3185
    },
    {
      "epoch": 1.6719077568134173,
      "grad_norm": 0.06392829120159149,
      "learning_rate": 0.000835691823899371,
      "loss": 0.4554,
      "num_input_tokens_seen": 2079016,
      "step": 3190
    },
    {
      "epoch": 1.6745283018867925,
      "grad_norm": 0.09330326318740845,
      "learning_rate": 0.0008370020964360588,
      "loss": 0.377,
      "num_input_tokens_seen": 2081960,
      "step": 3195
    },
    {
      "epoch": 1.6771488469601676,
      "grad_norm": 0.10519814491271973,
      "learning_rate": 0.0008383123689727464,
      "loss": 0.4096,
      "num_input_tokens_seen": 2084648,
      "step": 3200
    },
    {
      "epoch": 1.679769392033543,
      "grad_norm": 0.10176686942577362,
      "learning_rate": 0.000839622641509434,
      "loss": 0.4549,
      "num_input_tokens_seen": 2088840,
      "step": 3205
    },
    {
      "epoch": 1.6823899371069182,
      "grad_norm": 0.13175135850906372,
      "learning_rate": 0.0008409329140461216,
      "loss": 0.4993,
      "num_input_tokens_seen": 2091944,
      "step": 3210
    },
    {
      "epoch": 1.6850104821802936,
      "grad_norm": 0.1130446195602417,
      "learning_rate": 0.0008422431865828093,
      "loss": 0.5405,
      "num_input_tokens_seen": 2095144,
      "step": 3215
    },
    {
      "epoch": 1.6876310272536688,
      "grad_norm": 0.09734788537025452,
      "learning_rate": 0.0008435534591194969,
      "loss": 0.5443,
      "num_input_tokens_seen": 2098984,
      "step": 3220
    },
    {
      "epoch": 1.690251572327044,
      "grad_norm": 0.06836172938346863,
      "learning_rate": 0.0008448637316561845,
      "loss": 0.3469,
      "num_input_tokens_seen": 2101736,
      "step": 3225
    },
    {
      "epoch": 1.6928721174004193,
      "grad_norm": 0.08214569091796875,
      "learning_rate": 0.0008461740041928722,
      "loss": 0.5256,
      "num_input_tokens_seen": 2104904,
      "step": 3230
    },
    {
      "epoch": 1.6954926624737947,
      "grad_norm": 0.12698136270046234,
      "learning_rate": 0.0008474842767295598,
      "loss": 0.5294,
      "num_input_tokens_seen": 2108392,
      "step": 3235
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 0.08570567518472672,
      "learning_rate": 0.0008487945492662474,
      "loss": 0.6037,
      "num_input_tokens_seen": 2112360,
      "step": 3240
    },
    {
      "epoch": 1.700733752620545,
      "grad_norm": 0.1458934247493744,
      "learning_rate": 0.000850104821802935,
      "loss": 0.512,
      "num_input_tokens_seen": 2117992,
      "step": 3245
    },
    {
      "epoch": 1.7033542976939202,
      "grad_norm": 0.11601375043392181,
      "learning_rate": 0.0008514150943396226,
      "loss": 0.411,
      "num_input_tokens_seen": 2121096,
      "step": 3250
    },
    {
      "epoch": 1.7059748427672956,
      "grad_norm": 0.11036747694015503,
      "learning_rate": 0.0008527253668763103,
      "loss": 0.4941,
      "num_input_tokens_seen": 2125064,
      "step": 3255
    },
    {
      "epoch": 1.708595387840671,
      "grad_norm": 0.12293072789907455,
      "learning_rate": 0.0008540356394129979,
      "loss": 0.393,
      "num_input_tokens_seen": 2129032,
      "step": 3260
    },
    {
      "epoch": 1.7112159329140462,
      "grad_norm": 0.2633567452430725,
      "learning_rate": 0.0008553459119496856,
      "loss": 0.5836,
      "num_input_tokens_seen": 2131656,
      "step": 3265
    },
    {
      "epoch": 1.7138364779874213,
      "grad_norm": 1.0485820770263672,
      "learning_rate": 0.0008566561844863731,
      "loss": 0.5959,
      "num_input_tokens_seen": 2134760,
      "step": 3270
    },
    {
      "epoch": 1.7164570230607965,
      "grad_norm": 0.14652298390865326,
      "learning_rate": 0.0008579664570230608,
      "loss": 0.5868,
      "num_input_tokens_seen": 2137960,
      "step": 3275
    },
    {
      "epoch": 1.719077568134172,
      "grad_norm": 0.1308698058128357,
      "learning_rate": 0.0008592767295597484,
      "loss": 0.4612,
      "num_input_tokens_seen": 2140552,
      "step": 3280
    },
    {
      "epoch": 1.7216981132075473,
      "grad_norm": 0.1504572182893753,
      "learning_rate": 0.000860587002096436,
      "loss": 0.3742,
      "num_input_tokens_seen": 2143272,
      "step": 3285
    },
    {
      "epoch": 1.7243186582809225,
      "grad_norm": 0.06660829484462738,
      "learning_rate": 0.0008618972746331238,
      "loss": 0.5275,
      "num_input_tokens_seen": 2147304,
      "step": 3290
    },
    {
      "epoch": 1.7269392033542976,
      "grad_norm": 0.07939482480287552,
      "learning_rate": 0.0008632075471698114,
      "loss": 0.666,
      "num_input_tokens_seen": 2151624,
      "step": 3295
    },
    {
      "epoch": 1.7295597484276728,
      "grad_norm": 0.15162062644958496,
      "learning_rate": 0.000864517819706499,
      "loss": 0.5473,
      "num_input_tokens_seen": 2154248,
      "step": 3300
    },
    {
      "epoch": 1.7321802935010482,
      "grad_norm": 0.17145869135856628,
      "learning_rate": 0.0008658280922431866,
      "loss": 0.5689,
      "num_input_tokens_seen": 2157000,
      "step": 3305
    },
    {
      "epoch": 1.7348008385744236,
      "grad_norm": 0.2624160945415497,
      "learning_rate": 0.0008671383647798742,
      "loss": 0.4344,
      "num_input_tokens_seen": 2160904,
      "step": 3310
    },
    {
      "epoch": 1.7374213836477987,
      "grad_norm": 0.07931435108184814,
      "learning_rate": 0.0008684486373165619,
      "loss": 0.5471,
      "num_input_tokens_seen": 2163720,
      "step": 3315
    },
    {
      "epoch": 1.740041928721174,
      "grad_norm": 0.08898314833641052,
      "learning_rate": 0.0008697589098532495,
      "loss": 0.4304,
      "num_input_tokens_seen": 2167912,
      "step": 3320
    },
    {
      "epoch": 1.7426624737945493,
      "grad_norm": 0.09017103165388107,
      "learning_rate": 0.0008710691823899372,
      "loss": 0.4638,
      "num_input_tokens_seen": 2171592,
      "step": 3325
    },
    {
      "epoch": 1.7452830188679245,
      "grad_norm": 0.11272095888853073,
      "learning_rate": 0.0008723794549266247,
      "loss": 0.5429,
      "num_input_tokens_seen": 2173864,
      "step": 3330
    },
    {
      "epoch": 1.7479035639412999,
      "grad_norm": 0.1010701060295105,
      "learning_rate": 0.0008736897274633124,
      "loss": 0.4438,
      "num_input_tokens_seen": 2176744,
      "step": 3335
    },
    {
      "epoch": 1.750524109014675,
      "grad_norm": 0.07986171543598175,
      "learning_rate": 0.000875,
      "loss": 0.6511,
      "num_input_tokens_seen": 2179528,
      "step": 3340
    },
    {
      "epoch": 1.7531446540880502,
      "grad_norm": 0.1989065259695053,
      "learning_rate": 0.0008763102725366876,
      "loss": 0.4411,
      "num_input_tokens_seen": 2182856,
      "step": 3345
    },
    {
      "epoch": 1.7557651991614256,
      "grad_norm": 0.0832909494638443,
      "learning_rate": 0.0008776205450733753,
      "loss": 0.4344,
      "num_input_tokens_seen": 2185832,
      "step": 3350
    },
    {
      "epoch": 1.758385744234801,
      "grad_norm": 0.08214692026376724,
      "learning_rate": 0.0008789308176100629,
      "loss": 0.5057,
      "num_input_tokens_seen": 2189352,
      "step": 3355
    },
    {
      "epoch": 1.7610062893081762,
      "grad_norm": 0.16395634412765503,
      "learning_rate": 0.0008802410901467506,
      "loss": 0.4654,
      "num_input_tokens_seen": 2192264,
      "step": 3360
    },
    {
      "epoch": 1.7636268343815513,
      "grad_norm": 0.08822352439165115,
      "learning_rate": 0.0008815513626834381,
      "loss": 0.4776,
      "num_input_tokens_seen": 2195560,
      "step": 3365
    },
    {
      "epoch": 1.7662473794549265,
      "grad_norm": 0.08577209711074829,
      "learning_rate": 0.0008828616352201258,
      "loss": 0.4305,
      "num_input_tokens_seen": 2199080,
      "step": 3370
    },
    {
      "epoch": 1.7688679245283019,
      "grad_norm": 0.06759961694478989,
      "learning_rate": 0.0008841719077568134,
      "loss": 0.403,
      "num_input_tokens_seen": 2202408,
      "step": 3375
    },
    {
      "epoch": 1.7714884696016773,
      "grad_norm": 0.07581593096256256,
      "learning_rate": 0.000885482180293501,
      "loss": 0.5245,
      "num_input_tokens_seen": 2206056,
      "step": 3380
    },
    {
      "epoch": 1.7741090146750524,
      "grad_norm": 0.16021649539470673,
      "learning_rate": 0.0008867924528301887,
      "loss": 0.4139,
      "num_input_tokens_seen": 2208456,
      "step": 3385
    },
    {
      "epoch": 1.7767295597484276,
      "grad_norm": 0.06061306968331337,
      "learning_rate": 0.0008881027253668763,
      "loss": 0.3706,
      "num_input_tokens_seen": 2211848,
      "step": 3390
    },
    {
      "epoch": 1.7793501048218028,
      "grad_norm": 0.07280026376247406,
      "learning_rate": 0.000889412997903564,
      "loss": 0.402,
      "num_input_tokens_seen": 2215112,
      "step": 3395
    },
    {
      "epoch": 1.7819706498951782,
      "grad_norm": 0.11358582973480225,
      "learning_rate": 0.0008907232704402516,
      "loss": 0.5224,
      "num_input_tokens_seen": 2219528,
      "step": 3400
    },
    {
      "epoch": 1.7845911949685536,
      "grad_norm": 0.09189874678850174,
      "learning_rate": 0.0008920335429769392,
      "loss": 0.5433,
      "num_input_tokens_seen": 2223624,
      "step": 3405
    },
    {
      "epoch": 1.7872117400419287,
      "grad_norm": 0.07039090245962143,
      "learning_rate": 0.0008933438155136269,
      "loss": 0.4873,
      "num_input_tokens_seen": 2227464,
      "step": 3410
    },
    {
      "epoch": 1.789832285115304,
      "grad_norm": 0.0399029515683651,
      "learning_rate": 0.0008946540880503145,
      "loss": 0.4291,
      "num_input_tokens_seen": 2231240,
      "step": 3415
    },
    {
      "epoch": 1.7924528301886793,
      "grad_norm": 0.1427558809518814,
      "learning_rate": 0.0008959643605870022,
      "loss": 0.5872,
      "num_input_tokens_seen": 2234600,
      "step": 3420
    },
    {
      "epoch": 1.7950733752620545,
      "grad_norm": 0.05993777513504028,
      "learning_rate": 0.0008972746331236897,
      "loss": 0.5048,
      "num_input_tokens_seen": 2238056,
      "step": 3425
    },
    {
      "epoch": 1.7976939203354299,
      "grad_norm": 0.07162176072597504,
      "learning_rate": 0.0008985849056603774,
      "loss": 0.4224,
      "num_input_tokens_seen": 2241640,
      "step": 3430
    },
    {
      "epoch": 1.800314465408805,
      "grad_norm": 0.0806281790137291,
      "learning_rate": 0.000899895178197065,
      "loss": 0.5282,
      "num_input_tokens_seen": 2245832,
      "step": 3435
    },
    {
      "epoch": 1.8029350104821802,
      "grad_norm": 0.14014315605163574,
      "learning_rate": 0.0009012054507337526,
      "loss": 0.7093,
      "num_input_tokens_seen": 2248552,
      "step": 3440
    },
    {
      "epoch": 1.8055555555555556,
      "grad_norm": 0.11887964606285095,
      "learning_rate": 0.0009025157232704403,
      "loss": 0.5474,
      "num_input_tokens_seen": 2251432,
      "step": 3445
    },
    {
      "epoch": 1.808176100628931,
      "grad_norm": 0.1511143147945404,
      "learning_rate": 0.0009038259958071279,
      "loss": 0.5682,
      "num_input_tokens_seen": 2254408,
      "step": 3450
    },
    {
      "epoch": 1.8107966457023061,
      "grad_norm": 0.07858309149742126,
      "learning_rate": 0.0009051362683438156,
      "loss": 0.6698,
      "num_input_tokens_seen": 2257032,
      "step": 3455
    },
    {
      "epoch": 1.8134171907756813,
      "grad_norm": 0.1573200821876526,
      "learning_rate": 0.0009064465408805031,
      "loss": 0.435,
      "num_input_tokens_seen": 2259560,
      "step": 3460
    },
    {
      "epoch": 1.8160377358490565,
      "grad_norm": 0.08074133843183517,
      "learning_rate": 0.0009077568134171907,
      "loss": 0.5373,
      "num_input_tokens_seen": 2263240,
      "step": 3465
    },
    {
      "epoch": 1.8186582809224319,
      "grad_norm": 0.08898638188838959,
      "learning_rate": 0.0009090670859538784,
      "loss": 0.4613,
      "num_input_tokens_seen": 2267144,
      "step": 3470
    },
    {
      "epoch": 1.8212788259958073,
      "grad_norm": 0.2031051069498062,
      "learning_rate": 0.000910377358490566,
      "loss": 0.5479,
      "num_input_tokens_seen": 2270856,
      "step": 3475
    },
    {
      "epoch": 1.8238993710691824,
      "grad_norm": 0.22154080867767334,
      "learning_rate": 0.0009116876310272537,
      "loss": 0.5688,
      "num_input_tokens_seen": 2273672,
      "step": 3480
    },
    {
      "epoch": 1.8265199161425576,
      "grad_norm": 0.11358790099620819,
      "learning_rate": 0.0009129979035639413,
      "loss": 0.4809,
      "num_input_tokens_seen": 2276520,
      "step": 3485
    },
    {
      "epoch": 1.8291404612159328,
      "grad_norm": 0.09635080397129059,
      "learning_rate": 0.000914308176100629,
      "loss": 0.428,
      "num_input_tokens_seen": 2279368,
      "step": 3490
    },
    {
      "epoch": 1.8317610062893082,
      "grad_norm": 0.07172741740942001,
      "learning_rate": 0.0009156184486373166,
      "loss": 0.7151,
      "num_input_tokens_seen": 2281992,
      "step": 3495
    },
    {
      "epoch": 1.8343815513626835,
      "grad_norm": 0.06982380151748657,
      "learning_rate": 0.0009169287211740042,
      "loss": 0.4417,
      "num_input_tokens_seen": 2285576,
      "step": 3500
    },
    {
      "epoch": 1.8370020964360587,
      "grad_norm": 0.13095417618751526,
      "learning_rate": 0.0009182389937106919,
      "loss": 0.4197,
      "num_input_tokens_seen": 2288072,
      "step": 3505
    },
    {
      "epoch": 1.8396226415094339,
      "grad_norm": 0.06294064223766327,
      "learning_rate": 0.0009195492662473795,
      "loss": 0.381,
      "num_input_tokens_seen": 2291336,
      "step": 3510
    },
    {
      "epoch": 1.8422431865828093,
      "grad_norm": 0.08461927622556686,
      "learning_rate": 0.0009208595387840672,
      "loss": 0.5317,
      "num_input_tokens_seen": 2294152,
      "step": 3515
    },
    {
      "epoch": 1.8448637316561844,
      "grad_norm": 0.08476877212524414,
      "learning_rate": 0.0009221698113207547,
      "loss": 0.3897,
      "num_input_tokens_seen": 2296872,
      "step": 3520
    },
    {
      "epoch": 1.8474842767295598,
      "grad_norm": 0.08662398159503937,
      "learning_rate": 0.0009234800838574424,
      "loss": 0.4716,
      "num_input_tokens_seen": 2299528,
      "step": 3525
    },
    {
      "epoch": 1.850104821802935,
      "grad_norm": 0.15349288284778595,
      "learning_rate": 0.00092479035639413,
      "loss": 0.6484,
      "num_input_tokens_seen": 2301864,
      "step": 3530
    },
    {
      "epoch": 1.8527253668763102,
      "grad_norm": 0.04504687711596489,
      "learning_rate": 0.0009261006289308176,
      "loss": 0.4412,
      "num_input_tokens_seen": 2306088,
      "step": 3535
    },
    {
      "epoch": 1.8553459119496856,
      "grad_norm": 0.10282360017299652,
      "learning_rate": 0.0009274109014675053,
      "loss": 0.5363,
      "num_input_tokens_seen": 2309416,
      "step": 3540
    },
    {
      "epoch": 1.857966457023061,
      "grad_norm": 0.1003003790974617,
      "learning_rate": 0.0009287211740041929,
      "loss": 0.532,
      "num_input_tokens_seen": 2312328,
      "step": 3545
    },
    {
      "epoch": 1.8605870020964361,
      "grad_norm": 0.09905025362968445,
      "learning_rate": 0.0009300314465408806,
      "loss": 0.5517,
      "num_input_tokens_seen": 2314472,
      "step": 3550
    },
    {
      "epoch": 1.8632075471698113,
      "grad_norm": 0.07106028497219086,
      "learning_rate": 0.0009313417190775681,
      "loss": 0.3898,
      "num_input_tokens_seen": 2318664,
      "step": 3555
    },
    {
      "epoch": 1.8658280922431865,
      "grad_norm": 0.10611619800329208,
      "learning_rate": 0.0009326519916142557,
      "loss": 0.5736,
      "num_input_tokens_seen": 2323784,
      "step": 3560
    },
    {
      "epoch": 1.8684486373165619,
      "grad_norm": 0.10158761590719223,
      "learning_rate": 0.0009339622641509434,
      "loss": 0.5279,
      "num_input_tokens_seen": 2326536,
      "step": 3565
    },
    {
      "epoch": 1.8710691823899372,
      "grad_norm": 0.060628924518823624,
      "learning_rate": 0.000935272536687631,
      "loss": 0.4692,
      "num_input_tokens_seen": 2330440,
      "step": 3570
    },
    {
      "epoch": 1.8736897274633124,
      "grad_norm": 0.020615771412849426,
      "learning_rate": 0.0009365828092243187,
      "loss": 0.5676,
      "num_input_tokens_seen": 2335688,
      "step": 3575
    },
    {
      "epoch": 1.8763102725366876,
      "grad_norm": 0.08419796824455261,
      "learning_rate": 0.0009378930817610063,
      "loss": 0.4477,
      "num_input_tokens_seen": 2338568,
      "step": 3580
    },
    {
      "epoch": 1.8789308176100628,
      "grad_norm": 0.08628010749816895,
      "learning_rate": 0.0009392033542976939,
      "loss": 0.5171,
      "num_input_tokens_seen": 2342120,
      "step": 3585
    },
    {
      "epoch": 1.8815513626834381,
      "grad_norm": 0.05663001164793968,
      "learning_rate": 0.0009405136268343816,
      "loss": 0.4159,
      "num_input_tokens_seen": 2346184,
      "step": 3590
    },
    {
      "epoch": 1.8841719077568135,
      "grad_norm": 0.06957679986953735,
      "learning_rate": 0.0009418238993710692,
      "loss": 0.5966,
      "num_input_tokens_seen": 2348936,
      "step": 3595
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 0.09155772626399994,
      "learning_rate": 0.0009431341719077569,
      "loss": 0.4844,
      "num_input_tokens_seen": 2351976,
      "step": 3600
    },
    {
      "epoch": 1.8894129979035639,
      "grad_norm": 0.062229517847299576,
      "learning_rate": 0.0009444444444444445,
      "loss": 0.4954,
      "num_input_tokens_seen": 2357192,
      "step": 3605
    },
    {
      "epoch": 1.892033542976939,
      "grad_norm": 0.10126075148582458,
      "learning_rate": 0.0009457547169811322,
      "loss": 0.6317,
      "num_input_tokens_seen": 2360072,
      "step": 3610
    },
    {
      "epoch": 1.8946540880503144,
      "grad_norm": 0.07779504358768463,
      "learning_rate": 0.0009470649895178197,
      "loss": 0.4669,
      "num_input_tokens_seen": 2363656,
      "step": 3615
    },
    {
      "epoch": 1.8972746331236898,
      "grad_norm": 0.18746080994606018,
      "learning_rate": 0.0009483752620545073,
      "loss": 0.5419,
      "num_input_tokens_seen": 2366216,
      "step": 3620
    },
    {
      "epoch": 1.899895178197065,
      "grad_norm": 0.09995940327644348,
      "learning_rate": 0.000949685534591195,
      "loss": 0.5239,
      "num_input_tokens_seen": 2369448,
      "step": 3625
    },
    {
      "epoch": 1.9025157232704402,
      "grad_norm": 0.08168137818574905,
      "learning_rate": 0.0009509958071278826,
      "loss": 0.6532,
      "num_input_tokens_seen": 2372744,
      "step": 3630
    },
    {
      "epoch": 1.9051362683438156,
      "grad_norm": 0.09730476886034012,
      "learning_rate": 0.0009523060796645703,
      "loss": 0.4249,
      "num_input_tokens_seen": 2375080,
      "step": 3635
    },
    {
      "epoch": 1.9077568134171907,
      "grad_norm": 0.07151592522859573,
      "learning_rate": 0.0009536163522012579,
      "loss": 0.5104,
      "num_input_tokens_seen": 2377864,
      "step": 3640
    },
    {
      "epoch": 1.9103773584905661,
      "grad_norm": 0.08917061239480972,
      "learning_rate": 0.0009549266247379456,
      "loss": 0.4972,
      "num_input_tokens_seen": 2381224,
      "step": 3645
    },
    {
      "epoch": 1.9129979035639413,
      "grad_norm": 0.07791998982429504,
      "learning_rate": 0.0009562368972746331,
      "loss": 0.5206,
      "num_input_tokens_seen": 2384424,
      "step": 3650
    },
    {
      "epoch": 1.9156184486373165,
      "grad_norm": 0.12525926530361176,
      "learning_rate": 0.0009575471698113207,
      "loss": 0.4534,
      "num_input_tokens_seen": 2388936,
      "step": 3655
    },
    {
      "epoch": 1.9182389937106918,
      "grad_norm": 0.060787566006183624,
      "learning_rate": 0.0009588574423480084,
      "loss": 0.5267,
      "num_input_tokens_seen": 2392744,
      "step": 3660
    },
    {
      "epoch": 1.9208595387840672,
      "grad_norm": 0.08836202323436737,
      "learning_rate": 0.000960167714884696,
      "loss": 0.5947,
      "num_input_tokens_seen": 2395112,
      "step": 3665
    },
    {
      "epoch": 1.9234800838574424,
      "grad_norm": 0.06134742125868797,
      "learning_rate": 0.0009614779874213837,
      "loss": 0.4281,
      "num_input_tokens_seen": 2398344,
      "step": 3670
    },
    {
      "epoch": 1.9261006289308176,
      "grad_norm": 0.05426895245909691,
      "learning_rate": 0.0009627882599580712,
      "loss": 0.5068,
      "num_input_tokens_seen": 2401704,
      "step": 3675
    },
    {
      "epoch": 1.9287211740041927,
      "grad_norm": 0.11477774381637573,
      "learning_rate": 0.0009640985324947589,
      "loss": 0.3746,
      "num_input_tokens_seen": 2404168,
      "step": 3680
    },
    {
      "epoch": 1.9313417190775681,
      "grad_norm": 0.06332296878099442,
      "learning_rate": 0.0009654088050314465,
      "loss": 0.3896,
      "num_input_tokens_seen": 2407688,
      "step": 3685
    },
    {
      "epoch": 1.9339622641509435,
      "grad_norm": 0.04975622147321701,
      "learning_rate": 0.0009667190775681342,
      "loss": 0.4068,
      "num_input_tokens_seen": 2411528,
      "step": 3690
    },
    {
      "epoch": 1.9365828092243187,
      "grad_norm": 0.11527051776647568,
      "learning_rate": 0.0009680293501048219,
      "loss": 0.5376,
      "num_input_tokens_seen": 2414440,
      "step": 3695
    },
    {
      "epoch": 1.9392033542976939,
      "grad_norm": 0.07268032431602478,
      "learning_rate": 0.0009693396226415095,
      "loss": 0.4284,
      "num_input_tokens_seen": 2417448,
      "step": 3700
    },
    {
      "epoch": 1.941823899371069,
      "grad_norm": 0.11925869435071945,
      "learning_rate": 0.0009706498951781972,
      "loss": 0.4731,
      "num_input_tokens_seen": 2420680,
      "step": 3705
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 0.17098906636238098,
      "learning_rate": 0.0009719601677148847,
      "loss": 0.4777,
      "num_input_tokens_seen": 2423752,
      "step": 3710
    },
    {
      "epoch": 1.9470649895178198,
      "grad_norm": 0.07622753828763962,
      "learning_rate": 0.0009732704402515723,
      "loss": 0.5005,
      "num_input_tokens_seen": 2426216,
      "step": 3715
    },
    {
      "epoch": 1.949685534591195,
      "grad_norm": 0.0689966157078743,
      "learning_rate": 0.00097458071278826,
      "loss": 0.5479,
      "num_input_tokens_seen": 2429640,
      "step": 3720
    },
    {
      "epoch": 1.9523060796645701,
      "grad_norm": 0.08605426549911499,
      "learning_rate": 0.0009758909853249476,
      "loss": 0.4347,
      "num_input_tokens_seen": 2433224,
      "step": 3725
    },
    {
      "epoch": 1.9549266247379455,
      "grad_norm": 0.06444474309682846,
      "learning_rate": 0.0009772012578616353,
      "loss": 0.5512,
      "num_input_tokens_seen": 2437576,
      "step": 3730
    },
    {
      "epoch": 1.9575471698113207,
      "grad_norm": 0.11147456616163254,
      "learning_rate": 0.0009785115303983228,
      "loss": 0.4763,
      "num_input_tokens_seen": 2439976,
      "step": 3735
    },
    {
      "epoch": 1.960167714884696,
      "grad_norm": 0.0821961760520935,
      "learning_rate": 0.0009798218029350107,
      "loss": 0.4643,
      "num_input_tokens_seen": 2443272,
      "step": 3740
    },
    {
      "epoch": 1.9627882599580713,
      "grad_norm": 0.08453857153654099,
      "learning_rate": 0.0009811320754716981,
      "loss": 0.561,
      "num_input_tokens_seen": 2445832,
      "step": 3745
    },
    {
      "epoch": 1.9654088050314464,
      "grad_norm": 0.08184411376714706,
      "learning_rate": 0.0009824423480083858,
      "loss": 0.3808,
      "num_input_tokens_seen": 2448456,
      "step": 3750
    },
    {
      "epoch": 1.9680293501048218,
      "grad_norm": 0.09400838613510132,
      "learning_rate": 0.0009837526205450735,
      "loss": 0.554,
      "num_input_tokens_seen": 2451208,
      "step": 3755
    },
    {
      "epoch": 1.9706498951781972,
      "grad_norm": 0.08042267709970474,
      "learning_rate": 0.000985062893081761,
      "loss": 0.4463,
      "num_input_tokens_seen": 2454152,
      "step": 3760
    },
    {
      "epoch": 1.9732704402515724,
      "grad_norm": 0.26583439111709595,
      "learning_rate": 0.0009863731656184487,
      "loss": 0.3922,
      "num_input_tokens_seen": 2457736,
      "step": 3765
    },
    {
      "epoch": 1.9758909853249476,
      "grad_norm": 0.07687564939260483,
      "learning_rate": 0.0009876834381551364,
      "loss": 0.5966,
      "num_input_tokens_seen": 2461384,
      "step": 3770
    },
    {
      "epoch": 1.9785115303983227,
      "grad_norm": 0.09390607476234436,
      "learning_rate": 0.0009889937106918238,
      "loss": 0.3799,
      "num_input_tokens_seen": 2464776,
      "step": 3775
    },
    {
      "epoch": 1.9811320754716981,
      "grad_norm": 0.20160789787769318,
      "learning_rate": 0.0009903039832285115,
      "loss": 0.4618,
      "num_input_tokens_seen": 2467144,
      "step": 3780
    },
    {
      "epoch": 1.9837526205450735,
      "grad_norm": 0.0646854117512703,
      "learning_rate": 0.0009916142557651992,
      "loss": 0.599,
      "num_input_tokens_seen": 2469736,
      "step": 3785
    },
    {
      "epoch": 1.9863731656184487,
      "grad_norm": 0.06803485751152039,
      "learning_rate": 0.000992924528301887,
      "loss": 0.3779,
      "num_input_tokens_seen": 2474536,
      "step": 3790
    },
    {
      "epoch": 1.9889937106918238,
      "grad_norm": 0.07559171319007874,
      "learning_rate": 0.0009942348008385744,
      "loss": 0.5266,
      "num_input_tokens_seen": 2477704,
      "step": 3795
    },
    {
      "epoch": 1.991614255765199,
      "grad_norm": 0.06744515895843506,
      "learning_rate": 0.000995545073375262,
      "loss": 0.4644,
      "num_input_tokens_seen": 2480296,
      "step": 3800
    },
    {
      "epoch": 1.9942348008385744,
      "grad_norm": 0.09741527587175369,
      "learning_rate": 0.0009968553459119497,
      "loss": 0.5545,
      "num_input_tokens_seen": 2483368,
      "step": 3805
    },
    {
      "epoch": 1.9968553459119498,
      "grad_norm": 0.06935109198093414,
      "learning_rate": 0.0009981656184486372,
      "loss": 0.3954,
      "num_input_tokens_seen": 2486440,
      "step": 3810
    },
    {
      "epoch": 1.999475890985325,
      "grad_norm": 0.09183430671691895,
      "learning_rate": 0.000999475890985325,
      "loss": 0.3548,
      "num_input_tokens_seen": 2489352,
      "step": 3815
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.46972936391830444,
      "eval_runtime": 13.555,
      "eval_samples_per_second": 62.56,
      "eval_steps_per_second": 15.64,
      "num_input_tokens_seen": 2489456,
      "step": 3816
    },
    {
      "epoch": 2.0020964360587,
      "grad_norm": 0.16741476953029633,
      "learning_rate": 0.0009999999811730266,
      "loss": 0.5616,
      "num_input_tokens_seen": 2492112,
      "step": 3820
    },
    {
      "epoch": 2.0047169811320753,
      "grad_norm": 0.09701599180698395,
      "learning_rate": 0.000999999866119305,
      "loss": 0.3784,
      "num_input_tokens_seen": 2494800,
      "step": 3825
    },
    {
      "epoch": 2.007337526205451,
      "grad_norm": 0.11555616557598114,
      "learning_rate": 0.000999999646471316,
      "loss": 0.4886,
      "num_input_tokens_seen": 2498896,
      "step": 3830
    },
    {
      "epoch": 2.009958071278826,
      "grad_norm": 0.07225560396909714,
      "learning_rate": 0.0009999993222291053,
      "loss": 0.5382,
      "num_input_tokens_seen": 2502160,
      "step": 3835
    },
    {
      "epoch": 2.0125786163522013,
      "grad_norm": 0.07938838005065918,
      "learning_rate": 0.0009999988933927404,
      "loss": 0.4978,
      "num_input_tokens_seen": 2504752,
      "step": 3840
    },
    {
      "epoch": 2.0151991614255764,
      "grad_norm": 0.1038142740726471,
      "learning_rate": 0.0009999983599623115,
      "loss": 0.4899,
      "num_input_tokens_seen": 2507536,
      "step": 3845
    },
    {
      "epoch": 2.0178197064989516,
      "grad_norm": 0.06795326620340347,
      "learning_rate": 0.0009999977219379299,
      "loss": 0.4819,
      "num_input_tokens_seen": 2510352,
      "step": 3850
    },
    {
      "epoch": 2.020440251572327,
      "grad_norm": 0.08111339062452316,
      "learning_rate": 0.000999996979319729,
      "loss": 0.43,
      "num_input_tokens_seen": 2512784,
      "step": 3855
    },
    {
      "epoch": 2.0230607966457024,
      "grad_norm": 0.07354375720024109,
      "learning_rate": 0.0009999961321078645,
      "loss": 0.4149,
      "num_input_tokens_seen": 2516304,
      "step": 3860
    },
    {
      "epoch": 2.0256813417190775,
      "grad_norm": 0.11655111610889435,
      "learning_rate": 0.0009999951803025134,
      "loss": 0.5254,
      "num_input_tokens_seen": 2519184,
      "step": 3865
    },
    {
      "epoch": 2.0283018867924527,
      "grad_norm": 0.06891246885061264,
      "learning_rate": 0.0009999941239038747,
      "loss": 0.5622,
      "num_input_tokens_seen": 2522448,
      "step": 3870
    },
    {
      "epoch": 2.030922431865828,
      "grad_norm": 0.10467232763767242,
      "learning_rate": 0.0009999929629121696,
      "loss": 0.4468,
      "num_input_tokens_seen": 2525744,
      "step": 3875
    },
    {
      "epoch": 2.0335429769392035,
      "grad_norm": 0.08850888162851334,
      "learning_rate": 0.000999991697327641,
      "loss": 0.4294,
      "num_input_tokens_seen": 2528240,
      "step": 3880
    },
    {
      "epoch": 2.0361635220125787,
      "grad_norm": 0.11044721305370331,
      "learning_rate": 0.0009999903271505535,
      "loss": 0.4618,
      "num_input_tokens_seen": 2530928,
      "step": 3885
    },
    {
      "epoch": 2.038784067085954,
      "grad_norm": 0.07427483052015305,
      "learning_rate": 0.0009999888523811935,
      "loss": 0.4684,
      "num_input_tokens_seen": 2533936,
      "step": 3890
    },
    {
      "epoch": 2.041404612159329,
      "grad_norm": 0.100394606590271,
      "learning_rate": 0.00099998727301987,
      "loss": 0.5945,
      "num_input_tokens_seen": 2536720,
      "step": 3895
    },
    {
      "epoch": 2.0440251572327046,
      "grad_norm": 0.04568382725119591,
      "learning_rate": 0.000999985589066913,
      "loss": 0.4606,
      "num_input_tokens_seen": 2540656,
      "step": 3900
    },
    {
      "epoch": 2.04664570230608,
      "grad_norm": 0.06736352294683456,
      "learning_rate": 0.000999983800522675,
      "loss": 0.4367,
      "num_input_tokens_seen": 2545040,
      "step": 3905
    },
    {
      "epoch": 2.049266247379455,
      "grad_norm": 0.12414301186800003,
      "learning_rate": 0.00099998190738753,
      "loss": 0.718,
      "num_input_tokens_seen": 2547888,
      "step": 3910
    },
    {
      "epoch": 2.05188679245283,
      "grad_norm": 0.07695544511079788,
      "learning_rate": 0.000999979909661874,
      "loss": 0.3676,
      "num_input_tokens_seen": 2550256,
      "step": 3915
    },
    {
      "epoch": 2.0545073375262053,
      "grad_norm": 0.046180374920368195,
      "learning_rate": 0.000999977807346125,
      "loss": 0.3702,
      "num_input_tokens_seen": 2553712,
      "step": 3920
    },
    {
      "epoch": 2.057127882599581,
      "grad_norm": 0.0735442042350769,
      "learning_rate": 0.000999975600440723,
      "loss": 0.523,
      "num_input_tokens_seen": 2557392,
      "step": 3925
    },
    {
      "epoch": 2.059748427672956,
      "grad_norm": 0.07614301145076752,
      "learning_rate": 0.000999973288946129,
      "loss": 0.6425,
      "num_input_tokens_seen": 2560656,
      "step": 3930
    },
    {
      "epoch": 2.0623689727463312,
      "grad_norm": 0.08235225081443787,
      "learning_rate": 0.0009999708728628271,
      "loss": 0.4118,
      "num_input_tokens_seen": 2563984,
      "step": 3935
    },
    {
      "epoch": 2.0649895178197064,
      "grad_norm": 0.48397719860076904,
      "learning_rate": 0.0009999683521913227,
      "loss": 0.593,
      "num_input_tokens_seen": 2567088,
      "step": 3940
    },
    {
      "epoch": 2.0676100628930816,
      "grad_norm": 0.08321993798017502,
      "learning_rate": 0.000999965726932143,
      "loss": 0.4266,
      "num_input_tokens_seen": 2570288,
      "step": 3945
    },
    {
      "epoch": 2.070230607966457,
      "grad_norm": 0.10549290478229523,
      "learning_rate": 0.0009999629970858371,
      "loss": 0.5064,
      "num_input_tokens_seen": 2573136,
      "step": 3950
    },
    {
      "epoch": 2.0728511530398324,
      "grad_norm": 0.09410151839256287,
      "learning_rate": 0.000999960162652976,
      "loss": 0.5041,
      "num_input_tokens_seen": 2576912,
      "step": 3955
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 0.2179020196199417,
      "learning_rate": 0.0009999572236341527,
      "loss": 0.5134,
      "num_input_tokens_seen": 2579824,
      "step": 3960
    },
    {
      "epoch": 2.0780922431865827,
      "grad_norm": 0.09497710317373276,
      "learning_rate": 0.000999954180029982,
      "loss": 0.4804,
      "num_input_tokens_seen": 2583664,
      "step": 3965
    },
    {
      "epoch": 2.080712788259958,
      "grad_norm": 0.08071190863847733,
      "learning_rate": 0.0009999510318411007,
      "loss": 0.4763,
      "num_input_tokens_seen": 2587344,
      "step": 3970
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 0.08786571770906448,
      "learning_rate": 0.0009999477790681673,
      "loss": 0.4741,
      "num_input_tokens_seen": 2590800,
      "step": 3975
    },
    {
      "epoch": 2.0859538784067087,
      "grad_norm": 0.11985490471124649,
      "learning_rate": 0.000999944421711862,
      "loss": 0.6152,
      "num_input_tokens_seen": 2593872,
      "step": 3980
    },
    {
      "epoch": 2.088574423480084,
      "grad_norm": 0.0684785395860672,
      "learning_rate": 0.0009999409597728876,
      "loss": 0.5628,
      "num_input_tokens_seen": 2596976,
      "step": 3985
    },
    {
      "epoch": 2.091194968553459,
      "grad_norm": 0.0999487116932869,
      "learning_rate": 0.0009999373932519679,
      "loss": 0.4711,
      "num_input_tokens_seen": 2599824,
      "step": 3990
    },
    {
      "epoch": 2.0938155136268346,
      "grad_norm": 0.07008538395166397,
      "learning_rate": 0.0009999337221498492,
      "loss": 0.5345,
      "num_input_tokens_seen": 2602992,
      "step": 3995
    },
    {
      "epoch": 2.0964360587002098,
      "grad_norm": 0.10025153309106827,
      "learning_rate": 0.0009999299464672996,
      "loss": 0.3546,
      "num_input_tokens_seen": 2606448,
      "step": 4000
    },
    {
      "epoch": 2.099056603773585,
      "grad_norm": 0.08359622955322266,
      "learning_rate": 0.0009999260662051084,
      "loss": 0.3515,
      "num_input_tokens_seen": 2610864,
      "step": 4005
    },
    {
      "epoch": 2.10167714884696,
      "grad_norm": 0.06249988451600075,
      "learning_rate": 0.0009999220813640877,
      "loss": 0.4228,
      "num_input_tokens_seen": 2614256,
      "step": 4010
    },
    {
      "epoch": 2.1042976939203353,
      "grad_norm": 0.06265965849161148,
      "learning_rate": 0.000999917991945071,
      "loss": 0.6422,
      "num_input_tokens_seen": 2617072,
      "step": 4015
    },
    {
      "epoch": 2.106918238993711,
      "grad_norm": 0.11925460398197174,
      "learning_rate": 0.0009999137979489136,
      "loss": 0.4634,
      "num_input_tokens_seen": 2619792,
      "step": 4020
    },
    {
      "epoch": 2.109538784067086,
      "grad_norm": 0.09731762111186981,
      "learning_rate": 0.000999909499376493,
      "loss": 0.391,
      "num_input_tokens_seen": 2622448,
      "step": 4025
    },
    {
      "epoch": 2.1121593291404612,
      "grad_norm": 0.10700615495443344,
      "learning_rate": 0.0009999050962287085,
      "loss": 0.5344,
      "num_input_tokens_seen": 2625328,
      "step": 4030
    },
    {
      "epoch": 2.1147798742138364,
      "grad_norm": 0.08905218541622162,
      "learning_rate": 0.0009999005885064811,
      "loss": 0.5176,
      "num_input_tokens_seen": 2628336,
      "step": 4035
    },
    {
      "epoch": 2.1174004192872116,
      "grad_norm": 0.07599812746047974,
      "learning_rate": 0.0009998959762107537,
      "loss": 0.487,
      "num_input_tokens_seen": 2634128,
      "step": 4040
    },
    {
      "epoch": 2.120020964360587,
      "grad_norm": 0.0685771033167839,
      "learning_rate": 0.0009998912593424914,
      "loss": 0.417,
      "num_input_tokens_seen": 2638672,
      "step": 4045
    },
    {
      "epoch": 2.1226415094339623,
      "grad_norm": 0.04144611954689026,
      "learning_rate": 0.0009998864379026804,
      "loss": 0.5273,
      "num_input_tokens_seen": 2642768,
      "step": 4050
    },
    {
      "epoch": 2.1252620545073375,
      "grad_norm": 0.15551333129405975,
      "learning_rate": 0.00099988151189233,
      "loss": 0.4735,
      "num_input_tokens_seen": 2645552,
      "step": 4055
    },
    {
      "epoch": 2.1278825995807127,
      "grad_norm": 0.08341532945632935,
      "learning_rate": 0.00099987648131247,
      "loss": 0.4296,
      "num_input_tokens_seen": 2647728,
      "step": 4060
    },
    {
      "epoch": 2.130503144654088,
      "grad_norm": 0.14404654502868652,
      "learning_rate": 0.000999871346164153,
      "loss": 0.5692,
      "num_input_tokens_seen": 2650992,
      "step": 4065
    },
    {
      "epoch": 2.1331236897274635,
      "grad_norm": 0.11399997025728226,
      "learning_rate": 0.0009998661064484532,
      "loss": 0.522,
      "num_input_tokens_seen": 2653744,
      "step": 4070
    },
    {
      "epoch": 2.1357442348008386,
      "grad_norm": 0.25852710008621216,
      "learning_rate": 0.0009998607621664666,
      "loss": 0.298,
      "num_input_tokens_seen": 2656336,
      "step": 4075
    },
    {
      "epoch": 2.138364779874214,
      "grad_norm": 0.09817013144493103,
      "learning_rate": 0.0009998553133193113,
      "loss": 0.7396,
      "num_input_tokens_seen": 2659216,
      "step": 4080
    },
    {
      "epoch": 2.140985324947589,
      "grad_norm": 0.1284898966550827,
      "learning_rate": 0.0009998497599081272,
      "loss": 0.5666,
      "num_input_tokens_seen": 2662224,
      "step": 4085
    },
    {
      "epoch": 2.1436058700209646,
      "grad_norm": 0.07229320704936981,
      "learning_rate": 0.0009998441019340761,
      "loss": 0.3735,
      "num_input_tokens_seen": 2665936,
      "step": 4090
    },
    {
      "epoch": 2.1462264150943398,
      "grad_norm": 0.06471875309944153,
      "learning_rate": 0.000999838339398341,
      "loss": 0.3677,
      "num_input_tokens_seen": 2669648,
      "step": 4095
    },
    {
      "epoch": 2.148846960167715,
      "grad_norm": 0.13178007304668427,
      "learning_rate": 0.000999832472302128,
      "loss": 0.7111,
      "num_input_tokens_seen": 2673552,
      "step": 4100
    },
    {
      "epoch": 2.15146750524109,
      "grad_norm": 0.09492270648479462,
      "learning_rate": 0.0009998265006466642,
      "loss": 0.631,
      "num_input_tokens_seen": 2675824,
      "step": 4105
    },
    {
      "epoch": 2.1540880503144653,
      "grad_norm": 0.185032919049263,
      "learning_rate": 0.0009998204244331987,
      "loss": 0.5383,
      "num_input_tokens_seen": 2678256,
      "step": 4110
    },
    {
      "epoch": 2.156708595387841,
      "grad_norm": 0.05713917315006256,
      "learning_rate": 0.0009998142436630027,
      "loss": 0.4123,
      "num_input_tokens_seen": 2682224,
      "step": 4115
    },
    {
      "epoch": 2.159329140461216,
      "grad_norm": 0.0748058333992958,
      "learning_rate": 0.0009998079583373692,
      "loss": 0.3469,
      "num_input_tokens_seen": 2684848,
      "step": 4120
    },
    {
      "epoch": 2.161949685534591,
      "grad_norm": 0.06341387331485748,
      "learning_rate": 0.0009998015684576128,
      "loss": 0.5241,
      "num_input_tokens_seen": 2688816,
      "step": 4125
    },
    {
      "epoch": 2.1645702306079664,
      "grad_norm": 0.06370033323764801,
      "learning_rate": 0.0009997950740250703,
      "loss": 0.9151,
      "num_input_tokens_seen": 2691728,
      "step": 4130
    },
    {
      "epoch": 2.1671907756813416,
      "grad_norm": 0.1644037663936615,
      "learning_rate": 0.0009997884750411004,
      "loss": 0.5796,
      "num_input_tokens_seen": 2694608,
      "step": 4135
    },
    {
      "epoch": 2.169811320754717,
      "grad_norm": 0.05469399318099022,
      "learning_rate": 0.0009997817715070832,
      "loss": 0.5386,
      "num_input_tokens_seen": 2697872,
      "step": 4140
    },
    {
      "epoch": 2.1724318658280923,
      "grad_norm": 0.05438226833939552,
      "learning_rate": 0.0009997749634244213,
      "loss": 0.4297,
      "num_input_tokens_seen": 2701040,
      "step": 4145
    },
    {
      "epoch": 2.1750524109014675,
      "grad_norm": 0.13632188737392426,
      "learning_rate": 0.000999768050794539,
      "loss": 0.6311,
      "num_input_tokens_seen": 2703760,
      "step": 4150
    },
    {
      "epoch": 2.1776729559748427,
      "grad_norm": 0.053851909935474396,
      "learning_rate": 0.0009997610336188818,
      "loss": 0.4769,
      "num_input_tokens_seen": 2706480,
      "step": 4155
    },
    {
      "epoch": 2.180293501048218,
      "grad_norm": 0.0701640248298645,
      "learning_rate": 0.0009997539118989183,
      "loss": 0.3312,
      "num_input_tokens_seen": 2709488,
      "step": 4160
    },
    {
      "epoch": 2.1829140461215935,
      "grad_norm": 0.04927217215299606,
      "learning_rate": 0.0009997466856361376,
      "loss": 0.4199,
      "num_input_tokens_seen": 2712880,
      "step": 4165
    },
    {
      "epoch": 2.1855345911949686,
      "grad_norm": 0.08672579377889633,
      "learning_rate": 0.000999739354832052,
      "loss": 0.4885,
      "num_input_tokens_seen": 2716112,
      "step": 4170
    },
    {
      "epoch": 2.188155136268344,
      "grad_norm": 0.08901503682136536,
      "learning_rate": 0.0009997319194881944,
      "loss": 0.5274,
      "num_input_tokens_seen": 2719088,
      "step": 4175
    },
    {
      "epoch": 2.190775681341719,
      "grad_norm": 0.08702191710472107,
      "learning_rate": 0.0009997243796061204,
      "loss": 0.4615,
      "num_input_tokens_seen": 2722064,
      "step": 4180
    },
    {
      "epoch": 2.1933962264150946,
      "grad_norm": 0.09470000863075256,
      "learning_rate": 0.0009997167351874076,
      "loss": 0.3371,
      "num_input_tokens_seen": 2725072,
      "step": 4185
    },
    {
      "epoch": 2.1960167714884697,
      "grad_norm": 0.07662735134363174,
      "learning_rate": 0.0009997089862336548,
      "loss": 0.4337,
      "num_input_tokens_seen": 2728368,
      "step": 4190
    },
    {
      "epoch": 2.198637316561845,
      "grad_norm": 0.06557830423116684,
      "learning_rate": 0.000999701132746483,
      "loss": 0.5336,
      "num_input_tokens_seen": 2731984,
      "step": 4195
    },
    {
      "epoch": 2.20125786163522,
      "grad_norm": 0.09207457304000854,
      "learning_rate": 0.0009996931747275352,
      "loss": 0.4665,
      "num_input_tokens_seen": 2735632,
      "step": 4200
    },
    {
      "epoch": 2.2038784067085953,
      "grad_norm": 0.05295583978295326,
      "learning_rate": 0.000999685112178476,
      "loss": 0.5085,
      "num_input_tokens_seen": 2738928,
      "step": 4205
    },
    {
      "epoch": 2.2064989517819704,
      "grad_norm": 0.07615093886852264,
      "learning_rate": 0.0009996769451009922,
      "loss": 0.4811,
      "num_input_tokens_seen": 2743248,
      "step": 4210
    },
    {
      "epoch": 2.209119496855346,
      "grad_norm": 0.09741240739822388,
      "learning_rate": 0.000999668673496792,
      "loss": 0.6042,
      "num_input_tokens_seen": 2745744,
      "step": 4215
    },
    {
      "epoch": 2.211740041928721,
      "grad_norm": 0.07177629321813583,
      "learning_rate": 0.0009996602973676057,
      "loss": 0.4771,
      "num_input_tokens_seen": 2748432,
      "step": 4220
    },
    {
      "epoch": 2.2143605870020964,
      "grad_norm": 0.1504851132631302,
      "learning_rate": 0.0009996518167151858,
      "loss": 0.5978,
      "num_input_tokens_seen": 2751280,
      "step": 4225
    },
    {
      "epoch": 2.2169811320754715,
      "grad_norm": 0.09846767783164978,
      "learning_rate": 0.000999643231541306,
      "loss": 0.5655,
      "num_input_tokens_seen": 2754096,
      "step": 4230
    },
    {
      "epoch": 2.219601677148847,
      "grad_norm": 0.07341635227203369,
      "learning_rate": 0.0009996345418477625,
      "loss": 0.6119,
      "num_input_tokens_seen": 2757200,
      "step": 4235
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.11070431768894196,
      "learning_rate": 0.000999625747636373,
      "loss": 0.5338,
      "num_input_tokens_seen": 2760240,
      "step": 4240
    },
    {
      "epoch": 2.2248427672955975,
      "grad_norm": 0.1788286417722702,
      "learning_rate": 0.000999616848908977,
      "loss": 0.4632,
      "num_input_tokens_seen": 2763248,
      "step": 4245
    },
    {
      "epoch": 2.2274633123689727,
      "grad_norm": 0.050116438418626785,
      "learning_rate": 0.0009996078456674363,
      "loss": 0.488,
      "num_input_tokens_seen": 2766192,
      "step": 4250
    },
    {
      "epoch": 2.230083857442348,
      "grad_norm": 0.04793735221028328,
      "learning_rate": 0.000999598737913634,
      "loss": 0.4971,
      "num_input_tokens_seen": 2770160,
      "step": 4255
    },
    {
      "epoch": 2.2327044025157234,
      "grad_norm": 0.05358132719993591,
      "learning_rate": 0.0009995895256494755,
      "loss": 0.4129,
      "num_input_tokens_seen": 2773424,
      "step": 4260
    },
    {
      "epoch": 2.2353249475890986,
      "grad_norm": 0.0741594061255455,
      "learning_rate": 0.000999580208876888,
      "loss": 0.467,
      "num_input_tokens_seen": 2776240,
      "step": 4265
    },
    {
      "epoch": 2.237945492662474,
      "grad_norm": 0.12693287432193756,
      "learning_rate": 0.0009995707875978198,
      "loss": 0.489,
      "num_input_tokens_seen": 2779088,
      "step": 4270
    },
    {
      "epoch": 2.240566037735849,
      "grad_norm": 0.0808650404214859,
      "learning_rate": 0.0009995612618142428,
      "loss": 0.4194,
      "num_input_tokens_seen": 2782032,
      "step": 4275
    },
    {
      "epoch": 2.243186582809224,
      "grad_norm": 0.06595594435930252,
      "learning_rate": 0.000999551631528149,
      "loss": 0.6451,
      "num_input_tokens_seen": 2786000,
      "step": 4280
    },
    {
      "epoch": 2.2458071278825997,
      "grad_norm": 0.10713280737400055,
      "learning_rate": 0.000999541896741553,
      "loss": 0.3774,
      "num_input_tokens_seen": 2789584,
      "step": 4285
    },
    {
      "epoch": 2.248427672955975,
      "grad_norm": 0.09415583312511444,
      "learning_rate": 0.0009995320574564912,
      "loss": 0.5437,
      "num_input_tokens_seen": 2792976,
      "step": 4290
    },
    {
      "epoch": 2.25104821802935,
      "grad_norm": 0.09064682573080063,
      "learning_rate": 0.000999522113675022,
      "loss": 0.5111,
      "num_input_tokens_seen": 2795856,
      "step": 4295
    },
    {
      "epoch": 2.2536687631027252,
      "grad_norm": 0.10145796090364456,
      "learning_rate": 0.0009995120653992255,
      "loss": 0.4938,
      "num_input_tokens_seen": 2798384,
      "step": 4300
    },
    {
      "epoch": 2.2562893081761004,
      "grad_norm": 0.048888660967350006,
      "learning_rate": 0.0009995019126312035,
      "loss": 0.369,
      "num_input_tokens_seen": 2801552,
      "step": 4305
    },
    {
      "epoch": 2.258909853249476,
      "grad_norm": 0.15700706839561462,
      "learning_rate": 0.0009994916553730802,
      "loss": 0.5605,
      "num_input_tokens_seen": 2804816,
      "step": 4310
    },
    {
      "epoch": 2.261530398322851,
      "grad_norm": 0.20184287428855896,
      "learning_rate": 0.000999481293627001,
      "loss": 0.77,
      "num_input_tokens_seen": 2807536,
      "step": 4315
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 0.0515475757420063,
      "learning_rate": 0.0009994708273951337,
      "loss": 0.4729,
      "num_input_tokens_seen": 2812080,
      "step": 4320
    },
    {
      "epoch": 2.2667714884696015,
      "grad_norm": 0.20674680173397064,
      "learning_rate": 0.0009994602566796673,
      "loss": 0.5175,
      "num_input_tokens_seen": 2814768,
      "step": 4325
    },
    {
      "epoch": 2.269392033542977,
      "grad_norm": 0.13524845242500305,
      "learning_rate": 0.0009994495814828136,
      "loss": 0.5362,
      "num_input_tokens_seen": 2817328,
      "step": 4330
    },
    {
      "epoch": 2.2720125786163523,
      "grad_norm": 0.0687469094991684,
      "learning_rate": 0.0009994388018068055,
      "loss": 0.6179,
      "num_input_tokens_seen": 2820624,
      "step": 4335
    },
    {
      "epoch": 2.2746331236897275,
      "grad_norm": 0.08114948868751526,
      "learning_rate": 0.0009994279176538977,
      "loss": 0.396,
      "num_input_tokens_seen": 2824720,
      "step": 4340
    },
    {
      "epoch": 2.2772536687631026,
      "grad_norm": 0.06370213627815247,
      "learning_rate": 0.0009994169290263675,
      "loss": 0.5895,
      "num_input_tokens_seen": 2828560,
      "step": 4345
    },
    {
      "epoch": 2.279874213836478,
      "grad_norm": 0.06822328269481659,
      "learning_rate": 0.0009994058359265132,
      "loss": 0.5509,
      "num_input_tokens_seen": 2832080,
      "step": 4350
    },
    {
      "epoch": 2.2824947589098534,
      "grad_norm": 0.18411794304847717,
      "learning_rate": 0.0009993946383566558,
      "loss": 0.5834,
      "num_input_tokens_seen": 2834640,
      "step": 4355
    },
    {
      "epoch": 2.2851153039832286,
      "grad_norm": 0.07774952054023743,
      "learning_rate": 0.0009993833363191374,
      "loss": 0.4968,
      "num_input_tokens_seen": 2837136,
      "step": 4360
    },
    {
      "epoch": 2.2877358490566038,
      "grad_norm": 0.057716790586709976,
      "learning_rate": 0.0009993719298163222,
      "loss": 0.479,
      "num_input_tokens_seen": 2841040,
      "step": 4365
    },
    {
      "epoch": 2.290356394129979,
      "grad_norm": 0.09857672452926636,
      "learning_rate": 0.0009993604188505965,
      "loss": 0.5332,
      "num_input_tokens_seen": 2843984,
      "step": 4370
    },
    {
      "epoch": 2.2929769392033545,
      "grad_norm": 0.10427994281053543,
      "learning_rate": 0.000999348803424368,
      "loss": 0.6361,
      "num_input_tokens_seen": 2846832,
      "step": 4375
    },
    {
      "epoch": 2.2955974842767297,
      "grad_norm": 0.037421587854623795,
      "learning_rate": 0.000999337083540067,
      "loss": 0.4098,
      "num_input_tokens_seen": 2850512,
      "step": 4380
    },
    {
      "epoch": 2.298218029350105,
      "grad_norm": 0.07464276999235153,
      "learning_rate": 0.0009993252592001448,
      "loss": 0.3743,
      "num_input_tokens_seen": 2854800,
      "step": 4385
    },
    {
      "epoch": 2.30083857442348,
      "grad_norm": 0.124650739133358,
      "learning_rate": 0.0009993133304070747,
      "loss": 0.3949,
      "num_input_tokens_seen": 2857552,
      "step": 4390
    },
    {
      "epoch": 2.3034591194968552,
      "grad_norm": 0.09998011589050293,
      "learning_rate": 0.0009993012971633527,
      "loss": 0.4984,
      "num_input_tokens_seen": 2860432,
      "step": 4395
    },
    {
      "epoch": 2.3060796645702304,
      "grad_norm": 0.06403542309999466,
      "learning_rate": 0.0009992891594714952,
      "loss": 0.5719,
      "num_input_tokens_seen": 2863568,
      "step": 4400
    },
    {
      "epoch": 2.308700209643606,
      "grad_norm": 0.06300035864114761,
      "learning_rate": 0.0009992769173340422,
      "loss": 0.411,
      "num_input_tokens_seen": 2865968,
      "step": 4405
    },
    {
      "epoch": 2.311320754716981,
      "grad_norm": 0.09709267318248749,
      "learning_rate": 0.000999264570753554,
      "loss": 0.6423,
      "num_input_tokens_seen": 2868848,
      "step": 4410
    },
    {
      "epoch": 2.3139412997903563,
      "grad_norm": 0.07538067549467087,
      "learning_rate": 0.0009992521197326135,
      "loss": 0.425,
      "num_input_tokens_seen": 2871792,
      "step": 4415
    },
    {
      "epoch": 2.3165618448637315,
      "grad_norm": 0.05462957173585892,
      "learning_rate": 0.0009992395642738252,
      "loss": 0.491,
      "num_input_tokens_seen": 2874832,
      "step": 4420
    },
    {
      "epoch": 2.319182389937107,
      "grad_norm": 0.15277798473834991,
      "learning_rate": 0.0009992269043798158,
      "loss": 0.4954,
      "num_input_tokens_seen": 2877648,
      "step": 4425
    },
    {
      "epoch": 2.3218029350104823,
      "grad_norm": 0.09966146945953369,
      "learning_rate": 0.0009992141400532336,
      "loss": 0.5049,
      "num_input_tokens_seen": 2880016,
      "step": 4430
    },
    {
      "epoch": 2.3244234800838575,
      "grad_norm": 0.11548011004924774,
      "learning_rate": 0.0009992012712967484,
      "loss": 0.6229,
      "num_input_tokens_seen": 2883184,
      "step": 4435
    },
    {
      "epoch": 2.3270440251572326,
      "grad_norm": 0.08975134789943695,
      "learning_rate": 0.0009991882981130525,
      "loss": 0.4027,
      "num_input_tokens_seen": 2887408,
      "step": 4440
    },
    {
      "epoch": 2.329664570230608,
      "grad_norm": 0.031552933156490326,
      "learning_rate": 0.0009991752205048597,
      "loss": 0.2889,
      "num_input_tokens_seen": 2892912,
      "step": 4445
    },
    {
      "epoch": 2.3322851153039834,
      "grad_norm": 0.05612563714385033,
      "learning_rate": 0.0009991620384749058,
      "loss": 0.5234,
      "num_input_tokens_seen": 2895504,
      "step": 4450
    },
    {
      "epoch": 2.3349056603773586,
      "grad_norm": 0.06784063577651978,
      "learning_rate": 0.0009991487520259479,
      "loss": 0.5843,
      "num_input_tokens_seen": 2900208,
      "step": 4455
    },
    {
      "epoch": 2.3375262054507338,
      "grad_norm": 0.07202764600515366,
      "learning_rate": 0.000999135361160766,
      "loss": 0.403,
      "num_input_tokens_seen": 2903568,
      "step": 4460
    },
    {
      "epoch": 2.340146750524109,
      "grad_norm": 0.08297352492809296,
      "learning_rate": 0.0009991218658821608,
      "loss": 0.5729,
      "num_input_tokens_seen": 2906864,
      "step": 4465
    },
    {
      "epoch": 2.342767295597484,
      "grad_norm": 0.11451581865549088,
      "learning_rate": 0.0009991082661929556,
      "loss": 0.4312,
      "num_input_tokens_seen": 2910480,
      "step": 4470
    },
    {
      "epoch": 2.3453878406708597,
      "grad_norm": 0.07715889066457748,
      "learning_rate": 0.000999094562095995,
      "loss": 0.4303,
      "num_input_tokens_seen": 2913552,
      "step": 4475
    },
    {
      "epoch": 2.348008385744235,
      "grad_norm": 0.06055283173918724,
      "learning_rate": 0.0009990807535941461,
      "loss": 0.5476,
      "num_input_tokens_seen": 2916368,
      "step": 4480
    },
    {
      "epoch": 2.35062893081761,
      "grad_norm": 0.044763848185539246,
      "learning_rate": 0.0009990668406902976,
      "loss": 0.5049,
      "num_input_tokens_seen": 2918960,
      "step": 4485
    },
    {
      "epoch": 2.353249475890985,
      "grad_norm": 0.059303078800439835,
      "learning_rate": 0.0009990528233873594,
      "loss": 0.4902,
      "num_input_tokens_seen": 2922768,
      "step": 4490
    },
    {
      "epoch": 2.3558700209643604,
      "grad_norm": 0.057426437735557556,
      "learning_rate": 0.0009990387016882642,
      "loss": 0.5043,
      "num_input_tokens_seen": 2925840,
      "step": 4495
    },
    {
      "epoch": 2.358490566037736,
      "grad_norm": 0.07240378856658936,
      "learning_rate": 0.000999024475595966,
      "loss": 0.4153,
      "num_input_tokens_seen": 2929008,
      "step": 4500
    },
    {
      "epoch": 2.361111111111111,
      "grad_norm": 0.08123394101858139,
      "learning_rate": 0.0009990101451134406,
      "loss": 0.4942,
      "num_input_tokens_seen": 2931632,
      "step": 4505
    },
    {
      "epoch": 2.3637316561844863,
      "grad_norm": 0.06013911962509155,
      "learning_rate": 0.0009989957102436858,
      "loss": 0.558,
      "num_input_tokens_seen": 2935824,
      "step": 4510
    },
    {
      "epoch": 2.3663522012578615,
      "grad_norm": 0.10351747274398804,
      "learning_rate": 0.0009989811709897212,
      "loss": 0.3717,
      "num_input_tokens_seen": 2938832,
      "step": 4515
    },
    {
      "epoch": 2.368972746331237,
      "grad_norm": 0.05545517057180405,
      "learning_rate": 0.0009989665273545884,
      "loss": 0.4439,
      "num_input_tokens_seen": 2942096,
      "step": 4520
    },
    {
      "epoch": 2.3715932914046123,
      "grad_norm": 0.0905747339129448,
      "learning_rate": 0.0009989517793413507,
      "loss": 0.4734,
      "num_input_tokens_seen": 2945968,
      "step": 4525
    },
    {
      "epoch": 2.3742138364779874,
      "grad_norm": 0.04086661338806152,
      "learning_rate": 0.000998936926953093,
      "loss": 0.5304,
      "num_input_tokens_seen": 2949456,
      "step": 4530
    },
    {
      "epoch": 2.3768343815513626,
      "grad_norm": 0.06502846628427505,
      "learning_rate": 0.0009989219701929224,
      "loss": 0.4956,
      "num_input_tokens_seen": 2952592,
      "step": 4535
    },
    {
      "epoch": 2.379454926624738,
      "grad_norm": 0.05596686527132988,
      "learning_rate": 0.0009989069090639675,
      "loss": 0.3549,
      "num_input_tokens_seen": 2955760,
      "step": 4540
    },
    {
      "epoch": 2.3820754716981134,
      "grad_norm": 0.058458052575588226,
      "learning_rate": 0.0009988917435693793,
      "loss": 0.2994,
      "num_input_tokens_seen": 2958768,
      "step": 4545
    },
    {
      "epoch": 2.3846960167714886,
      "grad_norm": 0.22812820971012115,
      "learning_rate": 0.0009988764737123297,
      "loss": 0.394,
      "num_input_tokens_seen": 2961552,
      "step": 4550
    },
    {
      "epoch": 2.3873165618448637,
      "grad_norm": 0.07777221500873566,
      "learning_rate": 0.0009988610994960134,
      "loss": 0.4398,
      "num_input_tokens_seen": 2964464,
      "step": 4555
    },
    {
      "epoch": 2.389937106918239,
      "grad_norm": 0.07385165989398956,
      "learning_rate": 0.0009988456209236464,
      "loss": 0.525,
      "num_input_tokens_seen": 2966832,
      "step": 4560
    },
    {
      "epoch": 2.392557651991614,
      "grad_norm": 0.07492261379957199,
      "learning_rate": 0.0009988300379984668,
      "loss": 0.47,
      "num_input_tokens_seen": 2969968,
      "step": 4565
    },
    {
      "epoch": 2.3951781970649897,
      "grad_norm": 0.07893513143062592,
      "learning_rate": 0.0009988143507237341,
      "loss": 0.4392,
      "num_input_tokens_seen": 2973040,
      "step": 4570
    },
    {
      "epoch": 2.397798742138365,
      "grad_norm": 0.05448182299733162,
      "learning_rate": 0.00099879855910273,
      "loss": 0.5265,
      "num_input_tokens_seen": 2977520,
      "step": 4575
    },
    {
      "epoch": 2.40041928721174,
      "grad_norm": 0.08276235312223434,
      "learning_rate": 0.0009987826631387578,
      "loss": 0.4019,
      "num_input_tokens_seen": 2980880,
      "step": 4580
    },
    {
      "epoch": 2.403039832285115,
      "grad_norm": 0.06044970452785492,
      "learning_rate": 0.000998766662835143,
      "loss": 0.6019,
      "num_input_tokens_seen": 2983632,
      "step": 4585
    },
    {
      "epoch": 2.4056603773584904,
      "grad_norm": 0.07840324193239212,
      "learning_rate": 0.0009987505581952325,
      "loss": 0.4581,
      "num_input_tokens_seen": 2986256,
      "step": 4590
    },
    {
      "epoch": 2.408280922431866,
      "grad_norm": 0.054844971746206284,
      "learning_rate": 0.0009987343492223954,
      "loss": 0.4396,
      "num_input_tokens_seen": 2991056,
      "step": 4595
    },
    {
      "epoch": 2.410901467505241,
      "grad_norm": 0.05452904477715492,
      "learning_rate": 0.0009987180359200222,
      "loss": 0.3692,
      "num_input_tokens_seen": 2994320,
      "step": 4600
    },
    {
      "epoch": 2.4135220125786163,
      "grad_norm": 0.1397629678249359,
      "learning_rate": 0.0009987016182915257,
      "loss": 0.4061,
      "num_input_tokens_seen": 2997104,
      "step": 4605
    },
    {
      "epoch": 2.4161425576519915,
      "grad_norm": 0.08575969189405441,
      "learning_rate": 0.0009986850963403398,
      "loss": 0.4986,
      "num_input_tokens_seen": 3000112,
      "step": 4610
    },
    {
      "epoch": 2.418763102725367,
      "grad_norm": 0.056926388293504715,
      "learning_rate": 0.0009986684700699214,
      "loss": 0.4438,
      "num_input_tokens_seen": 3003568,
      "step": 4615
    },
    {
      "epoch": 2.4213836477987423,
      "grad_norm": 0.06675942987203598,
      "learning_rate": 0.000998651739483748,
      "loss": 0.4113,
      "num_input_tokens_seen": 3007696,
      "step": 4620
    },
    {
      "epoch": 2.4240041928721174,
      "grad_norm": 0.06661559641361237,
      "learning_rate": 0.0009986349045853196,
      "loss": 0.5354,
      "num_input_tokens_seen": 3011472,
      "step": 4625
    },
    {
      "epoch": 2.4266247379454926,
      "grad_norm": 0.13970547914505005,
      "learning_rate": 0.000998617965378158,
      "loss": 0.4428,
      "num_input_tokens_seen": 3014320,
      "step": 4630
    },
    {
      "epoch": 2.4292452830188678,
      "grad_norm": 0.09605412185192108,
      "learning_rate": 0.0009986009218658064,
      "loss": 0.6018,
      "num_input_tokens_seen": 3018224,
      "step": 4635
    },
    {
      "epoch": 2.431865828092243,
      "grad_norm": 0.09092273563146591,
      "learning_rate": 0.0009985837740518306,
      "loss": 0.5243,
      "num_input_tokens_seen": 3021072,
      "step": 4640
    },
    {
      "epoch": 2.4344863731656186,
      "grad_norm": 0.06834892928600311,
      "learning_rate": 0.0009985665219398173,
      "loss": 0.3662,
      "num_input_tokens_seen": 3023600,
      "step": 4645
    },
    {
      "epoch": 2.4371069182389937,
      "grad_norm": 0.08456269651651382,
      "learning_rate": 0.0009985491655333755,
      "loss": 0.4896,
      "num_input_tokens_seen": 3026768,
      "step": 4650
    },
    {
      "epoch": 2.439727463312369,
      "grad_norm": 0.04974367469549179,
      "learning_rate": 0.000998531704836136,
      "loss": 0.5012,
      "num_input_tokens_seen": 3033424,
      "step": 4655
    },
    {
      "epoch": 2.442348008385744,
      "grad_norm": 0.07849882543087006,
      "learning_rate": 0.0009985141398517513,
      "loss": 0.4783,
      "num_input_tokens_seen": 3035952,
      "step": 4660
    },
    {
      "epoch": 2.4449685534591197,
      "grad_norm": 0.14161206781864166,
      "learning_rate": 0.000998496470583896,
      "loss": 0.4844,
      "num_input_tokens_seen": 3039216,
      "step": 4665
    },
    {
      "epoch": 2.447589098532495,
      "grad_norm": 0.09063251316547394,
      "learning_rate": 0.0009984786970362663,
      "loss": 0.7001,
      "num_input_tokens_seen": 3042224,
      "step": 4670
    },
    {
      "epoch": 2.45020964360587,
      "grad_norm": 0.061660122126340866,
      "learning_rate": 0.00099846081921258,
      "loss": 0.4197,
      "num_input_tokens_seen": 3044752,
      "step": 4675
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 0.08370736986398697,
      "learning_rate": 0.000998442837116577,
      "loss": 0.5594,
      "num_input_tokens_seen": 3048336,
      "step": 4680
    },
    {
      "epoch": 2.4554507337526204,
      "grad_norm": 0.09608124941587448,
      "learning_rate": 0.0009984247507520193,
      "loss": 0.4458,
      "num_input_tokens_seen": 3050864,
      "step": 4685
    },
    {
      "epoch": 2.458071278825996,
      "grad_norm": 0.12709833681583405,
      "learning_rate": 0.0009984065601226896,
      "loss": 0.4659,
      "num_input_tokens_seen": 3053712,
      "step": 4690
    },
    {
      "epoch": 2.460691823899371,
      "grad_norm": 0.083387590944767,
      "learning_rate": 0.0009983882652323942,
      "loss": 0.6641,
      "num_input_tokens_seen": 3056944,
      "step": 4695
    },
    {
      "epoch": 2.4633123689727463,
      "grad_norm": 0.0630311369895935,
      "learning_rate": 0.0009983698660849592,
      "loss": 0.5122,
      "num_input_tokens_seen": 3060176,
      "step": 4700
    },
    {
      "epoch": 2.4659329140461215,
      "grad_norm": 0.10241593420505524,
      "learning_rate": 0.0009983513626842342,
      "loss": 0.7305,
      "num_input_tokens_seen": 3062352,
      "step": 4705
    },
    {
      "epoch": 2.468553459119497,
      "grad_norm": 0.09501153975725174,
      "learning_rate": 0.0009983327550340893,
      "loss": 0.4793,
      "num_input_tokens_seen": 3064848,
      "step": 4710
    },
    {
      "epoch": 2.4711740041928723,
      "grad_norm": 0.07590920478105545,
      "learning_rate": 0.0009983140431384177,
      "loss": 0.5454,
      "num_input_tokens_seen": 3068368,
      "step": 4715
    },
    {
      "epoch": 2.4737945492662474,
      "grad_norm": 0.14201542735099792,
      "learning_rate": 0.0009982952270011331,
      "loss": 0.6006,
      "num_input_tokens_seen": 3071536,
      "step": 4720
    },
    {
      "epoch": 2.4764150943396226,
      "grad_norm": 0.09051513671875,
      "learning_rate": 0.000998276306626172,
      "loss": 0.3713,
      "num_input_tokens_seen": 3074224,
      "step": 4725
    },
    {
      "epoch": 2.4790356394129978,
      "grad_norm": 0.10515929758548737,
      "learning_rate": 0.000998257282017492,
      "loss": 0.5426,
      "num_input_tokens_seen": 3077744,
      "step": 4730
    },
    {
      "epoch": 2.481656184486373,
      "grad_norm": 0.11151736974716187,
      "learning_rate": 0.0009982381531790732,
      "loss": 0.5018,
      "num_input_tokens_seen": 3080656,
      "step": 4735
    },
    {
      "epoch": 2.4842767295597485,
      "grad_norm": 0.050956785678863525,
      "learning_rate": 0.0009982189201149167,
      "loss": 0.4918,
      "num_input_tokens_seen": 3084624,
      "step": 4740
    },
    {
      "epoch": 2.4868972746331237,
      "grad_norm": 0.14949573576450348,
      "learning_rate": 0.0009981995828290465,
      "loss": 0.4389,
      "num_input_tokens_seen": 3087760,
      "step": 4745
    },
    {
      "epoch": 2.489517819706499,
      "grad_norm": 0.05605235695838928,
      "learning_rate": 0.0009981801413255068,
      "loss": 0.51,
      "num_input_tokens_seen": 3091440,
      "step": 4750
    },
    {
      "epoch": 2.492138364779874,
      "grad_norm": 0.112550750374794,
      "learning_rate": 0.0009981605956083657,
      "loss": 0.4885,
      "num_input_tokens_seen": 3094192,
      "step": 4755
    },
    {
      "epoch": 2.4947589098532497,
      "grad_norm": 0.08714986592531204,
      "learning_rate": 0.000998140945681711,
      "loss": 0.4191,
      "num_input_tokens_seen": 3097584,
      "step": 4760
    },
    {
      "epoch": 2.497379454926625,
      "grad_norm": 0.07728204131126404,
      "learning_rate": 0.0009981211915496536,
      "loss": 0.4948,
      "num_input_tokens_seen": 3101232,
      "step": 4765
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.08332377672195435,
      "learning_rate": 0.0009981013332163256,
      "loss": 0.3329,
      "num_input_tokens_seen": 3103888,
      "step": 4770
    },
    {
      "epoch": 2.502620545073375,
      "grad_norm": 0.08025306463241577,
      "learning_rate": 0.0009980813706858816,
      "loss": 0.3867,
      "num_input_tokens_seen": 3111408,
      "step": 4775
    },
    {
      "epoch": 2.5052410901467503,
      "grad_norm": 0.08211207389831543,
      "learning_rate": 0.000998061303962497,
      "loss": 0.406,
      "num_input_tokens_seen": 3114288,
      "step": 4780
    },
    {
      "epoch": 2.507861635220126,
      "grad_norm": 0.09536170959472656,
      "learning_rate": 0.00099804113305037,
      "loss": 0.5473,
      "num_input_tokens_seen": 3117296,
      "step": 4785
    },
    {
      "epoch": 2.510482180293501,
      "grad_norm": 0.09541606903076172,
      "learning_rate": 0.0009980208579537199,
      "loss": 0.3556,
      "num_input_tokens_seen": 3119792,
      "step": 4790
    },
    {
      "epoch": 2.5131027253668763,
      "grad_norm": 0.08248479664325714,
      "learning_rate": 0.000998000478676788,
      "loss": 0.5445,
      "num_input_tokens_seen": 3122544,
      "step": 4795
    },
    {
      "epoch": 2.5157232704402515,
      "grad_norm": 0.0867508128285408,
      "learning_rate": 0.0009979799952238373,
      "loss": 0.4258,
      "num_input_tokens_seen": 3125648,
      "step": 4800
    },
    {
      "epoch": 2.518343815513627,
      "grad_norm": 0.1071566566824913,
      "learning_rate": 0.000997959407599153,
      "loss": 0.4371,
      "num_input_tokens_seen": 3128400,
      "step": 4805
    },
    {
      "epoch": 2.5209643605870022,
      "grad_norm": 0.11864761263132095,
      "learning_rate": 0.000997938715807042,
      "loss": 0.4276,
      "num_input_tokens_seen": 3131280,
      "step": 4810
    },
    {
      "epoch": 2.5235849056603774,
      "grad_norm": 0.05896146222949028,
      "learning_rate": 0.000997917919851832,
      "loss": 0.506,
      "num_input_tokens_seen": 3134288,
      "step": 4815
    },
    {
      "epoch": 2.5262054507337526,
      "grad_norm": 0.11707305908203125,
      "learning_rate": 0.0009978970197378736,
      "loss": 0.5153,
      "num_input_tokens_seen": 3136784,
      "step": 4820
    },
    {
      "epoch": 2.5288259958071277,
      "grad_norm": 0.0993087962269783,
      "learning_rate": 0.0009978760154695392,
      "loss": 0.5077,
      "num_input_tokens_seen": 3139664,
      "step": 4825
    },
    {
      "epoch": 2.531446540880503,
      "grad_norm": 0.09366317838430405,
      "learning_rate": 0.0009978549070512226,
      "loss": 0.7152,
      "num_input_tokens_seen": 3142896,
      "step": 4830
    },
    {
      "epoch": 2.5340670859538785,
      "grad_norm": 0.0539751872420311,
      "learning_rate": 0.000997833694487339,
      "loss": 0.5313,
      "num_input_tokens_seen": 3146224,
      "step": 4835
    },
    {
      "epoch": 2.5366876310272537,
      "grad_norm": 0.07091009616851807,
      "learning_rate": 0.0009978123777823263,
      "loss": 0.5414,
      "num_input_tokens_seen": 3149136,
      "step": 4840
    },
    {
      "epoch": 2.539308176100629,
      "grad_norm": 0.10348072648048401,
      "learning_rate": 0.0009977909569406434,
      "loss": 0.5157,
      "num_input_tokens_seen": 3151984,
      "step": 4845
    },
    {
      "epoch": 2.541928721174004,
      "grad_norm": 0.0815887600183487,
      "learning_rate": 0.0009977694319667713,
      "loss": 0.5451,
      "num_input_tokens_seen": 3155440,
      "step": 4850
    },
    {
      "epoch": 2.5445492662473796,
      "grad_norm": 0.0357595793902874,
      "learning_rate": 0.0009977478028652131,
      "loss": 0.3719,
      "num_input_tokens_seen": 3159216,
      "step": 4855
    },
    {
      "epoch": 2.547169811320755,
      "grad_norm": 0.07152806967496872,
      "learning_rate": 0.000997726069640493,
      "loss": 0.594,
      "num_input_tokens_seen": 3162640,
      "step": 4860
    },
    {
      "epoch": 2.54979035639413,
      "grad_norm": 0.07776582986116409,
      "learning_rate": 0.0009977042322971577,
      "loss": 0.4393,
      "num_input_tokens_seen": 3165968,
      "step": 4865
    },
    {
      "epoch": 2.552410901467505,
      "grad_norm": 0.06644239276647568,
      "learning_rate": 0.000997682290839775,
      "loss": 0.4271,
      "num_input_tokens_seen": 3168688,
      "step": 4870
    },
    {
      "epoch": 2.5550314465408803,
      "grad_norm": 0.07916463166475296,
      "learning_rate": 0.0009976602452729348,
      "loss": 0.5711,
      "num_input_tokens_seen": 3172400,
      "step": 4875
    },
    {
      "epoch": 2.5576519916142555,
      "grad_norm": 0.10013331472873688,
      "learning_rate": 0.000997638095601249,
      "loss": 0.5878,
      "num_input_tokens_seen": 3175472,
      "step": 4880
    },
    {
      "epoch": 2.560272536687631,
      "grad_norm": 0.0973125770688057,
      "learning_rate": 0.000997615841829351,
      "loss": 0.7429,
      "num_input_tokens_seen": 3178032,
      "step": 4885
    },
    {
      "epoch": 2.5628930817610063,
      "grad_norm": 0.05062355473637581,
      "learning_rate": 0.000997593483961896,
      "loss": 0.3668,
      "num_input_tokens_seen": 3181744,
      "step": 4890
    },
    {
      "epoch": 2.5655136268343814,
      "grad_norm": 0.08634155988693237,
      "learning_rate": 0.0009975710220035607,
      "loss": 0.4541,
      "num_input_tokens_seen": 3184656,
      "step": 4895
    },
    {
      "epoch": 2.568134171907757,
      "grad_norm": 0.05933083966374397,
      "learning_rate": 0.0009975484559590444,
      "loss": 0.4705,
      "num_input_tokens_seen": 3187504,
      "step": 4900
    },
    {
      "epoch": 2.5707547169811322,
      "grad_norm": 0.08657482266426086,
      "learning_rate": 0.0009975257858330674,
      "loss": 0.5153,
      "num_input_tokens_seen": 3190608,
      "step": 4905
    },
    {
      "epoch": 2.5733752620545074,
      "grad_norm": 0.07736362516880035,
      "learning_rate": 0.0009975030116303723,
      "loss": 0.5102,
      "num_input_tokens_seen": 3193584,
      "step": 4910
    },
    {
      "epoch": 2.5759958071278826,
      "grad_norm": 0.0824592188000679,
      "learning_rate": 0.0009974801333557228,
      "loss": 0.6332,
      "num_input_tokens_seen": 3197488,
      "step": 4915
    },
    {
      "epoch": 2.5786163522012577,
      "grad_norm": 0.07375087589025497,
      "learning_rate": 0.000997457151013905,
      "loss": 0.5743,
      "num_input_tokens_seen": 3200592,
      "step": 4920
    },
    {
      "epoch": 2.581236897274633,
      "grad_norm": 0.09153006970882416,
      "learning_rate": 0.0009974340646097264,
      "loss": 0.4663,
      "num_input_tokens_seen": 3204208,
      "step": 4925
    },
    {
      "epoch": 2.5838574423480085,
      "grad_norm": 0.08202838152647018,
      "learning_rate": 0.0009974108741480166,
      "loss": 0.5792,
      "num_input_tokens_seen": 3208048,
      "step": 4930
    },
    {
      "epoch": 2.5864779874213837,
      "grad_norm": 0.04558040201663971,
      "learning_rate": 0.0009973875796336267,
      "loss": 0.4394,
      "num_input_tokens_seen": 3210576,
      "step": 4935
    },
    {
      "epoch": 2.589098532494759,
      "grad_norm": 0.0593661330640316,
      "learning_rate": 0.0009973641810714295,
      "loss": 0.3826,
      "num_input_tokens_seen": 3213648,
      "step": 4940
    },
    {
      "epoch": 2.591719077568134,
      "grad_norm": 0.05553669482469559,
      "learning_rate": 0.00099734067846632,
      "loss": 0.6814,
      "num_input_tokens_seen": 3216592,
      "step": 4945
    },
    {
      "epoch": 2.5943396226415096,
      "grad_norm": 0.07756442576646805,
      "learning_rate": 0.0009973170718232144,
      "loss": 0.4692,
      "num_input_tokens_seen": 3219760,
      "step": 4950
    },
    {
      "epoch": 2.596960167714885,
      "grad_norm": 0.07988253235816956,
      "learning_rate": 0.000997293361147051,
      "loss": 0.694,
      "num_input_tokens_seen": 3223600,
      "step": 4955
    },
    {
      "epoch": 2.59958071278826,
      "grad_norm": 0.09949373453855515,
      "learning_rate": 0.0009972695464427904,
      "loss": 0.5992,
      "num_input_tokens_seen": 3226416,
      "step": 4960
    },
    {
      "epoch": 2.602201257861635,
      "grad_norm": 0.05011788010597229,
      "learning_rate": 0.0009972456277154134,
      "loss": 0.3971,
      "num_input_tokens_seen": 3230288,
      "step": 4965
    },
    {
      "epoch": 2.6048218029350103,
      "grad_norm": 0.13938826322555542,
      "learning_rate": 0.000997221604969924,
      "loss": 0.5174,
      "num_input_tokens_seen": 3233104,
      "step": 4970
    },
    {
      "epoch": 2.6074423480083855,
      "grad_norm": 0.04082086682319641,
      "learning_rate": 0.0009971974782113475,
      "loss": 0.5211,
      "num_input_tokens_seen": 3237872,
      "step": 4975
    },
    {
      "epoch": 2.610062893081761,
      "grad_norm": 0.05375155061483383,
      "learning_rate": 0.0009971732474447308,
      "loss": 0.4539,
      "num_input_tokens_seen": 3242064,
      "step": 4980
    },
    {
      "epoch": 2.6126834381551363,
      "grad_norm": 0.09433335810899734,
      "learning_rate": 0.0009971489126751427,
      "loss": 0.4767,
      "num_input_tokens_seen": 3245712,
      "step": 4985
    },
    {
      "epoch": 2.6153039832285114,
      "grad_norm": 0.07638879120349884,
      "learning_rate": 0.0009971244739076742,
      "loss": 0.4936,
      "num_input_tokens_seen": 3248560,
      "step": 4990
    },
    {
      "epoch": 2.617924528301887,
      "grad_norm": 0.05977523326873779,
      "learning_rate": 0.000997099931147437,
      "loss": 0.4937,
      "num_input_tokens_seen": 3251440,
      "step": 4995
    },
    {
      "epoch": 2.620545073375262,
      "grad_norm": 0.07868889719247818,
      "learning_rate": 0.0009970752843995654,
      "loss": 0.4671,
      "num_input_tokens_seen": 3254640,
      "step": 5000
    },
    {
      "epoch": 2.6231656184486374,
      "grad_norm": 0.04928705841302872,
      "learning_rate": 0.0009970505336692153,
      "loss": 0.4113,
      "num_input_tokens_seen": 3258320,
      "step": 5005
    },
    {
      "epoch": 2.6257861635220126,
      "grad_norm": 0.05891483649611473,
      "learning_rate": 0.0009970256789615642,
      "loss": 0.4355,
      "num_input_tokens_seen": 3261168,
      "step": 5010
    },
    {
      "epoch": 2.6284067085953877,
      "grad_norm": 0.08256746828556061,
      "learning_rate": 0.0009970007202818115,
      "loss": 0.5616,
      "num_input_tokens_seen": 3264816,
      "step": 5015
    },
    {
      "epoch": 2.631027253668763,
      "grad_norm": 0.06643926352262497,
      "learning_rate": 0.000996975657635178,
      "loss": 0.3743,
      "num_input_tokens_seen": 3268080,
      "step": 5020
    },
    {
      "epoch": 2.6336477987421385,
      "grad_norm": 0.03664498031139374,
      "learning_rate": 0.000996950491026907,
      "loss": 0.5196,
      "num_input_tokens_seen": 3272528,
      "step": 5025
    },
    {
      "epoch": 2.6362683438155137,
      "grad_norm": 0.05567265301942825,
      "learning_rate": 0.0009969252204622624,
      "loss": 0.5071,
      "num_input_tokens_seen": 3275888,
      "step": 5030
    },
    {
      "epoch": 2.638888888888889,
      "grad_norm": 0.13159896433353424,
      "learning_rate": 0.0009968998459465312,
      "loss": 0.4195,
      "num_input_tokens_seen": 3281072,
      "step": 5035
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 0.06149822473526001,
      "learning_rate": 0.0009968743674850212,
      "loss": 0.4925,
      "num_input_tokens_seen": 3285808,
      "step": 5040
    },
    {
      "epoch": 2.6441299790356396,
      "grad_norm": 0.04954038932919502,
      "learning_rate": 0.0009968487850830622,
      "loss": 0.4312,
      "num_input_tokens_seen": 3289168,
      "step": 5045
    },
    {
      "epoch": 2.646750524109015,
      "grad_norm": 0.08261235803365707,
      "learning_rate": 0.0009968230987460055,
      "loss": 0.6515,
      "num_input_tokens_seen": 3292336,
      "step": 5050
    },
    {
      "epoch": 2.64937106918239,
      "grad_norm": 0.05420803278684616,
      "learning_rate": 0.0009967973084792246,
      "loss": 0.4366,
      "num_input_tokens_seen": 3295024,
      "step": 5055
    },
    {
      "epoch": 2.651991614255765,
      "grad_norm": 0.1681787222623825,
      "learning_rate": 0.0009967714142881145,
      "loss": 0.6597,
      "num_input_tokens_seen": 3297776,
      "step": 5060
    },
    {
      "epoch": 2.6546121593291403,
      "grad_norm": 0.061270587146282196,
      "learning_rate": 0.0009967454161780923,
      "loss": 0.3878,
      "num_input_tokens_seen": 3300560,
      "step": 5065
    },
    {
      "epoch": 2.6572327044025155,
      "grad_norm": 0.068723663687706,
      "learning_rate": 0.000996719314154596,
      "loss": 0.4905,
      "num_input_tokens_seen": 3304432,
      "step": 5070
    },
    {
      "epoch": 2.659853249475891,
      "grad_norm": 0.06424834579229355,
      "learning_rate": 0.0009966931082230862,
      "loss": 0.4738,
      "num_input_tokens_seen": 3307344,
      "step": 5075
    },
    {
      "epoch": 2.6624737945492662,
      "grad_norm": 0.05913976579904556,
      "learning_rate": 0.0009966667983890445,
      "loss": 0.5092,
      "num_input_tokens_seen": 3311056,
      "step": 5080
    },
    {
      "epoch": 2.6650943396226414,
      "grad_norm": 0.06100068241357803,
      "learning_rate": 0.000996640384657975,
      "loss": 0.4777,
      "num_input_tokens_seen": 3313712,
      "step": 5085
    },
    {
      "epoch": 2.667714884696017,
      "grad_norm": 0.04987926036119461,
      "learning_rate": 0.0009966138670354028,
      "loss": 0.4607,
      "num_input_tokens_seen": 3317232,
      "step": 5090
    },
    {
      "epoch": 2.670335429769392,
      "grad_norm": 0.06194008141756058,
      "learning_rate": 0.0009965872455268755,
      "loss": 0.617,
      "num_input_tokens_seen": 3320048,
      "step": 5095
    },
    {
      "epoch": 2.6729559748427674,
      "grad_norm": 0.05214774236083031,
      "learning_rate": 0.0009965605201379616,
      "loss": 0.4534,
      "num_input_tokens_seen": 3324720,
      "step": 5100
    },
    {
      "epoch": 2.6755765199161425,
      "grad_norm": 0.07320598512887955,
      "learning_rate": 0.000996533690874252,
      "loss": 0.4666,
      "num_input_tokens_seen": 3327696,
      "step": 5105
    },
    {
      "epoch": 2.6781970649895177,
      "grad_norm": 0.05058274045586586,
      "learning_rate": 0.0009965067577413593,
      "loss": 0.3645,
      "num_input_tokens_seen": 3330608,
      "step": 5110
    },
    {
      "epoch": 2.680817610062893,
      "grad_norm": 0.08355947583913803,
      "learning_rate": 0.0009964797207449173,
      "loss": 0.4875,
      "num_input_tokens_seen": 3333296,
      "step": 5115
    },
    {
      "epoch": 2.6834381551362685,
      "grad_norm": 0.10853370279073715,
      "learning_rate": 0.0009964525798905816,
      "loss": 0.4475,
      "num_input_tokens_seen": 3337552,
      "step": 5120
    },
    {
      "epoch": 2.6860587002096437,
      "grad_norm": 0.05127166584134102,
      "learning_rate": 0.0009964253351840303,
      "loss": 0.4689,
      "num_input_tokens_seen": 3341296,
      "step": 5125
    },
    {
      "epoch": 2.688679245283019,
      "grad_norm": 0.07796134799718857,
      "learning_rate": 0.000996397986630962,
      "loss": 0.5171,
      "num_input_tokens_seen": 3343920,
      "step": 5130
    },
    {
      "epoch": 2.691299790356394,
      "grad_norm": 0.08785861730575562,
      "learning_rate": 0.0009963705342370982,
      "loss": 0.4475,
      "num_input_tokens_seen": 3346736,
      "step": 5135
    },
    {
      "epoch": 2.6939203354297696,
      "grad_norm": 0.06244901940226555,
      "learning_rate": 0.000996342978008182,
      "loss": 0.4781,
      "num_input_tokens_seen": 3350320,
      "step": 5140
    },
    {
      "epoch": 2.6965408805031448,
      "grad_norm": 0.0831744372844696,
      "learning_rate": 0.000996315317949977,
      "loss": 0.4297,
      "num_input_tokens_seen": 3353552,
      "step": 5145
    },
    {
      "epoch": 2.69916142557652,
      "grad_norm": 0.14184214174747467,
      "learning_rate": 0.0009962875540682696,
      "loss": 0.518,
      "num_input_tokens_seen": 3357136,
      "step": 5150
    },
    {
      "epoch": 2.701781970649895,
      "grad_norm": 0.15313757956027985,
      "learning_rate": 0.0009962596863688682,
      "loss": 0.4417,
      "num_input_tokens_seen": 3360048,
      "step": 5155
    },
    {
      "epoch": 2.7044025157232703,
      "grad_norm": 0.0836501270532608,
      "learning_rate": 0.000996231714857602,
      "loss": 0.4198,
      "num_input_tokens_seen": 3363280,
      "step": 5160
    },
    {
      "epoch": 2.7070230607966455,
      "grad_norm": 0.12622539699077606,
      "learning_rate": 0.000996203639540322,
      "loss": 0.6311,
      "num_input_tokens_seen": 3366000,
      "step": 5165
    },
    {
      "epoch": 2.709643605870021,
      "grad_norm": 0.0631767213344574,
      "learning_rate": 0.0009961754604229018,
      "loss": 0.5092,
      "num_input_tokens_seen": 3369648,
      "step": 5170
    },
    {
      "epoch": 2.7122641509433962,
      "grad_norm": 0.1022714301943779,
      "learning_rate": 0.0009961471775112361,
      "loss": 0.4661,
      "num_input_tokens_seen": 3372528,
      "step": 5175
    },
    {
      "epoch": 2.7148846960167714,
      "grad_norm": 0.09015294164419174,
      "learning_rate": 0.000996118790811241,
      "loss": 0.5069,
      "num_input_tokens_seen": 3375728,
      "step": 5180
    },
    {
      "epoch": 2.717505241090147,
      "grad_norm": 0.15572670102119446,
      "learning_rate": 0.0009960903003288551,
      "loss": 0.435,
      "num_input_tokens_seen": 3378864,
      "step": 5185
    },
    {
      "epoch": 2.720125786163522,
      "grad_norm": 0.08400553464889526,
      "learning_rate": 0.0009960617060700378,
      "loss": 0.4216,
      "num_input_tokens_seen": 3381936,
      "step": 5190
    },
    {
      "epoch": 2.7227463312368974,
      "grad_norm": 0.0620720237493515,
      "learning_rate": 0.000996033008040771,
      "loss": 0.527,
      "num_input_tokens_seen": 3384464,
      "step": 5195
    },
    {
      "epoch": 2.7253668763102725,
      "grad_norm": 0.04497440904378891,
      "learning_rate": 0.0009960042062470583,
      "loss": 0.5652,
      "num_input_tokens_seen": 3387312,
      "step": 5200
    },
    {
      "epoch": 2.7279874213836477,
      "grad_norm": 0.06622759252786636,
      "learning_rate": 0.0009959753006949241,
      "loss": 0.4463,
      "num_input_tokens_seen": 3391376,
      "step": 5205
    },
    {
      "epoch": 2.730607966457023,
      "grad_norm": 0.0660894438624382,
      "learning_rate": 0.0009959462913904154,
      "loss": 0.4993,
      "num_input_tokens_seen": 3394352,
      "step": 5210
    },
    {
      "epoch": 2.7332285115303985,
      "grad_norm": 0.06055410951375961,
      "learning_rate": 0.0009959171783396007,
      "loss": 0.54,
      "num_input_tokens_seen": 3397456,
      "step": 5215
    },
    {
      "epoch": 2.7358490566037736,
      "grad_norm": 0.06393521279096603,
      "learning_rate": 0.00099588796154857,
      "loss": 0.5499,
      "num_input_tokens_seen": 3403792,
      "step": 5220
    },
    {
      "epoch": 2.738469601677149,
      "grad_norm": 0.1321006417274475,
      "learning_rate": 0.000995858641023435,
      "loss": 0.5079,
      "num_input_tokens_seen": 3407184,
      "step": 5225
    },
    {
      "epoch": 2.741090146750524,
      "grad_norm": 0.06740150600671768,
      "learning_rate": 0.0009958292167703293,
      "loss": 0.5329,
      "num_input_tokens_seen": 3410256,
      "step": 5230
    },
    {
      "epoch": 2.7437106918238996,
      "grad_norm": 0.0884270891547203,
      "learning_rate": 0.0009957996887954082,
      "loss": 0.5545,
      "num_input_tokens_seen": 3413168,
      "step": 5235
    },
    {
      "epoch": 2.7463312368972748,
      "grad_norm": 0.08786444365978241,
      "learning_rate": 0.0009957700571048486,
      "loss": 0.6493,
      "num_input_tokens_seen": 3415728,
      "step": 5240
    },
    {
      "epoch": 2.74895178197065,
      "grad_norm": 0.1409420520067215,
      "learning_rate": 0.0009957403217048493,
      "loss": 0.4034,
      "num_input_tokens_seen": 3419568,
      "step": 5245
    },
    {
      "epoch": 2.751572327044025,
      "grad_norm": 0.06124860793352127,
      "learning_rate": 0.0009957104826016302,
      "loss": 0.3652,
      "num_input_tokens_seen": 3422736,
      "step": 5250
    },
    {
      "epoch": 2.7541928721174003,
      "grad_norm": 0.06139567866921425,
      "learning_rate": 0.0009956805398014337,
      "loss": 0.4344,
      "num_input_tokens_seen": 3427472,
      "step": 5255
    },
    {
      "epoch": 2.7568134171907754,
      "grad_norm": 0.05654565244913101,
      "learning_rate": 0.0009956504933105231,
      "loss": 0.4141,
      "num_input_tokens_seen": 3430256,
      "step": 5260
    },
    {
      "epoch": 2.759433962264151,
      "grad_norm": 0.07855681329965591,
      "learning_rate": 0.000995620343135184,
      "loss": 0.5857,
      "num_input_tokens_seen": 3433328,
      "step": 5265
    },
    {
      "epoch": 2.762054507337526,
      "grad_norm": 0.09083158522844315,
      "learning_rate": 0.0009955900892817235,
      "loss": 0.4384,
      "num_input_tokens_seen": 3437456,
      "step": 5270
    },
    {
      "epoch": 2.7646750524109014,
      "grad_norm": 0.07887584716081619,
      "learning_rate": 0.0009955597317564703,
      "loss": 0.3807,
      "num_input_tokens_seen": 3440240,
      "step": 5275
    },
    {
      "epoch": 2.767295597484277,
      "grad_norm": 0.08431781828403473,
      "learning_rate": 0.0009955292705657749,
      "loss": 0.4722,
      "num_input_tokens_seen": 3443824,
      "step": 5280
    },
    {
      "epoch": 2.769916142557652,
      "grad_norm": 0.04013533145189285,
      "learning_rate": 0.0009954987057160093,
      "loss": 0.5572,
      "num_input_tokens_seen": 3447184,
      "step": 5285
    },
    {
      "epoch": 2.7725366876310273,
      "grad_norm": 0.0987238883972168,
      "learning_rate": 0.0009954680372135675,
      "loss": 0.4287,
      "num_input_tokens_seen": 3450320,
      "step": 5290
    },
    {
      "epoch": 2.7751572327044025,
      "grad_norm": 0.04226570576429367,
      "learning_rate": 0.000995437265064865,
      "loss": 0.5816,
      "num_input_tokens_seen": 3454640,
      "step": 5295
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 0.06675716489553452,
      "learning_rate": 0.0009954063892763387,
      "loss": 0.4634,
      "num_input_tokens_seen": 3458000,
      "step": 5300
    },
    {
      "epoch": 2.780398322851153,
      "grad_norm": 0.08876442164182663,
      "learning_rate": 0.0009953754098544479,
      "loss": 0.6013,
      "num_input_tokens_seen": 3461296,
      "step": 5305
    },
    {
      "epoch": 2.7830188679245285,
      "grad_norm": 0.0868031457066536,
      "learning_rate": 0.0009953443268056726,
      "loss": 0.4116,
      "num_input_tokens_seen": 3464336,
      "step": 5310
    },
    {
      "epoch": 2.7856394129979036,
      "grad_norm": 0.10040491819381714,
      "learning_rate": 0.0009953131401365155,
      "loss": 0.5837,
      "num_input_tokens_seen": 3467184,
      "step": 5315
    },
    {
      "epoch": 2.788259958071279,
      "grad_norm": 0.10573697090148926,
      "learning_rate": 0.0009952818498535003,
      "loss": 0.5457,
      "num_input_tokens_seen": 3470320,
      "step": 5320
    },
    {
      "epoch": 2.790880503144654,
      "grad_norm": 0.09950630366802216,
      "learning_rate": 0.0009952504559631726,
      "loss": 0.4116,
      "num_input_tokens_seen": 3474128,
      "step": 5325
    },
    {
      "epoch": 2.7935010482180296,
      "grad_norm": 0.08717279881238937,
      "learning_rate": 0.0009952189584720996,
      "loss": 0.6511,
      "num_input_tokens_seen": 3476336,
      "step": 5330
    },
    {
      "epoch": 2.7961215932914047,
      "grad_norm": 0.04779836907982826,
      "learning_rate": 0.0009951873573868701,
      "loss": 0.537,
      "num_input_tokens_seen": 3480016,
      "step": 5335
    },
    {
      "epoch": 2.79874213836478,
      "grad_norm": 0.18750202655792236,
      "learning_rate": 0.000995155652714095,
      "loss": 0.4226,
      "num_input_tokens_seen": 3482608,
      "step": 5340
    },
    {
      "epoch": 2.801362683438155,
      "grad_norm": 0.06746537238359451,
      "learning_rate": 0.0009951238444604064,
      "loss": 0.7051,
      "num_input_tokens_seen": 3485200,
      "step": 5345
    },
    {
      "epoch": 2.8039832285115303,
      "grad_norm": 0.0734463483095169,
      "learning_rate": 0.000995091932632458,
      "loss": 0.3597,
      "num_input_tokens_seen": 3488336,
      "step": 5350
    },
    {
      "epoch": 2.8066037735849054,
      "grad_norm": 0.06296215206384659,
      "learning_rate": 0.000995059917236926,
      "loss": 0.5557,
      "num_input_tokens_seen": 3492080,
      "step": 5355
    },
    {
      "epoch": 2.809224318658281,
      "grad_norm": 0.05205395445227623,
      "learning_rate": 0.000995027798280507,
      "loss": 0.4424,
      "num_input_tokens_seen": 3495952,
      "step": 5360
    },
    {
      "epoch": 2.811844863731656,
      "grad_norm": 0.0682181790471077,
      "learning_rate": 0.00099499557576992,
      "loss": 0.435,
      "num_input_tokens_seen": 3498224,
      "step": 5365
    },
    {
      "epoch": 2.8144654088050314,
      "grad_norm": 0.14458438754081726,
      "learning_rate": 0.000994963249711906,
      "loss": 0.5648,
      "num_input_tokens_seen": 3501872,
      "step": 5370
    },
    {
      "epoch": 2.8170859538784065,
      "grad_norm": 0.059713803231716156,
      "learning_rate": 0.000994930820113227,
      "loss": 0.4112,
      "num_input_tokens_seen": 3505008,
      "step": 5375
    },
    {
      "epoch": 2.819706498951782,
      "grad_norm": 0.06361255049705505,
      "learning_rate": 0.0009948982869806668,
      "loss": 0.3886,
      "num_input_tokens_seen": 3508016,
      "step": 5380
    },
    {
      "epoch": 2.8223270440251573,
      "grad_norm": 0.04160495474934578,
      "learning_rate": 0.0009948656503210311,
      "loss": 0.5118,
      "num_input_tokens_seen": 3512912,
      "step": 5385
    },
    {
      "epoch": 2.8249475890985325,
      "grad_norm": 0.057367824018001556,
      "learning_rate": 0.000994832910141147,
      "loss": 0.4644,
      "num_input_tokens_seen": 3516112,
      "step": 5390
    },
    {
      "epoch": 2.8275681341719077,
      "grad_norm": 0.05919116735458374,
      "learning_rate": 0.0009948000664478638,
      "loss": 0.5736,
      "num_input_tokens_seen": 3518800,
      "step": 5395
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 0.09388887137174606,
      "learning_rate": 0.0009947671192480515,
      "loss": 0.4028,
      "num_input_tokens_seen": 3522608,
      "step": 5400
    },
    {
      "epoch": 2.832809224318658,
      "grad_norm": 0.10817847400903702,
      "learning_rate": 0.0009947340685486023,
      "loss": 0.502,
      "num_input_tokens_seen": 3526512,
      "step": 5405
    },
    {
      "epoch": 2.8354297693920336,
      "grad_norm": 0.07789897918701172,
      "learning_rate": 0.0009947009143564303,
      "loss": 0.4383,
      "num_input_tokens_seen": 3529456,
      "step": 5410
    },
    {
      "epoch": 2.838050314465409,
      "grad_norm": 0.09729862213134766,
      "learning_rate": 0.0009946676566784708,
      "loss": 0.4385,
      "num_input_tokens_seen": 3532464,
      "step": 5415
    },
    {
      "epoch": 2.840670859538784,
      "grad_norm": 0.16652508080005646,
      "learning_rate": 0.000994634295521681,
      "loss": 0.4782,
      "num_input_tokens_seen": 3535824,
      "step": 5420
    },
    {
      "epoch": 2.8432914046121596,
      "grad_norm": 0.06612448394298553,
      "learning_rate": 0.0009946008308930397,
      "loss": 0.4993,
      "num_input_tokens_seen": 3539696,
      "step": 5425
    },
    {
      "epoch": 2.8459119496855347,
      "grad_norm": 0.07802990823984146,
      "learning_rate": 0.0009945672627995473,
      "loss": 0.4999,
      "num_input_tokens_seen": 3543408,
      "step": 5430
    },
    {
      "epoch": 2.84853249475891,
      "grad_norm": 0.16936172544956207,
      "learning_rate": 0.0009945335912482256,
      "loss": 0.3971,
      "num_input_tokens_seen": 3546864,
      "step": 5435
    },
    {
      "epoch": 2.851153039832285,
      "grad_norm": 0.05884429067373276,
      "learning_rate": 0.000994499816246119,
      "loss": 0.4553,
      "num_input_tokens_seen": 3549904,
      "step": 5440
    },
    {
      "epoch": 2.8537735849056602,
      "grad_norm": 0.06656364351511002,
      "learning_rate": 0.000994465937800292,
      "loss": 0.4611,
      "num_input_tokens_seen": 3553072,
      "step": 5445
    },
    {
      "epoch": 2.8563941299790354,
      "grad_norm": 0.05376653000712395,
      "learning_rate": 0.0009944319559178321,
      "loss": 0.3475,
      "num_input_tokens_seen": 3556784,
      "step": 5450
    },
    {
      "epoch": 2.859014675052411,
      "grad_norm": 0.03943787142634392,
      "learning_rate": 0.0009943978706058478,
      "loss": 0.5035,
      "num_input_tokens_seen": 3561232,
      "step": 5455
    },
    {
      "epoch": 2.861635220125786,
      "grad_norm": 0.040327418595552444,
      "learning_rate": 0.0009943636818714695,
      "loss": 0.5045,
      "num_input_tokens_seen": 3564048,
      "step": 5460
    },
    {
      "epoch": 2.8642557651991614,
      "grad_norm": 0.07014085352420807,
      "learning_rate": 0.0009943293897218487,
      "loss": 0.3893,
      "num_input_tokens_seen": 3567664,
      "step": 5465
    },
    {
      "epoch": 2.8668763102725365,
      "grad_norm": 0.049138229340314865,
      "learning_rate": 0.0009942949941641594,
      "loss": 0.4734,
      "num_input_tokens_seen": 3571344,
      "step": 5470
    },
    {
      "epoch": 2.869496855345912,
      "grad_norm": 0.1684853881597519,
      "learning_rate": 0.0009942604952055964,
      "loss": 0.4485,
      "num_input_tokens_seen": 3574128,
      "step": 5475
    },
    {
      "epoch": 2.8721174004192873,
      "grad_norm": 0.038810957223176956,
      "learning_rate": 0.0009942258928533768,
      "loss": 0.4769,
      "num_input_tokens_seen": 3577552,
      "step": 5480
    },
    {
      "epoch": 2.8747379454926625,
      "grad_norm": 0.07829365879297256,
      "learning_rate": 0.0009941911871147386,
      "loss": 0.672,
      "num_input_tokens_seen": 3581040,
      "step": 5485
    },
    {
      "epoch": 2.8773584905660377,
      "grad_norm": 0.10653910040855408,
      "learning_rate": 0.000994156377996942,
      "loss": 0.4455,
      "num_input_tokens_seen": 3583760,
      "step": 5490
    },
    {
      "epoch": 2.879979035639413,
      "grad_norm": 0.10256616771221161,
      "learning_rate": 0.0009941214655072692,
      "loss": 0.5032,
      "num_input_tokens_seen": 3587632,
      "step": 5495
    },
    {
      "epoch": 2.882599580712788,
      "grad_norm": 0.109367735683918,
      "learning_rate": 0.0009940864496530226,
      "loss": 0.4964,
      "num_input_tokens_seen": 3590960,
      "step": 5500
    },
    {
      "epoch": 2.8852201257861636,
      "grad_norm": 0.05739130824804306,
      "learning_rate": 0.000994051330441528,
      "loss": 0.3853,
      "num_input_tokens_seen": 3594672,
      "step": 5505
    },
    {
      "epoch": 2.8878406708595388,
      "grad_norm": 0.05850277096033096,
      "learning_rate": 0.0009940161078801312,
      "loss": 0.4308,
      "num_input_tokens_seen": 3597872,
      "step": 5510
    },
    {
      "epoch": 2.890461215932914,
      "grad_norm": 0.06588813662528992,
      "learning_rate": 0.0009939807819762008,
      "loss": 0.3603,
      "num_input_tokens_seen": 3600528,
      "step": 5515
    },
    {
      "epoch": 2.8930817610062896,
      "grad_norm": 0.06680869311094284,
      "learning_rate": 0.0009939453527371262,
      "loss": 0.5159,
      "num_input_tokens_seen": 3603344,
      "step": 5520
    },
    {
      "epoch": 2.8957023060796647,
      "grad_norm": 0.09097415208816528,
      "learning_rate": 0.0009939098201703193,
      "loss": 0.4488,
      "num_input_tokens_seen": 3606448,
      "step": 5525
    },
    {
      "epoch": 2.89832285115304,
      "grad_norm": 0.06710609048604965,
      "learning_rate": 0.0009938741842832129,
      "loss": 0.6639,
      "num_input_tokens_seen": 3609776,
      "step": 5530
    },
    {
      "epoch": 2.900943396226415,
      "grad_norm": 0.10288720577955246,
      "learning_rate": 0.0009938384450832614,
      "loss": 0.5016,
      "num_input_tokens_seen": 3612080,
      "step": 5535
    },
    {
      "epoch": 2.9035639412997902,
      "grad_norm": 0.06960073113441467,
      "learning_rate": 0.0009938026025779411,
      "loss": 0.5243,
      "num_input_tokens_seen": 3615504,
      "step": 5540
    },
    {
      "epoch": 2.9061844863731654,
      "grad_norm": 0.11858285963535309,
      "learning_rate": 0.0009937666567747501,
      "loss": 0.658,
      "num_input_tokens_seen": 3618096,
      "step": 5545
    },
    {
      "epoch": 2.908805031446541,
      "grad_norm": 0.09050600975751877,
      "learning_rate": 0.0009937306076812076,
      "loss": 0.4381,
      "num_input_tokens_seen": 3621296,
      "step": 5550
    },
    {
      "epoch": 2.911425576519916,
      "grad_norm": 0.09228192269802094,
      "learning_rate": 0.0009936944553048548,
      "loss": 0.4796,
      "num_input_tokens_seen": 3624784,
      "step": 5555
    },
    {
      "epoch": 2.9140461215932913,
      "grad_norm": 0.07737912237644196,
      "learning_rate": 0.0009936581996532543,
      "loss": 0.5642,
      "num_input_tokens_seen": 3628336,
      "step": 5560
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 0.08432413637638092,
      "learning_rate": 0.0009936218407339905,
      "loss": 0.5979,
      "num_input_tokens_seen": 3631280,
      "step": 5565
    },
    {
      "epoch": 2.919287211740042,
      "grad_norm": 0.047256480902433395,
      "learning_rate": 0.0009935853785546691,
      "loss": 0.6359,
      "num_input_tokens_seen": 3634448,
      "step": 5570
    },
    {
      "epoch": 2.9219077568134173,
      "grad_norm": 0.0716669037938118,
      "learning_rate": 0.0009935488131229177,
      "loss": 0.471,
      "num_input_tokens_seen": 3637072,
      "step": 5575
    },
    {
      "epoch": 2.9245283018867925,
      "grad_norm": 0.05349694937467575,
      "learning_rate": 0.000993512144446385,
      "loss": 0.6355,
      "num_input_tokens_seen": 3640880,
      "step": 5580
    },
    {
      "epoch": 2.9271488469601676,
      "grad_norm": 0.08887053281068802,
      "learning_rate": 0.000993475372532742,
      "loss": 0.5179,
      "num_input_tokens_seen": 3643312,
      "step": 5585
    },
    {
      "epoch": 2.929769392033543,
      "grad_norm": 0.06322536617517471,
      "learning_rate": 0.0009934384973896812,
      "loss": 0.4989,
      "num_input_tokens_seen": 3647152,
      "step": 5590
    },
    {
      "epoch": 2.932389937106918,
      "grad_norm": 0.10455873608589172,
      "learning_rate": 0.0009934015190249162,
      "loss": 0.5002,
      "num_input_tokens_seen": 3649904,
      "step": 5595
    },
    {
      "epoch": 2.9350104821802936,
      "grad_norm": 0.07274353504180908,
      "learning_rate": 0.0009933644374461822,
      "loss": 0.3973,
      "num_input_tokens_seen": 3653680,
      "step": 5600
    },
    {
      "epoch": 2.9376310272536688,
      "grad_norm": 0.07328981161117554,
      "learning_rate": 0.0009933272526612366,
      "loss": 0.5719,
      "num_input_tokens_seen": 3657328,
      "step": 5605
    },
    {
      "epoch": 2.940251572327044,
      "grad_norm": 0.060820322483778,
      "learning_rate": 0.0009932899646778578,
      "loss": 0.492,
      "num_input_tokens_seen": 3661008,
      "step": 5610
    },
    {
      "epoch": 2.9428721174004195,
      "grad_norm": 0.06608433276414871,
      "learning_rate": 0.0009932525735038464,
      "loss": 0.4808,
      "num_input_tokens_seen": 3664144,
      "step": 5615
    },
    {
      "epoch": 2.9454926624737947,
      "grad_norm": 0.11388228833675385,
      "learning_rate": 0.0009932150791470238,
      "loss": 0.4223,
      "num_input_tokens_seen": 3666800,
      "step": 5620
    },
    {
      "epoch": 2.94811320754717,
      "grad_norm": 0.070721834897995,
      "learning_rate": 0.0009931774816152334,
      "loss": 0.46,
      "num_input_tokens_seen": 3669936,
      "step": 5625
    },
    {
      "epoch": 2.950733752620545,
      "grad_norm": 0.059534475207328796,
      "learning_rate": 0.0009931397809163406,
      "loss": 0.3994,
      "num_input_tokens_seen": 3673584,
      "step": 5630
    },
    {
      "epoch": 2.95335429769392,
      "grad_norm": 0.06207121163606644,
      "learning_rate": 0.0009931019770582316,
      "loss": 0.4572,
      "num_input_tokens_seen": 3676816,
      "step": 5635
    },
    {
      "epoch": 2.9559748427672954,
      "grad_norm": 0.08462874591350555,
      "learning_rate": 0.0009930640700488143,
      "loss": 0.5818,
      "num_input_tokens_seen": 3679792,
      "step": 5640
    },
    {
      "epoch": 2.958595387840671,
      "grad_norm": 0.05591346696019173,
      "learning_rate": 0.000993026059896019,
      "loss": 0.4964,
      "num_input_tokens_seen": 3682704,
      "step": 5645
    },
    {
      "epoch": 2.961215932914046,
      "grad_norm": 0.10899454355239868,
      "learning_rate": 0.0009929879466077968,
      "loss": 0.4592,
      "num_input_tokens_seen": 3685968,
      "step": 5650
    },
    {
      "epoch": 2.9638364779874213,
      "grad_norm": 0.05324126034975052,
      "learning_rate": 0.0009929497301921202,
      "loss": 0.4911,
      "num_input_tokens_seen": 3688528,
      "step": 5655
    },
    {
      "epoch": 2.9664570230607965,
      "grad_norm": 0.0892677828669548,
      "learning_rate": 0.000992911410656984,
      "loss": 0.4989,
      "num_input_tokens_seen": 3692464,
      "step": 5660
    },
    {
      "epoch": 2.969077568134172,
      "grad_norm": 0.0651898980140686,
      "learning_rate": 0.000992872988010404,
      "loss": 0.3933,
      "num_input_tokens_seen": 3695920,
      "step": 5665
    },
    {
      "epoch": 2.9716981132075473,
      "grad_norm": 0.09600252658128738,
      "learning_rate": 0.0009928344622604183,
      "loss": 0.6394,
      "num_input_tokens_seen": 3699600,
      "step": 5670
    },
    {
      "epoch": 2.9743186582809225,
      "grad_norm": 1.718051791191101,
      "learning_rate": 0.0009927958334150853,
      "loss": 0.5498,
      "num_input_tokens_seen": 3703024,
      "step": 5675
    },
    {
      "epoch": 2.9769392033542976,
      "grad_norm": 0.03089650347828865,
      "learning_rate": 0.0009927571014824862,
      "loss": 0.3807,
      "num_input_tokens_seen": 3706544,
      "step": 5680
    },
    {
      "epoch": 2.979559748427673,
      "grad_norm": 0.07187272608280182,
      "learning_rate": 0.000992718266470723,
      "loss": 0.5953,
      "num_input_tokens_seen": 3709584,
      "step": 5685
    },
    {
      "epoch": 2.982180293501048,
      "grad_norm": 0.047910287976264954,
      "learning_rate": 0.00099267932838792,
      "loss": 0.5519,
      "num_input_tokens_seen": 3713616,
      "step": 5690
    },
    {
      "epoch": 2.9848008385744236,
      "grad_norm": 0.07233008742332458,
      "learning_rate": 0.000992640287242222,
      "loss": 0.5175,
      "num_input_tokens_seen": 3716656,
      "step": 5695
    },
    {
      "epoch": 2.9874213836477987,
      "grad_norm": 0.07983928918838501,
      "learning_rate": 0.0009926011430417961,
      "loss": 0.4071,
      "num_input_tokens_seen": 3719504,
      "step": 5700
    },
    {
      "epoch": 2.990041928721174,
      "grad_norm": 0.07813284546136856,
      "learning_rate": 0.0009925618957948312,
      "loss": 0.4531,
      "num_input_tokens_seen": 3722672,
      "step": 5705
    },
    {
      "epoch": 2.9926624737945495,
      "grad_norm": 0.06115923821926117,
      "learning_rate": 0.0009925225455095373,
      "loss": 0.4698,
      "num_input_tokens_seen": 3725104,
      "step": 5710
    },
    {
      "epoch": 2.9952830188679247,
      "grad_norm": 0.05781840905547142,
      "learning_rate": 0.0009924830921941455,
      "loss": 0.6194,
      "num_input_tokens_seen": 3728272,
      "step": 5715
    },
    {
      "epoch": 2.9979035639413,
      "grad_norm": 0.07138650864362717,
      "learning_rate": 0.0009924435358569096,
      "loss": 0.4295,
      "num_input_tokens_seen": 3731344,
      "step": 5720
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.4684428870677948,
      "eval_runtime": 13.5893,
      "eval_samples_per_second": 62.402,
      "eval_steps_per_second": 15.6,
      "num_input_tokens_seen": 3733736,
      "step": 5724
    },
    {
      "epoch": 3.000524109014675,
      "grad_norm": 0.09100688993930817,
      "learning_rate": 0.000992403876506104,
      "loss": 0.4733,
      "num_input_tokens_seen": 3734280,
      "step": 5725
    },
    {
      "epoch": 3.00314465408805,
      "grad_norm": 0.0811460018157959,
      "learning_rate": 0.000992364114150025,
      "loss": 0.4046,
      "num_input_tokens_seen": 3736776,
      "step": 5730
    },
    {
      "epoch": 3.0057651991614254,
      "grad_norm": 0.0645768791437149,
      "learning_rate": 0.0009923242487969908,
      "loss": 0.4667,
      "num_input_tokens_seen": 3739496,
      "step": 5735
    },
    {
      "epoch": 3.008385744234801,
      "grad_norm": 0.09815572947263718,
      "learning_rate": 0.0009922842804553403,
      "loss": 0.3941,
      "num_input_tokens_seen": 3741992,
      "step": 5740
    },
    {
      "epoch": 3.011006289308176,
      "grad_norm": 0.08909042924642563,
      "learning_rate": 0.0009922442091334345,
      "loss": 0.5477,
      "num_input_tokens_seen": 3744488,
      "step": 5745
    },
    {
      "epoch": 3.0136268343815513,
      "grad_norm": 0.06367643922567368,
      "learning_rate": 0.0009922040348396561,
      "loss": 0.5467,
      "num_input_tokens_seen": 3747688,
      "step": 5750
    },
    {
      "epoch": 3.0162473794549265,
      "grad_norm": 0.045467860996723175,
      "learning_rate": 0.000992163757582409,
      "loss": 0.4798,
      "num_input_tokens_seen": 3753064,
      "step": 5755
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 0.06720351427793503,
      "learning_rate": 0.0009921233773701188,
      "loss": 0.5391,
      "num_input_tokens_seen": 3756136,
      "step": 5760
    },
    {
      "epoch": 3.0214884696016773,
      "grad_norm": 0.0734126940369606,
      "learning_rate": 0.0009920828942112322,
      "loss": 0.5437,
      "num_input_tokens_seen": 3758888,
      "step": 5765
    },
    {
      "epoch": 3.0241090146750524,
      "grad_norm": 0.0739794448018074,
      "learning_rate": 0.0009920423081142184,
      "loss": 0.3819,
      "num_input_tokens_seen": 3761832,
      "step": 5770
    },
    {
      "epoch": 3.0267295597484276,
      "grad_norm": 0.06249677389860153,
      "learning_rate": 0.0009920016190875672,
      "loss": 0.4776,
      "num_input_tokens_seen": 3765032,
      "step": 5775
    },
    {
      "epoch": 3.029350104821803,
      "grad_norm": 0.08257205784320831,
      "learning_rate": 0.00099196082713979,
      "loss": 0.5643,
      "num_input_tokens_seen": 3767944,
      "step": 5780
    },
    {
      "epoch": 3.0319706498951784,
      "grad_norm": 0.09976755827665329,
      "learning_rate": 0.0009919199322794207,
      "loss": 0.4569,
      "num_input_tokens_seen": 3771112,
      "step": 5785
    },
    {
      "epoch": 3.0345911949685536,
      "grad_norm": 0.08522340655326843,
      "learning_rate": 0.0009918789345150136,
      "loss": 0.5821,
      "num_input_tokens_seen": 3773960,
      "step": 5790
    },
    {
      "epoch": 3.0372117400419287,
      "grad_norm": 0.06275424361228943,
      "learning_rate": 0.000991837833855145,
      "loss": 0.5608,
      "num_input_tokens_seen": 3777192,
      "step": 5795
    },
    {
      "epoch": 3.039832285115304,
      "grad_norm": 0.08463853597640991,
      "learning_rate": 0.000991796630308413,
      "loss": 0.5248,
      "num_input_tokens_seen": 3780488,
      "step": 5800
    },
    {
      "epoch": 3.042452830188679,
      "grad_norm": 0.03866233676671982,
      "learning_rate": 0.0009917553238834363,
      "loss": 0.4793,
      "num_input_tokens_seen": 3784328,
      "step": 5805
    },
    {
      "epoch": 3.0450733752620547,
      "grad_norm": 0.09611479938030243,
      "learning_rate": 0.0009917139145888562,
      "loss": 0.4975,
      "num_input_tokens_seen": 3786696,
      "step": 5810
    },
    {
      "epoch": 3.04769392033543,
      "grad_norm": 0.05059369280934334,
      "learning_rate": 0.000991672402433335,
      "loss": 0.5632,
      "num_input_tokens_seen": 3789768,
      "step": 5815
    },
    {
      "epoch": 3.050314465408805,
      "grad_norm": 0.07226280122995377,
      "learning_rate": 0.0009916307874255565,
      "loss": 0.5055,
      "num_input_tokens_seen": 3792936,
      "step": 5820
    },
    {
      "epoch": 3.05293501048218,
      "grad_norm": 0.06760697811841965,
      "learning_rate": 0.000991589069574226,
      "loss": 0.5436,
      "num_input_tokens_seen": 3796168,
      "step": 5825
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 0.10707798600196838,
      "learning_rate": 0.0009915472488880705,
      "loss": 0.4811,
      "num_input_tokens_seen": 3798696,
      "step": 5830
    },
    {
      "epoch": 3.058176100628931,
      "grad_norm": 0.13972143828868866,
      "learning_rate": 0.0009915053253758386,
      "loss": 0.4206,
      "num_input_tokens_seen": 3801992,
      "step": 5835
    },
    {
      "epoch": 3.060796645702306,
      "grad_norm": 0.04886050149798393,
      "learning_rate": 0.0009914632990462998,
      "loss": 0.3805,
      "num_input_tokens_seen": 3804648,
      "step": 5840
    },
    {
      "epoch": 3.0634171907756813,
      "grad_norm": 0.05108233168721199,
      "learning_rate": 0.0009914211699082458,
      "loss": 0.651,
      "num_input_tokens_seen": 3808296,
      "step": 5845
    },
    {
      "epoch": 3.0660377358490565,
      "grad_norm": 0.05194234848022461,
      "learning_rate": 0.0009913789379704897,
      "loss": 0.502,
      "num_input_tokens_seen": 3811688,
      "step": 5850
    },
    {
      "epoch": 3.068658280922432,
      "grad_norm": 0.04957089200615883,
      "learning_rate": 0.0009913366032418653,
      "loss": 0.45,
      "num_input_tokens_seen": 3817416,
      "step": 5855
    },
    {
      "epoch": 3.0712788259958073,
      "grad_norm": 0.07099737226963043,
      "learning_rate": 0.0009912941657312293,
      "loss": 0.4744,
      "num_input_tokens_seen": 3820168,
      "step": 5860
    },
    {
      "epoch": 3.0738993710691824,
      "grad_norm": 0.05506044998764992,
      "learning_rate": 0.0009912516254474586,
      "loss": 0.43,
      "num_input_tokens_seen": 3823720,
      "step": 5865
    },
    {
      "epoch": 3.0765199161425576,
      "grad_norm": 0.10908626765012741,
      "learning_rate": 0.0009912089823994525,
      "loss": 0.4877,
      "num_input_tokens_seen": 3827432,
      "step": 5870
    },
    {
      "epoch": 3.0791404612159328,
      "grad_norm": 0.07604335248470306,
      "learning_rate": 0.0009911662365961313,
      "loss": 0.5218,
      "num_input_tokens_seen": 3829832,
      "step": 5875
    },
    {
      "epoch": 3.0817610062893084,
      "grad_norm": 0.07598412036895752,
      "learning_rate": 0.000991123388046437,
      "loss": 0.4529,
      "num_input_tokens_seen": 3832040,
      "step": 5880
    },
    {
      "epoch": 3.0843815513626835,
      "grad_norm": 0.061136163771152496,
      "learning_rate": 0.0009910804367593328,
      "loss": 0.4934,
      "num_input_tokens_seen": 3834984,
      "step": 5885
    },
    {
      "epoch": 3.0870020964360587,
      "grad_norm": 0.06081462651491165,
      "learning_rate": 0.0009910373827438038,
      "loss": 0.5013,
      "num_input_tokens_seen": 3841800,
      "step": 5890
    },
    {
      "epoch": 3.089622641509434,
      "grad_norm": 0.07949317991733551,
      "learning_rate": 0.0009909942260088562,
      "loss": 0.4652,
      "num_input_tokens_seen": 3844456,
      "step": 5895
    },
    {
      "epoch": 3.092243186582809,
      "grad_norm": 0.07855772972106934,
      "learning_rate": 0.0009909509665635184,
      "loss": 0.3641,
      "num_input_tokens_seen": 3847912,
      "step": 5900
    },
    {
      "epoch": 3.0948637316561847,
      "grad_norm": 0.055068086832761765,
      "learning_rate": 0.0009909076044168394,
      "loss": 0.3501,
      "num_input_tokens_seen": 3851208,
      "step": 5905
    },
    {
      "epoch": 3.09748427672956,
      "grad_norm": 0.036729536950588226,
      "learning_rate": 0.00099086413957789,
      "loss": 0.3668,
      "num_input_tokens_seen": 3854920,
      "step": 5910
    },
    {
      "epoch": 3.100104821802935,
      "grad_norm": 0.06389323621988297,
      "learning_rate": 0.0009908205720557627,
      "loss": 0.477,
      "num_input_tokens_seen": 3858408,
      "step": 5915
    },
    {
      "epoch": 3.10272536687631,
      "grad_norm": 0.14579930901527405,
      "learning_rate": 0.0009907769018595713,
      "loss": 0.5255,
      "num_input_tokens_seen": 3861352,
      "step": 5920
    },
    {
      "epoch": 3.1053459119496853,
      "grad_norm": 0.08732905238866806,
      "learning_rate": 0.0009907331289984512,
      "loss": 0.3984,
      "num_input_tokens_seen": 3864680,
      "step": 5925
    },
    {
      "epoch": 3.107966457023061,
      "grad_norm": 0.06737829744815826,
      "learning_rate": 0.000990689253481559,
      "loss": 0.4598,
      "num_input_tokens_seen": 3867176,
      "step": 5930
    },
    {
      "epoch": 3.110587002096436,
      "grad_norm": 0.09614026546478271,
      "learning_rate": 0.000990645275318073,
      "loss": 0.465,
      "num_input_tokens_seen": 3869864,
      "step": 5935
    },
    {
      "epoch": 3.1132075471698113,
      "grad_norm": 0.053791437298059464,
      "learning_rate": 0.000990601194517193,
      "loss": 0.4659,
      "num_input_tokens_seen": 3872744,
      "step": 5940
    },
    {
      "epoch": 3.1158280922431865,
      "grad_norm": 0.09854235500097275,
      "learning_rate": 0.0009905570110881402,
      "loss": 0.4715,
      "num_input_tokens_seen": 3875880,
      "step": 5945
    },
    {
      "epoch": 3.1184486373165616,
      "grad_norm": 0.06205024942755699,
      "learning_rate": 0.0009905127250401573,
      "loss": 0.3938,
      "num_input_tokens_seen": 3879048,
      "step": 5950
    },
    {
      "epoch": 3.1210691823899372,
      "grad_norm": 0.060431044548749924,
      "learning_rate": 0.0009904683363825084,
      "loss": 0.3319,
      "num_input_tokens_seen": 3882376,
      "step": 5955
    },
    {
      "epoch": 3.1236897274633124,
      "grad_norm": 0.05719352141022682,
      "learning_rate": 0.0009904238451244791,
      "loss": 0.4006,
      "num_input_tokens_seen": 3885576,
      "step": 5960
    },
    {
      "epoch": 3.1263102725366876,
      "grad_norm": 0.10711083561182022,
      "learning_rate": 0.0009903792512753764,
      "loss": 0.459,
      "num_input_tokens_seen": 3888808,
      "step": 5965
    },
    {
      "epoch": 3.1289308176100628,
      "grad_norm": 0.09427393227815628,
      "learning_rate": 0.0009903345548445289,
      "loss": 0.469,
      "num_input_tokens_seen": 3891976,
      "step": 5970
    },
    {
      "epoch": 3.131551362683438,
      "grad_norm": 0.07318821549415588,
      "learning_rate": 0.0009902897558412864,
      "loss": 0.4712,
      "num_input_tokens_seen": 3895592,
      "step": 5975
    },
    {
      "epoch": 3.1341719077568135,
      "grad_norm": 0.04478585720062256,
      "learning_rate": 0.0009902448542750207,
      "loss": 0.6429,
      "num_input_tokens_seen": 3899336,
      "step": 5980
    },
    {
      "epoch": 3.1367924528301887,
      "grad_norm": 0.08824245631694794,
      "learning_rate": 0.0009901998501551245,
      "loss": 0.4216,
      "num_input_tokens_seen": 3902280,
      "step": 5985
    },
    {
      "epoch": 3.139412997903564,
      "grad_norm": 0.05094781145453453,
      "learning_rate": 0.000990154743491012,
      "loss": 0.4468,
      "num_input_tokens_seen": 3905672,
      "step": 5990
    },
    {
      "epoch": 3.142033542976939,
      "grad_norm": 0.05628031864762306,
      "learning_rate": 0.0009901095342921193,
      "loss": 0.4141,
      "num_input_tokens_seen": 3908392,
      "step": 5995
    },
    {
      "epoch": 3.1446540880503147,
      "grad_norm": 0.07300570607185364,
      "learning_rate": 0.0009900642225679035,
      "loss": 0.5446,
      "num_input_tokens_seen": 3912200,
      "step": 6000
    },
    {
      "epoch": 3.14727463312369,
      "grad_norm": 0.0839398130774498,
      "learning_rate": 0.000990018808327843,
      "loss": 0.5572,
      "num_input_tokens_seen": 3915304,
      "step": 6005
    },
    {
      "epoch": 3.149895178197065,
      "grad_norm": 0.10688567161560059,
      "learning_rate": 0.0009899732915814386,
      "loss": 0.5156,
      "num_input_tokens_seen": 3918952,
      "step": 6010
    },
    {
      "epoch": 3.15251572327044,
      "grad_norm": 0.04073898494243622,
      "learning_rate": 0.0009899276723382112,
      "loss": 0.6279,
      "num_input_tokens_seen": 3921864,
      "step": 6015
    },
    {
      "epoch": 3.1551362683438153,
      "grad_norm": 0.064847432076931,
      "learning_rate": 0.0009898819506077043,
      "loss": 0.4129,
      "num_input_tokens_seen": 3925256,
      "step": 6020
    },
    {
      "epoch": 3.157756813417191,
      "grad_norm": 0.06672724336385727,
      "learning_rate": 0.0009898361263994823,
      "loss": 0.4149,
      "num_input_tokens_seen": 3927432,
      "step": 6025
    },
    {
      "epoch": 3.160377358490566,
      "grad_norm": 0.07560684531927109,
      "learning_rate": 0.0009897901997231308,
      "loss": 0.4447,
      "num_input_tokens_seen": 3931368,
      "step": 6030
    },
    {
      "epoch": 3.1629979035639413,
      "grad_norm": 0.0683092549443245,
      "learning_rate": 0.0009897441705882576,
      "loss": 0.4313,
      "num_input_tokens_seen": 3934504,
      "step": 6035
    },
    {
      "epoch": 3.1656184486373165,
      "grad_norm": 0.07252301275730133,
      "learning_rate": 0.0009896980390044908,
      "loss": 0.4469,
      "num_input_tokens_seen": 3937384,
      "step": 6040
    },
    {
      "epoch": 3.1682389937106916,
      "grad_norm": 0.10376725345849991,
      "learning_rate": 0.0009896518049814812,
      "loss": 0.5967,
      "num_input_tokens_seen": 3940168,
      "step": 6045
    },
    {
      "epoch": 3.1708595387840672,
      "grad_norm": 0.0640687420964241,
      "learning_rate": 0.0009896054685289005,
      "loss": 0.4171,
      "num_input_tokens_seen": 3942504,
      "step": 6050
    },
    {
      "epoch": 3.1734800838574424,
      "grad_norm": 0.04166639968752861,
      "learning_rate": 0.0009895590296564412,
      "loss": 0.4972,
      "num_input_tokens_seen": 3945480,
      "step": 6055
    },
    {
      "epoch": 3.1761006289308176,
      "grad_norm": 0.08312332630157471,
      "learning_rate": 0.000989512488373818,
      "loss": 0.4607,
      "num_input_tokens_seen": 3948520,
      "step": 6060
    },
    {
      "epoch": 3.1787211740041927,
      "grad_norm": 0.03718854486942291,
      "learning_rate": 0.0009894658446907671,
      "loss": 0.5499,
      "num_input_tokens_seen": 3952904,
      "step": 6065
    },
    {
      "epoch": 3.181341719077568,
      "grad_norm": 0.04424664005637169,
      "learning_rate": 0.0009894190986170458,
      "loss": 0.4733,
      "num_input_tokens_seen": 3955816,
      "step": 6070
    },
    {
      "epoch": 3.1839622641509435,
      "grad_norm": 0.20605581998825073,
      "learning_rate": 0.0009893722501624323,
      "loss": 0.4823,
      "num_input_tokens_seen": 3958120,
      "step": 6075
    },
    {
      "epoch": 3.1865828092243187,
      "grad_norm": 0.10933102667331696,
      "learning_rate": 0.0009893252993367272,
      "loss": 0.53,
      "num_input_tokens_seen": 3960520,
      "step": 6080
    },
    {
      "epoch": 3.189203354297694,
      "grad_norm": 0.10733094066381454,
      "learning_rate": 0.000989278246149752,
      "loss": 0.5046,
      "num_input_tokens_seen": 3963208,
      "step": 6085
    },
    {
      "epoch": 3.191823899371069,
      "grad_norm": 0.10053511708974838,
      "learning_rate": 0.0009892310906113497,
      "loss": 0.5343,
      "num_input_tokens_seen": 3966568,
      "step": 6090
    },
    {
      "epoch": 3.1944444444444446,
      "grad_norm": 0.0871310755610466,
      "learning_rate": 0.0009891838327313847,
      "loss": 0.5541,
      "num_input_tokens_seen": 3970184,
      "step": 6095
    },
    {
      "epoch": 3.19706498951782,
      "grad_norm": 0.07011819630861282,
      "learning_rate": 0.000989136472519743,
      "loss": 0.5266,
      "num_input_tokens_seen": 3973192,
      "step": 6100
    },
    {
      "epoch": 3.199685534591195,
      "grad_norm": 0.07176215946674347,
      "learning_rate": 0.0009890890099863313,
      "loss": 0.4266,
      "num_input_tokens_seen": 3975656,
      "step": 6105
    },
    {
      "epoch": 3.20230607966457,
      "grad_norm": 0.039604559540748596,
      "learning_rate": 0.0009890414451410787,
      "loss": 0.4226,
      "num_input_tokens_seen": 3979240,
      "step": 6110
    },
    {
      "epoch": 3.2049266247379453,
      "grad_norm": 0.03823245316743851,
      "learning_rate": 0.0009889937779939348,
      "loss": 0.4542,
      "num_input_tokens_seen": 3983464,
      "step": 6115
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 0.07671971619129181,
      "learning_rate": 0.0009889460085548715,
      "loss": 0.5238,
      "num_input_tokens_seen": 3986568,
      "step": 6120
    },
    {
      "epoch": 3.210167714884696,
      "grad_norm": 0.05126567929983139,
      "learning_rate": 0.0009888981368338815,
      "loss": 0.4544,
      "num_input_tokens_seen": 3989224,
      "step": 6125
    },
    {
      "epoch": 3.2127882599580713,
      "grad_norm": 0.038751937448978424,
      "learning_rate": 0.0009888501628409789,
      "loss": 0.6113,
      "num_input_tokens_seen": 3992648,
      "step": 6130
    },
    {
      "epoch": 3.2154088050314464,
      "grad_norm": 0.028657792136073112,
      "learning_rate": 0.0009888020865861991,
      "loss": 0.4341,
      "num_input_tokens_seen": 3995528,
      "step": 6135
    },
    {
      "epoch": 3.2180293501048216,
      "grad_norm": 0.06660338491201401,
      "learning_rate": 0.0009887539080795996,
      "loss": 0.4694,
      "num_input_tokens_seen": 3998536,
      "step": 6140
    },
    {
      "epoch": 3.220649895178197,
      "grad_norm": 0.0668451339006424,
      "learning_rate": 0.0009887056273312584,
      "loss": 0.4904,
      "num_input_tokens_seen": 4001224,
      "step": 6145
    },
    {
      "epoch": 3.2232704402515724,
      "grad_norm": 0.05387014150619507,
      "learning_rate": 0.0009886572443512753,
      "loss": 0.4354,
      "num_input_tokens_seen": 4004328,
      "step": 6150
    },
    {
      "epoch": 3.2258909853249476,
      "grad_norm": 0.07249324023723602,
      "learning_rate": 0.0009886087591497717,
      "loss": 0.4629,
      "num_input_tokens_seen": 4007944,
      "step": 6155
    },
    {
      "epoch": 3.2285115303983227,
      "grad_norm": 0.09647758305072784,
      "learning_rate": 0.00098856017173689,
      "loss": 0.4849,
      "num_input_tokens_seen": 4011144,
      "step": 6160
    },
    {
      "epoch": 3.231132075471698,
      "grad_norm": 0.05753650143742561,
      "learning_rate": 0.0009885114821227942,
      "loss": 0.4174,
      "num_input_tokens_seen": 4014376,
      "step": 6165
    },
    {
      "epoch": 3.2337526205450735,
      "grad_norm": 0.0563916377723217,
      "learning_rate": 0.0009884626903176696,
      "loss": 0.489,
      "num_input_tokens_seen": 4018216,
      "step": 6170
    },
    {
      "epoch": 3.2363731656184487,
      "grad_norm": 0.09816310554742813,
      "learning_rate": 0.0009884137963317228,
      "loss": 0.5073,
      "num_input_tokens_seen": 4021032,
      "step": 6175
    },
    {
      "epoch": 3.238993710691824,
      "grad_norm": 0.05571044981479645,
      "learning_rate": 0.000988364800175182,
      "loss": 0.5944,
      "num_input_tokens_seen": 4024040,
      "step": 6180
    },
    {
      "epoch": 3.241614255765199,
      "grad_norm": 0.11038492619991302,
      "learning_rate": 0.0009883157018582966,
      "loss": 0.3505,
      "num_input_tokens_seen": 4027336,
      "step": 6185
    },
    {
      "epoch": 3.2442348008385746,
      "grad_norm": 0.09930907189846039,
      "learning_rate": 0.0009882665013913373,
      "loss": 0.5446,
      "num_input_tokens_seen": 4030472,
      "step": 6190
    },
    {
      "epoch": 3.24685534591195,
      "grad_norm": 0.046584438532590866,
      "learning_rate": 0.0009882171987845962,
      "loss": 0.4677,
      "num_input_tokens_seen": 4034312,
      "step": 6195
    },
    {
      "epoch": 3.249475890985325,
      "grad_norm": 0.03931175544857979,
      "learning_rate": 0.000988167794048387,
      "loss": 0.4686,
      "num_input_tokens_seen": 4038632,
      "step": 6200
    },
    {
      "epoch": 3.2520964360587,
      "grad_norm": 0.064809150993824,
      "learning_rate": 0.0009881182871930448,
      "loss": 0.5439,
      "num_input_tokens_seen": 4042824,
      "step": 6205
    },
    {
      "epoch": 3.2547169811320753,
      "grad_norm": 0.13739801943302155,
      "learning_rate": 0.0009880686782289256,
      "loss": 0.4668,
      "num_input_tokens_seen": 4045896,
      "step": 6210
    },
    {
      "epoch": 3.257337526205451,
      "grad_norm": 0.07217363268136978,
      "learning_rate": 0.000988018967166407,
      "loss": 0.5267,
      "num_input_tokens_seen": 4048712,
      "step": 6215
    },
    {
      "epoch": 3.259958071278826,
      "grad_norm": 0.073341503739357,
      "learning_rate": 0.0009879691540158884,
      "loss": 0.509,
      "num_input_tokens_seen": 4051176,
      "step": 6220
    },
    {
      "epoch": 3.2625786163522013,
      "grad_norm": 0.0647353008389473,
      "learning_rate": 0.0009879192387877895,
      "loss": 0.3601,
      "num_input_tokens_seen": 4054056,
      "step": 6225
    },
    {
      "epoch": 3.2651991614255764,
      "grad_norm": 0.0930057093501091,
      "learning_rate": 0.0009878692214925523,
      "loss": 0.4355,
      "num_input_tokens_seen": 4058120,
      "step": 6230
    },
    {
      "epoch": 3.2678197064989516,
      "grad_norm": 0.0867544412612915,
      "learning_rate": 0.00098781910214064,
      "loss": 0.5397,
      "num_input_tokens_seen": 4060840,
      "step": 6235
    },
    {
      "epoch": 3.270440251572327,
      "grad_norm": 0.03461115062236786,
      "learning_rate": 0.0009877688807425368,
      "loss": 0.543,
      "num_input_tokens_seen": 4064136,
      "step": 6240
    },
    {
      "epoch": 3.2730607966457024,
      "grad_norm": 0.05190691351890564,
      "learning_rate": 0.0009877185573087487,
      "loss": 0.4707,
      "num_input_tokens_seen": 4066824,
      "step": 6245
    },
    {
      "epoch": 3.2756813417190775,
      "grad_norm": 0.09212429076433182,
      "learning_rate": 0.0009876681318498025,
      "loss": 0.6259,
      "num_input_tokens_seen": 4069864,
      "step": 6250
    },
    {
      "epoch": 3.2783018867924527,
      "grad_norm": 0.039051804691553116,
      "learning_rate": 0.0009876176043762466,
      "loss": 0.3532,
      "num_input_tokens_seen": 4072936,
      "step": 6255
    },
    {
      "epoch": 3.280922431865828,
      "grad_norm": 0.053744349628686905,
      "learning_rate": 0.000987566974898651,
      "loss": 0.6389,
      "num_input_tokens_seen": 4075400,
      "step": 6260
    },
    {
      "epoch": 3.2835429769392035,
      "grad_norm": 0.09435269981622696,
      "learning_rate": 0.0009875162434276065,
      "loss": 0.5006,
      "num_input_tokens_seen": 4078696,
      "step": 6265
    },
    {
      "epoch": 3.2861635220125787,
      "grad_norm": 0.08597230166196823,
      "learning_rate": 0.000987465409973726,
      "loss": 0.4968,
      "num_input_tokens_seen": 4082824,
      "step": 6270
    },
    {
      "epoch": 3.288784067085954,
      "grad_norm": 0.061434898525476456,
      "learning_rate": 0.0009874144745476432,
      "loss": 0.5836,
      "num_input_tokens_seen": 4085448,
      "step": 6275
    },
    {
      "epoch": 3.291404612159329,
      "grad_norm": 0.09069880843162537,
      "learning_rate": 0.0009873634371600126,
      "loss": 0.527,
      "num_input_tokens_seen": 4088424,
      "step": 6280
    },
    {
      "epoch": 3.2940251572327046,
      "grad_norm": 0.0533517561852932,
      "learning_rate": 0.0009873122978215115,
      "loss": 0.5667,
      "num_input_tokens_seen": 4091752,
      "step": 6285
    },
    {
      "epoch": 3.29664570230608,
      "grad_norm": 0.08953108638525009,
      "learning_rate": 0.000987261056542837,
      "loss": 0.473,
      "num_input_tokens_seen": 4094568,
      "step": 6290
    },
    {
      "epoch": 3.299266247379455,
      "grad_norm": 0.06329204887151718,
      "learning_rate": 0.0009872097133347085,
      "loss": 0.5049,
      "num_input_tokens_seen": 4097352,
      "step": 6295
    },
    {
      "epoch": 3.30188679245283,
      "grad_norm": 0.0657682865858078,
      "learning_rate": 0.0009871582682078664,
      "loss": 0.3926,
      "num_input_tokens_seen": 4100328,
      "step": 6300
    },
    {
      "epoch": 3.3045073375262053,
      "grad_norm": 0.10409031808376312,
      "learning_rate": 0.0009871067211730722,
      "loss": 0.4389,
      "num_input_tokens_seen": 4102728,
      "step": 6305
    },
    {
      "epoch": 3.307127882599581,
      "grad_norm": 0.06250102072954178,
      "learning_rate": 0.0009870550722411093,
      "loss": 0.3858,
      "num_input_tokens_seen": 4105352,
      "step": 6310
    },
    {
      "epoch": 3.309748427672956,
      "grad_norm": 0.062027379870414734,
      "learning_rate": 0.000987003321422782,
      "loss": 0.4298,
      "num_input_tokens_seen": 4108744,
      "step": 6315
    },
    {
      "epoch": 3.3123689727463312,
      "grad_norm": 0.0593276247382164,
      "learning_rate": 0.0009869514687289155,
      "loss": 0.3658,
      "num_input_tokens_seen": 4112168,
      "step": 6320
    },
    {
      "epoch": 3.3149895178197064,
      "grad_norm": 0.054505862295627594,
      "learning_rate": 0.0009868995141703576,
      "loss": 0.6434,
      "num_input_tokens_seen": 4115240,
      "step": 6325
    },
    {
      "epoch": 3.3176100628930816,
      "grad_norm": 0.06411059945821762,
      "learning_rate": 0.000986847457757976,
      "loss": 0.4707,
      "num_input_tokens_seen": 4117864,
      "step": 6330
    },
    {
      "epoch": 3.320230607966457,
      "grad_norm": 0.07654447108507156,
      "learning_rate": 0.0009867952995026605,
      "loss": 0.4729,
      "num_input_tokens_seen": 4121960,
      "step": 6335
    },
    {
      "epoch": 3.3228511530398324,
      "grad_norm": 0.060268305242061615,
      "learning_rate": 0.0009867430394153221,
      "loss": 0.5743,
      "num_input_tokens_seen": 4125160,
      "step": 6340
    },
    {
      "epoch": 3.3254716981132075,
      "grad_norm": 0.06461812555789948,
      "learning_rate": 0.0009866906775068927,
      "loss": 0.539,
      "num_input_tokens_seen": 4128168,
      "step": 6345
    },
    {
      "epoch": 3.3280922431865827,
      "grad_norm": 0.08534812927246094,
      "learning_rate": 0.0009866382137883262,
      "loss": 0.5723,
      "num_input_tokens_seen": 4134632,
      "step": 6350
    },
    {
      "epoch": 3.330712788259958,
      "grad_norm": 0.03684196621179581,
      "learning_rate": 0.0009865856482705973,
      "loss": 0.5318,
      "num_input_tokens_seen": 4138088,
      "step": 6355
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.07410552352666855,
      "learning_rate": 0.0009865329809647019,
      "loss": 0.435,
      "num_input_tokens_seen": 4141608,
      "step": 6360
    },
    {
      "epoch": 3.3359538784067087,
      "grad_norm": 0.04783371835947037,
      "learning_rate": 0.0009864802118816575,
      "loss": 0.4097,
      "num_input_tokens_seen": 4144712,
      "step": 6365
    },
    {
      "epoch": 3.338574423480084,
      "grad_norm": 0.07309876382350922,
      "learning_rate": 0.0009864273410325028,
      "loss": 0.4182,
      "num_input_tokens_seen": 4147624,
      "step": 6370
    },
    {
      "epoch": 3.341194968553459,
      "grad_norm": 0.0505044125020504,
      "learning_rate": 0.000986374368428298,
      "loss": 0.3881,
      "num_input_tokens_seen": 4150728,
      "step": 6375
    },
    {
      "epoch": 3.3438155136268346,
      "grad_norm": 0.058200154453516006,
      "learning_rate": 0.000986321294080124,
      "loss": 0.4407,
      "num_input_tokens_seen": 4153928,
      "step": 6380
    },
    {
      "epoch": 3.3464360587002098,
      "grad_norm": 0.052826881408691406,
      "learning_rate": 0.0009862681179990838,
      "loss": 0.4131,
      "num_input_tokens_seen": 4157608,
      "step": 6385
    },
    {
      "epoch": 3.349056603773585,
      "grad_norm": 0.0639428123831749,
      "learning_rate": 0.0009862148401963008,
      "loss": 0.4505,
      "num_input_tokens_seen": 4160648,
      "step": 6390
    },
    {
      "epoch": 3.35167714884696,
      "grad_norm": 0.05313502624630928,
      "learning_rate": 0.0009861614606829201,
      "loss": 0.5034,
      "num_input_tokens_seen": 4163720,
      "step": 6395
    },
    {
      "epoch": 3.3542976939203353,
      "grad_norm": 0.11021380126476288,
      "learning_rate": 0.0009861079794701085,
      "loss": 0.4649,
      "num_input_tokens_seen": 4166920,
      "step": 6400
    },
    {
      "epoch": 3.3569182389937104,
      "grad_norm": 0.05679731070995331,
      "learning_rate": 0.000986054396569053,
      "loss": 0.3491,
      "num_input_tokens_seen": 4170024,
      "step": 6405
    },
    {
      "epoch": 3.359538784067086,
      "grad_norm": 0.06631094217300415,
      "learning_rate": 0.0009860007119909635,
      "loss": 0.3968,
      "num_input_tokens_seen": 4173288,
      "step": 6410
    },
    {
      "epoch": 3.3621593291404612,
      "grad_norm": 0.08088095486164093,
      "learning_rate": 0.0009859469257470692,
      "loss": 0.3658,
      "num_input_tokens_seen": 4175688,
      "step": 6415
    },
    {
      "epoch": 3.3647798742138364,
      "grad_norm": 0.09441743046045303,
      "learning_rate": 0.0009858930378486223,
      "loss": 0.4352,
      "num_input_tokens_seen": 4178504,
      "step": 6420
    },
    {
      "epoch": 3.3674004192872116,
      "grad_norm": 0.09636913985013962,
      "learning_rate": 0.000985839048306895,
      "loss": 0.4098,
      "num_input_tokens_seen": 4180840,
      "step": 6425
    },
    {
      "epoch": 3.370020964360587,
      "grad_norm": 0.05671549588441849,
      "learning_rate": 0.000985784957133182,
      "loss": 0.4578,
      "num_input_tokens_seen": 4184328,
      "step": 6430
    },
    {
      "epoch": 3.3726415094339623,
      "grad_norm": 0.06405793130397797,
      "learning_rate": 0.0009857307643387975,
      "loss": 0.5911,
      "num_input_tokens_seen": 4188072,
      "step": 6435
    },
    {
      "epoch": 3.3752620545073375,
      "grad_norm": 0.04881233721971512,
      "learning_rate": 0.000985676469935079,
      "loss": 0.4443,
      "num_input_tokens_seen": 4191848,
      "step": 6440
    },
    {
      "epoch": 3.3778825995807127,
      "grad_norm": 0.05966588482260704,
      "learning_rate": 0.0009856220739333837,
      "loss": 0.4261,
      "num_input_tokens_seen": 4194440,
      "step": 6445
    },
    {
      "epoch": 3.380503144654088,
      "grad_norm": 0.06371483951807022,
      "learning_rate": 0.000985567576345091,
      "loss": 0.5291,
      "num_input_tokens_seen": 4197256,
      "step": 6450
    },
    {
      "epoch": 3.3831236897274635,
      "grad_norm": 0.055057674646377563,
      "learning_rate": 0.0009855129771816006,
      "loss": 0.4742,
      "num_input_tokens_seen": 4200872,
      "step": 6455
    },
    {
      "epoch": 3.3857442348008386,
      "grad_norm": 0.04693373292684555,
      "learning_rate": 0.0009854582764543347,
      "loss": 0.4356,
      "num_input_tokens_seen": 4204232,
      "step": 6460
    },
    {
      "epoch": 3.388364779874214,
      "grad_norm": 0.07885543256998062,
      "learning_rate": 0.0009854034741747356,
      "loss": 0.4427,
      "num_input_tokens_seen": 4207304,
      "step": 6465
    },
    {
      "epoch": 3.390985324947589,
      "grad_norm": 0.06625408679246902,
      "learning_rate": 0.0009853485703542675,
      "loss": 0.4228,
      "num_input_tokens_seen": 4209832,
      "step": 6470
    },
    {
      "epoch": 3.3936058700209646,
      "grad_norm": 0.05585706606507301,
      "learning_rate": 0.0009852935650044158,
      "loss": 0.4525,
      "num_input_tokens_seen": 4213128,
      "step": 6475
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 0.1343909204006195,
      "learning_rate": 0.0009852384581366866,
      "loss": 0.3762,
      "num_input_tokens_seen": 4215976,
      "step": 6480
    },
    {
      "epoch": 3.398846960167715,
      "grad_norm": 0.09440387040376663,
      "learning_rate": 0.000985183249762608,
      "loss": 0.6441,
      "num_input_tokens_seen": 4218632,
      "step": 6485
    },
    {
      "epoch": 3.40146750524109,
      "grad_norm": 0.8846450448036194,
      "learning_rate": 0.000985127939893729,
      "loss": 0.5317,
      "num_input_tokens_seen": 4221416,
      "step": 6490
    },
    {
      "epoch": 3.4040880503144653,
      "grad_norm": 0.11099226027727127,
      "learning_rate": 0.0009850725285416194,
      "loss": 0.3771,
      "num_input_tokens_seen": 4224392,
      "step": 6495
    },
    {
      "epoch": 3.4067085953878404,
      "grad_norm": 0.07966355234384537,
      "learning_rate": 0.0009850170157178707,
      "loss": 0.4756,
      "num_input_tokens_seen": 4227176,
      "step": 6500
    },
    {
      "epoch": 3.409329140461216,
      "grad_norm": 0.0660334974527359,
      "learning_rate": 0.000984961401434096,
      "loss": 0.5219,
      "num_input_tokens_seen": 4229960,
      "step": 6505
    },
    {
      "epoch": 3.411949685534591,
      "grad_norm": 0.09014307707548141,
      "learning_rate": 0.0009849056857019285,
      "loss": 0.4575,
      "num_input_tokens_seen": 4232136,
      "step": 6510
    },
    {
      "epoch": 3.4145702306079664,
      "grad_norm": 0.06552863866090775,
      "learning_rate": 0.0009848498685330238,
      "loss": 0.5589,
      "num_input_tokens_seen": 4235624,
      "step": 6515
    },
    {
      "epoch": 3.4171907756813416,
      "grad_norm": 0.03946317732334137,
      "learning_rate": 0.000984793949939058,
      "loss": 0.5505,
      "num_input_tokens_seen": 4240616,
      "step": 6520
    },
    {
      "epoch": 3.419811320754717,
      "grad_norm": 0.0685093104839325,
      "learning_rate": 0.0009847379299317287,
      "loss": 0.4711,
      "num_input_tokens_seen": 4243528,
      "step": 6525
    },
    {
      "epoch": 3.4224318658280923,
      "grad_norm": 0.04114047810435295,
      "learning_rate": 0.0009846818085227549,
      "loss": 0.5071,
      "num_input_tokens_seen": 4247208,
      "step": 6530
    },
    {
      "epoch": 3.4250524109014675,
      "grad_norm": 0.10900962352752686,
      "learning_rate": 0.000984625585723876,
      "loss": 0.3639,
      "num_input_tokens_seen": 4252424,
      "step": 6535
    },
    {
      "epoch": 3.4276729559748427,
      "grad_norm": 0.08465880900621414,
      "learning_rate": 0.0009845692615468536,
      "loss": 0.5162,
      "num_input_tokens_seen": 4255752,
      "step": 6540
    },
    {
      "epoch": 3.430293501048218,
      "grad_norm": 0.05448804795742035,
      "learning_rate": 0.00098451283600347,
      "loss": 0.5301,
      "num_input_tokens_seen": 4258472,
      "step": 6545
    },
    {
      "epoch": 3.4329140461215935,
      "grad_norm": 0.08166711777448654,
      "learning_rate": 0.0009844563091055286,
      "loss": 0.4059,
      "num_input_tokens_seen": 4260904,
      "step": 6550
    },
    {
      "epoch": 3.4355345911949686,
      "grad_norm": 0.06712278723716736,
      "learning_rate": 0.0009843996808648542,
      "loss": 0.5742,
      "num_input_tokens_seen": 4264104,
      "step": 6555
    },
    {
      "epoch": 3.438155136268344,
      "grad_norm": 0.041991498321294785,
      "learning_rate": 0.000984342951293293,
      "loss": 0.5552,
      "num_input_tokens_seen": 4268648,
      "step": 6560
    },
    {
      "epoch": 3.440775681341719,
      "grad_norm": 0.10317469388246536,
      "learning_rate": 0.0009842861204027121,
      "loss": 0.5316,
      "num_input_tokens_seen": 4272776,
      "step": 6565
    },
    {
      "epoch": 3.4433962264150946,
      "grad_norm": 0.06927745789289474,
      "learning_rate": 0.0009842291882049999,
      "loss": 0.4802,
      "num_input_tokens_seen": 4275784,
      "step": 6570
    },
    {
      "epoch": 3.4460167714884697,
      "grad_norm": 0.06798280030488968,
      "learning_rate": 0.0009841721547120658,
      "loss": 0.5904,
      "num_input_tokens_seen": 4279400,
      "step": 6575
    },
    {
      "epoch": 3.448637316561845,
      "grad_norm": 0.056015726178884506,
      "learning_rate": 0.0009841150199358408,
      "loss": 0.4039,
      "num_input_tokens_seen": 4282920,
      "step": 6580
    },
    {
      "epoch": 3.45125786163522,
      "grad_norm": 0.05791166424751282,
      "learning_rate": 0.0009840577838882765,
      "loss": 0.5152,
      "num_input_tokens_seen": 4286856,
      "step": 6585
    },
    {
      "epoch": 3.4538784067085953,
      "grad_norm": 0.03660045191645622,
      "learning_rate": 0.0009840004465813464,
      "loss": 0.5301,
      "num_input_tokens_seen": 4290216,
      "step": 6590
    },
    {
      "epoch": 3.4564989517819704,
      "grad_norm": 0.12838156521320343,
      "learning_rate": 0.0009839430080270445,
      "loss": 0.6427,
      "num_input_tokens_seen": 4293768,
      "step": 6595
    },
    {
      "epoch": 3.459119496855346,
      "grad_norm": 0.028156893327832222,
      "learning_rate": 0.0009838854682373865,
      "loss": 0.4518,
      "num_input_tokens_seen": 4297288,
      "step": 6600
    },
    {
      "epoch": 3.461740041928721,
      "grad_norm": 0.10319481045007706,
      "learning_rate": 0.000983827827224409,
      "loss": 0.4622,
      "num_input_tokens_seen": 4299592,
      "step": 6605
    },
    {
      "epoch": 3.4643605870020964,
      "grad_norm": 0.28527313470840454,
      "learning_rate": 0.0009837700850001698,
      "loss": 0.4921,
      "num_input_tokens_seen": 4302600,
      "step": 6610
    },
    {
      "epoch": 3.4669811320754715,
      "grad_norm": 0.17643021047115326,
      "learning_rate": 0.000983712241576748,
      "loss": 0.6014,
      "num_input_tokens_seen": 4305640,
      "step": 6615
    },
    {
      "epoch": 3.469601677148847,
      "grad_norm": 0.0895598754286766,
      "learning_rate": 0.000983654296966244,
      "loss": 0.4482,
      "num_input_tokens_seen": 4308168,
      "step": 6620
    },
    {
      "epoch": 3.4722222222222223,
      "grad_norm": 0.08953770250082016,
      "learning_rate": 0.0009835962511807786,
      "loss": 0.4272,
      "num_input_tokens_seen": 4311208,
      "step": 6625
    },
    {
      "epoch": 3.4748427672955975,
      "grad_norm": 0.07544022053480148,
      "learning_rate": 0.0009835381042324948,
      "loss": 0.583,
      "num_input_tokens_seen": 4314312,
      "step": 6630
    },
    {
      "epoch": 3.4774633123689727,
      "grad_norm": 0.06269676238298416,
      "learning_rate": 0.0009834798561335558,
      "loss": 0.5285,
      "num_input_tokens_seen": 4316936,
      "step": 6635
    },
    {
      "epoch": 3.480083857442348,
      "grad_norm": 0.08605948835611343,
      "learning_rate": 0.000983421506896147,
      "loss": 0.4375,
      "num_input_tokens_seen": 4320232,
      "step": 6640
    },
    {
      "epoch": 3.4827044025157234,
      "grad_norm": 0.06639064103364944,
      "learning_rate": 0.000983363056532474,
      "loss": 0.5677,
      "num_input_tokens_seen": 4324104,
      "step": 6645
    },
    {
      "epoch": 3.4853249475890986,
      "grad_norm": 0.056130897253751755,
      "learning_rate": 0.000983304505054764,
      "loss": 0.6485,
      "num_input_tokens_seen": 4327304,
      "step": 6650
    },
    {
      "epoch": 3.487945492662474,
      "grad_norm": 0.08478730171918869,
      "learning_rate": 0.0009832458524752655,
      "loss": 0.5302,
      "num_input_tokens_seen": 4332008,
      "step": 6655
    },
    {
      "epoch": 3.490566037735849,
      "grad_norm": 0.06201785057783127,
      "learning_rate": 0.0009831870988062476,
      "loss": 0.4827,
      "num_input_tokens_seen": 4334728,
      "step": 6660
    },
    {
      "epoch": 3.4931865828092246,
      "grad_norm": 0.06691894680261612,
      "learning_rate": 0.0009831282440600014,
      "loss": 0.482,
      "num_input_tokens_seen": 4337416,
      "step": 6665
    },
    {
      "epoch": 3.4958071278825997,
      "grad_norm": 0.04887093976140022,
      "learning_rate": 0.0009830692882488383,
      "loss": 0.3768,
      "num_input_tokens_seen": 4340328,
      "step": 6670
    },
    {
      "epoch": 3.498427672955975,
      "grad_norm": 0.07142003625631332,
      "learning_rate": 0.000983010231385091,
      "loss": 0.4522,
      "num_input_tokens_seen": 4344456,
      "step": 6675
    },
    {
      "epoch": 3.50104821802935,
      "grad_norm": 0.0525413453578949,
      "learning_rate": 0.000982951073481114,
      "loss": 0.4429,
      "num_input_tokens_seen": 4347944,
      "step": 6680
    },
    {
      "epoch": 3.5036687631027252,
      "grad_norm": 0.053467195481061935,
      "learning_rate": 0.0009828918145492823,
      "loss": 0.5053,
      "num_input_tokens_seen": 4350760,
      "step": 6685
    },
    {
      "epoch": 3.5062893081761004,
      "grad_norm": 0.0512070469558239,
      "learning_rate": 0.000982832454601992,
      "loss": 0.5075,
      "num_input_tokens_seen": 4354760,
      "step": 6690
    },
    {
      "epoch": 3.508909853249476,
      "grad_norm": 0.07834021002054214,
      "learning_rate": 0.0009827729936516605,
      "loss": 0.4261,
      "num_input_tokens_seen": 4358120,
      "step": 6695
    },
    {
      "epoch": 3.511530398322851,
      "grad_norm": 0.04677791893482208,
      "learning_rate": 0.0009827134317107267,
      "loss": 0.4412,
      "num_input_tokens_seen": 4361352,
      "step": 6700
    },
    {
      "epoch": 3.5141509433962264,
      "grad_norm": 0.11636566370725632,
      "learning_rate": 0.0009826537687916501,
      "loss": 0.5416,
      "num_input_tokens_seen": 4363784,
      "step": 6705
    },
    {
      "epoch": 3.5167714884696015,
      "grad_norm": 0.03277561813592911,
      "learning_rate": 0.0009825940049069113,
      "loss": 0.451,
      "num_input_tokens_seen": 4367400,
      "step": 6710
    },
    {
      "epoch": 3.519392033542977,
      "grad_norm": 0.04858701676130295,
      "learning_rate": 0.0009825341400690126,
      "loss": 0.5494,
      "num_input_tokens_seen": 4372232,
      "step": 6715
    },
    {
      "epoch": 3.5220125786163523,
      "grad_norm": 0.04561154171824455,
      "learning_rate": 0.0009824741742904767,
      "loss": 0.4477,
      "num_input_tokens_seen": 4375560,
      "step": 6720
    },
    {
      "epoch": 3.5246331236897275,
      "grad_norm": 0.15785156190395355,
      "learning_rate": 0.000982414107583848,
      "loss": 0.4837,
      "num_input_tokens_seen": 4378216,
      "step": 6725
    },
    {
      "epoch": 3.5272536687631026,
      "grad_norm": 0.1286560595035553,
      "learning_rate": 0.0009823539399616914,
      "loss": 0.6063,
      "num_input_tokens_seen": 4380744,
      "step": 6730
    },
    {
      "epoch": 3.529874213836478,
      "grad_norm": 0.04499130696058273,
      "learning_rate": 0.0009822936714365938,
      "loss": 0.4331,
      "num_input_tokens_seen": 4384264,
      "step": 6735
    },
    {
      "epoch": 3.532494758909853,
      "grad_norm": 0.05660197511315346,
      "learning_rate": 0.0009822333020211623,
      "loss": 0.6028,
      "num_input_tokens_seen": 4388200,
      "step": 6740
    },
    {
      "epoch": 3.5351153039832286,
      "grad_norm": 0.10020502656698227,
      "learning_rate": 0.0009821728317280256,
      "loss": 0.4455,
      "num_input_tokens_seen": 4390760,
      "step": 6745
    },
    {
      "epoch": 3.5377358490566038,
      "grad_norm": 0.10690145939588547,
      "learning_rate": 0.0009821122605698336,
      "loss": 0.533,
      "num_input_tokens_seen": 4393416,
      "step": 6750
    },
    {
      "epoch": 3.540356394129979,
      "grad_norm": 0.04522063955664635,
      "learning_rate": 0.0009820515885592567,
      "loss": 0.4864,
      "num_input_tokens_seen": 4396392,
      "step": 6755
    },
    {
      "epoch": 3.5429769392033545,
      "grad_norm": 0.0498979352414608,
      "learning_rate": 0.0009819908157089872,
      "loss": 0.4414,
      "num_input_tokens_seen": 4399496,
      "step": 6760
    },
    {
      "epoch": 3.5455974842767297,
      "grad_norm": 0.04367312043905258,
      "learning_rate": 0.0009819299420317379,
      "loss": 0.4923,
      "num_input_tokens_seen": 4402440,
      "step": 6765
    },
    {
      "epoch": 3.548218029350105,
      "grad_norm": 0.09597344696521759,
      "learning_rate": 0.0009818689675402427,
      "loss": 0.5752,
      "num_input_tokens_seen": 4405672,
      "step": 6770
    },
    {
      "epoch": 3.55083857442348,
      "grad_norm": 0.12754902243614197,
      "learning_rate": 0.000981807892247257,
      "loss": 0.5709,
      "num_input_tokens_seen": 4408872,
      "step": 6775
    },
    {
      "epoch": 3.5534591194968552,
      "grad_norm": 0.09052684158086777,
      "learning_rate": 0.0009817467161655571,
      "loss": 0.4444,
      "num_input_tokens_seen": 4411112,
      "step": 6780
    },
    {
      "epoch": 3.5560796645702304,
      "grad_norm": 0.054650552570819855,
      "learning_rate": 0.0009816854393079402,
      "loss": 0.4056,
      "num_input_tokens_seen": 4414248,
      "step": 6785
    },
    {
      "epoch": 3.558700209643606,
      "grad_norm": 0.08726654201745987,
      "learning_rate": 0.0009816240616872247,
      "loss": 0.5767,
      "num_input_tokens_seen": 4417160,
      "step": 6790
    },
    {
      "epoch": 3.561320754716981,
      "grad_norm": 0.07599908858537674,
      "learning_rate": 0.0009815625833162502,
      "loss": 0.4323,
      "num_input_tokens_seen": 4421032,
      "step": 6795
    },
    {
      "epoch": 3.5639412997903563,
      "grad_norm": 0.06317907571792603,
      "learning_rate": 0.0009815010042078772,
      "loss": 0.6365,
      "num_input_tokens_seen": 4423848,
      "step": 6800
    },
    {
      "epoch": 3.5665618448637315,
      "grad_norm": 0.07292656600475311,
      "learning_rate": 0.0009814393243749873,
      "loss": 0.421,
      "num_input_tokens_seen": 4427144,
      "step": 6805
    },
    {
      "epoch": 3.569182389937107,
      "grad_norm": 0.0659596249461174,
      "learning_rate": 0.0009813775438304835,
      "loss": 0.5413,
      "num_input_tokens_seen": 4429896,
      "step": 6810
    },
    {
      "epoch": 3.5718029350104823,
      "grad_norm": 0.04988318309187889,
      "learning_rate": 0.0009813156625872893,
      "loss": 0.4846,
      "num_input_tokens_seen": 4435720,
      "step": 6815
    },
    {
      "epoch": 3.5744234800838575,
      "grad_norm": 0.048762399703264236,
      "learning_rate": 0.0009812536806583494,
      "loss": 0.4477,
      "num_input_tokens_seen": 4439048,
      "step": 6820
    },
    {
      "epoch": 3.5770440251572326,
      "grad_norm": 0.05194675177335739,
      "learning_rate": 0.0009811915980566302,
      "loss": 0.4957,
      "num_input_tokens_seen": 4441928,
      "step": 6825
    },
    {
      "epoch": 3.579664570230608,
      "grad_norm": 0.052353568375110626,
      "learning_rate": 0.0009811294147951182,
      "loss": 0.4197,
      "num_input_tokens_seen": 4445160,
      "step": 6830
    },
    {
      "epoch": 3.582285115303983,
      "grad_norm": 0.11009904742240906,
      "learning_rate": 0.000981067130886822,
      "loss": 0.5044,
      "num_input_tokens_seen": 4448328,
      "step": 6835
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 0.07277752459049225,
      "learning_rate": 0.00098100474634477,
      "loss": 0.3976,
      "num_input_tokens_seen": 4451816,
      "step": 6840
    },
    {
      "epoch": 3.5875262054507338,
      "grad_norm": 0.09892505407333374,
      "learning_rate": 0.0009809422611820127,
      "loss": 0.5091,
      "num_input_tokens_seen": 4454856,
      "step": 6845
    },
    {
      "epoch": 3.590146750524109,
      "grad_norm": 0.10072481632232666,
      "learning_rate": 0.0009808796754116212,
      "loss": 0.5015,
      "num_input_tokens_seen": 4457544,
      "step": 6850
    },
    {
      "epoch": 3.5927672955974845,
      "grad_norm": 0.06016822159290314,
      "learning_rate": 0.0009808169890466879,
      "loss": 0.5464,
      "num_input_tokens_seen": 4461160,
      "step": 6855
    },
    {
      "epoch": 3.5953878406708597,
      "grad_norm": 0.08984987437725067,
      "learning_rate": 0.000980754202100326,
      "loss": 0.3941,
      "num_input_tokens_seen": 4465000,
      "step": 6860
    },
    {
      "epoch": 3.598008385744235,
      "grad_norm": 0.08280809223651886,
      "learning_rate": 0.0009806913145856695,
      "loss": 0.5051,
      "num_input_tokens_seen": 4467944,
      "step": 6865
    },
    {
      "epoch": 3.60062893081761,
      "grad_norm": 0.07266128808259964,
      "learning_rate": 0.0009806283265158741,
      "loss": 0.546,
      "num_input_tokens_seen": 4471880,
      "step": 6870
    },
    {
      "epoch": 3.603249475890985,
      "grad_norm": 0.15294958651065826,
      "learning_rate": 0.0009805652379041162,
      "loss": 0.4415,
      "num_input_tokens_seen": 4474184,
      "step": 6875
    },
    {
      "epoch": 3.6058700209643604,
      "grad_norm": 0.08515094220638275,
      "learning_rate": 0.000980502048763593,
      "loss": 0.4141,
      "num_input_tokens_seen": 4476584,
      "step": 6880
    },
    {
      "epoch": 3.608490566037736,
      "grad_norm": 0.08371492475271225,
      "learning_rate": 0.0009804387591075233,
      "loss": 0.4449,
      "num_input_tokens_seen": 4479592,
      "step": 6885
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 0.16043971478939056,
      "learning_rate": 0.000980375368949146,
      "loss": 0.4872,
      "num_input_tokens_seen": 4482280,
      "step": 6890
    },
    {
      "epoch": 3.6137316561844863,
      "grad_norm": 0.10816337913274765,
      "learning_rate": 0.0009803118783017221,
      "loss": 0.4099,
      "num_input_tokens_seen": 4485224,
      "step": 6895
    },
    {
      "epoch": 3.6163522012578615,
      "grad_norm": 0.06475631892681122,
      "learning_rate": 0.0009802482871785329,
      "loss": 0.5568,
      "num_input_tokens_seen": 4488680,
      "step": 6900
    },
    {
      "epoch": 3.618972746331237,
      "grad_norm": 0.0698062926530838,
      "learning_rate": 0.000980184595592881,
      "loss": 0.4953,
      "num_input_tokens_seen": 4492104,
      "step": 6905
    },
    {
      "epoch": 3.6215932914046123,
      "grad_norm": 0.04275326430797577,
      "learning_rate": 0.00098012080355809,
      "loss": 0.4653,
      "num_input_tokens_seen": 4495240,
      "step": 6910
    },
    {
      "epoch": 3.6242138364779874,
      "grad_norm": 0.0907781571149826,
      "learning_rate": 0.0009800569110875043,
      "loss": 0.3641,
      "num_input_tokens_seen": 4498088,
      "step": 6915
    },
    {
      "epoch": 3.6268343815513626,
      "grad_norm": 0.05305417999625206,
      "learning_rate": 0.0009799929181944895,
      "loss": 0.4432,
      "num_input_tokens_seen": 4501224,
      "step": 6920
    },
    {
      "epoch": 3.629454926624738,
      "grad_norm": 0.0803600326180458,
      "learning_rate": 0.0009799288248924325,
      "loss": 0.5693,
      "num_input_tokens_seen": 4503784,
      "step": 6925
    },
    {
      "epoch": 3.632075471698113,
      "grad_norm": 0.07809603214263916,
      "learning_rate": 0.0009798646311947404,
      "loss": 0.4664,
      "num_input_tokens_seen": 4506568,
      "step": 6930
    },
    {
      "epoch": 3.6346960167714886,
      "grad_norm": 0.07083529233932495,
      "learning_rate": 0.0009798003371148422,
      "loss": 0.3817,
      "num_input_tokens_seen": 4509096,
      "step": 6935
    },
    {
      "epoch": 3.6373165618448637,
      "grad_norm": 0.05048134922981262,
      "learning_rate": 0.0009797359426661873,
      "loss": 0.366,
      "num_input_tokens_seen": 4512200,
      "step": 6940
    },
    {
      "epoch": 3.639937106918239,
      "grad_norm": 0.0624811090528965,
      "learning_rate": 0.0009796714478622463,
      "loss": 0.4411,
      "num_input_tokens_seen": 4514984,
      "step": 6945
    },
    {
      "epoch": 3.6425576519916145,
      "grad_norm": 0.055800389498472214,
      "learning_rate": 0.0009796068527165107,
      "loss": 0.4541,
      "num_input_tokens_seen": 4518568,
      "step": 6950
    },
    {
      "epoch": 3.6451781970649897,
      "grad_norm": 0.07580243051052094,
      "learning_rate": 0.0009795421572424935,
      "loss": 0.8077,
      "num_input_tokens_seen": 4521832,
      "step": 6955
    },
    {
      "epoch": 3.647798742138365,
      "grad_norm": 0.09984362870454788,
      "learning_rate": 0.0009794773614537275,
      "loss": 0.5581,
      "num_input_tokens_seen": 4525640,
      "step": 6960
    },
    {
      "epoch": 3.65041928721174,
      "grad_norm": 0.06998997926712036,
      "learning_rate": 0.0009794124653637677,
      "loss": 0.4386,
      "num_input_tokens_seen": 4529608,
      "step": 6965
    },
    {
      "epoch": 3.653039832285115,
      "grad_norm": 0.10325475037097931,
      "learning_rate": 0.00097934746898619,
      "loss": 0.5619,
      "num_input_tokens_seen": 4532328,
      "step": 6970
    },
    {
      "epoch": 3.6556603773584904,
      "grad_norm": 0.051526233553886414,
      "learning_rate": 0.00097928237233459,
      "loss": 0.505,
      "num_input_tokens_seen": 4535336,
      "step": 6975
    },
    {
      "epoch": 3.658280922431866,
      "grad_norm": 0.048677049577236176,
      "learning_rate": 0.0009792171754225859,
      "loss": 0.3757,
      "num_input_tokens_seen": 4538760,
      "step": 6980
    },
    {
      "epoch": 3.660901467505241,
      "grad_norm": 0.07046414166688919,
      "learning_rate": 0.0009791518782638158,
      "loss": 0.4245,
      "num_input_tokens_seen": 4546376,
      "step": 6985
    },
    {
      "epoch": 3.6635220125786163,
      "grad_norm": 0.09899493306875229,
      "learning_rate": 0.0009790864808719392,
      "loss": 0.4953,
      "num_input_tokens_seen": 4550376,
      "step": 6990
    },
    {
      "epoch": 3.6661425576519915,
      "grad_norm": 0.05240057408809662,
      "learning_rate": 0.0009790209832606365,
      "loss": 0.5907,
      "num_input_tokens_seen": 4553736,
      "step": 6995
    },
    {
      "epoch": 3.668763102725367,
      "grad_norm": 0.053665339946746826,
      "learning_rate": 0.000978955385443609,
      "loss": 0.4317,
      "num_input_tokens_seen": 4556936,
      "step": 7000
    },
    {
      "epoch": 3.6713836477987423,
      "grad_norm": 0.07537385076284409,
      "learning_rate": 0.0009788896874345792,
      "loss": 0.4601,
      "num_input_tokens_seen": 4559816,
      "step": 7005
    },
    {
      "epoch": 3.6740041928721174,
      "grad_norm": 0.08650009334087372,
      "learning_rate": 0.0009788238892472904,
      "loss": 0.4487,
      "num_input_tokens_seen": 4564520,
      "step": 7010
    },
    {
      "epoch": 3.6766247379454926,
      "grad_norm": 0.07963670045137405,
      "learning_rate": 0.0009787579908955063,
      "loss": 0.5149,
      "num_input_tokens_seen": 4567592,
      "step": 7015
    },
    {
      "epoch": 3.6792452830188678,
      "grad_norm": 0.05409886687994003,
      "learning_rate": 0.0009786919923930127,
      "loss": 0.4925,
      "num_input_tokens_seen": 4570408,
      "step": 7020
    },
    {
      "epoch": 3.681865828092243,
      "grad_norm": 0.08013308793306351,
      "learning_rate": 0.0009786258937536155,
      "loss": 0.4551,
      "num_input_tokens_seen": 4573320,
      "step": 7025
    },
    {
      "epoch": 3.6844863731656186,
      "grad_norm": 0.04562172666192055,
      "learning_rate": 0.0009785596949911418,
      "loss": 0.5497,
      "num_input_tokens_seen": 4576680,
      "step": 7030
    },
    {
      "epoch": 3.6871069182389937,
      "grad_norm": 0.11247482150793076,
      "learning_rate": 0.0009784933961194395,
      "loss": 0.4808,
      "num_input_tokens_seen": 4579368,
      "step": 7035
    },
    {
      "epoch": 3.689727463312369,
      "grad_norm": 0.053361572325229645,
      "learning_rate": 0.0009784269971523777,
      "loss": 0.611,
      "num_input_tokens_seen": 4582440,
      "step": 7040
    },
    {
      "epoch": 3.6923480083857445,
      "grad_norm": 0.06394883990287781,
      "learning_rate": 0.0009783604981038463,
      "loss": 0.4147,
      "num_input_tokens_seen": 4585128,
      "step": 7045
    },
    {
      "epoch": 3.6949685534591197,
      "grad_norm": 0.07013853639364243,
      "learning_rate": 0.000978293898987756,
      "loss": 0.4583,
      "num_input_tokens_seen": 4588168,
      "step": 7050
    },
    {
      "epoch": 3.697589098532495,
      "grad_norm": 0.1093866303563118,
      "learning_rate": 0.000978227199818039,
      "loss": 0.4511,
      "num_input_tokens_seen": 4591400,
      "step": 7055
    },
    {
      "epoch": 3.70020964360587,
      "grad_norm": 0.08755233883857727,
      "learning_rate": 0.0009781604006086474,
      "loss": 0.4781,
      "num_input_tokens_seen": 4594280,
      "step": 7060
    },
    {
      "epoch": 3.702830188679245,
      "grad_norm": 0.06553223729133606,
      "learning_rate": 0.0009780935013735553,
      "loss": 0.3747,
      "num_input_tokens_seen": 4598536,
      "step": 7065
    },
    {
      "epoch": 3.7054507337526204,
      "grad_norm": 0.07291843742132187,
      "learning_rate": 0.0009780265021267572,
      "loss": 0.5027,
      "num_input_tokens_seen": 4602440,
      "step": 7070
    },
    {
      "epoch": 3.708071278825996,
      "grad_norm": 0.05788036063313484,
      "learning_rate": 0.0009779594028822682,
      "loss": 0.4903,
      "num_input_tokens_seen": 4605352,
      "step": 7075
    },
    {
      "epoch": 3.710691823899371,
      "grad_norm": 0.05221452936530113,
      "learning_rate": 0.0009778922036541252,
      "loss": 0.3695,
      "num_input_tokens_seen": 4610152,
      "step": 7080
    },
    {
      "epoch": 3.7133123689727463,
      "grad_norm": 0.054633740335702896,
      "learning_rate": 0.0009778249044563852,
      "loss": 0.3967,
      "num_input_tokens_seen": 4612872,
      "step": 7085
    },
    {
      "epoch": 3.7159329140461215,
      "grad_norm": 0.04582147300243378,
      "learning_rate": 0.0009777575053031263,
      "loss": 0.5534,
      "num_input_tokens_seen": 4617064,
      "step": 7090
    },
    {
      "epoch": 3.718553459119497,
      "grad_norm": 0.1390591561794281,
      "learning_rate": 0.000977690006208448,
      "loss": 0.5525,
      "num_input_tokens_seen": 4621608,
      "step": 7095
    },
    {
      "epoch": 3.7211740041928723,
      "grad_norm": 0.12825602293014526,
      "learning_rate": 0.0009776224071864703,
      "loss": 0.4814,
      "num_input_tokens_seen": 4623880,
      "step": 7100
    },
    {
      "epoch": 3.7237945492662474,
      "grad_norm": 0.05068988353013992,
      "learning_rate": 0.000977554708251334,
      "loss": 0.4466,
      "num_input_tokens_seen": 4627080,
      "step": 7105
    },
    {
      "epoch": 3.7264150943396226,
      "grad_norm": 0.05098330229520798,
      "learning_rate": 0.0009774869094172007,
      "loss": 0.4294,
      "num_input_tokens_seen": 4630536,
      "step": 7110
    },
    {
      "epoch": 3.7290356394129978,
      "grad_norm": 0.10439322143793106,
      "learning_rate": 0.0009774190106982537,
      "loss": 0.5182,
      "num_input_tokens_seen": 4633256,
      "step": 7115
    },
    {
      "epoch": 3.731656184486373,
      "grad_norm": 0.05015382170677185,
      "learning_rate": 0.0009773510121086962,
      "loss": 0.4746,
      "num_input_tokens_seen": 4636104,
      "step": 7120
    },
    {
      "epoch": 3.7342767295597485,
      "grad_norm": 0.06644609570503235,
      "learning_rate": 0.0009772829136627528,
      "loss": 0.4587,
      "num_input_tokens_seen": 4639272,
      "step": 7125
    },
    {
      "epoch": 3.7368972746331237,
      "grad_norm": 0.07929859310388565,
      "learning_rate": 0.0009772147153746691,
      "loss": 0.3604,
      "num_input_tokens_seen": 4641800,
      "step": 7130
    },
    {
      "epoch": 3.739517819706499,
      "grad_norm": 0.051780957728624344,
      "learning_rate": 0.0009771464172587112,
      "loss": 0.5978,
      "num_input_tokens_seen": 4645160,
      "step": 7135
    },
    {
      "epoch": 3.742138364779874,
      "grad_norm": 0.0851329043507576,
      "learning_rate": 0.0009770780193291667,
      "loss": 0.3712,
      "num_input_tokens_seen": 4648488,
      "step": 7140
    },
    {
      "epoch": 3.7447589098532497,
      "grad_norm": 0.08531521260738373,
      "learning_rate": 0.000977009521600343,
      "loss": 0.4317,
      "num_input_tokens_seen": 4651944,
      "step": 7145
    },
    {
      "epoch": 3.747379454926625,
      "grad_norm": 0.08287575840950012,
      "learning_rate": 0.0009769409240865696,
      "loss": 0.5142,
      "num_input_tokens_seen": 4655816,
      "step": 7150
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.08123000711202621,
      "learning_rate": 0.0009768722268021959,
      "loss": 0.4404,
      "num_input_tokens_seen": 4658824,
      "step": 7155
    },
    {
      "epoch": 3.752620545073375,
      "grad_norm": 0.059862956404685974,
      "learning_rate": 0.000976803429761593,
      "loss": 0.5171,
      "num_input_tokens_seen": 4662440,
      "step": 7160
    },
    {
      "epoch": 3.7552410901467503,
      "grad_norm": 0.07570947706699371,
      "learning_rate": 0.000976734532979152,
      "loss": 0.4572,
      "num_input_tokens_seen": 4666280,
      "step": 7165
    },
    {
      "epoch": 3.757861635220126,
      "grad_norm": 0.04281763732433319,
      "learning_rate": 0.000976665536469286,
      "loss": 0.3967,
      "num_input_tokens_seen": 4669544,
      "step": 7170
    },
    {
      "epoch": 3.760482180293501,
      "grad_norm": 0.051875293254852295,
      "learning_rate": 0.0009765964402464276,
      "loss": 0.4521,
      "num_input_tokens_seen": 4672392,
      "step": 7175
    },
    {
      "epoch": 3.7631027253668763,
      "grad_norm": 0.05710861086845398,
      "learning_rate": 0.0009765272443250312,
      "loss": 0.5575,
      "num_input_tokens_seen": 4676872,
      "step": 7180
    },
    {
      "epoch": 3.7657232704402515,
      "grad_norm": 0.04349835589528084,
      "learning_rate": 0.0009764579487195717,
      "loss": 0.6078,
      "num_input_tokens_seen": 4680520,
      "step": 7185
    },
    {
      "epoch": 3.768343815513627,
      "grad_norm": 0.03512322157621384,
      "learning_rate": 0.0009763885534445452,
      "loss": 0.4637,
      "num_input_tokens_seen": 4683464,
      "step": 7190
    },
    {
      "epoch": 3.7709643605870022,
      "grad_norm": 0.060221754014492035,
      "learning_rate": 0.0009763190585144682,
      "loss": 0.4517,
      "num_input_tokens_seen": 4686728,
      "step": 7195
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 0.08129407465457916,
      "learning_rate": 0.0009762494639438783,
      "loss": 0.476,
      "num_input_tokens_seen": 4690376,
      "step": 7200
    },
    {
      "epoch": 3.7762054507337526,
      "grad_norm": 0.08272235095500946,
      "learning_rate": 0.0009761797697473339,
      "loss": 0.4421,
      "num_input_tokens_seen": 4695304,
      "step": 7205
    },
    {
      "epoch": 3.7788259958071277,
      "grad_norm": 0.09215036034584045,
      "learning_rate": 0.0009761099759394142,
      "loss": 0.5524,
      "num_input_tokens_seen": 4698216,
      "step": 7210
    },
    {
      "epoch": 3.781446540880503,
      "grad_norm": 0.0692506805062294,
      "learning_rate": 0.0009760400825347194,
      "loss": 0.5829,
      "num_input_tokens_seen": 4701512,
      "step": 7215
    },
    {
      "epoch": 3.7840670859538785,
      "grad_norm": 0.08365702629089355,
      "learning_rate": 0.00097597008954787,
      "loss": 0.5409,
      "num_input_tokens_seen": 4704744,
      "step": 7220
    },
    {
      "epoch": 3.7866876310272537,
      "grad_norm": 0.07668133825063705,
      "learning_rate": 0.0009758999969935083,
      "loss": 0.612,
      "num_input_tokens_seen": 4707880,
      "step": 7225
    },
    {
      "epoch": 3.789308176100629,
      "grad_norm": 0.0457402803003788,
      "learning_rate": 0.0009758298048862965,
      "loss": 0.4144,
      "num_input_tokens_seen": 4711080,
      "step": 7230
    },
    {
      "epoch": 3.791928721174004,
      "grad_norm": 0.035025108605623245,
      "learning_rate": 0.0009757595132409181,
      "loss": 0.4047,
      "num_input_tokens_seen": 4714760,
      "step": 7235
    },
    {
      "epoch": 3.7945492662473796,
      "grad_norm": 0.04578512907028198,
      "learning_rate": 0.0009756891220720771,
      "loss": 0.4132,
      "num_input_tokens_seen": 4718248,
      "step": 7240
    },
    {
      "epoch": 3.797169811320755,
      "grad_norm": 0.09916579723358154,
      "learning_rate": 0.0009756186313944988,
      "loss": 0.7935,
      "num_input_tokens_seen": 4721480,
      "step": 7245
    },
    {
      "epoch": 3.79979035639413,
      "grad_norm": 0.04685785621404648,
      "learning_rate": 0.0009755480412229291,
      "loss": 0.4794,
      "num_input_tokens_seen": 4724328,
      "step": 7250
    },
    {
      "epoch": 3.802410901467505,
      "grad_norm": 0.0759706199169159,
      "learning_rate": 0.0009754773515721343,
      "loss": 0.4169,
      "num_input_tokens_seen": 4727432,
      "step": 7255
    },
    {
      "epoch": 3.8050314465408803,
      "grad_norm": 0.03543663024902344,
      "learning_rate": 0.0009754065624569022,
      "loss": 0.5702,
      "num_input_tokens_seen": 4730952,
      "step": 7260
    },
    {
      "epoch": 3.8076519916142555,
      "grad_norm": 0.10474391281604767,
      "learning_rate": 0.000975335673892041,
      "loss": 0.5236,
      "num_input_tokens_seen": 4733704,
      "step": 7265
    },
    {
      "epoch": 3.810272536687631,
      "grad_norm": 0.05236850306391716,
      "learning_rate": 0.0009752646858923797,
      "loss": 0.4985,
      "num_input_tokens_seen": 4736424,
      "step": 7270
    },
    {
      "epoch": 3.8128930817610063,
      "grad_norm": 0.04038297384977341,
      "learning_rate": 0.0009751935984727683,
      "loss": 0.456,
      "num_input_tokens_seen": 4739944,
      "step": 7275
    },
    {
      "epoch": 3.8155136268343814,
      "grad_norm": 0.04912659898400307,
      "learning_rate": 0.0009751224116480772,
      "loss": 0.4036,
      "num_input_tokens_seen": 4743656,
      "step": 7280
    },
    {
      "epoch": 3.818134171907757,
      "grad_norm": 0.05510174110531807,
      "learning_rate": 0.0009750511254331982,
      "loss": 0.4539,
      "num_input_tokens_seen": 4746440,
      "step": 7285
    },
    {
      "epoch": 3.8207547169811322,
      "grad_norm": 0.08203623443841934,
      "learning_rate": 0.0009749797398430433,
      "loss": 0.4757,
      "num_input_tokens_seen": 4749672,
      "step": 7290
    },
    {
      "epoch": 3.8233752620545074,
      "grad_norm": 0.04359857365489006,
      "learning_rate": 0.0009749082548925459,
      "loss": 0.5758,
      "num_input_tokens_seen": 4753224,
      "step": 7295
    },
    {
      "epoch": 3.8259958071278826,
      "grad_norm": 0.06915987282991409,
      "learning_rate": 0.0009748366705966593,
      "loss": 0.4205,
      "num_input_tokens_seen": 4756072,
      "step": 7300
    },
    {
      "epoch": 3.8286163522012577,
      "grad_norm": 0.05264013633131981,
      "learning_rate": 0.0009747649869703588,
      "loss": 0.418,
      "num_input_tokens_seen": 4759368,
      "step": 7305
    },
    {
      "epoch": 3.831236897274633,
      "grad_norm": 0.04266665130853653,
      "learning_rate": 0.0009746932040286391,
      "loss": 0.5881,
      "num_input_tokens_seen": 4762280,
      "step": 7310
    },
    {
      "epoch": 3.8338574423480085,
      "grad_norm": 0.05769753083586693,
      "learning_rate": 0.000974621321786517,
      "loss": 0.4295,
      "num_input_tokens_seen": 4765544,
      "step": 7315
    },
    {
      "epoch": 3.8364779874213837,
      "grad_norm": 0.08490574359893799,
      "learning_rate": 0.000974549340259029,
      "loss": 0.4743,
      "num_input_tokens_seen": 4768424,
      "step": 7320
    },
    {
      "epoch": 3.839098532494759,
      "grad_norm": 0.07902824133634567,
      "learning_rate": 0.000974477259461233,
      "loss": 0.4627,
      "num_input_tokens_seen": 4771464,
      "step": 7325
    },
    {
      "epoch": 3.841719077568134,
      "grad_norm": 0.045254018157720566,
      "learning_rate": 0.0009744050794082074,
      "loss": 0.4337,
      "num_input_tokens_seen": 4775208,
      "step": 7330
    },
    {
      "epoch": 3.8443396226415096,
      "grad_norm": 0.07592577487230301,
      "learning_rate": 0.0009743328001150515,
      "loss": 0.3622,
      "num_input_tokens_seen": 4777768,
      "step": 7335
    },
    {
      "epoch": 3.846960167714885,
      "grad_norm": 0.039850011467933655,
      "learning_rate": 0.0009742604215968853,
      "loss": 0.4101,
      "num_input_tokens_seen": 4781864,
      "step": 7340
    },
    {
      "epoch": 3.84958071278826,
      "grad_norm": 0.07044629007577896,
      "learning_rate": 0.0009741879438688495,
      "loss": 0.405,
      "num_input_tokens_seen": 4785352,
      "step": 7345
    },
    {
      "epoch": 3.852201257861635,
      "grad_norm": 0.12438792735338211,
      "learning_rate": 0.0009741153669461058,
      "loss": 0.4652,
      "num_input_tokens_seen": 4787848,
      "step": 7350
    },
    {
      "epoch": 3.8548218029350103,
      "grad_norm": 0.07623899728059769,
      "learning_rate": 0.0009740426908438362,
      "loss": 0.3329,
      "num_input_tokens_seen": 4790952,
      "step": 7355
    },
    {
      "epoch": 3.8574423480083855,
      "grad_norm": 0.03044038824737072,
      "learning_rate": 0.0009739699155772439,
      "loss": 0.2973,
      "num_input_tokens_seen": 4794824,
      "step": 7360
    },
    {
      "epoch": 3.860062893081761,
      "grad_norm": 0.05770154669880867,
      "learning_rate": 0.0009738970411615525,
      "loss": 0.4267,
      "num_input_tokens_seen": 4798120,
      "step": 7365
    },
    {
      "epoch": 3.8626834381551363,
      "grad_norm": 0.0412030927836895,
      "learning_rate": 0.0009738240676120067,
      "loss": 0.4385,
      "num_input_tokens_seen": 4801704,
      "step": 7370
    },
    {
      "epoch": 3.8653039832285114,
      "grad_norm": 0.08096954971551895,
      "learning_rate": 0.0009737509949438717,
      "loss": 0.6351,
      "num_input_tokens_seen": 4804776,
      "step": 7375
    },
    {
      "epoch": 3.867924528301887,
      "grad_norm": 0.05037764832377434,
      "learning_rate": 0.0009736778231724333,
      "loss": 0.6406,
      "num_input_tokens_seen": 4807496,
      "step": 7380
    },
    {
      "epoch": 3.870545073375262,
      "grad_norm": 0.06015806272625923,
      "learning_rate": 0.0009736045523129982,
      "loss": 0.4659,
      "num_input_tokens_seen": 4810152,
      "step": 7385
    },
    {
      "epoch": 3.8731656184486374,
      "grad_norm": 0.048383478075265884,
      "learning_rate": 0.0009735311823808938,
      "loss": 0.3608,
      "num_input_tokens_seen": 4813736,
      "step": 7390
    },
    {
      "epoch": 3.8757861635220126,
      "grad_norm": 0.08621350675821304,
      "learning_rate": 0.0009734577133914687,
      "loss": 0.5668,
      "num_input_tokens_seen": 4816904,
      "step": 7395
    },
    {
      "epoch": 3.8784067085953877,
      "grad_norm": 0.05802103877067566,
      "learning_rate": 0.0009733841453600914,
      "loss": 0.4104,
      "num_input_tokens_seen": 4820168,
      "step": 7400
    },
    {
      "epoch": 3.881027253668763,
      "grad_norm": 0.059684570878744125,
      "learning_rate": 0.0009733104783021515,
      "loss": 0.6474,
      "num_input_tokens_seen": 4823176,
      "step": 7405
    },
    {
      "epoch": 3.8836477987421385,
      "grad_norm": 0.06122591346502304,
      "learning_rate": 0.0009732367122330593,
      "loss": 0.4528,
      "num_input_tokens_seen": 4826344,
      "step": 7410
    },
    {
      "epoch": 3.8862683438155137,
      "grad_norm": 0.055856090039014816,
      "learning_rate": 0.0009731628471682459,
      "loss": 0.4533,
      "num_input_tokens_seen": 4830120,
      "step": 7415
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.059150271117687225,
      "learning_rate": 0.000973088883123163,
      "loss": 0.4436,
      "num_input_tokens_seen": 4833576,
      "step": 7420
    },
    {
      "epoch": 3.891509433962264,
      "grad_norm": 0.06331046670675278,
      "learning_rate": 0.0009730148201132829,
      "loss": 0.3908,
      "num_input_tokens_seen": 4836776,
      "step": 7425
    },
    {
      "epoch": 3.8941299790356396,
      "grad_norm": 0.036323051899671555,
      "learning_rate": 0.0009729406581540991,
      "loss": 0.6002,
      "num_input_tokens_seen": 4840808,
      "step": 7430
    },
    {
      "epoch": 3.896750524109015,
      "grad_norm": 0.04415348172187805,
      "learning_rate": 0.0009728663972611251,
      "loss": 0.5569,
      "num_input_tokens_seen": 4847624,
      "step": 7435
    },
    {
      "epoch": 3.89937106918239,
      "grad_norm": 0.04500863701105118,
      "learning_rate": 0.0009727920374498955,
      "loss": 0.5714,
      "num_input_tokens_seen": 4850504,
      "step": 7440
    },
    {
      "epoch": 3.901991614255765,
      "grad_norm": 0.054009053856134415,
      "learning_rate": 0.0009727175787359656,
      "loss": 0.3912,
      "num_input_tokens_seen": 4853448,
      "step": 7445
    },
    {
      "epoch": 3.9046121593291403,
      "grad_norm": 0.07760302722454071,
      "learning_rate": 0.0009726430211349113,
      "loss": 0.4788,
      "num_input_tokens_seen": 4856040,
      "step": 7450
    },
    {
      "epoch": 3.9072327044025155,
      "grad_norm": 0.04815470427274704,
      "learning_rate": 0.0009725683646623291,
      "loss": 0.5021,
      "num_input_tokens_seen": 4859976,
      "step": 7455
    },
    {
      "epoch": 3.909853249475891,
      "grad_norm": 0.048037540167570114,
      "learning_rate": 0.0009724936093338365,
      "loss": 0.4916,
      "num_input_tokens_seen": 4863304,
      "step": 7460
    },
    {
      "epoch": 3.9124737945492662,
      "grad_norm": 0.06990426778793335,
      "learning_rate": 0.0009724187551650712,
      "loss": 0.6107,
      "num_input_tokens_seen": 4866568,
      "step": 7465
    },
    {
      "epoch": 3.9150943396226414,
      "grad_norm": 0.08727290481328964,
      "learning_rate": 0.0009723438021716919,
      "loss": 0.5308,
      "num_input_tokens_seen": 4870312,
      "step": 7470
    },
    {
      "epoch": 3.917714884696017,
      "grad_norm": 0.06200450658798218,
      "learning_rate": 0.0009722687503693782,
      "loss": 0.6377,
      "num_input_tokens_seen": 4873864,
      "step": 7475
    },
    {
      "epoch": 3.920335429769392,
      "grad_norm": 0.04702834039926529,
      "learning_rate": 0.0009721935997738296,
      "loss": 0.5481,
      "num_input_tokens_seen": 4877032,
      "step": 7480
    },
    {
      "epoch": 3.9229559748427674,
      "grad_norm": 0.05985569953918457,
      "learning_rate": 0.0009721183504007671,
      "loss": 0.5004,
      "num_input_tokens_seen": 4880136,
      "step": 7485
    },
    {
      "epoch": 3.9255765199161425,
      "grad_norm": 0.06989230960607529,
      "learning_rate": 0.0009720430022659319,
      "loss": 0.4855,
      "num_input_tokens_seen": 4883720,
      "step": 7490
    },
    {
      "epoch": 3.9281970649895177,
      "grad_norm": 0.08982371538877487,
      "learning_rate": 0.000971967555385086,
      "loss": 0.5803,
      "num_input_tokens_seen": 4886440,
      "step": 7495
    },
    {
      "epoch": 3.930817610062893,
      "grad_norm": 0.08365894109010696,
      "learning_rate": 0.000971892009774012,
      "loss": 0.4955,
      "num_input_tokens_seen": 4888936,
      "step": 7500
    },
    {
      "epoch": 3.9334381551362685,
      "grad_norm": 0.2529478669166565,
      "learning_rate": 0.0009718163654485133,
      "loss": 0.5093,
      "num_input_tokens_seen": 4890952,
      "step": 7505
    },
    {
      "epoch": 3.9360587002096437,
      "grad_norm": 0.10008673369884491,
      "learning_rate": 0.0009717406224244136,
      "loss": 0.5502,
      "num_input_tokens_seen": 4893736,
      "step": 7510
    },
    {
      "epoch": 3.938679245283019,
      "grad_norm": 0.051501188427209854,
      "learning_rate": 0.0009716647807175575,
      "loss": 0.5686,
      "num_input_tokens_seen": 4897384,
      "step": 7515
    },
    {
      "epoch": 3.941299790356394,
      "grad_norm": 0.05257077515125275,
      "learning_rate": 0.0009715888403438105,
      "loss": 0.4949,
      "num_input_tokens_seen": 4900808,
      "step": 7520
    },
    {
      "epoch": 3.9439203354297696,
      "grad_norm": 0.12736232578754425,
      "learning_rate": 0.0009715128013190581,
      "loss": 0.7748,
      "num_input_tokens_seen": 4903592,
      "step": 7525
    },
    {
      "epoch": 3.9465408805031448,
      "grad_norm": 0.05789962038397789,
      "learning_rate": 0.0009714366636592069,
      "loss": 0.4723,
      "num_input_tokens_seen": 4907048,
      "step": 7530
    },
    {
      "epoch": 3.94916142557652,
      "grad_norm": 0.0748300701379776,
      "learning_rate": 0.0009713604273801844,
      "loss": 0.484,
      "num_input_tokens_seen": 4909960,
      "step": 7535
    },
    {
      "epoch": 3.951781970649895,
      "grad_norm": 0.05204112455248833,
      "learning_rate": 0.0009712840924979378,
      "loss": 0.366,
      "num_input_tokens_seen": 4913768,
      "step": 7540
    },
    {
      "epoch": 3.9544025157232703,
      "grad_norm": 0.12814274430274963,
      "learning_rate": 0.0009712076590284357,
      "loss": 0.5326,
      "num_input_tokens_seen": 4917416,
      "step": 7545
    },
    {
      "epoch": 3.9570230607966455,
      "grad_norm": 0.05100652948021889,
      "learning_rate": 0.0009711311269876674,
      "loss": 0.3757,
      "num_input_tokens_seen": 4922056,
      "step": 7550
    },
    {
      "epoch": 3.959643605870021,
      "grad_norm": 0.10957393050193787,
      "learning_rate": 0.0009710544963916421,
      "loss": 0.5003,
      "num_input_tokens_seen": 4925320,
      "step": 7555
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 0.10832369327545166,
      "learning_rate": 0.0009709777672563903,
      "loss": 0.6686,
      "num_input_tokens_seen": 4928168,
      "step": 7560
    },
    {
      "epoch": 3.9648846960167714,
      "grad_norm": 0.0850154459476471,
      "learning_rate": 0.0009709009395979628,
      "loss": 0.4528,
      "num_input_tokens_seen": 4931176,
      "step": 7565
    },
    {
      "epoch": 3.967505241090147,
      "grad_norm": 0.05401380732655525,
      "learning_rate": 0.0009708240134324311,
      "loss": 0.4192,
      "num_input_tokens_seen": 4935304,
      "step": 7570
    },
    {
      "epoch": 3.970125786163522,
      "grad_norm": 0.059037186205387115,
      "learning_rate": 0.0009707469887758871,
      "loss": 0.5299,
      "num_input_tokens_seen": 4938600,
      "step": 7575
    },
    {
      "epoch": 3.9727463312368974,
      "grad_norm": 0.09470891952514648,
      "learning_rate": 0.0009706698656444437,
      "loss": 0.4068,
      "num_input_tokens_seen": 4941512,
      "step": 7580
    },
    {
      "epoch": 3.9753668763102725,
      "grad_norm": 0.04155610129237175,
      "learning_rate": 0.000970592644054234,
      "loss": 0.4638,
      "num_input_tokens_seen": 4944744,
      "step": 7585
    },
    {
      "epoch": 3.9779874213836477,
      "grad_norm": 0.11489833891391754,
      "learning_rate": 0.000970515324021412,
      "loss": 0.4887,
      "num_input_tokens_seen": 4948072,
      "step": 7590
    },
    {
      "epoch": 3.980607966457023,
      "grad_norm": 0.062172554433345795,
      "learning_rate": 0.0009704379055621523,
      "loss": 0.5549,
      "num_input_tokens_seen": 4951656,
      "step": 7595
    },
    {
      "epoch": 3.9832285115303985,
      "grad_norm": 0.03414211422204971,
      "learning_rate": 0.0009703603886926497,
      "loss": 0.3052,
      "num_input_tokens_seen": 4955400,
      "step": 7600
    },
    {
      "epoch": 3.9858490566037736,
      "grad_norm": 0.0464445985853672,
      "learning_rate": 0.0009702827734291198,
      "loss": 0.428,
      "num_input_tokens_seen": 4958568,
      "step": 7605
    },
    {
      "epoch": 3.988469601677149,
      "grad_norm": 0.06389942020177841,
      "learning_rate": 0.0009702050597877992,
      "loss": 0.4542,
      "num_input_tokens_seen": 4961608,
      "step": 7610
    },
    {
      "epoch": 3.991090146750524,
      "grad_norm": 0.08684727549552917,
      "learning_rate": 0.0009701272477849444,
      "loss": 0.4521,
      "num_input_tokens_seen": 4965288,
      "step": 7615
    },
    {
      "epoch": 3.9937106918238996,
      "grad_norm": 0.04708104208111763,
      "learning_rate": 0.0009700493374368327,
      "loss": 0.4247,
      "num_input_tokens_seen": 4968424,
      "step": 7620
    },
    {
      "epoch": 3.9963312368972748,
      "grad_norm": 0.055865395814180374,
      "learning_rate": 0.0009699713287597624,
      "loss": 0.4994,
      "num_input_tokens_seen": 4972200,
      "step": 7625
    },
    {
      "epoch": 3.99895178197065,
      "grad_norm": 0.05932066589593887,
      "learning_rate": 0.0009698932217700518,
      "loss": 0.515,
      "num_input_tokens_seen": 4975560,
      "step": 7630
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.4608750343322754,
      "eval_runtime": 13.6452,
      "eval_samples_per_second": 62.146,
      "eval_steps_per_second": 15.537,
      "num_input_tokens_seen": 4976128,
      "step": 7632
    },
    {
      "epoch": 4.001572327044025,
      "grad_norm": 0.04589091241359711,
      "learning_rate": 0.0009698150164840399,
      "loss": 0.4732,
      "num_input_tokens_seen": 4977984,
      "step": 7635
    },
    {
      "epoch": 4.0041928721174,
      "grad_norm": 0.07973460108041763,
      "learning_rate": 0.0009697367129180866,
      "loss": 0.3771,
      "num_input_tokens_seen": 4980896,
      "step": 7640
    },
    {
      "epoch": 4.006813417190775,
      "grad_norm": 0.08643874526023865,
      "learning_rate": 0.000969658311088572,
      "loss": 0.4093,
      "num_input_tokens_seen": 4984512,
      "step": 7645
    },
    {
      "epoch": 4.009433962264151,
      "grad_norm": 0.07449091970920563,
      "learning_rate": 0.0009695798110118969,
      "loss": 0.4575,
      "num_input_tokens_seen": 4988320,
      "step": 7650
    },
    {
      "epoch": 4.012054507337526,
      "grad_norm": 0.06445503979921341,
      "learning_rate": 0.0009695012127044824,
      "loss": 0.6352,
      "num_input_tokens_seen": 4991392,
      "step": 7655
    },
    {
      "epoch": 4.014675052410902,
      "grad_norm": 0.05741936340928078,
      "learning_rate": 0.0009694225161827707,
      "loss": 0.4822,
      "num_input_tokens_seen": 4995680,
      "step": 7660
    },
    {
      "epoch": 4.017295597484277,
      "grad_norm": 0.0753326490521431,
      "learning_rate": 0.0009693437214632241,
      "loss": 0.5754,
      "num_input_tokens_seen": 4999424,
      "step": 7665
    },
    {
      "epoch": 4.019916142557652,
      "grad_norm": 0.13220971822738647,
      "learning_rate": 0.0009692648285623256,
      "loss": 0.5263,
      "num_input_tokens_seen": 5002272,
      "step": 7670
    },
    {
      "epoch": 4.022536687631027,
      "grad_norm": 0.055223651230335236,
      "learning_rate": 0.0009691858374965784,
      "loss": 0.409,
      "num_input_tokens_seen": 5005888,
      "step": 7675
    },
    {
      "epoch": 4.0251572327044025,
      "grad_norm": 0.05475005879998207,
      "learning_rate": 0.0009691067482825069,
      "loss": 0.3479,
      "num_input_tokens_seen": 5009600,
      "step": 7680
    },
    {
      "epoch": 4.027777777777778,
      "grad_norm": 0.04111779108643532,
      "learning_rate": 0.0009690275609366554,
      "loss": 0.4309,
      "num_input_tokens_seen": 5013568,
      "step": 7685
    },
    {
      "epoch": 4.030398322851153,
      "grad_norm": 0.08893069624900818,
      "learning_rate": 0.0009689482754755891,
      "loss": 0.4858,
      "num_input_tokens_seen": 5016800,
      "step": 7690
    },
    {
      "epoch": 4.033018867924528,
      "grad_norm": 0.08924141526222229,
      "learning_rate": 0.0009688688919158938,
      "loss": 0.5316,
      "num_input_tokens_seen": 5020032,
      "step": 7695
    },
    {
      "epoch": 4.035639412997903,
      "grad_norm": 0.07703941315412521,
      "learning_rate": 0.0009687894102741754,
      "loss": 0.4836,
      "num_input_tokens_seen": 5023168,
      "step": 7700
    },
    {
      "epoch": 4.038259958071279,
      "grad_norm": 0.06003279983997345,
      "learning_rate": 0.0009687098305670605,
      "loss": 0.3718,
      "num_input_tokens_seen": 5027040,
      "step": 7705
    },
    {
      "epoch": 4.040880503144654,
      "grad_norm": 0.09782744944095612,
      "learning_rate": 0.0009686301528111964,
      "loss": 0.6156,
      "num_input_tokens_seen": 5030112,
      "step": 7710
    },
    {
      "epoch": 4.04350104821803,
      "grad_norm": 0.08671808242797852,
      "learning_rate": 0.0009685503770232507,
      "loss": 0.3784,
      "num_input_tokens_seen": 5033600,
      "step": 7715
    },
    {
      "epoch": 4.046121593291405,
      "grad_norm": 0.07864584773778915,
      "learning_rate": 0.0009684705032199117,
      "loss": 0.2836,
      "num_input_tokens_seen": 5036416,
      "step": 7720
    },
    {
      "epoch": 4.04874213836478,
      "grad_norm": 0.08171595633029938,
      "learning_rate": 0.0009683905314178881,
      "loss": 0.5517,
      "num_input_tokens_seen": 5039552,
      "step": 7725
    },
    {
      "epoch": 4.051362683438155,
      "grad_norm": 0.0913885086774826,
      "learning_rate": 0.000968310461633909,
      "loss": 0.4939,
      "num_input_tokens_seen": 5044032,
      "step": 7730
    },
    {
      "epoch": 4.05398322851153,
      "grad_norm": 0.060791756957769394,
      "learning_rate": 0.0009682302938847238,
      "loss": 0.3713,
      "num_input_tokens_seen": 5047232,
      "step": 7735
    },
    {
      "epoch": 4.056603773584905,
      "grad_norm": 0.054363347589969635,
      "learning_rate": 0.0009681500281871031,
      "loss": 0.5267,
      "num_input_tokens_seen": 5050624,
      "step": 7740
    },
    {
      "epoch": 4.059224318658281,
      "grad_norm": 0.14843913912773132,
      "learning_rate": 0.0009680696645578377,
      "loss": 0.4549,
      "num_input_tokens_seen": 5055008,
      "step": 7745
    },
    {
      "epoch": 4.061844863731656,
      "grad_norm": 0.09010645002126694,
      "learning_rate": 0.0009679892030137382,
      "loss": 0.4513,
      "num_input_tokens_seen": 5058240,
      "step": 7750
    },
    {
      "epoch": 4.064465408805032,
      "grad_norm": 0.09298838675022125,
      "learning_rate": 0.0009679086435716368,
      "loss": 0.3369,
      "num_input_tokens_seen": 5060864,
      "step": 7755
    },
    {
      "epoch": 4.067085953878407,
      "grad_norm": 0.06237221509218216,
      "learning_rate": 0.0009678279862483852,
      "loss": 0.5655,
      "num_input_tokens_seen": 5064384,
      "step": 7760
    },
    {
      "epoch": 4.069706498951782,
      "grad_norm": 0.047961074858903885,
      "learning_rate": 0.0009677472310608561,
      "loss": 0.4359,
      "num_input_tokens_seen": 5068096,
      "step": 7765
    },
    {
      "epoch": 4.072327044025157,
      "grad_norm": 0.0880693644285202,
      "learning_rate": 0.0009676663780259427,
      "loss": 0.524,
      "num_input_tokens_seen": 5070944,
      "step": 7770
    },
    {
      "epoch": 4.0749475890985325,
      "grad_norm": 0.0672934502363205,
      "learning_rate": 0.0009675854271605583,
      "loss": 0.5017,
      "num_input_tokens_seen": 5074624,
      "step": 7775
    },
    {
      "epoch": 4.077568134171908,
      "grad_norm": 0.10357329249382019,
      "learning_rate": 0.0009675043784816371,
      "loss": 0.5619,
      "num_input_tokens_seen": 5077216,
      "step": 7780
    },
    {
      "epoch": 4.080188679245283,
      "grad_norm": 0.07430513948202133,
      "learning_rate": 0.0009674232320061336,
      "loss": 0.5168,
      "num_input_tokens_seen": 5079872,
      "step": 7785
    },
    {
      "epoch": 4.082809224318658,
      "grad_norm": 0.11177793145179749,
      "learning_rate": 0.0009673419877510226,
      "loss": 0.5498,
      "num_input_tokens_seen": 5082592,
      "step": 7790
    },
    {
      "epoch": 4.085429769392033,
      "grad_norm": 0.0563947819173336,
      "learning_rate": 0.0009672606457332994,
      "loss": 0.4107,
      "num_input_tokens_seen": 5086144,
      "step": 7795
    },
    {
      "epoch": 4.088050314465409,
      "grad_norm": 0.07665712386369705,
      "learning_rate": 0.0009671792059699798,
      "loss": 0.3619,
      "num_input_tokens_seen": 5089696,
      "step": 7800
    },
    {
      "epoch": 4.090670859538784,
      "grad_norm": 0.06125263124704361,
      "learning_rate": 0.0009670976684781003,
      "loss": 0.4513,
      "num_input_tokens_seen": 5092672,
      "step": 7805
    },
    {
      "epoch": 4.09329140461216,
      "grad_norm": 0.05788576602935791,
      "learning_rate": 0.0009670160332747174,
      "loss": 0.4722,
      "num_input_tokens_seen": 5094848,
      "step": 7810
    },
    {
      "epoch": 4.095911949685535,
      "grad_norm": 0.07555203139781952,
      "learning_rate": 0.0009669343003769085,
      "loss": 0.3624,
      "num_input_tokens_seen": 5097408,
      "step": 7815
    },
    {
      "epoch": 4.09853249475891,
      "grad_norm": 0.08549541234970093,
      "learning_rate": 0.0009668524698017709,
      "loss": 0.4902,
      "num_input_tokens_seen": 5099872,
      "step": 7820
    },
    {
      "epoch": 4.101153039832285,
      "grad_norm": 0.09978635609149933,
      "learning_rate": 0.0009667705415664227,
      "loss": 0.5524,
      "num_input_tokens_seen": 5103392,
      "step": 7825
    },
    {
      "epoch": 4.10377358490566,
      "grad_norm": 0.11681987345218658,
      "learning_rate": 0.0009666885156880026,
      "loss": 0.4388,
      "num_input_tokens_seen": 5105696,
      "step": 7830
    },
    {
      "epoch": 4.106394129979035,
      "grad_norm": 0.11756402999162674,
      "learning_rate": 0.0009666063921836692,
      "loss": 0.5564,
      "num_input_tokens_seen": 5109312,
      "step": 7835
    },
    {
      "epoch": 4.109014675052411,
      "grad_norm": 0.1371457278728485,
      "learning_rate": 0.0009665241710706019,
      "loss": 0.5581,
      "num_input_tokens_seen": 5111840,
      "step": 7840
    },
    {
      "epoch": 4.111635220125786,
      "grad_norm": 0.07569771260023117,
      "learning_rate": 0.0009664418523660003,
      "loss": 0.5677,
      "num_input_tokens_seen": 5114752,
      "step": 7845
    },
    {
      "epoch": 4.114255765199162,
      "grad_norm": 0.1179710328578949,
      "learning_rate": 0.0009663594360870847,
      "loss": 0.5428,
      "num_input_tokens_seen": 5117760,
      "step": 7850
    },
    {
      "epoch": 4.116876310272537,
      "grad_norm": 0.03828177601099014,
      "learning_rate": 0.0009662769222510955,
      "loss": 0.5209,
      "num_input_tokens_seen": 5122176,
      "step": 7855
    },
    {
      "epoch": 4.119496855345912,
      "grad_norm": 0.08016887307167053,
      "learning_rate": 0.0009661943108752939,
      "loss": 0.8199,
      "num_input_tokens_seen": 5124256,
      "step": 7860
    },
    {
      "epoch": 4.122117400419287,
      "grad_norm": 0.13066495954990387,
      "learning_rate": 0.0009661116019769609,
      "loss": 0.5559,
      "num_input_tokens_seen": 5126912,
      "step": 7865
    },
    {
      "epoch": 4.1247379454926625,
      "grad_norm": 0.051645416766405106,
      "learning_rate": 0.0009660287955733986,
      "loss": 0.5484,
      "num_input_tokens_seen": 5130144,
      "step": 7870
    },
    {
      "epoch": 4.127358490566038,
      "grad_norm": 0.06686601042747498,
      "learning_rate": 0.0009659458916819289,
      "loss": 0.4658,
      "num_input_tokens_seen": 5133440,
      "step": 7875
    },
    {
      "epoch": 4.129979035639413,
      "grad_norm": 0.11832496523857117,
      "learning_rate": 0.0009658628903198945,
      "loss": 0.552,
      "num_input_tokens_seen": 5140064,
      "step": 7880
    },
    {
      "epoch": 4.132599580712788,
      "grad_norm": 0.06519630551338196,
      "learning_rate": 0.0009657797915046583,
      "loss": 0.5201,
      "num_input_tokens_seen": 5142720,
      "step": 7885
    },
    {
      "epoch": 4.135220125786163,
      "grad_norm": 0.06570505350828171,
      "learning_rate": 0.0009656965952536036,
      "loss": 0.4648,
      "num_input_tokens_seen": 5145632,
      "step": 7890
    },
    {
      "epoch": 4.137840670859539,
      "grad_norm": 0.07866798341274261,
      "learning_rate": 0.0009656133015841342,
      "loss": 0.4251,
      "num_input_tokens_seen": 5148224,
      "step": 7895
    },
    {
      "epoch": 4.140461215932914,
      "grad_norm": 0.07774636149406433,
      "learning_rate": 0.000965529910513674,
      "loss": 0.5702,
      "num_input_tokens_seen": 5152096,
      "step": 7900
    },
    {
      "epoch": 4.1430817610062896,
      "grad_norm": 0.032706908881664276,
      "learning_rate": 0.0009654464220596676,
      "loss": 0.5996,
      "num_input_tokens_seen": 5155968,
      "step": 7905
    },
    {
      "epoch": 4.145702306079665,
      "grad_norm": 0.04955239221453667,
      "learning_rate": 0.0009653628362395799,
      "loss": 0.4524,
      "num_input_tokens_seen": 5159168,
      "step": 7910
    },
    {
      "epoch": 4.14832285115304,
      "grad_norm": 0.04612930491566658,
      "learning_rate": 0.0009652791530708958,
      "loss": 0.3846,
      "num_input_tokens_seen": 5162816,
      "step": 7915
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 0.0909479558467865,
      "learning_rate": 0.0009651953725711212,
      "loss": 0.5538,
      "num_input_tokens_seen": 5167232,
      "step": 7920
    },
    {
      "epoch": 4.15356394129979,
      "grad_norm": 0.11229131370782852,
      "learning_rate": 0.0009651114947577818,
      "loss": 0.5103,
      "num_input_tokens_seen": 5170560,
      "step": 7925
    },
    {
      "epoch": 4.156184486373165,
      "grad_norm": 0.10074679553508759,
      "learning_rate": 0.0009650275196484239,
      "loss": 0.4012,
      "num_input_tokens_seen": 5173184,
      "step": 7930
    },
    {
      "epoch": 4.158805031446541,
      "grad_norm": 0.048728905618190765,
      "learning_rate": 0.0009649434472606144,
      "loss": 0.439,
      "num_input_tokens_seen": 5176288,
      "step": 7935
    },
    {
      "epoch": 4.161425576519916,
      "grad_norm": 0.07408004999160767,
      "learning_rate": 0.00096485927761194,
      "loss": 0.4282,
      "num_input_tokens_seen": 5181216,
      "step": 7940
    },
    {
      "epoch": 4.164046121593292,
      "grad_norm": 0.08759824931621552,
      "learning_rate": 0.0009647750107200082,
      "loss": 0.3656,
      "num_input_tokens_seen": 5184256,
      "step": 7945
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 0.04557780548930168,
      "learning_rate": 0.0009646906466024465,
      "loss": 0.3605,
      "num_input_tokens_seen": 5187616,
      "step": 7950
    },
    {
      "epoch": 4.169287211740042,
      "grad_norm": 0.0751553475856781,
      "learning_rate": 0.000964606185276903,
      "loss": 0.6301,
      "num_input_tokens_seen": 5189952,
      "step": 7955
    },
    {
      "epoch": 4.171907756813417,
      "grad_norm": 0.1527765542268753,
      "learning_rate": 0.0009645216267610461,
      "loss": 0.5209,
      "num_input_tokens_seen": 5192608,
      "step": 7960
    },
    {
      "epoch": 4.1745283018867925,
      "grad_norm": 0.06444250792264938,
      "learning_rate": 0.0009644369710725644,
      "loss": 0.4311,
      "num_input_tokens_seen": 5196256,
      "step": 7965
    },
    {
      "epoch": 4.177148846960168,
      "grad_norm": 0.05861902981996536,
      "learning_rate": 0.0009643522182291669,
      "loss": 0.4506,
      "num_input_tokens_seen": 5199840,
      "step": 7970
    },
    {
      "epoch": 4.179769392033543,
      "grad_norm": 0.0689849704504013,
      "learning_rate": 0.000964267368248583,
      "loss": 0.3182,
      "num_input_tokens_seen": 5202688,
      "step": 7975
    },
    {
      "epoch": 4.182389937106918,
      "grad_norm": 0.08227358758449554,
      "learning_rate": 0.0009641824211485623,
      "loss": 0.6421,
      "num_input_tokens_seen": 5205248,
      "step": 7980
    },
    {
      "epoch": 4.185010482180293,
      "grad_norm": 0.06936078518629074,
      "learning_rate": 0.0009640973769468747,
      "loss": 0.4044,
      "num_input_tokens_seen": 5207648,
      "step": 7985
    },
    {
      "epoch": 4.187631027253669,
      "grad_norm": 0.08380523324012756,
      "learning_rate": 0.0009640122356613105,
      "loss": 0.4176,
      "num_input_tokens_seen": 5210272,
      "step": 7990
    },
    {
      "epoch": 4.190251572327044,
      "grad_norm": 0.052739523351192474,
      "learning_rate": 0.0009639269973096805,
      "loss": 0.5846,
      "num_input_tokens_seen": 5214848,
      "step": 7995
    },
    {
      "epoch": 4.1928721174004195,
      "grad_norm": 0.06927201896905899,
      "learning_rate": 0.0009638416619098154,
      "loss": 0.4246,
      "num_input_tokens_seen": 5217888,
      "step": 8000
    },
    {
      "epoch": 4.195492662473795,
      "grad_norm": 0.06380534172058105,
      "learning_rate": 0.0009637562294795663,
      "loss": 0.5191,
      "num_input_tokens_seen": 5221952,
      "step": 8005
    },
    {
      "epoch": 4.19811320754717,
      "grad_norm": 0.09169501811265945,
      "learning_rate": 0.0009636707000368049,
      "loss": 0.5484,
      "num_input_tokens_seen": 5225120,
      "step": 8010
    },
    {
      "epoch": 4.200733752620545,
      "grad_norm": 0.12928567826747894,
      "learning_rate": 0.000963585073599423,
      "loss": 0.5317,
      "num_input_tokens_seen": 5228448,
      "step": 8015
    },
    {
      "epoch": 4.20335429769392,
      "grad_norm": 0.12298668175935745,
      "learning_rate": 0.0009634993501853323,
      "loss": 0.5416,
      "num_input_tokens_seen": 5230976,
      "step": 8020
    },
    {
      "epoch": 4.205974842767295,
      "grad_norm": 0.056666381657123566,
      "learning_rate": 0.0009634135298124656,
      "loss": 0.5286,
      "num_input_tokens_seen": 5234112,
      "step": 8025
    },
    {
      "epoch": 4.2085953878406706,
      "grad_norm": 0.05989782512187958,
      "learning_rate": 0.0009633276124987752,
      "loss": 0.6072,
      "num_input_tokens_seen": 5238592,
      "step": 8030
    },
    {
      "epoch": 4.211215932914046,
      "grad_norm": 0.06317538768053055,
      "learning_rate": 0.0009632415982622342,
      "loss": 0.454,
      "num_input_tokens_seen": 5241568,
      "step": 8035
    },
    {
      "epoch": 4.213836477987422,
      "grad_norm": 0.07959946244955063,
      "learning_rate": 0.0009631554871208359,
      "loss": 0.379,
      "num_input_tokens_seen": 5244704,
      "step": 8040
    },
    {
      "epoch": 4.216457023060797,
      "grad_norm": 0.05562129244208336,
      "learning_rate": 0.0009630692790925936,
      "loss": 0.3711,
      "num_input_tokens_seen": 5248000,
      "step": 8045
    },
    {
      "epoch": 4.219077568134172,
      "grad_norm": 0.044819533824920654,
      "learning_rate": 0.0009629829741955411,
      "loss": 0.3997,
      "num_input_tokens_seen": 5252096,
      "step": 8050
    },
    {
      "epoch": 4.221698113207547,
      "grad_norm": 0.04882779344916344,
      "learning_rate": 0.0009628965724477325,
      "loss": 0.5583,
      "num_input_tokens_seen": 5255360,
      "step": 8055
    },
    {
      "epoch": 4.2243186582809225,
      "grad_norm": 0.03679252415895462,
      "learning_rate": 0.0009628100738672419,
      "loss": 0.6131,
      "num_input_tokens_seen": 5259712,
      "step": 8060
    },
    {
      "epoch": 4.226939203354298,
      "grad_norm": 0.11150677502155304,
      "learning_rate": 0.0009627234784721637,
      "loss": 0.4479,
      "num_input_tokens_seen": 5262912,
      "step": 8065
    },
    {
      "epoch": 4.229559748427673,
      "grad_norm": 0.0990724116563797,
      "learning_rate": 0.0009626367862806129,
      "loss": 0.4879,
      "num_input_tokens_seen": 5265664,
      "step": 8070
    },
    {
      "epoch": 4.232180293501048,
      "grad_norm": 0.05672990903258324,
      "learning_rate": 0.0009625499973107246,
      "loss": 0.4378,
      "num_input_tokens_seen": 5268928,
      "step": 8075
    },
    {
      "epoch": 4.234800838574423,
      "grad_norm": 0.062334924936294556,
      "learning_rate": 0.0009624631115806537,
      "loss": 0.5138,
      "num_input_tokens_seen": 5272032,
      "step": 8080
    },
    {
      "epoch": 4.237421383647799,
      "grad_norm": 0.06879336386919022,
      "learning_rate": 0.0009623761291085761,
      "loss": 0.4155,
      "num_input_tokens_seen": 5275520,
      "step": 8085
    },
    {
      "epoch": 4.240041928721174,
      "grad_norm": 0.07819204777479172,
      "learning_rate": 0.0009622890499126873,
      "loss": 0.4939,
      "num_input_tokens_seen": 5278528,
      "step": 8090
    },
    {
      "epoch": 4.2426624737945495,
      "grad_norm": 0.16830633580684662,
      "learning_rate": 0.0009622018740112032,
      "loss": 0.647,
      "num_input_tokens_seen": 5281536,
      "step": 8095
    },
    {
      "epoch": 4.245283018867925,
      "grad_norm": 0.07666724175214767,
      "learning_rate": 0.0009621146014223603,
      "loss": 0.464,
      "num_input_tokens_seen": 5287168,
      "step": 8100
    },
    {
      "epoch": 4.2479035639413,
      "grad_norm": 0.07154251635074615,
      "learning_rate": 0.0009620272321644148,
      "loss": 0.4096,
      "num_input_tokens_seen": 5291008,
      "step": 8105
    },
    {
      "epoch": 4.250524109014675,
      "grad_norm": 0.07611490041017532,
      "learning_rate": 0.0009619397662556434,
      "loss": 0.3822,
      "num_input_tokens_seen": 5294144,
      "step": 8110
    },
    {
      "epoch": 4.25314465408805,
      "grad_norm": 0.09397440403699875,
      "learning_rate": 0.000961852203714343,
      "loss": 0.5843,
      "num_input_tokens_seen": 5297184,
      "step": 8115
    },
    {
      "epoch": 4.255765199161425,
      "grad_norm": 0.05423390492796898,
      "learning_rate": 0.0009617645445588307,
      "loss": 0.4539,
      "num_input_tokens_seen": 5300480,
      "step": 8120
    },
    {
      "epoch": 4.2583857442348005,
      "grad_norm": 0.0524229072034359,
      "learning_rate": 0.0009616767888074438,
      "loss": 0.486,
      "num_input_tokens_seen": 5303936,
      "step": 8125
    },
    {
      "epoch": 4.261006289308176,
      "grad_norm": 0.046444330364465714,
      "learning_rate": 0.0009615889364785397,
      "loss": 0.4166,
      "num_input_tokens_seen": 5306816,
      "step": 8130
    },
    {
      "epoch": 4.263626834381552,
      "grad_norm": 0.10198123753070831,
      "learning_rate": 0.000961500987590496,
      "loss": 0.4514,
      "num_input_tokens_seen": 5309792,
      "step": 8135
    },
    {
      "epoch": 4.266247379454927,
      "grad_norm": 0.05896982550621033,
      "learning_rate": 0.0009614129421617111,
      "loss": 0.5788,
      "num_input_tokens_seen": 5313056,
      "step": 8140
    },
    {
      "epoch": 4.268867924528302,
      "grad_norm": 0.0711526870727539,
      "learning_rate": 0.0009613248002106027,
      "loss": 0.3641,
      "num_input_tokens_seen": 5315488,
      "step": 8145
    },
    {
      "epoch": 4.271488469601677,
      "grad_norm": 0.09657667577266693,
      "learning_rate": 0.000961236561755609,
      "loss": 0.4101,
      "num_input_tokens_seen": 5318272,
      "step": 8150
    },
    {
      "epoch": 4.274109014675052,
      "grad_norm": 0.07299318164587021,
      "learning_rate": 0.0009611482268151888,
      "loss": 0.4947,
      "num_input_tokens_seen": 5321408,
      "step": 8155
    },
    {
      "epoch": 4.276729559748428,
      "grad_norm": 0.061610568314790726,
      "learning_rate": 0.0009610597954078206,
      "loss": 0.4156,
      "num_input_tokens_seen": 5325696,
      "step": 8160
    },
    {
      "epoch": 4.279350104821803,
      "grad_norm": 0.1563584804534912,
      "learning_rate": 0.0009609712675520031,
      "loss": 0.5661,
      "num_input_tokens_seen": 5328832,
      "step": 8165
    },
    {
      "epoch": 4.281970649895178,
      "grad_norm": 0.07999323308467865,
      "learning_rate": 0.0009608826432662556,
      "loss": 0.4966,
      "num_input_tokens_seen": 5332448,
      "step": 8170
    },
    {
      "epoch": 4.284591194968553,
      "grad_norm": 0.055772967636585236,
      "learning_rate": 0.0009607939225691172,
      "loss": 0.5179,
      "num_input_tokens_seen": 5335520,
      "step": 8175
    },
    {
      "epoch": 4.287211740041929,
      "grad_norm": 0.05767921358346939,
      "learning_rate": 0.0009607051054791472,
      "loss": 0.3562,
      "num_input_tokens_seen": 5339328,
      "step": 8180
    },
    {
      "epoch": 4.289832285115304,
      "grad_norm": 0.06435496360063553,
      "learning_rate": 0.000960616192014925,
      "loss": 0.4998,
      "num_input_tokens_seen": 5342368,
      "step": 8185
    },
    {
      "epoch": 4.2924528301886795,
      "grad_norm": 0.06685090065002441,
      "learning_rate": 0.0009605271821950506,
      "loss": 0.608,
      "num_input_tokens_seen": 5345632,
      "step": 8190
    },
    {
      "epoch": 4.295073375262055,
      "grad_norm": 0.0885709673166275,
      "learning_rate": 0.0009604380760381434,
      "loss": 0.5878,
      "num_input_tokens_seen": 5348608,
      "step": 8195
    },
    {
      "epoch": 4.29769392033543,
      "grad_norm": 0.08460699766874313,
      "learning_rate": 0.0009603488735628439,
      "loss": 0.406,
      "num_input_tokens_seen": 5351776,
      "step": 8200
    },
    {
      "epoch": 4.300314465408805,
      "grad_norm": 0.052529074251651764,
      "learning_rate": 0.0009602595747878118,
      "loss": 0.6108,
      "num_input_tokens_seen": 5354880,
      "step": 8205
    },
    {
      "epoch": 4.30293501048218,
      "grad_norm": 0.11417509615421295,
      "learning_rate": 0.0009601701797317278,
      "loss": 0.4517,
      "num_input_tokens_seen": 5357824,
      "step": 8210
    },
    {
      "epoch": 4.305555555555555,
      "grad_norm": 0.09454672038555145,
      "learning_rate": 0.0009600806884132917,
      "loss": 0.441,
      "num_input_tokens_seen": 5360512,
      "step": 8215
    },
    {
      "epoch": 4.3081761006289305,
      "grad_norm": 0.06514990329742432,
      "learning_rate": 0.0009599911008512248,
      "loss": 0.3925,
      "num_input_tokens_seen": 5364064,
      "step": 8220
    },
    {
      "epoch": 4.310796645702306,
      "grad_norm": 0.0857241153717041,
      "learning_rate": 0.0009599014170642674,
      "loss": 0.5446,
      "num_input_tokens_seen": 5367200,
      "step": 8225
    },
    {
      "epoch": 4.313417190775682,
      "grad_norm": 0.048395786434412,
      "learning_rate": 0.0009598116370711805,
      "loss": 0.5148,
      "num_input_tokens_seen": 5370272,
      "step": 8230
    },
    {
      "epoch": 4.316037735849057,
      "grad_norm": 0.10048344731330872,
      "learning_rate": 0.0009597217608907447,
      "loss": 0.4852,
      "num_input_tokens_seen": 5372896,
      "step": 8235
    },
    {
      "epoch": 4.318658280922432,
      "grad_norm": 0.060479141771793365,
      "learning_rate": 0.0009596317885417614,
      "loss": 0.5564,
      "num_input_tokens_seen": 5376800,
      "step": 8240
    },
    {
      "epoch": 4.321278825995807,
      "grad_norm": 0.06282056123018265,
      "learning_rate": 0.0009595417200430516,
      "loss": 0.6302,
      "num_input_tokens_seen": 5380288,
      "step": 8245
    },
    {
      "epoch": 4.323899371069182,
      "grad_norm": 0.09769164770841599,
      "learning_rate": 0.0009594515554134568,
      "loss": 0.4361,
      "num_input_tokens_seen": 5383072,
      "step": 8250
    },
    {
      "epoch": 4.326519916142558,
      "grad_norm": 0.0723399966955185,
      "learning_rate": 0.0009593612946718384,
      "loss": 0.6242,
      "num_input_tokens_seen": 5386400,
      "step": 8255
    },
    {
      "epoch": 4.329140461215933,
      "grad_norm": 0.06964773684740067,
      "learning_rate": 0.0009592709378370778,
      "loss": 0.5854,
      "num_input_tokens_seen": 5388832,
      "step": 8260
    },
    {
      "epoch": 4.331761006289308,
      "grad_norm": 0.13015946745872498,
      "learning_rate": 0.0009591804849280766,
      "loss": 0.4671,
      "num_input_tokens_seen": 5391968,
      "step": 8265
    },
    {
      "epoch": 4.334381551362683,
      "grad_norm": 0.06619153171777725,
      "learning_rate": 0.0009590899359637564,
      "loss": 0.3859,
      "num_input_tokens_seen": 5398624,
      "step": 8270
    },
    {
      "epoch": 4.337002096436059,
      "grad_norm": 0.04791460558772087,
      "learning_rate": 0.0009589992909630594,
      "loss": 0.4945,
      "num_input_tokens_seen": 5401824,
      "step": 8275
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 0.06773632764816284,
      "learning_rate": 0.0009589085499449471,
      "loss": 0.4005,
      "num_input_tokens_seen": 5404928,
      "step": 8280
    },
    {
      "epoch": 4.3422431865828095,
      "grad_norm": 0.05192112550139427,
      "learning_rate": 0.0009588177129284017,
      "loss": 0.4341,
      "num_input_tokens_seen": 5408320,
      "step": 8285
    },
    {
      "epoch": 4.344863731656185,
      "grad_norm": 0.07379284501075745,
      "learning_rate": 0.0009587267799324253,
      "loss": 0.5305,
      "num_input_tokens_seen": 5410752,
      "step": 8290
    },
    {
      "epoch": 4.34748427672956,
      "grad_norm": 0.07443712651729584,
      "learning_rate": 0.0009586357509760399,
      "loss": 0.3988,
      "num_input_tokens_seen": 5413312,
      "step": 8295
    },
    {
      "epoch": 4.350104821802935,
      "grad_norm": 0.05555155500769615,
      "learning_rate": 0.0009585446260782878,
      "loss": 0.5219,
      "num_input_tokens_seen": 5416992,
      "step": 8300
    },
    {
      "epoch": 4.35272536687631,
      "grad_norm": 0.043023306876420975,
      "learning_rate": 0.0009584534052582313,
      "loss": 0.3567,
      "num_input_tokens_seen": 5420672,
      "step": 8305
    },
    {
      "epoch": 4.355345911949685,
      "grad_norm": 0.1572924107313156,
      "learning_rate": 0.0009583620885349527,
      "loss": 0.4016,
      "num_input_tokens_seen": 5423360,
      "step": 8310
    },
    {
      "epoch": 4.3579664570230605,
      "grad_norm": 0.05776500329375267,
      "learning_rate": 0.0009582706759275546,
      "loss": 0.4424,
      "num_input_tokens_seen": 5426336,
      "step": 8315
    },
    {
      "epoch": 4.360587002096436,
      "grad_norm": 0.1425010859966278,
      "learning_rate": 0.0009581791674551592,
      "loss": 0.5122,
      "num_input_tokens_seen": 5429632,
      "step": 8320
    },
    {
      "epoch": 4.363207547169811,
      "grad_norm": 0.06150255724787712,
      "learning_rate": 0.000958087563136909,
      "loss": 0.4458,
      "num_input_tokens_seen": 5433312,
      "step": 8325
    },
    {
      "epoch": 4.365828092243187,
      "grad_norm": 0.056721031665802,
      "learning_rate": 0.000957995862991967,
      "loss": 0.5215,
      "num_input_tokens_seen": 5437248,
      "step": 8330
    },
    {
      "epoch": 4.368448637316562,
      "grad_norm": 0.06034320592880249,
      "learning_rate": 0.0009579040670395154,
      "loss": 0.3906,
      "num_input_tokens_seen": 5440384,
      "step": 8335
    },
    {
      "epoch": 4.371069182389937,
      "grad_norm": 0.10864824056625366,
      "learning_rate": 0.000957812175298757,
      "loss": 0.5543,
      "num_input_tokens_seen": 5444000,
      "step": 8340
    },
    {
      "epoch": 4.373689727463312,
      "grad_norm": 0.11304819583892822,
      "learning_rate": 0.0009577201877889145,
      "loss": 0.4722,
      "num_input_tokens_seen": 5447936,
      "step": 8345
    },
    {
      "epoch": 4.376310272536688,
      "grad_norm": 0.08155078440904617,
      "learning_rate": 0.0009576281045292308,
      "loss": 0.5375,
      "num_input_tokens_seen": 5451488,
      "step": 8350
    },
    {
      "epoch": 4.378930817610063,
      "grad_norm": 0.07544378936290741,
      "learning_rate": 0.0009575359255389686,
      "loss": 0.4727,
      "num_input_tokens_seen": 5455520,
      "step": 8355
    },
    {
      "epoch": 4.381551362683438,
      "grad_norm": 0.07916082441806793,
      "learning_rate": 0.0009574436508374104,
      "loss": 0.6311,
      "num_input_tokens_seen": 5458560,
      "step": 8360
    },
    {
      "epoch": 4.384171907756813,
      "grad_norm": 0.06933601200580597,
      "learning_rate": 0.0009573512804438594,
      "loss": 0.4555,
      "num_input_tokens_seen": 5461600,
      "step": 8365
    },
    {
      "epoch": 4.386792452830189,
      "grad_norm": 0.0639994665980339,
      "learning_rate": 0.0009572588143776381,
      "loss": 0.4022,
      "num_input_tokens_seen": 5465536,
      "step": 8370
    },
    {
      "epoch": 4.389412997903564,
      "grad_norm": 0.07995772361755371,
      "learning_rate": 0.0009571662526580897,
      "loss": 0.4393,
      "num_input_tokens_seen": 5468416,
      "step": 8375
    },
    {
      "epoch": 4.3920335429769395,
      "grad_norm": 0.1395753026008606,
      "learning_rate": 0.0009570735953045768,
      "loss": 0.5431,
      "num_input_tokens_seen": 5472000,
      "step": 8380
    },
    {
      "epoch": 4.394654088050315,
      "grad_norm": 0.06370563805103302,
      "learning_rate": 0.0009569808423364823,
      "loss": 0.3958,
      "num_input_tokens_seen": 5475168,
      "step": 8385
    },
    {
      "epoch": 4.39727463312369,
      "grad_norm": 0.10865437984466553,
      "learning_rate": 0.0009568879937732091,
      "loss": 0.4136,
      "num_input_tokens_seen": 5477760,
      "step": 8390
    },
    {
      "epoch": 4.399895178197065,
      "grad_norm": 0.07509278506040573,
      "learning_rate": 0.0009567950496341802,
      "loss": 0.4071,
      "num_input_tokens_seen": 5480544,
      "step": 8395
    },
    {
      "epoch": 4.40251572327044,
      "grad_norm": 0.06291668117046356,
      "learning_rate": 0.0009567020099388382,
      "loss": 0.3677,
      "num_input_tokens_seen": 5483776,
      "step": 8400
    },
    {
      "epoch": 4.405136268343815,
      "grad_norm": 0.059891387820243835,
      "learning_rate": 0.0009566088747066459,
      "loss": 0.436,
      "num_input_tokens_seen": 5485888,
      "step": 8405
    },
    {
      "epoch": 4.4077568134171905,
      "grad_norm": 0.06158573925495148,
      "learning_rate": 0.0009565156439570866,
      "loss": 0.6117,
      "num_input_tokens_seen": 5489056,
      "step": 8410
    },
    {
      "epoch": 4.410377358490566,
      "grad_norm": 0.05978408828377724,
      "learning_rate": 0.0009564223177096625,
      "loss": 0.5614,
      "num_input_tokens_seen": 5492224,
      "step": 8415
    },
    {
      "epoch": 4.412997903563941,
      "grad_norm": 0.07160869240760803,
      "learning_rate": 0.0009563288959838969,
      "loss": 0.4064,
      "num_input_tokens_seen": 5495872,
      "step": 8420
    },
    {
      "epoch": 4.415618448637317,
      "grad_norm": 0.03583051264286041,
      "learning_rate": 0.0009562353787993321,
      "loss": 0.365,
      "num_input_tokens_seen": 5499488,
      "step": 8425
    },
    {
      "epoch": 4.418238993710692,
      "grad_norm": 0.055510856211185455,
      "learning_rate": 0.0009561417661755312,
      "loss": 0.5166,
      "num_input_tokens_seen": 5503104,
      "step": 8430
    },
    {
      "epoch": 4.420859538784067,
      "grad_norm": 0.06359875202178955,
      "learning_rate": 0.0009560480581320768,
      "loss": 0.4738,
      "num_input_tokens_seen": 5506912,
      "step": 8435
    },
    {
      "epoch": 4.423480083857442,
      "grad_norm": 0.0880817249417305,
      "learning_rate": 0.0009559542546885714,
      "loss": 0.5446,
      "num_input_tokens_seen": 5509312,
      "step": 8440
    },
    {
      "epoch": 4.426100628930818,
      "grad_norm": 0.09183906763792038,
      "learning_rate": 0.0009558603558646378,
      "loss": 0.3778,
      "num_input_tokens_seen": 5512096,
      "step": 8445
    },
    {
      "epoch": 4.428721174004193,
      "grad_norm": 0.1407630890607834,
      "learning_rate": 0.0009557663616799185,
      "loss": 0.5273,
      "num_input_tokens_seen": 5515168,
      "step": 8450
    },
    {
      "epoch": 4.431341719077568,
      "grad_norm": 0.07404100894927979,
      "learning_rate": 0.0009556722721540759,
      "loss": 0.4961,
      "num_input_tokens_seen": 5518464,
      "step": 8455
    },
    {
      "epoch": 4.433962264150943,
      "grad_norm": 0.04491398483514786,
      "learning_rate": 0.0009555780873067927,
      "loss": 0.4526,
      "num_input_tokens_seen": 5521344,
      "step": 8460
    },
    {
      "epoch": 4.436582809224318,
      "grad_norm": 0.09034038335084915,
      "learning_rate": 0.000955483807157771,
      "loss": 0.5036,
      "num_input_tokens_seen": 5524416,
      "step": 8465
    },
    {
      "epoch": 4.439203354297694,
      "grad_norm": 0.0685165598988533,
      "learning_rate": 0.0009553894317267333,
      "loss": 0.5626,
      "num_input_tokens_seen": 5527488,
      "step": 8470
    },
    {
      "epoch": 4.4418238993710695,
      "grad_norm": 0.08429187536239624,
      "learning_rate": 0.0009552949610334219,
      "loss": 0.4634,
      "num_input_tokens_seen": 5530144,
      "step": 8475
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.06495928764343262,
      "learning_rate": 0.000955200395097599,
      "loss": 0.5941,
      "num_input_tokens_seen": 5533536,
      "step": 8480
    },
    {
      "epoch": 4.44706498951782,
      "grad_norm": 0.06275220960378647,
      "learning_rate": 0.0009551057339390464,
      "loss": 0.504,
      "num_input_tokens_seen": 5536032,
      "step": 8485
    },
    {
      "epoch": 4.449685534591195,
      "grad_norm": 0.047421645373106,
      "learning_rate": 0.0009550109775775666,
      "loss": 0.4364,
      "num_input_tokens_seen": 5539552,
      "step": 8490
    },
    {
      "epoch": 4.45230607966457,
      "grad_norm": 0.07108405977487564,
      "learning_rate": 0.0009549161260329811,
      "loss": 0.5009,
      "num_input_tokens_seen": 5542528,
      "step": 8495
    },
    {
      "epoch": 4.454926624737945,
      "grad_norm": 0.06535464525222778,
      "learning_rate": 0.0009548211793251322,
      "loss": 0.5861,
      "num_input_tokens_seen": 5545408,
      "step": 8500
    },
    {
      "epoch": 4.4575471698113205,
      "grad_norm": 0.07236534357070923,
      "learning_rate": 0.0009547261374738814,
      "loss": 0.3781,
      "num_input_tokens_seen": 5549120,
      "step": 8505
    },
    {
      "epoch": 4.460167714884696,
      "grad_norm": 0.08986755460500717,
      "learning_rate": 0.0009546310004991105,
      "loss": 0.5381,
      "num_input_tokens_seen": 5552352,
      "step": 8510
    },
    {
      "epoch": 4.462788259958071,
      "grad_norm": 0.04723511263728142,
      "learning_rate": 0.000954535768420721,
      "loss": 0.3616,
      "num_input_tokens_seen": 5555296,
      "step": 8515
    },
    {
      "epoch": 4.465408805031447,
      "grad_norm": 0.087918221950531,
      "learning_rate": 0.0009544404412586343,
      "loss": 0.3445,
      "num_input_tokens_seen": 5558720,
      "step": 8520
    },
    {
      "epoch": 4.468029350104822,
      "grad_norm": 0.070049948990345,
      "learning_rate": 0.0009543450190327917,
      "loss": 0.4878,
      "num_input_tokens_seen": 5562176,
      "step": 8525
    },
    {
      "epoch": 4.470649895178197,
      "grad_norm": 0.04717167094349861,
      "learning_rate": 0.0009542495017631547,
      "loss": 0.6273,
      "num_input_tokens_seen": 5565408,
      "step": 8530
    },
    {
      "epoch": 4.473270440251572,
      "grad_norm": 0.05601263418793678,
      "learning_rate": 0.0009541538894697043,
      "loss": 0.442,
      "num_input_tokens_seen": 5568288,
      "step": 8535
    },
    {
      "epoch": 4.475890985324948,
      "grad_norm": 0.05851681903004646,
      "learning_rate": 0.0009540581821724414,
      "loss": 0.7042,
      "num_input_tokens_seen": 5571744,
      "step": 8540
    },
    {
      "epoch": 4.478511530398323,
      "grad_norm": 0.04002220928668976,
      "learning_rate": 0.000953962379891387,
      "loss": 0.5257,
      "num_input_tokens_seen": 5575008,
      "step": 8545
    },
    {
      "epoch": 4.481132075471698,
      "grad_norm": 0.059086378663778305,
      "learning_rate": 0.0009538664826465818,
      "loss": 0.5989,
      "num_input_tokens_seen": 5577824,
      "step": 8550
    },
    {
      "epoch": 4.483752620545073,
      "grad_norm": 0.08810262382030487,
      "learning_rate": 0.0009537704904580864,
      "loss": 0.4643,
      "num_input_tokens_seen": 5581216,
      "step": 8555
    },
    {
      "epoch": 4.486373165618448,
      "grad_norm": 0.09024344384670258,
      "learning_rate": 0.0009536744033459815,
      "loss": 0.4837,
      "num_input_tokens_seen": 5586272,
      "step": 8560
    },
    {
      "epoch": 4.488993710691824,
      "grad_norm": 0.07004629075527191,
      "learning_rate": 0.0009535782213303669,
      "loss": 0.4891,
      "num_input_tokens_seen": 5588608,
      "step": 8565
    },
    {
      "epoch": 4.4916142557651995,
      "grad_norm": 0.08140536397695541,
      "learning_rate": 0.0009534819444313631,
      "loss": 0.3478,
      "num_input_tokens_seen": 5591296,
      "step": 8570
    },
    {
      "epoch": 4.494234800838575,
      "grad_norm": 0.10004029422998428,
      "learning_rate": 0.0009533855726691103,
      "loss": 0.4076,
      "num_input_tokens_seen": 5593568,
      "step": 8575
    },
    {
      "epoch": 4.49685534591195,
      "grad_norm": 0.12638092041015625,
      "learning_rate": 0.0009532891060637681,
      "loss": 0.4665,
      "num_input_tokens_seen": 5596480,
      "step": 8580
    },
    {
      "epoch": 4.499475890985325,
      "grad_norm": 0.04435793310403824,
      "learning_rate": 0.0009531925446355163,
      "loss": 0.5953,
      "num_input_tokens_seen": 5599968,
      "step": 8585
    },
    {
      "epoch": 4.5020964360587,
      "grad_norm": 0.05281359329819679,
      "learning_rate": 0.0009530958884045545,
      "loss": 0.464,
      "num_input_tokens_seen": 5603168,
      "step": 8590
    },
    {
      "epoch": 4.504716981132075,
      "grad_norm": 0.0423855260014534,
      "learning_rate": 0.000952999137391102,
      "loss": 0.6224,
      "num_input_tokens_seen": 5606464,
      "step": 8595
    },
    {
      "epoch": 4.5073375262054505,
      "grad_norm": 0.052419669926166534,
      "learning_rate": 0.0009529022916153982,
      "loss": 0.5292,
      "num_input_tokens_seen": 5609216,
      "step": 8600
    },
    {
      "epoch": 4.509958071278826,
      "grad_norm": 0.05399785190820694,
      "learning_rate": 0.0009528053510977017,
      "loss": 0.4374,
      "num_input_tokens_seen": 5612800,
      "step": 8605
    },
    {
      "epoch": 4.512578616352201,
      "grad_norm": 0.08867834508419037,
      "learning_rate": 0.0009527083158582919,
      "loss": 0.4316,
      "num_input_tokens_seen": 5615488,
      "step": 8610
    },
    {
      "epoch": 4.515199161425577,
      "grad_norm": 0.031594786792993546,
      "learning_rate": 0.0009526111859174671,
      "loss": 0.403,
      "num_input_tokens_seen": 5619264,
      "step": 8615
    },
    {
      "epoch": 4.517819706498952,
      "grad_norm": 0.08378195762634277,
      "learning_rate": 0.0009525139612955458,
      "loss": 0.401,
      "num_input_tokens_seen": 5623104,
      "step": 8620
    },
    {
      "epoch": 4.520440251572327,
      "grad_norm": 0.05780353397130966,
      "learning_rate": 0.0009524166420128664,
      "loss": 0.4837,
      "num_input_tokens_seen": 5626144,
      "step": 8625
    },
    {
      "epoch": 4.523060796645702,
      "grad_norm": 0.05638156458735466,
      "learning_rate": 0.0009523192280897867,
      "loss": 0.4715,
      "num_input_tokens_seen": 5629216,
      "step": 8630
    },
    {
      "epoch": 4.5256813417190775,
      "grad_norm": 0.07660569995641708,
      "learning_rate": 0.0009522217195466851,
      "loss": 0.542,
      "num_input_tokens_seen": 5632448,
      "step": 8635
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 0.05147381126880646,
      "learning_rate": 0.0009521241164039589,
      "loss": 0.5417,
      "num_input_tokens_seen": 5636704,
      "step": 8640
    },
    {
      "epoch": 4.530922431865828,
      "grad_norm": 0.035609178245067596,
      "learning_rate": 0.0009520264186820258,
      "loss": 0.4429,
      "num_input_tokens_seen": 5641408,
      "step": 8645
    },
    {
      "epoch": 4.533542976939203,
      "grad_norm": 0.07478778809309006,
      "learning_rate": 0.0009519286264013227,
      "loss": 0.4601,
      "num_input_tokens_seen": 5644096,
      "step": 8650
    },
    {
      "epoch": 4.536163522012579,
      "grad_norm": 0.05589250847697258,
      "learning_rate": 0.0009518307395823069,
      "loss": 0.4132,
      "num_input_tokens_seen": 5646720,
      "step": 8655
    },
    {
      "epoch": 4.538784067085954,
      "grad_norm": 0.05443554371595383,
      "learning_rate": 0.0009517327582454551,
      "loss": 0.5571,
      "num_input_tokens_seen": 5649696,
      "step": 8660
    },
    {
      "epoch": 4.5414046121593294,
      "grad_norm": 0.05901176854968071,
      "learning_rate": 0.000951634682411264,
      "loss": 0.404,
      "num_input_tokens_seen": 5653504,
      "step": 8665
    },
    {
      "epoch": 4.544025157232705,
      "grad_norm": 0.05560804530978203,
      "learning_rate": 0.0009515365121002498,
      "loss": 0.4892,
      "num_input_tokens_seen": 5656896,
      "step": 8670
    },
    {
      "epoch": 4.54664570230608,
      "grad_norm": 0.05749184265732765,
      "learning_rate": 0.0009514382473329487,
      "loss": 0.4828,
      "num_input_tokens_seen": 5660928,
      "step": 8675
    },
    {
      "epoch": 4.549266247379455,
      "grad_norm": 0.07058078795671463,
      "learning_rate": 0.0009513398881299164,
      "loss": 0.495,
      "num_input_tokens_seen": 5664064,
      "step": 8680
    },
    {
      "epoch": 4.55188679245283,
      "grad_norm": 0.1200486272573471,
      "learning_rate": 0.0009512414345117289,
      "loss": 0.5457,
      "num_input_tokens_seen": 5667232,
      "step": 8685
    },
    {
      "epoch": 4.554507337526205,
      "grad_norm": 0.07760611921548843,
      "learning_rate": 0.0009511428864989813,
      "loss": 0.4865,
      "num_input_tokens_seen": 5670176,
      "step": 8690
    },
    {
      "epoch": 4.5571278825995805,
      "grad_norm": 0.09407879412174225,
      "learning_rate": 0.0009510442441122886,
      "loss": 0.3779,
      "num_input_tokens_seen": 5673120,
      "step": 8695
    },
    {
      "epoch": 4.559748427672956,
      "grad_norm": 0.0706339105963707,
      "learning_rate": 0.0009509455073722859,
      "loss": 0.4978,
      "num_input_tokens_seen": 5675936,
      "step": 8700
    },
    {
      "epoch": 4.562368972746331,
      "grad_norm": 0.03853830695152283,
      "learning_rate": 0.0009508466762996277,
      "loss": 0.3701,
      "num_input_tokens_seen": 5679072,
      "step": 8705
    },
    {
      "epoch": 4.564989517819707,
      "grad_norm": 0.06552845239639282,
      "learning_rate": 0.0009507477509149883,
      "loss": 0.473,
      "num_input_tokens_seen": 5682112,
      "step": 8710
    },
    {
      "epoch": 4.567610062893082,
      "grad_norm": 0.05217764899134636,
      "learning_rate": 0.0009506487312390619,
      "loss": 0.5915,
      "num_input_tokens_seen": 5685696,
      "step": 8715
    },
    {
      "epoch": 4.570230607966457,
      "grad_norm": 0.059636421501636505,
      "learning_rate": 0.0009505496172925622,
      "loss": 0.4709,
      "num_input_tokens_seen": 5688096,
      "step": 8720
    },
    {
      "epoch": 4.572851153039832,
      "grad_norm": 0.06697054952383041,
      "learning_rate": 0.0009504504090962226,
      "loss": 0.449,
      "num_input_tokens_seen": 5691072,
      "step": 8725
    },
    {
      "epoch": 4.5754716981132075,
      "grad_norm": 0.05939861387014389,
      "learning_rate": 0.0009503511066707966,
      "loss": 0.499,
      "num_input_tokens_seen": 5694592,
      "step": 8730
    },
    {
      "epoch": 4.578092243186583,
      "grad_norm": 0.06376273930072784,
      "learning_rate": 0.0009502517100370568,
      "loss": 0.3839,
      "num_input_tokens_seen": 5697312,
      "step": 8735
    },
    {
      "epoch": 4.580712788259958,
      "grad_norm": 0.031952254474163055,
      "learning_rate": 0.0009501522192157961,
      "loss": 0.3425,
      "num_input_tokens_seen": 5701088,
      "step": 8740
    },
    {
      "epoch": 4.583333333333333,
      "grad_norm": 0.045850303024053574,
      "learning_rate": 0.0009500526342278266,
      "loss": 0.536,
      "num_input_tokens_seen": 5703904,
      "step": 8745
    },
    {
      "epoch": 4.585953878406709,
      "grad_norm": 0.055470675230026245,
      "learning_rate": 0.0009499529550939807,
      "loss": 0.4891,
      "num_input_tokens_seen": 5706816,
      "step": 8750
    },
    {
      "epoch": 4.588574423480084,
      "grad_norm": 0.094480000436306,
      "learning_rate": 0.0009498531818351098,
      "loss": 0.4834,
      "num_input_tokens_seen": 5709632,
      "step": 8755
    },
    {
      "epoch": 4.591194968553459,
      "grad_norm": 0.11806591600179672,
      "learning_rate": 0.0009497533144720854,
      "loss": 0.5069,
      "num_input_tokens_seen": 5713472,
      "step": 8760
    },
    {
      "epoch": 4.593815513626835,
      "grad_norm": 0.09739664196968079,
      "learning_rate": 0.0009496533530257988,
      "loss": 0.6281,
      "num_input_tokens_seen": 5716736,
      "step": 8765
    },
    {
      "epoch": 4.59643605870021,
      "grad_norm": 0.10237612575292587,
      "learning_rate": 0.0009495532975171605,
      "loss": 0.4181,
      "num_input_tokens_seen": 5720672,
      "step": 8770
    },
    {
      "epoch": 4.599056603773585,
      "grad_norm": 0.053454745560884476,
      "learning_rate": 0.0009494531479671014,
      "loss": 0.6879,
      "num_input_tokens_seen": 5724000,
      "step": 8775
    },
    {
      "epoch": 4.60167714884696,
      "grad_norm": 0.07417599856853485,
      "learning_rate": 0.0009493529043965712,
      "loss": 0.658,
      "num_input_tokens_seen": 5726784,
      "step": 8780
    },
    {
      "epoch": 4.604297693920335,
      "grad_norm": 0.04918177053332329,
      "learning_rate": 0.00094925256682654,
      "loss": 0.4756,
      "num_input_tokens_seen": 5730272,
      "step": 8785
    },
    {
      "epoch": 4.6069182389937104,
      "grad_norm": 0.10573069751262665,
      "learning_rate": 0.000949152135277997,
      "loss": 0.4768,
      "num_input_tokens_seen": 5733120,
      "step": 8790
    },
    {
      "epoch": 4.609538784067086,
      "grad_norm": 0.08252248913049698,
      "learning_rate": 0.0009490516097719515,
      "loss": 0.3087,
      "num_input_tokens_seen": 5737824,
      "step": 8795
    },
    {
      "epoch": 4.612159329140461,
      "grad_norm": 0.07524421811103821,
      "learning_rate": 0.0009489509903294324,
      "loss": 0.4023,
      "num_input_tokens_seen": 5740384,
      "step": 8800
    },
    {
      "epoch": 4.614779874213837,
      "grad_norm": 0.08075255155563354,
      "learning_rate": 0.000948850276971488,
      "loss": 0.4432,
      "num_input_tokens_seen": 5743232,
      "step": 8805
    },
    {
      "epoch": 4.617400419287212,
      "grad_norm": 0.05243588611483574,
      "learning_rate": 0.0009487494697191864,
      "loss": 0.5587,
      "num_input_tokens_seen": 5746432,
      "step": 8810
    },
    {
      "epoch": 4.620020964360587,
      "grad_norm": 0.08333833515644073,
      "learning_rate": 0.0009486485685936154,
      "loss": 0.5473,
      "num_input_tokens_seen": 5749696,
      "step": 8815
    },
    {
      "epoch": 4.622641509433962,
      "grad_norm": 0.0478074848651886,
      "learning_rate": 0.0009485475736158822,
      "loss": 0.4,
      "num_input_tokens_seen": 5753152,
      "step": 8820
    },
    {
      "epoch": 4.6252620545073375,
      "grad_norm": 0.058707937598228455,
      "learning_rate": 0.000948446484807114,
      "loss": 0.423,
      "num_input_tokens_seen": 5755712,
      "step": 8825
    },
    {
      "epoch": 4.627882599580713,
      "grad_norm": 0.06661152094602585,
      "learning_rate": 0.0009483453021884572,
      "loss": 0.4963,
      "num_input_tokens_seen": 5759040,
      "step": 8830
    },
    {
      "epoch": 4.630503144654088,
      "grad_norm": 0.09475835412740707,
      "learning_rate": 0.0009482440257810782,
      "loss": 0.5579,
      "num_input_tokens_seen": 5761408,
      "step": 8835
    },
    {
      "epoch": 4.633123689727463,
      "grad_norm": 0.0702674612402916,
      "learning_rate": 0.000948142655606163,
      "loss": 0.5347,
      "num_input_tokens_seen": 5764672,
      "step": 8840
    },
    {
      "epoch": 4.635744234800838,
      "grad_norm": 0.04429427161812782,
      "learning_rate": 0.0009480411916849168,
      "loss": 0.449,
      "num_input_tokens_seen": 5768160,
      "step": 8845
    },
    {
      "epoch": 4.638364779874214,
      "grad_norm": 0.16346682608127594,
      "learning_rate": 0.0009479396340385649,
      "loss": 0.461,
      "num_input_tokens_seen": 5770400,
      "step": 8850
    },
    {
      "epoch": 4.640985324947589,
      "grad_norm": 0.04753625765442848,
      "learning_rate": 0.0009478379826883519,
      "loss": 0.375,
      "num_input_tokens_seen": 5774080,
      "step": 8855
    },
    {
      "epoch": 4.643605870020965,
      "grad_norm": 0.12770149111747742,
      "learning_rate": 0.0009477362376555421,
      "loss": 0.5176,
      "num_input_tokens_seen": 5777184,
      "step": 8860
    },
    {
      "epoch": 4.64622641509434,
      "grad_norm": 0.08799320459365845,
      "learning_rate": 0.0009476343989614194,
      "loss": 0.5122,
      "num_input_tokens_seen": 5780736,
      "step": 8865
    },
    {
      "epoch": 4.648846960167715,
      "grad_norm": 0.056590646505355835,
      "learning_rate": 0.0009475324666272873,
      "loss": 0.5396,
      "num_input_tokens_seen": 5784320,
      "step": 8870
    },
    {
      "epoch": 4.65146750524109,
      "grad_norm": 0.06385891884565353,
      "learning_rate": 0.0009474304406744689,
      "loss": 0.4563,
      "num_input_tokens_seen": 5787840,
      "step": 8875
    },
    {
      "epoch": 4.654088050314465,
      "grad_norm": 0.06616915762424469,
      "learning_rate": 0.0009473283211243069,
      "loss": 0.6148,
      "num_input_tokens_seen": 5791040,
      "step": 8880
    },
    {
      "epoch": 4.65670859538784,
      "grad_norm": 0.20349039137363434,
      "learning_rate": 0.0009472261079981637,
      "loss": 0.7608,
      "num_input_tokens_seen": 5794560,
      "step": 8885
    },
    {
      "epoch": 4.659329140461216,
      "grad_norm": 0.097333624958992,
      "learning_rate": 0.0009471238013174206,
      "loss": 0.4817,
      "num_input_tokens_seen": 5797504,
      "step": 8890
    },
    {
      "epoch": 4.661949685534591,
      "grad_norm": 0.06172066926956177,
      "learning_rate": 0.0009470214011034795,
      "loss": 0.456,
      "num_input_tokens_seen": 5799808,
      "step": 8895
    },
    {
      "epoch": 4.664570230607967,
      "grad_norm": 0.05837142840027809,
      "learning_rate": 0.0009469189073777612,
      "loss": 0.3519,
      "num_input_tokens_seen": 5802752,
      "step": 8900
    },
    {
      "epoch": 4.667190775681342,
      "grad_norm": 0.05876993387937546,
      "learning_rate": 0.0009468163201617061,
      "loss": 0.4803,
      "num_input_tokens_seen": 5806144,
      "step": 8905
    },
    {
      "epoch": 4.669811320754717,
      "grad_norm": 0.05892123281955719,
      "learning_rate": 0.0009467136394767744,
      "loss": 0.5767,
      "num_input_tokens_seen": 5809536,
      "step": 8910
    },
    {
      "epoch": 4.672431865828092,
      "grad_norm": 0.06357603520154953,
      "learning_rate": 0.0009466108653444458,
      "loss": 0.4619,
      "num_input_tokens_seen": 5813088,
      "step": 8915
    },
    {
      "epoch": 4.6750524109014675,
      "grad_norm": 0.08226435631513596,
      "learning_rate": 0.0009465079977862193,
      "loss": 0.3398,
      "num_input_tokens_seen": 5815488,
      "step": 8920
    },
    {
      "epoch": 4.677672955974843,
      "grad_norm": 0.04482625424861908,
      "learning_rate": 0.0009464050368236137,
      "loss": 0.5981,
      "num_input_tokens_seen": 5818528,
      "step": 8925
    },
    {
      "epoch": 4.680293501048218,
      "grad_norm": 0.06080184131860733,
      "learning_rate": 0.0009463019824781674,
      "loss": 0.5419,
      "num_input_tokens_seen": 5822016,
      "step": 8930
    },
    {
      "epoch": 4.682914046121593,
      "grad_norm": 0.02957003377377987,
      "learning_rate": 0.0009461988347714377,
      "loss": 0.635,
      "num_input_tokens_seen": 5825952,
      "step": 8935
    },
    {
      "epoch": 4.685534591194968,
      "grad_norm": 0.04787308722734451,
      "learning_rate": 0.0009460955937250025,
      "loss": 0.3659,
      "num_input_tokens_seen": 5828416,
      "step": 8940
    },
    {
      "epoch": 4.688155136268344,
      "grad_norm": 0.06072263419628143,
      "learning_rate": 0.0009459922593604584,
      "loss": 0.4585,
      "num_input_tokens_seen": 5832352,
      "step": 8945
    },
    {
      "epoch": 4.690775681341719,
      "grad_norm": 0.05819239094853401,
      "learning_rate": 0.0009458888316994219,
      "loss": 0.5489,
      "num_input_tokens_seen": 5835488,
      "step": 8950
    },
    {
      "epoch": 4.693396226415095,
      "grad_norm": 0.06621179729700089,
      "learning_rate": 0.0009457853107635286,
      "loss": 0.5086,
      "num_input_tokens_seen": 5838624,
      "step": 8955
    },
    {
      "epoch": 4.69601677148847,
      "grad_norm": 0.051751501858234406,
      "learning_rate": 0.0009456816965744342,
      "loss": 0.4696,
      "num_input_tokens_seen": 5842592,
      "step": 8960
    },
    {
      "epoch": 4.698637316561845,
      "grad_norm": 0.06686946749687195,
      "learning_rate": 0.0009455779891538134,
      "loss": 0.5095,
      "num_input_tokens_seen": 5846080,
      "step": 8965
    },
    {
      "epoch": 4.70125786163522,
      "grad_norm": 0.11191636323928833,
      "learning_rate": 0.0009454741885233606,
      "loss": 0.5181,
      "num_input_tokens_seen": 5849536,
      "step": 8970
    },
    {
      "epoch": 4.703878406708595,
      "grad_norm": 0.07118020206689835,
      "learning_rate": 0.0009453702947047899,
      "loss": 0.4274,
      "num_input_tokens_seen": 5852032,
      "step": 8975
    },
    {
      "epoch": 4.70649895178197,
      "grad_norm": 0.10297800600528717,
      "learning_rate": 0.0009452663077198347,
      "loss": 0.5701,
      "num_input_tokens_seen": 5855072,
      "step": 8980
    },
    {
      "epoch": 4.709119496855346,
      "grad_norm": 0.04950203746557236,
      "learning_rate": 0.0009451622275902477,
      "loss": 0.519,
      "num_input_tokens_seen": 5858272,
      "step": 8985
    },
    {
      "epoch": 4.711740041928721,
      "grad_norm": 0.05171087384223938,
      "learning_rate": 0.0009450580543378013,
      "loss": 0.4722,
      "num_input_tokens_seen": 5861856,
      "step": 8990
    },
    {
      "epoch": 4.714360587002097,
      "grad_norm": 0.06882024556398392,
      "learning_rate": 0.0009449537879842875,
      "loss": 0.4775,
      "num_input_tokens_seen": 5864448,
      "step": 8995
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 0.05441726744174957,
      "learning_rate": 0.0009448494285515177,
      "loss": 0.4251,
      "num_input_tokens_seen": 5867744,
      "step": 9000
    },
    {
      "epoch": 4.719601677148847,
      "grad_norm": 0.09426817297935486,
      "learning_rate": 0.0009447449760613222,
      "loss": 0.3952,
      "num_input_tokens_seen": 5870336,
      "step": 9005
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 0.03928077220916748,
      "learning_rate": 0.0009446404305355519,
      "loss": 0.47,
      "num_input_tokens_seen": 5873728,
      "step": 9010
    },
    {
      "epoch": 4.7248427672955975,
      "grad_norm": 0.05045327916741371,
      "learning_rate": 0.0009445357919960762,
      "loss": 0.4263,
      "num_input_tokens_seen": 5876320,
      "step": 9015
    },
    {
      "epoch": 4.727463312368973,
      "grad_norm": 0.14632166922092438,
      "learning_rate": 0.0009444310604647844,
      "loss": 0.4311,
      "num_input_tokens_seen": 5879552,
      "step": 9020
    },
    {
      "epoch": 4.730083857442348,
      "grad_norm": 0.15566818416118622,
      "learning_rate": 0.000944326235963585,
      "loss": 0.4144,
      "num_input_tokens_seen": 5882432,
      "step": 9025
    },
    {
      "epoch": 4.732704402515723,
      "grad_norm": 0.11668270081281662,
      "learning_rate": 0.0009442213185144062,
      "loss": 0.5067,
      "num_input_tokens_seen": 5885248,
      "step": 9030
    },
    {
      "epoch": 4.735324947589098,
      "grad_norm": 0.04601726308465004,
      "learning_rate": 0.0009441163081391954,
      "loss": 0.4637,
      "num_input_tokens_seen": 5887680,
      "step": 9035
    },
    {
      "epoch": 4.737945492662474,
      "grad_norm": 0.062266550958156586,
      "learning_rate": 0.00094401120485992,
      "loss": 0.4707,
      "num_input_tokens_seen": 5890848,
      "step": 9040
    },
    {
      "epoch": 4.740566037735849,
      "grad_norm": 0.05255145579576492,
      "learning_rate": 0.0009439060086985658,
      "loss": 0.4611,
      "num_input_tokens_seen": 5893664,
      "step": 9045
    },
    {
      "epoch": 4.743186582809225,
      "grad_norm": 0.0451277531683445,
      "learning_rate": 0.000943800719677139,
      "loss": 0.5205,
      "num_input_tokens_seen": 5897216,
      "step": 9050
    },
    {
      "epoch": 4.7458071278826,
      "grad_norm": 0.08552230149507523,
      "learning_rate": 0.0009436953378176649,
      "loss": 0.4797,
      "num_input_tokens_seen": 5900992,
      "step": 9055
    },
    {
      "epoch": 4.748427672955975,
      "grad_norm": 0.05466673523187637,
      "learning_rate": 0.0009435898631421879,
      "loss": 0.5713,
      "num_input_tokens_seen": 5903904,
      "step": 9060
    },
    {
      "epoch": 4.75104821802935,
      "grad_norm": 0.06746350973844528,
      "learning_rate": 0.0009434842956727724,
      "loss": 0.5627,
      "num_input_tokens_seen": 5907008,
      "step": 9065
    },
    {
      "epoch": 4.753668763102725,
      "grad_norm": 0.06898508220911026,
      "learning_rate": 0.0009433786354315017,
      "loss": 0.4188,
      "num_input_tokens_seen": 5910112,
      "step": 9070
    },
    {
      "epoch": 4.7562893081761,
      "grad_norm": 0.05345109850168228,
      "learning_rate": 0.0009432728824404789,
      "loss": 0.4362,
      "num_input_tokens_seen": 5913216,
      "step": 9075
    },
    {
      "epoch": 4.758909853249476,
      "grad_norm": 0.05154608562588692,
      "learning_rate": 0.0009431670367218262,
      "loss": 0.469,
      "num_input_tokens_seen": 5916256,
      "step": 9080
    },
    {
      "epoch": 4.761530398322851,
      "grad_norm": 0.0702916607260704,
      "learning_rate": 0.0009430610982976852,
      "loss": 0.4846,
      "num_input_tokens_seen": 5919040,
      "step": 9085
    },
    {
      "epoch": 4.764150943396227,
      "grad_norm": 0.05147610232234001,
      "learning_rate": 0.0009429550671902171,
      "loss": 0.4371,
      "num_input_tokens_seen": 5922944,
      "step": 9090
    },
    {
      "epoch": 4.766771488469602,
      "grad_norm": 0.10509012639522552,
      "learning_rate": 0.0009428489434216028,
      "loss": 0.3677,
      "num_input_tokens_seen": 5926368,
      "step": 9095
    },
    {
      "epoch": 4.769392033542977,
      "grad_norm": 0.055707767605781555,
      "learning_rate": 0.0009427427270140415,
      "loss": 0.4331,
      "num_input_tokens_seen": 5929632,
      "step": 9100
    },
    {
      "epoch": 4.772012578616352,
      "grad_norm": 0.05082991346716881,
      "learning_rate": 0.0009426364179897529,
      "loss": 0.4749,
      "num_input_tokens_seen": 5932832,
      "step": 9105
    },
    {
      "epoch": 4.7746331236897275,
      "grad_norm": 0.05705252289772034,
      "learning_rate": 0.0009425300163709756,
      "loss": 0.4059,
      "num_input_tokens_seen": 5936608,
      "step": 9110
    },
    {
      "epoch": 4.777253668763103,
      "grad_norm": 0.12074942886829376,
      "learning_rate": 0.0009424235221799673,
      "loss": 0.5857,
      "num_input_tokens_seen": 5940640,
      "step": 9115
    },
    {
      "epoch": 4.779874213836478,
      "grad_norm": 0.08060827851295471,
      "learning_rate": 0.0009423169354390058,
      "loss": 0.4211,
      "num_input_tokens_seen": 5944416,
      "step": 9120
    },
    {
      "epoch": 4.782494758909853,
      "grad_norm": 0.17774562537670135,
      "learning_rate": 0.0009422102561703875,
      "loss": 0.4813,
      "num_input_tokens_seen": 5948352,
      "step": 9125
    },
    {
      "epoch": 4.785115303983228,
      "grad_norm": 0.056984540075063705,
      "learning_rate": 0.0009421034843964287,
      "loss": 0.4096,
      "num_input_tokens_seen": 5952000,
      "step": 9130
    },
    {
      "epoch": 4.787735849056604,
      "grad_norm": 0.12598612904548645,
      "learning_rate": 0.0009419966201394646,
      "loss": 0.4705,
      "num_input_tokens_seen": 5954528,
      "step": 9135
    },
    {
      "epoch": 4.790356394129979,
      "grad_norm": 0.09040778130292892,
      "learning_rate": 0.0009418896634218503,
      "loss": 0.4446,
      "num_input_tokens_seen": 5957664,
      "step": 9140
    },
    {
      "epoch": 4.7929769392033545,
      "grad_norm": 0.03928488865494728,
      "learning_rate": 0.0009417826142659596,
      "loss": 0.3677,
      "num_input_tokens_seen": 5960448,
      "step": 9145
    },
    {
      "epoch": 4.79559748427673,
      "grad_norm": 0.049854826182127,
      "learning_rate": 0.0009416754726941863,
      "loss": 0.4175,
      "num_input_tokens_seen": 5963072,
      "step": 9150
    },
    {
      "epoch": 4.798218029350105,
      "grad_norm": 0.06505906581878662,
      "learning_rate": 0.0009415682387289428,
      "loss": 0.3815,
      "num_input_tokens_seen": 5966240,
      "step": 9155
    },
    {
      "epoch": 4.80083857442348,
      "grad_norm": 0.05364169925451279,
      "learning_rate": 0.0009414609123926616,
      "loss": 0.3954,
      "num_input_tokens_seen": 5969664,
      "step": 9160
    },
    {
      "epoch": 4.803459119496855,
      "grad_norm": 0.07241550087928772,
      "learning_rate": 0.000941353493707794,
      "loss": 0.4527,
      "num_input_tokens_seen": 5972320,
      "step": 9165
    },
    {
      "epoch": 4.80607966457023,
      "grad_norm": 0.05691011995077133,
      "learning_rate": 0.0009412459826968107,
      "loss": 0.5375,
      "num_input_tokens_seen": 5975264,
      "step": 9170
    },
    {
      "epoch": 4.808700209643606,
      "grad_norm": 0.08962862193584442,
      "learning_rate": 0.0009411383793822017,
      "loss": 0.5284,
      "num_input_tokens_seen": 5978336,
      "step": 9175
    },
    {
      "epoch": 4.811320754716981,
      "grad_norm": 0.07736938446760178,
      "learning_rate": 0.0009410306837864766,
      "loss": 0.5014,
      "num_input_tokens_seen": 5981536,
      "step": 9180
    },
    {
      "epoch": 4.813941299790356,
      "grad_norm": 0.14244306087493896,
      "learning_rate": 0.000940922895932164,
      "loss": 0.6028,
      "num_input_tokens_seen": 5984640,
      "step": 9185
    },
    {
      "epoch": 4.816561844863732,
      "grad_norm": 0.10745087265968323,
      "learning_rate": 0.0009408150158418119,
      "loss": 0.5123,
      "num_input_tokens_seen": 5987168,
      "step": 9190
    },
    {
      "epoch": 4.819182389937107,
      "grad_norm": 0.03478652238845825,
      "learning_rate": 0.0009407070435379876,
      "loss": 0.5258,
      "num_input_tokens_seen": 5990880,
      "step": 9195
    },
    {
      "epoch": 4.821802935010482,
      "grad_norm": 0.06743571162223816,
      "learning_rate": 0.0009405989790432776,
      "loss": 0.4991,
      "num_input_tokens_seen": 5993664,
      "step": 9200
    },
    {
      "epoch": 4.8244234800838575,
      "grad_norm": 0.05272160843014717,
      "learning_rate": 0.0009404908223802877,
      "loss": 0.5406,
      "num_input_tokens_seen": 5997216,
      "step": 9205
    },
    {
      "epoch": 4.827044025157233,
      "grad_norm": 0.05652644857764244,
      "learning_rate": 0.0009403825735716433,
      "loss": 0.3222,
      "num_input_tokens_seen": 5999840,
      "step": 9210
    },
    {
      "epoch": 4.829664570230608,
      "grad_norm": 0.11129406839609146,
      "learning_rate": 0.0009402742326399887,
      "loss": 0.4543,
      "num_input_tokens_seen": 6003008,
      "step": 9215
    },
    {
      "epoch": 4.832285115303983,
      "grad_norm": 0.0651928186416626,
      "learning_rate": 0.0009401657996079873,
      "loss": 0.502,
      "num_input_tokens_seen": 6007552,
      "step": 9220
    },
    {
      "epoch": 4.834905660377358,
      "grad_norm": 0.06443551927804947,
      "learning_rate": 0.0009400572744983224,
      "loss": 0.5055,
      "num_input_tokens_seen": 6011040,
      "step": 9225
    },
    {
      "epoch": 4.837526205450734,
      "grad_norm": 0.05408147722482681,
      "learning_rate": 0.0009399486573336963,
      "loss": 0.385,
      "num_input_tokens_seen": 6013536,
      "step": 9230
    },
    {
      "epoch": 4.840146750524109,
      "grad_norm": 0.09644519537687302,
      "learning_rate": 0.00093983994813683,
      "loss": 0.4104,
      "num_input_tokens_seen": 6015808,
      "step": 9235
    },
    {
      "epoch": 4.8427672955974845,
      "grad_norm": 0.0670485869050026,
      "learning_rate": 0.0009397311469304648,
      "loss": 0.3601,
      "num_input_tokens_seen": 6019328,
      "step": 9240
    },
    {
      "epoch": 4.84538784067086,
      "grad_norm": 0.04477771371603012,
      "learning_rate": 0.0009396222537373602,
      "loss": 0.4222,
      "num_input_tokens_seen": 6022496,
      "step": 9245
    },
    {
      "epoch": 4.848008385744235,
      "grad_norm": 0.1082615926861763,
      "learning_rate": 0.0009395132685802956,
      "loss": 0.43,
      "num_input_tokens_seen": 6025408,
      "step": 9250
    },
    {
      "epoch": 4.85062893081761,
      "grad_norm": 0.05581658333539963,
      "learning_rate": 0.0009394041914820695,
      "loss": 0.5481,
      "num_input_tokens_seen": 6029056,
      "step": 9255
    },
    {
      "epoch": 4.853249475890985,
      "grad_norm": 0.07427355647087097,
      "learning_rate": 0.0009392950224654994,
      "loss": 0.3347,
      "num_input_tokens_seen": 6031584,
      "step": 9260
    },
    {
      "epoch": 4.85587002096436,
      "grad_norm": 0.07199940085411072,
      "learning_rate": 0.0009391857615534225,
      "loss": 0.5179,
      "num_input_tokens_seen": 6034432,
      "step": 9265
    },
    {
      "epoch": 4.8584905660377355,
      "grad_norm": 0.07058636844158173,
      "learning_rate": 0.0009390764087686948,
      "loss": 0.4154,
      "num_input_tokens_seen": 6037792,
      "step": 9270
    },
    {
      "epoch": 4.861111111111111,
      "grad_norm": 0.07158984243869781,
      "learning_rate": 0.0009389669641341916,
      "loss": 0.4852,
      "num_input_tokens_seen": 6040832,
      "step": 9275
    },
    {
      "epoch": 4.863731656184486,
      "grad_norm": 0.08348364382982254,
      "learning_rate": 0.0009388574276728075,
      "loss": 0.5968,
      "num_input_tokens_seen": 6044480,
      "step": 9280
    },
    {
      "epoch": 4.866352201257862,
      "grad_norm": 0.049542881548404694,
      "learning_rate": 0.0009387477994074562,
      "loss": 0.4293,
      "num_input_tokens_seen": 6047520,
      "step": 9285
    },
    {
      "epoch": 4.868972746331237,
      "grad_norm": 0.06520684063434601,
      "learning_rate": 0.0009386380793610707,
      "loss": 0.4375,
      "num_input_tokens_seen": 6051008,
      "step": 9290
    },
    {
      "epoch": 4.871593291404612,
      "grad_norm": 0.04586494714021683,
      "learning_rate": 0.0009385282675566034,
      "loss": 0.502,
      "num_input_tokens_seen": 6053216,
      "step": 9295
    },
    {
      "epoch": 4.8742138364779874,
      "grad_norm": 0.04707310348749161,
      "learning_rate": 0.0009384183640170255,
      "loss": 0.5634,
      "num_input_tokens_seen": 6056864,
      "step": 9300
    },
    {
      "epoch": 4.876834381551363,
      "grad_norm": 0.05499634891748428,
      "learning_rate": 0.0009383083687653275,
      "loss": 0.4259,
      "num_input_tokens_seen": 6059744,
      "step": 9305
    },
    {
      "epoch": 4.879454926624738,
      "grad_norm": 0.0856749415397644,
      "learning_rate": 0.0009381982818245193,
      "loss": 0.3996,
      "num_input_tokens_seen": 6062624,
      "step": 9310
    },
    {
      "epoch": 4.882075471698113,
      "grad_norm": 0.07556832581758499,
      "learning_rate": 0.0009380881032176299,
      "loss": 0.4641,
      "num_input_tokens_seen": 6066080,
      "step": 9315
    },
    {
      "epoch": 4.884696016771488,
      "grad_norm": 0.05593406409025192,
      "learning_rate": 0.0009379778329677071,
      "loss": 0.5504,
      "num_input_tokens_seen": 6069696,
      "step": 9320
    },
    {
      "epoch": 4.887316561844864,
      "grad_norm": 0.07476072758436203,
      "learning_rate": 0.0009378674710978184,
      "loss": 0.5805,
      "num_input_tokens_seen": 6073408,
      "step": 9325
    },
    {
      "epoch": 4.889937106918239,
      "grad_norm": 0.0670701190829277,
      "learning_rate": 0.0009377570176310503,
      "loss": 0.3156,
      "num_input_tokens_seen": 6078592,
      "step": 9330
    },
    {
      "epoch": 4.8925576519916145,
      "grad_norm": 0.059071727097034454,
      "learning_rate": 0.0009376464725905082,
      "loss": 0.3487,
      "num_input_tokens_seen": 6088128,
      "step": 9335
    },
    {
      "epoch": 4.89517819706499,
      "grad_norm": 0.12618286907672882,
      "learning_rate": 0.000937535835999317,
      "loss": 0.5041,
      "num_input_tokens_seen": 6090816,
      "step": 9340
    },
    {
      "epoch": 4.897798742138365,
      "grad_norm": 0.051674626767635345,
      "learning_rate": 0.0009374251078806206,
      "loss": 0.481,
      "num_input_tokens_seen": 6093856,
      "step": 9345
    },
    {
      "epoch": 4.90041928721174,
      "grad_norm": 0.057937297970056534,
      "learning_rate": 0.000937314288257582,
      "loss": 0.4495,
      "num_input_tokens_seen": 6097792,
      "step": 9350
    },
    {
      "epoch": 4.903039832285115,
      "grad_norm": 0.062447547912597656,
      "learning_rate": 0.0009372033771533835,
      "loss": 0.3237,
      "num_input_tokens_seen": 6100160,
      "step": 9355
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 0.10972990095615387,
      "learning_rate": 0.0009370923745912264,
      "loss": 0.572,
      "num_input_tokens_seen": 6103264,
      "step": 9360
    },
    {
      "epoch": 4.9082809224318655,
      "grad_norm": 0.05090043321251869,
      "learning_rate": 0.000936981280594331,
      "loss": 0.5739,
      "num_input_tokens_seen": 6105856,
      "step": 9365
    },
    {
      "epoch": 4.910901467505241,
      "grad_norm": 0.08810679614543915,
      "learning_rate": 0.0009368700951859371,
      "loss": 0.5515,
      "num_input_tokens_seen": 6108800,
      "step": 9370
    },
    {
      "epoch": 4.913522012578616,
      "grad_norm": 0.05772921442985535,
      "learning_rate": 0.0009367588183893035,
      "loss": 0.4227,
      "num_input_tokens_seen": 6111552,
      "step": 9375
    },
    {
      "epoch": 4.916142557651992,
      "grad_norm": 0.044315729290246964,
      "learning_rate": 0.0009366474502277076,
      "loss": 0.5367,
      "num_input_tokens_seen": 6115488,
      "step": 9380
    },
    {
      "epoch": 4.918763102725367,
      "grad_norm": 0.07554151862859726,
      "learning_rate": 0.0009365359907244469,
      "loss": 0.3995,
      "num_input_tokens_seen": 6119136,
      "step": 9385
    },
    {
      "epoch": 4.921383647798742,
      "grad_norm": 0.14291515946388245,
      "learning_rate": 0.0009364244399028372,
      "loss": 0.594,
      "num_input_tokens_seen": 6121696,
      "step": 9390
    },
    {
      "epoch": 4.924004192872117,
      "grad_norm": 0.07349108904600143,
      "learning_rate": 0.0009363127977862136,
      "loss": 0.411,
      "num_input_tokens_seen": 6124480,
      "step": 9395
    },
    {
      "epoch": 4.926624737945493,
      "grad_norm": 0.11152590066194534,
      "learning_rate": 0.0009362010643979304,
      "loss": 0.4497,
      "num_input_tokens_seen": 6127008,
      "step": 9400
    },
    {
      "epoch": 4.929245283018868,
      "grad_norm": 0.11632702499628067,
      "learning_rate": 0.000936089239761361,
      "loss": 0.299,
      "num_input_tokens_seen": 6130080,
      "step": 9405
    },
    {
      "epoch": 4.931865828092243,
      "grad_norm": 0.05165126547217369,
      "learning_rate": 0.0009359773238998978,
      "loss": 0.4615,
      "num_input_tokens_seen": 6133632,
      "step": 9410
    },
    {
      "epoch": 4.934486373165618,
      "grad_norm": 0.05948328599333763,
      "learning_rate": 0.0009358653168369524,
      "loss": 0.4579,
      "num_input_tokens_seen": 6136992,
      "step": 9415
    },
    {
      "epoch": 4.937106918238994,
      "grad_norm": 0.07114021480083466,
      "learning_rate": 0.0009357532185959554,
      "loss": 0.3844,
      "num_input_tokens_seen": 6139776,
      "step": 9420
    },
    {
      "epoch": 4.939727463312369,
      "grad_norm": 0.07402525842189789,
      "learning_rate": 0.0009356410292003562,
      "loss": 0.5224,
      "num_input_tokens_seen": 6142880,
      "step": 9425
    },
    {
      "epoch": 4.9423480083857445,
      "grad_norm": 0.07183165848255157,
      "learning_rate": 0.0009355287486736239,
      "loss": 0.4508,
      "num_input_tokens_seen": 6147200,
      "step": 9430
    },
    {
      "epoch": 4.94496855345912,
      "grad_norm": 0.08088783174753189,
      "learning_rate": 0.0009354163770392461,
      "loss": 0.4614,
      "num_input_tokens_seen": 6150016,
      "step": 9435
    },
    {
      "epoch": 4.947589098532495,
      "grad_norm": 0.052506010979413986,
      "learning_rate": 0.0009353039143207295,
      "loss": 0.3989,
      "num_input_tokens_seen": 6152960,
      "step": 9440
    },
    {
      "epoch": 4.95020964360587,
      "grad_norm": 0.06701838225126266,
      "learning_rate": 0.0009351913605416007,
      "loss": 0.3747,
      "num_input_tokens_seen": 6156544,
      "step": 9445
    },
    {
      "epoch": 4.952830188679245,
      "grad_norm": 0.095973901450634,
      "learning_rate": 0.000935078715725404,
      "loss": 0.4762,
      "num_input_tokens_seen": 6159584,
      "step": 9450
    },
    {
      "epoch": 4.95545073375262,
      "grad_norm": 0.07108544558286667,
      "learning_rate": 0.0009349659798957034,
      "loss": 0.4701,
      "num_input_tokens_seen": 6162752,
      "step": 9455
    },
    {
      "epoch": 4.9580712788259955,
      "grad_norm": 0.061616070568561554,
      "learning_rate": 0.0009348531530760823,
      "loss": 0.3267,
      "num_input_tokens_seen": 6166016,
      "step": 9460
    },
    {
      "epoch": 4.960691823899371,
      "grad_norm": 0.10552626848220825,
      "learning_rate": 0.0009347402352901426,
      "loss": 0.4365,
      "num_input_tokens_seen": 6169184,
      "step": 9465
    },
    {
      "epoch": 4.963312368972746,
      "grad_norm": 0.09633222967386246,
      "learning_rate": 0.0009346272265615057,
      "loss": 0.4662,
      "num_input_tokens_seen": 6172960,
      "step": 9470
    },
    {
      "epoch": 4.965932914046122,
      "grad_norm": 0.04340081289410591,
      "learning_rate": 0.0009345141269138112,
      "loss": 0.4414,
      "num_input_tokens_seen": 6176832,
      "step": 9475
    },
    {
      "epoch": 4.968553459119497,
      "grad_norm": 0.05963625758886337,
      "learning_rate": 0.0009344009363707186,
      "loss": 0.4619,
      "num_input_tokens_seen": 6179392,
      "step": 9480
    },
    {
      "epoch": 4.971174004192872,
      "grad_norm": 0.041338399052619934,
      "learning_rate": 0.000934287654955906,
      "loss": 0.4093,
      "num_input_tokens_seen": 6182208,
      "step": 9485
    },
    {
      "epoch": 4.973794549266247,
      "grad_norm": 0.07607361674308777,
      "learning_rate": 0.0009341742826930708,
      "loss": 0.5652,
      "num_input_tokens_seen": 6185376,
      "step": 9490
    },
    {
      "epoch": 4.976415094339623,
      "grad_norm": 0.0441717803478241,
      "learning_rate": 0.0009340608196059289,
      "loss": 0.3707,
      "num_input_tokens_seen": 6188896,
      "step": 9495
    },
    {
      "epoch": 4.979035639412998,
      "grad_norm": 0.08999891579151154,
      "learning_rate": 0.0009339472657182155,
      "loss": 0.4863,
      "num_input_tokens_seen": 6191840,
      "step": 9500
    },
    {
      "epoch": 4.981656184486373,
      "grad_norm": 0.06879865378141403,
      "learning_rate": 0.0009338336210536848,
      "loss": 0.4113,
      "num_input_tokens_seen": 6194784,
      "step": 9505
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.10070513188838959,
      "learning_rate": 0.0009337198856361102,
      "loss": 0.362,
      "num_input_tokens_seen": 6198048,
      "step": 9510
    },
    {
      "epoch": 4.986897274633124,
      "grad_norm": 0.1586475670337677,
      "learning_rate": 0.0009336060594892834,
      "loss": 0.4549,
      "num_input_tokens_seen": 6202688,
      "step": 9515
    },
    {
      "epoch": 4.989517819706499,
      "grad_norm": 0.0584392249584198,
      "learning_rate": 0.000933492142637016,
      "loss": 0.6448,
      "num_input_tokens_seen": 6205952,
      "step": 9520
    },
    {
      "epoch": 4.9921383647798745,
      "grad_norm": 0.055489808320999146,
      "learning_rate": 0.0009333781351031379,
      "loss": 0.3792,
      "num_input_tokens_seen": 6209504,
      "step": 9525
    },
    {
      "epoch": 4.99475890985325,
      "grad_norm": 0.0630263239145279,
      "learning_rate": 0.0009332640369114981,
      "loss": 0.5207,
      "num_input_tokens_seen": 6212928,
      "step": 9530
    },
    {
      "epoch": 4.997379454926625,
      "grad_norm": 0.0810672715306282,
      "learning_rate": 0.0009331498480859647,
      "loss": 0.4182,
      "num_input_tokens_seen": 6216672,
      "step": 9535
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.19640439748764038,
      "learning_rate": 0.0009330355686504247,
      "loss": 0.5856,
      "num_input_tokens_seen": 6219592,
      "step": 9540
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.4563165307044983,
      "eval_runtime": 13.6348,
      "eval_samples_per_second": 62.194,
      "eval_steps_per_second": 15.548,
      "num_input_tokens_seen": 6219592,
      "step": 9540
    },
    {
      "epoch": 5.002620545073375,
      "grad_norm": 0.11156320571899414,
      "learning_rate": 0.0009329211986287842,
      "loss": 0.4495,
      "num_input_tokens_seen": 6222216,
      "step": 9545
    },
    {
      "epoch": 5.00524109014675,
      "grad_norm": 0.0717056542634964,
      "learning_rate": 0.0009328067380449678,
      "loss": 0.4668,
      "num_input_tokens_seen": 6224872,
      "step": 9550
    },
    {
      "epoch": 5.0078616352201255,
      "grad_norm": 0.059935931116342545,
      "learning_rate": 0.0009326921869229197,
      "loss": 0.4595,
      "num_input_tokens_seen": 6228072,
      "step": 9555
    },
    {
      "epoch": 5.010482180293501,
      "grad_norm": 0.07761812955141068,
      "learning_rate": 0.0009325775452866023,
      "loss": 0.5466,
      "num_input_tokens_seen": 6230984,
      "step": 9560
    },
    {
      "epoch": 5.013102725366877,
      "grad_norm": 0.07515457272529602,
      "learning_rate": 0.0009324628131599977,
      "loss": 0.5312,
      "num_input_tokens_seen": 6234056,
      "step": 9565
    },
    {
      "epoch": 5.015723270440252,
      "grad_norm": 0.052474722266197205,
      "learning_rate": 0.0009323479905671064,
      "loss": 0.3146,
      "num_input_tokens_seen": 6237800,
      "step": 9570
    },
    {
      "epoch": 5.018343815513627,
      "grad_norm": 0.08264961838722229,
      "learning_rate": 0.0009322330775319479,
      "loss": 0.6394,
      "num_input_tokens_seen": 6240808,
      "step": 9575
    },
    {
      "epoch": 5.020964360587002,
      "grad_norm": 0.09706992655992508,
      "learning_rate": 0.0009321180740785607,
      "loss": 0.5262,
      "num_input_tokens_seen": 6244040,
      "step": 9580
    },
    {
      "epoch": 5.023584905660377,
      "grad_norm": 0.06960109621286392,
      "learning_rate": 0.0009320029802310023,
      "loss": 0.5418,
      "num_input_tokens_seen": 6246952,
      "step": 9585
    },
    {
      "epoch": 5.026205450733753,
      "grad_norm": 0.04873867705464363,
      "learning_rate": 0.0009318877960133489,
      "loss": 0.4981,
      "num_input_tokens_seen": 6250472,
      "step": 9590
    },
    {
      "epoch": 5.028825995807128,
      "grad_norm": 0.05446445569396019,
      "learning_rate": 0.0009317725214496959,
      "loss": 0.428,
      "num_input_tokens_seen": 6253640,
      "step": 9595
    },
    {
      "epoch": 5.031446540880503,
      "grad_norm": 0.09143280982971191,
      "learning_rate": 0.0009316571565641574,
      "loss": 0.4619,
      "num_input_tokens_seen": 6257160,
      "step": 9600
    },
    {
      "epoch": 5.034067085953878,
      "grad_norm": 0.06664638221263885,
      "learning_rate": 0.0009315417013808663,
      "loss": 0.4198,
      "num_input_tokens_seen": 6259688,
      "step": 9605
    },
    {
      "epoch": 5.036687631027253,
      "grad_norm": 0.07162821292877197,
      "learning_rate": 0.0009314261559239745,
      "loss": 0.444,
      "num_input_tokens_seen": 6263080,
      "step": 9610
    },
    {
      "epoch": 5.039308176100629,
      "grad_norm": 0.04780008643865585,
      "learning_rate": 0.000931310520217653,
      "loss": 0.2929,
      "num_input_tokens_seen": 6269320,
      "step": 9615
    },
    {
      "epoch": 5.0419287211740045,
      "grad_norm": 0.07592801749706268,
      "learning_rate": 0.0009311947942860912,
      "loss": 0.4562,
      "num_input_tokens_seen": 6272392,
      "step": 9620
    },
    {
      "epoch": 5.04454926624738,
      "grad_norm": 0.08538684248924255,
      "learning_rate": 0.0009310789781534979,
      "loss": 0.5658,
      "num_input_tokens_seen": 6275496,
      "step": 9625
    },
    {
      "epoch": 5.047169811320755,
      "grad_norm": 0.07854261994361877,
      "learning_rate": 0.0009309630718441003,
      "loss": 0.4048,
      "num_input_tokens_seen": 6277960,
      "step": 9630
    },
    {
      "epoch": 5.04979035639413,
      "grad_norm": 0.045498788356781006,
      "learning_rate": 0.0009308470753821445,
      "loss": 0.8634,
      "num_input_tokens_seen": 6280520,
      "step": 9635
    },
    {
      "epoch": 5.052410901467505,
      "grad_norm": 0.09755119681358337,
      "learning_rate": 0.0009307309887918962,
      "loss": 0.4665,
      "num_input_tokens_seen": 6283304,
      "step": 9640
    },
    {
      "epoch": 5.05503144654088,
      "grad_norm": 0.13527223467826843,
      "learning_rate": 0.000930614812097639,
      "loss": 0.5956,
      "num_input_tokens_seen": 6286536,
      "step": 9645
    },
    {
      "epoch": 5.0576519916142555,
      "grad_norm": 0.052103035151958466,
      "learning_rate": 0.0009304985453236758,
      "loss": 0.4682,
      "num_input_tokens_seen": 6289928,
      "step": 9650
    },
    {
      "epoch": 5.060272536687631,
      "grad_norm": 0.0661732479929924,
      "learning_rate": 0.0009303821884943284,
      "loss": 0.5219,
      "num_input_tokens_seen": 6293416,
      "step": 9655
    },
    {
      "epoch": 5.062893081761007,
      "grad_norm": 0.062492284923791885,
      "learning_rate": 0.0009302657416339371,
      "loss": 0.5118,
      "num_input_tokens_seen": 6297320,
      "step": 9660
    },
    {
      "epoch": 5.065513626834382,
      "grad_norm": 0.059056635946035385,
      "learning_rate": 0.0009301492047668615,
      "loss": 0.4236,
      "num_input_tokens_seen": 6300008,
      "step": 9665
    },
    {
      "epoch": 5.068134171907757,
      "grad_norm": 0.03353467583656311,
      "learning_rate": 0.0009300325779174796,
      "loss": 0.4153,
      "num_input_tokens_seen": 6304136,
      "step": 9670
    },
    {
      "epoch": 5.070754716981132,
      "grad_norm": 0.08251230418682098,
      "learning_rate": 0.0009299158611101885,
      "loss": 0.5683,
      "num_input_tokens_seen": 6306792,
      "step": 9675
    },
    {
      "epoch": 5.073375262054507,
      "grad_norm": 0.08074327558279037,
      "learning_rate": 0.000929799054369404,
      "loss": 0.529,
      "num_input_tokens_seen": 6309800,
      "step": 9680
    },
    {
      "epoch": 5.075995807127883,
      "grad_norm": 0.09930624812841415,
      "learning_rate": 0.0009296821577195606,
      "loss": 0.5332,
      "num_input_tokens_seen": 6312360,
      "step": 9685
    },
    {
      "epoch": 5.078616352201258,
      "grad_norm": 0.05106725916266441,
      "learning_rate": 0.000929565171185112,
      "loss": 0.516,
      "num_input_tokens_seen": 6316008,
      "step": 9690
    },
    {
      "epoch": 5.081236897274633,
      "grad_norm": 0.06080810725688934,
      "learning_rate": 0.0009294480947905304,
      "loss": 0.4436,
      "num_input_tokens_seen": 6318952,
      "step": 9695
    },
    {
      "epoch": 5.083857442348008,
      "grad_norm": 0.09858501702547073,
      "learning_rate": 0.0009293309285603067,
      "loss": 0.4908,
      "num_input_tokens_seen": 6322056,
      "step": 9700
    },
    {
      "epoch": 5.086477987421383,
      "grad_norm": 0.0850195437669754,
      "learning_rate": 0.0009292136725189506,
      "loss": 0.4345,
      "num_input_tokens_seen": 6325000,
      "step": 9705
    },
    {
      "epoch": 5.089098532494759,
      "grad_norm": 0.07535483688116074,
      "learning_rate": 0.0009290963266909912,
      "loss": 0.4089,
      "num_input_tokens_seen": 6329960,
      "step": 9710
    },
    {
      "epoch": 5.0917190775681345,
      "grad_norm": 0.0912756621837616,
      "learning_rate": 0.0009289788911009756,
      "loss": 0.3689,
      "num_input_tokens_seen": 6333640,
      "step": 9715
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 0.15934620797634125,
      "learning_rate": 0.0009288613657734699,
      "loss": 0.3953,
      "num_input_tokens_seen": 6336488,
      "step": 9720
    },
    {
      "epoch": 5.096960167714885,
      "grad_norm": 0.09577494114637375,
      "learning_rate": 0.0009287437507330594,
      "loss": 0.4734,
      "num_input_tokens_seen": 6340328,
      "step": 9725
    },
    {
      "epoch": 5.09958071278826,
      "grad_norm": 0.05751638114452362,
      "learning_rate": 0.0009286260460043475,
      "loss": 0.5456,
      "num_input_tokens_seen": 6343464,
      "step": 9730
    },
    {
      "epoch": 5.102201257861635,
      "grad_norm": 0.08888912200927734,
      "learning_rate": 0.0009285082516119567,
      "loss": 0.6591,
      "num_input_tokens_seen": 6349352,
      "step": 9735
    },
    {
      "epoch": 5.10482180293501,
      "grad_norm": 0.060160595923662186,
      "learning_rate": 0.0009283903675805285,
      "loss": 0.3976,
      "num_input_tokens_seen": 6351784,
      "step": 9740
    },
    {
      "epoch": 5.1074423480083855,
      "grad_norm": 0.07346680015325546,
      "learning_rate": 0.0009282723939347227,
      "loss": 0.6016,
      "num_input_tokens_seen": 6355368,
      "step": 9745
    },
    {
      "epoch": 5.110062893081761,
      "grad_norm": 0.07224461436271667,
      "learning_rate": 0.0009281543306992181,
      "loss": 0.5144,
      "num_input_tokens_seen": 6358312,
      "step": 9750
    },
    {
      "epoch": 5.112683438155136,
      "grad_norm": 0.10134021937847137,
      "learning_rate": 0.0009280361778987121,
      "loss": 0.4948,
      "num_input_tokens_seen": 6360776,
      "step": 9755
    },
    {
      "epoch": 5.115303983228512,
      "grad_norm": 0.05972234532237053,
      "learning_rate": 0.0009279179355579211,
      "loss": 0.5832,
      "num_input_tokens_seen": 6364424,
      "step": 9760
    },
    {
      "epoch": 5.117924528301887,
      "grad_norm": 0.06490366160869598,
      "learning_rate": 0.0009277996037015798,
      "loss": 0.4463,
      "num_input_tokens_seen": 6367496,
      "step": 9765
    },
    {
      "epoch": 5.120545073375262,
      "grad_norm": 0.12065612524747849,
      "learning_rate": 0.000927681182354442,
      "loss": 0.4318,
      "num_input_tokens_seen": 6370472,
      "step": 9770
    },
    {
      "epoch": 5.123165618448637,
      "grad_norm": 0.0589052177965641,
      "learning_rate": 0.0009275626715412802,
      "loss": 0.3693,
      "num_input_tokens_seen": 6373832,
      "step": 9775
    },
    {
      "epoch": 5.1257861635220126,
      "grad_norm": 0.05781720206141472,
      "learning_rate": 0.0009274440712868853,
      "loss": 0.4469,
      "num_input_tokens_seen": 6377672,
      "step": 9780
    },
    {
      "epoch": 5.128406708595388,
      "grad_norm": 0.09483715146780014,
      "learning_rate": 0.0009273253816160673,
      "loss": 0.4597,
      "num_input_tokens_seen": 6380712,
      "step": 9785
    },
    {
      "epoch": 5.131027253668763,
      "grad_norm": 0.11279866844415665,
      "learning_rate": 0.0009272066025536545,
      "loss": 0.4846,
      "num_input_tokens_seen": 6384776,
      "step": 9790
    },
    {
      "epoch": 5.133647798742138,
      "grad_norm": 0.10880224406719208,
      "learning_rate": 0.0009270877341244945,
      "loss": 0.4885,
      "num_input_tokens_seen": 6387560,
      "step": 9795
    },
    {
      "epoch": 5.136268343815513,
      "grad_norm": 0.08741100132465363,
      "learning_rate": 0.0009269687763534529,
      "loss": 0.601,
      "num_input_tokens_seen": 6390952,
      "step": 9800
    },
    {
      "epoch": 5.138888888888889,
      "grad_norm": 0.04659738764166832,
      "learning_rate": 0.0009268497292654143,
      "loss": 0.5004,
      "num_input_tokens_seen": 6394088,
      "step": 9805
    },
    {
      "epoch": 5.1415094339622645,
      "grad_norm": 0.042011961340904236,
      "learning_rate": 0.0009267305928852823,
      "loss": 0.3887,
      "num_input_tokens_seen": 6397064,
      "step": 9810
    },
    {
      "epoch": 5.14412997903564,
      "grad_norm": 0.09488141536712646,
      "learning_rate": 0.0009266113672379786,
      "loss": 0.6636,
      "num_input_tokens_seen": 6400584,
      "step": 9815
    },
    {
      "epoch": 5.146750524109015,
      "grad_norm": 0.07033082842826843,
      "learning_rate": 0.0009264920523484437,
      "loss": 0.3665,
      "num_input_tokens_seen": 6403976,
      "step": 9820
    },
    {
      "epoch": 5.14937106918239,
      "grad_norm": 0.08656206727027893,
      "learning_rate": 0.0009263726482416374,
      "loss": 0.4856,
      "num_input_tokens_seen": 6406536,
      "step": 9825
    },
    {
      "epoch": 5.151991614255765,
      "grad_norm": 0.06147002428770065,
      "learning_rate": 0.0009262531549425372,
      "loss": 0.4755,
      "num_input_tokens_seen": 6410312,
      "step": 9830
    },
    {
      "epoch": 5.15461215932914,
      "grad_norm": 0.10064250230789185,
      "learning_rate": 0.0009261335724761402,
      "loss": 0.4242,
      "num_input_tokens_seen": 6414184,
      "step": 9835
    },
    {
      "epoch": 5.1572327044025155,
      "grad_norm": 0.07135164737701416,
      "learning_rate": 0.0009260139008674612,
      "loss": 0.6985,
      "num_input_tokens_seen": 6417224,
      "step": 9840
    },
    {
      "epoch": 5.159853249475891,
      "grad_norm": 0.051722537726163864,
      "learning_rate": 0.0009258941401415344,
      "loss": 0.558,
      "num_input_tokens_seen": 6421128,
      "step": 9845
    },
    {
      "epoch": 5.162473794549266,
      "grad_norm": 0.06570906192064285,
      "learning_rate": 0.0009257742903234123,
      "loss": 0.4453,
      "num_input_tokens_seen": 6424232,
      "step": 9850
    },
    {
      "epoch": 5.165094339622642,
      "grad_norm": 0.07978320866823196,
      "learning_rate": 0.0009256543514381664,
      "loss": 0.4519,
      "num_input_tokens_seen": 6427464,
      "step": 9855
    },
    {
      "epoch": 5.167714884696017,
      "grad_norm": 0.06261777877807617,
      "learning_rate": 0.0009255343235108859,
      "loss": 0.4032,
      "num_input_tokens_seen": 6430504,
      "step": 9860
    },
    {
      "epoch": 5.170335429769392,
      "grad_norm": 0.06978512555360794,
      "learning_rate": 0.00092541420656668,
      "loss": 0.5041,
      "num_input_tokens_seen": 6434056,
      "step": 9865
    },
    {
      "epoch": 5.172955974842767,
      "grad_norm": 0.07397482544183731,
      "learning_rate": 0.0009252940006306753,
      "loss": 0.6509,
      "num_input_tokens_seen": 6437256,
      "step": 9870
    },
    {
      "epoch": 5.1755765199161425,
      "grad_norm": 0.0635354146361351,
      "learning_rate": 0.0009251737057280179,
      "loss": 0.474,
      "num_input_tokens_seen": 6441064,
      "step": 9875
    },
    {
      "epoch": 5.178197064989518,
      "grad_norm": 0.05341167375445366,
      "learning_rate": 0.0009250533218838717,
      "loss": 0.4038,
      "num_input_tokens_seen": 6444648,
      "step": 9880
    },
    {
      "epoch": 5.180817610062893,
      "grad_norm": 0.06893229484558105,
      "learning_rate": 0.0009249328491234199,
      "loss": 0.5103,
      "num_input_tokens_seen": 6447880,
      "step": 9885
    },
    {
      "epoch": 5.183438155136268,
      "grad_norm": 0.04019537940621376,
      "learning_rate": 0.0009248122874718638,
      "loss": 0.4401,
      "num_input_tokens_seen": 6450728,
      "step": 9890
    },
    {
      "epoch": 5.186058700209643,
      "grad_norm": 0.06495088338851929,
      "learning_rate": 0.0009246916369544238,
      "loss": 0.3819,
      "num_input_tokens_seen": 6453640,
      "step": 9895
    },
    {
      "epoch": 5.188679245283019,
      "grad_norm": 0.07405657321214676,
      "learning_rate": 0.0009245708975963386,
      "loss": 0.3705,
      "num_input_tokens_seen": 6456840,
      "step": 9900
    },
    {
      "epoch": 5.191299790356394,
      "grad_norm": 0.0847579687833786,
      "learning_rate": 0.0009244500694228653,
      "loss": 0.4282,
      "num_input_tokens_seen": 6459848,
      "step": 9905
    },
    {
      "epoch": 5.19392033542977,
      "grad_norm": 0.08485935628414154,
      "learning_rate": 0.0009243291524592799,
      "loss": 0.5484,
      "num_input_tokens_seen": 6463624,
      "step": 9910
    },
    {
      "epoch": 5.196540880503145,
      "grad_norm": 0.05587204173207283,
      "learning_rate": 0.0009242081467308766,
      "loss": 0.411,
      "num_input_tokens_seen": 6467176,
      "step": 9915
    },
    {
      "epoch": 5.19916142557652,
      "grad_norm": 0.08334653079509735,
      "learning_rate": 0.0009240870522629688,
      "loss": 0.4835,
      "num_input_tokens_seen": 6469896,
      "step": 9920
    },
    {
      "epoch": 5.201781970649895,
      "grad_norm": 0.06982061266899109,
      "learning_rate": 0.0009239658690808879,
      "loss": 0.5816,
      "num_input_tokens_seen": 6472904,
      "step": 9925
    },
    {
      "epoch": 5.20440251572327,
      "grad_norm": 0.07432735711336136,
      "learning_rate": 0.000923844597209984,
      "loss": 0.5077,
      "num_input_tokens_seen": 6477288,
      "step": 9930
    },
    {
      "epoch": 5.2070230607966455,
      "grad_norm": 0.06070728972554207,
      "learning_rate": 0.0009237232366756258,
      "loss": 0.4665,
      "num_input_tokens_seen": 6480488,
      "step": 9935
    },
    {
      "epoch": 5.209643605870021,
      "grad_norm": 0.10146468132734299,
      "learning_rate": 0.0009236017875032007,
      "loss": 0.3621,
      "num_input_tokens_seen": 6483368,
      "step": 9940
    },
    {
      "epoch": 5.212264150943396,
      "grad_norm": 0.13680024445056915,
      "learning_rate": 0.0009234802497181143,
      "loss": 0.5165,
      "num_input_tokens_seen": 6485928,
      "step": 9945
    },
    {
      "epoch": 5.214884696016772,
      "grad_norm": 0.06026388704776764,
      "learning_rate": 0.0009233586233457909,
      "loss": 0.5012,
      "num_input_tokens_seen": 6489768,
      "step": 9950
    },
    {
      "epoch": 5.217505241090147,
      "grad_norm": 0.09817184507846832,
      "learning_rate": 0.0009232369084116736,
      "loss": 0.4877,
      "num_input_tokens_seen": 6493352,
      "step": 9955
    },
    {
      "epoch": 5.220125786163522,
      "grad_norm": 0.06805992871522903,
      "learning_rate": 0.0009231151049412234,
      "loss": 0.4799,
      "num_input_tokens_seen": 6495944,
      "step": 9960
    },
    {
      "epoch": 5.222746331236897,
      "grad_norm": 0.06663099676370621,
      "learning_rate": 0.0009229932129599205,
      "loss": 0.4407,
      "num_input_tokens_seen": 6498472,
      "step": 9965
    },
    {
      "epoch": 5.2253668763102725,
      "grad_norm": 0.12045582383871078,
      "learning_rate": 0.0009228712324932634,
      "loss": 0.5053,
      "num_input_tokens_seen": 6502088,
      "step": 9970
    },
    {
      "epoch": 5.227987421383648,
      "grad_norm": 0.07292971014976501,
      "learning_rate": 0.0009227491635667685,
      "loss": 0.3549,
      "num_input_tokens_seen": 6505032,
      "step": 9975
    },
    {
      "epoch": 5.230607966457023,
      "grad_norm": 0.10192679613828659,
      "learning_rate": 0.0009226270062059717,
      "loss": 0.3512,
      "num_input_tokens_seen": 6508488,
      "step": 9980
    },
    {
      "epoch": 5.233228511530398,
      "grad_norm": 0.09526249766349792,
      "learning_rate": 0.0009225047604364267,
      "loss": 0.4555,
      "num_input_tokens_seen": 6510984,
      "step": 9985
    },
    {
      "epoch": 5.235849056603773,
      "grad_norm": 0.07802938669919968,
      "learning_rate": 0.0009223824262837062,
      "loss": 0.543,
      "num_input_tokens_seen": 6514376,
      "step": 9990
    },
    {
      "epoch": 5.238469601677149,
      "grad_norm": 0.08925095945596695,
      "learning_rate": 0.0009222600037734008,
      "loss": 0.5671,
      "num_input_tokens_seen": 6517192,
      "step": 9995
    },
    {
      "epoch": 5.241090146750524,
      "grad_norm": 0.07034717500209808,
      "learning_rate": 0.00092213749293112,
      "loss": 0.3985,
      "num_input_tokens_seen": 6520424,
      "step": 10000
    },
    {
      "epoch": 5.2437106918239,
      "grad_norm": 0.04395890608429909,
      "learning_rate": 0.0009220148937824917,
      "loss": 0.2486,
      "num_input_tokens_seen": 6523976,
      "step": 10005
    },
    {
      "epoch": 5.246331236897275,
      "grad_norm": 0.1331944316625595,
      "learning_rate": 0.0009218922063531623,
      "loss": 0.3861,
      "num_input_tokens_seen": 6526408,
      "step": 10010
    },
    {
      "epoch": 5.24895178197065,
      "grad_norm": 0.10098132491111755,
      "learning_rate": 0.0009217694306687963,
      "loss": 0.4884,
      "num_input_tokens_seen": 6529288,
      "step": 10015
    },
    {
      "epoch": 5.251572327044025,
      "grad_norm": 0.09716115146875381,
      "learning_rate": 0.0009216465667550774,
      "loss": 0.4861,
      "num_input_tokens_seen": 6532072,
      "step": 10020
    },
    {
      "epoch": 5.2541928721174,
      "grad_norm": 0.055884551256895065,
      "learning_rate": 0.0009215236146377071,
      "loss": 0.4291,
      "num_input_tokens_seen": 6536520,
      "step": 10025
    },
    {
      "epoch": 5.256813417190775,
      "grad_norm": 0.12153290212154388,
      "learning_rate": 0.0009214005743424056,
      "loss": 0.504,
      "num_input_tokens_seen": 6539880,
      "step": 10030
    },
    {
      "epoch": 5.259433962264151,
      "grad_norm": 0.1003350019454956,
      "learning_rate": 0.0009212774458949116,
      "loss": 0.472,
      "num_input_tokens_seen": 6542632,
      "step": 10035
    },
    {
      "epoch": 5.262054507337526,
      "grad_norm": 0.10136362165212631,
      "learning_rate": 0.000921154229320982,
      "loss": 0.4172,
      "num_input_tokens_seen": 6545416,
      "step": 10040
    },
    {
      "epoch": 5.264675052410902,
      "grad_norm": 0.03964899480342865,
      "learning_rate": 0.0009210309246463924,
      "loss": 0.3985,
      "num_input_tokens_seen": 6548680,
      "step": 10045
    },
    {
      "epoch": 5.267295597484277,
      "grad_norm": 0.05598973110318184,
      "learning_rate": 0.0009209075318969369,
      "loss": 0.4183,
      "num_input_tokens_seen": 6552744,
      "step": 10050
    },
    {
      "epoch": 5.269916142557652,
      "grad_norm": 0.0726013258099556,
      "learning_rate": 0.0009207840510984276,
      "loss": 0.381,
      "num_input_tokens_seen": 6557128,
      "step": 10055
    },
    {
      "epoch": 5.272536687631027,
      "grad_norm": 0.06299015134572983,
      "learning_rate": 0.0009206604822766953,
      "loss": 0.4698,
      "num_input_tokens_seen": 6565352,
      "step": 10060
    },
    {
      "epoch": 5.2751572327044025,
      "grad_norm": 0.11088854819536209,
      "learning_rate": 0.0009205368254575892,
      "loss": 0.4991,
      "num_input_tokens_seen": 6568680,
      "step": 10065
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 0.07280931621789932,
      "learning_rate": 0.000920413080666977,
      "loss": 0.4317,
      "num_input_tokens_seen": 6572520,
      "step": 10070
    },
    {
      "epoch": 5.280398322851153,
      "grad_norm": 0.057300686836242676,
      "learning_rate": 0.0009202892479307448,
      "loss": 0.5117,
      "num_input_tokens_seen": 6576232,
      "step": 10075
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 0.12315306067466736,
      "learning_rate": 0.0009201653272747967,
      "loss": 0.4328,
      "num_input_tokens_seen": 6579016,
      "step": 10080
    },
    {
      "epoch": 5.285639412997903,
      "grad_norm": 0.13002437353134155,
      "learning_rate": 0.0009200413187250558,
      "loss": 0.4604,
      "num_input_tokens_seen": 6581864,
      "step": 10085
    },
    {
      "epoch": 5.288259958071279,
      "grad_norm": 0.04627642035484314,
      "learning_rate": 0.000919917222307463,
      "loss": 0.5584,
      "num_input_tokens_seen": 6585256,
      "step": 10090
    },
    {
      "epoch": 5.290880503144654,
      "grad_norm": 0.050851695239543915,
      "learning_rate": 0.000919793038047978,
      "loss": 0.5544,
      "num_input_tokens_seen": 6587976,
      "step": 10095
    },
    {
      "epoch": 5.29350104821803,
      "grad_norm": 0.04613266512751579,
      "learning_rate": 0.0009196687659725787,
      "loss": 0.6123,
      "num_input_tokens_seen": 6591080,
      "step": 10100
    },
    {
      "epoch": 5.296121593291405,
      "grad_norm": 0.07291537523269653,
      "learning_rate": 0.0009195444061072612,
      "loss": 0.5077,
      "num_input_tokens_seen": 6594088,
      "step": 10105
    },
    {
      "epoch": 5.29874213836478,
      "grad_norm": 0.06749574095010757,
      "learning_rate": 0.0009194199584780405,
      "loss": 0.3958,
      "num_input_tokens_seen": 6597544,
      "step": 10110
    },
    {
      "epoch": 5.301362683438155,
      "grad_norm": 0.08000416308641434,
      "learning_rate": 0.0009192954231109496,
      "loss": 0.4619,
      "num_input_tokens_seen": 6600552,
      "step": 10115
    },
    {
      "epoch": 5.30398322851153,
      "grad_norm": 0.13461005687713623,
      "learning_rate": 0.0009191708000320396,
      "loss": 0.5052,
      "num_input_tokens_seen": 6602888,
      "step": 10120
    },
    {
      "epoch": 5.306603773584905,
      "grad_norm": 0.058422576636075974,
      "learning_rate": 0.0009190460892673805,
      "loss": 0.4318,
      "num_input_tokens_seen": 6606344,
      "step": 10125
    },
    {
      "epoch": 5.309224318658281,
      "grad_norm": 0.07449331879615784,
      "learning_rate": 0.0009189212908430601,
      "loss": 0.5541,
      "num_input_tokens_seen": 6609864,
      "step": 10130
    },
    {
      "epoch": 5.311844863731656,
      "grad_norm": 0.08293148875236511,
      "learning_rate": 0.0009187964047851851,
      "loss": 0.3992,
      "num_input_tokens_seen": 6612616,
      "step": 10135
    },
    {
      "epoch": 5.314465408805032,
      "grad_norm": 0.08640771359205246,
      "learning_rate": 0.0009186714311198801,
      "loss": 0.41,
      "num_input_tokens_seen": 6616072,
      "step": 10140
    },
    {
      "epoch": 5.317085953878407,
      "grad_norm": 0.06629449874162674,
      "learning_rate": 0.000918546369873288,
      "loss": 0.6494,
      "num_input_tokens_seen": 6619528,
      "step": 10145
    },
    {
      "epoch": 5.319706498951782,
      "grad_norm": 0.07774560898542404,
      "learning_rate": 0.0009184212210715704,
      "loss": 0.4586,
      "num_input_tokens_seen": 6622824,
      "step": 10150
    },
    {
      "epoch": 5.322327044025157,
      "grad_norm": 0.1294352114200592,
      "learning_rate": 0.0009182959847409072,
      "loss": 0.5121,
      "num_input_tokens_seen": 6626152,
      "step": 10155
    },
    {
      "epoch": 5.3249475890985325,
      "grad_norm": 0.06282245367765427,
      "learning_rate": 0.0009181706609074959,
      "loss": 0.4909,
      "num_input_tokens_seen": 6628808,
      "step": 10160
    },
    {
      "epoch": 5.327568134171908,
      "grad_norm": 0.08316423743963242,
      "learning_rate": 0.0009180452495975531,
      "loss": 0.4692,
      "num_input_tokens_seen": 6632296,
      "step": 10165
    },
    {
      "epoch": 5.330188679245283,
      "grad_norm": 0.06791456043720245,
      "learning_rate": 0.0009179197508373134,
      "loss": 0.5079,
      "num_input_tokens_seen": 6635400,
      "step": 10170
    },
    {
      "epoch": 5.332809224318658,
      "grad_norm": 0.0714358240365982,
      "learning_rate": 0.0009177941646530299,
      "loss": 0.5115,
      "num_input_tokens_seen": 6639432,
      "step": 10175
    },
    {
      "epoch": 5.335429769392033,
      "grad_norm": 0.0601314902305603,
      "learning_rate": 0.0009176684910709733,
      "loss": 0.4979,
      "num_input_tokens_seen": 6643336,
      "step": 10180
    },
    {
      "epoch": 5.338050314465409,
      "grad_norm": 0.09850703924894333,
      "learning_rate": 0.0009175427301174338,
      "loss": 0.5011,
      "num_input_tokens_seen": 6646088,
      "step": 10185
    },
    {
      "epoch": 5.340670859538784,
      "grad_norm": 0.09449709206819534,
      "learning_rate": 0.0009174168818187183,
      "loss": 0.3204,
      "num_input_tokens_seen": 6649384,
      "step": 10190
    },
    {
      "epoch": 5.34329140461216,
      "grad_norm": 0.10796729475259781,
      "learning_rate": 0.0009172909462011536,
      "loss": 0.4406,
      "num_input_tokens_seen": 6652424,
      "step": 10195
    },
    {
      "epoch": 5.345911949685535,
      "grad_norm": 0.12951965630054474,
      "learning_rate": 0.0009171649232910835,
      "loss": 0.4578,
      "num_input_tokens_seen": 6655464,
      "step": 10200
    },
    {
      "epoch": 5.34853249475891,
      "grad_norm": 0.07012820988893509,
      "learning_rate": 0.0009170388131148707,
      "loss": 0.5131,
      "num_input_tokens_seen": 6659112,
      "step": 10205
    },
    {
      "epoch": 5.351153039832285,
      "grad_norm": 0.1278323382139206,
      "learning_rate": 0.000916912615698896,
      "loss": 0.5255,
      "num_input_tokens_seen": 6663304,
      "step": 10210
    },
    {
      "epoch": 5.35377358490566,
      "grad_norm": 0.06310364603996277,
      "learning_rate": 0.0009167863310695585,
      "loss": 0.3428,
      "num_input_tokens_seen": 6666888,
      "step": 10215
    },
    {
      "epoch": 5.356394129979035,
      "grad_norm": 0.08095693588256836,
      "learning_rate": 0.0009166599592532756,
      "loss": 0.445,
      "num_input_tokens_seen": 6670504,
      "step": 10220
    },
    {
      "epoch": 5.359014675052411,
      "grad_norm": 0.052475687116384506,
      "learning_rate": 0.0009165335002764828,
      "loss": 0.4037,
      "num_input_tokens_seen": 6673832,
      "step": 10225
    },
    {
      "epoch": 5.361635220125786,
      "grad_norm": 0.11861336976289749,
      "learning_rate": 0.0009164069541656337,
      "loss": 0.4534,
      "num_input_tokens_seen": 6677160,
      "step": 10230
    },
    {
      "epoch": 5.364255765199162,
      "grad_norm": 0.04447439685463905,
      "learning_rate": 0.0009162803209472004,
      "loss": 0.4561,
      "num_input_tokens_seen": 6681160,
      "step": 10235
    },
    {
      "epoch": 5.366876310272537,
      "grad_norm": 0.07053130865097046,
      "learning_rate": 0.000916153600647673,
      "loss": 0.4473,
      "num_input_tokens_seen": 6684424,
      "step": 10240
    },
    {
      "epoch": 5.369496855345912,
      "grad_norm": 0.08325928449630737,
      "learning_rate": 0.0009160267932935602,
      "loss": 0.4511,
      "num_input_tokens_seen": 6687528,
      "step": 10245
    },
    {
      "epoch": 5.372117400419287,
      "grad_norm": 0.14910943806171417,
      "learning_rate": 0.0009158998989113885,
      "loss": 0.5175,
      "num_input_tokens_seen": 6690120,
      "step": 10250
    },
    {
      "epoch": 5.3747379454926625,
      "grad_norm": 0.14803481101989746,
      "learning_rate": 0.0009157729175277028,
      "loss": 0.5177,
      "num_input_tokens_seen": 6693992,
      "step": 10255
    },
    {
      "epoch": 5.377358490566038,
      "grad_norm": 0.10010114312171936,
      "learning_rate": 0.0009156458491690662,
      "loss": 0.4517,
      "num_input_tokens_seen": 6698568,
      "step": 10260
    },
    {
      "epoch": 5.379979035639413,
      "grad_norm": 0.06479546427726746,
      "learning_rate": 0.0009155186938620599,
      "loss": 0.4792,
      "num_input_tokens_seen": 6701320,
      "step": 10265
    },
    {
      "epoch": 5.382599580712788,
      "grad_norm": 0.10782650858163834,
      "learning_rate": 0.0009153914516332833,
      "loss": 0.4477,
      "num_input_tokens_seen": 6704168,
      "step": 10270
    },
    {
      "epoch": 5.385220125786163,
      "grad_norm": 0.09846757352352142,
      "learning_rate": 0.000915264122509354,
      "loss": 0.4924,
      "num_input_tokens_seen": 6707688,
      "step": 10275
    },
    {
      "epoch": 5.387840670859539,
      "grad_norm": 0.13731050491333008,
      "learning_rate": 0.0009151367065169078,
      "loss": 0.43,
      "num_input_tokens_seen": 6711048,
      "step": 10280
    },
    {
      "epoch": 5.390461215932914,
      "grad_norm": 0.2936151325702667,
      "learning_rate": 0.0009150092036825989,
      "loss": 0.4936,
      "num_input_tokens_seen": 6715016,
      "step": 10285
    },
    {
      "epoch": 5.3930817610062896,
      "grad_norm": 0.1277569681406021,
      "learning_rate": 0.0009148816140330991,
      "loss": 0.4604,
      "num_input_tokens_seen": 6717448,
      "step": 10290
    },
    {
      "epoch": 5.395702306079665,
      "grad_norm": 0.07491008192300797,
      "learning_rate": 0.000914753937595099,
      "loss": 0.3304,
      "num_input_tokens_seen": 6721032,
      "step": 10295
    },
    {
      "epoch": 5.39832285115304,
      "grad_norm": 0.09419108182191849,
      "learning_rate": 0.0009146261743953068,
      "loss": 0.4551,
      "num_input_tokens_seen": 6723304,
      "step": 10300
    },
    {
      "epoch": 5.400943396226415,
      "grad_norm": 0.06120631843805313,
      "learning_rate": 0.0009144983244604493,
      "loss": 0.4522,
      "num_input_tokens_seen": 6726344,
      "step": 10305
    },
    {
      "epoch": 5.40356394129979,
      "grad_norm": 0.04837194085121155,
      "learning_rate": 0.0009143703878172712,
      "loss": 0.4791,
      "num_input_tokens_seen": 6730248,
      "step": 10310
    },
    {
      "epoch": 5.406184486373165,
      "grad_norm": 0.06558210402727127,
      "learning_rate": 0.0009142423644925353,
      "loss": 0.5017,
      "num_input_tokens_seen": 6733352,
      "step": 10315
    },
    {
      "epoch": 5.408805031446541,
      "grad_norm": 0.1760401427745819,
      "learning_rate": 0.0009141142545130228,
      "loss": 0.3537,
      "num_input_tokens_seen": 6736104,
      "step": 10320
    },
    {
      "epoch": 5.411425576519916,
      "grad_norm": 0.05927221477031708,
      "learning_rate": 0.0009139860579055326,
      "loss": 0.4525,
      "num_input_tokens_seen": 6739336,
      "step": 10325
    },
    {
      "epoch": 5.414046121593292,
      "grad_norm": 0.0638372153043747,
      "learning_rate": 0.0009138577746968821,
      "loss": 0.4934,
      "num_input_tokens_seen": 6742216,
      "step": 10330
    },
    {
      "epoch": 5.416666666666667,
      "grad_norm": 0.06313929706811905,
      "learning_rate": 0.0009137294049139066,
      "loss": 0.3772,
      "num_input_tokens_seen": 6745320,
      "step": 10335
    },
    {
      "epoch": 5.419287211740042,
      "grad_norm": 0.07013420015573502,
      "learning_rate": 0.0009136009485834598,
      "loss": 0.3665,
      "num_input_tokens_seen": 6748328,
      "step": 10340
    },
    {
      "epoch": 5.421907756813417,
      "grad_norm": 0.07906433939933777,
      "learning_rate": 0.0009134724057324131,
      "loss": 0.4354,
      "num_input_tokens_seen": 6751240,
      "step": 10345
    },
    {
      "epoch": 5.4245283018867925,
      "grad_norm": 0.08213907480239868,
      "learning_rate": 0.0009133437763876562,
      "loss": 0.3864,
      "num_input_tokens_seen": 6754344,
      "step": 10350
    },
    {
      "epoch": 5.427148846960168,
      "grad_norm": 0.09790535271167755,
      "learning_rate": 0.000913215060576097,
      "loss": 0.4776,
      "num_input_tokens_seen": 6757320,
      "step": 10355
    },
    {
      "epoch": 5.429769392033543,
      "grad_norm": 0.07109267264604568,
      "learning_rate": 0.0009130862583246613,
      "loss": 0.3887,
      "num_input_tokens_seen": 6760744,
      "step": 10360
    },
    {
      "epoch": 5.432389937106918,
      "grad_norm": 0.05937211215496063,
      "learning_rate": 0.0009129573696602932,
      "loss": 0.5245,
      "num_input_tokens_seen": 6763848,
      "step": 10365
    },
    {
      "epoch": 5.435010482180293,
      "grad_norm": 0.10315272957086563,
      "learning_rate": 0.0009128283946099546,
      "loss": 0.5376,
      "num_input_tokens_seen": 6766792,
      "step": 10370
    },
    {
      "epoch": 5.437631027253669,
      "grad_norm": 0.08033576607704163,
      "learning_rate": 0.0009126993332006256,
      "loss": 0.4155,
      "num_input_tokens_seen": 6770024,
      "step": 10375
    },
    {
      "epoch": 5.440251572327044,
      "grad_norm": 0.08731776475906372,
      "learning_rate": 0.0009125701854593045,
      "loss": 0.4799,
      "num_input_tokens_seen": 6772584,
      "step": 10380
    },
    {
      "epoch": 5.4428721174004195,
      "grad_norm": 0.06682039052248001,
      "learning_rate": 0.0009124409514130073,
      "loss": 0.4307,
      "num_input_tokens_seen": 6776040,
      "step": 10385
    },
    {
      "epoch": 5.445492662473795,
      "grad_norm": 0.08214902132749557,
      "learning_rate": 0.0009123116310887686,
      "loss": 0.4464,
      "num_input_tokens_seen": 6780616,
      "step": 10390
    },
    {
      "epoch": 5.44811320754717,
      "grad_norm": 0.04882942885160446,
      "learning_rate": 0.0009121822245136404,
      "loss": 0.5346,
      "num_input_tokens_seen": 6783720,
      "step": 10395
    },
    {
      "epoch": 5.450733752620545,
      "grad_norm": 0.06982437521219254,
      "learning_rate": 0.0009120527317146934,
      "loss": 0.4638,
      "num_input_tokens_seen": 6786824,
      "step": 10400
    },
    {
      "epoch": 5.45335429769392,
      "grad_norm": 0.06024463474750519,
      "learning_rate": 0.0009119231527190158,
      "loss": 0.5359,
      "num_input_tokens_seen": 6790696,
      "step": 10405
    },
    {
      "epoch": 5.455974842767295,
      "grad_norm": 0.047284871339797974,
      "learning_rate": 0.0009117934875537141,
      "loss": 0.4444,
      "num_input_tokens_seen": 6794408,
      "step": 10410
    },
    {
      "epoch": 5.4585953878406706,
      "grad_norm": 0.0841098502278328,
      "learning_rate": 0.0009116637362459129,
      "loss": 0.4738,
      "num_input_tokens_seen": 6798184,
      "step": 10415
    },
    {
      "epoch": 5.461215932914046,
      "grad_norm": 0.0785265564918518,
      "learning_rate": 0.0009115338988227546,
      "loss": 0.4221,
      "num_input_tokens_seen": 6801928,
      "step": 10420
    },
    {
      "epoch": 5.463836477987422,
      "grad_norm": 0.06323014199733734,
      "learning_rate": 0.0009114039753113997,
      "loss": 0.4152,
      "num_input_tokens_seen": 6805640,
      "step": 10425
    },
    {
      "epoch": 5.466457023060797,
      "grad_norm": 0.06221801042556763,
      "learning_rate": 0.0009112739657390265,
      "loss": 0.3503,
      "num_input_tokens_seen": 6808680,
      "step": 10430
    },
    {
      "epoch": 5.469077568134172,
      "grad_norm": 0.05334554612636566,
      "learning_rate": 0.0009111438701328319,
      "loss": 0.4651,
      "num_input_tokens_seen": 6811752,
      "step": 10435
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 0.07761680334806442,
      "learning_rate": 0.0009110136885200302,
      "loss": 0.5068,
      "num_input_tokens_seen": 6814888,
      "step": 10440
    },
    {
      "epoch": 5.4743186582809225,
      "grad_norm": 0.045098237693309784,
      "learning_rate": 0.000910883420927854,
      "loss": 0.4504,
      "num_input_tokens_seen": 6819048,
      "step": 10445
    },
    {
      "epoch": 5.476939203354298,
      "grad_norm": 0.1584932804107666,
      "learning_rate": 0.0009107530673835536,
      "loss": 0.3449,
      "num_input_tokens_seen": 6822792,
      "step": 10450
    },
    {
      "epoch": 5.479559748427673,
      "grad_norm": 0.05096090957522392,
      "learning_rate": 0.0009106226279143979,
      "loss": 0.5606,
      "num_input_tokens_seen": 6826248,
      "step": 10455
    },
    {
      "epoch": 5.482180293501048,
      "grad_norm": 0.1170337051153183,
      "learning_rate": 0.0009104921025476728,
      "loss": 0.4173,
      "num_input_tokens_seen": 6830248,
      "step": 10460
    },
    {
      "epoch": 5.484800838574423,
      "grad_norm": 0.0627293512225151,
      "learning_rate": 0.0009103614913106832,
      "loss": 0.3953,
      "num_input_tokens_seen": 6833160,
      "step": 10465
    },
    {
      "epoch": 5.487421383647799,
      "grad_norm": 0.0918605774641037,
      "learning_rate": 0.0009102307942307511,
      "loss": 0.6306,
      "num_input_tokens_seen": 6836392,
      "step": 10470
    },
    {
      "epoch": 5.490041928721174,
      "grad_norm": 0.07679365575313568,
      "learning_rate": 0.000910100011335217,
      "loss": 0.4849,
      "num_input_tokens_seen": 6840872,
      "step": 10475
    },
    {
      "epoch": 5.4926624737945495,
      "grad_norm": 0.06384972482919693,
      "learning_rate": 0.0009099691426514392,
      "loss": 0.4711,
      "num_input_tokens_seen": 6843432,
      "step": 10480
    },
    {
      "epoch": 5.495283018867925,
      "grad_norm": 0.08546247333288193,
      "learning_rate": 0.0009098381882067941,
      "loss": 0.3772,
      "num_input_tokens_seen": 6846760,
      "step": 10485
    },
    {
      "epoch": 5.4979035639413,
      "grad_norm": 0.0854635089635849,
      "learning_rate": 0.0009097071480286756,
      "loss": 0.6064,
      "num_input_tokens_seen": 6849256,
      "step": 10490
    },
    {
      "epoch": 5.500524109014675,
      "grad_norm": 0.0748511552810669,
      "learning_rate": 0.0009095760221444959,
      "loss": 0.3687,
      "num_input_tokens_seen": 6853256,
      "step": 10495
    },
    {
      "epoch": 5.50314465408805,
      "grad_norm": 0.0582832507789135,
      "learning_rate": 0.000909444810581685,
      "loss": 0.5559,
      "num_input_tokens_seen": 6859688,
      "step": 10500
    },
    {
      "epoch": 5.505765199161425,
      "grad_norm": 0.08544477820396423,
      "learning_rate": 0.000909313513367691,
      "loss": 0.4111,
      "num_input_tokens_seen": 6863080,
      "step": 10505
    },
    {
      "epoch": 5.5083857442348005,
      "grad_norm": 0.06851579993963242,
      "learning_rate": 0.0009091821305299798,
      "loss": 0.4143,
      "num_input_tokens_seen": 6865704,
      "step": 10510
    },
    {
      "epoch": 5.511006289308176,
      "grad_norm": 0.07096036523580551,
      "learning_rate": 0.000909050662096035,
      "loss": 0.5288,
      "num_input_tokens_seen": 6868392,
      "step": 10515
    },
    {
      "epoch": 5.513626834381551,
      "grad_norm": 0.10618004202842712,
      "learning_rate": 0.0009089191080933583,
      "loss": 0.4456,
      "num_input_tokens_seen": 6871624,
      "step": 10520
    },
    {
      "epoch": 5.516247379454927,
      "grad_norm": 0.04742561653256416,
      "learning_rate": 0.0009087874685494695,
      "loss": 0.4149,
      "num_input_tokens_seen": 6875752,
      "step": 10525
    },
    {
      "epoch": 5.518867924528302,
      "grad_norm": 0.08810251951217651,
      "learning_rate": 0.0009086557434919059,
      "loss": 0.6435,
      "num_input_tokens_seen": 6878728,
      "step": 10530
    },
    {
      "epoch": 5.521488469601677,
      "grad_norm": 0.057500723749399185,
      "learning_rate": 0.0009085239329482231,
      "loss": 0.5396,
      "num_input_tokens_seen": 6882408,
      "step": 10535
    },
    {
      "epoch": 5.524109014675052,
      "grad_norm": 0.05792028456926346,
      "learning_rate": 0.0009083920369459941,
      "loss": 0.6065,
      "num_input_tokens_seen": 6885448,
      "step": 10540
    },
    {
      "epoch": 5.526729559748428,
      "grad_norm": 0.09032908082008362,
      "learning_rate": 0.00090826005551281,
      "loss": 0.5129,
      "num_input_tokens_seen": 6889192,
      "step": 10545
    },
    {
      "epoch": 5.529350104821803,
      "grad_norm": 0.09666340798139572,
      "learning_rate": 0.0009081279886762802,
      "loss": 0.5354,
      "num_input_tokens_seen": 6892424,
      "step": 10550
    },
    {
      "epoch": 5.531970649895178,
      "grad_norm": 0.05966253578662872,
      "learning_rate": 0.0009079958364640313,
      "loss": 0.4358,
      "num_input_tokens_seen": 6895368,
      "step": 10555
    },
    {
      "epoch": 5.534591194968553,
      "grad_norm": 0.050362903624773026,
      "learning_rate": 0.0009078635989037081,
      "loss": 0.4621,
      "num_input_tokens_seen": 6899784,
      "step": 10560
    },
    {
      "epoch": 5.537211740041929,
      "grad_norm": 0.075274258852005,
      "learning_rate": 0.000907731276022973,
      "loss": 0.4768,
      "num_input_tokens_seen": 6902504,
      "step": 10565
    },
    {
      "epoch": 5.539832285115304,
      "grad_norm": 0.05094173923134804,
      "learning_rate": 0.0009075988678495066,
      "loss": 0.4833,
      "num_input_tokens_seen": 6906184,
      "step": 10570
    },
    {
      "epoch": 5.5424528301886795,
      "grad_norm": 0.04330168291926384,
      "learning_rate": 0.0009074663744110073,
      "loss": 0.4664,
      "num_input_tokens_seen": 6909320,
      "step": 10575
    },
    {
      "epoch": 5.545073375262055,
      "grad_norm": 0.06429276615381241,
      "learning_rate": 0.0009073337957351911,
      "loss": 0.5179,
      "num_input_tokens_seen": 6912168,
      "step": 10580
    },
    {
      "epoch": 5.54769392033543,
      "grad_norm": 0.07023406773805618,
      "learning_rate": 0.0009072011318497919,
      "loss": 0.4882,
      "num_input_tokens_seen": 6915784,
      "step": 10585
    },
    {
      "epoch": 5.550314465408805,
      "grad_norm": 0.10675335675477982,
      "learning_rate": 0.0009070683827825615,
      "loss": 0.5077,
      "num_input_tokens_seen": 6918920,
      "step": 10590
    },
    {
      "epoch": 5.55293501048218,
      "grad_norm": 0.0739392563700676,
      "learning_rate": 0.0009069355485612695,
      "loss": 0.4326,
      "num_input_tokens_seen": 6921992,
      "step": 10595
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 0.1147344633936882,
      "learning_rate": 0.0009068026292137034,
      "loss": 0.4204,
      "num_input_tokens_seen": 6924840,
      "step": 10600
    },
    {
      "epoch": 5.5581761006289305,
      "grad_norm": 0.06823128461837769,
      "learning_rate": 0.0009066696247676682,
      "loss": 0.4159,
      "num_input_tokens_seen": 6927848,
      "step": 10605
    },
    {
      "epoch": 5.560796645702306,
      "grad_norm": 0.09992069005966187,
      "learning_rate": 0.0009065365352509871,
      "loss": 0.5864,
      "num_input_tokens_seen": 6930568,
      "step": 10610
    },
    {
      "epoch": 5.563417190775681,
      "grad_norm": 0.09840857982635498,
      "learning_rate": 0.0009064033606915008,
      "loss": 0.3766,
      "num_input_tokens_seen": 6933992,
      "step": 10615
    },
    {
      "epoch": 5.566037735849057,
      "grad_norm": 0.05842592939734459,
      "learning_rate": 0.000906270101117068,
      "loss": 0.4264,
      "num_input_tokens_seen": 6936680,
      "step": 10620
    },
    {
      "epoch": 5.568658280922432,
      "grad_norm": 0.10630831122398376,
      "learning_rate": 0.0009061367565555649,
      "loss": 0.4794,
      "num_input_tokens_seen": 6938792,
      "step": 10625
    },
    {
      "epoch": 5.571278825995807,
      "grad_norm": 0.05609438568353653,
      "learning_rate": 0.000906003327034886,
      "loss": 0.425,
      "num_input_tokens_seen": 6942664,
      "step": 10630
    },
    {
      "epoch": 5.573899371069182,
      "grad_norm": 0.10516680777072906,
      "learning_rate": 0.0009058698125829428,
      "loss": 0.4929,
      "num_input_tokens_seen": 6944968,
      "step": 10635
    },
    {
      "epoch": 5.576519916142558,
      "grad_norm": 0.06881435960531235,
      "learning_rate": 0.0009057362132276653,
      "loss": 0.4848,
      "num_input_tokens_seen": 6948264,
      "step": 10640
    },
    {
      "epoch": 5.579140461215933,
      "grad_norm": 0.09449811279773712,
      "learning_rate": 0.0009056025289970009,
      "loss": 0.3961,
      "num_input_tokens_seen": 6951304,
      "step": 10645
    },
    {
      "epoch": 5.581761006289308,
      "grad_norm": 0.047254715114831924,
      "learning_rate": 0.0009054687599189148,
      "loss": 0.414,
      "num_input_tokens_seen": 6954184,
      "step": 10650
    },
    {
      "epoch": 5.584381551362683,
      "grad_norm": 0.06753355264663696,
      "learning_rate": 0.0009053349060213899,
      "loss": 0.6058,
      "num_input_tokens_seen": 6956552,
      "step": 10655
    },
    {
      "epoch": 5.587002096436059,
      "grad_norm": 0.07156413048505783,
      "learning_rate": 0.0009052009673324269,
      "loss": 0.5398,
      "num_input_tokens_seen": 6959496,
      "step": 10660
    },
    {
      "epoch": 5.589622641509434,
      "grad_norm": 0.058000754565000534,
      "learning_rate": 0.0009050669438800445,
      "loss": 0.6328,
      "num_input_tokens_seen": 6962600,
      "step": 10665
    },
    {
      "epoch": 5.5922431865828095,
      "grad_norm": 0.07625769078731537,
      "learning_rate": 0.0009049328356922786,
      "loss": 0.4769,
      "num_input_tokens_seen": 6965160,
      "step": 10670
    },
    {
      "epoch": 5.594863731656185,
      "grad_norm": 0.05374633148312569,
      "learning_rate": 0.0009047986427971831,
      "loss": 0.45,
      "num_input_tokens_seen": 6967880,
      "step": 10675
    },
    {
      "epoch": 5.59748427672956,
      "grad_norm": 0.06645380705595016,
      "learning_rate": 0.0009046643652228297,
      "loss": 0.4558,
      "num_input_tokens_seen": 6971624,
      "step": 10680
    },
    {
      "epoch": 5.600104821802935,
      "grad_norm": 0.054332613945007324,
      "learning_rate": 0.0009045300029973079,
      "loss": 0.4963,
      "num_input_tokens_seen": 6974696,
      "step": 10685
    },
    {
      "epoch": 5.60272536687631,
      "grad_norm": 0.10415764153003693,
      "learning_rate": 0.0009043955561487245,
      "loss": 0.6451,
      "num_input_tokens_seen": 6977544,
      "step": 10690
    },
    {
      "epoch": 5.605345911949685,
      "grad_norm": 0.027206653729081154,
      "learning_rate": 0.0009042610247052044,
      "loss": 0.2922,
      "num_input_tokens_seen": 6983688,
      "step": 10695
    },
    {
      "epoch": 5.6079664570230605,
      "grad_norm": 0.062118884176015854,
      "learning_rate": 0.0009041264086948899,
      "loss": 0.4446,
      "num_input_tokens_seen": 6987112,
      "step": 10700
    },
    {
      "epoch": 5.610587002096436,
      "grad_norm": 0.08268912881612778,
      "learning_rate": 0.0009039917081459414,
      "loss": 0.6183,
      "num_input_tokens_seen": 6990312,
      "step": 10705
    },
    {
      "epoch": 5.613207547169811,
      "grad_norm": 0.08092840015888214,
      "learning_rate": 0.0009038569230865363,
      "loss": 0.4,
      "num_input_tokens_seen": 6992456,
      "step": 10710
    },
    {
      "epoch": 5.615828092243187,
      "grad_norm": 0.07099012285470963,
      "learning_rate": 0.0009037220535448705,
      "loss": 0.422,
      "num_input_tokens_seen": 6994792,
      "step": 10715
    },
    {
      "epoch": 5.618448637316562,
      "grad_norm": 0.07558080554008484,
      "learning_rate": 0.0009035870995491571,
      "loss": 0.4809,
      "num_input_tokens_seen": 6997960,
      "step": 10720
    },
    {
      "epoch": 5.621069182389937,
      "grad_norm": 0.11582443118095398,
      "learning_rate": 0.0009034520611276265,
      "loss": 0.5777,
      "num_input_tokens_seen": 7001640,
      "step": 10725
    },
    {
      "epoch": 5.623689727463312,
      "grad_norm": 0.06862039864063263,
      "learning_rate": 0.0009033169383085278,
      "loss": 0.4384,
      "num_input_tokens_seen": 7004456,
      "step": 10730
    },
    {
      "epoch": 5.626310272536688,
      "grad_norm": 0.13686466217041016,
      "learning_rate": 0.0009031817311201268,
      "loss": 0.5145,
      "num_input_tokens_seen": 7006888,
      "step": 10735
    },
    {
      "epoch": 5.628930817610063,
      "grad_norm": 0.056579213589429855,
      "learning_rate": 0.0009030464395907074,
      "loss": 0.4246,
      "num_input_tokens_seen": 7010600,
      "step": 10740
    },
    {
      "epoch": 5.631551362683438,
      "grad_norm": 0.08073300868272781,
      "learning_rate": 0.0009029110637485711,
      "loss": 0.4609,
      "num_input_tokens_seen": 7014312,
      "step": 10745
    },
    {
      "epoch": 5.634171907756813,
      "grad_norm": 0.048811543732881546,
      "learning_rate": 0.0009027756036220368,
      "loss": 0.3997,
      "num_input_tokens_seen": 7017896,
      "step": 10750
    },
    {
      "epoch": 5.636792452830189,
      "grad_norm": 0.07557635009288788,
      "learning_rate": 0.0009026400592394415,
      "loss": 0.5173,
      "num_input_tokens_seen": 7021256,
      "step": 10755
    },
    {
      "epoch": 5.639412997903564,
      "grad_norm": 0.04409310221672058,
      "learning_rate": 0.0009025044306291393,
      "loss": 0.4666,
      "num_input_tokens_seen": 7024968,
      "step": 10760
    },
    {
      "epoch": 5.6420335429769395,
      "grad_norm": 0.1202654242515564,
      "learning_rate": 0.0009023687178195022,
      "loss": 0.535,
      "num_input_tokens_seen": 7027528,
      "step": 10765
    },
    {
      "epoch": 5.644654088050315,
      "grad_norm": 0.10247913002967834,
      "learning_rate": 0.0009022329208389196,
      "loss": 0.4437,
      "num_input_tokens_seen": 7030632,
      "step": 10770
    },
    {
      "epoch": 5.64727463312369,
      "grad_norm": 0.04068814590573311,
      "learning_rate": 0.000902097039715799,
      "loss": 0.4171,
      "num_input_tokens_seen": 7034568,
      "step": 10775
    },
    {
      "epoch": 5.649895178197065,
      "grad_norm": 0.06930293887853622,
      "learning_rate": 0.0009019610744785651,
      "loss": 0.5121,
      "num_input_tokens_seen": 7037544,
      "step": 10780
    },
    {
      "epoch": 5.65251572327044,
      "grad_norm": 0.11749478429555893,
      "learning_rate": 0.0009018250251556603,
      "loss": 0.4691,
      "num_input_tokens_seen": 7041256,
      "step": 10785
    },
    {
      "epoch": 5.655136268343815,
      "grad_norm": 0.09934719651937485,
      "learning_rate": 0.0009016888917755445,
      "loss": 0.3817,
      "num_input_tokens_seen": 7043528,
      "step": 10790
    },
    {
      "epoch": 5.6577568134171905,
      "grad_norm": 0.06302771717309952,
      "learning_rate": 0.0009015526743666951,
      "loss": 0.4573,
      "num_input_tokens_seen": 7046696,
      "step": 10795
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 0.05309909209609032,
      "learning_rate": 0.0009014163729576074,
      "loss": 0.5839,
      "num_input_tokens_seen": 7050216,
      "step": 10800
    },
    {
      "epoch": 5.662997903563941,
      "grad_norm": 0.06714431941509247,
      "learning_rate": 0.0009012799875767943,
      "loss": 0.6432,
      "num_input_tokens_seen": 7053832,
      "step": 10805
    },
    {
      "epoch": 5.665618448637317,
      "grad_norm": 0.07254039496183395,
      "learning_rate": 0.0009011435182527856,
      "loss": 0.4565,
      "num_input_tokens_seen": 7056584,
      "step": 10810
    },
    {
      "epoch": 5.668238993710692,
      "grad_norm": 0.049457088112831116,
      "learning_rate": 0.0009010069650141295,
      "loss": 0.354,
      "num_input_tokens_seen": 7059240,
      "step": 10815
    },
    {
      "epoch": 5.670859538784067,
      "grad_norm": 0.10521192848682404,
      "learning_rate": 0.0009008703278893913,
      "loss": 0.5406,
      "num_input_tokens_seen": 7062280,
      "step": 10820
    },
    {
      "epoch": 5.673480083857442,
      "grad_norm": 0.050752926617860794,
      "learning_rate": 0.0009007336069071537,
      "loss": 0.5165,
      "num_input_tokens_seen": 7065288,
      "step": 10825
    },
    {
      "epoch": 5.676100628930818,
      "grad_norm": 0.06485725939273834,
      "learning_rate": 0.0009005968020960175,
      "loss": 0.5141,
      "num_input_tokens_seen": 7069160,
      "step": 10830
    },
    {
      "epoch": 5.678721174004193,
      "grad_norm": 0.06795477122068405,
      "learning_rate": 0.0009004599134846004,
      "loss": 0.5069,
      "num_input_tokens_seen": 7072552,
      "step": 10835
    },
    {
      "epoch": 5.681341719077568,
      "grad_norm": 0.07077471166849136,
      "learning_rate": 0.0009003229411015382,
      "loss": 0.3626,
      "num_input_tokens_seen": 7076040,
      "step": 10840
    },
    {
      "epoch": 5.683962264150943,
      "grad_norm": 0.07368505746126175,
      "learning_rate": 0.0009001858849754838,
      "loss": 0.5555,
      "num_input_tokens_seen": 7079368,
      "step": 10845
    },
    {
      "epoch": 5.686582809224319,
      "grad_norm": 0.0768175944685936,
      "learning_rate": 0.0009000487451351078,
      "loss": 0.3895,
      "num_input_tokens_seen": 7082632,
      "step": 10850
    },
    {
      "epoch": 5.689203354297694,
      "grad_norm": 0.0707940086722374,
      "learning_rate": 0.0008999115216090985,
      "loss": 0.4962,
      "num_input_tokens_seen": 7085192,
      "step": 10855
    },
    {
      "epoch": 5.6918238993710695,
      "grad_norm": 0.0991707295179367,
      "learning_rate": 0.0008997742144261612,
      "loss": 0.3954,
      "num_input_tokens_seen": 7088040,
      "step": 10860
    },
    {
      "epoch": 5.694444444444445,
      "grad_norm": 0.1224961057305336,
      "learning_rate": 0.000899636823615019,
      "loss": 0.5083,
      "num_input_tokens_seen": 7090664,
      "step": 10865
    },
    {
      "epoch": 5.69706498951782,
      "grad_norm": 0.07463034987449646,
      "learning_rate": 0.0008994993492044127,
      "loss": 0.4523,
      "num_input_tokens_seen": 7093224,
      "step": 10870
    },
    {
      "epoch": 5.699685534591195,
      "grad_norm": 0.06451240181922913,
      "learning_rate": 0.0008993617912231003,
      "loss": 0.4764,
      "num_input_tokens_seen": 7096008,
      "step": 10875
    },
    {
      "epoch": 5.70230607966457,
      "grad_norm": 0.08140119910240173,
      "learning_rate": 0.0008992241496998572,
      "loss": 0.4381,
      "num_input_tokens_seen": 7098760,
      "step": 10880
    },
    {
      "epoch": 5.704926624737945,
      "grad_norm": 0.13289262354373932,
      "learning_rate": 0.0008990864246634767,
      "loss": 0.5294,
      "num_input_tokens_seen": 7102120,
      "step": 10885
    },
    {
      "epoch": 5.7075471698113205,
      "grad_norm": 0.05214446038007736,
      "learning_rate": 0.0008989486161427691,
      "loss": 0.4552,
      "num_input_tokens_seen": 7105672,
      "step": 10890
    },
    {
      "epoch": 5.710167714884696,
      "grad_norm": 0.10238799452781677,
      "learning_rate": 0.0008988107241665624,
      "loss": 0.4505,
      "num_input_tokens_seen": 7109288,
      "step": 10895
    },
    {
      "epoch": 5.712788259958071,
      "grad_norm": 0.11310955882072449,
      "learning_rate": 0.0008986727487637022,
      "loss": 0.3771,
      "num_input_tokens_seen": 7112264,
      "step": 10900
    },
    {
      "epoch": 5.715408805031447,
      "grad_norm": 0.15028293430805206,
      "learning_rate": 0.0008985346899630513,
      "loss": 0.4049,
      "num_input_tokens_seen": 7114984,
      "step": 10905
    },
    {
      "epoch": 5.718029350104822,
      "grad_norm": 0.06323348730802536,
      "learning_rate": 0.0008983965477934899,
      "loss": 0.4605,
      "num_input_tokens_seen": 7117896,
      "step": 10910
    },
    {
      "epoch": 5.720649895178197,
      "grad_norm": 0.06442485004663467,
      "learning_rate": 0.0008982583222839158,
      "loss": 0.3843,
      "num_input_tokens_seen": 7121192,
      "step": 10915
    },
    {
      "epoch": 5.723270440251572,
      "grad_norm": 0.059316299855709076,
      "learning_rate": 0.0008981200134632444,
      "loss": 0.4532,
      "num_input_tokens_seen": 7124680,
      "step": 10920
    },
    {
      "epoch": 5.725890985324948,
      "grad_norm": 0.07558882981538773,
      "learning_rate": 0.000897981621360408,
      "loss": 0.4396,
      "num_input_tokens_seen": 7127624,
      "step": 10925
    },
    {
      "epoch": 5.728511530398323,
      "grad_norm": 0.06220080703496933,
      "learning_rate": 0.0008978431460043569,
      "loss": 0.4346,
      "num_input_tokens_seen": 7134792,
      "step": 10930
    },
    {
      "epoch": 5.731132075471698,
      "grad_norm": 0.0765901505947113,
      "learning_rate": 0.0008977045874240585,
      "loss": 0.3616,
      "num_input_tokens_seen": 7137448,
      "step": 10935
    },
    {
      "epoch": 5.733752620545073,
      "grad_norm": 0.14334513247013092,
      "learning_rate": 0.0008975659456484977,
      "loss": 0.5137,
      "num_input_tokens_seen": 7141384,
      "step": 10940
    },
    {
      "epoch": 5.736373165618449,
      "grad_norm": 0.13073836266994476,
      "learning_rate": 0.0008974272207066767,
      "loss": 0.5386,
      "num_input_tokens_seen": 7144296,
      "step": 10945
    },
    {
      "epoch": 5.738993710691824,
      "grad_norm": 0.08067308366298676,
      "learning_rate": 0.0008972884126276152,
      "loss": 0.5761,
      "num_input_tokens_seen": 7148328,
      "step": 10950
    },
    {
      "epoch": 5.7416142557651995,
      "grad_norm": 0.06655226647853851,
      "learning_rate": 0.0008971495214403502,
      "loss": 0.5339,
      "num_input_tokens_seen": 7151304,
      "step": 10955
    },
    {
      "epoch": 5.744234800838575,
      "grad_norm": 0.059309717267751694,
      "learning_rate": 0.0008970105471739364,
      "loss": 0.4166,
      "num_input_tokens_seen": 7155240,
      "step": 10960
    },
    {
      "epoch": 5.74685534591195,
      "grad_norm": 0.06291031837463379,
      "learning_rate": 0.0008968714898574454,
      "loss": 0.3637,
      "num_input_tokens_seen": 7159496,
      "step": 10965
    },
    {
      "epoch": 5.749475890985325,
      "grad_norm": 0.06970932334661484,
      "learning_rate": 0.0008967323495199665,
      "loss": 0.3857,
      "num_input_tokens_seen": 7163688,
      "step": 10970
    },
    {
      "epoch": 5.7520964360587,
      "grad_norm": 0.05516122281551361,
      "learning_rate": 0.0008965931261906061,
      "loss": 0.3779,
      "num_input_tokens_seen": 7166728,
      "step": 10975
    },
    {
      "epoch": 5.754716981132075,
      "grad_norm": 0.09492548555135727,
      "learning_rate": 0.0008964538198984885,
      "loss": 0.5295,
      "num_input_tokens_seen": 7169928,
      "step": 10980
    },
    {
      "epoch": 5.7573375262054505,
      "grad_norm": 0.07483864575624466,
      "learning_rate": 0.0008963144306727547,
      "loss": 0.5413,
      "num_input_tokens_seen": 7172936,
      "step": 10985
    },
    {
      "epoch": 5.759958071278826,
      "grad_norm": 0.15671490132808685,
      "learning_rate": 0.0008961749585425634,
      "loss": 0.4003,
      "num_input_tokens_seen": 7175240,
      "step": 10990
    },
    {
      "epoch": 5.762578616352201,
      "grad_norm": 0.05294308438897133,
      "learning_rate": 0.0008960354035370905,
      "loss": 0.4606,
      "num_input_tokens_seen": 7178216,
      "step": 10995
    },
    {
      "epoch": 5.765199161425577,
      "grad_norm": 0.09017486125230789,
      "learning_rate": 0.0008958957656855294,
      "loss": 0.502,
      "num_input_tokens_seen": 7182056,
      "step": 11000
    },
    {
      "epoch": 5.767819706498952,
      "grad_norm": 0.044109050184488297,
      "learning_rate": 0.0008957560450170907,
      "loss": 0.4518,
      "num_input_tokens_seen": 7186056,
      "step": 11005
    },
    {
      "epoch": 5.770440251572327,
      "grad_norm": 0.08066187053918839,
      "learning_rate": 0.0008956162415610025,
      "loss": 0.5896,
      "num_input_tokens_seen": 7188808,
      "step": 11010
    },
    {
      "epoch": 5.773060796645702,
      "grad_norm": 0.04610070958733559,
      "learning_rate": 0.00089547635534651,
      "loss": 0.4643,
      "num_input_tokens_seen": 7192328,
      "step": 11015
    },
    {
      "epoch": 5.7756813417190775,
      "grad_norm": 0.17625676095485687,
      "learning_rate": 0.0008953363864028758,
      "loss": 0.5348,
      "num_input_tokens_seen": 7195368,
      "step": 11020
    },
    {
      "epoch": 5.778301886792453,
      "grad_norm": 0.037027209997177124,
      "learning_rate": 0.0008951963347593796,
      "loss": 0.4765,
      "num_input_tokens_seen": 7199752,
      "step": 11025
    },
    {
      "epoch": 5.780922431865828,
      "grad_norm": 0.1287493109703064,
      "learning_rate": 0.0008950562004453191,
      "loss": 0.6353,
      "num_input_tokens_seen": 7203432,
      "step": 11030
    },
    {
      "epoch": 5.783542976939203,
      "grad_norm": 0.0639331266283989,
      "learning_rate": 0.0008949159834900082,
      "loss": 0.4104,
      "num_input_tokens_seen": 7205864,
      "step": 11035
    },
    {
      "epoch": 5.786163522012579,
      "grad_norm": 0.060445766896009445,
      "learning_rate": 0.0008947756839227792,
      "loss": 0.4322,
      "num_input_tokens_seen": 7208424,
      "step": 11040
    },
    {
      "epoch": 5.788784067085954,
      "grad_norm": 0.05733233690261841,
      "learning_rate": 0.0008946353017729809,
      "loss": 0.7984,
      "num_input_tokens_seen": 7211464,
      "step": 11045
    },
    {
      "epoch": 5.7914046121593294,
      "grad_norm": 0.0836600810289383,
      "learning_rate": 0.0008944948370699798,
      "loss": 0.5371,
      "num_input_tokens_seen": 7214472,
      "step": 11050
    },
    {
      "epoch": 5.794025157232705,
      "grad_norm": 0.06014174595475197,
      "learning_rate": 0.0008943542898431593,
      "loss": 0.4932,
      "num_input_tokens_seen": 7218184,
      "step": 11055
    },
    {
      "epoch": 5.79664570230608,
      "grad_norm": 0.06024546176195145,
      "learning_rate": 0.0008942136601219205,
      "loss": 0.4376,
      "num_input_tokens_seen": 7221480,
      "step": 11060
    },
    {
      "epoch": 5.799266247379455,
      "grad_norm": 0.08578911423683167,
      "learning_rate": 0.0008940729479356813,
      "loss": 0.3608,
      "num_input_tokens_seen": 7224424,
      "step": 11065
    },
    {
      "epoch": 5.80188679245283,
      "grad_norm": 0.07053779065608978,
      "learning_rate": 0.0008939321533138774,
      "loss": 0.5063,
      "num_input_tokens_seen": 7227144,
      "step": 11070
    },
    {
      "epoch": 5.804507337526205,
      "grad_norm": 0.08263245224952698,
      "learning_rate": 0.0008937912762859611,
      "loss": 0.4675,
      "num_input_tokens_seen": 7229768,
      "step": 11075
    },
    {
      "epoch": 5.8071278825995805,
      "grad_norm": 0.12588900327682495,
      "learning_rate": 0.0008936503168814024,
      "loss": 0.3752,
      "num_input_tokens_seen": 7233352,
      "step": 11080
    },
    {
      "epoch": 5.809748427672956,
      "grad_norm": 0.08724407106637955,
      "learning_rate": 0.0008935092751296885,
      "loss": 0.4638,
      "num_input_tokens_seen": 7236328,
      "step": 11085
    },
    {
      "epoch": 5.812368972746331,
      "grad_norm": 0.12811915576457977,
      "learning_rate": 0.0008933681510603235,
      "loss": 0.5808,
      "num_input_tokens_seen": 7238952,
      "step": 11090
    },
    {
      "epoch": 5.814989517819707,
      "grad_norm": 0.04950525239109993,
      "learning_rate": 0.0008932269447028292,
      "loss": 0.4309,
      "num_input_tokens_seen": 7242568,
      "step": 11095
    },
    {
      "epoch": 5.817610062893082,
      "grad_norm": 0.16692280769348145,
      "learning_rate": 0.0008930856560867442,
      "loss": 0.5837,
      "num_input_tokens_seen": 7245320,
      "step": 11100
    },
    {
      "epoch": 5.820230607966457,
      "grad_norm": 0.084649957716465,
      "learning_rate": 0.0008929442852416245,
      "loss": 0.4109,
      "num_input_tokens_seen": 7248264,
      "step": 11105
    },
    {
      "epoch": 5.822851153039832,
      "grad_norm": 0.08893614262342453,
      "learning_rate": 0.0008928028321970433,
      "loss": 0.5823,
      "num_input_tokens_seen": 7251176,
      "step": 11110
    },
    {
      "epoch": 5.8254716981132075,
      "grad_norm": 0.14817044138908386,
      "learning_rate": 0.0008926612969825909,
      "loss": 0.5001,
      "num_input_tokens_seen": 7253864,
      "step": 11115
    },
    {
      "epoch": 5.828092243186583,
      "grad_norm": 0.07538188993930817,
      "learning_rate": 0.000892519679627875,
      "loss": 0.3716,
      "num_input_tokens_seen": 7257704,
      "step": 11120
    },
    {
      "epoch": 5.830712788259958,
      "grad_norm": 0.06890647113323212,
      "learning_rate": 0.00089237798016252,
      "loss": 0.4161,
      "num_input_tokens_seen": 7260936,
      "step": 11125
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 0.058655545115470886,
      "learning_rate": 0.0008922361986161682,
      "loss": 0.4924,
      "num_input_tokens_seen": 7264456,
      "step": 11130
    },
    {
      "epoch": 5.835953878406709,
      "grad_norm": 0.06862232089042664,
      "learning_rate": 0.0008920943350184785,
      "loss": 0.6527,
      "num_input_tokens_seen": 7267624,
      "step": 11135
    },
    {
      "epoch": 5.838574423480084,
      "grad_norm": 0.04890800267457962,
      "learning_rate": 0.0008919523893991271,
      "loss": 0.4372,
      "num_input_tokens_seen": 7270760,
      "step": 11140
    },
    {
      "epoch": 5.841194968553459,
      "grad_norm": 0.07280600816011429,
      "learning_rate": 0.0008918103617878075,
      "loss": 0.461,
      "num_input_tokens_seen": 7273608,
      "step": 11145
    },
    {
      "epoch": 5.843815513626835,
      "grad_norm": 0.07160462439060211,
      "learning_rate": 0.0008916682522142302,
      "loss": 0.5293,
      "num_input_tokens_seen": 7276264,
      "step": 11150
    },
    {
      "epoch": 5.84643605870021,
      "grad_norm": 0.05544864386320114,
      "learning_rate": 0.000891526060708123,
      "loss": 0.4686,
      "num_input_tokens_seen": 7279336,
      "step": 11155
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 0.04640517011284828,
      "learning_rate": 0.0008913837872992306,
      "loss": 0.4798,
      "num_input_tokens_seen": 7282376,
      "step": 11160
    },
    {
      "epoch": 5.85167714884696,
      "grad_norm": 0.05426787957549095,
      "learning_rate": 0.000891241432017315,
      "loss": 0.4315,
      "num_input_tokens_seen": 7285448,
      "step": 11165
    },
    {
      "epoch": 5.854297693920335,
      "grad_norm": 0.09896861761808395,
      "learning_rate": 0.0008910989948921555,
      "loss": 0.5128,
      "num_input_tokens_seen": 7288648,
      "step": 11170
    },
    {
      "epoch": 5.8569182389937104,
      "grad_norm": 0.11425305157899857,
      "learning_rate": 0.000890956475953548,
      "loss": 0.3978,
      "num_input_tokens_seen": 7291432,
      "step": 11175
    },
    {
      "epoch": 5.859538784067086,
      "grad_norm": 0.06350655853748322,
      "learning_rate": 0.0008908138752313061,
      "loss": 0.312,
      "num_input_tokens_seen": 7296744,
      "step": 11180
    },
    {
      "epoch": 5.862159329140461,
      "grad_norm": 0.13068309426307678,
      "learning_rate": 0.0008906711927552601,
      "loss": 0.3891,
      "num_input_tokens_seen": 7299496,
      "step": 11185
    },
    {
      "epoch": 5.864779874213837,
      "grad_norm": 0.05243324860930443,
      "learning_rate": 0.0008905284285552575,
      "loss": 0.4399,
      "num_input_tokens_seen": 7302440,
      "step": 11190
    },
    {
      "epoch": 5.867400419287212,
      "grad_norm": 0.0605824813246727,
      "learning_rate": 0.0008903855826611631,
      "loss": 0.3457,
      "num_input_tokens_seen": 7305864,
      "step": 11195
    },
    {
      "epoch": 5.870020964360587,
      "grad_norm": 0.05818939581513405,
      "learning_rate": 0.0008902426551028586,
      "loss": 0.4865,
      "num_input_tokens_seen": 7309096,
      "step": 11200
    },
    {
      "epoch": 5.872641509433962,
      "grad_norm": 0.1034332811832428,
      "learning_rate": 0.0008900996459102428,
      "loss": 0.6475,
      "num_input_tokens_seen": 7312648,
      "step": 11205
    },
    {
      "epoch": 5.8752620545073375,
      "grad_norm": 0.07232688367366791,
      "learning_rate": 0.0008899565551132315,
      "loss": 0.4452,
      "num_input_tokens_seen": 7316264,
      "step": 11210
    },
    {
      "epoch": 5.877882599580713,
      "grad_norm": 0.04400108382105827,
      "learning_rate": 0.0008898133827417578,
      "loss": 0.5072,
      "num_input_tokens_seen": 7320584,
      "step": 11215
    },
    {
      "epoch": 5.880503144654088,
      "grad_norm": 0.058726996183395386,
      "learning_rate": 0.0008896701288257716,
      "loss": 0.6705,
      "num_input_tokens_seen": 7323880,
      "step": 11220
    },
    {
      "epoch": 5.883123689727463,
      "grad_norm": 0.10678305476903915,
      "learning_rate": 0.00088952679339524,
      "loss": 0.4695,
      "num_input_tokens_seen": 7326984,
      "step": 11225
    },
    {
      "epoch": 5.885744234800838,
      "grad_norm": 0.058385416865348816,
      "learning_rate": 0.0008893833764801473,
      "loss": 0.4332,
      "num_input_tokens_seen": 7330504,
      "step": 11230
    },
    {
      "epoch": 5.888364779874214,
      "grad_norm": 0.06721282750368118,
      "learning_rate": 0.0008892398781104945,
      "loss": 0.4837,
      "num_input_tokens_seen": 7333480,
      "step": 11235
    },
    {
      "epoch": 5.890985324947589,
      "grad_norm": 0.06028595566749573,
      "learning_rate": 0.0008890962983162997,
      "loss": 0.3579,
      "num_input_tokens_seen": 7336232,
      "step": 11240
    },
    {
      "epoch": 5.893605870020965,
      "grad_norm": 0.08060740679502487,
      "learning_rate": 0.0008889526371275987,
      "loss": 0.3555,
      "num_input_tokens_seen": 7338920,
      "step": 11245
    },
    {
      "epoch": 5.89622641509434,
      "grad_norm": 0.06459428369998932,
      "learning_rate": 0.0008888088945744432,
      "loss": 0.4063,
      "num_input_tokens_seen": 7341512,
      "step": 11250
    },
    {
      "epoch": 5.898846960167715,
      "grad_norm": 0.08869926631450653,
      "learning_rate": 0.0008886650706869027,
      "loss": 0.5801,
      "num_input_tokens_seen": 7345128,
      "step": 11255
    },
    {
      "epoch": 5.90146750524109,
      "grad_norm": 0.04239244386553764,
      "learning_rate": 0.0008885211654950636,
      "loss": 0.4342,
      "num_input_tokens_seen": 7348328,
      "step": 11260
    },
    {
      "epoch": 5.904088050314465,
      "grad_norm": 0.09088645130395889,
      "learning_rate": 0.0008883771790290291,
      "loss": 0.6321,
      "num_input_tokens_seen": 7351336,
      "step": 11265
    },
    {
      "epoch": 5.90670859538784,
      "grad_norm": 0.13089318573474884,
      "learning_rate": 0.0008882331113189197,
      "loss": 0.5116,
      "num_input_tokens_seen": 7355208,
      "step": 11270
    },
    {
      "epoch": 5.909329140461216,
      "grad_norm": 0.06841333210468292,
      "learning_rate": 0.0008880889623948724,
      "loss": 0.3714,
      "num_input_tokens_seen": 7357768,
      "step": 11275
    },
    {
      "epoch": 5.911949685534591,
      "grad_norm": 0.11551310122013092,
      "learning_rate": 0.000887944732287042,
      "loss": 0.4762,
      "num_input_tokens_seen": 7360808,
      "step": 11280
    },
    {
      "epoch": 5.914570230607967,
      "grad_norm": 0.08776749670505524,
      "learning_rate": 0.0008878004210255995,
      "loss": 0.4397,
      "num_input_tokens_seen": 7363208,
      "step": 11285
    },
    {
      "epoch": 5.917190775681342,
      "grad_norm": 0.06859584152698517,
      "learning_rate": 0.0008876560286407329,
      "loss": 0.378,
      "num_input_tokens_seen": 7366504,
      "step": 11290
    },
    {
      "epoch": 5.919811320754717,
      "grad_norm": 0.08293892443180084,
      "learning_rate": 0.000887511555162648,
      "loss": 0.4209,
      "num_input_tokens_seen": 7369800,
      "step": 11295
    },
    {
      "epoch": 5.922431865828092,
      "grad_norm": 0.08363088220357895,
      "learning_rate": 0.0008873670006215666,
      "loss": 0.4679,
      "num_input_tokens_seen": 7372680,
      "step": 11300
    },
    {
      "epoch": 5.9250524109014675,
      "grad_norm": 0.058114368468523026,
      "learning_rate": 0.0008872223650477281,
      "loss": 0.4905,
      "num_input_tokens_seen": 7377480,
      "step": 11305
    },
    {
      "epoch": 5.927672955974843,
      "grad_norm": 0.17598284780979156,
      "learning_rate": 0.0008870776484713882,
      "loss": 0.4177,
      "num_input_tokens_seen": 7380456,
      "step": 11310
    },
    {
      "epoch": 5.930293501048218,
      "grad_norm": 0.10520759969949722,
      "learning_rate": 0.0008869328509228205,
      "loss": 0.5529,
      "num_input_tokens_seen": 7383208,
      "step": 11315
    },
    {
      "epoch": 5.932914046121593,
      "grad_norm": 0.07427867501974106,
      "learning_rate": 0.0008867879724323147,
      "loss": 0.4184,
      "num_input_tokens_seen": 7385576,
      "step": 11320
    },
    {
      "epoch": 5.935534591194968,
      "grad_norm": 0.08491349220275879,
      "learning_rate": 0.0008866430130301777,
      "loss": 0.4013,
      "num_input_tokens_seen": 7388296,
      "step": 11325
    },
    {
      "epoch": 5.938155136268344,
      "grad_norm": 0.07630797475576401,
      "learning_rate": 0.0008864979727467335,
      "loss": 0.575,
      "num_input_tokens_seen": 7391784,
      "step": 11330
    },
    {
      "epoch": 5.940775681341719,
      "grad_norm": 0.0882171168923378,
      "learning_rate": 0.000886352851612323,
      "loss": 0.4394,
      "num_input_tokens_seen": 7395016,
      "step": 11335
    },
    {
      "epoch": 5.943396226415095,
      "grad_norm": 0.05303284525871277,
      "learning_rate": 0.0008862076496573034,
      "loss": 0.3806,
      "num_input_tokens_seen": 7398856,
      "step": 11340
    },
    {
      "epoch": 5.94601677148847,
      "grad_norm": 0.07858091592788696,
      "learning_rate": 0.0008860623669120496,
      "loss": 0.5034,
      "num_input_tokens_seen": 7401864,
      "step": 11345
    },
    {
      "epoch": 5.948637316561845,
      "grad_norm": 0.0944271832704544,
      "learning_rate": 0.0008859170034069533,
      "loss": 0.649,
      "num_input_tokens_seen": 7404360,
      "step": 11350
    },
    {
      "epoch": 5.95125786163522,
      "grad_norm": 0.08586662262678146,
      "learning_rate": 0.0008857715591724225,
      "loss": 0.5093,
      "num_input_tokens_seen": 7407464,
      "step": 11355
    },
    {
      "epoch": 5.953878406708595,
      "grad_norm": 0.08117777854204178,
      "learning_rate": 0.0008856260342388827,
      "loss": 0.5114,
      "num_input_tokens_seen": 7411144,
      "step": 11360
    },
    {
      "epoch": 5.95649895178197,
      "grad_norm": 0.05167759582400322,
      "learning_rate": 0.0008854804286367758,
      "loss": 0.4748,
      "num_input_tokens_seen": 7414376,
      "step": 11365
    },
    {
      "epoch": 5.959119496855346,
      "grad_norm": 0.06905199587345123,
      "learning_rate": 0.0008853347423965612,
      "loss": 0.5426,
      "num_input_tokens_seen": 7417672,
      "step": 11370
    },
    {
      "epoch": 5.961740041928721,
      "grad_norm": 0.09428885579109192,
      "learning_rate": 0.0008851889755487146,
      "loss": 0.4396,
      "num_input_tokens_seen": 7420392,
      "step": 11375
    },
    {
      "epoch": 5.964360587002097,
      "grad_norm": 0.05684882029891014,
      "learning_rate": 0.0008850431281237288,
      "loss": 0.4144,
      "num_input_tokens_seen": 7423880,
      "step": 11380
    },
    {
      "epoch": 5.966981132075472,
      "grad_norm": 0.054985471069812775,
      "learning_rate": 0.0008848972001521134,
      "loss": 0.5231,
      "num_input_tokens_seen": 7427624,
      "step": 11385
    },
    {
      "epoch": 5.969601677148847,
      "grad_norm": 0.1282189041376114,
      "learning_rate": 0.0008847511916643948,
      "loss": 0.4042,
      "num_input_tokens_seen": 7430536,
      "step": 11390
    },
    {
      "epoch": 5.972222222222222,
      "grad_norm": 0.07038356363773346,
      "learning_rate": 0.0008846051026911164,
      "loss": 0.5766,
      "num_input_tokens_seen": 7433192,
      "step": 11395
    },
    {
      "epoch": 5.9748427672955975,
      "grad_norm": 0.09220313280820847,
      "learning_rate": 0.0008844589332628383,
      "loss": 0.4277,
      "num_input_tokens_seen": 7436104,
      "step": 11400
    },
    {
      "epoch": 5.977463312368973,
      "grad_norm": 0.1391172707080841,
      "learning_rate": 0.0008843126834101376,
      "loss": 0.4061,
      "num_input_tokens_seen": 7439432,
      "step": 11405
    },
    {
      "epoch": 5.980083857442348,
      "grad_norm": 0.08158498257398605,
      "learning_rate": 0.000884166353163608,
      "loss": 0.4266,
      "num_input_tokens_seen": 7441960,
      "step": 11410
    },
    {
      "epoch": 5.982704402515723,
      "grad_norm": 0.05418136343359947,
      "learning_rate": 0.0008840199425538599,
      "loss": 0.3853,
      "num_input_tokens_seen": 7445576,
      "step": 11415
    },
    {
      "epoch": 5.985324947589098,
      "grad_norm": 0.0709351897239685,
      "learning_rate": 0.0008838734516115213,
      "loss": 0.4575,
      "num_input_tokens_seen": 7448104,
      "step": 11420
    },
    {
      "epoch": 5.987945492662474,
      "grad_norm": 0.06605210900306702,
      "learning_rate": 0.0008837268803672359,
      "loss": 0.5107,
      "num_input_tokens_seen": 7451976,
      "step": 11425
    },
    {
      "epoch": 5.990566037735849,
      "grad_norm": 0.0480300709605217,
      "learning_rate": 0.0008835802288516647,
      "loss": 0.4347,
      "num_input_tokens_seen": 7456744,
      "step": 11430
    },
    {
      "epoch": 5.993186582809225,
      "grad_norm": 0.0710815042257309,
      "learning_rate": 0.0008834334970954861,
      "loss": 0.5367,
      "num_input_tokens_seen": 7460392,
      "step": 11435
    },
    {
      "epoch": 5.9958071278826,
      "grad_norm": 0.059194739907979965,
      "learning_rate": 0.0008832866851293941,
      "loss": 0.4959,
      "num_input_tokens_seen": 7463496,
      "step": 11440
    },
    {
      "epoch": 5.998427672955975,
      "grad_norm": 0.12521101534366608,
      "learning_rate": 0.0008831397929841005,
      "loss": 0.5946,
      "num_input_tokens_seen": 7466376,
      "step": 11445
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.454883873462677,
      "eval_runtime": 13.5487,
      "eval_samples_per_second": 62.589,
      "eval_steps_per_second": 15.647,
      "num_input_tokens_seen": 7467968,
      "step": 11448
    },
    {
      "epoch": 6.00104821802935,
      "grad_norm": 0.0962064191699028,
      "learning_rate": 0.0008829928206903333,
      "loss": 0.5258,
      "num_input_tokens_seen": 7468896,
      "step": 11450
    },
    {
      "epoch": 6.003668763102725,
      "grad_norm": 0.04416278377175331,
      "learning_rate": 0.0008828457682788373,
      "loss": 0.4819,
      "num_input_tokens_seen": 7472640,
      "step": 11455
    },
    {
      "epoch": 6.0062893081761,
      "grad_norm": 0.06530338525772095,
      "learning_rate": 0.0008826986357803743,
      "loss": 0.434,
      "num_input_tokens_seen": 7476192,
      "step": 11460
    },
    {
      "epoch": 6.008909853249476,
      "grad_norm": 0.0628456249833107,
      "learning_rate": 0.0008825514232257228,
      "loss": 0.3968,
      "num_input_tokens_seen": 7479712,
      "step": 11465
    },
    {
      "epoch": 6.011530398322851,
      "grad_norm": 0.05958056077361107,
      "learning_rate": 0.000882404130645678,
      "loss": 0.4481,
      "num_input_tokens_seen": 7482816,
      "step": 11470
    },
    {
      "epoch": 6.014150943396227,
      "grad_norm": 0.06950097531080246,
      "learning_rate": 0.0008822567580710518,
      "loss": 0.5462,
      "num_input_tokens_seen": 7485760,
      "step": 11475
    },
    {
      "epoch": 6.016771488469602,
      "grad_norm": 0.07854973524808884,
      "learning_rate": 0.0008821093055326728,
      "loss": 0.5274,
      "num_input_tokens_seen": 7489440,
      "step": 11480
    },
    {
      "epoch": 6.019392033542977,
      "grad_norm": 0.07496190816164017,
      "learning_rate": 0.0008819617730613863,
      "loss": 0.2983,
      "num_input_tokens_seen": 7492704,
      "step": 11485
    },
    {
      "epoch": 6.022012578616352,
      "grad_norm": 0.08189673721790314,
      "learning_rate": 0.0008818141606880545,
      "loss": 0.4871,
      "num_input_tokens_seen": 7495776,
      "step": 11490
    },
    {
      "epoch": 6.0246331236897275,
      "grad_norm": 0.0823415145277977,
      "learning_rate": 0.0008816664684435564,
      "loss": 0.4376,
      "num_input_tokens_seen": 7498560,
      "step": 11495
    },
    {
      "epoch": 6.027253668763103,
      "grad_norm": 0.08228690177202225,
      "learning_rate": 0.0008815186963587873,
      "loss": 0.443,
      "num_input_tokens_seen": 7500992,
      "step": 11500
    },
    {
      "epoch": 6.029874213836478,
      "grad_norm": 0.10023830831050873,
      "learning_rate": 0.0008813708444646596,
      "loss": 0.5716,
      "num_input_tokens_seen": 7504192,
      "step": 11505
    },
    {
      "epoch": 6.032494758909853,
      "grad_norm": 0.07128185033798218,
      "learning_rate": 0.0008812229127921021,
      "loss": 0.4538,
      "num_input_tokens_seen": 7507552,
      "step": 11510
    },
    {
      "epoch": 6.035115303983228,
      "grad_norm": 0.13069121539592743,
      "learning_rate": 0.0008810749013720605,
      "loss": 0.5338,
      "num_input_tokens_seen": 7510592,
      "step": 11515
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 0.08890485018491745,
      "learning_rate": 0.0008809268102354969,
      "loss": 0.549,
      "num_input_tokens_seen": 7513664,
      "step": 11520
    },
    {
      "epoch": 6.040356394129979,
      "grad_norm": 0.19907230138778687,
      "learning_rate": 0.0008807786394133905,
      "loss": 0.4802,
      "num_input_tokens_seen": 7516320,
      "step": 11525
    },
    {
      "epoch": 6.0429769392033545,
      "grad_norm": 0.12197764217853546,
      "learning_rate": 0.0008806303889367371,
      "loss": 0.385,
      "num_input_tokens_seen": 7518944,
      "step": 11530
    },
    {
      "epoch": 6.04559748427673,
      "grad_norm": 0.15590713918209076,
      "learning_rate": 0.0008804820588365485,
      "loss": 0.4541,
      "num_input_tokens_seen": 7521376,
      "step": 11535
    },
    {
      "epoch": 6.048218029350105,
      "grad_norm": 0.045440707355737686,
      "learning_rate": 0.0008803336491438543,
      "loss": 0.4641,
      "num_input_tokens_seen": 7526432,
      "step": 11540
    },
    {
      "epoch": 6.05083857442348,
      "grad_norm": 0.024603810161352158,
      "learning_rate": 0.0008801851598896995,
      "loss": 0.3714,
      "num_input_tokens_seen": 7532064,
      "step": 11545
    },
    {
      "epoch": 6.053459119496855,
      "grad_norm": 0.12407837063074112,
      "learning_rate": 0.0008800365911051467,
      "loss": 0.4291,
      "num_input_tokens_seen": 7535360,
      "step": 11550
    },
    {
      "epoch": 6.05607966457023,
      "grad_norm": 0.055997416377067566,
      "learning_rate": 0.0008798879428212747,
      "loss": 0.3585,
      "num_input_tokens_seen": 7539232,
      "step": 11555
    },
    {
      "epoch": 6.058700209643606,
      "grad_norm": 0.09609520435333252,
      "learning_rate": 0.0008797392150691792,
      "loss": 0.5719,
      "num_input_tokens_seen": 7542464,
      "step": 11560
    },
    {
      "epoch": 6.061320754716981,
      "grad_norm": 0.11363624036312103,
      "learning_rate": 0.000879590407879972,
      "loss": 0.3821,
      "num_input_tokens_seen": 7546176,
      "step": 11565
    },
    {
      "epoch": 6.063941299790357,
      "grad_norm": 0.07611564546823502,
      "learning_rate": 0.0008794415212847822,
      "loss": 0.5764,
      "num_input_tokens_seen": 7549696,
      "step": 11570
    },
    {
      "epoch": 6.066561844863732,
      "grad_norm": 0.05334214121103287,
      "learning_rate": 0.000879292555314755,
      "loss": 0.3708,
      "num_input_tokens_seen": 7552544,
      "step": 11575
    },
    {
      "epoch": 6.069182389937107,
      "grad_norm": 0.08861038088798523,
      "learning_rate": 0.0008791435100010524,
      "loss": 0.374,
      "num_input_tokens_seen": 7555168,
      "step": 11580
    },
    {
      "epoch": 6.071802935010482,
      "grad_norm": 0.06649525463581085,
      "learning_rate": 0.000878994385374853,
      "loss": 0.6497,
      "num_input_tokens_seen": 7558528,
      "step": 11585
    },
    {
      "epoch": 6.0744234800838575,
      "grad_norm": 0.07958385348320007,
      "learning_rate": 0.0008788451814673521,
      "loss": 0.5115,
      "num_input_tokens_seen": 7561760,
      "step": 11590
    },
    {
      "epoch": 6.077044025157233,
      "grad_norm": 0.04790227860212326,
      "learning_rate": 0.0008786958983097613,
      "loss": 0.5464,
      "num_input_tokens_seen": 7565312,
      "step": 11595
    },
    {
      "epoch": 6.079664570230608,
      "grad_norm": 0.0736105814576149,
      "learning_rate": 0.0008785465359333088,
      "loss": 0.5093,
      "num_input_tokens_seen": 7568640,
      "step": 11600
    },
    {
      "epoch": 6.082285115303983,
      "grad_norm": 0.07114087045192719,
      "learning_rate": 0.0008783970943692398,
      "loss": 0.5056,
      "num_input_tokens_seen": 7572096,
      "step": 11605
    },
    {
      "epoch": 6.084905660377358,
      "grad_norm": 0.1039028912782669,
      "learning_rate": 0.0008782475736488156,
      "loss": 0.4786,
      "num_input_tokens_seen": 7574944,
      "step": 11610
    },
    {
      "epoch": 6.087526205450734,
      "grad_norm": 0.19719120860099792,
      "learning_rate": 0.0008780979738033142,
      "loss": 0.6351,
      "num_input_tokens_seen": 7578080,
      "step": 11615
    },
    {
      "epoch": 6.090146750524109,
      "grad_norm": 0.15024004876613617,
      "learning_rate": 0.0008779482948640305,
      "loss": 0.4412,
      "num_input_tokens_seen": 7580128,
      "step": 11620
    },
    {
      "epoch": 6.0927672955974845,
      "grad_norm": 0.08555465191602707,
      "learning_rate": 0.0008777985368622752,
      "loss": 0.5,
      "num_input_tokens_seen": 7583968,
      "step": 11625
    },
    {
      "epoch": 6.09538784067086,
      "grad_norm": 0.31020858883857727,
      "learning_rate": 0.0008776486998293762,
      "loss": 0.3776,
      "num_input_tokens_seen": 7586368,
      "step": 11630
    },
    {
      "epoch": 6.098008385744235,
      "grad_norm": 0.09559566527605057,
      "learning_rate": 0.0008774987837966775,
      "loss": 0.4767,
      "num_input_tokens_seen": 7589728,
      "step": 11635
    },
    {
      "epoch": 6.10062893081761,
      "grad_norm": 0.1497752070426941,
      "learning_rate": 0.0008773487887955401,
      "loss": 0.5784,
      "num_input_tokens_seen": 7592224,
      "step": 11640
    },
    {
      "epoch": 6.103249475890985,
      "grad_norm": 0.08461830765008926,
      "learning_rate": 0.000877198714857341,
      "loss": 0.6193,
      "num_input_tokens_seen": 7595648,
      "step": 11645
    },
    {
      "epoch": 6.10587002096436,
      "grad_norm": 0.0880059152841568,
      "learning_rate": 0.000877048562013474,
      "loss": 0.3699,
      "num_input_tokens_seen": 7598784,
      "step": 11650
    },
    {
      "epoch": 6.1084905660377355,
      "grad_norm": 0.09488954395055771,
      "learning_rate": 0.0008768983302953498,
      "loss": 0.3931,
      "num_input_tokens_seen": 7602112,
      "step": 11655
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 0.0981617420911789,
      "learning_rate": 0.0008767480197343945,
      "loss": 0.4241,
      "num_input_tokens_seen": 7608768,
      "step": 11660
    },
    {
      "epoch": 6.113731656184487,
      "grad_norm": 0.08520201593637466,
      "learning_rate": 0.0008765976303620515,
      "loss": 0.5998,
      "num_input_tokens_seen": 7612224,
      "step": 11665
    },
    {
      "epoch": 6.116352201257862,
      "grad_norm": 0.07887932658195496,
      "learning_rate": 0.0008764471622097809,
      "loss": 0.3795,
      "num_input_tokens_seen": 7616768,
      "step": 11670
    },
    {
      "epoch": 6.118972746331237,
      "grad_norm": 0.09126722812652588,
      "learning_rate": 0.0008762966153090588,
      "loss": 0.4578,
      "num_input_tokens_seen": 7619840,
      "step": 11675
    },
    {
      "epoch": 6.121593291404612,
      "grad_norm": 0.06081066280603409,
      "learning_rate": 0.0008761459896913776,
      "loss": 0.4886,
      "num_input_tokens_seen": 7623104,
      "step": 11680
    },
    {
      "epoch": 6.1242138364779874,
      "grad_norm": 0.09659025818109512,
      "learning_rate": 0.0008759952853882466,
      "loss": 0.5043,
      "num_input_tokens_seen": 7625856,
      "step": 11685
    },
    {
      "epoch": 6.126834381551363,
      "grad_norm": 0.07274021953344345,
      "learning_rate": 0.0008758445024311917,
      "loss": 0.3367,
      "num_input_tokens_seen": 7629600,
      "step": 11690
    },
    {
      "epoch": 6.129454926624738,
      "grad_norm": 0.11235588043928146,
      "learning_rate": 0.0008756936408517547,
      "loss": 0.5205,
      "num_input_tokens_seen": 7632960,
      "step": 11695
    },
    {
      "epoch": 6.132075471698113,
      "grad_norm": 0.11321710795164108,
      "learning_rate": 0.000875542700681494,
      "loss": 0.506,
      "num_input_tokens_seen": 7635840,
      "step": 11700
    },
    {
      "epoch": 6.134696016771488,
      "grad_norm": 0.06699983030557632,
      "learning_rate": 0.0008753916819519849,
      "loss": 0.591,
      "num_input_tokens_seen": 7639232,
      "step": 11705
    },
    {
      "epoch": 6.137316561844864,
      "grad_norm": 0.06248055025935173,
      "learning_rate": 0.0008752405846948186,
      "loss": 0.4359,
      "num_input_tokens_seen": 7645888,
      "step": 11710
    },
    {
      "epoch": 6.139937106918239,
      "grad_norm": 0.0831565260887146,
      "learning_rate": 0.0008750894089416029,
      "loss": 0.3697,
      "num_input_tokens_seen": 7648608,
      "step": 11715
    },
    {
      "epoch": 6.1425576519916145,
      "grad_norm": 0.12878815829753876,
      "learning_rate": 0.0008749381547239621,
      "loss": 0.5461,
      "num_input_tokens_seen": 7651072,
      "step": 11720
    },
    {
      "epoch": 6.14517819706499,
      "grad_norm": 0.06628201901912689,
      "learning_rate": 0.000874786822073537,
      "loss": 0.649,
      "num_input_tokens_seen": 7654304,
      "step": 11725
    },
    {
      "epoch": 6.147798742138365,
      "grad_norm": 0.055698469281196594,
      "learning_rate": 0.0008746354110219843,
      "loss": 0.4655,
      "num_input_tokens_seen": 7657568,
      "step": 11730
    },
    {
      "epoch": 6.15041928721174,
      "grad_norm": 0.06360884010791779,
      "learning_rate": 0.0008744839216009779,
      "loss": 0.3936,
      "num_input_tokens_seen": 7661056,
      "step": 11735
    },
    {
      "epoch": 6.153039832285115,
      "grad_norm": 0.07785719633102417,
      "learning_rate": 0.0008743323538422072,
      "loss": 0.3983,
      "num_input_tokens_seen": 7667360,
      "step": 11740
    },
    {
      "epoch": 6.15566037735849,
      "grad_norm": 0.07520198076963425,
      "learning_rate": 0.0008741807077773789,
      "loss": 0.4151,
      "num_input_tokens_seen": 7671200,
      "step": 11745
    },
    {
      "epoch": 6.1582809224318655,
      "grad_norm": 0.07074978202581406,
      "learning_rate": 0.0008740289834382153,
      "loss": 0.6407,
      "num_input_tokens_seen": 7674144,
      "step": 11750
    },
    {
      "epoch": 6.160901467505241,
      "grad_norm": 0.08338861167430878,
      "learning_rate": 0.0008738771808564555,
      "loss": 0.4225,
      "num_input_tokens_seen": 7677408,
      "step": 11755
    },
    {
      "epoch": 6.163522012578617,
      "grad_norm": 0.06326652318239212,
      "learning_rate": 0.000873725300063855,
      "loss": 0.6041,
      "num_input_tokens_seen": 7680160,
      "step": 11760
    },
    {
      "epoch": 6.166142557651992,
      "grad_norm": 0.08273881673812866,
      "learning_rate": 0.0008735733410921853,
      "loss": 0.4923,
      "num_input_tokens_seen": 7682848,
      "step": 11765
    },
    {
      "epoch": 6.168763102725367,
      "grad_norm": 0.09754101932048798,
      "learning_rate": 0.0008734213039732346,
      "loss": 0.481,
      "num_input_tokens_seen": 7685728,
      "step": 11770
    },
    {
      "epoch": 6.171383647798742,
      "grad_norm": 0.06061641499400139,
      "learning_rate": 0.0008732691887388073,
      "loss": 0.424,
      "num_input_tokens_seen": 7688960,
      "step": 11775
    },
    {
      "epoch": 6.174004192872117,
      "grad_norm": 0.0873623788356781,
      "learning_rate": 0.0008731169954207243,
      "loss": 0.5266,
      "num_input_tokens_seen": 7691872,
      "step": 11780
    },
    {
      "epoch": 6.176624737945493,
      "grad_norm": 0.1137629970908165,
      "learning_rate": 0.0008729647240508225,
      "loss": 0.4943,
      "num_input_tokens_seen": 7694336,
      "step": 11785
    },
    {
      "epoch": 6.179245283018868,
      "grad_norm": 0.09346560388803482,
      "learning_rate": 0.0008728123746609555,
      "loss": 0.4882,
      "num_input_tokens_seen": 7697280,
      "step": 11790
    },
    {
      "epoch": 6.181865828092243,
      "grad_norm": 0.09812332689762115,
      "learning_rate": 0.000872659947282993,
      "loss": 0.5042,
      "num_input_tokens_seen": 7700640,
      "step": 11795
    },
    {
      "epoch": 6.184486373165618,
      "grad_norm": 0.16140887141227722,
      "learning_rate": 0.000872507441948821,
      "loss": 0.463,
      "num_input_tokens_seen": 7704064,
      "step": 11800
    },
    {
      "epoch": 6.187106918238993,
      "grad_norm": 0.07230154424905777,
      "learning_rate": 0.0008723548586903418,
      "loss": 0.4987,
      "num_input_tokens_seen": 7707712,
      "step": 11805
    },
    {
      "epoch": 6.189727463312369,
      "grad_norm": 0.06834530830383301,
      "learning_rate": 0.0008722021975394745,
      "loss": 0.5091,
      "num_input_tokens_seen": 7710656,
      "step": 11810
    },
    {
      "epoch": 6.1923480083857445,
      "grad_norm": 0.0794316753745079,
      "learning_rate": 0.0008720494585281536,
      "loss": 0.5585,
      "num_input_tokens_seen": 7713664,
      "step": 11815
    },
    {
      "epoch": 6.19496855345912,
      "grad_norm": 0.11027948558330536,
      "learning_rate": 0.0008718966416883306,
      "loss": 0.4969,
      "num_input_tokens_seen": 7716000,
      "step": 11820
    },
    {
      "epoch": 6.197589098532495,
      "grad_norm": 0.07645252346992493,
      "learning_rate": 0.0008717437470519729,
      "loss": 0.5014,
      "num_input_tokens_seen": 7718432,
      "step": 11825
    },
    {
      "epoch": 6.20020964360587,
      "grad_norm": 0.14368432760238647,
      "learning_rate": 0.0008715907746510645,
      "loss": 0.4404,
      "num_input_tokens_seen": 7721216,
      "step": 11830
    },
    {
      "epoch": 6.202830188679245,
      "grad_norm": 0.07387110590934753,
      "learning_rate": 0.0008714377245176054,
      "loss": 0.4312,
      "num_input_tokens_seen": 7724128,
      "step": 11835
    },
    {
      "epoch": 6.20545073375262,
      "grad_norm": 0.11056379228830338,
      "learning_rate": 0.0008712845966836119,
      "loss": 0.369,
      "num_input_tokens_seen": 7727744,
      "step": 11840
    },
    {
      "epoch": 6.2080712788259955,
      "grad_norm": 0.07282586395740509,
      "learning_rate": 0.0008711313911811165,
      "loss": 0.5373,
      "num_input_tokens_seen": 7730880,
      "step": 11845
    },
    {
      "epoch": 6.210691823899371,
      "grad_norm": 0.16170504689216614,
      "learning_rate": 0.0008709781080421681,
      "loss": 0.4952,
      "num_input_tokens_seen": 7734048,
      "step": 11850
    },
    {
      "epoch": 6.213312368972747,
      "grad_norm": 0.14285717904567719,
      "learning_rate": 0.0008708247472988319,
      "loss": 0.4092,
      "num_input_tokens_seen": 7736832,
      "step": 11855
    },
    {
      "epoch": 6.215932914046122,
      "grad_norm": 0.07660763710737228,
      "learning_rate": 0.0008706713089831893,
      "loss": 0.4648,
      "num_input_tokens_seen": 7739808,
      "step": 11860
    },
    {
      "epoch": 6.218553459119497,
      "grad_norm": 0.09189681708812714,
      "learning_rate": 0.0008705177931273377,
      "loss": 0.4229,
      "num_input_tokens_seen": 7743072,
      "step": 11865
    },
    {
      "epoch": 6.221174004192872,
      "grad_norm": 0.07991020381450653,
      "learning_rate": 0.0008703641997633907,
      "loss": 0.4816,
      "num_input_tokens_seen": 7746336,
      "step": 11870
    },
    {
      "epoch": 6.223794549266247,
      "grad_norm": 0.07796722650527954,
      "learning_rate": 0.0008702105289234786,
      "loss": 0.5304,
      "num_input_tokens_seen": 7748864,
      "step": 11875
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 0.08395839482545853,
      "learning_rate": 0.0008700567806397473,
      "loss": 0.3853,
      "num_input_tokens_seen": 7751424,
      "step": 11880
    },
    {
      "epoch": 6.229035639412998,
      "grad_norm": 0.09665276110172272,
      "learning_rate": 0.0008699029549443595,
      "loss": 0.5361,
      "num_input_tokens_seen": 7754272,
      "step": 11885
    },
    {
      "epoch": 6.231656184486373,
      "grad_norm": 0.08035247027873993,
      "learning_rate": 0.0008697490518694935,
      "loss": 0.4135,
      "num_input_tokens_seen": 7757920,
      "step": 11890
    },
    {
      "epoch": 6.234276729559748,
      "grad_norm": 0.08330778777599335,
      "learning_rate": 0.0008695950714473442,
      "loss": 0.4202,
      "num_input_tokens_seen": 7761792,
      "step": 11895
    },
    {
      "epoch": 6.236897274633123,
      "grad_norm": 0.10025079548358917,
      "learning_rate": 0.0008694410137101224,
      "loss": 0.3798,
      "num_input_tokens_seen": 7764832,
      "step": 11900
    },
    {
      "epoch": 6.239517819706499,
      "grad_norm": 0.07018803060054779,
      "learning_rate": 0.0008692868786900554,
      "loss": 0.4562,
      "num_input_tokens_seen": 7768640,
      "step": 11905
    },
    {
      "epoch": 6.2421383647798745,
      "grad_norm": 0.07574202120304108,
      "learning_rate": 0.0008691326664193865,
      "loss": 0.351,
      "num_input_tokens_seen": 7771360,
      "step": 11910
    },
    {
      "epoch": 6.24475890985325,
      "grad_norm": 0.07888432592153549,
      "learning_rate": 0.000868978376930375,
      "loss": 0.3572,
      "num_input_tokens_seen": 7775104,
      "step": 11915
    },
    {
      "epoch": 6.247379454926625,
      "grad_norm": 0.09720655530691147,
      "learning_rate": 0.0008688240102552965,
      "loss": 0.4327,
      "num_input_tokens_seen": 7782720,
      "step": 11920
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.08563872426748276,
      "learning_rate": 0.000868669566426443,
      "loss": 0.4042,
      "num_input_tokens_seen": 7785568,
      "step": 11925
    },
    {
      "epoch": 6.252620545073375,
      "grad_norm": 0.08219907432794571,
      "learning_rate": 0.0008685150454761222,
      "loss": 0.49,
      "num_input_tokens_seen": 7789248,
      "step": 11930
    },
    {
      "epoch": 6.25524109014675,
      "grad_norm": 0.12731194496154785,
      "learning_rate": 0.000868360447436658,
      "loss": 0.4147,
      "num_input_tokens_seen": 7791744,
      "step": 11935
    },
    {
      "epoch": 6.2578616352201255,
      "grad_norm": 0.08578231930732727,
      "learning_rate": 0.0008682057723403907,
      "loss": 0.5974,
      "num_input_tokens_seen": 7795008,
      "step": 11940
    },
    {
      "epoch": 6.260482180293501,
      "grad_norm": 0.08258496224880219,
      "learning_rate": 0.0008680510202196765,
      "loss": 0.5317,
      "num_input_tokens_seen": 7798144,
      "step": 11945
    },
    {
      "epoch": 6.263102725366876,
      "grad_norm": 0.11816775053739548,
      "learning_rate": 0.0008678961911068879,
      "loss": 0.3955,
      "num_input_tokens_seen": 7800608,
      "step": 11950
    },
    {
      "epoch": 6.265723270440252,
      "grad_norm": 0.09175290167331696,
      "learning_rate": 0.0008677412850344133,
      "loss": 0.5911,
      "num_input_tokens_seen": 7804000,
      "step": 11955
    },
    {
      "epoch": 6.268343815513627,
      "grad_norm": 0.06791809946298599,
      "learning_rate": 0.0008675863020346573,
      "loss": 0.4244,
      "num_input_tokens_seen": 7807296,
      "step": 11960
    },
    {
      "epoch": 6.270964360587002,
      "grad_norm": 0.12633423507213593,
      "learning_rate": 0.0008674312421400405,
      "loss": 0.4356,
      "num_input_tokens_seen": 7810816,
      "step": 11965
    },
    {
      "epoch": 6.273584905660377,
      "grad_norm": 0.058131903409957886,
      "learning_rate": 0.0008672761053829997,
      "loss": 0.5008,
      "num_input_tokens_seen": 7814976,
      "step": 11970
    },
    {
      "epoch": 6.276205450733753,
      "grad_norm": 0.07986113429069519,
      "learning_rate": 0.0008671208917959878,
      "loss": 0.7242,
      "num_input_tokens_seen": 7818656,
      "step": 11975
    },
    {
      "epoch": 6.278825995807128,
      "grad_norm": 0.12191452085971832,
      "learning_rate": 0.0008669656014114737,
      "loss": 0.493,
      "num_input_tokens_seen": 7821664,
      "step": 11980
    },
    {
      "epoch": 6.281446540880503,
      "grad_norm": 0.05256790667772293,
      "learning_rate": 0.0008668102342619424,
      "loss": 0.3606,
      "num_input_tokens_seen": 7824800,
      "step": 11985
    },
    {
      "epoch": 6.284067085953878,
      "grad_norm": 0.06555308401584625,
      "learning_rate": 0.0008666547903798947,
      "loss": 0.5086,
      "num_input_tokens_seen": 7827456,
      "step": 11990
    },
    {
      "epoch": 6.286687631027253,
      "grad_norm": 0.080790676176548,
      "learning_rate": 0.000866499269797848,
      "loss": 0.4613,
      "num_input_tokens_seen": 7832064,
      "step": 11995
    },
    {
      "epoch": 6.289308176100629,
      "grad_norm": 0.12003755569458008,
      "learning_rate": 0.0008663436725483354,
      "loss": 0.4913,
      "num_input_tokens_seen": 7834976,
      "step": 12000
    },
    {
      "epoch": 6.2919287211740045,
      "grad_norm": 0.07931645959615707,
      "learning_rate": 0.0008661879986639058,
      "loss": 0.4015,
      "num_input_tokens_seen": 7837920,
      "step": 12005
    },
    {
      "epoch": 6.29454926624738,
      "grad_norm": 0.09822847694158554,
      "learning_rate": 0.0008660322481771246,
      "loss": 0.4642,
      "num_input_tokens_seen": 7840448,
      "step": 12010
    },
    {
      "epoch": 6.297169811320755,
      "grad_norm": 0.09187139570713043,
      "learning_rate": 0.0008658764211205731,
      "loss": 0.4557,
      "num_input_tokens_seen": 7843680,
      "step": 12015
    },
    {
      "epoch": 6.29979035639413,
      "grad_norm": 0.07747730612754822,
      "learning_rate": 0.0008657205175268484,
      "loss": 0.451,
      "num_input_tokens_seen": 7847264,
      "step": 12020
    },
    {
      "epoch": 6.302410901467505,
      "grad_norm": 0.0818730890750885,
      "learning_rate": 0.0008655645374285637,
      "loss": 0.4006,
      "num_input_tokens_seen": 7850144,
      "step": 12025
    },
    {
      "epoch": 6.30503144654088,
      "grad_norm": 0.09194612503051758,
      "learning_rate": 0.0008654084808583485,
      "loss": 0.4147,
      "num_input_tokens_seen": 7853696,
      "step": 12030
    },
    {
      "epoch": 6.3076519916142555,
      "grad_norm": 0.09470833092927933,
      "learning_rate": 0.0008652523478488478,
      "loss": 0.4479,
      "num_input_tokens_seen": 7856736,
      "step": 12035
    },
    {
      "epoch": 6.310272536687631,
      "grad_norm": 0.08766230195760727,
      "learning_rate": 0.000865096138432723,
      "loss": 0.454,
      "num_input_tokens_seen": 7860640,
      "step": 12040
    },
    {
      "epoch": 6.312893081761006,
      "grad_norm": 0.06459276378154755,
      "learning_rate": 0.0008649398526426512,
      "loss": 0.3651,
      "num_input_tokens_seen": 7863680,
      "step": 12045
    },
    {
      "epoch": 6.315513626834382,
      "grad_norm": 0.08061330020427704,
      "learning_rate": 0.0008647834905113256,
      "loss": 0.5772,
      "num_input_tokens_seen": 7866688,
      "step": 12050
    },
    {
      "epoch": 6.318134171907757,
      "grad_norm": 0.10740500688552856,
      "learning_rate": 0.0008646270520714556,
      "loss": 0.3919,
      "num_input_tokens_seen": 7869824,
      "step": 12055
    },
    {
      "epoch": 6.320754716981132,
      "grad_norm": 0.07503081113100052,
      "learning_rate": 0.0008644705373557663,
      "loss": 0.4257,
      "num_input_tokens_seen": 7872640,
      "step": 12060
    },
    {
      "epoch": 6.323375262054507,
      "grad_norm": 0.09431745857000351,
      "learning_rate": 0.0008643139463969985,
      "loss": 0.5007,
      "num_input_tokens_seen": 7875424,
      "step": 12065
    },
    {
      "epoch": 6.325995807127883,
      "grad_norm": 0.03894083946943283,
      "learning_rate": 0.0008641572792279095,
      "loss": 0.4794,
      "num_input_tokens_seen": 7879360,
      "step": 12070
    },
    {
      "epoch": 6.328616352201258,
      "grad_norm": 0.08834033459424973,
      "learning_rate": 0.0008640005358812723,
      "loss": 0.4035,
      "num_input_tokens_seen": 7882144,
      "step": 12075
    },
    {
      "epoch": 6.331236897274633,
      "grad_norm": 0.07477552443742752,
      "learning_rate": 0.0008638437163898756,
      "loss": 0.4214,
      "num_input_tokens_seen": 7886176,
      "step": 12080
    },
    {
      "epoch": 6.333857442348008,
      "grad_norm": 0.14704228937625885,
      "learning_rate": 0.0008636868207865244,
      "loss": 0.4354,
      "num_input_tokens_seen": 7889760,
      "step": 12085
    },
    {
      "epoch": 6.336477987421383,
      "grad_norm": 0.08089133352041245,
      "learning_rate": 0.0008635298491040394,
      "loss": 0.5271,
      "num_input_tokens_seen": 7892448,
      "step": 12090
    },
    {
      "epoch": 6.339098532494759,
      "grad_norm": 0.06592780351638794,
      "learning_rate": 0.0008633728013752574,
      "loss": 0.43,
      "num_input_tokens_seen": 7896608,
      "step": 12095
    },
    {
      "epoch": 6.3417190775681345,
      "grad_norm": 0.13086961209774017,
      "learning_rate": 0.000863215677633031,
      "loss": 0.4722,
      "num_input_tokens_seen": 7899584,
      "step": 12100
    },
    {
      "epoch": 6.34433962264151,
      "grad_norm": 0.08021949231624603,
      "learning_rate": 0.0008630584779102285,
      "loss": 0.5179,
      "num_input_tokens_seen": 7903008,
      "step": 12105
    },
    {
      "epoch": 6.346960167714885,
      "grad_norm": 0.13073870539665222,
      "learning_rate": 0.0008629012022397344,
      "loss": 0.4946,
      "num_input_tokens_seen": 7905696,
      "step": 12110
    },
    {
      "epoch": 6.34958071278826,
      "grad_norm": 0.08855185657739639,
      "learning_rate": 0.0008627438506544491,
      "loss": 0.5467,
      "num_input_tokens_seen": 7908832,
      "step": 12115
    },
    {
      "epoch": 6.352201257861635,
      "grad_norm": 0.17459645867347717,
      "learning_rate": 0.0008625864231872886,
      "loss": 0.4859,
      "num_input_tokens_seen": 7911520,
      "step": 12120
    },
    {
      "epoch": 6.35482180293501,
      "grad_norm": 0.10176931321620941,
      "learning_rate": 0.000862428919871185,
      "loss": 0.5447,
      "num_input_tokens_seen": 7914848,
      "step": 12125
    },
    {
      "epoch": 6.3574423480083855,
      "grad_norm": 0.07800455391407013,
      "learning_rate": 0.0008622713407390861,
      "loss": 0.4779,
      "num_input_tokens_seen": 7918400,
      "step": 12130
    },
    {
      "epoch": 6.360062893081761,
      "grad_norm": 0.101734958589077,
      "learning_rate": 0.0008621136858239557,
      "loss": 0.7211,
      "num_input_tokens_seen": 7921248,
      "step": 12135
    },
    {
      "epoch": 6.362683438155136,
      "grad_norm": 0.0524536594748497,
      "learning_rate": 0.0008619559551587736,
      "loss": 0.5338,
      "num_input_tokens_seen": 7926240,
      "step": 12140
    },
    {
      "epoch": 6.365303983228512,
      "grad_norm": 0.048070378601551056,
      "learning_rate": 0.0008617981487765349,
      "loss": 0.4684,
      "num_input_tokens_seen": 7930176,
      "step": 12145
    },
    {
      "epoch": 6.367924528301887,
      "grad_norm": 0.09759596735239029,
      "learning_rate": 0.0008616402667102512,
      "loss": 0.4651,
      "num_input_tokens_seen": 7933120,
      "step": 12150
    },
    {
      "epoch": 6.370545073375262,
      "grad_norm": 0.078939288854599,
      "learning_rate": 0.0008614823089929495,
      "loss": 0.5498,
      "num_input_tokens_seen": 7935552,
      "step": 12155
    },
    {
      "epoch": 6.373165618448637,
      "grad_norm": 0.07540196180343628,
      "learning_rate": 0.0008613242756576728,
      "loss": 0.3325,
      "num_input_tokens_seen": 7937824,
      "step": 12160
    },
    {
      "epoch": 6.3757861635220126,
      "grad_norm": 0.060838691890239716,
      "learning_rate": 0.0008611661667374798,
      "loss": 0.5816,
      "num_input_tokens_seen": 7940128,
      "step": 12165
    },
    {
      "epoch": 6.378406708595388,
      "grad_norm": 0.08957182615995407,
      "learning_rate": 0.0008610079822654451,
      "loss": 0.2921,
      "num_input_tokens_seen": 7943008,
      "step": 12170
    },
    {
      "epoch": 6.381027253668763,
      "grad_norm": 0.24135705828666687,
      "learning_rate": 0.0008608497222746591,
      "loss": 0.4181,
      "num_input_tokens_seen": 7945472,
      "step": 12175
    },
    {
      "epoch": 6.383647798742138,
      "grad_norm": 0.07281176000833511,
      "learning_rate": 0.000860691386798228,
      "loss": 0.4193,
      "num_input_tokens_seen": 7948608,
      "step": 12180
    },
    {
      "epoch": 6.386268343815513,
      "grad_norm": 0.08161981403827667,
      "learning_rate": 0.000860532975869274,
      "loss": 0.3705,
      "num_input_tokens_seen": 7951296,
      "step": 12185
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 0.06620174646377563,
      "learning_rate": 0.0008603744895209342,
      "loss": 0.3677,
      "num_input_tokens_seen": 7955488,
      "step": 12190
    },
    {
      "epoch": 6.3915094339622645,
      "grad_norm": 0.107461117208004,
      "learning_rate": 0.0008602159277863628,
      "loss": 0.4384,
      "num_input_tokens_seen": 7958880,
      "step": 12195
    },
    {
      "epoch": 6.39412997903564,
      "grad_norm": 0.12550297379493713,
      "learning_rate": 0.0008600572906987289,
      "loss": 0.4718,
      "num_input_tokens_seen": 7961888,
      "step": 12200
    },
    {
      "epoch": 6.396750524109015,
      "grad_norm": 0.059992868453264236,
      "learning_rate": 0.0008598985782912174,
      "loss": 0.4611,
      "num_input_tokens_seen": 7964896,
      "step": 12205
    },
    {
      "epoch": 6.39937106918239,
      "grad_norm": 0.09306295961141586,
      "learning_rate": 0.0008597397905970293,
      "loss": 0.4752,
      "num_input_tokens_seen": 7968032,
      "step": 12210
    },
    {
      "epoch": 6.401991614255765,
      "grad_norm": 0.08450910449028015,
      "learning_rate": 0.0008595809276493812,
      "loss": 0.3763,
      "num_input_tokens_seen": 7971040,
      "step": 12215
    },
    {
      "epoch": 6.40461215932914,
      "grad_norm": 0.09273796528577805,
      "learning_rate": 0.0008594219894815054,
      "loss": 0.4012,
      "num_input_tokens_seen": 7974560,
      "step": 12220
    },
    {
      "epoch": 6.4072327044025155,
      "grad_norm": 0.13135945796966553,
      "learning_rate": 0.0008592629761266497,
      "loss": 0.3975,
      "num_input_tokens_seen": 7978144,
      "step": 12225
    },
    {
      "epoch": 6.409853249475891,
      "grad_norm": 0.0858796164393425,
      "learning_rate": 0.0008591038876180784,
      "loss": 0.3665,
      "num_input_tokens_seen": 7981024,
      "step": 12230
    },
    {
      "epoch": 6.412473794549266,
      "grad_norm": 0.0473344512283802,
      "learning_rate": 0.0008589447239890705,
      "loss": 0.4919,
      "num_input_tokens_seen": 7984448,
      "step": 12235
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 0.08797916769981384,
      "learning_rate": 0.0008587854852729213,
      "loss": 0.4948,
      "num_input_tokens_seen": 7987520,
      "step": 12240
    },
    {
      "epoch": 6.417714884696017,
      "grad_norm": 0.11084766685962677,
      "learning_rate": 0.000858626171502942,
      "loss": 0.4648,
      "num_input_tokens_seen": 7990784,
      "step": 12245
    },
    {
      "epoch": 6.420335429769392,
      "grad_norm": 0.08159873634576797,
      "learning_rate": 0.0008584667827124589,
      "loss": 0.4822,
      "num_input_tokens_seen": 7993792,
      "step": 12250
    },
    {
      "epoch": 6.422955974842767,
      "grad_norm": 0.053084395825862885,
      "learning_rate": 0.0008583073189348145,
      "loss": 0.4911,
      "num_input_tokens_seen": 7996768,
      "step": 12255
    },
    {
      "epoch": 6.4255765199161425,
      "grad_norm": 0.12154089659452438,
      "learning_rate": 0.0008581477802033668,
      "loss": 0.3561,
      "num_input_tokens_seen": 7999424,
      "step": 12260
    },
    {
      "epoch": 6.428197064989518,
      "grad_norm": 0.049089353531599045,
      "learning_rate": 0.0008579881665514895,
      "loss": 0.3708,
      "num_input_tokens_seen": 8003456,
      "step": 12265
    },
    {
      "epoch": 6.430817610062893,
      "grad_norm": 0.080308698117733,
      "learning_rate": 0.0008578284780125718,
      "loss": 0.3894,
      "num_input_tokens_seen": 8007360,
      "step": 12270
    },
    {
      "epoch": 6.433438155136268,
      "grad_norm": 0.07975304871797562,
      "learning_rate": 0.0008576687146200189,
      "loss": 0.3258,
      "num_input_tokens_seen": 8010112,
      "step": 12275
    },
    {
      "epoch": 6.436058700209643,
      "grad_norm": 0.06998728215694427,
      "learning_rate": 0.0008575088764072511,
      "loss": 0.4395,
      "num_input_tokens_seen": 8013024,
      "step": 12280
    },
    {
      "epoch": 6.438679245283019,
      "grad_norm": 0.10419004410505295,
      "learning_rate": 0.0008573489634077054,
      "loss": 0.5921,
      "num_input_tokens_seen": 8015584,
      "step": 12285
    },
    {
      "epoch": 6.441299790356394,
      "grad_norm": 0.05462493374943733,
      "learning_rate": 0.0008571889756548333,
      "loss": 0.4583,
      "num_input_tokens_seen": 8018720,
      "step": 12290
    },
    {
      "epoch": 6.44392033542977,
      "grad_norm": 0.07271946966648102,
      "learning_rate": 0.0008570289131821025,
      "loss": 0.3981,
      "num_input_tokens_seen": 8021184,
      "step": 12295
    },
    {
      "epoch": 6.446540880503145,
      "grad_norm": 0.12421786785125732,
      "learning_rate": 0.0008568687760229963,
      "loss": 0.5335,
      "num_input_tokens_seen": 8024416,
      "step": 12300
    },
    {
      "epoch": 6.44916142557652,
      "grad_norm": 0.13558562099933624,
      "learning_rate": 0.0008567085642110135,
      "loss": 0.3656,
      "num_input_tokens_seen": 8028160,
      "step": 12305
    },
    {
      "epoch": 6.451781970649895,
      "grad_norm": 0.0854760929942131,
      "learning_rate": 0.0008565482777796687,
      "loss": 0.3174,
      "num_input_tokens_seen": 8031296,
      "step": 12310
    },
    {
      "epoch": 6.45440251572327,
      "grad_norm": 0.12598943710327148,
      "learning_rate": 0.0008563879167624919,
      "loss": 0.5499,
      "num_input_tokens_seen": 8034624,
      "step": 12315
    },
    {
      "epoch": 6.4570230607966455,
      "grad_norm": 0.05884164571762085,
      "learning_rate": 0.0008562274811930287,
      "loss": 0.439,
      "num_input_tokens_seen": 8038176,
      "step": 12320
    },
    {
      "epoch": 6.459643605870021,
      "grad_norm": 0.10203695297241211,
      "learning_rate": 0.0008560669711048406,
      "loss": 0.4652,
      "num_input_tokens_seen": 8041920,
      "step": 12325
    },
    {
      "epoch": 6.462264150943396,
      "grad_norm": 0.06769011169672012,
      "learning_rate": 0.0008559063865315042,
      "loss": 0.4552,
      "num_input_tokens_seen": 8044992,
      "step": 12330
    },
    {
      "epoch": 6.464884696016772,
      "grad_norm": 0.09152914583683014,
      "learning_rate": 0.0008557457275066122,
      "loss": 0.4647,
      "num_input_tokens_seen": 8048352,
      "step": 12335
    },
    {
      "epoch": 6.467505241090147,
      "grad_norm": 0.1279563009738922,
      "learning_rate": 0.0008555849940637725,
      "loss": 0.56,
      "num_input_tokens_seen": 8051168,
      "step": 12340
    },
    {
      "epoch": 6.470125786163522,
      "grad_norm": 0.08999256789684296,
      "learning_rate": 0.000855424186236609,
      "loss": 0.4892,
      "num_input_tokens_seen": 8054848,
      "step": 12345
    },
    {
      "epoch": 6.472746331236897,
      "grad_norm": 0.08143391460180283,
      "learning_rate": 0.0008552633040587605,
      "loss": 0.4054,
      "num_input_tokens_seen": 8057952,
      "step": 12350
    },
    {
      "epoch": 6.4753668763102725,
      "grad_norm": 0.06914786249399185,
      "learning_rate": 0.0008551023475638819,
      "loss": 0.4657,
      "num_input_tokens_seen": 8061344,
      "step": 12355
    },
    {
      "epoch": 6.477987421383648,
      "grad_norm": 0.08103613555431366,
      "learning_rate": 0.0008549413167856432,
      "loss": 0.4734,
      "num_input_tokens_seen": 8064704,
      "step": 12360
    },
    {
      "epoch": 6.480607966457023,
      "grad_norm": 0.0970177873969078,
      "learning_rate": 0.0008547802117577307,
      "loss": 0.5329,
      "num_input_tokens_seen": 8067360,
      "step": 12365
    },
    {
      "epoch": 6.483228511530398,
      "grad_norm": 0.08897486329078674,
      "learning_rate": 0.0008546190325138454,
      "loss": 0.5293,
      "num_input_tokens_seen": 8071520,
      "step": 12370
    },
    {
      "epoch": 6.485849056603773,
      "grad_norm": 0.08790402859449387,
      "learning_rate": 0.000854457779087704,
      "loss": 0.4592,
      "num_input_tokens_seen": 8075680,
      "step": 12375
    },
    {
      "epoch": 6.488469601677149,
      "grad_norm": 0.1527882218360901,
      "learning_rate": 0.0008542964515130392,
      "loss": 0.8798,
      "num_input_tokens_seen": 8077600,
      "step": 12380
    },
    {
      "epoch": 6.491090146750524,
      "grad_norm": 0.08147255331277847,
      "learning_rate": 0.0008541350498235989,
      "loss": 0.4477,
      "num_input_tokens_seen": 8079872,
      "step": 12385
    },
    {
      "epoch": 6.4937106918239,
      "grad_norm": 0.08261951059103012,
      "learning_rate": 0.0008539735740531462,
      "loss": 0.4257,
      "num_input_tokens_seen": 8082720,
      "step": 12390
    },
    {
      "epoch": 6.496331236897275,
      "grad_norm": 0.10463815182447433,
      "learning_rate": 0.0008538120242354602,
      "loss": 0.5171,
      "num_input_tokens_seen": 8085728,
      "step": 12395
    },
    {
      "epoch": 6.49895178197065,
      "grad_norm": 0.11087799817323685,
      "learning_rate": 0.0008536504004043353,
      "loss": 0.4551,
      "num_input_tokens_seen": 8089280,
      "step": 12400
    },
    {
      "epoch": 6.501572327044025,
      "grad_norm": 0.07101929187774658,
      "learning_rate": 0.0008534887025935811,
      "loss": 0.4527,
      "num_input_tokens_seen": 8092576,
      "step": 12405
    },
    {
      "epoch": 6.5041928721174,
      "grad_norm": 0.1009015366435051,
      "learning_rate": 0.0008533269308370233,
      "loss": 0.5061,
      "num_input_tokens_seen": 8096032,
      "step": 12410
    },
    {
      "epoch": 6.506813417190775,
      "grad_norm": 0.11393119394779205,
      "learning_rate": 0.0008531650851685025,
      "loss": 0.5204,
      "num_input_tokens_seen": 8099424,
      "step": 12415
    },
    {
      "epoch": 6.509433962264151,
      "grad_norm": 0.10666705667972565,
      "learning_rate": 0.000853003165621875,
      "loss": 0.4263,
      "num_input_tokens_seen": 8101952,
      "step": 12420
    },
    {
      "epoch": 6.512054507337526,
      "grad_norm": 0.14195743203163147,
      "learning_rate": 0.0008528411722310126,
      "loss": 0.4507,
      "num_input_tokens_seen": 8105472,
      "step": 12425
    },
    {
      "epoch": 6.514675052410902,
      "grad_norm": 0.07555488497018814,
      "learning_rate": 0.0008526791050298022,
      "loss": 0.5952,
      "num_input_tokens_seen": 8108256,
      "step": 12430
    },
    {
      "epoch": 6.517295597484277,
      "grad_norm": 0.10140352696180344,
      "learning_rate": 0.0008525169640521467,
      "loss": 0.4012,
      "num_input_tokens_seen": 8111008,
      "step": 12435
    },
    {
      "epoch": 6.519916142557652,
      "grad_norm": 0.11360366642475128,
      "learning_rate": 0.0008523547493319641,
      "loss": 0.5125,
      "num_input_tokens_seen": 8114208,
      "step": 12440
    },
    {
      "epoch": 6.522536687631027,
      "grad_norm": 0.06597685813903809,
      "learning_rate": 0.0008521924609031877,
      "loss": 0.3581,
      "num_input_tokens_seen": 8117376,
      "step": 12445
    },
    {
      "epoch": 6.5251572327044025,
      "grad_norm": 0.12313111871480942,
      "learning_rate": 0.0008520300987997667,
      "loss": 0.552,
      "num_input_tokens_seen": 8120256,
      "step": 12450
    },
    {
      "epoch": 6.527777777777778,
      "grad_norm": 0.2254895716905594,
      "learning_rate": 0.000851867663055665,
      "loss": 0.3039,
      "num_input_tokens_seen": 8122848,
      "step": 12455
    },
    {
      "epoch": 6.530398322851153,
      "grad_norm": 0.1196139007806778,
      "learning_rate": 0.0008517051537048625,
      "loss": 0.4057,
      "num_input_tokens_seen": 8126336,
      "step": 12460
    },
    {
      "epoch": 6.533018867924528,
      "grad_norm": 0.061468031257390976,
      "learning_rate": 0.0008515425707813544,
      "loss": 0.2915,
      "num_input_tokens_seen": 8129760,
      "step": 12465
    },
    {
      "epoch": 6.535639412997903,
      "grad_norm": 0.07177236676216125,
      "learning_rate": 0.0008513799143191512,
      "loss": 0.4261,
      "num_input_tokens_seen": 8133248,
      "step": 12470
    },
    {
      "epoch": 6.538259958071279,
      "grad_norm": 0.08549415320158005,
      "learning_rate": 0.0008512171843522786,
      "loss": 0.5014,
      "num_input_tokens_seen": 8136032,
      "step": 12475
    },
    {
      "epoch": 6.540880503144654,
      "grad_norm": 0.05885627120733261,
      "learning_rate": 0.000851054380914778,
      "loss": 0.3864,
      "num_input_tokens_seen": 8140224,
      "step": 12480
    },
    {
      "epoch": 6.54350104821803,
      "grad_norm": 0.06023002043366432,
      "learning_rate": 0.0008508915040407057,
      "loss": 0.4901,
      "num_input_tokens_seen": 8143424,
      "step": 12485
    },
    {
      "epoch": 6.546121593291405,
      "grad_norm": 0.07845747470855713,
      "learning_rate": 0.0008507285537641342,
      "loss": 0.4258,
      "num_input_tokens_seen": 8146336,
      "step": 12490
    },
    {
      "epoch": 6.54874213836478,
      "grad_norm": 0.12249557673931122,
      "learning_rate": 0.0008505655301191504,
      "loss": 0.5317,
      "num_input_tokens_seen": 8148768,
      "step": 12495
    },
    {
      "epoch": 6.551362683438155,
      "grad_norm": 0.09851841628551483,
      "learning_rate": 0.0008504024331398572,
      "loss": 0.3388,
      "num_input_tokens_seen": 8151840,
      "step": 12500
    },
    {
      "epoch": 6.55398322851153,
      "grad_norm": 0.05382048338651657,
      "learning_rate": 0.0008502392628603727,
      "loss": 0.4166,
      "num_input_tokens_seen": 8154560,
      "step": 12505
    },
    {
      "epoch": 6.556603773584905,
      "grad_norm": 0.05812652036547661,
      "learning_rate": 0.0008500760193148301,
      "loss": 0.4563,
      "num_input_tokens_seen": 8158144,
      "step": 12510
    },
    {
      "epoch": 6.559224318658281,
      "grad_norm": 0.13167937099933624,
      "learning_rate": 0.0008499127025373781,
      "loss": 0.578,
      "num_input_tokens_seen": 8160672,
      "step": 12515
    },
    {
      "epoch": 6.561844863731656,
      "grad_norm": 0.07311965525150299,
      "learning_rate": 0.0008497493125621807,
      "loss": 0.435,
      "num_input_tokens_seen": 8163744,
      "step": 12520
    },
    {
      "epoch": 6.564465408805032,
      "grad_norm": 0.08685647696256638,
      "learning_rate": 0.0008495858494234173,
      "loss": 0.6367,
      "num_input_tokens_seen": 8167104,
      "step": 12525
    },
    {
      "epoch": 6.567085953878407,
      "grad_norm": 0.08975008875131607,
      "learning_rate": 0.0008494223131552825,
      "loss": 0.4019,
      "num_input_tokens_seen": 8170304,
      "step": 12530
    },
    {
      "epoch": 6.569706498951782,
      "grad_norm": 0.15252803266048431,
      "learning_rate": 0.0008492587037919864,
      "loss": 0.4366,
      "num_input_tokens_seen": 8173184,
      "step": 12535
    },
    {
      "epoch": 6.572327044025157,
      "grad_norm": 0.0814368799328804,
      "learning_rate": 0.0008490950213677537,
      "loss": 0.4857,
      "num_input_tokens_seen": 8177472,
      "step": 12540
    },
    {
      "epoch": 6.5749475890985325,
      "grad_norm": 0.1043824553489685,
      "learning_rate": 0.0008489312659168254,
      "loss": 0.42,
      "num_input_tokens_seen": 8180480,
      "step": 12545
    },
    {
      "epoch": 6.577568134171908,
      "grad_norm": 0.05758801847696304,
      "learning_rate": 0.0008487674374734569,
      "loss": 0.3848,
      "num_input_tokens_seen": 8185088,
      "step": 12550
    },
    {
      "epoch": 6.580188679245283,
      "grad_norm": 0.06001896783709526,
      "learning_rate": 0.0008486035360719197,
      "loss": 0.5492,
      "num_input_tokens_seen": 8188384,
      "step": 12555
    },
    {
      "epoch": 6.582809224318658,
      "grad_norm": 0.10873952507972717,
      "learning_rate": 0.0008484395617464997,
      "loss": 0.5095,
      "num_input_tokens_seen": 8191360,
      "step": 12560
    },
    {
      "epoch": 6.585429769392033,
      "grad_norm": 0.09200447052717209,
      "learning_rate": 0.0008482755145314986,
      "loss": 0.3772,
      "num_input_tokens_seen": 8194816,
      "step": 12565
    },
    {
      "epoch": 6.588050314465409,
      "grad_norm": 0.10273735970258713,
      "learning_rate": 0.0008481113944612331,
      "loss": 0.3541,
      "num_input_tokens_seen": 8198016,
      "step": 12570
    },
    {
      "epoch": 6.590670859538784,
      "grad_norm": 0.09094121307134628,
      "learning_rate": 0.0008479472015700356,
      "loss": 0.464,
      "num_input_tokens_seen": 8200448,
      "step": 12575
    },
    {
      "epoch": 6.59329140461216,
      "grad_norm": 0.0909515768289566,
      "learning_rate": 0.0008477829358922529,
      "loss": 0.6162,
      "num_input_tokens_seen": 8203488,
      "step": 12580
    },
    {
      "epoch": 6.595911949685535,
      "grad_norm": 0.16003993153572083,
      "learning_rate": 0.0008476185974622477,
      "loss": 0.4729,
      "num_input_tokens_seen": 8206464,
      "step": 12585
    },
    {
      "epoch": 6.59853249475891,
      "grad_norm": 0.1124911904335022,
      "learning_rate": 0.0008474541863143978,
      "loss": 0.3953,
      "num_input_tokens_seen": 8212928,
      "step": 12590
    },
    {
      "epoch": 6.601153039832285,
      "grad_norm": 0.0759468674659729,
      "learning_rate": 0.0008472897024830961,
      "loss": 0.466,
      "num_input_tokens_seen": 8216128,
      "step": 12595
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 0.24854245781898499,
      "learning_rate": 0.0008471251460027508,
      "loss": 0.5521,
      "num_input_tokens_seen": 8218496,
      "step": 12600
    },
    {
      "epoch": 6.606394129979035,
      "grad_norm": 0.07633049041032791,
      "learning_rate": 0.0008469605169077852,
      "loss": 0.2868,
      "num_input_tokens_seen": 8221824,
      "step": 12605
    },
    {
      "epoch": 6.609014675052411,
      "grad_norm": 0.13244935870170593,
      "learning_rate": 0.0008467958152326377,
      "loss": 0.3995,
      "num_input_tokens_seen": 8225376,
      "step": 12610
    },
    {
      "epoch": 6.611635220125786,
      "grad_norm": 0.0921754315495491,
      "learning_rate": 0.0008466310410117622,
      "loss": 0.5321,
      "num_input_tokens_seen": 8228224,
      "step": 12615
    },
    {
      "epoch": 6.614255765199162,
      "grad_norm": 0.09396351873874664,
      "learning_rate": 0.0008464661942796275,
      "loss": 0.5081,
      "num_input_tokens_seen": 8230656,
      "step": 12620
    },
    {
      "epoch": 6.616876310272537,
      "grad_norm": 0.08349809795618057,
      "learning_rate": 0.0008463012750707175,
      "loss": 0.3925,
      "num_input_tokens_seen": 8233408,
      "step": 12625
    },
    {
      "epoch": 6.619496855345912,
      "grad_norm": 0.07144434005022049,
      "learning_rate": 0.0008461362834195316,
      "loss": 0.5182,
      "num_input_tokens_seen": 8237216,
      "step": 12630
    },
    {
      "epoch": 6.622117400419287,
      "grad_norm": 0.0901918113231659,
      "learning_rate": 0.0008459712193605842,
      "loss": 0.5346,
      "num_input_tokens_seen": 8240000,
      "step": 12635
    },
    {
      "epoch": 6.6247379454926625,
      "grad_norm": 0.10405346006155014,
      "learning_rate": 0.0008458060829284048,
      "loss": 0.4752,
      "num_input_tokens_seen": 8242848,
      "step": 12640
    },
    {
      "epoch": 6.627358490566038,
      "grad_norm": 0.10505606234073639,
      "learning_rate": 0.0008456408741575378,
      "loss": 0.4139,
      "num_input_tokens_seen": 8246304,
      "step": 12645
    },
    {
      "epoch": 6.629979035639413,
      "grad_norm": 0.05314522236585617,
      "learning_rate": 0.0008454755930825435,
      "loss": 0.4318,
      "num_input_tokens_seen": 8249984,
      "step": 12650
    },
    {
      "epoch": 6.632599580712788,
      "grad_norm": 0.09774128347635269,
      "learning_rate": 0.0008453102397379963,
      "loss": 0.5294,
      "num_input_tokens_seen": 8254240,
      "step": 12655
    },
    {
      "epoch": 6.635220125786163,
      "grad_norm": 0.0708106979727745,
      "learning_rate": 0.0008451448141584868,
      "loss": 0.401,
      "num_input_tokens_seen": 8257312,
      "step": 12660
    },
    {
      "epoch": 6.637840670859539,
      "grad_norm": 0.059299319982528687,
      "learning_rate": 0.0008449793163786194,
      "loss": 0.4558,
      "num_input_tokens_seen": 8261216,
      "step": 12665
    },
    {
      "epoch": 6.640461215932914,
      "grad_norm": 0.09654977917671204,
      "learning_rate": 0.0008448137464330149,
      "loss": 0.4347,
      "num_input_tokens_seen": 8264608,
      "step": 12670
    },
    {
      "epoch": 6.6430817610062896,
      "grad_norm": 0.17235901951789856,
      "learning_rate": 0.0008446481043563086,
      "loss": 0.4949,
      "num_input_tokens_seen": 8268384,
      "step": 12675
    },
    {
      "epoch": 6.645702306079665,
      "grad_norm": 0.06980688124895096,
      "learning_rate": 0.0008444823901831507,
      "loss": 0.4476,
      "num_input_tokens_seen": 8272000,
      "step": 12680
    },
    {
      "epoch": 6.64832285115304,
      "grad_norm": 0.0874103382229805,
      "learning_rate": 0.0008443166039482067,
      "loss": 0.3707,
      "num_input_tokens_seen": 8275776,
      "step": 12685
    },
    {
      "epoch": 6.650943396226415,
      "grad_norm": 0.12107431143522263,
      "learning_rate": 0.0008441507456861575,
      "loss": 0.4095,
      "num_input_tokens_seen": 8279008,
      "step": 12690
    },
    {
      "epoch": 6.65356394129979,
      "grad_norm": 0.0578635074198246,
      "learning_rate": 0.0008439848154316985,
      "loss": 0.4693,
      "num_input_tokens_seen": 8281952,
      "step": 12695
    },
    {
      "epoch": 6.656184486373165,
      "grad_norm": 0.04941476136445999,
      "learning_rate": 0.0008438188132195406,
      "loss": 0.3935,
      "num_input_tokens_seen": 8285472,
      "step": 12700
    },
    {
      "epoch": 6.658805031446541,
      "grad_norm": 0.12826964259147644,
      "learning_rate": 0.0008436527390844094,
      "loss": 0.5488,
      "num_input_tokens_seen": 8288416,
      "step": 12705
    },
    {
      "epoch": 6.661425576519916,
      "grad_norm": 0.05464569479227066,
      "learning_rate": 0.0008434865930610458,
      "loss": 0.4886,
      "num_input_tokens_seen": 8292064,
      "step": 12710
    },
    {
      "epoch": 6.664046121593291,
      "grad_norm": 0.0864122062921524,
      "learning_rate": 0.0008433203751842057,
      "loss": 0.4834,
      "num_input_tokens_seen": 8295008,
      "step": 12715
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.06255955249071121,
      "learning_rate": 0.0008431540854886598,
      "loss": 0.4277,
      "num_input_tokens_seen": 8298432,
      "step": 12720
    },
    {
      "epoch": 6.669287211740042,
      "grad_norm": 0.06215141713619232,
      "learning_rate": 0.0008429877240091941,
      "loss": 0.4358,
      "num_input_tokens_seen": 8302240,
      "step": 12725
    },
    {
      "epoch": 6.671907756813417,
      "grad_norm": 0.09343042969703674,
      "learning_rate": 0.0008428212907806097,
      "loss": 0.5249,
      "num_input_tokens_seen": 8304960,
      "step": 12730
    },
    {
      "epoch": 6.6745283018867925,
      "grad_norm": 0.1565130650997162,
      "learning_rate": 0.0008426547858377222,
      "loss": 0.3754,
      "num_input_tokens_seen": 8308320,
      "step": 12735
    },
    {
      "epoch": 6.677148846960168,
      "grad_norm": 0.08441407233476639,
      "learning_rate": 0.0008424882092153629,
      "loss": 0.4585,
      "num_input_tokens_seen": 8311360,
      "step": 12740
    },
    {
      "epoch": 6.679769392033543,
      "grad_norm": 0.10420465469360352,
      "learning_rate": 0.0008423215609483774,
      "loss": 0.4932,
      "num_input_tokens_seen": 8315680,
      "step": 12745
    },
    {
      "epoch": 6.682389937106918,
      "grad_norm": 0.08734307438135147,
      "learning_rate": 0.0008421548410716269,
      "loss": 0.5664,
      "num_input_tokens_seen": 8319200,
      "step": 12750
    },
    {
      "epoch": 6.685010482180293,
      "grad_norm": 0.09131255000829697,
      "learning_rate": 0.0008419880496199869,
      "loss": 0.4498,
      "num_input_tokens_seen": 8322816,
      "step": 12755
    },
    {
      "epoch": 6.687631027253669,
      "grad_norm": 0.08994173258543015,
      "learning_rate": 0.0008418211866283488,
      "loss": 0.4765,
      "num_input_tokens_seen": 8325408,
      "step": 12760
    },
    {
      "epoch": 6.690251572327044,
      "grad_norm": 0.060451533645391464,
      "learning_rate": 0.0008416542521316182,
      "loss": 0.3864,
      "num_input_tokens_seen": 8329408,
      "step": 12765
    },
    {
      "epoch": 6.6928721174004195,
      "grad_norm": 0.12187057733535767,
      "learning_rate": 0.0008414872461647156,
      "loss": 0.6115,
      "num_input_tokens_seen": 8332736,
      "step": 12770
    },
    {
      "epoch": 6.695492662473795,
      "grad_norm": 0.06014545261859894,
      "learning_rate": 0.0008413201687625772,
      "loss": 0.2981,
      "num_input_tokens_seen": 8338272,
      "step": 12775
    },
    {
      "epoch": 6.69811320754717,
      "grad_norm": 0.1537647694349289,
      "learning_rate": 0.0008411530199601534,
      "loss": 0.6451,
      "num_input_tokens_seen": 8341472,
      "step": 12780
    },
    {
      "epoch": 6.700733752620545,
      "grad_norm": 0.1015247330069542,
      "learning_rate": 0.0008409857997924097,
      "loss": 0.4336,
      "num_input_tokens_seen": 8344288,
      "step": 12785
    },
    {
      "epoch": 6.70335429769392,
      "grad_norm": 0.06417663395404816,
      "learning_rate": 0.0008408185082943271,
      "loss": 0.4324,
      "num_input_tokens_seen": 8347392,
      "step": 12790
    },
    {
      "epoch": 6.705974842767295,
      "grad_norm": 0.08683960884809494,
      "learning_rate": 0.0008406511455009007,
      "loss": 0.4294,
      "num_input_tokens_seen": 8350656,
      "step": 12795
    },
    {
      "epoch": 6.7085953878406706,
      "grad_norm": 0.0620427280664444,
      "learning_rate": 0.000840483711447141,
      "loss": 0.4348,
      "num_input_tokens_seen": 8353888,
      "step": 12800
    },
    {
      "epoch": 6.711215932914046,
      "grad_norm": 0.11711497604846954,
      "learning_rate": 0.0008403162061680736,
      "loss": 0.4652,
      "num_input_tokens_seen": 8356736,
      "step": 12805
    },
    {
      "epoch": 6.713836477987421,
      "grad_norm": 0.0758441910147667,
      "learning_rate": 0.0008401486296987382,
      "loss": 0.4642,
      "num_input_tokens_seen": 8360992,
      "step": 12810
    },
    {
      "epoch": 6.716457023060797,
      "grad_norm": 0.07545538246631622,
      "learning_rate": 0.0008399809820741901,
      "loss": 0.4635,
      "num_input_tokens_seen": 8364000,
      "step": 12815
    },
    {
      "epoch": 6.719077568134172,
      "grad_norm": 0.1012231707572937,
      "learning_rate": 0.0008398132633294993,
      "loss": 0.3957,
      "num_input_tokens_seen": 8366304,
      "step": 12820
    },
    {
      "epoch": 6.721698113207547,
      "grad_norm": 0.0674777701497078,
      "learning_rate": 0.0008396454734997506,
      "loss": 0.3767,
      "num_input_tokens_seen": 8369792,
      "step": 12825
    },
    {
      "epoch": 6.7243186582809225,
      "grad_norm": 0.10295267403125763,
      "learning_rate": 0.0008394776126200436,
      "loss": 0.4095,
      "num_input_tokens_seen": 8372384,
      "step": 12830
    },
    {
      "epoch": 6.726939203354298,
      "grad_norm": 0.07107633352279663,
      "learning_rate": 0.0008393096807254931,
      "loss": 0.4371,
      "num_input_tokens_seen": 8374720,
      "step": 12835
    },
    {
      "epoch": 6.729559748427673,
      "grad_norm": 0.10065519064664841,
      "learning_rate": 0.0008391416778512284,
      "loss": 0.5546,
      "num_input_tokens_seen": 8378144,
      "step": 12840
    },
    {
      "epoch": 6.732180293501048,
      "grad_norm": 0.08881823718547821,
      "learning_rate": 0.0008389736040323939,
      "loss": 0.4281,
      "num_input_tokens_seen": 8381536,
      "step": 12845
    },
    {
      "epoch": 6.734800838574423,
      "grad_norm": 0.06792403757572174,
      "learning_rate": 0.0008388054593041485,
      "loss": 0.4457,
      "num_input_tokens_seen": 8384416,
      "step": 12850
    },
    {
      "epoch": 6.737421383647799,
      "grad_norm": 0.0841895043849945,
      "learning_rate": 0.0008386372437016665,
      "loss": 0.3916,
      "num_input_tokens_seen": 8387616,
      "step": 12855
    },
    {
      "epoch": 6.740041928721174,
      "grad_norm": 0.10852161794900894,
      "learning_rate": 0.0008384689572601364,
      "loss": 0.4568,
      "num_input_tokens_seen": 8390208,
      "step": 12860
    },
    {
      "epoch": 6.7426624737945495,
      "grad_norm": 0.09966512024402618,
      "learning_rate": 0.000838300600014762,
      "loss": 0.4068,
      "num_input_tokens_seen": 8393792,
      "step": 12865
    },
    {
      "epoch": 6.745283018867925,
      "grad_norm": 0.08097311109304428,
      "learning_rate": 0.0008381321720007613,
      "loss": 0.5482,
      "num_input_tokens_seen": 8397120,
      "step": 12870
    },
    {
      "epoch": 6.7479035639413,
      "grad_norm": 0.11955004185438156,
      "learning_rate": 0.0008379636732533681,
      "loss": 0.5794,
      "num_input_tokens_seen": 8400160,
      "step": 12875
    },
    {
      "epoch": 6.750524109014675,
      "grad_norm": 0.09028202295303345,
      "learning_rate": 0.0008377951038078302,
      "loss": 0.6804,
      "num_input_tokens_seen": 8404032,
      "step": 12880
    },
    {
      "epoch": 6.75314465408805,
      "grad_norm": 0.08644835650920868,
      "learning_rate": 0.0008376264636994102,
      "loss": 0.4872,
      "num_input_tokens_seen": 8407392,
      "step": 12885
    },
    {
      "epoch": 6.755765199161425,
      "grad_norm": 0.15323171019554138,
      "learning_rate": 0.000837457752963386,
      "loss": 0.495,
      "num_input_tokens_seen": 8410656,
      "step": 12890
    },
    {
      "epoch": 6.7583857442348005,
      "grad_norm": 0.15990157425403595,
      "learning_rate": 0.0008372889716350496,
      "loss": 0.5307,
      "num_input_tokens_seen": 8413344,
      "step": 12895
    },
    {
      "epoch": 6.761006289308176,
      "grad_norm": 0.053266242146492004,
      "learning_rate": 0.0008371201197497084,
      "loss": 0.4827,
      "num_input_tokens_seen": 8416960,
      "step": 12900
    },
    {
      "epoch": 6.763626834381551,
      "grad_norm": 0.07844261080026627,
      "learning_rate": 0.0008369511973426842,
      "loss": 0.5753,
      "num_input_tokens_seen": 8420352,
      "step": 12905
    },
    {
      "epoch": 6.766247379454927,
      "grad_norm": 0.08548973500728607,
      "learning_rate": 0.0008367822044493138,
      "loss": 0.4976,
      "num_input_tokens_seen": 8423360,
      "step": 12910
    },
    {
      "epoch": 6.768867924528302,
      "grad_norm": 0.07555176317691803,
      "learning_rate": 0.0008366131411049483,
      "loss": 0.306,
      "num_input_tokens_seen": 8426816,
      "step": 12915
    },
    {
      "epoch": 6.771488469601677,
      "grad_norm": 0.08270000666379929,
      "learning_rate": 0.000836444007344954,
      "loss": 0.5961,
      "num_input_tokens_seen": 8429856,
      "step": 12920
    },
    {
      "epoch": 6.774109014675052,
      "grad_norm": 0.11537697911262512,
      "learning_rate": 0.0008362748032047119,
      "loss": 0.601,
      "num_input_tokens_seen": 8433056,
      "step": 12925
    },
    {
      "epoch": 6.776729559748428,
      "grad_norm": 0.0636945366859436,
      "learning_rate": 0.0008361055287196171,
      "loss": 0.5084,
      "num_input_tokens_seen": 8436448,
      "step": 12930
    },
    {
      "epoch": 6.779350104821803,
      "grad_norm": 0.06615664809942245,
      "learning_rate": 0.0008359361839250803,
      "loss": 0.3257,
      "num_input_tokens_seen": 8439200,
      "step": 12935
    },
    {
      "epoch": 6.781970649895178,
      "grad_norm": 0.14274559915065765,
      "learning_rate": 0.0008357667688565265,
      "loss": 0.4871,
      "num_input_tokens_seen": 8443072,
      "step": 12940
    },
    {
      "epoch": 6.784591194968553,
      "grad_norm": 0.05890651047229767,
      "learning_rate": 0.0008355972835493951,
      "loss": 0.4065,
      "num_input_tokens_seen": 8445952,
      "step": 12945
    },
    {
      "epoch": 6.787211740041929,
      "grad_norm": 0.05983971059322357,
      "learning_rate": 0.0008354277280391408,
      "loss": 0.5755,
      "num_input_tokens_seen": 8448768,
      "step": 12950
    },
    {
      "epoch": 6.789832285115304,
      "grad_norm": 0.09190454334020615,
      "learning_rate": 0.0008352581023612324,
      "loss": 0.33,
      "num_input_tokens_seen": 8451232,
      "step": 12955
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 0.09163102507591248,
      "learning_rate": 0.0008350884065511541,
      "loss": 0.4144,
      "num_input_tokens_seen": 8454688,
      "step": 12960
    },
    {
      "epoch": 6.795073375262055,
      "grad_norm": 0.11425977945327759,
      "learning_rate": 0.0008349186406444038,
      "loss": 0.3353,
      "num_input_tokens_seen": 8458048,
      "step": 12965
    },
    {
      "epoch": 6.79769392033543,
      "grad_norm": 0.098201684653759,
      "learning_rate": 0.0008347488046764949,
      "loss": 0.5017,
      "num_input_tokens_seen": 8460960,
      "step": 12970
    },
    {
      "epoch": 6.800314465408805,
      "grad_norm": 0.08147728443145752,
      "learning_rate": 0.0008345788986829551,
      "loss": 0.5133,
      "num_input_tokens_seen": 8464448,
      "step": 12975
    },
    {
      "epoch": 6.80293501048218,
      "grad_norm": 0.06871751695871353,
      "learning_rate": 0.0008344089226993268,
      "loss": 0.3469,
      "num_input_tokens_seen": 8468000,
      "step": 12980
    },
    {
      "epoch": 6.805555555555555,
      "grad_norm": 0.060036029666662216,
      "learning_rate": 0.000834238876761167,
      "loss": 0.4938,
      "num_input_tokens_seen": 8472288,
      "step": 12985
    },
    {
      "epoch": 6.8081761006289305,
      "grad_norm": 0.08604501932859421,
      "learning_rate": 0.0008340687609040473,
      "loss": 0.3875,
      "num_input_tokens_seen": 8475616,
      "step": 12990
    },
    {
      "epoch": 6.810796645702306,
      "grad_norm": 0.16364887356758118,
      "learning_rate": 0.0008338985751635542,
      "loss": 0.6011,
      "num_input_tokens_seen": 8477984,
      "step": 12995
    },
    {
      "epoch": 6.813417190775681,
      "grad_norm": 0.09293633699417114,
      "learning_rate": 0.0008337283195752883,
      "loss": 0.472,
      "num_input_tokens_seen": 8480864,
      "step": 13000
    },
    {
      "epoch": 6.816037735849057,
      "grad_norm": 0.06293078511953354,
      "learning_rate": 0.0008335579941748654,
      "loss": 0.5079,
      "num_input_tokens_seen": 8483680,
      "step": 13005
    },
    {
      "epoch": 6.818658280922432,
      "grad_norm": 0.06955090165138245,
      "learning_rate": 0.0008333875989979156,
      "loss": 0.4217,
      "num_input_tokens_seen": 8487296,
      "step": 13010
    },
    {
      "epoch": 6.821278825995807,
      "grad_norm": 0.08030242472887039,
      "learning_rate": 0.0008332171340800835,
      "loss": 0.3954,
      "num_input_tokens_seen": 8490528,
      "step": 13015
    },
    {
      "epoch": 6.823899371069182,
      "grad_norm": 0.09037308394908905,
      "learning_rate": 0.0008330465994570287,
      "loss": 0.4765,
      "num_input_tokens_seen": 8493536,
      "step": 13020
    },
    {
      "epoch": 6.826519916142558,
      "grad_norm": 0.15696077048778534,
      "learning_rate": 0.0008328759951644247,
      "loss": 0.5643,
      "num_input_tokens_seen": 8497472,
      "step": 13025
    },
    {
      "epoch": 6.829140461215933,
      "grad_norm": 0.09591522812843323,
      "learning_rate": 0.0008327053212379603,
      "loss": 0.5276,
      "num_input_tokens_seen": 8500288,
      "step": 13030
    },
    {
      "epoch": 6.831761006289308,
      "grad_norm": 0.10998334735631943,
      "learning_rate": 0.0008325345777133382,
      "loss": 0.4773,
      "num_input_tokens_seen": 8503136,
      "step": 13035
    },
    {
      "epoch": 6.834381551362683,
      "grad_norm": 0.06056833267211914,
      "learning_rate": 0.0008323637646262763,
      "loss": 0.3163,
      "num_input_tokens_seen": 8507168,
      "step": 13040
    },
    {
      "epoch": 6.837002096436059,
      "grad_norm": 0.10691709071397781,
      "learning_rate": 0.0008321928820125067,
      "loss": 0.4946,
      "num_input_tokens_seen": 8509760,
      "step": 13045
    },
    {
      "epoch": 6.839622641509434,
      "grad_norm": 0.057928334921598434,
      "learning_rate": 0.000832021929907776,
      "loss": 0.352,
      "num_input_tokens_seen": 8512672,
      "step": 13050
    },
    {
      "epoch": 6.8422431865828095,
      "grad_norm": 0.07998363673686981,
      "learning_rate": 0.0008318509083478454,
      "loss": 0.4384,
      "num_input_tokens_seen": 8516320,
      "step": 13055
    },
    {
      "epoch": 6.844863731656185,
      "grad_norm": 0.09305883198976517,
      "learning_rate": 0.0008316798173684907,
      "loss": 0.4656,
      "num_input_tokens_seen": 8519456,
      "step": 13060
    },
    {
      "epoch": 6.84748427672956,
      "grad_norm": 0.2144283801317215,
      "learning_rate": 0.0008315086570055024,
      "loss": 0.5708,
      "num_input_tokens_seen": 8522560,
      "step": 13065
    },
    {
      "epoch": 6.850104821802935,
      "grad_norm": 0.08070152252912521,
      "learning_rate": 0.000831337427294685,
      "loss": 0.4366,
      "num_input_tokens_seen": 8525792,
      "step": 13070
    },
    {
      "epoch": 6.85272536687631,
      "grad_norm": 0.12407448142766953,
      "learning_rate": 0.0008311661282718579,
      "loss": 0.4427,
      "num_input_tokens_seen": 8528704,
      "step": 13075
    },
    {
      "epoch": 6.855345911949685,
      "grad_norm": 0.09960034489631653,
      "learning_rate": 0.0008309947599728549,
      "loss": 0.4338,
      "num_input_tokens_seen": 8532000,
      "step": 13080
    },
    {
      "epoch": 6.8579664570230605,
      "grad_norm": 0.11514933407306671,
      "learning_rate": 0.0008308233224335243,
      "loss": 0.5273,
      "num_input_tokens_seen": 8535136,
      "step": 13085
    },
    {
      "epoch": 6.860587002096436,
      "grad_norm": 0.1237306296825409,
      "learning_rate": 0.0008306518156897289,
      "loss": 0.5353,
      "num_input_tokens_seen": 8538592,
      "step": 13090
    },
    {
      "epoch": 6.863207547169811,
      "grad_norm": 0.07698046416044235,
      "learning_rate": 0.0008304802397773459,
      "loss": 0.5044,
      "num_input_tokens_seen": 8541344,
      "step": 13095
    },
    {
      "epoch": 6.865828092243187,
      "grad_norm": 0.10289334505796432,
      "learning_rate": 0.0008303085947322672,
      "loss": 0.4897,
      "num_input_tokens_seen": 8544576,
      "step": 13100
    },
    {
      "epoch": 6.868448637316562,
      "grad_norm": 0.06611794233322144,
      "learning_rate": 0.0008301368805903988,
      "loss": 0.4547,
      "num_input_tokens_seen": 8547680,
      "step": 13105
    },
    {
      "epoch": 6.871069182389937,
      "grad_norm": 0.057211339473724365,
      "learning_rate": 0.0008299650973876611,
      "loss": 0.5323,
      "num_input_tokens_seen": 8550720,
      "step": 13110
    },
    {
      "epoch": 6.873689727463312,
      "grad_norm": 0.11332406848669052,
      "learning_rate": 0.0008297932451599898,
      "loss": 0.4107,
      "num_input_tokens_seen": 8554112,
      "step": 13115
    },
    {
      "epoch": 6.876310272536688,
      "grad_norm": 0.08523738384246826,
      "learning_rate": 0.0008296213239433339,
      "loss": 0.3433,
      "num_input_tokens_seen": 8557856,
      "step": 13120
    },
    {
      "epoch": 6.878930817610063,
      "grad_norm": 0.0870550200343132,
      "learning_rate": 0.0008294493337736577,
      "loss": 0.4905,
      "num_input_tokens_seen": 8560960,
      "step": 13125
    },
    {
      "epoch": 6.881551362683438,
      "grad_norm": 0.1171410083770752,
      "learning_rate": 0.0008292772746869392,
      "loss": 0.4376,
      "num_input_tokens_seen": 8564288,
      "step": 13130
    },
    {
      "epoch": 6.884171907756813,
      "grad_norm": 0.03774799406528473,
      "learning_rate": 0.0008291051467191717,
      "loss": 0.4508,
      "num_input_tokens_seen": 8568608,
      "step": 13135
    },
    {
      "epoch": 6.886792452830189,
      "grad_norm": 0.07576742768287659,
      "learning_rate": 0.000828932949906362,
      "loss": 0.4379,
      "num_input_tokens_seen": 8571872,
      "step": 13140
    },
    {
      "epoch": 6.889412997903564,
      "grad_norm": 0.11069905757904053,
      "learning_rate": 0.0008287606842845319,
      "loss": 0.5778,
      "num_input_tokens_seen": 8574336,
      "step": 13145
    },
    {
      "epoch": 6.8920335429769395,
      "grad_norm": 0.07110822200775146,
      "learning_rate": 0.0008285883498897174,
      "loss": 0.4574,
      "num_input_tokens_seen": 8577248,
      "step": 13150
    },
    {
      "epoch": 6.894654088050315,
      "grad_norm": 0.06785587221384048,
      "learning_rate": 0.0008284159467579686,
      "loss": 0.4605,
      "num_input_tokens_seen": 8581120,
      "step": 13155
    },
    {
      "epoch": 6.89727463312369,
      "grad_norm": 0.14420828223228455,
      "learning_rate": 0.0008282434749253507,
      "loss": 0.3867,
      "num_input_tokens_seen": 8585280,
      "step": 13160
    },
    {
      "epoch": 6.899895178197065,
      "grad_norm": 0.059095364063978195,
      "learning_rate": 0.0008280709344279426,
      "loss": 0.4382,
      "num_input_tokens_seen": 8590464,
      "step": 13165
    },
    {
      "epoch": 6.90251572327044,
      "grad_norm": 0.04628896340727806,
      "learning_rate": 0.0008278983253018377,
      "loss": 0.4316,
      "num_input_tokens_seen": 8593408,
      "step": 13170
    },
    {
      "epoch": 6.905136268343815,
      "grad_norm": 0.09239279478788376,
      "learning_rate": 0.0008277256475831443,
      "loss": 0.4278,
      "num_input_tokens_seen": 8596832,
      "step": 13175
    },
    {
      "epoch": 6.9077568134171905,
      "grad_norm": 0.07017997652292252,
      "learning_rate": 0.0008275529013079841,
      "loss": 0.3475,
      "num_input_tokens_seen": 8600000,
      "step": 13180
    },
    {
      "epoch": 6.910377358490566,
      "grad_norm": 0.12654094398021698,
      "learning_rate": 0.0008273800865124941,
      "loss": 0.4878,
      "num_input_tokens_seen": 8602560,
      "step": 13185
    },
    {
      "epoch": 6.912997903563941,
      "grad_norm": 0.09851802885532379,
      "learning_rate": 0.0008272072032328246,
      "loss": 0.5999,
      "num_input_tokens_seen": 8605088,
      "step": 13190
    },
    {
      "epoch": 6.915618448637317,
      "grad_norm": 0.09429441392421722,
      "learning_rate": 0.0008270342515051414,
      "loss": 0.5354,
      "num_input_tokens_seen": 8608704,
      "step": 13195
    },
    {
      "epoch": 6.918238993710692,
      "grad_norm": 0.055294789373874664,
      "learning_rate": 0.0008268612313656238,
      "loss": 0.4455,
      "num_input_tokens_seen": 8612320,
      "step": 13200
    },
    {
      "epoch": 6.920859538784067,
      "grad_norm": 0.08319047838449478,
      "learning_rate": 0.0008266881428504655,
      "loss": 0.5224,
      "num_input_tokens_seen": 8615360,
      "step": 13205
    },
    {
      "epoch": 6.923480083857442,
      "grad_norm": 0.07191596180200577,
      "learning_rate": 0.000826514985995875,
      "loss": 0.3791,
      "num_input_tokens_seen": 8618464,
      "step": 13210
    },
    {
      "epoch": 6.926100628930818,
      "grad_norm": 0.07215634733438492,
      "learning_rate": 0.0008263417608380743,
      "loss": 0.2977,
      "num_input_tokens_seen": 8621664,
      "step": 13215
    },
    {
      "epoch": 6.928721174004193,
      "grad_norm": 0.06206538900732994,
      "learning_rate": 0.0008261684674133003,
      "loss": 0.3941,
      "num_input_tokens_seen": 8624608,
      "step": 13220
    },
    {
      "epoch": 6.931341719077568,
      "grad_norm": 0.1299976408481598,
      "learning_rate": 0.0008259951057578041,
      "loss": 0.382,
      "num_input_tokens_seen": 8626880,
      "step": 13225
    },
    {
      "epoch": 6.933962264150943,
      "grad_norm": 0.1086968407034874,
      "learning_rate": 0.000825821675907851,
      "loss": 0.5649,
      "num_input_tokens_seen": 8630880,
      "step": 13230
    },
    {
      "epoch": 6.936582809224319,
      "grad_norm": 0.08479819446802139,
      "learning_rate": 0.0008256481778997205,
      "loss": 0.5179,
      "num_input_tokens_seen": 8633856,
      "step": 13235
    },
    {
      "epoch": 6.939203354297694,
      "grad_norm": 0.07141540944576263,
      "learning_rate": 0.0008254746117697062,
      "loss": 0.5288,
      "num_input_tokens_seen": 8636288,
      "step": 13240
    },
    {
      "epoch": 6.9418238993710695,
      "grad_norm": 0.15845836699008942,
      "learning_rate": 0.0008253009775541165,
      "loss": 0.6388,
      "num_input_tokens_seen": 8638880,
      "step": 13245
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 0.11585526913404465,
      "learning_rate": 0.0008251272752892733,
      "loss": 0.5228,
      "num_input_tokens_seen": 8642080,
      "step": 13250
    },
    {
      "epoch": 6.94706498951782,
      "grad_norm": 0.07380173355340958,
      "learning_rate": 0.0008249535050115135,
      "loss": 0.4319,
      "num_input_tokens_seen": 8644672,
      "step": 13255
    },
    {
      "epoch": 6.949685534591195,
      "grad_norm": 0.07845361530780792,
      "learning_rate": 0.0008247796667571879,
      "loss": 0.3994,
      "num_input_tokens_seen": 8648160,
      "step": 13260
    },
    {
      "epoch": 6.95230607966457,
      "grad_norm": 0.04875190183520317,
      "learning_rate": 0.000824605760562661,
      "loss": 0.4164,
      "num_input_tokens_seen": 8651168,
      "step": 13265
    },
    {
      "epoch": 6.954926624737945,
      "grad_norm": 0.05789134278893471,
      "learning_rate": 0.0008244317864643125,
      "loss": 0.4114,
      "num_input_tokens_seen": 8653984,
      "step": 13270
    },
    {
      "epoch": 6.9575471698113205,
      "grad_norm": 0.0924491435289383,
      "learning_rate": 0.0008242577444985355,
      "loss": 0.3767,
      "num_input_tokens_seen": 8656544,
      "step": 13275
    },
    {
      "epoch": 6.960167714884696,
      "grad_norm": 0.11323946714401245,
      "learning_rate": 0.0008240836347017375,
      "loss": 0.4178,
      "num_input_tokens_seen": 8659456,
      "step": 13280
    },
    {
      "epoch": 6.962788259958071,
      "grad_norm": 0.07424240559339523,
      "learning_rate": 0.0008239094571103408,
      "loss": 0.3962,
      "num_input_tokens_seen": 8662112,
      "step": 13285
    },
    {
      "epoch": 6.965408805031447,
      "grad_norm": 0.13530659675598145,
      "learning_rate": 0.0008237352117607808,
      "loss": 0.5381,
      "num_input_tokens_seen": 8665760,
      "step": 13290
    },
    {
      "epoch": 6.968029350104822,
      "grad_norm": 0.06713801622390747,
      "learning_rate": 0.000823560898689508,
      "loss": 0.5083,
      "num_input_tokens_seen": 8669152,
      "step": 13295
    },
    {
      "epoch": 6.970649895178197,
      "grad_norm": 0.15095387399196625,
      "learning_rate": 0.0008233865179329865,
      "loss": 0.4564,
      "num_input_tokens_seen": 8671648,
      "step": 13300
    },
    {
      "epoch": 6.973270440251572,
      "grad_norm": 0.07876161485910416,
      "learning_rate": 0.0008232120695276948,
      "loss": 0.5372,
      "num_input_tokens_seen": 8676192,
      "step": 13305
    },
    {
      "epoch": 6.975890985324948,
      "grad_norm": 0.11395036429166794,
      "learning_rate": 0.0008230375535101255,
      "loss": 0.5555,
      "num_input_tokens_seen": 8679008,
      "step": 13310
    },
    {
      "epoch": 6.978511530398323,
      "grad_norm": 0.07141033560037613,
      "learning_rate": 0.0008228629699167855,
      "loss": 0.4376,
      "num_input_tokens_seen": 8682848,
      "step": 13315
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 0.1251201033592224,
      "learning_rate": 0.0008226883187841957,
      "loss": 0.3494,
      "num_input_tokens_seen": 8685376,
      "step": 13320
    },
    {
      "epoch": 6.983752620545073,
      "grad_norm": 0.09192091226577759,
      "learning_rate": 0.0008225136001488908,
      "loss": 0.6178,
      "num_input_tokens_seen": 8688672,
      "step": 13325
    },
    {
      "epoch": 6.986373165618449,
      "grad_norm": 0.0698409304022789,
      "learning_rate": 0.0008223388140474205,
      "loss": 0.5177,
      "num_input_tokens_seen": 8692256,
      "step": 13330
    },
    {
      "epoch": 6.988993710691824,
      "grad_norm": 0.11629147082567215,
      "learning_rate": 0.0008221639605163476,
      "loss": 0.57,
      "num_input_tokens_seen": 8695968,
      "step": 13335
    },
    {
      "epoch": 6.9916142557651995,
      "grad_norm": 0.09004097431898117,
      "learning_rate": 0.0008219890395922495,
      "loss": 0.4332,
      "num_input_tokens_seen": 8699808,
      "step": 13340
    },
    {
      "epoch": 6.994234800838575,
      "grad_norm": 0.10073720663785934,
      "learning_rate": 0.0008218140513117178,
      "loss": 0.53,
      "num_input_tokens_seen": 8702432,
      "step": 13345
    },
    {
      "epoch": 6.99685534591195,
      "grad_norm": 0.11474945396184921,
      "learning_rate": 0.0008216389957113582,
      "loss": 0.5606,
      "num_input_tokens_seen": 8705792,
      "step": 13350
    },
    {
      "epoch": 6.999475890985325,
      "grad_norm": 0.06749685108661652,
      "learning_rate": 0.0008214638728277899,
      "loss": 0.4937,
      "num_input_tokens_seen": 8709056,
      "step": 13355
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.45616984367370605,
      "eval_runtime": 13.521,
      "eval_samples_per_second": 62.717,
      "eval_steps_per_second": 15.679,
      "num_input_tokens_seen": 8709168,
      "step": 13356
    },
    {
      "epoch": 7.0020964360587,
      "grad_norm": 0.05177436023950577,
      "learning_rate": 0.0008212886826976469,
      "loss": 0.5044,
      "num_input_tokens_seen": 8711536,
      "step": 13360
    },
    {
      "epoch": 7.004716981132075,
      "grad_norm": 0.0921458825469017,
      "learning_rate": 0.0008211134253575769,
      "loss": 0.5529,
      "num_input_tokens_seen": 8715728,
      "step": 13365
    },
    {
      "epoch": 7.0073375262054505,
      "grad_norm": 0.086473748087883,
      "learning_rate": 0.000820938100844242,
      "loss": 0.4851,
      "num_input_tokens_seen": 8719216,
      "step": 13370
    },
    {
      "epoch": 7.009958071278826,
      "grad_norm": 0.05812876671552658,
      "learning_rate": 0.0008207627091943177,
      "loss": 0.3688,
      "num_input_tokens_seen": 8722160,
      "step": 13375
    },
    {
      "epoch": 7.012578616352202,
      "grad_norm": 0.11480481922626495,
      "learning_rate": 0.000820587250444494,
      "loss": 0.3595,
      "num_input_tokens_seen": 8725584,
      "step": 13380
    },
    {
      "epoch": 7.015199161425577,
      "grad_norm": 0.1025649830698967,
      "learning_rate": 0.0008204117246314752,
      "loss": 0.4327,
      "num_input_tokens_seen": 8728272,
      "step": 13385
    },
    {
      "epoch": 7.017819706498952,
      "grad_norm": 0.06591147929430008,
      "learning_rate": 0.0008202361317919788,
      "loss": 0.4537,
      "num_input_tokens_seen": 8731568,
      "step": 13390
    },
    {
      "epoch": 7.020440251572327,
      "grad_norm": 0.17156007885932922,
      "learning_rate": 0.0008200604719627373,
      "loss": 0.4914,
      "num_input_tokens_seen": 8735120,
      "step": 13395
    },
    {
      "epoch": 7.023060796645702,
      "grad_norm": 0.07772296667098999,
      "learning_rate": 0.0008198847451804964,
      "loss": 0.3515,
      "num_input_tokens_seen": 8737552,
      "step": 13400
    },
    {
      "epoch": 7.0256813417190775,
      "grad_norm": 0.07613659650087357,
      "learning_rate": 0.0008197089514820162,
      "loss": 0.4118,
      "num_input_tokens_seen": 8741712,
      "step": 13405
    },
    {
      "epoch": 7.028301886792453,
      "grad_norm": 0.07714203745126724,
      "learning_rate": 0.0008195330909040708,
      "loss": 0.5109,
      "num_input_tokens_seen": 8744464,
      "step": 13410
    },
    {
      "epoch": 7.030922431865828,
      "grad_norm": 0.04743829369544983,
      "learning_rate": 0.0008193571634834483,
      "loss": 0.4089,
      "num_input_tokens_seen": 8748048,
      "step": 13415
    },
    {
      "epoch": 7.033542976939203,
      "grad_norm": 0.08470834046602249,
      "learning_rate": 0.0008191811692569506,
      "loss": 0.3926,
      "num_input_tokens_seen": 8751024,
      "step": 13420
    },
    {
      "epoch": 7.036163522012578,
      "grad_norm": 0.14211145043373108,
      "learning_rate": 0.0008190051082613936,
      "loss": 0.4449,
      "num_input_tokens_seen": 8753264,
      "step": 13425
    },
    {
      "epoch": 7.038784067085954,
      "grad_norm": 0.09016682952642441,
      "learning_rate": 0.0008188289805336074,
      "loss": 0.47,
      "num_input_tokens_seen": 8756240,
      "step": 13430
    },
    {
      "epoch": 7.0414046121593294,
      "grad_norm": 0.15210625529289246,
      "learning_rate": 0.0008186527861104359,
      "loss": 0.3895,
      "num_input_tokens_seen": 8759184,
      "step": 13435
    },
    {
      "epoch": 7.044025157232705,
      "grad_norm": 0.09258528053760529,
      "learning_rate": 0.0008184765250287369,
      "loss": 0.5598,
      "num_input_tokens_seen": 8762864,
      "step": 13440
    },
    {
      "epoch": 7.04664570230608,
      "grad_norm": 0.16178612411022186,
      "learning_rate": 0.0008183001973253823,
      "loss": 0.3384,
      "num_input_tokens_seen": 8767856,
      "step": 13445
    },
    {
      "epoch": 7.049266247379455,
      "grad_norm": 0.07522746175527573,
      "learning_rate": 0.0008181238030372576,
      "loss": 0.5116,
      "num_input_tokens_seen": 8771344,
      "step": 13450
    },
    {
      "epoch": 7.05188679245283,
      "grad_norm": 0.05553904175758362,
      "learning_rate": 0.0008179473422012628,
      "loss": 0.4563,
      "num_input_tokens_seen": 8775216,
      "step": 13455
    },
    {
      "epoch": 7.054507337526205,
      "grad_norm": 0.06574542075395584,
      "learning_rate": 0.0008177708148543114,
      "loss": 0.514,
      "num_input_tokens_seen": 8779536,
      "step": 13460
    },
    {
      "epoch": 7.0571278825995805,
      "grad_norm": 0.09401366114616394,
      "learning_rate": 0.0008175942210333307,
      "loss": 0.5847,
      "num_input_tokens_seen": 8782640,
      "step": 13465
    },
    {
      "epoch": 7.059748427672956,
      "grad_norm": 0.07985740900039673,
      "learning_rate": 0.0008174175607752626,
      "loss": 0.6584,
      "num_input_tokens_seen": 8786384,
      "step": 13470
    },
    {
      "epoch": 7.062368972746331,
      "grad_norm": 0.08363495767116547,
      "learning_rate": 0.0008172408341170617,
      "loss": 0.4124,
      "num_input_tokens_seen": 8789616,
      "step": 13475
    },
    {
      "epoch": 7.064989517819707,
      "grad_norm": 0.13015100359916687,
      "learning_rate": 0.0008170640410956976,
      "loss": 0.4984,
      "num_input_tokens_seen": 8792816,
      "step": 13480
    },
    {
      "epoch": 7.067610062893082,
      "grad_norm": 0.0738065242767334,
      "learning_rate": 0.0008168871817481536,
      "loss": 0.4206,
      "num_input_tokens_seen": 8796784,
      "step": 13485
    },
    {
      "epoch": 7.070230607966457,
      "grad_norm": 0.11113075166940689,
      "learning_rate": 0.0008167102561114261,
      "loss": 0.4806,
      "num_input_tokens_seen": 8800016,
      "step": 13490
    },
    {
      "epoch": 7.072851153039832,
      "grad_norm": 0.08530465513467789,
      "learning_rate": 0.0008165332642225265,
      "loss": 0.491,
      "num_input_tokens_seen": 8803088,
      "step": 13495
    },
    {
      "epoch": 7.0754716981132075,
      "grad_norm": 0.09297090023756027,
      "learning_rate": 0.0008163562061184791,
      "loss": 0.3977,
      "num_input_tokens_seen": 8806096,
      "step": 13500
    },
    {
      "epoch": 7.078092243186583,
      "grad_norm": 0.10221996158361435,
      "learning_rate": 0.0008161790818363227,
      "loss": 0.5086,
      "num_input_tokens_seen": 8808624,
      "step": 13505
    },
    {
      "epoch": 7.080712788259958,
      "grad_norm": 0.1020624116063118,
      "learning_rate": 0.0008160018914131094,
      "loss": 0.5875,
      "num_input_tokens_seen": 8811920,
      "step": 13510
    },
    {
      "epoch": 7.083333333333333,
      "grad_norm": 0.02069387212395668,
      "learning_rate": 0.0008158246348859057,
      "loss": 0.303,
      "num_input_tokens_seen": 8817904,
      "step": 13515
    },
    {
      "epoch": 7.085953878406708,
      "grad_norm": 0.06290017813444138,
      "learning_rate": 0.0008156473122917913,
      "loss": 0.4137,
      "num_input_tokens_seen": 8820496,
      "step": 13520
    },
    {
      "epoch": 7.088574423480084,
      "grad_norm": 0.08332987874746323,
      "learning_rate": 0.0008154699236678604,
      "loss": 0.5168,
      "num_input_tokens_seen": 8823120,
      "step": 13525
    },
    {
      "epoch": 7.091194968553459,
      "grad_norm": 0.08551155030727386,
      "learning_rate": 0.0008152924690512205,
      "loss": 0.4433,
      "num_input_tokens_seen": 8826192,
      "step": 13530
    },
    {
      "epoch": 7.093815513626835,
      "grad_norm": 0.07986601442098618,
      "learning_rate": 0.0008151149484789932,
      "loss": 0.4837,
      "num_input_tokens_seen": 8829456,
      "step": 13535
    },
    {
      "epoch": 7.09643605870021,
      "grad_norm": 0.11101774126291275,
      "learning_rate": 0.0008149373619883136,
      "loss": 0.5701,
      "num_input_tokens_seen": 8832112,
      "step": 13540
    },
    {
      "epoch": 7.099056603773585,
      "grad_norm": 0.1394275277853012,
      "learning_rate": 0.0008147597096163308,
      "loss": 0.5487,
      "num_input_tokens_seen": 8834832,
      "step": 13545
    },
    {
      "epoch": 7.10167714884696,
      "grad_norm": 0.05212678387761116,
      "learning_rate": 0.0008145819914002079,
      "loss": 0.5138,
      "num_input_tokens_seen": 8838480,
      "step": 13550
    },
    {
      "epoch": 7.104297693920335,
      "grad_norm": 0.15672613680362701,
      "learning_rate": 0.0008144042073771214,
      "loss": 0.3967,
      "num_input_tokens_seen": 8842128,
      "step": 13555
    },
    {
      "epoch": 7.1069182389937104,
      "grad_norm": 0.12060350924730301,
      "learning_rate": 0.0008142263575842615,
      "loss": 0.3805,
      "num_input_tokens_seen": 8845872,
      "step": 13560
    },
    {
      "epoch": 7.109538784067086,
      "grad_norm": 0.09619923681020737,
      "learning_rate": 0.0008140484420588323,
      "loss": 0.4291,
      "num_input_tokens_seen": 8848720,
      "step": 13565
    },
    {
      "epoch": 7.112159329140461,
      "grad_norm": 0.1433505415916443,
      "learning_rate": 0.000813870460838052,
      "loss": 0.3817,
      "num_input_tokens_seen": 8853072,
      "step": 13570
    },
    {
      "epoch": 7.114779874213837,
      "grad_norm": 0.061445631086826324,
      "learning_rate": 0.0008136924139591522,
      "loss": 0.4459,
      "num_input_tokens_seen": 8856208,
      "step": 13575
    },
    {
      "epoch": 7.117400419287212,
      "grad_norm": 0.08621474355459213,
      "learning_rate": 0.0008135143014593782,
      "loss": 0.4896,
      "num_input_tokens_seen": 8859600,
      "step": 13580
    },
    {
      "epoch": 7.120020964360587,
      "grad_norm": 0.1559056043624878,
      "learning_rate": 0.000813336123375989,
      "loss": 0.4631,
      "num_input_tokens_seen": 8862576,
      "step": 13585
    },
    {
      "epoch": 7.122641509433962,
      "grad_norm": 0.09162574261426926,
      "learning_rate": 0.0008131578797462575,
      "loss": 0.4266,
      "num_input_tokens_seen": 8866704,
      "step": 13590
    },
    {
      "epoch": 7.1252620545073375,
      "grad_norm": 0.1211182102560997,
      "learning_rate": 0.0008129795706074703,
      "loss": 0.3733,
      "num_input_tokens_seen": 8869584,
      "step": 13595
    },
    {
      "epoch": 7.127882599580713,
      "grad_norm": 0.07084301859140396,
      "learning_rate": 0.0008128011959969277,
      "loss": 0.4099,
      "num_input_tokens_seen": 8873040,
      "step": 13600
    },
    {
      "epoch": 7.130503144654088,
      "grad_norm": 0.1672949194908142,
      "learning_rate": 0.0008126227559519434,
      "loss": 0.5118,
      "num_input_tokens_seen": 8876208,
      "step": 13605
    },
    {
      "epoch": 7.133123689727463,
      "grad_norm": 0.07425224035978317,
      "learning_rate": 0.0008124442505098452,
      "loss": 0.4171,
      "num_input_tokens_seen": 8880240,
      "step": 13610
    },
    {
      "epoch": 7.135744234800838,
      "grad_norm": 0.16140444576740265,
      "learning_rate": 0.0008122656797079744,
      "loss": 0.4278,
      "num_input_tokens_seen": 8883280,
      "step": 13615
    },
    {
      "epoch": 7.138364779874214,
      "grad_norm": 0.06078909710049629,
      "learning_rate": 0.0008120870435836858,
      "loss": 0.4942,
      "num_input_tokens_seen": 8886192,
      "step": 13620
    },
    {
      "epoch": 7.140985324947589,
      "grad_norm": 0.08625829219818115,
      "learning_rate": 0.0008119083421743481,
      "loss": 0.3903,
      "num_input_tokens_seen": 8889456,
      "step": 13625
    },
    {
      "epoch": 7.143605870020965,
      "grad_norm": 0.1670597791671753,
      "learning_rate": 0.0008117295755173437,
      "loss": 0.3926,
      "num_input_tokens_seen": 8892528,
      "step": 13630
    },
    {
      "epoch": 7.14622641509434,
      "grad_norm": 0.10770602524280548,
      "learning_rate": 0.0008115507436500687,
      "loss": 0.4672,
      "num_input_tokens_seen": 8895824,
      "step": 13635
    },
    {
      "epoch": 7.148846960167715,
      "grad_norm": 0.1122632622718811,
      "learning_rate": 0.0008113718466099322,
      "loss": 0.4712,
      "num_input_tokens_seen": 8898544,
      "step": 13640
    },
    {
      "epoch": 7.15146750524109,
      "grad_norm": 0.09288862347602844,
      "learning_rate": 0.0008111928844343579,
      "loss": 0.577,
      "num_input_tokens_seen": 8902480,
      "step": 13645
    },
    {
      "epoch": 7.154088050314465,
      "grad_norm": 0.11052641272544861,
      "learning_rate": 0.0008110138571607823,
      "loss": 0.5884,
      "num_input_tokens_seen": 8905968,
      "step": 13650
    },
    {
      "epoch": 7.15670859538784,
      "grad_norm": 0.13357360661029816,
      "learning_rate": 0.0008108347648266563,
      "loss": 0.4214,
      "num_input_tokens_seen": 8909296,
      "step": 13655
    },
    {
      "epoch": 7.159329140461216,
      "grad_norm": 0.10566683113574982,
      "learning_rate": 0.0008106556074694434,
      "loss": 0.4116,
      "num_input_tokens_seen": 8911792,
      "step": 13660
    },
    {
      "epoch": 7.161949685534591,
      "grad_norm": 0.09233508259057999,
      "learning_rate": 0.0008104763851266216,
      "loss": 0.4895,
      "num_input_tokens_seen": 8915280,
      "step": 13665
    },
    {
      "epoch": 7.164570230607967,
      "grad_norm": 0.14383810758590698,
      "learning_rate": 0.0008102970978356822,
      "loss": 0.3875,
      "num_input_tokens_seen": 8918736,
      "step": 13670
    },
    {
      "epoch": 7.167190775681342,
      "grad_norm": 0.11478235572576523,
      "learning_rate": 0.00081011774563413,
      "loss": 0.517,
      "num_input_tokens_seen": 8921776,
      "step": 13675
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 0.056947242468595505,
      "learning_rate": 0.0008099383285594835,
      "loss": 0.5473,
      "num_input_tokens_seen": 8925264,
      "step": 13680
    },
    {
      "epoch": 7.172431865828092,
      "grad_norm": 0.07533730566501617,
      "learning_rate": 0.0008097588466492746,
      "loss": 0.5019,
      "num_input_tokens_seen": 8928400,
      "step": 13685
    },
    {
      "epoch": 7.1750524109014675,
      "grad_norm": 0.0719602033495903,
      "learning_rate": 0.0008095792999410487,
      "loss": 0.5617,
      "num_input_tokens_seen": 8931792,
      "step": 13690
    },
    {
      "epoch": 7.177672955974843,
      "grad_norm": 0.07614172250032425,
      "learning_rate": 0.0008093996884723653,
      "loss": 0.3799,
      "num_input_tokens_seen": 8934768,
      "step": 13695
    },
    {
      "epoch": 7.180293501048218,
      "grad_norm": 0.0932946503162384,
      "learning_rate": 0.0008092200122807969,
      "loss": 0.2672,
      "num_input_tokens_seen": 8937552,
      "step": 13700
    },
    {
      "epoch": 7.182914046121593,
      "grad_norm": 0.09117448329925537,
      "learning_rate": 0.0008090402714039295,
      "loss": 0.5046,
      "num_input_tokens_seen": 8941520,
      "step": 13705
    },
    {
      "epoch": 7.185534591194968,
      "grad_norm": 0.07370131462812424,
      "learning_rate": 0.0008088604658793632,
      "loss": 0.3223,
      "num_input_tokens_seen": 8945616,
      "step": 13710
    },
    {
      "epoch": 7.188155136268344,
      "grad_norm": 0.08694528043270111,
      "learning_rate": 0.0008086805957447111,
      "loss": 0.5364,
      "num_input_tokens_seen": 8948592,
      "step": 13715
    },
    {
      "epoch": 7.190775681341719,
      "grad_norm": 0.08188440650701523,
      "learning_rate": 0.0008085006610376,
      "loss": 0.4536,
      "num_input_tokens_seen": 8951344,
      "step": 13720
    },
    {
      "epoch": 7.193396226415095,
      "grad_norm": 0.12491223961114883,
      "learning_rate": 0.0008083206617956702,
      "loss": 0.4619,
      "num_input_tokens_seen": 8955056,
      "step": 13725
    },
    {
      "epoch": 7.19601677148847,
      "grad_norm": 0.07670099288225174,
      "learning_rate": 0.0008081405980565755,
      "loss": 0.3745,
      "num_input_tokens_seen": 8958128,
      "step": 13730
    },
    {
      "epoch": 7.198637316561845,
      "grad_norm": 0.07292961329221725,
      "learning_rate": 0.0008079604698579829,
      "loss": 0.4136,
      "num_input_tokens_seen": 8961840,
      "step": 13735
    },
    {
      "epoch": 7.20125786163522,
      "grad_norm": 0.08874966949224472,
      "learning_rate": 0.0008077802772375736,
      "loss": 0.5137,
      "num_input_tokens_seen": 8965040,
      "step": 13740
    },
    {
      "epoch": 7.203878406708595,
      "grad_norm": 0.12211589515209198,
      "learning_rate": 0.0008076000202330416,
      "loss": 0.4618,
      "num_input_tokens_seen": 8969200,
      "step": 13745
    },
    {
      "epoch": 7.20649895178197,
      "grad_norm": 0.1284874826669693,
      "learning_rate": 0.0008074196988820945,
      "loss": 0.6347,
      "num_input_tokens_seen": 8972624,
      "step": 13750
    },
    {
      "epoch": 7.209119496855346,
      "grad_norm": 0.09476202726364136,
      "learning_rate": 0.0008072393132224539,
      "loss": 0.4528,
      "num_input_tokens_seen": 8975728,
      "step": 13755
    },
    {
      "epoch": 7.211740041928721,
      "grad_norm": 0.07383695244789124,
      "learning_rate": 0.0008070588632918541,
      "loss": 0.5816,
      "num_input_tokens_seen": 8979312,
      "step": 13760
    },
    {
      "epoch": 7.214360587002097,
      "grad_norm": 0.0725034847855568,
      "learning_rate": 0.000806878349128043,
      "loss": 0.4455,
      "num_input_tokens_seen": 8982032,
      "step": 13765
    },
    {
      "epoch": 7.216981132075472,
      "grad_norm": 0.2296224683523178,
      "learning_rate": 0.0008066977707687826,
      "loss": 0.4512,
      "num_input_tokens_seen": 8984432,
      "step": 13770
    },
    {
      "epoch": 7.219601677148847,
      "grad_norm": 0.11624252796173096,
      "learning_rate": 0.0008065171282518473,
      "loss": 0.5145,
      "num_input_tokens_seen": 8988624,
      "step": 13775
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 0.10649978369474411,
      "learning_rate": 0.0008063364216150257,
      "loss": 0.6847,
      "num_input_tokens_seen": 8991664,
      "step": 13780
    },
    {
      "epoch": 7.2248427672955975,
      "grad_norm": 0.12526057660579681,
      "learning_rate": 0.0008061556508961199,
      "loss": 0.4462,
      "num_input_tokens_seen": 8994928,
      "step": 13785
    },
    {
      "epoch": 7.227463312368973,
      "grad_norm": 0.14201270043849945,
      "learning_rate": 0.0008059748161329443,
      "loss": 0.613,
      "num_input_tokens_seen": 8997872,
      "step": 13790
    },
    {
      "epoch": 7.230083857442348,
      "grad_norm": 0.0864005759358406,
      "learning_rate": 0.0008057939173633282,
      "loss": 0.4907,
      "num_input_tokens_seen": 9002384,
      "step": 13795
    },
    {
      "epoch": 7.232704402515723,
      "grad_norm": 0.173051655292511,
      "learning_rate": 0.0008056129546251132,
      "loss": 0.3982,
      "num_input_tokens_seen": 9005296,
      "step": 13800
    },
    {
      "epoch": 7.235324947589098,
      "grad_norm": 0.07444019615650177,
      "learning_rate": 0.0008054319279561546,
      "loss": 0.515,
      "num_input_tokens_seen": 9009968,
      "step": 13805
    },
    {
      "epoch": 7.237945492662474,
      "grad_norm": 0.17528393864631653,
      "learning_rate": 0.0008052508373943214,
      "loss": 0.5005,
      "num_input_tokens_seen": 9013776,
      "step": 13810
    },
    {
      "epoch": 7.240566037735849,
      "grad_norm": 0.10106755048036575,
      "learning_rate": 0.0008050696829774954,
      "loss": 0.3898,
      "num_input_tokens_seen": 9017072,
      "step": 13815
    },
    {
      "epoch": 7.243186582809225,
      "grad_norm": 0.11056758463382721,
      "learning_rate": 0.0008048884647435721,
      "loss": 0.7093,
      "num_input_tokens_seen": 9020272,
      "step": 13820
    },
    {
      "epoch": 7.2458071278826,
      "grad_norm": 0.08520101755857468,
      "learning_rate": 0.0008047071827304604,
      "loss": 0.4195,
      "num_input_tokens_seen": 9024528,
      "step": 13825
    },
    {
      "epoch": 7.248427672955975,
      "grad_norm": 0.08782754838466644,
      "learning_rate": 0.0008045258369760824,
      "loss": 0.3952,
      "num_input_tokens_seen": 9027376,
      "step": 13830
    },
    {
      "epoch": 7.25104821802935,
      "grad_norm": 0.12672726809978485,
      "learning_rate": 0.0008043444275183735,
      "loss": 0.5103,
      "num_input_tokens_seen": 9030832,
      "step": 13835
    },
    {
      "epoch": 7.253668763102725,
      "grad_norm": 0.08633465319871902,
      "learning_rate": 0.0008041629543952824,
      "loss": 0.5963,
      "num_input_tokens_seen": 9034064,
      "step": 13840
    },
    {
      "epoch": 7.2562893081761,
      "grad_norm": 0.11094009131193161,
      "learning_rate": 0.0008039814176447714,
      "loss": 0.4463,
      "num_input_tokens_seen": 9037072,
      "step": 13845
    },
    {
      "epoch": 7.258909853249476,
      "grad_norm": 0.046487241983413696,
      "learning_rate": 0.0008037998173048157,
      "loss": 0.3723,
      "num_input_tokens_seen": 9040624,
      "step": 13850
    },
    {
      "epoch": 7.261530398322851,
      "grad_norm": 0.1270788460969925,
      "learning_rate": 0.0008036181534134044,
      "loss": 0.4349,
      "num_input_tokens_seen": 9043792,
      "step": 13855
    },
    {
      "epoch": 7.264150943396227,
      "grad_norm": 0.10873308777809143,
      "learning_rate": 0.0008034364260085391,
      "loss": 0.4474,
      "num_input_tokens_seen": 9046640,
      "step": 13860
    },
    {
      "epoch": 7.266771488469602,
      "grad_norm": 0.12715595960617065,
      "learning_rate": 0.0008032546351282353,
      "loss": 0.4617,
      "num_input_tokens_seen": 9049328,
      "step": 13865
    },
    {
      "epoch": 7.269392033542977,
      "grad_norm": 0.07339238375425339,
      "learning_rate": 0.0008030727808105215,
      "loss": 0.4862,
      "num_input_tokens_seen": 9052944,
      "step": 13870
    },
    {
      "epoch": 7.272012578616352,
      "grad_norm": 0.12177667766809464,
      "learning_rate": 0.0008028908630934397,
      "loss": 0.3041,
      "num_input_tokens_seen": 9055344,
      "step": 13875
    },
    {
      "epoch": 7.2746331236897275,
      "grad_norm": 0.06668896973133087,
      "learning_rate": 0.0008027088820150447,
      "loss": 0.4101,
      "num_input_tokens_seen": 9058928,
      "step": 13880
    },
    {
      "epoch": 7.277253668763103,
      "grad_norm": 0.13631297647953033,
      "learning_rate": 0.0008025268376134054,
      "loss": 0.558,
      "num_input_tokens_seen": 9062384,
      "step": 13885
    },
    {
      "epoch": 7.279874213836478,
      "grad_norm": 0.09314925223588943,
      "learning_rate": 0.0008023447299266027,
      "loss": 0.4016,
      "num_input_tokens_seen": 9066288,
      "step": 13890
    },
    {
      "epoch": 7.282494758909853,
      "grad_norm": 0.10401374101638794,
      "learning_rate": 0.0008021625589927321,
      "loss": 0.483,
      "num_input_tokens_seen": 9070000,
      "step": 13895
    },
    {
      "epoch": 7.285115303983228,
      "grad_norm": 0.1041695699095726,
      "learning_rate": 0.0008019803248499013,
      "loss": 0.4258,
      "num_input_tokens_seen": 9072784,
      "step": 13900
    },
    {
      "epoch": 7.287735849056604,
      "grad_norm": 0.06184196099638939,
      "learning_rate": 0.0008017980275362318,
      "loss": 0.3632,
      "num_input_tokens_seen": 9075632,
      "step": 13905
    },
    {
      "epoch": 7.290356394129979,
      "grad_norm": 0.12158573418855667,
      "learning_rate": 0.000801615667089858,
      "loss": 0.5909,
      "num_input_tokens_seen": 9078448,
      "step": 13910
    },
    {
      "epoch": 7.2929769392033545,
      "grad_norm": 0.12112490087747574,
      "learning_rate": 0.0008014332435489275,
      "loss": 0.4789,
      "num_input_tokens_seen": 9082352,
      "step": 13915
    },
    {
      "epoch": 7.29559748427673,
      "grad_norm": 0.13665036857128143,
      "learning_rate": 0.0008012507569516016,
      "loss": 0.5357,
      "num_input_tokens_seen": 9085104,
      "step": 13920
    },
    {
      "epoch": 7.298218029350105,
      "grad_norm": 0.12568402290344238,
      "learning_rate": 0.0008010682073360541,
      "loss": 0.4528,
      "num_input_tokens_seen": 9087856,
      "step": 13925
    },
    {
      "epoch": 7.30083857442348,
      "grad_norm": 0.0651504248380661,
      "learning_rate": 0.0008008855947404724,
      "loss": 0.3554,
      "num_input_tokens_seen": 9091216,
      "step": 13930
    },
    {
      "epoch": 7.303459119496855,
      "grad_norm": 0.12088403105735779,
      "learning_rate": 0.000800702919203057,
      "loss": 0.4548,
      "num_input_tokens_seen": 9094256,
      "step": 13935
    },
    {
      "epoch": 7.30607966457023,
      "grad_norm": 0.04646396264433861,
      "learning_rate": 0.0008005201807620214,
      "loss": 0.3648,
      "num_input_tokens_seen": 9098192,
      "step": 13940
    },
    {
      "epoch": 7.308700209643606,
      "grad_norm": 0.10603929311037064,
      "learning_rate": 0.0008003373794555926,
      "loss": 0.4895,
      "num_input_tokens_seen": 9100432,
      "step": 13945
    },
    {
      "epoch": 7.311320754716981,
      "grad_norm": 0.06151574105024338,
      "learning_rate": 0.0008001545153220104,
      "loss": 0.4409,
      "num_input_tokens_seen": 9104016,
      "step": 13950
    },
    {
      "epoch": 7.313941299790357,
      "grad_norm": 0.11166474968194962,
      "learning_rate": 0.000799971588399528,
      "loss": 0.8389,
      "num_input_tokens_seen": 9107088,
      "step": 13955
    },
    {
      "epoch": 7.316561844863732,
      "grad_norm": 0.10623479634523392,
      "learning_rate": 0.0007997885987264115,
      "loss": 0.55,
      "num_input_tokens_seen": 9110064,
      "step": 13960
    },
    {
      "epoch": 7.319182389937107,
      "grad_norm": 0.0839163064956665,
      "learning_rate": 0.0007996055463409403,
      "loss": 0.5204,
      "num_input_tokens_seen": 9113040,
      "step": 13965
    },
    {
      "epoch": 7.321802935010482,
      "grad_norm": 0.06949745118618011,
      "learning_rate": 0.000799422431281407,
      "loss": 0.4462,
      "num_input_tokens_seen": 9116208,
      "step": 13970
    },
    {
      "epoch": 7.3244234800838575,
      "grad_norm": 0.09717851877212524,
      "learning_rate": 0.000799239253586117,
      "loss": 0.4334,
      "num_input_tokens_seen": 9119280,
      "step": 13975
    },
    {
      "epoch": 7.327044025157233,
      "grad_norm": 0.07560908794403076,
      "learning_rate": 0.0007990560132933891,
      "loss": 0.4759,
      "num_input_tokens_seen": 9124048,
      "step": 13980
    },
    {
      "epoch": 7.329664570230608,
      "grad_norm": 0.07116920500993729,
      "learning_rate": 0.0007988727104415549,
      "loss": 0.4615,
      "num_input_tokens_seen": 9128080,
      "step": 13985
    },
    {
      "epoch": 7.332285115303983,
      "grad_norm": 0.08072322607040405,
      "learning_rate": 0.0007986893450689594,
      "loss": 0.4807,
      "num_input_tokens_seen": 9131120,
      "step": 13990
    },
    {
      "epoch": 7.334905660377358,
      "grad_norm": 0.07028216123580933,
      "learning_rate": 0.0007985059172139606,
      "loss": 0.5207,
      "num_input_tokens_seen": 9133776,
      "step": 13995
    },
    {
      "epoch": 7.337526205450734,
      "grad_norm": 0.08431050926446915,
      "learning_rate": 0.0007983224269149296,
      "loss": 0.5255,
      "num_input_tokens_seen": 9137680,
      "step": 14000
    },
    {
      "epoch": 7.340146750524109,
      "grad_norm": 0.0637979656457901,
      "learning_rate": 0.00079813887421025,
      "loss": 0.4007,
      "num_input_tokens_seen": 9140912,
      "step": 14005
    },
    {
      "epoch": 7.3427672955974845,
      "grad_norm": 0.1691654920578003,
      "learning_rate": 0.0007979552591383195,
      "loss": 0.432,
      "num_input_tokens_seen": 9144432,
      "step": 14010
    },
    {
      "epoch": 7.34538784067086,
      "grad_norm": 0.15405796468257904,
      "learning_rate": 0.0007977715817375481,
      "loss": 0.5337,
      "num_input_tokens_seen": 9147504,
      "step": 14015
    },
    {
      "epoch": 7.348008385744235,
      "grad_norm": 0.07054365426301956,
      "learning_rate": 0.0007975878420463588,
      "loss": 0.4339,
      "num_input_tokens_seen": 9150256,
      "step": 14020
    },
    {
      "epoch": 7.35062893081761,
      "grad_norm": 0.07585826516151428,
      "learning_rate": 0.0007974040401031882,
      "loss": 0.4559,
      "num_input_tokens_seen": 9153552,
      "step": 14025
    },
    {
      "epoch": 7.353249475890985,
      "grad_norm": 0.109430693089962,
      "learning_rate": 0.0007972201759464851,
      "loss": 0.3517,
      "num_input_tokens_seen": 9156624,
      "step": 14030
    },
    {
      "epoch": 7.35587002096436,
      "grad_norm": 0.058893248438835144,
      "learning_rate": 0.0007970362496147121,
      "loss": 0.4204,
      "num_input_tokens_seen": 9159792,
      "step": 14035
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 0.08252420276403427,
      "learning_rate": 0.0007968522611463447,
      "loss": 0.4115,
      "num_input_tokens_seen": 9162960,
      "step": 14040
    },
    {
      "epoch": 7.361111111111111,
      "grad_norm": 0.08850686252117157,
      "learning_rate": 0.0007966682105798708,
      "loss": 0.4692,
      "num_input_tokens_seen": 9166416,
      "step": 14045
    },
    {
      "epoch": 7.363731656184487,
      "grad_norm": 0.1527671217918396,
      "learning_rate": 0.0007964840979537918,
      "loss": 0.5502,
      "num_input_tokens_seen": 9169712,
      "step": 14050
    },
    {
      "epoch": 7.366352201257862,
      "grad_norm": 0.09188425540924072,
      "learning_rate": 0.0007962999233066219,
      "loss": 0.4286,
      "num_input_tokens_seen": 9172208,
      "step": 14055
    },
    {
      "epoch": 7.368972746331237,
      "grad_norm": 0.09999595582485199,
      "learning_rate": 0.0007961156866768885,
      "loss": 0.4556,
      "num_input_tokens_seen": 9174640,
      "step": 14060
    },
    {
      "epoch": 7.371593291404612,
      "grad_norm": 0.08481183648109436,
      "learning_rate": 0.0007959313881031317,
      "loss": 0.4601,
      "num_input_tokens_seen": 9178416,
      "step": 14065
    },
    {
      "epoch": 7.3742138364779874,
      "grad_norm": 0.10623706877231598,
      "learning_rate": 0.0007957470276239048,
      "loss": 0.4582,
      "num_input_tokens_seen": 9181040,
      "step": 14070
    },
    {
      "epoch": 7.376834381551363,
      "grad_norm": 0.08490146696567535,
      "learning_rate": 0.0007955626052777735,
      "loss": 0.4726,
      "num_input_tokens_seen": 9184048,
      "step": 14075
    },
    {
      "epoch": 7.379454926624738,
      "grad_norm": 0.09312544018030167,
      "learning_rate": 0.0007953781211033173,
      "loss": 0.3861,
      "num_input_tokens_seen": 9187344,
      "step": 14080
    },
    {
      "epoch": 7.382075471698113,
      "grad_norm": 0.06678886711597443,
      "learning_rate": 0.000795193575139128,
      "loss": 0.3606,
      "num_input_tokens_seen": 9190608,
      "step": 14085
    },
    {
      "epoch": 7.384696016771488,
      "grad_norm": 0.12421488761901855,
      "learning_rate": 0.0007950089674238106,
      "loss": 0.414,
      "num_input_tokens_seen": 9194160,
      "step": 14090
    },
    {
      "epoch": 7.387316561844864,
      "grad_norm": 0.11429903656244278,
      "learning_rate": 0.0007948242979959828,
      "loss": 0.4101,
      "num_input_tokens_seen": 9196688,
      "step": 14095
    },
    {
      "epoch": 7.389937106918239,
      "grad_norm": 0.08468032628297806,
      "learning_rate": 0.0007946395668942754,
      "loss": 0.4712,
      "num_input_tokens_seen": 9200880,
      "step": 14100
    },
    {
      "epoch": 7.3925576519916145,
      "grad_norm": 0.09427148848772049,
      "learning_rate": 0.0007944547741573319,
      "loss": 0.4208,
      "num_input_tokens_seen": 9204720,
      "step": 14105
    },
    {
      "epoch": 7.39517819706499,
      "grad_norm": 0.13452686369419098,
      "learning_rate": 0.0007942699198238091,
      "loss": 0.4418,
      "num_input_tokens_seen": 9207184,
      "step": 14110
    },
    {
      "epoch": 7.397798742138365,
      "grad_norm": 0.09841493517160416,
      "learning_rate": 0.0007940850039323763,
      "loss": 0.4567,
      "num_input_tokens_seen": 9210224,
      "step": 14115
    },
    {
      "epoch": 7.40041928721174,
      "grad_norm": 0.13855519890785217,
      "learning_rate": 0.0007939000265217156,
      "loss": 0.3753,
      "num_input_tokens_seen": 9213776,
      "step": 14120
    },
    {
      "epoch": 7.403039832285115,
      "grad_norm": 0.11131665855646133,
      "learning_rate": 0.0007937149876305226,
      "loss": 0.474,
      "num_input_tokens_seen": 9216368,
      "step": 14125
    },
    {
      "epoch": 7.40566037735849,
      "grad_norm": 0.0778750628232956,
      "learning_rate": 0.000793529887297505,
      "loss": 0.69,
      "num_input_tokens_seen": 9220112,
      "step": 14130
    },
    {
      "epoch": 7.4082809224318655,
      "grad_norm": 0.12071818113327026,
      "learning_rate": 0.0007933447255613835,
      "loss": 0.4839,
      "num_input_tokens_seen": 9223120,
      "step": 14135
    },
    {
      "epoch": 7.410901467505241,
      "grad_norm": 0.1281144767999649,
      "learning_rate": 0.0007931595024608924,
      "loss": 0.5661,
      "num_input_tokens_seen": 9225744,
      "step": 14140
    },
    {
      "epoch": 7.413522012578617,
      "grad_norm": 0.07977798581123352,
      "learning_rate": 0.0007929742180347776,
      "loss": 0.4063,
      "num_input_tokens_seen": 9229200,
      "step": 14145
    },
    {
      "epoch": 7.416142557651992,
      "grad_norm": 0.06116437166929245,
      "learning_rate": 0.0007927888723217991,
      "loss": 0.4953,
      "num_input_tokens_seen": 9232752,
      "step": 14150
    },
    {
      "epoch": 7.418763102725367,
      "grad_norm": 0.09712297469377518,
      "learning_rate": 0.0007926034653607288,
      "loss": 0.3778,
      "num_input_tokens_seen": 9238704,
      "step": 14155
    },
    {
      "epoch": 7.421383647798742,
      "grad_norm": 0.08453070372343063,
      "learning_rate": 0.0007924179971903516,
      "loss": 0.6838,
      "num_input_tokens_seen": 9241648,
      "step": 14160
    },
    {
      "epoch": 7.424004192872117,
      "grad_norm": 0.12028225511312485,
      "learning_rate": 0.0007922324678494655,
      "loss": 0.4638,
      "num_input_tokens_seen": 9244944,
      "step": 14165
    },
    {
      "epoch": 7.426624737945493,
      "grad_norm": 0.15345773100852966,
      "learning_rate": 0.0007920468773768811,
      "loss": 0.4263,
      "num_input_tokens_seen": 9248144,
      "step": 14170
    },
    {
      "epoch": 7.429245283018868,
      "grad_norm": 0.06896977126598358,
      "learning_rate": 0.0007918612258114217,
      "loss": 0.5039,
      "num_input_tokens_seen": 9251792,
      "step": 14175
    },
    {
      "epoch": 7.431865828092243,
      "grad_norm": 0.06878423690795898,
      "learning_rate": 0.0007916755131919238,
      "loss": 0.527,
      "num_input_tokens_seen": 9255248,
      "step": 14180
    },
    {
      "epoch": 7.434486373165618,
      "grad_norm": 0.08485176414251328,
      "learning_rate": 0.000791489739557236,
      "loss": 0.4411,
      "num_input_tokens_seen": 9258096,
      "step": 14185
    },
    {
      "epoch": 7.437106918238994,
      "grad_norm": 0.13100385665893555,
      "learning_rate": 0.0007913039049462203,
      "loss": 0.4667,
      "num_input_tokens_seen": 9260816,
      "step": 14190
    },
    {
      "epoch": 7.439727463312369,
      "grad_norm": 0.12793397903442383,
      "learning_rate": 0.0007911180093977511,
      "loss": 0.417,
      "num_input_tokens_seen": 9263472,
      "step": 14195
    },
    {
      "epoch": 7.4423480083857445,
      "grad_norm": 0.05338159576058388,
      "learning_rate": 0.0007909320529507154,
      "loss": 0.4111,
      "num_input_tokens_seen": 9266832,
      "step": 14200
    },
    {
      "epoch": 7.44496855345912,
      "grad_norm": 0.14329054951667786,
      "learning_rate": 0.0007907460356440134,
      "loss": 0.4715,
      "num_input_tokens_seen": 9270512,
      "step": 14205
    },
    {
      "epoch": 7.447589098532495,
      "grad_norm": 0.2553615868091583,
      "learning_rate": 0.0007905599575165577,
      "loss": 0.5031,
      "num_input_tokens_seen": 9274032,
      "step": 14210
    },
    {
      "epoch": 7.45020964360587,
      "grad_norm": 0.11211032420396805,
      "learning_rate": 0.0007903738186072739,
      "loss": 0.4235,
      "num_input_tokens_seen": 9276720,
      "step": 14215
    },
    {
      "epoch": 7.452830188679245,
      "grad_norm": 0.10892169177532196,
      "learning_rate": 0.0007901876189550999,
      "loss": 0.3976,
      "num_input_tokens_seen": 9279504,
      "step": 14220
    },
    {
      "epoch": 7.45545073375262,
      "grad_norm": 0.061293572187423706,
      "learning_rate": 0.0007900013585989867,
      "loss": 0.4048,
      "num_input_tokens_seen": 9282864,
      "step": 14225
    },
    {
      "epoch": 7.4580712788259955,
      "grad_norm": 0.06949388980865479,
      "learning_rate": 0.0007898150375778979,
      "loss": 0.5498,
      "num_input_tokens_seen": 9285520,
      "step": 14230
    },
    {
      "epoch": 7.460691823899371,
      "grad_norm": 0.06997789442539215,
      "learning_rate": 0.0007896286559308095,
      "loss": 0.4932,
      "num_input_tokens_seen": 9289840,
      "step": 14235
    },
    {
      "epoch": 7.463312368972747,
      "grad_norm": 0.12180769443511963,
      "learning_rate": 0.0007894422136967105,
      "loss": 0.3892,
      "num_input_tokens_seen": 9292016,
      "step": 14240
    },
    {
      "epoch": 7.465932914046122,
      "grad_norm": 0.07124414294958115,
      "learning_rate": 0.0007892557109146026,
      "loss": 0.4387,
      "num_input_tokens_seen": 9295312,
      "step": 14245
    },
    {
      "epoch": 7.468553459119497,
      "grad_norm": 0.16823655366897583,
      "learning_rate": 0.0007890691476234999,
      "loss": 0.5697,
      "num_input_tokens_seen": 9298032,
      "step": 14250
    },
    {
      "epoch": 7.471174004192872,
      "grad_norm": 0.1057300940155983,
      "learning_rate": 0.0007888825238624294,
      "loss": 0.5247,
      "num_input_tokens_seen": 9301552,
      "step": 14255
    },
    {
      "epoch": 7.473794549266247,
      "grad_norm": 0.17313498258590698,
      "learning_rate": 0.0007886958396704307,
      "loss": 0.3697,
      "num_input_tokens_seen": 9304688,
      "step": 14260
    },
    {
      "epoch": 7.476415094339623,
      "grad_norm": 0.06424584984779358,
      "learning_rate": 0.0007885090950865559,
      "loss": 0.5378,
      "num_input_tokens_seen": 9307664,
      "step": 14265
    },
    {
      "epoch": 7.479035639412998,
      "grad_norm": 0.11591779440641403,
      "learning_rate": 0.0007883222901498701,
      "loss": 0.4533,
      "num_input_tokens_seen": 9310736,
      "step": 14270
    },
    {
      "epoch": 7.481656184486373,
      "grad_norm": 0.11336546391248703,
      "learning_rate": 0.0007881354248994503,
      "loss": 0.4625,
      "num_input_tokens_seen": 9315056,
      "step": 14275
    },
    {
      "epoch": 7.484276729559748,
      "grad_norm": 0.07655467092990875,
      "learning_rate": 0.0007879484993743869,
      "loss": 0.4263,
      "num_input_tokens_seen": 9318672,
      "step": 14280
    },
    {
      "epoch": 7.486897274633124,
      "grad_norm": 0.13827942311763763,
      "learning_rate": 0.0007877615136137827,
      "loss": 0.5667,
      "num_input_tokens_seen": 9325040,
      "step": 14285
    },
    {
      "epoch": 7.489517819706499,
      "grad_norm": 0.06632038205862045,
      "learning_rate": 0.0007875744676567527,
      "loss": 0.3758,
      "num_input_tokens_seen": 9328496,
      "step": 14290
    },
    {
      "epoch": 7.4921383647798745,
      "grad_norm": 0.05683545768260956,
      "learning_rate": 0.0007873873615424248,
      "loss": 0.4862,
      "num_input_tokens_seen": 9331824,
      "step": 14295
    },
    {
      "epoch": 7.49475890985325,
      "grad_norm": 0.07113119214773178,
      "learning_rate": 0.0007872001953099396,
      "loss": 0.3472,
      "num_input_tokens_seen": 9334768,
      "step": 14300
    },
    {
      "epoch": 7.497379454926625,
      "grad_norm": 0.07020923495292664,
      "learning_rate": 0.0007870129689984501,
      "loss": 0.5203,
      "num_input_tokens_seen": 9337616,
      "step": 14305
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.06862638890743256,
      "learning_rate": 0.000786825682647122,
      "loss": 0.523,
      "num_input_tokens_seen": 9340304,
      "step": 14310
    },
    {
      "epoch": 7.502620545073375,
      "grad_norm": 0.06351099908351898,
      "learning_rate": 0.0007866383362951332,
      "loss": 0.5174,
      "num_input_tokens_seen": 9343440,
      "step": 14315
    },
    {
      "epoch": 7.50524109014675,
      "grad_norm": 0.07435489445924759,
      "learning_rate": 0.0007864509299816746,
      "loss": 0.5171,
      "num_input_tokens_seen": 9346864,
      "step": 14320
    },
    {
      "epoch": 7.5078616352201255,
      "grad_norm": 0.11523985117673874,
      "learning_rate": 0.0007862634637459496,
      "loss": 0.5289,
      "num_input_tokens_seen": 9350384,
      "step": 14325
    },
    {
      "epoch": 7.510482180293501,
      "grad_norm": 0.1157648116350174,
      "learning_rate": 0.0007860759376271737,
      "loss": 0.4157,
      "num_input_tokens_seen": 9354256,
      "step": 14330
    },
    {
      "epoch": 7.513102725366876,
      "grad_norm": 0.08590184152126312,
      "learning_rate": 0.0007858883516645755,
      "loss": 0.411,
      "num_input_tokens_seen": 9357456,
      "step": 14335
    },
    {
      "epoch": 7.515723270440252,
      "grad_norm": 0.0653625875711441,
      "learning_rate": 0.0007857007058973957,
      "loss": 0.4099,
      "num_input_tokens_seen": 9360560,
      "step": 14340
    },
    {
      "epoch": 7.518343815513627,
      "grad_norm": 0.22529511153697968,
      "learning_rate": 0.0007855130003648876,
      "loss": 0.4908,
      "num_input_tokens_seen": 9363248,
      "step": 14345
    },
    {
      "epoch": 7.520964360587002,
      "grad_norm": 0.07123390585184097,
      "learning_rate": 0.0007853252351063171,
      "loss": 0.39,
      "num_input_tokens_seen": 9367088,
      "step": 14350
    },
    {
      "epoch": 7.523584905660377,
      "grad_norm": 0.05668541043996811,
      "learning_rate": 0.0007851374101609627,
      "loss": 0.4103,
      "num_input_tokens_seen": 9370960,
      "step": 14355
    },
    {
      "epoch": 7.526205450733753,
      "grad_norm": 0.11099474132061005,
      "learning_rate": 0.000784949525568115,
      "loss": 0.4452,
      "num_input_tokens_seen": 9373936,
      "step": 14360
    },
    {
      "epoch": 7.528825995807128,
      "grad_norm": 0.07499384135007858,
      "learning_rate": 0.0007847615813670776,
      "loss": 0.3592,
      "num_input_tokens_seen": 9377264,
      "step": 14365
    },
    {
      "epoch": 7.531446540880503,
      "grad_norm": 0.09777656197547913,
      "learning_rate": 0.000784573577597166,
      "loss": 0.5298,
      "num_input_tokens_seen": 9380240,
      "step": 14370
    },
    {
      "epoch": 7.534067085953878,
      "grad_norm": 0.062490466982126236,
      "learning_rate": 0.0007843855142977086,
      "loss": 0.3196,
      "num_input_tokens_seen": 9385136,
      "step": 14375
    },
    {
      "epoch": 7.536687631027254,
      "grad_norm": 0.0931195616722107,
      "learning_rate": 0.000784197391508046,
      "loss": 0.5362,
      "num_input_tokens_seen": 9387760,
      "step": 14380
    },
    {
      "epoch": 7.539308176100629,
      "grad_norm": 0.09149061888456345,
      "learning_rate": 0.0007840092092675313,
      "loss": 0.5478,
      "num_input_tokens_seen": 9391184,
      "step": 14385
    },
    {
      "epoch": 7.5419287211740045,
      "grad_norm": 0.05110584944486618,
      "learning_rate": 0.0007838209676155302,
      "loss": 0.4375,
      "num_input_tokens_seen": 9393776,
      "step": 14390
    },
    {
      "epoch": 7.54454926624738,
      "grad_norm": 0.07399223744869232,
      "learning_rate": 0.0007836326665914209,
      "loss": 0.5174,
      "num_input_tokens_seen": 9397008,
      "step": 14395
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 0.11446677893400192,
      "learning_rate": 0.0007834443062345932,
      "loss": 0.5739,
      "num_input_tokens_seen": 9400144,
      "step": 14400
    },
    {
      "epoch": 7.54979035639413,
      "grad_norm": 0.10857755690813065,
      "learning_rate": 0.0007832558865844507,
      "loss": 0.3379,
      "num_input_tokens_seen": 9403728,
      "step": 14405
    },
    {
      "epoch": 7.552410901467505,
      "grad_norm": 0.13685187697410583,
      "learning_rate": 0.0007830674076804083,
      "loss": 0.5834,
      "num_input_tokens_seen": 9408112,
      "step": 14410
    },
    {
      "epoch": 7.55503144654088,
      "grad_norm": 0.11073068529367447,
      "learning_rate": 0.0007828788695618934,
      "loss": 0.5586,
      "num_input_tokens_seen": 9411152,
      "step": 14415
    },
    {
      "epoch": 7.5576519916142555,
      "grad_norm": 0.07652726769447327,
      "learning_rate": 0.0007826902722683462,
      "loss": 0.3373,
      "num_input_tokens_seen": 9414416,
      "step": 14420
    },
    {
      "epoch": 7.560272536687631,
      "grad_norm": 0.10998315364122391,
      "learning_rate": 0.0007825016158392193,
      "loss": 0.4263,
      "num_input_tokens_seen": 9417168,
      "step": 14425
    },
    {
      "epoch": 7.562893081761006,
      "grad_norm": 0.06282546371221542,
      "learning_rate": 0.0007823129003139773,
      "loss": 0.3947,
      "num_input_tokens_seen": 9420880,
      "step": 14430
    },
    {
      "epoch": 7.565513626834382,
      "grad_norm": 0.06057996302843094,
      "learning_rate": 0.0007821241257320972,
      "loss": 0.4368,
      "num_input_tokens_seen": 9424144,
      "step": 14435
    },
    {
      "epoch": 7.568134171907757,
      "grad_norm": 0.0990133062005043,
      "learning_rate": 0.0007819352921330689,
      "loss": 0.4703,
      "num_input_tokens_seen": 9426448,
      "step": 14440
    },
    {
      "epoch": 7.570754716981132,
      "grad_norm": 0.10600601881742477,
      "learning_rate": 0.0007817463995563938,
      "loss": 0.5075,
      "num_input_tokens_seen": 9429456,
      "step": 14445
    },
    {
      "epoch": 7.573375262054507,
      "grad_norm": 0.09093093127012253,
      "learning_rate": 0.0007815574480415864,
      "loss": 0.5214,
      "num_input_tokens_seen": 9432720,
      "step": 14450
    },
    {
      "epoch": 7.575995807127883,
      "grad_norm": 0.09347035735845566,
      "learning_rate": 0.0007813684376281729,
      "loss": 0.5838,
      "num_input_tokens_seen": 9435344,
      "step": 14455
    },
    {
      "epoch": 7.578616352201258,
      "grad_norm": 0.07920762896537781,
      "learning_rate": 0.0007811793683556922,
      "loss": 0.5739,
      "num_input_tokens_seen": 9438416,
      "step": 14460
    },
    {
      "epoch": 7.581236897274633,
      "grad_norm": 0.09892641007900238,
      "learning_rate": 0.0007809902402636957,
      "loss": 0.4325,
      "num_input_tokens_seen": 9440912,
      "step": 14465
    },
    {
      "epoch": 7.583857442348008,
      "grad_norm": 0.0365714393556118,
      "learning_rate": 0.0007808010533917464,
      "loss": 0.2744,
      "num_input_tokens_seen": 9444272,
      "step": 14470
    },
    {
      "epoch": 7.586477987421384,
      "grad_norm": 0.08550161123275757,
      "learning_rate": 0.0007806118077794205,
      "loss": 0.5842,
      "num_input_tokens_seen": 9446704,
      "step": 14475
    },
    {
      "epoch": 7.589098532494759,
      "grad_norm": 0.17481175065040588,
      "learning_rate": 0.0007804225034663058,
      "loss": 0.4543,
      "num_input_tokens_seen": 9450128,
      "step": 14480
    },
    {
      "epoch": 7.5917190775681345,
      "grad_norm": 0.08878275752067566,
      "learning_rate": 0.0007802331404920024,
      "loss": 0.5578,
      "num_input_tokens_seen": 9453072,
      "step": 14485
    },
    {
      "epoch": 7.59433962264151,
      "grad_norm": 0.09526041895151138,
      "learning_rate": 0.0007800437188961232,
      "loss": 0.3653,
      "num_input_tokens_seen": 9455856,
      "step": 14490
    },
    {
      "epoch": 7.596960167714885,
      "grad_norm": 0.07581160962581635,
      "learning_rate": 0.0007798542387182929,
      "loss": 0.5641,
      "num_input_tokens_seen": 9458736,
      "step": 14495
    },
    {
      "epoch": 7.59958071278826,
      "grad_norm": 0.05957004800438881,
      "learning_rate": 0.0007796646999981488,
      "loss": 0.3246,
      "num_input_tokens_seen": 9461840,
      "step": 14500
    },
    {
      "epoch": 7.602201257861635,
      "grad_norm": 0.07565108686685562,
      "learning_rate": 0.0007794751027753397,
      "loss": 0.4102,
      "num_input_tokens_seen": 9468368,
      "step": 14505
    },
    {
      "epoch": 7.60482180293501,
      "grad_norm": 0.13891497254371643,
      "learning_rate": 0.0007792854470895278,
      "loss": 0.607,
      "num_input_tokens_seen": 9470928,
      "step": 14510
    },
    {
      "epoch": 7.6074423480083855,
      "grad_norm": 0.06474656611680984,
      "learning_rate": 0.0007790957329803865,
      "loss": 0.7042,
      "num_input_tokens_seen": 9475024,
      "step": 14515
    },
    {
      "epoch": 7.610062893081761,
      "grad_norm": 0.05658174306154251,
      "learning_rate": 0.0007789059604876019,
      "loss": 0.522,
      "num_input_tokens_seen": 9479408,
      "step": 14520
    },
    {
      "epoch": 7.612683438155136,
      "grad_norm": 0.07544215023517609,
      "learning_rate": 0.0007787161296508724,
      "loss": 0.3987,
      "num_input_tokens_seen": 9482480,
      "step": 14525
    },
    {
      "epoch": 7.615303983228512,
      "grad_norm": 0.1282845139503479,
      "learning_rate": 0.0007785262405099083,
      "loss": 0.3798,
      "num_input_tokens_seen": 9484816,
      "step": 14530
    },
    {
      "epoch": 7.617924528301887,
      "grad_norm": 0.14474676549434662,
      "learning_rate": 0.0007783362931044322,
      "loss": 0.5205,
      "num_input_tokens_seen": 9487728,
      "step": 14535
    },
    {
      "epoch": 7.620545073375262,
      "grad_norm": 0.12618517875671387,
      "learning_rate": 0.0007781462874741793,
      "loss": 0.6397,
      "num_input_tokens_seen": 9490448,
      "step": 14540
    },
    {
      "epoch": 7.623165618448637,
      "grad_norm": 0.07989520579576492,
      "learning_rate": 0.000777956223658896,
      "loss": 0.497,
      "num_input_tokens_seen": 9493360,
      "step": 14545
    },
    {
      "epoch": 7.6257861635220126,
      "grad_norm": 0.09482744336128235,
      "learning_rate": 0.000777766101698342,
      "loss": 0.3769,
      "num_input_tokens_seen": 9496560,
      "step": 14550
    },
    {
      "epoch": 7.628406708595388,
      "grad_norm": 0.1124315932393074,
      "learning_rate": 0.0007775759216322882,
      "loss": 0.3481,
      "num_input_tokens_seen": 9499632,
      "step": 14555
    },
    {
      "epoch": 7.631027253668763,
      "grad_norm": 0.059575241059064865,
      "learning_rate": 0.0007773856835005187,
      "loss": 0.4815,
      "num_input_tokens_seen": 9503696,
      "step": 14560
    },
    {
      "epoch": 7.633647798742138,
      "grad_norm": 0.07563942670822144,
      "learning_rate": 0.0007771953873428285,
      "loss": 0.547,
      "num_input_tokens_seen": 9506768,
      "step": 14565
    },
    {
      "epoch": 7.636268343815514,
      "grad_norm": 0.08145631104707718,
      "learning_rate": 0.0007770050331990259,
      "loss": 0.4116,
      "num_input_tokens_seen": 9509488,
      "step": 14570
    },
    {
      "epoch": 7.638888888888889,
      "grad_norm": 0.05301470682024956,
      "learning_rate": 0.0007768146211089304,
      "loss": 0.409,
      "num_input_tokens_seen": 9512336,
      "step": 14575
    },
    {
      "epoch": 7.6415094339622645,
      "grad_norm": 0.08656764030456543,
      "learning_rate": 0.0007766241511123744,
      "loss": 0.5688,
      "num_input_tokens_seen": 9515376,
      "step": 14580
    },
    {
      "epoch": 7.64412997903564,
      "grad_norm": 0.10380943864583969,
      "learning_rate": 0.0007764336232492018,
      "loss": 0.3843,
      "num_input_tokens_seen": 9518160,
      "step": 14585
    },
    {
      "epoch": 7.646750524109015,
      "grad_norm": 0.07913577556610107,
      "learning_rate": 0.0007762430375592689,
      "loss": 0.4344,
      "num_input_tokens_seen": 9520912,
      "step": 14590
    },
    {
      "epoch": 7.64937106918239,
      "grad_norm": 0.07004508376121521,
      "learning_rate": 0.0007760523940824441,
      "loss": 0.4269,
      "num_input_tokens_seen": 9523728,
      "step": 14595
    },
    {
      "epoch": 7.651991614255765,
      "grad_norm": 0.12171425670385361,
      "learning_rate": 0.0007758616928586077,
      "loss": 0.6168,
      "num_input_tokens_seen": 9526416,
      "step": 14600
    },
    {
      "epoch": 7.65461215932914,
      "grad_norm": 0.09062576293945312,
      "learning_rate": 0.0007756709339276527,
      "loss": 0.4775,
      "num_input_tokens_seen": 9529936,
      "step": 14605
    },
    {
      "epoch": 7.6572327044025155,
      "grad_norm": 0.06568481028079987,
      "learning_rate": 0.0007754801173294831,
      "loss": 0.4746,
      "num_input_tokens_seen": 9533232,
      "step": 14610
    },
    {
      "epoch": 7.659853249475891,
      "grad_norm": 0.05342532694339752,
      "learning_rate": 0.0007752892431040158,
      "loss": 0.4101,
      "num_input_tokens_seen": 9536656,
      "step": 14615
    },
    {
      "epoch": 7.662473794549266,
      "grad_norm": 0.08480194956064224,
      "learning_rate": 0.0007750983112911796,
      "loss": 0.5077,
      "num_input_tokens_seen": 9539344,
      "step": 14620
    },
    {
      "epoch": 7.665094339622642,
      "grad_norm": 0.06682360172271729,
      "learning_rate": 0.0007749073219309151,
      "loss": 0.5181,
      "num_input_tokens_seen": 9542832,
      "step": 14625
    },
    {
      "epoch": 7.667714884696017,
      "grad_norm": 0.12581293284893036,
      "learning_rate": 0.0007747162750631751,
      "loss": 0.5023,
      "num_input_tokens_seen": 9546640,
      "step": 14630
    },
    {
      "epoch": 7.670335429769392,
      "grad_norm": 0.0757158026099205,
      "learning_rate": 0.0007745251707279246,
      "loss": 0.3644,
      "num_input_tokens_seen": 9550000,
      "step": 14635
    },
    {
      "epoch": 7.672955974842767,
      "grad_norm": 0.07015331089496613,
      "learning_rate": 0.0007743340089651403,
      "loss": 0.4755,
      "num_input_tokens_seen": 9553424,
      "step": 14640
    },
    {
      "epoch": 7.6755765199161425,
      "grad_norm": 0.06345022469758987,
      "learning_rate": 0.0007741427898148111,
      "loss": 0.4975,
      "num_input_tokens_seen": 9556144,
      "step": 14645
    },
    {
      "epoch": 7.678197064989518,
      "grad_norm": 0.10976240038871765,
      "learning_rate": 0.0007739515133169379,
      "loss": 0.4264,
      "num_input_tokens_seen": 9559024,
      "step": 14650
    },
    {
      "epoch": 7.680817610062893,
      "grad_norm": 0.06640268862247467,
      "learning_rate": 0.0007737601795115334,
      "loss": 0.5298,
      "num_input_tokens_seen": 9562640,
      "step": 14655
    },
    {
      "epoch": 7.683438155136268,
      "grad_norm": 0.09421490877866745,
      "learning_rate": 0.0007735687884386226,
      "loss": 0.4243,
      "num_input_tokens_seen": 9565360,
      "step": 14660
    },
    {
      "epoch": 7.686058700209644,
      "grad_norm": 0.07054276019334793,
      "learning_rate": 0.0007733773401382424,
      "loss": 0.4878,
      "num_input_tokens_seen": 9568688,
      "step": 14665
    },
    {
      "epoch": 7.688679245283019,
      "grad_norm": 0.11577513813972473,
      "learning_rate": 0.0007731858346504414,
      "loss": 0.3909,
      "num_input_tokens_seen": 9571568,
      "step": 14670
    },
    {
      "epoch": 7.691299790356394,
      "grad_norm": 0.11562372744083405,
      "learning_rate": 0.0007729942720152805,
      "loss": 0.4515,
      "num_input_tokens_seen": 9574128,
      "step": 14675
    },
    {
      "epoch": 7.69392033542977,
      "grad_norm": 0.06898491084575653,
      "learning_rate": 0.0007728026522728324,
      "loss": 0.4586,
      "num_input_tokens_seen": 9577680,
      "step": 14680
    },
    {
      "epoch": 7.696540880503145,
      "grad_norm": 0.10019905120134354,
      "learning_rate": 0.0007726109754631817,
      "loss": 0.4292,
      "num_input_tokens_seen": 9580656,
      "step": 14685
    },
    {
      "epoch": 7.69916142557652,
      "grad_norm": 0.07088644802570343,
      "learning_rate": 0.000772419241626425,
      "loss": 0.481,
      "num_input_tokens_seen": 9584496,
      "step": 14690
    },
    {
      "epoch": 7.701781970649895,
      "grad_norm": 0.05916643887758255,
      "learning_rate": 0.000772227450802671,
      "loss": 0.5081,
      "num_input_tokens_seen": 9591056,
      "step": 14695
    },
    {
      "epoch": 7.70440251572327,
      "grad_norm": 0.10365781933069229,
      "learning_rate": 0.0007720356030320399,
      "loss": 0.4314,
      "num_input_tokens_seen": 9594320,
      "step": 14700
    },
    {
      "epoch": 7.7070230607966455,
      "grad_norm": 0.06564797461032867,
      "learning_rate": 0.0007718436983546642,
      "loss": 0.4557,
      "num_input_tokens_seen": 9597616,
      "step": 14705
    },
    {
      "epoch": 7.709643605870021,
      "grad_norm": 0.10222721099853516,
      "learning_rate": 0.0007716517368106882,
      "loss": 0.5053,
      "num_input_tokens_seen": 9601072,
      "step": 14710
    },
    {
      "epoch": 7.712264150943396,
      "grad_norm": 0.08421074599027634,
      "learning_rate": 0.000771459718440268,
      "loss": 0.4347,
      "num_input_tokens_seen": 9604752,
      "step": 14715
    },
    {
      "epoch": 7.714884696016772,
      "grad_norm": 0.09488148987293243,
      "learning_rate": 0.0007712676432835717,
      "loss": 0.4397,
      "num_input_tokens_seen": 9607600,
      "step": 14720
    },
    {
      "epoch": 7.717505241090147,
      "grad_norm": 0.0827852413058281,
      "learning_rate": 0.0007710755113807794,
      "loss": 0.4708,
      "num_input_tokens_seen": 9611184,
      "step": 14725
    },
    {
      "epoch": 7.720125786163522,
      "grad_norm": 0.09358076006174088,
      "learning_rate": 0.0007708833227720824,
      "loss": 0.4491,
      "num_input_tokens_seen": 9614288,
      "step": 14730
    },
    {
      "epoch": 7.722746331236897,
      "grad_norm": 0.12240190804004669,
      "learning_rate": 0.0007706910774976848,
      "loss": 0.4689,
      "num_input_tokens_seen": 9617680,
      "step": 14735
    },
    {
      "epoch": 7.7253668763102725,
      "grad_norm": 0.11920224130153656,
      "learning_rate": 0.0007704987755978021,
      "loss": 0.3888,
      "num_input_tokens_seen": 9620944,
      "step": 14740
    },
    {
      "epoch": 7.727987421383648,
      "grad_norm": 0.1111232191324234,
      "learning_rate": 0.0007703064171126615,
      "loss": 0.4512,
      "num_input_tokens_seen": 9627536,
      "step": 14745
    },
    {
      "epoch": 7.730607966457023,
      "grad_norm": 0.08462051302194595,
      "learning_rate": 0.0007701140020825022,
      "loss": 0.5572,
      "num_input_tokens_seen": 9631216,
      "step": 14750
    },
    {
      "epoch": 7.733228511530398,
      "grad_norm": 0.08340395241975784,
      "learning_rate": 0.0007699215305475753,
      "loss": 0.3761,
      "num_input_tokens_seen": 9634160,
      "step": 14755
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 0.07693679630756378,
      "learning_rate": 0.0007697290025481436,
      "loss": 0.4039,
      "num_input_tokens_seen": 9637456,
      "step": 14760
    },
    {
      "epoch": 7.738469601677149,
      "grad_norm": 0.06540918350219727,
      "learning_rate": 0.0007695364181244819,
      "loss": 0.4641,
      "num_input_tokens_seen": 9641520,
      "step": 14765
    },
    {
      "epoch": 7.741090146750524,
      "grad_norm": 0.06319954991340637,
      "learning_rate": 0.0007693437773168764,
      "loss": 0.4498,
      "num_input_tokens_seen": 9644912,
      "step": 14770
    },
    {
      "epoch": 7.7437106918239,
      "grad_norm": 0.06738138198852539,
      "learning_rate": 0.0007691510801656256,
      "loss": 0.403,
      "num_input_tokens_seen": 9647568,
      "step": 14775
    },
    {
      "epoch": 7.746331236897275,
      "grad_norm": 0.051479026675224304,
      "learning_rate": 0.0007689583267110395,
      "loss": 0.4109,
      "num_input_tokens_seen": 9651600,
      "step": 14780
    },
    {
      "epoch": 7.74895178197065,
      "grad_norm": 0.11594469845294952,
      "learning_rate": 0.0007687655169934398,
      "loss": 0.3317,
      "num_input_tokens_seen": 9654576,
      "step": 14785
    },
    {
      "epoch": 7.751572327044025,
      "grad_norm": 0.2002176195383072,
      "learning_rate": 0.0007685726510531603,
      "loss": 0.45,
      "num_input_tokens_seen": 9658000,
      "step": 14790
    },
    {
      "epoch": 7.7541928721174,
      "grad_norm": 0.05504612252116203,
      "learning_rate": 0.0007683797289305463,
      "loss": 0.4817,
      "num_input_tokens_seen": 9660880,
      "step": 14795
    },
    {
      "epoch": 7.756813417190775,
      "grad_norm": 0.09040787070989609,
      "learning_rate": 0.0007681867506659548,
      "loss": 0.3686,
      "num_input_tokens_seen": 9663920,
      "step": 14800
    },
    {
      "epoch": 7.759433962264151,
      "grad_norm": 0.09343988448381424,
      "learning_rate": 0.0007679937162997546,
      "loss": 0.4319,
      "num_input_tokens_seen": 9667792,
      "step": 14805
    },
    {
      "epoch": 7.762054507337526,
      "grad_norm": 0.09206807613372803,
      "learning_rate": 0.0007678006258723264,
      "loss": 0.4156,
      "num_input_tokens_seen": 9671120,
      "step": 14810
    },
    {
      "epoch": 7.764675052410902,
      "grad_norm": 0.04738384857773781,
      "learning_rate": 0.0007676074794240626,
      "loss": 0.383,
      "num_input_tokens_seen": 9674608,
      "step": 14815
    },
    {
      "epoch": 7.767295597484277,
      "grad_norm": 0.10836601257324219,
      "learning_rate": 0.000767414276995367,
      "loss": 0.429,
      "num_input_tokens_seen": 9677904,
      "step": 14820
    },
    {
      "epoch": 7.769916142557652,
      "grad_norm": 0.1091897040605545,
      "learning_rate": 0.0007672210186266555,
      "loss": 0.4427,
      "num_input_tokens_seen": 9681648,
      "step": 14825
    },
    {
      "epoch": 7.772536687631027,
      "grad_norm": 0.09590214490890503,
      "learning_rate": 0.0007670277043583556,
      "loss": 0.5371,
      "num_input_tokens_seen": 9684208,
      "step": 14830
    },
    {
      "epoch": 7.7751572327044025,
      "grad_norm": 0.06397537887096405,
      "learning_rate": 0.0007668343342309063,
      "loss": 0.5189,
      "num_input_tokens_seen": 9686800,
      "step": 14835
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 0.16552196443080902,
      "learning_rate": 0.0007666409082847586,
      "loss": 0.4272,
      "num_input_tokens_seen": 9689424,
      "step": 14840
    },
    {
      "epoch": 7.780398322851153,
      "grad_norm": 0.24022698402404785,
      "learning_rate": 0.0007664474265603747,
      "loss": 0.4142,
      "num_input_tokens_seen": 9692368,
      "step": 14845
    },
    {
      "epoch": 7.783018867924528,
      "grad_norm": 0.07132399082183838,
      "learning_rate": 0.0007662538890982291,
      "loss": 0.3748,
      "num_input_tokens_seen": 9695856,
      "step": 14850
    },
    {
      "epoch": 7.785639412997903,
      "grad_norm": 0.0740695521235466,
      "learning_rate": 0.0007660602959388075,
      "loss": 0.3737,
      "num_input_tokens_seen": 9699280,
      "step": 14855
    },
    {
      "epoch": 7.788259958071279,
      "grad_norm": 0.06610274314880371,
      "learning_rate": 0.0007658666471226073,
      "loss": 0.5622,
      "num_input_tokens_seen": 9701840,
      "step": 14860
    },
    {
      "epoch": 7.790880503144654,
      "grad_norm": 0.09071546792984009,
      "learning_rate": 0.0007656729426901377,
      "loss": 0.4659,
      "num_input_tokens_seen": 9704400,
      "step": 14865
    },
    {
      "epoch": 7.79350104821803,
      "grad_norm": 0.07158481329679489,
      "learning_rate": 0.0007654791826819194,
      "loss": 0.4112,
      "num_input_tokens_seen": 9707664,
      "step": 14870
    },
    {
      "epoch": 7.796121593291405,
      "grad_norm": 0.08489050716161728,
      "learning_rate": 0.0007652853671384847,
      "loss": 0.5004,
      "num_input_tokens_seen": 9710672,
      "step": 14875
    },
    {
      "epoch": 7.79874213836478,
      "grad_norm": 0.09292419254779816,
      "learning_rate": 0.0007650914961003781,
      "loss": 0.5313,
      "num_input_tokens_seen": 9713872,
      "step": 14880
    },
    {
      "epoch": 7.801362683438155,
      "grad_norm": 0.07540274411439896,
      "learning_rate": 0.0007648975696081546,
      "loss": 0.3952,
      "num_input_tokens_seen": 9717008,
      "step": 14885
    },
    {
      "epoch": 7.80398322851153,
      "grad_norm": 0.11906995624303818,
      "learning_rate": 0.0007647035877023816,
      "loss": 0.5802,
      "num_input_tokens_seen": 9720560,
      "step": 14890
    },
    {
      "epoch": 7.806603773584905,
      "grad_norm": 0.05956685543060303,
      "learning_rate": 0.0007645095504236381,
      "loss": 0.3912,
      "num_input_tokens_seen": 9723792,
      "step": 14895
    },
    {
      "epoch": 7.809224318658281,
      "grad_norm": 0.09541362524032593,
      "learning_rate": 0.0007643154578125142,
      "loss": 0.4077,
      "num_input_tokens_seen": 9727184,
      "step": 14900
    },
    {
      "epoch": 7.811844863731656,
      "grad_norm": 0.11619185656309128,
      "learning_rate": 0.0007641213099096121,
      "loss": 0.4479,
      "num_input_tokens_seen": 9730832,
      "step": 14905
    },
    {
      "epoch": 7.814465408805032,
      "grad_norm": 0.12898939847946167,
      "learning_rate": 0.0007639271067555452,
      "loss": 0.459,
      "num_input_tokens_seen": 9734640,
      "step": 14910
    },
    {
      "epoch": 7.817085953878407,
      "grad_norm": 0.10861827433109283,
      "learning_rate": 0.0007637328483909385,
      "loss": 0.3176,
      "num_input_tokens_seen": 9737520,
      "step": 14915
    },
    {
      "epoch": 7.819706498951782,
      "grad_norm": 0.13280142843723297,
      "learning_rate": 0.000763538534856429,
      "loss": 0.4551,
      "num_input_tokens_seen": 9740880,
      "step": 14920
    },
    {
      "epoch": 7.822327044025157,
      "grad_norm": 0.10186923295259476,
      "learning_rate": 0.0007633441661926643,
      "loss": 0.3479,
      "num_input_tokens_seen": 9744080,
      "step": 14925
    },
    {
      "epoch": 7.8249475890985325,
      "grad_norm": 0.08833721280097961,
      "learning_rate": 0.0007631497424403046,
      "loss": 0.4505,
      "num_input_tokens_seen": 9747184,
      "step": 14930
    },
    {
      "epoch": 7.827568134171908,
      "grad_norm": 0.13615299761295319,
      "learning_rate": 0.000762955263640021,
      "loss": 0.562,
      "num_input_tokens_seen": 9749648,
      "step": 14935
    },
    {
      "epoch": 7.830188679245283,
      "grad_norm": 0.1446366310119629,
      "learning_rate": 0.0007627607298324961,
      "loss": 0.5122,
      "num_input_tokens_seen": 9751984,
      "step": 14940
    },
    {
      "epoch": 7.832809224318658,
      "grad_norm": 0.07234140485525131,
      "learning_rate": 0.0007625661410584244,
      "loss": 0.4311,
      "num_input_tokens_seen": 9755152,
      "step": 14945
    },
    {
      "epoch": 7.835429769392033,
      "grad_norm": 0.11266711354255676,
      "learning_rate": 0.0007623714973585113,
      "loss": 0.4978,
      "num_input_tokens_seen": 9758448,
      "step": 14950
    },
    {
      "epoch": 7.838050314465409,
      "grad_norm": 0.07426180690526962,
      "learning_rate": 0.0007621767987734743,
      "loss": 0.443,
      "num_input_tokens_seen": 9761264,
      "step": 14955
    },
    {
      "epoch": 7.840670859538784,
      "grad_norm": 0.1273275762796402,
      "learning_rate": 0.000761982045344042,
      "loss": 0.5748,
      "num_input_tokens_seen": 9764976,
      "step": 14960
    },
    {
      "epoch": 7.84329140461216,
      "grad_norm": 0.08363983780145645,
      "learning_rate": 0.0007617872371109549,
      "loss": 0.5588,
      "num_input_tokens_seen": 9768208,
      "step": 14965
    },
    {
      "epoch": 7.845911949685535,
      "grad_norm": 0.09402448683977127,
      "learning_rate": 0.0007615923741149643,
      "loss": 0.4193,
      "num_input_tokens_seen": 9771248,
      "step": 14970
    },
    {
      "epoch": 7.84853249475891,
      "grad_norm": 0.08872047066688538,
      "learning_rate": 0.0007613974563968333,
      "loss": 0.4563,
      "num_input_tokens_seen": 9773776,
      "step": 14975
    },
    {
      "epoch": 7.851153039832285,
      "grad_norm": 0.09448891878128052,
      "learning_rate": 0.0007612024839973368,
      "loss": 0.5283,
      "num_input_tokens_seen": 9777360,
      "step": 14980
    },
    {
      "epoch": 7.85377358490566,
      "grad_norm": 0.09690208733081818,
      "learning_rate": 0.0007610074569572605,
      "loss": 0.428,
      "num_input_tokens_seen": 9779984,
      "step": 14985
    },
    {
      "epoch": 7.856394129979035,
      "grad_norm": 0.21474701166152954,
      "learning_rate": 0.0007608123753174019,
      "loss": 0.4526,
      "num_input_tokens_seen": 9782480,
      "step": 14990
    },
    {
      "epoch": 7.859014675052411,
      "grad_norm": 0.11495928466320038,
      "learning_rate": 0.00076061723911857,
      "loss": 0.5589,
      "num_input_tokens_seen": 9786608,
      "step": 14995
    },
    {
      "epoch": 7.861635220125786,
      "grad_norm": 0.11467918753623962,
      "learning_rate": 0.0007604220484015849,
      "loss": 0.517,
      "num_input_tokens_seen": 9790096,
      "step": 15000
    },
    {
      "epoch": 7.864255765199162,
      "grad_norm": 0.10869342088699341,
      "learning_rate": 0.0007602268032072784,
      "loss": 0.5425,
      "num_input_tokens_seen": 9793072,
      "step": 15005
    },
    {
      "epoch": 7.866876310272537,
      "grad_norm": 0.19591546058654785,
      "learning_rate": 0.0007600315035764933,
      "loss": 0.4501,
      "num_input_tokens_seen": 9796528,
      "step": 15010
    },
    {
      "epoch": 7.869496855345912,
      "grad_norm": 0.14542044699192047,
      "learning_rate": 0.0007598361495500844,
      "loss": 0.4606,
      "num_input_tokens_seen": 9800048,
      "step": 15015
    },
    {
      "epoch": 7.872117400419287,
      "grad_norm": 0.08077524602413177,
      "learning_rate": 0.0007596407411689173,
      "loss": 0.4187,
      "num_input_tokens_seen": 9802672,
      "step": 15020
    },
    {
      "epoch": 7.8747379454926625,
      "grad_norm": 0.1331639438867569,
      "learning_rate": 0.0007594452784738695,
      "loss": 0.5589,
      "num_input_tokens_seen": 9806352,
      "step": 15025
    },
    {
      "epoch": 7.877358490566038,
      "grad_norm": 0.1441786289215088,
      "learning_rate": 0.000759249761505829,
      "loss": 0.3869,
      "num_input_tokens_seen": 9809200,
      "step": 15030
    },
    {
      "epoch": 7.879979035639413,
      "grad_norm": 0.10799846053123474,
      "learning_rate": 0.0007590541903056963,
      "loss": 0.5659,
      "num_input_tokens_seen": 9811568,
      "step": 15035
    },
    {
      "epoch": 7.882599580712788,
      "grad_norm": 0.06778796017169952,
      "learning_rate": 0.0007588585649143825,
      "loss": 0.5339,
      "num_input_tokens_seen": 9814480,
      "step": 15040
    },
    {
      "epoch": 7.885220125786163,
      "grad_norm": 0.10102369636297226,
      "learning_rate": 0.0007586628853728099,
      "loss": 0.4519,
      "num_input_tokens_seen": 9817200,
      "step": 15045
    },
    {
      "epoch": 7.887840670859539,
      "grad_norm": 0.06974641233682632,
      "learning_rate": 0.0007584671517219128,
      "loss": 0.3437,
      "num_input_tokens_seen": 9820336,
      "step": 15050
    },
    {
      "epoch": 7.890461215932914,
      "grad_norm": 0.07332263141870499,
      "learning_rate": 0.0007582713640026364,
      "loss": 0.5144,
      "num_input_tokens_seen": 9823664,
      "step": 15055
    },
    {
      "epoch": 7.8930817610062896,
      "grad_norm": 0.07978585362434387,
      "learning_rate": 0.000758075522255937,
      "loss": 0.463,
      "num_input_tokens_seen": 9827280,
      "step": 15060
    },
    {
      "epoch": 7.895702306079665,
      "grad_norm": 0.10248321294784546,
      "learning_rate": 0.0007578796265227828,
      "loss": 0.5143,
      "num_input_tokens_seen": 9830192,
      "step": 15065
    },
    {
      "epoch": 7.89832285115304,
      "grad_norm": 0.13651375472545624,
      "learning_rate": 0.0007576836768441525,
      "loss": 0.4751,
      "num_input_tokens_seen": 9832816,
      "step": 15070
    },
    {
      "epoch": 7.900943396226415,
      "grad_norm": 0.08593332022428513,
      "learning_rate": 0.0007574876732610369,
      "loss": 0.3917,
      "num_input_tokens_seen": 9836080,
      "step": 15075
    },
    {
      "epoch": 7.90356394129979,
      "grad_norm": 0.11489742994308472,
      "learning_rate": 0.0007572916158144379,
      "loss": 0.3957,
      "num_input_tokens_seen": 9839152,
      "step": 15080
    },
    {
      "epoch": 7.906184486373165,
      "grad_norm": 0.11066685616970062,
      "learning_rate": 0.0007570955045453679,
      "loss": 0.3615,
      "num_input_tokens_seen": 9843120,
      "step": 15085
    },
    {
      "epoch": 7.908805031446541,
      "grad_norm": 0.11294703185558319,
      "learning_rate": 0.0007568993394948516,
      "loss": 0.4707,
      "num_input_tokens_seen": 9846608,
      "step": 15090
    },
    {
      "epoch": 7.911425576519916,
      "grad_norm": 0.11166775226593018,
      "learning_rate": 0.0007567031207039243,
      "loss": 0.4721,
      "num_input_tokens_seen": 9849552,
      "step": 15095
    },
    {
      "epoch": 7.914046121593291,
      "grad_norm": 0.13701382279396057,
      "learning_rate": 0.0007565068482136328,
      "loss": 0.4673,
      "num_input_tokens_seen": 9852336,
      "step": 15100
    },
    {
      "epoch": 7.916666666666667,
      "grad_norm": 0.08676021546125412,
      "learning_rate": 0.000756310522065035,
      "loss": 0.4546,
      "num_input_tokens_seen": 9855376,
      "step": 15105
    },
    {
      "epoch": 7.919287211740042,
      "grad_norm": 0.06874099373817444,
      "learning_rate": 0.0007561141422992002,
      "loss": 0.5025,
      "num_input_tokens_seen": 9858608,
      "step": 15110
    },
    {
      "epoch": 7.921907756813417,
      "grad_norm": 0.1265283077955246,
      "learning_rate": 0.0007559177089572086,
      "loss": 0.4434,
      "num_input_tokens_seen": 9861776,
      "step": 15115
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 0.1473974883556366,
      "learning_rate": 0.0007557212220801521,
      "loss": 0.4829,
      "num_input_tokens_seen": 9864144,
      "step": 15120
    },
    {
      "epoch": 7.927148846960168,
      "grad_norm": 0.13235920667648315,
      "learning_rate": 0.0007555246817091332,
      "loss": 0.472,
      "num_input_tokens_seen": 9866512,
      "step": 15125
    },
    {
      "epoch": 7.929769392033543,
      "grad_norm": 0.0940297394990921,
      "learning_rate": 0.0007553280878852663,
      "loss": 0.5297,
      "num_input_tokens_seen": 9869936,
      "step": 15130
    },
    {
      "epoch": 7.932389937106918,
      "grad_norm": 0.07940670847892761,
      "learning_rate": 0.0007551314406496762,
      "loss": 0.3653,
      "num_input_tokens_seen": 9873168,
      "step": 15135
    },
    {
      "epoch": 7.935010482180293,
      "grad_norm": 0.11377498507499695,
      "learning_rate": 0.0007549347400434994,
      "loss": 0.5807,
      "num_input_tokens_seen": 9875888,
      "step": 15140
    },
    {
      "epoch": 7.937631027253669,
      "grad_norm": 0.08372557163238525,
      "learning_rate": 0.0007547379861078835,
      "loss": 0.556,
      "num_input_tokens_seen": 9880432,
      "step": 15145
    },
    {
      "epoch": 7.940251572327044,
      "grad_norm": 0.1327015906572342,
      "learning_rate": 0.0007545411788839871,
      "loss": 0.4317,
      "num_input_tokens_seen": 9885136,
      "step": 15150
    },
    {
      "epoch": 7.9428721174004195,
      "grad_norm": 0.11188150942325592,
      "learning_rate": 0.0007543443184129799,
      "loss": 0.5031,
      "num_input_tokens_seen": 9887920,
      "step": 15155
    },
    {
      "epoch": 7.945492662473795,
      "grad_norm": 0.0919051468372345,
      "learning_rate": 0.000754147404736043,
      "loss": 0.4141,
      "num_input_tokens_seen": 9892464,
      "step": 15160
    },
    {
      "epoch": 7.94811320754717,
      "grad_norm": 0.13571061193943024,
      "learning_rate": 0.0007539504378943686,
      "loss": 0.4724,
      "num_input_tokens_seen": 9895344,
      "step": 15165
    },
    {
      "epoch": 7.950733752620545,
      "grad_norm": 0.0657375231385231,
      "learning_rate": 0.0007537534179291599,
      "loss": 0.4777,
      "num_input_tokens_seen": 9898032,
      "step": 15170
    },
    {
      "epoch": 7.95335429769392,
      "grad_norm": 0.1754390299320221,
      "learning_rate": 0.000753556344881631,
      "loss": 0.3327,
      "num_input_tokens_seen": 9902160,
      "step": 15175
    },
    {
      "epoch": 7.955974842767295,
      "grad_norm": 0.08245004713535309,
      "learning_rate": 0.0007533592187930076,
      "loss": 0.4158,
      "num_input_tokens_seen": 9906160,
      "step": 15180
    },
    {
      "epoch": 7.9585953878406706,
      "grad_norm": 0.10143247246742249,
      "learning_rate": 0.000753162039704526,
      "loss": 0.4292,
      "num_input_tokens_seen": 9909296,
      "step": 15185
    },
    {
      "epoch": 7.961215932914046,
      "grad_norm": 0.14034396409988403,
      "learning_rate": 0.000752964807657434,
      "loss": 0.475,
      "num_input_tokens_seen": 9911952,
      "step": 15190
    },
    {
      "epoch": 7.963836477987421,
      "grad_norm": 0.10728856921195984,
      "learning_rate": 0.0007527675226929902,
      "loss": 0.5491,
      "num_input_tokens_seen": 9914992,
      "step": 15195
    },
    {
      "epoch": 7.966457023060797,
      "grad_norm": 0.06810296326875687,
      "learning_rate": 0.0007525701848524643,
      "loss": 0.3313,
      "num_input_tokens_seen": 9919184,
      "step": 15200
    },
    {
      "epoch": 7.969077568134172,
      "grad_norm": 0.1384105384349823,
      "learning_rate": 0.0007523727941771372,
      "loss": 0.5871,
      "num_input_tokens_seen": 9921744,
      "step": 15205
    },
    {
      "epoch": 7.971698113207547,
      "grad_norm": 0.11324794590473175,
      "learning_rate": 0.0007521753507083009,
      "loss": 0.4679,
      "num_input_tokens_seen": 9924912,
      "step": 15210
    },
    {
      "epoch": 7.9743186582809225,
      "grad_norm": 0.08893167972564697,
      "learning_rate": 0.0007519778544872581,
      "loss": 0.3884,
      "num_input_tokens_seen": 9927728,
      "step": 15215
    },
    {
      "epoch": 7.976939203354298,
      "grad_norm": 0.09573788940906525,
      "learning_rate": 0.000751780305555323,
      "loss": 0.4286,
      "num_input_tokens_seen": 9930448,
      "step": 15220
    },
    {
      "epoch": 7.979559748427673,
      "grad_norm": 0.14715656638145447,
      "learning_rate": 0.0007515827039538202,
      "loss": 0.4813,
      "num_input_tokens_seen": 9934608,
      "step": 15225
    },
    {
      "epoch": 7.982180293501048,
      "grad_norm": 0.10647134482860565,
      "learning_rate": 0.0007513850497240861,
      "loss": 0.4153,
      "num_input_tokens_seen": 9937456,
      "step": 15230
    },
    {
      "epoch": 7.984800838574423,
      "grad_norm": 0.12252847105264664,
      "learning_rate": 0.0007511873429074676,
      "loss": 0.5064,
      "num_input_tokens_seen": 9940816,
      "step": 15235
    },
    {
      "epoch": 7.987421383647799,
      "grad_norm": 0.09699012339115143,
      "learning_rate": 0.0007509895835453224,
      "loss": 0.3927,
      "num_input_tokens_seen": 9944080,
      "step": 15240
    },
    {
      "epoch": 7.990041928721174,
      "grad_norm": 0.13463690876960754,
      "learning_rate": 0.0007507917716790201,
      "loss": 0.4209,
      "num_input_tokens_seen": 9946480,
      "step": 15245
    },
    {
      "epoch": 7.9926624737945495,
      "grad_norm": 0.09155893325805664,
      "learning_rate": 0.0007505939073499401,
      "loss": 0.4688,
      "num_input_tokens_seen": 9949872,
      "step": 15250
    },
    {
      "epoch": 7.995283018867925,
      "grad_norm": 0.10700905323028564,
      "learning_rate": 0.0007503959905994737,
      "loss": 0.4237,
      "num_input_tokens_seen": 9953520,
      "step": 15255
    },
    {
      "epoch": 7.9979035639413,
      "grad_norm": 0.09876551479101181,
      "learning_rate": 0.0007501980214690227,
      "loss": 0.4793,
      "num_input_tokens_seen": 9956624,
      "step": 15260
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.4539332985877991,
      "eval_runtime": 13.6095,
      "eval_samples_per_second": 62.309,
      "eval_steps_per_second": 15.577,
      "num_input_tokens_seen": 9958360,
      "step": 15264
    },
    {
      "epoch": 8.000524109014675,
      "grad_norm": 0.08682078123092651,
      "learning_rate": 0.00075,
      "loss": 0.448,
      "num_input_tokens_seen": 9958968,
      "step": 15265
    },
    {
      "epoch": 8.00314465408805,
      "grad_norm": 0.0889948382973671,
      "learning_rate": 0.0007498019262338295,
      "loss": 0.388,
      "num_input_tokens_seen": 9961624,
      "step": 15270
    },
    {
      "epoch": 8.005765199161425,
      "grad_norm": 0.14354367554187775,
      "learning_rate": 0.0007496038002119459,
      "loss": 0.5294,
      "num_input_tokens_seen": 9964600,
      "step": 15275
    },
    {
      "epoch": 8.0083857442348,
      "grad_norm": 0.12262675911188126,
      "learning_rate": 0.000749405621975795,
      "loss": 0.4519,
      "num_input_tokens_seen": 9967096,
      "step": 15280
    },
    {
      "epoch": 8.011006289308176,
      "grad_norm": 0.08615490049123764,
      "learning_rate": 0.0007492073915668334,
      "loss": 0.4108,
      "num_input_tokens_seen": 9969848,
      "step": 15285
    },
    {
      "epoch": 8.01362683438155,
      "grad_norm": 0.06651803106069565,
      "learning_rate": 0.0007490091090265283,
      "loss": 0.4293,
      "num_input_tokens_seen": 9973272,
      "step": 15290
    },
    {
      "epoch": 8.016247379454926,
      "grad_norm": 0.07416173070669174,
      "learning_rate": 0.0007488107743963587,
      "loss": 0.4019,
      "num_input_tokens_seen": 9975992,
      "step": 15295
    },
    {
      "epoch": 8.018867924528301,
      "grad_norm": 0.06265910714864731,
      "learning_rate": 0.0007486123877178136,
      "loss": 0.5395,
      "num_input_tokens_seen": 9979704,
      "step": 15300
    },
    {
      "epoch": 8.021488469601676,
      "grad_norm": 0.10584340989589691,
      "learning_rate": 0.0007484139490323932,
      "loss": 0.4772,
      "num_input_tokens_seen": 9982712,
      "step": 15305
    },
    {
      "epoch": 8.024109014675052,
      "grad_norm": 0.10789987444877625,
      "learning_rate": 0.000748215458381609,
      "loss": 0.3334,
      "num_input_tokens_seen": 9985560,
      "step": 15310
    },
    {
      "epoch": 8.026729559748428,
      "grad_norm": 0.1052044928073883,
      "learning_rate": 0.0007480169158069826,
      "loss": 0.3968,
      "num_input_tokens_seen": 9989272,
      "step": 15315
    },
    {
      "epoch": 8.029350104821804,
      "grad_norm": 0.10905996710062027,
      "learning_rate": 0.0007478183213500468,
      "loss": 0.3889,
      "num_input_tokens_seen": 9992088,
      "step": 15320
    },
    {
      "epoch": 8.031970649895179,
      "grad_norm": 0.11277889460325241,
      "learning_rate": 0.0007476196750523456,
      "loss": 0.4608,
      "num_input_tokens_seen": 9995032,
      "step": 15325
    },
    {
      "epoch": 8.034591194968554,
      "grad_norm": 0.19134528934955597,
      "learning_rate": 0.000747420976955433,
      "loss": 0.338,
      "num_input_tokens_seen": 9998008,
      "step": 15330
    },
    {
      "epoch": 8.03721174004193,
      "grad_norm": 0.05736108869314194,
      "learning_rate": 0.0007472222271008752,
      "loss": 0.5982,
      "num_input_tokens_seen": 10001464,
      "step": 15335
    },
    {
      "epoch": 8.039832285115304,
      "grad_norm": 0.0875384584069252,
      "learning_rate": 0.0007470234255302478,
      "loss": 0.427,
      "num_input_tokens_seen": 10004472,
      "step": 15340
    },
    {
      "epoch": 8.04245283018868,
      "grad_norm": 0.15239174664020538,
      "learning_rate": 0.0007468245722851379,
      "loss": 0.4309,
      "num_input_tokens_seen": 10007160,
      "step": 15345
    },
    {
      "epoch": 8.045073375262055,
      "grad_norm": 0.05685499310493469,
      "learning_rate": 0.0007466256674071434,
      "loss": 0.5246,
      "num_input_tokens_seen": 10010136,
      "step": 15350
    },
    {
      "epoch": 8.04769392033543,
      "grad_norm": 0.13663998246192932,
      "learning_rate": 0.000746426710937873,
      "loss": 0.4998,
      "num_input_tokens_seen": 10013560,
      "step": 15355
    },
    {
      "epoch": 8.050314465408805,
      "grad_norm": 0.09063687920570374,
      "learning_rate": 0.0007462277029189461,
      "loss": 0.4469,
      "num_input_tokens_seen": 10016280,
      "step": 15360
    },
    {
      "epoch": 8.05293501048218,
      "grad_norm": 0.12177053093910217,
      "learning_rate": 0.0007460286433919928,
      "loss": 0.6,
      "num_input_tokens_seen": 10019096,
      "step": 15365
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 0.0943535640835762,
      "learning_rate": 0.0007458295323986542,
      "loss": 0.511,
      "num_input_tokens_seen": 10022200,
      "step": 15370
    },
    {
      "epoch": 8.05817610062893,
      "grad_norm": 0.13441354036331177,
      "learning_rate": 0.0007456303699805817,
      "loss": 0.4542,
      "num_input_tokens_seen": 10025528,
      "step": 15375
    },
    {
      "epoch": 8.060796645702306,
      "grad_norm": 0.10113144665956497,
      "learning_rate": 0.0007454311561794386,
      "loss": 0.5148,
      "num_input_tokens_seen": 10030104,
      "step": 15380
    },
    {
      "epoch": 8.06341719077568,
      "grad_norm": 0.11422088742256165,
      "learning_rate": 0.0007452318910368972,
      "loss": 0.5267,
      "num_input_tokens_seen": 10033112,
      "step": 15385
    },
    {
      "epoch": 8.066037735849056,
      "grad_norm": 0.11458476632833481,
      "learning_rate": 0.0007450325745946422,
      "loss": 0.4152,
      "num_input_tokens_seen": 10036280,
      "step": 15390
    },
    {
      "epoch": 8.068658280922431,
      "grad_norm": 0.08362336456775665,
      "learning_rate": 0.0007448332068943681,
      "loss": 0.3682,
      "num_input_tokens_seen": 10039160,
      "step": 15395
    },
    {
      "epoch": 8.071278825995806,
      "grad_norm": 0.05659577250480652,
      "learning_rate": 0.0007446337879777802,
      "loss": 0.4283,
      "num_input_tokens_seen": 10043224,
      "step": 15400
    },
    {
      "epoch": 8.073899371069182,
      "grad_norm": 0.08855431526899338,
      "learning_rate": 0.0007444343178865948,
      "loss": 0.4139,
      "num_input_tokens_seen": 10045976,
      "step": 15405
    },
    {
      "epoch": 8.076519916142558,
      "grad_norm": 0.06988479942083359,
      "learning_rate": 0.0007442347966625389,
      "loss": 0.387,
      "num_input_tokens_seen": 10048824,
      "step": 15410
    },
    {
      "epoch": 8.079140461215934,
      "grad_norm": 0.08820968866348267,
      "learning_rate": 0.0007440352243473496,
      "loss": 0.4482,
      "num_input_tokens_seen": 10051512,
      "step": 15415
    },
    {
      "epoch": 8.081761006289309,
      "grad_norm": 0.10105746239423752,
      "learning_rate": 0.0007438356009827758,
      "loss": 0.5499,
      "num_input_tokens_seen": 10054712,
      "step": 15420
    },
    {
      "epoch": 8.084381551362684,
      "grad_norm": 0.10521382838487625,
      "learning_rate": 0.0007436359266105759,
      "loss": 0.5347,
      "num_input_tokens_seen": 10058232,
      "step": 15425
    },
    {
      "epoch": 8.08700209643606,
      "grad_norm": 0.05671608820557594,
      "learning_rate": 0.0007434362012725198,
      "loss": 0.4843,
      "num_input_tokens_seen": 10061880,
      "step": 15430
    },
    {
      "epoch": 8.089622641509434,
      "grad_norm": 0.057401761412620544,
      "learning_rate": 0.0007432364250103877,
      "loss": 0.4756,
      "num_input_tokens_seen": 10064984,
      "step": 15435
    },
    {
      "epoch": 8.09224318658281,
      "grad_norm": 0.08107790350914001,
      "learning_rate": 0.0007430365978659704,
      "loss": 0.4515,
      "num_input_tokens_seen": 10067384,
      "step": 15440
    },
    {
      "epoch": 8.094863731656185,
      "grad_norm": 0.058992013335227966,
      "learning_rate": 0.0007428367198810694,
      "loss": 0.4576,
      "num_input_tokens_seen": 10070776,
      "step": 15445
    },
    {
      "epoch": 8.09748427672956,
      "grad_norm": 0.13232672214508057,
      "learning_rate": 0.0007426367910974973,
      "loss": 0.3642,
      "num_input_tokens_seen": 10074104,
      "step": 15450
    },
    {
      "epoch": 8.100104821802935,
      "grad_norm": 0.08667711168527603,
      "learning_rate": 0.0007424368115570763,
      "loss": 0.3941,
      "num_input_tokens_seen": 10077784,
      "step": 15455
    },
    {
      "epoch": 8.10272536687631,
      "grad_norm": 0.1202806904911995,
      "learning_rate": 0.0007422367813016405,
      "loss": 0.5034,
      "num_input_tokens_seen": 10082392,
      "step": 15460
    },
    {
      "epoch": 8.105345911949685,
      "grad_norm": 0.09145142883062363,
      "learning_rate": 0.0007420367003730335,
      "loss": 0.398,
      "num_input_tokens_seen": 10085656,
      "step": 15465
    },
    {
      "epoch": 8.10796645702306,
      "grad_norm": 0.11872225999832153,
      "learning_rate": 0.0007418365688131103,
      "loss": 0.4781,
      "num_input_tokens_seen": 10089688,
      "step": 15470
    },
    {
      "epoch": 8.110587002096436,
      "grad_norm": 0.11055956780910492,
      "learning_rate": 0.0007416363866637356,
      "loss": 0.5415,
      "num_input_tokens_seen": 10092632,
      "step": 15475
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 0.13297750055789948,
      "learning_rate": 0.0007414361539667857,
      "loss": 0.4565,
      "num_input_tokens_seen": 10095384,
      "step": 15480
    },
    {
      "epoch": 8.115828092243186,
      "grad_norm": 0.14480197429656982,
      "learning_rate": 0.0007412358707641468,
      "loss": 0.5165,
      "num_input_tokens_seen": 10101272,
      "step": 15485
    },
    {
      "epoch": 8.118448637316561,
      "grad_norm": 0.09907440096139908,
      "learning_rate": 0.0007410355370977158,
      "loss": 0.5086,
      "num_input_tokens_seen": 10104920,
      "step": 15490
    },
    {
      "epoch": 8.121069182389936,
      "grad_norm": 0.07360955327749252,
      "learning_rate": 0.0007408351530094005,
      "loss": 0.4807,
      "num_input_tokens_seen": 10108440,
      "step": 15495
    },
    {
      "epoch": 8.123689727463312,
      "grad_norm": 0.12776872515678406,
      "learning_rate": 0.0007406347185411187,
      "loss": 0.4061,
      "num_input_tokens_seen": 10111672,
      "step": 15500
    },
    {
      "epoch": 8.126310272536688,
      "grad_norm": 0.054808564484119415,
      "learning_rate": 0.0007404342337347992,
      "loss": 0.5184,
      "num_input_tokens_seen": 10115448,
      "step": 15505
    },
    {
      "epoch": 8.128930817610064,
      "grad_norm": 0.05944914370775223,
      "learning_rate": 0.000740233698632381,
      "loss": 0.3763,
      "num_input_tokens_seen": 10118328,
      "step": 15510
    },
    {
      "epoch": 8.131551362683439,
      "grad_norm": 0.12194183468818665,
      "learning_rate": 0.0007400331132758136,
      "loss": 0.4825,
      "num_input_tokens_seen": 10122328,
      "step": 15515
    },
    {
      "epoch": 8.134171907756814,
      "grad_norm": 0.09455855935811996,
      "learning_rate": 0.0007398324777070576,
      "loss": 0.4991,
      "num_input_tokens_seen": 10125112,
      "step": 15520
    },
    {
      "epoch": 8.13679245283019,
      "grad_norm": 0.09130855649709702,
      "learning_rate": 0.0007396317919680832,
      "loss": 0.3635,
      "num_input_tokens_seen": 10128280,
      "step": 15525
    },
    {
      "epoch": 8.139412997903564,
      "grad_norm": 0.10677646100521088,
      "learning_rate": 0.0007394310561008717,
      "loss": 0.3623,
      "num_input_tokens_seen": 10131320,
      "step": 15530
    },
    {
      "epoch": 8.14203354297694,
      "grad_norm": 0.11228051781654358,
      "learning_rate": 0.0007392302701474151,
      "loss": 0.4633,
      "num_input_tokens_seen": 10135384,
      "step": 15535
    },
    {
      "epoch": 8.144654088050315,
      "grad_norm": 0.09972144663333893,
      "learning_rate": 0.0007390294341497151,
      "loss": 0.5788,
      "num_input_tokens_seen": 10138520,
      "step": 15540
    },
    {
      "epoch": 8.14727463312369,
      "grad_norm": 0.07884478569030762,
      "learning_rate": 0.0007388285481497845,
      "loss": 0.4522,
      "num_input_tokens_seen": 10141528,
      "step": 15545
    },
    {
      "epoch": 8.149895178197065,
      "grad_norm": 0.07109296321868896,
      "learning_rate": 0.0007386276121896465,
      "loss": 0.5004,
      "num_input_tokens_seen": 10144568,
      "step": 15550
    },
    {
      "epoch": 8.15251572327044,
      "grad_norm": 0.08996283262968063,
      "learning_rate": 0.0007384266263113342,
      "loss": 0.4567,
      "num_input_tokens_seen": 10148504,
      "step": 15555
    },
    {
      "epoch": 8.155136268343815,
      "grad_norm": 0.059551361948251724,
      "learning_rate": 0.0007382255905568919,
      "loss": 0.4814,
      "num_input_tokens_seen": 10151288,
      "step": 15560
    },
    {
      "epoch": 8.15775681341719,
      "grad_norm": 0.06375325471162796,
      "learning_rate": 0.000738024504968374,
      "loss": 0.4558,
      "num_input_tokens_seen": 10154744,
      "step": 15565
    },
    {
      "epoch": 8.160377358490566,
      "grad_norm": 0.09127175807952881,
      "learning_rate": 0.0007378233695878449,
      "loss": 0.358,
      "num_input_tokens_seen": 10158136,
      "step": 15570
    },
    {
      "epoch": 8.16299790356394,
      "grad_norm": 0.07538159191608429,
      "learning_rate": 0.0007376221844573803,
      "loss": 0.4109,
      "num_input_tokens_seen": 10161560,
      "step": 15575
    },
    {
      "epoch": 8.165618448637316,
      "grad_norm": 0.09032686799764633,
      "learning_rate": 0.0007374209496190656,
      "loss": 0.495,
      "num_input_tokens_seen": 10163960,
      "step": 15580
    },
    {
      "epoch": 8.168238993710691,
      "grad_norm": 0.1092090904712677,
      "learning_rate": 0.0007372196651149968,
      "loss": 0.3452,
      "num_input_tokens_seen": 10166776,
      "step": 15585
    },
    {
      "epoch": 8.170859538784066,
      "grad_norm": 0.10717881470918655,
      "learning_rate": 0.0007370183309872805,
      "loss": 0.3459,
      "num_input_tokens_seen": 10169592,
      "step": 15590
    },
    {
      "epoch": 8.173480083857442,
      "grad_norm": 0.08021748065948486,
      "learning_rate": 0.0007368169472780335,
      "loss": 0.4298,
      "num_input_tokens_seen": 10172760,
      "step": 15595
    },
    {
      "epoch": 8.176100628930818,
      "grad_norm": 0.05940985679626465,
      "learning_rate": 0.0007366155140293828,
      "loss": 0.4605,
      "num_input_tokens_seen": 10177496,
      "step": 15600
    },
    {
      "epoch": 8.178721174004194,
      "grad_norm": 0.16004176437854767,
      "learning_rate": 0.000736414031283466,
      "loss": 0.4404,
      "num_input_tokens_seen": 10181112,
      "step": 15605
    },
    {
      "epoch": 8.181341719077569,
      "grad_norm": 0.06824605911970139,
      "learning_rate": 0.000736212499082431,
      "loss": 0.4784,
      "num_input_tokens_seen": 10184632,
      "step": 15610
    },
    {
      "epoch": 8.183962264150944,
      "grad_norm": 0.08210420608520508,
      "learning_rate": 0.0007360109174684359,
      "loss": 0.4417,
      "num_input_tokens_seen": 10187576,
      "step": 15615
    },
    {
      "epoch": 8.18658280922432,
      "grad_norm": 0.1377170830965042,
      "learning_rate": 0.0007358092864836497,
      "loss": 0.5626,
      "num_input_tokens_seen": 10190648,
      "step": 15620
    },
    {
      "epoch": 8.189203354297694,
      "grad_norm": 0.06801638752222061,
      "learning_rate": 0.0007356076061702508,
      "loss": 0.3287,
      "num_input_tokens_seen": 10193944,
      "step": 15625
    },
    {
      "epoch": 8.19182389937107,
      "grad_norm": 0.06262906640768051,
      "learning_rate": 0.0007354058765704288,
      "loss": 0.4079,
      "num_input_tokens_seen": 10196824,
      "step": 15630
    },
    {
      "epoch": 8.194444444444445,
      "grad_norm": 0.10327622294425964,
      "learning_rate": 0.000735204097726383,
      "loss": 0.5183,
      "num_input_tokens_seen": 10200312,
      "step": 15635
    },
    {
      "epoch": 8.19706498951782,
      "grad_norm": 0.06625822186470032,
      "learning_rate": 0.0007350022696803233,
      "loss": 0.5734,
      "num_input_tokens_seen": 10203992,
      "step": 15640
    },
    {
      "epoch": 8.199685534591195,
      "grad_norm": 0.1397794783115387,
      "learning_rate": 0.0007348003924744699,
      "loss": 0.416,
      "num_input_tokens_seen": 10206904,
      "step": 15645
    },
    {
      "epoch": 8.20230607966457,
      "grad_norm": 0.06773652881383896,
      "learning_rate": 0.000734598466151053,
      "loss": 0.4541,
      "num_input_tokens_seen": 10209880,
      "step": 15650
    },
    {
      "epoch": 8.204926624737945,
      "grad_norm": 0.16382449865341187,
      "learning_rate": 0.0007343964907523135,
      "loss": 0.5031,
      "num_input_tokens_seen": 10215864,
      "step": 15655
    },
    {
      "epoch": 8.20754716981132,
      "grad_norm": 0.1479797512292862,
      "learning_rate": 0.0007341944663205023,
      "loss": 0.4471,
      "num_input_tokens_seen": 10218904,
      "step": 15660
    },
    {
      "epoch": 8.210167714884696,
      "grad_norm": 0.09997900575399399,
      "learning_rate": 0.0007339923928978805,
      "loss": 0.4026,
      "num_input_tokens_seen": 10225048,
      "step": 15665
    },
    {
      "epoch": 8.21278825995807,
      "grad_norm": 0.06084802746772766,
      "learning_rate": 0.0007337902705267197,
      "loss": 0.4522,
      "num_input_tokens_seen": 10227576,
      "step": 15670
    },
    {
      "epoch": 8.215408805031446,
      "grad_norm": 0.2505676746368408,
      "learning_rate": 0.0007335880992493016,
      "loss": 0.6135,
      "num_input_tokens_seen": 10230712,
      "step": 15675
    },
    {
      "epoch": 8.218029350104821,
      "grad_norm": 0.15114498138427734,
      "learning_rate": 0.000733385879107918,
      "loss": 0.544,
      "num_input_tokens_seen": 10233848,
      "step": 15680
    },
    {
      "epoch": 8.220649895178196,
      "grad_norm": 0.06761805713176727,
      "learning_rate": 0.000733183610144871,
      "loss": 0.4757,
      "num_input_tokens_seen": 10237816,
      "step": 15685
    },
    {
      "epoch": 8.223270440251572,
      "grad_norm": 0.07418669760227203,
      "learning_rate": 0.0007329812924024732,
      "loss": 0.4344,
      "num_input_tokens_seen": 10241272,
      "step": 15690
    },
    {
      "epoch": 8.225890985324948,
      "grad_norm": 0.11405172199010849,
      "learning_rate": 0.0007327789259230468,
      "loss": 0.4673,
      "num_input_tokens_seen": 10243864,
      "step": 15695
    },
    {
      "epoch": 8.228511530398324,
      "grad_norm": 0.06366896629333496,
      "learning_rate": 0.0007325765107489249,
      "loss": 0.3279,
      "num_input_tokens_seen": 10246712,
      "step": 15700
    },
    {
      "epoch": 8.231132075471699,
      "grad_norm": 0.051303569227457047,
      "learning_rate": 0.0007323740469224505,
      "loss": 0.5034,
      "num_input_tokens_seen": 10250648,
      "step": 15705
    },
    {
      "epoch": 8.233752620545074,
      "grad_norm": 0.1573975533246994,
      "learning_rate": 0.0007321715344859763,
      "loss": 0.4929,
      "num_input_tokens_seen": 10253624,
      "step": 15710
    },
    {
      "epoch": 8.23637316561845,
      "grad_norm": 0.15787078440189362,
      "learning_rate": 0.000731968973481866,
      "loss": 0.426,
      "num_input_tokens_seen": 10256536,
      "step": 15715
    },
    {
      "epoch": 8.238993710691824,
      "grad_norm": 0.21116787195205688,
      "learning_rate": 0.0007317663639524928,
      "loss": 0.4262,
      "num_input_tokens_seen": 10259448,
      "step": 15720
    },
    {
      "epoch": 8.2416142557652,
      "grad_norm": 0.14374001324176788,
      "learning_rate": 0.0007315637059402404,
      "loss": 0.4584,
      "num_input_tokens_seen": 10263768,
      "step": 15725
    },
    {
      "epoch": 8.244234800838575,
      "grad_norm": 0.1711031198501587,
      "learning_rate": 0.0007313609994875024,
      "loss": 0.4811,
      "num_input_tokens_seen": 10266712,
      "step": 15730
    },
    {
      "epoch": 8.24685534591195,
      "grad_norm": 0.0918717309832573,
      "learning_rate": 0.0007311582446366831,
      "loss": 0.6723,
      "num_input_tokens_seen": 10270712,
      "step": 15735
    },
    {
      "epoch": 8.249475890985325,
      "grad_norm": 0.1759357452392578,
      "learning_rate": 0.000730955441430196,
      "loss": 0.5871,
      "num_input_tokens_seen": 10274456,
      "step": 15740
    },
    {
      "epoch": 8.2520964360587,
      "grad_norm": 0.0644715204834938,
      "learning_rate": 0.0007307525899104655,
      "loss": 0.4464,
      "num_input_tokens_seen": 10278840,
      "step": 15745
    },
    {
      "epoch": 8.254716981132075,
      "grad_norm": 0.09908226132392883,
      "learning_rate": 0.0007305496901199259,
      "loss": 0.3343,
      "num_input_tokens_seen": 10282104,
      "step": 15750
    },
    {
      "epoch": 8.25733752620545,
      "grad_norm": 0.08396121859550476,
      "learning_rate": 0.0007303467421010211,
      "loss": 0.5292,
      "num_input_tokens_seen": 10285272,
      "step": 15755
    },
    {
      "epoch": 8.259958071278826,
      "grad_norm": 0.17382925748825073,
      "learning_rate": 0.0007301437458962061,
      "loss": 0.4461,
      "num_input_tokens_seen": 10288472,
      "step": 15760
    },
    {
      "epoch": 8.2625786163522,
      "grad_norm": 0.10809693485498428,
      "learning_rate": 0.0007299407015479448,
      "loss": 0.5011,
      "num_input_tokens_seen": 10291128,
      "step": 15765
    },
    {
      "epoch": 8.265199161425576,
      "grad_norm": 0.09272295236587524,
      "learning_rate": 0.0007297376090987122,
      "loss": 0.4603,
      "num_input_tokens_seen": 10295288,
      "step": 15770
    },
    {
      "epoch": 8.267819706498951,
      "grad_norm": 0.09505169838666916,
      "learning_rate": 0.0007295344685909926,
      "loss": 0.4338,
      "num_input_tokens_seen": 10298680,
      "step": 15775
    },
    {
      "epoch": 8.270440251572326,
      "grad_norm": 0.11324772983789444,
      "learning_rate": 0.000729331280067281,
      "loss": 0.4846,
      "num_input_tokens_seen": 10301496,
      "step": 15780
    },
    {
      "epoch": 8.273060796645701,
      "grad_norm": 0.1285163313150406,
      "learning_rate": 0.0007291280435700819,
      "loss": 0.3779,
      "num_input_tokens_seen": 10303992,
      "step": 15785
    },
    {
      "epoch": 8.275681341719078,
      "grad_norm": 0.07567115873098373,
      "learning_rate": 0.0007289247591419102,
      "loss": 0.4063,
      "num_input_tokens_seen": 10306808,
      "step": 15790
    },
    {
      "epoch": 8.278301886792454,
      "grad_norm": 0.04008674621582031,
      "learning_rate": 0.0007287214268252904,
      "loss": 0.4089,
      "num_input_tokens_seen": 10310584,
      "step": 15795
    },
    {
      "epoch": 8.280922431865829,
      "grad_norm": 0.05650902912020683,
      "learning_rate": 0.0007285180466627577,
      "loss": 0.3499,
      "num_input_tokens_seen": 10314872,
      "step": 15800
    },
    {
      "epoch": 8.283542976939204,
      "grad_norm": 0.08958634734153748,
      "learning_rate": 0.0007283146186968565,
      "loss": 0.3708,
      "num_input_tokens_seen": 10317624,
      "step": 15805
    },
    {
      "epoch": 8.286163522012579,
      "grad_norm": 0.12597563862800598,
      "learning_rate": 0.000728111142970142,
      "loss": 0.3894,
      "num_input_tokens_seen": 10320408,
      "step": 15810
    },
    {
      "epoch": 8.288784067085954,
      "grad_norm": 0.10498564690351486,
      "learning_rate": 0.0007279076195251788,
      "loss": 0.498,
      "num_input_tokens_seen": 10323320,
      "step": 15815
    },
    {
      "epoch": 8.29140461215933,
      "grad_norm": 0.07470580190420151,
      "learning_rate": 0.0007277040484045416,
      "loss": 0.3803,
      "num_input_tokens_seen": 10326776,
      "step": 15820
    },
    {
      "epoch": 8.294025157232705,
      "grad_norm": 0.1591021567583084,
      "learning_rate": 0.0007275004296508154,
      "loss": 0.2886,
      "num_input_tokens_seen": 10330584,
      "step": 15825
    },
    {
      "epoch": 8.29664570230608,
      "grad_norm": 0.10767313838005066,
      "learning_rate": 0.0007272967633065947,
      "loss": 0.4489,
      "num_input_tokens_seen": 10332824,
      "step": 15830
    },
    {
      "epoch": 8.299266247379455,
      "grad_norm": 0.15278826653957367,
      "learning_rate": 0.0007270930494144843,
      "loss": 0.5632,
      "num_input_tokens_seen": 10335480,
      "step": 15835
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 0.07333948463201523,
      "learning_rate": 0.0007268892880170988,
      "loss": 0.4803,
      "num_input_tokens_seen": 10338424,
      "step": 15840
    },
    {
      "epoch": 8.304507337526205,
      "grad_norm": 0.0816100537776947,
      "learning_rate": 0.0007266854791570625,
      "loss": 0.464,
      "num_input_tokens_seen": 10341784,
      "step": 15845
    },
    {
      "epoch": 8.30712788259958,
      "grad_norm": 0.09720376878976822,
      "learning_rate": 0.0007264816228770103,
      "loss": 0.4462,
      "num_input_tokens_seen": 10345144,
      "step": 15850
    },
    {
      "epoch": 8.309748427672956,
      "grad_norm": 0.07038898766040802,
      "learning_rate": 0.0007262777192195866,
      "loss": 0.4094,
      "num_input_tokens_seen": 10348984,
      "step": 15855
    },
    {
      "epoch": 8.31236897274633,
      "grad_norm": 0.1033870130777359,
      "learning_rate": 0.0007260737682274452,
      "loss": 0.4641,
      "num_input_tokens_seen": 10352152,
      "step": 15860
    },
    {
      "epoch": 8.314989517819706,
      "grad_norm": 0.19434359669685364,
      "learning_rate": 0.0007258697699432511,
      "loss": 0.5134,
      "num_input_tokens_seen": 10354392,
      "step": 15865
    },
    {
      "epoch": 8.317610062893081,
      "grad_norm": 0.08878914266824722,
      "learning_rate": 0.0007256657244096775,
      "loss": 0.7279,
      "num_input_tokens_seen": 10357368,
      "step": 15870
    },
    {
      "epoch": 8.320230607966456,
      "grad_norm": 0.13517649471759796,
      "learning_rate": 0.0007254616316694091,
      "loss": 0.4135,
      "num_input_tokens_seen": 10359864,
      "step": 15875
    },
    {
      "epoch": 8.322851153039831,
      "grad_norm": 0.08767963945865631,
      "learning_rate": 0.0007252574917651398,
      "loss": 0.5901,
      "num_input_tokens_seen": 10364536,
      "step": 15880
    },
    {
      "epoch": 8.325471698113208,
      "grad_norm": 0.10571492463350296,
      "learning_rate": 0.0007250533047395728,
      "loss": 0.4941,
      "num_input_tokens_seen": 10367864,
      "step": 15885
    },
    {
      "epoch": 8.328092243186584,
      "grad_norm": 0.04511900246143341,
      "learning_rate": 0.0007248490706354221,
      "loss": 0.5218,
      "num_input_tokens_seen": 10371736,
      "step": 15890
    },
    {
      "epoch": 8.330712788259959,
      "grad_norm": 0.07834286987781525,
      "learning_rate": 0.0007246447894954109,
      "loss": 0.3332,
      "num_input_tokens_seen": 10375192,
      "step": 15895
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 0.049104075878858566,
      "learning_rate": 0.0007244404613622727,
      "loss": 0.4251,
      "num_input_tokens_seen": 10383288,
      "step": 15900
    },
    {
      "epoch": 8.335953878406709,
      "grad_norm": 0.08888837695121765,
      "learning_rate": 0.0007242360862787507,
      "loss": 0.5489,
      "num_input_tokens_seen": 10387192,
      "step": 15905
    },
    {
      "epoch": 8.338574423480084,
      "grad_norm": 0.14239320158958435,
      "learning_rate": 0.0007240316642875975,
      "loss": 0.4212,
      "num_input_tokens_seen": 10389400,
      "step": 15910
    },
    {
      "epoch": 8.34119496855346,
      "grad_norm": 0.10052424669265747,
      "learning_rate": 0.0007238271954315759,
      "loss": 0.5033,
      "num_input_tokens_seen": 10392920,
      "step": 15915
    },
    {
      "epoch": 8.343815513626835,
      "grad_norm": 0.11743893474340439,
      "learning_rate": 0.0007236226797534589,
      "loss": 0.4463,
      "num_input_tokens_seen": 10396280,
      "step": 15920
    },
    {
      "epoch": 8.34643605870021,
      "grad_norm": 0.23776599764823914,
      "learning_rate": 0.0007234181172960283,
      "loss": 0.5118,
      "num_input_tokens_seen": 10399960,
      "step": 15925
    },
    {
      "epoch": 8.349056603773585,
      "grad_norm": 0.11278410255908966,
      "learning_rate": 0.0007232135081020766,
      "loss": 0.4551,
      "num_input_tokens_seen": 10403128,
      "step": 15930
    },
    {
      "epoch": 8.35167714884696,
      "grad_norm": 0.14267057180404663,
      "learning_rate": 0.0007230088522144054,
      "loss": 0.4246,
      "num_input_tokens_seen": 10405976,
      "step": 15935
    },
    {
      "epoch": 8.354297693920335,
      "grad_norm": 0.185385599732399,
      "learning_rate": 0.0007228041496758265,
      "loss": 0.4433,
      "num_input_tokens_seen": 10408632,
      "step": 15940
    },
    {
      "epoch": 8.35691823899371,
      "grad_norm": 0.08456125110387802,
      "learning_rate": 0.0007225994005291615,
      "loss": 0.5252,
      "num_input_tokens_seen": 10412216,
      "step": 15945
    },
    {
      "epoch": 8.359538784067086,
      "grad_norm": 0.23814301192760468,
      "learning_rate": 0.0007223946048172414,
      "loss": 0.444,
      "num_input_tokens_seen": 10414680,
      "step": 15950
    },
    {
      "epoch": 8.36215932914046,
      "grad_norm": 0.10620451718568802,
      "learning_rate": 0.0007221897625829072,
      "loss": 0.5205,
      "num_input_tokens_seen": 10417560,
      "step": 15955
    },
    {
      "epoch": 8.364779874213836,
      "grad_norm": 0.09386321157217026,
      "learning_rate": 0.0007219848738690095,
      "loss": 0.3483,
      "num_input_tokens_seen": 10420056,
      "step": 15960
    },
    {
      "epoch": 8.367400419287211,
      "grad_norm": 0.08552084863185883,
      "learning_rate": 0.0007217799387184089,
      "loss": 0.3547,
      "num_input_tokens_seen": 10423288,
      "step": 15965
    },
    {
      "epoch": 8.370020964360586,
      "grad_norm": 0.12724974751472473,
      "learning_rate": 0.0007215749571739752,
      "loss": 0.5003,
      "num_input_tokens_seen": 10425752,
      "step": 15970
    },
    {
      "epoch": 8.372641509433961,
      "grad_norm": 0.05913924425840378,
      "learning_rate": 0.0007213699292785883,
      "loss": 0.3969,
      "num_input_tokens_seen": 10429432,
      "step": 15975
    },
    {
      "epoch": 8.375262054507338,
      "grad_norm": 0.1030280590057373,
      "learning_rate": 0.0007211648550751377,
      "loss": 0.448,
      "num_input_tokens_seen": 10432856,
      "step": 15980
    },
    {
      "epoch": 8.377882599580714,
      "grad_norm": 0.11102644354104996,
      "learning_rate": 0.0007209597346065226,
      "loss": 0.5244,
      "num_input_tokens_seen": 10436600,
      "step": 15985
    },
    {
      "epoch": 8.380503144654089,
      "grad_norm": 0.09827805310487747,
      "learning_rate": 0.000720754567915652,
      "loss": 0.4207,
      "num_input_tokens_seen": 10439352,
      "step": 15990
    },
    {
      "epoch": 8.383123689727464,
      "grad_norm": 0.13123488426208496,
      "learning_rate": 0.000720549355045444,
      "loss": 0.4712,
      "num_input_tokens_seen": 10441976,
      "step": 15995
    },
    {
      "epoch": 8.385744234800839,
      "grad_norm": 0.10022435337305069,
      "learning_rate": 0.0007203440960388273,
      "loss": 0.4836,
      "num_input_tokens_seen": 10445272,
      "step": 16000
    },
    {
      "epoch": 8.388364779874214,
      "grad_norm": 0.07994692772626877,
      "learning_rate": 0.0007201387909387396,
      "loss": 0.5374,
      "num_input_tokens_seen": 10447864,
      "step": 16005
    },
    {
      "epoch": 8.39098532494759,
      "grad_norm": 0.0796879306435585,
      "learning_rate": 0.0007199334397881282,
      "loss": 0.3996,
      "num_input_tokens_seen": 10451992,
      "step": 16010
    },
    {
      "epoch": 8.393605870020965,
      "grad_norm": 0.10899140685796738,
      "learning_rate": 0.0007197280426299502,
      "loss": 0.524,
      "num_input_tokens_seen": 10454968,
      "step": 16015
    },
    {
      "epoch": 8.39622641509434,
      "grad_norm": 0.14102663099765778,
      "learning_rate": 0.0007195225995071727,
      "loss": 0.4876,
      "num_input_tokens_seen": 10458392,
      "step": 16020
    },
    {
      "epoch": 8.398846960167715,
      "grad_norm": 0.09093392640352249,
      "learning_rate": 0.0007193171104627718,
      "loss": 0.5741,
      "num_input_tokens_seen": 10461144,
      "step": 16025
    },
    {
      "epoch": 8.40146750524109,
      "grad_norm": 0.13391445577144623,
      "learning_rate": 0.0007191115755397333,
      "loss": 0.4064,
      "num_input_tokens_seen": 10463768,
      "step": 16030
    },
    {
      "epoch": 8.404088050314465,
      "grad_norm": 0.13549867272377014,
      "learning_rate": 0.0007189059947810531,
      "loss": 0.422,
      "num_input_tokens_seen": 10467128,
      "step": 16035
    },
    {
      "epoch": 8.40670859538784,
      "grad_norm": 0.2219201922416687,
      "learning_rate": 0.0007187003682297363,
      "loss": 0.442,
      "num_input_tokens_seen": 10469240,
      "step": 16040
    },
    {
      "epoch": 8.409329140461216,
      "grad_norm": 0.08587050437927246,
      "learning_rate": 0.0007184946959287972,
      "loss": 0.4523,
      "num_input_tokens_seen": 10472728,
      "step": 16045
    },
    {
      "epoch": 8.41194968553459,
      "grad_norm": 0.2658294141292572,
      "learning_rate": 0.0007182889779212609,
      "loss": 0.3999,
      "num_input_tokens_seen": 10474936,
      "step": 16050
    },
    {
      "epoch": 8.414570230607966,
      "grad_norm": 0.09164479374885559,
      "learning_rate": 0.0007180832142501605,
      "loss": 0.4685,
      "num_input_tokens_seen": 10478264,
      "step": 16055
    },
    {
      "epoch": 8.417190775681341,
      "grad_norm": 0.0767846554517746,
      "learning_rate": 0.0007178774049585397,
      "loss": 0.5175,
      "num_input_tokens_seen": 10481432,
      "step": 16060
    },
    {
      "epoch": 8.419811320754716,
      "grad_norm": 0.14863581955432892,
      "learning_rate": 0.0007176715500894515,
      "loss": 0.4244,
      "num_input_tokens_seen": 10483864,
      "step": 16065
    },
    {
      "epoch": 8.422431865828091,
      "grad_norm": 0.07972081005573273,
      "learning_rate": 0.0007174656496859584,
      "loss": 0.5176,
      "num_input_tokens_seen": 10487480,
      "step": 16070
    },
    {
      "epoch": 8.425052410901468,
      "grad_norm": 0.1308041661977768,
      "learning_rate": 0.0007172597037911323,
      "loss": 0.3553,
      "num_input_tokens_seen": 10490616,
      "step": 16075
    },
    {
      "epoch": 8.427672955974844,
      "grad_norm": 0.15330883860588074,
      "learning_rate": 0.0007170537124480548,
      "loss": 0.3708,
      "num_input_tokens_seen": 10493528,
      "step": 16080
    },
    {
      "epoch": 8.430293501048219,
      "grad_norm": 0.2190398871898651,
      "learning_rate": 0.0007168476756998169,
      "loss": 0.5056,
      "num_input_tokens_seen": 10496568,
      "step": 16085
    },
    {
      "epoch": 8.432914046121594,
      "grad_norm": 0.09563212841749191,
      "learning_rate": 0.0007166415935895194,
      "loss": 0.4647,
      "num_input_tokens_seen": 10499992,
      "step": 16090
    },
    {
      "epoch": 8.435534591194969,
      "grad_norm": 0.07294846326112747,
      "learning_rate": 0.0007164354661602718,
      "loss": 0.5007,
      "num_input_tokens_seen": 10502840,
      "step": 16095
    },
    {
      "epoch": 8.438155136268344,
      "grad_norm": 0.09168493002653122,
      "learning_rate": 0.000716229293455194,
      "loss": 0.4488,
      "num_input_tokens_seen": 10505752,
      "step": 16100
    },
    {
      "epoch": 8.44077568134172,
      "grad_norm": 0.09505162388086319,
      "learning_rate": 0.000716023075517415,
      "loss": 0.4627,
      "num_input_tokens_seen": 10508664,
      "step": 16105
    },
    {
      "epoch": 8.443396226415095,
      "grad_norm": 0.09875162690877914,
      "learning_rate": 0.000715816812390073,
      "loss": 0.506,
      "num_input_tokens_seen": 10511416,
      "step": 16110
    },
    {
      "epoch": 8.44601677148847,
      "grad_norm": 0.0814865231513977,
      "learning_rate": 0.0007156105041163161,
      "loss": 0.4333,
      "num_input_tokens_seen": 10513944,
      "step": 16115
    },
    {
      "epoch": 8.448637316561845,
      "grad_norm": 0.08958680182695389,
      "learning_rate": 0.0007154041507393015,
      "loss": 0.4324,
      "num_input_tokens_seen": 10517208,
      "step": 16120
    },
    {
      "epoch": 8.45125786163522,
      "grad_norm": 0.10321210324764252,
      "learning_rate": 0.000715197752302196,
      "loss": 0.4423,
      "num_input_tokens_seen": 10519704,
      "step": 16125
    },
    {
      "epoch": 8.453878406708595,
      "grad_norm": 0.06103438884019852,
      "learning_rate": 0.0007149913088481759,
      "loss": 0.5851,
      "num_input_tokens_seen": 10523160,
      "step": 16130
    },
    {
      "epoch": 8.45649895178197,
      "grad_norm": 0.1101788654923439,
      "learning_rate": 0.0007147848204204266,
      "loss": 0.4291,
      "num_input_tokens_seen": 10526168,
      "step": 16135
    },
    {
      "epoch": 8.459119496855346,
      "grad_norm": 0.09336010366678238,
      "learning_rate": 0.0007145782870621435,
      "loss": 0.3886,
      "num_input_tokens_seen": 10528824,
      "step": 16140
    },
    {
      "epoch": 8.46174004192872,
      "grad_norm": 0.07992511242628098,
      "learning_rate": 0.0007143717088165304,
      "loss": 0.4094,
      "num_input_tokens_seen": 10531640,
      "step": 16145
    },
    {
      "epoch": 8.464360587002096,
      "grad_norm": 0.13537807762622833,
      "learning_rate": 0.0007141650857268019,
      "loss": 0.3739,
      "num_input_tokens_seen": 10534808,
      "step": 16150
    },
    {
      "epoch": 8.466981132075471,
      "grad_norm": 0.119722381234169,
      "learning_rate": 0.0007139584178361807,
      "loss": 0.4562,
      "num_input_tokens_seen": 10537784,
      "step": 16155
    },
    {
      "epoch": 8.469601677148846,
      "grad_norm": 0.1331975907087326,
      "learning_rate": 0.0007137517051878994,
      "loss": 0.5133,
      "num_input_tokens_seen": 10540760,
      "step": 16160
    },
    {
      "epoch": 8.472222222222221,
      "grad_norm": 0.09458986669778824,
      "learning_rate": 0.0007135449478251998,
      "loss": 0.4194,
      "num_input_tokens_seen": 10543928,
      "step": 16165
    },
    {
      "epoch": 8.474842767295598,
      "grad_norm": 0.09110826253890991,
      "learning_rate": 0.0007133381457913336,
      "loss": 0.5887,
      "num_input_tokens_seen": 10546456,
      "step": 16170
    },
    {
      "epoch": 8.477463312368974,
      "grad_norm": 0.09692654013633728,
      "learning_rate": 0.0007131312991295611,
      "loss": 0.4573,
      "num_input_tokens_seen": 10549752,
      "step": 16175
    },
    {
      "epoch": 8.480083857442349,
      "grad_norm": 0.14150145649909973,
      "learning_rate": 0.0007129244078831525,
      "loss": 0.5054,
      "num_input_tokens_seen": 10553048,
      "step": 16180
    },
    {
      "epoch": 8.482704402515724,
      "grad_norm": 0.1725805401802063,
      "learning_rate": 0.0007127174720953866,
      "loss": 0.484,
      "num_input_tokens_seen": 10555768,
      "step": 16185
    },
    {
      "epoch": 8.485324947589099,
      "grad_norm": 0.12144071608781815,
      "learning_rate": 0.0007125104918095526,
      "loss": 0.5293,
      "num_input_tokens_seen": 10559448,
      "step": 16190
    },
    {
      "epoch": 8.487945492662474,
      "grad_norm": 0.09080374240875244,
      "learning_rate": 0.0007123034670689483,
      "loss": 0.6494,
      "num_input_tokens_seen": 10562008,
      "step": 16195
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 0.10940414667129517,
      "learning_rate": 0.0007120963979168804,
      "loss": 0.5054,
      "num_input_tokens_seen": 10564760,
      "step": 16200
    },
    {
      "epoch": 8.493186582809225,
      "grad_norm": 0.09282144159078598,
      "learning_rate": 0.000711889284396666,
      "loss": 0.5668,
      "num_input_tokens_seen": 10567512,
      "step": 16205
    },
    {
      "epoch": 8.4958071278826,
      "grad_norm": 0.12881849706172943,
      "learning_rate": 0.0007116821265516307,
      "loss": 0.3541,
      "num_input_tokens_seen": 10570104,
      "step": 16210
    },
    {
      "epoch": 8.498427672955975,
      "grad_norm": 0.08761733770370483,
      "learning_rate": 0.0007114749244251094,
      "loss": 0.5462,
      "num_input_tokens_seen": 10573304,
      "step": 16215
    },
    {
      "epoch": 8.50104821802935,
      "grad_norm": 0.11470959335565567,
      "learning_rate": 0.0007112676780604468,
      "loss": 0.3805,
      "num_input_tokens_seen": 10576408,
      "step": 16220
    },
    {
      "epoch": 8.503668763102725,
      "grad_norm": 0.0811779573559761,
      "learning_rate": 0.000711060387500996,
      "loss": 0.2928,
      "num_input_tokens_seen": 10579864,
      "step": 16225
    },
    {
      "epoch": 8.5062893081761,
      "grad_norm": 0.07312997430562973,
      "learning_rate": 0.0007108530527901199,
      "loss": 0.5064,
      "num_input_tokens_seen": 10582840,
      "step": 16230
    },
    {
      "epoch": 8.508909853249476,
      "grad_norm": 0.06066203489899635,
      "learning_rate": 0.000710645673971191,
      "loss": 0.4859,
      "num_input_tokens_seen": 10586392,
      "step": 16235
    },
    {
      "epoch": 8.51153039832285,
      "grad_norm": 0.0629240944981575,
      "learning_rate": 0.00071043825108759,
      "loss": 0.3797,
      "num_input_tokens_seen": 10591224,
      "step": 16240
    },
    {
      "epoch": 8.514150943396226,
      "grad_norm": 0.0920020043849945,
      "learning_rate": 0.0007102307841827079,
      "loss": 0.4673,
      "num_input_tokens_seen": 10594296,
      "step": 16245
    },
    {
      "epoch": 8.516771488469601,
      "grad_norm": 0.07529358565807343,
      "learning_rate": 0.0007100232732999443,
      "loss": 0.3882,
      "num_input_tokens_seen": 10597912,
      "step": 16250
    },
    {
      "epoch": 8.519392033542976,
      "grad_norm": 0.191665381193161,
      "learning_rate": 0.0007098157184827076,
      "loss": 0.4688,
      "num_input_tokens_seen": 10600632,
      "step": 16255
    },
    {
      "epoch": 8.522012578616351,
      "grad_norm": 0.08309086412191391,
      "learning_rate": 0.0007096081197744166,
      "loss": 0.4736,
      "num_input_tokens_seen": 10603672,
      "step": 16260
    },
    {
      "epoch": 8.524633123689728,
      "grad_norm": 0.08322712033987045,
      "learning_rate": 0.0007094004772184981,
      "loss": 0.3791,
      "num_input_tokens_seen": 10606648,
      "step": 16265
    },
    {
      "epoch": 8.527253668763104,
      "grad_norm": 0.0867699533700943,
      "learning_rate": 0.0007091927908583889,
      "loss": 0.4168,
      "num_input_tokens_seen": 10609432,
      "step": 16270
    },
    {
      "epoch": 8.529874213836479,
      "grad_norm": 0.10145829617977142,
      "learning_rate": 0.0007089850607375343,
      "loss": 0.4061,
      "num_input_tokens_seen": 10614424,
      "step": 16275
    },
    {
      "epoch": 8.532494758909854,
      "grad_norm": 0.13654647767543793,
      "learning_rate": 0.0007087772868993894,
      "loss": 0.5804,
      "num_input_tokens_seen": 10617784,
      "step": 16280
    },
    {
      "epoch": 8.535115303983229,
      "grad_norm": 0.1535218060016632,
      "learning_rate": 0.0007085694693874178,
      "loss": 0.4874,
      "num_input_tokens_seen": 10620216,
      "step": 16285
    },
    {
      "epoch": 8.537735849056604,
      "grad_norm": 0.07065701484680176,
      "learning_rate": 0.0007083616082450928,
      "loss": 0.4954,
      "num_input_tokens_seen": 10622808,
      "step": 16290
    },
    {
      "epoch": 8.54035639412998,
      "grad_norm": 0.062012262642383575,
      "learning_rate": 0.0007081537035158962,
      "loss": 0.4766,
      "num_input_tokens_seen": 10625912,
      "step": 16295
    },
    {
      "epoch": 8.542976939203355,
      "grad_norm": 0.10406924784183502,
      "learning_rate": 0.0007079457552433198,
      "loss": 0.4183,
      "num_input_tokens_seen": 10629016,
      "step": 16300
    },
    {
      "epoch": 8.54559748427673,
      "grad_norm": 0.08149073272943497,
      "learning_rate": 0.0007077377634708637,
      "loss": 0.594,
      "num_input_tokens_seen": 10632920,
      "step": 16305
    },
    {
      "epoch": 8.548218029350105,
      "grad_norm": 0.14024923741817474,
      "learning_rate": 0.0007075297282420375,
      "loss": 0.3736,
      "num_input_tokens_seen": 10635800,
      "step": 16310
    },
    {
      "epoch": 8.55083857442348,
      "grad_norm": 0.1228712946176529,
      "learning_rate": 0.0007073216496003598,
      "loss": 0.4201,
      "num_input_tokens_seen": 10638712,
      "step": 16315
    },
    {
      "epoch": 8.553459119496855,
      "grad_norm": 0.09308229386806488,
      "learning_rate": 0.0007071135275893584,
      "loss": 0.3644,
      "num_input_tokens_seen": 10641528,
      "step": 16320
    },
    {
      "epoch": 8.55607966457023,
      "grad_norm": 0.14767825603485107,
      "learning_rate": 0.0007069053622525696,
      "loss": 0.4117,
      "num_input_tokens_seen": 10644152,
      "step": 16325
    },
    {
      "epoch": 8.558700209643606,
      "grad_norm": 0.08013448119163513,
      "learning_rate": 0.0007066971536335395,
      "loss": 0.4439,
      "num_input_tokens_seen": 10647288,
      "step": 16330
    },
    {
      "epoch": 8.56132075471698,
      "grad_norm": 0.11402993649244308,
      "learning_rate": 0.0007064889017758234,
      "loss": 0.3941,
      "num_input_tokens_seen": 10650840,
      "step": 16335
    },
    {
      "epoch": 8.563941299790356,
      "grad_norm": 0.21921537816524506,
      "learning_rate": 0.0007062806067229845,
      "loss": 0.4103,
      "num_input_tokens_seen": 10654360,
      "step": 16340
    },
    {
      "epoch": 8.566561844863731,
      "grad_norm": 0.06991175562143326,
      "learning_rate": 0.0007060722685185961,
      "loss": 0.4901,
      "num_input_tokens_seen": 10659448,
      "step": 16345
    },
    {
      "epoch": 8.569182389937106,
      "grad_norm": 0.13645753264427185,
      "learning_rate": 0.00070586388720624,
      "loss": 0.4498,
      "num_input_tokens_seen": 10662520,
      "step": 16350
    },
    {
      "epoch": 8.571802935010481,
      "grad_norm": 0.21247294545173645,
      "learning_rate": 0.0007056554628295076,
      "loss": 0.4592,
      "num_input_tokens_seen": 10665176,
      "step": 16355
    },
    {
      "epoch": 8.574423480083858,
      "grad_norm": 0.09515949338674545,
      "learning_rate": 0.0007054469954319984,
      "loss": 0.4154,
      "num_input_tokens_seen": 10668984,
      "step": 16360
    },
    {
      "epoch": 8.577044025157234,
      "grad_norm": 0.05153496190905571,
      "learning_rate": 0.0007052384850573217,
      "loss": 0.3582,
      "num_input_tokens_seen": 10673112,
      "step": 16365
    },
    {
      "epoch": 8.579664570230609,
      "grad_norm": 0.16120192408561707,
      "learning_rate": 0.0007050299317490952,
      "loss": 0.3734,
      "num_input_tokens_seen": 10676504,
      "step": 16370
    },
    {
      "epoch": 8.582285115303984,
      "grad_norm": 0.054903604090213776,
      "learning_rate": 0.0007048213355509463,
      "loss": 0.2968,
      "num_input_tokens_seen": 10679480,
      "step": 16375
    },
    {
      "epoch": 8.584905660377359,
      "grad_norm": 0.2636769115924835,
      "learning_rate": 0.0007046126965065107,
      "loss": 0.4412,
      "num_input_tokens_seen": 10682360,
      "step": 16380
    },
    {
      "epoch": 8.587526205450734,
      "grad_norm": 0.07021120190620422,
      "learning_rate": 0.0007044040146594332,
      "loss": 0.4742,
      "num_input_tokens_seen": 10686232,
      "step": 16385
    },
    {
      "epoch": 8.59014675052411,
      "grad_norm": 0.09640637040138245,
      "learning_rate": 0.0007041952900533681,
      "loss": 0.3974,
      "num_input_tokens_seen": 10689592,
      "step": 16390
    },
    {
      "epoch": 8.592767295597485,
      "grad_norm": 0.09299764037132263,
      "learning_rate": 0.0007039865227319777,
      "loss": 0.4307,
      "num_input_tokens_seen": 10693272,
      "step": 16395
    },
    {
      "epoch": 8.59538784067086,
      "grad_norm": 0.10201088339090347,
      "learning_rate": 0.0007037777127389339,
      "loss": 0.3505,
      "num_input_tokens_seen": 10697560,
      "step": 16400
    },
    {
      "epoch": 8.598008385744235,
      "grad_norm": 0.07082455605268478,
      "learning_rate": 0.0007035688601179177,
      "loss": 0.6545,
      "num_input_tokens_seen": 10700664,
      "step": 16405
    },
    {
      "epoch": 8.60062893081761,
      "grad_norm": 0.1680172234773636,
      "learning_rate": 0.000703359964912618,
      "loss": 0.4669,
      "num_input_tokens_seen": 10704696,
      "step": 16410
    },
    {
      "epoch": 8.603249475890985,
      "grad_norm": 0.10611072182655334,
      "learning_rate": 0.000703151027166734,
      "loss": 0.527,
      "num_input_tokens_seen": 10707512,
      "step": 16415
    },
    {
      "epoch": 8.60587002096436,
      "grad_norm": 0.10092703253030777,
      "learning_rate": 0.0007029420469239726,
      "loss": 0.3472,
      "num_input_tokens_seen": 10711128,
      "step": 16420
    },
    {
      "epoch": 8.608490566037736,
      "grad_norm": 0.09713906049728394,
      "learning_rate": 0.0007027330242280503,
      "loss": 0.4928,
      "num_input_tokens_seen": 10714392,
      "step": 16425
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 0.14625006914138794,
      "learning_rate": 0.0007025239591226923,
      "loss": 0.5235,
      "num_input_tokens_seen": 10718008,
      "step": 16430
    },
    {
      "epoch": 8.613731656184486,
      "grad_norm": 0.06494001299142838,
      "learning_rate": 0.0007023148516516326,
      "loss": 0.4576,
      "num_input_tokens_seen": 10721240,
      "step": 16435
    },
    {
      "epoch": 8.616352201257861,
      "grad_norm": 0.08537974208593369,
      "learning_rate": 0.0007021057018586139,
      "loss": 0.5087,
      "num_input_tokens_seen": 10724312,
      "step": 16440
    },
    {
      "epoch": 8.618972746331236,
      "grad_norm": 0.17358919978141785,
      "learning_rate": 0.0007018965097873883,
      "loss": 0.4026,
      "num_input_tokens_seen": 10726872,
      "step": 16445
    },
    {
      "epoch": 8.621593291404611,
      "grad_norm": 0.09219559282064438,
      "learning_rate": 0.0007016872754817161,
      "loss": 0.4907,
      "num_input_tokens_seen": 10730808,
      "step": 16450
    },
    {
      "epoch": 8.624213836477988,
      "grad_norm": 0.08403996378183365,
      "learning_rate": 0.0007014779989853668,
      "loss": 0.4058,
      "num_input_tokens_seen": 10733848,
      "step": 16455
    },
    {
      "epoch": 8.626834381551364,
      "grad_norm": 0.0742608904838562,
      "learning_rate": 0.0007012686803421189,
      "loss": 0.3779,
      "num_input_tokens_seen": 10736952,
      "step": 16460
    },
    {
      "epoch": 8.629454926624739,
      "grad_norm": 0.06768450886011124,
      "learning_rate": 0.000701059319595759,
      "loss": 0.3781,
      "num_input_tokens_seen": 10740024,
      "step": 16465
    },
    {
      "epoch": 8.632075471698114,
      "grad_norm": 0.07557128369808197,
      "learning_rate": 0.0007008499167900833,
      "loss": 0.4825,
      "num_input_tokens_seen": 10742680,
      "step": 16470
    },
    {
      "epoch": 8.634696016771489,
      "grad_norm": 0.07107485085725784,
      "learning_rate": 0.0007006404719688966,
      "loss": 0.4356,
      "num_input_tokens_seen": 10746904,
      "step": 16475
    },
    {
      "epoch": 8.637316561844864,
      "grad_norm": 0.11723490804433823,
      "learning_rate": 0.000700430985176012,
      "loss": 0.4933,
      "num_input_tokens_seen": 10749656,
      "step": 16480
    },
    {
      "epoch": 8.63993710691824,
      "grad_norm": 0.12097787857055664,
      "learning_rate": 0.0007002214564552521,
      "loss": 0.6252,
      "num_input_tokens_seen": 10752280,
      "step": 16485
    },
    {
      "epoch": 8.642557651991615,
      "grad_norm": 0.0732109546661377,
      "learning_rate": 0.0007000118858504476,
      "loss": 0.4284,
      "num_input_tokens_seen": 10755672,
      "step": 16490
    },
    {
      "epoch": 8.64517819706499,
      "grad_norm": 0.07445777207612991,
      "learning_rate": 0.0006998022734054386,
      "loss": 0.4198,
      "num_input_tokens_seen": 10759256,
      "step": 16495
    },
    {
      "epoch": 8.647798742138365,
      "grad_norm": 0.15613868832588196,
      "learning_rate": 0.0006995926191640734,
      "loss": 0.5376,
      "num_input_tokens_seen": 10762840,
      "step": 16500
    },
    {
      "epoch": 8.65041928721174,
      "grad_norm": 0.0941404178738594,
      "learning_rate": 0.0006993829231702092,
      "loss": 0.5098,
      "num_input_tokens_seen": 10765592,
      "step": 16505
    },
    {
      "epoch": 8.653039832285115,
      "grad_norm": 0.13987086713314056,
      "learning_rate": 0.0006991731854677124,
      "loss": 0.3855,
      "num_input_tokens_seen": 10767800,
      "step": 16510
    },
    {
      "epoch": 8.65566037735849,
      "grad_norm": 0.09267036616802216,
      "learning_rate": 0.0006989634061004572,
      "loss": 0.477,
      "num_input_tokens_seen": 10775384,
      "step": 16515
    },
    {
      "epoch": 8.658280922431866,
      "grad_norm": 0.09312867373228073,
      "learning_rate": 0.0006987535851123276,
      "loss": 0.514,
      "num_input_tokens_seen": 10777752,
      "step": 16520
    },
    {
      "epoch": 8.66090146750524,
      "grad_norm": 0.07852691411972046,
      "learning_rate": 0.0006985437225472155,
      "loss": 0.2948,
      "num_input_tokens_seen": 10780376,
      "step": 16525
    },
    {
      "epoch": 8.663522012578616,
      "grad_norm": 0.12457644194364548,
      "learning_rate": 0.0006983338184490215,
      "loss": 0.442,
      "num_input_tokens_seen": 10784344,
      "step": 16530
    },
    {
      "epoch": 8.666142557651991,
      "grad_norm": 0.11264453828334808,
      "learning_rate": 0.0006981238728616557,
      "loss": 0.5352,
      "num_input_tokens_seen": 10787128,
      "step": 16535
    },
    {
      "epoch": 8.668763102725366,
      "grad_norm": 0.08854305744171143,
      "learning_rate": 0.0006979138858290358,
      "loss": 0.4184,
      "num_input_tokens_seen": 10790296,
      "step": 16540
    },
    {
      "epoch": 8.671383647798741,
      "grad_norm": 0.1251477301120758,
      "learning_rate": 0.0006977038573950889,
      "loss": 0.5034,
      "num_input_tokens_seen": 10793144,
      "step": 16545
    },
    {
      "epoch": 8.674004192872118,
      "grad_norm": 0.1403721570968628,
      "learning_rate": 0.0006974937876037508,
      "loss": 0.4708,
      "num_input_tokens_seen": 10796280,
      "step": 16550
    },
    {
      "epoch": 8.676624737945493,
      "grad_norm": 0.06809487193822861,
      "learning_rate": 0.0006972836764989652,
      "loss": 0.437,
      "num_input_tokens_seen": 10799000,
      "step": 16555
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 0.08049526065587997,
      "learning_rate": 0.0006970735241246853,
      "loss": 0.5282,
      "num_input_tokens_seen": 10802264,
      "step": 16560
    },
    {
      "epoch": 8.681865828092244,
      "grad_norm": 0.08272337913513184,
      "learning_rate": 0.0006968633305248724,
      "loss": 0.4274,
      "num_input_tokens_seen": 10805656,
      "step": 16565
    },
    {
      "epoch": 8.684486373165619,
      "grad_norm": 0.07470263540744781,
      "learning_rate": 0.0006966530957434968,
      "loss": 0.4322,
      "num_input_tokens_seen": 10809496,
      "step": 16570
    },
    {
      "epoch": 8.687106918238994,
      "grad_norm": 0.09958261996507645,
      "learning_rate": 0.0006964428198245372,
      "loss": 0.4919,
      "num_input_tokens_seen": 10812856,
      "step": 16575
    },
    {
      "epoch": 8.68972746331237,
      "grad_norm": 0.12061523646116257,
      "learning_rate": 0.0006962325028119806,
      "loss": 0.4442,
      "num_input_tokens_seen": 10815544,
      "step": 16580
    },
    {
      "epoch": 8.692348008385745,
      "grad_norm": 0.09840991348028183,
      "learning_rate": 0.0006960221447498232,
      "loss": 0.5021,
      "num_input_tokens_seen": 10818072,
      "step": 16585
    },
    {
      "epoch": 8.69496855345912,
      "grad_norm": 0.08964727818965912,
      "learning_rate": 0.0006958117456820696,
      "loss": 0.4146,
      "num_input_tokens_seen": 10821304,
      "step": 16590
    },
    {
      "epoch": 8.697589098532495,
      "grad_norm": 0.04855254665017128,
      "learning_rate": 0.0006956013056527326,
      "loss": 0.3609,
      "num_input_tokens_seen": 10824984,
      "step": 16595
    },
    {
      "epoch": 8.70020964360587,
      "grad_norm": 0.08926066756248474,
      "learning_rate": 0.0006953908247058341,
      "loss": 0.5898,
      "num_input_tokens_seen": 10828024,
      "step": 16600
    },
    {
      "epoch": 8.702830188679245,
      "grad_norm": 0.16916747391223907,
      "learning_rate": 0.0006951803028854041,
      "loss": 0.477,
      "num_input_tokens_seen": 10831640,
      "step": 16605
    },
    {
      "epoch": 8.70545073375262,
      "grad_norm": 0.05946515128016472,
      "learning_rate": 0.0006949697402354816,
      "loss": 0.5095,
      "num_input_tokens_seen": 10835160,
      "step": 16610
    },
    {
      "epoch": 8.708071278825996,
      "grad_norm": 0.1400000900030136,
      "learning_rate": 0.0006947591368001138,
      "loss": 0.3709,
      "num_input_tokens_seen": 10838232,
      "step": 16615
    },
    {
      "epoch": 8.71069182389937,
      "grad_norm": 0.09907994419336319,
      "learning_rate": 0.0006945484926233563,
      "loss": 0.3412,
      "num_input_tokens_seen": 10841240,
      "step": 16620
    },
    {
      "epoch": 8.713312368972746,
      "grad_norm": 0.09930083155632019,
      "learning_rate": 0.0006943378077492737,
      "loss": 0.4114,
      "num_input_tokens_seen": 10844312,
      "step": 16625
    },
    {
      "epoch": 8.715932914046121,
      "grad_norm": 0.08074131608009338,
      "learning_rate": 0.000694127082221939,
      "loss": 0.4007,
      "num_input_tokens_seen": 10847320,
      "step": 16630
    },
    {
      "epoch": 8.718553459119496,
      "grad_norm": 0.07595715671777725,
      "learning_rate": 0.0006939163160854334,
      "loss": 0.397,
      "num_input_tokens_seen": 10850648,
      "step": 16635
    },
    {
      "epoch": 8.721174004192871,
      "grad_norm": 0.07868009060621262,
      "learning_rate": 0.0006937055093838467,
      "loss": 0.465,
      "num_input_tokens_seen": 10853912,
      "step": 16640
    },
    {
      "epoch": 8.723794549266248,
      "grad_norm": 0.09835334122180939,
      "learning_rate": 0.0006934946621612774,
      "loss": 0.4048,
      "num_input_tokens_seen": 10856824,
      "step": 16645
    },
    {
      "epoch": 8.726415094339622,
      "grad_norm": 0.07952046394348145,
      "learning_rate": 0.0006932837744618322,
      "loss": 0.3943,
      "num_input_tokens_seen": 10860280,
      "step": 16650
    },
    {
      "epoch": 8.729035639412999,
      "grad_norm": 0.06606098264455795,
      "learning_rate": 0.0006930728463296265,
      "loss": 0.5311,
      "num_input_tokens_seen": 10863832,
      "step": 16655
    },
    {
      "epoch": 8.731656184486374,
      "grad_norm": 0.0942741110920906,
      "learning_rate": 0.0006928618778087842,
      "loss": 0.436,
      "num_input_tokens_seen": 10867576,
      "step": 16660
    },
    {
      "epoch": 8.734276729559749,
      "grad_norm": 0.07514185458421707,
      "learning_rate": 0.0006926508689434371,
      "loss": 0.4637,
      "num_input_tokens_seen": 10871384,
      "step": 16665
    },
    {
      "epoch": 8.736897274633124,
      "grad_norm": 0.11072710901498795,
      "learning_rate": 0.0006924398197777263,
      "loss": 0.5845,
      "num_input_tokens_seen": 10874584,
      "step": 16670
    },
    {
      "epoch": 8.7395178197065,
      "grad_norm": 0.07131238281726837,
      "learning_rate": 0.0006922287303558006,
      "loss": 0.411,
      "num_input_tokens_seen": 10878584,
      "step": 16675
    },
    {
      "epoch": 8.742138364779874,
      "grad_norm": 0.1163015067577362,
      "learning_rate": 0.0006920176007218175,
      "loss": 0.6605,
      "num_input_tokens_seen": 10881560,
      "step": 16680
    },
    {
      "epoch": 8.74475890985325,
      "grad_norm": 0.10277805477380753,
      "learning_rate": 0.000691806430919943,
      "loss": 0.5556,
      "num_input_tokens_seen": 10884600,
      "step": 16685
    },
    {
      "epoch": 8.747379454926625,
      "grad_norm": 0.06798457354307175,
      "learning_rate": 0.0006915952209943514,
      "loss": 0.355,
      "num_input_tokens_seen": 10888408,
      "step": 16690
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.07503052800893784,
      "learning_rate": 0.0006913839709892256,
      "loss": 0.5026,
      "num_input_tokens_seen": 10891736,
      "step": 16695
    },
    {
      "epoch": 8.752620545073375,
      "grad_norm": 0.09240412712097168,
      "learning_rate": 0.0006911726809487561,
      "loss": 0.3623,
      "num_input_tokens_seen": 10894648,
      "step": 16700
    },
    {
      "epoch": 8.75524109014675,
      "grad_norm": 0.10399191826581955,
      "learning_rate": 0.0006909613509171431,
      "loss": 0.3669,
      "num_input_tokens_seen": 10897624,
      "step": 16705
    },
    {
      "epoch": 8.757861635220126,
      "grad_norm": 0.09215571731328964,
      "learning_rate": 0.0006907499809385941,
      "loss": 0.4929,
      "num_input_tokens_seen": 10900696,
      "step": 16710
    },
    {
      "epoch": 8.7604821802935,
      "grad_norm": 0.10416636615991592,
      "learning_rate": 0.0006905385710573252,
      "loss": 0.5584,
      "num_input_tokens_seen": 10904280,
      "step": 16715
    },
    {
      "epoch": 8.763102725366876,
      "grad_norm": 0.09575121104717255,
      "learning_rate": 0.000690327121317561,
      "loss": 0.4943,
      "num_input_tokens_seen": 10907896,
      "step": 16720
    },
    {
      "epoch": 8.765723270440251,
      "grad_norm": 0.11113540828227997,
      "learning_rate": 0.0006901156317635342,
      "loss": 0.4179,
      "num_input_tokens_seen": 10911192,
      "step": 16725
    },
    {
      "epoch": 8.768343815513626,
      "grad_norm": 0.11151682585477829,
      "learning_rate": 0.0006899041024394864,
      "loss": 0.5898,
      "num_input_tokens_seen": 10914488,
      "step": 16730
    },
    {
      "epoch": 8.770964360587001,
      "grad_norm": 0.12220805138349533,
      "learning_rate": 0.000689692533389667,
      "loss": 0.4063,
      "num_input_tokens_seen": 10918456,
      "step": 16735
    },
    {
      "epoch": 8.773584905660378,
      "grad_norm": 0.07799280434846878,
      "learning_rate": 0.0006894809246583334,
      "loss": 0.4993,
      "num_input_tokens_seen": 10922296,
      "step": 16740
    },
    {
      "epoch": 8.776205450733752,
      "grad_norm": 0.06796310842037201,
      "learning_rate": 0.0006892692762897523,
      "loss": 0.5057,
      "num_input_tokens_seen": 10925400,
      "step": 16745
    },
    {
      "epoch": 8.778825995807129,
      "grad_norm": 0.09469863772392273,
      "learning_rate": 0.000689057588328198,
      "loss": 0.4409,
      "num_input_tokens_seen": 10928120,
      "step": 16750
    },
    {
      "epoch": 8.781446540880504,
      "grad_norm": 0.14847177267074585,
      "learning_rate": 0.0006888458608179528,
      "loss": 0.4844,
      "num_input_tokens_seen": 10932088,
      "step": 16755
    },
    {
      "epoch": 8.784067085953879,
      "grad_norm": 0.08098236471414566,
      "learning_rate": 0.0006886340938033085,
      "loss": 0.4878,
      "num_input_tokens_seen": 10935832,
      "step": 16760
    },
    {
      "epoch": 8.786687631027254,
      "grad_norm": 0.16893135011196136,
      "learning_rate": 0.0006884222873285632,
      "loss": 0.5775,
      "num_input_tokens_seen": 10938552,
      "step": 16765
    },
    {
      "epoch": 8.78930817610063,
      "grad_norm": 0.11986606568098068,
      "learning_rate": 0.0006882104414380254,
      "loss": 0.4456,
      "num_input_tokens_seen": 10942520,
      "step": 16770
    },
    {
      "epoch": 8.791928721174004,
      "grad_norm": 0.12302296608686447,
      "learning_rate": 0.0006879985561760105,
      "loss": 0.4608,
      "num_input_tokens_seen": 10946008,
      "step": 16775
    },
    {
      "epoch": 8.79454926624738,
      "grad_norm": 0.1269468069076538,
      "learning_rate": 0.0006877866315868422,
      "loss": 0.4634,
      "num_input_tokens_seen": 10949144,
      "step": 16780
    },
    {
      "epoch": 8.797169811320755,
      "grad_norm": 0.14023855328559875,
      "learning_rate": 0.0006875746677148531,
      "loss": 0.441,
      "num_input_tokens_seen": 10953400,
      "step": 16785
    },
    {
      "epoch": 8.79979035639413,
      "grad_norm": 0.09559710323810577,
      "learning_rate": 0.0006873626646043835,
      "loss": 0.4391,
      "num_input_tokens_seen": 10956472,
      "step": 16790
    },
    {
      "epoch": 8.802410901467505,
      "grad_norm": 0.09512533992528915,
      "learning_rate": 0.000687150622299782,
      "loss": 0.4621,
      "num_input_tokens_seen": 10960312,
      "step": 16795
    },
    {
      "epoch": 8.80503144654088,
      "grad_norm": 0.10108867287635803,
      "learning_rate": 0.0006869385408454056,
      "loss": 0.4399,
      "num_input_tokens_seen": 10962616,
      "step": 16800
    },
    {
      "epoch": 8.807651991614255,
      "grad_norm": 0.09283113479614258,
      "learning_rate": 0.0006867264202856188,
      "loss": 0.4287,
      "num_input_tokens_seen": 10966008,
      "step": 16805
    },
    {
      "epoch": 8.81027253668763,
      "grad_norm": 0.12326722592115402,
      "learning_rate": 0.0006865142606647954,
      "loss": 0.5532,
      "num_input_tokens_seen": 10970040,
      "step": 16810
    },
    {
      "epoch": 8.812893081761006,
      "grad_norm": 0.07661133259534836,
      "learning_rate": 0.0006863020620273166,
      "loss": 0.4824,
      "num_input_tokens_seen": 10972472,
      "step": 16815
    },
    {
      "epoch": 8.815513626834381,
      "grad_norm": 0.08192399144172668,
      "learning_rate": 0.0006860898244175716,
      "loss": 0.4036,
      "num_input_tokens_seen": 10976664,
      "step": 16820
    },
    {
      "epoch": 8.818134171907756,
      "grad_norm": 0.0715959370136261,
      "learning_rate": 0.0006858775478799586,
      "loss": 0.4299,
      "num_input_tokens_seen": 10979192,
      "step": 16825
    },
    {
      "epoch": 8.820754716981131,
      "grad_norm": 0.1425308734178543,
      "learning_rate": 0.0006856652324588831,
      "loss": 0.4022,
      "num_input_tokens_seen": 10981688,
      "step": 16830
    },
    {
      "epoch": 8.823375262054507,
      "grad_norm": 0.10178732872009277,
      "learning_rate": 0.000685452878198759,
      "loss": 0.4679,
      "num_input_tokens_seen": 10985176,
      "step": 16835
    },
    {
      "epoch": 8.825995807127882,
      "grad_norm": 0.06724434345960617,
      "learning_rate": 0.0006852404851440088,
      "loss": 0.4575,
      "num_input_tokens_seen": 10988536,
      "step": 16840
    },
    {
      "epoch": 8.828616352201259,
      "grad_norm": 0.15780630707740784,
      "learning_rate": 0.0006850280533390624,
      "loss": 0.3547,
      "num_input_tokens_seen": 10990936,
      "step": 16845
    },
    {
      "epoch": 8.831236897274634,
      "grad_norm": 0.09743385016918182,
      "learning_rate": 0.0006848155828283581,
      "loss": 0.4231,
      "num_input_tokens_seen": 10994200,
      "step": 16850
    },
    {
      "epoch": 8.833857442348009,
      "grad_norm": 0.08276718109846115,
      "learning_rate": 0.0006846030736563422,
      "loss": 0.4773,
      "num_input_tokens_seen": 10997240,
      "step": 16855
    },
    {
      "epoch": 8.836477987421384,
      "grad_norm": 0.07225465029478073,
      "learning_rate": 0.0006843905258674696,
      "loss": 0.3915,
      "num_input_tokens_seen": 11000568,
      "step": 16860
    },
    {
      "epoch": 8.83909853249476,
      "grad_norm": 0.06887646019458771,
      "learning_rate": 0.0006841779395062026,
      "loss": 0.351,
      "num_input_tokens_seen": 11005528,
      "step": 16865
    },
    {
      "epoch": 8.841719077568134,
      "grad_norm": 0.11553201824426651,
      "learning_rate": 0.0006839653146170116,
      "loss": 0.4933,
      "num_input_tokens_seen": 11009176,
      "step": 16870
    },
    {
      "epoch": 8.84433962264151,
      "grad_norm": 0.07750824093818665,
      "learning_rate": 0.0006837526512443758,
      "loss": 0.3892,
      "num_input_tokens_seen": 11012504,
      "step": 16875
    },
    {
      "epoch": 8.846960167714885,
      "grad_norm": 0.09655849635601044,
      "learning_rate": 0.0006835399494327818,
      "loss": 0.4692,
      "num_input_tokens_seen": 11015992,
      "step": 16880
    },
    {
      "epoch": 8.84958071278826,
      "grad_norm": 0.1069156602025032,
      "learning_rate": 0.0006833272092267241,
      "loss": 0.351,
      "num_input_tokens_seen": 11019928,
      "step": 16885
    },
    {
      "epoch": 8.852201257861635,
      "grad_norm": 0.07561156153678894,
      "learning_rate": 0.000683114430670706,
      "loss": 0.5283,
      "num_input_tokens_seen": 11023096,
      "step": 16890
    },
    {
      "epoch": 8.85482180293501,
      "grad_norm": 0.12881851196289062,
      "learning_rate": 0.0006829016138092378,
      "loss": 0.4906,
      "num_input_tokens_seen": 11026232,
      "step": 16895
    },
    {
      "epoch": 8.857442348008385,
      "grad_norm": 0.17399390041828156,
      "learning_rate": 0.0006826887586868388,
      "loss": 0.6377,
      "num_input_tokens_seen": 11029560,
      "step": 16900
    },
    {
      "epoch": 8.86006289308176,
      "grad_norm": 0.09825675189495087,
      "learning_rate": 0.0006824758653480356,
      "loss": 0.3963,
      "num_input_tokens_seen": 11032120,
      "step": 16905
    },
    {
      "epoch": 8.862683438155136,
      "grad_norm": 0.09616296738386154,
      "learning_rate": 0.0006822629338373632,
      "loss": 0.4663,
      "num_input_tokens_seen": 11034968,
      "step": 16910
    },
    {
      "epoch": 8.865303983228511,
      "grad_norm": 0.13862265646457672,
      "learning_rate": 0.0006820499641993644,
      "loss": 0.3137,
      "num_input_tokens_seen": 11038456,
      "step": 16915
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 0.11513625085353851,
      "learning_rate": 0.0006818369564785902,
      "loss": 0.4119,
      "num_input_tokens_seen": 11042712,
      "step": 16920
    },
    {
      "epoch": 8.870545073375261,
      "grad_norm": 0.11478215456008911,
      "learning_rate": 0.0006816239107195989,
      "loss": 0.5329,
      "num_input_tokens_seen": 11046008,
      "step": 16925
    },
    {
      "epoch": 8.873165618448636,
      "grad_norm": 0.1162056252360344,
      "learning_rate": 0.000681410826966958,
      "loss": 0.561,
      "num_input_tokens_seen": 11049336,
      "step": 16930
    },
    {
      "epoch": 8.875786163522012,
      "grad_norm": 0.07686225324869156,
      "learning_rate": 0.0006811977052652414,
      "loss": 0.4014,
      "num_input_tokens_seen": 11053272,
      "step": 16935
    },
    {
      "epoch": 8.878406708595389,
      "grad_norm": 0.11214864999055862,
      "learning_rate": 0.000680984545659032,
      "loss": 0.5075,
      "num_input_tokens_seen": 11057848,
      "step": 16940
    },
    {
      "epoch": 8.881027253668764,
      "grad_norm": 0.11429488658905029,
      "learning_rate": 0.0006807713481929207,
      "loss": 0.4637,
      "num_input_tokens_seen": 11060216,
      "step": 16945
    },
    {
      "epoch": 8.883647798742139,
      "grad_norm": 0.0808306410908699,
      "learning_rate": 0.0006805581129115055,
      "loss": 0.4613,
      "num_input_tokens_seen": 11063640,
      "step": 16950
    },
    {
      "epoch": 8.886268343815514,
      "grad_norm": 0.09210952371358871,
      "learning_rate": 0.0006803448398593931,
      "loss": 0.3782,
      "num_input_tokens_seen": 11066872,
      "step": 16955
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 0.11283509433269501,
      "learning_rate": 0.0006801315290811976,
      "loss": 0.4781,
      "num_input_tokens_seen": 11069624,
      "step": 16960
    },
    {
      "epoch": 8.891509433962264,
      "grad_norm": 0.08614158630371094,
      "learning_rate": 0.0006799181806215413,
      "loss": 0.4362,
      "num_input_tokens_seen": 11073048,
      "step": 16965
    },
    {
      "epoch": 8.89412997903564,
      "grad_norm": 0.07923497259616852,
      "learning_rate": 0.0006797047945250543,
      "loss": 0.3909,
      "num_input_tokens_seen": 11076312,
      "step": 16970
    },
    {
      "epoch": 8.896750524109015,
      "grad_norm": 0.0908903107047081,
      "learning_rate": 0.0006794913708363745,
      "loss": 0.531,
      "num_input_tokens_seen": 11079416,
      "step": 16975
    },
    {
      "epoch": 8.89937106918239,
      "grad_norm": 0.0815720185637474,
      "learning_rate": 0.0006792779096001475,
      "loss": 0.4879,
      "num_input_tokens_seen": 11082168,
      "step": 16980
    },
    {
      "epoch": 8.901991614255765,
      "grad_norm": 0.11498814076185226,
      "learning_rate": 0.0006790644108610273,
      "loss": 0.3524,
      "num_input_tokens_seen": 11084568,
      "step": 16985
    },
    {
      "epoch": 8.90461215932914,
      "grad_norm": 0.06989135593175888,
      "learning_rate": 0.0006788508746636751,
      "loss": 0.4118,
      "num_input_tokens_seen": 11087736,
      "step": 16990
    },
    {
      "epoch": 8.907232704402515,
      "grad_norm": 0.07608886063098907,
      "learning_rate": 0.0006786373010527605,
      "loss": 0.4507,
      "num_input_tokens_seen": 11090648,
      "step": 16995
    },
    {
      "epoch": 8.90985324947589,
      "grad_norm": 0.08645884692668915,
      "learning_rate": 0.0006784236900729603,
      "loss": 0.6216,
      "num_input_tokens_seen": 11093240,
      "step": 17000
    },
    {
      "epoch": 8.912473794549266,
      "grad_norm": 0.09473395347595215,
      "learning_rate": 0.0006782100417689599,
      "loss": 0.5673,
      "num_input_tokens_seen": 11096568,
      "step": 17005
    },
    {
      "epoch": 8.915094339622641,
      "grad_norm": 0.060873787850141525,
      "learning_rate": 0.0006779963561854517,
      "loss": 0.4246,
      "num_input_tokens_seen": 11099384,
      "step": 17010
    },
    {
      "epoch": 8.917714884696016,
      "grad_norm": 0.16185715794563293,
      "learning_rate": 0.0006777826333671367,
      "loss": 0.2988,
      "num_input_tokens_seen": 11101944,
      "step": 17015
    },
    {
      "epoch": 8.920335429769391,
      "grad_norm": 0.09565986692905426,
      "learning_rate": 0.0006775688733587227,
      "loss": 0.6237,
      "num_input_tokens_seen": 11104856,
      "step": 17020
    },
    {
      "epoch": 8.922955974842766,
      "grad_norm": 0.09823106974363327,
      "learning_rate": 0.0006773550762049265,
      "loss": 0.4562,
      "num_input_tokens_seen": 11108440,
      "step": 17025
    },
    {
      "epoch": 8.925576519916142,
      "grad_norm": 0.09804274886846542,
      "learning_rate": 0.0006771412419504716,
      "loss": 0.4157,
      "num_input_tokens_seen": 11112504,
      "step": 17030
    },
    {
      "epoch": 8.928197064989519,
      "grad_norm": 0.1358112394809723,
      "learning_rate": 0.00067692737064009,
      "loss": 0.5252,
      "num_input_tokens_seen": 11116024,
      "step": 17035
    },
    {
      "epoch": 8.930817610062894,
      "grad_norm": 0.1139194667339325,
      "learning_rate": 0.0006767134623185208,
      "loss": 0.4398,
      "num_input_tokens_seen": 11119128,
      "step": 17040
    },
    {
      "epoch": 8.933438155136269,
      "grad_norm": 0.09204061329364777,
      "learning_rate": 0.0006764995170305114,
      "loss": 0.547,
      "num_input_tokens_seen": 11122584,
      "step": 17045
    },
    {
      "epoch": 8.936058700209644,
      "grad_norm": 0.06669943779706955,
      "learning_rate": 0.0006762855348208166,
      "loss": 0.4577,
      "num_input_tokens_seen": 11125592,
      "step": 17050
    },
    {
      "epoch": 8.93867924528302,
      "grad_norm": 0.0827162116765976,
      "learning_rate": 0.0006760715157341991,
      "loss": 0.4626,
      "num_input_tokens_seen": 11128312,
      "step": 17055
    },
    {
      "epoch": 8.941299790356394,
      "grad_norm": 0.07396985590457916,
      "learning_rate": 0.0006758574598154292,
      "loss": 0.3858,
      "num_input_tokens_seen": 11132120,
      "step": 17060
    },
    {
      "epoch": 8.94392033542977,
      "grad_norm": 0.09983914345502853,
      "learning_rate": 0.0006756433671092852,
      "loss": 0.3501,
      "num_input_tokens_seen": 11135768,
      "step": 17065
    },
    {
      "epoch": 8.946540880503145,
      "grad_norm": 0.09078945964574814,
      "learning_rate": 0.0006754292376605524,
      "loss": 0.4352,
      "num_input_tokens_seen": 11139128,
      "step": 17070
    },
    {
      "epoch": 8.94916142557652,
      "grad_norm": 0.07483495026826859,
      "learning_rate": 0.0006752150715140247,
      "loss": 0.6466,
      "num_input_tokens_seen": 11142360,
      "step": 17075
    },
    {
      "epoch": 8.951781970649895,
      "grad_norm": 0.08225146681070328,
      "learning_rate": 0.0006750008687145028,
      "loss": 0.3752,
      "num_input_tokens_seen": 11144984,
      "step": 17080
    },
    {
      "epoch": 8.95440251572327,
      "grad_norm": 0.13013799488544464,
      "learning_rate": 0.0006747866293067958,
      "loss": 0.5026,
      "num_input_tokens_seen": 11147672,
      "step": 17085
    },
    {
      "epoch": 8.957023060796645,
      "grad_norm": 0.08838722854852676,
      "learning_rate": 0.00067457235333572,
      "loss": 0.5141,
      "num_input_tokens_seen": 11150872,
      "step": 17090
    },
    {
      "epoch": 8.95964360587002,
      "grad_norm": 0.0723438635468483,
      "learning_rate": 0.0006743580408460994,
      "loss": 0.4732,
      "num_input_tokens_seen": 11154360,
      "step": 17095
    },
    {
      "epoch": 8.962264150943396,
      "grad_norm": 0.11445499956607819,
      "learning_rate": 0.0006741436918827659,
      "loss": 0.5063,
      "num_input_tokens_seen": 11156984,
      "step": 17100
    },
    {
      "epoch": 8.964884696016771,
      "grad_norm": 0.10443729907274246,
      "learning_rate": 0.0006739293064905589,
      "loss": 0.6986,
      "num_input_tokens_seen": 11160696,
      "step": 17105
    },
    {
      "epoch": 8.967505241090146,
      "grad_norm": 0.07213644683361053,
      "learning_rate": 0.0006737148847143251,
      "loss": 0.5157,
      "num_input_tokens_seen": 11163800,
      "step": 17110
    },
    {
      "epoch": 8.970125786163521,
      "grad_norm": 0.09911330789327621,
      "learning_rate": 0.0006735004265989195,
      "loss": 0.5476,
      "num_input_tokens_seen": 11167480,
      "step": 17115
    },
    {
      "epoch": 8.972746331236896,
      "grad_norm": 0.12304742634296417,
      "learning_rate": 0.0006732859321892038,
      "loss": 0.4827,
      "num_input_tokens_seen": 11170232,
      "step": 17120
    },
    {
      "epoch": 8.975366876310272,
      "grad_norm": 0.10596770793199539,
      "learning_rate": 0.0006730714015300481,
      "loss": 0.6774,
      "num_input_tokens_seen": 11173400,
      "step": 17125
    },
    {
      "epoch": 8.977987421383649,
      "grad_norm": 0.11250264197587967,
      "learning_rate": 0.0006728568346663299,
      "loss": 0.45,
      "num_input_tokens_seen": 11176472,
      "step": 17130
    },
    {
      "epoch": 8.980607966457024,
      "grad_norm": 0.09188196808099747,
      "learning_rate": 0.0006726422316429337,
      "loss": 0.382,
      "num_input_tokens_seen": 11179288,
      "step": 17135
    },
    {
      "epoch": 8.983228511530399,
      "grad_norm": 0.12892979383468628,
      "learning_rate": 0.0006724275925047523,
      "loss": 0.4734,
      "num_input_tokens_seen": 11182136,
      "step": 17140
    },
    {
      "epoch": 8.985849056603774,
      "grad_norm": 0.08187852054834366,
      "learning_rate": 0.0006722129172966858,
      "loss": 0.4005,
      "num_input_tokens_seen": 11186296,
      "step": 17145
    },
    {
      "epoch": 8.98846960167715,
      "grad_norm": 0.0903630331158638,
      "learning_rate": 0.0006719982060636416,
      "loss": 0.3462,
      "num_input_tokens_seen": 11189304,
      "step": 17150
    },
    {
      "epoch": 8.991090146750524,
      "grad_norm": 0.11897910386323929,
      "learning_rate": 0.000671783458850535,
      "loss": 0.5255,
      "num_input_tokens_seen": 11192024,
      "step": 17155
    },
    {
      "epoch": 8.9937106918239,
      "grad_norm": 0.08859651535749435,
      "learning_rate": 0.0006715686757022886,
      "loss": 0.525,
      "num_input_tokens_seen": 11195160,
      "step": 17160
    },
    {
      "epoch": 8.996331236897275,
      "grad_norm": 0.12159152328968048,
      "learning_rate": 0.0006713538566638326,
      "loss": 0.5113,
      "num_input_tokens_seen": 11198264,
      "step": 17165
    },
    {
      "epoch": 8.99895178197065,
      "grad_norm": 0.054412100464105606,
      "learning_rate": 0.0006711390017801049,
      "loss": 0.4566,
      "num_input_tokens_seen": 11202840,
      "step": 17170
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.45358505845069885,
      "eval_runtime": 13.5148,
      "eval_samples_per_second": 62.746,
      "eval_steps_per_second": 15.687,
      "num_input_tokens_seen": 11204000,
      "step": 17172
    },
    {
      "epoch": 9.001572327044025,
      "grad_norm": 0.08094650506973267,
      "learning_rate": 0.0006709241110960502,
      "loss": 0.4921,
      "num_input_tokens_seen": 11206048,
      "step": 17175
    },
    {
      "epoch": 9.0041928721174,
      "grad_norm": 0.07270103693008423,
      "learning_rate": 0.0006707091846566216,
      "loss": 0.5509,
      "num_input_tokens_seen": 11210464,
      "step": 17180
    },
    {
      "epoch": 9.006813417190775,
      "grad_norm": 0.11419504135847092,
      "learning_rate": 0.0006704942225067791,
      "loss": 0.4248,
      "num_input_tokens_seen": 11213312,
      "step": 17185
    },
    {
      "epoch": 9.00943396226415,
      "grad_norm": 0.0688328668475151,
      "learning_rate": 0.0006702792246914902,
      "loss": 0.403,
      "num_input_tokens_seen": 11216704,
      "step": 17190
    },
    {
      "epoch": 9.012054507337526,
      "grad_norm": 0.09118975698947906,
      "learning_rate": 0.0006700641912557304,
      "loss": 0.4283,
      "num_input_tokens_seen": 11221248,
      "step": 17195
    },
    {
      "epoch": 9.014675052410901,
      "grad_norm": 0.09612806141376495,
      "learning_rate": 0.0006698491222444818,
      "loss": 0.434,
      "num_input_tokens_seen": 11223904,
      "step": 17200
    },
    {
      "epoch": 9.017295597484276,
      "grad_norm": 0.05645307898521423,
      "learning_rate": 0.0006696340177027346,
      "loss": 0.4039,
      "num_input_tokens_seen": 11227648,
      "step": 17205
    },
    {
      "epoch": 9.019916142557651,
      "grad_norm": 0.12805399298667908,
      "learning_rate": 0.0006694188776754863,
      "loss": 0.4214,
      "num_input_tokens_seen": 11230432,
      "step": 17210
    },
    {
      "epoch": 9.022536687631026,
      "grad_norm": 0.17932991683483124,
      "learning_rate": 0.0006692037022077415,
      "loss": 0.4424,
      "num_input_tokens_seen": 11233536,
      "step": 17215
    },
    {
      "epoch": 9.025157232704403,
      "grad_norm": 0.12532451748847961,
      "learning_rate": 0.0006689884913445126,
      "loss": 0.4811,
      "num_input_tokens_seen": 11236512,
      "step": 17220
    },
    {
      "epoch": 9.027777777777779,
      "grad_norm": 0.12061132490634918,
      "learning_rate": 0.0006687732451308193,
      "loss": 0.4125,
      "num_input_tokens_seen": 11238656,
      "step": 17225
    },
    {
      "epoch": 9.030398322851154,
      "grad_norm": 0.06408164650201797,
      "learning_rate": 0.0006685579636116886,
      "loss": 0.5283,
      "num_input_tokens_seen": 11242400,
      "step": 17230
    },
    {
      "epoch": 9.033018867924529,
      "grad_norm": 0.1348995566368103,
      "learning_rate": 0.0006683426468321547,
      "loss": 0.4596,
      "num_input_tokens_seen": 11245824,
      "step": 17235
    },
    {
      "epoch": 9.035639412997904,
      "grad_norm": 0.11905241757631302,
      "learning_rate": 0.0006681272948372598,
      "loss": 0.3665,
      "num_input_tokens_seen": 11248736,
      "step": 17240
    },
    {
      "epoch": 9.03825995807128,
      "grad_norm": 0.09602256119251251,
      "learning_rate": 0.000667911907672053,
      "loss": 0.5402,
      "num_input_tokens_seen": 11252096,
      "step": 17245
    },
    {
      "epoch": 9.040880503144654,
      "grad_norm": 0.10054950416088104,
      "learning_rate": 0.0006676964853815906,
      "loss": 0.434,
      "num_input_tokens_seen": 11258496,
      "step": 17250
    },
    {
      "epoch": 9.04350104821803,
      "grad_norm": 0.18389347195625305,
      "learning_rate": 0.0006674810280109367,
      "loss": 0.4036,
      "num_input_tokens_seen": 11260960,
      "step": 17255
    },
    {
      "epoch": 9.046121593291405,
      "grad_norm": 0.16966785490512848,
      "learning_rate": 0.0006672655356051625,
      "loss": 0.4207,
      "num_input_tokens_seen": 11264512,
      "step": 17260
    },
    {
      "epoch": 9.04874213836478,
      "grad_norm": 0.1006324291229248,
      "learning_rate": 0.0006670500082093465,
      "loss": 0.4955,
      "num_input_tokens_seen": 11268256,
      "step": 17265
    },
    {
      "epoch": 9.051362683438155,
      "grad_norm": 0.08755988627672195,
      "learning_rate": 0.0006668344458685745,
      "loss": 0.5992,
      "num_input_tokens_seen": 11271328,
      "step": 17270
    },
    {
      "epoch": 9.05398322851153,
      "grad_norm": 0.116339311003685,
      "learning_rate": 0.00066661884862794,
      "loss": 0.3778,
      "num_input_tokens_seen": 11274048,
      "step": 17275
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 0.06169823929667473,
      "learning_rate": 0.000666403216532543,
      "loss": 0.5216,
      "num_input_tokens_seen": 11277440,
      "step": 17280
    },
    {
      "epoch": 9.05922431865828,
      "grad_norm": 0.0766393393278122,
      "learning_rate": 0.0006661875496274916,
      "loss": 0.4623,
      "num_input_tokens_seen": 11281120,
      "step": 17285
    },
    {
      "epoch": 9.061844863731656,
      "grad_norm": 0.051424648612737656,
      "learning_rate": 0.0006659718479579008,
      "loss": 0.4526,
      "num_input_tokens_seen": 11284064,
      "step": 17290
    },
    {
      "epoch": 9.064465408805031,
      "grad_norm": 0.0899350568652153,
      "learning_rate": 0.0006657561115688929,
      "loss": 0.3997,
      "num_input_tokens_seen": 11288096,
      "step": 17295
    },
    {
      "epoch": 9.067085953878406,
      "grad_norm": 0.06051195412874222,
      "learning_rate": 0.0006655403405055977,
      "loss": 0.3312,
      "num_input_tokens_seen": 11292032,
      "step": 17300
    },
    {
      "epoch": 9.069706498951781,
      "grad_norm": 0.12277887761592865,
      "learning_rate": 0.0006653245348131517,
      "loss": 0.3861,
      "num_input_tokens_seen": 11294720,
      "step": 17305
    },
    {
      "epoch": 9.072327044025156,
      "grad_norm": 0.08402106910943985,
      "learning_rate": 0.0006651086945366991,
      "loss": 0.4658,
      "num_input_tokens_seen": 11298880,
      "step": 17310
    },
    {
      "epoch": 9.074947589098532,
      "grad_norm": 0.08289889246225357,
      "learning_rate": 0.0006648928197213914,
      "loss": 0.626,
      "num_input_tokens_seen": 11302336,
      "step": 17315
    },
    {
      "epoch": 9.077568134171909,
      "grad_norm": 0.12685786187648773,
      "learning_rate": 0.0006646769104123868,
      "loss": 0.3639,
      "num_input_tokens_seen": 11305216,
      "step": 17320
    },
    {
      "epoch": 9.080188679245284,
      "grad_norm": 0.10486709326505661,
      "learning_rate": 0.0006644609666548513,
      "loss": 0.5793,
      "num_input_tokens_seen": 11308512,
      "step": 17325
    },
    {
      "epoch": 9.082809224318659,
      "grad_norm": 0.10676731914281845,
      "learning_rate": 0.000664244988493958,
      "loss": 0.534,
      "num_input_tokens_seen": 11312416,
      "step": 17330
    },
    {
      "epoch": 9.085429769392034,
      "grad_norm": 0.08010406792163849,
      "learning_rate": 0.0006640289759748867,
      "loss": 0.3657,
      "num_input_tokens_seen": 11315904,
      "step": 17335
    },
    {
      "epoch": 9.08805031446541,
      "grad_norm": 0.018770670518279076,
      "learning_rate": 0.0006638129291428252,
      "loss": 0.3068,
      "num_input_tokens_seen": 11321664,
      "step": 17340
    },
    {
      "epoch": 9.090670859538784,
      "grad_norm": 0.08520646393299103,
      "learning_rate": 0.0006635968480429677,
      "loss": 0.3932,
      "num_input_tokens_seen": 11325728,
      "step": 17345
    },
    {
      "epoch": 9.09329140461216,
      "grad_norm": 0.09037657082080841,
      "learning_rate": 0.0006633807327205162,
      "loss": 0.4248,
      "num_input_tokens_seen": 11329344,
      "step": 17350
    },
    {
      "epoch": 9.095911949685535,
      "grad_norm": 0.13872405886650085,
      "learning_rate": 0.0006631645832206789,
      "loss": 0.4221,
      "num_input_tokens_seen": 11331840,
      "step": 17355
    },
    {
      "epoch": 9.09853249475891,
      "grad_norm": 0.10043752938508987,
      "learning_rate": 0.0006629483995886727,
      "loss": 0.5001,
      "num_input_tokens_seen": 11334912,
      "step": 17360
    },
    {
      "epoch": 9.101153039832285,
      "grad_norm": 0.1160997673869133,
      "learning_rate": 0.0006627321818697202,
      "loss": 0.4519,
      "num_input_tokens_seen": 11337344,
      "step": 17365
    },
    {
      "epoch": 9.10377358490566,
      "grad_norm": 0.07366121560335159,
      "learning_rate": 0.0006625159301090518,
      "loss": 0.5895,
      "num_input_tokens_seen": 11340384,
      "step": 17370
    },
    {
      "epoch": 9.106394129979035,
      "grad_norm": 0.18176054954528809,
      "learning_rate": 0.0006622996443519047,
      "loss": 0.4491,
      "num_input_tokens_seen": 11342976,
      "step": 17375
    },
    {
      "epoch": 9.10901467505241,
      "grad_norm": 0.2131030261516571,
      "learning_rate": 0.0006620833246435238,
      "loss": 0.4204,
      "num_input_tokens_seen": 11345504,
      "step": 17380
    },
    {
      "epoch": 9.111635220125786,
      "grad_norm": 0.07583621144294739,
      "learning_rate": 0.0006618669710291606,
      "loss": 0.4378,
      "num_input_tokens_seen": 11349216,
      "step": 17385
    },
    {
      "epoch": 9.114255765199161,
      "grad_norm": 0.08728117495775223,
      "learning_rate": 0.0006616505835540736,
      "loss": 0.4492,
      "num_input_tokens_seen": 11352576,
      "step": 17390
    },
    {
      "epoch": 9.116876310272536,
      "grad_norm": 0.09088297933340073,
      "learning_rate": 0.0006614341622635287,
      "loss": 0.3812,
      "num_input_tokens_seen": 11355104,
      "step": 17395
    },
    {
      "epoch": 9.119496855345911,
      "grad_norm": 0.1256064772605896,
      "learning_rate": 0.0006612177072027989,
      "loss": 0.5131,
      "num_input_tokens_seen": 11361504,
      "step": 17400
    },
    {
      "epoch": 9.122117400419286,
      "grad_norm": 0.07881557941436768,
      "learning_rate": 0.0006610012184171641,
      "loss": 0.4321,
      "num_input_tokens_seen": 11365120,
      "step": 17405
    },
    {
      "epoch": 9.124737945492662,
      "grad_norm": 0.09677232801914215,
      "learning_rate": 0.0006607846959519109,
      "loss": 0.5137,
      "num_input_tokens_seen": 11368800,
      "step": 17410
    },
    {
      "epoch": 9.127358490566039,
      "grad_norm": 0.10375039279460907,
      "learning_rate": 0.0006605681398523339,
      "loss": 0.678,
      "num_input_tokens_seen": 11371616,
      "step": 17415
    },
    {
      "epoch": 9.129979035639414,
      "grad_norm": 0.09179718792438507,
      "learning_rate": 0.0006603515501637338,
      "loss": 0.5186,
      "num_input_tokens_seen": 11374496,
      "step": 17420
    },
    {
      "epoch": 9.132599580712789,
      "grad_norm": 0.10920065641403198,
      "learning_rate": 0.0006601349269314187,
      "loss": 0.4227,
      "num_input_tokens_seen": 11378624,
      "step": 17425
    },
    {
      "epoch": 9.135220125786164,
      "grad_norm": 0.14311835169792175,
      "learning_rate": 0.0006599182702007042,
      "loss": 0.5231,
      "num_input_tokens_seen": 11381312,
      "step": 17430
    },
    {
      "epoch": 9.13784067085954,
      "grad_norm": 0.15138262510299683,
      "learning_rate": 0.0006597015800169116,
      "loss": 0.3695,
      "num_input_tokens_seen": 11383808,
      "step": 17435
    },
    {
      "epoch": 9.140461215932914,
      "grad_norm": 0.09867879003286362,
      "learning_rate": 0.0006594848564253705,
      "loss": 0.4707,
      "num_input_tokens_seen": 11387840,
      "step": 17440
    },
    {
      "epoch": 9.14308176100629,
      "grad_norm": 0.076818086206913,
      "learning_rate": 0.0006592680994714171,
      "loss": 0.3706,
      "num_input_tokens_seen": 11392800,
      "step": 17445
    },
    {
      "epoch": 9.145702306079665,
      "grad_norm": 0.11317351460456848,
      "learning_rate": 0.0006590513092003943,
      "loss": 0.3254,
      "num_input_tokens_seen": 11396224,
      "step": 17450
    },
    {
      "epoch": 9.14832285115304,
      "grad_norm": 0.0991118773818016,
      "learning_rate": 0.0006588344856576521,
      "loss": 0.4835,
      "num_input_tokens_seen": 11399456,
      "step": 17455
    },
    {
      "epoch": 9.150943396226415,
      "grad_norm": 0.09634196758270264,
      "learning_rate": 0.0006586176288885475,
      "loss": 0.4102,
      "num_input_tokens_seen": 11402304,
      "step": 17460
    },
    {
      "epoch": 9.15356394129979,
      "grad_norm": 0.09141942113637924,
      "learning_rate": 0.0006584007389384446,
      "loss": 0.4877,
      "num_input_tokens_seen": 11405440,
      "step": 17465
    },
    {
      "epoch": 9.156184486373165,
      "grad_norm": 0.0798366442322731,
      "learning_rate": 0.0006581838158527144,
      "loss": 0.4804,
      "num_input_tokens_seen": 11408256,
      "step": 17470
    },
    {
      "epoch": 9.15880503144654,
      "grad_norm": 0.057573866099119186,
      "learning_rate": 0.0006579668596767343,
      "loss": 0.3514,
      "num_input_tokens_seen": 11412288,
      "step": 17475
    },
    {
      "epoch": 9.161425576519916,
      "grad_norm": 0.15731969475746155,
      "learning_rate": 0.0006577498704558894,
      "loss": 0.4181,
      "num_input_tokens_seen": 11414816,
      "step": 17480
    },
    {
      "epoch": 9.164046121593291,
      "grad_norm": 0.07951829582452774,
      "learning_rate": 0.0006575328482355715,
      "loss": 0.4632,
      "num_input_tokens_seen": 11417632,
      "step": 17485
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 0.10453629493713379,
      "learning_rate": 0.0006573157930611788,
      "loss": 0.5855,
      "num_input_tokens_seen": 11420384,
      "step": 17490
    },
    {
      "epoch": 9.169287211740041,
      "grad_norm": 0.07047627121210098,
      "learning_rate": 0.000657098704978117,
      "loss": 0.5091,
      "num_input_tokens_seen": 11423648,
      "step": 17495
    },
    {
      "epoch": 9.171907756813416,
      "grad_norm": 0.06105698645114899,
      "learning_rate": 0.0006568815840317984,
      "loss": 0.3944,
      "num_input_tokens_seen": 11426976,
      "step": 17500
    },
    {
      "epoch": 9.174528301886792,
      "grad_norm": 0.12980155646800995,
      "learning_rate": 0.0006566644302676422,
      "loss": 0.3877,
      "num_input_tokens_seen": 11430304,
      "step": 17505
    },
    {
      "epoch": 9.177148846960169,
      "grad_norm": 0.17827948927879333,
      "learning_rate": 0.0006564472437310746,
      "loss": 0.4351,
      "num_input_tokens_seen": 11432704,
      "step": 17510
    },
    {
      "epoch": 9.179769392033544,
      "grad_norm": 0.07154131680727005,
      "learning_rate": 0.0006562300244675283,
      "loss": 0.3962,
      "num_input_tokens_seen": 11436448,
      "step": 17515
    },
    {
      "epoch": 9.182389937106919,
      "grad_norm": 0.13711880147457123,
      "learning_rate": 0.0006560127725224432,
      "loss": 0.4112,
      "num_input_tokens_seen": 11439872,
      "step": 17520
    },
    {
      "epoch": 9.185010482180294,
      "grad_norm": 0.23529593646526337,
      "learning_rate": 0.0006557954879412662,
      "loss": 0.5758,
      "num_input_tokens_seen": 11442304,
      "step": 17525
    },
    {
      "epoch": 9.18763102725367,
      "grad_norm": 0.08873890340328217,
      "learning_rate": 0.0006555781707694506,
      "loss": 0.6112,
      "num_input_tokens_seen": 11445856,
      "step": 17530
    },
    {
      "epoch": 9.190251572327044,
      "grad_norm": 0.1034906655550003,
      "learning_rate": 0.0006553608210524564,
      "loss": 0.4364,
      "num_input_tokens_seen": 11452288,
      "step": 17535
    },
    {
      "epoch": 9.19287211740042,
      "grad_norm": 0.10705482214689255,
      "learning_rate": 0.0006551434388357509,
      "loss": 0.5563,
      "num_input_tokens_seen": 11456032,
      "step": 17540
    },
    {
      "epoch": 9.195492662473795,
      "grad_norm": 0.08552296459674835,
      "learning_rate": 0.000654926024164808,
      "loss": 0.3992,
      "num_input_tokens_seen": 11459392,
      "step": 17545
    },
    {
      "epoch": 9.19811320754717,
      "grad_norm": 0.15522898733615875,
      "learning_rate": 0.0006547085770851084,
      "loss": 0.4816,
      "num_input_tokens_seen": 11462336,
      "step": 17550
    },
    {
      "epoch": 9.200733752620545,
      "grad_norm": 0.21013003587722778,
      "learning_rate": 0.0006544910976421395,
      "loss": 0.4476,
      "num_input_tokens_seen": 11464896,
      "step": 17555
    },
    {
      "epoch": 9.20335429769392,
      "grad_norm": 0.13142156600952148,
      "learning_rate": 0.0006542735858813953,
      "loss": 0.4232,
      "num_input_tokens_seen": 11472224,
      "step": 17560
    },
    {
      "epoch": 9.205974842767295,
      "grad_norm": 0.10853858292102814,
      "learning_rate": 0.0006540560418483771,
      "loss": 0.4926,
      "num_input_tokens_seen": 11474912,
      "step": 17565
    },
    {
      "epoch": 9.20859538784067,
      "grad_norm": 0.2045680433511734,
      "learning_rate": 0.0006538384655885925,
      "loss": 0.6496,
      "num_input_tokens_seen": 11478240,
      "step": 17570
    },
    {
      "epoch": 9.211215932914046,
      "grad_norm": 0.10694389790296555,
      "learning_rate": 0.000653620857147556,
      "loss": 0.3852,
      "num_input_tokens_seen": 11481216,
      "step": 17575
    },
    {
      "epoch": 9.213836477987421,
      "grad_norm": 0.10266074538230896,
      "learning_rate": 0.0006534032165707887,
      "loss": 0.4432,
      "num_input_tokens_seen": 11484384,
      "step": 17580
    },
    {
      "epoch": 9.216457023060796,
      "grad_norm": 0.1464919149875641,
      "learning_rate": 0.0006531855439038187,
      "loss": 0.5199,
      "num_input_tokens_seen": 11487104,
      "step": 17585
    },
    {
      "epoch": 9.219077568134171,
      "grad_norm": 0.12502747774124146,
      "learning_rate": 0.0006529678391921805,
      "loss": 0.377,
      "num_input_tokens_seen": 11489696,
      "step": 17590
    },
    {
      "epoch": 9.221698113207546,
      "grad_norm": 0.2511185109615326,
      "learning_rate": 0.0006527501024814155,
      "loss": 0.5347,
      "num_input_tokens_seen": 11492512,
      "step": 17595
    },
    {
      "epoch": 9.224318658280922,
      "grad_norm": 0.08454998582601547,
      "learning_rate": 0.0006525323338170718,
      "loss": 0.4237,
      "num_input_tokens_seen": 11495648,
      "step": 17600
    },
    {
      "epoch": 9.226939203354299,
      "grad_norm": 0.08037509769201279,
      "learning_rate": 0.000652314533244704,
      "loss": 0.4926,
      "num_input_tokens_seen": 11498976,
      "step": 17605
    },
    {
      "epoch": 9.229559748427674,
      "grad_norm": 0.11336810886859894,
      "learning_rate": 0.0006520967008098735,
      "loss": 0.403,
      "num_input_tokens_seen": 11501888,
      "step": 17610
    },
    {
      "epoch": 9.232180293501049,
      "grad_norm": 0.10895546525716782,
      "learning_rate": 0.0006518788365581485,
      "loss": 0.532,
      "num_input_tokens_seen": 11504544,
      "step": 17615
    },
    {
      "epoch": 9.234800838574424,
      "grad_norm": 0.0774364173412323,
      "learning_rate": 0.0006516609405351034,
      "loss": 0.391,
      "num_input_tokens_seen": 11507168,
      "step": 17620
    },
    {
      "epoch": 9.2374213836478,
      "grad_norm": 0.08231481909751892,
      "learning_rate": 0.00065144301278632,
      "loss": 0.385,
      "num_input_tokens_seen": 11510272,
      "step": 17625
    },
    {
      "epoch": 9.240041928721174,
      "grad_norm": 0.11630823463201523,
      "learning_rate": 0.0006512250533573862,
      "loss": 0.4587,
      "num_input_tokens_seen": 11513280,
      "step": 17630
    },
    {
      "epoch": 9.24266247379455,
      "grad_norm": 0.06985265016555786,
      "learning_rate": 0.0006510070622938962,
      "loss": 0.3657,
      "num_input_tokens_seen": 11516384,
      "step": 17635
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 0.10658679157495499,
      "learning_rate": 0.0006507890396414516,
      "loss": 0.5291,
      "num_input_tokens_seen": 11520672,
      "step": 17640
    },
    {
      "epoch": 9.2479035639413,
      "grad_norm": 0.12396284192800522,
      "learning_rate": 0.0006505709854456602,
      "loss": 0.4949,
      "num_input_tokens_seen": 11523456,
      "step": 17645
    },
    {
      "epoch": 9.250524109014675,
      "grad_norm": 0.13426744937896729,
      "learning_rate": 0.0006503528997521365,
      "loss": 0.3615,
      "num_input_tokens_seen": 11526240,
      "step": 17650
    },
    {
      "epoch": 9.25314465408805,
      "grad_norm": 0.08006512373685837,
      "learning_rate": 0.0006501347826065017,
      "loss": 0.4596,
      "num_input_tokens_seen": 11531040,
      "step": 17655
    },
    {
      "epoch": 9.255765199161425,
      "grad_norm": 0.09160688519477844,
      "learning_rate": 0.000649916634054383,
      "loss": 0.4566,
      "num_input_tokens_seen": 11533984,
      "step": 17660
    },
    {
      "epoch": 9.2583857442348,
      "grad_norm": 0.0988197848200798,
      "learning_rate": 0.0006496984541414147,
      "loss": 0.4464,
      "num_input_tokens_seen": 11537248,
      "step": 17665
    },
    {
      "epoch": 9.261006289308176,
      "grad_norm": 0.2081790715456009,
      "learning_rate": 0.0006494802429132378,
      "loss": 0.4544,
      "num_input_tokens_seen": 11540384,
      "step": 17670
    },
    {
      "epoch": 9.26362683438155,
      "grad_norm": 0.09969879686832428,
      "learning_rate": 0.0006492620004154993,
      "loss": 0.537,
      "num_input_tokens_seen": 11543424,
      "step": 17675
    },
    {
      "epoch": 9.266247379454926,
      "grad_norm": 0.09665560722351074,
      "learning_rate": 0.0006490437266938534,
      "loss": 0.4478,
      "num_input_tokens_seen": 11547136,
      "step": 17680
    },
    {
      "epoch": 9.268867924528301,
      "grad_norm": 0.17467382550239563,
      "learning_rate": 0.0006488254217939601,
      "loss": 0.5558,
      "num_input_tokens_seen": 11550528,
      "step": 17685
    },
    {
      "epoch": 9.271488469601676,
      "grad_norm": 0.0570111982524395,
      "learning_rate": 0.0006486070857614863,
      "loss": 0.3497,
      "num_input_tokens_seen": 11553440,
      "step": 17690
    },
    {
      "epoch": 9.274109014675052,
      "grad_norm": 0.08380302041769028,
      "learning_rate": 0.0006483887186421059,
      "loss": 0.3508,
      "num_input_tokens_seen": 11557184,
      "step": 17695
    },
    {
      "epoch": 9.276729559748428,
      "grad_norm": 0.12756238877773285,
      "learning_rate": 0.0006481703204814982,
      "loss": 0.4037,
      "num_input_tokens_seen": 11560416,
      "step": 17700
    },
    {
      "epoch": 9.279350104821804,
      "grad_norm": 0.09372636675834656,
      "learning_rate": 0.00064795189132535,
      "loss": 0.4584,
      "num_input_tokens_seen": 11563712,
      "step": 17705
    },
    {
      "epoch": 9.281970649895179,
      "grad_norm": 0.0685264989733696,
      "learning_rate": 0.000647733431219354,
      "loss": 0.5034,
      "num_input_tokens_seen": 11567776,
      "step": 17710
    },
    {
      "epoch": 9.284591194968554,
      "grad_norm": 0.08700624108314514,
      "learning_rate": 0.0006475149402092097,
      "loss": 0.4866,
      "num_input_tokens_seen": 11571456,
      "step": 17715
    },
    {
      "epoch": 9.28721174004193,
      "grad_norm": 0.09279298037290573,
      "learning_rate": 0.0006472964183406229,
      "loss": 0.453,
      "num_input_tokens_seen": 11575136,
      "step": 17720
    },
    {
      "epoch": 9.289832285115304,
      "grad_norm": 0.061131082475185394,
      "learning_rate": 0.0006470778656593055,
      "loss": 0.4626,
      "num_input_tokens_seen": 11578880,
      "step": 17725
    },
    {
      "epoch": 9.29245283018868,
      "grad_norm": 0.09127983450889587,
      "learning_rate": 0.0006468592822109769,
      "loss": 0.3178,
      "num_input_tokens_seen": 11581792,
      "step": 17730
    },
    {
      "epoch": 9.295073375262055,
      "grad_norm": 0.068997822701931,
      "learning_rate": 0.0006466406680413619,
      "loss": 0.5677,
      "num_input_tokens_seen": 11584704,
      "step": 17735
    },
    {
      "epoch": 9.29769392033543,
      "grad_norm": 0.16836753487586975,
      "learning_rate": 0.000646422023196192,
      "loss": 0.5597,
      "num_input_tokens_seen": 11587936,
      "step": 17740
    },
    {
      "epoch": 9.300314465408805,
      "grad_norm": 0.2874640226364136,
      "learning_rate": 0.0006462033477212054,
      "loss": 0.5186,
      "num_input_tokens_seen": 11591232,
      "step": 17745
    },
    {
      "epoch": 9.30293501048218,
      "grad_norm": 0.09239788353443146,
      "learning_rate": 0.0006459846416621462,
      "loss": 0.6321,
      "num_input_tokens_seen": 11594656,
      "step": 17750
    },
    {
      "epoch": 9.305555555555555,
      "grad_norm": 0.10134842246770859,
      "learning_rate": 0.0006457659050647657,
      "loss": 0.3889,
      "num_input_tokens_seen": 11598016,
      "step": 17755
    },
    {
      "epoch": 9.30817610062893,
      "grad_norm": 0.09986919164657593,
      "learning_rate": 0.0006455471379748208,
      "loss": 0.3637,
      "num_input_tokens_seen": 11601312,
      "step": 17760
    },
    {
      "epoch": 9.310796645702306,
      "grad_norm": 0.09306449443101883,
      "learning_rate": 0.000645328340438075,
      "loss": 0.5366,
      "num_input_tokens_seen": 11604896,
      "step": 17765
    },
    {
      "epoch": 9.31341719077568,
      "grad_norm": 0.09320754557847977,
      "learning_rate": 0.0006451095125002985,
      "loss": 0.4307,
      "num_input_tokens_seen": 11607840,
      "step": 17770
    },
    {
      "epoch": 9.316037735849056,
      "grad_norm": 0.12274452298879623,
      "learning_rate": 0.0006448906542072674,
      "loss": 0.4101,
      "num_input_tokens_seen": 11611392,
      "step": 17775
    },
    {
      "epoch": 9.318658280922431,
      "grad_norm": 0.08062182366847992,
      "learning_rate": 0.0006446717656047645,
      "loss": 0.4007,
      "num_input_tokens_seen": 11614752,
      "step": 17780
    },
    {
      "epoch": 9.321278825995806,
      "grad_norm": 0.11234714835882187,
      "learning_rate": 0.0006444528467385789,
      "loss": 0.4292,
      "num_input_tokens_seen": 11617728,
      "step": 17785
    },
    {
      "epoch": 9.323899371069182,
      "grad_norm": 0.08604279160499573,
      "learning_rate": 0.0006442338976545054,
      "loss": 0.4202,
      "num_input_tokens_seen": 11620736,
      "step": 17790
    },
    {
      "epoch": 9.326519916142558,
      "grad_norm": 0.09068909287452698,
      "learning_rate": 0.0006440149183983461,
      "loss": 0.5432,
      "num_input_tokens_seen": 11623744,
      "step": 17795
    },
    {
      "epoch": 9.329140461215934,
      "grad_norm": 0.11190130561590195,
      "learning_rate": 0.0006437959090159093,
      "loss": 0.4937,
      "num_input_tokens_seen": 11626496,
      "step": 17800
    },
    {
      "epoch": 9.331761006289309,
      "grad_norm": 0.10376586019992828,
      "learning_rate": 0.0006435768695530085,
      "loss": 0.3954,
      "num_input_tokens_seen": 11629504,
      "step": 17805
    },
    {
      "epoch": 9.334381551362684,
      "grad_norm": 0.15025927126407623,
      "learning_rate": 0.0006433578000554645,
      "loss": 0.4905,
      "num_input_tokens_seen": 11632384,
      "step": 17810
    },
    {
      "epoch": 9.33700209643606,
      "grad_norm": 0.13291466236114502,
      "learning_rate": 0.0006431387005691045,
      "loss": 0.3967,
      "num_input_tokens_seen": 11635616,
      "step": 17815
    },
    {
      "epoch": 9.339622641509434,
      "grad_norm": 0.11345231533050537,
      "learning_rate": 0.0006429195711397611,
      "loss": 0.4282,
      "num_input_tokens_seen": 11638912,
      "step": 17820
    },
    {
      "epoch": 9.34224318658281,
      "grad_norm": 0.11227905005216599,
      "learning_rate": 0.0006427004118132742,
      "loss": 0.4906,
      "num_input_tokens_seen": 11642432,
      "step": 17825
    },
    {
      "epoch": 9.344863731656185,
      "grad_norm": 0.12313897162675858,
      "learning_rate": 0.0006424812226354889,
      "loss": 0.4082,
      "num_input_tokens_seen": 11645120,
      "step": 17830
    },
    {
      "epoch": 9.34748427672956,
      "grad_norm": 0.11985563486814499,
      "learning_rate": 0.0006422620036522574,
      "loss": 0.4082,
      "num_input_tokens_seen": 11647456,
      "step": 17835
    },
    {
      "epoch": 9.350104821802935,
      "grad_norm": 0.07586375623941422,
      "learning_rate": 0.000642042754909438,
      "loss": 0.472,
      "num_input_tokens_seen": 11650976,
      "step": 17840
    },
    {
      "epoch": 9.35272536687631,
      "grad_norm": 0.12271080166101456,
      "learning_rate": 0.0006418234764528945,
      "loss": 0.4425,
      "num_input_tokens_seen": 11654304,
      "step": 17845
    },
    {
      "epoch": 9.355345911949685,
      "grad_norm": 0.10997079312801361,
      "learning_rate": 0.0006416041683284978,
      "loss": 0.5073,
      "num_input_tokens_seen": 11658592,
      "step": 17850
    },
    {
      "epoch": 9.35796645702306,
      "grad_norm": 0.143546462059021,
      "learning_rate": 0.0006413848305821246,
      "loss": 0.5415,
      "num_input_tokens_seen": 11662464,
      "step": 17855
    },
    {
      "epoch": 9.360587002096436,
      "grad_norm": 0.08491163700819016,
      "learning_rate": 0.0006411654632596578,
      "loss": 0.3913,
      "num_input_tokens_seen": 11665248,
      "step": 17860
    },
    {
      "epoch": 9.36320754716981,
      "grad_norm": 0.12689198553562164,
      "learning_rate": 0.0006409460664069867,
      "loss": 0.5148,
      "num_input_tokens_seen": 11668256,
      "step": 17865
    },
    {
      "epoch": 9.365828092243186,
      "grad_norm": 0.08738938719034195,
      "learning_rate": 0.0006407266400700064,
      "loss": 0.49,
      "num_input_tokens_seen": 11671264,
      "step": 17870
    },
    {
      "epoch": 9.368448637316561,
      "grad_norm": 0.12088426202535629,
      "learning_rate": 0.0006405071842946185,
      "loss": 0.6437,
      "num_input_tokens_seen": 11673728,
      "step": 17875
    },
    {
      "epoch": 9.371069182389936,
      "grad_norm": 0.09251347929239273,
      "learning_rate": 0.0006402876991267308,
      "loss": 0.3963,
      "num_input_tokens_seen": 11676960,
      "step": 17880
    },
    {
      "epoch": 9.373689727463312,
      "grad_norm": 0.15628333389759064,
      "learning_rate": 0.0006400681846122568,
      "loss": 0.5067,
      "num_input_tokens_seen": 11681440,
      "step": 17885
    },
    {
      "epoch": 9.376310272536688,
      "grad_norm": 0.1453104317188263,
      "learning_rate": 0.0006398486407971166,
      "loss": 0.4297,
      "num_input_tokens_seen": 11684544,
      "step": 17890
    },
    {
      "epoch": 9.378930817610064,
      "grad_norm": 0.0975276380777359,
      "learning_rate": 0.0006396290677272361,
      "loss": 0.4524,
      "num_input_tokens_seen": 11687584,
      "step": 17895
    },
    {
      "epoch": 9.381551362683439,
      "grad_norm": 0.07665035128593445,
      "learning_rate": 0.0006394094654485477,
      "loss": 0.3579,
      "num_input_tokens_seen": 11690944,
      "step": 17900
    },
    {
      "epoch": 9.384171907756814,
      "grad_norm": 0.13391005992889404,
      "learning_rate": 0.0006391898340069896,
      "loss": 0.4297,
      "num_input_tokens_seen": 11694112,
      "step": 17905
    },
    {
      "epoch": 9.38679245283019,
      "grad_norm": 0.0965605080127716,
      "learning_rate": 0.0006389701734485061,
      "loss": 0.4811,
      "num_input_tokens_seen": 11697376,
      "step": 17910
    },
    {
      "epoch": 9.389412997903564,
      "grad_norm": 0.1276337206363678,
      "learning_rate": 0.0006387504838190479,
      "loss": 0.3756,
      "num_input_tokens_seen": 11700832,
      "step": 17915
    },
    {
      "epoch": 9.39203354297694,
      "grad_norm": 0.10743089020252228,
      "learning_rate": 0.0006385307651645715,
      "loss": 0.4153,
      "num_input_tokens_seen": 11703520,
      "step": 17920
    },
    {
      "epoch": 9.394654088050315,
      "grad_norm": 0.07543767243623734,
      "learning_rate": 0.0006383110175310393,
      "loss": 0.4333,
      "num_input_tokens_seen": 11707008,
      "step": 17925
    },
    {
      "epoch": 9.39727463312369,
      "grad_norm": 0.09141096472740173,
      "learning_rate": 0.0006380912409644202,
      "loss": 0.3417,
      "num_input_tokens_seen": 11709408,
      "step": 17930
    },
    {
      "epoch": 9.399895178197065,
      "grad_norm": 0.11999539285898209,
      "learning_rate": 0.000637871435510689,
      "loss": 0.4003,
      "num_input_tokens_seen": 11712000,
      "step": 17935
    },
    {
      "epoch": 9.40251572327044,
      "grad_norm": 0.09534591436386108,
      "learning_rate": 0.0006376516012158262,
      "loss": 0.5248,
      "num_input_tokens_seen": 11716064,
      "step": 17940
    },
    {
      "epoch": 9.405136268343815,
      "grad_norm": 0.12817615270614624,
      "learning_rate": 0.000637431738125819,
      "loss": 0.3821,
      "num_input_tokens_seen": 11719776,
      "step": 17945
    },
    {
      "epoch": 9.40775681341719,
      "grad_norm": 0.11747851967811584,
      "learning_rate": 0.00063721184628666,
      "loss": 0.3775,
      "num_input_tokens_seen": 11722880,
      "step": 17950
    },
    {
      "epoch": 9.410377358490566,
      "grad_norm": 0.1110074445605278,
      "learning_rate": 0.0006369919257443484,
      "loss": 0.6267,
      "num_input_tokens_seen": 11725824,
      "step": 17955
    },
    {
      "epoch": 9.41299790356394,
      "grad_norm": 0.12427929788827896,
      "learning_rate": 0.0006367719765448886,
      "loss": 0.5323,
      "num_input_tokens_seen": 11728480,
      "step": 17960
    },
    {
      "epoch": 9.415618448637316,
      "grad_norm": 0.06888765096664429,
      "learning_rate": 0.0006365519987342916,
      "loss": 0.4175,
      "num_input_tokens_seen": 11731040,
      "step": 17965
    },
    {
      "epoch": 9.418238993710691,
      "grad_norm": 0.07922898977994919,
      "learning_rate": 0.0006363319923585746,
      "loss": 0.3584,
      "num_input_tokens_seen": 11733664,
      "step": 17970
    },
    {
      "epoch": 9.420859538784066,
      "grad_norm": 0.13559669256210327,
      "learning_rate": 0.0006361119574637599,
      "loss": 0.3899,
      "num_input_tokens_seen": 11737216,
      "step": 17975
    },
    {
      "epoch": 9.423480083857442,
      "grad_norm": 0.11646141111850739,
      "learning_rate": 0.0006358918940958767,
      "loss": 0.4147,
      "num_input_tokens_seen": 11740864,
      "step": 17980
    },
    {
      "epoch": 9.426100628930818,
      "grad_norm": 0.08183237165212631,
      "learning_rate": 0.0006356718023009597,
      "loss": 0.4078,
      "num_input_tokens_seen": 11743776,
      "step": 17985
    },
    {
      "epoch": 9.428721174004194,
      "grad_norm": 0.08128021657466888,
      "learning_rate": 0.0006354516821250492,
      "loss": 0.4311,
      "num_input_tokens_seen": 11747552,
      "step": 17990
    },
    {
      "epoch": 9.431341719077569,
      "grad_norm": 0.11364620178937912,
      "learning_rate": 0.0006352315336141923,
      "loss": 0.4625,
      "num_input_tokens_seen": 11750656,
      "step": 17995
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 0.12005054950714111,
      "learning_rate": 0.0006350113568144414,
      "loss": 0.3756,
      "num_input_tokens_seen": 11753888,
      "step": 18000
    },
    {
      "epoch": 9.43658280922432,
      "grad_norm": 0.20244741439819336,
      "learning_rate": 0.0006347911517718547,
      "loss": 0.4388,
      "num_input_tokens_seen": 11757184,
      "step": 18005
    },
    {
      "epoch": 9.439203354297694,
      "grad_norm": 0.11577392369508743,
      "learning_rate": 0.0006345709185324972,
      "loss": 0.3986,
      "num_input_tokens_seen": 11759936,
      "step": 18010
    },
    {
      "epoch": 9.44182389937107,
      "grad_norm": 0.10467736423015594,
      "learning_rate": 0.0006343506571424386,
      "loss": 0.4426,
      "num_input_tokens_seen": 11762816,
      "step": 18015
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 0.11468245834112167,
      "learning_rate": 0.0006341303676477552,
      "loss": 0.3908,
      "num_input_tokens_seen": 11765824,
      "step": 18020
    },
    {
      "epoch": 9.44706498951782,
      "grad_norm": 0.09325556457042694,
      "learning_rate": 0.0006339100500945293,
      "loss": 0.405,
      "num_input_tokens_seen": 11769056,
      "step": 18025
    },
    {
      "epoch": 9.449685534591195,
      "grad_norm": 0.09766270220279694,
      "learning_rate": 0.0006336897045288486,
      "loss": 0.3687,
      "num_input_tokens_seen": 11771808,
      "step": 18030
    },
    {
      "epoch": 9.45230607966457,
      "grad_norm": 0.08136346936225891,
      "learning_rate": 0.0006334693309968068,
      "loss": 0.3336,
      "num_input_tokens_seen": 11774720,
      "step": 18035
    },
    {
      "epoch": 9.454926624737945,
      "grad_norm": 0.1816672831773758,
      "learning_rate": 0.0006332489295445038,
      "loss": 0.4087,
      "num_input_tokens_seen": 11777664,
      "step": 18040
    },
    {
      "epoch": 9.45754716981132,
      "grad_norm": 0.09985342621803284,
      "learning_rate": 0.0006330285002180448,
      "loss": 0.3513,
      "num_input_tokens_seen": 11781696,
      "step": 18045
    },
    {
      "epoch": 9.460167714884696,
      "grad_norm": 0.10488422960042953,
      "learning_rate": 0.0006328080430635412,
      "loss": 0.4148,
      "num_input_tokens_seen": 11784672,
      "step": 18050
    },
    {
      "epoch": 9.46278825995807,
      "grad_norm": 0.10157729685306549,
      "learning_rate": 0.0006325875581271103,
      "loss": 0.4629,
      "num_input_tokens_seen": 11787424,
      "step": 18055
    },
    {
      "epoch": 9.465408805031446,
      "grad_norm": 0.1591539829969406,
      "learning_rate": 0.0006323670454548749,
      "loss": 0.4993,
      "num_input_tokens_seen": 11789824,
      "step": 18060
    },
    {
      "epoch": 9.468029350104821,
      "grad_norm": 0.09924926608800888,
      "learning_rate": 0.0006321465050929635,
      "loss": 0.2664,
      "num_input_tokens_seen": 11793120,
      "step": 18065
    },
    {
      "epoch": 9.470649895178196,
      "grad_norm": 0.09842345863580704,
      "learning_rate": 0.000631925937087511,
      "loss": 0.4125,
      "num_input_tokens_seen": 11798720,
      "step": 18070
    },
    {
      "epoch": 9.473270440251572,
      "grad_norm": 0.12055359035730362,
      "learning_rate": 0.0006317053414846576,
      "loss": 0.4333,
      "num_input_tokens_seen": 11801600,
      "step": 18075
    },
    {
      "epoch": 9.475890985324948,
      "grad_norm": 0.15390081703662872,
      "learning_rate": 0.0006314847183305492,
      "loss": 0.3446,
      "num_input_tokens_seen": 11804640,
      "step": 18080
    },
    {
      "epoch": 9.478511530398324,
      "grad_norm": 0.10614681243896484,
      "learning_rate": 0.0006312640676713378,
      "loss": 0.4281,
      "num_input_tokens_seen": 11807296,
      "step": 18085
    },
    {
      "epoch": 9.481132075471699,
      "grad_norm": 0.09433487057685852,
      "learning_rate": 0.0006310433895531811,
      "loss": 0.3211,
      "num_input_tokens_seen": 11809920,
      "step": 18090
    },
    {
      "epoch": 9.483752620545074,
      "grad_norm": 0.08113517612218857,
      "learning_rate": 0.0006308226840222422,
      "loss": 0.393,
      "num_input_tokens_seen": 11813440,
      "step": 18095
    },
    {
      "epoch": 9.48637316561845,
      "grad_norm": 0.09281765669584274,
      "learning_rate": 0.0006306019511246905,
      "loss": 0.4736,
      "num_input_tokens_seen": 11817440,
      "step": 18100
    },
    {
      "epoch": 9.488993710691824,
      "grad_norm": 0.13364063203334808,
      "learning_rate": 0.0006303811909067003,
      "loss": 0.508,
      "num_input_tokens_seen": 11820192,
      "step": 18105
    },
    {
      "epoch": 9.4916142557652,
      "grad_norm": 0.10030653327703476,
      "learning_rate": 0.0006301604034144527,
      "loss": 0.5523,
      "num_input_tokens_seen": 11824192,
      "step": 18110
    },
    {
      "epoch": 9.494234800838575,
      "grad_norm": 0.09484522044658661,
      "learning_rate": 0.0006299395886941336,
      "loss": 0.6074,
      "num_input_tokens_seen": 11826944,
      "step": 18115
    },
    {
      "epoch": 9.49685534591195,
      "grad_norm": 0.08304622769355774,
      "learning_rate": 0.0006297187467919347,
      "loss": 0.3949,
      "num_input_tokens_seen": 11830784,
      "step": 18120
    },
    {
      "epoch": 9.499475890985325,
      "grad_norm": 0.1655794382095337,
      "learning_rate": 0.0006294978777540542,
      "loss": 0.5939,
      "num_input_tokens_seen": 11833632,
      "step": 18125
    },
    {
      "epoch": 9.5020964360587,
      "grad_norm": 0.10910432785749435,
      "learning_rate": 0.000629276981626695,
      "loss": 0.5765,
      "num_input_tokens_seen": 11837056,
      "step": 18130
    },
    {
      "epoch": 9.504716981132075,
      "grad_norm": 0.12403620779514313,
      "learning_rate": 0.0006290560584560661,
      "loss": 0.3688,
      "num_input_tokens_seen": 11840736,
      "step": 18135
    },
    {
      "epoch": 9.50733752620545,
      "grad_norm": 0.10485371202230453,
      "learning_rate": 0.0006288351082883822,
      "loss": 0.6368,
      "num_input_tokens_seen": 11843328,
      "step": 18140
    },
    {
      "epoch": 9.509958071278826,
      "grad_norm": 0.062336407601833344,
      "learning_rate": 0.0006286141311698633,
      "loss": 0.5202,
      "num_input_tokens_seen": 11846784,
      "step": 18145
    },
    {
      "epoch": 9.5125786163522,
      "grad_norm": 0.07559780031442642,
      "learning_rate": 0.0006283931271467355,
      "loss": 0.4909,
      "num_input_tokens_seen": 11849984,
      "step": 18150
    },
    {
      "epoch": 9.515199161425576,
      "grad_norm": 0.1129050925374031,
      "learning_rate": 0.0006281720962652305,
      "loss": 0.4899,
      "num_input_tokens_seen": 11852992,
      "step": 18155
    },
    {
      "epoch": 9.517819706498951,
      "grad_norm": 0.16721490025520325,
      "learning_rate": 0.0006279510385715849,
      "loss": 0.4324,
      "num_input_tokens_seen": 11855776,
      "step": 18160
    },
    {
      "epoch": 9.520440251572326,
      "grad_norm": 0.06989776343107224,
      "learning_rate": 0.0006277299541120419,
      "loss": 0.3441,
      "num_input_tokens_seen": 11859008,
      "step": 18165
    },
    {
      "epoch": 9.523060796645701,
      "grad_norm": 0.12824220955371857,
      "learning_rate": 0.0006275088429328499,
      "loss": 0.3677,
      "num_input_tokens_seen": 11861632,
      "step": 18170
    },
    {
      "epoch": 9.525681341719078,
      "grad_norm": 0.09084419906139374,
      "learning_rate": 0.0006272877050802623,
      "loss": 0.3724,
      "num_input_tokens_seen": 11864864,
      "step": 18175
    },
    {
      "epoch": 9.528301886792454,
      "grad_norm": 0.08409487456083298,
      "learning_rate": 0.0006270665406005393,
      "loss": 0.5091,
      "num_input_tokens_seen": 11868224,
      "step": 18180
    },
    {
      "epoch": 9.530922431865829,
      "grad_norm": 0.09218794107437134,
      "learning_rate": 0.0006268453495399455,
      "loss": 0.4933,
      "num_input_tokens_seen": 11871072,
      "step": 18185
    },
    {
      "epoch": 9.533542976939204,
      "grad_norm": 0.10766283422708511,
      "learning_rate": 0.0006266241319447517,
      "loss": 0.4153,
      "num_input_tokens_seen": 11873888,
      "step": 18190
    },
    {
      "epoch": 9.536163522012579,
      "grad_norm": 0.14448298513889313,
      "learning_rate": 0.0006264028878612343,
      "loss": 0.3935,
      "num_input_tokens_seen": 11876608,
      "step": 18195
    },
    {
      "epoch": 9.538784067085954,
      "grad_norm": 0.07485415041446686,
      "learning_rate": 0.0006261816173356746,
      "loss": 0.4113,
      "num_input_tokens_seen": 11880864,
      "step": 18200
    },
    {
      "epoch": 9.54140461215933,
      "grad_norm": 0.11425630748271942,
      "learning_rate": 0.0006259603204143602,
      "loss": 0.4316,
      "num_input_tokens_seen": 11883520,
      "step": 18205
    },
    {
      "epoch": 9.544025157232705,
      "grad_norm": 0.08640793710947037,
      "learning_rate": 0.0006257389971435839,
      "loss": 0.3629,
      "num_input_tokens_seen": 11887168,
      "step": 18210
    },
    {
      "epoch": 9.54664570230608,
      "grad_norm": 0.11587425321340561,
      "learning_rate": 0.0006255176475696437,
      "loss": 0.6092,
      "num_input_tokens_seen": 11890976,
      "step": 18215
    },
    {
      "epoch": 9.549266247379455,
      "grad_norm": 0.09527276456356049,
      "learning_rate": 0.0006252962717388438,
      "loss": 0.4574,
      "num_input_tokens_seen": 11893888,
      "step": 18220
    },
    {
      "epoch": 9.55188679245283,
      "grad_norm": 0.08220777660608292,
      "learning_rate": 0.0006250748696974932,
      "loss": 0.437,
      "num_input_tokens_seen": 11896704,
      "step": 18225
    },
    {
      "epoch": 9.554507337526205,
      "grad_norm": 0.09200093150138855,
      "learning_rate": 0.0006248534414919067,
      "loss": 0.5422,
      "num_input_tokens_seen": 11900736,
      "step": 18230
    },
    {
      "epoch": 9.55712788259958,
      "grad_norm": 0.0832730084657669,
      "learning_rate": 0.0006246319871684047,
      "loss": 0.5102,
      "num_input_tokens_seen": 11904160,
      "step": 18235
    },
    {
      "epoch": 9.559748427672956,
      "grad_norm": 0.05378561466932297,
      "learning_rate": 0.000624410506773313,
      "loss": 0.3785,
      "num_input_tokens_seen": 11907200,
      "step": 18240
    },
    {
      "epoch": 9.56236897274633,
      "grad_norm": 0.06464916467666626,
      "learning_rate": 0.0006241890003529626,
      "loss": 0.3849,
      "num_input_tokens_seen": 11910368,
      "step": 18245
    },
    {
      "epoch": 9.564989517819706,
      "grad_norm": 0.09234380722045898,
      "learning_rate": 0.00062396746795369,
      "loss": 0.4808,
      "num_input_tokens_seen": 11913472,
      "step": 18250
    },
    {
      "epoch": 9.567610062893081,
      "grad_norm": 0.11222071945667267,
      "learning_rate": 0.0006237459096218375,
      "loss": 0.3612,
      "num_input_tokens_seen": 11917280,
      "step": 18255
    },
    {
      "epoch": 9.570230607966456,
      "grad_norm": 0.10155481845140457,
      "learning_rate": 0.0006235243254037525,
      "loss": 0.3867,
      "num_input_tokens_seen": 11919744,
      "step": 18260
    },
    {
      "epoch": 9.572851153039831,
      "grad_norm": 0.1353759467601776,
      "learning_rate": 0.0006233027153457878,
      "loss": 0.5394,
      "num_input_tokens_seen": 11924096,
      "step": 18265
    },
    {
      "epoch": 9.575471698113208,
      "grad_norm": 0.12209580838680267,
      "learning_rate": 0.0006230810794943017,
      "loss": 0.3567,
      "num_input_tokens_seen": 11927744,
      "step": 18270
    },
    {
      "epoch": 9.578092243186584,
      "grad_norm": 0.2968374788761139,
      "learning_rate": 0.0006228594178956581,
      "loss": 0.4368,
      "num_input_tokens_seen": 11930368,
      "step": 18275
    },
    {
      "epoch": 9.580712788259959,
      "grad_norm": 0.09021079540252686,
      "learning_rate": 0.0006226377305962259,
      "loss": 0.4779,
      "num_input_tokens_seen": 11933344,
      "step": 18280
    },
    {
      "epoch": 9.583333333333334,
      "grad_norm": 0.11069781333208084,
      "learning_rate": 0.0006224160176423796,
      "loss": 0.3595,
      "num_input_tokens_seen": 11936992,
      "step": 18285
    },
    {
      "epoch": 9.585953878406709,
      "grad_norm": 0.11076701432466507,
      "learning_rate": 0.0006221942790804989,
      "loss": 0.4649,
      "num_input_tokens_seen": 11940288,
      "step": 18290
    },
    {
      "epoch": 9.588574423480084,
      "grad_norm": 0.14414456486701965,
      "learning_rate": 0.0006219725149569691,
      "loss": 0.3841,
      "num_input_tokens_seen": 11943072,
      "step": 18295
    },
    {
      "epoch": 9.59119496855346,
      "grad_norm": 0.10394365340471268,
      "learning_rate": 0.0006217507253181809,
      "loss": 0.364,
      "num_input_tokens_seen": 11946016,
      "step": 18300
    },
    {
      "epoch": 9.593815513626835,
      "grad_norm": 0.1622346043586731,
      "learning_rate": 0.0006215289102105297,
      "loss": 0.4827,
      "num_input_tokens_seen": 11949280,
      "step": 18305
    },
    {
      "epoch": 9.59643605870021,
      "grad_norm": 0.07206673175096512,
      "learning_rate": 0.0006213070696804171,
      "loss": 0.4042,
      "num_input_tokens_seen": 11953216,
      "step": 18310
    },
    {
      "epoch": 9.599056603773585,
      "grad_norm": 0.09014561027288437,
      "learning_rate": 0.0006210852037742495,
      "loss": 0.432,
      "num_input_tokens_seen": 11956320,
      "step": 18315
    },
    {
      "epoch": 9.60167714884696,
      "grad_norm": 0.08849646151065826,
      "learning_rate": 0.0006208633125384385,
      "loss": 0.4475,
      "num_input_tokens_seen": 11959104,
      "step": 18320
    },
    {
      "epoch": 9.604297693920335,
      "grad_norm": 0.089805468916893,
      "learning_rate": 0.0006206413960194017,
      "loss": 0.4096,
      "num_input_tokens_seen": 11962208,
      "step": 18325
    },
    {
      "epoch": 9.60691823899371,
      "grad_norm": 0.1278567910194397,
      "learning_rate": 0.000620419454263561,
      "loss": 0.5802,
      "num_input_tokens_seen": 11964832,
      "step": 18330
    },
    {
      "epoch": 9.609538784067086,
      "grad_norm": 0.07896825671195984,
      "learning_rate": 0.0006201974873173443,
      "loss": 0.6494,
      "num_input_tokens_seen": 11968000,
      "step": 18335
    },
    {
      "epoch": 9.61215932914046,
      "grad_norm": 0.12205996364355087,
      "learning_rate": 0.0006199754952271847,
      "loss": 0.5862,
      "num_input_tokens_seen": 11970912,
      "step": 18340
    },
    {
      "epoch": 9.614779874213836,
      "grad_norm": 0.07925526797771454,
      "learning_rate": 0.00061975347803952,
      "loss": 0.3364,
      "num_input_tokens_seen": 11973472,
      "step": 18345
    },
    {
      "epoch": 9.617400419287211,
      "grad_norm": 0.11883335560560226,
      "learning_rate": 0.000619531435800794,
      "loss": 0.4695,
      "num_input_tokens_seen": 11976864,
      "step": 18350
    },
    {
      "epoch": 9.620020964360586,
      "grad_norm": 0.1211124137043953,
      "learning_rate": 0.0006193093685574553,
      "loss": 0.5272,
      "num_input_tokens_seen": 11979872,
      "step": 18355
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 0.06921296566724777,
      "learning_rate": 0.0006190872763559578,
      "loss": 0.4272,
      "num_input_tokens_seen": 11983168,
      "step": 18360
    },
    {
      "epoch": 9.625262054507338,
      "grad_norm": 0.09989932924509048,
      "learning_rate": 0.0006188651592427609,
      "loss": 0.4552,
      "num_input_tokens_seen": 11986432,
      "step": 18365
    },
    {
      "epoch": 9.627882599580714,
      "grad_norm": 0.11631099134683609,
      "learning_rate": 0.0006186430172643287,
      "loss": 0.2722,
      "num_input_tokens_seen": 11988512,
      "step": 18370
    },
    {
      "epoch": 9.630503144654089,
      "grad_norm": 0.11364596337080002,
      "learning_rate": 0.0006184208504671307,
      "loss": 0.6735,
      "num_input_tokens_seen": 11991424,
      "step": 18375
    },
    {
      "epoch": 9.633123689727464,
      "grad_norm": 0.10052991658449173,
      "learning_rate": 0.000618198658897642,
      "loss": 0.4499,
      "num_input_tokens_seen": 11994208,
      "step": 18380
    },
    {
      "epoch": 9.635744234800839,
      "grad_norm": 0.10961023718118668,
      "learning_rate": 0.0006179764426023422,
      "loss": 0.3819,
      "num_input_tokens_seen": 11997312,
      "step": 18385
    },
    {
      "epoch": 9.638364779874214,
      "grad_norm": 0.15845869481563568,
      "learning_rate": 0.0006177542016277167,
      "loss": 0.5764,
      "num_input_tokens_seen": 12000288,
      "step": 18390
    },
    {
      "epoch": 9.64098532494759,
      "grad_norm": 0.12113384902477264,
      "learning_rate": 0.0006175319360202556,
      "loss": 0.5718,
      "num_input_tokens_seen": 12003328,
      "step": 18395
    },
    {
      "epoch": 9.643605870020965,
      "grad_norm": 0.18900777399539948,
      "learning_rate": 0.0006173096458264542,
      "loss": 0.4819,
      "num_input_tokens_seen": 12006176,
      "step": 18400
    },
    {
      "epoch": 9.64622641509434,
      "grad_norm": 0.1098821684718132,
      "learning_rate": 0.0006170873310928133,
      "loss": 0.5552,
      "num_input_tokens_seen": 12009856,
      "step": 18405
    },
    {
      "epoch": 9.648846960167715,
      "grad_norm": 0.06804011017084122,
      "learning_rate": 0.0006168649918658387,
      "loss": 0.3497,
      "num_input_tokens_seen": 12013376,
      "step": 18410
    },
    {
      "epoch": 9.65146750524109,
      "grad_norm": 0.12452792376279831,
      "learning_rate": 0.0006166426281920407,
      "loss": 0.6078,
      "num_input_tokens_seen": 12016192,
      "step": 18415
    },
    {
      "epoch": 9.654088050314465,
      "grad_norm": 0.09119947999715805,
      "learning_rate": 0.000616420240117936,
      "loss": 0.5751,
      "num_input_tokens_seen": 12019680,
      "step": 18420
    },
    {
      "epoch": 9.65670859538784,
      "grad_norm": 0.06813053041696548,
      "learning_rate": 0.000616197827690045,
      "loss": 0.4015,
      "num_input_tokens_seen": 12023200,
      "step": 18425
    },
    {
      "epoch": 9.659329140461216,
      "grad_norm": 0.13375698029994965,
      "learning_rate": 0.0006159753909548943,
      "loss": 0.5367,
      "num_input_tokens_seen": 12026880,
      "step": 18430
    },
    {
      "epoch": 9.66194968553459,
      "grad_norm": 0.09013836830854416,
      "learning_rate": 0.0006157529299590148,
      "loss": 0.418,
      "num_input_tokens_seen": 12029472,
      "step": 18435
    },
    {
      "epoch": 9.664570230607966,
      "grad_norm": 0.11723936349153519,
      "learning_rate": 0.000615530444748943,
      "loss": 0.4814,
      "num_input_tokens_seen": 12032128,
      "step": 18440
    },
    {
      "epoch": 9.667190775681341,
      "grad_norm": 0.07440344244241714,
      "learning_rate": 0.0006153079353712201,
      "loss": 0.3827,
      "num_input_tokens_seen": 12034912,
      "step": 18445
    },
    {
      "epoch": 9.669811320754716,
      "grad_norm": 0.1080731749534607,
      "learning_rate": 0.0006150854018723926,
      "loss": 0.4152,
      "num_input_tokens_seen": 12037888,
      "step": 18450
    },
    {
      "epoch": 9.672431865828091,
      "grad_norm": 0.15916496515274048,
      "learning_rate": 0.0006148628442990122,
      "loss": 0.3922,
      "num_input_tokens_seen": 12040576,
      "step": 18455
    },
    {
      "epoch": 9.675052410901468,
      "grad_norm": 0.08788592368364334,
      "learning_rate": 0.0006146402626976351,
      "loss": 0.5451,
      "num_input_tokens_seen": 12044096,
      "step": 18460
    },
    {
      "epoch": 9.677672955974844,
      "grad_norm": 0.12198442220687866,
      "learning_rate": 0.0006144176571148228,
      "loss": 0.3695,
      "num_input_tokens_seen": 12047424,
      "step": 18465
    },
    {
      "epoch": 9.680293501048219,
      "grad_norm": 0.1002492755651474,
      "learning_rate": 0.0006141950275971422,
      "loss": 0.3486,
      "num_input_tokens_seen": 12050560,
      "step": 18470
    },
    {
      "epoch": 9.682914046121594,
      "grad_norm": 0.11740448325872421,
      "learning_rate": 0.0006139723741911644,
      "loss": 0.3983,
      "num_input_tokens_seen": 12053952,
      "step": 18475
    },
    {
      "epoch": 9.685534591194969,
      "grad_norm": 0.12544584274291992,
      "learning_rate": 0.0006137496969434664,
      "loss": 0.4093,
      "num_input_tokens_seen": 12056416,
      "step": 18480
    },
    {
      "epoch": 9.688155136268344,
      "grad_norm": 0.10500558465719223,
      "learning_rate": 0.0006135269959006294,
      "loss": 0.4725,
      "num_input_tokens_seen": 12060448,
      "step": 18485
    },
    {
      "epoch": 9.69077568134172,
      "grad_norm": 0.07190445810556412,
      "learning_rate": 0.00061330427110924,
      "loss": 0.5203,
      "num_input_tokens_seen": 12064960,
      "step": 18490
    },
    {
      "epoch": 9.693396226415095,
      "grad_norm": 0.07614414393901825,
      "learning_rate": 0.0006130815226158899,
      "loss": 0.3429,
      "num_input_tokens_seen": 12068192,
      "step": 18495
    },
    {
      "epoch": 9.69601677148847,
      "grad_norm": 0.16732434928417206,
      "learning_rate": 0.0006128587504671753,
      "loss": 0.3714,
      "num_input_tokens_seen": 12071200,
      "step": 18500
    },
    {
      "epoch": 9.698637316561845,
      "grad_norm": 0.16095376014709473,
      "learning_rate": 0.0006126359547096974,
      "loss": 0.4723,
      "num_input_tokens_seen": 12074112,
      "step": 18505
    },
    {
      "epoch": 9.70125786163522,
      "grad_norm": 0.30461394786834717,
      "learning_rate": 0.0006124131353900632,
      "loss": 0.5165,
      "num_input_tokens_seen": 12078048,
      "step": 18510
    },
    {
      "epoch": 9.703878406708595,
      "grad_norm": 0.11740636825561523,
      "learning_rate": 0.0006121902925548833,
      "loss": 0.5558,
      "num_input_tokens_seen": 12080896,
      "step": 18515
    },
    {
      "epoch": 9.70649895178197,
      "grad_norm": 0.08166959881782532,
      "learning_rate": 0.0006119674262507743,
      "loss": 0.3515,
      "num_input_tokens_seen": 12083744,
      "step": 18520
    },
    {
      "epoch": 9.709119496855346,
      "grad_norm": 0.0894090011715889,
      "learning_rate": 0.0006117445365243571,
      "loss": 0.4229,
      "num_input_tokens_seen": 12086496,
      "step": 18525
    },
    {
      "epoch": 9.71174004192872,
      "grad_norm": 0.0805690735578537,
      "learning_rate": 0.0006115216234222575,
      "loss": 0.3784,
      "num_input_tokens_seen": 12089728,
      "step": 18530
    },
    {
      "epoch": 9.714360587002096,
      "grad_norm": 0.14835940301418304,
      "learning_rate": 0.0006112986869911068,
      "loss": 0.588,
      "num_input_tokens_seen": 12093120,
      "step": 18535
    },
    {
      "epoch": 9.716981132075471,
      "grad_norm": 0.10832329839468002,
      "learning_rate": 0.0006110757272775405,
      "loss": 0.5321,
      "num_input_tokens_seen": 12095680,
      "step": 18540
    },
    {
      "epoch": 9.719601677148846,
      "grad_norm": 0.12017285078763962,
      "learning_rate": 0.0006108527443281992,
      "loss": 0.4555,
      "num_input_tokens_seen": 12099040,
      "step": 18545
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 0.05609551817178726,
      "learning_rate": 0.0006106297381897284,
      "loss": 0.3461,
      "num_input_tokens_seen": 12102112,
      "step": 18550
    },
    {
      "epoch": 9.724842767295598,
      "grad_norm": 0.10099728405475616,
      "learning_rate": 0.0006104067089087787,
      "loss": 0.4207,
      "num_input_tokens_seen": 12104544,
      "step": 18555
    },
    {
      "epoch": 9.727463312368974,
      "grad_norm": 0.11995375156402588,
      "learning_rate": 0.0006101836565320049,
      "loss": 0.4046,
      "num_input_tokens_seen": 12107712,
      "step": 18560
    },
    {
      "epoch": 9.730083857442349,
      "grad_norm": 0.08250979334115982,
      "learning_rate": 0.0006099605811060673,
      "loss": 0.3937,
      "num_input_tokens_seen": 12110528,
      "step": 18565
    },
    {
      "epoch": 9.732704402515724,
      "grad_norm": 0.1195971816778183,
      "learning_rate": 0.0006097374826776307,
      "loss": 0.4136,
      "num_input_tokens_seen": 12113600,
      "step": 18570
    },
    {
      "epoch": 9.735324947589099,
      "grad_norm": 0.07699618488550186,
      "learning_rate": 0.0006095143612933646,
      "loss": 0.4735,
      "num_input_tokens_seen": 12116896,
      "step": 18575
    },
    {
      "epoch": 9.737945492662474,
      "grad_norm": 0.06932520121335983,
      "learning_rate": 0.0006092912169999435,
      "loss": 0.3625,
      "num_input_tokens_seen": 12120832,
      "step": 18580
    },
    {
      "epoch": 9.74056603773585,
      "grad_norm": 0.05380554124712944,
      "learning_rate": 0.0006090680498440465,
      "loss": 0.3092,
      "num_input_tokens_seen": 12124704,
      "step": 18585
    },
    {
      "epoch": 9.743186582809225,
      "grad_norm": 0.13912983238697052,
      "learning_rate": 0.0006088448598723581,
      "loss": 0.3728,
      "num_input_tokens_seen": 12129760,
      "step": 18590
    },
    {
      "epoch": 9.7458071278826,
      "grad_norm": 0.21503517031669617,
      "learning_rate": 0.0006086216471315666,
      "loss": 0.4338,
      "num_input_tokens_seen": 12132096,
      "step": 18595
    },
    {
      "epoch": 9.748427672955975,
      "grad_norm": 0.09368952363729477,
      "learning_rate": 0.0006083984116683659,
      "loss": 0.5666,
      "num_input_tokens_seen": 12135008,
      "step": 18600
    },
    {
      "epoch": 9.75104821802935,
      "grad_norm": 0.13351304829120636,
      "learning_rate": 0.0006081751535294539,
      "loss": 0.4746,
      "num_input_tokens_seen": 12138144,
      "step": 18605
    },
    {
      "epoch": 9.753668763102725,
      "grad_norm": 0.19211892783641815,
      "learning_rate": 0.0006079518727615341,
      "loss": 0.5664,
      "num_input_tokens_seen": 12141568,
      "step": 18610
    },
    {
      "epoch": 9.7562893081761,
      "grad_norm": 0.09067568182945251,
      "learning_rate": 0.0006077285694113141,
      "loss": 0.5533,
      "num_input_tokens_seen": 12145248,
      "step": 18615
    },
    {
      "epoch": 9.758909853249476,
      "grad_norm": 0.0824904590845108,
      "learning_rate": 0.0006075052435255062,
      "loss": 0.4043,
      "num_input_tokens_seen": 12149472,
      "step": 18620
    },
    {
      "epoch": 9.76153039832285,
      "grad_norm": 0.07659609615802765,
      "learning_rate": 0.000607281895150828,
      "loss": 0.5295,
      "num_input_tokens_seen": 12152992,
      "step": 18625
    },
    {
      "epoch": 9.764150943396226,
      "grad_norm": 0.17156188189983368,
      "learning_rate": 0.0006070585243340013,
      "loss": 0.4199,
      "num_input_tokens_seen": 12155712,
      "step": 18630
    },
    {
      "epoch": 9.766771488469601,
      "grad_norm": 0.09780377149581909,
      "learning_rate": 0.0006068351311217526,
      "loss": 0.509,
      "num_input_tokens_seen": 12158688,
      "step": 18635
    },
    {
      "epoch": 9.769392033542976,
      "grad_norm": 0.14882643520832062,
      "learning_rate": 0.0006066117155608135,
      "loss": 0.7768,
      "num_input_tokens_seen": 12161824,
      "step": 18640
    },
    {
      "epoch": 9.772012578616351,
      "grad_norm": 0.10034777969121933,
      "learning_rate": 0.0006063882776979196,
      "loss": 0.3761,
      "num_input_tokens_seen": 12164992,
      "step": 18645
    },
    {
      "epoch": 9.774633123689728,
      "grad_norm": 0.09783245623111725,
      "learning_rate": 0.0006061648175798117,
      "loss": 0.4157,
      "num_input_tokens_seen": 12167808,
      "step": 18650
    },
    {
      "epoch": 9.777253668763104,
      "grad_norm": 0.09324688464403152,
      "learning_rate": 0.0006059413352532355,
      "loss": 0.2947,
      "num_input_tokens_seen": 12170752,
      "step": 18655
    },
    {
      "epoch": 9.779874213836479,
      "grad_norm": 0.16405430436134338,
      "learning_rate": 0.0006057178307649402,
      "loss": 0.3659,
      "num_input_tokens_seen": 12173664,
      "step": 18660
    },
    {
      "epoch": 9.782494758909854,
      "grad_norm": 0.0881318598985672,
      "learning_rate": 0.000605494304161681,
      "loss": 0.4209,
      "num_input_tokens_seen": 12176960,
      "step": 18665
    },
    {
      "epoch": 9.785115303983229,
      "grad_norm": 0.12354959547519684,
      "learning_rate": 0.0006052707554902169,
      "loss": 0.6249,
      "num_input_tokens_seen": 12180096,
      "step": 18670
    },
    {
      "epoch": 9.787735849056604,
      "grad_norm": 0.14208810031414032,
      "learning_rate": 0.0006050471847973117,
      "loss": 0.3821,
      "num_input_tokens_seen": 12183424,
      "step": 18675
    },
    {
      "epoch": 9.79035639412998,
      "grad_norm": 0.2954697906970978,
      "learning_rate": 0.000604823592129734,
      "loss": 0.3832,
      "num_input_tokens_seen": 12186592,
      "step": 18680
    },
    {
      "epoch": 9.792976939203355,
      "grad_norm": 0.12592048943042755,
      "learning_rate": 0.0006045999775342566,
      "loss": 0.5218,
      "num_input_tokens_seen": 12190240,
      "step": 18685
    },
    {
      "epoch": 9.79559748427673,
      "grad_norm": 0.1758650839328766,
      "learning_rate": 0.0006043763410576572,
      "loss": 0.4451,
      "num_input_tokens_seen": 12193024,
      "step": 18690
    },
    {
      "epoch": 9.798218029350105,
      "grad_norm": 0.07931248843669891,
      "learning_rate": 0.0006041526827467182,
      "loss": 0.4681,
      "num_input_tokens_seen": 12196128,
      "step": 18695
    },
    {
      "epoch": 9.80083857442348,
      "grad_norm": 0.08574016392230988,
      "learning_rate": 0.000603929002648226,
      "loss": 0.4223,
      "num_input_tokens_seen": 12198880,
      "step": 18700
    },
    {
      "epoch": 9.803459119496855,
      "grad_norm": 0.1086101159453392,
      "learning_rate": 0.0006037053008089723,
      "loss": 0.4131,
      "num_input_tokens_seen": 12201760,
      "step": 18705
    },
    {
      "epoch": 9.80607966457023,
      "grad_norm": 0.04840122163295746,
      "learning_rate": 0.0006034815772757528,
      "loss": 0.6875,
      "num_input_tokens_seen": 12206272,
      "step": 18710
    },
    {
      "epoch": 9.808700209643606,
      "grad_norm": 0.08213505148887634,
      "learning_rate": 0.0006032578320953678,
      "loss": 0.4297,
      "num_input_tokens_seen": 12209472,
      "step": 18715
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 0.11168914288282394,
      "learning_rate": 0.0006030340653146224,
      "loss": 0.4917,
      "num_input_tokens_seen": 12212256,
      "step": 18720
    },
    {
      "epoch": 9.813941299790356,
      "grad_norm": 0.07238415628671646,
      "learning_rate": 0.0006028102769803261,
      "loss": 0.5005,
      "num_input_tokens_seen": 12215456,
      "step": 18725
    },
    {
      "epoch": 9.816561844863731,
      "grad_norm": 0.06875133514404297,
      "learning_rate": 0.0006025864671392925,
      "loss": 0.4398,
      "num_input_tokens_seen": 12219936,
      "step": 18730
    },
    {
      "epoch": 9.819182389937106,
      "grad_norm": 0.12057217955589294,
      "learning_rate": 0.0006023626358383406,
      "loss": 0.4216,
      "num_input_tokens_seen": 12223200,
      "step": 18735
    },
    {
      "epoch": 9.821802935010481,
      "grad_norm": 0.13823236525058746,
      "learning_rate": 0.0006021387831242932,
      "loss": 0.4316,
      "num_input_tokens_seen": 12226432,
      "step": 18740
    },
    {
      "epoch": 9.824423480083858,
      "grad_norm": 0.07446273416280746,
      "learning_rate": 0.0006019149090439774,
      "loss": 0.4066,
      "num_input_tokens_seen": 12229120,
      "step": 18745
    },
    {
      "epoch": 9.827044025157234,
      "grad_norm": 0.044040609151124954,
      "learning_rate": 0.0006016910136442254,
      "loss": 0.3649,
      "num_input_tokens_seen": 12233248,
      "step": 18750
    },
    {
      "epoch": 9.829664570230609,
      "grad_norm": 0.05637265369296074,
      "learning_rate": 0.0006014670969718735,
      "loss": 0.5277,
      "num_input_tokens_seen": 12237696,
      "step": 18755
    },
    {
      "epoch": 9.832285115303984,
      "grad_norm": 0.11674612015485764,
      "learning_rate": 0.0006012431590737625,
      "loss": 0.3133,
      "num_input_tokens_seen": 12241504,
      "step": 18760
    },
    {
      "epoch": 9.834905660377359,
      "grad_norm": 0.2120673954486847,
      "learning_rate": 0.0006010191999967378,
      "loss": 0.4546,
      "num_input_tokens_seen": 12244832,
      "step": 18765
    },
    {
      "epoch": 9.837526205450734,
      "grad_norm": 0.13125476241111755,
      "learning_rate": 0.0006007952197876487,
      "loss": 0.425,
      "num_input_tokens_seen": 12247776,
      "step": 18770
    },
    {
      "epoch": 9.84014675052411,
      "grad_norm": 0.13294990360736847,
      "learning_rate": 0.0006005712184933498,
      "loss": 0.4224,
      "num_input_tokens_seen": 12250560,
      "step": 18775
    },
    {
      "epoch": 9.842767295597485,
      "grad_norm": 0.09935607761144638,
      "learning_rate": 0.0006003471961606992,
      "loss": 0.5179,
      "num_input_tokens_seen": 12253504,
      "step": 18780
    },
    {
      "epoch": 9.84538784067086,
      "grad_norm": 0.1274244487285614,
      "learning_rate": 0.0006001231528365601,
      "loss": 0.4299,
      "num_input_tokens_seen": 12256128,
      "step": 18785
    },
    {
      "epoch": 9.848008385744235,
      "grad_norm": 0.06703843176364899,
      "learning_rate": 0.0005998990885677996,
      "loss": 0.5638,
      "num_input_tokens_seen": 12260032,
      "step": 18790
    },
    {
      "epoch": 9.85062893081761,
      "grad_norm": 0.07643602788448334,
      "learning_rate": 0.0005996750034012895,
      "loss": 0.3921,
      "num_input_tokens_seen": 12263136,
      "step": 18795
    },
    {
      "epoch": 9.853249475890985,
      "grad_norm": 0.2622632682323456,
      "learning_rate": 0.0005994508973839058,
      "loss": 0.4975,
      "num_input_tokens_seen": 12266496,
      "step": 18800
    },
    {
      "epoch": 9.85587002096436,
      "grad_norm": 0.1338152438402176,
      "learning_rate": 0.0005992267705625289,
      "loss": 0.3803,
      "num_input_tokens_seen": 12268768,
      "step": 18805
    },
    {
      "epoch": 9.858490566037736,
      "grad_norm": 0.06733916699886322,
      "learning_rate": 0.0005990026229840439,
      "loss": 0.4624,
      "num_input_tokens_seen": 12272160,
      "step": 18810
    },
    {
      "epoch": 9.86111111111111,
      "grad_norm": 0.13199879229068756,
      "learning_rate": 0.0005987784546953394,
      "loss": 0.5475,
      "num_input_tokens_seen": 12275776,
      "step": 18815
    },
    {
      "epoch": 9.863731656184486,
      "grad_norm": 0.17509163916110992,
      "learning_rate": 0.0005985542657433091,
      "loss": 0.4295,
      "num_input_tokens_seen": 12278880,
      "step": 18820
    },
    {
      "epoch": 9.866352201257861,
      "grad_norm": 0.1031404510140419,
      "learning_rate": 0.000598330056174851,
      "loss": 0.4333,
      "num_input_tokens_seen": 12281536,
      "step": 18825
    },
    {
      "epoch": 9.868972746331236,
      "grad_norm": 0.07519888877868652,
      "learning_rate": 0.0005981058260368668,
      "loss": 0.5535,
      "num_input_tokens_seen": 12284416,
      "step": 18830
    },
    {
      "epoch": 9.871593291404611,
      "grad_norm": 0.0541648305952549,
      "learning_rate": 0.0005978815753762632,
      "loss": 0.4251,
      "num_input_tokens_seen": 12288224,
      "step": 18835
    },
    {
      "epoch": 9.874213836477988,
      "grad_norm": 0.09710607677698135,
      "learning_rate": 0.0005976573042399505,
      "loss": 0.38,
      "num_input_tokens_seen": 12291520,
      "step": 18840
    },
    {
      "epoch": 9.876834381551364,
      "grad_norm": 0.13575471937656403,
      "learning_rate": 0.000597433012674844,
      "loss": 0.5355,
      "num_input_tokens_seen": 12295264,
      "step": 18845
    },
    {
      "epoch": 9.879454926624739,
      "grad_norm": 0.07740969210863113,
      "learning_rate": 0.0005972087007278628,
      "loss": 0.4725,
      "num_input_tokens_seen": 12297984,
      "step": 18850
    },
    {
      "epoch": 9.882075471698114,
      "grad_norm": 0.1414135992527008,
      "learning_rate": 0.0005969843684459305,
      "loss": 0.4364,
      "num_input_tokens_seen": 12301184,
      "step": 18855
    },
    {
      "epoch": 9.884696016771489,
      "grad_norm": 0.0971233919262886,
      "learning_rate": 0.0005967600158759746,
      "loss": 0.4844,
      "num_input_tokens_seen": 12303968,
      "step": 18860
    },
    {
      "epoch": 9.887316561844864,
      "grad_norm": 0.10748016834259033,
      "learning_rate": 0.0005965356430649276,
      "loss": 0.3864,
      "num_input_tokens_seen": 12308256,
      "step": 18865
    },
    {
      "epoch": 9.88993710691824,
      "grad_norm": 0.09817912429571152,
      "learning_rate": 0.0005963112500597251,
      "loss": 0.5426,
      "num_input_tokens_seen": 12310816,
      "step": 18870
    },
    {
      "epoch": 9.892557651991615,
      "grad_norm": 0.07424537092447281,
      "learning_rate": 0.0005960868369073079,
      "loss": 0.4646,
      "num_input_tokens_seen": 12314752,
      "step": 18875
    },
    {
      "epoch": 9.89517819706499,
      "grad_norm": 0.1029123067855835,
      "learning_rate": 0.0005958624036546208,
      "loss": 0.4344,
      "num_input_tokens_seen": 12318272,
      "step": 18880
    },
    {
      "epoch": 9.897798742138365,
      "grad_norm": 0.09936170279979706,
      "learning_rate": 0.0005956379503486123,
      "loss": 0.5129,
      "num_input_tokens_seen": 12322336,
      "step": 18885
    },
    {
      "epoch": 9.90041928721174,
      "grad_norm": 0.07782087475061417,
      "learning_rate": 0.0005954134770362359,
      "loss": 0.4271,
      "num_input_tokens_seen": 12325760,
      "step": 18890
    },
    {
      "epoch": 9.903039832285115,
      "grad_norm": 0.11796512454748154,
      "learning_rate": 0.0005951889837644486,
      "loss": 0.5712,
      "num_input_tokens_seen": 12330304,
      "step": 18895
    },
    {
      "epoch": 9.90566037735849,
      "grad_norm": 0.13942411541938782,
      "learning_rate": 0.0005949644705802116,
      "loss": 0.4956,
      "num_input_tokens_seen": 12332960,
      "step": 18900
    },
    {
      "epoch": 9.908280922431866,
      "grad_norm": 0.08785610646009445,
      "learning_rate": 0.0005947399375304909,
      "loss": 0.5079,
      "num_input_tokens_seen": 12337536,
      "step": 18905
    },
    {
      "epoch": 9.91090146750524,
      "grad_norm": 0.1980201154947281,
      "learning_rate": 0.0005945153846622563,
      "loss": 0.4888,
      "num_input_tokens_seen": 12341120,
      "step": 18910
    },
    {
      "epoch": 9.913522012578616,
      "grad_norm": 0.10234901309013367,
      "learning_rate": 0.0005942908120224815,
      "loss": 0.4361,
      "num_input_tokens_seen": 12343904,
      "step": 18915
    },
    {
      "epoch": 9.916142557651991,
      "grad_norm": 0.1336854100227356,
      "learning_rate": 0.0005940662196581443,
      "loss": 0.5573,
      "num_input_tokens_seen": 12346912,
      "step": 18920
    },
    {
      "epoch": 9.918763102725366,
      "grad_norm": 0.17978161573410034,
      "learning_rate": 0.0005938416076162273,
      "loss": 0.3607,
      "num_input_tokens_seen": 12350272,
      "step": 18925
    },
    {
      "epoch": 9.921383647798741,
      "grad_norm": 0.2909812927246094,
      "learning_rate": 0.0005936169759437165,
      "loss": 0.536,
      "num_input_tokens_seen": 12352800,
      "step": 18930
    },
    {
      "epoch": 9.924004192872118,
      "grad_norm": 0.06889134645462036,
      "learning_rate": 0.0005933923246876024,
      "loss": 0.3854,
      "num_input_tokens_seen": 12355616,
      "step": 18935
    },
    {
      "epoch": 9.926624737945493,
      "grad_norm": 0.20294572412967682,
      "learning_rate": 0.0005931676538948795,
      "loss": 0.4273,
      "num_input_tokens_seen": 12358784,
      "step": 18940
    },
    {
      "epoch": 9.929245283018869,
      "grad_norm": 0.1157127320766449,
      "learning_rate": 0.0005929429636125463,
      "loss": 0.5075,
      "num_input_tokens_seen": 12361920,
      "step": 18945
    },
    {
      "epoch": 9.931865828092244,
      "grad_norm": 0.10965290665626526,
      "learning_rate": 0.0005927182538876055,
      "loss": 0.4482,
      "num_input_tokens_seen": 12365600,
      "step": 18950
    },
    {
      "epoch": 9.934486373165619,
      "grad_norm": 0.11111999303102493,
      "learning_rate": 0.0005924935247670637,
      "loss": 0.487,
      "num_input_tokens_seen": 12368000,
      "step": 18955
    },
    {
      "epoch": 9.937106918238994,
      "grad_norm": 0.09175273030996323,
      "learning_rate": 0.0005922687762979317,
      "loss": 0.4604,
      "num_input_tokens_seen": 12370592,
      "step": 18960
    },
    {
      "epoch": 9.93972746331237,
      "grad_norm": 0.10344359278678894,
      "learning_rate": 0.0005920440085272245,
      "loss": 0.4969,
      "num_input_tokens_seen": 12373760,
      "step": 18965
    },
    {
      "epoch": 9.942348008385745,
      "grad_norm": 0.11334064602851868,
      "learning_rate": 0.0005918192215019608,
      "loss": 0.4631,
      "num_input_tokens_seen": 12377600,
      "step": 18970
    },
    {
      "epoch": 9.94496855345912,
      "grad_norm": 0.1961027979850769,
      "learning_rate": 0.0005915944152691633,
      "loss": 0.387,
      "num_input_tokens_seen": 12381152,
      "step": 18975
    },
    {
      "epoch": 9.947589098532495,
      "grad_norm": 0.13103197515010834,
      "learning_rate": 0.0005913695898758593,
      "loss": 0.4809,
      "num_input_tokens_seen": 12384096,
      "step": 18980
    },
    {
      "epoch": 9.95020964360587,
      "grad_norm": 0.166547030210495,
      "learning_rate": 0.0005911447453690793,
      "loss": 0.4295,
      "num_input_tokens_seen": 12386848,
      "step": 18985
    },
    {
      "epoch": 9.952830188679245,
      "grad_norm": 0.15184731781482697,
      "learning_rate": 0.0005909198817958585,
      "loss": 0.4192,
      "num_input_tokens_seen": 12389344,
      "step": 18990
    },
    {
      "epoch": 9.95545073375262,
      "grad_norm": 0.07724333554506302,
      "learning_rate": 0.0005906949992032359,
      "loss": 0.4808,
      "num_input_tokens_seen": 12392640,
      "step": 18995
    },
    {
      "epoch": 9.958071278825996,
      "grad_norm": 0.3020954728126526,
      "learning_rate": 0.0005904700976382539,
      "loss": 0.4897,
      "num_input_tokens_seen": 12397248,
      "step": 19000
    },
    {
      "epoch": 9.96069182389937,
      "grad_norm": 0.11391711235046387,
      "learning_rate": 0.0005902451771479596,
      "loss": 0.4997,
      "num_input_tokens_seen": 12400352,
      "step": 19005
    },
    {
      "epoch": 9.963312368972746,
      "grad_norm": 0.11348896473646164,
      "learning_rate": 0.000590020237779404,
      "loss": 0.3841,
      "num_input_tokens_seen": 12402944,
      "step": 19010
    },
    {
      "epoch": 9.965932914046121,
      "grad_norm": 0.16243311762809753,
      "learning_rate": 0.0005897952795796415,
      "loss": 0.4477,
      "num_input_tokens_seen": 12405952,
      "step": 19015
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 0.08958359807729721,
      "learning_rate": 0.0005895703025957311,
      "loss": 0.6097,
      "num_input_tokens_seen": 12409376,
      "step": 19020
    },
    {
      "epoch": 9.971174004192871,
      "grad_norm": 0.10659455507993698,
      "learning_rate": 0.0005893453068747352,
      "loss": 0.4862,
      "num_input_tokens_seen": 12413216,
      "step": 19025
    },
    {
      "epoch": 9.973794549266248,
      "grad_norm": 0.11029645055532455,
      "learning_rate": 0.0005891202924637202,
      "loss": 0.3868,
      "num_input_tokens_seen": 12416512,
      "step": 19030
    },
    {
      "epoch": 9.976415094339622,
      "grad_norm": 0.06715545058250427,
      "learning_rate": 0.0005888952594097572,
      "loss": 0.4529,
      "num_input_tokens_seen": 12420192,
      "step": 19035
    },
    {
      "epoch": 9.979035639412999,
      "grad_norm": 0.4405798614025116,
      "learning_rate": 0.0005886702077599197,
      "loss": 0.5173,
      "num_input_tokens_seen": 12422464,
      "step": 19040
    },
    {
      "epoch": 9.981656184486374,
      "grad_norm": 0.10312025249004364,
      "learning_rate": 0.0005884451375612865,
      "loss": 0.5932,
      "num_input_tokens_seen": 12425248,
      "step": 19045
    },
    {
      "epoch": 9.984276729559749,
      "grad_norm": 0.13083156943321228,
      "learning_rate": 0.0005882200488609396,
      "loss": 0.4408,
      "num_input_tokens_seen": 12428704,
      "step": 19050
    },
    {
      "epoch": 9.986897274633124,
      "grad_norm": 0.13115350902080536,
      "learning_rate": 0.0005879949417059649,
      "loss": 0.5296,
      "num_input_tokens_seen": 12432480,
      "step": 19055
    },
    {
      "epoch": 9.9895178197065,
      "grad_norm": 0.10692878067493439,
      "learning_rate": 0.0005877698161434522,
      "loss": 0.3891,
      "num_input_tokens_seen": 12434912,
      "step": 19060
    },
    {
      "epoch": 9.992138364779874,
      "grad_norm": 0.18419738113880157,
      "learning_rate": 0.0005875446722204955,
      "loss": 0.405,
      "num_input_tokens_seen": 12438016,
      "step": 19065
    },
    {
      "epoch": 9.99475890985325,
      "grad_norm": 0.10560203343629837,
      "learning_rate": 0.0005873195099841919,
      "loss": 0.5152,
      "num_input_tokens_seen": 12440672,
      "step": 19070
    },
    {
      "epoch": 9.997379454926625,
      "grad_norm": 0.10719730705022812,
      "learning_rate": 0.0005870943294816431,
      "loss": 0.4651,
      "num_input_tokens_seen": 12443840,
      "step": 19075
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.14405983686447144,
      "learning_rate": 0.0005868691307599543,
      "loss": 0.4859,
      "num_input_tokens_seen": 12446408,
      "step": 19080
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.4540214240550995,
      "eval_runtime": 13.5699,
      "eval_samples_per_second": 62.491,
      "eval_steps_per_second": 15.623,
      "num_input_tokens_seen": 12446408,
      "step": 19080
    },
    {
      "epoch": 10.002620545073375,
      "grad_norm": 0.06634771823883057,
      "learning_rate": 0.0005866439138662343,
      "loss": 0.4087,
      "num_input_tokens_seen": 12450472,
      "step": 19085
    },
    {
      "epoch": 10.00524109014675,
      "grad_norm": 0.11630714684724808,
      "learning_rate": 0.000586418678847596,
      "loss": 0.3767,
      "num_input_tokens_seen": 12454056,
      "step": 19090
    },
    {
      "epoch": 10.007861635220126,
      "grad_norm": 0.10477319359779358,
      "learning_rate": 0.000586193425751156,
      "loss": 0.4916,
      "num_input_tokens_seen": 12457576,
      "step": 19095
    },
    {
      "epoch": 10.0104821802935,
      "grad_norm": 0.1654532104730606,
      "learning_rate": 0.0005859681546240348,
      "loss": 0.4718,
      "num_input_tokens_seen": 12461608,
      "step": 19100
    },
    {
      "epoch": 10.013102725366876,
      "grad_norm": 0.0729355588555336,
      "learning_rate": 0.0005857428655133562,
      "loss": 0.3696,
      "num_input_tokens_seen": 12464968,
      "step": 19105
    },
    {
      "epoch": 10.015723270440251,
      "grad_norm": 0.1635524034500122,
      "learning_rate": 0.0005855175584662485,
      "loss": 0.5033,
      "num_input_tokens_seen": 12468648,
      "step": 19110
    },
    {
      "epoch": 10.018343815513626,
      "grad_norm": 0.0874512642621994,
      "learning_rate": 0.0005852922335298433,
      "loss": 0.4887,
      "num_input_tokens_seen": 12471784,
      "step": 19115
    },
    {
      "epoch": 10.020964360587001,
      "grad_norm": 0.12874650955200195,
      "learning_rate": 0.0005850668907512757,
      "loss": 0.5805,
      "num_input_tokens_seen": 12474632,
      "step": 19120
    },
    {
      "epoch": 10.023584905660377,
      "grad_norm": 0.08755774050951004,
      "learning_rate": 0.0005848415301776852,
      "loss": 0.5325,
      "num_input_tokens_seen": 12477832,
      "step": 19125
    },
    {
      "epoch": 10.026205450733753,
      "grad_norm": 0.08611950278282166,
      "learning_rate": 0.0005846161518562144,
      "loss": 0.4696,
      "num_input_tokens_seen": 12482312,
      "step": 19130
    },
    {
      "epoch": 10.028825995807129,
      "grad_norm": 0.07569777965545654,
      "learning_rate": 0.00058439075583401,
      "loss": 0.4191,
      "num_input_tokens_seen": 12485768,
      "step": 19135
    },
    {
      "epoch": 10.031446540880504,
      "grad_norm": 0.09477218240499496,
      "learning_rate": 0.0005841653421582223,
      "loss": 0.3554,
      "num_input_tokens_seen": 12488776,
      "step": 19140
    },
    {
      "epoch": 10.034067085953879,
      "grad_norm": 0.096603624522686,
      "learning_rate": 0.0005839399108760052,
      "loss": 0.4598,
      "num_input_tokens_seen": 12491464,
      "step": 19145
    },
    {
      "epoch": 10.036687631027254,
      "grad_norm": 0.0809653177857399,
      "learning_rate": 0.0005837144620345165,
      "loss": 0.4704,
      "num_input_tokens_seen": 12494600,
      "step": 19150
    },
    {
      "epoch": 10.03930817610063,
      "grad_norm": 0.09892649203538895,
      "learning_rate": 0.0005834889956809173,
      "loss": 0.6288,
      "num_input_tokens_seen": 12498696,
      "step": 19155
    },
    {
      "epoch": 10.041928721174004,
      "grad_norm": 0.13214993476867676,
      "learning_rate": 0.0005832635118623726,
      "loss": 0.4602,
      "num_input_tokens_seen": 12501960,
      "step": 19160
    },
    {
      "epoch": 10.04454926624738,
      "grad_norm": 0.10132061690092087,
      "learning_rate": 0.0005830380106260513,
      "loss": 0.599,
      "num_input_tokens_seen": 12505224,
      "step": 19165
    },
    {
      "epoch": 10.047169811320755,
      "grad_norm": 0.08133473992347717,
      "learning_rate": 0.0005828124920191255,
      "loss": 0.4869,
      "num_input_tokens_seen": 12508744,
      "step": 19170
    },
    {
      "epoch": 10.04979035639413,
      "grad_norm": 0.11904203146696091,
      "learning_rate": 0.000582586956088771,
      "loss": 0.5589,
      "num_input_tokens_seen": 12511464,
      "step": 19175
    },
    {
      "epoch": 10.052410901467505,
      "grad_norm": 0.1304342895746231,
      "learning_rate": 0.0005823614028821677,
      "loss": 0.4835,
      "num_input_tokens_seen": 12513736,
      "step": 19180
    },
    {
      "epoch": 10.05503144654088,
      "grad_norm": 0.09764691442251205,
      "learning_rate": 0.0005821358324464985,
      "loss": 0.4955,
      "num_input_tokens_seen": 12516584,
      "step": 19185
    },
    {
      "epoch": 10.057651991614255,
      "grad_norm": 0.08899445831775665,
      "learning_rate": 0.0005819102448289501,
      "loss": 0.368,
      "num_input_tokens_seen": 12520680,
      "step": 19190
    },
    {
      "epoch": 10.06027253668763,
      "grad_norm": 0.19170211255550385,
      "learning_rate": 0.000581684640076713,
      "loss": 0.4923,
      "num_input_tokens_seen": 12523624,
      "step": 19195
    },
    {
      "epoch": 10.062893081761006,
      "grad_norm": 0.136613130569458,
      "learning_rate": 0.0005814590182369811,
      "loss": 0.416,
      "num_input_tokens_seen": 12526472,
      "step": 19200
    },
    {
      "epoch": 10.065513626834381,
      "grad_norm": 0.0974559634923935,
      "learning_rate": 0.000581233379356952,
      "loss": 0.5109,
      "num_input_tokens_seen": 12529768,
      "step": 19205
    },
    {
      "epoch": 10.068134171907756,
      "grad_norm": 0.09385141730308533,
      "learning_rate": 0.0005810077234838267,
      "loss": 0.4464,
      "num_input_tokens_seen": 12532680,
      "step": 19210
    },
    {
      "epoch": 10.070754716981131,
      "grad_norm": 0.0895027369260788,
      "learning_rate": 0.0005807820506648097,
      "loss": 0.4204,
      "num_input_tokens_seen": 12535944,
      "step": 19215
    },
    {
      "epoch": 10.073375262054507,
      "grad_norm": 0.14507046341896057,
      "learning_rate": 0.0005805563609471095,
      "loss": 0.5609,
      "num_input_tokens_seen": 12540008,
      "step": 19220
    },
    {
      "epoch": 10.075995807127883,
      "grad_norm": 0.16880330443382263,
      "learning_rate": 0.0005803306543779375,
      "loss": 0.3555,
      "num_input_tokens_seen": 12542952,
      "step": 19225
    },
    {
      "epoch": 10.078616352201259,
      "grad_norm": 0.15857365727424622,
      "learning_rate": 0.000580104931004509,
      "loss": 0.406,
      "num_input_tokens_seen": 12545672,
      "step": 19230
    },
    {
      "epoch": 10.081236897274634,
      "grad_norm": 0.27583274245262146,
      "learning_rate": 0.000579879190874043,
      "loss": 0.4189,
      "num_input_tokens_seen": 12548808,
      "step": 19235
    },
    {
      "epoch": 10.083857442348009,
      "grad_norm": 0.21567663550376892,
      "learning_rate": 0.0005796534340337614,
      "loss": 0.4869,
      "num_input_tokens_seen": 12551944,
      "step": 19240
    },
    {
      "epoch": 10.086477987421384,
      "grad_norm": 0.12406596541404724,
      "learning_rate": 0.0005794276605308902,
      "loss": 0.4789,
      "num_input_tokens_seen": 12555048,
      "step": 19245
    },
    {
      "epoch": 10.08909853249476,
      "grad_norm": 0.09523241221904755,
      "learning_rate": 0.0005792018704126586,
      "loss": 0.376,
      "num_input_tokens_seen": 12558312,
      "step": 19250
    },
    {
      "epoch": 10.091719077568134,
      "grad_norm": 0.05898785591125488,
      "learning_rate": 0.0005789760637262992,
      "loss": 0.4625,
      "num_input_tokens_seen": 12561992,
      "step": 19255
    },
    {
      "epoch": 10.09433962264151,
      "grad_norm": 0.06680694967508316,
      "learning_rate": 0.0005787502405190483,
      "loss": 0.4762,
      "num_input_tokens_seen": 12565544,
      "step": 19260
    },
    {
      "epoch": 10.096960167714885,
      "grad_norm": 0.15346716344356537,
      "learning_rate": 0.0005785244008381456,
      "loss": 0.5176,
      "num_input_tokens_seen": 12568232,
      "step": 19265
    },
    {
      "epoch": 10.09958071278826,
      "grad_norm": 0.20479413866996765,
      "learning_rate": 0.0005782985447308341,
      "loss": 0.4323,
      "num_input_tokens_seen": 12570568,
      "step": 19270
    },
    {
      "epoch": 10.102201257861635,
      "grad_norm": 0.08803833276033401,
      "learning_rate": 0.0005780726722443602,
      "loss": 0.4278,
      "num_input_tokens_seen": 12573448,
      "step": 19275
    },
    {
      "epoch": 10.10482180293501,
      "grad_norm": 0.11271852999925613,
      "learning_rate": 0.0005778467834259741,
      "loss": 0.3918,
      "num_input_tokens_seen": 12577256,
      "step": 19280
    },
    {
      "epoch": 10.107442348008385,
      "grad_norm": 0.10874661058187485,
      "learning_rate": 0.000577620878322929,
      "loss": 0.422,
      "num_input_tokens_seen": 12580808,
      "step": 19285
    },
    {
      "epoch": 10.11006289308176,
      "grad_norm": 0.04990832507610321,
      "learning_rate": 0.0005773949569824816,
      "loss": 0.4448,
      "num_input_tokens_seen": 12585896,
      "step": 19290
    },
    {
      "epoch": 10.112683438155136,
      "grad_norm": 0.08391886204481125,
      "learning_rate": 0.0005771690194518922,
      "loss": 0.4093,
      "num_input_tokens_seen": 12589000,
      "step": 19295
    },
    {
      "epoch": 10.115303983228511,
      "grad_norm": 0.08488965034484863,
      "learning_rate": 0.0005769430657784245,
      "loss": 0.4578,
      "num_input_tokens_seen": 12592104,
      "step": 19300
    },
    {
      "epoch": 10.117924528301886,
      "grad_norm": 0.05904199182987213,
      "learning_rate": 0.0005767170960093449,
      "loss": 0.3765,
      "num_input_tokens_seen": 12595240,
      "step": 19305
    },
    {
      "epoch": 10.120545073375261,
      "grad_norm": 0.1650860607624054,
      "learning_rate": 0.0005764911101919244,
      "loss": 0.3321,
      "num_input_tokens_seen": 12597480,
      "step": 19310
    },
    {
      "epoch": 10.123165618448636,
      "grad_norm": 0.07892763614654541,
      "learning_rate": 0.0005762651083734363,
      "loss": 0.5464,
      "num_input_tokens_seen": 12600328,
      "step": 19315
    },
    {
      "epoch": 10.125786163522013,
      "grad_norm": 0.228812575340271,
      "learning_rate": 0.0005760390906011574,
      "loss": 0.4362,
      "num_input_tokens_seen": 12602152,
      "step": 19320
    },
    {
      "epoch": 10.128406708595389,
      "grad_norm": 0.17374426126480103,
      "learning_rate": 0.0005758130569223684,
      "loss": 0.502,
      "num_input_tokens_seen": 12605352,
      "step": 19325
    },
    {
      "epoch": 10.131027253668764,
      "grad_norm": 0.09216538071632385,
      "learning_rate": 0.0005755870073843527,
      "loss": 0.3764,
      "num_input_tokens_seen": 12607784,
      "step": 19330
    },
    {
      "epoch": 10.133647798742139,
      "grad_norm": 0.1020294725894928,
      "learning_rate": 0.0005753609420343973,
      "loss": 0.6105,
      "num_input_tokens_seen": 12614056,
      "step": 19335
    },
    {
      "epoch": 10.136268343815514,
      "grad_norm": 0.1294758915901184,
      "learning_rate": 0.0005751348609197929,
      "loss": 0.5055,
      "num_input_tokens_seen": 12616552,
      "step": 19340
    },
    {
      "epoch": 10.13888888888889,
      "grad_norm": 0.13786731660366058,
      "learning_rate": 0.0005749087640878325,
      "loss": 0.484,
      "num_input_tokens_seen": 12619976,
      "step": 19345
    },
    {
      "epoch": 10.141509433962264,
      "grad_norm": 0.09171262383460999,
      "learning_rate": 0.0005746826515858135,
      "loss": 0.4316,
      "num_input_tokens_seen": 12623272,
      "step": 19350
    },
    {
      "epoch": 10.14412997903564,
      "grad_norm": 0.18189406394958496,
      "learning_rate": 0.0005744565234610356,
      "loss": 0.5098,
      "num_input_tokens_seen": 12625960,
      "step": 19355
    },
    {
      "epoch": 10.146750524109015,
      "grad_norm": 0.11338663846254349,
      "learning_rate": 0.0005742303797608023,
      "loss": 0.4213,
      "num_input_tokens_seen": 12630184,
      "step": 19360
    },
    {
      "epoch": 10.14937106918239,
      "grad_norm": 0.0652889832854271,
      "learning_rate": 0.0005740042205324208,
      "loss": 0.4466,
      "num_input_tokens_seen": 12633224,
      "step": 19365
    },
    {
      "epoch": 10.151991614255765,
      "grad_norm": 0.09054817259311676,
      "learning_rate": 0.0005737780458232004,
      "loss": 0.4566,
      "num_input_tokens_seen": 12636584,
      "step": 19370
    },
    {
      "epoch": 10.15461215932914,
      "grad_norm": 0.09271043539047241,
      "learning_rate": 0.0005735518556804546,
      "loss": 0.52,
      "num_input_tokens_seen": 12639944,
      "step": 19375
    },
    {
      "epoch": 10.157232704402515,
      "grad_norm": 0.12248951941728592,
      "learning_rate": 0.0005733256501514999,
      "loss": 0.5011,
      "num_input_tokens_seen": 12643624,
      "step": 19380
    },
    {
      "epoch": 10.15985324947589,
      "grad_norm": 0.06270179897546768,
      "learning_rate": 0.0005730994292836555,
      "loss": 0.3062,
      "num_input_tokens_seen": 12646792,
      "step": 19385
    },
    {
      "epoch": 10.162473794549266,
      "grad_norm": 0.140888512134552,
      "learning_rate": 0.0005728731931242445,
      "loss": 0.5125,
      "num_input_tokens_seen": 12650696,
      "step": 19390
    },
    {
      "epoch": 10.165094339622641,
      "grad_norm": 0.08457950502634048,
      "learning_rate": 0.0005726469417205928,
      "loss": 0.578,
      "num_input_tokens_seen": 12653512,
      "step": 19395
    },
    {
      "epoch": 10.167714884696016,
      "grad_norm": 0.10456304252147675,
      "learning_rate": 0.0005724206751200297,
      "loss": 0.4164,
      "num_input_tokens_seen": 12657160,
      "step": 19400
    },
    {
      "epoch": 10.170335429769391,
      "grad_norm": 0.08271681517362595,
      "learning_rate": 0.0005721943933698879,
      "loss": 0.4372,
      "num_input_tokens_seen": 12661096,
      "step": 19405
    },
    {
      "epoch": 10.172955974842766,
      "grad_norm": 0.06624303758144379,
      "learning_rate": 0.0005719680965175023,
      "loss": 0.3849,
      "num_input_tokens_seen": 12664168,
      "step": 19410
    },
    {
      "epoch": 10.175576519916143,
      "grad_norm": 0.1255820244550705,
      "learning_rate": 0.000571741784610212,
      "loss": 0.4152,
      "num_input_tokens_seen": 12667304,
      "step": 19415
    },
    {
      "epoch": 10.178197064989519,
      "grad_norm": 0.05470143258571625,
      "learning_rate": 0.0005715154576953589,
      "loss": 0.3756,
      "num_input_tokens_seen": 12670760,
      "step": 19420
    },
    {
      "epoch": 10.180817610062894,
      "grad_norm": 0.13345976173877716,
      "learning_rate": 0.0005712891158202879,
      "loss": 0.5103,
      "num_input_tokens_seen": 12673384,
      "step": 19425
    },
    {
      "epoch": 10.183438155136269,
      "grad_norm": 0.21085615456104279,
      "learning_rate": 0.0005710627590323473,
      "loss": 0.4509,
      "num_input_tokens_seen": 12676168,
      "step": 19430
    },
    {
      "epoch": 10.186058700209644,
      "grad_norm": 0.05735190585255623,
      "learning_rate": 0.0005708363873788882,
      "loss": 0.3837,
      "num_input_tokens_seen": 12678984,
      "step": 19435
    },
    {
      "epoch": 10.18867924528302,
      "grad_norm": 0.12310639768838882,
      "learning_rate": 0.0005706100009072649,
      "loss": 0.6023,
      "num_input_tokens_seen": 12682600,
      "step": 19440
    },
    {
      "epoch": 10.191299790356394,
      "grad_norm": 0.12692765891551971,
      "learning_rate": 0.000570383599664835,
      "loss": 0.3637,
      "num_input_tokens_seen": 12685704,
      "step": 19445
    },
    {
      "epoch": 10.19392033542977,
      "grad_norm": 0.10334959626197815,
      "learning_rate": 0.0005701571836989591,
      "loss": 0.3686,
      "num_input_tokens_seen": 12689256,
      "step": 19450
    },
    {
      "epoch": 10.196540880503145,
      "grad_norm": 0.0715106725692749,
      "learning_rate": 0.0005699307530570008,
      "loss": 0.4263,
      "num_input_tokens_seen": 12692968,
      "step": 19455
    },
    {
      "epoch": 10.19916142557652,
      "grad_norm": 0.11717797070741653,
      "learning_rate": 0.0005697043077863267,
      "loss": 0.3891,
      "num_input_tokens_seen": 12696712,
      "step": 19460
    },
    {
      "epoch": 10.201781970649895,
      "grad_norm": 0.1930541843175888,
      "learning_rate": 0.0005694778479343065,
      "loss": 0.5065,
      "num_input_tokens_seen": 12699848,
      "step": 19465
    },
    {
      "epoch": 10.20440251572327,
      "grad_norm": 0.09320088475942612,
      "learning_rate": 0.0005692513735483134,
      "loss": 0.4448,
      "num_input_tokens_seen": 12703176,
      "step": 19470
    },
    {
      "epoch": 10.207023060796645,
      "grad_norm": 0.06869140267372131,
      "learning_rate": 0.0005690248846757228,
      "loss": 0.5539,
      "num_input_tokens_seen": 12706280,
      "step": 19475
    },
    {
      "epoch": 10.20964360587002,
      "grad_norm": 0.11260127276182175,
      "learning_rate": 0.0005687983813639139,
      "loss": 0.5613,
      "num_input_tokens_seen": 12709224,
      "step": 19480
    },
    {
      "epoch": 10.212264150943396,
      "grad_norm": 0.07216088473796844,
      "learning_rate": 0.0005685718636602685,
      "loss": 0.3943,
      "num_input_tokens_seen": 12713064,
      "step": 19485
    },
    {
      "epoch": 10.214884696016771,
      "grad_norm": 0.12147360295057297,
      "learning_rate": 0.0005683453316121716,
      "loss": 0.4952,
      "num_input_tokens_seen": 12716424,
      "step": 19490
    },
    {
      "epoch": 10.217505241090146,
      "grad_norm": 0.10441254079341888,
      "learning_rate": 0.0005681187852670111,
      "loss": 0.4432,
      "num_input_tokens_seen": 12719816,
      "step": 19495
    },
    {
      "epoch": 10.220125786163521,
      "grad_norm": 0.14021043479442596,
      "learning_rate": 0.0005678922246721777,
      "loss": 0.5317,
      "num_input_tokens_seen": 12723624,
      "step": 19500
    },
    {
      "epoch": 10.222746331236896,
      "grad_norm": 0.13631503283977509,
      "learning_rate": 0.0005676656498750656,
      "loss": 0.4535,
      "num_input_tokens_seen": 12727240,
      "step": 19505
    },
    {
      "epoch": 10.225366876310272,
      "grad_norm": 0.19821996986865997,
      "learning_rate": 0.0005674390609230715,
      "loss": 0.4376,
      "num_input_tokens_seen": 12730792,
      "step": 19510
    },
    {
      "epoch": 10.227987421383649,
      "grad_norm": 0.19302432239055634,
      "learning_rate": 0.0005672124578635951,
      "loss": 0.4855,
      "num_input_tokens_seen": 12733928,
      "step": 19515
    },
    {
      "epoch": 10.230607966457024,
      "grad_norm": 0.09404604882001877,
      "learning_rate": 0.0005669858407440394,
      "loss": 0.389,
      "num_input_tokens_seen": 12737992,
      "step": 19520
    },
    {
      "epoch": 10.233228511530399,
      "grad_norm": 0.1417776495218277,
      "learning_rate": 0.00056675920961181,
      "loss": 0.463,
      "num_input_tokens_seen": 12740456,
      "step": 19525
    },
    {
      "epoch": 10.235849056603774,
      "grad_norm": 0.1875166893005371,
      "learning_rate": 0.0005665325645143155,
      "loss": 0.4908,
      "num_input_tokens_seen": 12743496,
      "step": 19530
    },
    {
      "epoch": 10.23846960167715,
      "grad_norm": 0.12832996249198914,
      "learning_rate": 0.0005663059054989679,
      "loss": 0.4009,
      "num_input_tokens_seen": 12746952,
      "step": 19535
    },
    {
      "epoch": 10.241090146750524,
      "grad_norm": 0.1415722668170929,
      "learning_rate": 0.0005660792326131809,
      "loss": 0.5417,
      "num_input_tokens_seen": 12749800,
      "step": 19540
    },
    {
      "epoch": 10.2437106918239,
      "grad_norm": 0.12305514514446259,
      "learning_rate": 0.0005658525459043725,
      "loss": 0.4496,
      "num_input_tokens_seen": 12752872,
      "step": 19545
    },
    {
      "epoch": 10.246331236897275,
      "grad_norm": 0.2631928324699402,
      "learning_rate": 0.000565625845419963,
      "loss": 0.539,
      "num_input_tokens_seen": 12755944,
      "step": 19550
    },
    {
      "epoch": 10.24895178197065,
      "grad_norm": 0.2323380559682846,
      "learning_rate": 0.000565399131207375,
      "loss": 0.4589,
      "num_input_tokens_seen": 12758344,
      "step": 19555
    },
    {
      "epoch": 10.251572327044025,
      "grad_norm": 0.06799039989709854,
      "learning_rate": 0.000565172403314035,
      "loss": 0.3256,
      "num_input_tokens_seen": 12762312,
      "step": 19560
    },
    {
      "epoch": 10.2541928721174,
      "grad_norm": 0.1185169368982315,
      "learning_rate": 0.0005649456617873717,
      "loss": 0.4321,
      "num_input_tokens_seen": 12765416,
      "step": 19565
    },
    {
      "epoch": 10.256813417190775,
      "grad_norm": 0.1112002357840538,
      "learning_rate": 0.0005647189066748169,
      "loss": 0.3391,
      "num_input_tokens_seen": 12768552,
      "step": 19570
    },
    {
      "epoch": 10.25943396226415,
      "grad_norm": 0.19645850360393524,
      "learning_rate": 0.0005644921380238052,
      "loss": 0.4139,
      "num_input_tokens_seen": 12770920,
      "step": 19575
    },
    {
      "epoch": 10.262054507337526,
      "grad_norm": 0.09383974224328995,
      "learning_rate": 0.000564265355881774,
      "loss": 0.4199,
      "num_input_tokens_seen": 12774184,
      "step": 19580
    },
    {
      "epoch": 10.264675052410901,
      "grad_norm": 0.12215368449687958,
      "learning_rate": 0.0005640385602961634,
      "loss": 0.4973,
      "num_input_tokens_seen": 12777288,
      "step": 19585
    },
    {
      "epoch": 10.267295597484276,
      "grad_norm": 0.1270664632320404,
      "learning_rate": 0.0005638117513144168,
      "loss": 0.4151,
      "num_input_tokens_seen": 12780136,
      "step": 19590
    },
    {
      "epoch": 10.269916142557651,
      "grad_norm": 0.09730140119791031,
      "learning_rate": 0.0005635849289839797,
      "loss": 0.4882,
      "num_input_tokens_seen": 12783144,
      "step": 19595
    },
    {
      "epoch": 10.272536687631026,
      "grad_norm": 0.09167308360338211,
      "learning_rate": 0.000563358093352301,
      "loss": 0.3431,
      "num_input_tokens_seen": 12785992,
      "step": 19600
    },
    {
      "epoch": 10.275157232704402,
      "grad_norm": 0.09973878413438797,
      "learning_rate": 0.0005631312444668318,
      "loss": 0.5355,
      "num_input_tokens_seen": 12789000,
      "step": 19605
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 0.08946266770362854,
      "learning_rate": 0.0005629043823750266,
      "loss": 0.4666,
      "num_input_tokens_seen": 12791976,
      "step": 19610
    },
    {
      "epoch": 10.280398322851154,
      "grad_norm": 0.1754351407289505,
      "learning_rate": 0.0005626775071243423,
      "loss": 0.4672,
      "num_input_tokens_seen": 12795976,
      "step": 19615
    },
    {
      "epoch": 10.283018867924529,
      "grad_norm": 0.08937634527683258,
      "learning_rate": 0.0005624506187622384,
      "loss": 0.5091,
      "num_input_tokens_seen": 12798984,
      "step": 19620
    },
    {
      "epoch": 10.285639412997904,
      "grad_norm": 0.0977378785610199,
      "learning_rate": 0.0005622237173361777,
      "loss": 0.4359,
      "num_input_tokens_seen": 12801864,
      "step": 19625
    },
    {
      "epoch": 10.28825995807128,
      "grad_norm": 0.15826719999313354,
      "learning_rate": 0.0005619968028936252,
      "loss": 0.4989,
      "num_input_tokens_seen": 12804968,
      "step": 19630
    },
    {
      "epoch": 10.290880503144654,
      "grad_norm": 0.14994119107723236,
      "learning_rate": 0.0005617698754820489,
      "loss": 0.4224,
      "num_input_tokens_seen": 12807848,
      "step": 19635
    },
    {
      "epoch": 10.29350104821803,
      "grad_norm": 0.11534398794174194,
      "learning_rate": 0.0005615429351489192,
      "loss": 0.3966,
      "num_input_tokens_seen": 12810632,
      "step": 19640
    },
    {
      "epoch": 10.296121593291405,
      "grad_norm": 0.11035341024398804,
      "learning_rate": 0.0005613159819417095,
      "loss": 0.4731,
      "num_input_tokens_seen": 12814024,
      "step": 19645
    },
    {
      "epoch": 10.29874213836478,
      "grad_norm": 0.11348988115787506,
      "learning_rate": 0.0005610890159078961,
      "loss": 0.4126,
      "num_input_tokens_seen": 12817800,
      "step": 19650
    },
    {
      "epoch": 10.301362683438155,
      "grad_norm": 0.16232584416866302,
      "learning_rate": 0.0005608620370949575,
      "loss": 0.4461,
      "num_input_tokens_seen": 12820456,
      "step": 19655
    },
    {
      "epoch": 10.30398322851153,
      "grad_norm": 0.14285598695278168,
      "learning_rate": 0.000560635045550375,
      "loss": 0.5345,
      "num_input_tokens_seen": 12822792,
      "step": 19660
    },
    {
      "epoch": 10.306603773584905,
      "grad_norm": 0.08932552486658096,
      "learning_rate": 0.0005604080413216328,
      "loss": 0.6005,
      "num_input_tokens_seen": 12826888,
      "step": 19665
    },
    {
      "epoch": 10.30922431865828,
      "grad_norm": 0.08800846338272095,
      "learning_rate": 0.0005601810244562175,
      "loss": 0.455,
      "num_input_tokens_seen": 12829896,
      "step": 19670
    },
    {
      "epoch": 10.311844863731656,
      "grad_norm": 0.11346165835857391,
      "learning_rate": 0.0005599539950016184,
      "loss": 0.5701,
      "num_input_tokens_seen": 12833480,
      "step": 19675
    },
    {
      "epoch": 10.314465408805031,
      "grad_norm": 0.18413779139518738,
      "learning_rate": 0.0005597269530053277,
      "loss": 0.4427,
      "num_input_tokens_seen": 12836424,
      "step": 19680
    },
    {
      "epoch": 10.317085953878406,
      "grad_norm": 0.16049088537693024,
      "learning_rate": 0.0005594998985148398,
      "loss": 0.5702,
      "num_input_tokens_seen": 12839464,
      "step": 19685
    },
    {
      "epoch": 10.319706498951781,
      "grad_norm": 0.09645064920186996,
      "learning_rate": 0.0005592728315776519,
      "loss": 0.3349,
      "num_input_tokens_seen": 12841896,
      "step": 19690
    },
    {
      "epoch": 10.322327044025156,
      "grad_norm": 0.11855101585388184,
      "learning_rate": 0.0005590457522412638,
      "loss": 0.4215,
      "num_input_tokens_seen": 12845768,
      "step": 19695
    },
    {
      "epoch": 10.324947589098532,
      "grad_norm": 0.11466817557811737,
      "learning_rate": 0.000558818660553178,
      "loss": 0.4407,
      "num_input_tokens_seen": 12848680,
      "step": 19700
    },
    {
      "epoch": 10.327568134171909,
      "grad_norm": 0.18121084570884705,
      "learning_rate": 0.0005585915565608995,
      "loss": 0.5227,
      "num_input_tokens_seen": 12851464,
      "step": 19705
    },
    {
      "epoch": 10.330188679245284,
      "grad_norm": 0.11794210225343704,
      "learning_rate": 0.0005583644403119357,
      "loss": 0.313,
      "num_input_tokens_seen": 12854504,
      "step": 19710
    },
    {
      "epoch": 10.332809224318659,
      "grad_norm": 0.11432146281003952,
      "learning_rate": 0.0005581373118537969,
      "loss": 0.3531,
      "num_input_tokens_seen": 12857736,
      "step": 19715
    },
    {
      "epoch": 10.335429769392034,
      "grad_norm": 0.19163377583026886,
      "learning_rate": 0.0005579101712339957,
      "loss": 0.4543,
      "num_input_tokens_seen": 12860360,
      "step": 19720
    },
    {
      "epoch": 10.33805031446541,
      "grad_norm": 0.11398757994174957,
      "learning_rate": 0.0005576830185000474,
      "loss": 0.4311,
      "num_input_tokens_seen": 12863016,
      "step": 19725
    },
    {
      "epoch": 10.340670859538784,
      "grad_norm": 0.305776447057724,
      "learning_rate": 0.0005574558536994698,
      "loss": 0.5543,
      "num_input_tokens_seen": 12865768,
      "step": 19730
    },
    {
      "epoch": 10.34329140461216,
      "grad_norm": 0.10487362742424011,
      "learning_rate": 0.0005572286768797829,
      "loss": 0.4884,
      "num_input_tokens_seen": 12869256,
      "step": 19735
    },
    {
      "epoch": 10.345911949685535,
      "grad_norm": 0.06870831549167633,
      "learning_rate": 0.0005570014880885098,
      "loss": 0.5015,
      "num_input_tokens_seen": 12872456,
      "step": 19740
    },
    {
      "epoch": 10.34853249475891,
      "grad_norm": 0.1596924066543579,
      "learning_rate": 0.0005567742873731757,
      "loss": 0.3754,
      "num_input_tokens_seen": 12875720,
      "step": 19745
    },
    {
      "epoch": 10.351153039832285,
      "grad_norm": 0.12075455486774445,
      "learning_rate": 0.0005565470747813085,
      "loss": 0.353,
      "num_input_tokens_seen": 12878088,
      "step": 19750
    },
    {
      "epoch": 10.35377358490566,
      "grad_norm": 0.2977828085422516,
      "learning_rate": 0.0005563198503604382,
      "loss": 0.4216,
      "num_input_tokens_seen": 12880616,
      "step": 19755
    },
    {
      "epoch": 10.356394129979035,
      "grad_norm": 0.16326624155044556,
      "learning_rate": 0.0005560926141580978,
      "loss": 0.5999,
      "num_input_tokens_seen": 12883144,
      "step": 19760
    },
    {
      "epoch": 10.35901467505241,
      "grad_norm": 0.13104894757270813,
      "learning_rate": 0.0005558653662218225,
      "loss": 0.3976,
      "num_input_tokens_seen": 12885832,
      "step": 19765
    },
    {
      "epoch": 10.361635220125786,
      "grad_norm": 0.13870492577552795,
      "learning_rate": 0.00055563810659915,
      "loss": 0.4544,
      "num_input_tokens_seen": 12889000,
      "step": 19770
    },
    {
      "epoch": 10.364255765199161,
      "grad_norm": 0.11964976042509079,
      "learning_rate": 0.0005554108353376201,
      "loss": 0.4338,
      "num_input_tokens_seen": 12892616,
      "step": 19775
    },
    {
      "epoch": 10.366876310272536,
      "grad_norm": 0.13567718863487244,
      "learning_rate": 0.0005551835524847757,
      "loss": 0.4801,
      "num_input_tokens_seen": 12895112,
      "step": 19780
    },
    {
      "epoch": 10.369496855345911,
      "grad_norm": 0.09674499183893204,
      "learning_rate": 0.0005549562580881618,
      "loss": 0.5193,
      "num_input_tokens_seen": 12899176,
      "step": 19785
    },
    {
      "epoch": 10.372117400419286,
      "grad_norm": 0.06379091739654541,
      "learning_rate": 0.0005547289521953257,
      "loss": 0.4878,
      "num_input_tokens_seen": 12902184,
      "step": 19790
    },
    {
      "epoch": 10.374737945492662,
      "grad_norm": 0.20542719960212708,
      "learning_rate": 0.000554501634853817,
      "loss": 0.4061,
      "num_input_tokens_seen": 12904712,
      "step": 19795
    },
    {
      "epoch": 10.377358490566039,
      "grad_norm": 0.10265086591243744,
      "learning_rate": 0.0005542743061111883,
      "loss": 0.5143,
      "num_input_tokens_seen": 12907560,
      "step": 19800
    },
    {
      "epoch": 10.379979035639414,
      "grad_norm": 0.08278947323560715,
      "learning_rate": 0.0005540469660149938,
      "loss": 0.4869,
      "num_input_tokens_seen": 12910568,
      "step": 19805
    },
    {
      "epoch": 10.382599580712789,
      "grad_norm": 0.19132278859615326,
      "learning_rate": 0.0005538196146127907,
      "loss": 0.6836,
      "num_input_tokens_seen": 12913608,
      "step": 19810
    },
    {
      "epoch": 10.385220125786164,
      "grad_norm": 0.09502747654914856,
      "learning_rate": 0.0005535922519521381,
      "loss": 0.3843,
      "num_input_tokens_seen": 12917288,
      "step": 19815
    },
    {
      "epoch": 10.38784067085954,
      "grad_norm": 0.1304320991039276,
      "learning_rate": 0.0005533648780805979,
      "loss": 0.4155,
      "num_input_tokens_seen": 12920232,
      "step": 19820
    },
    {
      "epoch": 10.390461215932914,
      "grad_norm": 0.12131669372320175,
      "learning_rate": 0.000553137493045734,
      "loss": 0.4693,
      "num_input_tokens_seen": 12923432,
      "step": 19825
    },
    {
      "epoch": 10.39308176100629,
      "grad_norm": 0.06921666860580444,
      "learning_rate": 0.0005529100968951127,
      "loss": 0.4587,
      "num_input_tokens_seen": 12927912,
      "step": 19830
    },
    {
      "epoch": 10.395702306079665,
      "grad_norm": 0.09701921045780182,
      "learning_rate": 0.0005526826896763029,
      "loss": 0.4263,
      "num_input_tokens_seen": 12931304,
      "step": 19835
    },
    {
      "epoch": 10.39832285115304,
      "grad_norm": 0.12051603198051453,
      "learning_rate": 0.0005524552714368753,
      "loss": 0.4839,
      "num_input_tokens_seen": 12934248,
      "step": 19840
    },
    {
      "epoch": 10.400943396226415,
      "grad_norm": 0.11203551292419434,
      "learning_rate": 0.0005522278422244035,
      "loss": 0.5065,
      "num_input_tokens_seen": 12937608,
      "step": 19845
    },
    {
      "epoch": 10.40356394129979,
      "grad_norm": 0.12658686935901642,
      "learning_rate": 0.0005520004020864631,
      "loss": 0.3711,
      "num_input_tokens_seen": 12940360,
      "step": 19850
    },
    {
      "epoch": 10.406184486373165,
      "grad_norm": 0.10554060339927673,
      "learning_rate": 0.0005517729510706315,
      "loss": 0.3607,
      "num_input_tokens_seen": 12943048,
      "step": 19855
    },
    {
      "epoch": 10.40880503144654,
      "grad_norm": 0.24416759610176086,
      "learning_rate": 0.0005515454892244892,
      "loss": 0.3144,
      "num_input_tokens_seen": 12945704,
      "step": 19860
    },
    {
      "epoch": 10.411425576519916,
      "grad_norm": 0.11560191214084625,
      "learning_rate": 0.0005513180165956189,
      "loss": 0.4257,
      "num_input_tokens_seen": 12948616,
      "step": 19865
    },
    {
      "epoch": 10.414046121593291,
      "grad_norm": 0.24316158890724182,
      "learning_rate": 0.0005510905332316049,
      "loss": 0.3442,
      "num_input_tokens_seen": 12951880,
      "step": 19870
    },
    {
      "epoch": 10.416666666666666,
      "grad_norm": 0.07974757254123688,
      "learning_rate": 0.0005508630391800341,
      "loss": 0.5075,
      "num_input_tokens_seen": 12954984,
      "step": 19875
    },
    {
      "epoch": 10.419287211740041,
      "grad_norm": 0.1282396763563156,
      "learning_rate": 0.0005506355344884959,
      "loss": 0.4155,
      "num_input_tokens_seen": 12957640,
      "step": 19880
    },
    {
      "epoch": 10.421907756813416,
      "grad_norm": 0.14442121982574463,
      "learning_rate": 0.0005504080192045813,
      "loss": 0.345,
      "num_input_tokens_seen": 12960488,
      "step": 19885
    },
    {
      "epoch": 10.424528301886792,
      "grad_norm": 0.11112987250089645,
      "learning_rate": 0.0005501804933758846,
      "loss": 0.5358,
      "num_input_tokens_seen": 12963560,
      "step": 19890
    },
    {
      "epoch": 10.427148846960169,
      "grad_norm": 0.07481998950242996,
      "learning_rate": 0.0005499529570500009,
      "loss": 0.4389,
      "num_input_tokens_seen": 12967208,
      "step": 19895
    },
    {
      "epoch": 10.429769392033544,
      "grad_norm": 0.0859396904706955,
      "learning_rate": 0.0005497254102745285,
      "loss": 0.3312,
      "num_input_tokens_seen": 12970248,
      "step": 19900
    },
    {
      "epoch": 10.432389937106919,
      "grad_norm": 0.20250414311885834,
      "learning_rate": 0.0005494978530970676,
      "loss": 0.4402,
      "num_input_tokens_seen": 12973480,
      "step": 19905
    },
    {
      "epoch": 10.435010482180294,
      "grad_norm": 0.06719926744699478,
      "learning_rate": 0.0005492702855652206,
      "loss": 0.4205,
      "num_input_tokens_seen": 12976904,
      "step": 19910
    },
    {
      "epoch": 10.43763102725367,
      "grad_norm": 0.215741366147995,
      "learning_rate": 0.000549042707726592,
      "loss": 0.528,
      "num_input_tokens_seen": 12979912,
      "step": 19915
    },
    {
      "epoch": 10.440251572327044,
      "grad_norm": 0.1324402242898941,
      "learning_rate": 0.0005488151196287882,
      "loss": 0.4177,
      "num_input_tokens_seen": 12982728,
      "step": 19920
    },
    {
      "epoch": 10.44287211740042,
      "grad_norm": 0.10492243617773056,
      "learning_rate": 0.0005485875213194184,
      "loss": 0.35,
      "num_input_tokens_seen": 12986472,
      "step": 19925
    },
    {
      "epoch": 10.445492662473795,
      "grad_norm": 0.10817260295152664,
      "learning_rate": 0.0005483599128460935,
      "loss": 0.3922,
      "num_input_tokens_seen": 12990600,
      "step": 19930
    },
    {
      "epoch": 10.44811320754717,
      "grad_norm": 0.11443884670734406,
      "learning_rate": 0.0005481322942564266,
      "loss": 0.5552,
      "num_input_tokens_seen": 12994216,
      "step": 19935
    },
    {
      "epoch": 10.450733752620545,
      "grad_norm": 0.08719495683908463,
      "learning_rate": 0.0005479046655980327,
      "loss": 0.4677,
      "num_input_tokens_seen": 12997256,
      "step": 19940
    },
    {
      "epoch": 10.45335429769392,
      "grad_norm": 0.09925319999456406,
      "learning_rate": 0.0005476770269185295,
      "loss": 0.4125,
      "num_input_tokens_seen": 13001384,
      "step": 19945
    },
    {
      "epoch": 10.455974842767295,
      "grad_norm": 0.12079760432243347,
      "learning_rate": 0.0005474493782655361,
      "loss": 0.6228,
      "num_input_tokens_seen": 13004232,
      "step": 19950
    },
    {
      "epoch": 10.45859538784067,
      "grad_norm": 0.1162102222442627,
      "learning_rate": 0.0005472217196866742,
      "loss": 0.4135,
      "num_input_tokens_seen": 13007208,
      "step": 19955
    },
    {
      "epoch": 10.461215932914046,
      "grad_norm": 0.06569571793079376,
      "learning_rate": 0.0005469940512295671,
      "loss": 0.4581,
      "num_input_tokens_seen": 13010696,
      "step": 19960
    },
    {
      "epoch": 10.463836477987421,
      "grad_norm": 0.12569540739059448,
      "learning_rate": 0.0005467663729418406,
      "loss": 0.5106,
      "num_input_tokens_seen": 13013416,
      "step": 19965
    },
    {
      "epoch": 10.466457023060796,
      "grad_norm": 0.0802546963095665,
      "learning_rate": 0.0005465386848711225,
      "loss": 0.4322,
      "num_input_tokens_seen": 13016904,
      "step": 19970
    },
    {
      "epoch": 10.469077568134171,
      "grad_norm": 0.11021362245082855,
      "learning_rate": 0.0005463109870650426,
      "loss": 0.4566,
      "num_input_tokens_seen": 13019496,
      "step": 19975
    },
    {
      "epoch": 10.471698113207546,
      "grad_norm": 0.1086183562874794,
      "learning_rate": 0.0005460832795712323,
      "loss": 0.503,
      "num_input_tokens_seen": 13022184,
      "step": 19980
    },
    {
      "epoch": 10.474318658280922,
      "grad_norm": 0.18876460194587708,
      "learning_rate": 0.0005458555624373258,
      "loss": 0.5204,
      "num_input_tokens_seen": 13024968,
      "step": 19985
    },
    {
      "epoch": 10.476939203354299,
      "grad_norm": 0.12830284237861633,
      "learning_rate": 0.0005456278357109589,
      "loss": 0.4252,
      "num_input_tokens_seen": 13028328,
      "step": 19990
    },
    {
      "epoch": 10.479559748427674,
      "grad_norm": 0.149796724319458,
      "learning_rate": 0.0005454000994397691,
      "loss": 0.4367,
      "num_input_tokens_seen": 13032040,
      "step": 19995
    },
    {
      "epoch": 10.482180293501049,
      "grad_norm": 0.14966809749603271,
      "learning_rate": 0.0005451723536713965,
      "loss": 0.4336,
      "num_input_tokens_seen": 13035944,
      "step": 20000
    },
    {
      "epoch": 10.484800838574424,
      "grad_norm": 0.14285051822662354,
      "learning_rate": 0.0005449445984534828,
      "loss": 0.5075,
      "num_input_tokens_seen": 13042856,
      "step": 20005
    },
    {
      "epoch": 10.4874213836478,
      "grad_norm": 0.1554693579673767,
      "learning_rate": 0.0005447168338336721,
      "loss": 0.4042,
      "num_input_tokens_seen": 13046792,
      "step": 20010
    },
    {
      "epoch": 10.490041928721174,
      "grad_norm": 0.05410531163215637,
      "learning_rate": 0.0005444890598596098,
      "loss": 0.3938,
      "num_input_tokens_seen": 13050920,
      "step": 20015
    },
    {
      "epoch": 10.49266247379455,
      "grad_norm": 0.08862295746803284,
      "learning_rate": 0.0005442612765789437,
      "loss": 0.4158,
      "num_input_tokens_seen": 13058664,
      "step": 20020
    },
    {
      "epoch": 10.495283018867925,
      "grad_norm": 0.06058031693100929,
      "learning_rate": 0.0005440334840393235,
      "loss": 0.4727,
      "num_input_tokens_seen": 13062952,
      "step": 20025
    },
    {
      "epoch": 10.4979035639413,
      "grad_norm": 0.0938226580619812,
      "learning_rate": 0.0005438056822884007,
      "loss": 0.2743,
      "num_input_tokens_seen": 13065768,
      "step": 20030
    },
    {
      "epoch": 10.500524109014675,
      "grad_norm": 0.0616169199347496,
      "learning_rate": 0.0005435778713738292,
      "loss": 0.5013,
      "num_input_tokens_seen": 13069160,
      "step": 20035
    },
    {
      "epoch": 10.50314465408805,
      "grad_norm": 0.08662635087966919,
      "learning_rate": 0.0005433500513432639,
      "loss": 0.3924,
      "num_input_tokens_seen": 13072168,
      "step": 20040
    },
    {
      "epoch": 10.505765199161425,
      "grad_norm": 0.09258244931697845,
      "learning_rate": 0.0005431222222443626,
      "loss": 0.3846,
      "num_input_tokens_seen": 13074888,
      "step": 20045
    },
    {
      "epoch": 10.5083857442348,
      "grad_norm": 0.1316518634557724,
      "learning_rate": 0.0005428943841247843,
      "loss": 0.6713,
      "num_input_tokens_seen": 13078280,
      "step": 20050
    },
    {
      "epoch": 10.511006289308176,
      "grad_norm": 0.09470216184854507,
      "learning_rate": 0.0005426665370321901,
      "loss": 0.5074,
      "num_input_tokens_seen": 13081544,
      "step": 20055
    },
    {
      "epoch": 10.51362683438155,
      "grad_norm": 0.11719361692667007,
      "learning_rate": 0.0005424386810142433,
      "loss": 0.4876,
      "num_input_tokens_seen": 13084360,
      "step": 20060
    },
    {
      "epoch": 10.516247379454926,
      "grad_norm": 0.12407553195953369,
      "learning_rate": 0.0005422108161186086,
      "loss": 0.5616,
      "num_input_tokens_seen": 13087464,
      "step": 20065
    },
    {
      "epoch": 10.518867924528301,
      "grad_norm": 0.06948684900999069,
      "learning_rate": 0.0005419829423929525,
      "loss": 0.3733,
      "num_input_tokens_seen": 13091112,
      "step": 20070
    },
    {
      "epoch": 10.521488469601676,
      "grad_norm": 0.2757301926612854,
      "learning_rate": 0.0005417550598849441,
      "loss": 0.5343,
      "num_input_tokens_seen": 13094824,
      "step": 20075
    },
    {
      "epoch": 10.524109014675052,
      "grad_norm": 0.09482786059379578,
      "learning_rate": 0.0005415271686422534,
      "loss": 0.393,
      "num_input_tokens_seen": 13097768,
      "step": 20080
    },
    {
      "epoch": 10.526729559748428,
      "grad_norm": 0.09673910588026047,
      "learning_rate": 0.0005412992687125527,
      "loss": 0.5072,
      "num_input_tokens_seen": 13100712,
      "step": 20085
    },
    {
      "epoch": 10.529350104821804,
      "grad_norm": 0.1034722626209259,
      "learning_rate": 0.0005410713601435164,
      "loss": 0.3739,
      "num_input_tokens_seen": 13103464,
      "step": 20090
    },
    {
      "epoch": 10.531970649895179,
      "grad_norm": 0.16156208515167236,
      "learning_rate": 0.0005408434429828199,
      "loss": 0.4082,
      "num_input_tokens_seen": 13106312,
      "step": 20095
    },
    {
      "epoch": 10.534591194968554,
      "grad_norm": 0.11639466136693954,
      "learning_rate": 0.0005406155172781412,
      "loss": 0.4246,
      "num_input_tokens_seen": 13108936,
      "step": 20100
    },
    {
      "epoch": 10.53721174004193,
      "grad_norm": 0.05522598326206207,
      "learning_rate": 0.0005403875830771596,
      "loss": 0.3905,
      "num_input_tokens_seen": 13113544,
      "step": 20105
    },
    {
      "epoch": 10.539832285115304,
      "grad_norm": 0.10926644504070282,
      "learning_rate": 0.0005401596404275564,
      "loss": 0.548,
      "num_input_tokens_seen": 13116648,
      "step": 20110
    },
    {
      "epoch": 10.54245283018868,
      "grad_norm": 0.06739144027233124,
      "learning_rate": 0.0005399316893770147,
      "loss": 0.2908,
      "num_input_tokens_seen": 13119816,
      "step": 20115
    },
    {
      "epoch": 10.545073375262055,
      "grad_norm": 0.08113288134336472,
      "learning_rate": 0.0005397037299732189,
      "loss": 0.7114,
      "num_input_tokens_seen": 13123656,
      "step": 20120
    },
    {
      "epoch": 10.54769392033543,
      "grad_norm": 0.11349621415138245,
      "learning_rate": 0.000539475762263856,
      "loss": 0.379,
      "num_input_tokens_seen": 13126120,
      "step": 20125
    },
    {
      "epoch": 10.550314465408805,
      "grad_norm": 0.08567003160715103,
      "learning_rate": 0.0005392477862966138,
      "loss": 0.5349,
      "num_input_tokens_seen": 13128968,
      "step": 20130
    },
    {
      "epoch": 10.55293501048218,
      "grad_norm": 0.21790820360183716,
      "learning_rate": 0.0005390198021191827,
      "loss": 0.3864,
      "num_input_tokens_seen": 13132168,
      "step": 20135
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 0.15278339385986328,
      "learning_rate": 0.000538791809779254,
      "loss": 0.5063,
      "num_input_tokens_seen": 13135560,
      "step": 20140
    },
    {
      "epoch": 10.55817610062893,
      "grad_norm": 0.07310017943382263,
      "learning_rate": 0.0005385638093245213,
      "loss": 0.472,
      "num_input_tokens_seen": 13139400,
      "step": 20145
    },
    {
      "epoch": 10.560796645702306,
      "grad_norm": 0.07912293076515198,
      "learning_rate": 0.0005383358008026795,
      "loss": 0.5146,
      "num_input_tokens_seen": 13144232,
      "step": 20150
    },
    {
      "epoch": 10.56341719077568,
      "grad_norm": 0.11879951506853104,
      "learning_rate": 0.0005381077842614257,
      "loss": 0.3399,
      "num_input_tokens_seen": 13147432,
      "step": 20155
    },
    {
      "epoch": 10.566037735849056,
      "grad_norm": 0.09915245324373245,
      "learning_rate": 0.0005378797597484582,
      "loss": 0.3536,
      "num_input_tokens_seen": 13150632,
      "step": 20160
    },
    {
      "epoch": 10.568658280922431,
      "grad_norm": 0.0806674212217331,
      "learning_rate": 0.0005376517273114771,
      "loss": 0.4286,
      "num_input_tokens_seen": 13154152,
      "step": 20165
    },
    {
      "epoch": 10.571278825995806,
      "grad_norm": 0.07228288799524307,
      "learning_rate": 0.0005374236869981841,
      "loss": 0.3226,
      "num_input_tokens_seen": 13160552,
      "step": 20170
    },
    {
      "epoch": 10.573899371069182,
      "grad_norm": 0.14438743889331818,
      "learning_rate": 0.0005371956388562827,
      "loss": 0.453,
      "num_input_tokens_seen": 13163208,
      "step": 20175
    },
    {
      "epoch": 10.576519916142558,
      "grad_norm": 0.11077705770730972,
      "learning_rate": 0.0005369675829334782,
      "loss": 0.5055,
      "num_input_tokens_seen": 13166056,
      "step": 20180
    },
    {
      "epoch": 10.579140461215934,
      "grad_norm": 0.10525841265916824,
      "learning_rate": 0.0005367395192774769,
      "loss": 0.4038,
      "num_input_tokens_seen": 13168712,
      "step": 20185
    },
    {
      "epoch": 10.581761006289309,
      "grad_norm": 0.14165657758712769,
      "learning_rate": 0.0005365114479359875,
      "loss": 0.4596,
      "num_input_tokens_seen": 13171400,
      "step": 20190
    },
    {
      "epoch": 10.584381551362684,
      "grad_norm": 0.1080063059926033,
      "learning_rate": 0.0005362833689567197,
      "loss": 0.4289,
      "num_input_tokens_seen": 13174888,
      "step": 20195
    },
    {
      "epoch": 10.58700209643606,
      "grad_norm": 0.15356554090976715,
      "learning_rate": 0.0005360552823873849,
      "loss": 0.4322,
      "num_input_tokens_seen": 13178920,
      "step": 20200
    },
    {
      "epoch": 10.589622641509434,
      "grad_norm": 0.06283508241176605,
      "learning_rate": 0.0005358271882756966,
      "loss": 0.4829,
      "num_input_tokens_seen": 13183176,
      "step": 20205
    },
    {
      "epoch": 10.59224318658281,
      "grad_norm": 0.14536139369010925,
      "learning_rate": 0.0005355990866693689,
      "loss": 0.396,
      "num_input_tokens_seen": 13186760,
      "step": 20210
    },
    {
      "epoch": 10.594863731656185,
      "grad_norm": 0.12468742579221725,
      "learning_rate": 0.0005353709776161187,
      "loss": 0.3737,
      "num_input_tokens_seen": 13189544,
      "step": 20215
    },
    {
      "epoch": 10.59748427672956,
      "grad_norm": 0.08908241242170334,
      "learning_rate": 0.0005351428611636636,
      "loss": 0.4441,
      "num_input_tokens_seen": 13193096,
      "step": 20220
    },
    {
      "epoch": 10.600104821802935,
      "grad_norm": 0.11426094174385071,
      "learning_rate": 0.0005349147373597226,
      "loss": 0.3536,
      "num_input_tokens_seen": 13195944,
      "step": 20225
    },
    {
      "epoch": 10.60272536687631,
      "grad_norm": 0.09734729677438736,
      "learning_rate": 0.000534686606252017,
      "loss": 0.3847,
      "num_input_tokens_seen": 13198760,
      "step": 20230
    },
    {
      "epoch": 10.605345911949685,
      "grad_norm": 0.1113787367939949,
      "learning_rate": 0.0005344584678882692,
      "loss": 0.4291,
      "num_input_tokens_seen": 13201704,
      "step": 20235
    },
    {
      "epoch": 10.60796645702306,
      "grad_norm": 0.11330113559961319,
      "learning_rate": 0.0005342303223162027,
      "loss": 0.4868,
      "num_input_tokens_seen": 13204872,
      "step": 20240
    },
    {
      "epoch": 10.610587002096436,
      "grad_norm": 0.10698744654655457,
      "learning_rate": 0.0005340021695835437,
      "loss": 0.4087,
      "num_input_tokens_seen": 13208200,
      "step": 20245
    },
    {
      "epoch": 10.61320754716981,
      "grad_norm": 0.09340961277484894,
      "learning_rate": 0.0005337740097380184,
      "loss": 0.3697,
      "num_input_tokens_seen": 13211848,
      "step": 20250
    },
    {
      "epoch": 10.615828092243186,
      "grad_norm": 0.09868079423904419,
      "learning_rate": 0.0005335458428273557,
      "loss": 0.4159,
      "num_input_tokens_seen": 13214760,
      "step": 20255
    },
    {
      "epoch": 10.618448637316561,
      "grad_norm": 0.13794684410095215,
      "learning_rate": 0.0005333176688992855,
      "loss": 0.3976,
      "num_input_tokens_seen": 13219016,
      "step": 20260
    },
    {
      "epoch": 10.621069182389936,
      "grad_norm": 0.10660549998283386,
      "learning_rate": 0.000533089488001539,
      "loss": 0.3467,
      "num_input_tokens_seen": 13221800,
      "step": 20265
    },
    {
      "epoch": 10.623689727463312,
      "grad_norm": 0.11085794866085052,
      "learning_rate": 0.0005328613001818492,
      "loss": 0.4171,
      "num_input_tokens_seen": 13224936,
      "step": 20270
    },
    {
      "epoch": 10.626310272536688,
      "grad_norm": 0.12815788388252258,
      "learning_rate": 0.0005326331054879502,
      "loss": 0.4354,
      "num_input_tokens_seen": 13227560,
      "step": 20275
    },
    {
      "epoch": 10.628930817610064,
      "grad_norm": 0.1142507791519165,
      "learning_rate": 0.0005324049039675778,
      "loss": 0.4019,
      "num_input_tokens_seen": 13230888,
      "step": 20280
    },
    {
      "epoch": 10.631551362683439,
      "grad_norm": 0.11179524660110474,
      "learning_rate": 0.0005321766956684693,
      "loss": 0.5,
      "num_input_tokens_seen": 13234152,
      "step": 20285
    },
    {
      "epoch": 10.634171907756814,
      "grad_norm": 0.1059517115354538,
      "learning_rate": 0.0005319484806383631,
      "loss": 0.4511,
      "num_input_tokens_seen": 13237352,
      "step": 20290
    },
    {
      "epoch": 10.63679245283019,
      "grad_norm": 0.11889486014842987,
      "learning_rate": 0.0005317202589249991,
      "loss": 0.4549,
      "num_input_tokens_seen": 13240264,
      "step": 20295
    },
    {
      "epoch": 10.639412997903564,
      "grad_norm": 0.12818235158920288,
      "learning_rate": 0.0005314920305761191,
      "loss": 0.4094,
      "num_input_tokens_seen": 13243528,
      "step": 20300
    },
    {
      "epoch": 10.64203354297694,
      "grad_norm": 0.12716595828533173,
      "learning_rate": 0.0005312637956394654,
      "loss": 0.5026,
      "num_input_tokens_seen": 13248712,
      "step": 20305
    },
    {
      "epoch": 10.644654088050315,
      "grad_norm": 0.09529899060726166,
      "learning_rate": 0.0005310355541627824,
      "loss": 0.4954,
      "num_input_tokens_seen": 13252712,
      "step": 20310
    },
    {
      "epoch": 10.64727463312369,
      "grad_norm": 0.15124844014644623,
      "learning_rate": 0.0005308073061938153,
      "loss": 0.4513,
      "num_input_tokens_seen": 13256392,
      "step": 20315
    },
    {
      "epoch": 10.649895178197065,
      "grad_norm": 0.1217249259352684,
      "learning_rate": 0.0005305790517803115,
      "loss": 0.4981,
      "num_input_tokens_seen": 13259240,
      "step": 20320
    },
    {
      "epoch": 10.65251572327044,
      "grad_norm": 0.07743892073631287,
      "learning_rate": 0.0005303507909700189,
      "loss": 0.452,
      "num_input_tokens_seen": 13263176,
      "step": 20325
    },
    {
      "epoch": 10.655136268343815,
      "grad_norm": 0.10812415927648544,
      "learning_rate": 0.000530122523810687,
      "loss": 0.5146,
      "num_input_tokens_seen": 13266696,
      "step": 20330
    },
    {
      "epoch": 10.65775681341719,
      "grad_norm": 0.12955071032047272,
      "learning_rate": 0.0005298942503500668,
      "loss": 0.3923,
      "num_input_tokens_seen": 13269512,
      "step": 20335
    },
    {
      "epoch": 10.660377358490566,
      "grad_norm": 0.08633538335561752,
      "learning_rate": 0.0005296659706359105,
      "loss": 0.3213,
      "num_input_tokens_seen": 13272072,
      "step": 20340
    },
    {
      "epoch": 10.66299790356394,
      "grad_norm": 0.09167780727148056,
      "learning_rate": 0.0005294376847159716,
      "loss": 0.4529,
      "num_input_tokens_seen": 13275848,
      "step": 20345
    },
    {
      "epoch": 10.665618448637316,
      "grad_norm": 0.4127943813800812,
      "learning_rate": 0.0005292093926380049,
      "loss": 0.5319,
      "num_input_tokens_seen": 13278440,
      "step": 20350
    },
    {
      "epoch": 10.668238993710691,
      "grad_norm": 0.07848412543535233,
      "learning_rate": 0.0005289810944497663,
      "loss": 0.3975,
      "num_input_tokens_seen": 13281576,
      "step": 20355
    },
    {
      "epoch": 10.670859538784066,
      "grad_norm": 0.1367436945438385,
      "learning_rate": 0.0005287527901990137,
      "loss": 0.4602,
      "num_input_tokens_seen": 13284168,
      "step": 20360
    },
    {
      "epoch": 10.673480083857442,
      "grad_norm": 0.13441169261932373,
      "learning_rate": 0.0005285244799335053,
      "loss": 0.4668,
      "num_input_tokens_seen": 13286760,
      "step": 20365
    },
    {
      "epoch": 10.676100628930818,
      "grad_norm": 0.23030853271484375,
      "learning_rate": 0.0005282961637010009,
      "loss": 0.5156,
      "num_input_tokens_seen": 13289672,
      "step": 20370
    },
    {
      "epoch": 10.678721174004194,
      "grad_norm": 0.0989859476685524,
      "learning_rate": 0.0005280678415492621,
      "loss": 0.5485,
      "num_input_tokens_seen": 13292616,
      "step": 20375
    },
    {
      "epoch": 10.681341719077569,
      "grad_norm": 0.07716835290193558,
      "learning_rate": 0.0005278395135260512,
      "loss": 0.5094,
      "num_input_tokens_seen": 13296424,
      "step": 20380
    },
    {
      "epoch": 10.683962264150944,
      "grad_norm": 0.15890315175056458,
      "learning_rate": 0.0005276111796791314,
      "loss": 0.3836,
      "num_input_tokens_seen": 13299048,
      "step": 20385
    },
    {
      "epoch": 10.68658280922432,
      "grad_norm": 0.07735336571931839,
      "learning_rate": 0.000527382840056268,
      "loss": 0.4258,
      "num_input_tokens_seen": 13302152,
      "step": 20390
    },
    {
      "epoch": 10.689203354297694,
      "grad_norm": 0.1008262112736702,
      "learning_rate": 0.0005271544947052267,
      "loss": 0.3379,
      "num_input_tokens_seen": 13304840,
      "step": 20395
    },
    {
      "epoch": 10.69182389937107,
      "grad_norm": 0.09629089385271072,
      "learning_rate": 0.000526926143673775,
      "loss": 0.3697,
      "num_input_tokens_seen": 13307976,
      "step": 20400
    },
    {
      "epoch": 10.694444444444445,
      "grad_norm": 0.09753414988517761,
      "learning_rate": 0.0005266977870096813,
      "loss": 0.4403,
      "num_input_tokens_seen": 13310856,
      "step": 20405
    },
    {
      "epoch": 10.69706498951782,
      "grad_norm": 0.1032775342464447,
      "learning_rate": 0.000526469424760715,
      "loss": 0.4412,
      "num_input_tokens_seen": 13313448,
      "step": 20410
    },
    {
      "epoch": 10.699685534591195,
      "grad_norm": 0.1664021760225296,
      "learning_rate": 0.0005262410569746471,
      "loss": 0.5126,
      "num_input_tokens_seen": 13316776,
      "step": 20415
    },
    {
      "epoch": 10.70230607966457,
      "grad_norm": 0.05783497542142868,
      "learning_rate": 0.0005260126836992495,
      "loss": 0.544,
      "num_input_tokens_seen": 13319560,
      "step": 20420
    },
    {
      "epoch": 10.704926624737945,
      "grad_norm": 0.11633848398923874,
      "learning_rate": 0.000525784304982295,
      "loss": 0.4471,
      "num_input_tokens_seen": 13322440,
      "step": 20425
    },
    {
      "epoch": 10.70754716981132,
      "grad_norm": 0.10018465667963028,
      "learning_rate": 0.0005255559208715585,
      "loss": 0.4027,
      "num_input_tokens_seen": 13325448,
      "step": 20430
    },
    {
      "epoch": 10.710167714884696,
      "grad_norm": 0.07933332026004791,
      "learning_rate": 0.0005253275314148145,
      "loss": 0.3599,
      "num_input_tokens_seen": 13328520,
      "step": 20435
    },
    {
      "epoch": 10.71278825995807,
      "grad_norm": 0.12836098670959473,
      "learning_rate": 0.0005250991366598401,
      "loss": 0.3529,
      "num_input_tokens_seen": 13331560,
      "step": 20440
    },
    {
      "epoch": 10.715408805031446,
      "grad_norm": 0.14558002352714539,
      "learning_rate": 0.0005248707366544126,
      "loss": 0.4653,
      "num_input_tokens_seen": 13334216,
      "step": 20445
    },
    {
      "epoch": 10.718029350104821,
      "grad_norm": 0.20939777791500092,
      "learning_rate": 0.0005246423314463106,
      "loss": 0.5188,
      "num_input_tokens_seen": 13337768,
      "step": 20450
    },
    {
      "epoch": 10.720649895178196,
      "grad_norm": 0.09189671277999878,
      "learning_rate": 0.000524413921083314,
      "loss": 0.4559,
      "num_input_tokens_seen": 13341064,
      "step": 20455
    },
    {
      "epoch": 10.723270440251572,
      "grad_norm": 0.09022945910692215,
      "learning_rate": 0.0005241855056132039,
      "loss": 0.3759,
      "num_input_tokens_seen": 13343688,
      "step": 20460
    },
    {
      "epoch": 10.725890985324948,
      "grad_norm": 0.1080949530005455,
      "learning_rate": 0.0005239570850837617,
      "loss": 0.4216,
      "num_input_tokens_seen": 13347048,
      "step": 20465
    },
    {
      "epoch": 10.728511530398324,
      "grad_norm": 0.06658397614955902,
      "learning_rate": 0.0005237286595427709,
      "loss": 0.4555,
      "num_input_tokens_seen": 13350952,
      "step": 20470
    },
    {
      "epoch": 10.731132075471699,
      "grad_norm": 0.17537625133991241,
      "learning_rate": 0.000523500229038015,
      "loss": 0.4652,
      "num_input_tokens_seen": 13353672,
      "step": 20475
    },
    {
      "epoch": 10.733752620545074,
      "grad_norm": 0.089531309902668,
      "learning_rate": 0.0005232717936172794,
      "loss": 0.5402,
      "num_input_tokens_seen": 13357544,
      "step": 20480
    },
    {
      "epoch": 10.73637316561845,
      "grad_norm": 0.10497741401195526,
      "learning_rate": 0.0005230433533283498,
      "loss": 0.3878,
      "num_input_tokens_seen": 13361544,
      "step": 20485
    },
    {
      "epoch": 10.738993710691824,
      "grad_norm": 0.06830678880214691,
      "learning_rate": 0.0005228149082190139,
      "loss": 0.5238,
      "num_input_tokens_seen": 13365480,
      "step": 20490
    },
    {
      "epoch": 10.7416142557652,
      "grad_norm": 0.09011723101139069,
      "learning_rate": 0.0005225864583370594,
      "loss": 0.3718,
      "num_input_tokens_seen": 13368072,
      "step": 20495
    },
    {
      "epoch": 10.744234800838575,
      "grad_norm": 0.08939450234174728,
      "learning_rate": 0.0005223580037302754,
      "loss": 0.4508,
      "num_input_tokens_seen": 13370920,
      "step": 20500
    },
    {
      "epoch": 10.74685534591195,
      "grad_norm": 0.15805982053279877,
      "learning_rate": 0.0005221295444464522,
      "loss": 0.729,
      "num_input_tokens_seen": 13373832,
      "step": 20505
    },
    {
      "epoch": 10.749475890985325,
      "grad_norm": 0.09933124482631683,
      "learning_rate": 0.0005219010805333807,
      "loss": 0.4051,
      "num_input_tokens_seen": 13376712,
      "step": 20510
    },
    {
      "epoch": 10.7520964360587,
      "grad_norm": 0.18415488302707672,
      "learning_rate": 0.000521672612038853,
      "loss": 0.4726,
      "num_input_tokens_seen": 13379496,
      "step": 20515
    },
    {
      "epoch": 10.754716981132075,
      "grad_norm": 0.07084916532039642,
      "learning_rate": 0.0005214441390106624,
      "loss": 0.3788,
      "num_input_tokens_seen": 13383688,
      "step": 20520
    },
    {
      "epoch": 10.75733752620545,
      "grad_norm": 0.08962199836969376,
      "learning_rate": 0.0005212156614966022,
      "loss": 0.4092,
      "num_input_tokens_seen": 13387112,
      "step": 20525
    },
    {
      "epoch": 10.759958071278826,
      "grad_norm": 0.06345626711845398,
      "learning_rate": 0.0005209871795444678,
      "loss": 0.4523,
      "num_input_tokens_seen": 13391048,
      "step": 20530
    },
    {
      "epoch": 10.7625786163522,
      "grad_norm": 0.08915531635284424,
      "learning_rate": 0.000520758693202055,
      "loss": 0.4284,
      "num_input_tokens_seen": 13393448,
      "step": 20535
    },
    {
      "epoch": 10.765199161425576,
      "grad_norm": 0.113924540579319,
      "learning_rate": 0.0005205302025171601,
      "loss": 0.542,
      "num_input_tokens_seen": 13396040,
      "step": 20540
    },
    {
      "epoch": 10.767819706498951,
      "grad_norm": 0.12490762770175934,
      "learning_rate": 0.0005203017075375812,
      "loss": 0.4808,
      "num_input_tokens_seen": 13398824,
      "step": 20545
    },
    {
      "epoch": 10.770440251572326,
      "grad_norm": 0.11022163927555084,
      "learning_rate": 0.0005200732083111169,
      "loss": 0.4327,
      "num_input_tokens_seen": 13403368,
      "step": 20550
    },
    {
      "epoch": 10.773060796645701,
      "grad_norm": 0.09683188796043396,
      "learning_rate": 0.000519844704885566,
      "loss": 0.4769,
      "num_input_tokens_seen": 13406376,
      "step": 20555
    },
    {
      "epoch": 10.775681341719078,
      "grad_norm": 0.12085988372564316,
      "learning_rate": 0.0005196161973087297,
      "loss": 0.5482,
      "num_input_tokens_seen": 13411208,
      "step": 20560
    },
    {
      "epoch": 10.778301886792454,
      "grad_norm": 0.08011049777269363,
      "learning_rate": 0.0005193876856284084,
      "loss": 0.3623,
      "num_input_tokens_seen": 13413704,
      "step": 20565
    },
    {
      "epoch": 10.780922431865829,
      "grad_norm": 0.11905010789632797,
      "learning_rate": 0.0005191591698924046,
      "loss": 0.4763,
      "num_input_tokens_seen": 13417128,
      "step": 20570
    },
    {
      "epoch": 10.783542976939204,
      "grad_norm": 0.14125092327594757,
      "learning_rate": 0.0005189306501485211,
      "loss": 0.4612,
      "num_input_tokens_seen": 13420136,
      "step": 20575
    },
    {
      "epoch": 10.786163522012579,
      "grad_norm": 0.21230696141719818,
      "learning_rate": 0.0005187021264445613,
      "loss": 0.5215,
      "num_input_tokens_seen": 13422472,
      "step": 20580
    },
    {
      "epoch": 10.788784067085954,
      "grad_norm": 0.11811331659555435,
      "learning_rate": 0.0005184735988283302,
      "loss": 0.3217,
      "num_input_tokens_seen": 13426280,
      "step": 20585
    },
    {
      "epoch": 10.79140461215933,
      "grad_norm": 0.1267276108264923,
      "learning_rate": 0.000518245067347633,
      "loss": 0.4168,
      "num_input_tokens_seen": 13429416,
      "step": 20590
    },
    {
      "epoch": 10.794025157232705,
      "grad_norm": 0.11515070497989655,
      "learning_rate": 0.0005180165320502756,
      "loss": 0.4658,
      "num_input_tokens_seen": 13432776,
      "step": 20595
    },
    {
      "epoch": 10.79664570230608,
      "grad_norm": 0.1952916979789734,
      "learning_rate": 0.0005177879929840655,
      "loss": 0.4354,
      "num_input_tokens_seen": 13435976,
      "step": 20600
    },
    {
      "epoch": 10.799266247379455,
      "grad_norm": 0.167893648147583,
      "learning_rate": 0.0005175594501968103,
      "loss": 0.4397,
      "num_input_tokens_seen": 13439208,
      "step": 20605
    },
    {
      "epoch": 10.80188679245283,
      "grad_norm": 0.10836885124444962,
      "learning_rate": 0.0005173309037363182,
      "loss": 0.3837,
      "num_input_tokens_seen": 13441928,
      "step": 20610
    },
    {
      "epoch": 10.804507337526205,
      "grad_norm": 0.11837267130613327,
      "learning_rate": 0.0005171023536503991,
      "loss": 0.4386,
      "num_input_tokens_seen": 13445000,
      "step": 20615
    },
    {
      "epoch": 10.80712788259958,
      "grad_norm": 0.2644480764865875,
      "learning_rate": 0.0005168737999868624,
      "loss": 0.532,
      "num_input_tokens_seen": 13449032,
      "step": 20620
    },
    {
      "epoch": 10.809748427672956,
      "grad_norm": 0.08234798908233643,
      "learning_rate": 0.0005166452427935193,
      "loss": 0.4447,
      "num_input_tokens_seen": 13452392,
      "step": 20625
    },
    {
      "epoch": 10.81236897274633,
      "grad_norm": 0.07690880447626114,
      "learning_rate": 0.0005164166821181814,
      "loss": 0.6985,
      "num_input_tokens_seen": 13455784,
      "step": 20630
    },
    {
      "epoch": 10.814989517819706,
      "grad_norm": 0.07506957650184631,
      "learning_rate": 0.0005161881180086607,
      "loss": 0.4136,
      "num_input_tokens_seen": 13459368,
      "step": 20635
    },
    {
      "epoch": 10.817610062893081,
      "grad_norm": 0.09912696480751038,
      "learning_rate": 0.0005159595505127705,
      "loss": 0.4203,
      "num_input_tokens_seen": 13462664,
      "step": 20640
    },
    {
      "epoch": 10.820230607966456,
      "grad_norm": 0.09429870545864105,
      "learning_rate": 0.0005157309796783246,
      "loss": 0.5018,
      "num_input_tokens_seen": 13464872,
      "step": 20645
    },
    {
      "epoch": 10.822851153039831,
      "grad_norm": 0.12895675003528595,
      "learning_rate": 0.000515502405553137,
      "loss": 0.3011,
      "num_input_tokens_seen": 13467624,
      "step": 20650
    },
    {
      "epoch": 10.825471698113208,
      "grad_norm": 0.14597629010677338,
      "learning_rate": 0.0005152738281850231,
      "loss": 0.527,
      "num_input_tokens_seen": 13470856,
      "step": 20655
    },
    {
      "epoch": 10.828092243186584,
      "grad_norm": 0.15582966804504395,
      "learning_rate": 0.0005150452476217986,
      "loss": 0.3851,
      "num_input_tokens_seen": 13474024,
      "step": 20660
    },
    {
      "epoch": 10.830712788259959,
      "grad_norm": 0.13587269186973572,
      "learning_rate": 0.0005148166639112799,
      "loss": 0.4898,
      "num_input_tokens_seen": 13478376,
      "step": 20665
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 0.12198081612586975,
      "learning_rate": 0.0005145880771012842,
      "loss": 0.4195,
      "num_input_tokens_seen": 13481480,
      "step": 20670
    },
    {
      "epoch": 10.835953878406709,
      "grad_norm": 0.1624118685722351,
      "learning_rate": 0.0005143594872396291,
      "loss": 0.376,
      "num_input_tokens_seen": 13484456,
      "step": 20675
    },
    {
      "epoch": 10.838574423480084,
      "grad_norm": 0.16287828981876373,
      "learning_rate": 0.0005141308943741333,
      "loss": 0.4716,
      "num_input_tokens_seen": 13487560,
      "step": 20680
    },
    {
      "epoch": 10.84119496855346,
      "grad_norm": 0.052609123289585114,
      "learning_rate": 0.0005139022985526154,
      "loss": 0.4594,
      "num_input_tokens_seen": 13491272,
      "step": 20685
    },
    {
      "epoch": 10.843815513626835,
      "grad_norm": 0.118387371301651,
      "learning_rate": 0.0005136736998228952,
      "loss": 0.3758,
      "num_input_tokens_seen": 13493992,
      "step": 20690
    },
    {
      "epoch": 10.84643605870021,
      "grad_norm": 0.14785130321979523,
      "learning_rate": 0.0005134450982327933,
      "loss": 0.4851,
      "num_input_tokens_seen": 13496968,
      "step": 20695
    },
    {
      "epoch": 10.849056603773585,
      "grad_norm": 0.12923480570316315,
      "learning_rate": 0.0005132164938301299,
      "loss": 0.5827,
      "num_input_tokens_seen": 13501032,
      "step": 20700
    },
    {
      "epoch": 10.85167714884696,
      "grad_norm": 0.14939087629318237,
      "learning_rate": 0.0005129878866627271,
      "loss": 0.6255,
      "num_input_tokens_seen": 13502984,
      "step": 20705
    },
    {
      "epoch": 10.854297693920335,
      "grad_norm": 0.0587441623210907,
      "learning_rate": 0.0005127592767784063,
      "loss": 0.4424,
      "num_input_tokens_seen": 13507816,
      "step": 20710
    },
    {
      "epoch": 10.85691823899371,
      "grad_norm": 0.10275546461343765,
      "learning_rate": 0.0005125306642249905,
      "loss": 0.4719,
      "num_input_tokens_seen": 13511368,
      "step": 20715
    },
    {
      "epoch": 10.859538784067086,
      "grad_norm": 0.1597221940755844,
      "learning_rate": 0.0005123020490503027,
      "loss": 0.506,
      "num_input_tokens_seen": 13515720,
      "step": 20720
    },
    {
      "epoch": 10.86215932914046,
      "grad_norm": 0.08071423321962357,
      "learning_rate": 0.0005120734313021665,
      "loss": 0.4911,
      "num_input_tokens_seen": 13519784,
      "step": 20725
    },
    {
      "epoch": 10.864779874213836,
      "grad_norm": 0.07933317869901657,
      "learning_rate": 0.0005118448110284063,
      "loss": 0.3886,
      "num_input_tokens_seen": 13523304,
      "step": 20730
    },
    {
      "epoch": 10.867400419287211,
      "grad_norm": 0.20840846002101898,
      "learning_rate": 0.0005116161882768469,
      "loss": 0.7238,
      "num_input_tokens_seen": 13526120,
      "step": 20735
    },
    {
      "epoch": 10.870020964360586,
      "grad_norm": 0.1365637481212616,
      "learning_rate": 0.0005113875630953131,
      "loss": 0.4086,
      "num_input_tokens_seen": 13529128,
      "step": 20740
    },
    {
      "epoch": 10.872641509433961,
      "grad_norm": 0.06927452236413956,
      "learning_rate": 0.0005111589355316314,
      "loss": 0.5197,
      "num_input_tokens_seen": 13532584,
      "step": 20745
    },
    {
      "epoch": 10.875262054507338,
      "grad_norm": 0.12221834063529968,
      "learning_rate": 0.0005109303056336274,
      "loss": 0.4355,
      "num_input_tokens_seen": 13536104,
      "step": 20750
    },
    {
      "epoch": 10.877882599580714,
      "grad_norm": 0.07963822782039642,
      "learning_rate": 0.0005107016734491281,
      "loss": 0.3561,
      "num_input_tokens_seen": 13539528,
      "step": 20755
    },
    {
      "epoch": 10.880503144654089,
      "grad_norm": 0.1346016228199005,
      "learning_rate": 0.0005104730390259611,
      "loss": 0.4899,
      "num_input_tokens_seen": 13542600,
      "step": 20760
    },
    {
      "epoch": 10.883123689727464,
      "grad_norm": 0.08473928272724152,
      "learning_rate": 0.0005102444024119537,
      "loss": 0.3258,
      "num_input_tokens_seen": 13545704,
      "step": 20765
    },
    {
      "epoch": 10.885744234800839,
      "grad_norm": 0.16317890584468842,
      "learning_rate": 0.0005100157636549342,
      "loss": 0.4315,
      "num_input_tokens_seen": 13549096,
      "step": 20770
    },
    {
      "epoch": 10.888364779874214,
      "grad_norm": 0.12063538283109665,
      "learning_rate": 0.0005097871228027312,
      "loss": 0.4966,
      "num_input_tokens_seen": 13552232,
      "step": 20775
    },
    {
      "epoch": 10.89098532494759,
      "grad_norm": 0.17967954277992249,
      "learning_rate": 0.0005095584799031736,
      "loss": 0.5201,
      "num_input_tokens_seen": 13554984,
      "step": 20780
    },
    {
      "epoch": 10.893605870020965,
      "grad_norm": 0.13005498051643372,
      "learning_rate": 0.0005093298350040912,
      "loss": 0.2687,
      "num_input_tokens_seen": 13561288,
      "step": 20785
    },
    {
      "epoch": 10.89622641509434,
      "grad_norm": 0.08527350425720215,
      "learning_rate": 0.0005091011881533139,
      "loss": 0.4371,
      "num_input_tokens_seen": 13564072,
      "step": 20790
    },
    {
      "epoch": 10.898846960167715,
      "grad_norm": 0.09243208914995193,
      "learning_rate": 0.0005088725393986716,
      "loss": 0.4344,
      "num_input_tokens_seen": 13566856,
      "step": 20795
    },
    {
      "epoch": 10.90146750524109,
      "grad_norm": 0.10868428647518158,
      "learning_rate": 0.0005086438887879955,
      "loss": 0.4712,
      "num_input_tokens_seen": 13570696,
      "step": 20800
    },
    {
      "epoch": 10.904088050314465,
      "grad_norm": 0.08991655707359314,
      "learning_rate": 0.0005084152363691164,
      "loss": 0.3336,
      "num_input_tokens_seen": 13573416,
      "step": 20805
    },
    {
      "epoch": 10.90670859538784,
      "grad_norm": 0.12320619076490402,
      "learning_rate": 0.0005081865821898658,
      "loss": 0.5357,
      "num_input_tokens_seen": 13576392,
      "step": 20810
    },
    {
      "epoch": 10.909329140461216,
      "grad_norm": 0.14794105291366577,
      "learning_rate": 0.0005079579262980756,
      "loss": 0.5184,
      "num_input_tokens_seen": 13579048,
      "step": 20815
    },
    {
      "epoch": 10.91194968553459,
      "grad_norm": 0.258886456489563,
      "learning_rate": 0.0005077292687415781,
      "loss": 0.6064,
      "num_input_tokens_seen": 13582920,
      "step": 20820
    },
    {
      "epoch": 10.914570230607966,
      "grad_norm": 0.09827021509408951,
      "learning_rate": 0.0005075006095682056,
      "loss": 0.3453,
      "num_input_tokens_seen": 13585576,
      "step": 20825
    },
    {
      "epoch": 10.917190775681341,
      "grad_norm": 0.16277608275413513,
      "learning_rate": 0.0005072719488257914,
      "loss": 0.4364,
      "num_input_tokens_seen": 13588360,
      "step": 20830
    },
    {
      "epoch": 10.919811320754716,
      "grad_norm": 0.08682871609926224,
      "learning_rate": 0.0005070432865621683,
      "loss": 0.5323,
      "num_input_tokens_seen": 13591272,
      "step": 20835
    },
    {
      "epoch": 10.922431865828091,
      "grad_norm": 0.11132581532001495,
      "learning_rate": 0.00050681462282517,
      "loss": 0.3697,
      "num_input_tokens_seen": 13594184,
      "step": 20840
    },
    {
      "epoch": 10.925052410901468,
      "grad_norm": 0.17360171675682068,
      "learning_rate": 0.0005065859576626304,
      "loss": 0.4749,
      "num_input_tokens_seen": 13597352,
      "step": 20845
    },
    {
      "epoch": 10.927672955974844,
      "grad_norm": 0.08534654974937439,
      "learning_rate": 0.0005063572911223836,
      "loss": 0.4917,
      "num_input_tokens_seen": 13600008,
      "step": 20850
    },
    {
      "epoch": 10.930293501048219,
      "grad_norm": 0.13525228202342987,
      "learning_rate": 0.0005061286232522639,
      "loss": 0.3877,
      "num_input_tokens_seen": 13603656,
      "step": 20855
    },
    {
      "epoch": 10.932914046121594,
      "grad_norm": 0.11135781556367874,
      "learning_rate": 0.0005058999541001064,
      "loss": 0.4937,
      "num_input_tokens_seen": 13607176,
      "step": 20860
    },
    {
      "epoch": 10.935534591194969,
      "grad_norm": 0.15299025177955627,
      "learning_rate": 0.0005056712837137456,
      "loss": 0.5866,
      "num_input_tokens_seen": 13610600,
      "step": 20865
    },
    {
      "epoch": 10.938155136268344,
      "grad_norm": 0.10316566377878189,
      "learning_rate": 0.000505442612141017,
      "loss": 0.4049,
      "num_input_tokens_seen": 13614472,
      "step": 20870
    },
    {
      "epoch": 10.94077568134172,
      "grad_norm": 0.1384267657995224,
      "learning_rate": 0.000505213939429756,
      "loss": 0.357,
      "num_input_tokens_seen": 13617384,
      "step": 20875
    },
    {
      "epoch": 10.943396226415095,
      "grad_norm": 0.10729413479566574,
      "learning_rate": 0.0005049852656277984,
      "loss": 0.4923,
      "num_input_tokens_seen": 13620968,
      "step": 20880
    },
    {
      "epoch": 10.94601677148847,
      "grad_norm": 0.12975682318210602,
      "learning_rate": 0.0005047565907829801,
      "loss": 0.5495,
      "num_input_tokens_seen": 13623752,
      "step": 20885
    },
    {
      "epoch": 10.948637316561845,
      "grad_norm": 0.09431643038988113,
      "learning_rate": 0.0005045279149431374,
      "loss": 0.4722,
      "num_input_tokens_seen": 13629000,
      "step": 20890
    },
    {
      "epoch": 10.95125786163522,
      "grad_norm": 0.10940831154584885,
      "learning_rate": 0.0005042992381561063,
      "loss": 0.4143,
      "num_input_tokens_seen": 13632200,
      "step": 20895
    },
    {
      "epoch": 10.953878406708595,
      "grad_norm": 0.13509215414524078,
      "learning_rate": 0.0005040705604697238,
      "loss": 0.4154,
      "num_input_tokens_seen": 13634536,
      "step": 20900
    },
    {
      "epoch": 10.95649895178197,
      "grad_norm": 0.09999363124370575,
      "learning_rate": 0.0005038418819318265,
      "loss": 0.4445,
      "num_input_tokens_seen": 13638760,
      "step": 20905
    },
    {
      "epoch": 10.959119496855346,
      "grad_norm": 0.10001721978187561,
      "learning_rate": 0.0005036132025902511,
      "loss": 0.4405,
      "num_input_tokens_seen": 13641256,
      "step": 20910
    },
    {
      "epoch": 10.96174004192872,
      "grad_norm": 0.11168947070837021,
      "learning_rate": 0.0005033845224928352,
      "loss": 0.4276,
      "num_input_tokens_seen": 13644808,
      "step": 20915
    },
    {
      "epoch": 10.964360587002096,
      "grad_norm": 0.08404374122619629,
      "learning_rate": 0.0005031558416874156,
      "loss": 0.4777,
      "num_input_tokens_seen": 13648168,
      "step": 20920
    },
    {
      "epoch": 10.966981132075471,
      "grad_norm": 0.09974592924118042,
      "learning_rate": 0.0005029271602218301,
      "loss": 0.5109,
      "num_input_tokens_seen": 13652520,
      "step": 20925
    },
    {
      "epoch": 10.969601677148846,
      "grad_norm": 0.10315366089344025,
      "learning_rate": 0.0005026984781439162,
      "loss": 0.4538,
      "num_input_tokens_seen": 13655656,
      "step": 20930
    },
    {
      "epoch": 10.972222222222221,
      "grad_norm": 0.1761312484741211,
      "learning_rate": 0.0005024697955015112,
      "loss": 0.5027,
      "num_input_tokens_seen": 13658440,
      "step": 20935
    },
    {
      "epoch": 10.974842767295598,
      "grad_norm": 0.14790409803390503,
      "learning_rate": 0.0005022411123424532,
      "loss": 0.3601,
      "num_input_tokens_seen": 13661160,
      "step": 20940
    },
    {
      "epoch": 10.977463312368974,
      "grad_norm": 0.08338770270347595,
      "learning_rate": 0.00050201242871458,
      "loss": 0.4323,
      "num_input_tokens_seen": 13664936,
      "step": 20945
    },
    {
      "epoch": 10.980083857442349,
      "grad_norm": 0.12353438884019852,
      "learning_rate": 0.0005017837446657298,
      "loss": 0.5121,
      "num_input_tokens_seen": 13668072,
      "step": 20950
    },
    {
      "epoch": 10.982704402515724,
      "grad_norm": 0.08851000666618347,
      "learning_rate": 0.0005015550602437405,
      "loss": 0.3919,
      "num_input_tokens_seen": 13670856,
      "step": 20955
    },
    {
      "epoch": 10.985324947589099,
      "grad_norm": 0.11818297952413559,
      "learning_rate": 0.0005013263754964504,
      "loss": 0.4154,
      "num_input_tokens_seen": 13674792,
      "step": 20960
    },
    {
      "epoch": 10.987945492662474,
      "grad_norm": 0.08901533484458923,
      "learning_rate": 0.0005010976904716977,
      "loss": 0.4789,
      "num_input_tokens_seen": 13677544,
      "step": 20965
    },
    {
      "epoch": 10.99056603773585,
      "grad_norm": 0.058083776384592056,
      "learning_rate": 0.0005008690052173206,
      "loss": 0.4509,
      "num_input_tokens_seen": 13681896,
      "step": 20970
    },
    {
      "epoch": 10.993186582809225,
      "grad_norm": 0.10954021662473679,
      "learning_rate": 0.0005006403197811575,
      "loss": 0.3438,
      "num_input_tokens_seen": 13685256,
      "step": 20975
    },
    {
      "epoch": 10.9958071278826,
      "grad_norm": 0.21089763939380646,
      "learning_rate": 0.0005004116342110469,
      "loss": 0.5057,
      "num_input_tokens_seen": 13688072,
      "step": 20980
    },
    {
      "epoch": 10.998427672955975,
      "grad_norm": 0.07261142879724503,
      "learning_rate": 0.0005001829485548272,
      "loss": 0.3744,
      "num_input_tokens_seen": 13690824,
      "step": 20985
    },
    {
      "epoch": 11.0,
      "eval_loss": 0.455534964799881,
      "eval_runtime": 13.5472,
      "eval_samples_per_second": 62.596,
      "eval_steps_per_second": 15.649,
      "num_input_tokens_seen": 13691904,
      "step": 20988
    },
    {
      "epoch": 11.00104821802935,
      "grad_norm": 0.09543800354003906,
      "learning_rate": 0.0004999542628603364,
      "loss": 0.5478,
      "num_input_tokens_seen": 13692896,
      "step": 20990
    },
    {
      "epoch": 11.003668763102725,
      "grad_norm": 0.11857408285140991,
      "learning_rate": 0.0004997255771754136,
      "loss": 0.3445,
      "num_input_tokens_seen": 13696160,
      "step": 20995
    },
    {
      "epoch": 11.0062893081761,
      "grad_norm": 0.14089855551719666,
      "learning_rate": 0.0004994968915478969,
      "loss": 0.5265,
      "num_input_tokens_seen": 13699136,
      "step": 21000
    },
    {
      "epoch": 11.008909853249476,
      "grad_norm": 0.09967416524887085,
      "learning_rate": 0.0004992682060256246,
      "loss": 0.404,
      "num_input_tokens_seen": 13702496,
      "step": 21005
    },
    {
      "epoch": 11.01153039832285,
      "grad_norm": 0.07210306078195572,
      "learning_rate": 0.0004990395206564353,
      "loss": 0.4461,
      "num_input_tokens_seen": 13706048,
      "step": 21010
    },
    {
      "epoch": 11.014150943396226,
      "grad_norm": 0.09723732620477676,
      "learning_rate": 0.0004988108354881674,
      "loss": 0.5214,
      "num_input_tokens_seen": 13708640,
      "step": 21015
    },
    {
      "epoch": 11.016771488469601,
      "grad_norm": 0.12364902347326279,
      "learning_rate": 0.0004985821505686591,
      "loss": 0.3506,
      "num_input_tokens_seen": 13712000,
      "step": 21020
    },
    {
      "epoch": 11.019392033542976,
      "grad_norm": 0.1204703226685524,
      "learning_rate": 0.0004983534659457489,
      "loss": 0.4265,
      "num_input_tokens_seen": 13715232,
      "step": 21025
    },
    {
      "epoch": 11.022012578616351,
      "grad_norm": 0.1365651935338974,
      "learning_rate": 0.0004981247816672745,
      "loss": 0.7968,
      "num_input_tokens_seen": 13718368,
      "step": 21030
    },
    {
      "epoch": 11.024633123689727,
      "grad_norm": 0.19311924278736115,
      "learning_rate": 0.0004978960977810746,
      "loss": 0.4526,
      "num_input_tokens_seen": 13721824,
      "step": 21035
    },
    {
      "epoch": 11.027253668763104,
      "grad_norm": 0.10422532260417938,
      "learning_rate": 0.0004976674143349869,
      "loss": 0.4727,
      "num_input_tokens_seen": 13724480,
      "step": 21040
    },
    {
      "epoch": 11.029874213836479,
      "grad_norm": 0.1399812251329422,
      "learning_rate": 0.0004974387313768497,
      "loss": 0.3922,
      "num_input_tokens_seen": 13726656,
      "step": 21045
    },
    {
      "epoch": 11.032494758909854,
      "grad_norm": 0.13458475470542908,
      "learning_rate": 0.0004972100489545007,
      "loss": 0.3869,
      "num_input_tokens_seen": 13729184,
      "step": 21050
    },
    {
      "epoch": 11.035115303983229,
      "grad_norm": 0.08934994041919708,
      "learning_rate": 0.0004969813671157774,
      "loss": 0.4385,
      "num_input_tokens_seen": 13733600,
      "step": 21055
    },
    {
      "epoch": 11.037735849056604,
      "grad_norm": 0.10189240425825119,
      "learning_rate": 0.0004967526859085177,
      "loss": 0.5345,
      "num_input_tokens_seen": 13736832,
      "step": 21060
    },
    {
      "epoch": 11.04035639412998,
      "grad_norm": 0.19461321830749512,
      "learning_rate": 0.0004965240053805591,
      "loss": 0.4574,
      "num_input_tokens_seen": 13739360,
      "step": 21065
    },
    {
      "epoch": 11.042976939203355,
      "grad_norm": 0.10124719887971878,
      "learning_rate": 0.000496295325579739,
      "loss": 0.3567,
      "num_input_tokens_seen": 13743296,
      "step": 21070
    },
    {
      "epoch": 11.04559748427673,
      "grad_norm": 0.07992177456617355,
      "learning_rate": 0.0004960666465538945,
      "loss": 0.3472,
      "num_input_tokens_seen": 13747072,
      "step": 21075
    },
    {
      "epoch": 11.048218029350105,
      "grad_norm": 0.10091342031955719,
      "learning_rate": 0.0004958379683508627,
      "loss": 0.4254,
      "num_input_tokens_seen": 13750208,
      "step": 21080
    },
    {
      "epoch": 11.05083857442348,
      "grad_norm": 0.13028305768966675,
      "learning_rate": 0.0004956092910184803,
      "loss": 0.4385,
      "num_input_tokens_seen": 13752928,
      "step": 21085
    },
    {
      "epoch": 11.053459119496855,
      "grad_norm": 0.16264928877353668,
      "learning_rate": 0.0004953806146045843,
      "loss": 0.3809,
      "num_input_tokens_seen": 13756736,
      "step": 21090
    },
    {
      "epoch": 11.05607966457023,
      "grad_norm": 0.17292524874210358,
      "learning_rate": 0.0004951519391570108,
      "loss": 0.4463,
      "num_input_tokens_seen": 13759520,
      "step": 21095
    },
    {
      "epoch": 11.058700209643606,
      "grad_norm": 0.11562454700469971,
      "learning_rate": 0.0004949232647235964,
      "loss": 0.4357,
      "num_input_tokens_seen": 13762592,
      "step": 21100
    },
    {
      "epoch": 11.06132075471698,
      "grad_norm": 0.09962384402751923,
      "learning_rate": 0.0004946945913521771,
      "loss": 0.5407,
      "num_input_tokens_seen": 13766464,
      "step": 21105
    },
    {
      "epoch": 11.063941299790356,
      "grad_norm": 0.026499124243855476,
      "learning_rate": 0.0004944659190905889,
      "loss": 0.4123,
      "num_input_tokens_seen": 13772672,
      "step": 21110
    },
    {
      "epoch": 11.066561844863731,
      "grad_norm": 0.11078742891550064,
      "learning_rate": 0.0004942372479866674,
      "loss": 0.4642,
      "num_input_tokens_seen": 13776000,
      "step": 21115
    },
    {
      "epoch": 11.069182389937106,
      "grad_norm": 0.20815375447273254,
      "learning_rate": 0.0004940085780882475,
      "loss": 0.7058,
      "num_input_tokens_seen": 13778464,
      "step": 21120
    },
    {
      "epoch": 11.071802935010481,
      "grad_norm": 0.0984833613038063,
      "learning_rate": 0.0004937799094431649,
      "loss": 0.4601,
      "num_input_tokens_seen": 13781536,
      "step": 21125
    },
    {
      "epoch": 11.074423480083857,
      "grad_norm": 0.11587077379226685,
      "learning_rate": 0.0004935512420992542,
      "loss": 0.3976,
      "num_input_tokens_seen": 13784800,
      "step": 21130
    },
    {
      "epoch": 11.077044025157234,
      "grad_norm": 0.048799026757478714,
      "learning_rate": 0.0004933225761043502,
      "loss": 0.478,
      "num_input_tokens_seen": 13788384,
      "step": 21135
    },
    {
      "epoch": 11.079664570230609,
      "grad_norm": 0.13526645302772522,
      "learning_rate": 0.0004930939115062872,
      "loss": 0.5572,
      "num_input_tokens_seen": 13791744,
      "step": 21140
    },
    {
      "epoch": 11.082285115303984,
      "grad_norm": 0.08070410043001175,
      "learning_rate": 0.0004928652483528989,
      "loss": 0.3569,
      "num_input_tokens_seen": 13798304,
      "step": 21145
    },
    {
      "epoch": 11.084905660377359,
      "grad_norm": 0.08142048865556717,
      "learning_rate": 0.0004926365866920193,
      "loss": 0.4119,
      "num_input_tokens_seen": 13801184,
      "step": 21150
    },
    {
      "epoch": 11.087526205450734,
      "grad_norm": 0.0743914470076561,
      "learning_rate": 0.0004924079265714817,
      "loss": 0.3981,
      "num_input_tokens_seen": 13805120,
      "step": 21155
    },
    {
      "epoch": 11.09014675052411,
      "grad_norm": 0.059151943773031235,
      "learning_rate": 0.0004921792680391193,
      "loss": 0.4001,
      "num_input_tokens_seen": 13808448,
      "step": 21160
    },
    {
      "epoch": 11.092767295597485,
      "grad_norm": 0.13688765466213226,
      "learning_rate": 0.0004919506111427649,
      "loss": 0.42,
      "num_input_tokens_seen": 13812320,
      "step": 21165
    },
    {
      "epoch": 11.09538784067086,
      "grad_norm": 0.053355537354946136,
      "learning_rate": 0.0004917219559302507,
      "loss": 0.3671,
      "num_input_tokens_seen": 13815168,
      "step": 21170
    },
    {
      "epoch": 11.098008385744235,
      "grad_norm": 0.19741778075695038,
      "learning_rate": 0.0004914933024494089,
      "loss": 0.4875,
      "num_input_tokens_seen": 13817696,
      "step": 21175
    },
    {
      "epoch": 11.10062893081761,
      "grad_norm": 0.11354990303516388,
      "learning_rate": 0.0004912646507480713,
      "loss": 0.5289,
      "num_input_tokens_seen": 13820320,
      "step": 21180
    },
    {
      "epoch": 11.103249475890985,
      "grad_norm": 0.17202730476856232,
      "learning_rate": 0.0004910360008740689,
      "loss": 0.4591,
      "num_input_tokens_seen": 13823680,
      "step": 21185
    },
    {
      "epoch": 11.10587002096436,
      "grad_norm": 0.19100473821163177,
      "learning_rate": 0.0004908073528752329,
      "loss": 0.4583,
      "num_input_tokens_seen": 13826016,
      "step": 21190
    },
    {
      "epoch": 11.108490566037736,
      "grad_norm": 0.23488107323646545,
      "learning_rate": 0.0004905787067993941,
      "loss": 0.5079,
      "num_input_tokens_seen": 13828416,
      "step": 21195
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 0.17962881922721863,
      "learning_rate": 0.000490350062694382,
      "loss": 0.4821,
      "num_input_tokens_seen": 13830976,
      "step": 21200
    },
    {
      "epoch": 11.113731656184486,
      "grad_norm": 0.10507511347532272,
      "learning_rate": 0.0004901214206080269,
      "loss": 0.493,
      "num_input_tokens_seen": 13833952,
      "step": 21205
    },
    {
      "epoch": 11.116352201257861,
      "grad_norm": 0.1329251378774643,
      "learning_rate": 0.0004898927805881577,
      "loss": 0.3507,
      "num_input_tokens_seen": 13836224,
      "step": 21210
    },
    {
      "epoch": 11.118972746331236,
      "grad_norm": 0.12790679931640625,
      "learning_rate": 0.0004896641426826036,
      "loss": 0.4081,
      "num_input_tokens_seen": 13840416,
      "step": 21215
    },
    {
      "epoch": 11.121593291404611,
      "grad_norm": 0.4002111852169037,
      "learning_rate": 0.0004894355069391931,
      "loss": 0.3769,
      "num_input_tokens_seen": 13843072,
      "step": 21220
    },
    {
      "epoch": 11.124213836477987,
      "grad_norm": 0.1139690950512886,
      "learning_rate": 0.0004892068734057538,
      "loss": 0.5689,
      "num_input_tokens_seen": 13845728,
      "step": 21225
    },
    {
      "epoch": 11.126834381551364,
      "grad_norm": 0.09356991946697235,
      "learning_rate": 0.0004889782421301134,
      "loss": 0.3716,
      "num_input_tokens_seen": 13849472,
      "step": 21230
    },
    {
      "epoch": 11.129454926624739,
      "grad_norm": 0.12663885951042175,
      "learning_rate": 0.0004887496131600991,
      "loss": 0.3957,
      "num_input_tokens_seen": 13852704,
      "step": 21235
    },
    {
      "epoch": 11.132075471698114,
      "grad_norm": 0.1240629106760025,
      "learning_rate": 0.0004885209865435373,
      "loss": 0.5054,
      "num_input_tokens_seen": 13856000,
      "step": 21240
    },
    {
      "epoch": 11.134696016771489,
      "grad_norm": 0.11546922475099564,
      "learning_rate": 0.0004882923623282543,
      "loss": 0.44,
      "num_input_tokens_seen": 13858944,
      "step": 21245
    },
    {
      "epoch": 11.137316561844864,
      "grad_norm": 0.13972344994544983,
      "learning_rate": 0.0004880637405620756,
      "loss": 0.4472,
      "num_input_tokens_seen": 13861504,
      "step": 21250
    },
    {
      "epoch": 11.13993710691824,
      "grad_norm": 0.08910083770751953,
      "learning_rate": 0.00048783512129282604,
      "loss": 0.4542,
      "num_input_tokens_seen": 13865088,
      "step": 21255
    },
    {
      "epoch": 11.142557651991615,
      "grad_norm": 0.14080917835235596,
      "learning_rate": 0.00048760650456833033,
      "loss": 0.459,
      "num_input_tokens_seen": 13867488,
      "step": 21260
    },
    {
      "epoch": 11.14517819706499,
      "grad_norm": 0.09468698501586914,
      "learning_rate": 0.0004873778904364125,
      "loss": 0.4798,
      "num_input_tokens_seen": 13870912,
      "step": 21265
    },
    {
      "epoch": 11.147798742138365,
      "grad_norm": 0.12406451255083084,
      "learning_rate": 0.00048714927894489597,
      "loss": 0.4265,
      "num_input_tokens_seen": 13874080,
      "step": 21270
    },
    {
      "epoch": 11.15041928721174,
      "grad_norm": 0.1093762218952179,
      "learning_rate": 0.00048692067014160357,
      "loss": 0.5401,
      "num_input_tokens_seen": 13877920,
      "step": 21275
    },
    {
      "epoch": 11.153039832285115,
      "grad_norm": 0.09313847124576569,
      "learning_rate": 0.00048669206407435806,
      "loss": 0.5061,
      "num_input_tokens_seen": 13880608,
      "step": 21280
    },
    {
      "epoch": 11.15566037735849,
      "grad_norm": 0.33507004380226135,
      "learning_rate": 0.00048646346079098075,
      "loss": 0.6067,
      "num_input_tokens_seen": 13883936,
      "step": 21285
    },
    {
      "epoch": 11.158280922431866,
      "grad_norm": 0.07077160477638245,
      "learning_rate": 0.0004862348603392929,
      "loss": 0.5501,
      "num_input_tokens_seen": 13887680,
      "step": 21290
    },
    {
      "epoch": 11.16090146750524,
      "grad_norm": 0.09633640944957733,
      "learning_rate": 0.0004860062627671154,
      "loss": 0.4716,
      "num_input_tokens_seen": 13890592,
      "step": 21295
    },
    {
      "epoch": 11.163522012578616,
      "grad_norm": 0.15079332888126373,
      "learning_rate": 0.0004857776681222679,
      "loss": 0.4795,
      "num_input_tokens_seen": 13892704,
      "step": 21300
    },
    {
      "epoch": 11.166142557651991,
      "grad_norm": 0.14948652684688568,
      "learning_rate": 0.00048554907645257024,
      "loss": 0.3469,
      "num_input_tokens_seen": 13895616,
      "step": 21305
    },
    {
      "epoch": 11.168763102725366,
      "grad_norm": 0.13276059925556183,
      "learning_rate": 0.0004853204878058407,
      "loss": 0.5298,
      "num_input_tokens_seen": 13898688,
      "step": 21310
    },
    {
      "epoch": 11.171383647798741,
      "grad_norm": 0.10049104690551758,
      "learning_rate": 0.0004850919022298977,
      "loss": 0.4224,
      "num_input_tokens_seen": 13901760,
      "step": 21315
    },
    {
      "epoch": 11.174004192872117,
      "grad_norm": 0.15571299195289612,
      "learning_rate": 0.0004848633197725586,
      "loss": 0.5675,
      "num_input_tokens_seen": 13904416,
      "step": 21320
    },
    {
      "epoch": 11.176624737945493,
      "grad_norm": 0.0860152319073677,
      "learning_rate": 0.0004846347404816404,
      "loss": 0.4411,
      "num_input_tokens_seen": 13908192,
      "step": 21325
    },
    {
      "epoch": 11.179245283018869,
      "grad_norm": 0.08232518285512924,
      "learning_rate": 0.0004844061644049591,
      "loss": 0.5655,
      "num_input_tokens_seen": 13911392,
      "step": 21330
    },
    {
      "epoch": 11.181865828092244,
      "grad_norm": 0.10891193896532059,
      "learning_rate": 0.0004841775915903305,
      "loss": 0.3238,
      "num_input_tokens_seen": 13914368,
      "step": 21335
    },
    {
      "epoch": 11.184486373165619,
      "grad_norm": 0.09485305845737457,
      "learning_rate": 0.00048394902208556917,
      "loss": 0.4058,
      "num_input_tokens_seen": 13916544,
      "step": 21340
    },
    {
      "epoch": 11.187106918238994,
      "grad_norm": 0.09032604843378067,
      "learning_rate": 0.0004837204559384892,
      "loss": 0.4949,
      "num_input_tokens_seen": 13919520,
      "step": 21345
    },
    {
      "epoch": 11.18972746331237,
      "grad_norm": 0.17461395263671875,
      "learning_rate": 0.0004834918931969042,
      "loss": 0.3755,
      "num_input_tokens_seen": 13922496,
      "step": 21350
    },
    {
      "epoch": 11.192348008385745,
      "grad_norm": 0.13275299966335297,
      "learning_rate": 0.0004832633339086267,
      "loss": 0.421,
      "num_input_tokens_seen": 13925440,
      "step": 21355
    },
    {
      "epoch": 11.19496855345912,
      "grad_norm": 0.1707422286272049,
      "learning_rate": 0.0004830347781214689,
      "loss": 0.4816,
      "num_input_tokens_seen": 13928544,
      "step": 21360
    },
    {
      "epoch": 11.197589098532495,
      "grad_norm": 0.12073438614606857,
      "learning_rate": 0.0004828062258832421,
      "loss": 0.4289,
      "num_input_tokens_seen": 13931200,
      "step": 21365
    },
    {
      "epoch": 11.20020964360587,
      "grad_norm": 0.11909956485033035,
      "learning_rate": 0.0004825776772417564,
      "loss": 0.4798,
      "num_input_tokens_seen": 13933984,
      "step": 21370
    },
    {
      "epoch": 11.202830188679245,
      "grad_norm": 0.10136862844228745,
      "learning_rate": 0.00048234913224482205,
      "loss": 0.4351,
      "num_input_tokens_seen": 13937184,
      "step": 21375
    },
    {
      "epoch": 11.20545073375262,
      "grad_norm": 0.09141670167446136,
      "learning_rate": 0.00048212059094024774,
      "loss": 0.4647,
      "num_input_tokens_seen": 13940864,
      "step": 21380
    },
    {
      "epoch": 11.208071278825996,
      "grad_norm": 0.1534283608198166,
      "learning_rate": 0.0004818920533758418,
      "loss": 0.532,
      "num_input_tokens_seen": 13943904,
      "step": 21385
    },
    {
      "epoch": 11.21069182389937,
      "grad_norm": 0.09081894159317017,
      "learning_rate": 0.00048166351959941185,
      "loss": 0.3944,
      "num_input_tokens_seen": 13948192,
      "step": 21390
    },
    {
      "epoch": 11.213312368972746,
      "grad_norm": 0.09849195927381516,
      "learning_rate": 0.0004814349896587645,
      "loss": 0.2956,
      "num_input_tokens_seen": 13950912,
      "step": 21395
    },
    {
      "epoch": 11.215932914046121,
      "grad_norm": 0.1525721698999405,
      "learning_rate": 0.0004812064636017053,
      "loss": 0.4247,
      "num_input_tokens_seen": 13953664,
      "step": 21400
    },
    {
      "epoch": 11.218553459119496,
      "grad_norm": 0.08314897865056992,
      "learning_rate": 0.0004809779414760396,
      "loss": 0.4284,
      "num_input_tokens_seen": 13956576,
      "step": 21405
    },
    {
      "epoch": 11.221174004192871,
      "grad_norm": 0.07844363898038864,
      "learning_rate": 0.00048074942332957136,
      "loss": 0.4299,
      "num_input_tokens_seen": 13959520,
      "step": 21410
    },
    {
      "epoch": 11.223794549266247,
      "grad_norm": 0.06346801668405533,
      "learning_rate": 0.00048052090921010435,
      "loss": 0.4793,
      "num_input_tokens_seen": 13963552,
      "step": 21415
    },
    {
      "epoch": 11.226415094339623,
      "grad_norm": 0.11204275488853455,
      "learning_rate": 0.00048029239916544094,
      "loss": 0.5664,
      "num_input_tokens_seen": 13967072,
      "step": 21420
    },
    {
      "epoch": 11.229035639412999,
      "grad_norm": 0.0900186076760292,
      "learning_rate": 0.0004800638932433826,
      "loss": 0.3735,
      "num_input_tokens_seen": 13970592,
      "step": 21425
    },
    {
      "epoch": 11.231656184486374,
      "grad_norm": 0.14474239945411682,
      "learning_rate": 0.00047983539149173043,
      "loss": 0.4891,
      "num_input_tokens_seen": 13973696,
      "step": 21430
    },
    {
      "epoch": 11.234276729559749,
      "grad_norm": 0.1215614303946495,
      "learning_rate": 0.0004796068939582843,
      "loss": 0.4737,
      "num_input_tokens_seen": 13977248,
      "step": 21435
    },
    {
      "epoch": 11.236897274633124,
      "grad_norm": 0.13033244013786316,
      "learning_rate": 0.00047937840069084333,
      "loss": 0.5296,
      "num_input_tokens_seen": 13980704,
      "step": 21440
    },
    {
      "epoch": 11.2395178197065,
      "grad_norm": 0.10284275561571121,
      "learning_rate": 0.0004791499117372056,
      "loss": 0.4996,
      "num_input_tokens_seen": 13982976,
      "step": 21445
    },
    {
      "epoch": 11.242138364779874,
      "grad_norm": 0.4729984402656555,
      "learning_rate": 0.0004789214271451686,
      "loss": 0.5918,
      "num_input_tokens_seen": 13985888,
      "step": 21450
    },
    {
      "epoch": 11.24475890985325,
      "grad_norm": 0.11738214641809464,
      "learning_rate": 0.00047869294696252854,
      "loss": 0.4126,
      "num_input_tokens_seen": 13990048,
      "step": 21455
    },
    {
      "epoch": 11.247379454926625,
      "grad_norm": 0.111255943775177,
      "learning_rate": 0.0004784644712370807,
      "loss": 0.3949,
      "num_input_tokens_seen": 13993152,
      "step": 21460
    },
    {
      "epoch": 11.25,
      "grad_norm": 0.1307547241449356,
      "learning_rate": 0.00047823600001661984,
      "loss": 0.5264,
      "num_input_tokens_seen": 13995776,
      "step": 21465
    },
    {
      "epoch": 11.252620545073375,
      "grad_norm": 0.07050418853759766,
      "learning_rate": 0.0004780075333489396,
      "loss": 0.599,
      "num_input_tokens_seen": 13999168,
      "step": 21470
    },
    {
      "epoch": 11.25524109014675,
      "grad_norm": 0.15295696258544922,
      "learning_rate": 0.00047777907128183233,
      "loss": 0.4078,
      "num_input_tokens_seen": 14002048,
      "step": 21475
    },
    {
      "epoch": 11.257861635220126,
      "grad_norm": 0.13005658984184265,
      "learning_rate": 0.0004775506138630901,
      "loss": 0.4416,
      "num_input_tokens_seen": 14004864,
      "step": 21480
    },
    {
      "epoch": 11.2604821802935,
      "grad_norm": 0.13650651276111603,
      "learning_rate": 0.0004773221611405032,
      "loss": 0.3311,
      "num_input_tokens_seen": 14008352,
      "step": 21485
    },
    {
      "epoch": 11.263102725366876,
      "grad_norm": 0.12534725666046143,
      "learning_rate": 0.0004770937131618614,
      "loss": 0.3936,
      "num_input_tokens_seen": 14011232,
      "step": 21490
    },
    {
      "epoch": 11.265723270440251,
      "grad_norm": 0.0873485803604126,
      "learning_rate": 0.0004768652699749535,
      "loss": 0.4375,
      "num_input_tokens_seen": 14014848,
      "step": 21495
    },
    {
      "epoch": 11.268343815513626,
      "grad_norm": 0.11699920147657394,
      "learning_rate": 0.00047663683162756717,
      "loss": 0.4174,
      "num_input_tokens_seen": 14017536,
      "step": 21500
    },
    {
      "epoch": 11.270964360587001,
      "grad_norm": 0.09157484024763107,
      "learning_rate": 0.00047640839816748926,
      "loss": 0.392,
      "num_input_tokens_seen": 14021376,
      "step": 21505
    },
    {
      "epoch": 11.273584905660377,
      "grad_norm": 0.08843233436346054,
      "learning_rate": 0.00047617996964250525,
      "loss": 0.4758,
      "num_input_tokens_seen": 14024224,
      "step": 21510
    },
    {
      "epoch": 11.276205450733753,
      "grad_norm": 0.11598388850688934,
      "learning_rate": 0.0004759515461003997,
      "loss": 0.3866,
      "num_input_tokens_seen": 14026848,
      "step": 21515
    },
    {
      "epoch": 11.278825995807129,
      "grad_norm": 0.07900834083557129,
      "learning_rate": 0.00047572312758895654,
      "loss": 0.3491,
      "num_input_tokens_seen": 14030720,
      "step": 21520
    },
    {
      "epoch": 11.281446540880504,
      "grad_norm": 0.09035750478506088,
      "learning_rate": 0.0004754947141559579,
      "loss": 0.4,
      "num_input_tokens_seen": 14033632,
      "step": 21525
    },
    {
      "epoch": 11.284067085953879,
      "grad_norm": 0.12007862329483032,
      "learning_rate": 0.0004752663058491856,
      "loss": 0.4431,
      "num_input_tokens_seen": 14036960,
      "step": 21530
    },
    {
      "epoch": 11.286687631027254,
      "grad_norm": 0.10141737759113312,
      "learning_rate": 0.0004750379027164201,
      "loss": 0.4644,
      "num_input_tokens_seen": 14039648,
      "step": 21535
    },
    {
      "epoch": 11.28930817610063,
      "grad_norm": 0.10650372505187988,
      "learning_rate": 0.00047480950480544037,
      "loss": 0.5615,
      "num_input_tokens_seen": 14043840,
      "step": 21540
    },
    {
      "epoch": 11.291928721174004,
      "grad_norm": 0.08989611268043518,
      "learning_rate": 0.00047458111216402483,
      "loss": 0.4903,
      "num_input_tokens_seen": 14047200,
      "step": 21545
    },
    {
      "epoch": 11.29454926624738,
      "grad_norm": 0.19916124641895294,
      "learning_rate": 0.00047435272483995074,
      "loss": 0.3946,
      "num_input_tokens_seen": 14050336,
      "step": 21550
    },
    {
      "epoch": 11.297169811320755,
      "grad_norm": 0.10569899529218674,
      "learning_rate": 0.00047412434288099386,
      "loss": 0.501,
      "num_input_tokens_seen": 14053312,
      "step": 21555
    },
    {
      "epoch": 11.29979035639413,
      "grad_norm": 0.09382320195436478,
      "learning_rate": 0.0004738959663349294,
      "loss": 0.3425,
      "num_input_tokens_seen": 14056256,
      "step": 21560
    },
    {
      "epoch": 11.302410901467505,
      "grad_norm": 0.14052844047546387,
      "learning_rate": 0.0004736675952495311,
      "loss": 0.5125,
      "num_input_tokens_seen": 14059168,
      "step": 21565
    },
    {
      "epoch": 11.30503144654088,
      "grad_norm": 0.07496501505374908,
      "learning_rate": 0.0004734392296725712,
      "loss": 0.3975,
      "num_input_tokens_seen": 14062720,
      "step": 21570
    },
    {
      "epoch": 11.307651991614255,
      "grad_norm": 0.11597047746181488,
      "learning_rate": 0.00047321086965182153,
      "loss": 0.4067,
      "num_input_tokens_seen": 14065952,
      "step": 21575
    },
    {
      "epoch": 11.31027253668763,
      "grad_norm": 0.1485333889722824,
      "learning_rate": 0.0004729825152350522,
      "loss": 0.4604,
      "num_input_tokens_seen": 14069664,
      "step": 21580
    },
    {
      "epoch": 11.312893081761006,
      "grad_norm": 0.1744222491979599,
      "learning_rate": 0.00047275416647003256,
      "loss": 0.4472,
      "num_input_tokens_seen": 14073024,
      "step": 21585
    },
    {
      "epoch": 11.315513626834381,
      "grad_norm": 0.10763806104660034,
      "learning_rate": 0.0004725258234045306,
      "loss": 0.4266,
      "num_input_tokens_seen": 14076576,
      "step": 21590
    },
    {
      "epoch": 11.318134171907756,
      "grad_norm": 0.17688363790512085,
      "learning_rate": 0.0004722974860863127,
      "loss": 0.5238,
      "num_input_tokens_seen": 14080128,
      "step": 21595
    },
    {
      "epoch": 11.320754716981131,
      "grad_norm": 0.06696989387273788,
      "learning_rate": 0.00047206915456314476,
      "loss": 0.4569,
      "num_input_tokens_seen": 14084320,
      "step": 21600
    },
    {
      "epoch": 11.323375262054507,
      "grad_norm": 0.09032764285802841,
      "learning_rate": 0.00047184082888279097,
      "loss": 0.3501,
      "num_input_tokens_seen": 14087008,
      "step": 21605
    },
    {
      "epoch": 11.325995807127883,
      "grad_norm": 0.10555711388587952,
      "learning_rate": 0.00047161250909301454,
      "loss": 0.3964,
      "num_input_tokens_seen": 14090208,
      "step": 21610
    },
    {
      "epoch": 11.328616352201259,
      "grad_norm": 0.08653095364570618,
      "learning_rate": 0.00047138419524157734,
      "loss": 0.3103,
      "num_input_tokens_seen": 14093184,
      "step": 21615
    },
    {
      "epoch": 11.331236897274634,
      "grad_norm": 0.13124671578407288,
      "learning_rate": 0.0004711558873762402,
      "loss": 0.4802,
      "num_input_tokens_seen": 14100064,
      "step": 21620
    },
    {
      "epoch": 11.333857442348009,
      "grad_norm": 0.09160646796226501,
      "learning_rate": 0.0004709275855447621,
      "loss": 0.425,
      "num_input_tokens_seen": 14103104,
      "step": 21625
    },
    {
      "epoch": 11.336477987421384,
      "grad_norm": 0.0757615864276886,
      "learning_rate": 0.0004706992897949016,
      "loss": 0.378,
      "num_input_tokens_seen": 14106496,
      "step": 21630
    },
    {
      "epoch": 11.33909853249476,
      "grad_norm": 0.09271959215402603,
      "learning_rate": 0.00047047100017441525,
      "loss": 0.5507,
      "num_input_tokens_seen": 14110368,
      "step": 21635
    },
    {
      "epoch": 11.341719077568134,
      "grad_norm": 0.09030288457870483,
      "learning_rate": 0.0004702427167310589,
      "loss": 0.6335,
      "num_input_tokens_seen": 14114944,
      "step": 21640
    },
    {
      "epoch": 11.34433962264151,
      "grad_norm": 0.1500084400177002,
      "learning_rate": 0.0004700144395125866,
      "loss": 0.39,
      "num_input_tokens_seen": 14118272,
      "step": 21645
    },
    {
      "epoch": 11.346960167714885,
      "grad_norm": 0.08677002042531967,
      "learning_rate": 0.00046978616856675173,
      "loss": 0.4282,
      "num_input_tokens_seen": 14121472,
      "step": 21650
    },
    {
      "epoch": 11.34958071278826,
      "grad_norm": 0.1288234293460846,
      "learning_rate": 0.0004695579039413055,
      "loss": 0.362,
      "num_input_tokens_seen": 14123808,
      "step": 21655
    },
    {
      "epoch": 11.352201257861635,
      "grad_norm": 0.24665264785289764,
      "learning_rate": 0.0004693296456839984,
      "loss": 0.4755,
      "num_input_tokens_seen": 14126144,
      "step": 21660
    },
    {
      "epoch": 11.35482180293501,
      "grad_norm": 0.08985782414674759,
      "learning_rate": 0.0004691013938425796,
      "loss": 0.4726,
      "num_input_tokens_seen": 14129888,
      "step": 21665
    },
    {
      "epoch": 11.357442348008385,
      "grad_norm": 0.12572525441646576,
      "learning_rate": 0.00046887314846479663,
      "loss": 0.3588,
      "num_input_tokens_seen": 14132448,
      "step": 21670
    },
    {
      "epoch": 11.36006289308176,
      "grad_norm": 0.08379598706960678,
      "learning_rate": 0.000468644909598396,
      "loss": 0.3821,
      "num_input_tokens_seen": 14136736,
      "step": 21675
    },
    {
      "epoch": 11.362683438155136,
      "grad_norm": 0.07543396204710007,
      "learning_rate": 0.00046841667729112264,
      "loss": 0.5074,
      "num_input_tokens_seen": 14139840,
      "step": 21680
    },
    {
      "epoch": 11.365303983228511,
      "grad_norm": 0.11612259596586227,
      "learning_rate": 0.00046818845159071987,
      "loss": 0.4101,
      "num_input_tokens_seen": 14142528,
      "step": 21685
    },
    {
      "epoch": 11.367924528301886,
      "grad_norm": 0.13425703346729279,
      "learning_rate": 0.0004679602325449302,
      "loss": 0.3693,
      "num_input_tokens_seen": 14145984,
      "step": 21690
    },
    {
      "epoch": 11.370545073375261,
      "grad_norm": 0.07535212486982346,
      "learning_rate": 0.00046773202020149423,
      "loss": 0.5398,
      "num_input_tokens_seen": 14149120,
      "step": 21695
    },
    {
      "epoch": 11.373165618448636,
      "grad_norm": 0.15135814249515533,
      "learning_rate": 0.0004675038146081516,
      "loss": 0.4304,
      "num_input_tokens_seen": 14151936,
      "step": 21700
    },
    {
      "epoch": 11.375786163522013,
      "grad_norm": 0.1385648101568222,
      "learning_rate": 0.0004672756158126403,
      "loss": 0.4676,
      "num_input_tokens_seen": 14155040,
      "step": 21705
    },
    {
      "epoch": 11.378406708595389,
      "grad_norm": 0.12477803230285645,
      "learning_rate": 0.0004670474238626968,
      "loss": 0.4607,
      "num_input_tokens_seen": 14158336,
      "step": 21710
    },
    {
      "epoch": 11.381027253668764,
      "grad_norm": 0.2402469366788864,
      "learning_rate": 0.00046681923880605614,
      "loss": 0.3963,
      "num_input_tokens_seen": 14161248,
      "step": 21715
    },
    {
      "epoch": 11.383647798742139,
      "grad_norm": 0.1310701072216034,
      "learning_rate": 0.0004665910606904523,
      "loss": 0.4603,
      "num_input_tokens_seen": 14164960,
      "step": 21720
    },
    {
      "epoch": 11.386268343815514,
      "grad_norm": 0.117458775639534,
      "learning_rate": 0.0004663628895636173,
      "loss": 0.3892,
      "num_input_tokens_seen": 14167584,
      "step": 21725
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 0.11822198331356049,
      "learning_rate": 0.0004661347254732822,
      "loss": 0.361,
      "num_input_tokens_seen": 14169888,
      "step": 21730
    },
    {
      "epoch": 11.391509433962264,
      "grad_norm": 0.09001264721155167,
      "learning_rate": 0.0004659065684671762,
      "loss": 0.4128,
      "num_input_tokens_seen": 14173024,
      "step": 21735
    },
    {
      "epoch": 11.39412997903564,
      "grad_norm": 0.0857696458697319,
      "learning_rate": 0.00046567841859302696,
      "loss": 0.3703,
      "num_input_tokens_seen": 14176576,
      "step": 21740
    },
    {
      "epoch": 11.396750524109015,
      "grad_norm": 0.13371972739696503,
      "learning_rate": 0.0004654502758985611,
      "loss": 0.4839,
      "num_input_tokens_seen": 14180064,
      "step": 21745
    },
    {
      "epoch": 11.39937106918239,
      "grad_norm": 0.14219626784324646,
      "learning_rate": 0.00046522214043150323,
      "loss": 0.4403,
      "num_input_tokens_seen": 14183456,
      "step": 21750
    },
    {
      "epoch": 11.401991614255765,
      "grad_norm": 0.0742873027920723,
      "learning_rate": 0.00046499401223957696,
      "loss": 0.3624,
      "num_input_tokens_seen": 14186752,
      "step": 21755
    },
    {
      "epoch": 11.40461215932914,
      "grad_norm": 0.10885842144489288,
      "learning_rate": 0.0004647658913705038,
      "loss": 0.3023,
      "num_input_tokens_seen": 14189280,
      "step": 21760
    },
    {
      "epoch": 11.407232704402515,
      "grad_norm": 0.10166852176189423,
      "learning_rate": 0.00046453777787200444,
      "loss": 0.4068,
      "num_input_tokens_seen": 14193120,
      "step": 21765
    },
    {
      "epoch": 11.40985324947589,
      "grad_norm": 0.13649298250675201,
      "learning_rate": 0.0004643096717917973,
      "loss": 0.5519,
      "num_input_tokens_seen": 14196352,
      "step": 21770
    },
    {
      "epoch": 11.412473794549266,
      "grad_norm": 0.1544647067785263,
      "learning_rate": 0.0004640815731775995,
      "loss": 0.4315,
      "num_input_tokens_seen": 14199456,
      "step": 21775
    },
    {
      "epoch": 11.415094339622641,
      "grad_norm": 0.16790871322155,
      "learning_rate": 0.0004638534820771267,
      "loss": 0.4353,
      "num_input_tokens_seen": 14201600,
      "step": 21780
    },
    {
      "epoch": 11.417714884696016,
      "grad_norm": 0.09171529859304428,
      "learning_rate": 0.0004636253985380934,
      "loss": 0.4487,
      "num_input_tokens_seen": 14204672,
      "step": 21785
    },
    {
      "epoch": 11.420335429769391,
      "grad_norm": 0.09538127481937408,
      "learning_rate": 0.00046339732260821173,
      "loss": 0.4887,
      "num_input_tokens_seen": 14208928,
      "step": 21790
    },
    {
      "epoch": 11.422955974842766,
      "grad_norm": 0.10258371382951736,
      "learning_rate": 0.0004631692543351924,
      "loss": 0.2897,
      "num_input_tokens_seen": 14213888,
      "step": 21795
    },
    {
      "epoch": 11.425576519916142,
      "grad_norm": 0.2529004216194153,
      "learning_rate": 0.00046294119376674486,
      "loss": 0.3318,
      "num_input_tokens_seen": 14216704,
      "step": 21800
    },
    {
      "epoch": 11.428197064989519,
      "grad_norm": 0.21396377682685852,
      "learning_rate": 0.0004627131409505768,
      "loss": 0.4706,
      "num_input_tokens_seen": 14219616,
      "step": 21805
    },
    {
      "epoch": 11.430817610062894,
      "grad_norm": 0.09575898945331573,
      "learning_rate": 0.0004624850959343944,
      "loss": 0.3301,
      "num_input_tokens_seen": 14222368,
      "step": 21810
    },
    {
      "epoch": 11.433438155136269,
      "grad_norm": 0.08810359239578247,
      "learning_rate": 0.00046225705876590175,
      "loss": 0.4301,
      "num_input_tokens_seen": 14229888,
      "step": 21815
    },
    {
      "epoch": 11.436058700209644,
      "grad_norm": 0.1333625316619873,
      "learning_rate": 0.00046202902949280206,
      "loss": 0.4259,
      "num_input_tokens_seen": 14233440,
      "step": 21820
    },
    {
      "epoch": 11.43867924528302,
      "grad_norm": 0.21365980803966522,
      "learning_rate": 0.00046180100816279614,
      "loss": 0.4504,
      "num_input_tokens_seen": 14235488,
      "step": 21825
    },
    {
      "epoch": 11.441299790356394,
      "grad_norm": 0.14006632566452026,
      "learning_rate": 0.0004615729948235833,
      "loss": 0.4566,
      "num_input_tokens_seen": 14240128,
      "step": 21830
    },
    {
      "epoch": 11.44392033542977,
      "grad_norm": 0.08380264788866043,
      "learning_rate": 0.00046134498952286174,
      "loss": 0.3331,
      "num_input_tokens_seen": 14243264,
      "step": 21835
    },
    {
      "epoch": 11.446540880503145,
      "grad_norm": 0.08234268426895142,
      "learning_rate": 0.00046111699230832727,
      "loss": 0.5301,
      "num_input_tokens_seen": 14247712,
      "step": 21840
    },
    {
      "epoch": 11.44916142557652,
      "grad_norm": 0.08845356851816177,
      "learning_rate": 0.00046088900322767434,
      "loss": 0.4621,
      "num_input_tokens_seen": 14251360,
      "step": 21845
    },
    {
      "epoch": 11.451781970649895,
      "grad_norm": 0.11961095780134201,
      "learning_rate": 0.0004606610223285961,
      "loss": 0.4184,
      "num_input_tokens_seen": 14254432,
      "step": 21850
    },
    {
      "epoch": 11.45440251572327,
      "grad_norm": 0.1672910898923874,
      "learning_rate": 0.00046043304965878284,
      "loss": 0.3695,
      "num_input_tokens_seen": 14257664,
      "step": 21855
    },
    {
      "epoch": 11.457023060796645,
      "grad_norm": 0.10839679837226868,
      "learning_rate": 0.00046020508526592425,
      "loss": 0.3804,
      "num_input_tokens_seen": 14260960,
      "step": 21860
    },
    {
      "epoch": 11.45964360587002,
      "grad_norm": 0.07738931477069855,
      "learning_rate": 0.0004599771291977079,
      "loss": 0.5566,
      "num_input_tokens_seen": 14264960,
      "step": 21865
    },
    {
      "epoch": 11.462264150943396,
      "grad_norm": 0.21557864546775818,
      "learning_rate": 0.0004597491815018194,
      "loss": 0.4551,
      "num_input_tokens_seen": 14267840,
      "step": 21870
    },
    {
      "epoch": 11.464884696016771,
      "grad_norm": 0.10966932028532028,
      "learning_rate": 0.0004595212422259431,
      "loss": 0.462,
      "num_input_tokens_seen": 14270368,
      "step": 21875
    },
    {
      "epoch": 11.467505241090146,
      "grad_norm": 0.09263760596513748,
      "learning_rate": 0.00045929331141776104,
      "loss": 0.3474,
      "num_input_tokens_seen": 14273248,
      "step": 21880
    },
    {
      "epoch": 11.470125786163521,
      "grad_norm": 0.09773094207048416,
      "learning_rate": 0.0004590653891249537,
      "loss": 0.4365,
      "num_input_tokens_seen": 14281472,
      "step": 21885
    },
    {
      "epoch": 11.472746331236896,
      "grad_norm": 0.12415476888418198,
      "learning_rate": 0.0004588374753952001,
      "loss": 0.5274,
      "num_input_tokens_seen": 14284864,
      "step": 21890
    },
    {
      "epoch": 11.475366876310272,
      "grad_norm": 0.08638659119606018,
      "learning_rate": 0.0004586095702761769,
      "loss": 0.4758,
      "num_input_tokens_seen": 14287968,
      "step": 21895
    },
    {
      "epoch": 11.477987421383649,
      "grad_norm": 0.11064141243696213,
      "learning_rate": 0.0004583816738155595,
      "loss": 0.5597,
      "num_input_tokens_seen": 14291392,
      "step": 21900
    },
    {
      "epoch": 11.480607966457024,
      "grad_norm": 0.12417688220739365,
      "learning_rate": 0.0004581537860610213,
      "loss": 0.4156,
      "num_input_tokens_seen": 14294432,
      "step": 21905
    },
    {
      "epoch": 11.483228511530399,
      "grad_norm": 0.08268732577562332,
      "learning_rate": 0.0004579259070602333,
      "loss": 0.3278,
      "num_input_tokens_seen": 14299328,
      "step": 21910
    },
    {
      "epoch": 11.485849056603774,
      "grad_norm": 0.13570396602153778,
      "learning_rate": 0.0004576980368608658,
      "loss": 0.3914,
      "num_input_tokens_seen": 14302272,
      "step": 21915
    },
    {
      "epoch": 11.48846960167715,
      "grad_norm": 0.06337308883666992,
      "learning_rate": 0.0004574701755105862,
      "loss": 0.453,
      "num_input_tokens_seen": 14306432,
      "step": 21920
    },
    {
      "epoch": 11.491090146750524,
      "grad_norm": 0.07752560824155807,
      "learning_rate": 0.00045724232305706066,
      "loss": 0.5852,
      "num_input_tokens_seen": 14311680,
      "step": 21925
    },
    {
      "epoch": 11.4937106918239,
      "grad_norm": 0.14413230121135712,
      "learning_rate": 0.00045701447954795343,
      "loss": 0.5088,
      "num_input_tokens_seen": 14314208,
      "step": 21930
    },
    {
      "epoch": 11.496331236897275,
      "grad_norm": 0.11250054836273193,
      "learning_rate": 0.0004567866450309268,
      "loss": 0.5988,
      "num_input_tokens_seen": 14317696,
      "step": 21935
    },
    {
      "epoch": 11.49895178197065,
      "grad_norm": 0.12519246339797974,
      "learning_rate": 0.00045655881955364086,
      "loss": 0.4862,
      "num_input_tokens_seen": 14320864,
      "step": 21940
    },
    {
      "epoch": 11.501572327044025,
      "grad_norm": 0.22792933881282806,
      "learning_rate": 0.0004563310031637543,
      "loss": 0.4587,
      "num_input_tokens_seen": 14323520,
      "step": 21945
    },
    {
      "epoch": 11.5041928721174,
      "grad_norm": 0.09028821438550949,
      "learning_rate": 0.00045610319590892365,
      "loss": 0.558,
      "num_input_tokens_seen": 14326400,
      "step": 21950
    },
    {
      "epoch": 11.506813417190775,
      "grad_norm": 0.2416311353445053,
      "learning_rate": 0.00045587539783680366,
      "loss": 0.3354,
      "num_input_tokens_seen": 14328992,
      "step": 21955
    },
    {
      "epoch": 11.50943396226415,
      "grad_norm": 0.11773132532835007,
      "learning_rate": 0.00045564760899504715,
      "loss": 0.459,
      "num_input_tokens_seen": 14331584,
      "step": 21960
    },
    {
      "epoch": 11.512054507337526,
      "grad_norm": 0.13302236795425415,
      "learning_rate": 0.00045541982943130466,
      "loss": 0.4596,
      "num_input_tokens_seen": 14335360,
      "step": 21965
    },
    {
      "epoch": 11.514675052410901,
      "grad_norm": 0.08664722740650177,
      "learning_rate": 0.00045519205919322534,
      "loss": 0.5468,
      "num_input_tokens_seen": 14339360,
      "step": 21970
    },
    {
      "epoch": 11.517295597484276,
      "grad_norm": 0.1462964117527008,
      "learning_rate": 0.00045496429832845594,
      "loss": 0.4828,
      "num_input_tokens_seen": 14342112,
      "step": 21975
    },
    {
      "epoch": 11.519916142557651,
      "grad_norm": 0.08596144616603851,
      "learning_rate": 0.0004547365468846416,
      "loss": 0.3619,
      "num_input_tokens_seen": 14344960,
      "step": 21980
    },
    {
      "epoch": 11.522536687631026,
      "grad_norm": 0.18008600175380707,
      "learning_rate": 0.0004545088049094252,
      "loss": 0.433,
      "num_input_tokens_seen": 14348384,
      "step": 21985
    },
    {
      "epoch": 11.525157232704402,
      "grad_norm": 0.2696985900402069,
      "learning_rate": 0.00045428107245044805,
      "loss": 0.5784,
      "num_input_tokens_seen": 14351616,
      "step": 21990
    },
    {
      "epoch": 11.527777777777779,
      "grad_norm": 0.1143055111169815,
      "learning_rate": 0.0004540533495553488,
      "loss": 0.4518,
      "num_input_tokens_seen": 14354848,
      "step": 21995
    },
    {
      "epoch": 11.530398322851154,
      "grad_norm": 0.18773898482322693,
      "learning_rate": 0.00045382563627176454,
      "loss": 0.4675,
      "num_input_tokens_seen": 14358496,
      "step": 22000
    },
    {
      "epoch": 11.533018867924529,
      "grad_norm": 0.1441454440355301,
      "learning_rate": 0.0004535979326473304,
      "loss": 0.4605,
      "num_input_tokens_seen": 14361728,
      "step": 22005
    },
    {
      "epoch": 11.535639412997904,
      "grad_norm": 0.1438806653022766,
      "learning_rate": 0.00045337023872967945,
      "loss": 0.3267,
      "num_input_tokens_seen": 14365248,
      "step": 22010
    },
    {
      "epoch": 11.53825995807128,
      "grad_norm": 0.12434892356395721,
      "learning_rate": 0.0004531425545664425,
      "loss": 0.4533,
      "num_input_tokens_seen": 14368416,
      "step": 22015
    },
    {
      "epoch": 11.540880503144654,
      "grad_norm": 0.0981118232011795,
      "learning_rate": 0.0004529148802052488,
      "loss": 0.5493,
      "num_input_tokens_seen": 14371424,
      "step": 22020
    },
    {
      "epoch": 11.54350104821803,
      "grad_norm": 0.07261136174201965,
      "learning_rate": 0.00045268721569372483,
      "loss": 0.4264,
      "num_input_tokens_seen": 14374592,
      "step": 22025
    },
    {
      "epoch": 11.546121593291405,
      "grad_norm": 0.1117531880736351,
      "learning_rate": 0.00045245956107949547,
      "loss": 0.4144,
      "num_input_tokens_seen": 14377504,
      "step": 22030
    },
    {
      "epoch": 11.54874213836478,
      "grad_norm": 0.11132162809371948,
      "learning_rate": 0.0004522319164101837,
      "loss": 0.4449,
      "num_input_tokens_seen": 14380960,
      "step": 22035
    },
    {
      "epoch": 11.551362683438155,
      "grad_norm": 0.0989522784948349,
      "learning_rate": 0.00045200428173340987,
      "loss": 0.4168,
      "num_input_tokens_seen": 14384608,
      "step": 22040
    },
    {
      "epoch": 11.55398322851153,
      "grad_norm": 0.11139316111803055,
      "learning_rate": 0.00045177665709679286,
      "loss": 0.3427,
      "num_input_tokens_seen": 14388160,
      "step": 22045
    },
    {
      "epoch": 11.556603773584905,
      "grad_norm": 0.1907111555337906,
      "learning_rate": 0.0004515490425479492,
      "loss": 0.53,
      "num_input_tokens_seen": 14391296,
      "step": 22050
    },
    {
      "epoch": 11.55922431865828,
      "grad_norm": 0.0858990028500557,
      "learning_rate": 0.0004513214381344928,
      "loss": 0.2976,
      "num_input_tokens_seen": 14394752,
      "step": 22055
    },
    {
      "epoch": 11.561844863731656,
      "grad_norm": 0.11069311201572418,
      "learning_rate": 0.00045109384390403635,
      "loss": 0.391,
      "num_input_tokens_seen": 14397376,
      "step": 22060
    },
    {
      "epoch": 11.564465408805031,
      "grad_norm": 0.1964409351348877,
      "learning_rate": 0.0004508662599041896,
      "loss": 0.4523,
      "num_input_tokens_seen": 14399872,
      "step": 22065
    },
    {
      "epoch": 11.567085953878406,
      "grad_norm": 0.10011068731546402,
      "learning_rate": 0.0004506386861825609,
      "loss": 0.4942,
      "num_input_tokens_seen": 14403552,
      "step": 22070
    },
    {
      "epoch": 11.569706498951781,
      "grad_norm": 0.09946881234645844,
      "learning_rate": 0.00045041112278675603,
      "loss": 0.4289,
      "num_input_tokens_seen": 14407552,
      "step": 22075
    },
    {
      "epoch": 11.572327044025156,
      "grad_norm": 0.08242810517549515,
      "learning_rate": 0.00045018356976437836,
      "loss": 0.4247,
      "num_input_tokens_seen": 14410912,
      "step": 22080
    },
    {
      "epoch": 11.574947589098532,
      "grad_norm": 0.13274291157722473,
      "learning_rate": 0.0004499560271630295,
      "loss": 0.4926,
      "num_input_tokens_seen": 14413440,
      "step": 22085
    },
    {
      "epoch": 11.577568134171909,
      "grad_norm": 0.07404975593090057,
      "learning_rate": 0.00044972849503030894,
      "loss": 0.3534,
      "num_input_tokens_seen": 14417568,
      "step": 22090
    },
    {
      "epoch": 11.580188679245284,
      "grad_norm": 0.08681471645832062,
      "learning_rate": 0.00044950097341381356,
      "loss": 0.4072,
      "num_input_tokens_seen": 14420736,
      "step": 22095
    },
    {
      "epoch": 11.582809224318659,
      "grad_norm": 0.11664872616529465,
      "learning_rate": 0.0004492734623611385,
      "loss": 0.4377,
      "num_input_tokens_seen": 14424224,
      "step": 22100
    },
    {
      "epoch": 11.585429769392034,
      "grad_norm": 0.10658174753189087,
      "learning_rate": 0.0004490459619198766,
      "loss": 0.3987,
      "num_input_tokens_seen": 14427264,
      "step": 22105
    },
    {
      "epoch": 11.58805031446541,
      "grad_norm": 0.19195567071437836,
      "learning_rate": 0.00044881847213761787,
      "loss": 0.5332,
      "num_input_tokens_seen": 14429504,
      "step": 22110
    },
    {
      "epoch": 11.590670859538784,
      "grad_norm": 0.10207679122686386,
      "learning_rate": 0.000448590993061951,
      "loss": 0.4922,
      "num_input_tokens_seen": 14432672,
      "step": 22115
    },
    {
      "epoch": 11.59329140461216,
      "grad_norm": 0.21139924228191376,
      "learning_rate": 0.00044836352474046177,
      "loss": 0.4552,
      "num_input_tokens_seen": 14435072,
      "step": 22120
    },
    {
      "epoch": 11.595911949685535,
      "grad_norm": 0.09821145236492157,
      "learning_rate": 0.00044813606722073415,
      "loss": 0.3688,
      "num_input_tokens_seen": 14439424,
      "step": 22125
    },
    {
      "epoch": 11.59853249475891,
      "grad_norm": 0.11416290700435638,
      "learning_rate": 0.00044790862055034945,
      "loss": 0.4703,
      "num_input_tokens_seen": 14441888,
      "step": 22130
    },
    {
      "epoch": 11.601153039832285,
      "grad_norm": 0.2166166454553604,
      "learning_rate": 0.00044768118477688727,
      "loss": 0.4735,
      "num_input_tokens_seen": 14444896,
      "step": 22135
    },
    {
      "epoch": 11.60377358490566,
      "grad_norm": 0.06759756058454514,
      "learning_rate": 0.00044745375994792425,
      "loss": 0.4262,
      "num_input_tokens_seen": 14447904,
      "step": 22140
    },
    {
      "epoch": 11.606394129979035,
      "grad_norm": 0.13668261468410492,
      "learning_rate": 0.00044722634611103507,
      "loss": 0.4711,
      "num_input_tokens_seen": 14450624,
      "step": 22145
    },
    {
      "epoch": 11.60901467505241,
      "grad_norm": 0.0811905786395073,
      "learning_rate": 0.0004469989433137923,
      "loss": 0.441,
      "num_input_tokens_seen": 14454240,
      "step": 22150
    },
    {
      "epoch": 11.611635220125786,
      "grad_norm": 0.10859562456607819,
      "learning_rate": 0.00044677155160376586,
      "loss": 0.3876,
      "num_input_tokens_seen": 14457088,
      "step": 22155
    },
    {
      "epoch": 11.614255765199161,
      "grad_norm": 0.17447005212306976,
      "learning_rate": 0.00044654417102852375,
      "loss": 0.4473,
      "num_input_tokens_seen": 14460608,
      "step": 22160
    },
    {
      "epoch": 11.616876310272536,
      "grad_norm": 0.06403497606515884,
      "learning_rate": 0.0004463168016356309,
      "loss": 0.5397,
      "num_input_tokens_seen": 14464064,
      "step": 22165
    },
    {
      "epoch": 11.619496855345911,
      "grad_norm": 0.11896849423646927,
      "learning_rate": 0.00044608944347265095,
      "loss": 0.4729,
      "num_input_tokens_seen": 14467584,
      "step": 22170
    },
    {
      "epoch": 11.622117400419286,
      "grad_norm": 0.10807820409536362,
      "learning_rate": 0.0004458620965871442,
      "loss": 0.4144,
      "num_input_tokens_seen": 14471328,
      "step": 22175
    },
    {
      "epoch": 11.624737945492662,
      "grad_norm": 0.0740790069103241,
      "learning_rate": 0.00044563476102666935,
      "loss": 0.291,
      "num_input_tokens_seen": 14475072,
      "step": 22180
    },
    {
      "epoch": 11.627358490566039,
      "grad_norm": 0.14182640612125397,
      "learning_rate": 0.0004454074368387821,
      "loss": 0.4281,
      "num_input_tokens_seen": 14477984,
      "step": 22185
    },
    {
      "epoch": 11.629979035639414,
      "grad_norm": 0.0481991283595562,
      "learning_rate": 0.0004451801240710365,
      "loss": 0.2446,
      "num_input_tokens_seen": 14481984,
      "step": 22190
    },
    {
      "epoch": 11.632599580712789,
      "grad_norm": 0.10488730669021606,
      "learning_rate": 0.0004449528227709834,
      "loss": 0.4411,
      "num_input_tokens_seen": 14485760,
      "step": 22195
    },
    {
      "epoch": 11.635220125786164,
      "grad_norm": 0.08807273954153061,
      "learning_rate": 0.0004447255329861716,
      "loss": 0.4907,
      "num_input_tokens_seen": 14489664,
      "step": 22200
    },
    {
      "epoch": 11.63784067085954,
      "grad_norm": 0.08206673711538315,
      "learning_rate": 0.0004444982547641479,
      "loss": 0.3189,
      "num_input_tokens_seen": 14493056,
      "step": 22205
    },
    {
      "epoch": 11.640461215932914,
      "grad_norm": 0.11979646235704422,
      "learning_rate": 0.00044427098815245594,
      "loss": 0.3221,
      "num_input_tokens_seen": 14495680,
      "step": 22210
    },
    {
      "epoch": 11.64308176100629,
      "grad_norm": 0.12120731920003891,
      "learning_rate": 0.00044404373319863763,
      "loss": 0.393,
      "num_input_tokens_seen": 14498592,
      "step": 22215
    },
    {
      "epoch": 11.645702306079665,
      "grad_norm": 0.09306160360574722,
      "learning_rate": 0.000443816489950232,
      "loss": 0.416,
      "num_input_tokens_seen": 14501888,
      "step": 22220
    },
    {
      "epoch": 11.64832285115304,
      "grad_norm": 0.13565295934677124,
      "learning_rate": 0.00044358925845477547,
      "loss": 0.4891,
      "num_input_tokens_seen": 14505088,
      "step": 22225
    },
    {
      "epoch": 11.650943396226415,
      "grad_norm": 0.07927641272544861,
      "learning_rate": 0.0004433620387598027,
      "loss": 0.4135,
      "num_input_tokens_seen": 14507744,
      "step": 22230
    },
    {
      "epoch": 11.65356394129979,
      "grad_norm": 0.17599189281463623,
      "learning_rate": 0.00044313483091284507,
      "loss": 0.4179,
      "num_input_tokens_seen": 14510560,
      "step": 22235
    },
    {
      "epoch": 11.656184486373165,
      "grad_norm": 0.09424713999032974,
      "learning_rate": 0.0004429076349614321,
      "loss": 0.4665,
      "num_input_tokens_seen": 14514432,
      "step": 22240
    },
    {
      "epoch": 11.65880503144654,
      "grad_norm": 0.11286137998104095,
      "learning_rate": 0.0004426804509530907,
      "loss": 0.4875,
      "num_input_tokens_seen": 14517664,
      "step": 22245
    },
    {
      "epoch": 11.661425576519916,
      "grad_norm": 0.11524947732686996,
      "learning_rate": 0.000442453278935345,
      "loss": 0.5195,
      "num_input_tokens_seen": 14520704,
      "step": 22250
    },
    {
      "epoch": 11.664046121593291,
      "grad_norm": 0.10905024409294128,
      "learning_rate": 0.0004422261189557167,
      "loss": 0.3605,
      "num_input_tokens_seen": 14525152,
      "step": 22255
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 0.07632412761449814,
      "learning_rate": 0.00044199897106172526,
      "loss": 0.4483,
      "num_input_tokens_seen": 14529824,
      "step": 22260
    },
    {
      "epoch": 11.669287211740041,
      "grad_norm": 0.17175881564617157,
      "learning_rate": 0.0004417718353008873,
      "loss": 0.4354,
      "num_input_tokens_seen": 14532416,
      "step": 22265
    },
    {
      "epoch": 11.671907756813416,
      "grad_norm": 0.09927878528833389,
      "learning_rate": 0.0004415447117207172,
      "loss": 0.5281,
      "num_input_tokens_seen": 14535776,
      "step": 22270
    },
    {
      "epoch": 11.674528301886792,
      "grad_norm": 0.11224659532308578,
      "learning_rate": 0.00044131760036872657,
      "loss": 0.5512,
      "num_input_tokens_seen": 14538624,
      "step": 22275
    },
    {
      "epoch": 11.677148846960169,
      "grad_norm": 0.11615828424692154,
      "learning_rate": 0.00044109050129242434,
      "loss": 0.3731,
      "num_input_tokens_seen": 14542112,
      "step": 22280
    },
    {
      "epoch": 11.679769392033544,
      "grad_norm": 0.15879575908184052,
      "learning_rate": 0.00044086341453931715,
      "loss": 0.3963,
      "num_input_tokens_seen": 14545024,
      "step": 22285
    },
    {
      "epoch": 11.682389937106919,
      "grad_norm": 0.0688292533159256,
      "learning_rate": 0.00044063634015690893,
      "loss": 0.4068,
      "num_input_tokens_seen": 14549440,
      "step": 22290
    },
    {
      "epoch": 11.685010482180294,
      "grad_norm": 0.09349813312292099,
      "learning_rate": 0.0004404092781927012,
      "loss": 0.431,
      "num_input_tokens_seen": 14552224,
      "step": 22295
    },
    {
      "epoch": 11.68763102725367,
      "grad_norm": 0.1432860642671585,
      "learning_rate": 0.00044018222869419255,
      "loss": 0.4176,
      "num_input_tokens_seen": 14555040,
      "step": 22300
    },
    {
      "epoch": 11.690251572327044,
      "grad_norm": 0.14383739233016968,
      "learning_rate": 0.00043995519170887936,
      "loss": 0.387,
      "num_input_tokens_seen": 14558016,
      "step": 22305
    },
    {
      "epoch": 11.69287211740042,
      "grad_norm": 0.08668031543493271,
      "learning_rate": 0.00043972816728425504,
      "loss": 0.4349,
      "num_input_tokens_seen": 14560640,
      "step": 22310
    },
    {
      "epoch": 11.695492662473795,
      "grad_norm": 0.12320089340209961,
      "learning_rate": 0.0004395011554678103,
      "loss": 0.3483,
      "num_input_tokens_seen": 14563264,
      "step": 22315
    },
    {
      "epoch": 11.69811320754717,
      "grad_norm": 0.15715482831001282,
      "learning_rate": 0.00043927415630703364,
      "loss": 0.4426,
      "num_input_tokens_seen": 14566432,
      "step": 22320
    },
    {
      "epoch": 11.700733752620545,
      "grad_norm": 0.09653566032648087,
      "learning_rate": 0.0004390471698494108,
      "loss": 0.4035,
      "num_input_tokens_seen": 14570176,
      "step": 22325
    },
    {
      "epoch": 11.70335429769392,
      "grad_norm": 0.07719025015830994,
      "learning_rate": 0.0004388201961424248,
      "loss": 0.5142,
      "num_input_tokens_seen": 14573536,
      "step": 22330
    },
    {
      "epoch": 11.705974842767295,
      "grad_norm": 0.11871805042028427,
      "learning_rate": 0.0004385932352335555,
      "loss": 0.3858,
      "num_input_tokens_seen": 14576672,
      "step": 22335
    },
    {
      "epoch": 11.70859538784067,
      "grad_norm": 0.08362738788127899,
      "learning_rate": 0.000438366287170281,
      "loss": 0.3979,
      "num_input_tokens_seen": 14579584,
      "step": 22340
    },
    {
      "epoch": 11.711215932914046,
      "grad_norm": 0.11988415569067001,
      "learning_rate": 0.0004381393520000759,
      "loss": 0.4608,
      "num_input_tokens_seen": 14582528,
      "step": 22345
    },
    {
      "epoch": 11.713836477987421,
      "grad_norm": 0.09066034108400345,
      "learning_rate": 0.0004379124297704127,
      "loss": 0.4578,
      "num_input_tokens_seen": 14586464,
      "step": 22350
    },
    {
      "epoch": 11.716457023060796,
      "grad_norm": 0.08970892429351807,
      "learning_rate": 0.0004376855205287609,
      "loss": 0.417,
      "num_input_tokens_seen": 14589440,
      "step": 22355
    },
    {
      "epoch": 11.719077568134171,
      "grad_norm": 0.10058770328760147,
      "learning_rate": 0.0004374586243225874,
      "loss": 0.4762,
      "num_input_tokens_seen": 14593280,
      "step": 22360
    },
    {
      "epoch": 11.721698113207546,
      "grad_norm": 0.1152944415807724,
      "learning_rate": 0.00043723174119935607,
      "loss": 0.4413,
      "num_input_tokens_seen": 14596800,
      "step": 22365
    },
    {
      "epoch": 11.724318658280922,
      "grad_norm": 0.13988764584064484,
      "learning_rate": 0.00043700487120652826,
      "loss": 0.5775,
      "num_input_tokens_seen": 14599552,
      "step": 22370
    },
    {
      "epoch": 11.726939203354299,
      "grad_norm": 0.1994173526763916,
      "learning_rate": 0.00043677801439156284,
      "loss": 0.42,
      "num_input_tokens_seen": 14601952,
      "step": 22375
    },
    {
      "epoch": 11.729559748427674,
      "grad_norm": 0.10458014160394669,
      "learning_rate": 0.00043655117080191545,
      "loss": 0.4109,
      "num_input_tokens_seen": 14606336,
      "step": 22380
    },
    {
      "epoch": 11.732180293501049,
      "grad_norm": 0.21555258333683014,
      "learning_rate": 0.00043632434048503934,
      "loss": 0.3297,
      "num_input_tokens_seen": 14608864,
      "step": 22385
    },
    {
      "epoch": 11.734800838574424,
      "grad_norm": 0.09782000631093979,
      "learning_rate": 0.00043609752348838484,
      "loss": 0.316,
      "num_input_tokens_seen": 14611488,
      "step": 22390
    },
    {
      "epoch": 11.7374213836478,
      "grad_norm": 0.12898162007331848,
      "learning_rate": 0.0004358707198593992,
      "loss": 0.3946,
      "num_input_tokens_seen": 14615296,
      "step": 22395
    },
    {
      "epoch": 11.740041928721174,
      "grad_norm": 0.0958310067653656,
      "learning_rate": 0.0004356439296455273,
      "loss": 0.4193,
      "num_input_tokens_seen": 14620096,
      "step": 22400
    },
    {
      "epoch": 11.74266247379455,
      "grad_norm": 0.1307821273803711,
      "learning_rate": 0.0004354171528942113,
      "loss": 0.6266,
      "num_input_tokens_seen": 14622912,
      "step": 22405
    },
    {
      "epoch": 11.745283018867925,
      "grad_norm": 0.11817014217376709,
      "learning_rate": 0.00043519038965288994,
      "loss": 0.5293,
      "num_input_tokens_seen": 14625568,
      "step": 22410
    },
    {
      "epoch": 11.7479035639413,
      "grad_norm": 0.14842738211154938,
      "learning_rate": 0.0004349636399689998,
      "loss": 0.5051,
      "num_input_tokens_seen": 14628576,
      "step": 22415
    },
    {
      "epoch": 11.750524109014675,
      "grad_norm": 0.1369665563106537,
      "learning_rate": 0.00043473690388997434,
      "loss": 0.4123,
      "num_input_tokens_seen": 14632096,
      "step": 22420
    },
    {
      "epoch": 11.75314465408805,
      "grad_norm": 0.2569872736930847,
      "learning_rate": 0.0004345101814632438,
      "loss": 0.2961,
      "num_input_tokens_seen": 14636192,
      "step": 22425
    },
    {
      "epoch": 11.755765199161425,
      "grad_norm": 0.2542794644832611,
      "learning_rate": 0.0004342834727362362,
      "loss": 0.4105,
      "num_input_tokens_seen": 14638944,
      "step": 22430
    },
    {
      "epoch": 11.7583857442348,
      "grad_norm": 0.22446192800998688,
      "learning_rate": 0.00043405677775637633,
      "loss": 0.4082,
      "num_input_tokens_seen": 14641088,
      "step": 22435
    },
    {
      "epoch": 11.761006289308176,
      "grad_norm": 0.11916735023260117,
      "learning_rate": 0.0004338300965710863,
      "loss": 0.339,
      "num_input_tokens_seen": 14644992,
      "step": 22440
    },
    {
      "epoch": 11.76362683438155,
      "grad_norm": 0.1508396863937378,
      "learning_rate": 0.0004336034292277853,
      "loss": 0.4648,
      "num_input_tokens_seen": 14648000,
      "step": 22445
    },
    {
      "epoch": 11.766247379454926,
      "grad_norm": 0.15040060877799988,
      "learning_rate": 0.00043337677577388916,
      "loss": 0.56,
      "num_input_tokens_seen": 14651168,
      "step": 22450
    },
    {
      "epoch": 11.768867924528301,
      "grad_norm": 0.07570187002420425,
      "learning_rate": 0.0004331501362568116,
      "loss": 0.463,
      "num_input_tokens_seen": 14654528,
      "step": 22455
    },
    {
      "epoch": 11.771488469601676,
      "grad_norm": 0.10311354696750641,
      "learning_rate": 0.00043292351072396273,
      "loss": 0.4502,
      "num_input_tokens_seen": 14657088,
      "step": 22460
    },
    {
      "epoch": 11.774109014675052,
      "grad_norm": 0.08169808238744736,
      "learning_rate": 0.0004326968992227503,
      "loss": 0.3865,
      "num_input_tokens_seen": 14660352,
      "step": 22465
    },
    {
      "epoch": 11.776729559748428,
      "grad_norm": 0.1064232885837555,
      "learning_rate": 0.0004324703018005786,
      "loss": 0.5278,
      "num_input_tokens_seen": 14664288,
      "step": 22470
    },
    {
      "epoch": 11.779350104821804,
      "grad_norm": 0.11457691341638565,
      "learning_rate": 0.00043224371850484955,
      "loss": 0.6907,
      "num_input_tokens_seen": 14668000,
      "step": 22475
    },
    {
      "epoch": 11.781970649895179,
      "grad_norm": 0.10156022012233734,
      "learning_rate": 0.0004320171493829614,
      "loss": 0.3951,
      "num_input_tokens_seen": 14671328,
      "step": 22480
    },
    {
      "epoch": 11.784591194968554,
      "grad_norm": 0.1657726764678955,
      "learning_rate": 0.00043179059448231004,
      "loss": 0.4493,
      "num_input_tokens_seen": 14674272,
      "step": 22485
    },
    {
      "epoch": 11.78721174004193,
      "grad_norm": 0.1479308307170868,
      "learning_rate": 0.0004315640538502881,
      "loss": 0.497,
      "num_input_tokens_seen": 14676896,
      "step": 22490
    },
    {
      "epoch": 11.789832285115304,
      "grad_norm": 0.48386016488075256,
      "learning_rate": 0.00043133752753428547,
      "loss": 0.4738,
      "num_input_tokens_seen": 14679840,
      "step": 22495
    },
    {
      "epoch": 11.79245283018868,
      "grad_norm": 0.10320916026830673,
      "learning_rate": 0.0004311110155816886,
      "loss": 0.444,
      "num_input_tokens_seen": 14683456,
      "step": 22500
    },
    {
      "epoch": 11.795073375262055,
      "grad_norm": 0.11250770092010498,
      "learning_rate": 0.00043088451803988153,
      "loss": 0.5423,
      "num_input_tokens_seen": 14686112,
      "step": 22505
    },
    {
      "epoch": 11.79769392033543,
      "grad_norm": 0.14208148419857025,
      "learning_rate": 0.00043065803495624465,
      "loss": 0.313,
      "num_input_tokens_seen": 14689024,
      "step": 22510
    },
    {
      "epoch": 11.800314465408805,
      "grad_norm": 0.14056508243083954,
      "learning_rate": 0.00043043156637815575,
      "loss": 0.4606,
      "num_input_tokens_seen": 14691712,
      "step": 22515
    },
    {
      "epoch": 11.80293501048218,
      "grad_norm": 0.14885959029197693,
      "learning_rate": 0.00043020511235298956,
      "loss": 0.5716,
      "num_input_tokens_seen": 14695104,
      "step": 22520
    },
    {
      "epoch": 11.805555555555555,
      "grad_norm": 0.2091265469789505,
      "learning_rate": 0.00042997867292811744,
      "loss": 0.4309,
      "num_input_tokens_seen": 14697952,
      "step": 22525
    },
    {
      "epoch": 11.80817610062893,
      "grad_norm": 0.1250302493572235,
      "learning_rate": 0.00042975224815090834,
      "loss": 0.3832,
      "num_input_tokens_seen": 14700704,
      "step": 22530
    },
    {
      "epoch": 11.810796645702306,
      "grad_norm": 0.13882794976234436,
      "learning_rate": 0.0004295258380687274,
      "loss": 0.4354,
      "num_input_tokens_seen": 14703488,
      "step": 22535
    },
    {
      "epoch": 11.81341719077568,
      "grad_norm": 0.162861168384552,
      "learning_rate": 0.000429299442728937,
      "loss": 0.5048,
      "num_input_tokens_seen": 14706752,
      "step": 22540
    },
    {
      "epoch": 11.816037735849056,
      "grad_norm": 0.1288725584745407,
      "learning_rate": 0.0004290730621788967,
      "loss": 0.3485,
      "num_input_tokens_seen": 14710176,
      "step": 22545
    },
    {
      "epoch": 11.818658280922431,
      "grad_norm": 0.10730842500925064,
      "learning_rate": 0.0004288466964659625,
      "loss": 0.5266,
      "num_input_tokens_seen": 14713376,
      "step": 22550
    },
    {
      "epoch": 11.821278825995806,
      "grad_norm": 0.10310423374176025,
      "learning_rate": 0.0004286203456374877,
      "loss": 0.4775,
      "num_input_tokens_seen": 14716864,
      "step": 22555
    },
    {
      "epoch": 11.823899371069182,
      "grad_norm": 0.12707029283046722,
      "learning_rate": 0.0004283940097408224,
      "loss": 0.3017,
      "num_input_tokens_seen": 14719776,
      "step": 22560
    },
    {
      "epoch": 11.826519916142558,
      "grad_norm": 0.08835718035697937,
      "learning_rate": 0.00042816768882331324,
      "loss": 0.4728,
      "num_input_tokens_seen": 14724512,
      "step": 22565
    },
    {
      "epoch": 11.829140461215934,
      "grad_norm": 0.14546659588813782,
      "learning_rate": 0.000427941382932304,
      "loss": 0.4436,
      "num_input_tokens_seen": 14727168,
      "step": 22570
    },
    {
      "epoch": 11.831761006289309,
      "grad_norm": 0.10552779585123062,
      "learning_rate": 0.0004277150921151354,
      "loss": 0.3258,
      "num_input_tokens_seen": 14730112,
      "step": 22575
    },
    {
      "epoch": 11.834381551362684,
      "grad_norm": 0.17119058966636658,
      "learning_rate": 0.0004274888164191448,
      "loss": 0.4424,
      "num_input_tokens_seen": 14733312,
      "step": 22580
    },
    {
      "epoch": 11.83700209643606,
      "grad_norm": 0.11171137541532516,
      "learning_rate": 0.00042726255589166666,
      "loss": 0.4505,
      "num_input_tokens_seen": 14737440,
      "step": 22585
    },
    {
      "epoch": 11.839622641509434,
      "grad_norm": 0.1316068023443222,
      "learning_rate": 0.0004270363105800321,
      "loss": 0.4736,
      "num_input_tokens_seen": 14740672,
      "step": 22590
    },
    {
      "epoch": 11.84224318658281,
      "grad_norm": 0.1519542783498764,
      "learning_rate": 0.0004268100805315688,
      "loss": 0.4578,
      "num_input_tokens_seen": 14743552,
      "step": 22595
    },
    {
      "epoch": 11.844863731656185,
      "grad_norm": 0.1081743910908699,
      "learning_rate": 0.00042658386579360165,
      "loss": 0.4449,
      "num_input_tokens_seen": 14746816,
      "step": 22600
    },
    {
      "epoch": 11.84748427672956,
      "grad_norm": 0.09854240715503693,
      "learning_rate": 0.00042635766641345213,
      "loss": 0.438,
      "num_input_tokens_seen": 14749696,
      "step": 22605
    },
    {
      "epoch": 11.850104821802935,
      "grad_norm": 0.07136220484972,
      "learning_rate": 0.0004261314824384388,
      "loss": 0.6201,
      "num_input_tokens_seen": 14753376,
      "step": 22610
    },
    {
      "epoch": 11.85272536687631,
      "grad_norm": 0.10766405612230301,
      "learning_rate": 0.00042590531391587666,
      "loss": 0.5553,
      "num_input_tokens_seen": 14756576,
      "step": 22615
    },
    {
      "epoch": 11.855345911949685,
      "grad_norm": 0.10848908126354218,
      "learning_rate": 0.00042567916089307733,
      "loss": 0.465,
      "num_input_tokens_seen": 14759552,
      "step": 22620
    },
    {
      "epoch": 11.85796645702306,
      "grad_norm": 0.0981898382306099,
      "learning_rate": 0.00042545302341734977,
      "loss": 0.4774,
      "num_input_tokens_seen": 14763424,
      "step": 22625
    },
    {
      "epoch": 11.860587002096436,
      "grad_norm": 0.20769836008548737,
      "learning_rate": 0.0004252269015359991,
      "loss": 0.4655,
      "num_input_tokens_seen": 14767392,
      "step": 22630
    },
    {
      "epoch": 11.86320754716981,
      "grad_norm": 0.13334780931472778,
      "learning_rate": 0.00042500079529632763,
      "loss": 0.5425,
      "num_input_tokens_seen": 14770272,
      "step": 22635
    },
    {
      "epoch": 11.865828092243186,
      "grad_norm": 0.15542635321617126,
      "learning_rate": 0.0004247747047456342,
      "loss": 0.4454,
      "num_input_tokens_seen": 14773664,
      "step": 22640
    },
    {
      "epoch": 11.868448637316561,
      "grad_norm": 0.20134443044662476,
      "learning_rate": 0.0004245486299312144,
      "loss": 0.3499,
      "num_input_tokens_seen": 14776640,
      "step": 22645
    },
    {
      "epoch": 11.871069182389936,
      "grad_norm": 0.11525148153305054,
      "learning_rate": 0.0004243225709003602,
      "loss": 0.4691,
      "num_input_tokens_seen": 14779584,
      "step": 22650
    },
    {
      "epoch": 11.873689727463312,
      "grad_norm": 0.12063571065664291,
      "learning_rate": 0.000424096527700361,
      "loss": 0.5427,
      "num_input_tokens_seen": 14782784,
      "step": 22655
    },
    {
      "epoch": 11.876310272536688,
      "grad_norm": 0.11164350807666779,
      "learning_rate": 0.0004238705003785021,
      "loss": 0.6191,
      "num_input_tokens_seen": 14787040,
      "step": 22660
    },
    {
      "epoch": 11.878930817610064,
      "grad_norm": 0.1562764197587967,
      "learning_rate": 0.0004236444889820661,
      "loss": 0.455,
      "num_input_tokens_seen": 14790560,
      "step": 22665
    },
    {
      "epoch": 11.881551362683439,
      "grad_norm": 0.11906570196151733,
      "learning_rate": 0.0004234184935583318,
      "loss": 0.5027,
      "num_input_tokens_seen": 14794976,
      "step": 22670
    },
    {
      "epoch": 11.884171907756814,
      "grad_norm": 0.08814739435911179,
      "learning_rate": 0.00042319251415457517,
      "loss": 0.4681,
      "num_input_tokens_seen": 14798368,
      "step": 22675
    },
    {
      "epoch": 11.88679245283019,
      "grad_norm": 0.08687734603881836,
      "learning_rate": 0.0004229665508180681,
      "loss": 0.4236,
      "num_input_tokens_seen": 14801536,
      "step": 22680
    },
    {
      "epoch": 11.889412997903564,
      "grad_norm": 0.18107767403125763,
      "learning_rate": 0.0004227406035960798,
      "loss": 0.3399,
      "num_input_tokens_seen": 14804544,
      "step": 22685
    },
    {
      "epoch": 11.89203354297694,
      "grad_norm": 0.13001030683517456,
      "learning_rate": 0.0004225146725358758,
      "loss": 0.3305,
      "num_input_tokens_seen": 14807328,
      "step": 22690
    },
    {
      "epoch": 11.894654088050315,
      "grad_norm": 0.1252908706665039,
      "learning_rate": 0.0004222887576847183,
      "loss": 0.4008,
      "num_input_tokens_seen": 14810464,
      "step": 22695
    },
    {
      "epoch": 11.89727463312369,
      "grad_norm": 0.16291958093643188,
      "learning_rate": 0.00042206285908986626,
      "loss": 0.4954,
      "num_input_tokens_seen": 14814080,
      "step": 22700
    },
    {
      "epoch": 11.899895178197065,
      "grad_norm": 0.1599496304988861,
      "learning_rate": 0.00042183697679857484,
      "loss": 0.5774,
      "num_input_tokens_seen": 14818368,
      "step": 22705
    },
    {
      "epoch": 11.90251572327044,
      "grad_norm": 0.07774186134338379,
      "learning_rate": 0.00042161111085809604,
      "loss": 0.4135,
      "num_input_tokens_seen": 14821344,
      "step": 22710
    },
    {
      "epoch": 11.905136268343815,
      "grad_norm": 0.09777122735977173,
      "learning_rate": 0.00042138526131567855,
      "loss": 0.356,
      "num_input_tokens_seen": 14824352,
      "step": 22715
    },
    {
      "epoch": 11.90775681341719,
      "grad_norm": 0.1266101449728012,
      "learning_rate": 0.0004211594282185677,
      "loss": 0.4461,
      "num_input_tokens_seen": 14828704,
      "step": 22720
    },
    {
      "epoch": 11.910377358490566,
      "grad_norm": 0.18085269629955292,
      "learning_rate": 0.0004209336116140048,
      "loss": 0.4768,
      "num_input_tokens_seen": 14831456,
      "step": 22725
    },
    {
      "epoch": 11.91299790356394,
      "grad_norm": 0.10634468495845795,
      "learning_rate": 0.00042070781154922857,
      "loss": 0.4083,
      "num_input_tokens_seen": 14833856,
      "step": 22730
    },
    {
      "epoch": 11.915618448637316,
      "grad_norm": 0.12067323178052902,
      "learning_rate": 0.0004204820280714734,
      "loss": 0.309,
      "num_input_tokens_seen": 14837120,
      "step": 22735
    },
    {
      "epoch": 11.918238993710691,
      "grad_norm": 0.11957793682813644,
      "learning_rate": 0.00042025626122797066,
      "loss": 0.4435,
      "num_input_tokens_seen": 14841056,
      "step": 22740
    },
    {
      "epoch": 11.920859538784066,
      "grad_norm": 0.07237277925014496,
      "learning_rate": 0.0004200305110659484,
      "loss": 0.4415,
      "num_input_tokens_seen": 14844064,
      "step": 22745
    },
    {
      "epoch": 11.923480083857442,
      "grad_norm": 0.18096928298473358,
      "learning_rate": 0.0004198047776326308,
      "loss": 0.4565,
      "num_input_tokens_seen": 14846656,
      "step": 22750
    },
    {
      "epoch": 11.926100628930818,
      "grad_norm": 0.09245671331882477,
      "learning_rate": 0.00041957906097523897,
      "loss": 0.5028,
      "num_input_tokens_seen": 14849248,
      "step": 22755
    },
    {
      "epoch": 11.928721174004194,
      "grad_norm": 0.09005538374185562,
      "learning_rate": 0.00041935336114099013,
      "loss": 0.4526,
      "num_input_tokens_seen": 14852384,
      "step": 22760
    },
    {
      "epoch": 11.931341719077569,
      "grad_norm": 0.13436150550842285,
      "learning_rate": 0.0004191276781770979,
      "loss": 0.3801,
      "num_input_tokens_seen": 14854720,
      "step": 22765
    },
    {
      "epoch": 11.933962264150944,
      "grad_norm": 0.16244852542877197,
      "learning_rate": 0.00041890201213077286,
      "loss": 0.4557,
      "num_input_tokens_seen": 14857696,
      "step": 22770
    },
    {
      "epoch": 11.93658280922432,
      "grad_norm": 0.1620258390903473,
      "learning_rate": 0.00041867636304922153,
      "loss": 0.5389,
      "num_input_tokens_seen": 14861216,
      "step": 22775
    },
    {
      "epoch": 11.939203354297694,
      "grad_norm": 0.0753367468714714,
      "learning_rate": 0.00041845073097964737,
      "loss": 0.452,
      "num_input_tokens_seen": 14864064,
      "step": 22780
    },
    {
      "epoch": 11.94182389937107,
      "grad_norm": 0.181405171751976,
      "learning_rate": 0.0004182251159692498,
      "loss": 0.4536,
      "num_input_tokens_seen": 14866784,
      "step": 22785
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 0.10442215949296951,
      "learning_rate": 0.0004179995180652253,
      "loss": 0.3728,
      "num_input_tokens_seen": 14869664,
      "step": 22790
    },
    {
      "epoch": 11.94706498951782,
      "grad_norm": 0.09983904659748077,
      "learning_rate": 0.00041777393731476587,
      "loss": 0.4623,
      "num_input_tokens_seen": 14872640,
      "step": 22795
    },
    {
      "epoch": 11.949685534591195,
      "grad_norm": 0.11255521327257156,
      "learning_rate": 0.0004175483737650608,
      "loss": 0.4321,
      "num_input_tokens_seen": 14876512,
      "step": 22800
    },
    {
      "epoch": 11.95230607966457,
      "grad_norm": 0.14846093952655792,
      "learning_rate": 0.0004173228274632951,
      "loss": 0.5124,
      "num_input_tokens_seen": 14879776,
      "step": 22805
    },
    {
      "epoch": 11.954926624737945,
      "grad_norm": 0.10466966778039932,
      "learning_rate": 0.00041709729845665084,
      "loss": 0.3505,
      "num_input_tokens_seen": 14883552,
      "step": 22810
    },
    {
      "epoch": 11.95754716981132,
      "grad_norm": 0.22858545184135437,
      "learning_rate": 0.0004168717867923061,
      "loss": 0.4351,
      "num_input_tokens_seen": 14885984,
      "step": 22815
    },
    {
      "epoch": 11.960167714884696,
      "grad_norm": 0.09673737734556198,
      "learning_rate": 0.00041664629251743487,
      "loss": 0.4873,
      "num_input_tokens_seen": 14888992,
      "step": 22820
    },
    {
      "epoch": 11.96278825995807,
      "grad_norm": 0.0951196700334549,
      "learning_rate": 0.0004164208156792084,
      "loss": 0.4068,
      "num_input_tokens_seen": 14891744,
      "step": 22825
    },
    {
      "epoch": 11.965408805031446,
      "grad_norm": 0.09863711148500443,
      "learning_rate": 0.00041619535632479374,
      "loss": 0.5988,
      "num_input_tokens_seen": 14894880,
      "step": 22830
    },
    {
      "epoch": 11.968029350104821,
      "grad_norm": 0.10054200142621994,
      "learning_rate": 0.00041596991450135445,
      "loss": 0.4284,
      "num_input_tokens_seen": 14897760,
      "step": 22835
    },
    {
      "epoch": 11.970649895178196,
      "grad_norm": 0.10136117786169052,
      "learning_rate": 0.0004157444902560504,
      "loss": 0.4491,
      "num_input_tokens_seen": 14900736,
      "step": 22840
    },
    {
      "epoch": 11.973270440251572,
      "grad_norm": 0.09538961946964264,
      "learning_rate": 0.00041551908363603786,
      "loss": 0.4006,
      "num_input_tokens_seen": 14904544,
      "step": 22845
    },
    {
      "epoch": 11.975890985324948,
      "grad_norm": 0.11408521234989166,
      "learning_rate": 0.00041529369468846914,
      "loss": 0.5122,
      "num_input_tokens_seen": 14907648,
      "step": 22850
    },
    {
      "epoch": 11.978511530398324,
      "grad_norm": 0.09930792450904846,
      "learning_rate": 0.00041506832346049303,
      "loss": 0.3354,
      "num_input_tokens_seen": 14910816,
      "step": 22855
    },
    {
      "epoch": 11.981132075471699,
      "grad_norm": 0.11390173435211182,
      "learning_rate": 0.00041484296999925476,
      "loss": 0.3654,
      "num_input_tokens_seen": 14913632,
      "step": 22860
    },
    {
      "epoch": 11.983752620545074,
      "grad_norm": 0.09521746635437012,
      "learning_rate": 0.00041461763435189555,
      "loss": 0.443,
      "num_input_tokens_seen": 14916768,
      "step": 22865
    },
    {
      "epoch": 11.98637316561845,
      "grad_norm": 0.08953731507062912,
      "learning_rate": 0.00041439231656555315,
      "loss": 0.3665,
      "num_input_tokens_seen": 14920896,
      "step": 22870
    },
    {
      "epoch": 11.988993710691824,
      "grad_norm": 0.08677609264850616,
      "learning_rate": 0.0004141670166873617,
      "loss": 0.4037,
      "num_input_tokens_seen": 14924192,
      "step": 22875
    },
    {
      "epoch": 11.9916142557652,
      "grad_norm": 0.06816968321800232,
      "learning_rate": 0.00041394173476445097,
      "loss": 0.4526,
      "num_input_tokens_seen": 14926784,
      "step": 22880
    },
    {
      "epoch": 11.994234800838575,
      "grad_norm": 0.13304492831230164,
      "learning_rate": 0.00041371647084394744,
      "loss": 0.499,
      "num_input_tokens_seen": 14930432,
      "step": 22885
    },
    {
      "epoch": 11.99685534591195,
      "grad_norm": 0.15913167595863342,
      "learning_rate": 0.00041349122497297394,
      "loss": 0.5295,
      "num_input_tokens_seen": 14933664,
      "step": 22890
    },
    {
      "epoch": 11.999475890985325,
      "grad_norm": 0.1119820699095726,
      "learning_rate": 0.0004132659971986491,
      "loss": 0.4353,
      "num_input_tokens_seen": 14937088,
      "step": 22895
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.45972055196762085,
      "eval_runtime": 13.5506,
      "eval_samples_per_second": 62.58,
      "eval_steps_per_second": 15.645,
      "num_input_tokens_seen": 14937216,
      "step": 22896
    },
    {
      "epoch": 12.0020964360587,
      "grad_norm": 0.1302689015865326,
      "learning_rate": 0.00041304078756808837,
      "loss": 0.4365,
      "num_input_tokens_seen": 14939712,
      "step": 22900
    },
    {
      "epoch": 12.004716981132075,
      "grad_norm": 0.1681492179632187,
      "learning_rate": 0.0004128155961284027,
      "loss": 0.461,
      "num_input_tokens_seen": 14942272,
      "step": 22905
    },
    {
      "epoch": 12.00733752620545,
      "grad_norm": 0.1266041100025177,
      "learning_rate": 0.0004125904229266996,
      "loss": 0.5216,
      "num_input_tokens_seen": 14945632,
      "step": 22910
    },
    {
      "epoch": 12.009958071278826,
      "grad_norm": 0.09357091784477234,
      "learning_rate": 0.0004123652680100828,
      "loss": 0.6418,
      "num_input_tokens_seen": 14949376,
      "step": 22915
    },
    {
      "epoch": 12.0125786163522,
      "grad_norm": 0.12398409098386765,
      "learning_rate": 0.0004121401314256521,
      "loss": 0.4902,
      "num_input_tokens_seen": 14951840,
      "step": 22920
    },
    {
      "epoch": 12.015199161425576,
      "grad_norm": 0.08281564712524414,
      "learning_rate": 0.00041191501322050363,
      "loss": 0.3243,
      "num_input_tokens_seen": 14956768,
      "step": 22925
    },
    {
      "epoch": 12.017819706498951,
      "grad_norm": 0.2151864916086197,
      "learning_rate": 0.0004116899134417296,
      "loss": 0.3588,
      "num_input_tokens_seen": 14960384,
      "step": 22930
    },
    {
      "epoch": 12.020440251572326,
      "grad_norm": 0.09787207841873169,
      "learning_rate": 0.00041146483213641793,
      "loss": 0.3649,
      "num_input_tokens_seen": 14963392,
      "step": 22935
    },
    {
      "epoch": 12.023060796645701,
      "grad_norm": 0.11873041838407516,
      "learning_rate": 0.00041123976935165355,
      "loss": 0.4234,
      "num_input_tokens_seen": 14967392,
      "step": 22940
    },
    {
      "epoch": 12.025681341719078,
      "grad_norm": 0.07839503139257431,
      "learning_rate": 0.0004110147251345165,
      "loss": 0.432,
      "num_input_tokens_seen": 14972704,
      "step": 22945
    },
    {
      "epoch": 12.028301886792454,
      "grad_norm": 0.10687728971242905,
      "learning_rate": 0.0004107896995320839,
      "loss": 0.383,
      "num_input_tokens_seen": 14975744,
      "step": 22950
    },
    {
      "epoch": 12.030922431865829,
      "grad_norm": 0.07378202676773071,
      "learning_rate": 0.0004105646925914284,
      "loss": 0.4466,
      "num_input_tokens_seen": 14979104,
      "step": 22955
    },
    {
      "epoch": 12.033542976939204,
      "grad_norm": 0.10677337646484375,
      "learning_rate": 0.00041033970435961904,
      "loss": 0.5187,
      "num_input_tokens_seen": 14981600,
      "step": 22960
    },
    {
      "epoch": 12.036163522012579,
      "grad_norm": 0.1484881341457367,
      "learning_rate": 0.00041011473488372044,
      "loss": 0.4048,
      "num_input_tokens_seen": 14985152,
      "step": 22965
    },
    {
      "epoch": 12.038784067085954,
      "grad_norm": 0.09953109174966812,
      "learning_rate": 0.0004098897842107939,
      "loss": 0.3256,
      "num_input_tokens_seen": 14989568,
      "step": 22970
    },
    {
      "epoch": 12.04140461215933,
      "grad_norm": 0.1278437376022339,
      "learning_rate": 0.00040966485238789634,
      "loss": 0.4446,
      "num_input_tokens_seen": 14992352,
      "step": 22975
    },
    {
      "epoch": 12.044025157232705,
      "grad_norm": 0.06349020451307297,
      "learning_rate": 0.00040943993946208126,
      "loss": 0.3498,
      "num_input_tokens_seen": 14995168,
      "step": 22980
    },
    {
      "epoch": 12.04664570230608,
      "grad_norm": 0.14481152594089508,
      "learning_rate": 0.00040921504548039765,
      "loss": 0.5716,
      "num_input_tokens_seen": 14998240,
      "step": 22985
    },
    {
      "epoch": 12.049266247379455,
      "grad_norm": 0.16627052426338196,
      "learning_rate": 0.00040899017048989073,
      "loss": 0.4735,
      "num_input_tokens_seen": 15001248,
      "step": 22990
    },
    {
      "epoch": 12.05188679245283,
      "grad_norm": 0.09220944344997406,
      "learning_rate": 0.00040876531453760186,
      "loss": 0.5159,
      "num_input_tokens_seen": 15004352,
      "step": 22995
    },
    {
      "epoch": 12.054507337526205,
      "grad_norm": 0.2085474282503128,
      "learning_rate": 0.00040854047767056824,
      "loss": 0.3923,
      "num_input_tokens_seen": 15007968,
      "step": 23000
    },
    {
      "epoch": 12.05712788259958,
      "grad_norm": 0.13074921071529388,
      "learning_rate": 0.00040831565993582335,
      "loss": 0.3887,
      "num_input_tokens_seen": 15011648,
      "step": 23005
    },
    {
      "epoch": 12.059748427672956,
      "grad_norm": 0.148638516664505,
      "learning_rate": 0.0004080908613803964,
      "loss": 0.4027,
      "num_input_tokens_seen": 15014176,
      "step": 23010
    },
    {
      "epoch": 12.06236897274633,
      "grad_norm": 0.09321694821119308,
      "learning_rate": 0.0004078660820513128,
      "loss": 0.4793,
      "num_input_tokens_seen": 15017184,
      "step": 23015
    },
    {
      "epoch": 12.064989517819706,
      "grad_norm": 0.15969863533973694,
      "learning_rate": 0.0004076413219955937,
      "loss": 0.3679,
      "num_input_tokens_seen": 15020416,
      "step": 23020
    },
    {
      "epoch": 12.067610062893081,
      "grad_norm": 0.11298465728759766,
      "learning_rate": 0.00040741658126025614,
      "loss": 0.3905,
      "num_input_tokens_seen": 15024800,
      "step": 23025
    },
    {
      "epoch": 12.070230607966456,
      "grad_norm": 0.19409878551959991,
      "learning_rate": 0.0004071918598923136,
      "loss": 0.3399,
      "num_input_tokens_seen": 15027456,
      "step": 23030
    },
    {
      "epoch": 12.072851153039831,
      "grad_norm": 0.1247546449303627,
      "learning_rate": 0.0004069671579387753,
      "loss": 0.5479,
      "num_input_tokens_seen": 15030208,
      "step": 23035
    },
    {
      "epoch": 12.075471698113208,
      "grad_norm": 0.14940744638442993,
      "learning_rate": 0.000406742475446646,
      "loss": 0.3814,
      "num_input_tokens_seen": 15033120,
      "step": 23040
    },
    {
      "epoch": 12.078092243186584,
      "grad_norm": 0.11548717319965363,
      "learning_rate": 0.0004065178124629272,
      "loss": 0.4861,
      "num_input_tokens_seen": 15036960,
      "step": 23045
    },
    {
      "epoch": 12.080712788259959,
      "grad_norm": 0.08447996526956558,
      "learning_rate": 0.0004062931690346153,
      "loss": 0.4285,
      "num_input_tokens_seen": 15040992,
      "step": 23050
    },
    {
      "epoch": 12.083333333333334,
      "grad_norm": 0.13066624104976654,
      "learning_rate": 0.0004060685452087034,
      "loss": 0.4371,
      "num_input_tokens_seen": 15044096,
      "step": 23055
    },
    {
      "epoch": 12.085953878406709,
      "grad_norm": 0.14069442451000214,
      "learning_rate": 0.00040584394103218026,
      "loss": 0.4651,
      "num_input_tokens_seen": 15047680,
      "step": 23060
    },
    {
      "epoch": 12.088574423480084,
      "grad_norm": 0.09867963194847107,
      "learning_rate": 0.0004056193565520304,
      "loss": 0.3653,
      "num_input_tokens_seen": 15051328,
      "step": 23065
    },
    {
      "epoch": 12.09119496855346,
      "grad_norm": 0.1375451236963272,
      "learning_rate": 0.0004053947918152344,
      "loss": 0.3715,
      "num_input_tokens_seen": 15054560,
      "step": 23070
    },
    {
      "epoch": 12.093815513626835,
      "grad_norm": 0.08046913892030716,
      "learning_rate": 0.0004051702468687688,
      "loss": 0.4491,
      "num_input_tokens_seen": 15057888,
      "step": 23075
    },
    {
      "epoch": 12.09643605870021,
      "grad_norm": 0.1334066540002823,
      "learning_rate": 0.0004049457217596055,
      "loss": 0.4924,
      "num_input_tokens_seen": 15061120,
      "step": 23080
    },
    {
      "epoch": 12.099056603773585,
      "grad_norm": 0.10567878931760788,
      "learning_rate": 0.0004047212165347129,
      "loss": 0.4407,
      "num_input_tokens_seen": 15064224,
      "step": 23085
    },
    {
      "epoch": 12.10167714884696,
      "grad_norm": 0.10379918664693832,
      "learning_rate": 0.00040449673124105464,
      "loss": 0.4178,
      "num_input_tokens_seen": 15067360,
      "step": 23090
    },
    {
      "epoch": 12.104297693920335,
      "grad_norm": 0.11455219238996506,
      "learning_rate": 0.00040427226592559064,
      "loss": 0.3697,
      "num_input_tokens_seen": 15070688,
      "step": 23095
    },
    {
      "epoch": 12.10691823899371,
      "grad_norm": 0.21611690521240234,
      "learning_rate": 0.00040404782063527684,
      "loss": 0.421,
      "num_input_tokens_seen": 15074304,
      "step": 23100
    },
    {
      "epoch": 12.109538784067086,
      "grad_norm": 0.1538778692483902,
      "learning_rate": 0.00040382339541706393,
      "loss": 0.4695,
      "num_input_tokens_seen": 15076768,
      "step": 23105
    },
    {
      "epoch": 12.11215932914046,
      "grad_norm": 0.15582790970802307,
      "learning_rate": 0.0004035989903178995,
      "loss": 0.4235,
      "num_input_tokens_seen": 15079232,
      "step": 23110
    },
    {
      "epoch": 12.114779874213836,
      "grad_norm": 0.10361269116401672,
      "learning_rate": 0.0004033746053847266,
      "loss": 0.356,
      "num_input_tokens_seen": 15082688,
      "step": 23115
    },
    {
      "epoch": 12.117400419287211,
      "grad_norm": 0.07695990800857544,
      "learning_rate": 0.00040315024066448374,
      "loss": 0.3729,
      "num_input_tokens_seen": 15086400,
      "step": 23120
    },
    {
      "epoch": 12.120020964360586,
      "grad_norm": 0.1016443520784378,
      "learning_rate": 0.00040292589620410577,
      "loss": 0.3867,
      "num_input_tokens_seen": 15089184,
      "step": 23125
    },
    {
      "epoch": 12.122641509433961,
      "grad_norm": 0.11160346865653992,
      "learning_rate": 0.0004027015720505229,
      "loss": 0.331,
      "num_input_tokens_seen": 15092384,
      "step": 23130
    },
    {
      "epoch": 12.125262054507338,
      "grad_norm": 0.1557365208864212,
      "learning_rate": 0.0004024772682506609,
      "loss": 0.4275,
      "num_input_tokens_seen": 15095776,
      "step": 23135
    },
    {
      "epoch": 12.127882599580714,
      "grad_norm": 0.14854510128498077,
      "learning_rate": 0.0004022529848514419,
      "loss": 0.5171,
      "num_input_tokens_seen": 15098624,
      "step": 23140
    },
    {
      "epoch": 12.130503144654089,
      "grad_norm": 0.10207197070121765,
      "learning_rate": 0.00040202872189978324,
      "loss": 0.4417,
      "num_input_tokens_seen": 15100992,
      "step": 23145
    },
    {
      "epoch": 12.133123689727464,
      "grad_norm": 0.11456894874572754,
      "learning_rate": 0.0004018044794425983,
      "loss": 0.339,
      "num_input_tokens_seen": 15104352,
      "step": 23150
    },
    {
      "epoch": 12.135744234800839,
      "grad_norm": 0.17928002774715424,
      "learning_rate": 0.00040158025752679596,
      "loss": 0.4115,
      "num_input_tokens_seen": 15107200,
      "step": 23155
    },
    {
      "epoch": 12.138364779874214,
      "grad_norm": 0.1612376868724823,
      "learning_rate": 0.0004013560561992811,
      "loss": 0.5506,
      "num_input_tokens_seen": 15111008,
      "step": 23160
    },
    {
      "epoch": 12.14098532494759,
      "grad_norm": 0.14394475519657135,
      "learning_rate": 0.0004011318755069537,
      "loss": 0.4476,
      "num_input_tokens_seen": 15113440,
      "step": 23165
    },
    {
      "epoch": 12.143605870020965,
      "grad_norm": 0.10949521511793137,
      "learning_rate": 0.00040090771549670994,
      "loss": 0.43,
      "num_input_tokens_seen": 15116256,
      "step": 23170
    },
    {
      "epoch": 12.14622641509434,
      "grad_norm": 0.0958627462387085,
      "learning_rate": 0.00040068357621544167,
      "loss": 0.4576,
      "num_input_tokens_seen": 15119296,
      "step": 23175
    },
    {
      "epoch": 12.148846960167715,
      "grad_norm": 0.08902879059314728,
      "learning_rate": 0.00040045945771003625,
      "loss": 0.3395,
      "num_input_tokens_seen": 15123584,
      "step": 23180
    },
    {
      "epoch": 12.15146750524109,
      "grad_norm": 0.10299093276262283,
      "learning_rate": 0.0004002353600273767,
      "loss": 0.4197,
      "num_input_tokens_seen": 15127008,
      "step": 23185
    },
    {
      "epoch": 12.154088050314465,
      "grad_norm": 0.10341963917016983,
      "learning_rate": 0.0004000112832143415,
      "loss": 0.4162,
      "num_input_tokens_seen": 15129728,
      "step": 23190
    },
    {
      "epoch": 12.15670859538784,
      "grad_norm": 0.09282957017421722,
      "learning_rate": 0.00039978722731780523,
      "loss": 0.479,
      "num_input_tokens_seen": 15133152,
      "step": 23195
    },
    {
      "epoch": 12.159329140461216,
      "grad_norm": 0.0813634991645813,
      "learning_rate": 0.00039956319238463754,
      "loss": 0.4822,
      "num_input_tokens_seen": 15136960,
      "step": 23200
    },
    {
      "epoch": 12.16194968553459,
      "grad_norm": 0.10740816593170166,
      "learning_rate": 0.00039933917846170436,
      "loss": 0.4559,
      "num_input_tokens_seen": 15140128,
      "step": 23205
    },
    {
      "epoch": 12.164570230607966,
      "grad_norm": 0.06669016182422638,
      "learning_rate": 0.0003991151855958665,
      "loss": 0.4251,
      "num_input_tokens_seen": 15144128,
      "step": 23210
    },
    {
      "epoch": 12.167190775681341,
      "grad_norm": 0.1811169981956482,
      "learning_rate": 0.00039889121383398113,
      "loss": 0.448,
      "num_input_tokens_seen": 15147968,
      "step": 23215
    },
    {
      "epoch": 12.169811320754716,
      "grad_norm": 0.11461876332759857,
      "learning_rate": 0.00039866726322290014,
      "loss": 0.3529,
      "num_input_tokens_seen": 15150976,
      "step": 23220
    },
    {
      "epoch": 12.172431865828091,
      "grad_norm": 0.1329086869955063,
      "learning_rate": 0.0003984433338094715,
      "loss": 0.6012,
      "num_input_tokens_seen": 15154368,
      "step": 23225
    },
    {
      "epoch": 12.175052410901468,
      "grad_norm": 0.16625849902629852,
      "learning_rate": 0.000398219425640539,
      "loss": 0.4658,
      "num_input_tokens_seen": 15157984,
      "step": 23230
    },
    {
      "epoch": 12.177672955974844,
      "grad_norm": 0.16752813756465912,
      "learning_rate": 0.0003979955387629413,
      "loss": 0.5087,
      "num_input_tokens_seen": 15162208,
      "step": 23235
    },
    {
      "epoch": 12.180293501048219,
      "grad_norm": 0.12895214557647705,
      "learning_rate": 0.0003977716732235133,
      "loss": 0.4283,
      "num_input_tokens_seen": 15165088,
      "step": 23240
    },
    {
      "epoch": 12.182914046121594,
      "grad_norm": 0.13995233178138733,
      "learning_rate": 0.000397547829069085,
      "loss": 0.3354,
      "num_input_tokens_seen": 15167200,
      "step": 23245
    },
    {
      "epoch": 12.185534591194969,
      "grad_norm": 0.11020921915769577,
      "learning_rate": 0.000397324006346482,
      "loss": 0.4986,
      "num_input_tokens_seen": 15170304,
      "step": 23250
    },
    {
      "epoch": 12.188155136268344,
      "grad_norm": 0.23926325142383575,
      "learning_rate": 0.0003971002051025253,
      "loss": 0.4516,
      "num_input_tokens_seen": 15173024,
      "step": 23255
    },
    {
      "epoch": 12.19077568134172,
      "grad_norm": 0.29370394349098206,
      "learning_rate": 0.000396876425384032,
      "loss": 0.5371,
      "num_input_tokens_seen": 15176960,
      "step": 23260
    },
    {
      "epoch": 12.193396226415095,
      "grad_norm": 0.12364084273576736,
      "learning_rate": 0.00039665266723781377,
      "loss": 0.4434,
      "num_input_tokens_seen": 15184576,
      "step": 23265
    },
    {
      "epoch": 12.19601677148847,
      "grad_norm": 0.11330728977918625,
      "learning_rate": 0.00039642893071067877,
      "loss": 0.4765,
      "num_input_tokens_seen": 15187840,
      "step": 23270
    },
    {
      "epoch": 12.198637316561845,
      "grad_norm": 0.08352181315422058,
      "learning_rate": 0.0003962052158494298,
      "loss": 0.3812,
      "num_input_tokens_seen": 15191072,
      "step": 23275
    },
    {
      "epoch": 12.20125786163522,
      "grad_norm": 0.12164551764726639,
      "learning_rate": 0.00039598152270086534,
      "loss": 0.4022,
      "num_input_tokens_seen": 15194592,
      "step": 23280
    },
    {
      "epoch": 12.203878406708595,
      "grad_norm": 0.17371566593647003,
      "learning_rate": 0.00039575785131177975,
      "loss": 0.4183,
      "num_input_tokens_seen": 15198240,
      "step": 23285
    },
    {
      "epoch": 12.20649895178197,
      "grad_norm": 0.1612064242362976,
      "learning_rate": 0.0003955342017289624,
      "loss": 0.4754,
      "num_input_tokens_seen": 15201056,
      "step": 23290
    },
    {
      "epoch": 12.209119496855346,
      "grad_norm": 0.13444536924362183,
      "learning_rate": 0.0003953105739991982,
      "loss": 0.4722,
      "num_input_tokens_seen": 15204320,
      "step": 23295
    },
    {
      "epoch": 12.21174004192872,
      "grad_norm": 0.1041954755783081,
      "learning_rate": 0.0003950869681692678,
      "loss": 0.4227,
      "num_input_tokens_seen": 15207264,
      "step": 23300
    },
    {
      "epoch": 12.214360587002096,
      "grad_norm": 0.15584617853164673,
      "learning_rate": 0.0003948633842859465,
      "loss": 0.3438,
      "num_input_tokens_seen": 15210112,
      "step": 23305
    },
    {
      "epoch": 12.216981132075471,
      "grad_norm": 0.07940918952226639,
      "learning_rate": 0.00039463982239600575,
      "loss": 0.3426,
      "num_input_tokens_seen": 15213504,
      "step": 23310
    },
    {
      "epoch": 12.219601677148846,
      "grad_norm": 0.10675273835659027,
      "learning_rate": 0.00039441628254621215,
      "loss": 0.458,
      "num_input_tokens_seen": 15215808,
      "step": 23315
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 0.1824306696653366,
      "learning_rate": 0.00039419276478332773,
      "loss": 0.5023,
      "num_input_tokens_seen": 15218304,
      "step": 23320
    },
    {
      "epoch": 12.224842767295598,
      "grad_norm": 0.1198733001947403,
      "learning_rate": 0.0003939692691541097,
      "loss": 0.5874,
      "num_input_tokens_seen": 15221024,
      "step": 23325
    },
    {
      "epoch": 12.227463312368974,
      "grad_norm": 0.12196478247642517,
      "learning_rate": 0.00039374579570531114,
      "loss": 0.4286,
      "num_input_tokens_seen": 15223872,
      "step": 23330
    },
    {
      "epoch": 12.230083857442349,
      "grad_norm": 0.13868479430675507,
      "learning_rate": 0.0003935223444836797,
      "loss": 0.3845,
      "num_input_tokens_seen": 15227168,
      "step": 23335
    },
    {
      "epoch": 12.232704402515724,
      "grad_norm": 0.17380836606025696,
      "learning_rate": 0.0003932989155359591,
      "loss": 0.44,
      "num_input_tokens_seen": 15230816,
      "step": 23340
    },
    {
      "epoch": 12.235324947589099,
      "grad_norm": 0.20676620304584503,
      "learning_rate": 0.000393075508908888,
      "loss": 0.4287,
      "num_input_tokens_seen": 15233568,
      "step": 23345
    },
    {
      "epoch": 12.237945492662474,
      "grad_norm": 0.1949271261692047,
      "learning_rate": 0.00039285212464920064,
      "loss": 0.3287,
      "num_input_tokens_seen": 15236064,
      "step": 23350
    },
    {
      "epoch": 12.24056603773585,
      "grad_norm": 0.12641240656375885,
      "learning_rate": 0.0003926287628036265,
      "loss": 0.3825,
      "num_input_tokens_seen": 15239168,
      "step": 23355
    },
    {
      "epoch": 12.243186582809225,
      "grad_norm": 0.08563205599784851,
      "learning_rate": 0.00039240542341889003,
      "loss": 0.393,
      "num_input_tokens_seen": 15241984,
      "step": 23360
    },
    {
      "epoch": 12.2458071278826,
      "grad_norm": 0.20750193297863007,
      "learning_rate": 0.0003921821065417116,
      "loss": 0.4415,
      "num_input_tokens_seen": 15244800,
      "step": 23365
    },
    {
      "epoch": 12.248427672955975,
      "grad_norm": 0.13038483262062073,
      "learning_rate": 0.0003919588122188063,
      "loss": 0.362,
      "num_input_tokens_seen": 15248416,
      "step": 23370
    },
    {
      "epoch": 12.25104821802935,
      "grad_norm": 0.2232559621334076,
      "learning_rate": 0.000391735540496885,
      "loss": 0.4136,
      "num_input_tokens_seen": 15251008,
      "step": 23375
    },
    {
      "epoch": 12.253668763102725,
      "grad_norm": 0.11572481691837311,
      "learning_rate": 0.00039151229142265337,
      "loss": 0.4758,
      "num_input_tokens_seen": 15256992,
      "step": 23380
    },
    {
      "epoch": 12.2562893081761,
      "grad_norm": 0.17795000970363617,
      "learning_rate": 0.00039128906504281295,
      "loss": 0.3518,
      "num_input_tokens_seen": 15259712,
      "step": 23385
    },
    {
      "epoch": 12.258909853249476,
      "grad_norm": 0.10922178626060486,
      "learning_rate": 0.00039106586140405964,
      "loss": 0.5392,
      "num_input_tokens_seen": 15262880,
      "step": 23390
    },
    {
      "epoch": 12.26153039832285,
      "grad_norm": 0.15186460316181183,
      "learning_rate": 0.00039084268055308537,
      "loss": 0.448,
      "num_input_tokens_seen": 15266048,
      "step": 23395
    },
    {
      "epoch": 12.264150943396226,
      "grad_norm": 0.19451487064361572,
      "learning_rate": 0.00039061952253657703,
      "loss": 0.4583,
      "num_input_tokens_seen": 15268928,
      "step": 23400
    },
    {
      "epoch": 12.266771488469601,
      "grad_norm": 0.11387000232934952,
      "learning_rate": 0.0003903963874012166,
      "loss": 0.5595,
      "num_input_tokens_seen": 15271872,
      "step": 23405
    },
    {
      "epoch": 12.269392033542976,
      "grad_norm": 0.07317297905683517,
      "learning_rate": 0.0003901732751936815,
      "loss": 0.4636,
      "num_input_tokens_seen": 15276704,
      "step": 23410
    },
    {
      "epoch": 12.272012578616351,
      "grad_norm": 0.11551272869110107,
      "learning_rate": 0.00038995018596064447,
      "loss": 0.4146,
      "num_input_tokens_seen": 15279648,
      "step": 23415
    },
    {
      "epoch": 12.274633123689728,
      "grad_norm": 0.22594131529331207,
      "learning_rate": 0.00038972711974877294,
      "loss": 0.4293,
      "num_input_tokens_seen": 15282176,
      "step": 23420
    },
    {
      "epoch": 12.277253668763104,
      "grad_norm": 0.14042891561985016,
      "learning_rate": 0.0003895040766047298,
      "loss": 0.3889,
      "num_input_tokens_seen": 15284640,
      "step": 23425
    },
    {
      "epoch": 12.279874213836479,
      "grad_norm": 0.13254369795322418,
      "learning_rate": 0.0003892810565751733,
      "loss": 0.3998,
      "num_input_tokens_seen": 15287712,
      "step": 23430
    },
    {
      "epoch": 12.282494758909854,
      "grad_norm": 0.10103964060544968,
      "learning_rate": 0.0003890580597067566,
      "loss": 0.3908,
      "num_input_tokens_seen": 15291296,
      "step": 23435
    },
    {
      "epoch": 12.285115303983229,
      "grad_norm": 0.10944667458534241,
      "learning_rate": 0.0003888350860461281,
      "loss": 0.4184,
      "num_input_tokens_seen": 15294464,
      "step": 23440
    },
    {
      "epoch": 12.287735849056604,
      "grad_norm": 0.10439293831586838,
      "learning_rate": 0.0003886121356399315,
      "loss": 0.2636,
      "num_input_tokens_seen": 15299392,
      "step": 23445
    },
    {
      "epoch": 12.29035639412998,
      "grad_norm": 0.203084334731102,
      "learning_rate": 0.0003883892085348052,
      "loss": 0.46,
      "num_input_tokens_seen": 15302496,
      "step": 23450
    },
    {
      "epoch": 12.292976939203355,
      "grad_norm": 0.10346361994743347,
      "learning_rate": 0.0003881663047773832,
      "loss": 0.5132,
      "num_input_tokens_seen": 15306464,
      "step": 23455
    },
    {
      "epoch": 12.29559748427673,
      "grad_norm": 0.16576117277145386,
      "learning_rate": 0.00038794342441429426,
      "loss": 0.394,
      "num_input_tokens_seen": 15309792,
      "step": 23460
    },
    {
      "epoch": 12.298218029350105,
      "grad_norm": 0.12746456265449524,
      "learning_rate": 0.00038772056749216267,
      "loss": 0.4492,
      "num_input_tokens_seen": 15312576,
      "step": 23465
    },
    {
      "epoch": 12.30083857442348,
      "grad_norm": 0.15400704741477966,
      "learning_rate": 0.00038749773405760744,
      "loss": 0.4247,
      "num_input_tokens_seen": 15315296,
      "step": 23470
    },
    {
      "epoch": 12.303459119496855,
      "grad_norm": 0.1138569712638855,
      "learning_rate": 0.00038727492415724265,
      "loss": 0.4625,
      "num_input_tokens_seen": 15318112,
      "step": 23475
    },
    {
      "epoch": 12.30607966457023,
      "grad_norm": 0.10675753653049469,
      "learning_rate": 0.00038705213783767767,
      "loss": 0.4699,
      "num_input_tokens_seen": 15320896,
      "step": 23480
    },
    {
      "epoch": 12.308700209643606,
      "grad_norm": 0.22709068655967712,
      "learning_rate": 0.00038682937514551686,
      "loss": 0.3165,
      "num_input_tokens_seen": 15323904,
      "step": 23485
    },
    {
      "epoch": 12.31132075471698,
      "grad_norm": 0.13853995501995087,
      "learning_rate": 0.0003866066361273596,
      "loss": 0.5175,
      "num_input_tokens_seen": 15326944,
      "step": 23490
    },
    {
      "epoch": 12.313941299790356,
      "grad_norm": 0.11835143715143204,
      "learning_rate": 0.00038638392082980056,
      "loss": 0.4365,
      "num_input_tokens_seen": 15331776,
      "step": 23495
    },
    {
      "epoch": 12.316561844863731,
      "grad_norm": 0.0837748721241951,
      "learning_rate": 0.0003861612292994292,
      "loss": 0.4143,
      "num_input_tokens_seen": 15334624,
      "step": 23500
    },
    {
      "epoch": 12.319182389937106,
      "grad_norm": 0.11403227597475052,
      "learning_rate": 0.0003859385615828297,
      "loss": 0.454,
      "num_input_tokens_seen": 15339072,
      "step": 23505
    },
    {
      "epoch": 12.321802935010481,
      "grad_norm": 0.13751451671123505,
      "learning_rate": 0.00038571591772658186,
      "loss": 0.3993,
      "num_input_tokens_seen": 15342240,
      "step": 23510
    },
    {
      "epoch": 12.324423480083858,
      "grad_norm": 0.11151199042797089,
      "learning_rate": 0.0003854932977772602,
      "loss": 0.4008,
      "num_input_tokens_seen": 15346528,
      "step": 23515
    },
    {
      "epoch": 12.327044025157234,
      "grad_norm": 0.10991551727056503,
      "learning_rate": 0.00038527070178143435,
      "loss": 0.4392,
      "num_input_tokens_seen": 15350144,
      "step": 23520
    },
    {
      "epoch": 12.329664570230609,
      "grad_norm": 0.1269962638616562,
      "learning_rate": 0.00038504812978566873,
      "loss": 0.3596,
      "num_input_tokens_seen": 15352832,
      "step": 23525
    },
    {
      "epoch": 12.332285115303984,
      "grad_norm": 0.14069944620132446,
      "learning_rate": 0.00038482558183652307,
      "loss": 0.4853,
      "num_input_tokens_seen": 15355392,
      "step": 23530
    },
    {
      "epoch": 12.334905660377359,
      "grad_norm": 0.14780083298683167,
      "learning_rate": 0.00038460305798055164,
      "loss": 0.3462,
      "num_input_tokens_seen": 15357504,
      "step": 23535
    },
    {
      "epoch": 12.337526205450734,
      "grad_norm": 0.12017841637134552,
      "learning_rate": 0.00038438055826430385,
      "loss": 0.3345,
      "num_input_tokens_seen": 15360672,
      "step": 23540
    },
    {
      "epoch": 12.34014675052411,
      "grad_norm": 0.23755155503749847,
      "learning_rate": 0.0003841580827343243,
      "loss": 0.628,
      "num_input_tokens_seen": 15364128,
      "step": 23545
    },
    {
      "epoch": 12.342767295597485,
      "grad_norm": 0.1687977910041809,
      "learning_rate": 0.00038393563143715215,
      "loss": 0.4646,
      "num_input_tokens_seen": 15367072,
      "step": 23550
    },
    {
      "epoch": 12.34538784067086,
      "grad_norm": 0.18552012741565704,
      "learning_rate": 0.00038371320441932195,
      "loss": 0.3447,
      "num_input_tokens_seen": 15370848,
      "step": 23555
    },
    {
      "epoch": 12.348008385744235,
      "grad_norm": 0.08054075390100479,
      "learning_rate": 0.00038349080172736267,
      "loss": 0.4041,
      "num_input_tokens_seen": 15375680,
      "step": 23560
    },
    {
      "epoch": 12.35062893081761,
      "grad_norm": 0.10317390412092209,
      "learning_rate": 0.00038326842340779833,
      "loss": 0.5243,
      "num_input_tokens_seen": 15378784,
      "step": 23565
    },
    {
      "epoch": 12.353249475890985,
      "grad_norm": 0.1212448850274086,
      "learning_rate": 0.00038304606950714803,
      "loss": 0.4355,
      "num_input_tokens_seen": 15381856,
      "step": 23570
    },
    {
      "epoch": 12.35587002096436,
      "grad_norm": 0.10031347721815109,
      "learning_rate": 0.0003828237400719259,
      "loss": 0.5033,
      "num_input_tokens_seen": 15385376,
      "step": 23575
    },
    {
      "epoch": 12.358490566037736,
      "grad_norm": 0.11558136343955994,
      "learning_rate": 0.00038260143514864037,
      "loss": 0.3806,
      "num_input_tokens_seen": 15389568,
      "step": 23580
    },
    {
      "epoch": 12.36111111111111,
      "grad_norm": 0.18614163994789124,
      "learning_rate": 0.0003823791547837955,
      "loss": 0.3465,
      "num_input_tokens_seen": 15392160,
      "step": 23585
    },
    {
      "epoch": 12.363731656184486,
      "grad_norm": 0.10922915488481522,
      "learning_rate": 0.0003821568990238894,
      "loss": 0.5233,
      "num_input_tokens_seen": 15395520,
      "step": 23590
    },
    {
      "epoch": 12.366352201257861,
      "grad_norm": 0.11684685200452805,
      "learning_rate": 0.0003819346679154155,
      "loss": 0.5261,
      "num_input_tokens_seen": 15398592,
      "step": 23595
    },
    {
      "epoch": 12.368972746331236,
      "grad_norm": 0.08944852650165558,
      "learning_rate": 0.0003817124615048623,
      "loss": 0.4551,
      "num_input_tokens_seen": 15401824,
      "step": 23600
    },
    {
      "epoch": 12.371593291404611,
      "grad_norm": 0.11568491160869598,
      "learning_rate": 0.00038149027983871243,
      "loss": 0.4618,
      "num_input_tokens_seen": 15405536,
      "step": 23605
    },
    {
      "epoch": 12.374213836477987,
      "grad_norm": 0.1247631087899208,
      "learning_rate": 0.00038126812296344415,
      "loss": 0.482,
      "num_input_tokens_seen": 15408800,
      "step": 23610
    },
    {
      "epoch": 12.376834381551364,
      "grad_norm": 0.12505994737148285,
      "learning_rate": 0.0003810459909255301,
      "loss": 0.5398,
      "num_input_tokens_seen": 15411392,
      "step": 23615
    },
    {
      "epoch": 12.379454926624739,
      "grad_norm": 0.2021995186805725,
      "learning_rate": 0.0003808238837714374,
      "loss": 0.5452,
      "num_input_tokens_seen": 15414144,
      "step": 23620
    },
    {
      "epoch": 12.382075471698114,
      "grad_norm": 0.12655924260616302,
      "learning_rate": 0.0003806018015476287,
      "loss": 0.525,
      "num_input_tokens_seen": 15417984,
      "step": 23625
    },
    {
      "epoch": 12.384696016771489,
      "grad_norm": 0.1150687113404274,
      "learning_rate": 0.0003803797443005609,
      "loss": 0.3711,
      "num_input_tokens_seen": 15421280,
      "step": 23630
    },
    {
      "epoch": 12.387316561844864,
      "grad_norm": 0.20292890071868896,
      "learning_rate": 0.0003801577120766859,
      "loss": 0.5364,
      "num_input_tokens_seen": 15423872,
      "step": 23635
    },
    {
      "epoch": 12.38993710691824,
      "grad_norm": 0.11118457466363907,
      "learning_rate": 0.0003799357049224505,
      "loss": 0.4102,
      "num_input_tokens_seen": 15426976,
      "step": 23640
    },
    {
      "epoch": 12.392557651991615,
      "grad_norm": 0.18553526699543,
      "learning_rate": 0.0003797137228842956,
      "loss": 0.4098,
      "num_input_tokens_seen": 15430240,
      "step": 23645
    },
    {
      "epoch": 12.39517819706499,
      "grad_norm": 0.10237590968608856,
      "learning_rate": 0.0003794917660086576,
      "loss": 0.3328,
      "num_input_tokens_seen": 15433664,
      "step": 23650
    },
    {
      "epoch": 12.397798742138365,
      "grad_norm": 0.14367644488811493,
      "learning_rate": 0.0003792698343419674,
      "loss": 0.4222,
      "num_input_tokens_seen": 15437088,
      "step": 23655
    },
    {
      "epoch": 12.40041928721174,
      "grad_norm": 0.16124066710472107,
      "learning_rate": 0.0003790479279306505,
      "loss": 0.4694,
      "num_input_tokens_seen": 15440096,
      "step": 23660
    },
    {
      "epoch": 12.403039832285115,
      "grad_norm": 0.11710210889577866,
      "learning_rate": 0.0003788260468211271,
      "loss": 0.4321,
      "num_input_tokens_seen": 15444352,
      "step": 23665
    },
    {
      "epoch": 12.40566037735849,
      "grad_norm": 0.24173645675182343,
      "learning_rate": 0.0003786041910598125,
      "loss": 0.4359,
      "num_input_tokens_seen": 15448736,
      "step": 23670
    },
    {
      "epoch": 12.408280922431866,
      "grad_norm": 0.08913233876228333,
      "learning_rate": 0.0003783823606931159,
      "loss": 0.3888,
      "num_input_tokens_seen": 15451712,
      "step": 23675
    },
    {
      "epoch": 12.41090146750524,
      "grad_norm": 0.13191260397434235,
      "learning_rate": 0.0003781605557674421,
      "loss": 0.3723,
      "num_input_tokens_seen": 15454272,
      "step": 23680
    },
    {
      "epoch": 12.413522012578616,
      "grad_norm": 0.08579491823911667,
      "learning_rate": 0.0003779387763291899,
      "loss": 0.4202,
      "num_input_tokens_seen": 15457856,
      "step": 23685
    },
    {
      "epoch": 12.416142557651991,
      "grad_norm": 0.1390186995267868,
      "learning_rate": 0.0003777170224247533,
      "loss": 0.3629,
      "num_input_tokens_seen": 15461408,
      "step": 23690
    },
    {
      "epoch": 12.418763102725366,
      "grad_norm": 0.07743368297815323,
      "learning_rate": 0.0003774952941005204,
      "loss": 0.4792,
      "num_input_tokens_seen": 15464992,
      "step": 23695
    },
    {
      "epoch": 12.421383647798741,
      "grad_norm": 0.106129489839077,
      "learning_rate": 0.00037727359140287455,
      "loss": 0.4734,
      "num_input_tokens_seen": 15468096,
      "step": 23700
    },
    {
      "epoch": 12.424004192872117,
      "grad_norm": 0.1487663984298706,
      "learning_rate": 0.00037705191437819316,
      "loss": 0.3784,
      "num_input_tokens_seen": 15470752,
      "step": 23705
    },
    {
      "epoch": 12.426624737945493,
      "grad_norm": 0.1176622286438942,
      "learning_rate": 0.00037683026307284853,
      "loss": 0.4823,
      "num_input_tokens_seen": 15473952,
      "step": 23710
    },
    {
      "epoch": 12.429245283018869,
      "grad_norm": 0.08219550549983978,
      "learning_rate": 0.0003766086375332077,
      "loss": 0.3424,
      "num_input_tokens_seen": 15477760,
      "step": 23715
    },
    {
      "epoch": 12.431865828092244,
      "grad_norm": 0.11519763618707657,
      "learning_rate": 0.0003763870378056321,
      "loss": 0.3968,
      "num_input_tokens_seen": 15480384,
      "step": 23720
    },
    {
      "epoch": 12.434486373165619,
      "grad_norm": 0.1628698855638504,
      "learning_rate": 0.00037616546393647824,
      "loss": 0.4313,
      "num_input_tokens_seen": 15482976,
      "step": 23725
    },
    {
      "epoch": 12.437106918238994,
      "grad_norm": 0.10786505788564682,
      "learning_rate": 0.0003759439159720962,
      "loss": 0.337,
      "num_input_tokens_seen": 15486016,
      "step": 23730
    },
    {
      "epoch": 12.43972746331237,
      "grad_norm": 0.12062808126211166,
      "learning_rate": 0.0003757223939588318,
      "loss": 0.3647,
      "num_input_tokens_seen": 15489472,
      "step": 23735
    },
    {
      "epoch": 12.442348008385745,
      "grad_norm": 0.20357084274291992,
      "learning_rate": 0.00037550089794302464,
      "loss": 0.5173,
      "num_input_tokens_seen": 15493120,
      "step": 23740
    },
    {
      "epoch": 12.44496855345912,
      "grad_norm": 0.09416180104017258,
      "learning_rate": 0.0003752794279710094,
      "loss": 0.4583,
      "num_input_tokens_seen": 15495968,
      "step": 23745
    },
    {
      "epoch": 12.447589098532495,
      "grad_norm": 0.16084907948970795,
      "learning_rate": 0.0003750579840891148,
      "loss": 0.477,
      "num_input_tokens_seen": 15498784,
      "step": 23750
    },
    {
      "epoch": 12.45020964360587,
      "grad_norm": 0.09343487024307251,
      "learning_rate": 0.0003748365663436647,
      "loss": 0.5424,
      "num_input_tokens_seen": 15502784,
      "step": 23755
    },
    {
      "epoch": 12.452830188679245,
      "grad_norm": 0.12918218970298767,
      "learning_rate": 0.0003746151747809769,
      "loss": 0.4203,
      "num_input_tokens_seen": 15505952,
      "step": 23760
    },
    {
      "epoch": 12.45545073375262,
      "grad_norm": 0.22188787162303925,
      "learning_rate": 0.000374393809447364,
      "loss": 0.415,
      "num_input_tokens_seen": 15509344,
      "step": 23765
    },
    {
      "epoch": 12.458071278825996,
      "grad_norm": 0.11267121881246567,
      "learning_rate": 0.0003741724703891333,
      "loss": 0.5151,
      "num_input_tokens_seen": 15512544,
      "step": 23770
    },
    {
      "epoch": 12.46069182389937,
      "grad_norm": 0.12022169679403305,
      "learning_rate": 0.00037395115765258616,
      "loss": 0.5129,
      "num_input_tokens_seen": 15519104,
      "step": 23775
    },
    {
      "epoch": 12.463312368972746,
      "grad_norm": 0.12031669169664383,
      "learning_rate": 0.0003737298712840188,
      "loss": 0.4578,
      "num_input_tokens_seen": 15522528,
      "step": 23780
    },
    {
      "epoch": 12.465932914046121,
      "grad_norm": 0.1309848576784134,
      "learning_rate": 0.000373508611329722,
      "loss": 0.4807,
      "num_input_tokens_seen": 15525440,
      "step": 23785
    },
    {
      "epoch": 12.468553459119496,
      "grad_norm": 0.0940946713089943,
      "learning_rate": 0.00037328737783598036,
      "loss": 0.4359,
      "num_input_tokens_seen": 15528448,
      "step": 23790
    },
    {
      "epoch": 12.471174004192871,
      "grad_norm": 0.10921554267406464,
      "learning_rate": 0.0003730661708490738,
      "loss": 0.3802,
      "num_input_tokens_seen": 15531392,
      "step": 23795
    },
    {
      "epoch": 12.473794549266247,
      "grad_norm": 0.11766868084669113,
      "learning_rate": 0.0003728449904152761,
      "loss": 0.295,
      "num_input_tokens_seen": 15534784,
      "step": 23800
    },
    {
      "epoch": 12.476415094339623,
      "grad_norm": 0.08898816257715225,
      "learning_rate": 0.00037262383658085563,
      "loss": 0.395,
      "num_input_tokens_seen": 15538144,
      "step": 23805
    },
    {
      "epoch": 12.479035639412999,
      "grad_norm": 0.06508205085992813,
      "learning_rate": 0.00037240270939207555,
      "loss": 0.3625,
      "num_input_tokens_seen": 15541600,
      "step": 23810
    },
    {
      "epoch": 12.481656184486374,
      "grad_norm": 0.12911316752433777,
      "learning_rate": 0.000372181608895193,
      "loss": 0.5324,
      "num_input_tokens_seen": 15545728,
      "step": 23815
    },
    {
      "epoch": 12.484276729559749,
      "grad_norm": 0.09668175131082535,
      "learning_rate": 0.00037196053513645957,
      "loss": 0.3684,
      "num_input_tokens_seen": 15548000,
      "step": 23820
    },
    {
      "epoch": 12.486897274633124,
      "grad_norm": 0.10635020583868027,
      "learning_rate": 0.00037173948816212146,
      "loss": 0.369,
      "num_input_tokens_seen": 15550496,
      "step": 23825
    },
    {
      "epoch": 12.4895178197065,
      "grad_norm": 0.09002199023962021,
      "learning_rate": 0.00037151846801841904,
      "loss": 0.4411,
      "num_input_tokens_seen": 15554304,
      "step": 23830
    },
    {
      "epoch": 12.492138364779874,
      "grad_norm": 0.0829613134264946,
      "learning_rate": 0.00037129747475158736,
      "loss": 0.432,
      "num_input_tokens_seen": 15557824,
      "step": 23835
    },
    {
      "epoch": 12.49475890985325,
      "grad_norm": 0.10295283794403076,
      "learning_rate": 0.0003710765084078558,
      "loss": 0.4265,
      "num_input_tokens_seen": 15560704,
      "step": 23840
    },
    {
      "epoch": 12.497379454926625,
      "grad_norm": 0.07051512598991394,
      "learning_rate": 0.0003708555690334477,
      "loss": 0.4471,
      "num_input_tokens_seen": 15564832,
      "step": 23845
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.09781340509653091,
      "learning_rate": 0.00037063465667458125,
      "loss": 0.3485,
      "num_input_tokens_seen": 15567904,
      "step": 23850
    },
    {
      "epoch": 12.502620545073375,
      "grad_norm": 0.1523115336894989,
      "learning_rate": 0.0003704137713774686,
      "loss": 0.5175,
      "num_input_tokens_seen": 15571168,
      "step": 23855
    },
    {
      "epoch": 12.50524109014675,
      "grad_norm": 0.1514555960893631,
      "learning_rate": 0.0003701929131883167,
      "loss": 0.3666,
      "num_input_tokens_seen": 15574144,
      "step": 23860
    },
    {
      "epoch": 12.507861635220126,
      "grad_norm": 0.14565008878707886,
      "learning_rate": 0.0003699720821533264,
      "loss": 0.4889,
      "num_input_tokens_seen": 15576160,
      "step": 23865
    },
    {
      "epoch": 12.5104821802935,
      "grad_norm": 0.17917101085186005,
      "learning_rate": 0.00036975127831869326,
      "loss": 0.4143,
      "num_input_tokens_seen": 15579424,
      "step": 23870
    },
    {
      "epoch": 12.513102725366876,
      "grad_norm": 0.0892876535654068,
      "learning_rate": 0.0003695305017306066,
      "loss": 0.2656,
      "num_input_tokens_seen": 15584032,
      "step": 23875
    },
    {
      "epoch": 12.515723270440251,
      "grad_norm": 0.2827084958553314,
      "learning_rate": 0.00036930975243525046,
      "loss": 0.498,
      "num_input_tokens_seen": 15588544,
      "step": 23880
    },
    {
      "epoch": 12.518343815513626,
      "grad_norm": 0.11093050986528397,
      "learning_rate": 0.00036908903047880304,
      "loss": 0.3786,
      "num_input_tokens_seen": 15594336,
      "step": 23885
    },
    {
      "epoch": 12.520964360587001,
      "grad_norm": 0.1877484917640686,
      "learning_rate": 0.00036886833590743707,
      "loss": 0.4795,
      "num_input_tokens_seen": 15597760,
      "step": 23890
    },
    {
      "epoch": 12.523584905660378,
      "grad_norm": 0.17628711462020874,
      "learning_rate": 0.00036864766876731913,
      "loss": 0.3953,
      "num_input_tokens_seen": 15601824,
      "step": 23895
    },
    {
      "epoch": 12.526205450733752,
      "grad_norm": 0.10677797347307205,
      "learning_rate": 0.00036842702910461054,
      "loss": 0.5386,
      "num_input_tokens_seen": 15605824,
      "step": 23900
    },
    {
      "epoch": 12.528825995807129,
      "grad_norm": 0.11840390413999557,
      "learning_rate": 0.0003682064169654663,
      "loss": 0.3989,
      "num_input_tokens_seen": 15609312,
      "step": 23905
    },
    {
      "epoch": 12.531446540880504,
      "grad_norm": 0.11753259599208832,
      "learning_rate": 0.00036798583239603587,
      "loss": 0.5887,
      "num_input_tokens_seen": 15613216,
      "step": 23910
    },
    {
      "epoch": 12.534067085953879,
      "grad_norm": 0.14928074181079865,
      "learning_rate": 0.0003677652754424634,
      "loss": 0.3321,
      "num_input_tokens_seen": 15616352,
      "step": 23915
    },
    {
      "epoch": 12.536687631027254,
      "grad_norm": 0.10469738394021988,
      "learning_rate": 0.0003675447461508865,
      "loss": 0.4605,
      "num_input_tokens_seen": 15619072,
      "step": 23920
    },
    {
      "epoch": 12.53930817610063,
      "grad_norm": 0.13938497006893158,
      "learning_rate": 0.00036732424456743784,
      "loss": 0.425,
      "num_input_tokens_seen": 15622144,
      "step": 23925
    },
    {
      "epoch": 12.541928721174004,
      "grad_norm": 0.1215309351682663,
      "learning_rate": 0.0003671037707382435,
      "loss": 0.427,
      "num_input_tokens_seen": 15625152,
      "step": 23930
    },
    {
      "epoch": 12.54454926624738,
      "grad_norm": 0.181486576795578,
      "learning_rate": 0.000366883324709424,
      "loss": 0.4709,
      "num_input_tokens_seen": 15628128,
      "step": 23935
    },
    {
      "epoch": 12.547169811320755,
      "grad_norm": 0.15164795517921448,
      "learning_rate": 0.00036666290652709446,
      "loss": 0.5115,
      "num_input_tokens_seen": 15631328,
      "step": 23940
    },
    {
      "epoch": 12.54979035639413,
      "grad_norm": 0.1473671942949295,
      "learning_rate": 0.0003664425162373635,
      "loss": 0.5097,
      "num_input_tokens_seen": 15634304,
      "step": 23945
    },
    {
      "epoch": 12.552410901467505,
      "grad_norm": 0.09018412977457047,
      "learning_rate": 0.0003662221538863346,
      "loss": 0.3046,
      "num_input_tokens_seen": 15637984,
      "step": 23950
    },
    {
      "epoch": 12.55503144654088,
      "grad_norm": 0.14803999662399292,
      "learning_rate": 0.0003660018195201049,
      "loss": 0.3661,
      "num_input_tokens_seen": 15640672,
      "step": 23955
    },
    {
      "epoch": 12.557651991614255,
      "grad_norm": 0.1722177416086197,
      "learning_rate": 0.0003657815131847657,
      "loss": 0.4151,
      "num_input_tokens_seen": 15644096,
      "step": 23960
    },
    {
      "epoch": 12.56027253668763,
      "grad_norm": 0.13780473172664642,
      "learning_rate": 0.0003655612349264027,
      "loss": 0.4729,
      "num_input_tokens_seen": 15647040,
      "step": 23965
    },
    {
      "epoch": 12.562893081761006,
      "grad_norm": 0.14369840919971466,
      "learning_rate": 0.0003653409847910957,
      "loss": 0.4148,
      "num_input_tokens_seen": 15649824,
      "step": 23970
    },
    {
      "epoch": 12.565513626834381,
      "grad_norm": 0.09292837977409363,
      "learning_rate": 0.0003651207628249182,
      "loss": 0.3434,
      "num_input_tokens_seen": 15652992,
      "step": 23975
    },
    {
      "epoch": 12.568134171907756,
      "grad_norm": 0.19818082451820374,
      "learning_rate": 0.0003649005690739386,
      "loss": 0.4445,
      "num_input_tokens_seen": 15655680,
      "step": 23980
    },
    {
      "epoch": 12.570754716981131,
      "grad_norm": 0.1339961439371109,
      "learning_rate": 0.0003646804035842187,
      "loss": 0.3347,
      "num_input_tokens_seen": 15658496,
      "step": 23985
    },
    {
      "epoch": 12.573375262054507,
      "grad_norm": 0.1258624941110611,
      "learning_rate": 0.0003644602664018143,
      "loss": 0.5978,
      "num_input_tokens_seen": 15661248,
      "step": 23990
    },
    {
      "epoch": 12.575995807127882,
      "grad_norm": 0.1122293621301651,
      "learning_rate": 0.000364240157572776,
      "loss": 0.4552,
      "num_input_tokens_seen": 15663840,
      "step": 23995
    },
    {
      "epoch": 12.578616352201259,
      "grad_norm": 0.119312584400177,
      "learning_rate": 0.0003640200771431478,
      "loss": 0.5045,
      "num_input_tokens_seen": 15667744,
      "step": 24000
    },
    {
      "epoch": 12.581236897274634,
      "grad_norm": 0.1589936763048172,
      "learning_rate": 0.0003638000251589683,
      "loss": 0.4629,
      "num_input_tokens_seen": 15670592,
      "step": 24005
    },
    {
      "epoch": 12.583857442348009,
      "grad_norm": 0.10755417495965958,
      "learning_rate": 0.0003635800016662696,
      "loss": 0.4574,
      "num_input_tokens_seen": 15674688,
      "step": 24010
    },
    {
      "epoch": 12.586477987421384,
      "grad_norm": 0.19194918870925903,
      "learning_rate": 0.00036336000671107816,
      "loss": 0.4387,
      "num_input_tokens_seen": 15677440,
      "step": 24015
    },
    {
      "epoch": 12.58909853249476,
      "grad_norm": 0.13759873807430267,
      "learning_rate": 0.00036314004033941445,
      "loss": 0.4743,
      "num_input_tokens_seen": 15680416,
      "step": 24020
    },
    {
      "epoch": 12.591719077568134,
      "grad_norm": 0.10809039324522018,
      "learning_rate": 0.00036292010259729283,
      "loss": 0.4472,
      "num_input_tokens_seen": 15684800,
      "step": 24025
    },
    {
      "epoch": 12.59433962264151,
      "grad_norm": 0.12461655586957932,
      "learning_rate": 0.000362700193530722,
      "loss": 0.4839,
      "num_input_tokens_seen": 15687936,
      "step": 24030
    },
    {
      "epoch": 12.596960167714885,
      "grad_norm": 0.1417551189661026,
      "learning_rate": 0.0003624803131857042,
      "loss": 0.3217,
      "num_input_tokens_seen": 15691104,
      "step": 24035
    },
    {
      "epoch": 12.59958071278826,
      "grad_norm": 0.1487068384885788,
      "learning_rate": 0.0003622604616082361,
      "loss": 0.4892,
      "num_input_tokens_seen": 15694432,
      "step": 24040
    },
    {
      "epoch": 12.602201257861635,
      "grad_norm": 0.10694319754838943,
      "learning_rate": 0.0003620406388443078,
      "loss": 0.4765,
      "num_input_tokens_seen": 15697504,
      "step": 24045
    },
    {
      "epoch": 12.60482180293501,
      "grad_norm": 0.21388062834739685,
      "learning_rate": 0.00036182084493990407,
      "loss": 0.4967,
      "num_input_tokens_seen": 15701824,
      "step": 24050
    },
    {
      "epoch": 12.607442348008385,
      "grad_norm": 0.10223579406738281,
      "learning_rate": 0.0003616010799410031,
      "loss": 0.4123,
      "num_input_tokens_seen": 15704736,
      "step": 24055
    },
    {
      "epoch": 12.61006289308176,
      "grad_norm": 0.12093371152877808,
      "learning_rate": 0.0003613813438935773,
      "loss": 0.3552,
      "num_input_tokens_seen": 15707744,
      "step": 24060
    },
    {
      "epoch": 12.612683438155136,
      "grad_norm": 0.12034275382757187,
      "learning_rate": 0.0003611616368435928,
      "loss": 0.5075,
      "num_input_tokens_seen": 15710208,
      "step": 24065
    },
    {
      "epoch": 12.615303983228511,
      "grad_norm": 0.0754864290356636,
      "learning_rate": 0.0003609419588370102,
      "loss": 0.4548,
      "num_input_tokens_seen": 15713792,
      "step": 24070
    },
    {
      "epoch": 12.617924528301886,
      "grad_norm": 0.21570716798305511,
      "learning_rate": 0.00036072230991978326,
      "loss": 0.4447,
      "num_input_tokens_seen": 15717792,
      "step": 24075
    },
    {
      "epoch": 12.620545073375261,
      "grad_norm": 0.11722779273986816,
      "learning_rate": 0.00036050269013785996,
      "loss": 0.4356,
      "num_input_tokens_seen": 15720896,
      "step": 24080
    },
    {
      "epoch": 12.623165618448636,
      "grad_norm": 0.143508642911911,
      "learning_rate": 0.0003602830995371825,
      "loss": 0.3074,
      "num_input_tokens_seen": 15723904,
      "step": 24085
    },
    {
      "epoch": 12.625786163522012,
      "grad_norm": 0.12856191396713257,
      "learning_rate": 0.0003600635381636866,
      "loss": 0.4754,
      "num_input_tokens_seen": 15727168,
      "step": 24090
    },
    {
      "epoch": 12.628406708595389,
      "grad_norm": 0.12840265035629272,
      "learning_rate": 0.0003598440060633022,
      "loss": 0.616,
      "num_input_tokens_seen": 15730688,
      "step": 24095
    },
    {
      "epoch": 12.631027253668764,
      "grad_norm": 0.10151837766170502,
      "learning_rate": 0.00035962450328195264,
      "loss": 0.5177,
      "num_input_tokens_seen": 15733664,
      "step": 24100
    },
    {
      "epoch": 12.633647798742139,
      "grad_norm": 0.11114746332168579,
      "learning_rate": 0.00035940502986555543,
      "loss": 0.4931,
      "num_input_tokens_seen": 15736352,
      "step": 24105
    },
    {
      "epoch": 12.636268343815514,
      "grad_norm": 0.14666897058486938,
      "learning_rate": 0.00035918558586002205,
      "loss": 0.3494,
      "num_input_tokens_seen": 15739648,
      "step": 24110
    },
    {
      "epoch": 12.63888888888889,
      "grad_norm": 0.14354175329208374,
      "learning_rate": 0.0003589661713112575,
      "loss": 0.4538,
      "num_input_tokens_seen": 15742464,
      "step": 24115
    },
    {
      "epoch": 12.641509433962264,
      "grad_norm": 0.10112129151821136,
      "learning_rate": 0.000358746786265161,
      "loss": 0.4652,
      "num_input_tokens_seen": 15745984,
      "step": 24120
    },
    {
      "epoch": 12.64412997903564,
      "grad_norm": 0.17515817284584045,
      "learning_rate": 0.0003585274307676254,
      "loss": 0.3703,
      "num_input_tokens_seen": 15748608,
      "step": 24125
    },
    {
      "epoch": 12.646750524109015,
      "grad_norm": 0.15566664934158325,
      "learning_rate": 0.00035830810486453725,
      "loss": 0.4569,
      "num_input_tokens_seen": 15751520,
      "step": 24130
    },
    {
      "epoch": 12.64937106918239,
      "grad_norm": 0.08398068696260452,
      "learning_rate": 0.00035808880860177696,
      "loss": 0.4089,
      "num_input_tokens_seen": 15754912,
      "step": 24135
    },
    {
      "epoch": 12.651991614255765,
      "grad_norm": 0.082011379301548,
      "learning_rate": 0.00035786954202521895,
      "loss": 0.6009,
      "num_input_tokens_seen": 15758688,
      "step": 24140
    },
    {
      "epoch": 12.65461215932914,
      "grad_norm": 0.1446646898984909,
      "learning_rate": 0.00035765030518073116,
      "loss": 0.3364,
      "num_input_tokens_seen": 15761248,
      "step": 24145
    },
    {
      "epoch": 12.657232704402515,
      "grad_norm": 0.12859933078289032,
      "learning_rate": 0.0003574310981141756,
      "loss": 0.5584,
      "num_input_tokens_seen": 15765088,
      "step": 24150
    },
    {
      "epoch": 12.65985324947589,
      "grad_norm": 0.14737479388713837,
      "learning_rate": 0.00035721192087140796,
      "loss": 0.4695,
      "num_input_tokens_seen": 15768896,
      "step": 24155
    },
    {
      "epoch": 12.662473794549266,
      "grad_norm": 0.12565986812114716,
      "learning_rate": 0.00035699277349827724,
      "loss": 0.3737,
      "num_input_tokens_seen": 15771584,
      "step": 24160
    },
    {
      "epoch": 12.665094339622641,
      "grad_norm": 0.15047740936279297,
      "learning_rate": 0.0003567736560406269,
      "loss": 0.5475,
      "num_input_tokens_seen": 15774080,
      "step": 24165
    },
    {
      "epoch": 12.667714884696016,
      "grad_norm": 0.12988917529582977,
      "learning_rate": 0.0003565545685442936,
      "loss": 0.5807,
      "num_input_tokens_seen": 15777280,
      "step": 24170
    },
    {
      "epoch": 12.670335429769391,
      "grad_norm": 0.14619480073451996,
      "learning_rate": 0.00035633551105510806,
      "loss": 0.5781,
      "num_input_tokens_seen": 15780544,
      "step": 24175
    },
    {
      "epoch": 12.672955974842766,
      "grad_norm": 0.1933140903711319,
      "learning_rate": 0.00035611648361889457,
      "loss": 0.5136,
      "num_input_tokens_seen": 15783680,
      "step": 24180
    },
    {
      "epoch": 12.675576519916142,
      "grad_norm": 0.17884144186973572,
      "learning_rate": 0.00035589748628147145,
      "loss": 0.3965,
      "num_input_tokens_seen": 15786752,
      "step": 24185
    },
    {
      "epoch": 12.678197064989519,
      "grad_norm": 0.1351318210363388,
      "learning_rate": 0.00035567851908865,
      "loss": 0.3907,
      "num_input_tokens_seen": 15790368,
      "step": 24190
    },
    {
      "epoch": 12.680817610062894,
      "grad_norm": 0.11174959689378738,
      "learning_rate": 0.0003554595820862358,
      "loss": 0.4791,
      "num_input_tokens_seen": 15793120,
      "step": 24195
    },
    {
      "epoch": 12.683438155136269,
      "grad_norm": 0.15858733654022217,
      "learning_rate": 0.00035524067532002803,
      "loss": 0.3627,
      "num_input_tokens_seen": 15797600,
      "step": 24200
    },
    {
      "epoch": 12.686058700209644,
      "grad_norm": 0.08821095526218414,
      "learning_rate": 0.00035502179883581956,
      "loss": 0.4338,
      "num_input_tokens_seen": 15802304,
      "step": 24205
    },
    {
      "epoch": 12.68867924528302,
      "grad_norm": 0.10500184446573257,
      "learning_rate": 0.0003548029526793969,
      "loss": 0.419,
      "num_input_tokens_seen": 15804672,
      "step": 24210
    },
    {
      "epoch": 12.691299790356394,
      "grad_norm": 0.14522193372249603,
      "learning_rate": 0.0003545841368965398,
      "loss": 0.4119,
      "num_input_tokens_seen": 15807168,
      "step": 24215
    },
    {
      "epoch": 12.69392033542977,
      "grad_norm": 0.19691279530525208,
      "learning_rate": 0.00035436535153302235,
      "loss": 0.3235,
      "num_input_tokens_seen": 15809824,
      "step": 24220
    },
    {
      "epoch": 12.696540880503145,
      "grad_norm": 0.13773873448371887,
      "learning_rate": 0.0003541465966346118,
      "loss": 0.5405,
      "num_input_tokens_seen": 15813408,
      "step": 24225
    },
    {
      "epoch": 12.69916142557652,
      "grad_norm": 0.1597684919834137,
      "learning_rate": 0.00035392787224706936,
      "loss": 0.4683,
      "num_input_tokens_seen": 15816096,
      "step": 24230
    },
    {
      "epoch": 12.701781970649895,
      "grad_norm": 0.0865032821893692,
      "learning_rate": 0.0003537091784161495,
      "loss": 0.319,
      "num_input_tokens_seen": 15819328,
      "step": 24235
    },
    {
      "epoch": 12.70440251572327,
      "grad_norm": 0.34106943011283875,
      "learning_rate": 0.0003534905151876007,
      "loss": 0.4072,
      "num_input_tokens_seen": 15822176,
      "step": 24240
    },
    {
      "epoch": 12.707023060796645,
      "grad_norm": 0.16849547624588013,
      "learning_rate": 0.0003532718826071646,
      "loss": 0.639,
      "num_input_tokens_seen": 15824768,
      "step": 24245
    },
    {
      "epoch": 12.70964360587002,
      "grad_norm": 0.0867711678147316,
      "learning_rate": 0.00035305328072057654,
      "loss": 0.5266,
      "num_input_tokens_seen": 15828608,
      "step": 24250
    },
    {
      "epoch": 12.712264150943396,
      "grad_norm": 0.17517709732055664,
      "learning_rate": 0.00035283470957356576,
      "loss": 0.4551,
      "num_input_tokens_seen": 15830880,
      "step": 24255
    },
    {
      "epoch": 12.714884696016771,
      "grad_norm": 0.11266857385635376,
      "learning_rate": 0.0003526161692118548,
      "loss": 0.4776,
      "num_input_tokens_seen": 15834112,
      "step": 24260
    },
    {
      "epoch": 12.717505241090146,
      "grad_norm": 0.07748597115278244,
      "learning_rate": 0.0003523976596811597,
      "loss": 0.3232,
      "num_input_tokens_seen": 15836832,
      "step": 24265
    },
    {
      "epoch": 12.720125786163521,
      "grad_norm": 0.18528464436531067,
      "learning_rate": 0.0003521791810271906,
      "loss": 0.3771,
      "num_input_tokens_seen": 15839840,
      "step": 24270
    },
    {
      "epoch": 12.722746331236896,
      "grad_norm": 0.12040577083826065,
      "learning_rate": 0.0003519607332956501,
      "loss": 0.4327,
      "num_input_tokens_seen": 15842848,
      "step": 24275
    },
    {
      "epoch": 12.725366876310272,
      "grad_norm": 0.11579152196645737,
      "learning_rate": 0.0003517423165322353,
      "loss": 0.4981,
      "num_input_tokens_seen": 15846496,
      "step": 24280
    },
    {
      "epoch": 12.727987421383649,
      "grad_norm": 0.24223370850086212,
      "learning_rate": 0.0003515239307826366,
      "loss": 0.4769,
      "num_input_tokens_seen": 15852832,
      "step": 24285
    },
    {
      "epoch": 12.730607966457024,
      "grad_norm": 0.0999043807387352,
      "learning_rate": 0.00035130557609253744,
      "loss": 0.4156,
      "num_input_tokens_seen": 15856416,
      "step": 24290
    },
    {
      "epoch": 12.733228511530399,
      "grad_norm": 0.13474655151367188,
      "learning_rate": 0.00035108725250761566,
      "loss": 0.4808,
      "num_input_tokens_seen": 15859104,
      "step": 24295
    },
    {
      "epoch": 12.735849056603774,
      "grad_norm": 0.10452654957771301,
      "learning_rate": 0.0003508689600735416,
      "loss": 0.3467,
      "num_input_tokens_seen": 15861792,
      "step": 24300
    },
    {
      "epoch": 12.73846960167715,
      "grad_norm": 0.11069319397211075,
      "learning_rate": 0.0003506506988359797,
      "loss": 0.5027,
      "num_input_tokens_seen": 15864320,
      "step": 24305
    },
    {
      "epoch": 12.741090146750524,
      "grad_norm": 0.15865899622440338,
      "learning_rate": 0.0003504324688405878,
      "loss": 0.2949,
      "num_input_tokens_seen": 15867168,
      "step": 24310
    },
    {
      "epoch": 12.7437106918239,
      "grad_norm": 0.16072578728199005,
      "learning_rate": 0.000350214270133017,
      "loss": 0.4304,
      "num_input_tokens_seen": 15870048,
      "step": 24315
    },
    {
      "epoch": 12.746331236897275,
      "grad_norm": 0.14677372574806213,
      "learning_rate": 0.00034999610275891204,
      "loss": 0.4234,
      "num_input_tokens_seen": 15873024,
      "step": 24320
    },
    {
      "epoch": 12.74895178197065,
      "grad_norm": 0.10764077305793762,
      "learning_rate": 0.0003497779667639113,
      "loss": 0.4444,
      "num_input_tokens_seen": 15875648,
      "step": 24325
    },
    {
      "epoch": 12.751572327044025,
      "grad_norm": 0.10367525368928909,
      "learning_rate": 0.0003495598621936458,
      "loss": 0.3312,
      "num_input_tokens_seen": 15878240,
      "step": 24330
    },
    {
      "epoch": 12.7541928721174,
      "grad_norm": 0.13486067950725555,
      "learning_rate": 0.00034934178909374084,
      "loss": 0.4918,
      "num_input_tokens_seen": 15881152,
      "step": 24335
    },
    {
      "epoch": 12.756813417190775,
      "grad_norm": 0.14502456784248352,
      "learning_rate": 0.0003491237475098148,
      "loss": 0.4564,
      "num_input_tokens_seen": 15884384,
      "step": 24340
    },
    {
      "epoch": 12.75943396226415,
      "grad_norm": 0.17470815777778625,
      "learning_rate": 0.00034890573748747945,
      "loss": 0.5981,
      "num_input_tokens_seen": 15887232,
      "step": 24345
    },
    {
      "epoch": 12.762054507337526,
      "grad_norm": 0.13295559585094452,
      "learning_rate": 0.00034868775907234017,
      "loss": 0.4067,
      "num_input_tokens_seen": 15889344,
      "step": 24350
    },
    {
      "epoch": 12.764675052410901,
      "grad_norm": 0.17049868404865265,
      "learning_rate": 0.0003484698123099956,
      "loss": 0.6466,
      "num_input_tokens_seen": 15892704,
      "step": 24355
    },
    {
      "epoch": 12.767295597484276,
      "grad_norm": 0.11315904557704926,
      "learning_rate": 0.00034825189724603723,
      "loss": 0.6598,
      "num_input_tokens_seen": 15895840,
      "step": 24360
    },
    {
      "epoch": 12.769916142557651,
      "grad_norm": 0.14910051226615906,
      "learning_rate": 0.0003480340139260509,
      "loss": 0.3946,
      "num_input_tokens_seen": 15898720,
      "step": 24365
    },
    {
      "epoch": 12.772536687631026,
      "grad_norm": 0.07933812588453293,
      "learning_rate": 0.000347816162395615,
      "loss": 0.3147,
      "num_input_tokens_seen": 15901408,
      "step": 24370
    },
    {
      "epoch": 12.775157232704402,
      "grad_norm": 0.27671366930007935,
      "learning_rate": 0.00034759834270030185,
      "loss": 0.4335,
      "num_input_tokens_seen": 15904160,
      "step": 24375
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 0.10242908447980881,
      "learning_rate": 0.0003473805548856768,
      "loss": 0.4615,
      "num_input_tokens_seen": 15907488,
      "step": 24380
    },
    {
      "epoch": 12.780398322851154,
      "grad_norm": 0.0868402048945427,
      "learning_rate": 0.00034716279899729826,
      "loss": 0.3963,
      "num_input_tokens_seen": 15910784,
      "step": 24385
    },
    {
      "epoch": 12.783018867924529,
      "grad_norm": 0.10893020778894424,
      "learning_rate": 0.00034694507508071864,
      "loss": 0.4673,
      "num_input_tokens_seen": 15914016,
      "step": 24390
    },
    {
      "epoch": 12.785639412997904,
      "grad_norm": 0.10547343641519547,
      "learning_rate": 0.00034672738318148303,
      "loss": 0.4457,
      "num_input_tokens_seen": 15916672,
      "step": 24395
    },
    {
      "epoch": 12.78825995807128,
      "grad_norm": 0.13719119131565094,
      "learning_rate": 0.00034650972334513043,
      "loss": 0.4169,
      "num_input_tokens_seen": 15920288,
      "step": 24400
    },
    {
      "epoch": 12.790880503144654,
      "grad_norm": 0.1027563139796257,
      "learning_rate": 0.00034629209561719244,
      "loss": 0.5569,
      "num_input_tokens_seen": 15924736,
      "step": 24405
    },
    {
      "epoch": 12.79350104821803,
      "grad_norm": 0.12411993741989136,
      "learning_rate": 0.0003460745000431946,
      "loss": 0.3611,
      "num_input_tokens_seen": 15928544,
      "step": 24410
    },
    {
      "epoch": 12.796121593291405,
      "grad_norm": 0.14918017387390137,
      "learning_rate": 0.00034585693666865525,
      "loss": 0.3296,
      "num_input_tokens_seen": 15932864,
      "step": 24415
    },
    {
      "epoch": 12.79874213836478,
      "grad_norm": 0.06460008770227432,
      "learning_rate": 0.00034563940553908605,
      "loss": 0.379,
      "num_input_tokens_seen": 15936512,
      "step": 24420
    },
    {
      "epoch": 12.801362683438155,
      "grad_norm": 0.11621135473251343,
      "learning_rate": 0.00034542190669999216,
      "loss": 0.452,
      "num_input_tokens_seen": 15939392,
      "step": 24425
    },
    {
      "epoch": 12.80398322851153,
      "grad_norm": 0.11130107939243317,
      "learning_rate": 0.000345204440196872,
      "loss": 0.3885,
      "num_input_tokens_seen": 15942272,
      "step": 24430
    },
    {
      "epoch": 12.806603773584905,
      "grad_norm": 0.15385472774505615,
      "learning_rate": 0.00034498700607521693,
      "loss": 0.5127,
      "num_input_tokens_seen": 15945280,
      "step": 24435
    },
    {
      "epoch": 12.80922431865828,
      "grad_norm": 0.33676403760910034,
      "learning_rate": 0.00034476960438051187,
      "loss": 0.4416,
      "num_input_tokens_seen": 15948704,
      "step": 24440
    },
    {
      "epoch": 12.811844863731656,
      "grad_norm": 0.09155430644750595,
      "learning_rate": 0.00034455223515823444,
      "loss": 0.3869,
      "num_input_tokens_seen": 15952320,
      "step": 24445
    },
    {
      "epoch": 12.814465408805031,
      "grad_norm": 0.10440824925899506,
      "learning_rate": 0.0003443348984538559,
      "loss": 0.3875,
      "num_input_tokens_seen": 15956576,
      "step": 24450
    },
    {
      "epoch": 12.817085953878406,
      "grad_norm": 0.09283944219350815,
      "learning_rate": 0.0003441175943128407,
      "loss": 0.3814,
      "num_input_tokens_seen": 15959232,
      "step": 24455
    },
    {
      "epoch": 12.819706498951781,
      "grad_norm": 0.08463060110807419,
      "learning_rate": 0.0003439003227806464,
      "loss": 0.4037,
      "num_input_tokens_seen": 15962720,
      "step": 24460
    },
    {
      "epoch": 12.822327044025156,
      "grad_norm": 0.10781931132078171,
      "learning_rate": 0.00034368308390272386,
      "loss": 0.3838,
      "num_input_tokens_seen": 15966240,
      "step": 24465
    },
    {
      "epoch": 12.824947589098532,
      "grad_norm": 0.13998891413211823,
      "learning_rate": 0.0003434658777245167,
      "loss": 0.5359,
      "num_input_tokens_seen": 15968800,
      "step": 24470
    },
    {
      "epoch": 12.827568134171909,
      "grad_norm": 0.1308983415365219,
      "learning_rate": 0.000343248704291462,
      "loss": 0.4668,
      "num_input_tokens_seen": 15971680,
      "step": 24475
    },
    {
      "epoch": 12.830188679245284,
      "grad_norm": 0.09049108624458313,
      "learning_rate": 0.00034303156364899016,
      "loss": 0.5136,
      "num_input_tokens_seen": 15975552,
      "step": 24480
    },
    {
      "epoch": 12.832809224318659,
      "grad_norm": 0.17605847120285034,
      "learning_rate": 0.00034281445584252425,
      "loss": 0.4632,
      "num_input_tokens_seen": 15978656,
      "step": 24485
    },
    {
      "epoch": 12.835429769392034,
      "grad_norm": 0.14985372126102448,
      "learning_rate": 0.00034259738091748103,
      "loss": 0.4688,
      "num_input_tokens_seen": 15981856,
      "step": 24490
    },
    {
      "epoch": 12.83805031446541,
      "grad_norm": 0.1682436615228653,
      "learning_rate": 0.00034238033891927013,
      "loss": 0.5216,
      "num_input_tokens_seen": 15984736,
      "step": 24495
    },
    {
      "epoch": 12.840670859538784,
      "grad_norm": 0.14736111462116241,
      "learning_rate": 0.00034216332989329387,
      "loss": 0.3733,
      "num_input_tokens_seen": 15987424,
      "step": 24500
    },
    {
      "epoch": 12.84329140461216,
      "grad_norm": 0.15706036984920502,
      "learning_rate": 0.00034194635388494835,
      "loss": 0.6837,
      "num_input_tokens_seen": 15991296,
      "step": 24505
    },
    {
      "epoch": 12.845911949685535,
      "grad_norm": 0.11616985499858856,
      "learning_rate": 0.00034172941093962237,
      "loss": 0.5857,
      "num_input_tokens_seen": 15993760,
      "step": 24510
    },
    {
      "epoch": 12.84853249475891,
      "grad_norm": 0.14200250804424286,
      "learning_rate": 0.00034151250110269805,
      "loss": 0.3421,
      "num_input_tokens_seen": 15996288,
      "step": 24515
    },
    {
      "epoch": 12.851153039832285,
      "grad_norm": 0.123048335313797,
      "learning_rate": 0.00034129562441955044,
      "loss": 0.4882,
      "num_input_tokens_seen": 15999328,
      "step": 24520
    },
    {
      "epoch": 12.85377358490566,
      "grad_norm": 0.16483359038829803,
      "learning_rate": 0.00034107878093554765,
      "loss": 0.4102,
      "num_input_tokens_seen": 16002336,
      "step": 24525
    },
    {
      "epoch": 12.856394129979035,
      "grad_norm": 0.0907866582274437,
      "learning_rate": 0.00034086197069605064,
      "loss": 0.2993,
      "num_input_tokens_seen": 16005440,
      "step": 24530
    },
    {
      "epoch": 12.85901467505241,
      "grad_norm": 0.11481958627700806,
      "learning_rate": 0.0003406451937464139,
      "loss": 0.5422,
      "num_input_tokens_seen": 16009376,
      "step": 24535
    },
    {
      "epoch": 12.861635220125786,
      "grad_norm": 0.16327646374702454,
      "learning_rate": 0.0003404284501319845,
      "loss": 0.4176,
      "num_input_tokens_seen": 16012768,
      "step": 24540
    },
    {
      "epoch": 12.864255765199161,
      "grad_norm": 0.11784915626049042,
      "learning_rate": 0.00034021173989810306,
      "loss": 0.4069,
      "num_input_tokens_seen": 16015136,
      "step": 24545
    },
    {
      "epoch": 12.866876310272536,
      "grad_norm": 0.141217902302742,
      "learning_rate": 0.0003399950630901025,
      "loss": 0.4653,
      "num_input_tokens_seen": 16018688,
      "step": 24550
    },
    {
      "epoch": 12.869496855345911,
      "grad_norm": 0.11249299347400665,
      "learning_rate": 0.0003397784197533095,
      "loss": 0.6755,
      "num_input_tokens_seen": 16021504,
      "step": 24555
    },
    {
      "epoch": 12.872117400419286,
      "grad_norm": 0.08874132484197617,
      "learning_rate": 0.0003395618099330431,
      "loss": 0.4244,
      "num_input_tokens_seen": 16024736,
      "step": 24560
    },
    {
      "epoch": 12.874737945492662,
      "grad_norm": 0.19437038898468018,
      "learning_rate": 0.0003393452336746156,
      "loss": 0.4307,
      "num_input_tokens_seen": 16027296,
      "step": 24565
    },
    {
      "epoch": 12.877358490566039,
      "grad_norm": 0.06877549737691879,
      "learning_rate": 0.00033912869102333237,
      "loss": 0.3542,
      "num_input_tokens_seen": 16029920,
      "step": 24570
    },
    {
      "epoch": 12.879979035639414,
      "grad_norm": 0.10556559264659882,
      "learning_rate": 0.00033891218202449164,
      "loss": 0.373,
      "num_input_tokens_seen": 16032672,
      "step": 24575
    },
    {
      "epoch": 12.882599580712789,
      "grad_norm": 0.29074808955192566,
      "learning_rate": 0.00033869570672338486,
      "loss": 0.3284,
      "num_input_tokens_seen": 16035040,
      "step": 24580
    },
    {
      "epoch": 12.885220125786164,
      "grad_norm": 0.09376557171344757,
      "learning_rate": 0.00033847926516529574,
      "loss": 0.3766,
      "num_input_tokens_seen": 16040064,
      "step": 24585
    },
    {
      "epoch": 12.88784067085954,
      "grad_norm": 0.10657884925603867,
      "learning_rate": 0.0003382628573955018,
      "loss": 0.4138,
      "num_input_tokens_seen": 16042784,
      "step": 24590
    },
    {
      "epoch": 12.890461215932914,
      "grad_norm": 0.07127305120229721,
      "learning_rate": 0.00033804648345927267,
      "loss": 0.3703,
      "num_input_tokens_seen": 16045952,
      "step": 24595
    },
    {
      "epoch": 12.89308176100629,
      "grad_norm": 0.16268904507160187,
      "learning_rate": 0.00033783014340187166,
      "loss": 0.4401,
      "num_input_tokens_seen": 16049248,
      "step": 24600
    },
    {
      "epoch": 12.895702306079665,
      "grad_norm": 0.10912831127643585,
      "learning_rate": 0.0003376138372685545,
      "loss": 0.5119,
      "num_input_tokens_seen": 16052576,
      "step": 24605
    },
    {
      "epoch": 12.89832285115304,
      "grad_norm": 0.17654578387737274,
      "learning_rate": 0.00033739756510457015,
      "loss": 0.6587,
      "num_input_tokens_seen": 16055936,
      "step": 24610
    },
    {
      "epoch": 12.900943396226415,
      "grad_norm": 0.21184904873371124,
      "learning_rate": 0.00033718132695516003,
      "loss": 0.5539,
      "num_input_tokens_seen": 16058176,
      "step": 24615
    },
    {
      "epoch": 12.90356394129979,
      "grad_norm": 0.12575411796569824,
      "learning_rate": 0.00033696512286555865,
      "loss": 0.4822,
      "num_input_tokens_seen": 16061568,
      "step": 24620
    },
    {
      "epoch": 12.906184486373165,
      "grad_norm": 0.1349324882030487,
      "learning_rate": 0.0003367489528809936,
      "loss": 0.5082,
      "num_input_tokens_seen": 16065536,
      "step": 24625
    },
    {
      "epoch": 12.90880503144654,
      "grad_norm": 0.09922154247760773,
      "learning_rate": 0.00033653281704668515,
      "loss": 0.4811,
      "num_input_tokens_seen": 16068192,
      "step": 24630
    },
    {
      "epoch": 12.911425576519916,
      "grad_norm": 0.16671998798847198,
      "learning_rate": 0.0003363167154078465,
      "loss": 0.4642,
      "num_input_tokens_seen": 16071008,
      "step": 24635
    },
    {
      "epoch": 12.914046121593291,
      "grad_norm": 0.10095272213220596,
      "learning_rate": 0.0003361006480096837,
      "loss": 0.4206,
      "num_input_tokens_seen": 16074688,
      "step": 24640
    },
    {
      "epoch": 12.916666666666666,
      "grad_norm": 0.19448238611221313,
      "learning_rate": 0.0003358846148973953,
      "loss": 0.3964,
      "num_input_tokens_seen": 16076928,
      "step": 24645
    },
    {
      "epoch": 12.919287211740041,
      "grad_norm": 0.1355290710926056,
      "learning_rate": 0.0003356686161161732,
      "loss": 0.5312,
      "num_input_tokens_seen": 16079904,
      "step": 24650
    },
    {
      "epoch": 12.921907756813416,
      "grad_norm": 0.08062653988599777,
      "learning_rate": 0.00033545265171120163,
      "loss": 0.3497,
      "num_input_tokens_seen": 16083168,
      "step": 24655
    },
    {
      "epoch": 12.924528301886792,
      "grad_norm": 0.1307508498430252,
      "learning_rate": 0.0003352367217276582,
      "loss": 0.5554,
      "num_input_tokens_seen": 16086592,
      "step": 24660
    },
    {
      "epoch": 12.927148846960169,
      "grad_norm": 0.1441798061132431,
      "learning_rate": 0.0003350208262107131,
      "loss": 0.4836,
      "num_input_tokens_seen": 16089088,
      "step": 24665
    },
    {
      "epoch": 12.929769392033544,
      "grad_norm": 0.14503853023052216,
      "learning_rate": 0.00033480496520552874,
      "loss": 0.7884,
      "num_input_tokens_seen": 16091936,
      "step": 24670
    },
    {
      "epoch": 12.932389937106919,
      "grad_norm": 0.3161413073539734,
      "learning_rate": 0.00033458913875726093,
      "loss": 0.4347,
      "num_input_tokens_seen": 16094336,
      "step": 24675
    },
    {
      "epoch": 12.935010482180294,
      "grad_norm": 0.13487930595874786,
      "learning_rate": 0.00033437334691105837,
      "loss": 0.4265,
      "num_input_tokens_seen": 16098304,
      "step": 24680
    },
    {
      "epoch": 12.93763102725367,
      "grad_norm": 0.09491321444511414,
      "learning_rate": 0.0003341575897120619,
      "loss": 0.4267,
      "num_input_tokens_seen": 16101632,
      "step": 24685
    },
    {
      "epoch": 12.940251572327044,
      "grad_norm": 0.11743921786546707,
      "learning_rate": 0.00033394186720540577,
      "loss": 0.4002,
      "num_input_tokens_seen": 16105024,
      "step": 24690
    },
    {
      "epoch": 12.94287211740042,
      "grad_norm": 0.10878322273492813,
      "learning_rate": 0.00033372617943621665,
      "loss": 0.3602,
      "num_input_tokens_seen": 16108096,
      "step": 24695
    },
    {
      "epoch": 12.945492662473795,
      "grad_norm": 0.12384265661239624,
      "learning_rate": 0.00033351052644961353,
      "loss": 0.4372,
      "num_input_tokens_seen": 16111648,
      "step": 24700
    },
    {
      "epoch": 12.94811320754717,
      "grad_norm": 0.13015632331371307,
      "learning_rate": 0.00033329490829070907,
      "loss": 0.4999,
      "num_input_tokens_seen": 16114752,
      "step": 24705
    },
    {
      "epoch": 12.950733752620545,
      "grad_norm": 0.15553395450115204,
      "learning_rate": 0.0003330793250046078,
      "loss": 0.5385,
      "num_input_tokens_seen": 16119616,
      "step": 24710
    },
    {
      "epoch": 12.95335429769392,
      "grad_norm": 0.10458353161811829,
      "learning_rate": 0.0003328637766364075,
      "loss": 0.468,
      "num_input_tokens_seen": 16123232,
      "step": 24715
    },
    {
      "epoch": 12.955974842767295,
      "grad_norm": 0.0860501304268837,
      "learning_rate": 0.00033264826323119824,
      "loss": 0.464,
      "num_input_tokens_seen": 16127552,
      "step": 24720
    },
    {
      "epoch": 12.95859538784067,
      "grad_norm": 0.1525387316942215,
      "learning_rate": 0.00033243278483406326,
      "loss": 0.4159,
      "num_input_tokens_seen": 16130368,
      "step": 24725
    },
    {
      "epoch": 12.961215932914046,
      "grad_norm": 0.18871574103832245,
      "learning_rate": 0.00033221734149007785,
      "loss": 0.4007,
      "num_input_tokens_seen": 16134208,
      "step": 24730
    },
    {
      "epoch": 12.963836477987421,
      "grad_norm": 0.09854216873645782,
      "learning_rate": 0.0003320019332443104,
      "loss": 0.4159,
      "num_input_tokens_seen": 16137120,
      "step": 24735
    },
    {
      "epoch": 12.966457023060796,
      "grad_norm": 0.08952595293521881,
      "learning_rate": 0.00033178656014182185,
      "loss": 0.4989,
      "num_input_tokens_seen": 16140576,
      "step": 24740
    },
    {
      "epoch": 12.969077568134171,
      "grad_norm": 0.157258078455925,
      "learning_rate": 0.00033157122222766597,
      "loss": 0.4891,
      "num_input_tokens_seen": 16143808,
      "step": 24745
    },
    {
      "epoch": 12.971698113207546,
      "grad_norm": 0.13118119537830353,
      "learning_rate": 0.0003313559195468889,
      "loss": 0.5541,
      "num_input_tokens_seen": 16147136,
      "step": 24750
    },
    {
      "epoch": 12.974318658280922,
      "grad_norm": 0.10318174958229065,
      "learning_rate": 0.0003311406521445293,
      "loss": 0.4189,
      "num_input_tokens_seen": 16150272,
      "step": 24755
    },
    {
      "epoch": 12.976939203354299,
      "grad_norm": 0.15731075406074524,
      "learning_rate": 0.000330925420065619,
      "loss": 0.506,
      "num_input_tokens_seen": 16153920,
      "step": 24760
    },
    {
      "epoch": 12.979559748427674,
      "grad_norm": 0.08721112459897995,
      "learning_rate": 0.0003307102233551817,
      "loss": 0.3731,
      "num_input_tokens_seen": 16157280,
      "step": 24765
    },
    {
      "epoch": 12.982180293501049,
      "grad_norm": 0.1764710247516632,
      "learning_rate": 0.0003304950620582344,
      "loss": 0.4751,
      "num_input_tokens_seen": 16161056,
      "step": 24770
    },
    {
      "epoch": 12.984800838574424,
      "grad_norm": 0.12457165122032166,
      "learning_rate": 0.0003302799362197864,
      "loss": 0.4683,
      "num_input_tokens_seen": 16163776,
      "step": 24775
    },
    {
      "epoch": 12.9874213836478,
      "grad_norm": 0.13062795996665955,
      "learning_rate": 0.00033006484588483947,
      "loss": 0.3902,
      "num_input_tokens_seen": 16166176,
      "step": 24780
    },
    {
      "epoch": 12.990041928721174,
      "grad_norm": 0.11611783504486084,
      "learning_rate": 0.00032984979109838803,
      "loss": 0.3613,
      "num_input_tokens_seen": 16168704,
      "step": 24785
    },
    {
      "epoch": 12.99266247379455,
      "grad_norm": 0.10469408333301544,
      "learning_rate": 0.000329634771905419,
      "loss": 0.4408,
      "num_input_tokens_seen": 16171584,
      "step": 24790
    },
    {
      "epoch": 12.995283018867925,
      "grad_norm": 0.16522656381130219,
      "learning_rate": 0.0003294197883509122,
      "loss": 0.5501,
      "num_input_tokens_seen": 16174656,
      "step": 24795
    },
    {
      "epoch": 12.9979035639413,
      "grad_norm": 0.09761796146631241,
      "learning_rate": 0.00032920484047983947,
      "loss": 0.5904,
      "num_input_tokens_seen": 16177344,
      "step": 24800
    },
    {
      "epoch": 13.0,
      "eval_loss": 0.45868632197380066,
      "eval_runtime": 13.6085,
      "eval_samples_per_second": 62.314,
      "eval_steps_per_second": 15.579,
      "num_input_tokens_seen": 16179624,
      "step": 24804
    },
    {
      "epoch": 13.000524109014675,
      "grad_norm": 0.13743223249912262,
      "learning_rate": 0.0003289899283371657,
      "loss": 0.4989,
      "num_input_tokens_seen": 16180040,
      "step": 24805
    },
    {
      "epoch": 13.00314465408805,
      "grad_norm": 0.2475561946630478,
      "learning_rate": 0.00032877505196784795,
      "loss": 0.4634,
      "num_input_tokens_seen": 16183176,
      "step": 24810
    },
    {
      "epoch": 13.005765199161425,
      "grad_norm": 0.1605468988418579,
      "learning_rate": 0.0003285602114168358,
      "loss": 0.4018,
      "num_input_tokens_seen": 16185928,
      "step": 24815
    },
    {
      "epoch": 13.0083857442348,
      "grad_norm": 0.11197611689567566,
      "learning_rate": 0.00032834540672907144,
      "loss": 0.5085,
      "num_input_tokens_seen": 16189160,
      "step": 24820
    },
    {
      "epoch": 13.011006289308176,
      "grad_norm": 0.11984051764011383,
      "learning_rate": 0.00032813063794948983,
      "loss": 0.2896,
      "num_input_tokens_seen": 16192392,
      "step": 24825
    },
    {
      "epoch": 13.01362683438155,
      "grad_norm": 0.07833502441644669,
      "learning_rate": 0.00032791590512301784,
      "loss": 0.4511,
      "num_input_tokens_seen": 16195752,
      "step": 24830
    },
    {
      "epoch": 13.016247379454926,
      "grad_norm": 0.09482721239328384,
      "learning_rate": 0.00032770120829457534,
      "loss": 0.3863,
      "num_input_tokens_seen": 16199176,
      "step": 24835
    },
    {
      "epoch": 13.018867924528301,
      "grad_norm": 0.12545455992221832,
      "learning_rate": 0.00032748654750907447,
      "loss": 0.5443,
      "num_input_tokens_seen": 16201768,
      "step": 24840
    },
    {
      "epoch": 13.021488469601676,
      "grad_norm": 0.15650303661823273,
      "learning_rate": 0.0003272719228114196,
      "loss": 0.391,
      "num_input_tokens_seen": 16204680,
      "step": 24845
    },
    {
      "epoch": 13.024109014675052,
      "grad_norm": 0.09711945801973343,
      "learning_rate": 0.0003270573342465079,
      "loss": 0.4469,
      "num_input_tokens_seen": 16208264,
      "step": 24850
    },
    {
      "epoch": 13.026729559748428,
      "grad_norm": 0.18712525069713593,
      "learning_rate": 0.00032684278185922877,
      "loss": 0.3676,
      "num_input_tokens_seen": 16210536,
      "step": 24855
    },
    {
      "epoch": 13.029350104821804,
      "grad_norm": 0.16795092821121216,
      "learning_rate": 0.0003266282656944643,
      "loss": 0.5692,
      "num_input_tokens_seen": 16214376,
      "step": 24860
    },
    {
      "epoch": 13.031970649895179,
      "grad_norm": 0.2082168459892273,
      "learning_rate": 0.0003264137857970888,
      "loss": 0.5512,
      "num_input_tokens_seen": 16217416,
      "step": 24865
    },
    {
      "epoch": 13.034591194968554,
      "grad_norm": 0.11734774708747864,
      "learning_rate": 0.0003261993422119688,
      "loss": 0.472,
      "num_input_tokens_seen": 16220584,
      "step": 24870
    },
    {
      "epoch": 13.03721174004193,
      "grad_norm": 0.13010883331298828,
      "learning_rate": 0.00032598493498396365,
      "loss": 0.3496,
      "num_input_tokens_seen": 16224040,
      "step": 24875
    },
    {
      "epoch": 13.039832285115304,
      "grad_norm": 0.17013880610466003,
      "learning_rate": 0.0003257705641579247,
      "loss": 0.4479,
      "num_input_tokens_seen": 16227016,
      "step": 24880
    },
    {
      "epoch": 13.04245283018868,
      "grad_norm": 0.1461832970380783,
      "learning_rate": 0.00032555622977869614,
      "loss": 0.4787,
      "num_input_tokens_seen": 16229768,
      "step": 24885
    },
    {
      "epoch": 13.045073375262055,
      "grad_norm": 0.07901003956794739,
      "learning_rate": 0.00032534193189111407,
      "loss": 0.3398,
      "num_input_tokens_seen": 16233224,
      "step": 24890
    },
    {
      "epoch": 13.04769392033543,
      "grad_norm": 0.12289204448461533,
      "learning_rate": 0.0003251276705400074,
      "loss": 0.3309,
      "num_input_tokens_seen": 16236264,
      "step": 24895
    },
    {
      "epoch": 13.050314465408805,
      "grad_norm": 0.09240709245204926,
      "learning_rate": 0.0003249134457701968,
      "loss": 0.3948,
      "num_input_tokens_seen": 16238984,
      "step": 24900
    },
    {
      "epoch": 13.05293501048218,
      "grad_norm": 0.1610741764307022,
      "learning_rate": 0.000324699257626496,
      "loss": 0.3767,
      "num_input_tokens_seen": 16241608,
      "step": 24905
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 0.17479805648326874,
      "learning_rate": 0.00032448510615371045,
      "loss": 0.4519,
      "num_input_tokens_seen": 16244328,
      "step": 24910
    },
    {
      "epoch": 13.05817610062893,
      "grad_norm": 0.08799177408218384,
      "learning_rate": 0.0003242709913966384,
      "loss": 0.484,
      "num_input_tokens_seen": 16247912,
      "step": 24915
    },
    {
      "epoch": 13.060796645702306,
      "grad_norm": 0.12052842974662781,
      "learning_rate": 0.00032405691340007,
      "loss": 0.4899,
      "num_input_tokens_seen": 16250920,
      "step": 24920
    },
    {
      "epoch": 13.06341719077568,
      "grad_norm": 0.12706723809242249,
      "learning_rate": 0.0003238428722087884,
      "loss": 0.4772,
      "num_input_tokens_seen": 16254728,
      "step": 24925
    },
    {
      "epoch": 13.066037735849056,
      "grad_norm": 0.1287400722503662,
      "learning_rate": 0.00032362886786756797,
      "loss": 0.3624,
      "num_input_tokens_seen": 16257640,
      "step": 24930
    },
    {
      "epoch": 13.068658280922431,
      "grad_norm": 0.2809236943721771,
      "learning_rate": 0.00032341490042117615,
      "loss": 0.5078,
      "num_input_tokens_seen": 16260808,
      "step": 24935
    },
    {
      "epoch": 13.071278825995806,
      "grad_norm": 0.11760123074054718,
      "learning_rate": 0.00032320096991437277,
      "loss": 0.3766,
      "num_input_tokens_seen": 16264264,
      "step": 24940
    },
    {
      "epoch": 13.073899371069182,
      "grad_norm": 0.11264221370220184,
      "learning_rate": 0.0003229870763919093,
      "loss": 0.4104,
      "num_input_tokens_seen": 16267528,
      "step": 24945
    },
    {
      "epoch": 13.076519916142558,
      "grad_norm": 0.19582019746303558,
      "learning_rate": 0.00032277321989853015,
      "loss": 0.5571,
      "num_input_tokens_seen": 16270088,
      "step": 24950
    },
    {
      "epoch": 13.079140461215934,
      "grad_norm": 0.12020466476678848,
      "learning_rate": 0.00032255940047897137,
      "loss": 0.5233,
      "num_input_tokens_seen": 16273320,
      "step": 24955
    },
    {
      "epoch": 13.081761006289309,
      "grad_norm": 0.25859206914901733,
      "learning_rate": 0.0003223456181779616,
      "loss": 0.4398,
      "num_input_tokens_seen": 16276232,
      "step": 24960
    },
    {
      "epoch": 13.084381551362684,
      "grad_norm": 0.08892641961574554,
      "learning_rate": 0.0003221318730402218,
      "loss": 0.5193,
      "num_input_tokens_seen": 16279272,
      "step": 24965
    },
    {
      "epoch": 13.08700209643606,
      "grad_norm": 0.10580413788557053,
      "learning_rate": 0.00032191816511046483,
      "loss": 0.4098,
      "num_input_tokens_seen": 16282024,
      "step": 24970
    },
    {
      "epoch": 13.089622641509434,
      "grad_norm": 0.18706177175045013,
      "learning_rate": 0.0003217044944333961,
      "loss": 0.4231,
      "num_input_tokens_seen": 16284808,
      "step": 24975
    },
    {
      "epoch": 13.09224318658281,
      "grad_norm": 0.10947602242231369,
      "learning_rate": 0.00032149086105371315,
      "loss": 0.64,
      "num_input_tokens_seen": 16288200,
      "step": 24980
    },
    {
      "epoch": 13.094863731656185,
      "grad_norm": 0.08405328541994095,
      "learning_rate": 0.00032127726501610554,
      "loss": 0.4141,
      "num_input_tokens_seen": 16291176,
      "step": 24985
    },
    {
      "epoch": 13.09748427672956,
      "grad_norm": 0.1723424792289734,
      "learning_rate": 0.00032106370636525493,
      "loss": 0.4835,
      "num_input_tokens_seen": 16293896,
      "step": 24990
    },
    {
      "epoch": 13.100104821802935,
      "grad_norm": 0.14498716592788696,
      "learning_rate": 0.00032085018514583564,
      "loss": 0.3679,
      "num_input_tokens_seen": 16297192,
      "step": 24995
    },
    {
      "epoch": 13.10272536687631,
      "grad_norm": 0.08531923592090607,
      "learning_rate": 0.00032063670140251366,
      "loss": 0.493,
      "num_input_tokens_seen": 16300680,
      "step": 25000
    },
    {
      "epoch": 13.105345911949685,
      "grad_norm": 0.1293623447418213,
      "learning_rate": 0.0003204232551799476,
      "loss": 0.2395,
      "num_input_tokens_seen": 16303080,
      "step": 25005
    },
    {
      "epoch": 13.10796645702306,
      "grad_norm": 0.13019227981567383,
      "learning_rate": 0.00032020984652278797,
      "loss": 0.469,
      "num_input_tokens_seen": 16305224,
      "step": 25010
    },
    {
      "epoch": 13.110587002096436,
      "grad_norm": 0.1277080476284027,
      "learning_rate": 0.000319996475475677,
      "loss": 0.5494,
      "num_input_tokens_seen": 16308168,
      "step": 25015
    },
    {
      "epoch": 13.11320754716981,
      "grad_norm": 0.08147195726633072,
      "learning_rate": 0.0003197831420832499,
      "loss": 0.4656,
      "num_input_tokens_seen": 16312072,
      "step": 25020
    },
    {
      "epoch": 13.115828092243186,
      "grad_norm": 0.12433242052793503,
      "learning_rate": 0.0003195698463901334,
      "loss": 0.4813,
      "num_input_tokens_seen": 16315112,
      "step": 25025
    },
    {
      "epoch": 13.118448637316561,
      "grad_norm": 0.1530289649963379,
      "learning_rate": 0.0003193565884409466,
      "loss": 0.5548,
      "num_input_tokens_seen": 16317704,
      "step": 25030
    },
    {
      "epoch": 13.121069182389936,
      "grad_norm": 0.12382163107395172,
      "learning_rate": 0.0003191433682803008,
      "loss": 0.4347,
      "num_input_tokens_seen": 16321288,
      "step": 25035
    },
    {
      "epoch": 13.123689727463312,
      "grad_norm": 0.09218622744083405,
      "learning_rate": 0.00031893018595279877,
      "loss": 0.322,
      "num_input_tokens_seen": 16324360,
      "step": 25040
    },
    {
      "epoch": 13.126310272536688,
      "grad_norm": 0.07516369223594666,
      "learning_rate": 0.00031871704150303624,
      "loss": 0.3643,
      "num_input_tokens_seen": 16327784,
      "step": 25045
    },
    {
      "epoch": 13.128930817610064,
      "grad_norm": 0.2805236876010895,
      "learning_rate": 0.00031850393497560034,
      "loss": 0.5527,
      "num_input_tokens_seen": 16331848,
      "step": 25050
    },
    {
      "epoch": 13.131551362683439,
      "grad_norm": 0.23075447976589203,
      "learning_rate": 0.00031829086641507054,
      "loss": 0.4775,
      "num_input_tokens_seen": 16335592,
      "step": 25055
    },
    {
      "epoch": 13.134171907756814,
      "grad_norm": 0.09219196438789368,
      "learning_rate": 0.00031807783586601865,
      "loss": 0.5506,
      "num_input_tokens_seen": 16340200,
      "step": 25060
    },
    {
      "epoch": 13.13679245283019,
      "grad_norm": 0.17963476479053497,
      "learning_rate": 0.00031786484337300805,
      "loss": 0.4837,
      "num_input_tokens_seen": 16343208,
      "step": 25065
    },
    {
      "epoch": 13.139412997903564,
      "grad_norm": 0.08993315696716309,
      "learning_rate": 0.0003176518889805941,
      "loss": 0.3508,
      "num_input_tokens_seen": 16346088,
      "step": 25070
    },
    {
      "epoch": 13.14203354297694,
      "grad_norm": 0.09306658059358597,
      "learning_rate": 0.0003174389727333248,
      "loss": 0.4582,
      "num_input_tokens_seen": 16349768,
      "step": 25075
    },
    {
      "epoch": 13.144654088050315,
      "grad_norm": 0.16056673228740692,
      "learning_rate": 0.00031722609467573946,
      "loss": 0.44,
      "num_input_tokens_seen": 16353960,
      "step": 25080
    },
    {
      "epoch": 13.14727463312369,
      "grad_norm": 0.12918223440647125,
      "learning_rate": 0.0003170132548523701,
      "loss": 0.579,
      "num_input_tokens_seen": 16356904,
      "step": 25085
    },
    {
      "epoch": 13.149895178197065,
      "grad_norm": 0.11743903160095215,
      "learning_rate": 0.00031680045330774003,
      "loss": 0.368,
      "num_input_tokens_seen": 16360040,
      "step": 25090
    },
    {
      "epoch": 13.15251572327044,
      "grad_norm": 0.3565417528152466,
      "learning_rate": 0.00031658769008636534,
      "loss": 0.4377,
      "num_input_tokens_seen": 16363496,
      "step": 25095
    },
    {
      "epoch": 13.155136268343815,
      "grad_norm": 0.1633545458316803,
      "learning_rate": 0.0003163749652327532,
      "loss": 0.4494,
      "num_input_tokens_seen": 16366376,
      "step": 25100
    },
    {
      "epoch": 13.15775681341719,
      "grad_norm": 0.08607733994722366,
      "learning_rate": 0.00031616227879140347,
      "loss": 0.5176,
      "num_input_tokens_seen": 16370184,
      "step": 25105
    },
    {
      "epoch": 13.160377358490566,
      "grad_norm": 0.11482041329145432,
      "learning_rate": 0.0003159496308068077,
      "loss": 0.4041,
      "num_input_tokens_seen": 16374152,
      "step": 25110
    },
    {
      "epoch": 13.16299790356394,
      "grad_norm": 0.1041068509221077,
      "learning_rate": 0.00031573702132344926,
      "loss": 0.4344,
      "num_input_tokens_seen": 16376904,
      "step": 25115
    },
    {
      "epoch": 13.165618448637316,
      "grad_norm": 0.1397431194782257,
      "learning_rate": 0.00031552445038580414,
      "loss": 0.3872,
      "num_input_tokens_seen": 16379784,
      "step": 25120
    },
    {
      "epoch": 13.168238993710691,
      "grad_norm": 0.09688485413789749,
      "learning_rate": 0.0003153119180383391,
      "loss": 0.4323,
      "num_input_tokens_seen": 16382888,
      "step": 25125
    },
    {
      "epoch": 13.170859538784066,
      "grad_norm": 0.11159984022378922,
      "learning_rate": 0.0003150994243255138,
      "loss": 0.3843,
      "num_input_tokens_seen": 16386472,
      "step": 25130
    },
    {
      "epoch": 13.173480083857442,
      "grad_norm": 0.2733440399169922,
      "learning_rate": 0.00031488696929177944,
      "loss": 0.564,
      "num_input_tokens_seen": 16389608,
      "step": 25135
    },
    {
      "epoch": 13.176100628930818,
      "grad_norm": 0.178640216588974,
      "learning_rate": 0.00031467455298157934,
      "loss": 0.353,
      "num_input_tokens_seen": 16393032,
      "step": 25140
    },
    {
      "epoch": 13.178721174004194,
      "grad_norm": 0.09422782808542252,
      "learning_rate": 0.0003144621754393483,
      "loss": 0.3613,
      "num_input_tokens_seen": 16396712,
      "step": 25145
    },
    {
      "epoch": 13.181341719077569,
      "grad_norm": 0.35924315452575684,
      "learning_rate": 0.0003142498367095137,
      "loss": 0.4972,
      "num_input_tokens_seen": 16399272,
      "step": 25150
    },
    {
      "epoch": 13.183962264150944,
      "grad_norm": 0.1288118064403534,
      "learning_rate": 0.00031403753683649396,
      "loss": 0.3959,
      "num_input_tokens_seen": 16402088,
      "step": 25155
    },
    {
      "epoch": 13.18658280922432,
      "grad_norm": 0.12365696579217911,
      "learning_rate": 0.0003138252758646999,
      "loss": 0.6086,
      "num_input_tokens_seen": 16406600,
      "step": 25160
    },
    {
      "epoch": 13.189203354297694,
      "grad_norm": 0.12622198462486267,
      "learning_rate": 0.00031361305383853413,
      "loss": 0.4022,
      "num_input_tokens_seen": 16409224,
      "step": 25165
    },
    {
      "epoch": 13.19182389937107,
      "grad_norm": 0.14887972176074982,
      "learning_rate": 0.00031340087080239105,
      "loss": 0.7269,
      "num_input_tokens_seen": 16412136,
      "step": 25170
    },
    {
      "epoch": 13.194444444444445,
      "grad_norm": 0.11691640317440033,
      "learning_rate": 0.0003131887268006571,
      "loss": 0.4119,
      "num_input_tokens_seen": 16416392,
      "step": 25175
    },
    {
      "epoch": 13.19706498951782,
      "grad_norm": 0.11492472141981125,
      "learning_rate": 0.00031297662187771026,
      "loss": 0.4705,
      "num_input_tokens_seen": 16419688,
      "step": 25180
    },
    {
      "epoch": 13.199685534591195,
      "grad_norm": 0.1274973452091217,
      "learning_rate": 0.0003127645560779203,
      "loss": 0.4492,
      "num_input_tokens_seen": 16422856,
      "step": 25185
    },
    {
      "epoch": 13.20230607966457,
      "grad_norm": 0.11699677258729935,
      "learning_rate": 0.0003125525294456492,
      "loss": 0.4466,
      "num_input_tokens_seen": 16425608,
      "step": 25190
    },
    {
      "epoch": 13.204926624737945,
      "grad_norm": 0.12796084582805634,
      "learning_rate": 0.0003123405420252503,
      "loss": 0.4108,
      "num_input_tokens_seen": 16429288,
      "step": 25195
    },
    {
      "epoch": 13.20754716981132,
      "grad_norm": 0.09706708788871765,
      "learning_rate": 0.00031212859386106927,
      "loss": 0.4758,
      "num_input_tokens_seen": 16432744,
      "step": 25200
    },
    {
      "epoch": 13.210167714884696,
      "grad_norm": 0.0672919824719429,
      "learning_rate": 0.00031191668499744286,
      "loss": 0.2952,
      "num_input_tokens_seen": 16435944,
      "step": 25205
    },
    {
      "epoch": 13.21278825995807,
      "grad_norm": 0.15093329548835754,
      "learning_rate": 0.00031170481547870046,
      "loss": 0.3296,
      "num_input_tokens_seen": 16438760,
      "step": 25210
    },
    {
      "epoch": 13.215408805031446,
      "grad_norm": 0.11768782883882523,
      "learning_rate": 0.00031149298534916215,
      "loss": 0.3922,
      "num_input_tokens_seen": 16442280,
      "step": 25215
    },
    {
      "epoch": 13.218029350104821,
      "grad_norm": 0.18018075823783875,
      "learning_rate": 0.00031128119465314094,
      "loss": 0.3508,
      "num_input_tokens_seen": 16445384,
      "step": 25220
    },
    {
      "epoch": 13.220649895178196,
      "grad_norm": 0.16795149445533752,
      "learning_rate": 0.0003110694434349406,
      "loss": 0.2925,
      "num_input_tokens_seen": 16448040,
      "step": 25225
    },
    {
      "epoch": 13.223270440251572,
      "grad_norm": 0.1490054875612259,
      "learning_rate": 0.0003108577317388574,
      "loss": 0.3971,
      "num_input_tokens_seen": 16451432,
      "step": 25230
    },
    {
      "epoch": 13.225890985324948,
      "grad_norm": 0.10726510733366013,
      "learning_rate": 0.00031064605960917896,
      "loss": 0.3711,
      "num_input_tokens_seen": 16455624,
      "step": 25235
    },
    {
      "epoch": 13.228511530398324,
      "grad_norm": 0.15869762003421783,
      "learning_rate": 0.00031043442709018446,
      "loss": 0.336,
      "num_input_tokens_seen": 16459112,
      "step": 25240
    },
    {
      "epoch": 13.231132075471699,
      "grad_norm": 0.09364533424377441,
      "learning_rate": 0.0003102228342261451,
      "loss": 0.4203,
      "num_input_tokens_seen": 16462312,
      "step": 25245
    },
    {
      "epoch": 13.233752620545074,
      "grad_norm": 0.18091905117034912,
      "learning_rate": 0.0003100112810613237,
      "loss": 0.5935,
      "num_input_tokens_seen": 16465128,
      "step": 25250
    },
    {
      "epoch": 13.23637316561845,
      "grad_norm": 0.07546059042215347,
      "learning_rate": 0.00030979976763997484,
      "loss": 0.386,
      "num_input_tokens_seen": 16469320,
      "step": 25255
    },
    {
      "epoch": 13.238993710691824,
      "grad_norm": 0.12678007781505585,
      "learning_rate": 0.0003095882940063447,
      "loss": 0.5704,
      "num_input_tokens_seen": 16472296,
      "step": 25260
    },
    {
      "epoch": 13.2416142557652,
      "grad_norm": 0.12455514818429947,
      "learning_rate": 0.0003093768602046712,
      "loss": 0.4911,
      "num_input_tokens_seen": 16475240,
      "step": 25265
    },
    {
      "epoch": 13.244234800838575,
      "grad_norm": 0.15970207750797272,
      "learning_rate": 0.00030916546627918375,
      "loss": 0.4707,
      "num_input_tokens_seen": 16477736,
      "step": 25270
    },
    {
      "epoch": 13.24685534591195,
      "grad_norm": 0.15288031101226807,
      "learning_rate": 0.00030895411227410354,
      "loss": 0.4881,
      "num_input_tokens_seen": 16480776,
      "step": 25275
    },
    {
      "epoch": 13.249475890985325,
      "grad_norm": 0.24665088951587677,
      "learning_rate": 0.0003087427982336435,
      "loss": 0.4188,
      "num_input_tokens_seen": 16484584,
      "step": 25280
    },
    {
      "epoch": 13.2520964360587,
      "grad_norm": 0.12778274714946747,
      "learning_rate": 0.00030853152420200804,
      "loss": 0.4129,
      "num_input_tokens_seen": 16487400,
      "step": 25285
    },
    {
      "epoch": 13.254716981132075,
      "grad_norm": 0.12223140150308609,
      "learning_rate": 0.0003083202902233932,
      "loss": 0.3981,
      "num_input_tokens_seen": 16490024,
      "step": 25290
    },
    {
      "epoch": 13.25733752620545,
      "grad_norm": 0.12859079241752625,
      "learning_rate": 0.00030810909634198727,
      "loss": 0.4537,
      "num_input_tokens_seen": 16493064,
      "step": 25295
    },
    {
      "epoch": 13.259958071278826,
      "grad_norm": 0.17318758368492126,
      "learning_rate": 0.00030789794260196893,
      "loss": 0.3641,
      "num_input_tokens_seen": 16496136,
      "step": 25300
    },
    {
      "epoch": 13.2625786163522,
      "grad_norm": 0.13319385051727295,
      "learning_rate": 0.00030768682904750935,
      "loss": 0.3836,
      "num_input_tokens_seen": 16500328,
      "step": 25305
    },
    {
      "epoch": 13.265199161425576,
      "grad_norm": 0.13389070332050323,
      "learning_rate": 0.00030747575572277107,
      "loss": 0.4545,
      "num_input_tokens_seen": 16503688,
      "step": 25310
    },
    {
      "epoch": 13.267819706498951,
      "grad_norm": 0.2967512905597687,
      "learning_rate": 0.0003072647226719083,
      "loss": 0.4427,
      "num_input_tokens_seen": 16507112,
      "step": 25315
    },
    {
      "epoch": 13.270440251572326,
      "grad_norm": 0.13786159455776215,
      "learning_rate": 0.0003070537299390669,
      "loss": 0.4518,
      "num_input_tokens_seen": 16510920,
      "step": 25320
    },
    {
      "epoch": 13.273060796645701,
      "grad_norm": 0.15908005833625793,
      "learning_rate": 0.0003068427775683838,
      "loss": 0.4569,
      "num_input_tokens_seen": 16513736,
      "step": 25325
    },
    {
      "epoch": 13.275681341719078,
      "grad_norm": 0.10084323585033417,
      "learning_rate": 0.00030663186560398793,
      "loss": 0.4766,
      "num_input_tokens_seen": 16516744,
      "step": 25330
    },
    {
      "epoch": 13.278301886792454,
      "grad_norm": 0.1732238233089447,
      "learning_rate": 0.0003064209940899998,
      "loss": 0.4835,
      "num_input_tokens_seen": 16519976,
      "step": 25335
    },
    {
      "epoch": 13.280922431865829,
      "grad_norm": 0.10732923448085785,
      "learning_rate": 0.00030621016307053105,
      "loss": 0.4234,
      "num_input_tokens_seen": 16523112,
      "step": 25340
    },
    {
      "epoch": 13.283542976939204,
      "grad_norm": 0.16969533264636993,
      "learning_rate": 0.0003059993725896855,
      "loss": 0.5294,
      "num_input_tokens_seen": 16526376,
      "step": 25345
    },
    {
      "epoch": 13.286163522012579,
      "grad_norm": 0.14430250227451324,
      "learning_rate": 0.0003057886226915581,
      "loss": 0.439,
      "num_input_tokens_seen": 16533160,
      "step": 25350
    },
    {
      "epoch": 13.288784067085954,
      "grad_norm": 0.12843205034732819,
      "learning_rate": 0.00030557791342023476,
      "loss": 0.3578,
      "num_input_tokens_seen": 16536680,
      "step": 25355
    },
    {
      "epoch": 13.29140461215933,
      "grad_norm": 0.1733255684375763,
      "learning_rate": 0.0003053672448197941,
      "loss": 0.394,
      "num_input_tokens_seen": 16539656,
      "step": 25360
    },
    {
      "epoch": 13.294025157232705,
      "grad_norm": 0.14685393869876862,
      "learning_rate": 0.0003051566169343052,
      "loss": 0.3916,
      "num_input_tokens_seen": 16542376,
      "step": 25365
    },
    {
      "epoch": 13.29664570230608,
      "grad_norm": 0.17988330125808716,
      "learning_rate": 0.00030494602980782917,
      "loss": 0.3956,
      "num_input_tokens_seen": 16545352,
      "step": 25370
    },
    {
      "epoch": 13.299266247379455,
      "grad_norm": 0.1657634824514389,
      "learning_rate": 0.00030473548348441845,
      "loss": 0.4939,
      "num_input_tokens_seen": 16548424,
      "step": 25375
    },
    {
      "epoch": 13.30188679245283,
      "grad_norm": 0.14078325033187866,
      "learning_rate": 0.000304524978008117,
      "loss": 0.426,
      "num_input_tokens_seen": 16551592,
      "step": 25380
    },
    {
      "epoch": 13.304507337526205,
      "grad_norm": 0.099464051425457,
      "learning_rate": 0.0003043145134229599,
      "loss": 0.3387,
      "num_input_tokens_seen": 16554568,
      "step": 25385
    },
    {
      "epoch": 13.30712788259958,
      "grad_norm": 0.09267593920230865,
      "learning_rate": 0.0003041040897729742,
      "loss": 0.4126,
      "num_input_tokens_seen": 16557416,
      "step": 25390
    },
    {
      "epoch": 13.309748427672956,
      "grad_norm": 0.16778349876403809,
      "learning_rate": 0.0003038937071021778,
      "loss": 0.4341,
      "num_input_tokens_seen": 16560424,
      "step": 25395
    },
    {
      "epoch": 13.31236897274633,
      "grad_norm": 0.06250389665365219,
      "learning_rate": 0.0003036833654545808,
      "loss": 0.3527,
      "num_input_tokens_seen": 16563880,
      "step": 25400
    },
    {
      "epoch": 13.314989517819706,
      "grad_norm": 0.16230624914169312,
      "learning_rate": 0.0003034730648741841,
      "loss": 0.3698,
      "num_input_tokens_seen": 16567432,
      "step": 25405
    },
    {
      "epoch": 13.317610062893081,
      "grad_norm": 0.1412886679172516,
      "learning_rate": 0.00030326280540497996,
      "loss": 0.5441,
      "num_input_tokens_seen": 16570888,
      "step": 25410
    },
    {
      "epoch": 13.320230607966456,
      "grad_norm": 0.2348252832889557,
      "learning_rate": 0.0003030525870909525,
      "loss": 0.4149,
      "num_input_tokens_seen": 16573384,
      "step": 25415
    },
    {
      "epoch": 13.322851153039831,
      "grad_norm": 0.14229246973991394,
      "learning_rate": 0.0003028424099760768,
      "loss": 0.4701,
      "num_input_tokens_seen": 16576584,
      "step": 25420
    },
    {
      "epoch": 13.325471698113208,
      "grad_norm": 0.16120603680610657,
      "learning_rate": 0.0003026322741043198,
      "loss": 0.4538,
      "num_input_tokens_seen": 16580584,
      "step": 25425
    },
    {
      "epoch": 13.328092243186584,
      "grad_norm": 0.15785425901412964,
      "learning_rate": 0.0003024221795196393,
      "loss": 0.5105,
      "num_input_tokens_seen": 16583848,
      "step": 25430
    },
    {
      "epoch": 13.330712788259959,
      "grad_norm": 0.210694819688797,
      "learning_rate": 0.0003022121262659849,
      "loss": 0.3254,
      "num_input_tokens_seen": 16587272,
      "step": 25435
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 0.1497471183538437,
      "learning_rate": 0.00030200211438729705,
      "loss": 0.4169,
      "num_input_tokens_seen": 16590088,
      "step": 25440
    },
    {
      "epoch": 13.335953878406709,
      "grad_norm": 0.25453370809555054,
      "learning_rate": 0.000301792143927508,
      "loss": 0.4165,
      "num_input_tokens_seen": 16593320,
      "step": 25445
    },
    {
      "epoch": 13.338574423480084,
      "grad_norm": 0.12974552810192108,
      "learning_rate": 0.0003015822149305411,
      "loss": 0.4449,
      "num_input_tokens_seen": 16596328,
      "step": 25450
    },
    {
      "epoch": 13.34119496855346,
      "grad_norm": 0.19734610617160797,
      "learning_rate": 0.0003013723274403113,
      "loss": 0.5919,
      "num_input_tokens_seen": 16599080,
      "step": 25455
    },
    {
      "epoch": 13.343815513626835,
      "grad_norm": 0.12821844220161438,
      "learning_rate": 0.00030116248150072444,
      "loss": 0.3897,
      "num_input_tokens_seen": 16602056,
      "step": 25460
    },
    {
      "epoch": 13.34643605870021,
      "grad_norm": 0.09711803495883942,
      "learning_rate": 0.00030095267715567816,
      "loss": 0.4653,
      "num_input_tokens_seen": 16606216,
      "step": 25465
    },
    {
      "epoch": 13.349056603773585,
      "grad_norm": 0.12395069003105164,
      "learning_rate": 0.00030074291444906085,
      "loss": 0.3857,
      "num_input_tokens_seen": 16609064,
      "step": 25470
    },
    {
      "epoch": 13.35167714884696,
      "grad_norm": 0.18960678577423096,
      "learning_rate": 0.00030053319342475247,
      "loss": 0.5212,
      "num_input_tokens_seen": 16611560,
      "step": 25475
    },
    {
      "epoch": 13.354297693920335,
      "grad_norm": 0.0918043702840805,
      "learning_rate": 0.00030032351412662465,
      "loss": 0.3568,
      "num_input_tokens_seen": 16614408,
      "step": 25480
    },
    {
      "epoch": 13.35691823899371,
      "grad_norm": 0.13447628915309906,
      "learning_rate": 0.0003001138765985394,
      "loss": 0.4103,
      "num_input_tokens_seen": 16617256,
      "step": 25485
    },
    {
      "epoch": 13.359538784067086,
      "grad_norm": 0.12547989189624786,
      "learning_rate": 0.00029990428088435097,
      "loss": 0.4433,
      "num_input_tokens_seen": 16620232,
      "step": 25490
    },
    {
      "epoch": 13.36215932914046,
      "grad_norm": 0.13411524891853333,
      "learning_rate": 0.000299694727027904,
      "loss": 0.4443,
      "num_input_tokens_seen": 16623912,
      "step": 25495
    },
    {
      "epoch": 13.364779874213836,
      "grad_norm": 0.2687835693359375,
      "learning_rate": 0.00029948521507303495,
      "loss": 0.5486,
      "num_input_tokens_seen": 16626440,
      "step": 25500
    },
    {
      "epoch": 13.367400419287211,
      "grad_norm": 0.12513284385204315,
      "learning_rate": 0.0002992757450635714,
      "loss": 0.5887,
      "num_input_tokens_seen": 16630440,
      "step": 25505
    },
    {
      "epoch": 13.370020964360586,
      "grad_norm": 0.13184590637683868,
      "learning_rate": 0.00029906631704333187,
      "loss": 0.4174,
      "num_input_tokens_seen": 16633352,
      "step": 25510
    },
    {
      "epoch": 13.372641509433961,
      "grad_norm": 0.23368576169013977,
      "learning_rate": 0.0002988569310561264,
      "loss": 0.4742,
      "num_input_tokens_seen": 16636392,
      "step": 25515
    },
    {
      "epoch": 13.375262054507338,
      "grad_norm": 0.17901204526424408,
      "learning_rate": 0.0002986475871457566,
      "loss": 0.42,
      "num_input_tokens_seen": 16639688,
      "step": 25520
    },
    {
      "epoch": 13.377882599580714,
      "grad_norm": 0.11298874020576477,
      "learning_rate": 0.000298438285356014,
      "loss": 0.3853,
      "num_input_tokens_seen": 16643176,
      "step": 25525
    },
    {
      "epoch": 13.380503144654089,
      "grad_norm": 0.13335402309894562,
      "learning_rate": 0.00029822902573068257,
      "loss": 0.4273,
      "num_input_tokens_seen": 16646952,
      "step": 25530
    },
    {
      "epoch": 13.383123689727464,
      "grad_norm": 0.14486414194107056,
      "learning_rate": 0.0002980198083135372,
      "loss": 0.4151,
      "num_input_tokens_seen": 16650792,
      "step": 25535
    },
    {
      "epoch": 13.385744234800839,
      "grad_norm": 0.16092312335968018,
      "learning_rate": 0.0002978106331483435,
      "loss": 0.3828,
      "num_input_tokens_seen": 16654568,
      "step": 25540
    },
    {
      "epoch": 13.388364779874214,
      "grad_norm": 0.14778928458690643,
      "learning_rate": 0.00029760150027885874,
      "loss": 0.5038,
      "num_input_tokens_seen": 16657608,
      "step": 25545
    },
    {
      "epoch": 13.39098532494759,
      "grad_norm": 0.2218049019575119,
      "learning_rate": 0.0002973924097488311,
      "loss": 0.4382,
      "num_input_tokens_seen": 16660360,
      "step": 25550
    },
    {
      "epoch": 13.393605870020965,
      "grad_norm": 0.10929979383945465,
      "learning_rate": 0.00029718336160199977,
      "loss": 0.3509,
      "num_input_tokens_seen": 16663048,
      "step": 25555
    },
    {
      "epoch": 13.39622641509434,
      "grad_norm": 0.15741942822933197,
      "learning_rate": 0.00029697435588209537,
      "loss": 0.3869,
      "num_input_tokens_seen": 16667080,
      "step": 25560
    },
    {
      "epoch": 13.398846960167715,
      "grad_norm": 0.16188998520374298,
      "learning_rate": 0.00029676539263283943,
      "loss": 0.4187,
      "num_input_tokens_seen": 16669608,
      "step": 25565
    },
    {
      "epoch": 13.40146750524109,
      "grad_norm": 0.07974326610565186,
      "learning_rate": 0.0002965564718979448,
      "loss": 0.4076,
      "num_input_tokens_seen": 16676968,
      "step": 25570
    },
    {
      "epoch": 13.404088050314465,
      "grad_norm": 0.10566426813602448,
      "learning_rate": 0.0002963475937211151,
      "loss": 0.4364,
      "num_input_tokens_seen": 16680552,
      "step": 25575
    },
    {
      "epoch": 13.40670859538784,
      "grad_norm": 0.1488226056098938,
      "learning_rate": 0.0002961387581460456,
      "loss": 0.3872,
      "num_input_tokens_seen": 16683880,
      "step": 25580
    },
    {
      "epoch": 13.409329140461216,
      "grad_norm": 0.1487998515367508,
      "learning_rate": 0.0002959299652164221,
      "loss": 0.5126,
      "num_input_tokens_seen": 16686312,
      "step": 25585
    },
    {
      "epoch": 13.41194968553459,
      "grad_norm": 0.16151447594165802,
      "learning_rate": 0.00029572121497592154,
      "loss": 0.4318,
      "num_input_tokens_seen": 16688776,
      "step": 25590
    },
    {
      "epoch": 13.414570230607966,
      "grad_norm": 0.14799517393112183,
      "learning_rate": 0.00029551250746821236,
      "loss": 0.3188,
      "num_input_tokens_seen": 16692104,
      "step": 25595
    },
    {
      "epoch": 13.417190775681341,
      "grad_norm": 0.159258171916008,
      "learning_rate": 0.0002953038427369537,
      "loss": 0.4166,
      "num_input_tokens_seen": 16694600,
      "step": 25600
    },
    {
      "epoch": 13.419811320754716,
      "grad_norm": 0.13248135149478912,
      "learning_rate": 0.000295095220825796,
      "loss": 0.4028,
      "num_input_tokens_seen": 16697384,
      "step": 25605
    },
    {
      "epoch": 13.422431865828091,
      "grad_norm": 0.1454167515039444,
      "learning_rate": 0.00029488664177838023,
      "loss": 0.4581,
      "num_input_tokens_seen": 16700776,
      "step": 25610
    },
    {
      "epoch": 13.425052410901468,
      "grad_norm": 0.12982752919197083,
      "learning_rate": 0.0002946781056383391,
      "loss": 0.4027,
      "num_input_tokens_seen": 16703272,
      "step": 25615
    },
    {
      "epoch": 13.427672955974844,
      "grad_norm": 0.07844717800617218,
      "learning_rate": 0.0002944696124492956,
      "loss": 0.3359,
      "num_input_tokens_seen": 16706440,
      "step": 25620
    },
    {
      "epoch": 13.430293501048219,
      "grad_norm": 0.13783031702041626,
      "learning_rate": 0.0002942611622548645,
      "loss": 0.4364,
      "num_input_tokens_seen": 16709256,
      "step": 25625
    },
    {
      "epoch": 13.432914046121594,
      "grad_norm": 0.11673683673143387,
      "learning_rate": 0.00029405275509865103,
      "loss": 0.3986,
      "num_input_tokens_seen": 16712712,
      "step": 25630
    },
    {
      "epoch": 13.435534591194969,
      "grad_norm": 0.12164591997861862,
      "learning_rate": 0.00029384439102425174,
      "loss": 0.4233,
      "num_input_tokens_seen": 16715464,
      "step": 25635
    },
    {
      "epoch": 13.438155136268344,
      "grad_norm": 0.18691201508045197,
      "learning_rate": 0.0002936360700752539,
      "loss": 0.6294,
      "num_input_tokens_seen": 16719304,
      "step": 25640
    },
    {
      "epoch": 13.44077568134172,
      "grad_norm": 0.11409438401460648,
      "learning_rate": 0.00029342779229523573,
      "loss": 0.4649,
      "num_input_tokens_seen": 16722760,
      "step": 25645
    },
    {
      "epoch": 13.443396226415095,
      "grad_norm": 0.14219354093074799,
      "learning_rate": 0.00029321955772776675,
      "loss": 0.4898,
      "num_input_tokens_seen": 16726152,
      "step": 25650
    },
    {
      "epoch": 13.44601677148847,
      "grad_norm": 0.22623379528522491,
      "learning_rate": 0.00029301136641640723,
      "loss": 0.4097,
      "num_input_tokens_seen": 16728808,
      "step": 25655
    },
    {
      "epoch": 13.448637316561845,
      "grad_norm": 0.20352056622505188,
      "learning_rate": 0.0002928032184047085,
      "loss": 0.4115,
      "num_input_tokens_seen": 16731240,
      "step": 25660
    },
    {
      "epoch": 13.45125786163522,
      "grad_norm": 0.09636776149272919,
      "learning_rate": 0.0002925951137362126,
      "loss": 0.5604,
      "num_input_tokens_seen": 16735112,
      "step": 25665
    },
    {
      "epoch": 13.453878406708595,
      "grad_norm": 0.11059249192476273,
      "learning_rate": 0.00029238705245445264,
      "loss": 0.4221,
      "num_input_tokens_seen": 16738568,
      "step": 25670
    },
    {
      "epoch": 13.45649895178197,
      "grad_norm": 0.21056069433689117,
      "learning_rate": 0.0002921790346029527,
      "loss": 0.4175,
      "num_input_tokens_seen": 16742056,
      "step": 25675
    },
    {
      "epoch": 13.459119496855346,
      "grad_norm": 0.16950304806232452,
      "learning_rate": 0.0002919710602252279,
      "loss": 0.4121,
      "num_input_tokens_seen": 16744616,
      "step": 25680
    },
    {
      "epoch": 13.46174004192872,
      "grad_norm": 0.20651927590370178,
      "learning_rate": 0.0002917631293647838,
      "loss": 0.5032,
      "num_input_tokens_seen": 16747464,
      "step": 25685
    },
    {
      "epoch": 13.464360587002096,
      "grad_norm": 0.06983882933855057,
      "learning_rate": 0.0002915552420651176,
      "loss": 0.4338,
      "num_input_tokens_seen": 16752104,
      "step": 25690
    },
    {
      "epoch": 13.466981132075471,
      "grad_norm": 0.06284525990486145,
      "learning_rate": 0.00029134739836971657,
      "loss": 0.3598,
      "num_input_tokens_seen": 16755720,
      "step": 25695
    },
    {
      "epoch": 13.469601677148846,
      "grad_norm": 0.1282622367143631,
      "learning_rate": 0.0002911395983220595,
      "loss": 0.5278,
      "num_input_tokens_seen": 16758600,
      "step": 25700
    },
    {
      "epoch": 13.472222222222221,
      "grad_norm": 0.197422593832016,
      "learning_rate": 0.0002909318419656154,
      "loss": 0.4056,
      "num_input_tokens_seen": 16761352,
      "step": 25705
    },
    {
      "epoch": 13.474842767295598,
      "grad_norm": 0.0782923698425293,
      "learning_rate": 0.00029072412934384484,
      "loss": 0.5266,
      "num_input_tokens_seen": 16765512,
      "step": 25710
    },
    {
      "epoch": 13.477463312368974,
      "grad_norm": 0.1923815757036209,
      "learning_rate": 0.00029051646050019884,
      "loss": 0.5572,
      "num_input_tokens_seen": 16769064,
      "step": 25715
    },
    {
      "epoch": 13.480083857442349,
      "grad_norm": 0.11910383403301239,
      "learning_rate": 0.00029030883547811966,
      "loss": 0.5136,
      "num_input_tokens_seen": 16772232,
      "step": 25720
    },
    {
      "epoch": 13.482704402515724,
      "grad_norm": 0.12106817960739136,
      "learning_rate": 0.0002901012543210397,
      "loss": 0.3606,
      "num_input_tokens_seen": 16776616,
      "step": 25725
    },
    {
      "epoch": 13.485324947589099,
      "grad_norm": 0.14950503408908844,
      "learning_rate": 0.00028989371707238264,
      "loss": 0.4349,
      "num_input_tokens_seen": 16779496,
      "step": 25730
    },
    {
      "epoch": 13.487945492662474,
      "grad_norm": 0.1225573718547821,
      "learning_rate": 0.00028968622377556285,
      "loss": 0.3513,
      "num_input_tokens_seen": 16783368,
      "step": 25735
    },
    {
      "epoch": 13.49056603773585,
      "grad_norm": 0.18331365287303925,
      "learning_rate": 0.00028947877447398567,
      "loss": 0.4589,
      "num_input_tokens_seen": 16785928,
      "step": 25740
    },
    {
      "epoch": 13.493186582809225,
      "grad_norm": 0.2123282551765442,
      "learning_rate": 0.00028927136921104733,
      "loss": 0.4017,
      "num_input_tokens_seen": 16789320,
      "step": 25745
    },
    {
      "epoch": 13.4958071278826,
      "grad_norm": 0.14156650006771088,
      "learning_rate": 0.00028906400803013446,
      "loss": 0.4961,
      "num_input_tokens_seen": 16793928,
      "step": 25750
    },
    {
      "epoch": 13.498427672955975,
      "grad_norm": 0.208187535405159,
      "learning_rate": 0.00028885669097462435,
      "loss": 0.4943,
      "num_input_tokens_seen": 16796872,
      "step": 25755
    },
    {
      "epoch": 13.50104821802935,
      "grad_norm": 0.11986960470676422,
      "learning_rate": 0.0002886494180878857,
      "loss": 0.3778,
      "num_input_tokens_seen": 16800552,
      "step": 25760
    },
    {
      "epoch": 13.503668763102725,
      "grad_norm": 0.1375109851360321,
      "learning_rate": 0.00028844218941327757,
      "loss": 0.413,
      "num_input_tokens_seen": 16803880,
      "step": 25765
    },
    {
      "epoch": 13.5062893081761,
      "grad_norm": 0.15807977318763733,
      "learning_rate": 0.0002882350049941498,
      "loss": 0.4172,
      "num_input_tokens_seen": 16806600,
      "step": 25770
    },
    {
      "epoch": 13.508909853249476,
      "grad_norm": 0.10062047094106674,
      "learning_rate": 0.0002880278648738432,
      "loss": 0.415,
      "num_input_tokens_seen": 16809576,
      "step": 25775
    },
    {
      "epoch": 13.51153039832285,
      "grad_norm": 0.11071797460317612,
      "learning_rate": 0.0002878207690956891,
      "loss": 0.4902,
      "num_input_tokens_seen": 16812776,
      "step": 25780
    },
    {
      "epoch": 13.514150943396226,
      "grad_norm": 0.18715821206569672,
      "learning_rate": 0.00028761371770300915,
      "loss": 0.3961,
      "num_input_tokens_seen": 16815496,
      "step": 25785
    },
    {
      "epoch": 13.516771488469601,
      "grad_norm": 0.2138691246509552,
      "learning_rate": 0.0002874067107391164,
      "loss": 0.4292,
      "num_input_tokens_seen": 16818984,
      "step": 25790
    },
    {
      "epoch": 13.519392033542976,
      "grad_norm": 0.10949201136827469,
      "learning_rate": 0.00028719974824731445,
      "loss": 0.3117,
      "num_input_tokens_seen": 16825096,
      "step": 25795
    },
    {
      "epoch": 13.522012578616351,
      "grad_norm": 0.22954989969730377,
      "learning_rate": 0.0002869928302708975,
      "loss": 0.5012,
      "num_input_tokens_seen": 16827880,
      "step": 25800
    },
    {
      "epoch": 13.524633123689728,
      "grad_norm": 0.15302135050296783,
      "learning_rate": 0.00028678595685315044,
      "loss": 0.5046,
      "num_input_tokens_seen": 16830504,
      "step": 25805
    },
    {
      "epoch": 13.527253668763104,
      "grad_norm": 0.18238604068756104,
      "learning_rate": 0.00028657912803734854,
      "loss": 0.3811,
      "num_input_tokens_seen": 16833128,
      "step": 25810
    },
    {
      "epoch": 13.529874213836479,
      "grad_norm": 0.10377337038516998,
      "learning_rate": 0.00028637234386675816,
      "loss": 0.4478,
      "num_input_tokens_seen": 16835592,
      "step": 25815
    },
    {
      "epoch": 13.532494758909854,
      "grad_norm": 0.21173186600208282,
      "learning_rate": 0.00028616560438463626,
      "loss": 0.5281,
      "num_input_tokens_seen": 16839752,
      "step": 25820
    },
    {
      "epoch": 13.535115303983229,
      "grad_norm": 0.11890047788619995,
      "learning_rate": 0.00028595890963423057,
      "loss": 0.4338,
      "num_input_tokens_seen": 16842568,
      "step": 25825
    },
    {
      "epoch": 13.537735849056604,
      "grad_norm": 0.223733589053154,
      "learning_rate": 0.0002857522596587789,
      "loss": 0.4221,
      "num_input_tokens_seen": 16845320,
      "step": 25830
    },
    {
      "epoch": 13.54035639412998,
      "grad_norm": 0.1472567915916443,
      "learning_rate": 0.00028554565450151045,
      "loss": 0.4213,
      "num_input_tokens_seen": 16848552,
      "step": 25835
    },
    {
      "epoch": 13.542976939203355,
      "grad_norm": 0.13554733991622925,
      "learning_rate": 0.00028533909420564417,
      "loss": 0.5695,
      "num_input_tokens_seen": 16853608,
      "step": 25840
    },
    {
      "epoch": 13.54559748427673,
      "grad_norm": 0.12007927894592285,
      "learning_rate": 0.00028513257881439047,
      "loss": 0.4006,
      "num_input_tokens_seen": 16856648,
      "step": 25845
    },
    {
      "epoch": 13.548218029350105,
      "grad_norm": 0.07565172016620636,
      "learning_rate": 0.00028492610837095003,
      "loss": 0.4623,
      "num_input_tokens_seen": 16861864,
      "step": 25850
    },
    {
      "epoch": 13.55083857442348,
      "grad_norm": 0.1726912260055542,
      "learning_rate": 0.00028471968291851407,
      "loss": 0.4228,
      "num_input_tokens_seen": 16865256,
      "step": 25855
    },
    {
      "epoch": 13.553459119496855,
      "grad_norm": 0.11843115091323853,
      "learning_rate": 0.0002845133025002645,
      "loss": 0.4133,
      "num_input_tokens_seen": 16868936,
      "step": 25860
    },
    {
      "epoch": 13.55607966457023,
      "grad_norm": 0.19408275187015533,
      "learning_rate": 0.0002843069671593734,
      "loss": 0.4377,
      "num_input_tokens_seen": 16872200,
      "step": 25865
    },
    {
      "epoch": 13.558700209643606,
      "grad_norm": 0.09508877992630005,
      "learning_rate": 0.000284100676939004,
      "loss": 0.4109,
      "num_input_tokens_seen": 16878824,
      "step": 25870
    },
    {
      "epoch": 13.56132075471698,
      "grad_norm": 0.185503751039505,
      "learning_rate": 0.0002838944318823099,
      "loss": 0.3637,
      "num_input_tokens_seen": 16883080,
      "step": 25875
    },
    {
      "epoch": 13.563941299790356,
      "grad_norm": 0.09682897478342056,
      "learning_rate": 0.0002836882320324354,
      "loss": 0.5524,
      "num_input_tokens_seen": 16887336,
      "step": 25880
    },
    {
      "epoch": 13.566561844863731,
      "grad_norm": 0.12668119370937347,
      "learning_rate": 0.00028348207743251477,
      "loss": 0.4727,
      "num_input_tokens_seen": 16890760,
      "step": 25885
    },
    {
      "epoch": 13.569182389937106,
      "grad_norm": 0.0790037289261818,
      "learning_rate": 0.0002832759681256735,
      "loss": 0.3744,
      "num_input_tokens_seen": 16893928,
      "step": 25890
    },
    {
      "epoch": 13.571802935010481,
      "grad_norm": 0.17710165679454803,
      "learning_rate": 0.000283069904155027,
      "loss": 0.3282,
      "num_input_tokens_seen": 16897032,
      "step": 25895
    },
    {
      "epoch": 13.574423480083858,
      "grad_norm": 0.19249504804611206,
      "learning_rate": 0.0002828638855636817,
      "loss": 0.3813,
      "num_input_tokens_seen": 16900648,
      "step": 25900
    },
    {
      "epoch": 13.577044025157234,
      "grad_norm": 0.19631537795066833,
      "learning_rate": 0.0002826579123947345,
      "loss": 0.538,
      "num_input_tokens_seen": 16903880,
      "step": 25905
    },
    {
      "epoch": 13.579664570230609,
      "grad_norm": 0.13369561731815338,
      "learning_rate": 0.0002824519846912723,
      "loss": 0.5017,
      "num_input_tokens_seen": 16907816,
      "step": 25910
    },
    {
      "epoch": 13.582285115303984,
      "grad_norm": 0.10306745767593384,
      "learning_rate": 0.00028224610249637293,
      "loss": 0.3594,
      "num_input_tokens_seen": 16910664,
      "step": 25915
    },
    {
      "epoch": 13.584905660377359,
      "grad_norm": 0.10085888206958771,
      "learning_rate": 0.00028204026585310483,
      "loss": 0.4773,
      "num_input_tokens_seen": 16914408,
      "step": 25920
    },
    {
      "epoch": 13.587526205450734,
      "grad_norm": 0.15773935616016388,
      "learning_rate": 0.0002818344748045264,
      "loss": 0.5621,
      "num_input_tokens_seen": 16918120,
      "step": 25925
    },
    {
      "epoch": 13.59014675052411,
      "grad_norm": 0.1637849360704422,
      "learning_rate": 0.0002816287293936868,
      "loss": 0.509,
      "num_input_tokens_seen": 16920680,
      "step": 25930
    },
    {
      "epoch": 13.592767295597485,
      "grad_norm": 0.0917583778500557,
      "learning_rate": 0.0002814230296636259,
      "loss": 0.4016,
      "num_input_tokens_seen": 16923592,
      "step": 25935
    },
    {
      "epoch": 13.59538784067086,
      "grad_norm": 0.12784601747989655,
      "learning_rate": 0.0002812173756573734,
      "loss": 0.3705,
      "num_input_tokens_seen": 16926920,
      "step": 25940
    },
    {
      "epoch": 13.598008385744235,
      "grad_norm": 0.1190318912267685,
      "learning_rate": 0.00028101176741794984,
      "loss": 0.4637,
      "num_input_tokens_seen": 16930056,
      "step": 25945
    },
    {
      "epoch": 13.60062893081761,
      "grad_norm": 0.24935072660446167,
      "learning_rate": 0.0002808062049883664,
      "loss": 0.4466,
      "num_input_tokens_seen": 16932424,
      "step": 25950
    },
    {
      "epoch": 13.603249475890985,
      "grad_norm": 0.12898705899715424,
      "learning_rate": 0.00028060068841162403,
      "loss": 0.368,
      "num_input_tokens_seen": 16936200,
      "step": 25955
    },
    {
      "epoch": 13.60587002096436,
      "grad_norm": 0.11502335965633392,
      "learning_rate": 0.00028039521773071477,
      "loss": 0.3936,
      "num_input_tokens_seen": 16939176,
      "step": 25960
    },
    {
      "epoch": 13.608490566037736,
      "grad_norm": 0.07343430817127228,
      "learning_rate": 0.0002801897929886204,
      "loss": 0.398,
      "num_input_tokens_seen": 16943048,
      "step": 25965
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 0.10235029458999634,
      "learning_rate": 0.0002799844142283135,
      "loss": 0.4201,
      "num_input_tokens_seen": 16946056,
      "step": 25970
    },
    {
      "epoch": 13.613731656184486,
      "grad_norm": 0.1377757042646408,
      "learning_rate": 0.00027977908149275733,
      "loss": 0.4705,
      "num_input_tokens_seen": 16948712,
      "step": 25975
    },
    {
      "epoch": 13.616352201257861,
      "grad_norm": 0.18456825613975525,
      "learning_rate": 0.00027957379482490476,
      "loss": 0.4597,
      "num_input_tokens_seen": 16952168,
      "step": 25980
    },
    {
      "epoch": 13.618972746331236,
      "grad_norm": 0.1310276836156845,
      "learning_rate": 0.0002793685542676997,
      "loss": 0.4165,
      "num_input_tokens_seen": 16955496,
      "step": 25985
    },
    {
      "epoch": 13.621593291404611,
      "grad_norm": 0.12846185266971588,
      "learning_rate": 0.0002791633598640758,
      "loss": 0.4514,
      "num_input_tokens_seen": 16958472,
      "step": 25990
    },
    {
      "epoch": 13.624213836477988,
      "grad_norm": 0.06551840156316757,
      "learning_rate": 0.0002789582116569576,
      "loss": 0.3884,
      "num_input_tokens_seen": 16961832,
      "step": 25995
    },
    {
      "epoch": 13.626834381551364,
      "grad_norm": 0.19271622598171234,
      "learning_rate": 0.0002787531096892597,
      "loss": 0.352,
      "num_input_tokens_seen": 16964552,
      "step": 26000
    },
    {
      "epoch": 13.629454926624739,
      "grad_norm": 0.10787578672170639,
      "learning_rate": 0.0002785480540038874,
      "loss": 0.3903,
      "num_input_tokens_seen": 16967688,
      "step": 26005
    },
    {
      "epoch": 13.632075471698114,
      "grad_norm": 0.16356605291366577,
      "learning_rate": 0.0002783430446437355,
      "loss": 0.456,
      "num_input_tokens_seen": 16971176,
      "step": 26010
    },
    {
      "epoch": 13.634696016771489,
      "grad_norm": 0.12444248795509338,
      "learning_rate": 0.00027813808165169006,
      "loss": 0.3898,
      "num_input_tokens_seen": 16974568,
      "step": 26015
    },
    {
      "epoch": 13.637316561844864,
      "grad_norm": 0.17844919860363007,
      "learning_rate": 0.00027793316507062667,
      "loss": 0.4794,
      "num_input_tokens_seen": 16978728,
      "step": 26020
    },
    {
      "epoch": 13.63993710691824,
      "grad_norm": 0.08201441913843155,
      "learning_rate": 0.00027772829494341163,
      "loss": 0.3134,
      "num_input_tokens_seen": 16981928,
      "step": 26025
    },
    {
      "epoch": 13.642557651991615,
      "grad_norm": 0.17048583924770355,
      "learning_rate": 0.00027752347131290147,
      "loss": 0.4302,
      "num_input_tokens_seen": 16984104,
      "step": 26030
    },
    {
      "epoch": 13.64517819706499,
      "grad_norm": 0.13757699728012085,
      "learning_rate": 0.0002773186942219431,
      "loss": 0.4544,
      "num_input_tokens_seen": 16986952,
      "step": 26035
    },
    {
      "epoch": 13.647798742138365,
      "grad_norm": 0.08344630151987076,
      "learning_rate": 0.0002771139637133735,
      "loss": 0.3274,
      "num_input_tokens_seen": 16993160,
      "step": 26040
    },
    {
      "epoch": 13.65041928721174,
      "grad_norm": 0.11125148087739944,
      "learning_rate": 0.00027690927983001966,
      "loss": 0.4322,
      "num_input_tokens_seen": 16995560,
      "step": 26045
    },
    {
      "epoch": 13.653039832285115,
      "grad_norm": 0.15770123898983002,
      "learning_rate": 0.00027670464261469937,
      "loss": 0.4747,
      "num_input_tokens_seen": 16999016,
      "step": 26050
    },
    {
      "epoch": 13.65566037735849,
      "grad_norm": 0.09232719242572784,
      "learning_rate": 0.0002765000521102203,
      "loss": 0.3408,
      "num_input_tokens_seen": 17001960,
      "step": 26055
    },
    {
      "epoch": 13.658280922431866,
      "grad_norm": 0.1640058159828186,
      "learning_rate": 0.0002762955083593807,
      "loss": 0.5073,
      "num_input_tokens_seen": 17004936,
      "step": 26060
    },
    {
      "epoch": 13.66090146750524,
      "grad_norm": 0.13161572813987732,
      "learning_rate": 0.0002760910114049686,
      "loss": 0.3646,
      "num_input_tokens_seen": 17008584,
      "step": 26065
    },
    {
      "epoch": 13.663522012578616,
      "grad_norm": 0.2285350263118744,
      "learning_rate": 0.0002758865612897623,
      "loss": 0.463,
      "num_input_tokens_seen": 17011048,
      "step": 26070
    },
    {
      "epoch": 13.666142557651991,
      "grad_norm": 0.1581227034330368,
      "learning_rate": 0.00027568215805653045,
      "loss": 0.3482,
      "num_input_tokens_seen": 17013384,
      "step": 26075
    },
    {
      "epoch": 13.668763102725366,
      "grad_norm": 0.16501012444496155,
      "learning_rate": 0.00027547780174803205,
      "loss": 0.4047,
      "num_input_tokens_seen": 17016584,
      "step": 26080
    },
    {
      "epoch": 13.671383647798741,
      "grad_norm": 0.14281414449214935,
      "learning_rate": 0.0002752734924070161,
      "loss": 0.4789,
      "num_input_tokens_seen": 17019368,
      "step": 26085
    },
    {
      "epoch": 13.674004192872118,
      "grad_norm": 0.12635833024978638,
      "learning_rate": 0.00027506923007622177,
      "loss": 0.3794,
      "num_input_tokens_seen": 17022184,
      "step": 26090
    },
    {
      "epoch": 13.676624737945493,
      "grad_norm": 0.10268573462963104,
      "learning_rate": 0.00027486501479837846,
      "loss": 0.4117,
      "num_input_tokens_seen": 17025288,
      "step": 26095
    },
    {
      "epoch": 13.679245283018869,
      "grad_norm": 0.1379212886095047,
      "learning_rate": 0.0002746608466162053,
      "loss": 0.3838,
      "num_input_tokens_seen": 17028424,
      "step": 26100
    },
    {
      "epoch": 13.681865828092244,
      "grad_norm": 0.12980087101459503,
      "learning_rate": 0.0002744567255724123,
      "loss": 0.3276,
      "num_input_tokens_seen": 17031720,
      "step": 26105
    },
    {
      "epoch": 13.684486373165619,
      "grad_norm": 0.10167760401964188,
      "learning_rate": 0.0002742526517096992,
      "loss": 0.5021,
      "num_input_tokens_seen": 17034888,
      "step": 26110
    },
    {
      "epoch": 13.687106918238994,
      "grad_norm": 0.13985180854797363,
      "learning_rate": 0.000274048625070756,
      "loss": 0.4406,
      "num_input_tokens_seen": 17041576,
      "step": 26115
    },
    {
      "epoch": 13.68972746331237,
      "grad_norm": 0.14980703592300415,
      "learning_rate": 0.0002738446456982627,
      "loss": 0.4191,
      "num_input_tokens_seen": 17044424,
      "step": 26120
    },
    {
      "epoch": 13.692348008385745,
      "grad_norm": 0.08688553422689438,
      "learning_rate": 0.0002736407136348892,
      "loss": 0.3577,
      "num_input_tokens_seen": 17047976,
      "step": 26125
    },
    {
      "epoch": 13.69496855345912,
      "grad_norm": 0.10317174345254898,
      "learning_rate": 0.0002734368289232959,
      "loss": 0.5922,
      "num_input_tokens_seen": 17051208,
      "step": 26130
    },
    {
      "epoch": 13.697589098532495,
      "grad_norm": 0.13116323947906494,
      "learning_rate": 0.00027323299160613335,
      "loss": 0.3889,
      "num_input_tokens_seen": 17053640,
      "step": 26135
    },
    {
      "epoch": 13.70020964360587,
      "grad_norm": 0.2521916329860687,
      "learning_rate": 0.00027302920172604173,
      "loss": 0.3653,
      "num_input_tokens_seen": 17056968,
      "step": 26140
    },
    {
      "epoch": 13.702830188679245,
      "grad_norm": 0.271331787109375,
      "learning_rate": 0.00027282545932565214,
      "loss": 0.4593,
      "num_input_tokens_seen": 17059688,
      "step": 26145
    },
    {
      "epoch": 13.70545073375262,
      "grad_norm": 0.2059740275144577,
      "learning_rate": 0.00027262176444758435,
      "loss": 0.4668,
      "num_input_tokens_seen": 17064712,
      "step": 26150
    },
    {
      "epoch": 13.708071278825996,
      "grad_norm": 0.13177216053009033,
      "learning_rate": 0.00027241811713444935,
      "loss": 0.3835,
      "num_input_tokens_seen": 17067528,
      "step": 26155
    },
    {
      "epoch": 13.71069182389937,
      "grad_norm": 0.11327420175075531,
      "learning_rate": 0.0002722145174288478,
      "loss": 0.3736,
      "num_input_tokens_seen": 17070952,
      "step": 26160
    },
    {
      "epoch": 13.713312368972746,
      "grad_norm": 0.11388333886861801,
      "learning_rate": 0.0002720109653733706,
      "loss": 0.5258,
      "num_input_tokens_seen": 17073992,
      "step": 26165
    },
    {
      "epoch": 13.715932914046121,
      "grad_norm": 0.09662523865699768,
      "learning_rate": 0.00027180746101059856,
      "loss": 0.4275,
      "num_input_tokens_seen": 17078376,
      "step": 26170
    },
    {
      "epoch": 13.718553459119496,
      "grad_norm": 0.17050229012966156,
      "learning_rate": 0.0002716040043831023,
      "loss": 0.4863,
      "num_input_tokens_seen": 17082952,
      "step": 26175
    },
    {
      "epoch": 13.721174004192871,
      "grad_norm": 0.14744959771633148,
      "learning_rate": 0.0002714005955334424,
      "loss": 0.4571,
      "num_input_tokens_seen": 17085640,
      "step": 26180
    },
    {
      "epoch": 13.723794549266248,
      "grad_norm": 0.07840133458375931,
      "learning_rate": 0.00027119723450417,
      "loss": 0.3215,
      "num_input_tokens_seen": 17089096,
      "step": 26185
    },
    {
      "epoch": 13.726415094339622,
      "grad_norm": 0.2603253722190857,
      "learning_rate": 0.0002709939213378258,
      "loss": 0.5921,
      "num_input_tokens_seen": 17091752,
      "step": 26190
    },
    {
      "epoch": 13.729035639412999,
      "grad_norm": 0.08971889317035675,
      "learning_rate": 0.0002707906560769408,
      "loss": 0.4604,
      "num_input_tokens_seen": 17094824,
      "step": 26195
    },
    {
      "epoch": 13.731656184486374,
      "grad_norm": 0.10903657227754593,
      "learning_rate": 0.0002705874387640354,
      "loss": 0.579,
      "num_input_tokens_seen": 17097992,
      "step": 26200
    },
    {
      "epoch": 13.734276729559749,
      "grad_norm": 0.14976944029331207,
      "learning_rate": 0.00027038426944162074,
      "loss": 0.544,
      "num_input_tokens_seen": 17100968,
      "step": 26205
    },
    {
      "epoch": 13.736897274633124,
      "grad_norm": 0.18233239650726318,
      "learning_rate": 0.0002701811481521971,
      "loss": 0.456,
      "num_input_tokens_seen": 17103496,
      "step": 26210
    },
    {
      "epoch": 13.7395178197065,
      "grad_norm": 0.12340754270553589,
      "learning_rate": 0.0002699780749382554,
      "loss": 0.4135,
      "num_input_tokens_seen": 17106920,
      "step": 26215
    },
    {
      "epoch": 13.742138364779874,
      "grad_norm": 0.1493319720029831,
      "learning_rate": 0.0002697750498422761,
      "loss": 0.378,
      "num_input_tokens_seen": 17110472,
      "step": 26220
    },
    {
      "epoch": 13.74475890985325,
      "grad_norm": 0.08962471038103104,
      "learning_rate": 0.0002695720729067301,
      "loss": 0.4785,
      "num_input_tokens_seen": 17113640,
      "step": 26225
    },
    {
      "epoch": 13.747379454926625,
      "grad_norm": 0.12059714645147324,
      "learning_rate": 0.00026936914417407756,
      "loss": 0.5544,
      "num_input_tokens_seen": 17116808,
      "step": 26230
    },
    {
      "epoch": 13.75,
      "grad_norm": 0.09851338714361191,
      "learning_rate": 0.0002691662636867687,
      "loss": 0.4384,
      "num_input_tokens_seen": 17119144,
      "step": 26235
    },
    {
      "epoch": 13.752620545073375,
      "grad_norm": 0.16496439278125763,
      "learning_rate": 0.000268963431487244,
      "loss": 0.3803,
      "num_input_tokens_seen": 17122056,
      "step": 26240
    },
    {
      "epoch": 13.75524109014675,
      "grad_norm": 0.12388157099485397,
      "learning_rate": 0.0002687606476179336,
      "loss": 0.3674,
      "num_input_tokens_seen": 17125352,
      "step": 26245
    },
    {
      "epoch": 13.757861635220126,
      "grad_norm": 0.20163293182849884,
      "learning_rate": 0.0002685579121212578,
      "loss": 0.3969,
      "num_input_tokens_seen": 17128456,
      "step": 26250
    },
    {
      "epoch": 13.7604821802935,
      "grad_norm": 0.1390807330608368,
      "learning_rate": 0.00026835522503962616,
      "loss": 0.3178,
      "num_input_tokens_seen": 17131496,
      "step": 26255
    },
    {
      "epoch": 13.763102725366876,
      "grad_norm": 0.10219433903694153,
      "learning_rate": 0.00026815258641543896,
      "loss": 0.3043,
      "num_input_tokens_seen": 17135528,
      "step": 26260
    },
    {
      "epoch": 13.765723270440251,
      "grad_norm": 0.1322212964296341,
      "learning_rate": 0.0002679499962910853,
      "loss": 0.4483,
      "num_input_tokens_seen": 17138728,
      "step": 26265
    },
    {
      "epoch": 13.768343815513626,
      "grad_norm": 0.13329677283763885,
      "learning_rate": 0.0002677474547089451,
      "loss": 0.4558,
      "num_input_tokens_seen": 17142600,
      "step": 26270
    },
    {
      "epoch": 13.770964360587001,
      "grad_norm": 0.25527873635292053,
      "learning_rate": 0.000267544961711388,
      "loss": 0.4988,
      "num_input_tokens_seen": 17145448,
      "step": 26275
    },
    {
      "epoch": 13.773584905660378,
      "grad_norm": 0.1078430712223053,
      "learning_rate": 0.0002673425173407726,
      "loss": 0.336,
      "num_input_tokens_seen": 17148712,
      "step": 26280
    },
    {
      "epoch": 13.776205450733752,
      "grad_norm": 0.17229005694389343,
      "learning_rate": 0.00026714012163944826,
      "loss": 0.3821,
      "num_input_tokens_seen": 17151144,
      "step": 26285
    },
    {
      "epoch": 13.778825995807129,
      "grad_norm": 0.12412898242473602,
      "learning_rate": 0.0002669377746497541,
      "loss": 0.3646,
      "num_input_tokens_seen": 17153864,
      "step": 26290
    },
    {
      "epoch": 13.781446540880504,
      "grad_norm": 0.175087571144104,
      "learning_rate": 0.0002667354764140184,
      "loss": 0.4495,
      "num_input_tokens_seen": 17156968,
      "step": 26295
    },
    {
      "epoch": 13.784067085953879,
      "grad_norm": 0.15582311153411865,
      "learning_rate": 0.00026653322697455987,
      "loss": 0.4992,
      "num_input_tokens_seen": 17160520,
      "step": 26300
    },
    {
      "epoch": 13.786687631027254,
      "grad_norm": 0.13973087072372437,
      "learning_rate": 0.00026633102637368684,
      "loss": 0.3798,
      "num_input_tokens_seen": 17163368,
      "step": 26305
    },
    {
      "epoch": 13.78930817610063,
      "grad_norm": 0.15143144130706787,
      "learning_rate": 0.00026612887465369704,
      "loss": 0.5994,
      "num_input_tokens_seen": 17166472,
      "step": 26310
    },
    {
      "epoch": 13.791928721174004,
      "grad_norm": 0.11650248616933823,
      "learning_rate": 0.0002659267718568786,
      "loss": 0.54,
      "num_input_tokens_seen": 17170088,
      "step": 26315
    },
    {
      "epoch": 13.79454926624738,
      "grad_norm": 0.0913144052028656,
      "learning_rate": 0.0002657247180255091,
      "loss": 0.4276,
      "num_input_tokens_seen": 17172840,
      "step": 26320
    },
    {
      "epoch": 13.797169811320755,
      "grad_norm": 0.10137942433357239,
      "learning_rate": 0.0002655227132018558,
      "loss": 0.381,
      "num_input_tokens_seen": 17175752,
      "step": 26325
    },
    {
      "epoch": 13.79979035639413,
      "grad_norm": 0.1152055561542511,
      "learning_rate": 0.00026532075742817594,
      "loss": 0.4879,
      "num_input_tokens_seen": 17178504,
      "step": 26330
    },
    {
      "epoch": 13.802410901467505,
      "grad_norm": 0.1051362082362175,
      "learning_rate": 0.0002651188507467161,
      "loss": 0.467,
      "num_input_tokens_seen": 17180648,
      "step": 26335
    },
    {
      "epoch": 13.80503144654088,
      "grad_norm": 0.1394907385110855,
      "learning_rate": 0.00026491699319971303,
      "loss": 0.5361,
      "num_input_tokens_seen": 17184648,
      "step": 26340
    },
    {
      "epoch": 13.807651991614255,
      "grad_norm": 0.16272523999214172,
      "learning_rate": 0.00026471518482939317,
      "loss": 0.5441,
      "num_input_tokens_seen": 17188264,
      "step": 26345
    },
    {
      "epoch": 13.81027253668763,
      "grad_norm": 0.09913373738527298,
      "learning_rate": 0.0002645134256779722,
      "loss": 0.3345,
      "num_input_tokens_seen": 17191464,
      "step": 26350
    },
    {
      "epoch": 13.812893081761006,
      "grad_norm": 0.14978636801242828,
      "learning_rate": 0.00026431171578765624,
      "loss": 0.4181,
      "num_input_tokens_seen": 17194792,
      "step": 26355
    },
    {
      "epoch": 13.815513626834381,
      "grad_norm": 0.18738028407096863,
      "learning_rate": 0.0002641100552006402,
      "loss": 0.4172,
      "num_input_tokens_seen": 17197864,
      "step": 26360
    },
    {
      "epoch": 13.818134171907756,
      "grad_norm": 0.17769382894039154,
      "learning_rate": 0.0002639084439591095,
      "loss": 0.4742,
      "num_input_tokens_seen": 17200840,
      "step": 26365
    },
    {
      "epoch": 13.820754716981131,
      "grad_norm": 0.11389845609664917,
      "learning_rate": 0.00026370688210523873,
      "loss": 0.3871,
      "num_input_tokens_seen": 17204296,
      "step": 26370
    },
    {
      "epoch": 13.823375262054507,
      "grad_norm": 0.0952211543917656,
      "learning_rate": 0.00026350536968119275,
      "loss": 0.4273,
      "num_input_tokens_seen": 17206920,
      "step": 26375
    },
    {
      "epoch": 13.825995807127882,
      "grad_norm": 0.13173487782478333,
      "learning_rate": 0.0002633039067291252,
      "loss": 0.4938,
      "num_input_tokens_seen": 17209672,
      "step": 26380
    },
    {
      "epoch": 13.828616352201259,
      "grad_norm": 0.16944147646427155,
      "learning_rate": 0.00026310249329118007,
      "loss": 0.605,
      "num_input_tokens_seen": 17212296,
      "step": 26385
    },
    {
      "epoch": 13.831236897274634,
      "grad_norm": 0.1464463174343109,
      "learning_rate": 0.0002629011294094905,
      "loss": 0.5555,
      "num_input_tokens_seen": 17215880,
      "step": 26390
    },
    {
      "epoch": 13.833857442348009,
      "grad_norm": 0.1406737118959427,
      "learning_rate": 0.0002626998151261798,
      "loss": 0.4642,
      "num_input_tokens_seen": 17218760,
      "step": 26395
    },
    {
      "epoch": 13.836477987421384,
      "grad_norm": 0.18416835367679596,
      "learning_rate": 0.0002624985504833604,
      "loss": 0.5778,
      "num_input_tokens_seen": 17222408,
      "step": 26400
    },
    {
      "epoch": 13.83909853249476,
      "grad_norm": 0.127702534198761,
      "learning_rate": 0.0002622973355231349,
      "loss": 0.5037,
      "num_input_tokens_seen": 17225800,
      "step": 26405
    },
    {
      "epoch": 13.841719077568134,
      "grad_norm": 0.1329980492591858,
      "learning_rate": 0.00026209617028759497,
      "loss": 0.3418,
      "num_input_tokens_seen": 17229256,
      "step": 26410
    },
    {
      "epoch": 13.84433962264151,
      "grad_norm": 0.20398041605949402,
      "learning_rate": 0.00026189505481882184,
      "loss": 0.5377,
      "num_input_tokens_seen": 17232104,
      "step": 26415
    },
    {
      "epoch": 13.846960167714885,
      "grad_norm": 0.10987204313278198,
      "learning_rate": 0.00026169398915888687,
      "loss": 0.6781,
      "num_input_tokens_seen": 17235528,
      "step": 26420
    },
    {
      "epoch": 13.84958071278826,
      "grad_norm": 0.1493348777294159,
      "learning_rate": 0.0002614929733498506,
      "loss": 0.3971,
      "num_input_tokens_seen": 17238088,
      "step": 26425
    },
    {
      "epoch": 13.852201257861635,
      "grad_norm": 0.08356615155935287,
      "learning_rate": 0.0002612920074337634,
      "loss": 0.4537,
      "num_input_tokens_seen": 17241288,
      "step": 26430
    },
    {
      "epoch": 13.85482180293501,
      "grad_norm": 0.08784984052181244,
      "learning_rate": 0.00026109109145266496,
      "loss": 0.3835,
      "num_input_tokens_seen": 17244840,
      "step": 26435
    },
    {
      "epoch": 13.857442348008385,
      "grad_norm": 0.09819117188453674,
      "learning_rate": 0.00026089022544858445,
      "loss": 0.4609,
      "num_input_tokens_seen": 17247688,
      "step": 26440
    },
    {
      "epoch": 13.86006289308176,
      "grad_norm": 0.19111086428165436,
      "learning_rate": 0.00026068940946354075,
      "loss": 0.5057,
      "num_input_tokens_seen": 17252360,
      "step": 26445
    },
    {
      "epoch": 13.862683438155136,
      "grad_norm": 0.1600389927625656,
      "learning_rate": 0.0002604886435395425,
      "loss": 0.5158,
      "num_input_tokens_seen": 17255368,
      "step": 26450
    },
    {
      "epoch": 13.865303983228511,
      "grad_norm": 0.24718089401721954,
      "learning_rate": 0.00026028792771858744,
      "loss": 0.4853,
      "num_input_tokens_seen": 17257832,
      "step": 26455
    },
    {
      "epoch": 13.867924528301886,
      "grad_norm": 0.13044342398643494,
      "learning_rate": 0.00026008726204266333,
      "loss": 0.3186,
      "num_input_tokens_seen": 17261480,
      "step": 26460
    },
    {
      "epoch": 13.870545073375261,
      "grad_norm": 0.09314427524805069,
      "learning_rate": 0.00025988664655374693,
      "loss": 0.5021,
      "num_input_tokens_seen": 17264360,
      "step": 26465
    },
    {
      "epoch": 13.873165618448636,
      "grad_norm": 0.10281798988580704,
      "learning_rate": 0.00025968608129380455,
      "loss": 0.3804,
      "num_input_tokens_seen": 17268104,
      "step": 26470
    },
    {
      "epoch": 13.875786163522012,
      "grad_norm": 0.2101934254169464,
      "learning_rate": 0.00025948556630479234,
      "loss": 0.4037,
      "num_input_tokens_seen": 17271048,
      "step": 26475
    },
    {
      "epoch": 13.878406708595389,
      "grad_norm": 0.18816697597503662,
      "learning_rate": 0.0002592851016286557,
      "loss": 0.4066,
      "num_input_tokens_seen": 17274056,
      "step": 26480
    },
    {
      "epoch": 13.881027253668764,
      "grad_norm": 0.1782318353652954,
      "learning_rate": 0.0002590846873073298,
      "loss": 0.4393,
      "num_input_tokens_seen": 17277192,
      "step": 26485
    },
    {
      "epoch": 13.883647798742139,
      "grad_norm": 0.13427948951721191,
      "learning_rate": 0.0002588843233827387,
      "loss": 0.4274,
      "num_input_tokens_seen": 17281128,
      "step": 26490
    },
    {
      "epoch": 13.886268343815514,
      "grad_norm": 0.11275335401296616,
      "learning_rate": 0.0002586840098967963,
      "loss": 0.3752,
      "num_input_tokens_seen": 17283848,
      "step": 26495
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 0.1857282519340515,
      "learning_rate": 0.00025848374689140587,
      "loss": 0.3575,
      "num_input_tokens_seen": 17287208,
      "step": 26500
    },
    {
      "epoch": 13.891509433962264,
      "grad_norm": 0.13365253806114197,
      "learning_rate": 0.0002582835344084602,
      "loss": 0.374,
      "num_input_tokens_seen": 17290696,
      "step": 26505
    },
    {
      "epoch": 13.89412997903564,
      "grad_norm": 0.12781211733818054,
      "learning_rate": 0.00025808337248984175,
      "loss": 0.4262,
      "num_input_tokens_seen": 17294600,
      "step": 26510
    },
    {
      "epoch": 13.896750524109015,
      "grad_norm": 0.13816775381565094,
      "learning_rate": 0.00025788326117742185,
      "loss": 0.4677,
      "num_input_tokens_seen": 17297768,
      "step": 26515
    },
    {
      "epoch": 13.89937106918239,
      "grad_norm": 0.15394893288612366,
      "learning_rate": 0.00025768320051306127,
      "loss": 0.4232,
      "num_input_tokens_seen": 17300808,
      "step": 26520
    },
    {
      "epoch": 13.901991614255765,
      "grad_norm": 0.10966223478317261,
      "learning_rate": 0.00025748319053861063,
      "loss": 0.4454,
      "num_input_tokens_seen": 17303432,
      "step": 26525
    },
    {
      "epoch": 13.90461215932914,
      "grad_norm": 0.10554639250040054,
      "learning_rate": 0.0002572832312959098,
      "loss": 0.3939,
      "num_input_tokens_seen": 17306152,
      "step": 26530
    },
    {
      "epoch": 13.907232704402515,
      "grad_norm": 0.11596523970365524,
      "learning_rate": 0.0002570833228267879,
      "loss": 0.4091,
      "num_input_tokens_seen": 17308840,
      "step": 26535
    },
    {
      "epoch": 13.90985324947589,
      "grad_norm": 0.17843903601169586,
      "learning_rate": 0.00025688346517306366,
      "loss": 0.5132,
      "num_input_tokens_seen": 17311944,
      "step": 26540
    },
    {
      "epoch": 13.912473794549266,
      "grad_norm": 0.09793847799301147,
      "learning_rate": 0.00025668365837654497,
      "loss": 0.4254,
      "num_input_tokens_seen": 17314888,
      "step": 26545
    },
    {
      "epoch": 13.915094339622641,
      "grad_norm": 0.23504860699176788,
      "learning_rate": 0.0002564839024790288,
      "loss": 0.506,
      "num_input_tokens_seen": 17318248,
      "step": 26550
    },
    {
      "epoch": 13.917714884696016,
      "grad_norm": 0.07001612335443497,
      "learning_rate": 0.000256284197522302,
      "loss": 0.4366,
      "num_input_tokens_seen": 17321320,
      "step": 26555
    },
    {
      "epoch": 13.920335429769391,
      "grad_norm": 0.10046915709972382,
      "learning_rate": 0.00025608454354814075,
      "loss": 0.4063,
      "num_input_tokens_seen": 17324680,
      "step": 26560
    },
    {
      "epoch": 13.922955974842766,
      "grad_norm": 0.0863179862499237,
      "learning_rate": 0.0002558849405983104,
      "loss": 0.405,
      "num_input_tokens_seen": 17328072,
      "step": 26565
    },
    {
      "epoch": 13.925576519916142,
      "grad_norm": 0.1864691823720932,
      "learning_rate": 0.00025568538871456527,
      "loss": 0.4629,
      "num_input_tokens_seen": 17331272,
      "step": 26570
    },
    {
      "epoch": 13.928197064989519,
      "grad_norm": 0.1026180312037468,
      "learning_rate": 0.00025548588793864976,
      "loss": 0.3954,
      "num_input_tokens_seen": 17334920,
      "step": 26575
    },
    {
      "epoch": 13.930817610062894,
      "grad_norm": 0.16198274493217468,
      "learning_rate": 0.0002552864383122967,
      "loss": 0.4863,
      "num_input_tokens_seen": 17338664,
      "step": 26580
    },
    {
      "epoch": 13.933438155136269,
      "grad_norm": 0.17573782801628113,
      "learning_rate": 0.0002550870398772289,
      "loss": 0.5157,
      "num_input_tokens_seen": 17341000,
      "step": 26585
    },
    {
      "epoch": 13.936058700209644,
      "grad_norm": 0.10039006173610687,
      "learning_rate": 0.0002548876926751584,
      "loss": 0.3602,
      "num_input_tokens_seen": 17344456,
      "step": 26590
    },
    {
      "epoch": 13.93867924528302,
      "grad_norm": 0.17585189640522003,
      "learning_rate": 0.0002546883967477861,
      "loss": 0.537,
      "num_input_tokens_seen": 17347976,
      "step": 26595
    },
    {
      "epoch": 13.941299790356394,
      "grad_norm": 0.1489306092262268,
      "learning_rate": 0.00025448915213680245,
      "loss": 0.4811,
      "num_input_tokens_seen": 17351336,
      "step": 26600
    },
    {
      "epoch": 13.94392033542977,
      "grad_norm": 0.06738505512475967,
      "learning_rate": 0.0002542899588838875,
      "loss": 0.3862,
      "num_input_tokens_seen": 17356744,
      "step": 26605
    },
    {
      "epoch": 13.946540880503145,
      "grad_norm": 0.08697648346424103,
      "learning_rate": 0.0002540908170307097,
      "loss": 0.4415,
      "num_input_tokens_seen": 17359240,
      "step": 26610
    },
    {
      "epoch": 13.94916142557652,
      "grad_norm": 0.20860083401203156,
      "learning_rate": 0.00025389172661892753,
      "loss": 0.4592,
      "num_input_tokens_seen": 17363592,
      "step": 26615
    },
    {
      "epoch": 13.951781970649895,
      "grad_norm": 0.1982254981994629,
      "learning_rate": 0.00025369268769018856,
      "loss": 0.6062,
      "num_input_tokens_seen": 17365992,
      "step": 26620
    },
    {
      "epoch": 13.95440251572327,
      "grad_norm": 0.1158304512500763,
      "learning_rate": 0.00025349370028612914,
      "loss": 0.4416,
      "num_input_tokens_seen": 17368520,
      "step": 26625
    },
    {
      "epoch": 13.957023060796645,
      "grad_norm": 0.11016975343227386,
      "learning_rate": 0.0002532947644483755,
      "loss": 0.3878,
      "num_input_tokens_seen": 17371880,
      "step": 26630
    },
    {
      "epoch": 13.95964360587002,
      "grad_norm": 0.13498102128505707,
      "learning_rate": 0.00025309588021854237,
      "loss": 0.3673,
      "num_input_tokens_seen": 17374600,
      "step": 26635
    },
    {
      "epoch": 13.962264150943396,
      "grad_norm": 0.14053286612033844,
      "learning_rate": 0.0002528970476382343,
      "loss": 0.3481,
      "num_input_tokens_seen": 17378600,
      "step": 26640
    },
    {
      "epoch": 13.964884696016771,
      "grad_norm": 0.17347081005573273,
      "learning_rate": 0.00025269826674904493,
      "loss": 0.4077,
      "num_input_tokens_seen": 17381416,
      "step": 26645
    },
    {
      "epoch": 13.967505241090146,
      "grad_norm": 0.14569635689258575,
      "learning_rate": 0.0002524995375925566,
      "loss": 0.416,
      "num_input_tokens_seen": 17384264,
      "step": 26650
    },
    {
      "epoch": 13.970125786163521,
      "grad_norm": 0.14344340562820435,
      "learning_rate": 0.0002523008602103414,
      "loss": 0.5194,
      "num_input_tokens_seen": 17387912,
      "step": 26655
    },
    {
      "epoch": 13.972746331236896,
      "grad_norm": 0.2690078318119049,
      "learning_rate": 0.00025210223464396055,
      "loss": 0.3366,
      "num_input_tokens_seen": 17390600,
      "step": 26660
    },
    {
      "epoch": 13.975366876310272,
      "grad_norm": 0.07959900796413422,
      "learning_rate": 0.00025190366093496376,
      "loss": 0.4129,
      "num_input_tokens_seen": 17393704,
      "step": 26665
    },
    {
      "epoch": 13.977987421383649,
      "grad_norm": 0.11542534083127975,
      "learning_rate": 0.0002517051391248909,
      "loss": 0.5076,
      "num_input_tokens_seen": 17396808,
      "step": 26670
    },
    {
      "epoch": 13.980607966457024,
      "grad_norm": 0.11459436267614365,
      "learning_rate": 0.0002515066692552701,
      "loss": 0.4579,
      "num_input_tokens_seen": 17400040,
      "step": 26675
    },
    {
      "epoch": 13.983228511530399,
      "grad_norm": 0.0696704238653183,
      "learning_rate": 0.00025130825136761916,
      "loss": 0.3365,
      "num_input_tokens_seen": 17403624,
      "step": 26680
    },
    {
      "epoch": 13.985849056603774,
      "grad_norm": 0.11503296345472336,
      "learning_rate": 0.0002511098855034447,
      "loss": 0.6446,
      "num_input_tokens_seen": 17406760,
      "step": 26685
    },
    {
      "epoch": 13.98846960167715,
      "grad_norm": 0.07372309267520905,
      "learning_rate": 0.00025091157170424296,
      "loss": 0.4397,
      "num_input_tokens_seen": 17409832,
      "step": 26690
    },
    {
      "epoch": 13.991090146750524,
      "grad_norm": 0.12198126316070557,
      "learning_rate": 0.0002507133100114984,
      "loss": 0.4179,
      "num_input_tokens_seen": 17413672,
      "step": 26695
    },
    {
      "epoch": 13.9937106918239,
      "grad_norm": 0.09636210650205612,
      "learning_rate": 0.0002505151004666857,
      "loss": 0.5487,
      "num_input_tokens_seen": 17416936,
      "step": 26700
    },
    {
      "epoch": 13.996331236897275,
      "grad_norm": 0.10239371657371521,
      "learning_rate": 0.0002503169431112674,
      "loss": 0.4638,
      "num_input_tokens_seen": 17420520,
      "step": 26705
    },
    {
      "epoch": 13.99895178197065,
      "grad_norm": 0.07861364632844925,
      "learning_rate": 0.0002501188379866961,
      "loss": 0.3804,
      "num_input_tokens_seen": 17424840,
      "step": 26710
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.4591861665248871,
      "eval_runtime": 13.5803,
      "eval_samples_per_second": 62.443,
      "eval_steps_per_second": 15.611,
      "num_input_tokens_seen": 17425368,
      "step": 26712
    },
    {
      "epoch": 14.001572327044025,
      "grad_norm": 0.18276505172252655,
      "learning_rate": 0.0002499207851344133,
      "loss": 0.3925,
      "num_input_tokens_seen": 17427448,
      "step": 26715
    },
    {
      "epoch": 14.0041928721174,
      "grad_norm": 0.1446499079465866,
      "learning_rate": 0.00024972278459584903,
      "loss": 0.4476,
      "num_input_tokens_seen": 17430008,
      "step": 26720
    },
    {
      "epoch": 14.006813417190775,
      "grad_norm": 0.11057281494140625,
      "learning_rate": 0.00024952483641242304,
      "loss": 0.5064,
      "num_input_tokens_seen": 17433080,
      "step": 26725
    },
    {
      "epoch": 14.00943396226415,
      "grad_norm": 0.1483139842748642,
      "learning_rate": 0.0002493269406255435,
      "loss": 0.4679,
      "num_input_tokens_seen": 17436088,
      "step": 26730
    },
    {
      "epoch": 14.012054507337526,
      "grad_norm": 0.14779382944107056,
      "learning_rate": 0.0002491290972766082,
      "loss": 0.3463,
      "num_input_tokens_seen": 17438680,
      "step": 26735
    },
    {
      "epoch": 14.014675052410901,
      "grad_norm": 0.14528676867485046,
      "learning_rate": 0.0002489313064070037,
      "loss": 0.4037,
      "num_input_tokens_seen": 17442424,
      "step": 26740
    },
    {
      "epoch": 14.017295597484276,
      "grad_norm": 0.20193038880825043,
      "learning_rate": 0.00024873356805810566,
      "loss": 0.4141,
      "num_input_tokens_seen": 17445784,
      "step": 26745
    },
    {
      "epoch": 14.019916142557651,
      "grad_norm": 0.12414872646331787,
      "learning_rate": 0.00024853588227127864,
      "loss": 0.3221,
      "num_input_tokens_seen": 17448280,
      "step": 26750
    },
    {
      "epoch": 14.022536687631026,
      "grad_norm": 0.18916012346744537,
      "learning_rate": 0.000248338249087876,
      "loss": 0.4558,
      "num_input_tokens_seen": 17452248,
      "step": 26755
    },
    {
      "epoch": 14.025157232704403,
      "grad_norm": 0.09667158126831055,
      "learning_rate": 0.0002481406685492405,
      "loss": 0.3421,
      "num_input_tokens_seen": 17455672,
      "step": 26760
    },
    {
      "epoch": 14.027777777777779,
      "grad_norm": 0.0951257273554802,
      "learning_rate": 0.00024794314069670383,
      "loss": 0.3263,
      "num_input_tokens_seen": 17458328,
      "step": 26765
    },
    {
      "epoch": 14.030398322851154,
      "grad_norm": 0.15660755336284637,
      "learning_rate": 0.0002477456655715865,
      "loss": 0.474,
      "num_input_tokens_seen": 17460856,
      "step": 26770
    },
    {
      "epoch": 14.033018867924529,
      "grad_norm": 0.10506119579076767,
      "learning_rate": 0.0002475482432151982,
      "loss": 0.3763,
      "num_input_tokens_seen": 17464440,
      "step": 26775
    },
    {
      "epoch": 14.035639412997904,
      "grad_norm": 0.13171572983264923,
      "learning_rate": 0.00024735087366883733,
      "loss": 0.48,
      "num_input_tokens_seen": 17467512,
      "step": 26780
    },
    {
      "epoch": 14.03825995807128,
      "grad_norm": 0.21453121304512024,
      "learning_rate": 0.00024715355697379115,
      "loss": 0.4154,
      "num_input_tokens_seen": 17471224,
      "step": 26785
    },
    {
      "epoch": 14.040880503144654,
      "grad_norm": 0.09627760201692581,
      "learning_rate": 0.0002469562931713362,
      "loss": 0.3371,
      "num_input_tokens_seen": 17474616,
      "step": 26790
    },
    {
      "epoch": 14.04350104821803,
      "grad_norm": 0.09736017882823944,
      "learning_rate": 0.00024675908230273785,
      "loss": 0.3785,
      "num_input_tokens_seen": 17477880,
      "step": 26795
    },
    {
      "epoch": 14.046121593291405,
      "grad_norm": 0.14310245215892792,
      "learning_rate": 0.00024656192440925055,
      "loss": 0.5621,
      "num_input_tokens_seen": 17481016,
      "step": 26800
    },
    {
      "epoch": 14.04874213836478,
      "grad_norm": 0.09365909546613693,
      "learning_rate": 0.0002463648195321173,
      "loss": 0.358,
      "num_input_tokens_seen": 17484568,
      "step": 26805
    },
    {
      "epoch": 14.051362683438155,
      "grad_norm": 0.15773916244506836,
      "learning_rate": 0.00024616776771257,
      "loss": 0.3875,
      "num_input_tokens_seen": 17491032,
      "step": 26810
    },
    {
      "epoch": 14.05398322851153,
      "grad_norm": 0.22852925956249237,
      "learning_rate": 0.00024597076899182977,
      "loss": 0.3318,
      "num_input_tokens_seen": 17494136,
      "step": 26815
    },
    {
      "epoch": 14.056603773584905,
      "grad_norm": 0.1666318029165268,
      "learning_rate": 0.0002457738234111066,
      "loss": 0.5217,
      "num_input_tokens_seen": 17496920,
      "step": 26820
    },
    {
      "epoch": 14.05922431865828,
      "grad_norm": 0.14272496104240417,
      "learning_rate": 0.00024557693101159937,
      "loss": 0.3004,
      "num_input_tokens_seen": 17500024,
      "step": 26825
    },
    {
      "epoch": 14.061844863731656,
      "grad_norm": 0.14434026181697845,
      "learning_rate": 0.00024538009183449553,
      "loss": 0.4132,
      "num_input_tokens_seen": 17503416,
      "step": 26830
    },
    {
      "epoch": 14.064465408805031,
      "grad_norm": 0.16823571920394897,
      "learning_rate": 0.0002451833059209715,
      "loss": 0.3423,
      "num_input_tokens_seen": 17506648,
      "step": 26835
    },
    {
      "epoch": 14.067085953878406,
      "grad_norm": 0.19622229039669037,
      "learning_rate": 0.00024498657331219274,
      "loss": 0.3787,
      "num_input_tokens_seen": 17509496,
      "step": 26840
    },
    {
      "epoch": 14.069706498951781,
      "grad_norm": 0.09876818209886551,
      "learning_rate": 0.0002447898940493135,
      "loss": 0.4157,
      "num_input_tokens_seen": 17513464,
      "step": 26845
    },
    {
      "epoch": 14.072327044025156,
      "grad_norm": 0.15772725641727448,
      "learning_rate": 0.00024459326817347684,
      "loss": 0.2986,
      "num_input_tokens_seen": 17516568,
      "step": 26850
    },
    {
      "epoch": 14.074947589098532,
      "grad_norm": 0.1381009817123413,
      "learning_rate": 0.0002443966957258148,
      "loss": 0.5198,
      "num_input_tokens_seen": 17520440,
      "step": 26855
    },
    {
      "epoch": 14.077568134171909,
      "grad_norm": 0.18808172643184662,
      "learning_rate": 0.000244200176747448,
      "loss": 0.4157,
      "num_input_tokens_seen": 17523224,
      "step": 26860
    },
    {
      "epoch": 14.080188679245284,
      "grad_norm": 0.07652987539768219,
      "learning_rate": 0.00024400371127948568,
      "loss": 0.4103,
      "num_input_tokens_seen": 17530008,
      "step": 26865
    },
    {
      "epoch": 14.082809224318659,
      "grad_norm": 0.13239043951034546,
      "learning_rate": 0.00024380729936302636,
      "loss": 0.4075,
      "num_input_tokens_seen": 17533048,
      "step": 26870
    },
    {
      "epoch": 14.085429769392034,
      "grad_norm": 0.11795571446418762,
      "learning_rate": 0.00024361094103915725,
      "loss": 0.3429,
      "num_input_tokens_seen": 17535608,
      "step": 26875
    },
    {
      "epoch": 14.08805031446541,
      "grad_norm": 0.08140157163143158,
      "learning_rate": 0.00024341463634895444,
      "loss": 0.3687,
      "num_input_tokens_seen": 17539096,
      "step": 26880
    },
    {
      "epoch": 14.090670859538784,
      "grad_norm": 0.12341146916151047,
      "learning_rate": 0.0002432183853334824,
      "loss": 0.5294,
      "num_input_tokens_seen": 17542936,
      "step": 26885
    },
    {
      "epoch": 14.09329140461216,
      "grad_norm": 0.14887602627277374,
      "learning_rate": 0.00024302218803379445,
      "loss": 0.5257,
      "num_input_tokens_seen": 17545688,
      "step": 26890
    },
    {
      "epoch": 14.095911949685535,
      "grad_norm": 0.10071398317813873,
      "learning_rate": 0.000242826044490933,
      "loss": 0.4288,
      "num_input_tokens_seen": 17548536,
      "step": 26895
    },
    {
      "epoch": 14.09853249475891,
      "grad_norm": 0.12343429774045944,
      "learning_rate": 0.00024262995474592903,
      "loss": 0.3789,
      "num_input_tokens_seen": 17552056,
      "step": 26900
    },
    {
      "epoch": 14.101153039832285,
      "grad_norm": 0.10343031585216522,
      "learning_rate": 0.0002424339188398025,
      "loss": 0.3553,
      "num_input_tokens_seen": 17555512,
      "step": 26905
    },
    {
      "epoch": 14.10377358490566,
      "grad_norm": 0.24394746124744415,
      "learning_rate": 0.00024223793681356148,
      "loss": 0.3996,
      "num_input_tokens_seen": 17559320,
      "step": 26910
    },
    {
      "epoch": 14.106394129979035,
      "grad_norm": 0.11488091945648193,
      "learning_rate": 0.00024204200870820358,
      "loss": 0.3149,
      "num_input_tokens_seen": 17562616,
      "step": 26915
    },
    {
      "epoch": 14.10901467505241,
      "grad_norm": 0.13171514868736267,
      "learning_rate": 0.0002418461345647143,
      "loss": 0.4558,
      "num_input_tokens_seen": 17566200,
      "step": 26920
    },
    {
      "epoch": 14.111635220125786,
      "grad_norm": 0.12368959188461304,
      "learning_rate": 0.00024165031442406853,
      "loss": 0.3815,
      "num_input_tokens_seen": 17569496,
      "step": 26925
    },
    {
      "epoch": 14.114255765199161,
      "grad_norm": 0.17683330178260803,
      "learning_rate": 0.00024145454832722956,
      "loss": 0.4644,
      "num_input_tokens_seen": 17571832,
      "step": 26930
    },
    {
      "epoch": 14.116876310272536,
      "grad_norm": 0.18070806562900543,
      "learning_rate": 0.00024125883631514967,
      "loss": 0.4277,
      "num_input_tokens_seen": 17574744,
      "step": 26935
    },
    {
      "epoch": 14.119496855345911,
      "grad_norm": 0.10973707586526871,
      "learning_rate": 0.00024106317842876907,
      "loss": 0.4537,
      "num_input_tokens_seen": 17578232,
      "step": 26940
    },
    {
      "epoch": 14.122117400419286,
      "grad_norm": 0.1428249478340149,
      "learning_rate": 0.0002408675747090177,
      "loss": 0.4758,
      "num_input_tokens_seen": 17581688,
      "step": 26945
    },
    {
      "epoch": 14.124737945492662,
      "grad_norm": 0.13206467032432556,
      "learning_rate": 0.00024067202519681313,
      "loss": 0.4128,
      "num_input_tokens_seen": 17585336,
      "step": 26950
    },
    {
      "epoch": 14.127358490566039,
      "grad_norm": 0.11939214169979095,
      "learning_rate": 0.00024047652993306235,
      "loss": 0.4044,
      "num_input_tokens_seen": 17588120,
      "step": 26955
    },
    {
      "epoch": 14.129979035639414,
      "grad_norm": 0.10674396902322769,
      "learning_rate": 0.00024028108895866084,
      "loss": 0.421,
      "num_input_tokens_seen": 17591928,
      "step": 26960
    },
    {
      "epoch": 14.132599580712789,
      "grad_norm": 0.08046822994947433,
      "learning_rate": 0.00024008570231449239,
      "loss": 0.3263,
      "num_input_tokens_seen": 17596120,
      "step": 26965
    },
    {
      "epoch": 14.135220125786164,
      "grad_norm": 0.13548478484153748,
      "learning_rate": 0.00023989037004142966,
      "loss": 0.5023,
      "num_input_tokens_seen": 17599416,
      "step": 26970
    },
    {
      "epoch": 14.13784067085954,
      "grad_norm": 0.09293042123317719,
      "learning_rate": 0.0002396950921803343,
      "loss": 0.4866,
      "num_input_tokens_seen": 17603192,
      "step": 26975
    },
    {
      "epoch": 14.140461215932914,
      "grad_norm": 0.16757260262966156,
      "learning_rate": 0.0002394998687720557,
      "loss": 0.4279,
      "num_input_tokens_seen": 17606520,
      "step": 26980
    },
    {
      "epoch": 14.14308176100629,
      "grad_norm": 0.181559219956398,
      "learning_rate": 0.0002393046998574328,
      "loss": 0.3894,
      "num_input_tokens_seen": 17609112,
      "step": 26985
    },
    {
      "epoch": 14.145702306079665,
      "grad_norm": 0.21731045842170715,
      "learning_rate": 0.0002391095854772924,
      "loss": 0.4584,
      "num_input_tokens_seen": 17612440,
      "step": 26990
    },
    {
      "epoch": 14.14832285115304,
      "grad_norm": 0.08077992498874664,
      "learning_rate": 0.00023891452567245027,
      "loss": 0.3622,
      "num_input_tokens_seen": 17615384,
      "step": 26995
    },
    {
      "epoch": 14.150943396226415,
      "grad_norm": 0.12013047933578491,
      "learning_rate": 0.00023871952048371093,
      "loss": 0.3356,
      "num_input_tokens_seen": 17618456,
      "step": 27000
    },
    {
      "epoch": 14.15356394129979,
      "grad_norm": 0.13268449902534485,
      "learning_rate": 0.00023852456995186684,
      "loss": 0.7076,
      "num_input_tokens_seen": 17621720,
      "step": 27005
    },
    {
      "epoch": 14.156184486373165,
      "grad_norm": 0.09911465644836426,
      "learning_rate": 0.0002383296741176997,
      "loss": 0.4553,
      "num_input_tokens_seen": 17625112,
      "step": 27010
    },
    {
      "epoch": 14.15880503144654,
      "grad_norm": 0.17698000371456146,
      "learning_rate": 0.00023813483302197964,
      "loss": 0.5749,
      "num_input_tokens_seen": 17628824,
      "step": 27015
    },
    {
      "epoch": 14.161425576519916,
      "grad_norm": 0.29148828983306885,
      "learning_rate": 0.00023794004670546476,
      "loss": 0.3204,
      "num_input_tokens_seen": 17632568,
      "step": 27020
    },
    {
      "epoch": 14.164046121593291,
      "grad_norm": 0.09020198881626129,
      "learning_rate": 0.00023774531520890242,
      "loss": 0.3653,
      "num_input_tokens_seen": 17635928,
      "step": 27025
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 0.14075708389282227,
      "learning_rate": 0.00023755063857302833,
      "loss": 0.4855,
      "num_input_tokens_seen": 17639832,
      "step": 27030
    },
    {
      "epoch": 14.169287211740041,
      "grad_norm": 0.21489495038986206,
      "learning_rate": 0.00023735601683856628,
      "loss": 0.396,
      "num_input_tokens_seen": 17643160,
      "step": 27035
    },
    {
      "epoch": 14.171907756813416,
      "grad_norm": 0.09223420172929764,
      "learning_rate": 0.0002371614500462293,
      "loss": 0.5298,
      "num_input_tokens_seen": 17646040,
      "step": 27040
    },
    {
      "epoch": 14.174528301886792,
      "grad_norm": 0.2496863305568695,
      "learning_rate": 0.00023696693823671816,
      "loss": 0.392,
      "num_input_tokens_seen": 17648600,
      "step": 27045
    },
    {
      "epoch": 14.177148846960169,
      "grad_norm": 0.20716950297355652,
      "learning_rate": 0.00023677248145072272,
      "loss": 0.4408,
      "num_input_tokens_seen": 17652024,
      "step": 27050
    },
    {
      "epoch": 14.179769392033544,
      "grad_norm": 0.27219465374946594,
      "learning_rate": 0.0002365780797289211,
      "loss": 0.3969,
      "num_input_tokens_seen": 17655160,
      "step": 27055
    },
    {
      "epoch": 14.182389937106919,
      "grad_norm": 0.15881027281284332,
      "learning_rate": 0.00023638373311198014,
      "loss": 0.4465,
      "num_input_tokens_seen": 17658712,
      "step": 27060
    },
    {
      "epoch": 14.185010482180294,
      "grad_norm": 0.17079955339431763,
      "learning_rate": 0.00023618944164055468,
      "loss": 0.3525,
      "num_input_tokens_seen": 17661432,
      "step": 27065
    },
    {
      "epoch": 14.18763102725367,
      "grad_norm": 0.10441554337739944,
      "learning_rate": 0.00023599520535528813,
      "loss": 0.5085,
      "num_input_tokens_seen": 17664408,
      "step": 27070
    },
    {
      "epoch": 14.190251572327044,
      "grad_norm": 0.16559331119060516,
      "learning_rate": 0.00023580102429681266,
      "loss": 0.3629,
      "num_input_tokens_seen": 17667608,
      "step": 27075
    },
    {
      "epoch": 14.19287211740042,
      "grad_norm": 0.13276296854019165,
      "learning_rate": 0.00023560689850574885,
      "loss": 0.4342,
      "num_input_tokens_seen": 17670872,
      "step": 27080
    },
    {
      "epoch": 14.195492662473795,
      "grad_norm": 0.1122281402349472,
      "learning_rate": 0.00023541282802270563,
      "loss": 0.4748,
      "num_input_tokens_seen": 17673368,
      "step": 27085
    },
    {
      "epoch": 14.19811320754717,
      "grad_norm": 0.10692740231752396,
      "learning_rate": 0.00023521881288828005,
      "loss": 0.4059,
      "num_input_tokens_seen": 17675800,
      "step": 27090
    },
    {
      "epoch": 14.200733752620545,
      "grad_norm": 0.1275417059659958,
      "learning_rate": 0.00023502485314305828,
      "loss": 0.355,
      "num_input_tokens_seen": 17679288,
      "step": 27095
    },
    {
      "epoch": 14.20335429769392,
      "grad_norm": 0.22926288843154907,
      "learning_rate": 0.00023483094882761397,
      "loss": 0.4886,
      "num_input_tokens_seen": 17682392,
      "step": 27100
    },
    {
      "epoch": 14.205974842767295,
      "grad_norm": 0.17984454333782196,
      "learning_rate": 0.00023463709998250998,
      "loss": 0.4809,
      "num_input_tokens_seen": 17685176,
      "step": 27105
    },
    {
      "epoch": 14.20859538784067,
      "grad_norm": 0.2248874306678772,
      "learning_rate": 0.00023444330664829728,
      "loss": 0.3818,
      "num_input_tokens_seen": 17687768,
      "step": 27110
    },
    {
      "epoch": 14.211215932914046,
      "grad_norm": 0.12554794549942017,
      "learning_rate": 0.00023424956886551536,
      "loss": 0.3166,
      "num_input_tokens_seen": 17690712,
      "step": 27115
    },
    {
      "epoch": 14.213836477987421,
      "grad_norm": 0.12038902193307877,
      "learning_rate": 0.0002340558866746918,
      "loss": 0.6865,
      "num_input_tokens_seen": 17693944,
      "step": 27120
    },
    {
      "epoch": 14.216457023060796,
      "grad_norm": 0.10979607701301575,
      "learning_rate": 0.0002338622601163425,
      "loss": 0.379,
      "num_input_tokens_seen": 17697528,
      "step": 27125
    },
    {
      "epoch": 14.219077568134171,
      "grad_norm": 0.17736050486564636,
      "learning_rate": 0.00023366868923097213,
      "loss": 0.4568,
      "num_input_tokens_seen": 17700472,
      "step": 27130
    },
    {
      "epoch": 14.221698113207546,
      "grad_norm": 0.13535459339618683,
      "learning_rate": 0.0002334751740590735,
      "loss": 0.3836,
      "num_input_tokens_seen": 17704024,
      "step": 27135
    },
    {
      "epoch": 14.224318658280922,
      "grad_norm": 0.16610610485076904,
      "learning_rate": 0.00023328171464112796,
      "loss": 0.3723,
      "num_input_tokens_seen": 17706840,
      "step": 27140
    },
    {
      "epoch": 14.226939203354299,
      "grad_norm": 0.14073042571544647,
      "learning_rate": 0.00023308831101760485,
      "loss": 0.565,
      "num_input_tokens_seen": 17709592,
      "step": 27145
    },
    {
      "epoch": 14.229559748427674,
      "grad_norm": 0.09983041882514954,
      "learning_rate": 0.00023289496322896174,
      "loss": 0.4555,
      "num_input_tokens_seen": 17713112,
      "step": 27150
    },
    {
      "epoch": 14.232180293501049,
      "grad_norm": 0.16469505429267883,
      "learning_rate": 0.00023270167131564512,
      "loss": 0.5168,
      "num_input_tokens_seen": 17716056,
      "step": 27155
    },
    {
      "epoch": 14.234800838574424,
      "grad_norm": 0.16130781173706055,
      "learning_rate": 0.0002325084353180893,
      "loss": 0.4273,
      "num_input_tokens_seen": 17719704,
      "step": 27160
    },
    {
      "epoch": 14.2374213836478,
      "grad_norm": 0.1654834747314453,
      "learning_rate": 0.00023231525527671716,
      "loss": 0.418,
      "num_input_tokens_seen": 17722456,
      "step": 27165
    },
    {
      "epoch": 14.240041928721174,
      "grad_norm": 0.09090162813663483,
      "learning_rate": 0.00023212213123193986,
      "loss": 0.3973,
      "num_input_tokens_seen": 17725400,
      "step": 27170
    },
    {
      "epoch": 14.24266247379455,
      "grad_norm": 0.1284588724374771,
      "learning_rate": 0.0002319290632241566,
      "loss": 0.5558,
      "num_input_tokens_seen": 17728440,
      "step": 27175
    },
    {
      "epoch": 14.245283018867925,
      "grad_norm": 0.3882398009300232,
      "learning_rate": 0.0002317360512937548,
      "loss": 0.3618,
      "num_input_tokens_seen": 17731192,
      "step": 27180
    },
    {
      "epoch": 14.2479035639413,
      "grad_norm": 0.09777311980724335,
      "learning_rate": 0.00023154309548111058,
      "loss": 0.3425,
      "num_input_tokens_seen": 17735096,
      "step": 27185
    },
    {
      "epoch": 14.250524109014675,
      "grad_norm": 0.13528449833393097,
      "learning_rate": 0.000231350195826588,
      "loss": 0.4678,
      "num_input_tokens_seen": 17739000,
      "step": 27190
    },
    {
      "epoch": 14.25314465408805,
      "grad_norm": 0.14764735102653503,
      "learning_rate": 0.00023115735237053982,
      "loss": 0.3757,
      "num_input_tokens_seen": 17741784,
      "step": 27195
    },
    {
      "epoch": 14.255765199161425,
      "grad_norm": 0.2963281273841858,
      "learning_rate": 0.00023096456515330632,
      "loss": 0.5084,
      "num_input_tokens_seen": 17744440,
      "step": 27200
    },
    {
      "epoch": 14.2583857442348,
      "grad_norm": 0.12290485203266144,
      "learning_rate": 0.0002307718342152163,
      "loss": 0.6026,
      "num_input_tokens_seen": 17747352,
      "step": 27205
    },
    {
      "epoch": 14.261006289308176,
      "grad_norm": 0.09445343911647797,
      "learning_rate": 0.0002305791595965871,
      "loss": 0.4028,
      "num_input_tokens_seen": 17750104,
      "step": 27210
    },
    {
      "epoch": 14.26362683438155,
      "grad_norm": 0.1389102339744568,
      "learning_rate": 0.00023038654133772395,
      "loss": 0.3945,
      "num_input_tokens_seen": 17752728,
      "step": 27215
    },
    {
      "epoch": 14.266247379454926,
      "grad_norm": 0.09474895149469376,
      "learning_rate": 0.00023019397947892062,
      "loss": 0.3533,
      "num_input_tokens_seen": 17756248,
      "step": 27220
    },
    {
      "epoch": 14.268867924528301,
      "grad_norm": 0.1315271407365799,
      "learning_rate": 0.00023000147406045853,
      "loss": 0.3934,
      "num_input_tokens_seen": 17759704,
      "step": 27225
    },
    {
      "epoch": 14.271488469601676,
      "grad_norm": 0.12919814884662628,
      "learning_rate": 0.00022980902512260786,
      "loss": 0.4363,
      "num_input_tokens_seen": 17763064,
      "step": 27230
    },
    {
      "epoch": 14.274109014675052,
      "grad_norm": 0.10214187949895859,
      "learning_rate": 0.00022961663270562656,
      "loss": 0.4902,
      "num_input_tokens_seen": 17767384,
      "step": 27235
    },
    {
      "epoch": 14.276729559748428,
      "grad_norm": 0.1387891173362732,
      "learning_rate": 0.0002294242968497609,
      "loss": 0.4956,
      "num_input_tokens_seen": 17770328,
      "step": 27240
    },
    {
      "epoch": 14.279350104821804,
      "grad_norm": 0.17735570669174194,
      "learning_rate": 0.00022923201759524552,
      "loss": 0.401,
      "num_input_tokens_seen": 17774072,
      "step": 27245
    },
    {
      "epoch": 14.281970649895179,
      "grad_norm": 0.2380700260400772,
      "learning_rate": 0.00022903979498230316,
      "loss": 0.2827,
      "num_input_tokens_seen": 17776344,
      "step": 27250
    },
    {
      "epoch": 14.284591194968554,
      "grad_norm": 0.1114635244011879,
      "learning_rate": 0.00022884762905114436,
      "loss": 0.4809,
      "num_input_tokens_seen": 17780536,
      "step": 27255
    },
    {
      "epoch": 14.28721174004193,
      "grad_norm": 0.07819311320781708,
      "learning_rate": 0.00022865551984196792,
      "loss": 0.4658,
      "num_input_tokens_seen": 17784024,
      "step": 27260
    },
    {
      "epoch": 14.289832285115304,
      "grad_norm": 0.149290069937706,
      "learning_rate": 0.0002284634673949611,
      "loss": 0.5698,
      "num_input_tokens_seen": 17788248,
      "step": 27265
    },
    {
      "epoch": 14.29245283018868,
      "grad_norm": 0.20543096959590912,
      "learning_rate": 0.00022827147175029906,
      "loss": 0.4158,
      "num_input_tokens_seen": 17790872,
      "step": 27270
    },
    {
      "epoch": 14.295073375262055,
      "grad_norm": 0.12743061780929565,
      "learning_rate": 0.0002280795329481452,
      "loss": 0.4105,
      "num_input_tokens_seen": 17793912,
      "step": 27275
    },
    {
      "epoch": 14.29769392033543,
      "grad_norm": 0.11559204012155533,
      "learning_rate": 0.00022788765102865066,
      "loss": 0.4106,
      "num_input_tokens_seen": 17797336,
      "step": 27280
    },
    {
      "epoch": 14.300314465408805,
      "grad_norm": 0.17653875052928925,
      "learning_rate": 0.00022769582603195533,
      "loss": 0.499,
      "num_input_tokens_seen": 17800728,
      "step": 27285
    },
    {
      "epoch": 14.30293501048218,
      "grad_norm": 0.19998209178447723,
      "learning_rate": 0.00022750405799818634,
      "loss": 0.6324,
      "num_input_tokens_seen": 17803448,
      "step": 27290
    },
    {
      "epoch": 14.305555555555555,
      "grad_norm": 0.08140391856431961,
      "learning_rate": 0.00022731234696745967,
      "loss": 0.3636,
      "num_input_tokens_seen": 17807896,
      "step": 27295
    },
    {
      "epoch": 14.30817610062893,
      "grad_norm": 0.09142924845218658,
      "learning_rate": 0.0002271206929798792,
      "loss": 0.4193,
      "num_input_tokens_seen": 17811032,
      "step": 27300
    },
    {
      "epoch": 14.310796645702306,
      "grad_norm": 0.07193213701248169,
      "learning_rate": 0.00022692909607553642,
      "loss": 0.4644,
      "num_input_tokens_seen": 17815640,
      "step": 27305
    },
    {
      "epoch": 14.31341719077568,
      "grad_norm": 0.08709776401519775,
      "learning_rate": 0.00022673755629451132,
      "loss": 0.3895,
      "num_input_tokens_seen": 17818744,
      "step": 27310
    },
    {
      "epoch": 14.316037735849056,
      "grad_norm": 0.139583021402359,
      "learning_rate": 0.00022654607367687213,
      "loss": 0.3785,
      "num_input_tokens_seen": 17821432,
      "step": 27315
    },
    {
      "epoch": 14.318658280922431,
      "grad_norm": 0.2046259045600891,
      "learning_rate": 0.00022635464826267442,
      "loss": 0.45,
      "num_input_tokens_seen": 17824312,
      "step": 27320
    },
    {
      "epoch": 14.321278825995806,
      "grad_norm": 0.09396625310182571,
      "learning_rate": 0.00022616328009196236,
      "loss": 0.315,
      "num_input_tokens_seen": 17827288,
      "step": 27325
    },
    {
      "epoch": 14.323899371069182,
      "grad_norm": 0.14985565841197968,
      "learning_rate": 0.0002259719692047682,
      "loss": 0.3779,
      "num_input_tokens_seen": 17830392,
      "step": 27330
    },
    {
      "epoch": 14.326519916142558,
      "grad_norm": 0.1253538727760315,
      "learning_rate": 0.00022578071564111163,
      "loss": 0.4083,
      "num_input_tokens_seen": 17833432,
      "step": 27335
    },
    {
      "epoch": 14.329140461215934,
      "grad_norm": 0.13149599730968475,
      "learning_rate": 0.00022558951944100087,
      "loss": 0.4311,
      "num_input_tokens_seen": 17837112,
      "step": 27340
    },
    {
      "epoch": 14.331761006289309,
      "grad_norm": 0.11600250750780106,
      "learning_rate": 0.00022539838064443223,
      "loss": 0.3599,
      "num_input_tokens_seen": 17840088,
      "step": 27345
    },
    {
      "epoch": 14.334381551362684,
      "grad_norm": 0.19407574832439423,
      "learning_rate": 0.00022520729929138933,
      "loss": 0.3837,
      "num_input_tokens_seen": 17843000,
      "step": 27350
    },
    {
      "epoch": 14.33700209643606,
      "grad_norm": 0.19755974411964417,
      "learning_rate": 0.0002250162754218446,
      "loss": 0.447,
      "num_input_tokens_seen": 17846872,
      "step": 27355
    },
    {
      "epoch": 14.339622641509434,
      "grad_norm": 0.10419458150863647,
      "learning_rate": 0.00022482530907575767,
      "loss": 0.4453,
      "num_input_tokens_seen": 17850232,
      "step": 27360
    },
    {
      "epoch": 14.34224318658281,
      "grad_norm": 0.08201885968446732,
      "learning_rate": 0.00022463440029307674,
      "loss": 0.6233,
      "num_input_tokens_seen": 17854360,
      "step": 27365
    },
    {
      "epoch": 14.344863731656185,
      "grad_norm": 0.1139906644821167,
      "learning_rate": 0.00022444354911373794,
      "loss": 0.4312,
      "num_input_tokens_seen": 17858040,
      "step": 27370
    },
    {
      "epoch": 14.34748427672956,
      "grad_norm": 0.22066816687583923,
      "learning_rate": 0.00022425275557766473,
      "loss": 0.5198,
      "num_input_tokens_seen": 17861400,
      "step": 27375
    },
    {
      "epoch": 14.350104821802935,
      "grad_norm": 0.1958269625902176,
      "learning_rate": 0.00022406201972476937,
      "loss": 0.6471,
      "num_input_tokens_seen": 17865176,
      "step": 27380
    },
    {
      "epoch": 14.35272536687631,
      "grad_norm": 0.1419280618429184,
      "learning_rate": 0.00022387134159495122,
      "loss": 0.5096,
      "num_input_tokens_seen": 17867544,
      "step": 27385
    },
    {
      "epoch": 14.355345911949685,
      "grad_norm": 0.15022258460521698,
      "learning_rate": 0.00022368072122809813,
      "loss": 0.3584,
      "num_input_tokens_seen": 17870584,
      "step": 27390
    },
    {
      "epoch": 14.35796645702306,
      "grad_norm": 0.15658727288246155,
      "learning_rate": 0.0002234901586640858,
      "loss": 0.4,
      "num_input_tokens_seen": 17873560,
      "step": 27395
    },
    {
      "epoch": 14.360587002096436,
      "grad_norm": 0.12462262064218521,
      "learning_rate": 0.00022329965394277785,
      "loss": 0.5054,
      "num_input_tokens_seen": 17876824,
      "step": 27400
    },
    {
      "epoch": 14.36320754716981,
      "grad_norm": 0.16364245116710663,
      "learning_rate": 0.00022310920710402532,
      "loss": 0.458,
      "num_input_tokens_seen": 17879928,
      "step": 27405
    },
    {
      "epoch": 14.365828092243186,
      "grad_norm": 0.12460837513208389,
      "learning_rate": 0.00022291881818766796,
      "loss": 0.5034,
      "num_input_tokens_seen": 17883672,
      "step": 27410
    },
    {
      "epoch": 14.368448637316561,
      "grad_norm": 0.1236337199807167,
      "learning_rate": 0.00022272848723353252,
      "loss": 0.3752,
      "num_input_tokens_seen": 17887032,
      "step": 27415
    },
    {
      "epoch": 14.371069182389936,
      "grad_norm": 0.23962680995464325,
      "learning_rate": 0.00022253821428143422,
      "loss": 0.4214,
      "num_input_tokens_seen": 17889816,
      "step": 27420
    },
    {
      "epoch": 14.373689727463312,
      "grad_norm": 0.13544896245002747,
      "learning_rate": 0.0002223479993711761,
      "loss": 0.4439,
      "num_input_tokens_seen": 17893016,
      "step": 27425
    },
    {
      "epoch": 14.376310272536688,
      "grad_norm": 0.10638643801212311,
      "learning_rate": 0.00022215784254254906,
      "loss": 0.3837,
      "num_input_tokens_seen": 17896056,
      "step": 27430
    },
    {
      "epoch": 14.378930817610064,
      "grad_norm": 0.14156770706176758,
      "learning_rate": 0.00022196774383533157,
      "loss": 0.5243,
      "num_input_tokens_seen": 17899224,
      "step": 27435
    },
    {
      "epoch": 14.381551362683439,
      "grad_norm": 0.14405189454555511,
      "learning_rate": 0.0002217777032892899,
      "loss": 0.5312,
      "num_input_tokens_seen": 17902840,
      "step": 27440
    },
    {
      "epoch": 14.384171907756814,
      "grad_norm": 0.24109479784965515,
      "learning_rate": 0.00022158772094417863,
      "loss": 0.5216,
      "num_input_tokens_seen": 17905272,
      "step": 27445
    },
    {
      "epoch": 14.38679245283019,
      "grad_norm": 0.12690740823745728,
      "learning_rate": 0.00022139779683973983,
      "loss": 0.4644,
      "num_input_tokens_seen": 17907992,
      "step": 27450
    },
    {
      "epoch": 14.389412997903564,
      "grad_norm": 0.10523621737957001,
      "learning_rate": 0.00022120793101570364,
      "loss": 0.3629,
      "num_input_tokens_seen": 17910904,
      "step": 27455
    },
    {
      "epoch": 14.39203354297694,
      "grad_norm": 0.11427538841962814,
      "learning_rate": 0.00022101812351178762,
      "loss": 0.3779,
      "num_input_tokens_seen": 17914904,
      "step": 27460
    },
    {
      "epoch": 14.394654088050315,
      "grad_norm": 0.1403488963842392,
      "learning_rate": 0.00022082837436769725,
      "loss": 0.4284,
      "num_input_tokens_seen": 17918104,
      "step": 27465
    },
    {
      "epoch": 14.39727463312369,
      "grad_norm": 0.11531523615121841,
      "learning_rate": 0.00022063868362312596,
      "loss": 0.4381,
      "num_input_tokens_seen": 17921720,
      "step": 27470
    },
    {
      "epoch": 14.399895178197065,
      "grad_norm": 0.11240822821855545,
      "learning_rate": 0.00022044905131775495,
      "loss": 0.4691,
      "num_input_tokens_seen": 17924984,
      "step": 27475
    },
    {
      "epoch": 14.40251572327044,
      "grad_norm": 0.1546717882156372,
      "learning_rate": 0.00022025947749125313,
      "loss": 0.4297,
      "num_input_tokens_seen": 17927672,
      "step": 27480
    },
    {
      "epoch": 14.405136268343815,
      "grad_norm": 0.07895123213529587,
      "learning_rate": 0.0002200699621832773,
      "loss": 0.5454,
      "num_input_tokens_seen": 17931128,
      "step": 27485
    },
    {
      "epoch": 14.40775681341719,
      "grad_norm": 0.14158006012439728,
      "learning_rate": 0.0002198805054334718,
      "loss": 0.4346,
      "num_input_tokens_seen": 17934936,
      "step": 27490
    },
    {
      "epoch": 14.410377358490566,
      "grad_norm": 0.19862067699432373,
      "learning_rate": 0.00021969110728146862,
      "loss": 0.3197,
      "num_input_tokens_seen": 17937432,
      "step": 27495
    },
    {
      "epoch": 14.41299790356394,
      "grad_norm": 0.24423637986183167,
      "learning_rate": 0.00021950176776688784,
      "loss": 0.4141,
      "num_input_tokens_seen": 17939800,
      "step": 27500
    },
    {
      "epoch": 14.415618448637316,
      "grad_norm": 0.13461288809776306,
      "learning_rate": 0.0002193124869293372,
      "loss": 0.4581,
      "num_input_tokens_seen": 17942808,
      "step": 27505
    },
    {
      "epoch": 14.418238993710691,
      "grad_norm": 0.17019139230251312,
      "learning_rate": 0.00021912326480841223,
      "loss": 0.4366,
      "num_input_tokens_seen": 17945592,
      "step": 27510
    },
    {
      "epoch": 14.420859538784066,
      "grad_norm": 0.1991034746170044,
      "learning_rate": 0.0002189341014436958,
      "loss": 0.5102,
      "num_input_tokens_seen": 17948536,
      "step": 27515
    },
    {
      "epoch": 14.423480083857442,
      "grad_norm": 0.12815585732460022,
      "learning_rate": 0.00021874499687475857,
      "loss": 0.4377,
      "num_input_tokens_seen": 17952408,
      "step": 27520
    },
    {
      "epoch": 14.426100628930818,
      "grad_norm": 0.18995125591754913,
      "learning_rate": 0.00021855595114115935,
      "loss": 0.4317,
      "num_input_tokens_seen": 17955896,
      "step": 27525
    },
    {
      "epoch": 14.428721174004194,
      "grad_norm": 0.14901891350746155,
      "learning_rate": 0.00021836696428244418,
      "loss": 0.4883,
      "num_input_tokens_seen": 17963128,
      "step": 27530
    },
    {
      "epoch": 14.431341719077569,
      "grad_norm": 0.13178427517414093,
      "learning_rate": 0.0002181780363381473,
      "loss": 0.4481,
      "num_input_tokens_seen": 17965688,
      "step": 27535
    },
    {
      "epoch": 14.433962264150944,
      "grad_norm": 0.16926202178001404,
      "learning_rate": 0.00021798916734779,
      "loss": 0.3266,
      "num_input_tokens_seen": 17968888,
      "step": 27540
    },
    {
      "epoch": 14.43658280922432,
      "grad_norm": 0.16151022911071777,
      "learning_rate": 0.00021780035735088132,
      "loss": 0.3941,
      "num_input_tokens_seen": 17972120,
      "step": 27545
    },
    {
      "epoch": 14.439203354297694,
      "grad_norm": 0.14081111550331116,
      "learning_rate": 0.0002176116063869184,
      "loss": 0.3652,
      "num_input_tokens_seen": 17975608,
      "step": 27550
    },
    {
      "epoch": 14.44182389937107,
      "grad_norm": 0.12490250170230865,
      "learning_rate": 0.00021742291449538582,
      "loss": 0.5321,
      "num_input_tokens_seen": 17979128,
      "step": 27555
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 0.1846802532672882,
      "learning_rate": 0.00021723428171575566,
      "loss": 0.4673,
      "num_input_tokens_seen": 17982456,
      "step": 27560
    },
    {
      "epoch": 14.44706498951782,
      "grad_norm": 0.10356488823890686,
      "learning_rate": 0.00021704570808748802,
      "loss": 0.4984,
      "num_input_tokens_seen": 17985720,
      "step": 27565
    },
    {
      "epoch": 14.449685534591195,
      "grad_norm": 0.12320739030838013,
      "learning_rate": 0.00021685719365003008,
      "loss": 0.4086,
      "num_input_tokens_seen": 17989272,
      "step": 27570
    },
    {
      "epoch": 14.45230607966457,
      "grad_norm": 0.14965221285820007,
      "learning_rate": 0.00021666873844281682,
      "loss": 0.5114,
      "num_input_tokens_seen": 17992408,
      "step": 27575
    },
    {
      "epoch": 14.454926624737945,
      "grad_norm": 0.0929509699344635,
      "learning_rate": 0.0002164803425052711,
      "loss": 0.4045,
      "num_input_tokens_seen": 17996568,
      "step": 27580
    },
    {
      "epoch": 14.45754716981132,
      "grad_norm": 0.11197733134031296,
      "learning_rate": 0.0002162920058768031,
      "loss": 0.3825,
      "num_input_tokens_seen": 17999832,
      "step": 27585
    },
    {
      "epoch": 14.460167714884696,
      "grad_norm": 0.10499437153339386,
      "learning_rate": 0.000216103728596811,
      "loss": 0.5266,
      "num_input_tokens_seen": 18003704,
      "step": 27590
    },
    {
      "epoch": 14.46278825995807,
      "grad_norm": 0.2249618023633957,
      "learning_rate": 0.00021591551070467975,
      "loss": 0.3689,
      "num_input_tokens_seen": 18006328,
      "step": 27595
    },
    {
      "epoch": 14.465408805031446,
      "grad_norm": 0.12489625066518784,
      "learning_rate": 0.0002157273522397829,
      "loss": 0.4523,
      "num_input_tokens_seen": 18010008,
      "step": 27600
    },
    {
      "epoch": 14.468029350104821,
      "grad_norm": 0.21163147687911987,
      "learning_rate": 0.0002155392532414806,
      "loss": 0.4956,
      "num_input_tokens_seen": 18013880,
      "step": 27605
    },
    {
      "epoch": 14.470649895178196,
      "grad_norm": 0.16059130430221558,
      "learning_rate": 0.00021535121374912126,
      "loss": 0.4188,
      "num_input_tokens_seen": 18016824,
      "step": 27610
    },
    {
      "epoch": 14.473270440251572,
      "grad_norm": 0.1584390550851822,
      "learning_rate": 0.00021516323380204073,
      "loss": 0.4129,
      "num_input_tokens_seen": 18020472,
      "step": 27615
    },
    {
      "epoch": 14.475890985324948,
      "grad_norm": 0.1728876382112503,
      "learning_rate": 0.00021497531343956188,
      "loss": 0.3598,
      "num_input_tokens_seen": 18023128,
      "step": 27620
    },
    {
      "epoch": 14.478511530398324,
      "grad_norm": 0.22253744304180145,
      "learning_rate": 0.00021478745270099591,
      "loss": 0.47,
      "num_input_tokens_seen": 18026872,
      "step": 27625
    },
    {
      "epoch": 14.481132075471699,
      "grad_norm": 0.11814632266759872,
      "learning_rate": 0.0002145996516256408,
      "loss": 0.5318,
      "num_input_tokens_seen": 18030200,
      "step": 27630
    },
    {
      "epoch": 14.483752620545074,
      "grad_norm": 0.13407692313194275,
      "learning_rate": 0.00021441191025278256,
      "loss": 0.325,
      "num_input_tokens_seen": 18033528,
      "step": 27635
    },
    {
      "epoch": 14.48637316561845,
      "grad_norm": 0.17637017369270325,
      "learning_rate": 0.00021422422862169455,
      "loss": 0.4335,
      "num_input_tokens_seen": 18036088,
      "step": 27640
    },
    {
      "epoch": 14.488993710691824,
      "grad_norm": 0.07861385494470596,
      "learning_rate": 0.00021403660677163773,
      "loss": 0.4405,
      "num_input_tokens_seen": 18040408,
      "step": 27645
    },
    {
      "epoch": 14.4916142557652,
      "grad_norm": 0.1293877363204956,
      "learning_rate": 0.00021384904474186018,
      "loss": 0.3386,
      "num_input_tokens_seen": 18043096,
      "step": 27650
    },
    {
      "epoch": 14.494234800838575,
      "grad_norm": 0.08635668456554413,
      "learning_rate": 0.00021366154257159808,
      "loss": 0.4441,
      "num_input_tokens_seen": 18046232,
      "step": 27655
    },
    {
      "epoch": 14.49685534591195,
      "grad_norm": 0.2911568880081177,
      "learning_rate": 0.00021347410030007435,
      "loss": 0.4572,
      "num_input_tokens_seen": 18048792,
      "step": 27660
    },
    {
      "epoch": 14.499475890985325,
      "grad_norm": 0.14551621675491333,
      "learning_rate": 0.0002132867179665,
      "loss": 0.4399,
      "num_input_tokens_seen": 18052696,
      "step": 27665
    },
    {
      "epoch": 14.5020964360587,
      "grad_norm": 0.09500988572835922,
      "learning_rate": 0.00021309939561007341,
      "loss": 0.3349,
      "num_input_tokens_seen": 18055256,
      "step": 27670
    },
    {
      "epoch": 14.504716981132075,
      "grad_norm": 0.1114833801984787,
      "learning_rate": 0.00021291213326997998,
      "loss": 0.4181,
      "num_input_tokens_seen": 18057784,
      "step": 27675
    },
    {
      "epoch": 14.50733752620545,
      "grad_norm": 0.18912391364574432,
      "learning_rate": 0.00021272493098539296,
      "loss": 0.5568,
      "num_input_tokens_seen": 18060920,
      "step": 27680
    },
    {
      "epoch": 14.509958071278826,
      "grad_norm": 0.11359734833240509,
      "learning_rate": 0.0002125377887954732,
      "loss": 0.4279,
      "num_input_tokens_seen": 18064184,
      "step": 27685
    },
    {
      "epoch": 14.5125786163522,
      "grad_norm": 0.15317191183567047,
      "learning_rate": 0.00021235070673936824,
      "loss": 0.5101,
      "num_input_tokens_seen": 18067672,
      "step": 27690
    },
    {
      "epoch": 14.515199161425576,
      "grad_norm": 0.20218458771705627,
      "learning_rate": 0.00021216368485621394,
      "loss": 0.4108,
      "num_input_tokens_seen": 18071032,
      "step": 27695
    },
    {
      "epoch": 14.517819706498951,
      "grad_norm": 0.19507066905498505,
      "learning_rate": 0.00021197672318513282,
      "loss": 0.671,
      "num_input_tokens_seen": 18074200,
      "step": 27700
    },
    {
      "epoch": 14.520440251572326,
      "grad_norm": 0.1604124754667282,
      "learning_rate": 0.00021178982176523525,
      "loss": 0.4102,
      "num_input_tokens_seen": 18076824,
      "step": 27705
    },
    {
      "epoch": 14.523060796645701,
      "grad_norm": 0.26696136593818665,
      "learning_rate": 0.0002116029806356189,
      "loss": 0.458,
      "num_input_tokens_seen": 18079512,
      "step": 27710
    },
    {
      "epoch": 14.525681341719078,
      "grad_norm": 0.15966109931468964,
      "learning_rate": 0.00021141619983536893,
      "loss": 0.4078,
      "num_input_tokens_seen": 18082424,
      "step": 27715
    },
    {
      "epoch": 14.528301886792454,
      "grad_norm": 0.07779937237501144,
      "learning_rate": 0.00021122947940355747,
      "loss": 0.3927,
      "num_input_tokens_seen": 18085400,
      "step": 27720
    },
    {
      "epoch": 14.530922431865829,
      "grad_norm": 0.154611274600029,
      "learning_rate": 0.00021104281937924462,
      "loss": 0.5814,
      "num_input_tokens_seen": 18088664,
      "step": 27725
    },
    {
      "epoch": 14.533542976939204,
      "grad_norm": 0.25420430302619934,
      "learning_rate": 0.00021085621980147716,
      "loss": 0.4301,
      "num_input_tokens_seen": 18091512,
      "step": 27730
    },
    {
      "epoch": 14.536163522012579,
      "grad_norm": 0.14919723570346832,
      "learning_rate": 0.00021066968070928982,
      "loss": 0.6106,
      "num_input_tokens_seen": 18093752,
      "step": 27735
    },
    {
      "epoch": 14.538784067085954,
      "grad_norm": 0.15795205533504486,
      "learning_rate": 0.00021048320214170463,
      "loss": 0.4995,
      "num_input_tokens_seen": 18096440,
      "step": 27740
    },
    {
      "epoch": 14.54140461215933,
      "grad_norm": 0.16224625706672668,
      "learning_rate": 0.00021029678413773034,
      "loss": 0.3567,
      "num_input_tokens_seen": 18099384,
      "step": 27745
    },
    {
      "epoch": 14.544025157232705,
      "grad_norm": 0.09494581818580627,
      "learning_rate": 0.0002101104267363639,
      "loss": 0.4003,
      "num_input_tokens_seen": 18102744,
      "step": 27750
    },
    {
      "epoch": 14.54664570230608,
      "grad_norm": 0.19528381526470184,
      "learning_rate": 0.00020992412997658877,
      "loss": 0.4727,
      "num_input_tokens_seen": 18105720,
      "step": 27755
    },
    {
      "epoch": 14.549266247379455,
      "grad_norm": 0.14139409363269806,
      "learning_rate": 0.0002097378938973763,
      "loss": 0.348,
      "num_input_tokens_seen": 18108536,
      "step": 27760
    },
    {
      "epoch": 14.55188679245283,
      "grad_norm": 0.3315965533256531,
      "learning_rate": 0.0002095517185376849,
      "loss": 0.4787,
      "num_input_tokens_seen": 18110904,
      "step": 27765
    },
    {
      "epoch": 14.554507337526205,
      "grad_norm": 0.13580477237701416,
      "learning_rate": 0.0002093656039364606,
      "loss": 0.3921,
      "num_input_tokens_seen": 18113944,
      "step": 27770
    },
    {
      "epoch": 14.55712788259958,
      "grad_norm": 0.08944575488567352,
      "learning_rate": 0.00020917955013263618,
      "loss": 0.469,
      "num_input_tokens_seen": 18117304,
      "step": 27775
    },
    {
      "epoch": 14.559748427672956,
      "grad_norm": 0.10786809027194977,
      "learning_rate": 0.00020899355716513186,
      "loss": 0.4926,
      "num_input_tokens_seen": 18119704,
      "step": 27780
    },
    {
      "epoch": 14.56236897274633,
      "grad_norm": 0.09452961385250092,
      "learning_rate": 0.00020880762507285544,
      "loss": 0.5208,
      "num_input_tokens_seen": 18124312,
      "step": 27785
    },
    {
      "epoch": 14.564989517819706,
      "grad_norm": 0.2021232694387436,
      "learning_rate": 0.00020862175389470172,
      "loss": 0.3929,
      "num_input_tokens_seen": 18128408,
      "step": 27790
    },
    {
      "epoch": 14.567610062893081,
      "grad_norm": 0.16524191200733185,
      "learning_rate": 0.00020843594366955288,
      "loss": 0.3652,
      "num_input_tokens_seen": 18131416,
      "step": 27795
    },
    {
      "epoch": 14.570230607966456,
      "grad_norm": 0.1146249920129776,
      "learning_rate": 0.0002082501944362784,
      "loss": 0.4895,
      "num_input_tokens_seen": 18134872,
      "step": 27800
    },
    {
      "epoch": 14.572851153039831,
      "grad_norm": 0.08757217228412628,
      "learning_rate": 0.00020806450623373486,
      "loss": 0.3076,
      "num_input_tokens_seen": 18138776,
      "step": 27805
    },
    {
      "epoch": 14.575471698113208,
      "grad_norm": 0.15409570932388306,
      "learning_rate": 0.00020787887910076586,
      "loss": 0.4591,
      "num_input_tokens_seen": 18141752,
      "step": 27810
    },
    {
      "epoch": 14.578092243186584,
      "grad_norm": 0.1602240353822708,
      "learning_rate": 0.00020769331307620258,
      "loss": 0.3184,
      "num_input_tokens_seen": 18144152,
      "step": 27815
    },
    {
      "epoch": 14.580712788259959,
      "grad_norm": 0.16604946553707123,
      "learning_rate": 0.0002075078081988635,
      "loss": 0.4383,
      "num_input_tokens_seen": 18147192,
      "step": 27820
    },
    {
      "epoch": 14.583333333333334,
      "grad_norm": 0.06678672879934311,
      "learning_rate": 0.00020732236450755415,
      "loss": 0.3873,
      "num_input_tokens_seen": 18150456,
      "step": 27825
    },
    {
      "epoch": 14.585953878406709,
      "grad_norm": 0.12842345237731934,
      "learning_rate": 0.00020713698204106708,
      "loss": 0.4103,
      "num_input_tokens_seen": 18153464,
      "step": 27830
    },
    {
      "epoch": 14.588574423480084,
      "grad_norm": 0.11949802190065384,
      "learning_rate": 0.00020695166083818206,
      "loss": 0.5051,
      "num_input_tokens_seen": 18157336,
      "step": 27835
    },
    {
      "epoch": 14.59119496855346,
      "grad_norm": 0.10988243669271469,
      "learning_rate": 0.0002067664009376663,
      "loss": 0.3511,
      "num_input_tokens_seen": 18161016,
      "step": 27840
    },
    {
      "epoch": 14.593815513626835,
      "grad_norm": 0.1232398971915245,
      "learning_rate": 0.0002065812023782741,
      "loss": 0.4972,
      "num_input_tokens_seen": 18163480,
      "step": 27845
    },
    {
      "epoch": 14.59643605870021,
      "grad_norm": 0.17293310165405273,
      "learning_rate": 0.00020639606519874703,
      "loss": 0.4087,
      "num_input_tokens_seen": 18166584,
      "step": 27850
    },
    {
      "epoch": 14.599056603773585,
      "grad_norm": 0.12689387798309326,
      "learning_rate": 0.00020621098943781352,
      "loss": 0.371,
      "num_input_tokens_seen": 18169144,
      "step": 27855
    },
    {
      "epoch": 14.60167714884696,
      "grad_norm": 0.1353386491537094,
      "learning_rate": 0.0002060259751341891,
      "loss": 0.4787,
      "num_input_tokens_seen": 18172184,
      "step": 27860
    },
    {
      "epoch": 14.604297693920335,
      "grad_norm": 0.12292983382940292,
      "learning_rate": 0.00020584102232657688,
      "loss": 0.4043,
      "num_input_tokens_seen": 18174840,
      "step": 27865
    },
    {
      "epoch": 14.60691823899371,
      "grad_norm": 0.11128430813550949,
      "learning_rate": 0.0002056561310536668,
      "loss": 0.353,
      "num_input_tokens_seen": 18178584,
      "step": 27870
    },
    {
      "epoch": 14.609538784067086,
      "grad_norm": 0.1698535531759262,
      "learning_rate": 0.0002054713013541361,
      "loss": 0.423,
      "num_input_tokens_seen": 18184792,
      "step": 27875
    },
    {
      "epoch": 14.61215932914046,
      "grad_norm": 0.07561512291431427,
      "learning_rate": 0.00020528653326664915,
      "loss": 0.4417,
      "num_input_tokens_seen": 18188248,
      "step": 27880
    },
    {
      "epoch": 14.614779874213836,
      "grad_norm": 0.12229868769645691,
      "learning_rate": 0.00020510182682985717,
      "loss": 0.3315,
      "num_input_tokens_seen": 18191160,
      "step": 27885
    },
    {
      "epoch": 14.617400419287211,
      "grad_norm": 0.1102495938539505,
      "learning_rate": 0.00020491718208239847,
      "loss": 0.346,
      "num_input_tokens_seen": 18195000,
      "step": 27890
    },
    {
      "epoch": 14.620020964360586,
      "grad_norm": 0.1728360652923584,
      "learning_rate": 0.00020473259906289877,
      "loss": 0.3484,
      "num_input_tokens_seen": 18198232,
      "step": 27895
    },
    {
      "epoch": 14.622641509433961,
      "grad_norm": 0.16866301000118256,
      "learning_rate": 0.00020454807780997077,
      "loss": 0.5374,
      "num_input_tokens_seen": 18200856,
      "step": 27900
    },
    {
      "epoch": 14.625262054507338,
      "grad_norm": 0.19154296815395355,
      "learning_rate": 0.0002043636183622144,
      "loss": 0.4136,
      "num_input_tokens_seen": 18203256,
      "step": 27905
    },
    {
      "epoch": 14.627882599580714,
      "grad_norm": 0.26388412714004517,
      "learning_rate": 0.00020417922075821626,
      "loss": 0.526,
      "num_input_tokens_seen": 18206008,
      "step": 27910
    },
    {
      "epoch": 14.630503144654089,
      "grad_norm": 0.13799405097961426,
      "learning_rate": 0.00020399488503655005,
      "loss": 0.4936,
      "num_input_tokens_seen": 18208984,
      "step": 27915
    },
    {
      "epoch": 14.633123689727464,
      "grad_norm": 0.1620902717113495,
      "learning_rate": 0.00020381061123577688,
      "loss": 0.3058,
      "num_input_tokens_seen": 18213848,
      "step": 27920
    },
    {
      "epoch": 14.635744234800839,
      "grad_norm": 0.15377824008464813,
      "learning_rate": 0.00020362639939444473,
      "loss": 0.4368,
      "num_input_tokens_seen": 18217496,
      "step": 27925
    },
    {
      "epoch": 14.638364779874214,
      "grad_norm": 0.2028040885925293,
      "learning_rate": 0.0002034422495510888,
      "loss": 0.4145,
      "num_input_tokens_seen": 18220536,
      "step": 27930
    },
    {
      "epoch": 14.64098532494759,
      "grad_norm": 0.21220791339874268,
      "learning_rate": 0.00020325816174423077,
      "loss": 0.5382,
      "num_input_tokens_seen": 18224888,
      "step": 27935
    },
    {
      "epoch": 14.643605870020965,
      "grad_norm": 0.06290535628795624,
      "learning_rate": 0.00020307413601238012,
      "loss": 0.3083,
      "num_input_tokens_seen": 18229368,
      "step": 27940
    },
    {
      "epoch": 14.64622641509434,
      "grad_norm": 0.13786379992961884,
      "learning_rate": 0.00020289017239403246,
      "loss": 0.4429,
      "num_input_tokens_seen": 18232760,
      "step": 27945
    },
    {
      "epoch": 14.648846960167715,
      "grad_norm": 0.11187241971492767,
      "learning_rate": 0.0002027062709276712,
      "loss": 0.4543,
      "num_input_tokens_seen": 18235576,
      "step": 27950
    },
    {
      "epoch": 14.65146750524109,
      "grad_norm": 0.09206999838352203,
      "learning_rate": 0.00020252243165176632,
      "loss": 0.4403,
      "num_input_tokens_seen": 18239224,
      "step": 27955
    },
    {
      "epoch": 14.654088050314465,
      "grad_norm": 0.17921002209186554,
      "learning_rate": 0.00020233865460477508,
      "loss": 0.4001,
      "num_input_tokens_seen": 18241912,
      "step": 27960
    },
    {
      "epoch": 14.65670859538784,
      "grad_norm": 0.10300428420305252,
      "learning_rate": 0.00020215493982514128,
      "loss": 0.365,
      "num_input_tokens_seen": 18244632,
      "step": 27965
    },
    {
      "epoch": 14.659329140461216,
      "grad_norm": 0.10921639949083328,
      "learning_rate": 0.00020197128735129622,
      "loss": 0.4279,
      "num_input_tokens_seen": 18247768,
      "step": 27970
    },
    {
      "epoch": 14.66194968553459,
      "grad_norm": 0.22145038843154907,
      "learning_rate": 0.0002017876972216575,
      "loss": 0.3793,
      "num_input_tokens_seen": 18250456,
      "step": 27975
    },
    {
      "epoch": 14.664570230607966,
      "grad_norm": 0.132878378033638,
      "learning_rate": 0.00020160416947463046,
      "loss": 0.4435,
      "num_input_tokens_seen": 18253688,
      "step": 27980
    },
    {
      "epoch": 14.667190775681341,
      "grad_norm": 0.09507710486650467,
      "learning_rate": 0.00020142070414860702,
      "loss": 0.3702,
      "num_input_tokens_seen": 18256952,
      "step": 27985
    },
    {
      "epoch": 14.669811320754716,
      "grad_norm": 0.10126682370901108,
      "learning_rate": 0.0002012373012819657,
      "loss": 0.3632,
      "num_input_tokens_seen": 18263064,
      "step": 27990
    },
    {
      "epoch": 14.672431865828091,
      "grad_norm": 0.17297354340553284,
      "learning_rate": 0.00020105396091307247,
      "loss": 0.3704,
      "num_input_tokens_seen": 18265848,
      "step": 27995
    },
    {
      "epoch": 14.675052410901468,
      "grad_norm": 0.1347460150718689,
      "learning_rate": 0.0002008706830802803,
      "loss": 0.5836,
      "num_input_tokens_seen": 18269208,
      "step": 28000
    },
    {
      "epoch": 14.677672955974844,
      "grad_norm": 0.09499268978834152,
      "learning_rate": 0.00020068746782192837,
      "loss": 0.402,
      "num_input_tokens_seen": 18274136,
      "step": 28005
    },
    {
      "epoch": 14.680293501048219,
      "grad_norm": 0.08147899806499481,
      "learning_rate": 0.00020050431517634366,
      "loss": 0.3425,
      "num_input_tokens_seen": 18277016,
      "step": 28010
    },
    {
      "epoch": 14.682914046121594,
      "grad_norm": 0.1360003650188446,
      "learning_rate": 0.00020032122518183915,
      "loss": 0.3709,
      "num_input_tokens_seen": 18279640,
      "step": 28015
    },
    {
      "epoch": 14.685534591194969,
      "grad_norm": 0.2162550836801529,
      "learning_rate": 0.00020013819787671538,
      "loss": 0.3969,
      "num_input_tokens_seen": 18281816,
      "step": 28020
    },
    {
      "epoch": 14.688155136268344,
      "grad_norm": 0.10590807348489761,
      "learning_rate": 0.00019995523329925985,
      "loss": 0.5475,
      "num_input_tokens_seen": 18284568,
      "step": 28025
    },
    {
      "epoch": 14.69077568134172,
      "grad_norm": 0.10506623983383179,
      "learning_rate": 0.00019977233148774627,
      "loss": 0.4128,
      "num_input_tokens_seen": 18288792,
      "step": 28030
    },
    {
      "epoch": 14.693396226415095,
      "grad_norm": 0.13670983910560608,
      "learning_rate": 0.00019958949248043573,
      "loss": 0.4173,
      "num_input_tokens_seen": 18292344,
      "step": 28035
    },
    {
      "epoch": 14.69601677148847,
      "grad_norm": 0.12382956594228745,
      "learning_rate": 0.00019940671631557628,
      "loss": 0.4024,
      "num_input_tokens_seen": 18296184,
      "step": 28040
    },
    {
      "epoch": 14.698637316561845,
      "grad_norm": 0.1441928744316101,
      "learning_rate": 0.0001992240030314022,
      "loss": 0.4967,
      "num_input_tokens_seen": 18299960,
      "step": 28045
    },
    {
      "epoch": 14.70125786163522,
      "grad_norm": 0.1157400980591774,
      "learning_rate": 0.00019904135266613532,
      "loss": 0.4293,
      "num_input_tokens_seen": 18304024,
      "step": 28050
    },
    {
      "epoch": 14.703878406708595,
      "grad_norm": 0.10004269331693649,
      "learning_rate": 0.00019885876525798407,
      "loss": 0.4687,
      "num_input_tokens_seen": 18306808,
      "step": 28055
    },
    {
      "epoch": 14.70649895178197,
      "grad_norm": 0.31737130880355835,
      "learning_rate": 0.0001986762408451434,
      "loss": 0.3956,
      "num_input_tokens_seen": 18308984,
      "step": 28060
    },
    {
      "epoch": 14.709119496855346,
      "grad_norm": 0.18423685431480408,
      "learning_rate": 0.00019849377946579562,
      "loss": 0.4069,
      "num_input_tokens_seen": 18311416,
      "step": 28065
    },
    {
      "epoch": 14.71174004192872,
      "grad_norm": 0.21906977891921997,
      "learning_rate": 0.00019831138115810926,
      "loss": 0.5485,
      "num_input_tokens_seen": 18314552,
      "step": 28070
    },
    {
      "epoch": 14.714360587002096,
      "grad_norm": 0.1101025938987732,
      "learning_rate": 0.00019812904596024005,
      "loss": 0.355,
      "num_input_tokens_seen": 18317752,
      "step": 28075
    },
    {
      "epoch": 14.716981132075471,
      "grad_norm": 0.11281199753284454,
      "learning_rate": 0.00019794677391033055,
      "loss": 0.4653,
      "num_input_tokens_seen": 18320888,
      "step": 28080
    },
    {
      "epoch": 14.719601677148846,
      "grad_norm": 0.15147750079631805,
      "learning_rate": 0.00019776456504651014,
      "loss": 0.585,
      "num_input_tokens_seen": 18324600,
      "step": 28085
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 0.14195804297924042,
      "learning_rate": 0.0001975824194068946,
      "loss": 0.3826,
      "num_input_tokens_seen": 18328184,
      "step": 28090
    },
    {
      "epoch": 14.724842767295598,
      "grad_norm": 0.14058047533035278,
      "learning_rate": 0.00019740033702958653,
      "loss": 0.3754,
      "num_input_tokens_seen": 18330680,
      "step": 28095
    },
    {
      "epoch": 14.727463312368974,
      "grad_norm": 0.16191555559635162,
      "learning_rate": 0.0001972183179526758,
      "loss": 0.4737,
      "num_input_tokens_seen": 18334552,
      "step": 28100
    },
    {
      "epoch": 14.730083857442349,
      "grad_norm": 0.15690481662750244,
      "learning_rate": 0.0001970363622142386,
      "loss": 0.4519,
      "num_input_tokens_seen": 18338168,
      "step": 28105
    },
    {
      "epoch": 14.732704402515724,
      "grad_norm": 0.08982089161872864,
      "learning_rate": 0.00019685446985233824,
      "loss": 0.3977,
      "num_input_tokens_seen": 18341528,
      "step": 28110
    },
    {
      "epoch": 14.735324947589099,
      "grad_norm": 0.1174715906381607,
      "learning_rate": 0.00019667264090502412,
      "loss": 0.4254,
      "num_input_tokens_seen": 18344920,
      "step": 28115
    },
    {
      "epoch": 14.737945492662474,
      "grad_norm": 0.08734872937202454,
      "learning_rate": 0.0001964908754103331,
      "loss": 0.337,
      "num_input_tokens_seen": 18348312,
      "step": 28120
    },
    {
      "epoch": 14.74056603773585,
      "grad_norm": 0.1232527643442154,
      "learning_rate": 0.00019630917340628822,
      "loss": 0.3855,
      "num_input_tokens_seen": 18351160,
      "step": 28125
    },
    {
      "epoch": 14.743186582809225,
      "grad_norm": 0.32258135080337524,
      "learning_rate": 0.00019612753493089959,
      "loss": 0.5121,
      "num_input_tokens_seen": 18354104,
      "step": 28130
    },
    {
      "epoch": 14.7458071278826,
      "grad_norm": 0.1789444088935852,
      "learning_rate": 0.00019594596002216387,
      "loss": 0.42,
      "num_input_tokens_seen": 18357784,
      "step": 28135
    },
    {
      "epoch": 14.748427672955975,
      "grad_norm": 0.14756283164024353,
      "learning_rate": 0.00019576444871806465,
      "loss": 0.5088,
      "num_input_tokens_seen": 18360984,
      "step": 28140
    },
    {
      "epoch": 14.75104821802935,
      "grad_norm": 0.11598232388496399,
      "learning_rate": 0.0001955830010565719,
      "loss": 0.4218,
      "num_input_tokens_seen": 18365080,
      "step": 28145
    },
    {
      "epoch": 14.753668763102725,
      "grad_norm": 0.15114271640777588,
      "learning_rate": 0.00019540161707564213,
      "loss": 0.4746,
      "num_input_tokens_seen": 18367800,
      "step": 28150
    },
    {
      "epoch": 14.7562893081761,
      "grad_norm": 0.2934889793395996,
      "learning_rate": 0.000195220296813219,
      "loss": 0.6041,
      "num_input_tokens_seen": 18370392,
      "step": 28155
    },
    {
      "epoch": 14.758909853249476,
      "grad_norm": 0.08947190642356873,
      "learning_rate": 0.00019503904030723274,
      "loss": 0.4578,
      "num_input_tokens_seen": 18373560,
      "step": 28160
    },
    {
      "epoch": 14.76153039832285,
      "grad_norm": 0.14664055407047272,
      "learning_rate": 0.00019485784759560004,
      "loss": 0.3485,
      "num_input_tokens_seen": 18378296,
      "step": 28165
    },
    {
      "epoch": 14.764150943396226,
      "grad_norm": 0.15747885406017303,
      "learning_rate": 0.0001946767187162247,
      "loss": 0.441,
      "num_input_tokens_seen": 18381016,
      "step": 28170
    },
    {
      "epoch": 14.766771488469601,
      "grad_norm": 0.12495668977499008,
      "learning_rate": 0.00019449565370699618,
      "loss": 0.4791,
      "num_input_tokens_seen": 18384248,
      "step": 28175
    },
    {
      "epoch": 14.769392033542976,
      "grad_norm": 0.392414927482605,
      "learning_rate": 0.00019431465260579152,
      "loss": 0.5008,
      "num_input_tokens_seen": 18386648,
      "step": 28180
    },
    {
      "epoch": 14.772012578616351,
      "grad_norm": 0.11998410522937775,
      "learning_rate": 0.00019413371545047404,
      "loss": 0.3914,
      "num_input_tokens_seen": 18390232,
      "step": 28185
    },
    {
      "epoch": 14.774633123689728,
      "grad_norm": 0.14074839651584625,
      "learning_rate": 0.00019395284227889382,
      "loss": 0.4522,
      "num_input_tokens_seen": 18393624,
      "step": 28190
    },
    {
      "epoch": 14.777253668763104,
      "grad_norm": 0.11712069809436798,
      "learning_rate": 0.00019377203312888764,
      "loss": 0.3622,
      "num_input_tokens_seen": 18397176,
      "step": 28195
    },
    {
      "epoch": 14.779874213836479,
      "grad_norm": 0.13888780772686005,
      "learning_rate": 0.0001935912880382784,
      "loss": 0.4624,
      "num_input_tokens_seen": 18400472,
      "step": 28200
    },
    {
      "epoch": 14.782494758909854,
      "grad_norm": 0.13950634002685547,
      "learning_rate": 0.00019341060704487596,
      "loss": 0.4039,
      "num_input_tokens_seen": 18403224,
      "step": 28205
    },
    {
      "epoch": 14.785115303983229,
      "grad_norm": 0.1270095854997635,
      "learning_rate": 0.00019322999018647668,
      "loss": 0.311,
      "num_input_tokens_seen": 18406200,
      "step": 28210
    },
    {
      "epoch": 14.787735849056604,
      "grad_norm": 0.18995995819568634,
      "learning_rate": 0.00019304943750086362,
      "loss": 0.4322,
      "num_input_tokens_seen": 18409432,
      "step": 28215
    },
    {
      "epoch": 14.79035639412998,
      "grad_norm": 0.11194268614053726,
      "learning_rate": 0.0001928689490258066,
      "loss": 0.3906,
      "num_input_tokens_seen": 18411736,
      "step": 28220
    },
    {
      "epoch": 14.792976939203355,
      "grad_norm": 0.10522574931383133,
      "learning_rate": 0.00019268852479906146,
      "loss": 0.4403,
      "num_input_tokens_seen": 18415448,
      "step": 28225
    },
    {
      "epoch": 14.79559748427673,
      "grad_norm": 0.10841953754425049,
      "learning_rate": 0.0001925081648583708,
      "loss": 0.5206,
      "num_input_tokens_seen": 18419032,
      "step": 28230
    },
    {
      "epoch": 14.798218029350105,
      "grad_norm": 0.13009676337242126,
      "learning_rate": 0.00019232786924146394,
      "loss": 0.4659,
      "num_input_tokens_seen": 18422136,
      "step": 28235
    },
    {
      "epoch": 14.80083857442348,
      "grad_norm": 0.15079016983509064,
      "learning_rate": 0.00019214763798605677,
      "loss": 0.5776,
      "num_input_tokens_seen": 18424696,
      "step": 28240
    },
    {
      "epoch": 14.803459119496855,
      "grad_norm": 0.14446842670440674,
      "learning_rate": 0.00019196747112985158,
      "loss": 0.5047,
      "num_input_tokens_seen": 18427704,
      "step": 28245
    },
    {
      "epoch": 14.80607966457023,
      "grad_norm": 0.16567620635032654,
      "learning_rate": 0.00019178736871053737,
      "loss": 0.3606,
      "num_input_tokens_seen": 18430328,
      "step": 28250
    },
    {
      "epoch": 14.808700209643606,
      "grad_norm": 0.2456701248884201,
      "learning_rate": 0.00019160733076578935,
      "loss": 0.513,
      "num_input_tokens_seen": 18432920,
      "step": 28255
    },
    {
      "epoch": 14.81132075471698,
      "grad_norm": 0.16563309729099274,
      "learning_rate": 0.0001914273573332692,
      "loss": 0.547,
      "num_input_tokens_seen": 18435896,
      "step": 28260
    },
    {
      "epoch": 14.813941299790356,
      "grad_norm": 0.18553322553634644,
      "learning_rate": 0.00019124744845062553,
      "loss": 0.4331,
      "num_input_tokens_seen": 18438712,
      "step": 28265
    },
    {
      "epoch": 14.816561844863731,
      "grad_norm": 0.16481846570968628,
      "learning_rate": 0.00019106760415549324,
      "loss": 0.45,
      "num_input_tokens_seen": 18441592,
      "step": 28270
    },
    {
      "epoch": 14.819182389937106,
      "grad_norm": 0.13009564578533173,
      "learning_rate": 0.0001908878244854938,
      "loss": 0.3922,
      "num_input_tokens_seen": 18444120,
      "step": 28275
    },
    {
      "epoch": 14.821802935010481,
      "grad_norm": 0.1611488312482834,
      "learning_rate": 0.00019070810947823498,
      "loss": 0.3579,
      "num_input_tokens_seen": 18447640,
      "step": 28280
    },
    {
      "epoch": 14.824423480083858,
      "grad_norm": 0.053402770310640335,
      "learning_rate": 0.0001905284591713109,
      "loss": 0.4888,
      "num_input_tokens_seen": 18452312,
      "step": 28285
    },
    {
      "epoch": 14.827044025157234,
      "grad_norm": 0.11704213917255402,
      "learning_rate": 0.00019034887360230246,
      "loss": 0.4001,
      "num_input_tokens_seen": 18455704,
      "step": 28290
    },
    {
      "epoch": 14.829664570230609,
      "grad_norm": 0.10315924137830734,
      "learning_rate": 0.000190169352808777,
      "loss": 0.3922,
      "num_input_tokens_seen": 18459384,
      "step": 28295
    },
    {
      "epoch": 14.832285115303984,
      "grad_norm": 0.15914741158485413,
      "learning_rate": 0.00018998989682828844,
      "loss": 0.5439,
      "num_input_tokens_seen": 18463512,
      "step": 28300
    },
    {
      "epoch": 14.834905660377359,
      "grad_norm": 0.23156510293483734,
      "learning_rate": 0.00018981050569837643,
      "loss": 0.5382,
      "num_input_tokens_seen": 18466520,
      "step": 28305
    },
    {
      "epoch": 14.837526205450734,
      "grad_norm": 0.11727398633956909,
      "learning_rate": 0.000189631179456568,
      "loss": 0.3706,
      "num_input_tokens_seen": 18469240,
      "step": 28310
    },
    {
      "epoch": 14.84014675052411,
      "grad_norm": 0.09502197802066803,
      "learning_rate": 0.00018945191814037578,
      "loss": 0.4329,
      "num_input_tokens_seen": 18473272,
      "step": 28315
    },
    {
      "epoch": 14.842767295597485,
      "grad_norm": 0.09947096556425095,
      "learning_rate": 0.00018927272178729938,
      "loss": 0.4565,
      "num_input_tokens_seen": 18478392,
      "step": 28320
    },
    {
      "epoch": 14.84538784067086,
      "grad_norm": 0.12374579161405563,
      "learning_rate": 0.00018909359043482477,
      "loss": 0.3881,
      "num_input_tokens_seen": 18481048,
      "step": 28325
    },
    {
      "epoch": 14.848008385744235,
      "grad_norm": 0.1338832825422287,
      "learning_rate": 0.00018891452412042386,
      "loss": 0.5028,
      "num_input_tokens_seen": 18484376,
      "step": 28330
    },
    {
      "epoch": 14.85062893081761,
      "grad_norm": 0.10448021441698074,
      "learning_rate": 0.00018873552288155542,
      "loss": 0.3035,
      "num_input_tokens_seen": 18487928,
      "step": 28335
    },
    {
      "epoch": 14.853249475890985,
      "grad_norm": 0.10640571266412735,
      "learning_rate": 0.00018855658675566471,
      "loss": 0.5628,
      "num_input_tokens_seen": 18491192,
      "step": 28340
    },
    {
      "epoch": 14.85587002096436,
      "grad_norm": 0.22769396007061005,
      "learning_rate": 0.00018837771578018275,
      "loss": 0.3837,
      "num_input_tokens_seen": 18494744,
      "step": 28345
    },
    {
      "epoch": 14.858490566037736,
      "grad_norm": 0.11048311740159988,
      "learning_rate": 0.00018819890999252747,
      "loss": 0.4937,
      "num_input_tokens_seen": 18498584,
      "step": 28350
    },
    {
      "epoch": 14.86111111111111,
      "grad_norm": 0.14332108199596405,
      "learning_rate": 0.0001880201694301032,
      "loss": 0.4963,
      "num_input_tokens_seen": 18501400,
      "step": 28355
    },
    {
      "epoch": 14.863731656184486,
      "grad_norm": 0.19385536015033722,
      "learning_rate": 0.00018784149413030005,
      "loss": 0.4233,
      "num_input_tokens_seen": 18503864,
      "step": 28360
    },
    {
      "epoch": 14.866352201257861,
      "grad_norm": 0.12629041075706482,
      "learning_rate": 0.00018766288413049503,
      "loss": 0.3782,
      "num_input_tokens_seen": 18506712,
      "step": 28365
    },
    {
      "epoch": 14.868972746331236,
      "grad_norm": 0.15711979568004608,
      "learning_rate": 0.00018748433946805148,
      "loss": 0.4684,
      "num_input_tokens_seen": 18510360,
      "step": 28370
    },
    {
      "epoch": 14.871593291404611,
      "grad_norm": 0.10129322111606598,
      "learning_rate": 0.00018730586018031854,
      "loss": 0.4308,
      "num_input_tokens_seen": 18513432,
      "step": 28375
    },
    {
      "epoch": 14.874213836477988,
      "grad_norm": 0.13089972734451294,
      "learning_rate": 0.00018712744630463248,
      "loss": 0.4509,
      "num_input_tokens_seen": 18516632,
      "step": 28380
    },
    {
      "epoch": 14.876834381551364,
      "grad_norm": 0.09797253459692001,
      "learning_rate": 0.000186949097878315,
      "loss": 0.5006,
      "num_input_tokens_seen": 18521624,
      "step": 28385
    },
    {
      "epoch": 14.879454926624739,
      "grad_norm": 0.10312981903553009,
      "learning_rate": 0.0001867708149386748,
      "loss": 0.4871,
      "num_input_tokens_seen": 18524760,
      "step": 28390
    },
    {
      "epoch": 14.882075471698114,
      "grad_norm": 0.22922493517398834,
      "learning_rate": 0.0001865925975230068,
      "loss": 0.3779,
      "num_input_tokens_seen": 18527416,
      "step": 28395
    },
    {
      "epoch": 14.884696016771489,
      "grad_norm": 0.13904350996017456,
      "learning_rate": 0.0001864144456685916,
      "loss": 0.4869,
      "num_input_tokens_seen": 18531032,
      "step": 28400
    },
    {
      "epoch": 14.887316561844864,
      "grad_norm": 0.09475377947092056,
      "learning_rate": 0.00018623635941269706,
      "loss": 0.4433,
      "num_input_tokens_seen": 18534744,
      "step": 28405
    },
    {
      "epoch": 14.88993710691824,
      "grad_norm": 0.1383856236934662,
      "learning_rate": 0.00018605833879257628,
      "loss": 0.45,
      "num_input_tokens_seen": 18538392,
      "step": 28410
    },
    {
      "epoch": 14.892557651991615,
      "grad_norm": 0.1786930114030838,
      "learning_rate": 0.0001858803838454694,
      "loss": 0.3744,
      "num_input_tokens_seen": 18541400,
      "step": 28415
    },
    {
      "epoch": 14.89517819706499,
      "grad_norm": 0.49855583906173706,
      "learning_rate": 0.0001857024946086026,
      "loss": 0.4207,
      "num_input_tokens_seen": 18544696,
      "step": 28420
    },
    {
      "epoch": 14.897798742138365,
      "grad_norm": 0.08453790843486786,
      "learning_rate": 0.00018552467111918836,
      "loss": 0.3215,
      "num_input_tokens_seen": 18547864,
      "step": 28425
    },
    {
      "epoch": 14.90041928721174,
      "grad_norm": 0.1426129788160324,
      "learning_rate": 0.00018534691341442507,
      "loss": 0.4216,
      "num_input_tokens_seen": 18550584,
      "step": 28430
    },
    {
      "epoch": 14.903039832285115,
      "grad_norm": 0.269414484500885,
      "learning_rate": 0.0001851692215314979,
      "loss": 0.5193,
      "num_input_tokens_seen": 18553624,
      "step": 28435
    },
    {
      "epoch": 14.90566037735849,
      "grad_norm": 0.17905959486961365,
      "learning_rate": 0.0001849915955075776,
      "loss": 0.5083,
      "num_input_tokens_seen": 18556472,
      "step": 28440
    },
    {
      "epoch": 14.908280922431866,
      "grad_norm": 0.13470299541950226,
      "learning_rate": 0.0001848140353798217,
      "loss": 0.3134,
      "num_input_tokens_seen": 18558840,
      "step": 28445
    },
    {
      "epoch": 14.91090146750524,
      "grad_norm": 0.1736757755279541,
      "learning_rate": 0.00018463654118537382,
      "loss": 0.421,
      "num_input_tokens_seen": 18562008,
      "step": 28450
    },
    {
      "epoch": 14.913522012578616,
      "grad_norm": 0.11875315755605698,
      "learning_rate": 0.00018445911296136386,
      "loss": 0.5235,
      "num_input_tokens_seen": 18564952,
      "step": 28455
    },
    {
      "epoch": 14.916142557651991,
      "grad_norm": 0.12814514338970184,
      "learning_rate": 0.00018428175074490754,
      "loss": 0.4262,
      "num_input_tokens_seen": 18567800,
      "step": 28460
    },
    {
      "epoch": 14.918763102725366,
      "grad_norm": 0.16568681597709656,
      "learning_rate": 0.00018410445457310687,
      "loss": 0.5019,
      "num_input_tokens_seen": 18571096,
      "step": 28465
    },
    {
      "epoch": 14.921383647798741,
      "grad_norm": 0.10459603369235992,
      "learning_rate": 0.00018392722448305044,
      "loss": 0.3308,
      "num_input_tokens_seen": 18574136,
      "step": 28470
    },
    {
      "epoch": 14.924004192872118,
      "grad_norm": 0.2655940353870392,
      "learning_rate": 0.00018375006051181265,
      "loss": 0.4541,
      "num_input_tokens_seen": 18576664,
      "step": 28475
    },
    {
      "epoch": 14.926624737945493,
      "grad_norm": 0.10463084280490875,
      "learning_rate": 0.0001835729626964544,
      "loss": 0.4104,
      "num_input_tokens_seen": 18580824,
      "step": 28480
    },
    {
      "epoch": 14.929245283018869,
      "grad_norm": 0.14010567963123322,
      "learning_rate": 0.00018339593107402242,
      "loss": 0.3758,
      "num_input_tokens_seen": 18583864,
      "step": 28485
    },
    {
      "epoch": 14.931865828092244,
      "grad_norm": 0.18558554351329803,
      "learning_rate": 0.00018321896568154945,
      "loss": 0.419,
      "num_input_tokens_seen": 18586680,
      "step": 28490
    },
    {
      "epoch": 14.934486373165619,
      "grad_norm": 0.07148153334856033,
      "learning_rate": 0.00018304206655605476,
      "loss": 0.338,
      "num_input_tokens_seen": 18590776,
      "step": 28495
    },
    {
      "epoch": 14.937106918238994,
      "grad_norm": 0.16480474174022675,
      "learning_rate": 0.00018286523373454378,
      "loss": 0.3391,
      "num_input_tokens_seen": 18594616,
      "step": 28500
    },
    {
      "epoch": 14.93972746331237,
      "grad_norm": 0.11775679886341095,
      "learning_rate": 0.00018268846725400783,
      "loss": 0.5165,
      "num_input_tokens_seen": 18597528,
      "step": 28505
    },
    {
      "epoch": 14.942348008385745,
      "grad_norm": 0.16984373331069946,
      "learning_rate": 0.00018251176715142458,
      "loss": 0.2367,
      "num_input_tokens_seen": 18599992,
      "step": 28510
    },
    {
      "epoch": 14.94496855345912,
      "grad_norm": 0.23424464464187622,
      "learning_rate": 0.00018233513346375756,
      "loss": 0.3931,
      "num_input_tokens_seen": 18602872,
      "step": 28515
    },
    {
      "epoch": 14.947589098532495,
      "grad_norm": 0.10904381424188614,
      "learning_rate": 0.00018215856622795634,
      "loss": 0.4069,
      "num_input_tokens_seen": 18606360,
      "step": 28520
    },
    {
      "epoch": 14.95020964360587,
      "grad_norm": 0.14368946850299835,
      "learning_rate": 0.00018198206548095697,
      "loss": 0.3629,
      "num_input_tokens_seen": 18608824,
      "step": 28525
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 0.07976909726858139,
      "learning_rate": 0.00018180563125968135,
      "loss": 0.4171,
      "num_input_tokens_seen": 18611864,
      "step": 28530
    },
    {
      "epoch": 14.95545073375262,
      "grad_norm": 0.0875006914138794,
      "learning_rate": 0.00018162926360103766,
      "loss": 0.5268,
      "num_input_tokens_seen": 18615480,
      "step": 28535
    },
    {
      "epoch": 14.958071278825996,
      "grad_norm": 0.12641891837120056,
      "learning_rate": 0.0001814529625419199,
      "loss": 0.4742,
      "num_input_tokens_seen": 18618392,
      "step": 28540
    },
    {
      "epoch": 14.96069182389937,
      "grad_norm": 0.17059749364852905,
      "learning_rate": 0.00018127672811920798,
      "loss": 0.5148,
      "num_input_tokens_seen": 18622200,
      "step": 28545
    },
    {
      "epoch": 14.963312368972746,
      "grad_norm": 0.08987563848495483,
      "learning_rate": 0.00018110056036976834,
      "loss": 0.4374,
      "num_input_tokens_seen": 18625080,
      "step": 28550
    },
    {
      "epoch": 14.965932914046121,
      "grad_norm": 0.13279829919338226,
      "learning_rate": 0.00018092445933045332,
      "loss": 0.4768,
      "num_input_tokens_seen": 18627992,
      "step": 28555
    },
    {
      "epoch": 14.968553459119496,
      "grad_norm": 0.12992343306541443,
      "learning_rate": 0.00018074842503810118,
      "loss": 0.4182,
      "num_input_tokens_seen": 18631192,
      "step": 28560
    },
    {
      "epoch": 14.971174004192871,
      "grad_norm": 0.10877405852079391,
      "learning_rate": 0.00018057245752953666,
      "loss": 0.3927,
      "num_input_tokens_seen": 18634904,
      "step": 28565
    },
    {
      "epoch": 14.973794549266248,
      "grad_norm": 0.17023244500160217,
      "learning_rate": 0.00018039655684156948,
      "loss": 0.4503,
      "num_input_tokens_seen": 18637784,
      "step": 28570
    },
    {
      "epoch": 14.976415094339622,
      "grad_norm": 0.3357425630092621,
      "learning_rate": 0.00018022072301099635,
      "loss": 0.4223,
      "num_input_tokens_seen": 18640120,
      "step": 28575
    },
    {
      "epoch": 14.979035639412999,
      "grad_norm": 0.21173331141471863,
      "learning_rate": 0.00018004495607459975,
      "loss": 0.5831,
      "num_input_tokens_seen": 18643672,
      "step": 28580
    },
    {
      "epoch": 14.981656184486374,
      "grad_norm": 0.1617693305015564,
      "learning_rate": 0.00017986925606914805,
      "loss": 0.5325,
      "num_input_tokens_seen": 18646872,
      "step": 28585
    },
    {
      "epoch": 14.984276729559749,
      "grad_norm": 0.17308884859085083,
      "learning_rate": 0.00017969362303139598,
      "loss": 0.3273,
      "num_input_tokens_seen": 18650744,
      "step": 28590
    },
    {
      "epoch": 14.986897274633124,
      "grad_norm": 0.14660851657390594,
      "learning_rate": 0.00017951805699808365,
      "loss": 0.437,
      "num_input_tokens_seen": 18654296,
      "step": 28595
    },
    {
      "epoch": 14.9895178197065,
      "grad_norm": 0.1781870275735855,
      "learning_rate": 0.00017934255800593747,
      "loss": 0.4013,
      "num_input_tokens_seen": 18657560,
      "step": 28600
    },
    {
      "epoch": 14.992138364779874,
      "grad_norm": 0.1360103040933609,
      "learning_rate": 0.0001791671260916698,
      "loss": 0.4373,
      "num_input_tokens_seen": 18660728,
      "step": 28605
    },
    {
      "epoch": 14.99475890985325,
      "grad_norm": 0.15372170507907867,
      "learning_rate": 0.00017899176129197914,
      "loss": 0.413,
      "num_input_tokens_seen": 18663352,
      "step": 28610
    },
    {
      "epoch": 14.997379454926625,
      "grad_norm": 0.26496344804763794,
      "learning_rate": 0.00017881646364354998,
      "loss": 0.4405,
      "num_input_tokens_seen": 18665816,
      "step": 28615
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.2532367408275604,
      "learning_rate": 0.00017864123318305213,
      "loss": 0.3276,
      "num_input_tokens_seen": 18668536,
      "step": 28620
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.4626333713531494,
      "eval_runtime": 13.5541,
      "eval_samples_per_second": 62.564,
      "eval_steps_per_second": 15.641,
      "num_input_tokens_seen": 18668536,
      "step": 28620
    },
    {
      "epoch": 15.002620545073375,
      "grad_norm": 0.13385237753391266,
      "learning_rate": 0.00017846606994714222,
      "loss": 0.6707,
      "num_input_tokens_seen": 18672024,
      "step": 28625
    },
    {
      "epoch": 15.00524109014675,
      "grad_norm": 0.13367651402950287,
      "learning_rate": 0.0001782909739724621,
      "loss": 0.5198,
      "num_input_tokens_seen": 18674712,
      "step": 28630
    },
    {
      "epoch": 15.007861635220126,
      "grad_norm": 0.1267087459564209,
      "learning_rate": 0.0001781159452956399,
      "loss": 0.4104,
      "num_input_tokens_seen": 18682488,
      "step": 28635
    },
    {
      "epoch": 15.0104821802935,
      "grad_norm": 0.09600579738616943,
      "learning_rate": 0.00017794098395328977,
      "loss": 0.4259,
      "num_input_tokens_seen": 18685816,
      "step": 28640
    },
    {
      "epoch": 15.013102725366876,
      "grad_norm": 0.1568574160337448,
      "learning_rate": 0.00017776608998201172,
      "loss": 0.4407,
      "num_input_tokens_seen": 18688760,
      "step": 28645
    },
    {
      "epoch": 15.015723270440251,
      "grad_norm": 0.10495852679014206,
      "learning_rate": 0.00017759126341839132,
      "loss": 0.3249,
      "num_input_tokens_seen": 18691160,
      "step": 28650
    },
    {
      "epoch": 15.018343815513626,
      "grad_norm": 0.14689108729362488,
      "learning_rate": 0.0001774165042990003,
      "loss": 0.4684,
      "num_input_tokens_seen": 18694616,
      "step": 28655
    },
    {
      "epoch": 15.020964360587001,
      "grad_norm": 0.09991160780191422,
      "learning_rate": 0.00017724181266039635,
      "loss": 0.3643,
      "num_input_tokens_seen": 18697656,
      "step": 28660
    },
    {
      "epoch": 15.023584905660377,
      "grad_norm": 0.11971664428710938,
      "learning_rate": 0.00017706718853912292,
      "loss": 0.3651,
      "num_input_tokens_seen": 18700536,
      "step": 28665
    },
    {
      "epoch": 15.026205450733753,
      "grad_norm": 0.2281801700592041,
      "learning_rate": 0.00017689263197170964,
      "loss": 0.5224,
      "num_input_tokens_seen": 18703960,
      "step": 28670
    },
    {
      "epoch": 15.028825995807129,
      "grad_norm": 0.17912322282791138,
      "learning_rate": 0.00017671814299467136,
      "loss": 0.3571,
      "num_input_tokens_seen": 18706744,
      "step": 28675
    },
    {
      "epoch": 15.031446540880504,
      "grad_norm": 0.0833624005317688,
      "learning_rate": 0.00017654372164450954,
      "loss": 0.4294,
      "num_input_tokens_seen": 18710424,
      "step": 28680
    },
    {
      "epoch": 15.034067085953879,
      "grad_norm": 0.0891890898346901,
      "learning_rate": 0.0001763693679577108,
      "loss": 0.5382,
      "num_input_tokens_seen": 18715640,
      "step": 28685
    },
    {
      "epoch": 15.036687631027254,
      "grad_norm": 0.08414801955223083,
      "learning_rate": 0.00017619508197074812,
      "loss": 0.5208,
      "num_input_tokens_seen": 18718680,
      "step": 28690
    },
    {
      "epoch": 15.03930817610063,
      "grad_norm": 0.09282302856445312,
      "learning_rate": 0.0001760208637200803,
      "loss": 0.4411,
      "num_input_tokens_seen": 18721688,
      "step": 28695
    },
    {
      "epoch": 15.041928721174004,
      "grad_norm": 0.1300269067287445,
      "learning_rate": 0.00017584671324215152,
      "loss": 0.3974,
      "num_input_tokens_seen": 18724920,
      "step": 28700
    },
    {
      "epoch": 15.04454926624738,
      "grad_norm": 0.3095894157886505,
      "learning_rate": 0.00017567263057339212,
      "loss": 0.566,
      "num_input_tokens_seen": 18728376,
      "step": 28705
    },
    {
      "epoch": 15.047169811320755,
      "grad_norm": 0.13643451035022736,
      "learning_rate": 0.00017549861575021848,
      "loss": 0.3245,
      "num_input_tokens_seen": 18731704,
      "step": 28710
    },
    {
      "epoch": 15.04979035639413,
      "grad_norm": 0.14643216133117676,
      "learning_rate": 0.00017532466880903214,
      "loss": 0.4922,
      "num_input_tokens_seen": 18734968,
      "step": 28715
    },
    {
      "epoch": 15.052410901467505,
      "grad_norm": 0.1915867179632187,
      "learning_rate": 0.00017515078978622102,
      "loss": 0.4778,
      "num_input_tokens_seen": 18738680,
      "step": 28720
    },
    {
      "epoch": 15.05503144654088,
      "grad_norm": 0.11400086432695389,
      "learning_rate": 0.0001749769787181587,
      "loss": 0.3965,
      "num_input_tokens_seen": 18742136,
      "step": 28725
    },
    {
      "epoch": 15.057651991614255,
      "grad_norm": 0.12993624806404114,
      "learning_rate": 0.00017480323564120426,
      "loss": 0.3345,
      "num_input_tokens_seen": 18745016,
      "step": 28730
    },
    {
      "epoch": 15.06027253668763,
      "grad_norm": 0.14799652993679047,
      "learning_rate": 0.00017462956059170287,
      "loss": 0.4448,
      "num_input_tokens_seen": 18748920,
      "step": 28735
    },
    {
      "epoch": 15.062893081761006,
      "grad_norm": 0.17464414238929749,
      "learning_rate": 0.00017445595360598553,
      "loss": 0.4326,
      "num_input_tokens_seen": 18753560,
      "step": 28740
    },
    {
      "epoch": 15.065513626834381,
      "grad_norm": 0.11573896557092667,
      "learning_rate": 0.0001742824147203686,
      "loss": 0.5243,
      "num_input_tokens_seen": 18756952,
      "step": 28745
    },
    {
      "epoch": 15.068134171907756,
      "grad_norm": 0.09878142178058624,
      "learning_rate": 0.00017410894397115463,
      "loss": 0.4811,
      "num_input_tokens_seen": 18760472,
      "step": 28750
    },
    {
      "epoch": 15.070754716981131,
      "grad_norm": 0.1885787546634674,
      "learning_rate": 0.00017393554139463147,
      "loss": 0.4364,
      "num_input_tokens_seen": 18763928,
      "step": 28755
    },
    {
      "epoch": 15.073375262054507,
      "grad_norm": 0.18804247677326202,
      "learning_rate": 0.00017376220702707308,
      "loss": 0.4327,
      "num_input_tokens_seen": 18767512,
      "step": 28760
    },
    {
      "epoch": 15.075995807127883,
      "grad_norm": 0.15214727818965912,
      "learning_rate": 0.00017358894090473926,
      "loss": 0.5443,
      "num_input_tokens_seen": 18774040,
      "step": 28765
    },
    {
      "epoch": 15.078616352201259,
      "grad_norm": 0.21224819123744965,
      "learning_rate": 0.00017341574306387492,
      "loss": 0.3806,
      "num_input_tokens_seen": 18777336,
      "step": 28770
    },
    {
      "epoch": 15.081236897274634,
      "grad_norm": 0.11002187430858612,
      "learning_rate": 0.00017324261354071147,
      "loss": 0.3143,
      "num_input_tokens_seen": 18780504,
      "step": 28775
    },
    {
      "epoch": 15.083857442348009,
      "grad_norm": 0.0849493220448494,
      "learning_rate": 0.00017306955237146522,
      "loss": 0.623,
      "num_input_tokens_seen": 18783800,
      "step": 28780
    },
    {
      "epoch": 15.086477987421384,
      "grad_norm": 0.19400599598884583,
      "learning_rate": 0.0001728965595923388,
      "loss": 0.4155,
      "num_input_tokens_seen": 18786968,
      "step": 28785
    },
    {
      "epoch": 15.08909853249476,
      "grad_norm": 0.165838360786438,
      "learning_rate": 0.00017272363523952033,
      "loss": 0.3829,
      "num_input_tokens_seen": 18789720,
      "step": 28790
    },
    {
      "epoch": 15.091719077568134,
      "grad_norm": 0.23573388159275055,
      "learning_rate": 0.0001725507793491838,
      "loss": 0.3937,
      "num_input_tokens_seen": 18792952,
      "step": 28795
    },
    {
      "epoch": 15.09433962264151,
      "grad_norm": 0.14205235242843628,
      "learning_rate": 0.00017237799195748827,
      "loss": 0.374,
      "num_input_tokens_seen": 18795480,
      "step": 28800
    },
    {
      "epoch": 15.096960167714885,
      "grad_norm": 0.22756075859069824,
      "learning_rate": 0.00017220527310057927,
      "loss": 0.2897,
      "num_input_tokens_seen": 18798648,
      "step": 28805
    },
    {
      "epoch": 15.09958071278826,
      "grad_norm": 0.17150670289993286,
      "learning_rate": 0.0001720326228145873,
      "loss": 0.4896,
      "num_input_tokens_seen": 18802680,
      "step": 28810
    },
    {
      "epoch": 15.102201257861635,
      "grad_norm": 0.1605619341135025,
      "learning_rate": 0.00017186004113562903,
      "loss": 0.4537,
      "num_input_tokens_seen": 18805112,
      "step": 28815
    },
    {
      "epoch": 15.10482180293501,
      "grad_norm": 0.13677209615707397,
      "learning_rate": 0.00017168752809980655,
      "loss": 0.4025,
      "num_input_tokens_seen": 18809272,
      "step": 28820
    },
    {
      "epoch": 15.107442348008385,
      "grad_norm": 0.12895269691944122,
      "learning_rate": 0.00017151508374320767,
      "loss": 0.401,
      "num_input_tokens_seen": 18812440,
      "step": 28825
    },
    {
      "epoch": 15.11006289308176,
      "grad_norm": 0.1292998194694519,
      "learning_rate": 0.00017134270810190583,
      "loss": 0.5515,
      "num_input_tokens_seen": 18815288,
      "step": 28830
    },
    {
      "epoch": 15.112683438155136,
      "grad_norm": 0.18750719726085663,
      "learning_rate": 0.0001711704012119597,
      "loss": 0.5263,
      "num_input_tokens_seen": 18818424,
      "step": 28835
    },
    {
      "epoch": 15.115303983228511,
      "grad_norm": 0.0995907187461853,
      "learning_rate": 0.0001709981631094142,
      "loss": 0.4733,
      "num_input_tokens_seen": 18821144,
      "step": 28840
    },
    {
      "epoch": 15.117924528301886,
      "grad_norm": 0.13228817284107208,
      "learning_rate": 0.00017082599383029946,
      "loss": 0.3621,
      "num_input_tokens_seen": 18824632,
      "step": 28845
    },
    {
      "epoch": 15.120545073375261,
      "grad_norm": 0.16525079309940338,
      "learning_rate": 0.0001706538934106316,
      "loss": 0.4647,
      "num_input_tokens_seen": 18827224,
      "step": 28850
    },
    {
      "epoch": 15.123165618448636,
      "grad_norm": 0.11392122507095337,
      "learning_rate": 0.00017048186188641185,
      "loss": 0.4375,
      "num_input_tokens_seen": 18830072,
      "step": 28855
    },
    {
      "epoch": 15.125786163522013,
      "grad_norm": 0.18393704295158386,
      "learning_rate": 0.0001703098992936271,
      "loss": 0.5073,
      "num_input_tokens_seen": 18832856,
      "step": 28860
    },
    {
      "epoch": 15.128406708595389,
      "grad_norm": 0.41592860221862793,
      "learning_rate": 0.00017013800566825012,
      "loss": 0.6184,
      "num_input_tokens_seen": 18835544,
      "step": 28865
    },
    {
      "epoch": 15.131027253668764,
      "grad_norm": 0.10000120103359222,
      "learning_rate": 0.00016996618104623905,
      "loss": 0.4407,
      "num_input_tokens_seen": 18838360,
      "step": 28870
    },
    {
      "epoch": 15.133647798742139,
      "grad_norm": 0.10440567135810852,
      "learning_rate": 0.00016979442546353774,
      "loss": 0.5716,
      "num_input_tokens_seen": 18842168,
      "step": 28875
    },
    {
      "epoch": 15.136268343815514,
      "grad_norm": 0.18211017549037933,
      "learning_rate": 0.00016962273895607555,
      "loss": 0.4662,
      "num_input_tokens_seen": 18845496,
      "step": 28880
    },
    {
      "epoch": 15.13888888888889,
      "grad_norm": 0.12175018340349197,
      "learning_rate": 0.00016945112155976723,
      "loss": 0.3317,
      "num_input_tokens_seen": 18850232,
      "step": 28885
    },
    {
      "epoch": 15.141509433962264,
      "grad_norm": 0.16617554426193237,
      "learning_rate": 0.000169279573310513,
      "loss": 0.4561,
      "num_input_tokens_seen": 18852536,
      "step": 28890
    },
    {
      "epoch": 15.14412997903564,
      "grad_norm": 0.11003720760345459,
      "learning_rate": 0.00016910809424419904,
      "loss": 0.4835,
      "num_input_tokens_seen": 18855544,
      "step": 28895
    },
    {
      "epoch": 15.146750524109015,
      "grad_norm": 0.16356161236763,
      "learning_rate": 0.00016893668439669674,
      "loss": 0.5213,
      "num_input_tokens_seen": 18858584,
      "step": 28900
    },
    {
      "epoch": 15.14937106918239,
      "grad_norm": 0.1213255524635315,
      "learning_rate": 0.0001687653438038632,
      "loss": 0.4261,
      "num_input_tokens_seen": 18861688,
      "step": 28905
    },
    {
      "epoch": 15.151991614255765,
      "grad_norm": 0.18635322153568268,
      "learning_rate": 0.00016859407250154084,
      "loss": 0.4456,
      "num_input_tokens_seen": 18864056,
      "step": 28910
    },
    {
      "epoch": 15.15461215932914,
      "grad_norm": 0.11538216471672058,
      "learning_rate": 0.0001684228705255575,
      "loss": 0.4171,
      "num_input_tokens_seen": 18867960,
      "step": 28915
    },
    {
      "epoch": 15.157232704402515,
      "grad_norm": 0.18541967868804932,
      "learning_rate": 0.00016825173791172677,
      "loss": 0.3193,
      "num_input_tokens_seen": 18870328,
      "step": 28920
    },
    {
      "epoch": 15.15985324947589,
      "grad_norm": 0.12133210897445679,
      "learning_rate": 0.0001680806746958477,
      "loss": 0.4287,
      "num_input_tokens_seen": 18873976,
      "step": 28925
    },
    {
      "epoch": 15.162473794549266,
      "grad_norm": 0.1157635748386383,
      "learning_rate": 0.00016790968091370494,
      "loss": 0.2876,
      "num_input_tokens_seen": 18877112,
      "step": 28930
    },
    {
      "epoch": 15.165094339622641,
      "grad_norm": 0.22391951084136963,
      "learning_rate": 0.00016773875660106824,
      "loss": 0.4337,
      "num_input_tokens_seen": 18879480,
      "step": 28935
    },
    {
      "epoch": 15.167714884696016,
      "grad_norm": 0.174841970205307,
      "learning_rate": 0.0001675679017936928,
      "loss": 0.3511,
      "num_input_tokens_seen": 18882072,
      "step": 28940
    },
    {
      "epoch": 15.170335429769391,
      "grad_norm": 0.08081687241792679,
      "learning_rate": 0.00016739711652731976,
      "loss": 0.5126,
      "num_input_tokens_seen": 18884760,
      "step": 28945
    },
    {
      "epoch": 15.172955974842766,
      "grad_norm": 0.10236389189958572,
      "learning_rate": 0.00016722640083767542,
      "loss": 0.4085,
      "num_input_tokens_seen": 18887736,
      "step": 28950
    },
    {
      "epoch": 15.175576519916143,
      "grad_norm": 0.1764647215604782,
      "learning_rate": 0.00016705575476047153,
      "loss": 0.3887,
      "num_input_tokens_seen": 18890584,
      "step": 28955
    },
    {
      "epoch": 15.178197064989519,
      "grad_norm": 0.13900883495807648,
      "learning_rate": 0.0001668851783314055,
      "loss": 0.3876,
      "num_input_tokens_seen": 18893208,
      "step": 28960
    },
    {
      "epoch": 15.180817610062894,
      "grad_norm": 0.1149090975522995,
      "learning_rate": 0.00016671467158615977,
      "loss": 0.3772,
      "num_input_tokens_seen": 18895864,
      "step": 28965
    },
    {
      "epoch": 15.183438155136269,
      "grad_norm": 0.15987586975097656,
      "learning_rate": 0.00016654423456040224,
      "loss": 0.3856,
      "num_input_tokens_seen": 18898712,
      "step": 28970
    },
    {
      "epoch": 15.186058700209644,
      "grad_norm": 0.14033623039722443,
      "learning_rate": 0.0001663738672897866,
      "loss": 0.4494,
      "num_input_tokens_seen": 18901624,
      "step": 28975
    },
    {
      "epoch": 15.18867924528302,
      "grad_norm": 0.17207522690296173,
      "learning_rate": 0.0001662035698099517,
      "loss": 0.3284,
      "num_input_tokens_seen": 18905176,
      "step": 28980
    },
    {
      "epoch": 15.191299790356394,
      "grad_norm": 0.21003630757331848,
      "learning_rate": 0.00016603334215652206,
      "loss": 0.277,
      "num_input_tokens_seen": 18907864,
      "step": 28985
    },
    {
      "epoch": 15.19392033542977,
      "grad_norm": 0.10201890021562576,
      "learning_rate": 0.0001658631843651069,
      "loss": 0.3376,
      "num_input_tokens_seen": 18911064,
      "step": 28990
    },
    {
      "epoch": 15.196540880503145,
      "grad_norm": 0.1218399778008461,
      "learning_rate": 0.00016569309647130177,
      "loss": 0.5244,
      "num_input_tokens_seen": 18914136,
      "step": 28995
    },
    {
      "epoch": 15.19916142557652,
      "grad_norm": 0.1165149137377739,
      "learning_rate": 0.00016552307851068666,
      "loss": 0.3902,
      "num_input_tokens_seen": 18917048,
      "step": 29000
    },
    {
      "epoch": 15.201781970649895,
      "grad_norm": 0.10505588352680206,
      "learning_rate": 0.00016535313051882765,
      "loss": 0.5331,
      "num_input_tokens_seen": 18919864,
      "step": 29005
    },
    {
      "epoch": 15.20440251572327,
      "grad_norm": 0.11758586764335632,
      "learning_rate": 0.00016518325253127604,
      "loss": 0.4084,
      "num_input_tokens_seen": 18922968,
      "step": 29010
    },
    {
      "epoch": 15.207023060796645,
      "grad_norm": 0.15166790783405304,
      "learning_rate": 0.00016501344458356803,
      "loss": 0.3807,
      "num_input_tokens_seen": 18926328,
      "step": 29015
    },
    {
      "epoch": 15.20964360587002,
      "grad_norm": 0.19648432731628418,
      "learning_rate": 0.00016484370671122588,
      "loss": 0.405,
      "num_input_tokens_seen": 18929368,
      "step": 29020
    },
    {
      "epoch": 15.212264150943396,
      "grad_norm": 0.14721116423606873,
      "learning_rate": 0.00016467403894975641,
      "loss": 0.3928,
      "num_input_tokens_seen": 18932056,
      "step": 29025
    },
    {
      "epoch": 15.214884696016771,
      "grad_norm": 0.1331866979598999,
      "learning_rate": 0.0001645044413346524,
      "loss": 0.5303,
      "num_input_tokens_seen": 18935896,
      "step": 29030
    },
    {
      "epoch": 15.217505241090146,
      "grad_norm": 0.17097124457359314,
      "learning_rate": 0.00016433491390139177,
      "loss": 0.3798,
      "num_input_tokens_seen": 18938968,
      "step": 29035
    },
    {
      "epoch": 15.220125786163521,
      "grad_norm": 0.0938403457403183,
      "learning_rate": 0.00016416545668543786,
      "loss": 0.4882,
      "num_input_tokens_seen": 18943000,
      "step": 29040
    },
    {
      "epoch": 15.222746331236896,
      "grad_norm": 0.11825662106275558,
      "learning_rate": 0.0001639960697222388,
      "loss": 0.4827,
      "num_input_tokens_seen": 18946104,
      "step": 29045
    },
    {
      "epoch": 15.225366876310272,
      "grad_norm": 0.14826668798923492,
      "learning_rate": 0.00016382675304722877,
      "loss": 0.4119,
      "num_input_tokens_seen": 18950328,
      "step": 29050
    },
    {
      "epoch": 15.227987421383649,
      "grad_norm": 0.18636542558670044,
      "learning_rate": 0.00016365750669582662,
      "loss": 0.3502,
      "num_input_tokens_seen": 18953144,
      "step": 29055
    },
    {
      "epoch": 15.230607966457024,
      "grad_norm": 0.16964301466941833,
      "learning_rate": 0.00016348833070343682,
      "loss": 0.5097,
      "num_input_tokens_seen": 18960248,
      "step": 29060
    },
    {
      "epoch": 15.233228511530399,
      "grad_norm": 0.17803286015987396,
      "learning_rate": 0.00016331922510544932,
      "loss": 0.3738,
      "num_input_tokens_seen": 18962744,
      "step": 29065
    },
    {
      "epoch": 15.235849056603774,
      "grad_norm": 0.09249129146337509,
      "learning_rate": 0.0001631501899372387,
      "loss": 0.3708,
      "num_input_tokens_seen": 18965816,
      "step": 29070
    },
    {
      "epoch": 15.23846960167715,
      "grad_norm": 0.20237195491790771,
      "learning_rate": 0.0001629812252341653,
      "loss": 0.3965,
      "num_input_tokens_seen": 18969560,
      "step": 29075
    },
    {
      "epoch": 15.241090146750524,
      "grad_norm": 0.2455739676952362,
      "learning_rate": 0.0001628123310315749,
      "loss": 0.2507,
      "num_input_tokens_seen": 18972792,
      "step": 29080
    },
    {
      "epoch": 15.2437106918239,
      "grad_norm": 0.23347076773643494,
      "learning_rate": 0.00016264350736479777,
      "loss": 0.3178,
      "num_input_tokens_seen": 18975640,
      "step": 29085
    },
    {
      "epoch": 15.246331236897275,
      "grad_norm": 0.13806328177452087,
      "learning_rate": 0.00016247475426915038,
      "loss": 0.3639,
      "num_input_tokens_seen": 18978840,
      "step": 29090
    },
    {
      "epoch": 15.24895178197065,
      "grad_norm": 0.18638229370117188,
      "learning_rate": 0.0001623060717799335,
      "loss": 0.3854,
      "num_input_tokens_seen": 18981464,
      "step": 29095
    },
    {
      "epoch": 15.251572327044025,
      "grad_norm": 0.13577885925769806,
      "learning_rate": 0.0001621374599324338,
      "loss": 0.3662,
      "num_input_tokens_seen": 18985368,
      "step": 29100
    },
    {
      "epoch": 15.2541928721174,
      "grad_norm": 0.13172349333763123,
      "learning_rate": 0.00016196891876192292,
      "loss": 0.4752,
      "num_input_tokens_seen": 18988120,
      "step": 29105
    },
    {
      "epoch": 15.256813417190775,
      "grad_norm": 0.12636969983577728,
      "learning_rate": 0.00016180044830365798,
      "loss": 0.5296,
      "num_input_tokens_seen": 18991640,
      "step": 29110
    },
    {
      "epoch": 15.25943396226415,
      "grad_norm": 0.12056180834770203,
      "learning_rate": 0.00016163204859288066,
      "loss": 0.4803,
      "num_input_tokens_seen": 18994552,
      "step": 29115
    },
    {
      "epoch": 15.262054507337526,
      "grad_norm": 0.11575037986040115,
      "learning_rate": 0.00016146371966481866,
      "loss": 0.5611,
      "num_input_tokens_seen": 18999224,
      "step": 29120
    },
    {
      "epoch": 15.264675052410901,
      "grad_norm": 0.0891474038362503,
      "learning_rate": 0.00016129546155468411,
      "loss": 0.6806,
      "num_input_tokens_seen": 19002648,
      "step": 29125
    },
    {
      "epoch": 15.267295597484276,
      "grad_norm": 0.18438221514225006,
      "learning_rate": 0.00016112727429767483,
      "loss": 0.3748,
      "num_input_tokens_seen": 19005976,
      "step": 29130
    },
    {
      "epoch": 15.269916142557651,
      "grad_norm": 0.1346215158700943,
      "learning_rate": 0.00016095915792897392,
      "loss": 0.4514,
      "num_input_tokens_seen": 19009144,
      "step": 29135
    },
    {
      "epoch": 15.272536687631026,
      "grad_norm": 0.12595337629318237,
      "learning_rate": 0.000160791112483749,
      "loss": 0.3547,
      "num_input_tokens_seen": 19011896,
      "step": 29140
    },
    {
      "epoch": 15.275157232704402,
      "grad_norm": 0.14107993245124817,
      "learning_rate": 0.00016062313799715362,
      "loss": 0.569,
      "num_input_tokens_seen": 19015256,
      "step": 29145
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 0.24640385806560516,
      "learning_rate": 0.00016045523450432576,
      "loss": 0.5013,
      "num_input_tokens_seen": 19018360,
      "step": 29150
    },
    {
      "epoch": 15.280398322851154,
      "grad_norm": 0.11992021650075912,
      "learning_rate": 0.0001602874020403891,
      "loss": 0.2841,
      "num_input_tokens_seen": 19022136,
      "step": 29155
    },
    {
      "epoch": 15.283018867924529,
      "grad_norm": 0.19338077306747437,
      "learning_rate": 0.00016011964064045232,
      "loss": 0.4597,
      "num_input_tokens_seen": 19024920,
      "step": 29160
    },
    {
      "epoch": 15.285639412997904,
      "grad_norm": 0.12302298843860626,
      "learning_rate": 0.00015995195033960936,
      "loss": 0.3531,
      "num_input_tokens_seen": 19029464,
      "step": 29165
    },
    {
      "epoch": 15.28825995807128,
      "grad_norm": 0.10943569988012314,
      "learning_rate": 0.00015978433117293884,
      "loss": 0.4728,
      "num_input_tokens_seen": 19033272,
      "step": 29170
    },
    {
      "epoch": 15.290880503144654,
      "grad_norm": 0.1723390817642212,
      "learning_rate": 0.0001596167831755047,
      "loss": 0.4482,
      "num_input_tokens_seen": 19035736,
      "step": 29175
    },
    {
      "epoch": 15.29350104821803,
      "grad_norm": 0.14086514711380005,
      "learning_rate": 0.00015944930638235626,
      "loss": 0.4761,
      "num_input_tokens_seen": 19038904,
      "step": 29180
    },
    {
      "epoch": 15.296121593291405,
      "grad_norm": 0.13552746176719666,
      "learning_rate": 0.00015928190082852773,
      "loss": 0.3973,
      "num_input_tokens_seen": 19042424,
      "step": 29185
    },
    {
      "epoch": 15.29874213836478,
      "grad_norm": 0.13387441635131836,
      "learning_rate": 0.0001591145665490384,
      "loss": 0.2993,
      "num_input_tokens_seen": 19045784,
      "step": 29190
    },
    {
      "epoch": 15.301362683438155,
      "grad_norm": 0.16042622923851013,
      "learning_rate": 0.00015894730357889287,
      "loss": 0.4085,
      "num_input_tokens_seen": 19049016,
      "step": 29195
    },
    {
      "epoch": 15.30398322851153,
      "grad_norm": 0.20692673325538635,
      "learning_rate": 0.0001587801119530805,
      "loss": 0.3686,
      "num_input_tokens_seen": 19052536,
      "step": 29200
    },
    {
      "epoch": 15.306603773584905,
      "grad_norm": 0.13917042315006256,
      "learning_rate": 0.00015861299170657572,
      "loss": 0.3254,
      "num_input_tokens_seen": 19055480,
      "step": 29205
    },
    {
      "epoch": 15.30922431865828,
      "grad_norm": 0.095188669860363,
      "learning_rate": 0.00015844594287433834,
      "loss": 0.4599,
      "num_input_tokens_seen": 19059416,
      "step": 29210
    },
    {
      "epoch": 15.311844863731656,
      "grad_norm": 0.21040914952754974,
      "learning_rate": 0.00015827896549131304,
      "loss": 0.3451,
      "num_input_tokens_seen": 19062360,
      "step": 29215
    },
    {
      "epoch": 15.314465408805031,
      "grad_norm": 0.09646442532539368,
      "learning_rate": 0.0001581120595924298,
      "loss": 0.3943,
      "num_input_tokens_seen": 19065400,
      "step": 29220
    },
    {
      "epoch": 15.317085953878406,
      "grad_norm": 0.12331584841012955,
      "learning_rate": 0.00015794522521260317,
      "loss": 0.435,
      "num_input_tokens_seen": 19068376,
      "step": 29225
    },
    {
      "epoch": 15.319706498951781,
      "grad_norm": 0.18356642127037048,
      "learning_rate": 0.00015777846238673293,
      "loss": 0.312,
      "num_input_tokens_seen": 19071032,
      "step": 29230
    },
    {
      "epoch": 15.322327044025156,
      "grad_norm": 0.17780178785324097,
      "learning_rate": 0.00015761177114970404,
      "loss": 0.4379,
      "num_input_tokens_seen": 19073944,
      "step": 29235
    },
    {
      "epoch": 15.324947589098532,
      "grad_norm": 0.09934958070516586,
      "learning_rate": 0.0001574451515363865,
      "loss": 0.408,
      "num_input_tokens_seen": 19076760,
      "step": 29240
    },
    {
      "epoch": 15.327568134171909,
      "grad_norm": 0.13011349737644196,
      "learning_rate": 0.00015727860358163536,
      "loss": 0.5459,
      "num_input_tokens_seen": 19079480,
      "step": 29245
    },
    {
      "epoch": 15.330188679245284,
      "grad_norm": 0.12721878290176392,
      "learning_rate": 0.00015711212732029035,
      "loss": 0.5369,
      "num_input_tokens_seen": 19081944,
      "step": 29250
    },
    {
      "epoch": 15.332809224318659,
      "grad_norm": 0.14272837340831757,
      "learning_rate": 0.00015694572278717627,
      "loss": 0.4147,
      "num_input_tokens_seen": 19085272,
      "step": 29255
    },
    {
      "epoch": 15.335429769392034,
      "grad_norm": 0.07931757718324661,
      "learning_rate": 0.00015677939001710323,
      "loss": 0.4487,
      "num_input_tokens_seen": 19088152,
      "step": 29260
    },
    {
      "epoch": 15.33805031446541,
      "grad_norm": 0.1328994184732437,
      "learning_rate": 0.00015661312904486613,
      "loss": 0.5447,
      "num_input_tokens_seen": 19091576,
      "step": 29265
    },
    {
      "epoch": 15.340670859538784,
      "grad_norm": 0.09490392357110977,
      "learning_rate": 0.00015644693990524483,
      "loss": 0.5069,
      "num_input_tokens_seen": 19095672,
      "step": 29270
    },
    {
      "epoch": 15.34329140461216,
      "grad_norm": 0.1341313272714615,
      "learning_rate": 0.00015628082263300448,
      "loss": 0.4931,
      "num_input_tokens_seen": 19099416,
      "step": 29275
    },
    {
      "epoch": 15.345911949685535,
      "grad_norm": 0.2248154580593109,
      "learning_rate": 0.00015611477726289465,
      "loss": 0.41,
      "num_input_tokens_seen": 19102296,
      "step": 29280
    },
    {
      "epoch": 15.34853249475891,
      "grad_norm": 0.09786306321620941,
      "learning_rate": 0.00015594880382964994,
      "loss": 0.3951,
      "num_input_tokens_seen": 19105432,
      "step": 29285
    },
    {
      "epoch": 15.351153039832285,
      "grad_norm": 0.3095405101776123,
      "learning_rate": 0.00015578290236799037,
      "loss": 0.327,
      "num_input_tokens_seen": 19108632,
      "step": 29290
    },
    {
      "epoch": 15.35377358490566,
      "grad_norm": 0.16999532282352448,
      "learning_rate": 0.00015561707291262061,
      "loss": 0.5696,
      "num_input_tokens_seen": 19111832,
      "step": 29295
    },
    {
      "epoch": 15.356394129979035,
      "grad_norm": 0.13705629110336304,
      "learning_rate": 0.00015545131549823043,
      "loss": 0.3717,
      "num_input_tokens_seen": 19115288,
      "step": 29300
    },
    {
      "epoch": 15.35901467505241,
      "grad_norm": 0.10296806693077087,
      "learning_rate": 0.0001552856301594942,
      "loss": 0.3701,
      "num_input_tokens_seen": 19118456,
      "step": 29305
    },
    {
      "epoch": 15.361635220125786,
      "grad_norm": 0.15418386459350586,
      "learning_rate": 0.00015512001693107126,
      "loss": 0.3763,
      "num_input_tokens_seen": 19121272,
      "step": 29310
    },
    {
      "epoch": 15.364255765199161,
      "grad_norm": 0.10987994819879532,
      "learning_rate": 0.0001549544758476062,
      "loss": 0.347,
      "num_input_tokens_seen": 19124440,
      "step": 29315
    },
    {
      "epoch": 15.366876310272536,
      "grad_norm": 0.15494009852409363,
      "learning_rate": 0.00015478900694372832,
      "loss": 0.4427,
      "num_input_tokens_seen": 19127672,
      "step": 29320
    },
    {
      "epoch": 15.369496855345911,
      "grad_norm": 0.20877009630203247,
      "learning_rate": 0.00015462361025405196,
      "loss": 0.4926,
      "num_input_tokens_seen": 19130776,
      "step": 29325
    },
    {
      "epoch": 15.372117400419286,
      "grad_norm": 0.20521411299705505,
      "learning_rate": 0.0001544582858131759,
      "loss": 0.4259,
      "num_input_tokens_seen": 19133720,
      "step": 29330
    },
    {
      "epoch": 15.374737945492662,
      "grad_norm": 0.20480690896511078,
      "learning_rate": 0.00015429303365568448,
      "loss": 0.4652,
      "num_input_tokens_seen": 19137112,
      "step": 29335
    },
    {
      "epoch": 15.377358490566039,
      "grad_norm": 0.13752688467502594,
      "learning_rate": 0.00015412785381614625,
      "loss": 0.4168,
      "num_input_tokens_seen": 19140408,
      "step": 29340
    },
    {
      "epoch": 15.379979035639414,
      "grad_norm": 0.1517147570848465,
      "learning_rate": 0.000153962746329115,
      "loss": 0.4216,
      "num_input_tokens_seen": 19143672,
      "step": 29345
    },
    {
      "epoch": 15.382599580712789,
      "grad_norm": 0.14765091240406036,
      "learning_rate": 0.0001537977112291295,
      "loss": 0.3544,
      "num_input_tokens_seen": 19147288,
      "step": 29350
    },
    {
      "epoch": 15.385220125786164,
      "grad_norm": 0.13530094921588898,
      "learning_rate": 0.00015363274855071323,
      "loss": 0.5045,
      "num_input_tokens_seen": 19149688,
      "step": 29355
    },
    {
      "epoch": 15.38784067085954,
      "grad_norm": 0.11878439784049988,
      "learning_rate": 0.0001534678583283743,
      "loss": 0.4403,
      "num_input_tokens_seen": 19153752,
      "step": 29360
    },
    {
      "epoch": 15.390461215932914,
      "grad_norm": 0.11604972928762436,
      "learning_rate": 0.00015330304059660605,
      "loss": 0.5249,
      "num_input_tokens_seen": 19157208,
      "step": 29365
    },
    {
      "epoch": 15.39308176100629,
      "grad_norm": 0.12918208539485931,
      "learning_rate": 0.00015313829538988628,
      "loss": 0.506,
      "num_input_tokens_seen": 19161656,
      "step": 29370
    },
    {
      "epoch": 15.395702306079665,
      "grad_norm": 0.11643408238887787,
      "learning_rate": 0.00015297362274267783,
      "loss": 0.4393,
      "num_input_tokens_seen": 19164824,
      "step": 29375
    },
    {
      "epoch": 15.39832285115304,
      "grad_norm": 0.11849524825811386,
      "learning_rate": 0.00015280902268942866,
      "loss": 0.424,
      "num_input_tokens_seen": 19167480,
      "step": 29380
    },
    {
      "epoch": 15.400943396226415,
      "grad_norm": 0.17812858521938324,
      "learning_rate": 0.00015264449526457074,
      "loss": 0.3932,
      "num_input_tokens_seen": 19169752,
      "step": 29385
    },
    {
      "epoch": 15.40356394129979,
      "grad_norm": 0.16544191539287567,
      "learning_rate": 0.00015248004050252178,
      "loss": 0.5358,
      "num_input_tokens_seen": 19172696,
      "step": 29390
    },
    {
      "epoch": 15.406184486373165,
      "grad_norm": 0.1264103651046753,
      "learning_rate": 0.00015231565843768348,
      "loss": 0.4283,
      "num_input_tokens_seen": 19176600,
      "step": 29395
    },
    {
      "epoch": 15.40880503144654,
      "grad_norm": 0.16112704575061798,
      "learning_rate": 0.00015215134910444284,
      "loss": 0.3074,
      "num_input_tokens_seen": 19179000,
      "step": 29400
    },
    {
      "epoch": 15.411425576519916,
      "grad_norm": 0.11362826824188232,
      "learning_rate": 0.00015198711253717163,
      "loss": 0.4889,
      "num_input_tokens_seen": 19182264,
      "step": 29405
    },
    {
      "epoch": 15.414046121593291,
      "grad_norm": 0.1468636840581894,
      "learning_rate": 0.00015182294877022596,
      "loss": 0.4093,
      "num_input_tokens_seen": 19185240,
      "step": 29410
    },
    {
      "epoch": 15.416666666666666,
      "grad_norm": 0.0700279101729393,
      "learning_rate": 0.00015165885783794726,
      "loss": 0.3703,
      "num_input_tokens_seen": 19188920,
      "step": 29415
    },
    {
      "epoch": 15.419287211740041,
      "grad_norm": 0.18384946882724762,
      "learning_rate": 0.0001514948397746615,
      "loss": 0.4462,
      "num_input_tokens_seen": 19191672,
      "step": 29420
    },
    {
      "epoch": 15.421907756813416,
      "grad_norm": 0.16924230754375458,
      "learning_rate": 0.0001513308946146792,
      "loss": 0.3886,
      "num_input_tokens_seen": 19194744,
      "step": 29425
    },
    {
      "epoch": 15.424528301886792,
      "grad_norm": 0.14361907541751862,
      "learning_rate": 0.0001511670223922959,
      "loss": 0.5069,
      "num_input_tokens_seen": 19197848,
      "step": 29430
    },
    {
      "epoch": 15.427148846960169,
      "grad_norm": 0.09366191923618317,
      "learning_rate": 0.000151003223141792,
      "loss": 0.3027,
      "num_input_tokens_seen": 19201432,
      "step": 29435
    },
    {
      "epoch": 15.429769392033544,
      "grad_norm": 0.06539838016033173,
      "learning_rate": 0.00015083949689743202,
      "loss": 0.3631,
      "num_input_tokens_seen": 19205304,
      "step": 29440
    },
    {
      "epoch": 15.432389937106919,
      "grad_norm": 0.28807857632637024,
      "learning_rate": 0.00015067584369346594,
      "loss": 0.4936,
      "num_input_tokens_seen": 19208888,
      "step": 29445
    },
    {
      "epoch": 15.435010482180294,
      "grad_norm": 0.18958815932273865,
      "learning_rate": 0.00015051226356412818,
      "loss": 0.4229,
      "num_input_tokens_seen": 19212216,
      "step": 29450
    },
    {
      "epoch": 15.43763102725367,
      "grad_norm": 0.15161462128162384,
      "learning_rate": 0.00015034875654363754,
      "loss": 0.5184,
      "num_input_tokens_seen": 19215736,
      "step": 29455
    },
    {
      "epoch": 15.440251572327044,
      "grad_norm": 0.06952966004610062,
      "learning_rate": 0.00015018532266619818,
      "loss": 0.4074,
      "num_input_tokens_seen": 19219224,
      "step": 29460
    },
    {
      "epoch": 15.44287211740042,
      "grad_norm": 0.10198888927698135,
      "learning_rate": 0.00015002196196599828,
      "loss": 0.4812,
      "num_input_tokens_seen": 19222424,
      "step": 29465
    },
    {
      "epoch": 15.445492662473795,
      "grad_norm": 0.16147567331790924,
      "learning_rate": 0.00014985867447721114,
      "loss": 0.4863,
      "num_input_tokens_seen": 19225784,
      "step": 29470
    },
    {
      "epoch": 15.44811320754717,
      "grad_norm": 0.2032189965248108,
      "learning_rate": 0.00014969546023399467,
      "loss": 0.3264,
      "num_input_tokens_seen": 19228600,
      "step": 29475
    },
    {
      "epoch": 15.450733752620545,
      "grad_norm": 0.14212174713611603,
      "learning_rate": 0.00014953231927049165,
      "loss": 0.4071,
      "num_input_tokens_seen": 19231864,
      "step": 29480
    },
    {
      "epoch": 15.45335429769392,
      "grad_norm": 0.16982455551624298,
      "learning_rate": 0.00014936925162082904,
      "loss": 0.4995,
      "num_input_tokens_seen": 19235480,
      "step": 29485
    },
    {
      "epoch": 15.455974842767295,
      "grad_norm": 0.18456588685512543,
      "learning_rate": 0.00014920625731911863,
      "loss": 0.5062,
      "num_input_tokens_seen": 19238904,
      "step": 29490
    },
    {
      "epoch": 15.45859538784067,
      "grad_norm": 0.11212984472513199,
      "learning_rate": 0.00014904333639945716,
      "loss": 0.3622,
      "num_input_tokens_seen": 19242040,
      "step": 29495
    },
    {
      "epoch": 15.461215932914046,
      "grad_norm": 0.08364130556583405,
      "learning_rate": 0.00014888048889592575,
      "loss": 0.4051,
      "num_input_tokens_seen": 19245336,
      "step": 29500
    },
    {
      "epoch": 15.463836477987421,
      "grad_norm": 0.09478573501110077,
      "learning_rate": 0.00014871771484259046,
      "loss": 0.4213,
      "num_input_tokens_seen": 19250264,
      "step": 29505
    },
    {
      "epoch": 15.466457023060796,
      "grad_norm": 0.16044668853282928,
      "learning_rate": 0.00014855501427350144,
      "loss": 0.7696,
      "num_input_tokens_seen": 19253592,
      "step": 29510
    },
    {
      "epoch": 15.469077568134171,
      "grad_norm": 0.20244638621807098,
      "learning_rate": 0.00014839238722269405,
      "loss": 0.375,
      "num_input_tokens_seen": 19256728,
      "step": 29515
    },
    {
      "epoch": 15.471698113207546,
      "grad_norm": 0.20712848007678986,
      "learning_rate": 0.00014822983372418774,
      "loss": 0.4203,
      "num_input_tokens_seen": 19259640,
      "step": 29520
    },
    {
      "epoch": 15.474318658280922,
      "grad_norm": 0.1673084795475006,
      "learning_rate": 0.00014806735381198705,
      "loss": 0.3626,
      "num_input_tokens_seen": 19263128,
      "step": 29525
    },
    {
      "epoch": 15.476939203354299,
      "grad_norm": 0.13797900080680847,
      "learning_rate": 0.00014790494752008087,
      "loss": 0.3694,
      "num_input_tokens_seen": 19267928,
      "step": 29530
    },
    {
      "epoch": 15.479559748427674,
      "grad_norm": 0.1799703985452652,
      "learning_rate": 0.00014774261488244295,
      "loss": 0.4243,
      "num_input_tokens_seen": 19270552,
      "step": 29535
    },
    {
      "epoch": 15.482180293501049,
      "grad_norm": 0.09992212057113647,
      "learning_rate": 0.00014758035593303114,
      "loss": 0.3924,
      "num_input_tokens_seen": 19274072,
      "step": 29540
    },
    {
      "epoch": 15.484800838574424,
      "grad_norm": 0.08412076532840729,
      "learning_rate": 0.00014741817070578817,
      "loss": 0.3936,
      "num_input_tokens_seen": 19277048,
      "step": 29545
    },
    {
      "epoch": 15.4874213836478,
      "grad_norm": 0.2297218292951584,
      "learning_rate": 0.0001472560592346414,
      "loss": 0.419,
      "num_input_tokens_seen": 19279576,
      "step": 29550
    },
    {
      "epoch": 15.490041928721174,
      "grad_norm": 0.2496759444475174,
      "learning_rate": 0.0001470940215535027,
      "loss": 0.4058,
      "num_input_tokens_seen": 19282200,
      "step": 29555
    },
    {
      "epoch": 15.49266247379455,
      "grad_norm": 0.13556408882141113,
      "learning_rate": 0.00014693205769626878,
      "loss": 0.51,
      "num_input_tokens_seen": 19285336,
      "step": 29560
    },
    {
      "epoch": 15.495283018867925,
      "grad_norm": 0.1369592249393463,
      "learning_rate": 0.00014677016769682028,
      "loss": 0.4565,
      "num_input_tokens_seen": 19288344,
      "step": 29565
    },
    {
      "epoch": 15.4979035639413,
      "grad_norm": 0.19885413348674774,
      "learning_rate": 0.00014660835158902268,
      "loss": 0.366,
      "num_input_tokens_seen": 19291928,
      "step": 29570
    },
    {
      "epoch": 15.500524109014675,
      "grad_norm": 0.3547385334968567,
      "learning_rate": 0.00014644660940672628,
      "loss": 0.3503,
      "num_input_tokens_seen": 19295736,
      "step": 29575
    },
    {
      "epoch": 15.50314465408805,
      "grad_norm": 0.22494499385356903,
      "learning_rate": 0.0001462849411837656,
      "loss": 0.4004,
      "num_input_tokens_seen": 19298744,
      "step": 29580
    },
    {
      "epoch": 15.505765199161425,
      "grad_norm": 0.18251363933086395,
      "learning_rate": 0.00014612334695395985,
      "loss": 0.4137,
      "num_input_tokens_seen": 19301368,
      "step": 29585
    },
    {
      "epoch": 15.5083857442348,
      "grad_norm": 0.09349897503852844,
      "learning_rate": 0.0001459618267511128,
      "loss": 0.3588,
      "num_input_tokens_seen": 19304952,
      "step": 29590
    },
    {
      "epoch": 15.511006289308176,
      "grad_norm": 0.12396366894245148,
      "learning_rate": 0.00014580038060901256,
      "loss": 0.49,
      "num_input_tokens_seen": 19308088,
      "step": 29595
    },
    {
      "epoch": 15.51362683438155,
      "grad_norm": 0.17301417887210846,
      "learning_rate": 0.00014563900856143157,
      "loss": 0.4128,
      "num_input_tokens_seen": 19310840,
      "step": 29600
    },
    {
      "epoch": 15.516247379454926,
      "grad_norm": 0.16379490494728088,
      "learning_rate": 0.0001454777106421273,
      "loss": 0.4632,
      "num_input_tokens_seen": 19313720,
      "step": 29605
    },
    {
      "epoch": 15.518867924528301,
      "grad_norm": 0.16684642434120178,
      "learning_rate": 0.00014531648688484127,
      "loss": 0.5195,
      "num_input_tokens_seen": 19318424,
      "step": 29610
    },
    {
      "epoch": 15.521488469601676,
      "grad_norm": 0.16792622208595276,
      "learning_rate": 0.0001451553373233,
      "loss": 0.3347,
      "num_input_tokens_seen": 19322488,
      "step": 29615
    },
    {
      "epoch": 15.524109014675052,
      "grad_norm": 0.14739060401916504,
      "learning_rate": 0.00014499426199121379,
      "loss": 0.6154,
      "num_input_tokens_seen": 19325592,
      "step": 29620
    },
    {
      "epoch": 15.526729559748428,
      "grad_norm": 0.2083694338798523,
      "learning_rate": 0.0001448332609222777,
      "loss": 0.453,
      "num_input_tokens_seen": 19328408,
      "step": 29625
    },
    {
      "epoch": 15.529350104821804,
      "grad_norm": 0.2075255811214447,
      "learning_rate": 0.0001446723341501715,
      "loss": 0.5547,
      "num_input_tokens_seen": 19330776,
      "step": 29630
    },
    {
      "epoch": 15.531970649895179,
      "grad_norm": 0.15228089690208435,
      "learning_rate": 0.00014451148170855922,
      "loss": 0.4228,
      "num_input_tokens_seen": 19334360,
      "step": 29635
    },
    {
      "epoch": 15.534591194968554,
      "grad_norm": 0.1779707968235016,
      "learning_rate": 0.0001443507036310895,
      "loss": 0.3782,
      "num_input_tokens_seen": 19336664,
      "step": 29640
    },
    {
      "epoch": 15.53721174004193,
      "grad_norm": 0.10426324605941772,
      "learning_rate": 0.00014418999995139493,
      "loss": 0.3817,
      "num_input_tokens_seen": 19339608,
      "step": 29645
    },
    {
      "epoch": 15.539832285115304,
      "grad_norm": 0.12205319851636887,
      "learning_rate": 0.00014402937070309325,
      "loss": 0.4162,
      "num_input_tokens_seen": 19342712,
      "step": 29650
    },
    {
      "epoch": 15.54245283018868,
      "grad_norm": 0.15766854584217072,
      "learning_rate": 0.0001438688159197859,
      "loss": 0.3923,
      "num_input_tokens_seen": 19346296,
      "step": 29655
    },
    {
      "epoch": 15.545073375262055,
      "grad_norm": 0.09216087311506271,
      "learning_rate": 0.0001437083356350593,
      "loss": 0.3106,
      "num_input_tokens_seen": 19349656,
      "step": 29660
    },
    {
      "epoch": 15.54769392033543,
      "grad_norm": 0.1781277358531952,
      "learning_rate": 0.0001435479298824841,
      "loss": 0.4044,
      "num_input_tokens_seen": 19352280,
      "step": 29665
    },
    {
      "epoch": 15.550314465408805,
      "grad_norm": 0.1348106861114502,
      "learning_rate": 0.00014338759869561557,
      "loss": 0.4575,
      "num_input_tokens_seen": 19355128,
      "step": 29670
    },
    {
      "epoch": 15.55293501048218,
      "grad_norm": 0.23597802221775055,
      "learning_rate": 0.00014322734210799287,
      "loss": 0.5723,
      "num_input_tokens_seen": 19359096,
      "step": 29675
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 0.14579591155052185,
      "learning_rate": 0.0001430671601531398,
      "loss": 0.4145,
      "num_input_tokens_seen": 19361784,
      "step": 29680
    },
    {
      "epoch": 15.55817610062893,
      "grad_norm": 0.20718178153038025,
      "learning_rate": 0.00014290705286456473,
      "loss": 0.3651,
      "num_input_tokens_seen": 19364824,
      "step": 29685
    },
    {
      "epoch": 15.560796645702306,
      "grad_norm": 0.2201717048883438,
      "learning_rate": 0.00014274702027576024,
      "loss": 0.5448,
      "num_input_tokens_seen": 19368056,
      "step": 29690
    },
    {
      "epoch": 15.56341719077568,
      "grad_norm": 0.09090803563594818,
      "learning_rate": 0.00014258706242020354,
      "loss": 0.3964,
      "num_input_tokens_seen": 19371448,
      "step": 29695
    },
    {
      "epoch": 15.566037735849056,
      "grad_norm": 0.11258033663034439,
      "learning_rate": 0.00014242717933135558,
      "loss": 0.544,
      "num_input_tokens_seen": 19375448,
      "step": 29700
    },
    {
      "epoch": 15.568658280922431,
      "grad_norm": 0.09749285876750946,
      "learning_rate": 0.0001422673710426625,
      "loss": 0.4148,
      "num_input_tokens_seen": 19379256,
      "step": 29705
    },
    {
      "epoch": 15.571278825995806,
      "grad_norm": 0.1199333518743515,
      "learning_rate": 0.00014210763758755403,
      "loss": 0.4235,
      "num_input_tokens_seen": 19383032,
      "step": 29710
    },
    {
      "epoch": 15.573899371069182,
      "grad_norm": 0.11518338322639465,
      "learning_rate": 0.00014194797899944468,
      "loss": 0.4117,
      "num_input_tokens_seen": 19387640,
      "step": 29715
    },
    {
      "epoch": 15.576519916142558,
      "grad_norm": 0.20647792518138885,
      "learning_rate": 0.00014178839531173344,
      "loss": 0.4698,
      "num_input_tokens_seen": 19389944,
      "step": 29720
    },
    {
      "epoch": 15.579140461215934,
      "grad_norm": 0.12123239785432816,
      "learning_rate": 0.00014162888655780303,
      "loss": 0.4523,
      "num_input_tokens_seen": 19393464,
      "step": 29725
    },
    {
      "epoch": 15.581761006289309,
      "grad_norm": 0.1998816728591919,
      "learning_rate": 0.00014146945277102108,
      "loss": 0.3196,
      "num_input_tokens_seen": 19395864,
      "step": 29730
    },
    {
      "epoch": 15.584381551362684,
      "grad_norm": 0.2237349897623062,
      "learning_rate": 0.00014131009398473938,
      "loss": 0.4547,
      "num_input_tokens_seen": 19398552,
      "step": 29735
    },
    {
      "epoch": 15.58700209643606,
      "grad_norm": 0.11937402933835983,
      "learning_rate": 0.00014115081023229376,
      "loss": 0.3847,
      "num_input_tokens_seen": 19401688,
      "step": 29740
    },
    {
      "epoch": 15.589622641509434,
      "grad_norm": 0.16639108955860138,
      "learning_rate": 0.00014099160154700462,
      "loss": 0.3791,
      "num_input_tokens_seen": 19404472,
      "step": 29745
    },
    {
      "epoch": 15.59224318658281,
      "grad_norm": 0.1204974576830864,
      "learning_rate": 0.00014083246796217684,
      "loss": 0.4237,
      "num_input_tokens_seen": 19407544,
      "step": 29750
    },
    {
      "epoch": 15.594863731656185,
      "grad_norm": 0.12000639736652374,
      "learning_rate": 0.00014067340951109898,
      "loss": 0.4341,
      "num_input_tokens_seen": 19410488,
      "step": 29755
    },
    {
      "epoch": 15.59748427672956,
      "grad_norm": 0.21444976329803467,
      "learning_rate": 0.00014051442622704447,
      "loss": 0.5962,
      "num_input_tokens_seen": 19414360,
      "step": 29760
    },
    {
      "epoch": 15.600104821802935,
      "grad_norm": 0.1413598656654358,
      "learning_rate": 0.0001403555181432709,
      "loss": 0.4197,
      "num_input_tokens_seen": 19418072,
      "step": 29765
    },
    {
      "epoch": 15.60272536687631,
      "grad_norm": 0.11493348330259323,
      "learning_rate": 0.00014019668529301977,
      "loss": 0.4136,
      "num_input_tokens_seen": 19421016,
      "step": 29770
    },
    {
      "epoch": 15.605345911949685,
      "grad_norm": 0.07733004540205002,
      "learning_rate": 0.00014003792770951733,
      "loss": 0.4794,
      "num_input_tokens_seen": 19424568,
      "step": 29775
    },
    {
      "epoch": 15.60796645702306,
      "grad_norm": 0.13061049580574036,
      "learning_rate": 0.00013987924542597363,
      "loss": 0.4373,
      "num_input_tokens_seen": 19427352,
      "step": 29780
    },
    {
      "epoch": 15.610587002096436,
      "grad_norm": 0.19554206728935242,
      "learning_rate": 0.0001397206384755833,
      "loss": 0.3717,
      "num_input_tokens_seen": 19430328,
      "step": 29785
    },
    {
      "epoch": 15.61320754716981,
      "grad_norm": 0.07564565539360046,
      "learning_rate": 0.00013956210689152526,
      "loss": 0.3413,
      "num_input_tokens_seen": 19433848,
      "step": 29790
    },
    {
      "epoch": 15.615828092243186,
      "grad_norm": 0.27403387427330017,
      "learning_rate": 0.0001394036507069622,
      "loss": 0.4863,
      "num_input_tokens_seen": 19436888,
      "step": 29795
    },
    {
      "epoch": 15.618448637316561,
      "grad_norm": 0.1614564210176468,
      "learning_rate": 0.00013924526995504172,
      "loss": 0.5428,
      "num_input_tokens_seen": 19441176,
      "step": 29800
    },
    {
      "epoch": 15.621069182389936,
      "grad_norm": 0.22484013438224792,
      "learning_rate": 0.00013908696466889486,
      "loss": 0.4256,
      "num_input_tokens_seen": 19444664,
      "step": 29805
    },
    {
      "epoch": 15.623689727463312,
      "grad_norm": 0.13958492875099182,
      "learning_rate": 0.0001389287348816375,
      "loss": 0.4056,
      "num_input_tokens_seen": 19447736,
      "step": 29810
    },
    {
      "epoch": 15.626310272536688,
      "grad_norm": 0.125222846865654,
      "learning_rate": 0.00013877058062636944,
      "loss": 0.4536,
      "num_input_tokens_seen": 19451384,
      "step": 29815
    },
    {
      "epoch": 15.628930817610064,
      "grad_norm": 0.11259657144546509,
      "learning_rate": 0.00013861250193617498,
      "loss": 0.4234,
      "num_input_tokens_seen": 19454552,
      "step": 29820
    },
    {
      "epoch": 15.631551362683439,
      "grad_norm": 0.15151670575141907,
      "learning_rate": 0.00013845449884412203,
      "loss": 0.3785,
      "num_input_tokens_seen": 19457336,
      "step": 29825
    },
    {
      "epoch": 15.634171907756814,
      "grad_norm": 0.19560028612613678,
      "learning_rate": 0.00013829657138326334,
      "loss": 0.4393,
      "num_input_tokens_seen": 19460344,
      "step": 29830
    },
    {
      "epoch": 15.63679245283019,
      "grad_norm": 0.18494616448879242,
      "learning_rate": 0.00013813871958663526,
      "loss": 0.4828,
      "num_input_tokens_seen": 19462744,
      "step": 29835
    },
    {
      "epoch": 15.639412997903564,
      "grad_norm": 0.12897460162639618,
      "learning_rate": 0.0001379809434872587,
      "loss": 0.4212,
      "num_input_tokens_seen": 19465880,
      "step": 29840
    },
    {
      "epoch": 15.64203354297694,
      "grad_norm": 0.15777961909770966,
      "learning_rate": 0.0001378232431181386,
      "loss": 0.4901,
      "num_input_tokens_seen": 19468632,
      "step": 29845
    },
    {
      "epoch": 15.644654088050315,
      "grad_norm": 0.13486507534980774,
      "learning_rate": 0.00013766561851226428,
      "loss": 0.4199,
      "num_input_tokens_seen": 19471736,
      "step": 29850
    },
    {
      "epoch": 15.64727463312369,
      "grad_norm": 0.09829393029212952,
      "learning_rate": 0.00013750806970260886,
      "loss": 0.3944,
      "num_input_tokens_seen": 19474424,
      "step": 29855
    },
    {
      "epoch": 15.649895178197065,
      "grad_norm": 0.10515277087688446,
      "learning_rate": 0.00013735059672212959,
      "loss": 0.3601,
      "num_input_tokens_seen": 19477400,
      "step": 29860
    },
    {
      "epoch": 15.65251572327044,
      "grad_norm": 0.16627664864063263,
      "learning_rate": 0.00013719319960376808,
      "loss": 0.4007,
      "num_input_tokens_seen": 19480024,
      "step": 29865
    },
    {
      "epoch": 15.655136268343815,
      "grad_norm": 0.15370218455791473,
      "learning_rate": 0.00013703587838045017,
      "loss": 0.449,
      "num_input_tokens_seen": 19483256,
      "step": 29870
    },
    {
      "epoch": 15.65775681341719,
      "grad_norm": 0.1576152890920639,
      "learning_rate": 0.00013687863308508568,
      "loss": 0.3758,
      "num_input_tokens_seen": 19486392,
      "step": 29875
    },
    {
      "epoch": 15.660377358490566,
      "grad_norm": 0.12589707970619202,
      "learning_rate": 0.00013672146375056849,
      "loss": 0.4031,
      "num_input_tokens_seen": 19489464,
      "step": 29880
    },
    {
      "epoch": 15.66299790356394,
      "grad_norm": 0.2508959472179413,
      "learning_rate": 0.00013656437040977643,
      "loss": 0.4179,
      "num_input_tokens_seen": 19492952,
      "step": 29885
    },
    {
      "epoch": 15.665618448637316,
      "grad_norm": 0.15592384338378906,
      "learning_rate": 0.00013640735309557178,
      "loss": 0.4969,
      "num_input_tokens_seen": 19495928,
      "step": 29890
    },
    {
      "epoch": 15.668238993710691,
      "grad_norm": 0.08944158256053925,
      "learning_rate": 0.00013625041184080078,
      "loss": 0.4177,
      "num_input_tokens_seen": 19499800,
      "step": 29895
    },
    {
      "epoch": 15.670859538784066,
      "grad_norm": 0.1328129768371582,
      "learning_rate": 0.00013609354667829378,
      "loss": 0.3613,
      "num_input_tokens_seen": 19504408,
      "step": 29900
    },
    {
      "epoch": 15.673480083857442,
      "grad_norm": 0.09443587809801102,
      "learning_rate": 0.00013593675764086521,
      "loss": 0.3516,
      "num_input_tokens_seen": 19507576,
      "step": 29905
    },
    {
      "epoch": 15.676100628930818,
      "grad_norm": 0.2154245376586914,
      "learning_rate": 0.00013578004476131356,
      "loss": 0.6138,
      "num_input_tokens_seen": 19510712,
      "step": 29910
    },
    {
      "epoch": 15.678721174004194,
      "grad_norm": 0.12016832828521729,
      "learning_rate": 0.00013562340807242112,
      "loss": 0.344,
      "num_input_tokens_seen": 19514168,
      "step": 29915
    },
    {
      "epoch": 15.681341719077569,
      "grad_norm": 0.15956540405750275,
      "learning_rate": 0.00013546684760695472,
      "loss": 0.3582,
      "num_input_tokens_seen": 19517304,
      "step": 29920
    },
    {
      "epoch": 15.683962264150944,
      "grad_norm": 0.10551378130912781,
      "learning_rate": 0.00013531036339766494,
      "loss": 0.4013,
      "num_input_tokens_seen": 19520504,
      "step": 29925
    },
    {
      "epoch": 15.68658280922432,
      "grad_norm": 0.04775470495223999,
      "learning_rate": 0.00013515395547728675,
      "loss": 0.3786,
      "num_input_tokens_seen": 19527000,
      "step": 29930
    },
    {
      "epoch": 15.689203354297694,
      "grad_norm": 0.10710529237985611,
      "learning_rate": 0.00013499762387853866,
      "loss": 0.4399,
      "num_input_tokens_seen": 19530008,
      "step": 29935
    },
    {
      "epoch": 15.69182389937107,
      "grad_norm": 0.1060333251953125,
      "learning_rate": 0.00013484136863412333,
      "loss": 0.3225,
      "num_input_tokens_seen": 19533208,
      "step": 29940
    },
    {
      "epoch": 15.694444444444445,
      "grad_norm": 0.12303679436445236,
      "learning_rate": 0.00013468518977672773,
      "loss": 0.6403,
      "num_input_tokens_seen": 19536248,
      "step": 29945
    },
    {
      "epoch": 15.69706498951782,
      "grad_norm": 0.07539606839418411,
      "learning_rate": 0.0001345290873390227,
      "loss": 0.3863,
      "num_input_tokens_seen": 19539480,
      "step": 29950
    },
    {
      "epoch": 15.699685534591195,
      "grad_norm": 0.20376217365264893,
      "learning_rate": 0.00013437306135366322,
      "loss": 0.4757,
      "num_input_tokens_seen": 19542584,
      "step": 29955
    },
    {
      "epoch": 15.70230607966457,
      "grad_norm": 0.18767918646335602,
      "learning_rate": 0.00013421711185328806,
      "loss": 0.5484,
      "num_input_tokens_seen": 19546296,
      "step": 29960
    },
    {
      "epoch": 15.704926624737945,
      "grad_norm": 0.14301538467407227,
      "learning_rate": 0.00013406123887051984,
      "loss": 0.4585,
      "num_input_tokens_seen": 19549400,
      "step": 29965
    },
    {
      "epoch": 15.70754716981132,
      "grad_norm": 0.132441908121109,
      "learning_rate": 0.0001339054424379656,
      "loss": 0.4628,
      "num_input_tokens_seen": 19552984,
      "step": 29970
    },
    {
      "epoch": 15.710167714884696,
      "grad_norm": 0.15033431351184845,
      "learning_rate": 0.00013374972258821621,
      "loss": 0.5381,
      "num_input_tokens_seen": 19555448,
      "step": 29975
    },
    {
      "epoch": 15.71278825995807,
      "grad_norm": 0.14932547509670258,
      "learning_rate": 0.00013359407935384642,
      "loss": 0.3856,
      "num_input_tokens_seen": 19558136,
      "step": 29980
    },
    {
      "epoch": 15.715408805031446,
      "grad_norm": 0.10520284622907639,
      "learning_rate": 0.0001334385127674152,
      "loss": 0.3508,
      "num_input_tokens_seen": 19560824,
      "step": 29985
    },
    {
      "epoch": 15.718029350104821,
      "grad_norm": 0.12325097620487213,
      "learning_rate": 0.00013328302286146515,
      "loss": 0.3116,
      "num_input_tokens_seen": 19564472,
      "step": 29990
    },
    {
      "epoch": 15.720649895178196,
      "grad_norm": 0.19639572501182556,
      "learning_rate": 0.00013312760966852283,
      "loss": 0.5351,
      "num_input_tokens_seen": 19567960,
      "step": 29995
    },
    {
      "epoch": 15.723270440251572,
      "grad_norm": 0.2251555323600769,
      "learning_rate": 0.00013297227322109912,
      "loss": 0.4008,
      "num_input_tokens_seen": 19570616,
      "step": 30000
    },
    {
      "epoch": 15.725890985324948,
      "grad_norm": 0.24368155002593994,
      "learning_rate": 0.00013281701355168852,
      "loss": 0.5636,
      "num_input_tokens_seen": 19573816,
      "step": 30005
    },
    {
      "epoch": 15.728511530398324,
      "grad_norm": 0.108958899974823,
      "learning_rate": 0.00013266183069276982,
      "loss": 0.4705,
      "num_input_tokens_seen": 19577272,
      "step": 30010
    },
    {
      "epoch": 15.731132075471699,
      "grad_norm": 0.1406196504831314,
      "learning_rate": 0.00013250672467680514,
      "loss": 0.3418,
      "num_input_tokens_seen": 19580408,
      "step": 30015
    },
    {
      "epoch": 15.733752620545074,
      "grad_norm": 0.1486373096704483,
      "learning_rate": 0.00013235169553624127,
      "loss": 0.4097,
      "num_input_tokens_seen": 19583640,
      "step": 30020
    },
    {
      "epoch": 15.73637316561845,
      "grad_norm": 0.11231029778718948,
      "learning_rate": 0.00013219674330350816,
      "loss": 0.3944,
      "num_input_tokens_seen": 19587128,
      "step": 30025
    },
    {
      "epoch": 15.738993710691824,
      "grad_norm": 0.12923547625541687,
      "learning_rate": 0.00013204186801102024,
      "loss": 0.4226,
      "num_input_tokens_seen": 19590584,
      "step": 30030
    },
    {
      "epoch": 15.7416142557652,
      "grad_norm": 0.2292691320180893,
      "learning_rate": 0.0001318870696911758,
      "loss": 0.4225,
      "num_input_tokens_seen": 19594264,
      "step": 30035
    },
    {
      "epoch": 15.744234800838575,
      "grad_norm": 0.4074522852897644,
      "learning_rate": 0.00013173234837635656,
      "loss": 0.4624,
      "num_input_tokens_seen": 19597176,
      "step": 30040
    },
    {
      "epoch": 15.74685534591195,
      "grad_norm": 0.1230587288737297,
      "learning_rate": 0.00013157770409892878,
      "loss": 0.5331,
      "num_input_tokens_seen": 19599896,
      "step": 30045
    },
    {
      "epoch": 15.749475890985325,
      "grad_norm": 0.16348765790462494,
      "learning_rate": 0.00013142313689124197,
      "loss": 0.4131,
      "num_input_tokens_seen": 19602904,
      "step": 30050
    },
    {
      "epoch": 15.7520964360587,
      "grad_norm": 0.09581664949655533,
      "learning_rate": 0.00013126864678562994,
      "loss": 0.4576,
      "num_input_tokens_seen": 19606456,
      "step": 30055
    },
    {
      "epoch": 15.754716981132075,
      "grad_norm": 0.10323064774274826,
      "learning_rate": 0.00013111423381441034,
      "loss": 0.3397,
      "num_input_tokens_seen": 19609304,
      "step": 30060
    },
    {
      "epoch": 15.75733752620545,
      "grad_norm": 0.10117333382368088,
      "learning_rate": 0.00013095989800988468,
      "loss": 0.4078,
      "num_input_tokens_seen": 19613304,
      "step": 30065
    },
    {
      "epoch": 15.759958071278826,
      "grad_norm": 0.14700786769390106,
      "learning_rate": 0.000130805639404338,
      "loss": 0.4266,
      "num_input_tokens_seen": 19616952,
      "step": 30070
    },
    {
      "epoch": 15.7625786163522,
      "grad_norm": 0.28309696912765503,
      "learning_rate": 0.0001306514580300397,
      "loss": 0.3883,
      "num_input_tokens_seen": 19620056,
      "step": 30075
    },
    {
      "epoch": 15.765199161425576,
      "grad_norm": 0.13338115811347961,
      "learning_rate": 0.00013049735391924244,
      "loss": 0.3649,
      "num_input_tokens_seen": 19623096,
      "step": 30080
    },
    {
      "epoch": 15.767819706498951,
      "grad_norm": 0.1343255341053009,
      "learning_rate": 0.00013034332710418328,
      "loss": 0.6299,
      "num_input_tokens_seen": 19625624,
      "step": 30085
    },
    {
      "epoch": 15.770440251572326,
      "grad_norm": 0.20880882441997528,
      "learning_rate": 0.00013018937761708298,
      "loss": 0.4441,
      "num_input_tokens_seen": 19628312,
      "step": 30090
    },
    {
      "epoch": 15.773060796645701,
      "grad_norm": 0.21439456939697266,
      "learning_rate": 0.0001300355054901457,
      "loss": 0.3688,
      "num_input_tokens_seen": 19632056,
      "step": 30095
    },
    {
      "epoch": 15.775681341719078,
      "grad_norm": 0.1932409256696701,
      "learning_rate": 0.0001298817107555599,
      "loss": 0.4152,
      "num_input_tokens_seen": 19637560,
      "step": 30100
    },
    {
      "epoch": 15.778301886792454,
      "grad_norm": 0.17459364235401154,
      "learning_rate": 0.0001297279934454978,
      "loss": 0.4423,
      "num_input_tokens_seen": 19640600,
      "step": 30105
    },
    {
      "epoch": 15.780922431865829,
      "grad_norm": 0.08969135582447052,
      "learning_rate": 0.000129574353592115,
      "loss": 0.4034,
      "num_input_tokens_seen": 19643608,
      "step": 30110
    },
    {
      "epoch": 15.783542976939204,
      "grad_norm": 0.16696813702583313,
      "learning_rate": 0.00012942079122755162,
      "loss": 0.5074,
      "num_input_tokens_seen": 19646040,
      "step": 30115
    },
    {
      "epoch": 15.786163522012579,
      "grad_norm": 0.16968786716461182,
      "learning_rate": 0.00012926730638393075,
      "loss": 0.4496,
      "num_input_tokens_seen": 19648600,
      "step": 30120
    },
    {
      "epoch": 15.788784067085954,
      "grad_norm": 0.11532247811555862,
      "learning_rate": 0.0001291138990933598,
      "loss": 0.3166,
      "num_input_tokens_seen": 19652280,
      "step": 30125
    },
    {
      "epoch": 15.79140461215933,
      "grad_norm": 0.14871463179588318,
      "learning_rate": 0.00012896056938792994,
      "loss": 0.4502,
      "num_input_tokens_seen": 19655736,
      "step": 30130
    },
    {
      "epoch": 15.794025157232705,
      "grad_norm": 0.12205573171377182,
      "learning_rate": 0.000128807317299716,
      "loss": 0.4979,
      "num_input_tokens_seen": 19658456,
      "step": 30135
    },
    {
      "epoch": 15.79664570230608,
      "grad_norm": 0.10947472602128983,
      "learning_rate": 0.00012865414286077637,
      "loss": 0.4034,
      "num_input_tokens_seen": 19661688,
      "step": 30140
    },
    {
      "epoch": 15.799266247379455,
      "grad_norm": 0.18668943643569946,
      "learning_rate": 0.00012850104610315365,
      "loss": 0.4492,
      "num_input_tokens_seen": 19665400,
      "step": 30145
    },
    {
      "epoch": 15.80188679245283,
      "grad_norm": 0.13627296686172485,
      "learning_rate": 0.00012834802705887372,
      "loss": 0.4033,
      "num_input_tokens_seen": 19668472,
      "step": 30150
    },
    {
      "epoch": 15.804507337526205,
      "grad_norm": 0.17507211863994598,
      "learning_rate": 0.00012819508575994643,
      "loss": 0.3303,
      "num_input_tokens_seen": 19672088,
      "step": 30155
    },
    {
      "epoch": 15.80712788259958,
      "grad_norm": 0.11824383586645126,
      "learning_rate": 0.0001280422222383656,
      "loss": 0.4049,
      "num_input_tokens_seen": 19675864,
      "step": 30160
    },
    {
      "epoch": 15.809748427672956,
      "grad_norm": 0.10108453035354614,
      "learning_rate": 0.00012788943652610818,
      "loss": 0.3846,
      "num_input_tokens_seen": 19680216,
      "step": 30165
    },
    {
      "epoch": 15.81236897274633,
      "grad_norm": 0.17769873142242432,
      "learning_rate": 0.0001277367286551356,
      "loss": 0.5586,
      "num_input_tokens_seen": 19683480,
      "step": 30170
    },
    {
      "epoch": 15.814989517819706,
      "grad_norm": 0.11426595598459244,
      "learning_rate": 0.00012758409865739213,
      "loss": 0.5224,
      "num_input_tokens_seen": 19687160,
      "step": 30175
    },
    {
      "epoch": 15.817610062893081,
      "grad_norm": 0.07881252467632294,
      "learning_rate": 0.00012743154656480655,
      "loss": 0.5075,
      "num_input_tokens_seen": 19690424,
      "step": 30180
    },
    {
      "epoch": 15.820230607966456,
      "grad_norm": 0.08513443171977997,
      "learning_rate": 0.00012727907240929094,
      "loss": 0.414,
      "num_input_tokens_seen": 19693656,
      "step": 30185
    },
    {
      "epoch": 15.822851153039831,
      "grad_norm": 0.16564840078353882,
      "learning_rate": 0.00012712667622274127,
      "loss": 0.5463,
      "num_input_tokens_seen": 19696920,
      "step": 30190
    },
    {
      "epoch": 15.825471698113208,
      "grad_norm": 0.14120516180992126,
      "learning_rate": 0.000126974358037037,
      "loss": 0.3071,
      "num_input_tokens_seen": 19700088,
      "step": 30195
    },
    {
      "epoch": 15.828092243186584,
      "grad_norm": 0.1396002620458603,
      "learning_rate": 0.00012682211788404114,
      "loss": 0.3957,
      "num_input_tokens_seen": 19703320,
      "step": 30200
    },
    {
      "epoch": 15.830712788259959,
      "grad_norm": 0.2703501582145691,
      "learning_rate": 0.0001266699557956008,
      "loss": 0.4318,
      "num_input_tokens_seen": 19706168,
      "step": 30205
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 0.18732662498950958,
      "learning_rate": 0.00012651787180354657,
      "loss": 0.3359,
      "num_input_tokens_seen": 19708984,
      "step": 30210
    },
    {
      "epoch": 15.835953878406709,
      "grad_norm": 0.14766062796115875,
      "learning_rate": 0.00012636586593969262,
      "loss": 0.4321,
      "num_input_tokens_seen": 19711928,
      "step": 30215
    },
    {
      "epoch": 15.838574423480084,
      "grad_norm": 0.11890235543251038,
      "learning_rate": 0.000126213938235837,
      "loss": 0.4331,
      "num_input_tokens_seen": 19715640,
      "step": 30220
    },
    {
      "epoch": 15.84119496855346,
      "grad_norm": 0.1018020510673523,
      "learning_rate": 0.0001260620887237611,
      "loss": 0.3873,
      "num_input_tokens_seen": 19718616,
      "step": 30225
    },
    {
      "epoch": 15.843815513626835,
      "grad_norm": 0.1408502608537674,
      "learning_rate": 0.00012591031743523006,
      "loss": 0.3962,
      "num_input_tokens_seen": 19721688,
      "step": 30230
    },
    {
      "epoch": 15.84643605870021,
      "grad_norm": 0.14703086018562317,
      "learning_rate": 0.00012575862440199277,
      "loss": 0.4294,
      "num_input_tokens_seen": 19725272,
      "step": 30235
    },
    {
      "epoch": 15.849056603773585,
      "grad_norm": 0.09858440607786179,
      "learning_rate": 0.00012560700965578165,
      "loss": 0.3444,
      "num_input_tokens_seen": 19727960,
      "step": 30240
    },
    {
      "epoch": 15.85167714884696,
      "grad_norm": 0.20645059645175934,
      "learning_rate": 0.0001254554732283129,
      "loss": 0.4998,
      "num_input_tokens_seen": 19730904,
      "step": 30245
    },
    {
      "epoch": 15.854297693920335,
      "grad_norm": 0.11738118529319763,
      "learning_rate": 0.0001253040151512862,
      "loss": 0.5072,
      "num_input_tokens_seen": 19734552,
      "step": 30250
    },
    {
      "epoch": 15.85691823899371,
      "grad_norm": 0.11146173626184464,
      "learning_rate": 0.0001251526354563846,
      "loss": 0.3913,
      "num_input_tokens_seen": 19737400,
      "step": 30255
    },
    {
      "epoch": 15.859538784067086,
      "grad_norm": 0.11786989867687225,
      "learning_rate": 0.0001250013341752752,
      "loss": 0.39,
      "num_input_tokens_seen": 19740792,
      "step": 30260
    },
    {
      "epoch": 15.86215932914046,
      "grad_norm": 0.10250793397426605,
      "learning_rate": 0.00012485011133960843,
      "loss": 0.4468,
      "num_input_tokens_seen": 19744248,
      "step": 30265
    },
    {
      "epoch": 15.864779874213836,
      "grad_norm": 0.07986263185739517,
      "learning_rate": 0.00012469896698101863,
      "loss": 0.3047,
      "num_input_tokens_seen": 19746744,
      "step": 30270
    },
    {
      "epoch": 15.867400419287211,
      "grad_norm": 0.16583207249641418,
      "learning_rate": 0.00012454790113112325,
      "loss": 0.4376,
      "num_input_tokens_seen": 19749688,
      "step": 30275
    },
    {
      "epoch": 15.870020964360586,
      "grad_norm": 0.1361740082502365,
      "learning_rate": 0.00012439691382152345,
      "loss": 0.3633,
      "num_input_tokens_seen": 19753208,
      "step": 30280
    },
    {
      "epoch": 15.872641509433961,
      "grad_norm": 0.10927627980709076,
      "learning_rate": 0.00012424600508380412,
      "loss": 0.4952,
      "num_input_tokens_seen": 19756568,
      "step": 30285
    },
    {
      "epoch": 15.875262054507338,
      "grad_norm": 0.17838948965072632,
      "learning_rate": 0.00012409517494953377,
      "loss": 0.4182,
      "num_input_tokens_seen": 19759576,
      "step": 30290
    },
    {
      "epoch": 15.877882599580714,
      "grad_norm": 0.16228222846984863,
      "learning_rate": 0.00012394442345026418,
      "loss": 0.3334,
      "num_input_tokens_seen": 19763608,
      "step": 30295
    },
    {
      "epoch": 15.880503144654089,
      "grad_norm": 0.10526114702224731,
      "learning_rate": 0.00012379375061753118,
      "loss": 0.6235,
      "num_input_tokens_seen": 19766808,
      "step": 30300
    },
    {
      "epoch": 15.883123689727464,
      "grad_norm": 0.1097661629319191,
      "learning_rate": 0.00012364315648285353,
      "loss": 0.3229,
      "num_input_tokens_seen": 19769752,
      "step": 30305
    },
    {
      "epoch": 15.885744234800839,
      "grad_norm": 0.16450995206832886,
      "learning_rate": 0.00012349264107773363,
      "loss": 0.5912,
      "num_input_tokens_seen": 19772920,
      "step": 30310
    },
    {
      "epoch": 15.888364779874214,
      "grad_norm": 0.13199691474437714,
      "learning_rate": 0.00012334220443365785,
      "loss": 0.4032,
      "num_input_tokens_seen": 19775512,
      "step": 30315
    },
    {
      "epoch": 15.89098532494759,
      "grad_norm": 0.1111537516117096,
      "learning_rate": 0.00012319184658209575,
      "loss": 0.5166,
      "num_input_tokens_seen": 19778968,
      "step": 30320
    },
    {
      "epoch": 15.893605870020965,
      "grad_norm": 0.14482031762599945,
      "learning_rate": 0.00012304156755450063,
      "loss": 0.3613,
      "num_input_tokens_seen": 19781656,
      "step": 30325
    },
    {
      "epoch": 15.89622641509434,
      "grad_norm": 0.1880887895822525,
      "learning_rate": 0.00012289136738230906,
      "loss": 0.472,
      "num_input_tokens_seen": 19785112,
      "step": 30330
    },
    {
      "epoch": 15.898846960167715,
      "grad_norm": 0.11728495359420776,
      "learning_rate": 0.00012274124609694092,
      "loss": 0.4482,
      "num_input_tokens_seen": 19788088,
      "step": 30335
    },
    {
      "epoch": 15.90146750524109,
      "grad_norm": 0.1962834596633911,
      "learning_rate": 0.00012259120372980014,
      "loss": 0.3019,
      "num_input_tokens_seen": 19791352,
      "step": 30340
    },
    {
      "epoch": 15.904088050314465,
      "grad_norm": 0.180393248796463,
      "learning_rate": 0.0001224412403122739,
      "loss": 0.4165,
      "num_input_tokens_seen": 19795288,
      "step": 30345
    },
    {
      "epoch": 15.90670859538784,
      "grad_norm": 0.14006277918815613,
      "learning_rate": 0.0001222913558757328,
      "loss": 0.4378,
      "num_input_tokens_seen": 19798328,
      "step": 30350
    },
    {
      "epoch": 15.909329140461216,
      "grad_norm": 0.13077200949192047,
      "learning_rate": 0.00012214155045153085,
      "loss": 0.3209,
      "num_input_tokens_seen": 19801432,
      "step": 30355
    },
    {
      "epoch": 15.91194968553459,
      "grad_norm": 0.18932020664215088,
      "learning_rate": 0.00012199182407100584,
      "loss": 0.5068,
      "num_input_tokens_seen": 19804280,
      "step": 30360
    },
    {
      "epoch": 15.914570230607966,
      "grad_norm": 0.08055941760540009,
      "learning_rate": 0.00012184217676547855,
      "loss": 0.3706,
      "num_input_tokens_seen": 19807768,
      "step": 30365
    },
    {
      "epoch": 15.917190775681341,
      "grad_norm": 0.11632480472326279,
      "learning_rate": 0.00012169260856625358,
      "loss": 0.41,
      "num_input_tokens_seen": 19810616,
      "step": 30370
    },
    {
      "epoch": 15.919811320754716,
      "grad_norm": 0.1199265718460083,
      "learning_rate": 0.000121543119504619,
      "loss": 0.364,
      "num_input_tokens_seen": 19816472,
      "step": 30375
    },
    {
      "epoch": 15.922431865828091,
      "grad_norm": 0.13946335017681122,
      "learning_rate": 0.00012139370961184626,
      "loss": 0.3635,
      "num_input_tokens_seen": 19819928,
      "step": 30380
    },
    {
      "epoch": 15.925052410901468,
      "grad_norm": 0.14925384521484375,
      "learning_rate": 0.00012124437891918994,
      "loss": 0.4202,
      "num_input_tokens_seen": 19823704,
      "step": 30385
    },
    {
      "epoch": 15.927672955974844,
      "grad_norm": 0.1774970293045044,
      "learning_rate": 0.0001210951274578887,
      "loss": 0.4521,
      "num_input_tokens_seen": 19826456,
      "step": 30390
    },
    {
      "epoch": 15.930293501048219,
      "grad_norm": 0.0989966168999672,
      "learning_rate": 0.00012094595525916379,
      "loss": 0.4113,
      "num_input_tokens_seen": 19829848,
      "step": 30395
    },
    {
      "epoch": 15.932914046121594,
      "grad_norm": 0.1681496649980545,
      "learning_rate": 0.0001207968623542206,
      "loss": 0.4449,
      "num_input_tokens_seen": 19832856,
      "step": 30400
    },
    {
      "epoch": 15.935534591194969,
      "grad_norm": 0.12104509770870209,
      "learning_rate": 0.00012064784877424778,
      "loss": 0.38,
      "num_input_tokens_seen": 19835704,
      "step": 30405
    },
    {
      "epoch": 15.938155136268344,
      "grad_norm": 0.17429307103157043,
      "learning_rate": 0.00012049891455041695,
      "loss": 0.3946,
      "num_input_tokens_seen": 19839352,
      "step": 30410
    },
    {
      "epoch": 15.94077568134172,
      "grad_norm": 0.21822631359100342,
      "learning_rate": 0.00012035005971388379,
      "loss": 0.4124,
      "num_input_tokens_seen": 19842840,
      "step": 30415
    },
    {
      "epoch": 15.943396226415095,
      "grad_norm": 0.1021294817328453,
      "learning_rate": 0.0001202012842957867,
      "loss": 0.3755,
      "num_input_tokens_seen": 19846808,
      "step": 30420
    },
    {
      "epoch": 15.94601677148847,
      "grad_norm": 0.09499458223581314,
      "learning_rate": 0.00012005258832724797,
      "loss": 0.4183,
      "num_input_tokens_seen": 19849816,
      "step": 30425
    },
    {
      "epoch": 15.948637316561845,
      "grad_norm": 0.14405983686447144,
      "learning_rate": 0.00011990397183937324,
      "loss": 0.3338,
      "num_input_tokens_seen": 19852568,
      "step": 30430
    },
    {
      "epoch": 15.95125786163522,
      "grad_norm": 0.13846901059150696,
      "learning_rate": 0.00011975543486325108,
      "loss": 0.2887,
      "num_input_tokens_seen": 19857976,
      "step": 30435
    },
    {
      "epoch": 15.953878406708595,
      "grad_norm": 0.17773815989494324,
      "learning_rate": 0.00011960697742995392,
      "loss": 0.4413,
      "num_input_tokens_seen": 19861432,
      "step": 30440
    },
    {
      "epoch": 15.95649895178197,
      "grad_norm": 0.09339766204357147,
      "learning_rate": 0.00011945859957053746,
      "loss": 0.4873,
      "num_input_tokens_seen": 19865144,
      "step": 30445
    },
    {
      "epoch": 15.959119496855346,
      "grad_norm": 0.10739342868328094,
      "learning_rate": 0.00011931030131604036,
      "loss": 0.3584,
      "num_input_tokens_seen": 19867928,
      "step": 30450
    },
    {
      "epoch": 15.96174004192872,
      "grad_norm": 0.12263453751802444,
      "learning_rate": 0.00011916208269748507,
      "loss": 0.3943,
      "num_input_tokens_seen": 19870456,
      "step": 30455
    },
    {
      "epoch": 15.964360587002096,
      "grad_norm": 0.18874628841876984,
      "learning_rate": 0.00011901394374587743,
      "loss": 0.3862,
      "num_input_tokens_seen": 19872824,
      "step": 30460
    },
    {
      "epoch": 15.966981132075471,
      "grad_norm": 0.1339806765317917,
      "learning_rate": 0.00011886588449220609,
      "loss": 0.3379,
      "num_input_tokens_seen": 19875256,
      "step": 30465
    },
    {
      "epoch": 15.969601677148846,
      "grad_norm": 0.3302920460700989,
      "learning_rate": 0.00011871790496744351,
      "loss": 0.4733,
      "num_input_tokens_seen": 19878328,
      "step": 30470
    },
    {
      "epoch": 15.972222222222221,
      "grad_norm": 0.3060268759727478,
      "learning_rate": 0.00011857000520254546,
      "loss": 0.4774,
      "num_input_tokens_seen": 19881784,
      "step": 30475
    },
    {
      "epoch": 15.974842767295598,
      "grad_norm": 0.2320416122674942,
      "learning_rate": 0.0001184221852284506,
      "loss": 0.4487,
      "num_input_tokens_seen": 19885144,
      "step": 30480
    },
    {
      "epoch": 15.977463312368974,
      "grad_norm": 0.15566125512123108,
      "learning_rate": 0.00011827444507608143,
      "loss": 0.5031,
      "num_input_tokens_seen": 19888120,
      "step": 30485
    },
    {
      "epoch": 15.980083857442349,
      "grad_norm": 0.12290009111166,
      "learning_rate": 0.00011812678477634325,
      "loss": 0.4972,
      "num_input_tokens_seen": 19891352,
      "step": 30490
    },
    {
      "epoch": 15.982704402515724,
      "grad_norm": 0.17360088229179382,
      "learning_rate": 0.0001179792043601251,
      "loss": 0.4906,
      "num_input_tokens_seen": 19894872,
      "step": 30495
    },
    {
      "epoch": 15.985324947589099,
      "grad_norm": 0.060830555856227875,
      "learning_rate": 0.00011783170385829905,
      "loss": 0.3853,
      "num_input_tokens_seen": 19898904,
      "step": 30500
    },
    {
      "epoch": 15.987945492662474,
      "grad_norm": 0.11555638909339905,
      "learning_rate": 0.00011768428330172071,
      "loss": 0.4504,
      "num_input_tokens_seen": 19902840,
      "step": 30505
    },
    {
      "epoch": 15.99056603773585,
      "grad_norm": 0.14610633254051208,
      "learning_rate": 0.00011753694272122856,
      "loss": 0.3715,
      "num_input_tokens_seen": 19906168,
      "step": 30510
    },
    {
      "epoch": 15.993186582809225,
      "grad_norm": 0.17831425368785858,
      "learning_rate": 0.00011738968214764456,
      "loss": 0.6486,
      "num_input_tokens_seen": 19908920,
      "step": 30515
    },
    {
      "epoch": 15.9958071278826,
      "grad_norm": 0.18820500373840332,
      "learning_rate": 0.00011724250161177391,
      "loss": 0.5054,
      "num_input_tokens_seen": 19912216,
      "step": 30520
    },
    {
      "epoch": 15.998427672955975,
      "grad_norm": 0.15700814127922058,
      "learning_rate": 0.00011709540114440525,
      "loss": 0.458,
      "num_input_tokens_seen": 19914904,
      "step": 30525
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.46345266699790955,
      "eval_runtime": 13.5311,
      "eval_samples_per_second": 62.67,
      "eval_steps_per_second": 15.668,
      "num_input_tokens_seen": 19916008,
      "step": 30528
    },
    {
      "epoch": 16.00104821802935,
      "grad_norm": 0.17702975869178772,
      "learning_rate": 0.00011694838077631043,
      "loss": 0.4557,
      "num_input_tokens_seen": 19917032,
      "step": 30530
    },
    {
      "epoch": 16.003668763102727,
      "grad_norm": 0.15280389785766602,
      "learning_rate": 0.0001168014405382441,
      "loss": 0.5499,
      "num_input_tokens_seen": 19919976,
      "step": 30535
    },
    {
      "epoch": 16.0062893081761,
      "grad_norm": 0.2194783240556717,
      "learning_rate": 0.0001166545804609448,
      "loss": 0.4634,
      "num_input_tokens_seen": 19922312,
      "step": 30540
    },
    {
      "epoch": 16.008909853249477,
      "grad_norm": 0.2462158501148224,
      "learning_rate": 0.00011650780057513367,
      "loss": 0.5355,
      "num_input_tokens_seen": 19924392,
      "step": 30545
    },
    {
      "epoch": 16.01153039832285,
      "grad_norm": 0.15317374467849731,
      "learning_rate": 0.00011636110091151553,
      "loss": 0.4273,
      "num_input_tokens_seen": 19927528,
      "step": 30550
    },
    {
      "epoch": 16.014150943396228,
      "grad_norm": 0.09154576808214188,
      "learning_rate": 0.00011621448150077834,
      "loss": 0.4002,
      "num_input_tokens_seen": 19930920,
      "step": 30555
    },
    {
      "epoch": 16.0167714884696,
      "grad_norm": 0.12198320776224136,
      "learning_rate": 0.0001160679423735933,
      "loss": 0.3461,
      "num_input_tokens_seen": 19934152,
      "step": 30560
    },
    {
      "epoch": 16.019392033542978,
      "grad_norm": 0.22019396722316742,
      "learning_rate": 0.00011592148356061455,
      "loss": 0.3317,
      "num_input_tokens_seen": 19938216,
      "step": 30565
    },
    {
      "epoch": 16.02201257861635,
      "grad_norm": 0.094613216817379,
      "learning_rate": 0.00011577510509247951,
      "loss": 0.5036,
      "num_input_tokens_seen": 19942088,
      "step": 30570
    },
    {
      "epoch": 16.02463312368973,
      "grad_norm": 0.16358411312103271,
      "learning_rate": 0.00011562880699980904,
      "loss": 0.7358,
      "num_input_tokens_seen": 19944744,
      "step": 30575
    },
    {
      "epoch": 16.0272536687631,
      "grad_norm": 0.10773386806249619,
      "learning_rate": 0.00011548258931320704,
      "loss": 0.4188,
      "num_input_tokens_seen": 19948360,
      "step": 30580
    },
    {
      "epoch": 16.02987421383648,
      "grad_norm": 0.1731019765138626,
      "learning_rate": 0.00011533645206326049,
      "loss": 0.4158,
      "num_input_tokens_seen": 19951016,
      "step": 30585
    },
    {
      "epoch": 16.032494758909852,
      "grad_norm": 0.17363691329956055,
      "learning_rate": 0.00011519039528053999,
      "loss": 0.4045,
      "num_input_tokens_seen": 19954376,
      "step": 30590
    },
    {
      "epoch": 16.03511530398323,
      "grad_norm": 0.1870976686477661,
      "learning_rate": 0.00011504441899559837,
      "loss": 0.4313,
      "num_input_tokens_seen": 19957864,
      "step": 30595
    },
    {
      "epoch": 16.037735849056602,
      "grad_norm": 0.1265399008989334,
      "learning_rate": 0.00011489852323897249,
      "loss": 0.3966,
      "num_input_tokens_seen": 19961256,
      "step": 30600
    },
    {
      "epoch": 16.04035639412998,
      "grad_norm": 0.10803086310625076,
      "learning_rate": 0.0001147527080411821,
      "loss": 0.3763,
      "num_input_tokens_seen": 19964744,
      "step": 30605
    },
    {
      "epoch": 16.042976939203353,
      "grad_norm": 0.08638317883014679,
      "learning_rate": 0.00011460697343273002,
      "loss": 0.4569,
      "num_input_tokens_seen": 19968232,
      "step": 30610
    },
    {
      "epoch": 16.04559748427673,
      "grad_norm": 0.10140921175479889,
      "learning_rate": 0.00011446131944410249,
      "loss": 0.3914,
      "num_input_tokens_seen": 19971176,
      "step": 30615
    },
    {
      "epoch": 16.048218029350103,
      "grad_norm": 0.1606733351945877,
      "learning_rate": 0.00011431574610576844,
      "loss": 0.4242,
      "num_input_tokens_seen": 19974888,
      "step": 30620
    },
    {
      "epoch": 16.05083857442348,
      "grad_norm": 0.11033955961465836,
      "learning_rate": 0.00011417025344818005,
      "loss": 0.3762,
      "num_input_tokens_seen": 19977864,
      "step": 30625
    },
    {
      "epoch": 16.053459119496857,
      "grad_norm": 0.16784439980983734,
      "learning_rate": 0.00011402484150177289,
      "loss": 0.4874,
      "num_input_tokens_seen": 19981640,
      "step": 30630
    },
    {
      "epoch": 16.05607966457023,
      "grad_norm": 0.1220802590250969,
      "learning_rate": 0.00011387951029696542,
      "loss": 0.5064,
      "num_input_tokens_seen": 19985640,
      "step": 30635
    },
    {
      "epoch": 16.058700209643607,
      "grad_norm": 0.12946969270706177,
      "learning_rate": 0.00011373425986415941,
      "loss": 0.3329,
      "num_input_tokens_seen": 19988776,
      "step": 30640
    },
    {
      "epoch": 16.06132075471698,
      "grad_norm": 0.17847421765327454,
      "learning_rate": 0.00011358909023373953,
      "loss": 0.4331,
      "num_input_tokens_seen": 19991976,
      "step": 30645
    },
    {
      "epoch": 16.063941299790358,
      "grad_norm": 0.19230735301971436,
      "learning_rate": 0.00011344400143607342,
      "loss": 0.3977,
      "num_input_tokens_seen": 19995048,
      "step": 30650
    },
    {
      "epoch": 16.06656184486373,
      "grad_norm": 0.09157437086105347,
      "learning_rate": 0.00011329899350151212,
      "loss": 0.4625,
      "num_input_tokens_seen": 19997992,
      "step": 30655
    },
    {
      "epoch": 16.069182389937108,
      "grad_norm": 0.19865529239177704,
      "learning_rate": 0.00011315406646038973,
      "loss": 0.3509,
      "num_input_tokens_seen": 20000520,
      "step": 30660
    },
    {
      "epoch": 16.07180293501048,
      "grad_norm": 0.1669904589653015,
      "learning_rate": 0.0001130092203430232,
      "loss": 0.5848,
      "num_input_tokens_seen": 20003176,
      "step": 30665
    },
    {
      "epoch": 16.07442348008386,
      "grad_norm": 0.13461807370185852,
      "learning_rate": 0.00011286445517971289,
      "loss": 0.5221,
      "num_input_tokens_seen": 20006664,
      "step": 30670
    },
    {
      "epoch": 16.07704402515723,
      "grad_norm": 0.12025267630815506,
      "learning_rate": 0.00011271977100074188,
      "loss": 0.4228,
      "num_input_tokens_seen": 20009320,
      "step": 30675
    },
    {
      "epoch": 16.07966457023061,
      "grad_norm": 0.3383345603942871,
      "learning_rate": 0.00011257516783637633,
      "loss": 0.3244,
      "num_input_tokens_seen": 20011656,
      "step": 30680
    },
    {
      "epoch": 16.082285115303982,
      "grad_norm": 0.11650027334690094,
      "learning_rate": 0.00011243064571686573,
      "loss": 0.4309,
      "num_input_tokens_seen": 20015304,
      "step": 30685
    },
    {
      "epoch": 16.08490566037736,
      "grad_norm": 0.1378200203180313,
      "learning_rate": 0.00011228620467244238,
      "loss": 0.3775,
      "num_input_tokens_seen": 20018504,
      "step": 30690
    },
    {
      "epoch": 16.087526205450732,
      "grad_norm": 0.11323405057191849,
      "learning_rate": 0.00011214184473332183,
      "loss": 0.3622,
      "num_input_tokens_seen": 20021096,
      "step": 30695
    },
    {
      "epoch": 16.09014675052411,
      "grad_norm": 0.15665562450885773,
      "learning_rate": 0.0001119975659297025,
      "loss": 0.4632,
      "num_input_tokens_seen": 20023912,
      "step": 30700
    },
    {
      "epoch": 16.092767295597483,
      "grad_norm": 0.08783762902021408,
      "learning_rate": 0.00011185336829176568,
      "loss": 0.4637,
      "num_input_tokens_seen": 20026792,
      "step": 30705
    },
    {
      "epoch": 16.09538784067086,
      "grad_norm": 0.22132311761379242,
      "learning_rate": 0.00011170925184967601,
      "loss": 0.4326,
      "num_input_tokens_seen": 20028840,
      "step": 30710
    },
    {
      "epoch": 16.098008385744233,
      "grad_norm": 0.09129929542541504,
      "learning_rate": 0.00011156521663358094,
      "loss": 0.4113,
      "num_input_tokens_seen": 20032072,
      "step": 30715
    },
    {
      "epoch": 16.10062893081761,
      "grad_norm": 0.12718802690505981,
      "learning_rate": 0.00011142126267361124,
      "loss": 0.375,
      "num_input_tokens_seen": 20034824,
      "step": 30720
    },
    {
      "epoch": 16.103249475890987,
      "grad_norm": 0.08188159018754959,
      "learning_rate": 0.00011127738999988007,
      "loss": 0.3139,
      "num_input_tokens_seen": 20038792,
      "step": 30725
    },
    {
      "epoch": 16.10587002096436,
      "grad_norm": 0.13429316878318787,
      "learning_rate": 0.00011113359864248429,
      "loss": 0.3267,
      "num_input_tokens_seen": 20042024,
      "step": 30730
    },
    {
      "epoch": 16.108490566037737,
      "grad_norm": 0.21684662997722626,
      "learning_rate": 0.00011098988863150317,
      "loss": 0.5344,
      "num_input_tokens_seen": 20045576,
      "step": 30735
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 0.10318909585475922,
      "learning_rate": 0.0001108462599969992,
      "loss": 0.3486,
      "num_input_tokens_seen": 20048872,
      "step": 30740
    },
    {
      "epoch": 16.113731656184488,
      "grad_norm": 0.16548524796962738,
      "learning_rate": 0.000110702712769018,
      "loss": 0.477,
      "num_input_tokens_seen": 20053096,
      "step": 30745
    },
    {
      "epoch": 16.11635220125786,
      "grad_norm": 0.13731379806995392,
      "learning_rate": 0.00011055924697758801,
      "loss": 0.4297,
      "num_input_tokens_seen": 20056296,
      "step": 30750
    },
    {
      "epoch": 16.118972746331238,
      "grad_norm": 0.23215575516223907,
      "learning_rate": 0.00011041586265272052,
      "loss": 0.4966,
      "num_input_tokens_seen": 20059368,
      "step": 30755
    },
    {
      "epoch": 16.12159329140461,
      "grad_norm": 0.18686401844024658,
      "learning_rate": 0.00011027255982441003,
      "loss": 0.3069,
      "num_input_tokens_seen": 20062472,
      "step": 30760
    },
    {
      "epoch": 16.12421383647799,
      "grad_norm": 0.15798790752887726,
      "learning_rate": 0.00011012933852263369,
      "loss": 0.3563,
      "num_input_tokens_seen": 20066024,
      "step": 30765
    },
    {
      "epoch": 16.12683438155136,
      "grad_norm": 0.15736964344978333,
      "learning_rate": 0.00010998619877735184,
      "loss": 0.369,
      "num_input_tokens_seen": 20068968,
      "step": 30770
    },
    {
      "epoch": 16.12945492662474,
      "grad_norm": 0.13373002409934998,
      "learning_rate": 0.0001098431406185078,
      "loss": 0.4256,
      "num_input_tokens_seen": 20071720,
      "step": 30775
    },
    {
      "epoch": 16.132075471698112,
      "grad_norm": 0.1329547017812729,
      "learning_rate": 0.00010970016407602751,
      "loss": 0.641,
      "num_input_tokens_seen": 20074984,
      "step": 30780
    },
    {
      "epoch": 16.13469601677149,
      "grad_norm": 0.10236281156539917,
      "learning_rate": 0.00010955726917982023,
      "loss": 0.5468,
      "num_input_tokens_seen": 20078760,
      "step": 30785
    },
    {
      "epoch": 16.137316561844862,
      "grad_norm": 0.22074049711227417,
      "learning_rate": 0.00010941445595977767,
      "loss": 0.3812,
      "num_input_tokens_seen": 20081480,
      "step": 30790
    },
    {
      "epoch": 16.13993710691824,
      "grad_norm": 0.14628082513809204,
      "learning_rate": 0.00010927172444577494,
      "loss": 0.4524,
      "num_input_tokens_seen": 20084072,
      "step": 30795
    },
    {
      "epoch": 16.142557651991613,
      "grad_norm": 0.1404198557138443,
      "learning_rate": 0.00010912907466766985,
      "loss": 0.5374,
      "num_input_tokens_seen": 20088488,
      "step": 30800
    },
    {
      "epoch": 16.14517819706499,
      "grad_norm": 0.09443037956953049,
      "learning_rate": 0.00010898650665530302,
      "loss": 0.4173,
      "num_input_tokens_seen": 20092008,
      "step": 30805
    },
    {
      "epoch": 16.147798742138363,
      "grad_norm": 0.16816122829914093,
      "learning_rate": 0.000108844020438498,
      "loss": 0.4724,
      "num_input_tokens_seen": 20094664,
      "step": 30810
    },
    {
      "epoch": 16.15041928721174,
      "grad_norm": 0.20449943840503693,
      "learning_rate": 0.00010870161604706152,
      "loss": 0.3784,
      "num_input_tokens_seen": 20101640,
      "step": 30815
    },
    {
      "epoch": 16.153039832285117,
      "grad_norm": 0.16353926062583923,
      "learning_rate": 0.00010855929351078264,
      "loss": 0.4792,
      "num_input_tokens_seen": 20104840,
      "step": 30820
    },
    {
      "epoch": 16.15566037735849,
      "grad_norm": 0.08306337893009186,
      "learning_rate": 0.00010841705285943382,
      "loss": 0.4991,
      "num_input_tokens_seen": 20108456,
      "step": 30825
    },
    {
      "epoch": 16.158280922431867,
      "grad_norm": 0.1267094910144806,
      "learning_rate": 0.0001082748941227702,
      "loss": 0.5275,
      "num_input_tokens_seen": 20110792,
      "step": 30830
    },
    {
      "epoch": 16.16090146750524,
      "grad_norm": 0.23688729107379913,
      "learning_rate": 0.00010813281733052959,
      "loss": 0.3189,
      "num_input_tokens_seen": 20113256,
      "step": 30835
    },
    {
      "epoch": 16.163522012578618,
      "grad_norm": 0.17959046363830566,
      "learning_rate": 0.00010799082251243292,
      "loss": 0.3421,
      "num_input_tokens_seen": 20116424,
      "step": 30840
    },
    {
      "epoch": 16.16614255765199,
      "grad_norm": 0.14434632658958435,
      "learning_rate": 0.00010784890969818407,
      "loss": 0.4724,
      "num_input_tokens_seen": 20120200,
      "step": 30845
    },
    {
      "epoch": 16.168763102725368,
      "grad_norm": 0.13160698115825653,
      "learning_rate": 0.00010770707891746928,
      "loss": 0.4551,
      "num_input_tokens_seen": 20122664,
      "step": 30850
    },
    {
      "epoch": 16.17138364779874,
      "grad_norm": 0.1341925859451294,
      "learning_rate": 0.00010756533019995817,
      "loss": 0.4967,
      "num_input_tokens_seen": 20125224,
      "step": 30855
    },
    {
      "epoch": 16.17400419287212,
      "grad_norm": 0.12287598103284836,
      "learning_rate": 0.0001074236635753027,
      "loss": 0.4081,
      "num_input_tokens_seen": 20128968,
      "step": 30860
    },
    {
      "epoch": 16.17662473794549,
      "grad_norm": 0.2223501056432724,
      "learning_rate": 0.00010728207907313809,
      "loss": 0.3962,
      "num_input_tokens_seen": 20131656,
      "step": 30865
    },
    {
      "epoch": 16.17924528301887,
      "grad_norm": 0.19532756507396698,
      "learning_rate": 0.0001071405767230822,
      "loss": 0.4583,
      "num_input_tokens_seen": 20134696,
      "step": 30870
    },
    {
      "epoch": 16.181865828092242,
      "grad_norm": 0.1962505578994751,
      "learning_rate": 0.0001069991565547358,
      "loss": 0.5193,
      "num_input_tokens_seen": 20137192,
      "step": 30875
    },
    {
      "epoch": 16.18448637316562,
      "grad_norm": 0.13258537650108337,
      "learning_rate": 0.00010685781859768223,
      "loss": 0.3216,
      "num_input_tokens_seen": 20140712,
      "step": 30880
    },
    {
      "epoch": 16.187106918238992,
      "grad_norm": 0.181433767080307,
      "learning_rate": 0.00010671656288148768,
      "loss": 0.5536,
      "num_input_tokens_seen": 20143496,
      "step": 30885
    },
    {
      "epoch": 16.18972746331237,
      "grad_norm": 0.11321533471345901,
      "learning_rate": 0.00010657538943570138,
      "loss": 0.3488,
      "num_input_tokens_seen": 20148296,
      "step": 30890
    },
    {
      "epoch": 16.192348008385743,
      "grad_norm": 0.13622814416885376,
      "learning_rate": 0.00010643429828985518,
      "loss": 0.4,
      "num_input_tokens_seen": 20152168,
      "step": 30895
    },
    {
      "epoch": 16.19496855345912,
      "grad_norm": 0.09698141366243362,
      "learning_rate": 0.0001062932894734639,
      "loss": 0.4881,
      "num_input_tokens_seen": 20155784,
      "step": 30900
    },
    {
      "epoch": 16.197589098532493,
      "grad_norm": 0.1190309226512909,
      "learning_rate": 0.00010615236301602476,
      "loss": 0.5365,
      "num_input_tokens_seen": 20159112,
      "step": 30905
    },
    {
      "epoch": 16.20020964360587,
      "grad_norm": 0.1477579027414322,
      "learning_rate": 0.00010601151894701794,
      "loss": 0.3815,
      "num_input_tokens_seen": 20162152,
      "step": 30910
    },
    {
      "epoch": 16.202830188679247,
      "grad_norm": 0.16317443549633026,
      "learning_rate": 0.0001058707572959065,
      "loss": 0.4943,
      "num_input_tokens_seen": 20165032,
      "step": 30915
    },
    {
      "epoch": 16.20545073375262,
      "grad_norm": 0.15983687341213226,
      "learning_rate": 0.00010573007809213614,
      "loss": 0.409,
      "num_input_tokens_seen": 20168136,
      "step": 30920
    },
    {
      "epoch": 16.208071278825997,
      "grad_norm": 0.26350516080856323,
      "learning_rate": 0.00010558948136513536,
      "loss": 0.3598,
      "num_input_tokens_seen": 20171368,
      "step": 30925
    },
    {
      "epoch": 16.21069182389937,
      "grad_norm": 0.19060014188289642,
      "learning_rate": 0.00010544896714431557,
      "loss": 0.2471,
      "num_input_tokens_seen": 20174664,
      "step": 30930
    },
    {
      "epoch": 16.213312368972748,
      "grad_norm": 0.17855200171470642,
      "learning_rate": 0.0001053085354590706,
      "loss": 0.6162,
      "num_input_tokens_seen": 20177224,
      "step": 30935
    },
    {
      "epoch": 16.21593291404612,
      "grad_norm": 0.14711830019950867,
      "learning_rate": 0.00010516818633877695,
      "loss": 0.3922,
      "num_input_tokens_seen": 20180200,
      "step": 30940
    },
    {
      "epoch": 16.218553459119498,
      "grad_norm": 0.1607459932565689,
      "learning_rate": 0.00010502791981279425,
      "loss": 0.3642,
      "num_input_tokens_seen": 20183592,
      "step": 30945
    },
    {
      "epoch": 16.22117400419287,
      "grad_norm": 0.23694635927677155,
      "learning_rate": 0.00010488773591046469,
      "loss": 0.4273,
      "num_input_tokens_seen": 20186760,
      "step": 30950
    },
    {
      "epoch": 16.22379454926625,
      "grad_norm": 0.12278096377849579,
      "learning_rate": 0.0001047476346611132,
      "loss": 0.4017,
      "num_input_tokens_seen": 20189864,
      "step": 30955
    },
    {
      "epoch": 16.22641509433962,
      "grad_norm": 0.10947147756814957,
      "learning_rate": 0.00010460761609404724,
      "loss": 0.4602,
      "num_input_tokens_seen": 20193096,
      "step": 30960
    },
    {
      "epoch": 16.229035639413,
      "grad_norm": 0.14077673852443695,
      "learning_rate": 0.00010446768023855701,
      "loss": 0.396,
      "num_input_tokens_seen": 20196136,
      "step": 30965
    },
    {
      "epoch": 16.231656184486372,
      "grad_norm": 0.14699581265449524,
      "learning_rate": 0.00010432782712391559,
      "loss": 0.4545,
      "num_input_tokens_seen": 20199880,
      "step": 30970
    },
    {
      "epoch": 16.23427672955975,
      "grad_norm": 0.07833945006132126,
      "learning_rate": 0.00010418805677937871,
      "loss": 0.4351,
      "num_input_tokens_seen": 20203752,
      "step": 30975
    },
    {
      "epoch": 16.236897274633122,
      "grad_norm": 0.16479068994522095,
      "learning_rate": 0.00010404836923418465,
      "loss": 0.3442,
      "num_input_tokens_seen": 20206760,
      "step": 30980
    },
    {
      "epoch": 16.2395178197065,
      "grad_norm": 0.12340966612100601,
      "learning_rate": 0.00010390876451755477,
      "loss": 0.4399,
      "num_input_tokens_seen": 20209096,
      "step": 30985
    },
    {
      "epoch": 16.242138364779873,
      "grad_norm": 0.09470245242118835,
      "learning_rate": 0.0001037692426586922,
      "loss": 0.3111,
      "num_input_tokens_seen": 20212520,
      "step": 30990
    },
    {
      "epoch": 16.24475890985325,
      "grad_norm": 0.13103827834129333,
      "learning_rate": 0.0001036298036867837,
      "loss": 0.3587,
      "num_input_tokens_seen": 20215720,
      "step": 30995
    },
    {
      "epoch": 16.247379454926623,
      "grad_norm": 0.13331273198127747,
      "learning_rate": 0.00010349044763099819,
      "loss": 0.317,
      "num_input_tokens_seen": 20219048,
      "step": 31000
    },
    {
      "epoch": 16.25,
      "grad_norm": 0.1427554041147232,
      "learning_rate": 0.00010335117452048742,
      "loss": 0.5751,
      "num_input_tokens_seen": 20222440,
      "step": 31005
    },
    {
      "epoch": 16.252620545073377,
      "grad_norm": 0.13830667734146118,
      "learning_rate": 0.00010321198438438589,
      "loss": 0.4434,
      "num_input_tokens_seen": 20225224,
      "step": 31010
    },
    {
      "epoch": 16.25524109014675,
      "grad_norm": 0.19670940935611725,
      "learning_rate": 0.00010307287725181036,
      "loss": 0.4626,
      "num_input_tokens_seen": 20228680,
      "step": 31015
    },
    {
      "epoch": 16.257861635220127,
      "grad_norm": 0.263519287109375,
      "learning_rate": 0.00010293385315186049,
      "loss": 0.5654,
      "num_input_tokens_seen": 20231176,
      "step": 31020
    },
    {
      "epoch": 16.2604821802935,
      "grad_norm": 0.18690066039562225,
      "learning_rate": 0.00010279491211361853,
      "loss": 0.4239,
      "num_input_tokens_seen": 20233672,
      "step": 31025
    },
    {
      "epoch": 16.263102725366878,
      "grad_norm": 0.14893442392349243,
      "learning_rate": 0.00010265605416614938,
      "loss": 0.3132,
      "num_input_tokens_seen": 20236424,
      "step": 31030
    },
    {
      "epoch": 16.26572327044025,
      "grad_norm": 0.16952770948410034,
      "learning_rate": 0.00010251727933850069,
      "loss": 0.4503,
      "num_input_tokens_seen": 20239336,
      "step": 31035
    },
    {
      "epoch": 16.268343815513628,
      "grad_norm": 0.14199857413768768,
      "learning_rate": 0.00010237858765970231,
      "loss": 0.4309,
      "num_input_tokens_seen": 20242056,
      "step": 31040
    },
    {
      "epoch": 16.270964360587,
      "grad_norm": 0.09557023644447327,
      "learning_rate": 0.00010223997915876726,
      "loss": 0.3691,
      "num_input_tokens_seen": 20245384,
      "step": 31045
    },
    {
      "epoch": 16.27358490566038,
      "grad_norm": 0.1651749461889267,
      "learning_rate": 0.00010210145386469049,
      "loss": 0.4096,
      "num_input_tokens_seen": 20248936,
      "step": 31050
    },
    {
      "epoch": 16.27620545073375,
      "grad_norm": 0.12370415031909943,
      "learning_rate": 0.00010196301180645012,
      "loss": 0.3856,
      "num_input_tokens_seen": 20253416,
      "step": 31055
    },
    {
      "epoch": 16.27882599580713,
      "grad_norm": 0.12651558220386505,
      "learning_rate": 0.0001018246530130067,
      "loss": 0.4377,
      "num_input_tokens_seen": 20256968,
      "step": 31060
    },
    {
      "epoch": 16.281446540880502,
      "grad_norm": 0.2031250149011612,
      "learning_rate": 0.00010168637751330328,
      "loss": 0.5599,
      "num_input_tokens_seen": 20259752,
      "step": 31065
    },
    {
      "epoch": 16.28406708595388,
      "grad_norm": 0.16524510085582733,
      "learning_rate": 0.00010154818533626552,
      "loss": 0.4506,
      "num_input_tokens_seen": 20262600,
      "step": 31070
    },
    {
      "epoch": 16.286687631027252,
      "grad_norm": 0.10959690809249878,
      "learning_rate": 0.00010141007651080153,
      "loss": 0.3701,
      "num_input_tokens_seen": 20265896,
      "step": 31075
    },
    {
      "epoch": 16.28930817610063,
      "grad_norm": 0.11981530487537384,
      "learning_rate": 0.00010127205106580212,
      "loss": 0.4934,
      "num_input_tokens_seen": 20269224,
      "step": 31080
    },
    {
      "epoch": 16.291928721174003,
      "grad_norm": 0.09798921644687653,
      "learning_rate": 0.0001011341090301407,
      "loss": 0.3253,
      "num_input_tokens_seen": 20272072,
      "step": 31085
    },
    {
      "epoch": 16.29454926624738,
      "grad_norm": 0.2597713768482208,
      "learning_rate": 0.00010099625043267336,
      "loss": 0.4369,
      "num_input_tokens_seen": 20275400,
      "step": 31090
    },
    {
      "epoch": 16.297169811320753,
      "grad_norm": 0.15358762443065643,
      "learning_rate": 0.00010085847530223812,
      "loss": 0.4208,
      "num_input_tokens_seen": 20279560,
      "step": 31095
    },
    {
      "epoch": 16.29979035639413,
      "grad_norm": 0.1403001993894577,
      "learning_rate": 0.00010072078366765641,
      "loss": 0.4424,
      "num_input_tokens_seen": 20282728,
      "step": 31100
    },
    {
      "epoch": 16.302410901467507,
      "grad_norm": 0.15166859328746796,
      "learning_rate": 0.00010058317555773139,
      "loss": 0.5126,
      "num_input_tokens_seen": 20285224,
      "step": 31105
    },
    {
      "epoch": 16.30503144654088,
      "grad_norm": 0.13872021436691284,
      "learning_rate": 0.00010044565100124925,
      "loss": 0.3601,
      "num_input_tokens_seen": 20288552,
      "step": 31110
    },
    {
      "epoch": 16.307651991614257,
      "grad_norm": 0.15595826506614685,
      "learning_rate": 0.00010030821002697871,
      "loss": 0.4361,
      "num_input_tokens_seen": 20292776,
      "step": 31115
    },
    {
      "epoch": 16.31027253668763,
      "grad_norm": 0.1730518639087677,
      "learning_rate": 0.00010017085266367054,
      "loss": 0.5731,
      "num_input_tokens_seen": 20295208,
      "step": 31120
    },
    {
      "epoch": 16.312893081761008,
      "grad_norm": 0.1257336288690567,
      "learning_rate": 0.00010003357894005854,
      "loss": 0.4165,
      "num_input_tokens_seen": 20297960,
      "step": 31125
    },
    {
      "epoch": 16.31551362683438,
      "grad_norm": 0.14451588690280914,
      "learning_rate": 9.989638888485885e-05,
      "loss": 0.5239,
      "num_input_tokens_seen": 20300968,
      "step": 31130
    },
    {
      "epoch": 16.318134171907758,
      "grad_norm": 0.2240147888660431,
      "learning_rate": 9.975928252676991e-05,
      "loss": 0.4736,
      "num_input_tokens_seen": 20304648,
      "step": 31135
    },
    {
      "epoch": 16.32075471698113,
      "grad_norm": 0.20577311515808105,
      "learning_rate": 9.962225989447288e-05,
      "loss": 0.4894,
      "num_input_tokens_seen": 20308488,
      "step": 31140
    },
    {
      "epoch": 16.32337526205451,
      "grad_norm": 0.13381646573543549,
      "learning_rate": 9.948532101663144e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 20311560,
      "step": 31145
    },
    {
      "epoch": 16.32599580712788,
      "grad_norm": 0.10750774294137955,
      "learning_rate": 9.934846592189134e-05,
      "loss": 0.4023,
      "num_input_tokens_seen": 20315592,
      "step": 31150
    },
    {
      "epoch": 16.32861635220126,
      "grad_norm": 0.11819221079349518,
      "learning_rate": 9.921169463888153e-05,
      "loss": 0.4931,
      "num_input_tokens_seen": 20319304,
      "step": 31155
    },
    {
      "epoch": 16.331236897274632,
      "grad_norm": 0.14673495292663574,
      "learning_rate": 9.907500719621253e-05,
      "loss": 0.3734,
      "num_input_tokens_seen": 20324264,
      "step": 31160
    },
    {
      "epoch": 16.33385744234801,
      "grad_norm": 0.21808405220508575,
      "learning_rate": 9.893840362247809e-05,
      "loss": 0.3387,
      "num_input_tokens_seen": 20326696,
      "step": 31165
    },
    {
      "epoch": 16.336477987421382,
      "grad_norm": 0.12390564382076263,
      "learning_rate": 9.880188394625417e-05,
      "loss": 0.5011,
      "num_input_tokens_seen": 20330248,
      "step": 31170
    },
    {
      "epoch": 16.33909853249476,
      "grad_norm": 0.14061203598976135,
      "learning_rate": 9.86654481960989e-05,
      "loss": 0.3321,
      "num_input_tokens_seen": 20332552,
      "step": 31175
    },
    {
      "epoch": 16.341719077568133,
      "grad_norm": 0.13240456581115723,
      "learning_rate": 9.852909640055325e-05,
      "loss": 0.479,
      "num_input_tokens_seen": 20335304,
      "step": 31180
    },
    {
      "epoch": 16.34433962264151,
      "grad_norm": 0.2259160876274109,
      "learning_rate": 9.839282858814047e-05,
      "loss": 0.5273,
      "num_input_tokens_seen": 20337992,
      "step": 31185
    },
    {
      "epoch": 16.346960167714883,
      "grad_norm": 0.16872936487197876,
      "learning_rate": 9.825664478736607e-05,
      "loss": 0.4119,
      "num_input_tokens_seen": 20340680,
      "step": 31190
    },
    {
      "epoch": 16.34958071278826,
      "grad_norm": 0.19364559650421143,
      "learning_rate": 9.812054502671835e-05,
      "loss": 0.3976,
      "num_input_tokens_seen": 20343848,
      "step": 31195
    },
    {
      "epoch": 16.352201257861637,
      "grad_norm": 0.1241489127278328,
      "learning_rate": 9.798452933466761e-05,
      "loss": 0.3512,
      "num_input_tokens_seen": 20346792,
      "step": 31200
    },
    {
      "epoch": 16.35482180293501,
      "grad_norm": 0.17571651935577393,
      "learning_rate": 9.784859773966693e-05,
      "loss": 0.4072,
      "num_input_tokens_seen": 20349544,
      "step": 31205
    },
    {
      "epoch": 16.357442348008387,
      "grad_norm": 0.17155256867408752,
      "learning_rate": 9.771275027015159e-05,
      "loss": 0.4851,
      "num_input_tokens_seen": 20352904,
      "step": 31210
    },
    {
      "epoch": 16.36006289308176,
      "grad_norm": 0.1882844865322113,
      "learning_rate": 9.757698695453954e-05,
      "loss": 0.3348,
      "num_input_tokens_seen": 20356424,
      "step": 31215
    },
    {
      "epoch": 16.362683438155138,
      "grad_norm": 0.15023833513259888,
      "learning_rate": 9.74413078212305e-05,
      "loss": 0.4245,
      "num_input_tokens_seen": 20360264,
      "step": 31220
    },
    {
      "epoch": 16.36530398322851,
      "grad_norm": 0.14208075404167175,
      "learning_rate": 9.730571289860746e-05,
      "loss": 0.4175,
      "num_input_tokens_seen": 20363848,
      "step": 31225
    },
    {
      "epoch": 16.367924528301888,
      "grad_norm": 0.192727729678154,
      "learning_rate": 9.717020221503493e-05,
      "loss": 0.408,
      "num_input_tokens_seen": 20366280,
      "step": 31230
    },
    {
      "epoch": 16.37054507337526,
      "grad_norm": 0.10160087049007416,
      "learning_rate": 9.703477579886038e-05,
      "loss": 0.2796,
      "num_input_tokens_seen": 20370568,
      "step": 31235
    },
    {
      "epoch": 16.37316561844864,
      "grad_norm": 0.1625789999961853,
      "learning_rate": 9.689943367841347e-05,
      "loss": 0.4014,
      "num_input_tokens_seen": 20373928,
      "step": 31240
    },
    {
      "epoch": 16.37578616352201,
      "grad_norm": 0.2169084995985031,
      "learning_rate": 9.676417588200632e-05,
      "loss": 0.4361,
      "num_input_tokens_seen": 20376488,
      "step": 31245
    },
    {
      "epoch": 16.37840670859539,
      "grad_norm": 0.15398156642913818,
      "learning_rate": 9.662900243793321e-05,
      "loss": 0.3711,
      "num_input_tokens_seen": 20379176,
      "step": 31250
    },
    {
      "epoch": 16.381027253668762,
      "grad_norm": 0.24806365370750427,
      "learning_rate": 9.649391337447084e-05,
      "loss": 0.4894,
      "num_input_tokens_seen": 20381736,
      "step": 31255
    },
    {
      "epoch": 16.38364779874214,
      "grad_norm": 0.13321194052696228,
      "learning_rate": 9.635890871987829e-05,
      "loss": 0.3595,
      "num_input_tokens_seen": 20385064,
      "step": 31260
    },
    {
      "epoch": 16.386268343815512,
      "grad_norm": 0.09755979478359222,
      "learning_rate": 9.622398850239705e-05,
      "loss": 0.45,
      "num_input_tokens_seen": 20388072,
      "step": 31265
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 0.2885207533836365,
      "learning_rate": 9.608915275025104e-05,
      "loss": 0.5099,
      "num_input_tokens_seen": 20391080,
      "step": 31270
    },
    {
      "epoch": 16.391509433962263,
      "grad_norm": 0.2563965618610382,
      "learning_rate": 9.595440149164619e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 20394024,
      "step": 31275
    },
    {
      "epoch": 16.39412997903564,
      "grad_norm": 0.07187259197235107,
      "learning_rate": 9.581973475477085e-05,
      "loss": 0.3565,
      "num_input_tokens_seen": 20397672,
      "step": 31280
    },
    {
      "epoch": 16.396750524109013,
      "grad_norm": 0.07925773411989212,
      "learning_rate": 9.568515256779587e-05,
      "loss": 0.5242,
      "num_input_tokens_seen": 20401352,
      "step": 31285
    },
    {
      "epoch": 16.39937106918239,
      "grad_norm": 0.17077040672302246,
      "learning_rate": 9.555065495887433e-05,
      "loss": 0.4035,
      "num_input_tokens_seen": 20404008,
      "step": 31290
    },
    {
      "epoch": 16.401991614255767,
      "grad_norm": 0.12431541085243225,
      "learning_rate": 9.541624195614152e-05,
      "loss": 0.5858,
      "num_input_tokens_seen": 20407496,
      "step": 31295
    },
    {
      "epoch": 16.40461215932914,
      "grad_norm": 0.2089712917804718,
      "learning_rate": 9.528191358771532e-05,
      "loss": 0.4291,
      "num_input_tokens_seen": 20410696,
      "step": 31300
    },
    {
      "epoch": 16.407232704402517,
      "grad_norm": 0.23374852538108826,
      "learning_rate": 9.514766988169549e-05,
      "loss": 0.4721,
      "num_input_tokens_seen": 20413736,
      "step": 31305
    },
    {
      "epoch": 16.40985324947589,
      "grad_norm": 0.21561405062675476,
      "learning_rate": 9.501351086616422e-05,
      "loss": 0.3869,
      "num_input_tokens_seen": 20417160,
      "step": 31310
    },
    {
      "epoch": 16.412473794549268,
      "grad_norm": 0.08651239424943924,
      "learning_rate": 9.487943656918613e-05,
      "loss": 0.4378,
      "num_input_tokens_seen": 20420520,
      "step": 31315
    },
    {
      "epoch": 16.41509433962264,
      "grad_norm": 0.12694910168647766,
      "learning_rate": 9.474544701880805e-05,
      "loss": 0.4157,
      "num_input_tokens_seen": 20423272,
      "step": 31320
    },
    {
      "epoch": 16.417714884696018,
      "grad_norm": 0.22053155303001404,
      "learning_rate": 9.461154224305923e-05,
      "loss": 0.409,
      "num_input_tokens_seen": 20426216,
      "step": 31325
    },
    {
      "epoch": 16.42033542976939,
      "grad_norm": 0.17952744662761688,
      "learning_rate": 9.447772226995082e-05,
      "loss": 0.4478,
      "num_input_tokens_seen": 20429512,
      "step": 31330
    },
    {
      "epoch": 16.42295597484277,
      "grad_norm": 0.0840093195438385,
      "learning_rate": 9.434398712747639e-05,
      "loss": 0.4889,
      "num_input_tokens_seen": 20433064,
      "step": 31335
    },
    {
      "epoch": 16.42557651991614,
      "grad_norm": 0.3078885078430176,
      "learning_rate": 9.421033684361185e-05,
      "loss": 0.466,
      "num_input_tokens_seen": 20436232,
      "step": 31340
    },
    {
      "epoch": 16.42819706498952,
      "grad_norm": 0.16232164204120636,
      "learning_rate": 9.407677144631533e-05,
      "loss": 0.4761,
      "num_input_tokens_seen": 20439688,
      "step": 31345
    },
    {
      "epoch": 16.430817610062892,
      "grad_norm": 0.1642945110797882,
      "learning_rate": 9.394329096352732e-05,
      "loss": 0.6238,
      "num_input_tokens_seen": 20442888,
      "step": 31350
    },
    {
      "epoch": 16.43343815513627,
      "grad_norm": 0.1825721710920334,
      "learning_rate": 9.380989542317037e-05,
      "loss": 0.4736,
      "num_input_tokens_seen": 20445928,
      "step": 31355
    },
    {
      "epoch": 16.436058700209642,
      "grad_norm": 0.14947426319122314,
      "learning_rate": 9.367658485314907e-05,
      "loss": 0.352,
      "num_input_tokens_seen": 20448552,
      "step": 31360
    },
    {
      "epoch": 16.43867924528302,
      "grad_norm": 0.13125385344028473,
      "learning_rate": 9.354335928135066e-05,
      "loss": 0.4178,
      "num_input_tokens_seen": 20451304,
      "step": 31365
    },
    {
      "epoch": 16.441299790356393,
      "grad_norm": 0.15528149902820587,
      "learning_rate": 9.341021873564432e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 20454408,
      "step": 31370
    },
    {
      "epoch": 16.44392033542977,
      "grad_norm": 0.12241031229496002,
      "learning_rate": 9.327716324388164e-05,
      "loss": 0.381,
      "num_input_tokens_seen": 20457384,
      "step": 31375
    },
    {
      "epoch": 16.446540880503143,
      "grad_norm": 0.1317904144525528,
      "learning_rate": 9.314419283389641e-05,
      "loss": 0.3917,
      "num_input_tokens_seen": 20460456,
      "step": 31380
    },
    {
      "epoch": 16.44916142557652,
      "grad_norm": 0.14658832550048828,
      "learning_rate": 9.30113075335044e-05,
      "loss": 0.3997,
      "num_input_tokens_seen": 20463336,
      "step": 31385
    },
    {
      "epoch": 16.451781970649897,
      "grad_norm": 0.1828823685646057,
      "learning_rate": 9.287850737050352e-05,
      "loss": 0.4723,
      "num_input_tokens_seen": 20465928,
      "step": 31390
    },
    {
      "epoch": 16.45440251572327,
      "grad_norm": 0.21090538799762726,
      "learning_rate": 9.274579237267422e-05,
      "loss": 0.4146,
      "num_input_tokens_seen": 20468584,
      "step": 31395
    },
    {
      "epoch": 16.457023060796647,
      "grad_norm": 0.1646708846092224,
      "learning_rate": 9.261316256777897e-05,
      "loss": 0.4563,
      "num_input_tokens_seen": 20471336,
      "step": 31400
    },
    {
      "epoch": 16.45964360587002,
      "grad_norm": 0.22487783432006836,
      "learning_rate": 9.24806179835625e-05,
      "loss": 0.3636,
      "num_input_tokens_seen": 20474696,
      "step": 31405
    },
    {
      "epoch": 16.462264150943398,
      "grad_norm": 0.15487627685070038,
      "learning_rate": 9.234815864775137e-05,
      "loss": 0.4297,
      "num_input_tokens_seen": 20477832,
      "step": 31410
    },
    {
      "epoch": 16.46488469601677,
      "grad_norm": 0.1606740802526474,
      "learning_rate": 9.221578458805485e-05,
      "loss": 0.5195,
      "num_input_tokens_seen": 20481160,
      "step": 31415
    },
    {
      "epoch": 16.467505241090148,
      "grad_norm": 0.15179754793643951,
      "learning_rate": 9.208349583216385e-05,
      "loss": 0.4483,
      "num_input_tokens_seen": 20484648,
      "step": 31420
    },
    {
      "epoch": 16.47012578616352,
      "grad_norm": 0.3021206855773926,
      "learning_rate": 9.195129240775174e-05,
      "loss": 0.2617,
      "num_input_tokens_seen": 20486696,
      "step": 31425
    },
    {
      "epoch": 16.4727463312369,
      "grad_norm": 0.17072169482707977,
      "learning_rate": 9.181917434247417e-05,
      "loss": 0.4182,
      "num_input_tokens_seen": 20489640,
      "step": 31430
    },
    {
      "epoch": 16.47536687631027,
      "grad_norm": 0.1186550110578537,
      "learning_rate": 9.168714166396835e-05,
      "loss": 0.5267,
      "num_input_tokens_seen": 20492392,
      "step": 31435
    },
    {
      "epoch": 16.47798742138365,
      "grad_norm": 0.15627023577690125,
      "learning_rate": 9.155519439985438e-05,
      "loss": 0.3923,
      "num_input_tokens_seen": 20494984,
      "step": 31440
    },
    {
      "epoch": 16.480607966457022,
      "grad_norm": 0.1705167442560196,
      "learning_rate": 9.142333257773383e-05,
      "loss": 0.409,
      "num_input_tokens_seen": 20497832,
      "step": 31445
    },
    {
      "epoch": 16.4832285115304,
      "grad_norm": 0.09059387445449829,
      "learning_rate": 9.12915562251908e-05,
      "loss": 0.3543,
      "num_input_tokens_seen": 20500488,
      "step": 31450
    },
    {
      "epoch": 16.485849056603772,
      "grad_norm": 0.15917344391345978,
      "learning_rate": 9.115986536979149e-05,
      "loss": 0.4616,
      "num_input_tokens_seen": 20503688,
      "step": 31455
    },
    {
      "epoch": 16.48846960167715,
      "grad_norm": 0.1050930917263031,
      "learning_rate": 9.10282600390841e-05,
      "loss": 0.3292,
      "num_input_tokens_seen": 20507816,
      "step": 31460
    },
    {
      "epoch": 16.491090146750523,
      "grad_norm": 0.15112867951393127,
      "learning_rate": 9.08967402605988e-05,
      "loss": 0.5437,
      "num_input_tokens_seen": 20510920,
      "step": 31465
    },
    {
      "epoch": 16.4937106918239,
      "grad_norm": 0.13697397708892822,
      "learning_rate": 9.07653060618483e-05,
      "loss": 0.4037,
      "num_input_tokens_seen": 20514408,
      "step": 31470
    },
    {
      "epoch": 16.496331236897273,
      "grad_norm": 0.16511420905590057,
      "learning_rate": 9.063395747032676e-05,
      "loss": 0.3909,
      "num_input_tokens_seen": 20518472,
      "step": 31475
    },
    {
      "epoch": 16.49895178197065,
      "grad_norm": 0.17187589406967163,
      "learning_rate": 9.050269451351112e-05,
      "loss": 0.5352,
      "num_input_tokens_seen": 20521064,
      "step": 31480
    },
    {
      "epoch": 16.501572327044027,
      "grad_norm": 0.2495919018983841,
      "learning_rate": 9.037151721886006e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 20523784,
      "step": 31485
    },
    {
      "epoch": 16.5041928721174,
      "grad_norm": 0.08941759169101715,
      "learning_rate": 9.024042561381424e-05,
      "loss": 0.4598,
      "num_input_tokens_seen": 20527976,
      "step": 31490
    },
    {
      "epoch": 16.506813417190777,
      "grad_norm": 0.12697528302669525,
      "learning_rate": 9.010941972579656e-05,
      "loss": 0.5437,
      "num_input_tokens_seen": 20531784,
      "step": 31495
    },
    {
      "epoch": 16.50943396226415,
      "grad_norm": 0.2074061781167984,
      "learning_rate": 8.99784995822121e-05,
      "loss": 0.4616,
      "num_input_tokens_seen": 20535048,
      "step": 31500
    },
    {
      "epoch": 16.512054507337528,
      "grad_norm": 0.1385708898305893,
      "learning_rate": 8.984766521044769e-05,
      "loss": 0.5221,
      "num_input_tokens_seen": 20537672,
      "step": 31505
    },
    {
      "epoch": 16.5146750524109,
      "grad_norm": 0.15980376303195953,
      "learning_rate": 8.971691663787252e-05,
      "loss": 0.3371,
      "num_input_tokens_seen": 20540264,
      "step": 31510
    },
    {
      "epoch": 16.517295597484278,
      "grad_norm": 0.11953779309988022,
      "learning_rate": 8.958625389183756e-05,
      "loss": 0.3705,
      "num_input_tokens_seen": 20543208,
      "step": 31515
    },
    {
      "epoch": 16.51991614255765,
      "grad_norm": 0.1999165117740631,
      "learning_rate": 8.9455676999676e-05,
      "loss": 0.401,
      "num_input_tokens_seen": 20546632,
      "step": 31520
    },
    {
      "epoch": 16.52253668763103,
      "grad_norm": 0.1478559821844101,
      "learning_rate": 8.932518598870309e-05,
      "loss": 0.5345,
      "num_input_tokens_seen": 20549256,
      "step": 31525
    },
    {
      "epoch": 16.5251572327044,
      "grad_norm": 0.11767039448022842,
      "learning_rate": 8.919478088621614e-05,
      "loss": 0.5228,
      "num_input_tokens_seen": 20551880,
      "step": 31530
    },
    {
      "epoch": 16.52777777777778,
      "grad_norm": 0.16233330965042114,
      "learning_rate": 8.906446171949422e-05,
      "loss": 0.4004,
      "num_input_tokens_seen": 20555368,
      "step": 31535
    },
    {
      "epoch": 16.530398322851152,
      "grad_norm": 0.107028067111969,
      "learning_rate": 8.893422851579885e-05,
      "loss": 0.3146,
      "num_input_tokens_seen": 20557864,
      "step": 31540
    },
    {
      "epoch": 16.53301886792453,
      "grad_norm": 0.12255271524190903,
      "learning_rate": 8.8804081302373e-05,
      "loss": 0.3762,
      "num_input_tokens_seen": 20561800,
      "step": 31545
    },
    {
      "epoch": 16.535639412997902,
      "grad_norm": 0.1091788038611412,
      "learning_rate": 8.867402010644221e-05,
      "loss": 0.3875,
      "num_input_tokens_seen": 20565128,
      "step": 31550
    },
    {
      "epoch": 16.53825995807128,
      "grad_norm": 0.18333861231803894,
      "learning_rate": 8.854404495521389e-05,
      "loss": 0.3315,
      "num_input_tokens_seen": 20568648,
      "step": 31555
    },
    {
      "epoch": 16.540880503144653,
      "grad_norm": 0.11236733198165894,
      "learning_rate": 8.841415587587709e-05,
      "loss": 0.3787,
      "num_input_tokens_seen": 20571816,
      "step": 31560
    },
    {
      "epoch": 16.54350104821803,
      "grad_norm": 0.13145269453525543,
      "learning_rate": 8.828435289560344e-05,
      "loss": 0.5082,
      "num_input_tokens_seen": 20576136,
      "step": 31565
    },
    {
      "epoch": 16.546121593291403,
      "grad_norm": 0.18999364972114563,
      "learning_rate": 8.815463604154588e-05,
      "loss": 0.4798,
      "num_input_tokens_seen": 20578472,
      "step": 31570
    },
    {
      "epoch": 16.54874213836478,
      "grad_norm": 0.12257598340511322,
      "learning_rate": 8.80250053408399e-05,
      "loss": 0.5097,
      "num_input_tokens_seen": 20581800,
      "step": 31575
    },
    {
      "epoch": 16.551362683438157,
      "grad_norm": 0.11426733434200287,
      "learning_rate": 8.789546082060273e-05,
      "loss": 0.6167,
      "num_input_tokens_seen": 20584296,
      "step": 31580
    },
    {
      "epoch": 16.55398322851153,
      "grad_norm": 0.23077820241451263,
      "learning_rate": 8.776600250793371e-05,
      "loss": 0.4248,
      "num_input_tokens_seen": 20587368,
      "step": 31585
    },
    {
      "epoch": 16.556603773584907,
      "grad_norm": 0.1898655891418457,
      "learning_rate": 8.763663042991399e-05,
      "loss": 0.4979,
      "num_input_tokens_seen": 20590504,
      "step": 31590
    },
    {
      "epoch": 16.55922431865828,
      "grad_norm": 0.12208834290504456,
      "learning_rate": 8.75073446136066e-05,
      "loss": 0.45,
      "num_input_tokens_seen": 20593448,
      "step": 31595
    },
    {
      "epoch": 16.561844863731658,
      "grad_norm": 0.4195410907268524,
      "learning_rate": 8.737814508605674e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 20596904,
      "step": 31600
    },
    {
      "epoch": 16.56446540880503,
      "grad_norm": 0.2016170173883438,
      "learning_rate": 8.724903187429145e-05,
      "loss": 0.421,
      "num_input_tokens_seen": 20600040,
      "step": 31605
    },
    {
      "epoch": 16.567085953878408,
      "grad_norm": 0.24632839858531952,
      "learning_rate": 8.71200050053198e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 20603112,
      "step": 31610
    },
    {
      "epoch": 16.56970649895178,
      "grad_norm": 0.18140849471092224,
      "learning_rate": 8.699106450613287e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 20606760,
      "step": 31615
    },
    {
      "epoch": 16.572327044025158,
      "grad_norm": 0.21955369412899017,
      "learning_rate": 8.686221040370334e-05,
      "loss": 0.4056,
      "num_input_tokens_seen": 20609896,
      "step": 31620
    },
    {
      "epoch": 16.57494758909853,
      "grad_norm": 0.6087865829467773,
      "learning_rate": 8.673344272498596e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 20612712,
      "step": 31625
    },
    {
      "epoch": 16.57756813417191,
      "grad_norm": 0.156467005610466,
      "learning_rate": 8.660476149691759e-05,
      "loss": 0.6805,
      "num_input_tokens_seen": 20616488,
      "step": 31630
    },
    {
      "epoch": 16.580188679245282,
      "grad_norm": 0.14153523743152618,
      "learning_rate": 8.647616674641684e-05,
      "loss": 0.5352,
      "num_input_tokens_seen": 20619944,
      "step": 31635
    },
    {
      "epoch": 16.58280922431866,
      "grad_norm": 0.12363384664058685,
      "learning_rate": 8.63476585003844e-05,
      "loss": 0.4306,
      "num_input_tokens_seen": 20622952,
      "step": 31640
    },
    {
      "epoch": 16.585429769392032,
      "grad_norm": 0.13894511759281158,
      "learning_rate": 8.621923678570259e-05,
      "loss": 0.335,
      "num_input_tokens_seen": 20626248,
      "step": 31645
    },
    {
      "epoch": 16.58805031446541,
      "grad_norm": 0.10633653402328491,
      "learning_rate": 8.609090162923567e-05,
      "loss": 0.3216,
      "num_input_tokens_seen": 20630088,
      "step": 31650
    },
    {
      "epoch": 16.590670859538783,
      "grad_norm": 0.23089277744293213,
      "learning_rate": 8.596265305783002e-05,
      "loss": 0.4351,
      "num_input_tokens_seen": 20632840,
      "step": 31655
    },
    {
      "epoch": 16.59329140461216,
      "grad_norm": 0.1477123498916626,
      "learning_rate": 8.583449109831375e-05,
      "loss": 0.4244,
      "num_input_tokens_seen": 20635720,
      "step": 31660
    },
    {
      "epoch": 16.595911949685533,
      "grad_norm": 0.25109508633613586,
      "learning_rate": 8.570641577749705e-05,
      "loss": 0.3579,
      "num_input_tokens_seen": 20639400,
      "step": 31665
    },
    {
      "epoch": 16.59853249475891,
      "grad_norm": 0.10376718640327454,
      "learning_rate": 8.557842712217162e-05,
      "loss": 0.4073,
      "num_input_tokens_seen": 20643656,
      "step": 31670
    },
    {
      "epoch": 16.601153039832283,
      "grad_norm": 0.1384209245443344,
      "learning_rate": 8.545052515911112e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 20646504,
      "step": 31675
    },
    {
      "epoch": 16.60377358490566,
      "grad_norm": 0.18376553058624268,
      "learning_rate": 8.532270991507136e-05,
      "loss": 0.5047,
      "num_input_tokens_seen": 20650120,
      "step": 31680
    },
    {
      "epoch": 16.606394129979037,
      "grad_norm": 0.22181963920593262,
      "learning_rate": 8.519498141678983e-05,
      "loss": 0.3785,
      "num_input_tokens_seen": 20653064,
      "step": 31685
    },
    {
      "epoch": 16.60901467505241,
      "grad_norm": 0.08537405729293823,
      "learning_rate": 8.506733969098579e-05,
      "loss": 0.4029,
      "num_input_tokens_seen": 20657160,
      "step": 31690
    },
    {
      "epoch": 16.611635220125788,
      "grad_norm": 0.13043177127838135,
      "learning_rate": 8.49397847643606e-05,
      "loss": 0.3342,
      "num_input_tokens_seen": 20659912,
      "step": 31695
    },
    {
      "epoch": 16.61425576519916,
      "grad_norm": 0.13391980528831482,
      "learning_rate": 8.481231666359723e-05,
      "loss": 0.3204,
      "num_input_tokens_seen": 20662312,
      "step": 31700
    },
    {
      "epoch": 16.616876310272538,
      "grad_norm": 0.20726709067821503,
      "learning_rate": 8.468493541536031e-05,
      "loss": 0.4215,
      "num_input_tokens_seen": 20666248,
      "step": 31705
    },
    {
      "epoch": 16.61949685534591,
      "grad_norm": 0.19303934276103973,
      "learning_rate": 8.455764104629681e-05,
      "loss": 0.4125,
      "num_input_tokens_seen": 20670888,
      "step": 31710
    },
    {
      "epoch": 16.622117400419288,
      "grad_norm": 0.1530061960220337,
      "learning_rate": 8.443043358303515e-05,
      "loss": 0.4029,
      "num_input_tokens_seen": 20673320,
      "step": 31715
    },
    {
      "epoch": 16.62473794549266,
      "grad_norm": 0.13032032549381256,
      "learning_rate": 8.430331305218585e-05,
      "loss": 0.3218,
      "num_input_tokens_seen": 20676552,
      "step": 31720
    },
    {
      "epoch": 16.62735849056604,
      "grad_norm": 0.11581093817949295,
      "learning_rate": 8.417627948034096e-05,
      "loss": 0.4071,
      "num_input_tokens_seen": 20678792,
      "step": 31725
    },
    {
      "epoch": 16.629979035639412,
      "grad_norm": 0.14619946479797363,
      "learning_rate": 8.404933289407424e-05,
      "loss": 0.3573,
      "num_input_tokens_seen": 20683240,
      "step": 31730
    },
    {
      "epoch": 16.63259958071279,
      "grad_norm": 0.15476295351982117,
      "learning_rate": 8.392247331994174e-05,
      "loss": 0.514,
      "num_input_tokens_seen": 20686536,
      "step": 31735
    },
    {
      "epoch": 16.635220125786162,
      "grad_norm": 0.14878585934638977,
      "learning_rate": 8.37957007844809e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 20689832,
      "step": 31740
    },
    {
      "epoch": 16.63784067085954,
      "grad_norm": 0.1719713807106018,
      "learning_rate": 8.366901531421134e-05,
      "loss": 0.492,
      "num_input_tokens_seen": 20692648,
      "step": 31745
    },
    {
      "epoch": 16.640461215932913,
      "grad_norm": 0.24345487356185913,
      "learning_rate": 8.354241693563385e-05,
      "loss": 0.4656,
      "num_input_tokens_seen": 20695048,
      "step": 31750
    },
    {
      "epoch": 16.64308176100629,
      "grad_norm": 0.14065800607204437,
      "learning_rate": 8.341590567523166e-05,
      "loss": 0.3932,
      "num_input_tokens_seen": 20698216,
      "step": 31755
    },
    {
      "epoch": 16.645702306079663,
      "grad_norm": 0.07180841267108917,
      "learning_rate": 8.328948155946924e-05,
      "loss": 0.4787,
      "num_input_tokens_seen": 20702024,
      "step": 31760
    },
    {
      "epoch": 16.64832285115304,
      "grad_norm": 0.15682175755500793,
      "learning_rate": 8.316314461479318e-05,
      "loss": 0.3199,
      "num_input_tokens_seen": 20705032,
      "step": 31765
    },
    {
      "epoch": 16.650943396226417,
      "grad_norm": 0.08493208885192871,
      "learning_rate": 8.303689486763177e-05,
      "loss": 0.3523,
      "num_input_tokens_seen": 20708072,
      "step": 31770
    },
    {
      "epoch": 16.65356394129979,
      "grad_norm": 0.14770253002643585,
      "learning_rate": 8.291073234439512e-05,
      "loss": 0.4122,
      "num_input_tokens_seen": 20711432,
      "step": 31775
    },
    {
      "epoch": 16.656184486373167,
      "grad_norm": 0.28713107109069824,
      "learning_rate": 8.27846570714747e-05,
      "loss": 0.5054,
      "num_input_tokens_seen": 20715112,
      "step": 31780
    },
    {
      "epoch": 16.65880503144654,
      "grad_norm": 0.32303494215011597,
      "learning_rate": 8.265866907524427e-05,
      "loss": 0.4,
      "num_input_tokens_seen": 20718024,
      "step": 31785
    },
    {
      "epoch": 16.661425576519918,
      "grad_norm": 0.17594635486602783,
      "learning_rate": 8.253276838205892e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 20721192,
      "step": 31790
    },
    {
      "epoch": 16.66404612159329,
      "grad_norm": 0.19791758060455322,
      "learning_rate": 8.240695501825568e-05,
      "loss": 0.3292,
      "num_input_tokens_seen": 20723656,
      "step": 31795
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 0.1321822851896286,
      "learning_rate": 8.228122901015345e-05,
      "loss": 0.3287,
      "num_input_tokens_seen": 20726088,
      "step": 31800
    },
    {
      "epoch": 16.66928721174004,
      "grad_norm": 0.1311774104833603,
      "learning_rate": 8.21555903840524e-05,
      "loss": 0.3251,
      "num_input_tokens_seen": 20728840,
      "step": 31805
    },
    {
      "epoch": 16.671907756813418,
      "grad_norm": 0.11701270937919617,
      "learning_rate": 8.203003916623491e-05,
      "loss": 0.2877,
      "num_input_tokens_seen": 20732424,
      "step": 31810
    },
    {
      "epoch": 16.67452830188679,
      "grad_norm": 0.09684299677610397,
      "learning_rate": 8.190457538296464e-05,
      "loss": 0.32,
      "num_input_tokens_seen": 20738184,
      "step": 31815
    },
    {
      "epoch": 16.67714884696017,
      "grad_norm": 0.07071655243635178,
      "learning_rate": 8.177919906048736e-05,
      "loss": 0.3413,
      "num_input_tokens_seen": 20743464,
      "step": 31820
    },
    {
      "epoch": 16.679769392033542,
      "grad_norm": 0.09851280599832535,
      "learning_rate": 8.165391022503044e-05,
      "loss": 0.3967,
      "num_input_tokens_seen": 20746536,
      "step": 31825
    },
    {
      "epoch": 16.68238993710692,
      "grad_norm": 0.2024935781955719,
      "learning_rate": 8.152870890280261e-05,
      "loss": 0.4931,
      "num_input_tokens_seen": 20749896,
      "step": 31830
    },
    {
      "epoch": 16.685010482180292,
      "grad_norm": 0.16077940165996552,
      "learning_rate": 8.140359511999473e-05,
      "loss": 0.3837,
      "num_input_tokens_seen": 20753448,
      "step": 31835
    },
    {
      "epoch": 16.68763102725367,
      "grad_norm": 0.16186809539794922,
      "learning_rate": 8.127856890277923e-05,
      "loss": 0.4779,
      "num_input_tokens_seen": 20757032,
      "step": 31840
    },
    {
      "epoch": 16.690251572327043,
      "grad_norm": 0.10579606890678406,
      "learning_rate": 8.115363027730998e-05,
      "loss": 0.3452,
      "num_input_tokens_seen": 20760136,
      "step": 31845
    },
    {
      "epoch": 16.69287211740042,
      "grad_norm": 0.16167518496513367,
      "learning_rate": 8.102877926972286e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 20763144,
      "step": 31850
    },
    {
      "epoch": 16.695492662473793,
      "grad_norm": 0.11918742954730988,
      "learning_rate": 8.090401590613533e-05,
      "loss": 0.4614,
      "num_input_tokens_seen": 20766248,
      "step": 31855
    },
    {
      "epoch": 16.69811320754717,
      "grad_norm": 0.23100773990154266,
      "learning_rate": 8.077934021264627e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 20769160,
      "step": 31860
    },
    {
      "epoch": 16.700733752620543,
      "grad_norm": 0.1383388638496399,
      "learning_rate": 8.065475221533652e-05,
      "loss": 0.3697,
      "num_input_tokens_seen": 20773384,
      "step": 31865
    },
    {
      "epoch": 16.70335429769392,
      "grad_norm": 0.13074511289596558,
      "learning_rate": 8.053025194026858e-05,
      "loss": 0.5252,
      "num_input_tokens_seen": 20777416,
      "step": 31870
    },
    {
      "epoch": 16.705974842767297,
      "grad_norm": 0.15550073981285095,
      "learning_rate": 8.040583941348623e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 20780808,
      "step": 31875
    },
    {
      "epoch": 16.70859538784067,
      "grad_norm": 0.11698853224515915,
      "learning_rate": 8.028151466101541e-05,
      "loss": 0.2852,
      "num_input_tokens_seen": 20784040,
      "step": 31880
    },
    {
      "epoch": 16.711215932914047,
      "grad_norm": 0.09518007934093475,
      "learning_rate": 8.015727770886321e-05,
      "loss": 0.4147,
      "num_input_tokens_seen": 20787144,
      "step": 31885
    },
    {
      "epoch": 16.71383647798742,
      "grad_norm": 0.11471238732337952,
      "learning_rate": 8.00331285830187e-05,
      "loss": 0.4174,
      "num_input_tokens_seen": 20790408,
      "step": 31890
    },
    {
      "epoch": 16.716457023060798,
      "grad_norm": 0.1362512707710266,
      "learning_rate": 7.990906730945247e-05,
      "loss": 0.4499,
      "num_input_tokens_seen": 20793768,
      "step": 31895
    },
    {
      "epoch": 16.71907756813417,
      "grad_norm": 0.05763031914830208,
      "learning_rate": 7.978509391411681e-05,
      "loss": 0.3787,
      "num_input_tokens_seen": 20799624,
      "step": 31900
    },
    {
      "epoch": 16.721698113207548,
      "grad_norm": 0.14889217913150787,
      "learning_rate": 7.966120842294544e-05,
      "loss": 0.4669,
      "num_input_tokens_seen": 20802152,
      "step": 31905
    },
    {
      "epoch": 16.72431865828092,
      "grad_norm": 0.1491522192955017,
      "learning_rate": 7.953741086185368e-05,
      "loss": 0.4239,
      "num_input_tokens_seen": 20804776,
      "step": 31910
    },
    {
      "epoch": 16.7269392033543,
      "grad_norm": 0.15321208536624908,
      "learning_rate": 7.941370125673864e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 20807880,
      "step": 31915
    },
    {
      "epoch": 16.729559748427672,
      "grad_norm": 0.16052579879760742,
      "learning_rate": 7.9290079633479e-05,
      "loss": 0.4212,
      "num_input_tokens_seen": 20810568,
      "step": 31920
    },
    {
      "epoch": 16.73218029350105,
      "grad_norm": 0.07622429728507996,
      "learning_rate": 7.916654601793516e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 20815368,
      "step": 31925
    },
    {
      "epoch": 16.734800838574422,
      "grad_norm": 0.156732976436615,
      "learning_rate": 7.904310043594859e-05,
      "loss": 0.5207,
      "num_input_tokens_seen": 20818472,
      "step": 31930
    },
    {
      "epoch": 16.7374213836478,
      "grad_norm": 0.13259819149971008,
      "learning_rate": 7.891974291334303e-05,
      "loss": 0.4371,
      "num_input_tokens_seen": 20822120,
      "step": 31935
    },
    {
      "epoch": 16.740041928721173,
      "grad_norm": 0.15338370203971863,
      "learning_rate": 7.879647347592322e-05,
      "loss": 0.4202,
      "num_input_tokens_seen": 20825160,
      "step": 31940
    },
    {
      "epoch": 16.74266247379455,
      "grad_norm": 0.10380114614963531,
      "learning_rate": 7.867329214947578e-05,
      "loss": 0.6133,
      "num_input_tokens_seen": 20830056,
      "step": 31945
    },
    {
      "epoch": 16.745283018867923,
      "grad_norm": 0.138589546084404,
      "learning_rate": 7.855019895976889e-05,
      "loss": 0.4937,
      "num_input_tokens_seen": 20833032,
      "step": 31950
    },
    {
      "epoch": 16.7479035639413,
      "grad_norm": 0.12026527523994446,
      "learning_rate": 7.842719393255232e-05,
      "loss": 0.4166,
      "num_input_tokens_seen": 20836200,
      "step": 31955
    },
    {
      "epoch": 16.750524109014677,
      "grad_norm": 0.13840840756893158,
      "learning_rate": 7.830427709355725e-05,
      "loss": 0.4358,
      "num_input_tokens_seen": 20839784,
      "step": 31960
    },
    {
      "epoch": 16.75314465408805,
      "grad_norm": 0.1002127006649971,
      "learning_rate": 7.818144846849634e-05,
      "loss": 0.3612,
      "num_input_tokens_seen": 20842952,
      "step": 31965
    },
    {
      "epoch": 16.755765199161427,
      "grad_norm": 0.16148626804351807,
      "learning_rate": 7.805870808306403e-05,
      "loss": 0.4599,
      "num_input_tokens_seen": 20846344,
      "step": 31970
    },
    {
      "epoch": 16.7583857442348,
      "grad_norm": 0.23543429374694824,
      "learning_rate": 7.793605596293618e-05,
      "loss": 0.3897,
      "num_input_tokens_seen": 20849512,
      "step": 31975
    },
    {
      "epoch": 16.761006289308177,
      "grad_norm": 0.2012590616941452,
      "learning_rate": 7.781349213377048e-05,
      "loss": 0.5701,
      "num_input_tokens_seen": 20852744,
      "step": 31980
    },
    {
      "epoch": 16.76362683438155,
      "grad_norm": 0.1493479609489441,
      "learning_rate": 7.769101662120559e-05,
      "loss": 0.4866,
      "num_input_tokens_seen": 20856392,
      "step": 31985
    },
    {
      "epoch": 16.766247379454928,
      "grad_norm": 0.18195569515228271,
      "learning_rate": 7.756862945086196e-05,
      "loss": 0.4396,
      "num_input_tokens_seen": 20859048,
      "step": 31990
    },
    {
      "epoch": 16.7688679245283,
      "grad_norm": 0.14191576838493347,
      "learning_rate": 7.744633064834172e-05,
      "loss": 0.4158,
      "num_input_tokens_seen": 20861992,
      "step": 31995
    },
    {
      "epoch": 16.771488469601678,
      "grad_norm": 0.18159237504005432,
      "learning_rate": 7.732412023922836e-05,
      "loss": 0.4378,
      "num_input_tokens_seen": 20864904,
      "step": 32000
    },
    {
      "epoch": 16.77410901467505,
      "grad_norm": 0.17937050759792328,
      "learning_rate": 7.720199824908692e-05,
      "loss": 0.4499,
      "num_input_tokens_seen": 20867976,
      "step": 32005
    },
    {
      "epoch": 16.77672955974843,
      "grad_norm": 0.14386042952537537,
      "learning_rate": 7.707996470346402e-05,
      "loss": 0.4257,
      "num_input_tokens_seen": 20871464,
      "step": 32010
    },
    {
      "epoch": 16.779350104821802,
      "grad_norm": 0.11963314563035965,
      "learning_rate": 7.695801962788756e-05,
      "loss": 0.4788,
      "num_input_tokens_seen": 20874984,
      "step": 32015
    },
    {
      "epoch": 16.78197064989518,
      "grad_norm": 0.13527972996234894,
      "learning_rate": 7.683616304786695e-05,
      "loss": 0.3319,
      "num_input_tokens_seen": 20877832,
      "step": 32020
    },
    {
      "epoch": 16.784591194968552,
      "grad_norm": 0.18975062668323517,
      "learning_rate": 7.671439498889332e-05,
      "loss": 0.5321,
      "num_input_tokens_seen": 20881096,
      "step": 32025
    },
    {
      "epoch": 16.78721174004193,
      "grad_norm": 0.13499616086483002,
      "learning_rate": 7.65927154764392e-05,
      "loss": 0.4571,
      "num_input_tokens_seen": 20884680,
      "step": 32030
    },
    {
      "epoch": 16.789832285115303,
      "grad_norm": 0.1774340718984604,
      "learning_rate": 7.647112453595862e-05,
      "loss": 0.5452,
      "num_input_tokens_seen": 20888008,
      "step": 32035
    },
    {
      "epoch": 16.79245283018868,
      "grad_norm": 0.10875166207551956,
      "learning_rate": 7.634962219288688e-05,
      "loss": 0.3803,
      "num_input_tokens_seen": 20891240,
      "step": 32040
    },
    {
      "epoch": 16.795073375262053,
      "grad_norm": 0.14611126482486725,
      "learning_rate": 7.622820847264083e-05,
      "loss": 0.333,
      "num_input_tokens_seen": 20894664,
      "step": 32045
    },
    {
      "epoch": 16.79769392033543,
      "grad_norm": 0.13660907745361328,
      "learning_rate": 7.610688340061894e-05,
      "loss": 0.4919,
      "num_input_tokens_seen": 20897608,
      "step": 32050
    },
    {
      "epoch": 16.800314465408803,
      "grad_norm": 0.16968007385730743,
      "learning_rate": 7.598564700220101e-05,
      "loss": 0.4242,
      "num_input_tokens_seen": 20901096,
      "step": 32055
    },
    {
      "epoch": 16.80293501048218,
      "grad_norm": 0.21500101685523987,
      "learning_rate": 7.586449930274842e-05,
      "loss": 0.3958,
      "num_input_tokens_seen": 20904520,
      "step": 32060
    },
    {
      "epoch": 16.805555555555557,
      "grad_norm": 0.1455705612897873,
      "learning_rate": 7.574344032760367e-05,
      "loss": 0.3797,
      "num_input_tokens_seen": 20907592,
      "step": 32065
    },
    {
      "epoch": 16.80817610062893,
      "grad_norm": 0.10741720348596573,
      "learning_rate": 7.562247010209111e-05,
      "loss": 0.3992,
      "num_input_tokens_seen": 20911752,
      "step": 32070
    },
    {
      "epoch": 16.810796645702307,
      "grad_norm": 0.1199188381433487,
      "learning_rate": 7.550158865151618e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 20914728,
      "step": 32075
    },
    {
      "epoch": 16.81341719077568,
      "grad_norm": 0.17234420776367188,
      "learning_rate": 7.538079600116593e-05,
      "loss": 0.4088,
      "num_input_tokens_seen": 20917704,
      "step": 32080
    },
    {
      "epoch": 16.816037735849058,
      "grad_norm": 0.13499800860881805,
      "learning_rate": 7.526009217630886e-05,
      "loss": 0.4396,
      "num_input_tokens_seen": 20920808,
      "step": 32085
    },
    {
      "epoch": 16.81865828092243,
      "grad_norm": 0.16106879711151123,
      "learning_rate": 7.513947720219494e-05,
      "loss": 0.4045,
      "num_input_tokens_seen": 20923528,
      "step": 32090
    },
    {
      "epoch": 16.821278825995808,
      "grad_norm": 0.13495974242687225,
      "learning_rate": 7.501895110405533e-05,
      "loss": 0.3851,
      "num_input_tokens_seen": 20926376,
      "step": 32095
    },
    {
      "epoch": 16.82389937106918,
      "grad_norm": 0.10221949219703674,
      "learning_rate": 7.489851390710262e-05,
      "loss": 0.3906,
      "num_input_tokens_seen": 20929672,
      "step": 32100
    },
    {
      "epoch": 16.82651991614256,
      "grad_norm": 0.15593478083610535,
      "learning_rate": 7.477816563653095e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 20932616,
      "step": 32105
    },
    {
      "epoch": 16.829140461215932,
      "grad_norm": 0.3370157480239868,
      "learning_rate": 7.465790631751584e-05,
      "loss": 0.3254,
      "num_input_tokens_seen": 20935400,
      "step": 32110
    },
    {
      "epoch": 16.83176100628931,
      "grad_norm": 0.10782403498888016,
      "learning_rate": 7.453773597521429e-05,
      "loss": 0.358,
      "num_input_tokens_seen": 20938824,
      "step": 32115
    },
    {
      "epoch": 16.834381551362682,
      "grad_norm": 0.12689267098903656,
      "learning_rate": 7.44176546347643e-05,
      "loss": 0.5666,
      "num_input_tokens_seen": 20941960,
      "step": 32120
    },
    {
      "epoch": 16.83700209643606,
      "grad_norm": 0.21717365086078644,
      "learning_rate": 7.429766232128583e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 20948712,
      "step": 32125
    },
    {
      "epoch": 16.839622641509433,
      "grad_norm": 0.16951647400856018,
      "learning_rate": 7.417775905987956e-05,
      "loss": 0.4714,
      "num_input_tokens_seen": 20951592,
      "step": 32130
    },
    {
      "epoch": 16.84224318658281,
      "grad_norm": 0.14538921415805817,
      "learning_rate": 7.405794487562811e-05,
      "loss": 0.5156,
      "num_input_tokens_seen": 20955880,
      "step": 32135
    },
    {
      "epoch": 16.844863731656183,
      "grad_norm": 0.1260954886674881,
      "learning_rate": 7.393821979359528e-05,
      "loss": 0.401,
      "num_input_tokens_seen": 20959112,
      "step": 32140
    },
    {
      "epoch": 16.84748427672956,
      "grad_norm": 0.1235063448548317,
      "learning_rate": 7.3818583838826e-05,
      "loss": 0.2918,
      "num_input_tokens_seen": 20961992,
      "step": 32145
    },
    {
      "epoch": 16.850104821802937,
      "grad_norm": 0.12254821509122849,
      "learning_rate": 7.369903703634684e-05,
      "loss": 0.3319,
      "num_input_tokens_seen": 20964776,
      "step": 32150
    },
    {
      "epoch": 16.85272536687631,
      "grad_norm": 0.1383436620235443,
      "learning_rate": 7.357957941116572e-05,
      "loss": 0.4395,
      "num_input_tokens_seen": 20967944,
      "step": 32155
    },
    {
      "epoch": 16.855345911949687,
      "grad_norm": 0.1283493936061859,
      "learning_rate": 7.346021098827166e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 20971976,
      "step": 32160
    },
    {
      "epoch": 16.85796645702306,
      "grad_norm": 0.11580763757228851,
      "learning_rate": 7.334093179263518e-05,
      "loss": 0.391,
      "num_input_tokens_seen": 20975208,
      "step": 32165
    },
    {
      "epoch": 16.860587002096437,
      "grad_norm": 0.2923060357570648,
      "learning_rate": 7.322174184920837e-05,
      "loss": 0.4364,
      "num_input_tokens_seen": 20977736,
      "step": 32170
    },
    {
      "epoch": 16.86320754716981,
      "grad_norm": 0.1397806704044342,
      "learning_rate": 7.310264118292404e-05,
      "loss": 0.4207,
      "num_input_tokens_seen": 20981704,
      "step": 32175
    },
    {
      "epoch": 16.865828092243188,
      "grad_norm": 0.1430586576461792,
      "learning_rate": 7.298362981869705e-05,
      "loss": 0.5,
      "num_input_tokens_seen": 20985992,
      "step": 32180
    },
    {
      "epoch": 16.86844863731656,
      "grad_norm": 0.13749103248119354,
      "learning_rate": 7.286470778142284e-05,
      "loss": 0.5073,
      "num_input_tokens_seen": 20989544,
      "step": 32185
    },
    {
      "epoch": 16.871069182389938,
      "grad_norm": 0.28392449021339417,
      "learning_rate": 7.274587509597886e-05,
      "loss": 0.4781,
      "num_input_tokens_seen": 20996392,
      "step": 32190
    },
    {
      "epoch": 16.87368972746331,
      "grad_norm": 0.18204770982265472,
      "learning_rate": 7.262713178722346e-05,
      "loss": 0.4355,
      "num_input_tokens_seen": 20999400,
      "step": 32195
    },
    {
      "epoch": 16.87631027253669,
      "grad_norm": 0.18113519251346588,
      "learning_rate": 7.250847787999625e-05,
      "loss": 0.3671,
      "num_input_tokens_seen": 21001704,
      "step": 32200
    },
    {
      "epoch": 16.878930817610062,
      "grad_norm": 0.16268271207809448,
      "learning_rate": 7.238991339911844e-05,
      "loss": 0.4802,
      "num_input_tokens_seen": 21004872,
      "step": 32205
    },
    {
      "epoch": 16.88155136268344,
      "grad_norm": 0.1531711369752884,
      "learning_rate": 7.227143836939237e-05,
      "loss": 0.4991,
      "num_input_tokens_seen": 21011656,
      "step": 32210
    },
    {
      "epoch": 16.884171907756812,
      "grad_norm": 0.16534742712974548,
      "learning_rate": 7.215305281560153e-05,
      "loss": 0.4114,
      "num_input_tokens_seen": 21014472,
      "step": 32215
    },
    {
      "epoch": 16.88679245283019,
      "grad_norm": 0.23383700847625732,
      "learning_rate": 7.203475676251104e-05,
      "loss": 0.4183,
      "num_input_tokens_seen": 21017960,
      "step": 32220
    },
    {
      "epoch": 16.889412997903563,
      "grad_norm": 0.12821900844573975,
      "learning_rate": 7.191655023486682e-05,
      "loss": 0.4141,
      "num_input_tokens_seen": 21021288,
      "step": 32225
    },
    {
      "epoch": 16.89203354297694,
      "grad_norm": 0.11517009139060974,
      "learning_rate": 7.179843325739644e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 21024872,
      "step": 32230
    },
    {
      "epoch": 16.894654088050313,
      "grad_norm": 0.2428339272737503,
      "learning_rate": 7.168040585480861e-05,
      "loss": 0.3936,
      "num_input_tokens_seen": 21028008,
      "step": 32235
    },
    {
      "epoch": 16.89727463312369,
      "grad_norm": 0.1133684441447258,
      "learning_rate": 7.156246805179351e-05,
      "loss": 0.3015,
      "num_input_tokens_seen": 21030920,
      "step": 32240
    },
    {
      "epoch": 16.899895178197063,
      "grad_norm": 0.0788903534412384,
      "learning_rate": 7.144461987302208e-05,
      "loss": 0.4926,
      "num_input_tokens_seen": 21034248,
      "step": 32245
    },
    {
      "epoch": 16.90251572327044,
      "grad_norm": 0.15356788039207458,
      "learning_rate": 7.132686134314714e-05,
      "loss": 0.453,
      "num_input_tokens_seen": 21037672,
      "step": 32250
    },
    {
      "epoch": 16.905136268343817,
      "grad_norm": 0.16898782551288605,
      "learning_rate": 7.120919248680208e-05,
      "loss": 0.4737,
      "num_input_tokens_seen": 21041192,
      "step": 32255
    },
    {
      "epoch": 16.90775681341719,
      "grad_norm": 0.15874961018562317,
      "learning_rate": 7.109161332860203e-05,
      "loss": 0.3878,
      "num_input_tokens_seen": 21044488,
      "step": 32260
    },
    {
      "epoch": 16.910377358490567,
      "grad_norm": 0.18717683851718903,
      "learning_rate": 7.097412389314322e-05,
      "loss": 0.4858,
      "num_input_tokens_seen": 21047752,
      "step": 32265
    },
    {
      "epoch": 16.91299790356394,
      "grad_norm": 0.06411871314048767,
      "learning_rate": 7.085672420500322e-05,
      "loss": 0.3893,
      "num_input_tokens_seen": 21050344,
      "step": 32270
    },
    {
      "epoch": 16.915618448637318,
      "grad_norm": 0.10839046537876129,
      "learning_rate": 7.073941428874064e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 21053288,
      "step": 32275
    },
    {
      "epoch": 16.91823899371069,
      "grad_norm": 0.2226969599723816,
      "learning_rate": 7.062219416889514e-05,
      "loss": 0.4256,
      "num_input_tokens_seen": 21056904,
      "step": 32280
    },
    {
      "epoch": 16.920859538784068,
      "grad_norm": 0.16869007050991058,
      "learning_rate": 7.050506386998806e-05,
      "loss": 0.3571,
      "num_input_tokens_seen": 21059848,
      "step": 32285
    },
    {
      "epoch": 16.92348008385744,
      "grad_norm": 0.13943538069725037,
      "learning_rate": 7.038802341652172e-05,
      "loss": 0.3908,
      "num_input_tokens_seen": 21062472,
      "step": 32290
    },
    {
      "epoch": 16.92610062893082,
      "grad_norm": 0.04186463728547096,
      "learning_rate": 7.027107283297967e-05,
      "loss": 0.4149,
      "num_input_tokens_seen": 21067752,
      "step": 32295
    },
    {
      "epoch": 16.928721174004192,
      "grad_norm": 0.42286577820777893,
      "learning_rate": 7.015421214382661e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 21070632,
      "step": 32300
    },
    {
      "epoch": 16.93134171907757,
      "grad_norm": 0.15171824395656586,
      "learning_rate": 7.003744137350827e-05,
      "loss": 0.4817,
      "num_input_tokens_seen": 21074568,
      "step": 32305
    },
    {
      "epoch": 16.933962264150942,
      "grad_norm": 0.11607662588357925,
      "learning_rate": 6.992076054645197e-05,
      "loss": 0.3873,
      "num_input_tokens_seen": 21077288,
      "step": 32310
    },
    {
      "epoch": 16.93658280922432,
      "grad_norm": 0.13153834640979767,
      "learning_rate": 6.980416968706594e-05,
      "loss": 0.3678,
      "num_input_tokens_seen": 21079912,
      "step": 32315
    },
    {
      "epoch": 16.939203354297693,
      "grad_norm": 0.241196870803833,
      "learning_rate": 6.968766881973965e-05,
      "loss": 0.5207,
      "num_input_tokens_seen": 21084040,
      "step": 32320
    },
    {
      "epoch": 16.94182389937107,
      "grad_norm": 0.1843496412038803,
      "learning_rate": 6.957125796884395e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 21087528,
      "step": 32325
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 0.2017507553100586,
      "learning_rate": 6.945493715873046e-05,
      "loss": 0.3694,
      "num_input_tokens_seen": 21090888,
      "step": 32330
    },
    {
      "epoch": 16.94706498951782,
      "grad_norm": 0.21902619302272797,
      "learning_rate": 6.933870641373208e-05,
      "loss": 0.4694,
      "num_input_tokens_seen": 21094792,
      "step": 32335
    },
    {
      "epoch": 16.949685534591197,
      "grad_norm": 0.17529867589473724,
      "learning_rate": 6.92225657581631e-05,
      "loss": 0.3855,
      "num_input_tokens_seen": 21097736,
      "step": 32340
    },
    {
      "epoch": 16.95230607966457,
      "grad_norm": 0.1056528091430664,
      "learning_rate": 6.910651521631877e-05,
      "loss": 0.3058,
      "num_input_tokens_seen": 21100776,
      "step": 32345
    },
    {
      "epoch": 16.954926624737947,
      "grad_norm": 0.13054506480693817,
      "learning_rate": 6.899055481247568e-05,
      "loss": 0.426,
      "num_input_tokens_seen": 21104040,
      "step": 32350
    },
    {
      "epoch": 16.95754716981132,
      "grad_norm": 0.14816395938396454,
      "learning_rate": 6.88746845708913e-05,
      "loss": 0.3484,
      "num_input_tokens_seen": 21106856,
      "step": 32355
    },
    {
      "epoch": 16.960167714884697,
      "grad_norm": 0.22880223393440247,
      "learning_rate": 6.87589045158043e-05,
      "loss": 0.4181,
      "num_input_tokens_seen": 21109768,
      "step": 32360
    },
    {
      "epoch": 16.96278825995807,
      "grad_norm": 0.1306333690881729,
      "learning_rate": 6.864321467143459e-05,
      "loss": 0.4037,
      "num_input_tokens_seen": 21113512,
      "step": 32365
    },
    {
      "epoch": 16.965408805031448,
      "grad_norm": 0.17553429305553436,
      "learning_rate": 6.852761506198319e-05,
      "loss": 0.4385,
      "num_input_tokens_seen": 21117544,
      "step": 32370
    },
    {
      "epoch": 16.96802935010482,
      "grad_norm": 0.1111622080206871,
      "learning_rate": 6.841210571163231e-05,
      "loss": 0.3988,
      "num_input_tokens_seen": 21121864,
      "step": 32375
    },
    {
      "epoch": 16.970649895178198,
      "grad_norm": 0.11613044887781143,
      "learning_rate": 6.829668664454513e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 21124328,
      "step": 32380
    },
    {
      "epoch": 16.97327044025157,
      "grad_norm": 0.14381609857082367,
      "learning_rate": 6.818135788486584e-05,
      "loss": 0.3302,
      "num_input_tokens_seen": 21127336,
      "step": 32385
    },
    {
      "epoch": 16.97589098532495,
      "grad_norm": 0.10388179868459702,
      "learning_rate": 6.806611945672004e-05,
      "loss": 0.3343,
      "num_input_tokens_seen": 21129896,
      "step": 32390
    },
    {
      "epoch": 16.978511530398322,
      "grad_norm": 0.18251623213291168,
      "learning_rate": 6.79509713842143e-05,
      "loss": 0.4391,
      "num_input_tokens_seen": 21133512,
      "step": 32395
    },
    {
      "epoch": 16.9811320754717,
      "grad_norm": 0.11036914587020874,
      "learning_rate": 6.783591369143626e-05,
      "loss": 0.4058,
      "num_input_tokens_seen": 21135816,
      "step": 32400
    },
    {
      "epoch": 16.983752620545072,
      "grad_norm": 0.12408167868852615,
      "learning_rate": 6.77209464024548e-05,
      "loss": 0.2922,
      "num_input_tokens_seen": 21138536,
      "step": 32405
    },
    {
      "epoch": 16.98637316561845,
      "grad_norm": 0.16448728740215302,
      "learning_rate": 6.760606954131965e-05,
      "loss": 0.4355,
      "num_input_tokens_seen": 21141160,
      "step": 32410
    },
    {
      "epoch": 16.988993710691823,
      "grad_norm": 0.17082342505455017,
      "learning_rate": 6.749128313206165e-05,
      "loss": 0.3992,
      "num_input_tokens_seen": 21143624,
      "step": 32415
    },
    {
      "epoch": 16.9916142557652,
      "grad_norm": 0.1194399744272232,
      "learning_rate": 6.737658719869288e-05,
      "loss": 0.4917,
      "num_input_tokens_seen": 21147528,
      "step": 32420
    },
    {
      "epoch": 16.994234800838573,
      "grad_norm": 0.12665824592113495,
      "learning_rate": 6.726198176520642e-05,
      "loss": 0.4281,
      "num_input_tokens_seen": 21150376,
      "step": 32425
    },
    {
      "epoch": 16.99685534591195,
      "grad_norm": 0.2550720274448395,
      "learning_rate": 6.71474668555766e-05,
      "loss": 0.464,
      "num_input_tokens_seen": 21155016,
      "step": 32430
    },
    {
      "epoch": 16.999475890985323,
      "grad_norm": 0.1098184883594513,
      "learning_rate": 6.703304249375836e-05,
      "loss": 0.2553,
      "num_input_tokens_seen": 21157992,
      "step": 32435
    },
    {
      "epoch": 17.0,
      "eval_loss": 0.46461933851242065,
      "eval_runtime": 13.5646,
      "eval_samples_per_second": 62.516,
      "eval_steps_per_second": 15.629,
      "num_input_tokens_seen": 21158120,
      "step": 32436
    },
    {
      "epoch": 17.0020964360587,
      "grad_norm": 0.10452546924352646,
      "learning_rate": 6.691870870368815e-05,
      "loss": 0.3022,
      "num_input_tokens_seen": 21160712,
      "step": 32440
    },
    {
      "epoch": 17.004716981132077,
      "grad_norm": 0.11369583010673523,
      "learning_rate": 6.680446550928316e-05,
      "loss": 0.386,
      "num_input_tokens_seen": 21163528,
      "step": 32445
    },
    {
      "epoch": 17.00733752620545,
      "grad_norm": 0.224717915058136,
      "learning_rate": 6.669031293444177e-05,
      "loss": 0.5215,
      "num_input_tokens_seen": 21167304,
      "step": 32450
    },
    {
      "epoch": 17.009958071278827,
      "grad_norm": 0.12938514351844788,
      "learning_rate": 6.65762510030436e-05,
      "loss": 0.4331,
      "num_input_tokens_seen": 21170792,
      "step": 32455
    },
    {
      "epoch": 17.0125786163522,
      "grad_norm": 0.1341233104467392,
      "learning_rate": 6.646227973894886e-05,
      "loss": 0.3956,
      "num_input_tokens_seen": 21174408,
      "step": 32460
    },
    {
      "epoch": 17.015199161425578,
      "grad_norm": 0.09129804372787476,
      "learning_rate": 6.634839916599921e-05,
      "loss": 0.4642,
      "num_input_tokens_seen": 21178248,
      "step": 32465
    },
    {
      "epoch": 17.01781970649895,
      "grad_norm": 0.1476486772298813,
      "learning_rate": 6.623460930801701e-05,
      "loss": 0.3157,
      "num_input_tokens_seen": 21181064,
      "step": 32470
    },
    {
      "epoch": 17.020440251572328,
      "grad_norm": 0.14440107345581055,
      "learning_rate": 6.61209101888059e-05,
      "loss": 0.4595,
      "num_input_tokens_seen": 21184296,
      "step": 32475
    },
    {
      "epoch": 17.0230607966457,
      "grad_norm": 0.3723008334636688,
      "learning_rate": 6.600730183215043e-05,
      "loss": 0.4821,
      "num_input_tokens_seen": 21187176,
      "step": 32480
    },
    {
      "epoch": 17.02568134171908,
      "grad_norm": 0.21481585502624512,
      "learning_rate": 6.589378426181624e-05,
      "loss": 0.3277,
      "num_input_tokens_seen": 21189800,
      "step": 32485
    },
    {
      "epoch": 17.028301886792452,
      "grad_norm": 0.11847139894962311,
      "learning_rate": 6.578035750154976e-05,
      "loss": 0.4992,
      "num_input_tokens_seen": 21196328,
      "step": 32490
    },
    {
      "epoch": 17.03092243186583,
      "grad_norm": 0.14419110119342804,
      "learning_rate": 6.566702157507875e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 21199400,
      "step": 32495
    },
    {
      "epoch": 17.033542976939202,
      "grad_norm": 0.09918327629566193,
      "learning_rate": 6.555377650611155e-05,
      "loss": 0.3607,
      "num_input_tokens_seen": 21202248,
      "step": 32500
    },
    {
      "epoch": 17.03616352201258,
      "grad_norm": 0.19980011880397797,
      "learning_rate": 6.544062231833792e-05,
      "loss": 0.5154,
      "num_input_tokens_seen": 21205448,
      "step": 32505
    },
    {
      "epoch": 17.038784067085953,
      "grad_norm": 0.09319867193698883,
      "learning_rate": 6.532755903542847e-05,
      "loss": 0.3544,
      "num_input_tokens_seen": 21209576,
      "step": 32510
    },
    {
      "epoch": 17.04140461215933,
      "grad_norm": 0.1431431621313095,
      "learning_rate": 6.521458668103458e-05,
      "loss": 0.4569,
      "num_input_tokens_seen": 21212904,
      "step": 32515
    },
    {
      "epoch": 17.044025157232703,
      "grad_norm": 0.12940676510334015,
      "learning_rate": 6.510170527878889e-05,
      "loss": 0.4073,
      "num_input_tokens_seen": 21216104,
      "step": 32520
    },
    {
      "epoch": 17.04664570230608,
      "grad_norm": 0.12328775227069855,
      "learning_rate": 6.498891485230491e-05,
      "loss": 0.4958,
      "num_input_tokens_seen": 21220104,
      "step": 32525
    },
    {
      "epoch": 17.049266247379453,
      "grad_norm": 0.18756842613220215,
      "learning_rate": 6.487621542517703e-05,
      "loss": 0.4923,
      "num_input_tokens_seen": 21222920,
      "step": 32530
    },
    {
      "epoch": 17.05188679245283,
      "grad_norm": 0.12307632714509964,
      "learning_rate": 6.476360702098078e-05,
      "loss": 0.387,
      "num_input_tokens_seen": 21226408,
      "step": 32535
    },
    {
      "epoch": 17.054507337526207,
      "grad_norm": 0.21159134805202484,
      "learning_rate": 6.465108966327243e-05,
      "loss": 0.5131,
      "num_input_tokens_seen": 21230184,
      "step": 32540
    },
    {
      "epoch": 17.05712788259958,
      "grad_norm": 0.17715546488761902,
      "learning_rate": 6.453866337558939e-05,
      "loss": 0.3885,
      "num_input_tokens_seen": 21233288,
      "step": 32545
    },
    {
      "epoch": 17.059748427672957,
      "grad_norm": 0.15956540405750275,
      "learning_rate": 6.442632818145011e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 21235720,
      "step": 32550
    },
    {
      "epoch": 17.06236897274633,
      "grad_norm": 0.18334928154945374,
      "learning_rate": 6.431408410435352e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 21238408,
      "step": 32555
    },
    {
      "epoch": 17.064989517819708,
      "grad_norm": 0.08726285398006439,
      "learning_rate": 6.420193116778e-05,
      "loss": 0.3644,
      "num_input_tokens_seen": 21242632,
      "step": 32560
    },
    {
      "epoch": 17.06761006289308,
      "grad_norm": 0.11888047307729721,
      "learning_rate": 6.408986939519074e-05,
      "loss": 0.3888,
      "num_input_tokens_seen": 21246056,
      "step": 32565
    },
    {
      "epoch": 17.070230607966458,
      "grad_norm": 0.17504338920116425,
      "learning_rate": 6.397789881002752e-05,
      "loss": 0.5273,
      "num_input_tokens_seen": 21249032,
      "step": 32570
    },
    {
      "epoch": 17.07285115303983,
      "grad_norm": 0.17936451733112335,
      "learning_rate": 6.386601943571352e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 21251816,
      "step": 32575
    },
    {
      "epoch": 17.07547169811321,
      "grad_norm": 0.07429858297109604,
      "learning_rate": 6.375423129565266e-05,
      "loss": 0.3586,
      "num_input_tokens_seen": 21256200,
      "step": 32580
    },
    {
      "epoch": 17.078092243186582,
      "grad_norm": 0.22023043036460876,
      "learning_rate": 6.36425344132296e-05,
      "loss": 0.4003,
      "num_input_tokens_seen": 21260392,
      "step": 32585
    },
    {
      "epoch": 17.08071278825996,
      "grad_norm": 0.12041507661342621,
      "learning_rate": 6.353092881181016e-05,
      "loss": 0.4628,
      "num_input_tokens_seen": 21264296,
      "step": 32590
    },
    {
      "epoch": 17.083333333333332,
      "grad_norm": 0.1000756174325943,
      "learning_rate": 6.341941451474082e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 21267848,
      "step": 32595
    },
    {
      "epoch": 17.08595387840671,
      "grad_norm": 0.11405497789382935,
      "learning_rate": 6.330799154534921e-05,
      "loss": 0.3766,
      "num_input_tokens_seen": 21271560,
      "step": 32600
    },
    {
      "epoch": 17.088574423480082,
      "grad_norm": 0.27862030267715454,
      "learning_rate": 6.319665992694368e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 21274760,
      "step": 32605
    },
    {
      "epoch": 17.09119496855346,
      "grad_norm": 0.18050868809223175,
      "learning_rate": 6.308541968281373e-05,
      "loss": 0.3702,
      "num_input_tokens_seen": 21277352,
      "step": 32610
    },
    {
      "epoch": 17.093815513626833,
      "grad_norm": 0.2215503454208374,
      "learning_rate": 6.297427083622936e-05,
      "loss": 0.5639,
      "num_input_tokens_seen": 21280168,
      "step": 32615
    },
    {
      "epoch": 17.09643605870021,
      "grad_norm": 0.16711832582950592,
      "learning_rate": 6.28632134104416e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 21283688,
      "step": 32620
    },
    {
      "epoch": 17.099056603773583,
      "grad_norm": 0.1555449515581131,
      "learning_rate": 6.275224742868247e-05,
      "loss": 0.442,
      "num_input_tokens_seen": 21286920,
      "step": 32625
    },
    {
      "epoch": 17.10167714884696,
      "grad_norm": 0.1341559886932373,
      "learning_rate": 6.264137291416477e-05,
      "loss": 0.4369,
      "num_input_tokens_seen": 21290600,
      "step": 32630
    },
    {
      "epoch": 17.104297693920337,
      "grad_norm": 0.14660100638866425,
      "learning_rate": 6.253058989008226e-05,
      "loss": 0.4711,
      "num_input_tokens_seen": 21293032,
      "step": 32635
    },
    {
      "epoch": 17.10691823899371,
      "grad_norm": 0.17274317145347595,
      "learning_rate": 6.241989837960949e-05,
      "loss": 0.4518,
      "num_input_tokens_seen": 21295976,
      "step": 32640
    },
    {
      "epoch": 17.109538784067087,
      "grad_norm": 0.08982937783002853,
      "learning_rate": 6.230929840590177e-05,
      "loss": 0.4509,
      "num_input_tokens_seen": 21299816,
      "step": 32645
    },
    {
      "epoch": 17.11215932914046,
      "grad_norm": 0.2734285891056061,
      "learning_rate": 6.219878999209533e-05,
      "loss": 0.4198,
      "num_input_tokens_seen": 21302696,
      "step": 32650
    },
    {
      "epoch": 17.114779874213838,
      "grad_norm": 0.13535374402999878,
      "learning_rate": 6.208837316130733e-05,
      "loss": 0.4378,
      "num_input_tokens_seen": 21305288,
      "step": 32655
    },
    {
      "epoch": 17.11740041928721,
      "grad_norm": 0.17365321516990662,
      "learning_rate": 6.197804793663564e-05,
      "loss": 0.3509,
      "num_input_tokens_seen": 21308808,
      "step": 32660
    },
    {
      "epoch": 17.120020964360588,
      "grad_norm": 0.19016236066818237,
      "learning_rate": 6.18678143411593e-05,
      "loss": 0.394,
      "num_input_tokens_seen": 21312008,
      "step": 32665
    },
    {
      "epoch": 17.12264150943396,
      "grad_norm": 0.2777344286441803,
      "learning_rate": 6.175767239793767e-05,
      "loss": 0.4111,
      "num_input_tokens_seen": 21314280,
      "step": 32670
    },
    {
      "epoch": 17.12526205450734,
      "grad_norm": 0.1449241191148758,
      "learning_rate": 6.164762213001112e-05,
      "loss": 0.3724,
      "num_input_tokens_seen": 21316808,
      "step": 32675
    },
    {
      "epoch": 17.127882599580712,
      "grad_norm": 0.2110089808702469,
      "learning_rate": 6.153766356040107e-05,
      "loss": 0.418,
      "num_input_tokens_seen": 21320712,
      "step": 32680
    },
    {
      "epoch": 17.13050314465409,
      "grad_norm": 0.15167084336280823,
      "learning_rate": 6.142779671210951e-05,
      "loss": 0.4167,
      "num_input_tokens_seen": 21324968,
      "step": 32685
    },
    {
      "epoch": 17.133123689727462,
      "grad_norm": 0.14734341204166412,
      "learning_rate": 6.131802160811956e-05,
      "loss": 0.4208,
      "num_input_tokens_seen": 21328360,
      "step": 32690
    },
    {
      "epoch": 17.13574423480084,
      "grad_norm": 0.1755588799715042,
      "learning_rate": 6.120833827139466e-05,
      "loss": 0.4293,
      "num_input_tokens_seen": 21331336,
      "step": 32695
    },
    {
      "epoch": 17.138364779874212,
      "grad_norm": 0.13264623284339905,
      "learning_rate": 6.109874672487936e-05,
      "loss": 0.5429,
      "num_input_tokens_seen": 21334344,
      "step": 32700
    },
    {
      "epoch": 17.14098532494759,
      "grad_norm": 0.1385704129934311,
      "learning_rate": 6.098924699149894e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 21340360,
      "step": 32705
    },
    {
      "epoch": 17.143605870020963,
      "grad_norm": 0.20703205466270447,
      "learning_rate": 6.087983909415962e-05,
      "loss": 0.4121,
      "num_input_tokens_seen": 21343176,
      "step": 32710
    },
    {
      "epoch": 17.14622641509434,
      "grad_norm": 0.15366561710834503,
      "learning_rate": 6.077052305574815e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 21346024,
      "step": 32715
    },
    {
      "epoch": 17.148846960167713,
      "grad_norm": 0.0773748829960823,
      "learning_rate": 6.066129889913241e-05,
      "loss": 0.4041,
      "num_input_tokens_seen": 21350120,
      "step": 32720
    },
    {
      "epoch": 17.15146750524109,
      "grad_norm": 0.10408826917409897,
      "learning_rate": 6.0552166647160676e-05,
      "loss": 0.5483,
      "num_input_tokens_seen": 21353608,
      "step": 32725
    },
    {
      "epoch": 17.154088050314467,
      "grad_norm": 0.1141272559762001,
      "learning_rate": 6.044312632266208e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 21358376,
      "step": 32730
    },
    {
      "epoch": 17.15670859538784,
      "grad_norm": 0.2290128767490387,
      "learning_rate": 6.03341779484467e-05,
      "loss": 0.3568,
      "num_input_tokens_seen": 21361128,
      "step": 32735
    },
    {
      "epoch": 17.159329140461217,
      "grad_norm": 0.17703039944171906,
      "learning_rate": 6.022532154730537e-05,
      "loss": 0.4453,
      "num_input_tokens_seen": 21364584,
      "step": 32740
    },
    {
      "epoch": 17.16194968553459,
      "grad_norm": 0.10489712655544281,
      "learning_rate": 6.011655714200964e-05,
      "loss": 0.3939,
      "num_input_tokens_seen": 21368168,
      "step": 32745
    },
    {
      "epoch": 17.164570230607968,
      "grad_norm": 0.08304214477539062,
      "learning_rate": 6.0007884755311715e-05,
      "loss": 0.3776,
      "num_input_tokens_seen": 21371752,
      "step": 32750
    },
    {
      "epoch": 17.16719077568134,
      "grad_norm": 0.1882385015487671,
      "learning_rate": 5.989930440994451e-05,
      "loss": 0.5052,
      "num_input_tokens_seen": 21374632,
      "step": 32755
    },
    {
      "epoch": 17.169811320754718,
      "grad_norm": 0.08972078561782837,
      "learning_rate": 5.979081612862186e-05,
      "loss": 0.4309,
      "num_input_tokens_seen": 21377608,
      "step": 32760
    },
    {
      "epoch": 17.17243186582809,
      "grad_norm": 0.12054939568042755,
      "learning_rate": 5.968241993403834e-05,
      "loss": 0.4814,
      "num_input_tokens_seen": 21381192,
      "step": 32765
    },
    {
      "epoch": 17.17505241090147,
      "grad_norm": 0.1699236035346985,
      "learning_rate": 5.957411584886924e-05,
      "loss": 0.59,
      "num_input_tokens_seen": 21384168,
      "step": 32770
    },
    {
      "epoch": 17.177672955974842,
      "grad_norm": 0.23477011919021606,
      "learning_rate": 5.946590389577033e-05,
      "loss": 0.4088,
      "num_input_tokens_seen": 21387240,
      "step": 32775
    },
    {
      "epoch": 17.18029350104822,
      "grad_norm": 0.09835123270750046,
      "learning_rate": 5.935778409737857e-05,
      "loss": 0.3617,
      "num_input_tokens_seen": 21390216,
      "step": 32780
    },
    {
      "epoch": 17.182914046121592,
      "grad_norm": 0.18353533744812012,
      "learning_rate": 5.92497564763112e-05,
      "loss": 0.4383,
      "num_input_tokens_seen": 21393544,
      "step": 32785
    },
    {
      "epoch": 17.18553459119497,
      "grad_norm": 0.14068907499313354,
      "learning_rate": 5.914182105516641e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 21397288,
      "step": 32790
    },
    {
      "epoch": 17.188155136268342,
      "grad_norm": 0.11121035367250443,
      "learning_rate": 5.903397785652304e-05,
      "loss": 0.3427,
      "num_input_tokens_seen": 21399912,
      "step": 32795
    },
    {
      "epoch": 17.19077568134172,
      "grad_norm": 0.18641775846481323,
      "learning_rate": 5.8926226902940804e-05,
      "loss": 0.4168,
      "num_input_tokens_seen": 21402472,
      "step": 32800
    },
    {
      "epoch": 17.193396226415093,
      "grad_norm": 0.11298584192991257,
      "learning_rate": 5.88185682169598e-05,
      "loss": 0.3751,
      "num_input_tokens_seen": 21405352,
      "step": 32805
    },
    {
      "epoch": 17.19601677148847,
      "grad_norm": 0.14838142693042755,
      "learning_rate": 5.871100182110117e-05,
      "loss": 0.3594,
      "num_input_tokens_seen": 21408008,
      "step": 32810
    },
    {
      "epoch": 17.198637316561843,
      "grad_norm": 0.1708192080259323,
      "learning_rate": 5.8603527737866314e-05,
      "loss": 0.5386,
      "num_input_tokens_seen": 21411880,
      "step": 32815
    },
    {
      "epoch": 17.20125786163522,
      "grad_norm": 0.18859629333019257,
      "learning_rate": 5.849614598973779e-05,
      "loss": 0.4141,
      "num_input_tokens_seen": 21414504,
      "step": 32820
    },
    {
      "epoch": 17.203878406708597,
      "grad_norm": 0.21117328107357025,
      "learning_rate": 5.8388856599178644e-05,
      "loss": 0.394,
      "num_input_tokens_seen": 21417608,
      "step": 32825
    },
    {
      "epoch": 17.20649895178197,
      "grad_norm": 0.14363126456737518,
      "learning_rate": 5.8281659588632464e-05,
      "loss": 0.4806,
      "num_input_tokens_seen": 21420712,
      "step": 32830
    },
    {
      "epoch": 17.209119496855347,
      "grad_norm": 0.14485885202884674,
      "learning_rate": 5.817455498052382e-05,
      "loss": 0.6107,
      "num_input_tokens_seen": 21423400,
      "step": 32835
    },
    {
      "epoch": 17.21174004192872,
      "grad_norm": 0.20286265015602112,
      "learning_rate": 5.806754279725751e-05,
      "loss": 0.493,
      "num_input_tokens_seen": 21426888,
      "step": 32840
    },
    {
      "epoch": 17.214360587002098,
      "grad_norm": 0.1764051616191864,
      "learning_rate": 5.796062306121947e-05,
      "loss": 0.4502,
      "num_input_tokens_seen": 21429736,
      "step": 32845
    },
    {
      "epoch": 17.21698113207547,
      "grad_norm": 0.18448859453201294,
      "learning_rate": 5.785379579477607e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 21433320,
      "step": 32850
    },
    {
      "epoch": 17.219601677148848,
      "grad_norm": 0.1824064701795578,
      "learning_rate": 5.774706102027427e-05,
      "loss": 0.3902,
      "num_input_tokens_seen": 21436584,
      "step": 32855
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 0.1351414918899536,
      "learning_rate": 5.7640418760041776e-05,
      "loss": 0.5389,
      "num_input_tokens_seen": 21439528,
      "step": 32860
    },
    {
      "epoch": 17.2248427672956,
      "grad_norm": 0.1746259182691574,
      "learning_rate": 5.753386903638713e-05,
      "loss": 0.3843,
      "num_input_tokens_seen": 21442824,
      "step": 32865
    },
    {
      "epoch": 17.22746331236897,
      "grad_norm": 0.15130648016929626,
      "learning_rate": 5.7427411871599e-05,
      "loss": 0.463,
      "num_input_tokens_seen": 21446024,
      "step": 32870
    },
    {
      "epoch": 17.23008385744235,
      "grad_norm": 0.15716566145420074,
      "learning_rate": 5.7321047287947235e-05,
      "loss": 0.3175,
      "num_input_tokens_seen": 21448744,
      "step": 32875
    },
    {
      "epoch": 17.232704402515722,
      "grad_norm": 0.17260615527629852,
      "learning_rate": 5.72147753076821e-05,
      "loss": 0.4293,
      "num_input_tokens_seen": 21451688,
      "step": 32880
    },
    {
      "epoch": 17.2353249475891,
      "grad_norm": 0.16851313412189484,
      "learning_rate": 5.7108595953034306e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 21453864,
      "step": 32885
    },
    {
      "epoch": 17.237945492662472,
      "grad_norm": 0.15852101147174835,
      "learning_rate": 5.7002509246215415e-05,
      "loss": 0.4352,
      "num_input_tokens_seen": 21456936,
      "step": 32890
    },
    {
      "epoch": 17.24056603773585,
      "grad_norm": 0.1632077693939209,
      "learning_rate": 5.68965152094178e-05,
      "loss": 0.3835,
      "num_input_tokens_seen": 21460424,
      "step": 32895
    },
    {
      "epoch": 17.243186582809223,
      "grad_norm": 0.16383743286132812,
      "learning_rate": 5.679061386481388e-05,
      "loss": 0.41,
      "num_input_tokens_seen": 21463016,
      "step": 32900
    },
    {
      "epoch": 17.2458071278826,
      "grad_norm": 0.09722322225570679,
      "learning_rate": 5.6684805234557216e-05,
      "loss": 0.3866,
      "num_input_tokens_seen": 21466632,
      "step": 32905
    },
    {
      "epoch": 17.248427672955973,
      "grad_norm": 0.14162641763687134,
      "learning_rate": 5.65790893407816e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 21469544,
      "step": 32910
    },
    {
      "epoch": 17.25104821802935,
      "grad_norm": 0.15440981090068817,
      "learning_rate": 5.647346620560168e-05,
      "loss": 0.351,
      "num_input_tokens_seen": 21472040,
      "step": 32915
    },
    {
      "epoch": 17.253668763102727,
      "grad_norm": 0.1420236974954605,
      "learning_rate": 5.6367935851112605e-05,
      "loss": 0.3674,
      "num_input_tokens_seen": 21475784,
      "step": 32920
    },
    {
      "epoch": 17.2562893081761,
      "grad_norm": 0.10804670304059982,
      "learning_rate": 5.626249829939023e-05,
      "loss": 0.3636,
      "num_input_tokens_seen": 21479656,
      "step": 32925
    },
    {
      "epoch": 17.258909853249477,
      "grad_norm": 0.14517448842525482,
      "learning_rate": 5.6157153572490795e-05,
      "loss": 0.3983,
      "num_input_tokens_seen": 21482824,
      "step": 32930
    },
    {
      "epoch": 17.26153039832285,
      "grad_norm": 0.29027870297431946,
      "learning_rate": 5.6051901692451115e-05,
      "loss": 0.4061,
      "num_input_tokens_seen": 21485672,
      "step": 32935
    },
    {
      "epoch": 17.264150943396228,
      "grad_norm": 0.19138745963573456,
      "learning_rate": 5.5946742681288744e-05,
      "loss": 0.3703,
      "num_input_tokens_seen": 21488616,
      "step": 32940
    },
    {
      "epoch": 17.2667714884696,
      "grad_norm": 0.14404982328414917,
      "learning_rate": 5.5841676561001805e-05,
      "loss": 0.3906,
      "num_input_tokens_seen": 21490952,
      "step": 32945
    },
    {
      "epoch": 17.269392033542978,
      "grad_norm": 0.13037708401679993,
      "learning_rate": 5.573670335356901e-05,
      "loss": 0.4118,
      "num_input_tokens_seen": 21493832,
      "step": 32950
    },
    {
      "epoch": 17.27201257861635,
      "grad_norm": 0.24058601260185242,
      "learning_rate": 5.563182308094933e-05,
      "loss": 0.3978,
      "num_input_tokens_seen": 21496136,
      "step": 32955
    },
    {
      "epoch": 17.27463312368973,
      "grad_norm": 0.1498773843050003,
      "learning_rate": 5.552703576508272e-05,
      "loss": 0.4674,
      "num_input_tokens_seen": 21499432,
      "step": 32960
    },
    {
      "epoch": 17.2772536687631,
      "grad_norm": 0.0920444056391716,
      "learning_rate": 5.5422341427889324e-05,
      "loss": 0.5719,
      "num_input_tokens_seen": 21502472,
      "step": 32965
    },
    {
      "epoch": 17.27987421383648,
      "grad_norm": 0.10920968651771545,
      "learning_rate": 5.531774009127011e-05,
      "loss": 0.3526,
      "num_input_tokens_seen": 21506088,
      "step": 32970
    },
    {
      "epoch": 17.282494758909852,
      "grad_norm": 0.13674934208393097,
      "learning_rate": 5.521323177710647e-05,
      "loss": 0.3815,
      "num_input_tokens_seen": 21508904,
      "step": 32975
    },
    {
      "epoch": 17.28511530398323,
      "grad_norm": 0.11527694761753082,
      "learning_rate": 5.510881650726046e-05,
      "loss": 0.3538,
      "num_input_tokens_seen": 21511624,
      "step": 32980
    },
    {
      "epoch": 17.287735849056602,
      "grad_norm": 0.0991603434085846,
      "learning_rate": 5.50044943035744e-05,
      "loss": 0.3559,
      "num_input_tokens_seen": 21514504,
      "step": 32985
    },
    {
      "epoch": 17.29035639412998,
      "grad_norm": 0.2321290820837021,
      "learning_rate": 5.490026518787128e-05,
      "loss": 0.5067,
      "num_input_tokens_seen": 21517256,
      "step": 32990
    },
    {
      "epoch": 17.292976939203353,
      "grad_norm": 0.10225740820169449,
      "learning_rate": 5.47961291819547e-05,
      "loss": 0.3926,
      "num_input_tokens_seen": 21520648,
      "step": 32995
    },
    {
      "epoch": 17.29559748427673,
      "grad_norm": 0.11932485550642014,
      "learning_rate": 5.46920863076088e-05,
      "loss": 0.3751,
      "num_input_tokens_seen": 21523688,
      "step": 33000
    },
    {
      "epoch": 17.298218029350103,
      "grad_norm": 0.15847986936569214,
      "learning_rate": 5.4588136586598056e-05,
      "loss": 0.4624,
      "num_input_tokens_seen": 21527016,
      "step": 33005
    },
    {
      "epoch": 17.30083857442348,
      "grad_norm": 0.14785662293434143,
      "learning_rate": 5.44842800406678e-05,
      "loss": 0.3704,
      "num_input_tokens_seen": 21530408,
      "step": 33010
    },
    {
      "epoch": 17.303459119496857,
      "grad_norm": 0.12391319870948792,
      "learning_rate": 5.438051669154326e-05,
      "loss": 0.4828,
      "num_input_tokens_seen": 21533288,
      "step": 33015
    },
    {
      "epoch": 17.30607966457023,
      "grad_norm": 0.17971129715442657,
      "learning_rate": 5.427684656093074e-05,
      "loss": 0.4014,
      "num_input_tokens_seen": 21536520,
      "step": 33020
    },
    {
      "epoch": 17.308700209643607,
      "grad_norm": 0.11751685291528702,
      "learning_rate": 5.417326967051683e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 21539304,
      "step": 33025
    },
    {
      "epoch": 17.31132075471698,
      "grad_norm": 0.21919922530651093,
      "learning_rate": 5.4069786041968663e-05,
      "loss": 0.3808,
      "num_input_tokens_seen": 21542472,
      "step": 33030
    },
    {
      "epoch": 17.313941299790358,
      "grad_norm": 0.1370256096124649,
      "learning_rate": 5.39663956969339e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 21545736,
      "step": 33035
    },
    {
      "epoch": 17.31656184486373,
      "grad_norm": 0.1260453462600708,
      "learning_rate": 5.3863098657040534e-05,
      "loss": 0.5184,
      "num_input_tokens_seen": 21548744,
      "step": 33040
    },
    {
      "epoch": 17.319182389937108,
      "grad_norm": 0.16790710389614105,
      "learning_rate": 5.375989494389705e-05,
      "loss": 0.527,
      "num_input_tokens_seen": 21551592,
      "step": 33045
    },
    {
      "epoch": 17.32180293501048,
      "grad_norm": 0.2028276026248932,
      "learning_rate": 5.365678457909257e-05,
      "loss": 0.4336,
      "num_input_tokens_seen": 21554216,
      "step": 33050
    },
    {
      "epoch": 17.32442348008386,
      "grad_norm": 0.1518654078245163,
      "learning_rate": 5.3553767584196555e-05,
      "loss": 0.3176,
      "num_input_tokens_seen": 21556776,
      "step": 33055
    },
    {
      "epoch": 17.32704402515723,
      "grad_norm": 0.07850047945976257,
      "learning_rate": 5.3450843980759166e-05,
      "loss": 0.4414,
      "num_input_tokens_seen": 21559976,
      "step": 33060
    },
    {
      "epoch": 17.32966457023061,
      "grad_norm": 0.15518464148044586,
      "learning_rate": 5.3348013790310736e-05,
      "loss": 0.4347,
      "num_input_tokens_seen": 21562792,
      "step": 33065
    },
    {
      "epoch": 17.332285115303982,
      "grad_norm": 0.11513829976320267,
      "learning_rate": 5.324527703436199e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 21565736,
      "step": 33070
    },
    {
      "epoch": 17.33490566037736,
      "grad_norm": 0.13521626591682434,
      "learning_rate": 5.314263373440448e-05,
      "loss": 0.3595,
      "num_input_tokens_seen": 21569864,
      "step": 33075
    },
    {
      "epoch": 17.337526205450732,
      "grad_norm": 0.12567049264907837,
      "learning_rate": 5.304008391190995e-05,
      "loss": 0.2607,
      "num_input_tokens_seen": 21572776,
      "step": 33080
    },
    {
      "epoch": 17.34014675052411,
      "grad_norm": 0.12987373769283295,
      "learning_rate": 5.293762758833071e-05,
      "loss": 0.5977,
      "num_input_tokens_seen": 21576584,
      "step": 33085
    },
    {
      "epoch": 17.342767295597483,
      "grad_norm": 0.1420259326696396,
      "learning_rate": 5.283526478509953e-05,
      "loss": 0.4244,
      "num_input_tokens_seen": 21579496,
      "step": 33090
    },
    {
      "epoch": 17.34538784067086,
      "grad_norm": 0.13781410455703735,
      "learning_rate": 5.273299552362942e-05,
      "loss": 0.4943,
      "num_input_tokens_seen": 21582120,
      "step": 33095
    },
    {
      "epoch": 17.348008385744233,
      "grad_norm": 0.15489032864570618,
      "learning_rate": 5.263081982531381e-05,
      "loss": 0.479,
      "num_input_tokens_seen": 21585224,
      "step": 33100
    },
    {
      "epoch": 17.35062893081761,
      "grad_norm": 0.23026913404464722,
      "learning_rate": 5.2528737711526885e-05,
      "loss": 0.3173,
      "num_input_tokens_seen": 21588264,
      "step": 33105
    },
    {
      "epoch": 17.353249475890987,
      "grad_norm": 0.17486748099327087,
      "learning_rate": 5.2426749203623005e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 21591464,
      "step": 33110
    },
    {
      "epoch": 17.35587002096436,
      "grad_norm": 0.19935907423496246,
      "learning_rate": 5.23248543229371e-05,
      "loss": 0.4028,
      "num_input_tokens_seen": 21594024,
      "step": 33115
    },
    {
      "epoch": 17.358490566037737,
      "grad_norm": 0.1969933956861496,
      "learning_rate": 5.222305309078429e-05,
      "loss": 0.4915,
      "num_input_tokens_seen": 21596776,
      "step": 33120
    },
    {
      "epoch": 17.36111111111111,
      "grad_norm": 0.19442203640937805,
      "learning_rate": 5.2121345528460205e-05,
      "loss": 0.2863,
      "num_input_tokens_seen": 21600392,
      "step": 33125
    },
    {
      "epoch": 17.363731656184488,
      "grad_norm": 0.1957457959651947,
      "learning_rate": 5.201973165724094e-05,
      "loss": 0.5997,
      "num_input_tokens_seen": 21603880,
      "step": 33130
    },
    {
      "epoch": 17.36635220125786,
      "grad_norm": 0.13630926609039307,
      "learning_rate": 5.1918211498382996e-05,
      "loss": 0.4048,
      "num_input_tokens_seen": 21606600,
      "step": 33135
    },
    {
      "epoch": 17.368972746331238,
      "grad_norm": 0.10675479471683502,
      "learning_rate": 5.181678507312332e-05,
      "loss": 0.5321,
      "num_input_tokens_seen": 21610856,
      "step": 33140
    },
    {
      "epoch": 17.37159329140461,
      "grad_norm": 0.11301186680793762,
      "learning_rate": 5.171545240267894e-05,
      "loss": 0.3787,
      "num_input_tokens_seen": 21615016,
      "step": 33145
    },
    {
      "epoch": 17.37421383647799,
      "grad_norm": 0.1195489838719368,
      "learning_rate": 5.161421350824774e-05,
      "loss": 0.4646,
      "num_input_tokens_seen": 21618920,
      "step": 33150
    },
    {
      "epoch": 17.37683438155136,
      "grad_norm": 0.14708003401756287,
      "learning_rate": 5.151306841100756e-05,
      "loss": 0.3266,
      "num_input_tokens_seen": 21621960,
      "step": 33155
    },
    {
      "epoch": 17.37945492662474,
      "grad_norm": 0.219651997089386,
      "learning_rate": 5.141201713211679e-05,
      "loss": 0.4023,
      "num_input_tokens_seen": 21624584,
      "step": 33160
    },
    {
      "epoch": 17.382075471698112,
      "grad_norm": 0.1494217813014984,
      "learning_rate": 5.1311059692714324e-05,
      "loss": 0.3914,
      "num_input_tokens_seen": 21628200,
      "step": 33165
    },
    {
      "epoch": 17.38469601677149,
      "grad_norm": 0.10705491155385971,
      "learning_rate": 5.1210196113919306e-05,
      "loss": 0.3089,
      "num_input_tokens_seen": 21630792,
      "step": 33170
    },
    {
      "epoch": 17.387316561844862,
      "grad_norm": 0.14488032460212708,
      "learning_rate": 5.110942641683114e-05,
      "loss": 0.5198,
      "num_input_tokens_seen": 21633480,
      "step": 33175
    },
    {
      "epoch": 17.38993710691824,
      "grad_norm": 0.18362818658351898,
      "learning_rate": 5.10087506225298e-05,
      "loss": 0.4214,
      "num_input_tokens_seen": 21636584,
      "step": 33180
    },
    {
      "epoch": 17.392557651991613,
      "grad_norm": 0.1438598781824112,
      "learning_rate": 5.090816875207543e-05,
      "loss": 0.4284,
      "num_input_tokens_seen": 21639624,
      "step": 33185
    },
    {
      "epoch": 17.39517819706499,
      "grad_norm": 0.3069605529308319,
      "learning_rate": 5.0807680826508604e-05,
      "loss": 0.5439,
      "num_input_tokens_seen": 21642248,
      "step": 33190
    },
    {
      "epoch": 17.397798742138363,
      "grad_norm": 0.5026648044586182,
      "learning_rate": 5.070728686685039e-05,
      "loss": 0.5228,
      "num_input_tokens_seen": 21645576,
      "step": 33195
    },
    {
      "epoch": 17.40041928721174,
      "grad_norm": 0.2140810340642929,
      "learning_rate": 5.0606986894101925e-05,
      "loss": 0.5018,
      "num_input_tokens_seen": 21648712,
      "step": 33200
    },
    {
      "epoch": 17.403039832285117,
      "grad_norm": 0.17660674452781677,
      "learning_rate": 5.0506780929244875e-05,
      "loss": 0.4804,
      "num_input_tokens_seen": 21651624,
      "step": 33205
    },
    {
      "epoch": 17.40566037735849,
      "grad_norm": 0.1591964215040207,
      "learning_rate": 5.040666899324109e-05,
      "loss": 0.5293,
      "num_input_tokens_seen": 21655624,
      "step": 33210
    },
    {
      "epoch": 17.408280922431867,
      "grad_norm": 0.1612844467163086,
      "learning_rate": 5.0306651107032916e-05,
      "loss": 0.4831,
      "num_input_tokens_seen": 21658632,
      "step": 33215
    },
    {
      "epoch": 17.41090146750524,
      "grad_norm": 0.11944432556629181,
      "learning_rate": 5.020672729154308e-05,
      "loss": 0.3506,
      "num_input_tokens_seen": 21661736,
      "step": 33220
    },
    {
      "epoch": 17.413522012578618,
      "grad_norm": 0.13498631119728088,
      "learning_rate": 5.010689756767428e-05,
      "loss": 0.4854,
      "num_input_tokens_seen": 21664776,
      "step": 33225
    },
    {
      "epoch": 17.41614255765199,
      "grad_norm": 0.10373208671808243,
      "learning_rate": 5.000716195630983e-05,
      "loss": 0.4165,
      "num_input_tokens_seen": 21667752,
      "step": 33230
    },
    {
      "epoch": 17.418763102725368,
      "grad_norm": 0.0808224231004715,
      "learning_rate": 4.990752047831343e-05,
      "loss": 0.5431,
      "num_input_tokens_seen": 21672168,
      "step": 33235
    },
    {
      "epoch": 17.42138364779874,
      "grad_norm": 0.14667677879333496,
      "learning_rate": 4.980797315452873e-05,
      "loss": 0.4762,
      "num_input_tokens_seen": 21674472,
      "step": 33240
    },
    {
      "epoch": 17.42400419287212,
      "grad_norm": 0.20022638142108917,
      "learning_rate": 4.9708520005779965e-05,
      "loss": 0.6127,
      "num_input_tokens_seen": 21677608,
      "step": 33245
    },
    {
      "epoch": 17.42662473794549,
      "grad_norm": 0.16644947230815887,
      "learning_rate": 4.960916105287172e-05,
      "loss": 0.4198,
      "num_input_tokens_seen": 21680552,
      "step": 33250
    },
    {
      "epoch": 17.42924528301887,
      "grad_norm": 0.14524157345294952,
      "learning_rate": 4.950989631658859e-05,
      "loss": 0.4861,
      "num_input_tokens_seen": 21682696,
      "step": 33255
    },
    {
      "epoch": 17.431865828092242,
      "grad_norm": 0.1542394608259201,
      "learning_rate": 4.941072581769562e-05,
      "loss": 0.4762,
      "num_input_tokens_seen": 21686920,
      "step": 33260
    },
    {
      "epoch": 17.43448637316562,
      "grad_norm": 0.1786799430847168,
      "learning_rate": 4.931164957693834e-05,
      "loss": 0.3148,
      "num_input_tokens_seen": 21689416,
      "step": 33265
    },
    {
      "epoch": 17.437106918238992,
      "grad_norm": 0.10342435538768768,
      "learning_rate": 4.921266761504217e-05,
      "loss": 0.2847,
      "num_input_tokens_seen": 21693480,
      "step": 33270
    },
    {
      "epoch": 17.43972746331237,
      "grad_norm": 0.23202964663505554,
      "learning_rate": 4.9113779952713155e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 21696296,
      "step": 33275
    },
    {
      "epoch": 17.442348008385743,
      "grad_norm": 0.14813436567783356,
      "learning_rate": 4.90149866106373e-05,
      "loss": 0.3671,
      "num_input_tokens_seen": 21702088,
      "step": 33280
    },
    {
      "epoch": 17.44496855345912,
      "grad_norm": 0.1780284345149994,
      "learning_rate": 4.8916287609481137e-05,
      "loss": 0.4839,
      "num_input_tokens_seen": 21705160,
      "step": 33285
    },
    {
      "epoch": 17.447589098532493,
      "grad_norm": 0.12187013030052185,
      "learning_rate": 4.8817682969891366e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 21708328,
      "step": 33290
    },
    {
      "epoch": 17.45020964360587,
      "grad_norm": 0.10513915866613388,
      "learning_rate": 4.8719172712495095e-05,
      "loss": 0.5671,
      "num_input_tokens_seen": 21713768,
      "step": 33295
    },
    {
      "epoch": 17.452830188679247,
      "grad_norm": 0.33555230498313904,
      "learning_rate": 4.862075685789935e-05,
      "loss": 0.3934,
      "num_input_tokens_seen": 21717192,
      "step": 33300
    },
    {
      "epoch": 17.45545073375262,
      "grad_norm": 0.13690894842147827,
      "learning_rate": 4.8522435426691604e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 21721384,
      "step": 33305
    },
    {
      "epoch": 17.458071278825997,
      "grad_norm": 0.16012883186340332,
      "learning_rate": 4.842420843943962e-05,
      "loss": 0.4155,
      "num_input_tokens_seen": 21724392,
      "step": 33310
    },
    {
      "epoch": 17.46069182389937,
      "grad_norm": 0.2431144416332245,
      "learning_rate": 4.8326075916691415e-05,
      "loss": 0.5012,
      "num_input_tokens_seen": 21727208,
      "step": 33315
    },
    {
      "epoch": 17.463312368972748,
      "grad_norm": 0.15638430416584015,
      "learning_rate": 4.822803787897523e-05,
      "loss": 0.3648,
      "num_input_tokens_seen": 21729960,
      "step": 33320
    },
    {
      "epoch": 17.46593291404612,
      "grad_norm": 0.10580471158027649,
      "learning_rate": 4.813009434679932e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 21734216,
      "step": 33325
    },
    {
      "epoch": 17.468553459119498,
      "grad_norm": 0.28009000420570374,
      "learning_rate": 4.803224534065259e-05,
      "loss": 0.5188,
      "num_input_tokens_seen": 21738152,
      "step": 33330
    },
    {
      "epoch": 17.47117400419287,
      "grad_norm": 0.1363658308982849,
      "learning_rate": 4.793449088100371e-05,
      "loss": 0.483,
      "num_input_tokens_seen": 21741576,
      "step": 33335
    },
    {
      "epoch": 17.47379454926625,
      "grad_norm": 0.11666643619537354,
      "learning_rate": 4.783683098830188e-05,
      "loss": 0.328,
      "num_input_tokens_seen": 21745224,
      "step": 33340
    },
    {
      "epoch": 17.47641509433962,
      "grad_norm": 0.10181738436222076,
      "learning_rate": 4.7739265682976485e-05,
      "loss": 0.5175,
      "num_input_tokens_seen": 21748424,
      "step": 33345
    },
    {
      "epoch": 17.479035639413,
      "grad_norm": 0.23225459456443787,
      "learning_rate": 4.764179498543714e-05,
      "loss": 0.4269,
      "num_input_tokens_seen": 21751336,
      "step": 33350
    },
    {
      "epoch": 17.481656184486372,
      "grad_norm": 0.12775297462940216,
      "learning_rate": 4.754441891607347e-05,
      "loss": 0.3598,
      "num_input_tokens_seen": 21754568,
      "step": 33355
    },
    {
      "epoch": 17.48427672955975,
      "grad_norm": 0.18740713596343994,
      "learning_rate": 4.7447137495255424e-05,
      "loss": 0.3643,
      "num_input_tokens_seen": 21757832,
      "step": 33360
    },
    {
      "epoch": 17.486897274633122,
      "grad_norm": 0.1422543078660965,
      "learning_rate": 4.7349950743333206e-05,
      "loss": 0.3252,
      "num_input_tokens_seen": 21760968,
      "step": 33365
    },
    {
      "epoch": 17.4895178197065,
      "grad_norm": 0.12733355164527893,
      "learning_rate": 4.725285868063722e-05,
      "loss": 0.3677,
      "num_input_tokens_seen": 21765160,
      "step": 33370
    },
    {
      "epoch": 17.492138364779873,
      "grad_norm": 0.1791674941778183,
      "learning_rate": 4.715586132747801e-05,
      "loss": 0.5912,
      "num_input_tokens_seen": 21768584,
      "step": 33375
    },
    {
      "epoch": 17.49475890985325,
      "grad_norm": 0.16663475334644318,
      "learning_rate": 4.705895870414634e-05,
      "loss": 0.4237,
      "num_input_tokens_seen": 21771304,
      "step": 33380
    },
    {
      "epoch": 17.497379454926623,
      "grad_norm": 0.15371817350387573,
      "learning_rate": 4.696215083091299e-05,
      "loss": 0.508,
      "num_input_tokens_seen": 21774568,
      "step": 33385
    },
    {
      "epoch": 17.5,
      "grad_norm": 0.1550389975309372,
      "learning_rate": 4.686543772802915e-05,
      "loss": 0.5091,
      "num_input_tokens_seen": 21777608,
      "step": 33390
    },
    {
      "epoch": 17.502620545073377,
      "grad_norm": 0.17151056230068207,
      "learning_rate": 4.6768819415726085e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 21781288,
      "step": 33395
    },
    {
      "epoch": 17.50524109014675,
      "grad_norm": 0.16700859367847443,
      "learning_rate": 4.667229591421529e-05,
      "loss": 0.4878,
      "num_input_tokens_seen": 21784168,
      "step": 33400
    },
    {
      "epoch": 17.507861635220127,
      "grad_norm": 0.1812911480665207,
      "learning_rate": 4.657586724368845e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 21786760,
      "step": 33405
    },
    {
      "epoch": 17.5104821802935,
      "grad_norm": 0.1628728210926056,
      "learning_rate": 4.647953342431726e-05,
      "loss": 0.4018,
      "num_input_tokens_seen": 21789352,
      "step": 33410
    },
    {
      "epoch": 17.513102725366878,
      "grad_norm": 0.18665966391563416,
      "learning_rate": 4.63832944762535e-05,
      "loss": 0.4819,
      "num_input_tokens_seen": 21792296,
      "step": 33415
    },
    {
      "epoch": 17.51572327044025,
      "grad_norm": 0.15593834221363068,
      "learning_rate": 4.628715041962944e-05,
      "loss": 0.3734,
      "num_input_tokens_seen": 21794824,
      "step": 33420
    },
    {
      "epoch": 17.518343815513628,
      "grad_norm": 0.12675009667873383,
      "learning_rate": 4.619110127455728e-05,
      "loss": 0.4251,
      "num_input_tokens_seen": 21797928,
      "step": 33425
    },
    {
      "epoch": 17.520964360587,
      "grad_norm": 0.15085168182849884,
      "learning_rate": 4.6095147061129494e-05,
      "loss": 0.6305,
      "num_input_tokens_seen": 21801352,
      "step": 33430
    },
    {
      "epoch": 17.52358490566038,
      "grad_norm": 0.21530397236347198,
      "learning_rate": 4.599928779941853e-05,
      "loss": 0.4136,
      "num_input_tokens_seen": 21805768,
      "step": 33435
    },
    {
      "epoch": 17.52620545073375,
      "grad_norm": 0.1968393176794052,
      "learning_rate": 4.590352350947691e-05,
      "loss": 0.3404,
      "num_input_tokens_seen": 21808648,
      "step": 33440
    },
    {
      "epoch": 17.52882599580713,
      "grad_norm": 0.24227222800254822,
      "learning_rate": 4.580785421133759e-05,
      "loss": 0.4684,
      "num_input_tokens_seen": 21811080,
      "step": 33445
    },
    {
      "epoch": 17.531446540880502,
      "grad_norm": 0.1534217745065689,
      "learning_rate": 4.571227992501348e-05,
      "loss": 0.336,
      "num_input_tokens_seen": 21814568,
      "step": 33450
    },
    {
      "epoch": 17.53406708595388,
      "grad_norm": 0.19787031412124634,
      "learning_rate": 4.561680067049773e-05,
      "loss": 0.4212,
      "num_input_tokens_seen": 21817416,
      "step": 33455
    },
    {
      "epoch": 17.536687631027252,
      "grad_norm": 0.16617423295974731,
      "learning_rate": 4.552141646776331e-05,
      "loss": 0.4055,
      "num_input_tokens_seen": 21820648,
      "step": 33460
    },
    {
      "epoch": 17.53930817610063,
      "grad_norm": 0.23649226129055023,
      "learning_rate": 4.542612733676365e-05,
      "loss": 0.4766,
      "num_input_tokens_seen": 21823560,
      "step": 33465
    },
    {
      "epoch": 17.541928721174003,
      "grad_norm": 0.15266239643096924,
      "learning_rate": 4.533093329743204e-05,
      "loss": 0.5102,
      "num_input_tokens_seen": 21827208,
      "step": 33470
    },
    {
      "epoch": 17.54454926624738,
      "grad_norm": 0.23795604705810547,
      "learning_rate": 4.523583436968204e-05,
      "loss": 0.5166,
      "num_input_tokens_seen": 21830120,
      "step": 33475
    },
    {
      "epoch": 17.547169811320753,
      "grad_norm": 0.16930507123470306,
      "learning_rate": 4.514083057340723e-05,
      "loss": 0.43,
      "num_input_tokens_seen": 21833064,
      "step": 33480
    },
    {
      "epoch": 17.54979035639413,
      "grad_norm": 0.17518292367458344,
      "learning_rate": 4.504592192848151e-05,
      "loss": 0.4397,
      "num_input_tokens_seen": 21835752,
      "step": 33485
    },
    {
      "epoch": 17.552410901467507,
      "grad_norm": 0.15118171274662018,
      "learning_rate": 4.495110845475847e-05,
      "loss": 0.5432,
      "num_input_tokens_seen": 21838664,
      "step": 33490
    },
    {
      "epoch": 17.55503144654088,
      "grad_norm": 0.18226398527622223,
      "learning_rate": 4.4856390172072045e-05,
      "loss": 0.4565,
      "num_input_tokens_seen": 21842184,
      "step": 33495
    },
    {
      "epoch": 17.557651991614257,
      "grad_norm": 0.1987518072128296,
      "learning_rate": 4.476176710023622e-05,
      "loss": 0.4242,
      "num_input_tokens_seen": 21845224,
      "step": 33500
    },
    {
      "epoch": 17.56027253668763,
      "grad_norm": 0.1263689547777176,
      "learning_rate": 4.4667239259045035e-05,
      "loss": 0.6302,
      "num_input_tokens_seen": 21848552,
      "step": 33505
    },
    {
      "epoch": 17.562893081761008,
      "grad_norm": 0.20766302943229675,
      "learning_rate": 4.457280666827285e-05,
      "loss": 0.4343,
      "num_input_tokens_seen": 21850856,
      "step": 33510
    },
    {
      "epoch": 17.56551362683438,
      "grad_norm": 0.18377967178821564,
      "learning_rate": 4.447846934767358e-05,
      "loss": 0.3459,
      "num_input_tokens_seen": 21856104,
      "step": 33515
    },
    {
      "epoch": 17.568134171907758,
      "grad_norm": 0.16561049222946167,
      "learning_rate": 4.4384227316981774e-05,
      "loss": 0.3168,
      "num_input_tokens_seen": 21858440,
      "step": 33520
    },
    {
      "epoch": 17.57075471698113,
      "grad_norm": 0.12771975994110107,
      "learning_rate": 4.4290080595911555e-05,
      "loss": 0.7186,
      "num_input_tokens_seen": 21861064,
      "step": 33525
    },
    {
      "epoch": 17.57337526205451,
      "grad_norm": 0.2624497711658478,
      "learning_rate": 4.419602920415749e-05,
      "loss": 0.2812,
      "num_input_tokens_seen": 21863848,
      "step": 33530
    },
    {
      "epoch": 17.57599580712788,
      "grad_norm": 0.19000273942947388,
      "learning_rate": 4.410207316139408e-05,
      "loss": 0.4325,
      "num_input_tokens_seen": 21867208,
      "step": 33535
    },
    {
      "epoch": 17.57861635220126,
      "grad_norm": 0.16534212231636047,
      "learning_rate": 4.400821248727577e-05,
      "loss": 0.3927,
      "num_input_tokens_seen": 21871944,
      "step": 33540
    },
    {
      "epoch": 17.581236897274632,
      "grad_norm": 0.13553903996944427,
      "learning_rate": 4.391444720143717e-05,
      "loss": 0.2418,
      "num_input_tokens_seen": 21874728,
      "step": 33545
    },
    {
      "epoch": 17.58385744234801,
      "grad_norm": 0.2871432602405548,
      "learning_rate": 4.382077732349299e-05,
      "loss": 0.3814,
      "num_input_tokens_seen": 21877320,
      "step": 33550
    },
    {
      "epoch": 17.586477987421382,
      "grad_norm": 0.1995096057653427,
      "learning_rate": 4.372720287303772e-05,
      "loss": 0.3933,
      "num_input_tokens_seen": 21880488,
      "step": 33555
    },
    {
      "epoch": 17.58909853249476,
      "grad_norm": 0.10880474001169205,
      "learning_rate": 4.3633723869646204e-05,
      "loss": 0.4753,
      "num_input_tokens_seen": 21883080,
      "step": 33560
    },
    {
      "epoch": 17.591719077568133,
      "grad_norm": 0.16321271657943726,
      "learning_rate": 4.3540340332873184e-05,
      "loss": 0.5058,
      "num_input_tokens_seen": 21886536,
      "step": 33565
    },
    {
      "epoch": 17.59433962264151,
      "grad_norm": 0.10388030111789703,
      "learning_rate": 4.3447052282253375e-05,
      "loss": 0.4338,
      "num_input_tokens_seen": 21889896,
      "step": 33570
    },
    {
      "epoch": 17.596960167714883,
      "grad_norm": 0.11114795506000519,
      "learning_rate": 4.335385973730166e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 21893256,
      "step": 33575
    },
    {
      "epoch": 17.59958071278826,
      "grad_norm": 0.2090616226196289,
      "learning_rate": 4.3260762717512726e-05,
      "loss": 0.432,
      "num_input_tokens_seen": 21896328,
      "step": 33580
    },
    {
      "epoch": 17.602201257861637,
      "grad_norm": 0.2071024775505066,
      "learning_rate": 4.316776124236144e-05,
      "loss": 0.5007,
      "num_input_tokens_seen": 21898856,
      "step": 33585
    },
    {
      "epoch": 17.60482180293501,
      "grad_norm": 0.17988599836826324,
      "learning_rate": 4.307485533130279e-05,
      "loss": 0.4096,
      "num_input_tokens_seen": 21901992,
      "step": 33590
    },
    {
      "epoch": 17.607442348008387,
      "grad_norm": 0.1957206428050995,
      "learning_rate": 4.298204500377145e-05,
      "loss": 0.3407,
      "num_input_tokens_seen": 21904904,
      "step": 33595
    },
    {
      "epoch": 17.61006289308176,
      "grad_norm": 0.09787601977586746,
      "learning_rate": 4.288933027918235e-05,
      "loss": 0.3532,
      "num_input_tokens_seen": 21908808,
      "step": 33600
    },
    {
      "epoch": 17.612683438155138,
      "grad_norm": 0.14640679955482483,
      "learning_rate": 4.279671117693046e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 21911720,
      "step": 33605
    },
    {
      "epoch": 17.61530398322851,
      "grad_norm": 0.15277568995952606,
      "learning_rate": 4.270418771639045e-05,
      "loss": 0.3514,
      "num_input_tokens_seen": 21914024,
      "step": 33610
    },
    {
      "epoch": 17.617924528301888,
      "grad_norm": 0.12309790402650833,
      "learning_rate": 4.2611759916917345e-05,
      "loss": 0.3577,
      "num_input_tokens_seen": 21916648,
      "step": 33615
    },
    {
      "epoch": 17.62054507337526,
      "grad_norm": 0.08663200587034225,
      "learning_rate": 4.251942779784579e-05,
      "loss": 0.4317,
      "num_input_tokens_seen": 21920136,
      "step": 33620
    },
    {
      "epoch": 17.62316561844864,
      "grad_norm": 0.1554468721151352,
      "learning_rate": 4.242719137849077e-05,
      "loss": 0.3757,
      "num_input_tokens_seen": 21924168,
      "step": 33625
    },
    {
      "epoch": 17.62578616352201,
      "grad_norm": 0.2953473627567291,
      "learning_rate": 4.233505067814702e-05,
      "loss": 0.435,
      "num_input_tokens_seen": 21931656,
      "step": 33630
    },
    {
      "epoch": 17.62840670859539,
      "grad_norm": 0.15242040157318115,
      "learning_rate": 4.224300571608947e-05,
      "loss": 0.3862,
      "num_input_tokens_seen": 21934184,
      "step": 33635
    },
    {
      "epoch": 17.631027253668762,
      "grad_norm": 0.12131956964731216,
      "learning_rate": 4.215105651157264e-05,
      "loss": 0.4582,
      "num_input_tokens_seen": 21938856,
      "step": 33640
    },
    {
      "epoch": 17.63364779874214,
      "grad_norm": 0.07840970158576965,
      "learning_rate": 4.2059203083831434e-05,
      "loss": 0.3665,
      "num_input_tokens_seen": 21941736,
      "step": 33645
    },
    {
      "epoch": 17.636268343815512,
      "grad_norm": 0.15630032122135162,
      "learning_rate": 4.196744545208042e-05,
      "loss": 0.4745,
      "num_input_tokens_seen": 21945256,
      "step": 33650
    },
    {
      "epoch": 17.63888888888889,
      "grad_norm": 0.21306033432483673,
      "learning_rate": 4.187578363551431e-05,
      "loss": 0.4169,
      "num_input_tokens_seen": 21947432,
      "step": 33655
    },
    {
      "epoch": 17.641509433962263,
      "grad_norm": 0.18654093146324158,
      "learning_rate": 4.17842176533077e-05,
      "loss": 0.5223,
      "num_input_tokens_seen": 21950664,
      "step": 33660
    },
    {
      "epoch": 17.64412997903564,
      "grad_norm": 0.19870871305465698,
      "learning_rate": 4.1692747524615274e-05,
      "loss": 0.4849,
      "num_input_tokens_seen": 21953704,
      "step": 33665
    },
    {
      "epoch": 17.646750524109013,
      "grad_norm": 0.09708160161972046,
      "learning_rate": 4.160137326857138e-05,
      "loss": 0.3503,
      "num_input_tokens_seen": 21956936,
      "step": 33670
    },
    {
      "epoch": 17.64937106918239,
      "grad_norm": 0.140177920460701,
      "learning_rate": 4.151009490429042e-05,
      "loss": 0.4268,
      "num_input_tokens_seen": 21960200,
      "step": 33675
    },
    {
      "epoch": 17.651991614255767,
      "grad_norm": 0.12359487265348434,
      "learning_rate": 4.141891245086687e-05,
      "loss": 0.3981,
      "num_input_tokens_seen": 21962824,
      "step": 33680
    },
    {
      "epoch": 17.65461215932914,
      "grad_norm": 0.1436990350484848,
      "learning_rate": 4.132782592737511e-05,
      "loss": 0.4149,
      "num_input_tokens_seen": 21966504,
      "step": 33685
    },
    {
      "epoch": 17.657232704402517,
      "grad_norm": 0.14292171597480774,
      "learning_rate": 4.123683535286943e-05,
      "loss": 0.3953,
      "num_input_tokens_seen": 21970664,
      "step": 33690
    },
    {
      "epoch": 17.65985324947589,
      "grad_norm": 0.08264091610908508,
      "learning_rate": 4.114594074638395e-05,
      "loss": 0.3442,
      "num_input_tokens_seen": 21974312,
      "step": 33695
    },
    {
      "epoch": 17.662473794549268,
      "grad_norm": 0.1709204614162445,
      "learning_rate": 4.105514212693268e-05,
      "loss": 0.5069,
      "num_input_tokens_seen": 21977288,
      "step": 33700
    },
    {
      "epoch": 17.66509433962264,
      "grad_norm": 0.2582954168319702,
      "learning_rate": 4.096443951350976e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 21979720,
      "step": 33705
    },
    {
      "epoch": 17.667714884696018,
      "grad_norm": 0.10702043771743774,
      "learning_rate": 4.0873832925089116e-05,
      "loss": 0.4035,
      "num_input_tokens_seen": 21982600,
      "step": 33710
    },
    {
      "epoch": 17.67033542976939,
      "grad_norm": 0.12309285253286362,
      "learning_rate": 4.0783322380624656e-05,
      "loss": 0.4224,
      "num_input_tokens_seen": 21986184,
      "step": 33715
    },
    {
      "epoch": 17.67295597484277,
      "grad_norm": 0.14543794095516205,
      "learning_rate": 4.069290789905023e-05,
      "loss": 0.4869,
      "num_input_tokens_seen": 21989352,
      "step": 33720
    },
    {
      "epoch": 17.67557651991614,
      "grad_norm": 0.1356494277715683,
      "learning_rate": 4.060258949927942e-05,
      "loss": 0.4093,
      "num_input_tokens_seen": 21992072,
      "step": 33725
    },
    {
      "epoch": 17.67819706498952,
      "grad_norm": 0.17489586770534515,
      "learning_rate": 4.051236720020568e-05,
      "loss": 0.4825,
      "num_input_tokens_seen": 21996200,
      "step": 33730
    },
    {
      "epoch": 17.680817610062892,
      "grad_norm": 0.16480329632759094,
      "learning_rate": 4.042224102070263e-05,
      "loss": 0.4365,
      "num_input_tokens_seen": 21998728,
      "step": 33735
    },
    {
      "epoch": 17.68343815513627,
      "grad_norm": 0.1273222267627716,
      "learning_rate": 4.033221097962364e-05,
      "loss": 0.446,
      "num_input_tokens_seen": 22001480,
      "step": 33740
    },
    {
      "epoch": 17.686058700209642,
      "grad_norm": 0.14953888952732086,
      "learning_rate": 4.0242277095802025e-05,
      "loss": 0.5611,
      "num_input_tokens_seen": 22005800,
      "step": 33745
    },
    {
      "epoch": 17.68867924528302,
      "grad_norm": 0.2080511748790741,
      "learning_rate": 4.015243938805086e-05,
      "loss": 0.3337,
      "num_input_tokens_seen": 22009096,
      "step": 33750
    },
    {
      "epoch": 17.691299790356393,
      "grad_norm": 0.1340065449476242,
      "learning_rate": 4.006269787516304e-05,
      "loss": 0.3987,
      "num_input_tokens_seen": 22012072,
      "step": 33755
    },
    {
      "epoch": 17.69392033542977,
      "grad_norm": 0.1328224539756775,
      "learning_rate": 3.9973052575911606e-05,
      "loss": 0.5173,
      "num_input_tokens_seen": 22015304,
      "step": 33760
    },
    {
      "epoch": 17.696540880503143,
      "grad_norm": 0.0858159288764,
      "learning_rate": 3.988350350904929e-05,
      "loss": 0.3618,
      "num_input_tokens_seen": 22019016,
      "step": 33765
    },
    {
      "epoch": 17.69916142557652,
      "grad_norm": 0.11569040268659592,
      "learning_rate": 3.979405069330882e-05,
      "loss": 0.4613,
      "num_input_tokens_seen": 22021992,
      "step": 33770
    },
    {
      "epoch": 17.701781970649897,
      "grad_norm": 0.31439679861068726,
      "learning_rate": 3.970469414740269e-05,
      "loss": 0.3605,
      "num_input_tokens_seen": 22024808,
      "step": 33775
    },
    {
      "epoch": 17.70440251572327,
      "grad_norm": 0.17107927799224854,
      "learning_rate": 3.961543389002309e-05,
      "loss": 0.447,
      "num_input_tokens_seen": 22027336,
      "step": 33780
    },
    {
      "epoch": 17.707023060796647,
      "grad_norm": 0.21614156663417816,
      "learning_rate": 3.952626993984243e-05,
      "loss": 0.3993,
      "num_input_tokens_seen": 22030440,
      "step": 33785
    },
    {
      "epoch": 17.70964360587002,
      "grad_norm": 0.15985003113746643,
      "learning_rate": 3.943720231551268e-05,
      "loss": 0.5759,
      "num_input_tokens_seen": 22033160,
      "step": 33790
    },
    {
      "epoch": 17.712264150943398,
      "grad_norm": 0.11064603179693222,
      "learning_rate": 3.934823103566587e-05,
      "loss": 0.264,
      "num_input_tokens_seen": 22035720,
      "step": 33795
    },
    {
      "epoch": 17.71488469601677,
      "grad_norm": 0.22966845333576202,
      "learning_rate": 3.925935611891379e-05,
      "loss": 0.5306,
      "num_input_tokens_seen": 22038664,
      "step": 33800
    },
    {
      "epoch": 17.717505241090148,
      "grad_norm": 0.13125106692314148,
      "learning_rate": 3.917057758384801e-05,
      "loss": 0.3585,
      "num_input_tokens_seen": 22041512,
      "step": 33805
    },
    {
      "epoch": 17.72012578616352,
      "grad_norm": 0.119649738073349,
      "learning_rate": 3.9081895449039896e-05,
      "loss": 0.4144,
      "num_input_tokens_seen": 22045320,
      "step": 33810
    },
    {
      "epoch": 17.7227463312369,
      "grad_norm": 0.2063623070716858,
      "learning_rate": 3.899330973304083e-05,
      "loss": 0.318,
      "num_input_tokens_seen": 22048232,
      "step": 33815
    },
    {
      "epoch": 17.72536687631027,
      "grad_norm": 0.13310997188091278,
      "learning_rate": 3.8904820454381875e-05,
      "loss": 0.3757,
      "num_input_tokens_seen": 22051528,
      "step": 33820
    },
    {
      "epoch": 17.72798742138365,
      "grad_norm": 0.1702127903699875,
      "learning_rate": 3.881642763157417e-05,
      "loss": 0.4555,
      "num_input_tokens_seen": 22054248,
      "step": 33825
    },
    {
      "epoch": 17.730607966457022,
      "grad_norm": 0.14864879846572876,
      "learning_rate": 3.872813128310826e-05,
      "loss": 0.3298,
      "num_input_tokens_seen": 22058600,
      "step": 33830
    },
    {
      "epoch": 17.7332285115304,
      "grad_norm": 0.1681671142578125,
      "learning_rate": 3.8639931427454866e-05,
      "loss": 0.3919,
      "num_input_tokens_seen": 22061896,
      "step": 33835
    },
    {
      "epoch": 17.735849056603772,
      "grad_norm": 0.13472893834114075,
      "learning_rate": 3.855182808306429e-05,
      "loss": 0.5423,
      "num_input_tokens_seen": 22065192,
      "step": 33840
    },
    {
      "epoch": 17.73846960167715,
      "grad_norm": 0.1422896385192871,
      "learning_rate": 3.8463821268366784e-05,
      "loss": 0.4837,
      "num_input_tokens_seen": 22068296,
      "step": 33845
    },
    {
      "epoch": 17.741090146750523,
      "grad_norm": 0.11315283924341202,
      "learning_rate": 3.837591100177246e-05,
      "loss": 0.4747,
      "num_input_tokens_seen": 22071976,
      "step": 33850
    },
    {
      "epoch": 17.7437106918239,
      "grad_norm": 0.15829837322235107,
      "learning_rate": 3.828809730167099e-05,
      "loss": 0.5166,
      "num_input_tokens_seen": 22075656,
      "step": 33855
    },
    {
      "epoch": 17.746331236897273,
      "grad_norm": 0.1288633495569229,
      "learning_rate": 3.820038018643218e-05,
      "loss": 0.374,
      "num_input_tokens_seen": 22079816,
      "step": 33860
    },
    {
      "epoch": 17.74895178197065,
      "grad_norm": 0.212275892496109,
      "learning_rate": 3.8112759674405254e-05,
      "loss": 0.4275,
      "num_input_tokens_seen": 22082472,
      "step": 33865
    },
    {
      "epoch": 17.751572327044027,
      "grad_norm": 0.12439291924238205,
      "learning_rate": 3.802523578391953e-05,
      "loss": 0.432,
      "num_input_tokens_seen": 22086024,
      "step": 33870
    },
    {
      "epoch": 17.7541928721174,
      "grad_norm": 0.14853499829769135,
      "learning_rate": 3.793780853328399e-05,
      "loss": 0.4463,
      "num_input_tokens_seen": 22090024,
      "step": 33875
    },
    {
      "epoch": 17.756813417190777,
      "grad_norm": 0.16585634648799896,
      "learning_rate": 3.785047794078744e-05,
      "loss": 0.3431,
      "num_input_tokens_seen": 22092680,
      "step": 33880
    },
    {
      "epoch": 17.75943396226415,
      "grad_norm": 0.36588573455810547,
      "learning_rate": 3.776324402469844e-05,
      "loss": 0.4919,
      "num_input_tokens_seen": 22095112,
      "step": 33885
    },
    {
      "epoch": 17.762054507337528,
      "grad_norm": 0.17210809886455536,
      "learning_rate": 3.767610680326533e-05,
      "loss": 0.4714,
      "num_input_tokens_seen": 22098696,
      "step": 33890
    },
    {
      "epoch": 17.7646750524109,
      "grad_norm": 0.20711708068847656,
      "learning_rate": 3.7589066294716145e-05,
      "loss": 0.4554,
      "num_input_tokens_seen": 22102024,
      "step": 33895
    },
    {
      "epoch": 17.767295597484278,
      "grad_norm": 0.15260641276836395,
      "learning_rate": 3.750212251725882e-05,
      "loss": 0.4158,
      "num_input_tokens_seen": 22105064,
      "step": 33900
    },
    {
      "epoch": 17.76991614255765,
      "grad_norm": 0.127075657248497,
      "learning_rate": 3.741527548908113e-05,
      "loss": 0.549,
      "num_input_tokens_seen": 22108968,
      "step": 33905
    },
    {
      "epoch": 17.77253668763103,
      "grad_norm": 0.11703795939683914,
      "learning_rate": 3.732852522835023e-05,
      "loss": 0.3701,
      "num_input_tokens_seen": 22113288,
      "step": 33910
    },
    {
      "epoch": 17.7751572327044,
      "grad_norm": 0.10425010323524475,
      "learning_rate": 3.724187175321342e-05,
      "loss": 0.4783,
      "num_input_tokens_seen": 22117768,
      "step": 33915
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 0.1768600344657898,
      "learning_rate": 3.71553150817977e-05,
      "loss": 0.3971,
      "num_input_tokens_seen": 22120936,
      "step": 33920
    },
    {
      "epoch": 17.780398322851152,
      "grad_norm": 0.10680646449327469,
      "learning_rate": 3.70688552322096e-05,
      "loss": 0.3334,
      "num_input_tokens_seen": 22124616,
      "step": 33925
    },
    {
      "epoch": 17.78301886792453,
      "grad_norm": 0.31549376249313354,
      "learning_rate": 3.698249222253569e-05,
      "loss": 0.3985,
      "num_input_tokens_seen": 22128072,
      "step": 33930
    },
    {
      "epoch": 17.785639412997902,
      "grad_norm": 0.2209773063659668,
      "learning_rate": 3.6896226070841863e-05,
      "loss": 0.3996,
      "num_input_tokens_seen": 22131272,
      "step": 33935
    },
    {
      "epoch": 17.78825995807128,
      "grad_norm": 0.12325587123632431,
      "learning_rate": 3.681005679517424e-05,
      "loss": 0.3389,
      "num_input_tokens_seen": 22134888,
      "step": 33940
    },
    {
      "epoch": 17.790880503144653,
      "grad_norm": 0.18095846474170685,
      "learning_rate": 3.6723984413558455e-05,
      "loss": 0.3487,
      "num_input_tokens_seen": 22137480,
      "step": 33945
    },
    {
      "epoch": 17.79350104821803,
      "grad_norm": 0.26339828968048096,
      "learning_rate": 3.663800894399971e-05,
      "loss": 0.4273,
      "num_input_tokens_seen": 22141000,
      "step": 33950
    },
    {
      "epoch": 17.796121593291403,
      "grad_norm": 0.1249522939324379,
      "learning_rate": 3.655213040448318e-05,
      "loss": 0.3113,
      "num_input_tokens_seen": 22143752,
      "step": 33955
    },
    {
      "epoch": 17.79874213836478,
      "grad_norm": 0.12665405869483948,
      "learning_rate": 3.6466348812973723e-05,
      "loss": 0.4087,
      "num_input_tokens_seen": 22147272,
      "step": 33960
    },
    {
      "epoch": 17.801362683438157,
      "grad_norm": 0.11527835577726364,
      "learning_rate": 3.6380664187415815e-05,
      "loss": 0.3059,
      "num_input_tokens_seen": 22151080,
      "step": 33965
    },
    {
      "epoch": 17.80398322851153,
      "grad_norm": 0.501395046710968,
      "learning_rate": 3.6295076545733663e-05,
      "loss": 0.3676,
      "num_input_tokens_seen": 22154664,
      "step": 33970
    },
    {
      "epoch": 17.806603773584907,
      "grad_norm": 0.3029007315635681,
      "learning_rate": 3.6209585905831344e-05,
      "loss": 0.383,
      "num_input_tokens_seen": 22157608,
      "step": 33975
    },
    {
      "epoch": 17.80922431865828,
      "grad_norm": 0.2518995702266693,
      "learning_rate": 3.6124192285592385e-05,
      "loss": 0.4763,
      "num_input_tokens_seen": 22160136,
      "step": 33980
    },
    {
      "epoch": 17.811844863731658,
      "grad_norm": 0.1948797106742859,
      "learning_rate": 3.6038895702880324e-05,
      "loss": 0.3736,
      "num_input_tokens_seen": 22162376,
      "step": 33985
    },
    {
      "epoch": 17.81446540880503,
      "grad_norm": 0.1837143748998642,
      "learning_rate": 3.595369617553806e-05,
      "loss": 0.6033,
      "num_input_tokens_seen": 22166376,
      "step": 33990
    },
    {
      "epoch": 17.817085953878408,
      "grad_norm": 0.185873880982399,
      "learning_rate": 3.58685937213884e-05,
      "loss": 0.367,
      "num_input_tokens_seen": 22169128,
      "step": 33995
    },
    {
      "epoch": 17.81970649895178,
      "grad_norm": 0.14832808077335358,
      "learning_rate": 3.5783588358233863e-05,
      "loss": 0.476,
      "num_input_tokens_seen": 22173544,
      "step": 34000
    },
    {
      "epoch": 17.822327044025158,
      "grad_norm": 0.11154938489198685,
      "learning_rate": 3.569868010385663e-05,
      "loss": 0.4014,
      "num_input_tokens_seen": 22176712,
      "step": 34005
    },
    {
      "epoch": 17.82494758909853,
      "grad_norm": 0.18277160823345184,
      "learning_rate": 3.5613868976018535e-05,
      "loss": 0.3747,
      "num_input_tokens_seen": 22179816,
      "step": 34010
    },
    {
      "epoch": 17.82756813417191,
      "grad_norm": 0.23718123137950897,
      "learning_rate": 3.5529154992460955e-05,
      "loss": 0.3955,
      "num_input_tokens_seen": 22182888,
      "step": 34015
    },
    {
      "epoch": 17.830188679245282,
      "grad_norm": 0.11532042175531387,
      "learning_rate": 3.544453817090521e-05,
      "loss": 0.4791,
      "num_input_tokens_seen": 22186472,
      "step": 34020
    },
    {
      "epoch": 17.83280922431866,
      "grad_norm": 0.14772240817546844,
      "learning_rate": 3.536001852905207e-05,
      "loss": 0.3967,
      "num_input_tokens_seen": 22189320,
      "step": 34025
    },
    {
      "epoch": 17.835429769392032,
      "grad_norm": 0.15394826233386993,
      "learning_rate": 3.527559608458225e-05,
      "loss": 0.4389,
      "num_input_tokens_seen": 22192200,
      "step": 34030
    },
    {
      "epoch": 17.83805031446541,
      "grad_norm": 0.2022319883108139,
      "learning_rate": 3.5191270855155886e-05,
      "loss": 0.3336,
      "num_input_tokens_seen": 22194696,
      "step": 34035
    },
    {
      "epoch": 17.840670859538783,
      "grad_norm": 0.1490752249956131,
      "learning_rate": 3.510704285841293e-05,
      "loss": 0.3677,
      "num_input_tokens_seen": 22197192,
      "step": 34040
    },
    {
      "epoch": 17.84329140461216,
      "grad_norm": 0.16454456746578217,
      "learning_rate": 3.502291211197267e-05,
      "loss": 0.3994,
      "num_input_tokens_seen": 22199976,
      "step": 34045
    },
    {
      "epoch": 17.845911949685533,
      "grad_norm": 0.19568684697151184,
      "learning_rate": 3.493887863343448e-05,
      "loss": 0.4316,
      "num_input_tokens_seen": 22203176,
      "step": 34050
    },
    {
      "epoch": 17.84853249475891,
      "grad_norm": 0.14567452669143677,
      "learning_rate": 3.485494244037718e-05,
      "loss": 0.3028,
      "num_input_tokens_seen": 22207368,
      "step": 34055
    },
    {
      "epoch": 17.851153039832283,
      "grad_norm": 0.19128061830997467,
      "learning_rate": 3.477110355035934e-05,
      "loss": 0.4525,
      "num_input_tokens_seen": 22210376,
      "step": 34060
    },
    {
      "epoch": 17.85377358490566,
      "grad_norm": 0.08718050271272659,
      "learning_rate": 3.468736198091904e-05,
      "loss": 0.3839,
      "num_input_tokens_seen": 22214856,
      "step": 34065
    },
    {
      "epoch": 17.856394129979037,
      "grad_norm": 0.11549176275730133,
      "learning_rate": 3.4603717749573935e-05,
      "loss": 0.367,
      "num_input_tokens_seen": 22218504,
      "step": 34070
    },
    {
      "epoch": 17.85901467505241,
      "grad_norm": 0.14431162178516388,
      "learning_rate": 3.452017087382159e-05,
      "loss": 0.399,
      "num_input_tokens_seen": 22221384,
      "step": 34075
    },
    {
      "epoch": 17.861635220125788,
      "grad_norm": 0.1673527956008911,
      "learning_rate": 3.443672137113901e-05,
      "loss": 0.4946,
      "num_input_tokens_seen": 22225480,
      "step": 34080
    },
    {
      "epoch": 17.86425576519916,
      "grad_norm": 0.1913977712392807,
      "learning_rate": 3.4353369258983025e-05,
      "loss": 0.4887,
      "num_input_tokens_seen": 22228488,
      "step": 34085
    },
    {
      "epoch": 17.866876310272538,
      "grad_norm": 0.11475817114114761,
      "learning_rate": 3.4270114554789735e-05,
      "loss": 0.4372,
      "num_input_tokens_seen": 22231912,
      "step": 34090
    },
    {
      "epoch": 17.86949685534591,
      "grad_norm": 0.1891220360994339,
      "learning_rate": 3.41869572759751e-05,
      "loss": 0.4489,
      "num_input_tokens_seen": 22235144,
      "step": 34095
    },
    {
      "epoch": 17.872117400419288,
      "grad_norm": 0.20191951096057892,
      "learning_rate": 3.410389743993475e-05,
      "loss": 0.5227,
      "num_input_tokens_seen": 22238088,
      "step": 34100
    },
    {
      "epoch": 17.87473794549266,
      "grad_norm": 0.11132737994194031,
      "learning_rate": 3.40209350640438e-05,
      "loss": 0.5143,
      "num_input_tokens_seen": 22241800,
      "step": 34105
    },
    {
      "epoch": 17.87735849056604,
      "grad_norm": 0.13166949152946472,
      "learning_rate": 3.3938070165657084e-05,
      "loss": 0.4693,
      "num_input_tokens_seen": 22245512,
      "step": 34110
    },
    {
      "epoch": 17.879979035639412,
      "grad_norm": 0.12465142458677292,
      "learning_rate": 3.385530276210902e-05,
      "loss": 0.478,
      "num_input_tokens_seen": 22248328,
      "step": 34115
    },
    {
      "epoch": 17.88259958071279,
      "grad_norm": 0.16300764679908752,
      "learning_rate": 3.377263287071358e-05,
      "loss": 0.5029,
      "num_input_tokens_seen": 22252744,
      "step": 34120
    },
    {
      "epoch": 17.885220125786162,
      "grad_norm": 0.15260373055934906,
      "learning_rate": 3.369006050876428e-05,
      "loss": 0.3428,
      "num_input_tokens_seen": 22257544,
      "step": 34125
    },
    {
      "epoch": 17.88784067085954,
      "grad_norm": 0.23381036520004272,
      "learning_rate": 3.3607585693534394e-05,
      "loss": 0.4121,
      "num_input_tokens_seen": 22260456,
      "step": 34130
    },
    {
      "epoch": 17.890461215932913,
      "grad_norm": 0.16583652794361115,
      "learning_rate": 3.352520844227663e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 22264328,
      "step": 34135
    },
    {
      "epoch": 17.89308176100629,
      "grad_norm": 0.16065376996994019,
      "learning_rate": 3.344292877222355e-05,
      "loss": 0.3397,
      "num_input_tokens_seen": 22267080,
      "step": 34140
    },
    {
      "epoch": 17.895702306079663,
      "grad_norm": 0.10058461874723434,
      "learning_rate": 3.336074670058703e-05,
      "loss": 0.3141,
      "num_input_tokens_seen": 22270248,
      "step": 34145
    },
    {
      "epoch": 17.89832285115304,
      "grad_norm": 0.1045636236667633,
      "learning_rate": 3.327866224455844e-05,
      "loss": 0.4327,
      "num_input_tokens_seen": 22273960,
      "step": 34150
    },
    {
      "epoch": 17.900943396226417,
      "grad_norm": 0.29398539662361145,
      "learning_rate": 3.3196675421309165e-05,
      "loss": 0.4219,
      "num_input_tokens_seen": 22277256,
      "step": 34155
    },
    {
      "epoch": 17.90356394129979,
      "grad_norm": 0.11590835452079773,
      "learning_rate": 3.3114786247989744e-05,
      "loss": 0.4711,
      "num_input_tokens_seen": 22280424,
      "step": 34160
    },
    {
      "epoch": 17.906184486373167,
      "grad_norm": 0.14755958318710327,
      "learning_rate": 3.303299474173066e-05,
      "loss": 0.3693,
      "num_input_tokens_seen": 22284808,
      "step": 34165
    },
    {
      "epoch": 17.90880503144654,
      "grad_norm": 0.15704549849033356,
      "learning_rate": 3.2951300919641544e-05,
      "loss": 0.3745,
      "num_input_tokens_seen": 22287912,
      "step": 34170
    },
    {
      "epoch": 17.911425576519918,
      "grad_norm": 0.11532143503427505,
      "learning_rate": 3.2869704798811964e-05,
      "loss": 0.4772,
      "num_input_tokens_seen": 22291112,
      "step": 34175
    },
    {
      "epoch": 17.91404612159329,
      "grad_norm": 0.13158079981803894,
      "learning_rate": 3.2788206396310746e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 22294280,
      "step": 34180
    },
    {
      "epoch": 17.916666666666668,
      "grad_norm": 0.10596740990877151,
      "learning_rate": 3.270680572918655e-05,
      "loss": 0.3222,
      "num_input_tokens_seen": 22297640,
      "step": 34185
    },
    {
      "epoch": 17.91928721174004,
      "grad_norm": 0.15706273913383484,
      "learning_rate": 3.262550281446741e-05,
      "loss": 0.3496,
      "num_input_tokens_seen": 22300712,
      "step": 34190
    },
    {
      "epoch": 17.921907756813418,
      "grad_norm": 0.08543519675731659,
      "learning_rate": 3.2544297669161115e-05,
      "loss": 0.3601,
      "num_input_tokens_seen": 22304072,
      "step": 34195
    },
    {
      "epoch": 17.92452830188679,
      "grad_norm": 0.22694525122642517,
      "learning_rate": 3.2463190310254633e-05,
      "loss": 0.5652,
      "num_input_tokens_seen": 22307336,
      "step": 34200
    },
    {
      "epoch": 17.92714884696017,
      "grad_norm": 0.15277411043643951,
      "learning_rate": 3.238218075471483e-05,
      "loss": 0.3937,
      "num_input_tokens_seen": 22310216,
      "step": 34205
    },
    {
      "epoch": 17.929769392033542,
      "grad_norm": 0.1395018845796585,
      "learning_rate": 3.230126901948788e-05,
      "loss": 0.3836,
      "num_input_tokens_seen": 22314216,
      "step": 34210
    },
    {
      "epoch": 17.93238993710692,
      "grad_norm": 0.20977240800857544,
      "learning_rate": 3.222045512149968e-05,
      "loss": 0.3365,
      "num_input_tokens_seen": 22317416,
      "step": 34215
    },
    {
      "epoch": 17.935010482180292,
      "grad_norm": 0.23092345893383026,
      "learning_rate": 3.213973907765566e-05,
      "loss": 0.3958,
      "num_input_tokens_seen": 22320456,
      "step": 34220
    },
    {
      "epoch": 17.93763102725367,
      "grad_norm": 0.14649470150470734,
      "learning_rate": 3.205912090484048e-05,
      "loss": 0.3224,
      "num_input_tokens_seen": 22323048,
      "step": 34225
    },
    {
      "epoch": 17.940251572327043,
      "grad_norm": 0.1508500576019287,
      "learning_rate": 3.1978600619918754e-05,
      "loss": 0.3407,
      "num_input_tokens_seen": 22326568,
      "step": 34230
    },
    {
      "epoch": 17.94287211740042,
      "grad_norm": 0.12128621339797974,
      "learning_rate": 3.189817823973423e-05,
      "loss": 0.444,
      "num_input_tokens_seen": 22329864,
      "step": 34235
    },
    {
      "epoch": 17.945492662473793,
      "grad_norm": 0.21074913442134857,
      "learning_rate": 3.18178537811104e-05,
      "loss": 0.6116,
      "num_input_tokens_seen": 22333000,
      "step": 34240
    },
    {
      "epoch": 17.94811320754717,
      "grad_norm": 0.18299512565135956,
      "learning_rate": 3.1737627260850335e-05,
      "loss": 0.3624,
      "num_input_tokens_seen": 22336328,
      "step": 34245
    },
    {
      "epoch": 17.950733752620543,
      "grad_norm": 0.1754014939069748,
      "learning_rate": 3.1657498695736364e-05,
      "loss": 0.3301,
      "num_input_tokens_seen": 22338920,
      "step": 34250
    },
    {
      "epoch": 17.95335429769392,
      "grad_norm": 0.21287506818771362,
      "learning_rate": 3.157746810253054e-05,
      "loss": 0.6485,
      "num_input_tokens_seen": 22342568,
      "step": 34255
    },
    {
      "epoch": 17.955974842767297,
      "grad_norm": 0.1933109164237976,
      "learning_rate": 3.149753549797446e-05,
      "loss": 0.4342,
      "num_input_tokens_seen": 22345512,
      "step": 34260
    },
    {
      "epoch": 17.95859538784067,
      "grad_norm": 0.4112900495529175,
      "learning_rate": 3.141770089878887e-05,
      "loss": 0.4907,
      "num_input_tokens_seen": 22348488,
      "step": 34265
    },
    {
      "epoch": 17.961215932914047,
      "grad_norm": 0.1300780326128006,
      "learning_rate": 3.1337964321674354e-05,
      "loss": 0.3645,
      "num_input_tokens_seen": 22351848,
      "step": 34270
    },
    {
      "epoch": 17.96383647798742,
      "grad_norm": 0.44122010469436646,
      "learning_rate": 3.125832578331106e-05,
      "loss": 0.3634,
      "num_input_tokens_seen": 22354856,
      "step": 34275
    },
    {
      "epoch": 17.966457023060798,
      "grad_norm": 0.09106679260730743,
      "learning_rate": 3.117878530035823e-05,
      "loss": 0.2964,
      "num_input_tokens_seen": 22358152,
      "step": 34280
    },
    {
      "epoch": 17.96907756813417,
      "grad_norm": 0.15022294223308563,
      "learning_rate": 3.1099342889454875e-05,
      "loss": 0.4622,
      "num_input_tokens_seen": 22361704,
      "step": 34285
    },
    {
      "epoch": 17.971698113207548,
      "grad_norm": 0.11534631997346878,
      "learning_rate": 3.1019998567219597e-05,
      "loss": 0.5186,
      "num_input_tokens_seen": 22364680,
      "step": 34290
    },
    {
      "epoch": 17.97431865828092,
      "grad_norm": 0.22862833738327026,
      "learning_rate": 3.0940752350250125e-05,
      "loss": 0.3527,
      "num_input_tokens_seen": 22367752,
      "step": 34295
    },
    {
      "epoch": 17.9769392033543,
      "grad_norm": 0.17947135865688324,
      "learning_rate": 3.0861604255124074e-05,
      "loss": 0.406,
      "num_input_tokens_seen": 22370632,
      "step": 34300
    },
    {
      "epoch": 17.979559748427672,
      "grad_norm": 0.13353286683559418,
      "learning_rate": 3.0782554298398045e-05,
      "loss": 0.3948,
      "num_input_tokens_seen": 22373224,
      "step": 34305
    },
    {
      "epoch": 17.98218029350105,
      "grad_norm": 0.15335504710674286,
      "learning_rate": 3.070360249660859e-05,
      "loss": 0.4143,
      "num_input_tokens_seen": 22379816,
      "step": 34310
    },
    {
      "epoch": 17.984800838574422,
      "grad_norm": 0.1426427811384201,
      "learning_rate": 3.0624748866271535e-05,
      "loss": 0.3397,
      "num_input_tokens_seen": 22383112,
      "step": 34315
    },
    {
      "epoch": 17.9874213836478,
      "grad_norm": 0.20095856487751007,
      "learning_rate": 3.0545993423882036e-05,
      "loss": 0.5399,
      "num_input_tokens_seen": 22386472,
      "step": 34320
    },
    {
      "epoch": 17.990041928721173,
      "grad_norm": 0.15760233998298645,
      "learning_rate": 3.046733618591496e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 22388968,
      "step": 34325
    },
    {
      "epoch": 17.99266247379455,
      "grad_norm": 0.2052348554134369,
      "learning_rate": 3.0388777168824434e-05,
      "loss": 0.3535,
      "num_input_tokens_seen": 22392008,
      "step": 34330
    },
    {
      "epoch": 17.995283018867923,
      "grad_norm": 0.23329801857471466,
      "learning_rate": 3.0310316389044022e-05,
      "loss": 0.3516,
      "num_input_tokens_seen": 22394984,
      "step": 34335
    },
    {
      "epoch": 17.9979035639413,
      "grad_norm": 0.14605844020843506,
      "learning_rate": 3.0231953862986995e-05,
      "loss": 0.4614,
      "num_input_tokens_seen": 22398376,
      "step": 34340
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.46658554673194885,
      "eval_runtime": 13.6084,
      "eval_samples_per_second": 62.315,
      "eval_steps_per_second": 15.579,
      "num_input_tokens_seen": 22400368,
      "step": 34344
    },
    {
      "epoch": 18.000524109014677,
      "grad_norm": 0.23443610966205597,
      "learning_rate": 3.0153689607045842e-05,
      "loss": 0.6094,
      "num_input_tokens_seen": 22400816,
      "step": 34345
    },
    {
      "epoch": 18.00314465408805,
      "grad_norm": 0.0818527564406395,
      "learning_rate": 3.0075523637592474e-05,
      "loss": 0.4229,
      "num_input_tokens_seen": 22403888,
      "step": 34350
    },
    {
      "epoch": 18.005765199161427,
      "grad_norm": 0.18983164429664612,
      "learning_rate": 2.999745597097847e-05,
      "loss": 0.4744,
      "num_input_tokens_seen": 22408208,
      "step": 34355
    },
    {
      "epoch": 18.0083857442348,
      "grad_norm": 0.1254458874464035,
      "learning_rate": 2.9919486623534497e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 22411856,
      "step": 34360
    },
    {
      "epoch": 18.011006289308177,
      "grad_norm": 0.11917602270841599,
      "learning_rate": 2.9841615611571005e-05,
      "loss": 0.3884,
      "num_input_tokens_seen": 22415248,
      "step": 34365
    },
    {
      "epoch": 18.01362683438155,
      "grad_norm": 0.10835645347833633,
      "learning_rate": 2.9763842951377628e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 22418416,
      "step": 34370
    },
    {
      "epoch": 18.016247379454928,
      "grad_norm": 0.1567777842283249,
      "learning_rate": 2.968616865922369e-05,
      "loss": 0.403,
      "num_input_tokens_seen": 22421456,
      "step": 34375
    },
    {
      "epoch": 18.0188679245283,
      "grad_norm": 0.1637333780527115,
      "learning_rate": 2.960859275135758e-05,
      "loss": 0.2892,
      "num_input_tokens_seen": 22424464,
      "step": 34380
    },
    {
      "epoch": 18.021488469601678,
      "grad_norm": 0.12197622656822205,
      "learning_rate": 2.9531115244007367e-05,
      "loss": 0.3506,
      "num_input_tokens_seen": 22427152,
      "step": 34385
    },
    {
      "epoch": 18.02410901467505,
      "grad_norm": 0.11516262590885162,
      "learning_rate": 2.945373615338037e-05,
      "loss": 0.3666,
      "num_input_tokens_seen": 22431696,
      "step": 34390
    },
    {
      "epoch": 18.02672955974843,
      "grad_norm": 0.1864306479692459,
      "learning_rate": 2.9376455495663534e-05,
      "loss": 0.5684,
      "num_input_tokens_seen": 22434704,
      "step": 34395
    },
    {
      "epoch": 18.029350104821802,
      "grad_norm": 0.22194121778011322,
      "learning_rate": 2.9299273287023144e-05,
      "loss": 0.4663,
      "num_input_tokens_seen": 22437776,
      "step": 34400
    },
    {
      "epoch": 18.03197064989518,
      "grad_norm": 0.13178056478500366,
      "learning_rate": 2.922218954360473e-05,
      "loss": 0.3512,
      "num_input_tokens_seen": 22440912,
      "step": 34405
    },
    {
      "epoch": 18.034591194968552,
      "grad_norm": 0.1829199641942978,
      "learning_rate": 2.9145204281533278e-05,
      "loss": 0.4478,
      "num_input_tokens_seen": 22444112,
      "step": 34410
    },
    {
      "epoch": 18.03721174004193,
      "grad_norm": 0.14135222136974335,
      "learning_rate": 2.9068317516913244e-05,
      "loss": 0.4255,
      "num_input_tokens_seen": 22448080,
      "step": 34415
    },
    {
      "epoch": 18.039832285115303,
      "grad_norm": 0.6254279017448425,
      "learning_rate": 2.8991529265828542e-05,
      "loss": 0.412,
      "num_input_tokens_seen": 22450800,
      "step": 34420
    },
    {
      "epoch": 18.04245283018868,
      "grad_norm": 0.17113783955574036,
      "learning_rate": 2.891483954434243e-05,
      "loss": 0.484,
      "num_input_tokens_seen": 22453648,
      "step": 34425
    },
    {
      "epoch": 18.045073375262053,
      "grad_norm": 0.18951278924942017,
      "learning_rate": 2.8838248368497465e-05,
      "loss": 0.3794,
      "num_input_tokens_seen": 22458480,
      "step": 34430
    },
    {
      "epoch": 18.04769392033543,
      "grad_norm": 0.08040226995944977,
      "learning_rate": 2.8761755754315667e-05,
      "loss": 0.4342,
      "num_input_tokens_seen": 22461840,
      "step": 34435
    },
    {
      "epoch": 18.050314465408807,
      "grad_norm": 0.1418941169977188,
      "learning_rate": 2.8685361717798286e-05,
      "loss": 0.2968,
      "num_input_tokens_seen": 22465104,
      "step": 34440
    },
    {
      "epoch": 18.05293501048218,
      "grad_norm": 0.1356164962053299,
      "learning_rate": 2.8609066274926265e-05,
      "loss": 0.4147,
      "num_input_tokens_seen": 22468976,
      "step": 34445
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 0.15544140338897705,
      "learning_rate": 2.8532869441659615e-05,
      "loss": 0.3654,
      "num_input_tokens_seen": 22471888,
      "step": 34450
    },
    {
      "epoch": 18.05817610062893,
      "grad_norm": 0.12873126566410065,
      "learning_rate": 2.8456771233937973e-05,
      "loss": 0.4113,
      "num_input_tokens_seen": 22474640,
      "step": 34455
    },
    {
      "epoch": 18.060796645702307,
      "grad_norm": 0.11499877274036407,
      "learning_rate": 2.838077166768016e-05,
      "loss": 0.5413,
      "num_input_tokens_seen": 22477168,
      "step": 34460
    },
    {
      "epoch": 18.06341719077568,
      "grad_norm": 0.19486837089061737,
      "learning_rate": 2.8304870758784296e-05,
      "loss": 0.5416,
      "num_input_tokens_seen": 22480240,
      "step": 34465
    },
    {
      "epoch": 18.066037735849058,
      "grad_norm": 0.18571525812149048,
      "learning_rate": 2.822906852312812e-05,
      "loss": 0.4499,
      "num_input_tokens_seen": 22483344,
      "step": 34470
    },
    {
      "epoch": 18.06865828092243,
      "grad_norm": 0.4890564978122711,
      "learning_rate": 2.8153364976568563e-05,
      "loss": 0.3865,
      "num_input_tokens_seen": 22485840,
      "step": 34475
    },
    {
      "epoch": 18.071278825995808,
      "grad_norm": 0.16459627449512482,
      "learning_rate": 2.8077760134941955e-05,
      "loss": 0.4024,
      "num_input_tokens_seen": 22489392,
      "step": 34480
    },
    {
      "epoch": 18.07389937106918,
      "grad_norm": 0.16877758502960205,
      "learning_rate": 2.800225401406392e-05,
      "loss": 0.3471,
      "num_input_tokens_seen": 22491696,
      "step": 34485
    },
    {
      "epoch": 18.07651991614256,
      "grad_norm": 0.21953625977039337,
      "learning_rate": 2.7926846629729607e-05,
      "loss": 0.4333,
      "num_input_tokens_seen": 22494448,
      "step": 34490
    },
    {
      "epoch": 18.079140461215932,
      "grad_norm": 0.19522152841091156,
      "learning_rate": 2.7851537997713174e-05,
      "loss": 0.4007,
      "num_input_tokens_seen": 22497264,
      "step": 34495
    },
    {
      "epoch": 18.08176100628931,
      "grad_norm": 0.10775262862443924,
      "learning_rate": 2.7776328133768458e-05,
      "loss": 0.3384,
      "num_input_tokens_seen": 22500080,
      "step": 34500
    },
    {
      "epoch": 18.084381551362682,
      "grad_norm": 0.16130802035331726,
      "learning_rate": 2.770121705362849e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 22503152,
      "step": 34505
    },
    {
      "epoch": 18.08700209643606,
      "grad_norm": 0.18981514871120453,
      "learning_rate": 2.7626204773005704e-05,
      "loss": 0.4776,
      "num_input_tokens_seen": 22506288,
      "step": 34510
    },
    {
      "epoch": 18.089622641509433,
      "grad_norm": 0.22861571609973907,
      "learning_rate": 2.7551291307591765e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 22508496,
      "step": 34515
    },
    {
      "epoch": 18.09224318658281,
      "grad_norm": 0.1487826555967331,
      "learning_rate": 2.7476476673057636e-05,
      "loss": 0.4567,
      "num_input_tokens_seen": 22512496,
      "step": 34520
    },
    {
      "epoch": 18.094863731656183,
      "grad_norm": 0.09353939443826675,
      "learning_rate": 2.740176088505375e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 22515888,
      "step": 34525
    },
    {
      "epoch": 18.09748427672956,
      "grad_norm": 0.18621210753917694,
      "learning_rate": 2.7327143959209765e-05,
      "loss": 0.3824,
      "num_input_tokens_seen": 22518928,
      "step": 34530
    },
    {
      "epoch": 18.100104821802937,
      "grad_norm": 0.14335116744041443,
      "learning_rate": 2.725262591113481e-05,
      "loss": 0.3789,
      "num_input_tokens_seen": 22523120,
      "step": 34535
    },
    {
      "epoch": 18.10272536687631,
      "grad_norm": 0.2072412371635437,
      "learning_rate": 2.7178206756417078e-05,
      "loss": 0.3732,
      "num_input_tokens_seen": 22525392,
      "step": 34540
    },
    {
      "epoch": 18.105345911949687,
      "grad_norm": 0.32504168152809143,
      "learning_rate": 2.7103886510624344e-05,
      "loss": 0.3853,
      "num_input_tokens_seen": 22527920,
      "step": 34545
    },
    {
      "epoch": 18.10796645702306,
      "grad_norm": 0.1146218329668045,
      "learning_rate": 2.7029665189303387e-05,
      "loss": 0.4668,
      "num_input_tokens_seen": 22531472,
      "step": 34550
    },
    {
      "epoch": 18.110587002096437,
      "grad_norm": 0.17236119508743286,
      "learning_rate": 2.6955542807980515e-05,
      "loss": 0.4821,
      "num_input_tokens_seen": 22534672,
      "step": 34555
    },
    {
      "epoch": 18.11320754716981,
      "grad_norm": 0.12666764855384827,
      "learning_rate": 2.688151938216138e-05,
      "loss": 0.4549,
      "num_input_tokens_seen": 22537424,
      "step": 34560
    },
    {
      "epoch": 18.115828092243188,
      "grad_norm": 0.15113408863544464,
      "learning_rate": 2.6807594927330703e-05,
      "loss": 0.5083,
      "num_input_tokens_seen": 22540944,
      "step": 34565
    },
    {
      "epoch": 18.11844863731656,
      "grad_norm": 0.13064752519130707,
      "learning_rate": 2.6733769458952727e-05,
      "loss": 0.3702,
      "num_input_tokens_seen": 22544368,
      "step": 34570
    },
    {
      "epoch": 18.121069182389938,
      "grad_norm": 0.16103491187095642,
      "learning_rate": 2.6660042992470934e-05,
      "loss": 0.4153,
      "num_input_tokens_seen": 22547568,
      "step": 34575
    },
    {
      "epoch": 18.12368972746331,
      "grad_norm": 0.1623591184616089,
      "learning_rate": 2.658641554330793e-05,
      "loss": 0.4046,
      "num_input_tokens_seen": 22551504,
      "step": 34580
    },
    {
      "epoch": 18.12631027253669,
      "grad_norm": 0.11972303688526154,
      "learning_rate": 2.6512887126865782e-05,
      "loss": 0.409,
      "num_input_tokens_seen": 22554416,
      "step": 34585
    },
    {
      "epoch": 18.128930817610062,
      "grad_norm": 0.1455005407333374,
      "learning_rate": 2.6439457758525908e-05,
      "loss": 0.2697,
      "num_input_tokens_seen": 22557232,
      "step": 34590
    },
    {
      "epoch": 18.13155136268344,
      "grad_norm": 0.07625190913677216,
      "learning_rate": 2.6366127453648748e-05,
      "loss": 0.3729,
      "num_input_tokens_seen": 22560976,
      "step": 34595
    },
    {
      "epoch": 18.134171907756812,
      "grad_norm": 0.16863928735256195,
      "learning_rate": 2.6292896227574303e-05,
      "loss": 0.5576,
      "num_input_tokens_seen": 22564272,
      "step": 34600
    },
    {
      "epoch": 18.13679245283019,
      "grad_norm": 0.13762477040290833,
      "learning_rate": 2.6219764095621547e-05,
      "loss": 0.6959,
      "num_input_tokens_seen": 22567280,
      "step": 34605
    },
    {
      "epoch": 18.139412997903563,
      "grad_norm": 0.13662651181221008,
      "learning_rate": 2.6146731073088958e-05,
      "loss": 0.4199,
      "num_input_tokens_seen": 22571120,
      "step": 34610
    },
    {
      "epoch": 18.14203354297694,
      "grad_norm": 0.1028747409582138,
      "learning_rate": 2.607379717525432e-05,
      "loss": 0.38,
      "num_input_tokens_seen": 22574672,
      "step": 34615
    },
    {
      "epoch": 18.144654088050313,
      "grad_norm": 0.19633778929710388,
      "learning_rate": 2.6000962417374373e-05,
      "loss": 0.4761,
      "num_input_tokens_seen": 22577616,
      "step": 34620
    },
    {
      "epoch": 18.14727463312369,
      "grad_norm": 0.2795371115207672,
      "learning_rate": 2.5928226814685485e-05,
      "loss": 0.4897,
      "num_input_tokens_seen": 22579888,
      "step": 34625
    },
    {
      "epoch": 18.149895178197063,
      "grad_norm": 0.23409375548362732,
      "learning_rate": 2.585559038240304e-05,
      "loss": 0.4687,
      "num_input_tokens_seen": 22582896,
      "step": 34630
    },
    {
      "epoch": 18.15251572327044,
      "grad_norm": 0.20373988151550293,
      "learning_rate": 2.5783053135721714e-05,
      "loss": 0.3422,
      "num_input_tokens_seen": 22585616,
      "step": 34635
    },
    {
      "epoch": 18.155136268343817,
      "grad_norm": 0.185830757021904,
      "learning_rate": 2.571061508981565e-05,
      "loss": 0.5253,
      "num_input_tokens_seen": 22588496,
      "step": 34640
    },
    {
      "epoch": 18.15775681341719,
      "grad_norm": 0.16111938655376434,
      "learning_rate": 2.5638276259837778e-05,
      "loss": 0.4101,
      "num_input_tokens_seen": 22590896,
      "step": 34645
    },
    {
      "epoch": 18.160377358490567,
      "grad_norm": 0.32372573018074036,
      "learning_rate": 2.5566036660920778e-05,
      "loss": 0.7325,
      "num_input_tokens_seen": 22593104,
      "step": 34650
    },
    {
      "epoch": 18.16299790356394,
      "grad_norm": 0.12139274924993515,
      "learning_rate": 2.5493896308176223e-05,
      "loss": 0.4124,
      "num_input_tokens_seen": 22596656,
      "step": 34655
    },
    {
      "epoch": 18.165618448637318,
      "grad_norm": 0.10443613678216934,
      "learning_rate": 2.542185521669521e-05,
      "loss": 0.4359,
      "num_input_tokens_seen": 22599312,
      "step": 34660
    },
    {
      "epoch": 18.16823899371069,
      "grad_norm": 0.08826583623886108,
      "learning_rate": 2.534991340154774e-05,
      "loss": 0.3336,
      "num_input_tokens_seen": 22603440,
      "step": 34665
    },
    {
      "epoch": 18.170859538784068,
      "grad_norm": 0.2027357816696167,
      "learning_rate": 2.5278070877783332e-05,
      "loss": 0.4386,
      "num_input_tokens_seen": 22606512,
      "step": 34670
    },
    {
      "epoch": 18.17348008385744,
      "grad_norm": 0.1227668896317482,
      "learning_rate": 2.520632766043052e-05,
      "loss": 0.5036,
      "num_input_tokens_seen": 22609584,
      "step": 34675
    },
    {
      "epoch": 18.17610062893082,
      "grad_norm": 0.1627541035413742,
      "learning_rate": 2.513468376449729e-05,
      "loss": 0.4587,
      "num_input_tokens_seen": 22613712,
      "step": 34680
    },
    {
      "epoch": 18.178721174004192,
      "grad_norm": 0.11513673514127731,
      "learning_rate": 2.506313920497061e-05,
      "loss": 0.3843,
      "num_input_tokens_seen": 22617008,
      "step": 34685
    },
    {
      "epoch": 18.18134171907757,
      "grad_norm": 0.2836589217185974,
      "learning_rate": 2.4991693996816888e-05,
      "loss": 0.282,
      "num_input_tokens_seen": 22620112,
      "step": 34690
    },
    {
      "epoch": 18.183962264150942,
      "grad_norm": 0.14982949197292328,
      "learning_rate": 2.4920348154981677e-05,
      "loss": 0.3383,
      "num_input_tokens_seen": 22622608,
      "step": 34695
    },
    {
      "epoch": 18.18658280922432,
      "grad_norm": 0.2120458334684372,
      "learning_rate": 2.4849101694389477e-05,
      "loss": 0.5241,
      "num_input_tokens_seen": 22626416,
      "step": 34700
    },
    {
      "epoch": 18.189203354297693,
      "grad_norm": 0.19551613926887512,
      "learning_rate": 2.4777954629944478e-05,
      "loss": 0.326,
      "num_input_tokens_seen": 22629328,
      "step": 34705
    },
    {
      "epoch": 18.19182389937107,
      "grad_norm": 0.19315205514431,
      "learning_rate": 2.4706906976529718e-05,
      "loss": 0.5246,
      "num_input_tokens_seen": 22632144,
      "step": 34710
    },
    {
      "epoch": 18.194444444444443,
      "grad_norm": 0.13152571022510529,
      "learning_rate": 2.4635958749007648e-05,
      "loss": 0.4111,
      "num_input_tokens_seen": 22634992,
      "step": 34715
    },
    {
      "epoch": 18.19706498951782,
      "grad_norm": 0.10618185251951218,
      "learning_rate": 2.456510996221978e-05,
      "loss": 0.3829,
      "num_input_tokens_seen": 22638256,
      "step": 34720
    },
    {
      "epoch": 18.199685534591197,
      "grad_norm": 0.15246321260929108,
      "learning_rate": 2.4494360630986756e-05,
      "loss": 0.4402,
      "num_input_tokens_seen": 22641712,
      "step": 34725
    },
    {
      "epoch": 18.20230607966457,
      "grad_norm": 0.09207676351070404,
      "learning_rate": 2.4423710770108687e-05,
      "loss": 0.4189,
      "num_input_tokens_seen": 22645904,
      "step": 34730
    },
    {
      "epoch": 18.204926624737947,
      "grad_norm": 0.2518055737018585,
      "learning_rate": 2.435316039436464e-05,
      "loss": 0.4479,
      "num_input_tokens_seen": 22648944,
      "step": 34735
    },
    {
      "epoch": 18.20754716981132,
      "grad_norm": 0.19282864034175873,
      "learning_rate": 2.428270951851297e-05,
      "loss": 0.4419,
      "num_input_tokens_seen": 22651888,
      "step": 34740
    },
    {
      "epoch": 18.210167714884697,
      "grad_norm": 0.21503517031669617,
      "learning_rate": 2.421235815729128e-05,
      "loss": 0.4568,
      "num_input_tokens_seen": 22655120,
      "step": 34745
    },
    {
      "epoch": 18.21278825995807,
      "grad_norm": 0.24514459073543549,
      "learning_rate": 2.414210632541619e-05,
      "loss": 0.7387,
      "num_input_tokens_seen": 22658928,
      "step": 34750
    },
    {
      "epoch": 18.215408805031448,
      "grad_norm": 0.23394513130187988,
      "learning_rate": 2.40719540375835e-05,
      "loss": 0.4457,
      "num_input_tokens_seen": 22662288,
      "step": 34755
    },
    {
      "epoch": 18.21802935010482,
      "grad_norm": 0.297480970621109,
      "learning_rate": 2.4001901308468353e-05,
      "loss": 0.3886,
      "num_input_tokens_seen": 22664848,
      "step": 34760
    },
    {
      "epoch": 18.220649895178198,
      "grad_norm": 0.10893955826759338,
      "learning_rate": 2.3931948152724982e-05,
      "loss": 0.3537,
      "num_input_tokens_seen": 22667376,
      "step": 34765
    },
    {
      "epoch": 18.22327044025157,
      "grad_norm": 0.1625664085149765,
      "learning_rate": 2.386209458498684e-05,
      "loss": 0.4802,
      "num_input_tokens_seen": 22671440,
      "step": 34770
    },
    {
      "epoch": 18.22589098532495,
      "grad_norm": 0.11515478789806366,
      "learning_rate": 2.3792340619866458e-05,
      "loss": 0.4087,
      "num_input_tokens_seen": 22675408,
      "step": 34775
    },
    {
      "epoch": 18.228511530398322,
      "grad_norm": 0.1638508290052414,
      "learning_rate": 2.3722686271955507e-05,
      "loss": 0.3748,
      "num_input_tokens_seen": 22678224,
      "step": 34780
    },
    {
      "epoch": 18.2311320754717,
      "grad_norm": 0.14118489623069763,
      "learning_rate": 2.365313155582488e-05,
      "loss": 0.3309,
      "num_input_tokens_seen": 22681712,
      "step": 34785
    },
    {
      "epoch": 18.233752620545072,
      "grad_norm": 0.13083061575889587,
      "learning_rate": 2.358367648602472e-05,
      "loss": 0.2974,
      "num_input_tokens_seen": 22685008,
      "step": 34790
    },
    {
      "epoch": 18.23637316561845,
      "grad_norm": 0.19989018142223358,
      "learning_rate": 2.3514321077084234e-05,
      "loss": 0.4638,
      "num_input_tokens_seen": 22688368,
      "step": 34795
    },
    {
      "epoch": 18.238993710691823,
      "grad_norm": 0.18579570949077606,
      "learning_rate": 2.3445065343511763e-05,
      "loss": 0.4309,
      "num_input_tokens_seen": 22691184,
      "step": 34800
    },
    {
      "epoch": 18.2416142557652,
      "grad_norm": 0.12207461148500443,
      "learning_rate": 2.3375909299794717e-05,
      "loss": 0.4477,
      "num_input_tokens_seen": 22695024,
      "step": 34805
    },
    {
      "epoch": 18.244234800838573,
      "grad_norm": 0.21863703429698944,
      "learning_rate": 2.330685296039986e-05,
      "loss": 0.4103,
      "num_input_tokens_seen": 22699088,
      "step": 34810
    },
    {
      "epoch": 18.24685534591195,
      "grad_norm": 0.15528081357479095,
      "learning_rate": 2.3237896339772914e-05,
      "loss": 0.3671,
      "num_input_tokens_seen": 22702480,
      "step": 34815
    },
    {
      "epoch": 18.249475890985323,
      "grad_norm": 0.2373822182416916,
      "learning_rate": 2.3169039452338892e-05,
      "loss": 0.2956,
      "num_input_tokens_seen": 22705840,
      "step": 34820
    },
    {
      "epoch": 18.2520964360587,
      "grad_norm": 0.1753515601158142,
      "learning_rate": 2.3100282312501886e-05,
      "loss": 0.4418,
      "num_input_tokens_seen": 22708752,
      "step": 34825
    },
    {
      "epoch": 18.254716981132077,
      "grad_norm": 0.23714616894721985,
      "learning_rate": 2.3031624934645113e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 22711504,
      "step": 34830
    },
    {
      "epoch": 18.25733752620545,
      "grad_norm": 0.15465296804904938,
      "learning_rate": 2.296306733313075e-05,
      "loss": 0.6007,
      "num_input_tokens_seen": 22714384,
      "step": 34835
    },
    {
      "epoch": 18.259958071278827,
      "grad_norm": 0.27659374475479126,
      "learning_rate": 2.289460952230038e-05,
      "loss": 0.447,
      "num_input_tokens_seen": 22718192,
      "step": 34840
    },
    {
      "epoch": 18.2625786163522,
      "grad_norm": 0.24372872710227966,
      "learning_rate": 2.2826251516474604e-05,
      "loss": 0.4613,
      "num_input_tokens_seen": 22721616,
      "step": 34845
    },
    {
      "epoch": 18.265199161425578,
      "grad_norm": 0.13333852589130402,
      "learning_rate": 2.2757993329953152e-05,
      "loss": 0.5123,
      "num_input_tokens_seen": 22724624,
      "step": 34850
    },
    {
      "epoch": 18.26781970649895,
      "grad_norm": 0.1106642335653305,
      "learning_rate": 2.2689834977014822e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 22727600,
      "step": 34855
    },
    {
      "epoch": 18.270440251572328,
      "grad_norm": 0.09266278892755508,
      "learning_rate": 2.2621776471917598e-05,
      "loss": 0.2946,
      "num_input_tokens_seen": 22730608,
      "step": 34860
    },
    {
      "epoch": 18.2730607966457,
      "grad_norm": 0.16729256510734558,
      "learning_rate": 2.255381782889848e-05,
      "loss": 0.4198,
      "num_input_tokens_seen": 22733424,
      "step": 34865
    },
    {
      "epoch": 18.27568134171908,
      "grad_norm": 0.11026950925588608,
      "learning_rate": 2.2485959062173654e-05,
      "loss": 0.397,
      "num_input_tokens_seen": 22737008,
      "step": 34870
    },
    {
      "epoch": 18.278301886792452,
      "grad_norm": 0.2208625227212906,
      "learning_rate": 2.2418200185938485e-05,
      "loss": 0.4014,
      "num_input_tokens_seen": 22740272,
      "step": 34875
    },
    {
      "epoch": 18.28092243186583,
      "grad_norm": 0.08858656138181686,
      "learning_rate": 2.235054121436725e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 22744464,
      "step": 34880
    },
    {
      "epoch": 18.283542976939202,
      "grad_norm": 0.1578589230775833,
      "learning_rate": 2.2282982161613562e-05,
      "loss": 0.4315,
      "num_input_tokens_seen": 22747152,
      "step": 34885
    },
    {
      "epoch": 18.28616352201258,
      "grad_norm": 0.16493381559848785,
      "learning_rate": 2.221552304180985e-05,
      "loss": 0.4046,
      "num_input_tokens_seen": 22751408,
      "step": 34890
    },
    {
      "epoch": 18.288784067085953,
      "grad_norm": 0.20180492103099823,
      "learning_rate": 2.2148163869067816e-05,
      "loss": 0.4814,
      "num_input_tokens_seen": 22755568,
      "step": 34895
    },
    {
      "epoch": 18.29140461215933,
      "grad_norm": 0.1287577748298645,
      "learning_rate": 2.2080904657478305e-05,
      "loss": 0.3868,
      "num_input_tokens_seen": 22758576,
      "step": 34900
    },
    {
      "epoch": 18.294025157232703,
      "grad_norm": 0.1373993456363678,
      "learning_rate": 2.201374542111123e-05,
      "loss": 0.4292,
      "num_input_tokens_seen": 22762128,
      "step": 34905
    },
    {
      "epoch": 18.29664570230608,
      "grad_norm": 0.18446561694145203,
      "learning_rate": 2.1946686174015407e-05,
      "loss": 0.4068,
      "num_input_tokens_seen": 22764784,
      "step": 34910
    },
    {
      "epoch": 18.299266247379457,
      "grad_norm": 0.17769773304462433,
      "learning_rate": 2.1879726930218946e-05,
      "loss": 0.426,
      "num_input_tokens_seen": 22768304,
      "step": 34915
    },
    {
      "epoch": 18.30188679245283,
      "grad_norm": 0.11726228892803192,
      "learning_rate": 2.1812867703728866e-05,
      "loss": 0.3971,
      "num_input_tokens_seen": 22771632,
      "step": 34920
    },
    {
      "epoch": 18.304507337526207,
      "grad_norm": 0.13938429951667786,
      "learning_rate": 2.1746108508531426e-05,
      "loss": 0.3411,
      "num_input_tokens_seen": 22774928,
      "step": 34925
    },
    {
      "epoch": 18.30712788259958,
      "grad_norm": 0.1745249330997467,
      "learning_rate": 2.1679449358591952e-05,
      "loss": 0.3557,
      "num_input_tokens_seen": 22778640,
      "step": 34930
    },
    {
      "epoch": 18.309748427672957,
      "grad_norm": 0.1256585419178009,
      "learning_rate": 2.1612890267854624e-05,
      "loss": 0.3278,
      "num_input_tokens_seen": 22781936,
      "step": 34935
    },
    {
      "epoch": 18.31236897274633,
      "grad_norm": 0.15876458585262299,
      "learning_rate": 2.1546431250242914e-05,
      "loss": 0.3724,
      "num_input_tokens_seen": 22784528,
      "step": 34940
    },
    {
      "epoch": 18.314989517819708,
      "grad_norm": 0.11352559924125671,
      "learning_rate": 2.1480072319659427e-05,
      "loss": 0.3077,
      "num_input_tokens_seen": 22790704,
      "step": 34945
    },
    {
      "epoch": 18.31761006289308,
      "grad_norm": 0.10428372025489807,
      "learning_rate": 2.1413813489985447e-05,
      "loss": 0.4542,
      "num_input_tokens_seen": 22794320,
      "step": 34950
    },
    {
      "epoch": 18.320230607966458,
      "grad_norm": 0.09294899553060532,
      "learning_rate": 2.1347654775081716e-05,
      "loss": 0.4753,
      "num_input_tokens_seen": 22798544,
      "step": 34955
    },
    {
      "epoch": 18.32285115303983,
      "grad_norm": 0.18966306746006012,
      "learning_rate": 2.1281596188787834e-05,
      "loss": 0.3972,
      "num_input_tokens_seen": 22801392,
      "step": 34960
    },
    {
      "epoch": 18.32547169811321,
      "grad_norm": 0.10119341313838959,
      "learning_rate": 2.121563774492252e-05,
      "loss": 0.4025,
      "num_input_tokens_seen": 22804880,
      "step": 34965
    },
    {
      "epoch": 18.328092243186582,
      "grad_norm": 0.13554252684116364,
      "learning_rate": 2.114977945728358e-05,
      "loss": 0.608,
      "num_input_tokens_seen": 22808112,
      "step": 34970
    },
    {
      "epoch": 18.33071278825996,
      "grad_norm": 0.15146151185035706,
      "learning_rate": 2.1084021339647707e-05,
      "loss": 0.4494,
      "num_input_tokens_seen": 22814544,
      "step": 34975
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 0.14876778423786163,
      "learning_rate": 2.1018363405770792e-05,
      "loss": 0.3492,
      "num_input_tokens_seen": 22817200,
      "step": 34980
    },
    {
      "epoch": 18.33595387840671,
      "grad_norm": 0.22333897650241852,
      "learning_rate": 2.095280566938784e-05,
      "loss": 0.3357,
      "num_input_tokens_seen": 22820048,
      "step": 34985
    },
    {
      "epoch": 18.338574423480082,
      "grad_norm": 0.168918177485466,
      "learning_rate": 2.0887348144212615e-05,
      "loss": 0.3754,
      "num_input_tokens_seen": 22823280,
      "step": 34990
    },
    {
      "epoch": 18.34119496855346,
      "grad_norm": 0.14043912291526794,
      "learning_rate": 2.08219908439381e-05,
      "loss": 0.4762,
      "num_input_tokens_seen": 22826224,
      "step": 34995
    },
    {
      "epoch": 18.343815513626833,
      "grad_norm": 0.14757387340068817,
      "learning_rate": 2.075673378223647e-05,
      "loss": 0.3158,
      "num_input_tokens_seen": 22829360,
      "step": 35000
    },
    {
      "epoch": 18.34643605870021,
      "grad_norm": 0.23211580514907837,
      "learning_rate": 2.069157697275853e-05,
      "loss": 0.3997,
      "num_input_tokens_seen": 22832688,
      "step": 35005
    },
    {
      "epoch": 18.349056603773583,
      "grad_norm": 0.09832059592008591,
      "learning_rate": 2.0626520429134543e-05,
      "loss": 0.3555,
      "num_input_tokens_seen": 22836176,
      "step": 35010
    },
    {
      "epoch": 18.35167714884696,
      "grad_norm": 0.14694538712501526,
      "learning_rate": 2.0561564164973458e-05,
      "loss": 0.3097,
      "num_input_tokens_seen": 22839088,
      "step": 35015
    },
    {
      "epoch": 18.354297693920337,
      "grad_norm": 0.19071711599826813,
      "learning_rate": 2.04967081938634e-05,
      "loss": 0.4021,
      "num_input_tokens_seen": 22842384,
      "step": 35020
    },
    {
      "epoch": 18.35691823899371,
      "grad_norm": 0.13607248663902283,
      "learning_rate": 2.043195252937152e-05,
      "loss": 0.4408,
      "num_input_tokens_seen": 22845232,
      "step": 35025
    },
    {
      "epoch": 18.359538784067087,
      "grad_norm": 0.19592465460300446,
      "learning_rate": 2.0367297185044043e-05,
      "loss": 0.4412,
      "num_input_tokens_seen": 22848016,
      "step": 35030
    },
    {
      "epoch": 18.36215932914046,
      "grad_norm": 0.1881827861070633,
      "learning_rate": 2.030274217440603e-05,
      "loss": 0.5107,
      "num_input_tokens_seen": 22850672,
      "step": 35035
    },
    {
      "epoch": 18.364779874213838,
      "grad_norm": 0.14707684516906738,
      "learning_rate": 2.0238287510961628e-05,
      "loss": 0.2978,
      "num_input_tokens_seen": 22853584,
      "step": 35040
    },
    {
      "epoch": 18.36740041928721,
      "grad_norm": 0.07163728028535843,
      "learning_rate": 2.017393320819405e-05,
      "loss": 0.3747,
      "num_input_tokens_seen": 22857424,
      "step": 35045
    },
    {
      "epoch": 18.370020964360588,
      "grad_norm": 0.1442745327949524,
      "learning_rate": 2.010967927956553e-05,
      "loss": 0.3845,
      "num_input_tokens_seen": 22861456,
      "step": 35050
    },
    {
      "epoch": 18.37264150943396,
      "grad_norm": 0.1702989935874939,
      "learning_rate": 2.00455257385172e-05,
      "loss": 0.3705,
      "num_input_tokens_seen": 22864752,
      "step": 35055
    },
    {
      "epoch": 18.37526205450734,
      "grad_norm": 0.146615669131279,
      "learning_rate": 1.9981472598469386e-05,
      "loss": 0.3471,
      "num_input_tokens_seen": 22868624,
      "step": 35060
    },
    {
      "epoch": 18.377882599580712,
      "grad_norm": 0.13223972916603088,
      "learning_rate": 1.9917519872821142e-05,
      "loss": 0.4729,
      "num_input_tokens_seen": 22871760,
      "step": 35065
    },
    {
      "epoch": 18.38050314465409,
      "grad_norm": 0.12972864508628845,
      "learning_rate": 1.9853667574950605e-05,
      "loss": 0.3713,
      "num_input_tokens_seen": 22875248,
      "step": 35070
    },
    {
      "epoch": 18.383123689727462,
      "grad_norm": 0.1382351815700531,
      "learning_rate": 1.978991571821498e-05,
      "loss": 0.4612,
      "num_input_tokens_seen": 22878640,
      "step": 35075
    },
    {
      "epoch": 18.38574423480084,
      "grad_norm": 0.12740397453308105,
      "learning_rate": 1.972626431595048e-05,
      "loss": 0.377,
      "num_input_tokens_seen": 22881584,
      "step": 35080
    },
    {
      "epoch": 18.388364779874212,
      "grad_norm": 0.15781532227993011,
      "learning_rate": 1.9662713381472295e-05,
      "loss": 0.3947,
      "num_input_tokens_seen": 22884368,
      "step": 35085
    },
    {
      "epoch": 18.39098532494759,
      "grad_norm": 0.20466601848602295,
      "learning_rate": 1.959926292807451e-05,
      "loss": 0.509,
      "num_input_tokens_seen": 22886896,
      "step": 35090
    },
    {
      "epoch": 18.393605870020963,
      "grad_norm": 0.11229072511196136,
      "learning_rate": 1.9535912969030178e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 22890448,
      "step": 35095
    },
    {
      "epoch": 18.39622641509434,
      "grad_norm": 0.09069973975419998,
      "learning_rate": 1.947266351759136e-05,
      "loss": 0.4035,
      "num_input_tokens_seen": 22894224,
      "step": 35100
    },
    {
      "epoch": 18.398846960167713,
      "grad_norm": 0.1466568559408188,
      "learning_rate": 1.940951458698925e-05,
      "loss": 0.5141,
      "num_input_tokens_seen": 22897680,
      "step": 35105
    },
    {
      "epoch": 18.40146750524109,
      "grad_norm": 0.18473921716213226,
      "learning_rate": 1.9346466190433842e-05,
      "loss": 0.4072,
      "num_input_tokens_seen": 22900880,
      "step": 35110
    },
    {
      "epoch": 18.404088050314467,
      "grad_norm": 0.19000670313835144,
      "learning_rate": 1.9283518341114136e-05,
      "loss": 0.4257,
      "num_input_tokens_seen": 22903984,
      "step": 35115
    },
    {
      "epoch": 18.40670859538784,
      "grad_norm": 0.17471705377101898,
      "learning_rate": 1.9220671052198047e-05,
      "loss": 0.5028,
      "num_input_tokens_seen": 22907216,
      "step": 35120
    },
    {
      "epoch": 18.409329140461217,
      "grad_norm": 0.1592024266719818,
      "learning_rate": 1.9157924336832556e-05,
      "loss": 0.5816,
      "num_input_tokens_seen": 22910352,
      "step": 35125
    },
    {
      "epoch": 18.41194968553459,
      "grad_norm": 0.1277400106191635,
      "learning_rate": 1.909527820814355e-05,
      "loss": 0.6077,
      "num_input_tokens_seen": 22913776,
      "step": 35130
    },
    {
      "epoch": 18.414570230607968,
      "grad_norm": 0.15919266641139984,
      "learning_rate": 1.9032732679235886e-05,
      "loss": 0.5534,
      "num_input_tokens_seen": 22917584,
      "step": 35135
    },
    {
      "epoch": 18.41719077568134,
      "grad_norm": 0.24422258138656616,
      "learning_rate": 1.8970287763193428e-05,
      "loss": 0.4127,
      "num_input_tokens_seen": 22920592,
      "step": 35140
    },
    {
      "epoch": 18.419811320754718,
      "grad_norm": 0.15110361576080322,
      "learning_rate": 1.8907943473078892e-05,
      "loss": 0.4117,
      "num_input_tokens_seen": 22923248,
      "step": 35145
    },
    {
      "epoch": 18.42243186582809,
      "grad_norm": 0.2185227870941162,
      "learning_rate": 1.884569982193396e-05,
      "loss": 0.3661,
      "num_input_tokens_seen": 22926832,
      "step": 35150
    },
    {
      "epoch": 18.42505241090147,
      "grad_norm": 0.2458072006702423,
      "learning_rate": 1.8783556822779267e-05,
      "loss": 0.4296,
      "num_input_tokens_seen": 22930256,
      "step": 35155
    },
    {
      "epoch": 18.427672955974842,
      "grad_norm": 0.1814337521791458,
      "learning_rate": 1.8721514488614532e-05,
      "loss": 0.5727,
      "num_input_tokens_seen": 22933552,
      "step": 35160
    },
    {
      "epoch": 18.43029350104822,
      "grad_norm": 0.2644719183444977,
      "learning_rate": 1.8659572832418315e-05,
      "loss": 0.3944,
      "num_input_tokens_seen": 22936112,
      "step": 35165
    },
    {
      "epoch": 18.432914046121592,
      "grad_norm": 0.36345478892326355,
      "learning_rate": 1.8597731867148026e-05,
      "loss": 0.5773,
      "num_input_tokens_seen": 22939376,
      "step": 35170
    },
    {
      "epoch": 18.43553459119497,
      "grad_norm": 0.11548806726932526,
      "learning_rate": 1.8535991605740043e-05,
      "loss": 0.3923,
      "num_input_tokens_seen": 22942416,
      "step": 35175
    },
    {
      "epoch": 18.438155136268342,
      "grad_norm": 0.11924194544553757,
      "learning_rate": 1.8474352061109757e-05,
      "loss": 0.4708,
      "num_input_tokens_seen": 22945520,
      "step": 35180
    },
    {
      "epoch": 18.44077568134172,
      "grad_norm": 0.1716402918100357,
      "learning_rate": 1.8412813246151515e-05,
      "loss": 0.4677,
      "num_input_tokens_seen": 22948240,
      "step": 35185
    },
    {
      "epoch": 18.443396226415093,
      "grad_norm": 0.1293991208076477,
      "learning_rate": 1.8351375173738584e-05,
      "loss": 0.4811,
      "num_input_tokens_seen": 22951536,
      "step": 35190
    },
    {
      "epoch": 18.44601677148847,
      "grad_norm": 0.1591542810201645,
      "learning_rate": 1.829003785672295e-05,
      "loss": 0.4009,
      "num_input_tokens_seen": 22955280,
      "step": 35195
    },
    {
      "epoch": 18.448637316561843,
      "grad_norm": 0.08600656688213348,
      "learning_rate": 1.8228801307935806e-05,
      "loss": 0.4267,
      "num_input_tokens_seen": 22958576,
      "step": 35200
    },
    {
      "epoch": 18.45125786163522,
      "grad_norm": 0.15130619704723358,
      "learning_rate": 1.8167665540187063e-05,
      "loss": 0.4438,
      "num_input_tokens_seen": 22961328,
      "step": 35205
    },
    {
      "epoch": 18.453878406708597,
      "grad_norm": 0.10706930607557297,
      "learning_rate": 1.8106630566265604e-05,
      "loss": 0.3014,
      "num_input_tokens_seen": 22964880,
      "step": 35210
    },
    {
      "epoch": 18.45649895178197,
      "grad_norm": 0.1806456446647644,
      "learning_rate": 1.8045696398939326e-05,
      "loss": 0.405,
      "num_input_tokens_seen": 22968848,
      "step": 35215
    },
    {
      "epoch": 18.459119496855347,
      "grad_norm": 0.12899667024612427,
      "learning_rate": 1.7984863050955036e-05,
      "loss": 0.5101,
      "num_input_tokens_seen": 22972944,
      "step": 35220
    },
    {
      "epoch": 18.46174004192872,
      "grad_norm": 0.15491105616092682,
      "learning_rate": 1.7924130535038162e-05,
      "loss": 0.4701,
      "num_input_tokens_seen": 22976208,
      "step": 35225
    },
    {
      "epoch": 18.464360587002098,
      "grad_norm": 0.1912110298871994,
      "learning_rate": 1.7863498863893433e-05,
      "loss": 0.4033,
      "num_input_tokens_seen": 22978832,
      "step": 35230
    },
    {
      "epoch": 18.46698113207547,
      "grad_norm": 0.19270800054073334,
      "learning_rate": 1.7802968050204203e-05,
      "loss": 0.5057,
      "num_input_tokens_seen": 22982096,
      "step": 35235
    },
    {
      "epoch": 18.469601677148848,
      "grad_norm": 0.17906993627548218,
      "learning_rate": 1.7742538106632844e-05,
      "loss": 0.5859,
      "num_input_tokens_seen": 22986032,
      "step": 35240
    },
    {
      "epoch": 18.47222222222222,
      "grad_norm": 0.12538950145244598,
      "learning_rate": 1.7682209045820684e-05,
      "loss": 0.3565,
      "num_input_tokens_seen": 22989168,
      "step": 35245
    },
    {
      "epoch": 18.4748427672956,
      "grad_norm": 0.11903142929077148,
      "learning_rate": 1.76219808803878e-05,
      "loss": 0.4738,
      "num_input_tokens_seen": 22992784,
      "step": 35250
    },
    {
      "epoch": 18.47746331236897,
      "grad_norm": 0.09884186089038849,
      "learning_rate": 1.7561853622933278e-05,
      "loss": 0.3464,
      "num_input_tokens_seen": 22995824,
      "step": 35255
    },
    {
      "epoch": 18.48008385744235,
      "grad_norm": 0.2165519744157791,
      "learning_rate": 1.7501827286035e-05,
      "loss": 0.4793,
      "num_input_tokens_seen": 22998928,
      "step": 35260
    },
    {
      "epoch": 18.482704402515722,
      "grad_norm": 0.09234408289194107,
      "learning_rate": 1.7441901882249754e-05,
      "loss": 0.3338,
      "num_input_tokens_seen": 23001424,
      "step": 35265
    },
    {
      "epoch": 18.4853249475891,
      "grad_norm": 0.1671629697084427,
      "learning_rate": 1.7382077424113464e-05,
      "loss": 0.3866,
      "num_input_tokens_seen": 23004784,
      "step": 35270
    },
    {
      "epoch": 18.487945492662472,
      "grad_norm": 0.19568270444869995,
      "learning_rate": 1.7322353924140498e-05,
      "loss": 0.3582,
      "num_input_tokens_seen": 23008048,
      "step": 35275
    },
    {
      "epoch": 18.49056603773585,
      "grad_norm": 0.16937020421028137,
      "learning_rate": 1.7262731394824372e-05,
      "loss": 0.2952,
      "num_input_tokens_seen": 23010416,
      "step": 35280
    },
    {
      "epoch": 18.493186582809223,
      "grad_norm": 0.25061720609664917,
      "learning_rate": 1.7203209848637603e-05,
      "loss": 0.4329,
      "num_input_tokens_seen": 23013616,
      "step": 35285
    },
    {
      "epoch": 18.4958071278826,
      "grad_norm": 0.15144263207912445,
      "learning_rate": 1.7143789298031175e-05,
      "loss": 0.4018,
      "num_input_tokens_seen": 23017808,
      "step": 35290
    },
    {
      "epoch": 18.498427672955973,
      "grad_norm": 0.12199393659830093,
      "learning_rate": 1.708446975543537e-05,
      "loss": 0.3369,
      "num_input_tokens_seen": 23020720,
      "step": 35295
    },
    {
      "epoch": 18.50104821802935,
      "grad_norm": 0.23349842429161072,
      "learning_rate": 1.7025251233259098e-05,
      "loss": 0.3027,
      "num_input_tokens_seen": 23024560,
      "step": 35300
    },
    {
      "epoch": 18.503668763102727,
      "grad_norm": 0.11287684738636017,
      "learning_rate": 1.6966133743890166e-05,
      "loss": 0.4093,
      "num_input_tokens_seen": 23027856,
      "step": 35305
    },
    {
      "epoch": 18.5062893081761,
      "grad_norm": 0.22907668352127075,
      "learning_rate": 1.690711729969535e-05,
      "loss": 0.3726,
      "num_input_tokens_seen": 23031024,
      "step": 35310
    },
    {
      "epoch": 18.508909853249477,
      "grad_norm": 0.14416445791721344,
      "learning_rate": 1.684820191302022e-05,
      "loss": 0.3525,
      "num_input_tokens_seen": 23034192,
      "step": 35315
    },
    {
      "epoch": 18.51153039832285,
      "grad_norm": 0.16566286981105804,
      "learning_rate": 1.6789387596189087e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 23037680,
      "step": 35320
    },
    {
      "epoch": 18.514150943396228,
      "grad_norm": 0.1465466469526291,
      "learning_rate": 1.6730674361505382e-05,
      "loss": 0.4078,
      "num_input_tokens_seen": 23040784,
      "step": 35325
    },
    {
      "epoch": 18.5167714884696,
      "grad_norm": 0.11581508815288544,
      "learning_rate": 1.6672062221251117e-05,
      "loss": 0.3935,
      "num_input_tokens_seen": 23043888,
      "step": 35330
    },
    {
      "epoch": 18.519392033542978,
      "grad_norm": 0.2302982360124588,
      "learning_rate": 1.6613551187687314e-05,
      "loss": 0.4532,
      "num_input_tokens_seen": 23046928,
      "step": 35335
    },
    {
      "epoch": 18.52201257861635,
      "grad_norm": 0.10915984958410263,
      "learning_rate": 1.6555141273053907e-05,
      "loss": 0.34,
      "num_input_tokens_seen": 23052016,
      "step": 35340
    },
    {
      "epoch": 18.52463312368973,
      "grad_norm": 0.17871415615081787,
      "learning_rate": 1.6496832489569457e-05,
      "loss": 0.4066,
      "num_input_tokens_seen": 23055632,
      "step": 35345
    },
    {
      "epoch": 18.5272536687631,
      "grad_norm": 0.142666757106781,
      "learning_rate": 1.643862484943165e-05,
      "loss": 0.4199,
      "num_input_tokens_seen": 23058640,
      "step": 35350
    },
    {
      "epoch": 18.52987421383648,
      "grad_norm": 0.21005111932754517,
      "learning_rate": 1.6380518364816687e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 23061488,
      "step": 35355
    },
    {
      "epoch": 18.532494758909852,
      "grad_norm": 0.1941126435995102,
      "learning_rate": 1.632251304787985e-05,
      "loss": 0.3365,
      "num_input_tokens_seen": 23064016,
      "step": 35360
    },
    {
      "epoch": 18.53511530398323,
      "grad_norm": 0.11774710565805435,
      "learning_rate": 1.6264608910755153e-05,
      "loss": 0.3891,
      "num_input_tokens_seen": 23066576,
      "step": 35365
    },
    {
      "epoch": 18.537735849056602,
      "grad_norm": 0.09727490693330765,
      "learning_rate": 1.6206805965555627e-05,
      "loss": 0.4386,
      "num_input_tokens_seen": 23069936,
      "step": 35370
    },
    {
      "epoch": 18.54035639412998,
      "grad_norm": 0.14849790930747986,
      "learning_rate": 1.614910422437288e-05,
      "loss": 0.7166,
      "num_input_tokens_seen": 23073392,
      "step": 35375
    },
    {
      "epoch": 18.542976939203353,
      "grad_norm": 0.16805247962474823,
      "learning_rate": 1.6091503699277477e-05,
      "loss": 0.4112,
      "num_input_tokens_seen": 23079568,
      "step": 35380
    },
    {
      "epoch": 18.54559748427673,
      "grad_norm": 0.22398722171783447,
      "learning_rate": 1.6034004402318726e-05,
      "loss": 0.329,
      "num_input_tokens_seen": 23082480,
      "step": 35385
    },
    {
      "epoch": 18.548218029350103,
      "grad_norm": 0.15994171798229218,
      "learning_rate": 1.5976606345524836e-05,
      "loss": 0.3841,
      "num_input_tokens_seen": 23085328,
      "step": 35390
    },
    {
      "epoch": 18.55083857442348,
      "grad_norm": 0.12140487134456635,
      "learning_rate": 1.5919309540902927e-05,
      "loss": 0.447,
      "num_input_tokens_seen": 23088208,
      "step": 35395
    },
    {
      "epoch": 18.553459119496857,
      "grad_norm": 0.13693268597126007,
      "learning_rate": 1.5862114000438797e-05,
      "loss": 0.3579,
      "num_input_tokens_seen": 23091376,
      "step": 35400
    },
    {
      "epoch": 18.55607966457023,
      "grad_norm": 0.1962408572435379,
      "learning_rate": 1.5805019736097104e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 23094832,
      "step": 35405
    },
    {
      "epoch": 18.558700209643607,
      "grad_norm": 0.1320924013853073,
      "learning_rate": 1.5748026759821232e-05,
      "loss": 0.347,
      "num_input_tokens_seen": 23097808,
      "step": 35410
    },
    {
      "epoch": 18.56132075471698,
      "grad_norm": 0.4133259952068329,
      "learning_rate": 1.5691135083533537e-05,
      "loss": 0.5002,
      "num_input_tokens_seen": 23100464,
      "step": 35415
    },
    {
      "epoch": 18.563941299790358,
      "grad_norm": 0.09558404982089996,
      "learning_rate": 1.5634344719135052e-05,
      "loss": 0.3637,
      "num_input_tokens_seen": 23103760,
      "step": 35420
    },
    {
      "epoch": 18.56656184486373,
      "grad_norm": 0.18039532005786896,
      "learning_rate": 1.5577655678505776e-05,
      "loss": 0.4174,
      "num_input_tokens_seen": 23106832,
      "step": 35425
    },
    {
      "epoch": 18.569182389937108,
      "grad_norm": 0.1355999857187271,
      "learning_rate": 1.5521067973504442e-05,
      "loss": 0.4753,
      "num_input_tokens_seen": 23110832,
      "step": 35430
    },
    {
      "epoch": 18.57180293501048,
      "grad_norm": 0.2126908302307129,
      "learning_rate": 1.546458161596831e-05,
      "loss": 0.3034,
      "num_input_tokens_seen": 23113968,
      "step": 35435
    },
    {
      "epoch": 18.57442348008386,
      "grad_norm": 0.2747897207736969,
      "learning_rate": 1.5408196617713866e-05,
      "loss": 0.4055,
      "num_input_tokens_seen": 23116944,
      "step": 35440
    },
    {
      "epoch": 18.57704402515723,
      "grad_norm": 0.2102476805448532,
      "learning_rate": 1.5351912990536175e-05,
      "loss": 0.4864,
      "num_input_tokens_seen": 23120112,
      "step": 35445
    },
    {
      "epoch": 18.57966457023061,
      "grad_norm": 0.12639573216438293,
      "learning_rate": 1.5295730746209103e-05,
      "loss": 0.4781,
      "num_input_tokens_seen": 23124176,
      "step": 35450
    },
    {
      "epoch": 18.582285115303982,
      "grad_norm": 0.14307278394699097,
      "learning_rate": 1.5239649896485463e-05,
      "loss": 0.378,
      "num_input_tokens_seen": 23126832,
      "step": 35455
    },
    {
      "epoch": 18.58490566037736,
      "grad_norm": 0.10959102213382721,
      "learning_rate": 1.5183670453096598e-05,
      "loss": 0.4183,
      "num_input_tokens_seen": 23129904,
      "step": 35460
    },
    {
      "epoch": 18.587526205450732,
      "grad_norm": 0.11696252226829529,
      "learning_rate": 1.5127792427752696e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 23133200,
      "step": 35465
    },
    {
      "epoch": 18.59014675052411,
      "grad_norm": 0.21598637104034424,
      "learning_rate": 1.507201583214296e-05,
      "loss": 0.3552,
      "num_input_tokens_seen": 23136496,
      "step": 35470
    },
    {
      "epoch": 18.592767295597483,
      "grad_norm": 0.17139574885368347,
      "learning_rate": 1.5016340677935169e-05,
      "loss": 0.349,
      "num_input_tokens_seen": 23139472,
      "step": 35475
    },
    {
      "epoch": 18.59538784067086,
      "grad_norm": 0.21488995850086212,
      "learning_rate": 1.4960766976775953e-05,
      "loss": 0.3955,
      "num_input_tokens_seen": 23142384,
      "step": 35480
    },
    {
      "epoch": 18.598008385744233,
      "grad_norm": 0.14198674261569977,
      "learning_rate": 1.4905294740290677e-05,
      "loss": 0.4373,
      "num_input_tokens_seen": 23145328,
      "step": 35485
    },
    {
      "epoch": 18.60062893081761,
      "grad_norm": 0.2269202619791031,
      "learning_rate": 1.4849923980083391e-05,
      "loss": 0.4296,
      "num_input_tokens_seen": 23148912,
      "step": 35490
    },
    {
      "epoch": 18.603249475890987,
      "grad_norm": 0.21772244572639465,
      "learning_rate": 1.4794654707737164e-05,
      "loss": 0.4428,
      "num_input_tokens_seen": 23151632,
      "step": 35495
    },
    {
      "epoch": 18.60587002096436,
      "grad_norm": 0.17154867947101593,
      "learning_rate": 1.4739486934813696e-05,
      "loss": 0.3877,
      "num_input_tokens_seen": 23154032,
      "step": 35500
    },
    {
      "epoch": 18.608490566037737,
      "grad_norm": 0.1361662596464157,
      "learning_rate": 1.468442067285336e-05,
      "loss": 0.4348,
      "num_input_tokens_seen": 23156976,
      "step": 35505
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 0.18370738625526428,
      "learning_rate": 1.4629455933375501e-05,
      "loss": 0.4224,
      "num_input_tokens_seen": 23159472,
      "step": 35510
    },
    {
      "epoch": 18.613731656184488,
      "grad_norm": 0.14899003505706787,
      "learning_rate": 1.4574592727878089e-05,
      "loss": 0.5039,
      "num_input_tokens_seen": 23162672,
      "step": 35515
    },
    {
      "epoch": 18.61635220125786,
      "grad_norm": 0.18513897061347961,
      "learning_rate": 1.4519831067837774e-05,
      "loss": 0.4714,
      "num_input_tokens_seen": 23165456,
      "step": 35520
    },
    {
      "epoch": 18.618972746331238,
      "grad_norm": 0.26495492458343506,
      "learning_rate": 1.4465170964710172e-05,
      "loss": 0.4144,
      "num_input_tokens_seen": 23168656,
      "step": 35525
    },
    {
      "epoch": 18.62159329140461,
      "grad_norm": 0.17430385947227478,
      "learning_rate": 1.441061242992947e-05,
      "loss": 0.4298,
      "num_input_tokens_seen": 23171856,
      "step": 35530
    },
    {
      "epoch": 18.62421383647799,
      "grad_norm": 0.14890186488628387,
      "learning_rate": 1.4356155474908871e-05,
      "loss": 0.3604,
      "num_input_tokens_seen": 23175856,
      "step": 35535
    },
    {
      "epoch": 18.62683438155136,
      "grad_norm": 0.161901593208313,
      "learning_rate": 1.4301800111039986e-05,
      "loss": 0.4788,
      "num_input_tokens_seen": 23178928,
      "step": 35540
    },
    {
      "epoch": 18.62945492662474,
      "grad_norm": 0.13783320784568787,
      "learning_rate": 1.4247546349693386e-05,
      "loss": 0.472,
      "num_input_tokens_seen": 23182384,
      "step": 35545
    },
    {
      "epoch": 18.632075471698112,
      "grad_norm": 0.15049658715724945,
      "learning_rate": 1.4193394202218268e-05,
      "loss": 0.2986,
      "num_input_tokens_seen": 23186192,
      "step": 35550
    },
    {
      "epoch": 18.63469601677149,
      "grad_norm": 0.1212732270359993,
      "learning_rate": 1.413934367994274e-05,
      "loss": 0.4596,
      "num_input_tokens_seen": 23191056,
      "step": 35555
    },
    {
      "epoch": 18.637316561844862,
      "grad_norm": 0.19104504585266113,
      "learning_rate": 1.408539479417359e-05,
      "loss": 0.4185,
      "num_input_tokens_seen": 23194160,
      "step": 35560
    },
    {
      "epoch": 18.63993710691824,
      "grad_norm": 0.15363502502441406,
      "learning_rate": 1.4031547556196178e-05,
      "loss": 0.3927,
      "num_input_tokens_seen": 23197392,
      "step": 35565
    },
    {
      "epoch": 18.642557651991613,
      "grad_norm": 0.13942040503025055,
      "learning_rate": 1.3977801977274828e-05,
      "loss": 0.3783,
      "num_input_tokens_seen": 23201008,
      "step": 35570
    },
    {
      "epoch": 18.64517819706499,
      "grad_norm": 0.10542018711566925,
      "learning_rate": 1.3924158068652437e-05,
      "loss": 0.3483,
      "num_input_tokens_seen": 23208176,
      "step": 35575
    },
    {
      "epoch": 18.647798742138363,
      "grad_norm": 0.15469114482402802,
      "learning_rate": 1.3870615841550693e-05,
      "loss": 0.4237,
      "num_input_tokens_seen": 23210768,
      "step": 35580
    },
    {
      "epoch": 18.65041928721174,
      "grad_norm": 0.14723636209964752,
      "learning_rate": 1.3817175307170138e-05,
      "loss": 0.655,
      "num_input_tokens_seen": 23214480,
      "step": 35585
    },
    {
      "epoch": 18.653039832285117,
      "grad_norm": 0.3848152458667755,
      "learning_rate": 1.3763836476689828e-05,
      "loss": 0.38,
      "num_input_tokens_seen": 23217296,
      "step": 35590
    },
    {
      "epoch": 18.65566037735849,
      "grad_norm": 0.14407195150852203,
      "learning_rate": 1.3710599361267617e-05,
      "loss": 0.3501,
      "num_input_tokens_seen": 23220272,
      "step": 35595
    },
    {
      "epoch": 18.658280922431867,
      "grad_norm": 0.1187802404165268,
      "learning_rate": 1.3657463972040207e-05,
      "loss": 0.3514,
      "num_input_tokens_seen": 23224912,
      "step": 35600
    },
    {
      "epoch": 18.66090146750524,
      "grad_norm": 0.13565626740455627,
      "learning_rate": 1.3604430320122762e-05,
      "loss": 0.3482,
      "num_input_tokens_seen": 23227600,
      "step": 35605
    },
    {
      "epoch": 18.663522012578618,
      "grad_norm": 0.2465268075466156,
      "learning_rate": 1.355149841660941e-05,
      "loss": 0.5449,
      "num_input_tokens_seen": 23230576,
      "step": 35610
    },
    {
      "epoch": 18.66614255765199,
      "grad_norm": 0.22550685703754425,
      "learning_rate": 1.3498668272572955e-05,
      "loss": 0.5222,
      "num_input_tokens_seen": 23233360,
      "step": 35615
    },
    {
      "epoch": 18.668763102725368,
      "grad_norm": 0.11687405407428741,
      "learning_rate": 1.3445939899064729e-05,
      "loss": 0.6085,
      "num_input_tokens_seen": 23237328,
      "step": 35620
    },
    {
      "epoch": 18.67138364779874,
      "grad_norm": 0.1569223552942276,
      "learning_rate": 1.3393313307115019e-05,
      "loss": 0.3555,
      "num_input_tokens_seen": 23240592,
      "step": 35625
    },
    {
      "epoch": 18.67400419287212,
      "grad_norm": 0.10969287902116776,
      "learning_rate": 1.3340788507732626e-05,
      "loss": 0.4175,
      "num_input_tokens_seen": 23243472,
      "step": 35630
    },
    {
      "epoch": 18.67662473794549,
      "grad_norm": 0.11299540847539902,
      "learning_rate": 1.328836551190521e-05,
      "loss": 0.3939,
      "num_input_tokens_seen": 23246032,
      "step": 35635
    },
    {
      "epoch": 18.67924528301887,
      "grad_norm": 0.10294940322637558,
      "learning_rate": 1.323604433059905e-05,
      "loss": 0.3871,
      "num_input_tokens_seen": 23248304,
      "step": 35640
    },
    {
      "epoch": 18.681865828092242,
      "grad_norm": 0.14268240332603455,
      "learning_rate": 1.3183824974759063e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 23251632,
      "step": 35645
    },
    {
      "epoch": 18.68448637316562,
      "grad_norm": 0.15562033653259277,
      "learning_rate": 1.3131707455309006e-05,
      "loss": 0.3709,
      "num_input_tokens_seen": 23255216,
      "step": 35650
    },
    {
      "epoch": 18.687106918238992,
      "grad_norm": 0.22195208072662354,
      "learning_rate": 1.307969178315127e-05,
      "loss": 0.5033,
      "num_input_tokens_seen": 23258736,
      "step": 35655
    },
    {
      "epoch": 18.68972746331237,
      "grad_norm": 0.15552395582199097,
      "learning_rate": 1.3027777969166932e-05,
      "loss": 0.3184,
      "num_input_tokens_seen": 23261872,
      "step": 35660
    },
    {
      "epoch": 18.692348008385743,
      "grad_norm": 0.11034548282623291,
      "learning_rate": 1.2975966024215746e-05,
      "loss": 0.5256,
      "num_input_tokens_seen": 23265008,
      "step": 35665
    },
    {
      "epoch": 18.69496855345912,
      "grad_norm": 0.12731851637363434,
      "learning_rate": 1.2924255959136267e-05,
      "loss": 0.3927,
      "num_input_tokens_seen": 23268912,
      "step": 35670
    },
    {
      "epoch": 18.697589098532493,
      "grad_norm": 0.1985291689634323,
      "learning_rate": 1.2872647784745561e-05,
      "loss": 0.2467,
      "num_input_tokens_seen": 23271344,
      "step": 35675
    },
    {
      "epoch": 18.70020964360587,
      "grad_norm": 0.09715043008327484,
      "learning_rate": 1.282114151183944e-05,
      "loss": 0.4069,
      "num_input_tokens_seen": 23274960,
      "step": 35680
    },
    {
      "epoch": 18.702830188679247,
      "grad_norm": 0.2058853656053543,
      "learning_rate": 1.2769737151192562e-05,
      "loss": 0.4416,
      "num_input_tokens_seen": 23278160,
      "step": 35685
    },
    {
      "epoch": 18.70545073375262,
      "grad_norm": 0.0992412194609642,
      "learning_rate": 1.2718434713558047e-05,
      "loss": 0.4615,
      "num_input_tokens_seen": 23282192,
      "step": 35690
    },
    {
      "epoch": 18.708071278825997,
      "grad_norm": 0.1431523859500885,
      "learning_rate": 1.2667234209667755e-05,
      "loss": 0.3885,
      "num_input_tokens_seen": 23285648,
      "step": 35695
    },
    {
      "epoch": 18.71069182389937,
      "grad_norm": 0.3024989068508148,
      "learning_rate": 1.2616135650232286e-05,
      "loss": 0.6108,
      "num_input_tokens_seen": 23288304,
      "step": 35700
    },
    {
      "epoch": 18.713312368972748,
      "grad_norm": 0.15091446042060852,
      "learning_rate": 1.2565139045940866e-05,
      "loss": 0.384,
      "num_input_tokens_seen": 23291408,
      "step": 35705
    },
    {
      "epoch": 18.71593291404612,
      "grad_norm": 0.1409914493560791,
      "learning_rate": 1.2514244407461462e-05,
      "loss": 0.4041,
      "num_input_tokens_seen": 23294448,
      "step": 35710
    },
    {
      "epoch": 18.718553459119498,
      "grad_norm": 0.1561100035905838,
      "learning_rate": 1.2463451745440501e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 23297232,
      "step": 35715
    },
    {
      "epoch": 18.72117400419287,
      "grad_norm": 0.19882787764072418,
      "learning_rate": 1.241276107050343e-05,
      "loss": 0.4629,
      "num_input_tokens_seen": 23301264,
      "step": 35720
    },
    {
      "epoch": 18.72379454926625,
      "grad_norm": 0.16173101961612701,
      "learning_rate": 1.236217239325399e-05,
      "loss": 0.4489,
      "num_input_tokens_seen": 23304720,
      "step": 35725
    },
    {
      "epoch": 18.72641509433962,
      "grad_norm": 0.08085206896066666,
      "learning_rate": 1.2311685724274768e-05,
      "loss": 0.4643,
      "num_input_tokens_seen": 23308592,
      "step": 35730
    },
    {
      "epoch": 18.729035639413,
      "grad_norm": 0.18408246338367462,
      "learning_rate": 1.2261301074127096e-05,
      "loss": 0.8562,
      "num_input_tokens_seen": 23312080,
      "step": 35735
    },
    {
      "epoch": 18.731656184486372,
      "grad_norm": 0.2089003473520279,
      "learning_rate": 1.2211018453350874e-05,
      "loss": 0.6608,
      "num_input_tokens_seen": 23314800,
      "step": 35740
    },
    {
      "epoch": 18.73427672955975,
      "grad_norm": 0.1339966058731079,
      "learning_rate": 1.2160837872464581e-05,
      "loss": 0.5304,
      "num_input_tokens_seen": 23317840,
      "step": 35745
    },
    {
      "epoch": 18.736897274633122,
      "grad_norm": 0.3155742883682251,
      "learning_rate": 1.2110759341965428e-05,
      "loss": 0.4351,
      "num_input_tokens_seen": 23321200,
      "step": 35750
    },
    {
      "epoch": 18.7395178197065,
      "grad_norm": 0.2328176647424698,
      "learning_rate": 1.2060782872329256e-05,
      "loss": 0.5064,
      "num_input_tokens_seen": 23323824,
      "step": 35755
    },
    {
      "epoch": 18.742138364779873,
      "grad_norm": 0.24447116255760193,
      "learning_rate": 1.2010908474010595e-05,
      "loss": 0.4459,
      "num_input_tokens_seen": 23326928,
      "step": 35760
    },
    {
      "epoch": 18.74475890985325,
      "grad_norm": 0.18438993394374847,
      "learning_rate": 1.1961136157442654e-05,
      "loss": 0.5196,
      "num_input_tokens_seen": 23330576,
      "step": 35765
    },
    {
      "epoch": 18.747379454926623,
      "grad_norm": 0.10710815340280533,
      "learning_rate": 1.1911465933037214e-05,
      "loss": 0.4839,
      "num_input_tokens_seen": 23334832,
      "step": 35770
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.14147137105464935,
      "learning_rate": 1.1861897811184686e-05,
      "loss": 0.3525,
      "num_input_tokens_seen": 23337488,
      "step": 35775
    },
    {
      "epoch": 18.752620545073377,
      "grad_norm": 0.15045952796936035,
      "learning_rate": 1.1812431802254109e-05,
      "loss": 0.3354,
      "num_input_tokens_seen": 23341296,
      "step": 35780
    },
    {
      "epoch": 18.75524109014675,
      "grad_norm": 0.21089282631874084,
      "learning_rate": 1.1763067916593262e-05,
      "loss": 0.4981,
      "num_input_tokens_seen": 23348144,
      "step": 35785
    },
    {
      "epoch": 18.757861635220127,
      "grad_norm": 0.15439800918102264,
      "learning_rate": 1.1713806164528496e-05,
      "loss": 0.3961,
      "num_input_tokens_seen": 23352080,
      "step": 35790
    },
    {
      "epoch": 18.7604821802935,
      "grad_norm": 0.1404285728931427,
      "learning_rate": 1.1664646556364844e-05,
      "loss": 0.2556,
      "num_input_tokens_seen": 23355536,
      "step": 35795
    },
    {
      "epoch": 18.763102725366878,
      "grad_norm": 0.1799292415380478,
      "learning_rate": 1.161558910238597e-05,
      "loss": 0.3476,
      "num_input_tokens_seen": 23358160,
      "step": 35800
    },
    {
      "epoch": 18.76572327044025,
      "grad_norm": 0.14913471043109894,
      "learning_rate": 1.1566633812854e-05,
      "loss": 0.5274,
      "num_input_tokens_seen": 23361296,
      "step": 35805
    },
    {
      "epoch": 18.768343815513628,
      "grad_norm": 0.1388961523771286,
      "learning_rate": 1.151778069800985e-05,
      "loss": 0.4236,
      "num_input_tokens_seen": 23365264,
      "step": 35810
    },
    {
      "epoch": 18.770964360587,
      "grad_norm": 0.17331549525260925,
      "learning_rate": 1.1469029768073125e-05,
      "loss": 0.3163,
      "num_input_tokens_seen": 23368208,
      "step": 35815
    },
    {
      "epoch": 18.77358490566038,
      "grad_norm": 0.2037079781293869,
      "learning_rate": 1.1420381033241889e-05,
      "loss": 0.4608,
      "num_input_tokens_seen": 23371248,
      "step": 35820
    },
    {
      "epoch": 18.77620545073375,
      "grad_norm": 0.16713646054267883,
      "learning_rate": 1.1371834503693002e-05,
      "loss": 0.3427,
      "num_input_tokens_seen": 23374960,
      "step": 35825
    },
    {
      "epoch": 18.77882599580713,
      "grad_norm": 0.12153371423482895,
      "learning_rate": 1.1323390189581784e-05,
      "loss": 0.3504,
      "num_input_tokens_seen": 23378832,
      "step": 35830
    },
    {
      "epoch": 18.781446540880502,
      "grad_norm": 0.22639180719852448,
      "learning_rate": 1.127504810104213e-05,
      "loss": 0.514,
      "num_input_tokens_seen": 23382032,
      "step": 35835
    },
    {
      "epoch": 18.78406708595388,
      "grad_norm": 0.1389073133468628,
      "learning_rate": 1.1226808248186782e-05,
      "loss": 0.3865,
      "num_input_tokens_seen": 23385456,
      "step": 35840
    },
    {
      "epoch": 18.786687631027252,
      "grad_norm": 0.13668517768383026,
      "learning_rate": 1.1178670641106891e-05,
      "loss": 0.4484,
      "num_input_tokens_seen": 23388592,
      "step": 35845
    },
    {
      "epoch": 18.78930817610063,
      "grad_norm": 0.12234590202569962,
      "learning_rate": 1.1130635289872403e-05,
      "loss": 0.4082,
      "num_input_tokens_seen": 23391824,
      "step": 35850
    },
    {
      "epoch": 18.791928721174003,
      "grad_norm": 0.18029604852199554,
      "learning_rate": 1.1082702204531725e-05,
      "loss": 0.4023,
      "num_input_tokens_seen": 23394896,
      "step": 35855
    },
    {
      "epoch": 18.79454926624738,
      "grad_norm": 0.19807368516921997,
      "learning_rate": 1.1034871395111778e-05,
      "loss": 0.3763,
      "num_input_tokens_seen": 23399888,
      "step": 35860
    },
    {
      "epoch": 18.797169811320753,
      "grad_norm": 0.1694711595773697,
      "learning_rate": 1.0987142871618394e-05,
      "loss": 0.3961,
      "num_input_tokens_seen": 23402832,
      "step": 35865
    },
    {
      "epoch": 18.79979035639413,
      "grad_norm": 0.13666170835494995,
      "learning_rate": 1.0939516644035696e-05,
      "loss": 0.5302,
      "num_input_tokens_seen": 23405968,
      "step": 35870
    },
    {
      "epoch": 18.802410901467507,
      "grad_norm": 0.16181814670562744,
      "learning_rate": 1.0891992722326716e-05,
      "loss": 0.5617,
      "num_input_tokens_seen": 23408656,
      "step": 35875
    },
    {
      "epoch": 18.80503144654088,
      "grad_norm": 0.16426897048950195,
      "learning_rate": 1.0844571116432778e-05,
      "loss": 0.59,
      "num_input_tokens_seen": 23411952,
      "step": 35880
    },
    {
      "epoch": 18.807651991614257,
      "grad_norm": 0.23400206863880157,
      "learning_rate": 1.0797251836274003e-05,
      "loss": 0.4504,
      "num_input_tokens_seen": 23414896,
      "step": 35885
    },
    {
      "epoch": 18.81027253668763,
      "grad_norm": 0.17571599781513214,
      "learning_rate": 1.0750034891748972e-05,
      "loss": 0.5086,
      "num_input_tokens_seen": 23418224,
      "step": 35890
    },
    {
      "epoch": 18.812893081761008,
      "grad_norm": 0.22916871309280396,
      "learning_rate": 1.0702920292735009e-05,
      "loss": 0.3793,
      "num_input_tokens_seen": 23421008,
      "step": 35895
    },
    {
      "epoch": 18.81551362683438,
      "grad_norm": 0.10953430831432343,
      "learning_rate": 1.0655908049087893e-05,
      "loss": 0.3846,
      "num_input_tokens_seen": 23423792,
      "step": 35900
    },
    {
      "epoch": 18.818134171907758,
      "grad_norm": 0.15694624185562134,
      "learning_rate": 1.0608998170642149e-05,
      "loss": 0.4634,
      "num_input_tokens_seen": 23426320,
      "step": 35905
    },
    {
      "epoch": 18.82075471698113,
      "grad_norm": 0.1807015836238861,
      "learning_rate": 1.0562190667210703e-05,
      "loss": 0.4991,
      "num_input_tokens_seen": 23429840,
      "step": 35910
    },
    {
      "epoch": 18.82337526205451,
      "grad_norm": 0.3495023250579834,
      "learning_rate": 1.0515485548585113e-05,
      "loss": 0.3364,
      "num_input_tokens_seen": 23432720,
      "step": 35915
    },
    {
      "epoch": 18.82599580712788,
      "grad_norm": 0.10924609005451202,
      "learning_rate": 1.0468882824535676e-05,
      "loss": 0.3472,
      "num_input_tokens_seen": 23436592,
      "step": 35920
    },
    {
      "epoch": 18.82861635220126,
      "grad_norm": 0.3456616699695587,
      "learning_rate": 1.0422382504811034e-05,
      "loss": 0.3523,
      "num_input_tokens_seen": 23439856,
      "step": 35925
    },
    {
      "epoch": 18.831236897274632,
      "grad_norm": 0.14582547545433044,
      "learning_rate": 1.0375984599138633e-05,
      "loss": 0.3457,
      "num_input_tokens_seen": 23442832,
      "step": 35930
    },
    {
      "epoch": 18.83385744234801,
      "grad_norm": 0.1236053928732872,
      "learning_rate": 1.0329689117224261e-05,
      "loss": 0.3674,
      "num_input_tokens_seen": 23446416,
      "step": 35935
    },
    {
      "epoch": 18.836477987421382,
      "grad_norm": 0.1636594533920288,
      "learning_rate": 1.0283496068752507e-05,
      "loss": 0.399,
      "num_input_tokens_seen": 23449104,
      "step": 35940
    },
    {
      "epoch": 18.83909853249476,
      "grad_norm": 0.15776365995407104,
      "learning_rate": 1.0237405463386418e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 23452272,
      "step": 35945
    },
    {
      "epoch": 18.841719077568133,
      "grad_norm": 0.22129185497760773,
      "learning_rate": 1.0191417310767503e-05,
      "loss": 0.4437,
      "num_input_tokens_seen": 23455088,
      "step": 35950
    },
    {
      "epoch": 18.84433962264151,
      "grad_norm": 0.15562233328819275,
      "learning_rate": 1.0145531620516179e-05,
      "loss": 0.4573,
      "num_input_tokens_seen": 23458704,
      "step": 35955
    },
    {
      "epoch": 18.846960167714883,
      "grad_norm": 0.10578373074531555,
      "learning_rate": 1.009974840223099e-05,
      "loss": 0.4269,
      "num_input_tokens_seen": 23462704,
      "step": 35960
    },
    {
      "epoch": 18.84958071278826,
      "grad_norm": 0.14477699995040894,
      "learning_rate": 1.0054067665489386e-05,
      "loss": 0.3427,
      "num_input_tokens_seen": 23465328,
      "step": 35965
    },
    {
      "epoch": 18.852201257861637,
      "grad_norm": 0.150782510638237,
      "learning_rate": 1.0008489419847278e-05,
      "loss": 0.4781,
      "num_input_tokens_seen": 23468912,
      "step": 35970
    },
    {
      "epoch": 18.85482180293501,
      "grad_norm": 0.1264631748199463,
      "learning_rate": 9.963013674839038e-06,
      "loss": 0.3724,
      "num_input_tokens_seen": 23471984,
      "step": 35975
    },
    {
      "epoch": 18.857442348008387,
      "grad_norm": 0.15364572405815125,
      "learning_rate": 9.91764043997767e-06,
      "loss": 0.3482,
      "num_input_tokens_seen": 23475888,
      "step": 35980
    },
    {
      "epoch": 18.86006289308176,
      "grad_norm": 0.16153918206691742,
      "learning_rate": 9.872369724754804e-06,
      "loss": 0.3959,
      "num_input_tokens_seen": 23478704,
      "step": 35985
    },
    {
      "epoch": 18.862683438155138,
      "grad_norm": 0.27794530987739563,
      "learning_rate": 9.827201538640473e-06,
      "loss": 0.3791,
      "num_input_tokens_seen": 23481136,
      "step": 35990
    },
    {
      "epoch": 18.86530398322851,
      "grad_norm": 0.1447695642709732,
      "learning_rate": 9.782135891083455e-06,
      "loss": 0.3275,
      "num_input_tokens_seen": 23483408,
      "step": 35995
    },
    {
      "epoch": 18.867924528301888,
      "grad_norm": 0.13563445210456848,
      "learning_rate": 9.737172791510873e-06,
      "loss": 0.4428,
      "num_input_tokens_seen": 23487568,
      "step": 36000
    },
    {
      "epoch": 18.87054507337526,
      "grad_norm": 0.15855486690998077,
      "learning_rate": 9.692312249328483e-06,
      "loss": 0.486,
      "num_input_tokens_seen": 23490512,
      "step": 36005
    },
    {
      "epoch": 18.87316561844864,
      "grad_norm": 0.22668346762657166,
      "learning_rate": 9.647554273920722e-06,
      "loss": 0.4898,
      "num_input_tokens_seen": 23493680,
      "step": 36010
    },
    {
      "epoch": 18.87578616352201,
      "grad_norm": 0.15781551599502563,
      "learning_rate": 9.602898874650323e-06,
      "loss": 0.4379,
      "num_input_tokens_seen": 23497680,
      "step": 36015
    },
    {
      "epoch": 18.87840670859539,
      "grad_norm": 0.14633765816688538,
      "learning_rate": 9.558346060858759e-06,
      "loss": 0.4638,
      "num_input_tokens_seen": 23500464,
      "step": 36020
    },
    {
      "epoch": 18.881027253668762,
      "grad_norm": 0.21105468273162842,
      "learning_rate": 9.51389584186596e-06,
      "loss": 0.3628,
      "num_input_tokens_seen": 23503184,
      "step": 36025
    },
    {
      "epoch": 18.88364779874214,
      "grad_norm": 0.19060726463794708,
      "learning_rate": 9.469548226970326e-06,
      "loss": 0.4666,
      "num_input_tokens_seen": 23506064,
      "step": 36030
    },
    {
      "epoch": 18.886268343815512,
      "grad_norm": 0.15158161520957947,
      "learning_rate": 9.425303225448989e-06,
      "loss": 0.4821,
      "num_input_tokens_seen": 23508848,
      "step": 36035
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 0.22663912177085876,
      "learning_rate": 9.381160846557435e-06,
      "loss": 0.4835,
      "num_input_tokens_seen": 23512432,
      "step": 36040
    },
    {
      "epoch": 18.891509433962263,
      "grad_norm": 0.5198838710784912,
      "learning_rate": 9.337121099529722e-06,
      "loss": 0.5042,
      "num_input_tokens_seen": 23515312,
      "step": 36045
    },
    {
      "epoch": 18.89412997903564,
      "grad_norm": 0.16166867315769196,
      "learning_rate": 9.293183993578535e-06,
      "loss": 0.4751,
      "num_input_tokens_seen": 23519024,
      "step": 36050
    },
    {
      "epoch": 18.896750524109013,
      "grad_norm": 0.13990768790245056,
      "learning_rate": 9.249349537894968e-06,
      "loss": 0.3286,
      "num_input_tokens_seen": 23522032,
      "step": 36055
    },
    {
      "epoch": 18.89937106918239,
      "grad_norm": 0.25880008935928345,
      "learning_rate": 9.205617741648686e-06,
      "loss": 0.5296,
      "num_input_tokens_seen": 23524912,
      "step": 36060
    },
    {
      "epoch": 18.901991614255767,
      "grad_norm": 0.14834266901016235,
      "learning_rate": 9.161988613987982e-06,
      "loss": 0.3009,
      "num_input_tokens_seen": 23527984,
      "step": 36065
    },
    {
      "epoch": 18.90461215932914,
      "grad_norm": 0.11460419744253159,
      "learning_rate": 9.118462164039387e-06,
      "loss": 0.4853,
      "num_input_tokens_seen": 23531792,
      "step": 36070
    },
    {
      "epoch": 18.907232704402517,
      "grad_norm": 0.13937945663928986,
      "learning_rate": 9.07503840090823e-06,
      "loss": 0.415,
      "num_input_tokens_seen": 23535504,
      "step": 36075
    },
    {
      "epoch": 18.90985324947589,
      "grad_norm": 0.11538148671388626,
      "learning_rate": 9.031717333678303e-06,
      "loss": 0.4045,
      "num_input_tokens_seen": 23538864,
      "step": 36080
    },
    {
      "epoch": 18.912473794549268,
      "grad_norm": 0.20038676261901855,
      "learning_rate": 8.988498971411851e-06,
      "loss": 0.4879,
      "num_input_tokens_seen": 23542736,
      "step": 36085
    },
    {
      "epoch": 18.91509433962264,
      "grad_norm": 0.1364116370677948,
      "learning_rate": 8.945383323149647e-06,
      "loss": 0.4748,
      "num_input_tokens_seen": 23547504,
      "step": 36090
    },
    {
      "epoch": 18.917714884696018,
      "grad_norm": 0.3032207190990448,
      "learning_rate": 8.902370397911031e-06,
      "loss": 0.3808,
      "num_input_tokens_seen": 23550224,
      "step": 36095
    },
    {
      "epoch": 18.92033542976939,
      "grad_norm": 0.11618948727846146,
      "learning_rate": 8.859460204693748e-06,
      "loss": 0.5029,
      "num_input_tokens_seen": 23553488,
      "step": 36100
    },
    {
      "epoch": 18.92295597484277,
      "grad_norm": 0.1477808803319931,
      "learning_rate": 8.816652752474175e-06,
      "loss": 0.5101,
      "num_input_tokens_seen": 23556816,
      "step": 36105
    },
    {
      "epoch": 18.92557651991614,
      "grad_norm": 0.20151996612548828,
      "learning_rate": 8.773948050207148e-06,
      "loss": 0.3153,
      "num_input_tokens_seen": 23559280,
      "step": 36110
    },
    {
      "epoch": 18.92819706498952,
      "grad_norm": 0.14749880135059357,
      "learning_rate": 8.731346106826021e-06,
      "loss": 0.5044,
      "num_input_tokens_seen": 23562416,
      "step": 36115
    },
    {
      "epoch": 18.930817610062892,
      "grad_norm": 0.1437956541776657,
      "learning_rate": 8.6888469312425e-06,
      "loss": 0.3342,
      "num_input_tokens_seen": 23566032,
      "step": 36120
    },
    {
      "epoch": 18.93343815513627,
      "grad_norm": 0.1186174750328064,
      "learning_rate": 8.646450532347083e-06,
      "loss": 0.484,
      "num_input_tokens_seen": 23568880,
      "step": 36125
    },
    {
      "epoch": 18.936058700209642,
      "grad_norm": 0.14700326323509216,
      "learning_rate": 8.604156919008565e-06,
      "loss": 0.3792,
      "num_input_tokens_seen": 23571216,
      "step": 36130
    },
    {
      "epoch": 18.93867924528302,
      "grad_norm": 0.1169910579919815,
      "learning_rate": 8.561966100074258e-06,
      "loss": 0.5367,
      "num_input_tokens_seen": 23575216,
      "step": 36135
    },
    {
      "epoch": 18.941299790356393,
      "grad_norm": 0.2410590946674347,
      "learning_rate": 8.519878084370048e-06,
      "loss": 0.401,
      "num_input_tokens_seen": 23578576,
      "step": 36140
    },
    {
      "epoch": 18.94392033542977,
      "grad_norm": 0.12715834379196167,
      "learning_rate": 8.477892880700222e-06,
      "loss": 0.3816,
      "num_input_tokens_seen": 23580848,
      "step": 36145
    },
    {
      "epoch": 18.946540880503143,
      "grad_norm": 0.15721352398395538,
      "learning_rate": 8.436010497847646e-06,
      "loss": 0.5002,
      "num_input_tokens_seen": 23583440,
      "step": 36150
    },
    {
      "epoch": 18.94916142557652,
      "grad_norm": 0.1470145285129547,
      "learning_rate": 8.39423094457359e-06,
      "loss": 0.3324,
      "num_input_tokens_seen": 23586640,
      "step": 36155
    },
    {
      "epoch": 18.951781970649897,
      "grad_norm": 0.17082619667053223,
      "learning_rate": 8.352554229617892e-06,
      "loss": 0.3794,
      "num_input_tokens_seen": 23589968,
      "step": 36160
    },
    {
      "epoch": 18.95440251572327,
      "grad_norm": 0.19889923930168152,
      "learning_rate": 8.310980361698861e-06,
      "loss": 0.3504,
      "num_input_tokens_seen": 23592368,
      "step": 36165
    },
    {
      "epoch": 18.957023060796647,
      "grad_norm": 0.27517959475517273,
      "learning_rate": 8.26950934951326e-06,
      "loss": 0.3765,
      "num_input_tokens_seen": 23595248,
      "step": 36170
    },
    {
      "epoch": 18.95964360587002,
      "grad_norm": 0.17953389883041382,
      "learning_rate": 8.22814120173626e-06,
      "loss": 0.4357,
      "num_input_tokens_seen": 23598384,
      "step": 36175
    },
    {
      "epoch": 18.962264150943398,
      "grad_norm": 0.10081231594085693,
      "learning_rate": 8.186875927021775e-06,
      "loss": 0.481,
      "num_input_tokens_seen": 23602128,
      "step": 36180
    },
    {
      "epoch": 18.96488469601677,
      "grad_norm": 0.08780104666948318,
      "learning_rate": 8.145713534001897e-06,
      "loss": 0.316,
      "num_input_tokens_seen": 23605232,
      "step": 36185
    },
    {
      "epoch": 18.967505241090148,
      "grad_norm": 0.22651220858097076,
      "learning_rate": 8.104654031287406e-06,
      "loss": 0.3828,
      "num_input_tokens_seen": 23608272,
      "step": 36190
    },
    {
      "epoch": 18.97012578616352,
      "grad_norm": 0.188573956489563,
      "learning_rate": 8.063697427467486e-06,
      "loss": 0.3777,
      "num_input_tokens_seen": 23610576,
      "step": 36195
    },
    {
      "epoch": 18.9727463312369,
      "grad_norm": 0.17426425218582153,
      "learning_rate": 8.022843731109675e-06,
      "loss": 0.4814,
      "num_input_tokens_seen": 23614000,
      "step": 36200
    },
    {
      "epoch": 18.97536687631027,
      "grad_norm": 0.1278482973575592,
      "learning_rate": 7.982092950760245e-06,
      "loss": 0.3688,
      "num_input_tokens_seen": 23617264,
      "step": 36205
    },
    {
      "epoch": 18.97798742138365,
      "grad_norm": 0.17091882228851318,
      "learning_rate": 7.941445094943711e-06,
      "loss": 0.4542,
      "num_input_tokens_seen": 23620240,
      "step": 36210
    },
    {
      "epoch": 18.980607966457022,
      "grad_norm": 0.12938013672828674,
      "learning_rate": 7.900900172163107e-06,
      "loss": 0.3813,
      "num_input_tokens_seen": 23623216,
      "step": 36215
    },
    {
      "epoch": 18.9832285115304,
      "grad_norm": 0.276632159948349,
      "learning_rate": 7.860458190900144e-06,
      "loss": 0.3715,
      "num_input_tokens_seen": 23626064,
      "step": 36220
    },
    {
      "epoch": 18.985849056603772,
      "grad_norm": 0.16482488811016083,
      "learning_rate": 7.820119159614669e-06,
      "loss": 0.4644,
      "num_input_tokens_seen": 23629360,
      "step": 36225
    },
    {
      "epoch": 18.98846960167715,
      "grad_norm": 0.205451101064682,
      "learning_rate": 7.779883086745098e-06,
      "loss": 0.4304,
      "num_input_tokens_seen": 23631792,
      "step": 36230
    },
    {
      "epoch": 18.991090146750523,
      "grad_norm": 0.10335464030504227,
      "learning_rate": 7.739749980708533e-06,
      "loss": 0.3842,
      "num_input_tokens_seen": 23636400,
      "step": 36235
    },
    {
      "epoch": 18.9937106918239,
      "grad_norm": 0.20487594604492188,
      "learning_rate": 7.6997198499002e-06,
      "loss": 0.4036,
      "num_input_tokens_seen": 23639056,
      "step": 36240
    },
    {
      "epoch": 18.996331236897273,
      "grad_norm": 0.16909894347190857,
      "learning_rate": 7.659792702694068e-06,
      "loss": 0.3463,
      "num_input_tokens_seen": 23642672,
      "step": 36245
    },
    {
      "epoch": 18.99895178197065,
      "grad_norm": 0.24762073159217834,
      "learning_rate": 7.619968547442346e-06,
      "loss": 0.402,
      "num_input_tokens_seen": 23644848,
      "step": 36250
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.46669089794158936,
      "eval_runtime": 13.5774,
      "eval_samples_per_second": 62.457,
      "eval_steps_per_second": 15.614,
      "num_input_tokens_seen": 23645440,
      "step": 36252
    },
    {
      "epoch": 19.001572327044027,
      "grad_norm": 0.19042640924453735,
      "learning_rate": 7.580247392475926e-06,
      "loss": 0.3034,
      "num_input_tokens_seen": 23647040,
      "step": 36255
    },
    {
      "epoch": 19.0041928721174,
      "grad_norm": 0.15793539583683014,
      "learning_rate": 7.540629246103825e-06,
      "loss": 0.4126,
      "num_input_tokens_seen": 23649760,
      "step": 36260
    },
    {
      "epoch": 19.006813417190777,
      "grad_norm": 0.14914608001708984,
      "learning_rate": 7.501114116613861e-06,
      "loss": 0.3458,
      "num_input_tokens_seen": 23652672,
      "step": 36265
    },
    {
      "epoch": 19.00943396226415,
      "grad_norm": 0.15995018184185028,
      "learning_rate": 7.461702012272087e-06,
      "loss": 0.4088,
      "num_input_tokens_seen": 23655584,
      "step": 36270
    },
    {
      "epoch": 19.012054507337528,
      "grad_norm": 0.1691301017999649,
      "learning_rate": 7.422392941323075e-06,
      "loss": 0.4572,
      "num_input_tokens_seen": 23658496,
      "step": 36275
    },
    {
      "epoch": 19.0146750524109,
      "grad_norm": 0.20682434737682343,
      "learning_rate": 7.383186911989858e-06,
      "loss": 0.3665,
      "num_input_tokens_seen": 23660928,
      "step": 36280
    },
    {
      "epoch": 19.017295597484278,
      "grad_norm": 0.22637610137462616,
      "learning_rate": 7.344083932473822e-06,
      "loss": 0.6045,
      "num_input_tokens_seen": 23664448,
      "step": 36285
    },
    {
      "epoch": 19.01991614255765,
      "grad_norm": 0.11398707330226898,
      "learning_rate": 7.305084010954976e-06,
      "loss": 0.5478,
      "num_input_tokens_seen": 23668128,
      "step": 36290
    },
    {
      "epoch": 19.02253668763103,
      "grad_norm": 0.14556099474430084,
      "learning_rate": 7.26618715559152e-06,
      "loss": 0.4284,
      "num_input_tokens_seen": 23672128,
      "step": 36295
    },
    {
      "epoch": 19.0251572327044,
      "grad_norm": 0.12102926522493362,
      "learning_rate": 7.227393374520386e-06,
      "loss": 0.3928,
      "num_input_tokens_seen": 23675392,
      "step": 36300
    },
    {
      "epoch": 19.02777777777778,
      "grad_norm": 0.1450113207101822,
      "learning_rate": 7.18870267585664e-06,
      "loss": 0.471,
      "num_input_tokens_seen": 23679136,
      "step": 36305
    },
    {
      "epoch": 19.030398322851152,
      "grad_norm": 0.2879179120063782,
      "learning_rate": 7.150115067694085e-06,
      "loss": 0.4341,
      "num_input_tokens_seen": 23681920,
      "step": 36310
    },
    {
      "epoch": 19.03301886792453,
      "grad_norm": 0.10900984704494476,
      "learning_rate": 7.111630558104653e-06,
      "loss": 0.3693,
      "num_input_tokens_seen": 23685824,
      "step": 36315
    },
    {
      "epoch": 19.035639412997902,
      "grad_norm": 0.1599484533071518,
      "learning_rate": 7.073249155138961e-06,
      "loss": 0.4818,
      "num_input_tokens_seen": 23689344,
      "step": 36320
    },
    {
      "epoch": 19.03825995807128,
      "grad_norm": 0.22445368766784668,
      "learning_rate": 7.0349708668259736e-06,
      "loss": 0.3727,
      "num_input_tokens_seen": 23692384,
      "step": 36325
    },
    {
      "epoch": 19.040880503144653,
      "grad_norm": 0.16719357669353485,
      "learning_rate": 6.996795701173009e-06,
      "loss": 0.3992,
      "num_input_tokens_seen": 23695936,
      "step": 36330
    },
    {
      "epoch": 19.04350104821803,
      "grad_norm": 0.1879802644252777,
      "learning_rate": 6.958723666165901e-06,
      "loss": 0.4489,
      "num_input_tokens_seen": 23698848,
      "step": 36335
    },
    {
      "epoch": 19.046121593291403,
      "grad_norm": 0.14649337530136108,
      "learning_rate": 6.9207547697689446e-06,
      "loss": 0.3901,
      "num_input_tokens_seen": 23702464,
      "step": 36340
    },
    {
      "epoch": 19.04874213836478,
      "grad_norm": 0.17741291224956512,
      "learning_rate": 6.882889019924676e-06,
      "loss": 0.3496,
      "num_input_tokens_seen": 23705344,
      "step": 36345
    },
    {
      "epoch": 19.051362683438157,
      "grad_norm": 0.12400217354297638,
      "learning_rate": 6.84512642455426e-06,
      "loss": 0.4499,
      "num_input_tokens_seen": 23708992,
      "step": 36350
    },
    {
      "epoch": 19.05398322851153,
      "grad_norm": 0.16790078580379486,
      "learning_rate": 6.8074669915572095e-06,
      "loss": 0.3377,
      "num_input_tokens_seen": 23711808,
      "step": 36355
    },
    {
      "epoch": 19.056603773584907,
      "grad_norm": 0.11126475781202316,
      "learning_rate": 6.769910728811391e-06,
      "loss": 0.5293,
      "num_input_tokens_seen": 23714720,
      "step": 36360
    },
    {
      "epoch": 19.05922431865828,
      "grad_norm": 0.1739240139722824,
      "learning_rate": 6.73245764417324e-06,
      "loss": 0.5391,
      "num_input_tokens_seen": 23717984,
      "step": 36365
    },
    {
      "epoch": 19.061844863731658,
      "grad_norm": 0.1352824866771698,
      "learning_rate": 6.695107745477435e-06,
      "loss": 0.437,
      "num_input_tokens_seen": 23721184,
      "step": 36370
    },
    {
      "epoch": 19.06446540880503,
      "grad_norm": 0.09799793362617493,
      "learning_rate": 6.657861040537117e-06,
      "loss": 0.4458,
      "num_input_tokens_seen": 23724384,
      "step": 36375
    },
    {
      "epoch": 19.067085953878408,
      "grad_norm": 0.17092002928256989,
      "learning_rate": 6.620717537143994e-06,
      "loss": 0.4599,
      "num_input_tokens_seen": 23727680,
      "step": 36380
    },
    {
      "epoch": 19.06970649895178,
      "grad_norm": 0.29858115315437317,
      "learning_rate": 6.583677243067965e-06,
      "loss": 0.5164,
      "num_input_tokens_seen": 23730464,
      "step": 36385
    },
    {
      "epoch": 19.072327044025158,
      "grad_norm": 0.13133454322814941,
      "learning_rate": 6.546740166057441e-06,
      "loss": 0.3298,
      "num_input_tokens_seen": 23733888,
      "step": 36390
    },
    {
      "epoch": 19.07494758909853,
      "grad_norm": 0.1756727546453476,
      "learning_rate": 6.5099063138392975e-06,
      "loss": 0.4311,
      "num_input_tokens_seen": 23736736,
      "step": 36395
    },
    {
      "epoch": 19.07756813417191,
      "grad_norm": 0.112906813621521,
      "learning_rate": 6.473175694118705e-06,
      "loss": 0.4452,
      "num_input_tokens_seen": 23739872,
      "step": 36400
    },
    {
      "epoch": 19.080188679245282,
      "grad_norm": 0.17417100071907043,
      "learning_rate": 6.436548314579349e-06,
      "loss": 0.3497,
      "num_input_tokens_seen": 23742592,
      "step": 36405
    },
    {
      "epoch": 19.08280922431866,
      "grad_norm": 0.08419182896614075,
      "learning_rate": 6.400024182883158e-06,
      "loss": 0.5368,
      "num_input_tokens_seen": 23747296,
      "step": 36410
    },
    {
      "epoch": 19.085429769392032,
      "grad_norm": 0.10413379967212677,
      "learning_rate": 6.363603306670629e-06,
      "loss": 0.4247,
      "num_input_tokens_seen": 23751040,
      "step": 36415
    },
    {
      "epoch": 19.08805031446541,
      "grad_norm": 0.1976833939552307,
      "learning_rate": 6.327285693560614e-06,
      "loss": 0.4725,
      "num_input_tokens_seen": 23754976,
      "step": 36420
    },
    {
      "epoch": 19.090670859538783,
      "grad_norm": 0.23613153398036957,
      "learning_rate": 6.2910713511503125e-06,
      "loss": 0.5075,
      "num_input_tokens_seen": 23758464,
      "step": 36425
    },
    {
      "epoch": 19.09329140461216,
      "grad_norm": 0.23298057913780212,
      "learning_rate": 6.254960287015332e-06,
      "loss": 0.4728,
      "num_input_tokens_seen": 23762880,
      "step": 36430
    },
    {
      "epoch": 19.095911949685533,
      "grad_norm": 0.2806166708469391,
      "learning_rate": 6.218952508709741e-06,
      "loss": 0.5596,
      "num_input_tokens_seen": 23765664,
      "step": 36435
    },
    {
      "epoch": 19.09853249475891,
      "grad_norm": 0.1458229273557663,
      "learning_rate": 6.183048023765903e-06,
      "loss": 0.573,
      "num_input_tokens_seen": 23770016,
      "step": 36440
    },
    {
      "epoch": 19.101153039832287,
      "grad_norm": 0.09532785415649414,
      "learning_rate": 6.147246839694698e-06,
      "loss": 0.4412,
      "num_input_tokens_seen": 23773408,
      "step": 36445
    },
    {
      "epoch": 19.10377358490566,
      "grad_norm": 0.13694369792938232,
      "learning_rate": 6.111548963985247e-06,
      "loss": 0.3591,
      "num_input_tokens_seen": 23776672,
      "step": 36450
    },
    {
      "epoch": 19.106394129979037,
      "grad_norm": 0.24574467539787292,
      "learning_rate": 6.075954404105188e-06,
      "loss": 0.3477,
      "num_input_tokens_seen": 23779264,
      "step": 36455
    },
    {
      "epoch": 19.10901467505241,
      "grad_norm": 0.08748237788677216,
      "learning_rate": 6.040463167500509e-06,
      "loss": 0.4582,
      "num_input_tokens_seen": 23783104,
      "step": 36460
    },
    {
      "epoch": 19.111635220125788,
      "grad_norm": 0.23842979967594147,
      "learning_rate": 6.005075261595494e-06,
      "loss": 0.3454,
      "num_input_tokens_seen": 23785856,
      "step": 36465
    },
    {
      "epoch": 19.11425576519916,
      "grad_norm": 0.17994484305381775,
      "learning_rate": 5.969790693792998e-06,
      "loss": 0.3494,
      "num_input_tokens_seen": 23789344,
      "step": 36470
    },
    {
      "epoch": 19.116876310272538,
      "grad_norm": 0.14311842620372772,
      "learning_rate": 5.9346094714740615e-06,
      "loss": 0.4377,
      "num_input_tokens_seen": 23793056,
      "step": 36475
    },
    {
      "epoch": 19.11949685534591,
      "grad_norm": 0.3928828835487366,
      "learning_rate": 5.8995316019982425e-06,
      "loss": 0.496,
      "num_input_tokens_seen": 23795552,
      "step": 36480
    },
    {
      "epoch": 19.122117400419288,
      "grad_norm": 0.15062859654426575,
      "learning_rate": 5.8645570927034485e-06,
      "loss": 0.3878,
      "num_input_tokens_seen": 23797920,
      "step": 36485
    },
    {
      "epoch": 19.12473794549266,
      "grad_norm": 0.13447488844394684,
      "learning_rate": 5.8296859509058275e-06,
      "loss": 0.4982,
      "num_input_tokens_seen": 23800864,
      "step": 36490
    },
    {
      "epoch": 19.12735849056604,
      "grad_norm": 0.13821864128112793,
      "learning_rate": 5.794918183900155e-06,
      "loss": 0.4003,
      "num_input_tokens_seen": 23804480,
      "step": 36495
    },
    {
      "epoch": 19.129979035639412,
      "grad_norm": 0.174325630068779,
      "learning_rate": 5.760253798959447e-06,
      "loss": 0.3456,
      "num_input_tokens_seen": 23807296,
      "step": 36500
    },
    {
      "epoch": 19.13259958071279,
      "grad_norm": 0.15129604935646057,
      "learning_rate": 5.725692803335015e-06,
      "loss": 0.3211,
      "num_input_tokens_seen": 23810432,
      "step": 36505
    },
    {
      "epoch": 19.135220125786162,
      "grad_norm": 0.2687546908855438,
      "learning_rate": 5.691235204256739e-06,
      "loss": 0.3634,
      "num_input_tokens_seen": 23813088,
      "step": 36510
    },
    {
      "epoch": 19.13784067085954,
      "grad_norm": 0.176686093211174,
      "learning_rate": 5.65688100893258e-06,
      "loss": 0.6378,
      "num_input_tokens_seen": 23815968,
      "step": 36515
    },
    {
      "epoch": 19.140461215932913,
      "grad_norm": 0.17562466859817505,
      "learning_rate": 5.622630224549174e-06,
      "loss": 0.419,
      "num_input_tokens_seen": 23819392,
      "step": 36520
    },
    {
      "epoch": 19.14308176100629,
      "grad_norm": 0.15851271152496338,
      "learning_rate": 5.588482858271404e-06,
      "loss": 0.532,
      "num_input_tokens_seen": 23822144,
      "step": 36525
    },
    {
      "epoch": 19.145702306079663,
      "grad_norm": 0.11959642916917801,
      "learning_rate": 5.554438917242444e-06,
      "loss": 0.4649,
      "num_input_tokens_seen": 23826144,
      "step": 36530
    },
    {
      "epoch": 19.14832285115304,
      "grad_norm": 0.4240926504135132,
      "learning_rate": 5.520498408583985e-06,
      "loss": 0.4344,
      "num_input_tokens_seen": 23829536,
      "step": 36535
    },
    {
      "epoch": 19.150943396226417,
      "grad_norm": 0.1500963717699051,
      "learning_rate": 5.486661339395904e-06,
      "loss": 0.3224,
      "num_input_tokens_seen": 23833408,
      "step": 36540
    },
    {
      "epoch": 19.15356394129979,
      "grad_norm": 0.1941530853509903,
      "learning_rate": 5.452927716756595e-06,
      "loss": 0.335,
      "num_input_tokens_seen": 23835808,
      "step": 36545
    },
    {
      "epoch": 19.156184486373167,
      "grad_norm": 0.1469784379005432,
      "learning_rate": 5.419297547722635e-06,
      "loss": 0.5183,
      "num_input_tokens_seen": 23838624,
      "step": 36550
    },
    {
      "epoch": 19.15880503144654,
      "grad_norm": 0.1516275852918625,
      "learning_rate": 5.385770839329229e-06,
      "loss": 0.3105,
      "num_input_tokens_seen": 23841312,
      "step": 36555
    },
    {
      "epoch": 19.161425576519918,
      "grad_norm": 0.21673019230365753,
      "learning_rate": 5.352347598589713e-06,
      "loss": 0.3949,
      "num_input_tokens_seen": 23844416,
      "step": 36560
    },
    {
      "epoch": 19.16404612159329,
      "grad_norm": 0.09661135077476501,
      "learning_rate": 5.319027832495826e-06,
      "loss": 0.3408,
      "num_input_tokens_seen": 23847744,
      "step": 36565
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 0.22025166451931,
      "learning_rate": 5.285811548017661e-06,
      "loss": 0.5108,
      "num_input_tokens_seen": 23850144,
      "step": 36570
    },
    {
      "epoch": 19.16928721174004,
      "grad_norm": 0.09271887689828873,
      "learning_rate": 5.252698752103713e-06,
      "loss": 0.556,
      "num_input_tokens_seen": 23853952,
      "step": 36575
    },
    {
      "epoch": 19.171907756813418,
      "grad_norm": 0.12100416421890259,
      "learning_rate": 5.219689451680832e-06,
      "loss": 0.4346,
      "num_input_tokens_seen": 23857376,
      "step": 36580
    },
    {
      "epoch": 19.17452830188679,
      "grad_norm": 0.14842109382152557,
      "learning_rate": 5.186783653654214e-06,
      "loss": 0.4513,
      "num_input_tokens_seen": 23861056,
      "step": 36585
    },
    {
      "epoch": 19.17714884696017,
      "grad_norm": 0.1690065860748291,
      "learning_rate": 5.153981364907245e-06,
      "loss": 0.335,
      "num_input_tokens_seen": 23863968,
      "step": 36590
    },
    {
      "epoch": 19.179769392033542,
      "grad_norm": 0.13454726338386536,
      "learning_rate": 5.121282592301935e-06,
      "loss": 0.3893,
      "num_input_tokens_seen": 23866912,
      "step": 36595
    },
    {
      "epoch": 19.18238993710692,
      "grad_norm": 0.1556791365146637,
      "learning_rate": 5.088687342678422e-06,
      "loss": 0.3297,
      "num_input_tokens_seen": 23869856,
      "step": 36600
    },
    {
      "epoch": 19.185010482180292,
      "grad_norm": 0.10847029089927673,
      "learning_rate": 5.056195622855253e-06,
      "loss": 0.4246,
      "num_input_tokens_seen": 23873088,
      "step": 36605
    },
    {
      "epoch": 19.18763102725367,
      "grad_norm": 0.12866957485675812,
      "learning_rate": 5.023807439629324e-06,
      "loss": 0.3707,
      "num_input_tokens_seen": 23876160,
      "step": 36610
    },
    {
      "epoch": 19.190251572327043,
      "grad_norm": 0.21324023604393005,
      "learning_rate": 4.991522799775938e-06,
      "loss": 0.473,
      "num_input_tokens_seen": 23878880,
      "step": 36615
    },
    {
      "epoch": 19.19287211740042,
      "grad_norm": 0.18277961015701294,
      "learning_rate": 4.9593417100485816e-06,
      "loss": 0.347,
      "num_input_tokens_seen": 23882912,
      "step": 36620
    },
    {
      "epoch": 19.195492662473793,
      "grad_norm": 0.13307446241378784,
      "learning_rate": 4.927264177179258e-06,
      "loss": 0.3699,
      "num_input_tokens_seen": 23886560,
      "step": 36625
    },
    {
      "epoch": 19.19811320754717,
      "grad_norm": 0.26142770051956177,
      "learning_rate": 4.895290207878156e-06,
      "loss": 0.5331,
      "num_input_tokens_seen": 23889728,
      "step": 36630
    },
    {
      "epoch": 19.200733752620547,
      "grad_norm": 0.1134132444858551,
      "learning_rate": 4.863419808833924e-06,
      "loss": 0.4265,
      "num_input_tokens_seen": 23893408,
      "step": 36635
    },
    {
      "epoch": 19.20335429769392,
      "grad_norm": 0.12425082921981812,
      "learning_rate": 4.831652986713453e-06,
      "loss": 0.3684,
      "num_input_tokens_seen": 23896960,
      "step": 36640
    },
    {
      "epoch": 19.205974842767297,
      "grad_norm": 0.16025711596012115,
      "learning_rate": 4.799989748161926e-06,
      "loss": 0.3784,
      "num_input_tokens_seen": 23899936,
      "step": 36645
    },
    {
      "epoch": 19.20859538784067,
      "grad_norm": 0.16886281967163086,
      "learning_rate": 4.768430099803101e-06,
      "loss": 0.3593,
      "num_input_tokens_seen": 23903360,
      "step": 36650
    },
    {
      "epoch": 19.211215932914047,
      "grad_norm": 0.30184102058410645,
      "learning_rate": 4.736974048238696e-06,
      "loss": 0.3543,
      "num_input_tokens_seen": 23906784,
      "step": 36655
    },
    {
      "epoch": 19.21383647798742,
      "grad_norm": 0.10907553881406784,
      "learning_rate": 4.705621600049115e-06,
      "loss": 0.2881,
      "num_input_tokens_seen": 23910240,
      "step": 36660
    },
    {
      "epoch": 19.216457023060798,
      "grad_norm": 0.1008915826678276,
      "learning_rate": 4.674372761792889e-06,
      "loss": 0.4108,
      "num_input_tokens_seen": 23913888,
      "step": 36665
    },
    {
      "epoch": 19.21907756813417,
      "grad_norm": 0.11776474863290787,
      "learning_rate": 4.6432275400069e-06,
      "loss": 0.3626,
      "num_input_tokens_seen": 23917312,
      "step": 36670
    },
    {
      "epoch": 19.221698113207548,
      "grad_norm": 0.2968358099460602,
      "learning_rate": 4.6121859412063264e-06,
      "loss": 0.3997,
      "num_input_tokens_seen": 23920032,
      "step": 36675
    },
    {
      "epoch": 19.22431865828092,
      "grad_norm": 0.09630090743303299,
      "learning_rate": 4.581247971884861e-06,
      "loss": 0.4931,
      "num_input_tokens_seen": 23924000,
      "step": 36680
    },
    {
      "epoch": 19.2269392033543,
      "grad_norm": 0.13761568069458008,
      "learning_rate": 4.550413638514217e-06,
      "loss": 0.3236,
      "num_input_tokens_seen": 23928128,
      "step": 36685
    },
    {
      "epoch": 19.229559748427672,
      "grad_norm": 0.13339683413505554,
      "learning_rate": 4.519682947544679e-06,
      "loss": 0.427,
      "num_input_tokens_seen": 23931712,
      "step": 36690
    },
    {
      "epoch": 19.23218029350105,
      "grad_norm": 0.15010899305343628,
      "learning_rate": 4.489055905404715e-06,
      "loss": 0.4843,
      "num_input_tokens_seen": 23934912,
      "step": 36695
    },
    {
      "epoch": 19.234800838574422,
      "grad_norm": 0.11357567459344864,
      "learning_rate": 4.4585325185012014e-06,
      "loss": 0.507,
      "num_input_tokens_seen": 23938560,
      "step": 36700
    },
    {
      "epoch": 19.2374213836478,
      "grad_norm": 0.1408793181180954,
      "learning_rate": 4.428112793219197e-06,
      "loss": 0.4427,
      "num_input_tokens_seen": 23941440,
      "step": 36705
    },
    {
      "epoch": 19.240041928721173,
      "grad_norm": 0.15696509182453156,
      "learning_rate": 4.397796735922277e-06,
      "loss": 0.4182,
      "num_input_tokens_seen": 23944736,
      "step": 36710
    },
    {
      "epoch": 19.24266247379455,
      "grad_norm": 0.11324654519557953,
      "learning_rate": 4.367584352952092e-06,
      "loss": 0.6641,
      "num_input_tokens_seen": 23949376,
      "step": 36715
    },
    {
      "epoch": 19.245283018867923,
      "grad_norm": 0.10185562074184418,
      "learning_rate": 4.337475650628808e-06,
      "loss": 0.4654,
      "num_input_tokens_seen": 23952640,
      "step": 36720
    },
    {
      "epoch": 19.2479035639413,
      "grad_norm": 0.18359637260437012,
      "learning_rate": 4.307470635250832e-06,
      "loss": 0.4095,
      "num_input_tokens_seen": 23956992,
      "step": 36725
    },
    {
      "epoch": 19.250524109014677,
      "grad_norm": 0.16933949291706085,
      "learning_rate": 4.277569313094809e-06,
      "loss": 0.3438,
      "num_input_tokens_seen": 23960192,
      "step": 36730
    },
    {
      "epoch": 19.25314465408805,
      "grad_norm": 0.12308559566736221,
      "learning_rate": 4.247771690415791e-06,
      "loss": 0.2995,
      "num_input_tokens_seen": 23963040,
      "step": 36735
    },
    {
      "epoch": 19.255765199161427,
      "grad_norm": 0.14959891140460968,
      "learning_rate": 4.218077773447071e-06,
      "loss": 0.4326,
      "num_input_tokens_seen": 23965664,
      "step": 36740
    },
    {
      "epoch": 19.2583857442348,
      "grad_norm": 0.18094003200531006,
      "learning_rate": 4.1884875684003455e-06,
      "loss": 0.4281,
      "num_input_tokens_seen": 23968992,
      "step": 36745
    },
    {
      "epoch": 19.261006289308177,
      "grad_norm": 0.10897189378738403,
      "learning_rate": 4.159001081465497e-06,
      "loss": 0.767,
      "num_input_tokens_seen": 23972032,
      "step": 36750
    },
    {
      "epoch": 19.26362683438155,
      "grad_norm": 0.20997944474220276,
      "learning_rate": 4.129618318810702e-06,
      "loss": 0.536,
      "num_input_tokens_seen": 23975040,
      "step": 36755
    },
    {
      "epoch": 19.266247379454928,
      "grad_norm": 0.1545051485300064,
      "learning_rate": 4.100339286582655e-06,
      "loss": 0.4195,
      "num_input_tokens_seen": 23977408,
      "step": 36760
    },
    {
      "epoch": 19.2688679245283,
      "grad_norm": 0.182264044880867,
      "learning_rate": 4.071163990906068e-06,
      "loss": 0.4315,
      "num_input_tokens_seen": 23980032,
      "step": 36765
    },
    {
      "epoch": 19.271488469601678,
      "grad_norm": 0.15345537662506104,
      "learning_rate": 4.042092437884115e-06,
      "loss": 0.377,
      "num_input_tokens_seen": 23982720,
      "step": 36770
    },
    {
      "epoch": 19.27410901467505,
      "grad_norm": 0.2492908090353012,
      "learning_rate": 4.013124633598264e-06,
      "loss": 0.4636,
      "num_input_tokens_seen": 23985984,
      "step": 36775
    },
    {
      "epoch": 19.27672955974843,
      "grad_norm": 0.12557627260684967,
      "learning_rate": 3.984260584108168e-06,
      "loss": 0.5292,
      "num_input_tokens_seen": 23989504,
      "step": 36780
    },
    {
      "epoch": 19.279350104821802,
      "grad_norm": 0.19536907970905304,
      "learning_rate": 3.955500295451942e-06,
      "loss": 0.4322,
      "num_input_tokens_seen": 23992160,
      "step": 36785
    },
    {
      "epoch": 19.28197064989518,
      "grad_norm": 0.14626851677894592,
      "learning_rate": 3.926843773645883e-06,
      "loss": 0.5027,
      "num_input_tokens_seen": 23998816,
      "step": 36790
    },
    {
      "epoch": 19.284591194968552,
      "grad_norm": 0.09079428017139435,
      "learning_rate": 3.8982910246846415e-06,
      "loss": 0.4209,
      "num_input_tokens_seen": 24003040,
      "step": 36795
    },
    {
      "epoch": 19.28721174004193,
      "grad_norm": 0.12371115386486053,
      "learning_rate": 3.869842054541051e-06,
      "loss": 0.459,
      "num_input_tokens_seen": 24006272,
      "step": 36800
    },
    {
      "epoch": 19.289832285115303,
      "grad_norm": 0.16324865818023682,
      "learning_rate": 3.84149686916635e-06,
      "loss": 0.3771,
      "num_input_tokens_seen": 24009792,
      "step": 36805
    },
    {
      "epoch": 19.29245283018868,
      "grad_norm": 0.16641703248023987,
      "learning_rate": 3.8132554744900183e-06,
      "loss": 0.3331,
      "num_input_tokens_seen": 24012960,
      "step": 36810
    },
    {
      "epoch": 19.295073375262053,
      "grad_norm": 0.16804909706115723,
      "learning_rate": 3.7851178764198302e-06,
      "loss": 0.5595,
      "num_input_tokens_seen": 24016416,
      "step": 36815
    },
    {
      "epoch": 19.29769392033543,
      "grad_norm": 0.2588146924972534,
      "learning_rate": 3.7570840808419104e-06,
      "loss": 0.3412,
      "num_input_tokens_seen": 24019520,
      "step": 36820
    },
    {
      "epoch": 19.300314465408803,
      "grad_norm": 0.2465135008096695,
      "learning_rate": 3.729154093620568e-06,
      "loss": 0.4084,
      "num_input_tokens_seen": 24023168,
      "step": 36825
    },
    {
      "epoch": 19.30293501048218,
      "grad_norm": 0.1428520232439041,
      "learning_rate": 3.7013279205984073e-06,
      "loss": 0.4144,
      "num_input_tokens_seen": 24026368,
      "step": 36830
    },
    {
      "epoch": 19.305555555555557,
      "grad_norm": 0.3073488771915436,
      "learning_rate": 3.6736055675963275e-06,
      "loss": 0.3441,
      "num_input_tokens_seen": 24028960,
      "step": 36835
    },
    {
      "epoch": 19.30817610062893,
      "grad_norm": 0.13041916489601135,
      "learning_rate": 3.645987040413634e-06,
      "loss": 0.4784,
      "num_input_tokens_seen": 24031392,
      "step": 36840
    },
    {
      "epoch": 19.310796645702307,
      "grad_norm": 0.21217408776283264,
      "learning_rate": 3.6184723448277056e-06,
      "loss": 0.4131,
      "num_input_tokens_seen": 24034368,
      "step": 36845
    },
    {
      "epoch": 19.31341719077568,
      "grad_norm": 0.10445310175418854,
      "learning_rate": 3.5910614865943826e-06,
      "loss": 0.4571,
      "num_input_tokens_seen": 24037472,
      "step": 36850
    },
    {
      "epoch": 19.316037735849058,
      "grad_norm": 0.19768331944942474,
      "learning_rate": 3.563754471447689e-06,
      "loss": 0.4573,
      "num_input_tokens_seen": 24040672,
      "step": 36855
    },
    {
      "epoch": 19.31865828092243,
      "grad_norm": 0.11555106937885284,
      "learning_rate": 3.5365513050998334e-06,
      "loss": 0.4154,
      "num_input_tokens_seen": 24044192,
      "step": 36860
    },
    {
      "epoch": 19.321278825995808,
      "grad_norm": 0.19039304554462433,
      "learning_rate": 3.5094519932415414e-06,
      "loss": 0.4202,
      "num_input_tokens_seen": 24047520,
      "step": 36865
    },
    {
      "epoch": 19.32389937106918,
      "grad_norm": 0.12550142407417297,
      "learning_rate": 3.4824565415416123e-06,
      "loss": 0.534,
      "num_input_tokens_seen": 24050944,
      "step": 36870
    },
    {
      "epoch": 19.32651991614256,
      "grad_norm": 0.18607334792613983,
      "learning_rate": 3.4555649556471946e-06,
      "loss": 0.3959,
      "num_input_tokens_seen": 24053440,
      "step": 36875
    },
    {
      "epoch": 19.329140461215932,
      "grad_norm": 0.34029197692871094,
      "learning_rate": 3.4287772411837338e-06,
      "loss": 0.4353,
      "num_input_tokens_seen": 24056480,
      "step": 36880
    },
    {
      "epoch": 19.33176100628931,
      "grad_norm": 0.1559806764125824,
      "learning_rate": 3.402093403754858e-06,
      "loss": 0.3426,
      "num_input_tokens_seen": 24060768,
      "step": 36885
    },
    {
      "epoch": 19.334381551362682,
      "grad_norm": 0.16573090851306915,
      "learning_rate": 3.375513448942602e-06,
      "loss": 0.4913,
      "num_input_tokens_seen": 24063200,
      "step": 36890
    },
    {
      "epoch": 19.33700209643606,
      "grad_norm": 0.24478760361671448,
      "learning_rate": 3.349037382307074e-06,
      "loss": 0.5868,
      "num_input_tokens_seen": 24066272,
      "step": 36895
    },
    {
      "epoch": 19.339622641509433,
      "grad_norm": 0.14545965194702148,
      "learning_rate": 3.322665209386899e-06,
      "loss": 0.4101,
      "num_input_tokens_seen": 24068960,
      "step": 36900
    },
    {
      "epoch": 19.34224318658281,
      "grad_norm": 0.14537721872329712,
      "learning_rate": 3.29639693569872e-06,
      "loss": 0.5584,
      "num_input_tokens_seen": 24071936,
      "step": 36905
    },
    {
      "epoch": 19.344863731656183,
      "grad_norm": 0.3185994029045105,
      "learning_rate": 3.270232566737641e-06,
      "loss": 0.4329,
      "num_input_tokens_seen": 24074752,
      "step": 36910
    },
    {
      "epoch": 19.34748427672956,
      "grad_norm": 0.107883021235466,
      "learning_rate": 3.244172107976895e-06,
      "loss": 0.4745,
      "num_input_tokens_seen": 24078240,
      "step": 36915
    },
    {
      "epoch": 19.350104821802937,
      "grad_norm": 0.14704132080078125,
      "learning_rate": 3.2182155648680657e-06,
      "loss": 0.4198,
      "num_input_tokens_seen": 24081920,
      "step": 36920
    },
    {
      "epoch": 19.35272536687631,
      "grad_norm": 0.16822564601898193,
      "learning_rate": 3.1923629428409205e-06,
      "loss": 0.4922,
      "num_input_tokens_seen": 24085952,
      "step": 36925
    },
    {
      "epoch": 19.355345911949687,
      "grad_norm": 0.18555565178394318,
      "learning_rate": 3.166614247303634e-06,
      "loss": 0.4111,
      "num_input_tokens_seen": 24088672,
      "step": 36930
    },
    {
      "epoch": 19.35796645702306,
      "grad_norm": 0.10831964761018753,
      "learning_rate": 3.140969483642453e-06,
      "loss": 0.4622,
      "num_input_tokens_seen": 24092064,
      "step": 36935
    },
    {
      "epoch": 19.360587002096437,
      "grad_norm": 0.11805298179388046,
      "learning_rate": 3.1154286572219747e-06,
      "loss": 0.4965,
      "num_input_tokens_seen": 24095328,
      "step": 36940
    },
    {
      "epoch": 19.36320754716981,
      "grad_norm": 0.17596395313739777,
      "learning_rate": 3.089991773385037e-06,
      "loss": 0.4752,
      "num_input_tokens_seen": 24098560,
      "step": 36945
    },
    {
      "epoch": 19.365828092243188,
      "grad_norm": 0.1108972355723381,
      "learning_rate": 3.064658837452772e-06,
      "loss": 0.4193,
      "num_input_tokens_seen": 24101696,
      "step": 36950
    },
    {
      "epoch": 19.36844863731656,
      "grad_norm": 0.19741258025169373,
      "learning_rate": 3.0394298547246068e-06,
      "loss": 0.5001,
      "num_input_tokens_seen": 24104448,
      "step": 36955
    },
    {
      "epoch": 19.371069182389938,
      "grad_norm": 0.1385761797428131,
      "learning_rate": 3.0143048304779875e-06,
      "loss": 0.4235,
      "num_input_tokens_seen": 24107520,
      "step": 36960
    },
    {
      "epoch": 19.37368972746331,
      "grad_norm": 0.2739775478839874,
      "learning_rate": 2.989283769968987e-06,
      "loss": 0.372,
      "num_input_tokens_seen": 24110304,
      "step": 36965
    },
    {
      "epoch": 19.37631027253669,
      "grad_norm": 0.1986979991197586,
      "learning_rate": 2.964366678431585e-06,
      "loss": 0.5129,
      "num_input_tokens_seen": 24113792,
      "step": 36970
    },
    {
      "epoch": 19.378930817610062,
      "grad_norm": 0.2926880717277527,
      "learning_rate": 2.9395535610781678e-06,
      "loss": 0.4426,
      "num_input_tokens_seen": 24116224,
      "step": 36975
    },
    {
      "epoch": 19.38155136268344,
      "grad_norm": 0.11669619381427765,
      "learning_rate": 2.9148444230994166e-06,
      "loss": 0.3788,
      "num_input_tokens_seen": 24120480,
      "step": 36980
    },
    {
      "epoch": 19.384171907756812,
      "grad_norm": 0.4413863718509674,
      "learning_rate": 2.890239269664141e-06,
      "loss": 0.5368,
      "num_input_tokens_seen": 24123424,
      "step": 36985
    },
    {
      "epoch": 19.38679245283019,
      "grad_norm": 0.18648508191108704,
      "learning_rate": 2.8657381059194466e-06,
      "loss": 0.3593,
      "num_input_tokens_seen": 24126528,
      "step": 36990
    },
    {
      "epoch": 19.389412997903563,
      "grad_norm": 0.16783668100833893,
      "learning_rate": 2.8413409369907885e-06,
      "loss": 0.3643,
      "num_input_tokens_seen": 24131360,
      "step": 36995
    },
    {
      "epoch": 19.39203354297694,
      "grad_norm": 0.2144777625799179,
      "learning_rate": 2.817047767981695e-06,
      "loss": 0.4317,
      "num_input_tokens_seen": 24133920,
      "step": 37000
    },
    {
      "epoch": 19.394654088050313,
      "grad_norm": 0.14322297275066376,
      "learning_rate": 2.7928586039740466e-06,
      "loss": 0.3721,
      "num_input_tokens_seen": 24137536,
      "step": 37005
    },
    {
      "epoch": 19.39727463312369,
      "grad_norm": 0.21808305382728577,
      "learning_rate": 2.7687734500279615e-06,
      "loss": 0.3816,
      "num_input_tokens_seen": 24140448,
      "step": 37010
    },
    {
      "epoch": 19.399895178197063,
      "grad_norm": 0.18457892537117004,
      "learning_rate": 2.744792311181743e-06,
      "loss": 0.4815,
      "num_input_tokens_seen": 24143488,
      "step": 37015
    },
    {
      "epoch": 19.40251572327044,
      "grad_norm": 0.35829463601112366,
      "learning_rate": 2.720915192451989e-06,
      "loss": 0.3899,
      "num_input_tokens_seen": 24146944,
      "step": 37020
    },
    {
      "epoch": 19.405136268343817,
      "grad_norm": 0.13576769828796387,
      "learning_rate": 2.697142098833538e-06,
      "loss": 0.4235,
      "num_input_tokens_seen": 24150944,
      "step": 37025
    },
    {
      "epoch": 19.40775681341719,
      "grad_norm": 0.3406977653503418,
      "learning_rate": 2.6734730352993563e-06,
      "loss": 0.4184,
      "num_input_tokens_seen": 24153504,
      "step": 37030
    },
    {
      "epoch": 19.410377358490567,
      "grad_norm": 0.10944721847772598,
      "learning_rate": 2.649908006800872e-06,
      "loss": 0.4093,
      "num_input_tokens_seen": 24156480,
      "step": 37035
    },
    {
      "epoch": 19.41299790356394,
      "grad_norm": 0.19271166622638702,
      "learning_rate": 2.626447018267586e-06,
      "loss": 0.5618,
      "num_input_tokens_seen": 24160192,
      "step": 37040
    },
    {
      "epoch": 19.415618448637318,
      "grad_norm": 0.16385991871356964,
      "learning_rate": 2.603090074607184e-06,
      "loss": 0.4577,
      "num_input_tokens_seen": 24162560,
      "step": 37045
    },
    {
      "epoch": 19.41823899371069,
      "grad_norm": 0.1203693374991417,
      "learning_rate": 2.579837180705813e-06,
      "loss": 0.4331,
      "num_input_tokens_seen": 24165568,
      "step": 37050
    },
    {
      "epoch": 19.420859538784068,
      "grad_norm": 0.10570314526557922,
      "learning_rate": 2.556688341427582e-06,
      "loss": 0.3426,
      "num_input_tokens_seen": 24168608,
      "step": 37055
    },
    {
      "epoch": 19.42348008385744,
      "grad_norm": 0.12440130114555359,
      "learning_rate": 2.5336435616150066e-06,
      "loss": 0.5014,
      "num_input_tokens_seen": 24171552,
      "step": 37060
    },
    {
      "epoch": 19.42610062893082,
      "grad_norm": 0.20311735570430756,
      "learning_rate": 2.510702846088786e-06,
      "loss": 0.3937,
      "num_input_tokens_seen": 24174336,
      "step": 37065
    },
    {
      "epoch": 19.428721174004192,
      "grad_norm": 0.10246216505765915,
      "learning_rate": 2.487866199647915e-06,
      "loss": 0.3085,
      "num_input_tokens_seen": 24178400,
      "step": 37070
    },
    {
      "epoch": 19.43134171907757,
      "grad_norm": 0.16050122678279877,
      "learning_rate": 2.4651336270695156e-06,
      "loss": 0.3661,
      "num_input_tokens_seen": 24181312,
      "step": 37075
    },
    {
      "epoch": 19.433962264150942,
      "grad_norm": 0.14399631321430206,
      "learning_rate": 2.442505133108952e-06,
      "loss": 0.4408,
      "num_input_tokens_seen": 24184544,
      "step": 37080
    },
    {
      "epoch": 19.43658280922432,
      "grad_norm": 0.20437216758728027,
      "learning_rate": 2.419980722499937e-06,
      "loss": 0.2861,
      "num_input_tokens_seen": 24186720,
      "step": 37085
    },
    {
      "epoch": 19.439203354297693,
      "grad_norm": 0.21532325446605682,
      "learning_rate": 2.397560399954202e-06,
      "loss": 0.4594,
      "num_input_tokens_seen": 24189984,
      "step": 37090
    },
    {
      "epoch": 19.44182389937107,
      "grad_norm": 0.17689622938632965,
      "learning_rate": 2.3752441701618833e-06,
      "loss": 0.3654,
      "num_input_tokens_seen": 24192448,
      "step": 37095
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 0.11409201472997665,
      "learning_rate": 2.3530320377913027e-06,
      "loss": 0.3748,
      "num_input_tokens_seen": 24196128,
      "step": 37100
    },
    {
      "epoch": 19.44706498951782,
      "grad_norm": 0.12437096238136292,
      "learning_rate": 2.3309240074890213e-06,
      "loss": 0.3611,
      "num_input_tokens_seen": 24199104,
      "step": 37105
    },
    {
      "epoch": 19.449685534591197,
      "grad_norm": 0.13802452385425568,
      "learning_rate": 2.3089200838796176e-06,
      "loss": 0.54,
      "num_input_tokens_seen": 24202176,
      "step": 37110
    },
    {
      "epoch": 19.45230607966457,
      "grad_norm": 0.1569560468196869,
      "learning_rate": 2.2870202715662426e-06,
      "loss": 0.4897,
      "num_input_tokens_seen": 24205824,
      "step": 37115
    },
    {
      "epoch": 19.454926624737947,
      "grad_norm": 0.12873193621635437,
      "learning_rate": 2.265224575130009e-06,
      "loss": 0.5309,
      "num_input_tokens_seen": 24208640,
      "step": 37120
    },
    {
      "epoch": 19.45754716981132,
      "grad_norm": 0.025524798780679703,
      "learning_rate": 2.2435329991303268e-06,
      "loss": 0.2571,
      "num_input_tokens_seen": 24215456,
      "step": 37125
    },
    {
      "epoch": 19.460167714884697,
      "grad_norm": 0.1289217472076416,
      "learning_rate": 2.2219455481047868e-06,
      "loss": 0.4262,
      "num_input_tokens_seen": 24218752,
      "step": 37130
    },
    {
      "epoch": 19.46278825995807,
      "grad_norm": 0.13333268463611603,
      "learning_rate": 2.2004622265693886e-06,
      "loss": 0.3992,
      "num_input_tokens_seen": 24221632,
      "step": 37135
    },
    {
      "epoch": 19.465408805031448,
      "grad_norm": 0.16573898494243622,
      "learning_rate": 2.179083039018037e-06,
      "loss": 0.4954,
      "num_input_tokens_seen": 24225824,
      "step": 37140
    },
    {
      "epoch": 19.46802935010482,
      "grad_norm": 0.17575669288635254,
      "learning_rate": 2.157807989923044e-06,
      "loss": 0.4513,
      "num_input_tokens_seen": 24228960,
      "step": 37145
    },
    {
      "epoch": 19.470649895178198,
      "grad_norm": 0.1288762390613556,
      "learning_rate": 2.1366370837349603e-06,
      "loss": 0.374,
      "num_input_tokens_seen": 24232352,
      "step": 37150
    },
    {
      "epoch": 19.47327044025157,
      "grad_norm": 0.1227479949593544,
      "learning_rate": 2.1155703248825207e-06,
      "loss": 0.3531,
      "num_input_tokens_seen": 24235456,
      "step": 37155
    },
    {
      "epoch": 19.47589098532495,
      "grad_norm": 0.21030694246292114,
      "learning_rate": 2.094607717772534e-06,
      "loss": 0.4298,
      "num_input_tokens_seen": 24238016,
      "step": 37160
    },
    {
      "epoch": 19.478511530398322,
      "grad_norm": 0.17031127214431763,
      "learning_rate": 2.0737492667902702e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 24240672,
      "step": 37165
    },
    {
      "epoch": 19.4811320754717,
      "grad_norm": 0.2826606035232544,
      "learning_rate": 2.0529949762989608e-06,
      "loss": 0.3173,
      "num_input_tokens_seen": 24243520,
      "step": 37170
    },
    {
      "epoch": 19.483752620545072,
      "grad_norm": 0.23951447010040283,
      "learning_rate": 2.032344850640244e-06,
      "loss": 0.4524,
      "num_input_tokens_seen": 24246784,
      "step": 37175
    },
    {
      "epoch": 19.48637316561845,
      "grad_norm": 0.14520183205604553,
      "learning_rate": 2.011798894133887e-06,
      "loss": 0.6324,
      "num_input_tokens_seen": 24250848,
      "step": 37180
    },
    {
      "epoch": 19.488993710691823,
      "grad_norm": 0.1132698729634285,
      "learning_rate": 1.9913571110777852e-06,
      "loss": 0.4488,
      "num_input_tokens_seen": 24254208,
      "step": 37185
    },
    {
      "epoch": 19.4916142557652,
      "grad_norm": 0.17722132802009583,
      "learning_rate": 1.971019505748295e-06,
      "loss": 0.4439,
      "num_input_tokens_seen": 24260896,
      "step": 37190
    },
    {
      "epoch": 19.494234800838573,
      "grad_norm": 0.47176969051361084,
      "learning_rate": 1.9507860823996803e-06,
      "loss": 0.4328,
      "num_input_tokens_seen": 24263232,
      "step": 37195
    },
    {
      "epoch": 19.49685534591195,
      "grad_norm": 0.12904874980449677,
      "learning_rate": 1.9306568452645e-06,
      "loss": 0.4888,
      "num_input_tokens_seen": 24266784,
      "step": 37200
    },
    {
      "epoch": 19.499475890985323,
      "grad_norm": 0.12440747022628784,
      "learning_rate": 1.910631798553664e-06,
      "loss": 0.357,
      "num_input_tokens_seen": 24270112,
      "step": 37205
    },
    {
      "epoch": 19.5020964360587,
      "grad_norm": 0.14340320229530334,
      "learning_rate": 1.8907109464562088e-06,
      "loss": 0.4867,
      "num_input_tokens_seen": 24273184,
      "step": 37210
    },
    {
      "epoch": 19.504716981132077,
      "grad_norm": 0.18497063219547272,
      "learning_rate": 1.870894293139247e-06,
      "loss": 0.4502,
      "num_input_tokens_seen": 24276224,
      "step": 37215
    },
    {
      "epoch": 19.50733752620545,
      "grad_norm": 0.12734949588775635,
      "learning_rate": 1.8511818427482396e-06,
      "loss": 0.3666,
      "num_input_tokens_seen": 24281024,
      "step": 37220
    },
    {
      "epoch": 19.509958071278827,
      "grad_norm": 0.13789039850234985,
      "learning_rate": 1.8315735994068327e-06,
      "loss": 0.5454,
      "num_input_tokens_seen": 24283840,
      "step": 37225
    },
    {
      "epoch": 19.5125786163522,
      "grad_norm": 0.14601218700408936,
      "learning_rate": 1.8120695672168009e-06,
      "loss": 0.248,
      "num_input_tokens_seen": 24286240,
      "step": 37230
    },
    {
      "epoch": 19.515199161425578,
      "grad_norm": 0.20602668821811676,
      "learning_rate": 1.792669750258158e-06,
      "loss": 0.3818,
      "num_input_tokens_seen": 24288992,
      "step": 37235
    },
    {
      "epoch": 19.51781970649895,
      "grad_norm": 0.15880419313907623,
      "learning_rate": 1.7733741525892134e-06,
      "loss": 0.4034,
      "num_input_tokens_seen": 24292288,
      "step": 37240
    },
    {
      "epoch": 19.520440251572328,
      "grad_norm": 0.12169241905212402,
      "learning_rate": 1.7541827782462937e-06,
      "loss": 0.4658,
      "num_input_tokens_seen": 24295584,
      "step": 37245
    },
    {
      "epoch": 19.5230607966457,
      "grad_norm": 0.13487966358661652,
      "learning_rate": 1.7350956312440768e-06,
      "loss": 0.3764,
      "num_input_tokens_seen": 24298400,
      "step": 37250
    },
    {
      "epoch": 19.52568134171908,
      "grad_norm": 0.2037704586982727,
      "learning_rate": 1.716112715575313e-06,
      "loss": 0.3428,
      "num_input_tokens_seen": 24301280,
      "step": 37255
    },
    {
      "epoch": 19.528301886792452,
      "grad_norm": 0.14583608508110046,
      "learning_rate": 1.6972340352110481e-06,
      "loss": 0.4553,
      "num_input_tokens_seen": 24303968,
      "step": 37260
    },
    {
      "epoch": 19.53092243186583,
      "grad_norm": 0.12670989334583282,
      "learning_rate": 1.6784595941004565e-06,
      "loss": 0.4346,
      "num_input_tokens_seen": 24307456,
      "step": 37265
    },
    {
      "epoch": 19.533542976939202,
      "grad_norm": 0.21231414377689362,
      "learning_rate": 1.659789396171063e-06,
      "loss": 0.3948,
      "num_input_tokens_seen": 24310912,
      "step": 37270
    },
    {
      "epoch": 19.53616352201258,
      "grad_norm": 0.12798026204109192,
      "learning_rate": 1.6412234453282993e-06,
      "loss": 0.4151,
      "num_input_tokens_seen": 24313728,
      "step": 37275
    },
    {
      "epoch": 19.538784067085953,
      "grad_norm": 0.1797236055135727,
      "learning_rate": 1.622761745456003e-06,
      "loss": 0.4073,
      "num_input_tokens_seen": 24316928,
      "step": 37280
    },
    {
      "epoch": 19.54140461215933,
      "grad_norm": 0.2312169224023819,
      "learning_rate": 1.6044043004161958e-06,
      "loss": 0.366,
      "num_input_tokens_seen": 24319552,
      "step": 37285
    },
    {
      "epoch": 19.544025157232703,
      "grad_norm": 0.17148177325725555,
      "learning_rate": 1.5861511140489725e-06,
      "loss": 0.4135,
      "num_input_tokens_seen": 24321984,
      "step": 37290
    },
    {
      "epoch": 19.54664570230608,
      "grad_norm": 0.21846795082092285,
      "learning_rate": 1.5680021901727237e-06,
      "loss": 0.4859,
      "num_input_tokens_seen": 24324928,
      "step": 37295
    },
    {
      "epoch": 19.549266247379457,
      "grad_norm": 0.11720463633537292,
      "learning_rate": 1.5499575325840232e-06,
      "loss": 0.3087,
      "num_input_tokens_seen": 24328064,
      "step": 37300
    },
    {
      "epoch": 19.55188679245283,
      "grad_norm": 0.3088264465332031,
      "learning_rate": 1.5320171450576293e-06,
      "loss": 0.5207,
      "num_input_tokens_seen": 24331136,
      "step": 37305
    },
    {
      "epoch": 19.554507337526207,
      "grad_norm": 0.21190820634365082,
      "learning_rate": 1.5141810313463733e-06,
      "loss": 0.4399,
      "num_input_tokens_seen": 24334112,
      "step": 37310
    },
    {
      "epoch": 19.55712788259958,
      "grad_norm": 0.07981634140014648,
      "learning_rate": 1.4964491951814374e-06,
      "loss": 0.3635,
      "num_input_tokens_seen": 24338048,
      "step": 37315
    },
    {
      "epoch": 19.559748427672957,
      "grad_norm": 0.1074913889169693,
      "learning_rate": 1.4788216402720766e-06,
      "loss": 0.5321,
      "num_input_tokens_seen": 24341632,
      "step": 37320
    },
    {
      "epoch": 19.56236897274633,
      "grad_norm": 0.1408797800540924,
      "learning_rate": 1.4612983703058413e-06,
      "loss": 0.4173,
      "num_input_tokens_seen": 24345088,
      "step": 37325
    },
    {
      "epoch": 19.564989517819708,
      "grad_norm": 0.09152025729417801,
      "learning_rate": 1.4438793889483549e-06,
      "loss": 0.4175,
      "num_input_tokens_seen": 24348736,
      "step": 37330
    },
    {
      "epoch": 19.56761006289308,
      "grad_norm": 0.10966866463422775,
      "learning_rate": 1.4265646998434246e-06,
      "loss": 0.3316,
      "num_input_tokens_seen": 24351776,
      "step": 37335
    },
    {
      "epoch": 19.570230607966458,
      "grad_norm": 0.11037307977676392,
      "learning_rate": 1.409354306613153e-06,
      "loss": 0.3668,
      "num_input_tokens_seen": 24355296,
      "step": 37340
    },
    {
      "epoch": 19.57285115303983,
      "grad_norm": 0.1199977919459343,
      "learning_rate": 1.3922482128577718e-06,
      "loss": 0.4171,
      "num_input_tokens_seen": 24359168,
      "step": 37345
    },
    {
      "epoch": 19.57547169811321,
      "grad_norm": 0.12981891632080078,
      "learning_rate": 1.3752464221556404e-06,
      "loss": 0.4846,
      "num_input_tokens_seen": 24362720,
      "step": 37350
    },
    {
      "epoch": 19.578092243186582,
      "grad_norm": 0.108600914478302,
      "learning_rate": 1.358348938063303e-06,
      "loss": 0.382,
      "num_input_tokens_seen": 24365824,
      "step": 37355
    },
    {
      "epoch": 19.58071278825996,
      "grad_norm": 0.17466020584106445,
      "learning_rate": 1.341555764115543e-06,
      "loss": 0.35,
      "num_input_tokens_seen": 24368928,
      "step": 37360
    },
    {
      "epoch": 19.583333333333332,
      "grad_norm": 0.19627457857131958,
      "learning_rate": 1.3248669038253835e-06,
      "loss": 0.3859,
      "num_input_tokens_seen": 24372192,
      "step": 37365
    },
    {
      "epoch": 19.58595387840671,
      "grad_norm": 0.21451617777347565,
      "learning_rate": 1.3082823606838656e-06,
      "loss": 0.4425,
      "num_input_tokens_seen": 24374912,
      "step": 37370
    },
    {
      "epoch": 19.588574423480082,
      "grad_norm": 0.14616863429546356,
      "learning_rate": 1.2918021381603251e-06,
      "loss": 0.3914,
      "num_input_tokens_seen": 24378176,
      "step": 37375
    },
    {
      "epoch": 19.59119496855346,
      "grad_norm": 0.19953610002994537,
      "learning_rate": 1.275426239702171e-06,
      "loss": 0.3741,
      "num_input_tokens_seen": 24381760,
      "step": 37380
    },
    {
      "epoch": 19.593815513626833,
      "grad_norm": 0.13500091433525085,
      "learning_rate": 1.2591546687351073e-06,
      "loss": 0.3935,
      "num_input_tokens_seen": 24386144,
      "step": 37385
    },
    {
      "epoch": 19.59643605870021,
      "grad_norm": 0.09011843800544739,
      "learning_rate": 1.2429874286629673e-06,
      "loss": 0.4078,
      "num_input_tokens_seen": 24389600,
      "step": 37390
    },
    {
      "epoch": 19.599056603773583,
      "grad_norm": 0.3450719118118286,
      "learning_rate": 1.2269245228677116e-06,
      "loss": 0.4075,
      "num_input_tokens_seen": 24392160,
      "step": 37395
    },
    {
      "epoch": 19.60167714884696,
      "grad_norm": 0.15198703110218048,
      "learning_rate": 1.210965954709542e-06,
      "loss": 0.4063,
      "num_input_tokens_seen": 24395584,
      "step": 37400
    },
    {
      "epoch": 19.604297693920337,
      "grad_norm": 0.27539119124412537,
      "learning_rate": 1.1951117275268431e-06,
      "loss": 0.3625,
      "num_input_tokens_seen": 24397696,
      "step": 37405
    },
    {
      "epoch": 19.60691823899371,
      "grad_norm": 0.1255144327878952,
      "learning_rate": 1.1793618446360732e-06,
      "loss": 0.4169,
      "num_input_tokens_seen": 24401696,
      "step": 37410
    },
    {
      "epoch": 19.609538784067087,
      "grad_norm": 0.21553881466388702,
      "learning_rate": 1.1637163093319303e-06,
      "loss": 0.4148,
      "num_input_tokens_seen": 24404416,
      "step": 37415
    },
    {
      "epoch": 19.61215932914046,
      "grad_norm": 0.19873909652233124,
      "learning_rate": 1.1481751248874072e-06,
      "loss": 0.37,
      "num_input_tokens_seen": 24407008,
      "step": 37420
    },
    {
      "epoch": 19.614779874213838,
      "grad_norm": 0.11564186215400696,
      "learning_rate": 1.1327382945533482e-06,
      "loss": 0.3025,
      "num_input_tokens_seen": 24410208,
      "step": 37425
    },
    {
      "epoch": 19.61740041928721,
      "grad_norm": 0.08316364139318466,
      "learning_rate": 1.1174058215591143e-06,
      "loss": 0.3042,
      "num_input_tokens_seen": 24415168,
      "step": 37430
    },
    {
      "epoch": 19.620020964360588,
      "grad_norm": 0.16748355329036713,
      "learning_rate": 1.1021777091119732e-06,
      "loss": 0.3679,
      "num_input_tokens_seen": 24417984,
      "step": 37435
    },
    {
      "epoch": 19.62264150943396,
      "grad_norm": 0.22301895916461945,
      "learning_rate": 1.0870539603975994e-06,
      "loss": 0.4822,
      "num_input_tokens_seen": 24421376,
      "step": 37440
    },
    {
      "epoch": 19.62526205450734,
      "grad_norm": 0.08289226144552231,
      "learning_rate": 1.0720345785795727e-06,
      "loss": 0.2813,
      "num_input_tokens_seen": 24424544,
      "step": 37445
    },
    {
      "epoch": 19.627882599580712,
      "grad_norm": 0.19036202132701874,
      "learning_rate": 1.0571195667998802e-06,
      "loss": 0.3549,
      "num_input_tokens_seen": 24427488,
      "step": 37450
    },
    {
      "epoch": 19.63050314465409,
      "grad_norm": 0.19171369075775146,
      "learning_rate": 1.042308928178526e-06,
      "loss": 0.4368,
      "num_input_tokens_seen": 24430336,
      "step": 37455
    },
    {
      "epoch": 19.633123689727462,
      "grad_norm": 0.17349457740783691,
      "learning_rate": 1.0276026658137538e-06,
      "loss": 0.4106,
      "num_input_tokens_seen": 24435136,
      "step": 37460
    },
    {
      "epoch": 19.63574423480084,
      "grad_norm": 0.2674713432788849,
      "learning_rate": 1.013000782781881e-06,
      "loss": 0.4952,
      "num_input_tokens_seen": 24438848,
      "step": 37465
    },
    {
      "epoch": 19.638364779874212,
      "grad_norm": 0.12826736271381378,
      "learning_rate": 9.985032821375195e-07,
      "loss": 0.4517,
      "num_input_tokens_seen": 24441248,
      "step": 37470
    },
    {
      "epoch": 19.64098532494759,
      "grad_norm": 0.1977064609527588,
      "learning_rate": 9.841101669134101e-07,
      "loss": 0.3931,
      "num_input_tokens_seen": 24443840,
      "step": 37475
    },
    {
      "epoch": 19.643605870020963,
      "grad_norm": 0.14116492867469788,
      "learning_rate": 9.69821440120311e-07,
      "loss": 0.4657,
      "num_input_tokens_seen": 24447200,
      "step": 37480
    },
    {
      "epoch": 19.64622641509434,
      "grad_norm": 0.5814397931098938,
      "learning_rate": 9.556371047473866e-07,
      "loss": 0.5455,
      "num_input_tokens_seen": 24450368,
      "step": 37485
    },
    {
      "epoch": 19.648846960167717,
      "grad_norm": 0.1350126564502716,
      "learning_rate": 9.415571637617082e-07,
      "loss": 0.3123,
      "num_input_tokens_seen": 24452992,
      "step": 37490
    },
    {
      "epoch": 19.65146750524109,
      "grad_norm": 0.1671357899904251,
      "learning_rate": 9.275816201087528e-07,
      "loss": 0.477,
      "num_input_tokens_seen": 24455264,
      "step": 37495
    },
    {
      "epoch": 19.654088050314467,
      "grad_norm": 0.17672429978847504,
      "learning_rate": 9.137104767120153e-07,
      "loss": 0.4983,
      "num_input_tokens_seen": 24458304,
      "step": 37500
    },
    {
      "epoch": 19.65670859538784,
      "grad_norm": 0.1675829291343689,
      "learning_rate": 8.999437364731189e-07,
      "loss": 0.486,
      "num_input_tokens_seen": 24461824,
      "step": 37505
    },
    {
      "epoch": 19.659329140461217,
      "grad_norm": 0.20969349145889282,
      "learning_rate": 8.862814022720378e-07,
      "loss": 0.3793,
      "num_input_tokens_seen": 24464736,
      "step": 37510
    },
    {
      "epoch": 19.66194968553459,
      "grad_norm": 0.17408446967601776,
      "learning_rate": 8.727234769666526e-07,
      "loss": 0.3724,
      "num_input_tokens_seen": 24467264,
      "step": 37515
    },
    {
      "epoch": 19.664570230607968,
      "grad_norm": 0.13882414996623993,
      "learning_rate": 8.592699633931389e-07,
      "loss": 0.4585,
      "num_input_tokens_seen": 24470912,
      "step": 37520
    },
    {
      "epoch": 19.66719077568134,
      "grad_norm": 0.11801744252443314,
      "learning_rate": 8.459208643659122e-07,
      "loss": 0.4497,
      "num_input_tokens_seen": 24474304,
      "step": 37525
    },
    {
      "epoch": 19.669811320754718,
      "grad_norm": 0.25720998644828796,
      "learning_rate": 8.326761826773499e-07,
      "loss": 0.3752,
      "num_input_tokens_seen": 24477280,
      "step": 37530
    },
    {
      "epoch": 19.67243186582809,
      "grad_norm": 0.2093113660812378,
      "learning_rate": 8.195359210981246e-07,
      "loss": 0.4012,
      "num_input_tokens_seen": 24479456,
      "step": 37535
    },
    {
      "epoch": 19.67505241090147,
      "grad_norm": 0.24812081456184387,
      "learning_rate": 8.065000823770929e-07,
      "loss": 0.4812,
      "num_input_tokens_seen": 24482624,
      "step": 37540
    },
    {
      "epoch": 19.677672955974842,
      "grad_norm": 0.20114213228225708,
      "learning_rate": 7.935686692410737e-07,
      "loss": 0.3898,
      "num_input_tokens_seen": 24485792,
      "step": 37545
    },
    {
      "epoch": 19.68029350104822,
      "grad_norm": 0.20504792034626007,
      "learning_rate": 7.807416843952364e-07,
      "loss": 0.3794,
      "num_input_tokens_seen": 24488576,
      "step": 37550
    },
    {
      "epoch": 19.682914046121592,
      "grad_norm": 0.24890021979808807,
      "learning_rate": 7.68019130522879e-07,
      "loss": 0.5857,
      "num_input_tokens_seen": 24491136,
      "step": 37555
    },
    {
      "epoch": 19.68553459119497,
      "grad_norm": 0.13579563796520233,
      "learning_rate": 7.554010102853726e-07,
      "loss": 0.6165,
      "num_input_tokens_seen": 24494528,
      "step": 37560
    },
    {
      "epoch": 19.688155136268342,
      "grad_norm": 0.09883920103311539,
      "learning_rate": 7.428873263223279e-07,
      "loss": 0.5233,
      "num_input_tokens_seen": 24498816,
      "step": 37565
    },
    {
      "epoch": 19.69077568134172,
      "grad_norm": 0.1498717963695526,
      "learning_rate": 7.304780812513734e-07,
      "loss": 0.3905,
      "num_input_tokens_seen": 24502368,
      "step": 37570
    },
    {
      "epoch": 19.693396226415093,
      "grad_norm": 0.1877356320619583,
      "learning_rate": 7.181732776684325e-07,
      "loss": 0.4011,
      "num_input_tokens_seen": 24505440,
      "step": 37575
    },
    {
      "epoch": 19.69601677148847,
      "grad_norm": 0.115076445043087,
      "learning_rate": 7.059729181475572e-07,
      "loss": 0.4615,
      "num_input_tokens_seen": 24508512,
      "step": 37580
    },
    {
      "epoch": 19.698637316561843,
      "grad_norm": 0.12388212233781815,
      "learning_rate": 6.938770052409282e-07,
      "loss": 0.3764,
      "num_input_tokens_seen": 24511360,
      "step": 37585
    },
    {
      "epoch": 19.70125786163522,
      "grad_norm": 0.11069686710834503,
      "learning_rate": 6.81885541478855e-07,
      "loss": 0.4547,
      "num_input_tokens_seen": 24514048,
      "step": 37590
    },
    {
      "epoch": 19.703878406708597,
      "grad_norm": 0.29645398259162903,
      "learning_rate": 6.699985293697197e-07,
      "loss": 0.3693,
      "num_input_tokens_seen": 24517088,
      "step": 37595
    },
    {
      "epoch": 19.70649895178197,
      "grad_norm": 0.1831020563840866,
      "learning_rate": 6.582159714003111e-07,
      "loss": 0.4787,
      "num_input_tokens_seen": 24519808,
      "step": 37600
    },
    {
      "epoch": 19.709119496855347,
      "grad_norm": 0.22776445746421814,
      "learning_rate": 6.465378700352687e-07,
      "loss": 0.4893,
      "num_input_tokens_seen": 24524320,
      "step": 37605
    },
    {
      "epoch": 19.71174004192872,
      "grad_norm": 0.1517411768436432,
      "learning_rate": 6.349642277176382e-07,
      "loss": 0.4393,
      "num_input_tokens_seen": 24528192,
      "step": 37610
    },
    {
      "epoch": 19.714360587002098,
      "grad_norm": 0.19382672011852264,
      "learning_rate": 6.23495046868372e-07,
      "loss": 0.4694,
      "num_input_tokens_seen": 24530880,
      "step": 37615
    },
    {
      "epoch": 19.71698113207547,
      "grad_norm": 0.14850811660289764,
      "learning_rate": 6.121303298868286e-07,
      "loss": 0.4592,
      "num_input_tokens_seen": 24533952,
      "step": 37620
    },
    {
      "epoch": 19.719601677148848,
      "grad_norm": 0.15656247735023499,
      "learning_rate": 6.008700791502175e-07,
      "loss": 0.4091,
      "num_input_tokens_seen": 24537152,
      "step": 37625
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 0.3217153549194336,
      "learning_rate": 5.8971429701421e-07,
      "loss": 0.2973,
      "num_input_tokens_seen": 24539776,
      "step": 37630
    },
    {
      "epoch": 19.7248427672956,
      "grad_norm": 0.1238572895526886,
      "learning_rate": 5.786629858123283e-07,
      "loss": 0.4464,
      "num_input_tokens_seen": 24542944,
      "step": 37635
    },
    {
      "epoch": 19.72746331236897,
      "grad_norm": 0.1695779412984848,
      "learning_rate": 5.677161478565008e-07,
      "loss": 0.3889,
      "num_input_tokens_seen": 24545504,
      "step": 37640
    },
    {
      "epoch": 19.73008385744235,
      "grad_norm": 0.2282288372516632,
      "learning_rate": 5.56873785436618e-07,
      "loss": 0.4629,
      "num_input_tokens_seen": 24548736,
      "step": 37645
    },
    {
      "epoch": 19.732704402515722,
      "grad_norm": 0.20901957154273987,
      "learning_rate": 5.4613590082081e-07,
      "loss": 0.6219,
      "num_input_tokens_seen": 24552000,
      "step": 37650
    },
    {
      "epoch": 19.7353249475891,
      "grad_norm": 0.12380103021860123,
      "learning_rate": 5.355024962552801e-07,
      "loss": 0.376,
      "num_input_tokens_seen": 24555968,
      "step": 37655
    },
    {
      "epoch": 19.737945492662472,
      "grad_norm": 0.1409524530172348,
      "learning_rate": 5.249735739644157e-07,
      "loss": 0.4779,
      "num_input_tokens_seen": 24559104,
      "step": 37660
    },
    {
      "epoch": 19.74056603773585,
      "grad_norm": 0.2844904065132141,
      "learning_rate": 5.145491361508436e-07,
      "loss": 0.4605,
      "num_input_tokens_seen": 24561824,
      "step": 37665
    },
    {
      "epoch": 19.743186582809223,
      "grad_norm": 0.4963788390159607,
      "learning_rate": 5.042291849950975e-07,
      "loss": 0.3957,
      "num_input_tokens_seen": 24564864,
      "step": 37670
    },
    {
      "epoch": 19.7458071278826,
      "grad_norm": 0.06340683251619339,
      "learning_rate": 4.940137226560615e-07,
      "loss": 0.3753,
      "num_input_tokens_seen": 24568224,
      "step": 37675
    },
    {
      "epoch": 19.748427672955973,
      "grad_norm": 0.09945216774940491,
      "learning_rate": 4.839027512706928e-07,
      "loss": 0.4078,
      "num_input_tokens_seen": 24573312,
      "step": 37680
    },
    {
      "epoch": 19.75104821802935,
      "grad_norm": 0.11434116214513779,
      "learning_rate": 4.7389627295407743e-07,
      "loss": 0.3993,
      "num_input_tokens_seen": 24577248,
      "step": 37685
    },
    {
      "epoch": 19.753668763102727,
      "grad_norm": 0.12329350411891937,
      "learning_rate": 4.6399428979948534e-07,
      "loss": 0.3869,
      "num_input_tokens_seen": 24580544,
      "step": 37690
    },
    {
      "epoch": 19.7562893081761,
      "grad_norm": 0.13676801323890686,
      "learning_rate": 4.541968038782596e-07,
      "loss": 0.4324,
      "num_input_tokens_seen": 24582944,
      "step": 37695
    },
    {
      "epoch": 19.758909853249477,
      "grad_norm": 0.11743058264255524,
      "learning_rate": 4.445038172399829e-07,
      "loss": 0.5885,
      "num_input_tokens_seen": 24585728,
      "step": 37700
    },
    {
      "epoch": 19.76153039832285,
      "grad_norm": 0.14951163530349731,
      "learning_rate": 4.3491533191225563e-07,
      "loss": 0.4863,
      "num_input_tokens_seen": 24588640,
      "step": 37705
    },
    {
      "epoch": 19.764150943396228,
      "grad_norm": 0.1220066100358963,
      "learning_rate": 4.254313499009177e-07,
      "loss": 0.4548,
      "num_input_tokens_seen": 24592064,
      "step": 37710
    },
    {
      "epoch": 19.7667714884696,
      "grad_norm": 0.10869356244802475,
      "learning_rate": 4.1605187318982664e-07,
      "loss": 0.3957,
      "num_input_tokens_seen": 24598944,
      "step": 37715
    },
    {
      "epoch": 19.769392033542978,
      "grad_norm": 0.18809233605861664,
      "learning_rate": 4.067769037411906e-07,
      "loss": 0.3797,
      "num_input_tokens_seen": 24601152,
      "step": 37720
    },
    {
      "epoch": 19.77201257861635,
      "grad_norm": 0.17263756692409515,
      "learning_rate": 3.9760644349517984e-07,
      "loss": 0.4028,
      "num_input_tokens_seen": 24603904,
      "step": 37725
    },
    {
      "epoch": 19.77463312368973,
      "grad_norm": 0.18664014339447021,
      "learning_rate": 3.885404943700932e-07,
      "loss": 0.3788,
      "num_input_tokens_seen": 24606240,
      "step": 37730
    },
    {
      "epoch": 19.7772536687631,
      "grad_norm": 0.181952103972435,
      "learning_rate": 3.795790582624692e-07,
      "loss": 0.445,
      "num_input_tokens_seen": 24609088,
      "step": 37735
    },
    {
      "epoch": 19.77987421383648,
      "grad_norm": 0.16141915321350098,
      "learning_rate": 3.707221370469749e-07,
      "loss": 0.3871,
      "num_input_tokens_seen": 24612320,
      "step": 37740
    },
    {
      "epoch": 19.782494758909852,
      "grad_norm": 0.11064077168703079,
      "learning_rate": 3.6196973257629494e-07,
      "loss": 0.4046,
      "num_input_tokens_seen": 24617664,
      "step": 37745
    },
    {
      "epoch": 19.78511530398323,
      "grad_norm": 0.1435154676437378,
      "learning_rate": 3.533218466813537e-07,
      "loss": 0.3933,
      "num_input_tokens_seen": 24622304,
      "step": 37750
    },
    {
      "epoch": 19.787735849056602,
      "grad_norm": 0.17389971017837524,
      "learning_rate": 3.447784811712595e-07,
      "loss": 0.3953,
      "num_input_tokens_seen": 24625088,
      "step": 37755
    },
    {
      "epoch": 19.79035639412998,
      "grad_norm": 0.1319841593503952,
      "learning_rate": 3.363396378331385e-07,
      "loss": 0.4197,
      "num_input_tokens_seen": 24628544,
      "step": 37760
    },
    {
      "epoch": 19.792976939203353,
      "grad_norm": 0.30264168977737427,
      "learning_rate": 3.280053184323006e-07,
      "loss": 0.5636,
      "num_input_tokens_seen": 24631776,
      "step": 37765
    },
    {
      "epoch": 19.79559748427673,
      "grad_norm": 0.19069649279117584,
      "learning_rate": 3.1977552471218476e-07,
      "loss": 0.4295,
      "num_input_tokens_seen": 24634976,
      "step": 37770
    },
    {
      "epoch": 19.798218029350103,
      "grad_norm": 0.17445531487464905,
      "learning_rate": 3.116502583943581e-07,
      "loss": 0.49,
      "num_input_tokens_seen": 24637696,
      "step": 37775
    },
    {
      "epoch": 19.80083857442348,
      "grad_norm": 0.18367622792720795,
      "learning_rate": 3.036295211785722e-07,
      "loss": 0.3671,
      "num_input_tokens_seen": 24640864,
      "step": 37780
    },
    {
      "epoch": 19.803459119496857,
      "grad_norm": 0.06431667506694794,
      "learning_rate": 2.957133147425961e-07,
      "loss": 0.2871,
      "num_input_tokens_seen": 24646784,
      "step": 37785
    },
    {
      "epoch": 19.80607966457023,
      "grad_norm": 0.1373807191848755,
      "learning_rate": 2.879016407425494e-07,
      "loss": 0.3882,
      "num_input_tokens_seen": 24650048,
      "step": 37790
    },
    {
      "epoch": 19.808700209643607,
      "grad_norm": 0.16593942046165466,
      "learning_rate": 2.8019450081240295e-07,
      "loss": 0.3888,
      "num_input_tokens_seen": 24653728,
      "step": 37795
    },
    {
      "epoch": 19.81132075471698,
      "grad_norm": 0.19640259444713593,
      "learning_rate": 2.7259189656447803e-07,
      "loss": 0.3346,
      "num_input_tokens_seen": 24656288,
      "step": 37800
    },
    {
      "epoch": 19.813941299790358,
      "grad_norm": 0.13000473380088806,
      "learning_rate": 2.650938295891692e-07,
      "loss": 0.3878,
      "num_input_tokens_seen": 24659744,
      "step": 37805
    },
    {
      "epoch": 19.81656184486373,
      "grad_norm": 0.08978027105331421,
      "learning_rate": 2.5770030145494395e-07,
      "loss": 0.5245,
      "num_input_tokens_seen": 24663488,
      "step": 37810
    },
    {
      "epoch": 19.819182389937108,
      "grad_norm": 0.15311941504478455,
      "learning_rate": 2.504113137083985e-07,
      "loss": 0.3635,
      "num_input_tokens_seen": 24666272,
      "step": 37815
    },
    {
      "epoch": 19.82180293501048,
      "grad_norm": 0.10379710793495178,
      "learning_rate": 2.4322686787442425e-07,
      "loss": 0.2474,
      "num_input_tokens_seen": 24670400,
      "step": 37820
    },
    {
      "epoch": 19.82442348008386,
      "grad_norm": 0.20628918707370758,
      "learning_rate": 2.3614696545581904e-07,
      "loss": 0.3938,
      "num_input_tokens_seen": 24673280,
      "step": 37825
    },
    {
      "epoch": 19.82704402515723,
      "grad_norm": 0.16837771236896515,
      "learning_rate": 2.2917160793367585e-07,
      "loss": 0.4644,
      "num_input_tokens_seen": 24676320,
      "step": 37830
    },
    {
      "epoch": 19.82966457023061,
      "grad_norm": 0.2505894601345062,
      "learning_rate": 2.2230079676716086e-07,
      "loss": 0.3086,
      "num_input_tokens_seen": 24680960,
      "step": 37835
    },
    {
      "epoch": 19.832285115303982,
      "grad_norm": 0.17526890337467194,
      "learning_rate": 2.1553453339356875e-07,
      "loss": 0.4518,
      "num_input_tokens_seen": 24685792,
      "step": 37840
    },
    {
      "epoch": 19.83490566037736,
      "grad_norm": 0.17281851172447205,
      "learning_rate": 2.0887281922826738e-07,
      "loss": 0.3705,
      "num_input_tokens_seen": 24688448,
      "step": 37845
    },
    {
      "epoch": 19.837526205450732,
      "grad_norm": 0.12684383988380432,
      "learning_rate": 2.023156556648642e-07,
      "loss": 0.3311,
      "num_input_tokens_seen": 24691552,
      "step": 37850
    },
    {
      "epoch": 19.84014675052411,
      "grad_norm": 0.24551069736480713,
      "learning_rate": 1.9586304407503975e-07,
      "loss": 0.3933,
      "num_input_tokens_seen": 24694368,
      "step": 37855
    },
    {
      "epoch": 19.842767295597483,
      "grad_norm": 0.2053845226764679,
      "learning_rate": 1.8951498580860315e-07,
      "loss": 0.4543,
      "num_input_tokens_seen": 24696608,
      "step": 37860
    },
    {
      "epoch": 19.84538784067086,
      "grad_norm": 0.13585183024406433,
      "learning_rate": 1.832714821934922e-07,
      "loss": 0.3981,
      "num_input_tokens_seen": 24699872,
      "step": 37865
    },
    {
      "epoch": 19.848008385744233,
      "grad_norm": 0.1327725201845169,
      "learning_rate": 1.7713253453577328e-07,
      "loss": 0.3778,
      "num_input_tokens_seen": 24704608,
      "step": 37870
    },
    {
      "epoch": 19.85062893081761,
      "grad_norm": 0.16145828366279602,
      "learning_rate": 1.7109814411964132e-07,
      "loss": 0.4473,
      "num_input_tokens_seen": 24708544,
      "step": 37875
    },
    {
      "epoch": 19.853249475890987,
      "grad_norm": 0.19670705497264862,
      "learning_rate": 1.651683122074754e-07,
      "loss": 0.4523,
      "num_input_tokens_seen": 24710880,
      "step": 37880
    },
    {
      "epoch": 19.85587002096436,
      "grad_norm": 0.12058696895837784,
      "learning_rate": 1.5934304003961675e-07,
      "loss": 0.3786,
      "num_input_tokens_seen": 24715008,
      "step": 37885
    },
    {
      "epoch": 19.858490566037737,
      "grad_norm": 0.20549120008945465,
      "learning_rate": 1.5362232883475713e-07,
      "loss": 0.4264,
      "num_input_tokens_seen": 24718240,
      "step": 37890
    },
    {
      "epoch": 19.86111111111111,
      "grad_norm": 0.31511804461479187,
      "learning_rate": 1.4800617978949492e-07,
      "loss": 0.3491,
      "num_input_tokens_seen": 24720800,
      "step": 37895
    },
    {
      "epoch": 19.863731656184488,
      "grad_norm": 0.13419052958488464,
      "learning_rate": 1.424945940787792e-07,
      "loss": 0.4266,
      "num_input_tokens_seen": 24723392,
      "step": 37900
    },
    {
      "epoch": 19.86635220125786,
      "grad_norm": 0.16583633422851562,
      "learning_rate": 1.3708757285552098e-07,
      "loss": 0.4186,
      "num_input_tokens_seen": 24725792,
      "step": 37905
    },
    {
      "epoch": 19.868972746331238,
      "grad_norm": 0.29735830426216125,
      "learning_rate": 1.3178511725076004e-07,
      "loss": 0.3973,
      "num_input_tokens_seen": 24728672,
      "step": 37910
    },
    {
      "epoch": 19.87159329140461,
      "grad_norm": 0.16215915977954865,
      "learning_rate": 1.265872283738312e-07,
      "loss": 0.4687,
      "num_input_tokens_seen": 24732384,
      "step": 37915
    },
    {
      "epoch": 19.87421383647799,
      "grad_norm": 0.358022004365921,
      "learning_rate": 1.2149390731192033e-07,
      "loss": 0.389,
      "num_input_tokens_seen": 24735744,
      "step": 37920
    },
    {
      "epoch": 19.87683438155136,
      "grad_norm": 0.2141668200492859,
      "learning_rate": 1.1650515513061955e-07,
      "loss": 0.4022,
      "num_input_tokens_seen": 24738496,
      "step": 37925
    },
    {
      "epoch": 19.87945492662474,
      "grad_norm": 0.10764685273170471,
      "learning_rate": 1.1162097287342743e-07,
      "loss": 0.3821,
      "num_input_tokens_seen": 24742080,
      "step": 37930
    },
    {
      "epoch": 19.882075471698112,
      "grad_norm": 0.14543677866458893,
      "learning_rate": 1.0684136156213775e-07,
      "loss": 0.3836,
      "num_input_tokens_seen": 24744512,
      "step": 37935
    },
    {
      "epoch": 19.88469601677149,
      "grad_norm": 0.11368519812822342,
      "learning_rate": 1.0216632219650634e-07,
      "loss": 0.3552,
      "num_input_tokens_seen": 24748224,
      "step": 37940
    },
    {
      "epoch": 19.887316561844862,
      "grad_norm": 0.27930089831352234,
      "learning_rate": 9.759585575458418e-08,
      "loss": 0.5227,
      "num_input_tokens_seen": 24751232,
      "step": 37945
    },
    {
      "epoch": 19.88993710691824,
      "grad_norm": 0.09355738013982773,
      "learning_rate": 9.312996319238432e-08,
      "loss": 0.3906,
      "num_input_tokens_seen": 24754208,
      "step": 37950
    },
    {
      "epoch": 19.892557651991613,
      "grad_norm": 0.19356176257133484,
      "learning_rate": 8.876864544421493e-08,
      "loss": 0.414,
      "num_input_tokens_seen": 24757440,
      "step": 37955
    },
    {
      "epoch": 19.89517819706499,
      "grad_norm": 0.22088783979415894,
      "learning_rate": 8.451190342229077e-08,
      "loss": 0.4331,
      "num_input_tokens_seen": 24759808,
      "step": 37960
    },
    {
      "epoch": 19.897798742138363,
      "grad_norm": 0.1959412395954132,
      "learning_rate": 8.035973801717722e-08,
      "loss": 0.3844,
      "num_input_tokens_seen": 24761952,
      "step": 37965
    },
    {
      "epoch": 19.90041928721174,
      "grad_norm": 0.08224489539861679,
      "learning_rate": 7.631215009740178e-08,
      "loss": 0.5027,
      "num_input_tokens_seen": 24765280,
      "step": 37970
    },
    {
      "epoch": 19.903039832285117,
      "grad_norm": 0.19808271527290344,
      "learning_rate": 7.236914050973153e-08,
      "loss": 0.4227,
      "num_input_tokens_seen": 24768320,
      "step": 37975
    },
    {
      "epoch": 19.90566037735849,
      "grad_norm": 0.3963340222835541,
      "learning_rate": 6.853071007895117e-08,
      "loss": 0.4272,
      "num_input_tokens_seen": 24770912,
      "step": 37980
    },
    {
      "epoch": 19.908280922431867,
      "grad_norm": 0.19390518963336945,
      "learning_rate": 6.479685960797399e-08,
      "loss": 0.5197,
      "num_input_tokens_seen": 24774176,
      "step": 37985
    },
    {
      "epoch": 19.91090146750524,
      "grad_norm": 0.14693881571292877,
      "learning_rate": 6.116758987800841e-08,
      "loss": 0.4546,
      "num_input_tokens_seen": 24777120,
      "step": 37990
    },
    {
      "epoch": 19.913522012578618,
      "grad_norm": 0.17677158117294312,
      "learning_rate": 5.7642901648113924e-08,
      "loss": 0.3814,
      "num_input_tokens_seen": 24779712,
      "step": 37995
    },
    {
      "epoch": 19.91614255765199,
      "grad_norm": 0.1954173445701599,
      "learning_rate": 5.422279565570065e-08,
      "loss": 0.4565,
      "num_input_tokens_seen": 24783200,
      "step": 38000
    },
    {
      "epoch": 19.918763102725368,
      "grad_norm": 0.13731749355793,
      "learning_rate": 5.090727261619632e-08,
      "loss": 0.533,
      "num_input_tokens_seen": 24786592,
      "step": 38005
    },
    {
      "epoch": 19.92138364779874,
      "grad_norm": 0.20658667385578156,
      "learning_rate": 4.7696333223212765e-08,
      "loss": 0.3519,
      "num_input_tokens_seen": 24789664,
      "step": 38010
    },
    {
      "epoch": 19.92400419287212,
      "grad_norm": 0.19306719303131104,
      "learning_rate": 4.4589978148323884e-08,
      "loss": 0.4375,
      "num_input_tokens_seen": 24792288,
      "step": 38015
    },
    {
      "epoch": 19.92662473794549,
      "grad_norm": 0.1261039674282074,
      "learning_rate": 4.1588208041454244e-08,
      "loss": 0.3561,
      "num_input_tokens_seen": 24796000,
      "step": 38020
    },
    {
      "epoch": 19.92924528301887,
      "grad_norm": 0.13297304511070251,
      "learning_rate": 3.8691023530545985e-08,
      "loss": 0.3805,
      "num_input_tokens_seen": 24798976,
      "step": 38025
    },
    {
      "epoch": 19.931865828092242,
      "grad_norm": 0.22517029941082,
      "learning_rate": 3.589842522155884e-08,
      "loss": 0.3813,
      "num_input_tokens_seen": 24801760,
      "step": 38030
    },
    {
      "epoch": 19.93448637316562,
      "grad_norm": 0.15852969884872437,
      "learning_rate": 3.321041369874767e-08,
      "loss": 0.4208,
      "num_input_tokens_seen": 24805376,
      "step": 38035
    },
    {
      "epoch": 19.937106918238992,
      "grad_norm": 0.20382773876190186,
      "learning_rate": 3.0626989524384916e-08,
      "loss": 0.4813,
      "num_input_tokens_seen": 24809440,
      "step": 38040
    },
    {
      "epoch": 19.93972746331237,
      "grad_norm": 0.1704576164484024,
      "learning_rate": 2.8148153238927167e-08,
      "loss": 0.4518,
      "num_input_tokens_seen": 24813472,
      "step": 38045
    },
    {
      "epoch": 19.942348008385743,
      "grad_norm": 0.13498327136039734,
      "learning_rate": 2.5773905360904072e-08,
      "loss": 0.4683,
      "num_input_tokens_seen": 24816608,
      "step": 38050
    },
    {
      "epoch": 19.94496855345912,
      "grad_norm": 0.1554742455482483,
      "learning_rate": 2.3504246386918392e-08,
      "loss": 0.4927,
      "num_input_tokens_seen": 24819776,
      "step": 38055
    },
    {
      "epoch": 19.947589098532493,
      "grad_norm": 0.18718905746936798,
      "learning_rate": 2.133917679186803e-08,
      "loss": 0.3691,
      "num_input_tokens_seen": 24822560,
      "step": 38060
    },
    {
      "epoch": 19.95020964360587,
      "grad_norm": 0.1688859760761261,
      "learning_rate": 1.9278697028557447e-08,
      "loss": 0.3015,
      "num_input_tokens_seen": 24825600,
      "step": 38065
    },
    {
      "epoch": 19.952830188679247,
      "grad_norm": 0.16020479798316956,
      "learning_rate": 1.7322807528086238e-08,
      "loss": 0.4665,
      "num_input_tokens_seen": 24828224,
      "step": 38070
    },
    {
      "epoch": 19.95545073375262,
      "grad_norm": 0.13883472979068756,
      "learning_rate": 1.547150869957159e-08,
      "loss": 0.3996,
      "num_input_tokens_seen": 24830848,
      "step": 38075
    },
    {
      "epoch": 19.958071278825997,
      "grad_norm": 0.21889746189117432,
      "learning_rate": 1.3724800930314806e-08,
      "loss": 0.4331,
      "num_input_tokens_seen": 24834304,
      "step": 38080
    },
    {
      "epoch": 19.96069182389937,
      "grad_norm": 0.10661504417657852,
      "learning_rate": 1.2082684585634773e-08,
      "loss": 0.3117,
      "num_input_tokens_seen": 24838112,
      "step": 38085
    },
    {
      "epoch": 19.963312368972748,
      "grad_norm": 0.13833199441432953,
      "learning_rate": 1.0545160009145516e-08,
      "loss": 0.3964,
      "num_input_tokens_seen": 24842240,
      "step": 38090
    },
    {
      "epoch": 19.96593291404612,
      "grad_norm": 0.15689346194267273,
      "learning_rate": 9.112227522423133e-09,
      "loss": 0.3396,
      "num_input_tokens_seen": 24845824,
      "step": 38095
    },
    {
      "epoch": 19.968553459119498,
      "grad_norm": 0.20959697663784027,
      "learning_rate": 7.783887425172331e-09,
      "loss": 0.5092,
      "num_input_tokens_seen": 24848704,
      "step": 38100
    },
    {
      "epoch": 19.97117400419287,
      "grad_norm": 0.16224069893360138,
      "learning_rate": 6.560139995392955e-09,
      "loss": 0.2943,
      "num_input_tokens_seen": 24853120,
      "step": 38105
    },
    {
      "epoch": 19.97379454926625,
      "grad_norm": 0.11844205111265182,
      "learning_rate": 5.440985488935901e-09,
      "loss": 0.5035,
      "num_input_tokens_seen": 24856800,
      "step": 38110
    },
    {
      "epoch": 19.97641509433962,
      "grad_norm": 0.20280981063842773,
      "learning_rate": 4.426424140058227e-09,
      "loss": 0.4863,
      "num_input_tokens_seen": 24859392,
      "step": 38115
    },
    {
      "epoch": 19.979035639413,
      "grad_norm": 0.152194544672966,
      "learning_rate": 3.5164561608680424e-09,
      "loss": 0.4299,
      "num_input_tokens_seen": 24863328,
      "step": 38120
    },
    {
      "epoch": 19.981656184486372,
      "grad_norm": 0.11542458832263947,
      "learning_rate": 2.7110817417685953e-09,
      "loss": 0.3743,
      "num_input_tokens_seen": 24866496,
      "step": 38125
    },
    {
      "epoch": 19.98427672955975,
      "grad_norm": 0.13419221341609955,
      "learning_rate": 2.010301051291741e-09,
      "loss": 0.3776,
      "num_input_tokens_seen": 24871328,
      "step": 38130
    },
    {
      "epoch": 19.986897274633122,
      "grad_norm": 0.17621798813343048,
      "learning_rate": 1.4141142359314074e-09,
      "loss": 0.3589,
      "num_input_tokens_seen": 24874336,
      "step": 38135
    },
    {
      "epoch": 19.9895178197065,
      "grad_norm": 0.1716824322938919,
      "learning_rate": 9.22521420476663e-10,
      "loss": 0.3865,
      "num_input_tokens_seen": 24877024,
      "step": 38140
    },
    {
      "epoch": 19.992138364779873,
      "grad_norm": 0.3087512254714966,
      "learning_rate": 5.355227077341596e-10,
      "loss": 0.4515,
      "num_input_tokens_seen": 24880096,
      "step": 38145
    },
    {
      "epoch": 19.99475890985325,
      "grad_norm": 0.18653440475463867,
      "learning_rate": 2.5311817863915566e-10,
      "loss": 0.4372,
      "num_input_tokens_seen": 24883200,
      "step": 38150
    },
    {
      "epoch": 19.997379454926623,
      "grad_norm": 0.29159706830978394,
      "learning_rate": 7.53078923110273e-11,
      "loss": 0.4077,
      "num_input_tokens_seen": 24885664,
      "step": 38155
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.022653376683592796,
      "learning_rate": 2.0918859422458526e-12,
      "loss": 0.301,
      "num_input_tokens_seen": 24887720,
      "step": 38160
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.4668559730052948,
      "eval_runtime": 13.5762,
      "eval_samples_per_second": 62.462,
      "eval_steps_per_second": 15.616,
      "num_input_tokens_seen": 24887720,
      "step": 38160
    },
    {
      "epoch": 20.0,
      "num_input_tokens_seen": 24887720,
      "step": 38160,
      "total_flos": 1.120682783146967e+18,
      "train_loss": 0.4599470546722912,
      "train_runtime": 5871.1936,
      "train_samples_per_second": 25.988,
      "train_steps_per_second": 6.5
    }
  ],
  "logging_steps": 5,
  "max_steps": 38160,
  "num_input_tokens_seen": 24887720,
  "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": 1.120682783146967e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}